Deep Bible
Back to Questions
Deep dive · 13 min read
Scholars debate

David's census: did God or Satan move him?

Two biblical accounts of the same event start with two different instigators. 2 Samuel 24:1 says the LORD's anger moved David to count Israel. 1 Chronicles 21:1 says Satan stood against Israel and incited him. The chapters that follow disagree on numbers, plague duration, and the price of the threshing floor as well. Three positions on what is going on.

What's at stake

The two accounts of David's census are the cleanest case of textual divergence in the Hebrew Bible. Same event, same king, same plague, same threshing floor. The opening verse identifies a different mover. The casualty count and the purchase price come out differently. The Chronicler is clearly working from Samuel (or a source close to Samuel) and rewriting. The question is why, and what each rewriting commits the reader to about how the satan figure developed across the Second Temple period. Readers have been asking some version of this since the Targumists tried to harmonize the two verses in late antiquity. Three positions cover most of the modern range, and they sit on different assumptions about composition history, divine agency, and the literary career of the word satan.

What the texts say

2 Samuel 24:1 opens with a covenant frame. 'And again the anger of the LORD was kindled against Israel, and he moved David against them to say, Go, number Israel and Judah.' The verse does not give the reason for the anger. The 'again' points backward, possibly to the famine of chapter 21, possibly to a longer pattern. The instigator is YHWH, and the act is presented from inside that anger.

1 Chronicles 21:1 narrates the same moment with a different opening. 'And Satan stood up against Israel, and provoked David to number Israel.' The Chronicler keeps the rest of the chapter close to Samuel but changes the agent at the top. The Hebrew is satan (without the article in most manuscripts), and in the Chronicles context it functions as a proper name or near-proper name for an adversary figure. The change is not a slip. The Chronicler has reworked the entire chapter, and the satan substitution sits at the head of the rewrite.

What follows in both accounts is the same basic arc. Joab protests. David presses. The count is taken. David's heart smites him. Gad the prophet gives him three options. David chooses the plague. Seventy thousand die. The angel of the LORD is stopped at the threshing floor of Araunah (Samuel) or Ornan (Chronicles) the Jebusite. David buys the floor, builds an altar, offers sacrifice, the plague is stayed. The site becomes the location of the Temple (a connection Chronicles makes explicit at 1 Chr 22:1, Samuel leaves implicit). Inside that shared arc, every quantitative detail differs.

The three positions

How the divergence has been read

Three families of reading, each with its own defenders and its own evidence about what changed between Samuel and Chronicles and why.

By the time Chronicles was composed (c. 400-350 BCE), the satan had emerged as a distinct adversary figure in late Persian and early Hellenistic Jewish thought. The Chronicler rewrote 2 Samuel 24:1 to attribute the incitement to that figure rather than to YHWH directly, reflecting an updated theological vocabulary. Job 1-2 and Zechariah 3 sit on the same trajectory.
Held by
  • Peggy L. Day, An Adversary in Heaven (HSM 43; Scholars Press, 1988)
  • Ryan E. Stokes, 'The Devil Made David Do It... or Did He?' (JBL 128, 2009)
  • Sara Japhet, I and II Chronicles (OTL; Westminster John Knox, 1993)
  • Gary N. Knoppers, 1 Chronicles 10-29 (Anchor Bible; Doubleday, 2004)
  • Ralph W. Klein, 1 Chronicles (Hermeneia; Fortress, 2006)
  • Henrike Frey-Anthes, Unheilsmächte und Schutzgenien (OBO 227; 2007)
Evidence
  • Job 1-2 (likely 6th-5th c. BCE) and Zechariah 3 (late 6th c. BCE) use ha-satan ('the satan') as a title for an adversary figure in the heavenly court. By 1 Chr 21:1 the word appears without the article in most manuscripts, moving toward a proper name
  • The Chronicler's broader editorial pattern is to soften or reframe theologically difficult moments in Samuel-Kings. The Bathsheba episode is omitted entirely (compare 1 Chr 20:1-3 with 2 Sam 11-12). The substitution at 1 Chr 21:1 fits the same editorial profile
  • Second Temple texts after Chronicles continue the trajectory. Jubilees (2nd c. BCE) names Mastema as the adversary who incites and tests; the Dead Sea Scrolls (1QM, 4Q280, 4Q286) develop Belial as the corresponding figure; the New Testament inherits Satan as a personal adversary
  • The Chronicler's parallel substitution at 1 Chr 21:15 (where the angel of the LORD is described in terms that diverge from 2 Sam 24:16) shows the chapter has been theologically reworked across multiple agents, not just at the opening verse
  • The Septuagint translators of Samuel (3rd-2nd c. BCE) felt the difficulty of 2 Sam 24:1 and partly smoothed it (the LXX reads 'and the anger of the LORD again was kindled against Israel, and Satan provoked David against them'). The harmonization shows the change was felt as theologically meaningful within a few centuries
Challenges
  • The Hebrew at 1 Chr 21:1 lacks the article, but the word satan also lacks the article when it functions as a common noun ('an adversary') elsewhere. Whether the Chronicler intends a proper name or an unnamed human or angelic adversary is not settled by the syntax alone
  • The theological-development reading assumes a relatively linear trajectory from common-noun satan (1 Sam 29:4; 1 Kgs 11:14) through the-satan (Job, Zechariah) to Satan as proper name. The actual development was probably more diffuse, and the Chronicles verse may sit at a transitional point rather than after the transition
  • The position commits to a late date for Chronicles (after Job and Zechariah's final forms), which is the majority view but is not without challenges from those who date Chronicles earlier

The two accounts side by side

The opening verse is the famous divergence, but it is not the only one. Reading the chapters in parallel surfaces a pattern of differences across cause, numbers, plague options, purchase, and closing frame. Each divergence is what every position above has to account for.

