Westminster Theological Journal 20 (1958) 46-70.

        Copyright © 1958 by Westminster Theological Seminary, cited with permission.   





                                                THIRD ARTICLE



                                           MEREDITH G. KLINE


                              II. Ha-BI-ru--HEBREW RELATIONS


            A fascination with the possibilities of illuminating Hebrew

origins has characterized studies of the ha-BI-ru. As observed

at the outset, popular theory has it that the Hebrews were

one offshoot of the ha-BI-ru. This theory may start with

the supposition that the ha-BI-ru were a social class or an

ethnic group. Although some form of either approach can be

developed without the assumption that the terms ha-BI-ru

and 'Ibri can be equated phonetically or at least semantically

they are greatly strengthened if such equation can be estab-

lished. It is necessary in this connection to survey the usage

of 'Ibrim in the Old Testament and to face the question of

the phonetic relation of ha-BI-ru and 'Ibri.



A. The Usage of 'Ibrim in the Old Testament.


Support for the view that the term ha-BI-ru denotes a

larger whole from which the biblical Hebrews originated has

been claimed in the usage of the term 'Ibrim in the Old

Testament. There is no doubt that the gentilic 'Ibri is

ordinarily used in the Old Testament as an ethnicon for

Abraham and his descendants of the Isaac-Jacob line.178 In a


178 The word is found almost exclusively in a few clusters which suggests

that particular circumstances account for its employment. One such

group appears in the narrative of the Egyptian sojourn and bondage; a

second in the record of Israelite-Philistine relationships during the days of

Samuel and Saul; and a third in a series of texts dealing with the manumis-

sion of Hebrew servants. There are besides only the isolated appearances

in Genesis 14:13 and Jonah 1:9. The great majority of these are instances

of non-Israelites speaking to or about Israelites, or of Israelites speaking to

foreigners, or of declarations of God destined for foreigners. Where it is


HA-BI-RU                                          47


few passages, however, some have judged that 'Ibrim is used

in a non-Israelite or even appellative sense and that in such

texts an original, wider (i. e., ha-BI-ru) connotation emerges.

These passages must be examined.



1. The 'Ebed 'Ibri Legislation.


In the legislation of Exod. 21:2 and Deut. 15:12 and in

the references to these laws in Jer. 34:9, 14 the term ‘Ibri has

been thought to denote not the ethnic character of the servant

but a particular variety of servanthood. J. Lewy develops

this theory on the basis of his interpretation of the term

ha-Bl-ru in the Nuzu contracts as an appellative meaning

"foreign-servant", and his judgment that the parallels between

the status of the ha-BI-ru servants and the 'ebed ‘Ibri of

Exod. 21:2 (and the associated passages) are so close and

numerous as to indicate identical institutions and identity

of meaning for ha-BI-ru and 'Ibri.179


the Israelite author who employs the term he is often adapting his ter-

minology to the usage in the context. In several passages a contrast is

drawn between Israelites and other ethnic groups.

It has been suggested that ‘Ibri uniformly possesses a peculiar connota-

tion. For example, DeVaux (RB 55, 1948, pp. 344 ff.) maintains that it

has a derogatory nuance and finds the common element in the fact that

the 'Ibrim are strangers in the milieu, while Kraeling (AJSL 58, 1941

pp. 237 ff.) suggests that 'Ibri is an alternate for "Israelite" in situations

where the designee is not a free citizen in a free community or on free soil.

The latter formulation seems to be successful in unravelling a strand

common to all the 'Ibri contexts but it remains uncertain whether such a

nuance necessarily attached to the employment of the word. Cf. Green-

berg, op. cit., p. 92.

   179 HUCA XIV, 1939, pp. 587 ff.; XV, 1940, pp. 47 ff. Cf. his note in

Bottero, op. cit., pp. 163-4, where he translates ha-BI-ru as "resident

alien". Lewy supports his thesis with the considerations that the ha-BI-ru

are present in the Mitannian orbit in the period during which the 'Ibrim

became a nation and that the whole area in question had been unified

under the Hyksos with the result that the same technical terms and

analogous institutions are found throughout. He holds that this social-

legal appellative usage of Ibri represents the earliest stage (noting its

appearance in the first paragraph of Israel's Book of the Covenant) but

that later the term was used in an ethnic sense for the descendants of the

"Hebrews par excellence". Cf. supra WTJ XIX, pp. 183, 184.



But is the situation on the Nuzu side clearly as Lewy has

reconstructed it? There are texts180 in which the person(s)

concerned is not designated as an ha-BI-ru and yet the essen-

tial clauses of the contract are those characteristic of the

contracts where the persons are labeled as ha-BI-ru. It is,

therefore, difficult to insist that we are dealing with a specif-

ically ha-BI-ru type of servanthood.181 While, therefore,

ha-BI-ru are found in the great majority of these contracts,

they are not necessarily involved in all of them,182 and one

may not assume then the existence in the Nuzu area of a

specifically ha-BI-ru brand of slavery.

Moreover, even if Lewy's view of the Nuzu evidence were

to be adopted, the biblical evidence would contradict the

translation of ‘Ibri as "foreign-servant" in the ‘ebed ‘Ibri

legislation. For the biblical law is patently not dealing with

foreign servants but with those who were their masters'

brethren. The Deut. 15:12 expansion of the original state-

ment reads, "If thy brother183 a Hebrew man, or a Hebrew

woman, be sold unto thee"; while Jeremiah, further expanding

it urges "that every man should let go free his man-servant

and every man his maid-servant, that is a Hebrew or He-

brewess ; that none should make bondmen of them, namely,

of a Jew, his brother" (34:9, cf. vs. 14). While one may then

recognize the instructive parallels in the conditions of servant-

hood at Nuzu and in the biblical legislation, it is impossible

to hold that ‘Ibri is in this legislation a technical term for a


   180 JEN VI, 610, 611, 613 (cf. JEN V, 456:9-23); JEN V, 446, 449,

457 and 462.

   181 An alternate interpretation has been advocated in the present study.

See supra WTJ XIX, pp. 179, 180, 183, 184.

   182 Especially relevant is the figure of Attilammu the Assyrian in the

servant contract JEN VI, 613:2. Even when this text in abbreviated form

is included in the Sammelurkunde JEN V, 456 between two contracts in

which the persons are specifically designated as ha-BI-ru (i. e., in a situation

where there would be a tendency to uniformity), Attilammu is not

described as an ha-BI-ru. It is further to be observed in connection with

the use of as-su-ra-a-a-u for Attilammu in JEN VI, 613 that when ha-BI-ru

from Ashur are so described it is as sa-mat as-su-ur.

   183 Note the clear distinction drawn in verse 3 between "the foreigner"

and "thy brother" in the law of the seventh year release with respect

to debt.

HA-BI-RU                                          49


specific type of servanthood184 and least of all for the

idea of "foreign-servant". Its usage is rather ethnic, as



2. The ‘Ibrim in I Samuel 13 and 14.


It has been affirmed that the 'Ibrim here (cf. 13:3, 7, 19;

14:11, 21) are quite clearly non-Israelites.185 The proper

interpretation of these verses is, indeed, difficult; nevertheless,

to distinguish between the ‘Ibrim and the Israelites would

be at odds with the decisive evidence in this context of their

identity. Thus, in 13:3, 4,  Myrib;fihA and lxerAW;yi-lkA  are obvious

equivalents (cf. Ufm;wA lxerAW;yi-lkAv; :Myrib;fihA Ufm;w;yi).186 More-

over, it is apparently in reference to the hiding of those de-

scribed in 13:6 as the "men of Israel" that the Philistines say,

"Behold, the ‘Ibrim are coming out of the holes where they had

hid themselves" (14:11b). Again, the equivalence of Myrib;fihA

with the inhabitants lxerAW;yi Cr,x, lkoB; and with lxerAW;yi-lkA

in 13:19, 20 is evident.

