The Cities of Refuge
The purpose of this page is to consider the Cities of Refuge identified in Joshua 20 from several points of view. The main part is a geographical study of the cities named. This suggests a composition date in the Middle or Late Bronze Age, and certainly before the division of the monarchy after Solomon's death. The different geographical extent of land said to be under Israelite control renders the suggestion that the Joshua and Pentateuchal accounts were composed late in the divided monarchy period improbable.
The principle of having specific locations at which a person wrongly accused of murder could seek refuge was established in Exodus 21:12,13. "Anyone who strikes a man and kills him shall surely be put to death. However, if he does not do it intentionally, but God lets it happen, he is to flee to a place I will designate."
Numbers 35:6,11-15 identifies 6 such cities amongst a total of 48 to be used for the support of the Levites. Three of these were to be in Canaan and three in trans-Jordan, but the locations are still not specified. "Six of the towns you give to the Levites will be cities of refuge, to which a person who has killed someone may flee... select some towns to be your cities of refuge, to which a person who has killed someone accidentally may flee... Give three on this side of the Jordan and three in Canaan as cities of refuge..."
Deuteronomy 4:41-43 records Moses identifying the towns to the east of the Jordan. At this stage the various rulers in transJordan had been defeated, and Israel was poised ready to cross into Canaan. Hence it is sensible for Moses to name this group but not the Canaanite cities. "Then Moses set aside three cities east of the Jordan... Bezer... Ramoth... Golan..."
Deuteronomy 19:1-13 repeats the concept, and includes more details of the practical administration involved (for example, the elders of the person's home town were to be involved in the decision as to his guilt or innocence). Finally, in Joshua 20 all six cities are named. This takes place when the initial conquest had been completed, and the division of the land between the tribes was being confirmed. "Tell the Israelites to designate the cities of refuge... So they set apart Kedesh... Shechem... Kiriath Arba (that is, Hebron)... On the east side of the Jordan... Bezer... Ramoth... Golan..."
The locations of the rest of the Levitical support-cities were decided at the same time (Joshua 21). The allocations are repeated in 1 Chronicles 6.
The places identified in Joshua 20 are as follows:
The cities in Canaan are given as:
The trans-Jordan cities are:
- Kedesh in Galilee in the hill country of Naphtali,
- Shechem in the hill country of Ephraim, and
- Kiriath Arba (Hebron) in the hill country of Judah.
- Bezer in the desert on the plateau in the tribe of Reuben (the town is near Heshbon, to the north-east),
- Ramoth in Gilead in the tribe of Gad, and
- Golan in Bashan in the tribe of Manasseh.
It is important to ascertain what periods of time it would be realistic for these places to be chosen for the Israelites. Were they, for example, places which the Israelites had secured during the early entrance into the land? At the opposite extreme, at what point were they lost to the Israelites to other nations? The following analysis shows key stages in the territorial changes for Israel.
The transJordan area was disputed by a variety of nations at different times. At the time of conquest, Moab and Edom were left untouched by the Israelites. Numbers 21 and other related verses speak of two Amorite kings - Og of Bashan, ruling from Ashtaroth, and Sihon, reigning over Gilead from Heshbon. They controlled the area immediately to the east of the Jordan River. Moab had recently lost territory up to the Arnon river to Sihon, so the area on which Israel camped, though called the plains of Moab, was under the control of Sihon. The Ammonites controlled a self-contained area to the east of Gilead.
|Before the Conquest||After the Conquest|
The battles fought by Israel in transJordan were against first Sihon (near Heshbon), then after a northward march Og (at Edrei). This resulted in the complete capture of Gilead and Bashan, but only up to the Ammonite border which was fortified. The southern border was the Arnon River, and to the north the region extended to Mount Hermon. The three tribes who elected to remain in this area (Gad, Reuben and part of Manasseh) rebuilt cities including Dibon and Heshbon.
