Millions living in the wilderness

The Bible is full of numbers that appear to be too big, and new research has found a likely reason for this.


There is clearly something strange about the many big numbers in Numbers and elsewhere in the Bible. This is confirmed when we look at the consequences of the numbers as they are. We discover, for example, that an average family had 44 sons – without counting daughters!1 Also the descendants of Levi (one of Jacob’s 12 sons) didn’t make up 1/12 of the population (as we’d expect) but were only 1/45th of the total – i.e. the tribe of Levi was insignificantly tiny.2

These large numbers of non-Levites come from the lists of military-aged men in the other tribes. One possible explanation is that the Hebrew word for ‘a thousand’ (eleph) can also mean ‘an army unit’ or ‘a clan’ (e.g. Num 10:4). So these lists might not be counting how many thousands of men are in each tribe but counting the number of army units – which would likely contain far fewer than 1000 people each. But this explanation has problems because, for example, the tribe of Judah had “74,600” men of military age (Num 1:27): if the word for 'thousand' meant 'clan'or a ‘military unit’, this says they had 74.6 clans – but what is “0.6” of a clan?3

A more recent theory is that large numbers in the OT should all be divided by ten, because at some point they were all multiplied by ten. This theory initially sounds arbitrary and unconvincing, but this kind of multiplying became a normal style of writing in the ancient world. You can probably think of a myriad problems with this kind of theory, but we actually do similar things with verbal numbers. The previous sentence is an example: we use the word 'myriad' without considering that it actually means ‘thousand’. There are dozens of different ways we do this, so you’ve probably seen this kind of exaggeration a million times (!).


The trouble is that once this kind of exaggeration becomes normal, you have to join in or it sounds like you are minimizing things. When German propaganda during World War II boasted shooting down several planes every day, the Allies could have responded by pointing out that the claimed kills exceeded their total number of planes - but instead they broadcast equally inflated figures of their own.4

Even in modern news, if a journalist reported that there were 'several' casualties after a bombing it comes across as downplaying the incident, but if he said 'a multitude’ or 'myriad' casualties it sounds like a proper expression of the magnitude of the tragedy. However, when it comes to actual numbers, in our culture we expect accuracy, so if a journalist said there were 150 casualties instead of 15 this would be considered inaccurate.

The only occasion when we might inflate actual numbers is when we convert things into modern values. For example the King James Bible translates the numbers accurately in the Parable of the Two Debtors: “the one owed five hundred pence, and the other fifty” (Luke 7:41 – translating ‘denarius’ as ‘pence’). This made sense in the 17th century when laborers were paid a few pence per day. However, we would retell it as “the one owed $50,000, and the other $5000” – i.e. multiplying the numbers by a hundred. We would do this on the basis that a denarius was the laborer's daily wage, which now is about $100. In other words, in order for the story to make sense today, the original numbers have to be larger.

Trojan war

Greek and Roman historians faced a similar problem when they recorded ancient battles that occurred long before their time. Massive population expansions meant that armies in their days were much larger than the ones in previous centuries. So when they reported important battles from the past, the actual numbers made it appear like a minor skirmish. For example, the thousand ships that launched to rescue Helen of Troy represented the combined forces of all the Greek nations. This was an impressively large number of ships when Homer wrote about it, centuries after the event, but it was an almost impossibly huge number in the 12th century BC when Troy actually fell. So most historians assume that Homer exaggerated to make the story sound correct at the time.5

These ancient historians were helping their readers in the same way that modern Bible translations help to convey the parable of the debtors – by multiplying the numbers so that the story made sense to the current generation. But unfortunately, once you start doing this to some numbers, you have to do it to them all. If there are ten times more ships, there must be ten times more soldiers and ten times more deaths.

What if this has been done in the Bible? What if the scribes decided to make the numbers in the Bible look more convincing by multiplying everything by 10? This audacious proposal was put forward by Kevin P. Edgecomb.6 Normally I would simply reject an idea like this on the grounds that scribes did not make this kind of change to the meaning of the text. However, this simple idea has remarkable results which verify the text rather than undermine it.

Multiplied by ten

If we assume that the numbers of fighting men have all been multiplied by ten, many inconsistencies quickly disappear. First, of course, it explains why every number ends in “0”! Second, the population of the Exodus population becomes 240,000 people with 60,000 fighting men.7 This is a formidable army for that time, without being unreasonably large.

