Censuses in the Wilderness

Exegetical Approaches

Multiple Censuses

There were two or more distinct censuses during the first two years of the Israelites' sojourn through the wilderness.

Purpose of census – According to Rashi, there was no practical need for the various censuses. Rather, Hashem repeatedly counts the Children of Israel out of His extraordinary love for them,1 much like a shepherd continuously monitors the size of his flock.2 Thus, Hashem counted the Israelites upon their departure from Egypt, and then again a few months later in Shemot 30 to see how many had died in the aftermath of the sin of the Golden Calf. He later also counted them for a third time, in the second month of the second year, when His presence began to dwell in the Tabernacle amongst the nation.3
Above the age of twenty – If the census is Hashem's way of showing His love for the Children of Israel, it is not clear why it should have only included only those aged twenty and up, rather than all members of the nation.4
"כָּל יֹצֵא צָבָא" – The description of those who were counted as "יֹצֵא צָבָא" also poses a difficulty for Rashi, as he disconnects the census from any military goals. Ramban, though, suggests that the word "צָבָא" might not refer to an army, but rather to a gathering of the congregation.5
When was the census of Shemot 30 commanded? According to Rashi, the command to count the people is recorded in achronological order, and actually took place only after the atonement for the sin of the Golden Calf in Shemot 34.6 See Purpose of the Mishkan for elaboration.
Were the Levites included in the original census? Rashi maintains that the Levites were not counted in any of the censuses. This is a logical outgrowth of his understanding that the census of Shemot was necessitated by the sin of the Golden Calf. Since the Levites (according to Rashi) did not participate in the sin, they were also unaffected by the ensuing punishment, and thus did not need to be recounted.7
Identical tallies – To account for this phenomenon, Rashi is compelled to make a number of major assumptions:
  • For the purpose of both the census in the first year of the wilderness and the one in the second year, people's ages were calculated not by their biological age, but rather by how old they were on the first day of the new year,8 which for this purpose (only) was counted from Tishrei.9 As such, no one turned twenty in the period between the census of Shemot 30 (which, according to Rashi, took place in Tishrei after Yom HaKippurim – see above) and that of Bemidbar 1-4 (which took place seven months later in Iyyar).
  • The Levites were not included in either census – see above.
  • There were no deaths during the period between the censuses.10
Additionally, if one accepts all of these assumptions, it should then have been self-evident that there would be no change in the number of Israelites. Ostensibly, this would have made the census in the second year redundant. However, this does not trouble Rashi, as he views the counting as a manifestation of Hashem's love rather than as arising from any practical need.
Half-shekels? According to Rashi, the census in Bemidbar 1 was no different than the earlier one in Shemot 30, and it too utilized half-shekels, even though the fact is not mentioned explicitly. See Half Shekels – For Census or Tabernacle?.
"כְּשֵׁשׁ מֵאוֹת אֶלֶף רַגְלִי" – The tally in Shemot 12 was not merely an estimate, but the result of an actual census as well. This census, too, resulted from Hashem's love of the nation.
Census of Bemidbar 26 in the fortieth year – Rashi maintains that, similarly here, Hashem counted the nation here since there had just been a plague11 and He wanted to know how many people remained.12
Chronology of the opening chapters of Bemidbar – According to Rashi and Lekach Tov, the census of Bemidbar 1 took place only in the second month of the second year.  Since Bemidbar 7 and 9 are dated to the first month of that year, they derive from here the principle that the Torah is not always written in chronological order ("אין מוקדם ומאוחר בתורה").  See Chronology of Bemidbar 1 – 10 for elaboration.

One Full and One Partial Census

The census in Shemot 30 was a general census which merely provided the total number of Israelites, while the census of Bemidbar 1 was a far more comprehensive one, which collected data about individuals, their families, and tribal affiliations.

Purpose of census – All of these commentators connect the census of Bemidbar 1 to the originally planned imminent entry into the land of Israel. In preparation for the conquest, it was important to know not just the total number of Israelites but to also divide them according to family and tribe.13 The main goal of the census of Shemot 30, in contrast,14 was to raise funds for the construction of the Tabernacle and the count itself was only secondary.15 Thus, the census in Shemot 30 included only a collection of half-shekels without a headcount and naming of individuals.
Above the age of twenty – The census of Bemidbar included only those who were old enough to wage war.
"כָּל יֹצֵא צָבָא" – This phrase, too, supports the notion that the census was military in nature.
When was the census of Shemot 30? While most of these commentators suggest that the census of Shemot 30 is in its proper place and took place before the sin of the Golden Calf, the Hoil Moshe maintains that it occurred afterwards.
Were the Levites included in the original census? According to most of these commentators, the Levites were included in the earlier census of Shemot 30, since at that point they had not yet been singled out for any special role. Hoil Moshe, though, asserts that they were not included since, according to his chronological reconstruction, they had already been consecrated to serve in the Tabernacle.16
Identical tallies
  • Intentional – The Netziv suggests that after the first census, the total of 603,550 was set as the necessary number for the army, and for God's presence to dwell amongst the nation. Thus, during the second census, the people were not counted to see how many they were, but to ensure that they met the proper quota.17
  • Coincidence – The other commentators suggest that the identical numbers were a coincidence, but they differ in their understandings of the details of how this worked:
    • Deaths match those coming of age – Ramban (first explanation) proposes that by happenstance the number of men who turned twenty equaled the number of men who had died.18
    • Levites included or omitted – Ramban (second proposal) and Abarbanel explain that the coincidence was possible because the Levites were included in the first census, but not the second. This would allow for approximately 22,000 people to turn twenty in the intervening months.19
    • Levites replace firstborns – According to Shadal, the Levites were included in the first census, but the firstborns were not.20 As these two groups were close in number, the omission of the Levites from the second count did not have any significant affect on the census21 and by complete chance it turned out that with the small discrepancy between Levites and firstborns, the number of deaths equaled the number of men turning twenty.
Half-shekels? Ramban maintains that both censuses took place via a half-shekel donation, while the other exegetes22 assert that this was a requirement only for the census of Shemot 30.23
"כְּשֵׁשׁ מֵאוֹת אֶלֶף רַגְלִי" – Ramban emphasizes that this number is an estimate rather than an exact count. He also suggests that, unlike the later censuses, it might include all males above the age of thirteen.
Chronology of the opening chapters of Bemidbar – According to this position, the census of Bemidbar 1 took place only in the second month of the second year.  Since Bemidbar 7 and 9 are dated to the first month of that year, most of these commentators assume that the opening chapters of Bemidbar are not in chronological order.  See, however, Chronology of Bemidbar 1 – 10 for Abarbanel's novel approach.
Census of Bemidbar 26 in the fortieth year – Ramban, Abarbanel, and the Netziv connect this census to the conquest and division of the land as well.24 Ramban notes that in this count, there is a division by families, and not by heads, as that was what was necessary to determine the tribal inheritances.