Judges 3 recounts how Ehud led the Israelites while Eglon king of Moab had power over them. Moab was supported by Amalekites and Ammonites, but the territory concerned was in the area of Jericho. Presumably the alliance was raiding along either side of the Dead Sea. This would affect the Reubenite portion of the land and perhaps Bezer, but there is no indication that areas of land were actually lost to the invaders. Indeed, Jephthah's speech in Judges 10 suggests the reverse. Perhaps the Moabite alliance simply sent in raiders to claim tribute without any attempt to seize land. The oppression tackled by Deborah and Barak was centred in the north. Kedesh would be most strongly affected, but since it is reported as Barak's home town it must be assumed that once again the oppressors did not seek to claim land and expel the Israelites living there. Gideon, from the region of Manasseh west of the Jordan, fought against invaders from Midian. Once again, although the Midianites roamed freely through the land, the Israelites do not seem to have been moved away from their homes.
|Judges 3 to 5||Judges 6 to 8||Judges 10 to 11|
Judges 10 relates how the king of Ammon oppressed the Israelites in Gilead, and sent raiding parties west of the Jordan into the territories of Judah, Benjamin and Ephraim. Judges 11 indicates that - yet again - this oppression did not prevent the Israelites living there, though their life was evidently harsh. Jephthah - a Gileadite - contests the Ammonite claim that Gilead was traditionally theirs (Jdg. 11:13), on the grounds that Israel had captured it from Sihon and had dwelt in the area for 300 years. This strongly indicates that the transJordan territories of Israel had retained their essential shape in the period between the conquest and the time of Jephthah. His war against the Ammonites was successful in driving them back.
The remainder of Judges is concerned either with the Philistines on the western border, or with internal disputes amongst the tribes of Israel. Because of this, the territorial developments in transJordan is less clear. However, to summarise the above, during Judges up to the end of chapter 11 (Jephthah), all of the cities of refuge remained occupied by Israelites. The least secure of the cities was Bezer, close to Heshbon, as this area was overrun on several occasions by other neighbouring countries dominating the land.
With the books of Samuel more details emerge again. Saul's area of influence was centred on the Jordan valley but strong in Gilead - for example the people of Jabesh Gilead took considerable trouble to honour his body after death. To the north and west he struggled continually with Canaanite and Philistine kings. The Canaanites had two major spurs of territory jutting into the area under his control, one around Megiddo and the other including Jerusalem. The Battle of Gilboa (at which he and Jonathan died) can be seen as a failed attempt to secure the northern border. If the placenames listed through the books of Samuel are investigated, we find that since the time of Jephthah there had been a considerable contraction of territory. The western border had been pushed back from the sea. Parts of the north were controlled by others. The main change in transJordan was the area formerly controlled by Reuben. Moab had annexed the lands which had been theirs before first the Ammonites and then the Israelites had taken them, and now controlled the region north of the Arnon Gorge to the east of the Dead Sea. Dibon and Heshbon - and so Bezer - were no longer in Israelite hands. Moabite Mizpah, where David sent his parents there for shelter with the permission of the Moabite king (1 Sam. 22) was in this area. The Aramaeans appear to have expanded back into Bashan, and there is no suggestion that the area to the east of Galilee was still considered Israelite land. Thus Golan was most likely lost at this stage. At this time, then, Bezer and probably Golan would not be available to be chosen as Cities of Refuge, and Kedesh was in somewhat of an isolated position, largely cut off from the rest of Israel by the strong city-state of Hazor.