However, being reasonable doesn’t make it right – we need a way to corroborate this. One way to check is by calculating the average size of a family. When we didn’t take multiplication into account, the average family in Israel contained 44 sons, but when we assume all the military numbers are multiplied by ten, the average family has 4.4 sons.8 We can confirm this independently by examining the genealogies recorded in the Bible. Listing all the families throughout Israel’s history produces an average of three sons per family.9 This is not an exact corroboration but it is close enough, especially as this average includes later periods when families tended to be smaller.

Another corroboration is found in Judges 5:8 where the number of fighters in Israel is given as 40,000. This number occurs in a poem so wasn’t multiplied because that would have ruined the poetry. Now, according to the multiplying theory, Israel’s army was actually 60,000 men in the Wilderness years, so it isn’t surprising that it had declined 33% by the time of the Judges. However it would be extraordinary if instead Israel’s numbers had fallen from 600,000 (Num 1:46; 2:32; Exod 12:36) – a decline of 93%!

There are further corroborations throughout the Old Testament which is full of unrealistically large numbers that make sense when they are un-multiplied. For example we read of enormous weights like the 1,700 shekels of gold from 300 gold earrings worn by defeated Ishmaelite soldiers (Judg 8:4, 26). This meant that they each wore 65g of gold (the weight of three ice cubes) hanging from each ear – and they wore them during a battle! One tenth of this weight would represent a significantly impressive ear-ring which they were realistically able to wear.

Distances and casualties

Another example is the respectful distance of 2000 cubits between the Ark of the Covenant and the Israelites who followed it through the wilderness (Joshua 3:3-4). This means they walked about half a mile behind the Ark. This distance means they would often lose sight of it when they went round a rock face or over even a slight hill! One tenth would be 100 meters – a very respectful though practical distance.

Fatalities of war and disasters also had to be multiplied, so that the readers could see how disastrous they were. So when a city wall collapsed on an army, the number it killed was 27,000 (1 Kgs 20:30). This is nine times the number of people who were killed when the Twin Towers fell! None of these small ancient cities would have a wall big enough to fall on that number. One tenth (2700) is still a huge number, but it is possible.

The time-span of God's covenant with his people is described as “a thousand generations” (Ps 105:8) - but this would be about 30,000 years, which means that Jesus shouldn’t be expected to return for another 25,000 years. One tenth of this (3000 years) means he could come back any time from now, because the earliest Psalms were written about a millennium before Christ.

And although Solomon was a great king, his 40,000 chariot horses are an exorbitant number (1Kgs 4:26).10 Even if the stables were compacted in 40 rows of tiny stalls with narrow corridors in a building 100 meters wide (as wide as Buckingham Palace), they would stretch back for a kilometer11 – i.e. from one wall of ancient Jerusalem all the way to the opposite wall of the city. One tenth of this number would still be an extraordinary military force for that time.

Updated text

When we see huge numbers like this, it is tempting to simply dismiss them as unrealistic exaggerations, misunderstandings or mistakes, but there are too many and they are too consistent. Or we might regard them as miraculous – including the ability of Midianites to hang so much gold from their ears! However, all these numbers make historic sense if we assume they have been multiplied by ten.

Our un-multiplication in the last example is actually corroborated in a parallel text at 2 Chronicles 9:25: the number of Solomon's chariot horses is recorded there as one tenth of the number in 1 Kings 4:26: a figure of 4,000. It was this unmultiplied figure that inspired Kevin Edgecomb to come up with this theory.

But who would have multiplied all these numbers? Edgecomb calls him ‘Multiplier’ and thinks he was one of the final editors of the official Hebrew Bible in around 300 BC. He suggests the motive was to write numbers in the normal style of the time.12

For example, according to a Greek historian in the 5th century BC, Xerxes had an army of 1.7 million (Herodotus 7.60). No one really believed that the number was meant to impress. The problem for editors like the Multiplier was that the figures in the Bible looked paltry next to numbers like these, so he regarded this multiplying as part of the task of translating the text into the language of the time.

This multiplication wasn’t meant to fool anyone – the fact that everyone did it, meant that everyone was expected to do it. Thankfully, he did it in a way that preserved the original numbers because he multiplied everything by ten. And probably this was intentional. The absurdity that this sometimes created would alert the careful reader, who expected this kind of thing in any case. The one place where he failed - when he left the number of Solomon's horses unmultiplied in 2 Chronicles 9:25 – may have been deliberate. Perhaps he wanted to leave a blatantly clear signal concerning what he had done. After all, it is difficult to believe that he simply missed this verse.

We should be cautious about this theory because, although it solves a lot of problems, it also involves a large number of intentional changes to the text. However, perhaps we should regard this as a single global change – i.e. updating the whole text in line with the newer way of expressing large numbers in histories, by adding a zero to them all.