To find, then, in the ‘Ibrim of 13:7 a group ethnically

distinct from the "men of Israel" in 13:6 would involve for

the term ‘Ibrim a change from its contextual significance too

abrupt to be plausible. Verses 6 and 7 are concerned with

two groups of Israelites. Verse 6 refers to those excused by

Saul from military service (cf. vs. 2).187 These hide in the

hills and caves west of Jordan. Verse 7 refers to certain of

the selected troops who were with Saul at Gilgal near the

Jordan. These, deserting, cross over the river to the land of

Gad and Gilead east of Jordan.188


    184 The 'ebed in the phrase ‘ebed ‘Ibri (Exod. 21:2) would then be tau-

tological, and Alt feels obliged to exscind it from the text.

    185 Cf. e. g., A. Guillaume, PEQ, 1946, p. 68.

    186 The LXX rendering of the end of verse 3, h]qeth<kasin oi[ dou?loi

(as though the Hebrew were Myrbfh vfwp) seems to be a conjectural emenda-

tion occasioned by the fact that Myrib;fihA comes somewhat unexpectedly

on the lips of Saul.

    187 13:4b does not describe a regathering of those sent home but simply

indicates the new location of Saul and his chosen army at Gilgal.

    188 There were originally 3000 chosen by Saul (13:2), but after the

approach of the Philistines in force and Samuel's delay there were only

600 left (13:11, 15; 14:2).



In 14:21 it is not necessary to follow the English versions in

regarding the ‘Ibrim as men who had been serving in the

Philistine army. Even if such a translation were adopted, it

would still be gratuitous to identify these ‘Ibrim as non-

Israelites for they might be Israelite turn-coats.

But verse 21 may be translated : "Now the Hebrews were

towards the Philistines as formerly when189 they went up

with them in the camp round about;190 both they were with

the Israelites who were with Saul and Jonathan and...".

The antecedent of Mm.Afi, "with them", appears to be "Saul

and all the people (or army)" of verse 20. Another possibility

is to regard "the Philistines" as the antecedent of "them"

but to translate the preposition "against".191 In either case

this passage would contain no mention of ‘Ibrim as having

served in Philistine forces. Verses 21 and 22 rather distinguish

as two elements swelling the unexpectedly triumphant rem-

nants of Saul's army those who had deserted after being

selected by Saul to encamp against the Philistines (vs. 21)

and those who, after being dismissed by Saul,192 were fright-

ened into hiding by the alarming course of the conflict (vs. 22).

This distinction in 14:21, 22 is the same as that found in

13:6, 7a. Indeed, the terminology in the two passages is

deliberately made to correspond. ‘Ibrim is used in both

13:7a and 14:21 for the deserters; and "men of Israel" in

13:6 and 14:22 for the people who hid in the hill-country of

Ephraim. The ‘Ibrim of 14:21 will then be the deserting

soldiers of Saul who had crossed over193 the Jordan but now

resume their former position in the Israelite ranks against

the Philistines.


    189 Cf. Brown, Driver and Briggs, Hebrew and English Lexicon of the Old

Testament (Oxford, 1952) under rwAxE 4b (a).

    190 Is this an allusion to the circumstance that the original three Israelite

positions at Bethel, Michmash, and Gibeah surrounded the Philistine

garrison at Geba? If the Massoretic text and accentuation (bybisA) stand,

the next clause will be a pseudo-verbal construction (as translated above).

The LXX and Syraic would read MGa Ubb;sA, "they also turned", which would

provide a parallel to Mga UqB;d;y.ava (vs. 22).

    191 Cf. Brown, Driver and Briggs, op. cit., under Mfi lc.

    192 For a similar military development see Judg. 7:3-7, 23, 24.

    193 The use of  Urb;fA. in 13:7a suggests the possibility of Myrib;fohAv;, "those

who passed over", as the original in 14:21 (cf. the participle, MyxiB;Hat;mi.ha,

HA-BI-RU                                          51


3. Abraham the ‘Ibri (Gen. 14:13).


Is ‘Ibri in this its earliest biblical appearance used eth-

nically? This question may be dealt with in connection with

an inquiry into the origin of the term ‘Ibri. Broad contextual

considerations indicate that in his use of ‘Ibri in Gen. 14:13,

the author had in mind ‘Eber of the line of Shem (cf. Gen.

10:21, 24, 25; 11:14-i 7).194 The direct descent of Abraham

from ‘Eber had already been traced in the genealogy of

Gen. 11:10-26. Moreover, the departure from the stereotyped

presentation of the genealogical data in Gen. 10 to describe

Shem as "the father of all the children of ‘Eber" (vs. 21)195

is most readily accounted for as an anticipation of the author's

imminent concentration (cf. Gen. 11:27 ff.) upon the Semitic

Eberites par excellence, i. e., the "Hebrews" whom Yahweh

chose to be the channel of revelation and redemption. In

Gen. 14:13 then, ‘Ibri is a patronymic, applied in this isolated

way to Abraham perhaps to contrast him with the many other

ethnic elements which play a role in this context.

On the other hand, many regard this usage of ‘Ibri as

appellative and then find their interpretations of the term

ha-BI-ru reflected in it.196  The appellative view is ancient,

for the LXX renders yrib;fihA as o[ pera<thj;197 Aquila, as

perai~thj; Jerome, as transeuphratensis; and the prevailing

view of the rabbis a generation after Aquila was that yrib;fihA


in the corresponding member of 14:21). Such a change in the Massoretic

pointing would support a corresponding change to Myrib;fov; in 13:7a. If

the Massoretic Myrib;fiv;. is original, the author perhaps employed this

designation of the Israelites to produce a word play with Urb;fA.

    194 yrib;fiibri) is the gentilic formation of rbAfaeber).

    195 Cf. also the additional remark in Gen. 10:25.

    196 For example, W. F. Albright, JAOS 48, 1928, pp. 183 ff., once found

in both the idea of "mercenary"; and DeVaux, op. cit., pp. 337 ff., that of

"stranger". Kraeling, op. cit., held that ‘Ibri is used to underscore Abra-

ham's role as a sojourner who pays tribute to Melchizedek.

    197 Parzen, AJSL 49, pp. 254 ff., is mistaken in his opinion that the

LXX actually found rbfh in the Hebrew text. Noth, "Erwagungen zur

Hebraerfrage", in Festschrift Otto Procksch (Leipzig, 1934), pp. 99 ff., is

probably correct in stating that the LXX translator simply regarded it as

desirable at this first appearance of ‘Ibri to indicate what was, in his

opinion, its significance.



designated Abraham as "from the other side of the river".198

All of these derived 'Ibri from the substantive meaning "the

other side" rather than from the verb ‘br.199 In line with this

view of the etymology is the emphasis in Joshua 24:2, 3 on

Abraham's origin "beyond the River". But these facts are

far from possessing the weight of the more immediate con-

textual considerations cited above. Here too then ‘Ibri is

not appellative but ethnic.