Only One Census

There was only a single census during the first two years in the wilderness. This approach subdivides as to when this census transpired:

In the Second Year

Shemot 30 did not constitute a command to immediately count the nation, and there was only a single census which took place in the second year as described in Bemidbar 1.

"כִּי תִשָּׂא" – According to this position, the command in Shemot 30, "When you count..." was not a command to immediately take a census, but only a general command dictating what one should do in future censuses.25
Half-shekels
  • Prevention of plague – Chizkuni asserts that the shekel donations were a necessary contribution to prevent plagues in future censuses, but their initial collection in Shemot 30 did not constitute a census in its own right.
  • Contribution to Mishkan – According to R. Yosef Bekhor Shor and the GR"A, the command to give half-shekels was wholly unconnected to the census and solely for the building of and service of the Mishkan. R. Yosef Bekhor Shor suggests that the shekalim were not even counted; the number of half-shekels totaled in Shemot 38 is the Torah's omniscient parenthetical statement regarding the future total which had not yet occurred.
Purpose of census – According to R. Yosef Bekhor Shor and Chizkuni the count was necessary since the nation was soon to go to war.
Above the age of twenty / "כָּל יֹצֵא צָבָא" – The count was limited to the men eligible to fight in the wars of conquest.
Identical tallies – This problem is the main motivating factor leading R. Yosef Bekhor Shor to posit that there must have been only one census. He explains that the precise number of half-shekel coins collected was only known after the census of Bemidbar 1,26 and so there was really only one total.
Were the Levites included in the original census? As there was merely one census which took place only after the Levites were already chosen to serve in the Tabernacle, they were not included.
"כְּשֵׁשׁ מֵאוֹת אֶלֶף רַגְלִי" – R. Yosef Bekhor Shor states that this was not a census, just an approximation of the number of Israelites leaving Egypt.27
Census of Bemidbar 26 in the fortieth year – R. Yosef Bekhor Shor views this census as parallel to the one in Bemidbar 1. In each, the Children of Israel are about to enter the land and embark on the wars of conquest and therefore need to be counted.

Spanning Both the First and Second Years

The censuses described in Shemot 30 and Bemidbar 1 were both part of a single extended process which began when the Tabernacle was being constructed and continued through the second month of the second year.

Two in one - Cassuto proposes that the census had two parts. At the time of the construction of the Tabernacle, the people gave their half-shekels, names and other information, but it was only in the second month of the following year that the data was analyzed and all of the necessary calculations were made.
Purpose of census – Cassuto does not address the specific goal of the census. Since he posits that there was only one, there is no issue of redundancy, and one could suggest that it was related either to the establishment of the Tabernacle when it began, or to the upcoming travels and conquest which were to take place right after it was finished.
Parallels – Cassuto points to documentation of other censuses in the Ancient Near East and to David's census described in Shemuel II 24:8 as evidence that such counts were a long drawn out process that could take many months.28
Identical tallies – Since the "two" counts were really one and the same, it is not astonishing that the final totals were identical.
Above the age of twenty / "כָּל יֹצֵא צָבָא" – Cassuto agrees that the count included only those of age to go to war, but points out that in other Ancient Near Eastern cultures, too, censuses generally included only those who were of military fighting age. The reference to war, then, does not necessarily connote that the purpose of the census was related to the imminent conquest.
When was the census of Shemot 30? According to Cassuto, the chapters are written in chronological order, and the beginning of the census took place before the sin of the Golden Calf.
Were the Levites included in the original census? As there is but one census, the Levites were not included, as becomes clear from the description in Bemidbar.
Half-shekels? The census began with a half-shekel count (as described in Shemot 30) and concluded with a more detailed analysis of families and individuals (as found in Bemidbar 1).
"כְּשֵׁשׁ מֵאוֹת אֶלֶף רַגְלִי" – Cassuto views this number as an estimate and not the result of an actual census.
Chronology of the opening chapters of Bemidbar – According to Cassuto, Bemidbar 1 describes the culmination of a process which began already in the first year and only ended in the second month of the second year.  This reading also allows for an innovative understanding which would preserve chronological order in the opening chapters of Bemidbar.  See Chronology of Bemidbar 1 – 10 for elaboration.
×