|Saul||Early David||Late David &|
The early part of David's reign was based in the south, at Hebron. Ish-bosheth, Saul's son, was reigning as his successor and controlled Gilead, Ephraim, and thence north towards Galilee. During this period, David and the prototype Judah controlled only one of the Cities of Refuge. Three were in Ish-bosheth's territory, and the remaining two outside Israelite land. Once David's western border was secure against the Philistines, and he had ensured internal stability after Ish-bosheth's death, he turned his attention east. 2 Samuel 8 describes how the transJordan areas of Edom, Moab, Ammon and Aram came under his control. It may be assumed that Gilead and Bashan were also secure at this time. Solomon continued to exert firm control over the whole of transJordan. 1 Kings 4 lists his district governors. Three of these were in transJordan, though only in the central and northern parts of Bashan and Gilead. Moab is not mentioned, so conceivably it had the status more of an occupied territory rather than part of the homeland. So even though all six towns were under Israelite control until Solomon's death, it is not clear that Bezer would be considered a suitable location for a City of Refuge if the choice was being made at that time.
However, the troubles of the divided monarchy resulted in unsettled times. Regarding the Cities of Refuge west of the Jordan, Hebron was in the territory of Judah and the others in Israel. After the northern kingdom was destroyed and largely taken into exile by the Assyrians (722 BCE), Hezekiah appears to have taken over parts of the central area of land immediately to the north of Judah. However, the northern parts of Israel - including Kedesh - were lost.
The entire transJordan area became debatable and from time to time large parts of this were lost. In the north, the Aramaeans under Rezon began to show hostility during Solomon's reign (1 Kings 11:23-25, c. 975-935 BCE). Asa of Judah (1 Kings 15:18-20, c. 900) sent tribute to Aram to persuade Ben-Hadad to attack Israel. This led to various towns in the vicinity of Galilee (the tribal areas of Dan and Naphtali) being taken. Ahab of Israel (1 Kings 20, c. 875-855) retaliated, but evidently with limited success as Asa's son Jehosaphat (c. 875-850) and Ahab attacked again as a joint venture (1 Kings 22), having lamented that Ramoth-Gilead was still in Aramaean hands. Ahab died in this venture. So this city must have been captured from Israel after Solomon's death in c. 935 and before Jehosaphat's accession in c. 875. According to Josephus, Omri had taken it back from Ben-Hadad I which would place the original loss to Aram before 890. The loss of the Golan area must have been prior to this. Just over a decade later another military alliance between Israel and Judah tried again to take the city. By this time Hazael had taken the throne of Aram. Ahaziah of Judah, grandson of Jehosaphat, and Joram of Israel, son of Ahab, attacked (2 Kings 8:28-29). Joram was wounded and went to Jezreel to recuperate, but 2 Kings 9:14 indicates that the city had been recaptured (Josephus suggests that Jehu in fact took the city after Joram's departure). Jehu, the army commander, conspired against Joram, seized power and killed both Joram and Ahaziah. 2 Kings 10.32-33 indicates that transJordan was lost in or shortly after Jehu's reign (c. 840-815). This situation continued until Jehoash and Jeroboam II of Israel (2 Kings 13, c.800-785 and 2 Kings 14, c.785-745) recovered parts of the land captured by the Aramaeans. Within a few years, however (Pekah, 2 Kings 15, c.735-730) the Assyrians under Tiglath-pileser III (c. 745-725) overran Syria, the northern part of Israel around Galilee, and transJordan.
|900-850 BCE||850-800 BCE||800-700 BCE|
Turning now to the southern parts of transJordan, Moab rebelled against Israel after the death of Ahab (2 Kings 1.1, c. 850 BCE). Joram of Israel (c. 855-840) and Jehosaphat of Judah (c. 875-850) set out to put down the rebellion, led by Mesha king of Moab (2 Kings 3). Although their attack is recorded as a success, they were forced to withdraw to their own land because of the vehemence of feeling against them. The Moabite (or Mesha) stone (dated before 830 BCE) confirms that Mesha had indeed taken this region for his own. The region was debated over the years after this, and many of the points noted above in connection with Gilead apply here. The key difference is that Moab started this period as an occupied rather than Israelite territory. Isaiah (c.760-690, ch. 15) and Jeremiah (c. 645-575, ch. 48) both regard Dibon and Heshbon as Moabite rather than Israelite.