4. Conclusion.


It has appeared from this study that, the term 'Ibrim in

the Old Testament has uniformly an ethnic meaning and

denotes descendants of Eber in the line of Abraham-Isaac-

Jacob exclusively. Deriving from the eponymous ancestor

'Eber the term is probably early;200 in particular, its applica-

tion to Abraham need not be proleptic. To judge from

its characteristic association with foreigners in the biblical

contexts and the general avoidance of it by the Israelites,

it possibly originated outside the line of Abraham. Orig-

inally it may have been of wider application than is the

usage in the Old Testament, denoting other descendants of

Eber than the Abrahamites. This is perhaps suggested by

the use of 'Eber in Gen. 10:21 and Num. 24:24.201 In that


    199 Greenberg, op. cit., p. 5, n. 24, directs attention to the evidence for

this in Beresit Rabba 42, 8. A minority opinion of the rabbis was that

Abraham was called the 'Ibri because he was a descendant of 'Eber.

    199 This appears to be so even in the LXX, although later Patristic

writings in treating the LXX rendering derived it from a verbal base.

(cf. Greenberg, ibid.).

    200 Kraeling, op. cit., offers the strange hypothesis that "Hebrews"

is a secondarily personalized form of a geographical name, i. e., "Overites"

from  rhAnA.ha rbAfa adopted by the Israelites as late as the early monarchy in

an attempt to orientate themselves to the world in which they had just

become prominent. The usage would thus be that of the first millennium

even when applied to the Patriarchs. H. H. Rowley counters: (a) in the

early monarchy, consciousness of being from over the Euphrates is not

apparent among the Hebrews; (b) the term disappeared almost completely

from the Old Testament with the establishment of the monarchy; (c) The

Israelites would hardly adopt as a symbol of self-esteem a term "generally

employed in a pejorative sense". PEQ, 1942, pp. 41-53; From Joseph to

Joshua, 1952, pp. 54-5; cf. further O'Callaghan's criticism in Aram

Naharaim p. 216, n. 4.

    201 The validity of conclusions based on the tradition of descent from

HA-BI-RU                                          53


case the appearance of such gentilic but non-Abrahamic

‘Ibrim in some non-biblical text of the patriarchal age need

not come altogether unexpectedly.

Do the ha-BI-ru qualify? According to the conclusions

already reached in this study concerning the probable ge-

ographical and ethnic origins of the ha-BI-ru they do not

qualify as Semitic let alone Eberite kin of the Hebrews.202

On the other hand, a final judgment on this larger issue is


Eber is challenged by DeVaux's contention (op. cit.) that there are diver-

gent views within the Old Testament. He grants that the composer(s)

of the biblical genealogies derives ‘Ibri from the ancestor ‘Eber, but finds

in the reference to Jacob as a "wandering Aramean" (Deut. 26:5) a

conflicting tradition of Aramaic origin (cf. Gen. 10:22-24). DeVaux

believes the latter to be further supported by the description of Laban,

grandson of Abraham's brother Nahor, as an "Aramean" (Gen. 31:20).

According to the record, however, the term "Aramean" could have been

applied to both Jacob and Laban in virtue of their long residence in

Paddan-aram and so construed would say nothing about their lineage.

DeVaux also insists, but unnecessarily, on identifying the Aram of Gen.

10:22 and the Aram of Gen. 22:21, which would then bring the two passages

into hopeless confusion. Finally, DeVaux appeals to the prophetic denun-

ciation of Jerusalem in Ezek. 16:3, "your origin and your nativity are of

the land of the Canaanite; the Amorite was your father and the Hittite

your mother". Actually, as is apparent from the context (cf. especially

vss. 45 ff.), Ezekiel is using a scathing figure to say that from the first

Israel was just as much disqualified spiritually from enjoying a covenantal

relationship with Yahweh as were her despised heathen neighbors--the

point being that Israel's election must be attributed solely to the principle

of divine grace. But even if Ezekiel were speaking of literal racial inter-

mixture, the reference would be not to Abraham's family origins but to the

subsequent mingling of the racial strain of his descendants with those of

the inhabitants of Canaan. DeVaux's view is that the Hebrews and ha-

BI-ru were of common Aramaean descent. Starting with the notion that

the ha-BI-ru were desert nomads, DeVaux seeks to relate the ha-BI-ru

to the Aramaeans by a partial identification of them with proto-Aramaean

nomadic Ahlamu.

     202 Greenberg, op. cit, pp. 93 ff., provides an example of how the biblical

usage of ‘Ibrim can be regarded as consistently ethnic, and ha-BI-ru be

deemed an appellative for a social class, and yet the terms be equated

and the Hebrews derived from the ha-BI-ru. He suggests that Abraham

was an ha-BI-ru, but this epithet as applied to Abraham's descendants

became an ethnicon. Later biblical genealogists, unaware of this, invented

the ancestor 'Eber, man of many descendants, in order to explain at one

stroke the known kinship of the Hebrews to other Semitic tribes and the

origin of their name!



bound to be seriously affected by one's opinion on the phonetic

question of whether the term ha-BI-ru can be equated with

the term 'Ibri (and so be derived from 'Eber).203


B. Phonetic Relation of Ha-BI-ru to 'Ibri.


1. Consonants. The common cuneiform spelling of the name

is ha-BI-ru the final u being, according to the usual assump-

tion, the nominative case ending, which yields as the grammat-

ical relations require to other case or gentilic endings.204 In this

cuneiform rendering the identity of the first two radicals is

ambiguous. The initial consonant is ambiguous because

Accadian h may represent other letters than Hebrew H;205

among them, Hebrew f.206 The second is ambiguous because


    203 In addition to the supposed phonetic equivalence of ha-BI-ru and

'Ibri, support has been sought for the derivation of the Hebrews from the

ha-BI-ru by appeal to certain parallels in the careers of the two. But the

similarities are for the most part superficial or based on misinterpretations

of the data on one side or the other. For a recent popular example see

H. Orlinsky, Ancient Israel, 1954; cf. DeVaux RB 55, 1948, pp. 342 ff.;

H. H. Rowley From Joseph to Joshua, 1952, p. 53, n. 1. Items like the

following have been or might be mentioned: (a) In each case there is a

westward movement about the Fertile Crescent. (But this cannot be

demonstrated for the ha-BI-ru and, in the case of the Hebrews, it applies

not to the group as such but only to Abraham.) (b) The chronological

span of the use of the terms ha-BI-ru and 'Ibri is roughly the same. (c)

Both groups move in the Hurrian cultural orbit and exhibit the influence

of this fact. (d) The military activity of Abraham the Hebrew in Genesis

14 and the attack of Simeon and Levi on Shechem are comparable to

ha-BI-ru razzias. (But this involves a superficial estimate of both biblical

instances.) (e) The ha-BI-ru mercenary activity is paralleled by the

Hebrews in the Philistine army. (But this is a misinterpretation of the

biblical data.) (f) Both groups are in Egypt forced into the corvee.

(g) The ha-BI-ru are frequently strangers in the milieu and such are the

Hebrew patriarchs in Canaan. (h) Both groups deprive Egypt of its

holdings in Canaan by military operations during the Amarna Age.

    204 Cf. supra, WTJ XIX, pp. 9-11.

    205 Indeed, as A. Ungnad observes, "Bisweilen wird h fur 3 gebraucht"

(Grammatik des Akkadischen, 1949, p. 9).

    206 In the Canaanite glosses in the Tell el Amarna tablets are found, for

example: hu-ul-lu (EA 296:38) = lfo (cf. XXX) ; and hi-na-ia (EA 144:17) =

ynayfa (cf. XXXX). Cf. E. A. Speiser, Ethnic Movements in the Near East in

the Second Millennium B.C., 1933, p. 39.

HA-BI-RU                                          55


BI represents among other values that of pi as well as that

of bi in all periods of the cuneiform literature.