The Old Testament does contain other global updates because, like English, Hebrew was a living language that constantly changed. For example, in older Hebrew the word for “she” used to be spelled differently (הוא, HWA). When the spelling changed (to היא, HYA), the Bible text looked old fashioned, so the whole Hebrew Bible was updated several centuries before Christ. An exception was made for Genesis to Deuteronomy – this was allowed to remain looking traditionally archaic – so every occurrence of “she” had the old spelling in those books.13

This is similar to the way that the King James version was updated periodically. The first printing had “For God so loued ye world, that he gaue his only begotten Sonne”. Later versions standardized the use of “v” and changed the “y” (i.e. the letter “thorn”) to “th”. None of this was done to change the meaning, but to faithfully convey the meaning of the text to the next generation of readers.14 When the RSV changed “thou” to “you”, they deliberately kept the archaic “thou” when addressing God – rather like the Hebrew editors allowed the books of Moses to look archaic, out of respect.


• Calculations based on numbers in the Exodus generation result in nonsense, such as 44 sons per mother. • These and other numbers make sense if they are all divided by ten • Other histories of the time enlarged ancient numbers to make them sound right in an age when the population was much larger Therefore it is likely that an editor added a zero to all the large numbers to make them suitably impressive, like other ancient historians did. He didn’t aim to fool anyone, so he left indications of what he had done to preserve the originals.

1^ The number of firstborn sons for all the tribes except Levi are 22,273 (Num 3:43) so if the total males were 990,000 (as calculated above), the average number of boys born in each family is 990,000/22,273 = 44.45.
2^ The total males in Levi were 22,000 (Num 3:39) so if the total males were 990,000 (as calculated above) their proportion of the population is 22,000/990,000 = 1/45.
3^ This theory is explored at For a more complex version see
4^ See
5^ Actually Homer says 1186 ships. Thucydides 1.10 (in late 5th century BC) assumed this was an exaggeration – see A few modern authors have considered it might be accurate – see
6^ See He suggests a mix of 1/10 and 1/100 but in this chapter I assume 1/10 throughout.
7^ This isn’t simply 1/10 of the previous calculation because the figures for the Levites weren’t exaggerated like the numbers of fighting men. The previous calculation for an average age of 51 remains the same because it was based on Levites. When the numbers of fighting men in Numbers 1-2 is divided by ten, the total is 60355. The total males non-Levites of all ages is therefore (603,55x51.28) / (51.28-20) = 98,945. The males in Levi are 22,000 so the total males is 120,945, making a population of about 2402,000.
8^ The un-multiplied number of males is 98,945 and the total number of firstborn is 22,273 (Num 3:43), so the number of sons in an average family is 98,945/22,273 = 4.44.
9^ The Bible genealogies include 208 families with two sons or more - see the data at . Of these, 73 families had 2 sons, 70 had 3 or 4 sons, and 65 had 5 or more sons – i.e. an average sized family had 3.5 sons. The number of families with only one son are excluded because most genealogies consist of a simple list of ancestors. If we did take these into account, the average is reduced to about three sons per family.
10^ The NIV translates the ancient Greek Septuagint which has already unmultiplied this number; see the NIV footnote for the original Hebrew
11^ If each horse had a stall of only 2 meters by 1 meter with a one-meter corridor space between back-to-back stalls – i.e. (2x1x40,000) for horses plus (1x40,000/2) for human access, this comes to 100,000 m2.
12^ He points to Marco de Odorico, "The Use of Numbers and Quantifications in the Assyrian Royal Inscriptions." State Archives of Assyria Studies 3. Helskinki: The Neo-Assyrian Text Corpus Project, 1995. See also Fouts (
13^ Actually there are a few places the new form occurs also in the Books of Moses: Exo.1:16; Lev.5:11; 11:39; 13:6; 16:31; 21:9.
14^ See

This was previously published in a similar form in Christianity magazine

Your comments can start a discussion

Share this page on social media and your comments could start a discussion among your friends. Any link you create this way will continue working even after this month when the topic will no longer be available on this site. So new visitors to your discussion will still be able to read the discussion topic so long as they use your social media link.
  • On Facebook the topic, then go to your Facebook page to add your comment.
    If you want me to see your comments, mention "David Instone-Brewer"
  • On Twitter tweet the topic, then go to your Twitter account to read it.
    If you want me to see your comments, mention "@DavidIBrewer"

Subscribe to each new monthly release

● To follow on Twitter: 
● To follow by Email:        
● On Facebook, first "Like" it:
Then, to ensure you see the post each month, in "Following" tick "See first"
("Default" means Facebook decides whether to show it to you or not).