To summarise the Biblical data then, it is only in the period from the conquest to Jephthah that all six Cities of Refuge were firmly in the hands of the Israelites. Their various fortunes are summarised in the table below. Another interesting point concerns the towns chosen. Hebron, Shechem, and Ramoth Gilead were all key towns throughout the history of Israel. The other three, as events transpired, were quite minor places. Golan and Bezer are not mentioned after their original apportionment, and Kedesh is only referred to in one other place, in 2 Kings 15 when Tiglath-pileser captures the region. Bezer and Kedesh are both close to other much larger towns - Heshbon and Hazor respectively - which if the selection had been made at a later date might have seemed far more obvious choices. So the cities named strongly support the idea that the composition of the Pentateuch and Joshua narratives were early. Some of the places picked were not in Israelite hands for periods of time, and in some cases other nearby towns became much more prominent.
|1075|| || ||
|| || || |
||Israelite occupied||Israelite||Israelite occupied|
||Israelite occupied||Israelite||Israelite occupied|
||Israelite occupied||Israelite||Israelite occupied|
In this table, a red entry
indicates that the town in question was in the hands of a nation hostile to Israel. A black entry
indicates that, though the town was under Israelite control, it seems unlikely that the place would be considered a suitable location for a City of Refuge. The most probable time is evidently in the time between the conquest and Jephthah.
This section reviews other sources of evidence to see to what extent the above picture is mirrored outside the Old Testament.
Assyrian texts from the time of Ashur-rabi II (conventionally 1012-972, and the various NC models usually place him a few years earlier) indicate that a king of Aram gained control of Pethor and Mutleinu either side of the Euphrates for a short time. This may reflect David's clashes with Hadadezer son of Rehob, king of Zobah, since he is reported to have extended his kingdom to the Euphrates to subdue Toi king of Hamath. David's treaty with Toi led to the defeat of Hadadezer.
A little later, Hezion seized Damascus (c. 925-915?), followed by his son Tabrimmon (915-900?) and grandson Ben-Hadad (900-860?). The sequence is listed on the Melqat Stele, though the ancestry of Ben-Hadad is difficult to read and sometimes disputed. Two possible readings are "The monument which Bar-Hadad, son of Tab-Rimmon son of Hadyan, king of Aram, set up for his lord, Melqart
", or alternatively "The statue which Bar-Hadad, son of Ezra, the king, the Rehobite, king of Aram, set up for his lord Melkart
". In this latter case the stele is taken to refer David's adversary - the dating of the stele is not sufficiently accurate to discriminate between the two.
The Ben-Hadad who (in the OT) fought Ahab and was murdered by Hazael is probably a second of that name, though some people argue for a very long reign (ended 843). The records of Shalmaneser III (conventionally and NC 858-824) list several battles against Adad-Idri of Damascus in 853, 849, 848 and 845. The names Adad-Idri and Ben-Hadad are sufficiently similar there is no real reason for doubting these are the same individual. Adad-Idri was one of the main opponents of Assyria at the battle of Qarqar (853), together with Urhileni of Hamath - Ahab of Israel also supplied troops.
From 841, Shalmaneser names his opponent in Damascus as Hazael, calling him the "son of a nobody
" and alluding to the murder of his predecessor. They fought again in 837, but Hazael was evidently a sufficiently strong and difficult enemy that there is then a 30 year gap in the Assyrian assault. Hazael was evidently left on his own by former allies to face Assyria, which may explain his subsequent bitter and unrelenting assault on Israel. The next
Assyrian attack was by Adad-nirari III (conventionally and in NC 810-783) who refers to Hazael as "Mari" (Aramaic for "lord"). 2 Kings 13.5 refers to a "deliverer sent by God
" and this may refer to Adad-nirari III. Assyrian spoil from Damascus from this era includes 2 pieces of ivory inscribed with "belonging to our lord Hazael
". According to the Zakur stele, Hazaelís son Ben-Hadad acceded c. 796 and reigned until c. 770. Zakur usurped the throne of Hamath and together with allies then defeated Ben-Hadad. For a time, Jeroboam II seems to have exerted control over Damascus, until the appearance of Rezin (perhaps after Jeroboamís death in 753). Tiglath-pileser III (conventionally and NC 744-727) lists Rezin as giving tribute in 738, but in 732 he defeated and slew Rezin and deported the people to Qir. According to the OT this was at the request of Ahaz of Judah (c.737-722)
In summary, the Assyrian records of the time between David and Ahaz support
the general impression given by the OT. The same rulers are attested, as is their
general character as being militarily strong and competent.