Further evidence is available, however, for in some cases

other signs of the cuneiform syllabary are used to write this

name and, moreover, the name has appeared in other systems

of writing, syllabic and alphabetic. From Ras Shamra207

comes the form 'prm written in the alphabetic cuneiform

common in texts from that site, in which the 'Ayin is distinct

from other gutturals and the b is distinct from p. This form

is, therefore, unambiguous. But the question has been raised

whether this form, in particular the second consonant, is

original or secondary. If the phonetic equivalence of 'prm

and 'Ibrim were to be maintained, the primacy of the p would

still he favored by the fact that Ugaritic often preserves a

more primitive Semitic form than does the Hebrew.208 On

the other hand there is evidence of an original b becoming p

in Ugaritic.209

In Egyptian hieroglyphics appears the form 'pr.w which

is also without ambiguity. But here again the question arises

as to whether the p is primary or secondary. It can be shown

that Egyptian p may represent foreign, including Semitic, b,

especially when the b is immediately preceded or followed by l


    207 Virolleaud, Syria 21, 1940, p. 132, pl. 8 and p. 134, pl. 10.

    208 So Kraeling, AJSL 58, 1941, pp. 237 ff. Cf. W. F. Albright, BASOR

77, 1940, pp. 32-3; DeVaux, RB 55, 1948, p. 342, n. 3. In an effort to

show that it is "quite possible that the isolated Ugaritic as well as the

Egyptian 'pr are secondary forms due to Hurrian influence" J. Lewy

observes that "the population of Ugarit included Hurrian elements and

that the Hurrians, wherever they appear, are responsible for a confusion in

the rendering of Semitic b and p because their scribes did not distinguish

between voiced and voiceless stops" (HUCA 15, 1940, p. 48, n. 7). C. H.

Gordon, however, informs me that the Ugaritic scribes who wrote the

tablets bearing 'prm carefully distinguish p and b. J. W. Jack (PEQ, 1940,

p. 101) attributes the Ugaritic spelling to Egyptian influence at Ugarit.

309 There are, e. g., the variants lbs/lps and nbk/npk. Cf. Greenberg,

op. cit., p. 90, n. 24. For evidence of confusion in Ugaritic between b

and p, and that in the very name ha-BI-ru, attention has been called to

the Ugaritic text 124:14, 15 (Gordon, Ugaritic Manual, 1955). Cf. Virol-

leaud, Syria XV, 1934, p. 317 n., and La Legende de Keret, 1936, p. 74;

and H. H. Rowley, From Joseph to Joshua, 1950, p. 50. Actually, the

text has nothing to do with the ha-BI-ru or with the Hebrews (as suggested

by Virolleaud).



or r.210  Such, however, is not the rule211, and, as Kraeling

observes,212 in the case of the 'pr.w, a people present in Egypt

itself, it is difficult to assume an error of hearing on the

part of the scribe.

The spelling ha-BIR-a-a is found twice in Babylonian

documents of the 12th and 11th centuries B.C.213 Commenting

on this form, B. Landsberger observes that "b nicht p als

mittlerer Radikal steht durch die Schreibung ha-bir-a-a (IV

R 34 Nr. 2, 5) fest".214 In signs, however, of the variety

consonant-vowel-consonant there is not only vocalic var-

iability but flexibility of both consonants within the limits of

their type.215


    210 For the evidence see B. Gunn apud Speiser, op. cit., p. 38, n. Cf. J. A.

Wilson, AJSL 49, 4, pp. 275 ff. W. F. Albright (JAOS 48, 1928, pp.

183 ff.) argues that the equation of Egyptian 'pr with 'eber is difficult

since Egyptian of the New Empire regularly transcribes Semitic b by

Egyptian b. As for Egyptian hrp for Can. harb (Heb. hereb), he says that

it only shows there was the same tendency for a final vowelless sonant

stop following a consonant to become voiceless that there is in the modern

Arabic dialect of Egypt; but the b in 'eber is medial and cannot have been

pronounced as a voiceless p. It should be noticed, however, that in some

instances of the use of Egyptian p for foreign b, the b is medial: thus,

isbr varies with ispr ("whip") and Kpn (O. K. Kbn) = Can. Gbl ("Byblos").

211 Gunn op. cit., p. 38, n.: "There are many cases (36 counted) in which

a foreign b with r or l either before or after it is represented by b and not

by p in the Egyptian writings". Wilson op. cit., pp. 275 ff. affirms that

the most straightforward equation is 'pr =rpf.

    212 Op. cit., pp. 237 ff.

    213 Rawlinson, Cuneiform Inscriptions of Western Asia, IV, 34:2, 5; and

Hilprecht, Old Babylonian Inscriptions, I, 2, pl. 66, no. 149, 22.

    214 ZA, N. F. 1, 1923, p. 214, n. 1.

    215 See the remarks of C. H. Gordon, Orientalia 19, 1950, pp. 91 ff. There

is specific evidence that BIR was used (though not commonly) for pir in

the neo-Assyrian period and possibly (the evidence is doubtful) in the

middle-Assyrian period. Cf. Von Soden, Das Akkadische Syllabar, 1948,

p. 73, no. 237. Bottero, op. cit., p. 132 urges against reading pir here the

absence of specific Babylonian evidence for this value to date, plus the

availability of the sign UD (pir). However, he acknowledges (p. 156)

that this form is not decisive for a root 'br. It may be additionally noted

that J. Lewy in defense of reading the second radical as b appeals to the

occurrence of the god "dHa-bi-ru in an Assyrian text (Keilschrifttexte aus

Assur verschiedenen Inhalts, no. 42), i. e., in a text in which ha-bi-ru can

hardly stand for *ha-pi-ru" (HUCA 15, 1940, p. 48, n. 7). Bottero (op. cit.,

p. 135) agrees on the grounds that in the neo-Assyrian era one normally

HA-BI-RU                                          57


By way of conclusion, there can be no doubt that the

Ugaritic and Egyptian forms of the name definitely require

that the consonant represented in the cuneiform syllable ha

be read as 'Ayin.216 They also strongly support an original p.

While there is a possibility that 'br is primary, it is highly

probable that 'pr is the original form. In fact, unless it can

be shown that ha-BI-ru is to be equated with the biblical

'Ibri there is no unquestionable evidence for 'br as even a

secondary form.217


2. Vowels. That the first vowel is A-type and the second

is I-type is obvious from the cuneiform, ha-BI-ru;211 but it is

more difficult to determine the length of these vowels. This

question requires examination before one attempts to draw

conclusions concerning the possibilities of phonetic equation

with 'Ibri.


used PI to signify pi. For evidence that BI = pi in all periods see Von

Soden, ibid., p. 53 no. 140. Also J. W. Jack states, "In the Hittite doc-

uments, for instance, habiru clearly has bi" (PEQ, 1940, p. 102). E.

Laroche (in Bottero, op. cit., p. 71, n. 2) argues, "D'apres le systeme en

usage a Boghazkoy, ha-bi-ri note une pronunciation habiri (sonore inter-

vocalique non geminee) ". But ha-ab-bi-ri appears twice. Moreover, P.

Sturtevant maintains that in cuneiform Hittite "the Akkadian distinction

between ... p and b did not exist", adding, "To all intents, therefore,

Hittite has dispensed with the means of writing b" (Comparative Grammar

of the Hittite Language, 1933, p. 66). Similarly, J. Friedrich, Hethitisches

Elementarbuch I, 1940,.p. 6(21). Accordingly, even the form ha-ab-bi-ri

(KBo V, 9, IV, 12) is quite ambiguous, as it would also be in Akkadian

cuneiform where AB stands in all periods for both ap and ab. Greenberg

(op. cit., p. 90, n. 20) suggests the possibility that a Hittite scribe utilized

a native convention, doubling the labial to indicate a sound heard by

him asp. Also ambiguous is the sign BAD (bi or pi) used in the Alishar text.

