The Torah presents a lot of numbers: seven days of creation, 10 plagues on Egypt, 12 tribes of Israel. Big numbers, too: the children of Israel are counted during various censuses, money and goods donated to the tabernacle and Temple are tallied, animals sacrificed to the Temple are totaled up. In all this reckoning, the numbers don’t always remain consistent across different biblical accounts of the same events. On today’s daf, the rabbis interrogate some of those inconsistencies.
Kontrokos the officer asked Rabban Yohanan ben Zakkai: In the individual count of the families of Levites you find 22,300 (Numbers 3:22–34). But in the collective count of Levites you find a total of 22,000 (Numbers 3:39). Where did the 300 go?
Kontrokos is a non-Jewish political figure who notices an inconsistency in census reports about the number of Levites. Verses 22–34 of the third chapter of Numbers tally up the male Levites among different clans, finding 7,500 in one set, 6,200 in another and 8,600 in a third. These three numbers add to 22,300. But then, just a few lines later, a summarizing verse reports the total as 22,000. The math is off — 300 Levites are missing.
Rabban Yohanan ben Zakkai said to him: Those 300 Levites were firstborns, and a firstborn Levite cannot abrogate the sanctity of a firstborn.
The rabbis assume that firstborns were, in some sense, the original Levites. That is, God’s original plan was that all firstborns would be dedicated to divine service. In the wake of the golden calf incident, however, God changed course and chose, instead, to have the Levites dedicated to divine service. At this moment, God needed to perform a one-for-one swap, and the Levites redeemed — that is, replaced — the firstborns across the people. Therefore, the number of Levites needed to be the same as the number of firstborns. However, Rabbi Yohanan ben Zakkai explains, some of those Levites were also firstborns — 300, to be exact. Therefore, the number of Levites recruited to divine service, elected to redeem the firstborns of Israel, was exactly the number of Levites who were not already firstborns — 22,000. The other 300 Levites were themselves firstborns and already part of the divine service. First math problem solved.
But wait, there’s more:
Additionally, Kontrokos asked Rabban Yohanan ben Zakkai: With regard to the collection of silver, you find 201 talents and 11 maneh, as it is written: “A beka a head, that is, half a shekel, of the shekel of the tabernacle 603,550 men” (Exodus 38:26). But with regard to the giving of the silver, you find only 100 talents, as it is written: “The 100 talents of silver were for casting the sockets of the sanctuary and the sockets for the curtain” (Exodus 38:28). Now, was Moses your teacher a thief, or was he a gambler, or was he not expert in accounting? He gave half and took half, and he did not return even a complete half!
When the Israelites collected metals with which to build the tabernacle, Kontrokos points out, we see a major fiscal discrepancy. And, as with the counting of the Levites, this discrepancy isn’t hidden — the numbers appear in adjacent verses. A talent is made up of 3,000 shekels. Therefore, 603,500 shekels should come to just over 200 talents. And yet, the text reports, in the next verse, that Moses used 100 talents to actually construct the sockets. Kontrokos sees this discrepancy as evidence that Moses was either a thief who stole the money, or a gambler who gambled it away. Again, Rabbi Yohanan ben Zakkai has a ready explanation:
He said to him: Moses, our teacher, was a trustworthy treasurer and was expert in accounting, but a maneh of the tabernacle was double.
Rabban Yohanan ben Zakkai explains that in the days before there were standardized weights and measures, the tabernacle calculated the weight of a maneh differently than other organizations. A maneh in the tabernacle weighed twice a maneh in other contexts. Therefore, the 200 manehs collected were the 100 manehs used in the construction. This discussion is not only about math. It’s also about how our presuppositions and prior biases shape how we encounter data. If you approach a topic, a text or a person assuming bad faith and ill will, you will find plenty of evidence for both. And if you approach with goodwill and even reverence, you are likely to find ways to make the story consistent. There is a fine line between genuine but critical goodwill and naivety and blindness to reality, but today’s daf reminds us that when we try to understand anything, the first data point we need to name and analyze is actually ourselves and what we are bringing to the table.
Read all of Bekhorot 5 on Sefaria.
This piece originally appeared in a My Jewish Learning Daf Yomi email newsletter sent on September 23, 2026. If you are interested in receiving the newsletter, sign up here.
An Open Door for Everyone. On Sukkot, guests are welcomed into the sukkah — a tradition of hospitality that's central to the holiday. My Jewish Learning holds that same door open all year, welcoming anyone curious about Jewish life and community. Your gift keeps it open.