The chief source here is the Moabite Stone, usually dated a little before 830 BCE. This refers to the triumph of Mesha ben Chemosh, king of Moab, and tells of how he threw off the yoke of Israel and thus honoured his god Chemosh. Part of the text reads:
"As for Omri King of Israel, he humbled Moab many years [literally 'days'], for Chemosh was angry at his land. Ans his son followed him and he also said 'I will humble Moab'. In my time he spoke but I have triumphed over him and over his house, while Israel has perished for ever! Now Omri had occupied the land of Medeba and [Israel] had dwelt there in his time and helf the time of his son [Ahab], 40 years, but Chemosh dwelt there in my time.
This suggests that freedom had been attained prior to Ahab's death, which differs from the claim in 2 Kings 1.1 that Moab rebelled after Ahab's death. However, Ahab was struggling with the Aramaeans in the later years of his reign and most likely did lose effective control over Moab. Mesha would naturally mark this as the start of his freedom, whereas Israel did not reckon it until the failed attempt of Joram to reassert control. The Stone continues with a description of numerous towns having been built or rebuilt, "I have built Aroer... I have built Bezer, for it lay in ruins... And I have built Beth-Medeba...
". Bezer had clearly declined in importance by Mesha's time. About Ataroth he says, "And the men of Gad lived in the land of Atarot from ancient times; and the king of Israel built Atarot for himself, and I fought against the city and captured it. And I killed all the people of the city as a sacrifice for Kemosh and for Moab.
Again, this account supports the OT account, but has been written from the opposite perspective. Israel lost control over Moab, and in particular lost the area round Bezer, eaely in the divided monarchy.
The Amarna letters
The Amarna letters provide an illuminating insight into Canaan. The conventional and New Chronologies differ considerably in their historical placement of the Amarna period. Conventionally it is in the mid-fourteenth century BCE, say 1350-1330, in the early part of Judges. In NC terms it is rather later, from about 1020-1000, in the early united monarchy period. The picture of Canaan given in these letters is one of individual city states, all nominally loyal to Egypt but in practice with some intent on establishing their own power-base at the expense of others. This is dealt with in more detail elsewhere. There are several main groups of protagonists. In the coastal plain and parts of the Jezreel valley are rulers of cities who have Indo-European names - Shuwardata of Gath, Widia of Ashkelon, and Biridiya of Megiddo, for example. Gath is ruled by Canaanites with Semitic names - Milkilu, then later Yapuhu and Addadanu. Jerusalem is independent, ruled by Abdi-Heba, Heba being the name of a Hurrian goddess. To the north, Syria (Amurru) is dominated by aggressive rulers Abdiashirta and his son Aziru. Towards the Mediterranean coast west of this, Hazor and Acco are also independent and strong city-states. The central hill country area is increasingly dominated by Labayu, who despite claims of loyalty to Pharaoh is clearly acting in his own interests. The language he writes in, though using Akkadian characters, is laced with Canaanite expressions with a strong degree of similarity to Ugaritic. TransJordan does not feature in the Amarna archive - the furthest east location being Pella, in the Jordan valley - but the general picture of the coastal region and north supports the OT picture of these regions in late Judges and the books of Samuel.
Other data sources - Egyptian campaign lists, Under preparation.
Conclusions under preparation.