2,6 Cf. Bottero, op. cit., p. 154.

    217 Speiser (op. cit., p. 40), writing at a time when he did not have the

benefit of the Ugaritic evidence, begged the question of the phonetic

equation with 'Ibri in concluding, "The second consonant is ambiguous

both in cuneiform and in Egyptian, but not so in Hebrew: since the latter

has b, the labial must be read as voiced in cuneiform, while the voiceless

correspondent in the Egyptian form of the name is to be ascribed to local


    218 As far as it goes the Egyptian data is compatible. Gunn (op. cit.,

p. 38, n.) concludes from a survey of the evidence that "we seem to have

the alternatives 'apar, 'apir, 'apur, with a possible indication in" the

Beth-shan stele of Seti I "in favor of 'apir".



a. The A-Vowel: According to Gustavs,219 the form ha-

AB-BI-ri220 shows that the a is short. He explains the doubling

of the middle radical on the ground that consonants in

Akkadian are often doubled after an accented short vowel .221

This possibility, however, rests on the doubtful opinion that

the following I-vowel is short, for otherwise the penult would

receive the accent.222 Another possible explanation of the

doubling of the middle radical, although the phenomenon is

rare and late, is that it indicates that the preceding vowel is


Other unusual forms have appeared which suggest that the

A-vowel is long. One is ha-a-BI-ri-ia-as.224 Another is ha-

a-BI-i-ri-a[n?] (cf. ha-a-BI-i-ri-ia-an).225 Finally, from Alalah

comes the form ha-a'-BI-ru.226

b. The I-Vowel: Inasmuch as short unaccented vowels

between single consonants often drop out227 and the name


    219 ZAW, N. F. 3, 1926, pp. 28 f.

    220 KBo V, 9, IV, 12. Cf. also ha-AB-BI-ri-ia-an (KUB XXXV, 43,

III, 31).

    221 Cf. Ungnad, op. cit., p. 18 (6p); W. Von Soden, Grundriss der Ak-

kadischen Grammatik, 1952, p. 21 (20g).

   222 Cf. Von Soden, op. cit., p. 37 (38 f).

   223 Cf. Ungnad, op. cit., p. 7 (3d).

   224 HT 6, 18. This text is a variant of KUB IX, 34, IV. Greenberg

(op. cit., p. 90, n. 20) comments, "Were this writing not unique and not

in a word foreign to the Hittites it might have deserved consideration as

indicative of a participial form".

   225 KUB XXXI, 14 (XXXIV, 62), 10; and KUB XXXV, 49, I, 6 ff.

(cf. IV, 15).

   226 AT 58:29. E. A. Speiser (JAOS 74, 1954, p. 24) observes that the

main purpose of this unique form may be to indicate a form like *Habiru.

He suggests that even if the sign be given its value ah4 instead of a' the h

might be a graphic device signifying a long vowel or stressed syllable.

Cf. Greenberg (op. cit., p. 20): "Assuming that the scribe was West Semitic

he may have noted that his alephs became long vowels in Akkadian:

hence, by a sort of back analogy he may have converted what he took to

be a long vowel into an aleph". Wiseman (in Bottero, op. cit., p. 37)

"The word is unusually written ha-'a-bi-ru. This may be either a case of

HAR=AB4 or, as I am inclined to think, a case of the scribe erasing by

the three small horizontal strokes of the stylus".

    227 Cf. Ungnad, op. cit., pp. 12, 13 (5c). The possibility that the i is

short but accented is obviated by the fact that were it short, the antepenult

with its long a (as maintained above) would receive the accent.

HA-BI-RU                              59


ha-BI-ru is never found without the i, it would seem that

this i is long.228

Further support for this is found in the spelling ha-BI-i-ra229

used for the Nuzu personal name (assuming this name may be

identified with our ha-BI-ru). There are also the forms noted

above: ha-a-BI-i-ri-a[n?] and ha-a-BI-i-ri-ia-an.

c. Conclusion: The vocalization is largely a question of

how much weight to attach to the exceptional spellings.

Quite possibly they require two long vowels, producing the

(apparently non-Semitic) form, 'apir. Perhaps only one vowel

is long. It would be precarious, however, to assume that

every indication of a long vowel is misleading and to adopt

the form 'apir --or still less likely--'abir.

3. The Hebrew Equivalent. The difference in middle radicals

between ha-BI-ru (read as ha-pi-ru) and 'Ibri would not be

an insuperable obstacle for the phonetic equation of the two.

There are a few examples of a shift in Hebrew from p to b.230

Nevertheless, this shift is not the rule23l and the difference in

labials must be regarded as a serious difficulty in the case for


            If we allow the consonantal equation and examine the

vowels it will be found that the difficulties increase and the

equation can be regarded as at best a bare possibility. The

following are the possible vowel combinations of ha-BI-ru

(reading bi for the moment and listing the more probable

combinations first) along with their normal Hebrew gentilic

equivalents: 'abir, yriybiOf; 'abir, yriybifE; 'abir, yrib;Of; 'abir,

yribefE; and 'abr, yrib;fa.

Attempts have been made, however, to derive 'Ibri from

one or other of these vowel combinations. The most plausible

efforts are those which assume two short vowels, 'abir .232


    228 So C. H. Gordon (Orientalia 21, 1952, p. 382, n. 2) : "That the i is

long follows from the fact that it is not dropt to become *hapru".

    229 JEN 228:29.

    230 dpr-dbr, "drive"; parzillu, 511 ; dispu, wbd. Cf. W. F. Albright,

BASOR 77, 1940, p. 33; H. H. Rowley, PEQ, 1940, p. 92; DeVaux, RB

55, p. 342.

    231 Cf., e. g., rpAfa, rpefo,  rpAKo, rpAse, rpAxa.

    232 J. Lewy (op. cit.), assuming the form Habiru, suggests that it "is



Speiser suggests that "the form qitl may go back to an older

qatil" with the restriction that such forms derive from stative,

not transitive, verbs.233 In line with this, attention has been

called to the derivation of late Canaanite milk, "king", from

older malik, "prince”.234 "Whatever validity there may be in

the theory of a qatil to qitl shift,235 it must be remembered that

such is not the dominant tendency. Moreover, the degree of

plausibility in applying such a principle in the present case

is greatly diminished by the following considerations: a) The

combination of two short vowels ('abir) is one of the less

likely possibilities; b) The supposed shift from 'abir to 'ibr

did not occur according to our evidence in extra-biblical

documents either earlier than, or contemporary with, the

appearances of 'Ibri in the Bible. It is necessary to assume

that the shift took place first and only with the Hebrew

authors. And if we may not assume that the Hebrew form is

based on a previous shift to ‘ibr elsewhere, then proof is

required within the Hebrew language itself, and not merely,

for example, from inner-Canaanite developments, of a shift

from qatil to qitl.236


to rbAfe and yrib;fi as the Akkadian proper name Zakiru(m) [for references

see, e. g., A. T. Clay, Personal Names from Cuneiform Inscriptions of the

Cassite Period (New Haven, 1912) p.- 145] is to rkAze and  yrik;zi (Ex. 6:21,

etc.) ". There is, however, no evidence that the Hebrew form rkAze represents

the Akkadian Zakiru.

