|How many men were in Israel's army?|
In my opinion the difference between the 800,000 Israelites in 2 Sam and the 1,100,000 Israelites in 1 Chr cannot be attributed to one figure being rounded off. The figures are already rounded off to the nearest 100,000. One can of course a priori deny that there might be different ways of reckoning the total. Or, one can do what historians do when they find divergent details for parallel accounts, which is to explore the possibilities and try to see if a reasonable reconciliation can be provided. If one takes the latter approach, which is a general principle followed by historians, then the chore becomes to find a way of harmonizing the accounts without resorting to means that stretch credulity and probability.
To be quite blunt, the burden of proof falls on the person claiming that there is necessarily only one way of reckoning things out. And, until solid evidence is unearthed for this position, such a person would be wise to avoid dogmatism regarding the claim. The student of Scripture who is faced with a discrepancy must play the responsible historian, presenting reasonable hypotheses that can possibly explain the divergent details. We first examine the divergent numbers for Israel.
In his Encyclopedia of Bible Difficulties Gleason Archer presents one reasonable (but of course not conclusively correct) way of harmonizing the 800,000 and the 1,100,000 numbers in the accounts.
A possible solution may be found along these lines. So far as Israel (ie, the tribes north of Judah) is concerned, the 1 Chr figure includes all the available men of fighting age, whether battle seasoned or not. But from 2 Sam 24 we learn that Joab's report gave a subtotal of "mighty men" ('ish hayil), ie, battle-seasoned troops, consisting of 800,000 veterans. But in addition there may have been 300,000 more men of military age who served in the reserves but had not yet been involved in field combat. These two contingents would make up a total of 1,100,000 men -- as 1 Chr reports them, with employing the term 'ish hayil. (pp. 188-9)
Archer's possible solution here is conjectural. Notice that he speaks in the subjunctive voice here: "A possible solution...", "But in addition there may be...". We cannot answer definitively if his solution is right or if his solution is wrong. Dogmatism either way is unwarranted and dishonest. But is his solution unreasonable? Does it fly in the face of common sense? Does it assume many improbable things to make its case? I certainly don't think so.
His argument hangs on the fact that 2 Sam includes the descriptive Hebrew 'ish hayil whereas the Chronicler does not use the term. In turn this rests on the part of whether one can reasonably in the light of Hebrew idiom press an intended distinction here between the usage of the term in 2 Sam and the absence of it in 1 Chr. Note how the KJV brings out the distinction with the word "valiant" which appears in 2 Sam but not 1 Chr. The KJV is more faithful to the Hebrew here.
Until solid evidence can be presented against the noting of the idiomatic 'ish hayil being present in 2 Sam and absent in 1 Chr, plausibly suggesting that the 800,000 in 2 Sam is but a (large) subset of the 1.1 million mentioned in 1 Chr, cannot be confuted as a reasonable possibility for harmonization.
There is another possibility that will be reasonable after examination. The reader should re-read 1 Chr 27. Notice here that there are 12 divisions of 24,000 men each, giving a total of 288,000 men. It is possible that the Chronicler counts these men whereas the author of 2 Sam does not. Notice that the 800,000 men in 2 Sam were included in a census, as David wanted to know how many men there were for fighting. Yet, as the numbers of divisions were apparently fixed at 24,000 per division, one would presumably not need to take a census of groups whose sizes are intrinsically defined by a priori fixed numbers.
It is not requiring too much to state that it is reasonably possible that the author of 2 Sam did not include these 288,000 while the (different) author of 1 Chr did. With two different authors writing apart from each other at non-identical times, it is not at all specious to assert a reasonable plausibility to a different mode of reckoning in reporting the census. Thus my suggested reconciliation at the very least cannot be called credulous, and at the best it can be called quite reasonable. Yet, the constant refrain should be sounded again: dogmatism on both sides is to be avoided here. We don't have all the facts -- in fact we are missing a great deal of them.
This fact of the absence of all the facts should keep us humble. Apologists can not be overly zealous in claiming that the divergent numbers have in fact been reconciled, but neither can critics of the Bible be overly zealous in claiming that there is a necessary contradiction here.
In my attempted reconciliation, then, if we add the 288,000 men to the 800,000 in 2 Sam, we arrive at 1,088,000 men, which is quite close to the round number of 1,100,000 which the Chronicler presents. Unless a text explicitly states that exact numbers are being conveyed, it is most unreasonable to still claim discrepancy here. A responsible historian, recognizing that 1,100,000 is a number that is a rounded number, will be satisfied with the proximity of this number to 1,088,000. Those who still wish to press the argument for an error need to prove that the text intended in both 2 Sam and 1 Chr to convey an exact figure. (See our essay on copyist errors and estimations.)
The relative difference between the numbers 1,100,000 and 1,088,000 is but a scant (1,100,000-1,088,000)/(1,088,000)=.01, that is, there is only a 1% difference in the two figures, and it is clear that one of the figures (1,100,000) is rounded. For those who still insist on arguing the point, it should be pointed out that such continuing argument is really a refusal to face the harmonizing nature of my suggested reconciliation, for the propensity to round sizable figures off is a quite common idiom.
Now we examine the difference between the tallies of Judah found in 2 Sam (500,000) and 1 Chr (470,000). P. 37 of Arndt's Does the Bible Contradict Itself? says: "Evidently the account in 1 Chr is more exact than the other. The writer of 2 Sam contents himself with stating the number of warriors in round figures. Here, then, there is no discrepancy." Arndt's solution is quite reasonable as 500,000 is a rounded figure -- it is most unlikely that a count of men in Judah should yield an exact figure of 500,000.
In fact, the relative difference between 470,000 and 500,000 is (500-470)/470=.06, a mere six percent difference, quite small for those who don't force an exacting idiom of numerical reporting on the text, which itself makes no such claim to total and absolute precision. A reasonable person should recognize Arndt's solution as a probable and satisfying one, for it is not unreasonable to believe that the Chronicler rounded off to the nearest ten-thousand, whereas the author of 2 Sam rounded off more crudely, to the nearest hundred-thousand.
Another solution that is reasonable but different from the above is that given by Archer:
So far as Judah was concerned, 2 Sam 24 gives the round figure of 500,000, which was 30,000 more than the corresponding item in 1 Chr 21. Now it should be observed that 1 Chr 21:6 makes it clear that Joab did not complete the numbering, for he did not get around to a census of the tribe of Benjamin (nor that of Levi, either) before David came under conviction about completing the census at all. Joab was glad to desist when he saw the king's change of heart. The procedure for conducting the census had been to start with the Transjordanian tribes (2 Sam 24:5) and then shift to the northernmost tribe of Dan and work southward back toward Jerusalem (v. 7). This meant that the numbering of Benjamin would have come fast. Hence Benjamin was not included with the total for Israel or that for Judah, either.Again, it must be noted that such a solution is conjectural, but there is nothing here that stretches credibility and requires sizable leaps of faith.
Thus two reasonable harmonizations of the divergent numbers have been given. (For a related issue, see here.)