It should be noted that the numbers given in the census at the beginning of Numbers are also disputed (which would affect the numbers available for Midian's army here).
The most probable solution at this point is to understand that the
numbers given here are mixtures. Since the Hebrew word translated
“thousand” (‘lp) looks the same as the word translated “military
division,” a number like 74,600 (v. 4) may be read as 74 military
divisions, (totaling) 600 men. The total in verse 32 would originally
have been written 598 military divisions (‘lp), 5 thousand (‘lp) and 5
hundred men. But at some point in the transmission of the text the two
words were confused and added together to make 603 thousand. If this
solution is correct, the size of the Israelite group that left Egypt
would have been about 20,000.
Victor Harold Matthews, Mark W. Chavalas and John H. Walton, The IVP
Bible Background Commentary: Old Testament, electronic ed. (Downers
Grove, IL: InterVarsity Press, 2000), Nu 2:32.
According to R. Dennis Cole,
Numbers 31 is said to be dependent on Deuteronomy 20, Judges 8, and 1
Samuel 30. Yet, as Wenham and Ashley have so astutely pointed out,
none of these arguments is conclusive; and in fact if one hypothesizes
that the chapter derives from the Mosaic era, the three passages above
derive from Numbers 25 and 31 sequentially and logically. In a general
sense Gray may have been correct in describing Numbers 31 as midrashic
(a literary tool for teaching certain lessons or principles) in its
literary structure, but that need not imply that it should be
associated with midrashic exegesis of the late postexilic era (second
century B.C. and later). The basis of the passage is material that
should be taken as historically reliable, which was fashioned into a
literary composition that had the function of providing case law
precedents for future holy war endeavors for this generation that
would enter the land and for generations to come. J. Milgrom concluded
that “the assembled evidence clearly points to the historic reality
that Midian was the most powerful and menacing enemy that Israel had
to encounter during its migration into Canaan.”
Critics have observed
that several components in the account seem incredulous, including the
annihilation of all the males of Midian, the enormous numbers of
various animals seized in the plunder, and that none was missing or
lost from Israel’s battalions. G. Wenham and Ashley have provided
answers to these and other questions regarding the content and
character of the narrative. First, it is an overstatement of the data
in the narrative to suggest that the report of vv. 7–8, that
Israelites “killed all the males of Midian” including the kings (or
tribal chiefs), implies that every male of every Midianite tribe from
the Transjordan to Arabia to Sinai was exterminated in this one
campaign. Obviously this was not the case since the Midianites are
well attested in the biblical and ancient Near Eastern texts. Taken in
the historical context of this being a divinely directed follow-up
campaign after the sinful Baal Peor incident (25:16–18; 31:3–8), this
crusade was directed at the tribes or clans of Midianites who dwelled
in the central and northern Transjordan highlands, in the vicinity of
the lands of the Moabites, Ammonites, and Amorites. The Midianites of
the southern regions, such as those of Moses in-laws, were on better
terms with the Israelites or were not involved on this occasion.
Second, the large numbers of animals taken as spoils of war seem
incredulous. The totals are much higher by comparison with those
confiscated in the campaign of Thutmose III of Egypt ca. 1460 B.C.
during his campaign against Megiddo and other northern Canaanite
cities. The Karnak temple account lists booty of 1,929 cattle, 2,000
goats, 20,500 sheep, and 2,503 slaves (men, women, and children),
along with a variety of physical objects such as gold bowls and ebony
statues. G. Wenham suggests an adjustment should be made to the
numbers by analogy with the two census summaries of 1:1–46 and
26:1–51. Taking the alternative meaning of the word for thousand
(ʾelep), that of “clans” or “battalions of troops,” as perhaps herds
or flocks in this analogy, the totals of the animals might be
interpreted as 67,500 or even fewer sheep (vs. 675,00 in v. 32), 3,600
or fewer cattle taken by the men of war as their share (vs. 36,000 in
v. 38), et cetera. Yet within the text there is consistency in the
resultant numbers of animals and persons provided as gifts to the Lord
(and hence the priests) on the basis of the 1:500 ratio delineated in
v. 28. Of 32,000 persons captured, 32 (1/500 of the warriors’
half=16,000) were presented as the Lord’s tribute (v. 40).
The
testimony that not a man was missing from those who went out to war
against the Midianites (v. 49) seems exaggerated on the surface, yet
such claims are not unknown from the Bible or the texts of the ancient
Near East. The account of Gideon’s night raid against the Midianites
suggests that all three hundred men survived the initial confrontation
and continued their efforts by pursuing their enemies down into the
Jordan Valley and beyond (Judg 7:7, 16, 19–22; 8:4). The Persian King
Cyrus, for example, who captured the city of Babylon only after
conquering the rest of the Babylonian kingdom, claims to have captured
this seemingly impregnable city “without any battle” in 539/538 B.C.
The victory over the Midianites was a remarkable one indeed, but with
the providential direction and protection of the armies of Israel,
such was definitely not out of the realm of possibility.
R. Dennis Cole, vol. 3B, Numbers, The New American Commentary
(Nashville: Broadman & Holman Publishers, 2000), 490-92.
Thus it is possible that the numbers are incorrect. They have either been confounded, embellished, or the meanings of some of the words have been lost or changed. This is not a satisfactory position for some, but I'm sure someone else will chime in with an alternate one ;)