   233 Op. cit., p. 40, n. 96. Cf. T. J. Meek, Hebrew Origins, 1936, p. 7.

Similarly Bauer-Leander (Grammatik, 459), on the basis of a possible

relation of adjectival qatil and abstract qitl: e. g.., sapil-sipl, "base-


    234 So, e. g., Albright, Archaeology of Palestine and the Bible (New York,

1935), p. 206, and Bohl, Kanaander and Hebraer 1911, p. 85. In an earlier

article (JBL 43, 1924, pp. 389 ff.), Albright stated that Hebrew 'Eber for 'Ibr

stands by epenthesis for *'Apir, adding that the philological process is

familiar in all the Semitic languages; e. g., Arab. bi'sa from ba'isa. Cf. the

alternation of ma-si-ri and mi-is-ri in syllabic texts from Ugarit.

    235 DeVaux (op. cit.) goes to the extreme of describing the passing of

‘apir into 'ipr as "normal".

    236 The qatil type of noun does appear at times in Hebrew like a segholate;

cf. Gesenius, Hebrew Grammar, 1910, 93 hh, ii. Most of these are of the

getel-type which is usually the A-type but is sometimes the I-type (e. g.,

bcAq,,  rtAy,, fmaD,); but lz,Ge (Eccles. 5:7; Ezek. 18:18) is also found and that is

clearly I-type. This phenomenon is, however, confined to the construct

HA-BI-RU                                          61


Conclusion: The complete phonetic equation of ha-BI-ru

and ‘Ibri is at most a bare possibility. If a difference in

morphology were to be allowed while identity of denotation

was assumed the difference in the vowels could be explained237

and only the labial problem would remain as a phonetic

obstacle for the theory of common derivation. Even that

assumption, however, is implausible in dealing as we are

not with appellatives but proper names. The phonetic situa-

tion, therefore, is such as would weaken an otherwise strong

case for tracing Hebrew origins to the ha-BI-ru, not such as

to strengthen a theory already feeble.


C. Amarna Age Encounter.


In spite of the negative conclusions reached thus far the

investigation of ha-BI-ru--Hebrew relationships is not much

ado about nothing. For history apparently did witness an

ha-BI-ru--Hebrew encounter.

How is the ha-BI-ru activity in Palestine as reflected in

the Amarna letters to be integrated with the Israelite con-

quest of their promised land as described in the books of

Joshua and Judges? That is the question.


1. Conquest. The Amarna activity of the ha-BI-ru has

been identified by some with the Hebrew Conquest, more

specifically, with its first phase led by Joshua. But quite

apart from all the aforementioned obstacles to any identifica-

tion of the two groups, the Conquest under Joshua differed

from the Amarna military operations of the ha-BI-ru even in

broadest outline and fundamental character.

(a) The Hebrew conquerors were a people which had long

been in Egypt and were newly arrived in Canaan. The

Ugaritic and Alalah evidence reveals that the ha-BI-ru were


state. This restriction would not, of course, be significant so far as the

gentilic form yrib;fi is concerned. It becomes significant though when

account is taken of the derivation of yrib;fi from the patronymic rbAfa which

is found in the absolute state.

    237 Albright compares a development of gentilic ‘Ibri from an appellative

ha-BI-ru to Lewi, "Levite", probably derived from *lawiyu, "person

pledged for a debt or vow"; Qeni, "kenite", from qain, "smith"; or hopshi,

"free-man", from hupshu.



in Syria for a long while before the Hebrew Conquest (on any

view of its date). Moreover, since in Syria the ha-BI-ru had

long enjoyed permanent settlements of their own in well-

regulated, peace-time integration with the local population

and authorities, while the Amarna letters show the ha-BI-ru

in Palestine to be on the move, quartered here and there,

without absolute loyalty to any one party, it seems clear

that the Amarna ha-BI-ru were in Canaan as professional

militarists to exploit the anarchy there for their northern


(b) Also in conflict with this picture of the ha-BI-ru

operating in relatively small, detached companies and fighting

as mercenaries with no apparent national aspirations of their

own as ha-BI-ru is the biblical picture of the Hebrew Conquest

as an invasion by a united multitude, advancing in their own

name in a concerted effort to achieve a common national goal.

(c) The natives of Canaan were to the Israelites an enemy

to be exterminated; the acceptance of them as allies would

directly contravene Israel's purposes.238 But the ha-BI-ru

had no special antipathy for the Canaanites as such. Quite

the contrary, the Canaanites were their employers, and for

the most part the ha-BI-ru are found abetting the attempts

of those Canaanites who strove to gain independence from

Egyptian domination. Complaints are frequently heard from

the loyalists that Canaanite rebels are going over to the

cause of the SA-GAZ.

(d) The goal of Israel in Canaan with respect to the land

was to gain possession, and agreeably their general policy in

dealing with cities was to exterminate the population and

seize the spoil but to refrain from destroying the cities by fire.

The ha-BI-ru, however, after conquering and plundering,

frequently set the city on fire,239 apparently having no designs

to acquire territory or to build an empire.

The difference between the two movements can also be

traced in matters of detail.


    238 Cf. Josh. 11:19. Nothing underscores this more than the anomalous

character of the Gibeonite alliance. It should not be overlooked, however,

that after the days of Joshua's leadership the original determination gave

way frequently to a fraternizing attitude (e. g., Judg. 3:5-6).

   239 So repeatedly in EA 185.

HA-BI-RU                                          63


(a) Names: None of the names of the Israelite leaders

is found in the Amarna letters.240 Moreover, where the names

of the rulers of specific Canaanite cities can be checked (as at

Jerusalem, Lachish, Gezer, and Hazor) there is in every case

disagreement between the Bible and the Amarna texts.

(b) Numbers: In the pleas of the loyalists for military

assistance it appears that Egyptian support in the form of

fifty or so men will be adequate to turn the tide of battle. It

seems unlikely then that these Canaanite kings were con-

fronted with an assault on the scale of Joshua's army.241

(c) Places: The ha-BI-ru operated successfully in Phoenicia

and Syria, but neither the Conquest under Joshua nor later

tribal efforts penetrated that far.242

(d) Military Technology: The Israelites made no use of

chariotry,243 whereas chariots were a standard division of the

ha-BI-ru corps at Alalah and in Palestine.244


2. Pre-Conquest. An alternative must be found then to

identifying the biblical Conquest under Joshua with the

Amarna disclosures. The procedure of the majority of scholars

is to place Joshua after the Amarna events. Thus Meek,


    240 Proposals to equate Joshua with Yashuia and Benjamin with Benenima

(or Ben-elima) are phonetically impossible. Furthermore the Amarna

men were pro-Egyptian.

    241 Cf. Exod. 12:37; 38:26; Num. 1:46; 2:32; 26:51. At the same time it

should not be overlooked that even fifty professional soldiers might

provide adequate leadership to defend a walled garrison. Moreover, there

are larger requests like that of Rib-Addi (EA 71:23-24) for fifty pair of

horses and 200 infantry as a merely defensive measure.

    242 The way in which this argument is developed by Rowley (op. cit.,

pp. 42 ff.) is an illuminating exhibition of rewriting history to one's taste.

He argues that the exploits of Joshua were mainly if not entirely confined

to the central districts while the ha-BI-ru trouble was in the south and

north and only at Shechem in the center. It will be recognized that this

is the precise opposite of the prima facie biblical account, according to

which Joshua's campaigns were notably in the south (Josh. 10) and in

the north (Josh. 11:1-14). Rowley rejects Joshua 10 in favor of the

supposedly conflicting account in Judges 1; and Joshua 11, in favor of

the supposed variant in Judges 4. According to the record itself, Judges 1

records events after the death of Joshua and the events of Judges 4 fall

well over a century after those of Joshua 11.

