I had a question regarding Numbers 3. Why does there seem to be an
inconsistency between the sum of the males counted in the Levite clans
(7,500+8,600+6,200=22,300) in ch3, vs. 21-34, and the amount listed in
ch3, v. 39 (22,000)? It would seem to pose a problem when (in vs. 40-51)
the people of Israel are to “redeem the 273 firstborn sons of Israel who
are in excess of the number of Levites,” with five silver pieces for each.
Wouldn’t the sum from the first count put the number of Levite males above
the firstborn Israelites (22,273) by 27? I’m aware, as the footnote in my
bible mentions, that “3:28 Some Greek manuscripts read 8,300; see total in
3:39.”. However, this doesn’t explain why the translators chose to include
the “8,600” rather than the number that would have helped v.39 add up, so
to speak. Do you know of the manuscript evidence for choosing one number
over the other, even though it seems to not add up? I did a couple of
searches over the Web using bible commentaries but haven’t come upon
anything that deals directly with these passages. What am I missing here
that seems to be obvious enough to others that it does not merit mention
in the sources I’ve looked over? Any feedback or direction you can provide
would be of great help.

Have you read my book “Reasons for Belief”? I deal with the general
question of numbers in the Old Testament in this book, although I do not
deal with the specific question of Numbers 3. Let me give a general
comment, and then deal with the specific passage.

First, you should know that the Jews represented numbers using Hebrew
letters. Their system was somewhat like the Roman numerals. The problem is
that some of the Hebrew letters used to represent different numbers are
very similar to one another. For example, the Hebrew letters kaleth and
resh look almost identical, yet they represent different numbers. For this
reason, when copies of the Hebrew were made, it was very easy for
accidental errors to slip into the manuscripts. Therefore, one would
expect that our Hebrew Bibles will have a significant quantity of number
copying errors, which would reduce our general confidence in the numbers
in the Old Testament. For example, if one saw in a particular passage that
there were 500 soldiers, it would be hard to rule out that the original
might have been 530 or even 5000. One would be well advised to not base
any strong claims based on the numbers found in the Old Testament.
Exceptions to this rule would include the command to circumcise on the
eighth day, or the forty years in the wilderness, both of which are found
in numerous places in the Old Testament, making it extremely unlikely that
those numbers were not in the original.

By contrast to numbers, mistakes in spelling of words in the Hebrew are
fairly easy to detect and to correct. Using an example in English, if one
was making a copy of a manuscript and saw the letters thougt, it would be
easy to discover the mistake and correct it, especially if the context
implied that the word thought fit in the passage. Number copying errors
are not so easily detected or corrected, and once they occur, it is
difficult to establish from either context or spelling rule which of two
different manuscripts is correct.

About the specific example you raise, it is very possible that a scribe
made a mistake in copying Numbers 3. It would be my guess that the most
likely place that a scribe might have made a mistake would be in 3:23,28
or 34. The footnote you find in your Bible says that some manuscripts of
the Septuagint have 8300 in Numbers 3:28. Given that the Septuagint is
Greek translation from the Hebrew made in about 200 BC, my first guess
(and probably the guess of those who created the NIV translation) is that
the Septuagint manuscripts with 8300 represent an attempt by a scribe to
“improve” or “correct” the apparent error in the Hebrew manuscript he was
copying. It is possible, but unlikely that the Septuagint is the correct
original from which the accepted Hebrew Masoretic text represents a later
copying error. For this reason, what the origin of the different totals in
the two halves of Numbers is, will be difficult to reconstruct. This will
have to remain one of the fairly small number of copying mistakes we have
in our Hebrew Old Testaments which will be difficult to correct. Clearly
however, no important teaching or doctrine is effected by this copying
error.

John Oakes, PhD