   243 Cf., e. g., Josh. 11:9.

   244 Cf. EA 87:21; 197:2-11.



though he believes the Amarna ha-BI-ru and Joshua's cam-

paign belong to one movement, specifies that "the Amarna

account marks the beginning of the movement, while the

Old Testament account has to do largely with its final ac-

complishment".245 An odd quirk of Meek's view is that the

Exodus from Egypt under Moses follows Joshua by more

than a century.

Albright, though he posits an earlier, pre-Amarna exodus

from Egypt and entry into Canaan on the part of the Joseph

tribes and finds their presence in central Palestine before the

major Hebrew arrival reflected in the ha-BI-ru of the Amarna

letters, dates the (second) exodus (i. e., Moses leading out

the Leah tribes) and the campaigning of Joshua in the 13th

century, long after the Amarna correspondence.246

To cite one further variety of this approach, there is

Rowley's intricate reconstruction. He also espouses a theory

of a two-fold entry into the land, according to which certain

Hebrew groups, notably Judah, press northward from Kadesh

c. 1400 B.C. (these Rowley would identify with the ha-BI-ru

of the Amarna letters) while kindred tribes, including Asher,

Zebulon, and Dan, exert pressure in the north (these, Rowley

conjectures, are the SA-GAZ of the Amarna letters). But

the exodus from Egypt under Moses and the entry of Joshua

into central Palestine he dates late in the 13th century B. C.247

It will be observed that all these efforts to locate Joshua

after the Amarna episode involve drastic recasting of the

biblical data--the rejection not merely of points of detail

but of the biblical history in its basic structure. It requires

some ingenuity, indeed, to produce one of these elaborate

creations by weaving together a host of miscellaneous data

sublimated from their original contexts, but the result is

fiction not history. Under the mask of a claim of controlling

the biblical sources by means of archaeological and extra-

biblical sources an almost totally undisciplined biblical ex-

egesis has been introduced. But why the penchant for the

hasty rejection of the Old Testament source in favor of


   245 Op. cit.

   246 BASOR 58, 1935, pp. 10 ff.

   247 See Rowley, op. cit., esp. pp. 140 ff. for a survey of the various views.

HA-BI-RU                                          65


interpretations of archaeological evidence which are them-

selves so uncertain and disputed at countless points?


3. Post-Conquest. There is another alternative for the

integration of the Amarna and the biblical histories. It is

the reverse of those just surveyed in that it locates the Con-

quest under Joshua before rather than after the Amarna

letters, at least before those of Abdi-Hepa.248 This is in


   248 The historian is at this juncture always embroiled in the complex

question of the date of the Exodus. Aware of the difficulties of the early

date (i. e., locating Joshua in or before the Armarna Age) and not aware of

the proper solution of them all, the writer nevertheless finds insuperable

the difficulties of a later date. Relevant as the problem is, limitations of

space allow only brief comment on a few salient points: a) The case

presented by H. H. Rowley (in From Joseph to Joshua) against a Hebrew

entry into Egypt in the Hyksos period has not been answered. If valid,

that majority of scholars which is certainly correct in dating the patriarchal

period early in the second millennium B.C. rather than (with Rowley)

in the middle of it must date the beginning of the sojourn before the Hyksos

period, not (with Rowley) after it. And that, in turn, virtually necessitates

the early date of the Exodus. b) Advocates of a 19th dynasty Exodus

constantly appeal to the archaeological evidences of royal building opera-

tions at the sites of Pithom and Raamses. G. E. Wright, for a recent

example, states, "We now know that if there is any historical value at all

to the store-city tradition in Exodus (and there is no reason to doubt its

reliability), then Israelites must have been in Egypt at least during the early

part of the reign of Rameses II" (Biblical Archaeology (Philadelphia and

London, 1957), p. 60. Italics his.) That is a curiously misleading state-

ment. Is it not rather the case that, if one has no reason to doubt the

reliability of the record in Exodus 1:11 that Pharaoh forced the Israelites

to build Pithom and Raamses as store-cities, he cannot possibly identify

that pharaoh with Ramses II? For it is inconceivable that anyone should

have described the magnificent operations of Ramses II at these sites,

transforming one of them into the capital of Egypt, in the "store-cities"

terms of Exodus 1:11. The Hebrew building and the Hebrew Exodus

must then precede Ramses II. c) Albright has dated the destruction of

Canaanite Bethel, Lachish, and Debir, all by conflagration, in the 13th

century B.C., and would identify this destruction with Joshua's campaigns

as evidence of a late Exodus. Such a deduction does not do justice to the

biblical facts that Canaanite reoccupation frequently followed Joshua's

conquest of Canaanite cities and that destruction by fire was exceptional

in Joshua's campaigns. (Apparently only Jericho and Ai among the

southern cities were burned and only Hazor was burned in the Galilean

campaign. Josh. 11 .13.) The evidence of these Palestinian excavations,

therefore, actually requires a date for Joshua considerably earlier than the



precise agreement with the chronological data in Judges 11:26

and I Kings 6:1 and assumes a fairly brief period for Joshua's

campaigns which also agrees with the biblical record.249

Even more compatible with this view than with the iden-

tification of Joshua's campaigns and the Amarna activity are

certain facts which have long constituted a popular argument

in favor of the latter view.251 Giving it a somewhat different

turn than the advocates of identification, the argument is as

follows: Precisely those cities which appear in the Amarna

letters as under Canaanite control, whether pro-Egyptian or

rebel (and, therefore, likely allied to the SA-GAZ), are those

which were not permanently dispossessed either by Joshua251

or the early tribal efforts after the death of Joshua.252


13th century fall of these cities. A propos of Josh. 11:13, Yadin's recent

report of the second season of excavations at Hazor is of interest (cf.

Biblical Archaeologist, XX, 1957, pp. 34 ff.). In addition to the latest

Canaanite city which was destroyed in the 13th century (perhaps then,

according to an early Exodus, in the days of Deborah, cf. Judges 4 and 5),

remains were found of a 14th century city "approximately in the el-Amarna

period" (p. 44) and of an earlier city of the Middle Bronze Age which

"was effectively destroyed by fire, most probably by one of the Egyptian

pharaohs of the New Kingdom, Amenophis II or more probably Thut-

mose III" (p. 44). The supposition that a pharaoh of the New Kingdom

captured Hazor is questionable; for in spite of their many campaigns into

Canaan their ignorance of the techniques of siege warfare made the

capture of a fortified city a rarity. But according to the early date of the

Exodus, Joshua was a contemporary of Amenophis II and as for Hazor,

"that did Joshua burn".

   249 Josh. 14:7 and 10 indicate that the initial phase was completed

within five years of the entry into Canaan.

   250 Cf., e. g., Olmstead, History of Palestine and Syria (New York, 1931),

pp. 196-197; Meek, op. cit., p. 20.

   251 Joshua 10 and 11.

   252 The situation at Shechem is problematic. Nothing is said about an

Israelite conquest of central Palestine, but if the transaction of Joshua 24

implies Israelite control of Shechem, they subsequently lost their foothold,

for Labaya ruled Shechem some thirty years after the Israelite entry

(cf. EA 289:22 ff.). Similarly, if Albright (BASOR 87, 1942, p. 38) is

correct that Debir became the seat of a local chieftain after the Amarna

period, not only Joshua's raid but even Othniel's capture of that city

(Josh. 15:15-17; cf. Judg. 1:11 ff.) failed to be permanently effective.

Again, though Joshua's raid had depopulated Lachish and Gezer, these

cities fell again into Canaanite hands according to EA 287:14-15, whether

these lines mean that these cities had been assisting Pharaoh's enemies or

HA-BI-RU                                          67


Albright has concluded that in southern Palestine of the

Amarna period the main city-states were Gezer, Lachish,

Jerusalem, and Hebron-Keilah.253 In the period of Joshua

there are in this area five additional city-states: Jarmuth,

Makkedah, Libnah, Debir, and Eglon, with still others like

Jericho, Bethel and Gibeon nearby. Albright then theorizes

that from c. 1375-1250 there had been a gradual reduction in

the power of the city-states combined with an increase in

their number, which he attributes to a settled Egyptian policy

of divide et impera. This decrease in the power of the Cana-

anite city-states is then judged to have aided Israel in her

Conquest. Indeed, this is seized upon as compelling evidence

that the Hebrew Conquest was late.

It will be recognized that this reconstruction of the 14th

century situation in southern Palestine is based in part on

silences in the Amarna letters. Such a procedure is precarious,

however, for the silences might readily be accounted for by

the fact that the authors of the Amarna letters simply had no

occasion to mention the towns in question. To the extent,

however, that there may actually have been fewer city-states

in the Amarna period than in Joshua's day, a more plausible

explanation would be that between Joshua and the Amarna

situation the Israelites had been encroaching on the territory

of the old Canaanite city-states, reducing their number by


Furthermore, the spontaneous confederation of Canaanite

kings described in Joshua 10 is difficult to explain if it be

supposed that Joshua's campaigns were contemporary with

or subsequent to the ha-BI-ru activity of the Amarna letters.

For these letters graphically exhibit the mutual distrust and

growing antagonism among the Canaanite kings during this

period. Is it not apparent that neither in the midst of, nor

soon after, such intrigues and civil strife could a king of

Jerusalem so easily consolidate the surrounding city-states for


were to provide for Pharaoh's archers. Such developments indicate that

Israel's permanent acquisition of territory in Canaan was a gradual

process only initiated by Joshua's campaigns.

   253 Besides these, Jarmuth was a minor independency and an Egyptian

garrison and official were stationed at Eglon. BASOR 87, 1942, pp. 37-38.

Cf. Wright, op. cit., pp. 75, 76.



a joint military venture against a common foe? Abdi-Hepa's

futile efforts during the struggle with the ha-Bi-ru is a witness

that a king of Jerusalem would find such a task impossible.

Again a more plausible reconstruction is that the collapse of

the five-city alliance against Joshua terminated the southern

confederation and prepared for the Canaanite disunity ev-

idenced in the Amarna letters.

If Joshua is to be placed before the Amarna period, the

problem still remains of synchronizing the later Israelite tribal

efforts to take actual possession of their allotted inheritances

(i. e., the Book of judges) with the Amarna ha-BI-ru move-

ments. The arguments already presented against the pos-

sibility of identifying the ha-BI-ru with the Israelites of

Joshua's day for the most part hold against any such iden-

tification at this point as well. However, in view of the known

tendency of the authors of the Amarna letters to stigmatize

the cause of all enemies (or at least all accused of disloyalty

to Egypt) with the SA-GAZ label, we ought not to be too

dogmatic in denying the possibility that some Hebrew

activity might be hidden in the Amarna letters under that


More significant is the fact that on the chronology followed

here the first oppression of Israel in Canaan254 falls in the late

second and in the third decade of the 14th century B.C. This

corresponds with part of the era of the ha-BI-ru in Canaan.255

Israel's first oppressor was "Cushan-rishathaim king of Aram

Naharaim".256 The area designated by "Aram Naharaim"

would include within its southwestern limits the region about

Alalah (and probably still farther south) which was a strong

ha-BI-ru center in the 14th century B. C.257 Though styled


   254 Judg. 3:9-10.

   255 part of this era corresponds to the career of Labaya which can be

dated in the second and third decades of the 14th century on either

Albright's or Knudtzon's reading of the date on the hieratic docket on

Labaya's letter, EA 254.

   256 Judg. 3:8. It is possible that the additional MyitafAw;ri, "double wicked-

ness", was appended by Cushan's victims, perhaps as a pun on Myirahana Mraxa.

Cf. Burney, The Book of Judges, 1920, pp. 65-66.

    257 Cf. O'Callaghan, Aram Naharaim, esp. pp. 131-145; cf. p. 122.

HA-BI-RU                                          69


melek, Cushan-rishathaim need not have been more than one

strong chieftain among several in Aram Naharaim.251

Moreover, the name Cushan is attested in this area both as

the name of a geographical district and as a personal name.

That there was a district in northern Syria in the 13th and

12th centuries B.C. called Qusana-ruma, is known from the

list of Ramses III.259 Still more pertinent is the 15th century

tablet from Alalah260 which contains the personal name

ku-sa-an.261  This tablet is a fragment of a census list of

unspecified purpose, on which 43 personal names remain

along with the phrase found on the left edge, "owner of a

chariot". The list then might well be one of the numerous

military lists and probably includes the names of several


Within the framework of synchronization proposed here

for Hebrew and ha-BI-ru careers, it is difficult to dissociate the

oppression of Israel by Cushan-rishathaim from the ha-BI-ru

menace of the Amarna letters. The facts rather suggest that

elements of the ha-BI-ru corps from Syria active in southern

Canaan as the terror of the loyalist Canaanite city kings began

in time to raid the settlements of the more recently arrived

Israelites. The Israelites were becoming, like the Egyptians,

too dominating a power in Palestine to suit the interests which

the ha-BI-ru were engaged to further. It appears then that it

was from plundering ha-BI-ru mercenaries that Othniel

delivered oppressed Israel.262

If so, the ha-BI-ru, certainly not the kin of Israel, were

actually Israel's foe--the first oppressors of Israel in Canaan.

And then, far from offering a Canaanite version of the Hebrew


   258 Such is the usage elsewhere in judges. Thus Jabin of Hazor is called

"king of Canaan" (Judg. 4:2; cf. 4:23, 24), though he was but one of

several Canaanite kings (cf. Judg. 5:19). So also, O'Callaghan, op. cit.,

p. 123.

   259 Cf. W. Edgerton, J. Wilson, Historical Records of Ramesses III,

pl. 101, p. 110.

   260 Wiseman, AT 154.

   261 Ibid., p. 140. 36 names end in -an (ibid., p. 10).

   262 Since Othniel is associated with the south, this first oppression

probably centered there.



march of conquest, the Amarna letters dealing with the

ha-BI-ru are a Canaanite portrait of the first scourge employed

by Yahweh to chastise the Israelites for their failure to

prosecute the mandate of conquest.

It is not difficult to surmise what verdict the biblical

historians would have given if they had left to us their inter-

pretation of the data of the ha-BI-ru oppression of the

theocratic people in the early 14th century and the almost

total disappearance of the ha-BI-ru as a social-political entity

by about the close of that century. Surely they would have

judged that the brief Amarna Age encounter with Israel was

for the ha-BI-ru a crucial hour of more than ordinary political

decision. It was an encounter that sealed their destined fall.


Westminster Theological Seminary, Philadelphia



This material is cited with gracious permission from:

            Westminster Theological Seminary

            2960 W. Church Rd.

            Glenside, PA  19038


Please report any errors to Ted Hildebrandt at:  thildebrandt@gordon.edu