2 Samuel 24 versus 1 Chronicles 21

Same event, narrated by two writers. The differences are concentrated at predictable seams: the agent of causation, the demographic numbers, the price of the floor, and what the chapter says about the site at its close.

2 Samuel 24
v. 1: instigator
'And again the anger of the LORD was kindled against Israel, and he moved David against them to say, Go, number Israel and Judah.' YHWH is the explicit mover; no adversary figure is named.
v. 9: census total
'Eight hundred thousand valiant men that drew the sword in Israel; and the men of Judah were five hundred thousand men.' Total: 1,300,000.
v. 13: plague option
Gad offers seven years of famine, three months of fleeing before enemies, or three days of pestilence. (Some LXX manuscripts read 'three' for the famine option, matching Chronicles.)
v. 24: purchase price
David buys the threshing floor and the oxen for fifty shekels of silver. Araunah the Jebusite is the seller.
v. 25: closing frame
David builds an altar, offers sacrifice, the LORD answers, the plague is stayed. The chapter ends. No explicit identification of the site as the future Temple location.
1 Chronicles 21
v. 1: instigator
'And Satan stood up against Israel, and provoked David to number Israel.' The adversary figure is the named mover; no anger of YHWH is mentioned at the chapter's opening.
v. 5: census total
'And all they of Israel were a thousand thousand and a hundred thousand men that drew sword: and Judah was four hundred threescore and ten thousand men.' Total: 1,570,000. The Israel figure is 300,000 higher and the Judah figure 30,000 lower.
v. 12: plague option
Gad offers three years of famine, three months of being defeated by enemies, or three days of pestilence. The famine option is shorter than in 2 Sam 24 MT and matches the LXX reading of Samuel.
v. 25: purchase price
David weighs out 'six hundred shekels of gold by weight' for the site. The seller is Ornan (the same name as Araunah, in Chronicles' spelling). The order of magnitude is significantly higher than the Samuel figure.
vv. 26-22:1: closing frame
David builds the altar, fire from heaven answers, the plague is stayed, and David declares: 'This is the house of the LORD God, and this is the altar of the burnt offering for Israel' (1 Chr 22:1). The threshing floor is explicitly identified as the future Temple site.

The price difference (50 shekels of silver versus 600 shekels of gold) has its own discussion history. The traditional harmonization, going back to the Targum, reads the Samuel figure as the price of the floor and the oxen alone, and the Chronicles figure as the price of the entire surrounding site that became the Temple precinct. The two-traditions reading takes the divergence as evidence the Chronicler was working with additional source material or calculating from later Temple-period valuations. The numerical inflation in Chronicles (also visible in the army totals) fits a wider pattern in the book that some scholars read as schematic and others read as access to additional records.

The career of the satan

Whichever position one ends up holding on 1 Chr 21:1, the chapter sits inside a longer story about how the word satan develops across the Hebrew Bible and the Second Temple period. The trajectory matters because the development reading depends on it directly, and the other two positions have to give some account of it as well.

Stages in the use of satan from common noun to proper name. Green entries treat satan as a common noun or generic adversary; amber entries develop the figure as a specific adversary in the divine court or beyond.

Common noun / generic adversary
Specific adversary figure
1000 BCE
1 Samuel 29:4
The Philistine commanders worry David might become 'a satan' (an adversary) against them in battle. Pure common noun: a human enemy in a military sense.
0% along range
960 BCE
2 Samuel 19:22
David rebukes the sons of Zeruiah for becoming 'a satan' to him on the day of his return to Jerusalem. Common noun again, a human adversary inside the royal court.
4% along range
930 BCE
1 Kings 11:14, 23, 25
YHWH raises up Hadad the Edomite and Rezon son of Eliada as 'a satan' (adversary) against Solomon. The word still functions as common noun, but now describes an adversary YHWH is using providentially.
7% along range
550 BCE
Job 1-2 (likely 6th c. BCE)
Ha-satan ('the satan,' with the article) appears as a member of the heavenly court who tests Job after dialogue with YHWH. A defined role, not yet a proper name.
44% along range
520 BCE
Zechariah 3:1-2
Ha-satan stands at the right hand of the high priest Joshua to accuse him; the angel of the LORD rebukes him. The adversary role is judicial in the heavenly court.
47% along range
400 BCE
1 Chronicles 21:1
Satan (mostly without the article in the manuscripts) incites David. The move from common-noun adversary to a near-proper-name agent who acts on his own initiative.
58% along range
160 BCE
Jubilees and 1 Enoch (Maccabean era)
Mastema (in Jubilees) and a fallen-angel tradition (in 1 Enoch) develop the adversary figure as a chief opponent of God's purposes. Belial appears in the Dead Sea Scrolls in the same role.
82% along range
100 BCE
Dead Sea Scrolls (1QM, 4Q280, 4Q286)
Belial is the named adversary of the sons of light in the War Scroll and is cursed in the community's liturgies. The figure is now thoroughly personified.
87% along range
30 CE
New Testament Satan
Satan is a proper name for the adversary (Matt 4:1-11; Luke 22:31-32; Rev 12:9). The figure 'the great dragon, that old serpent, called the Devil, and Satan' is treated as the same agent across the New Testament writers.
100% along range

The timeline is what the development reading takes as decisive. The word starts as a common noun for any adversary, narrows to a defined heavenly-court role in the exile and early Persian period, and is treated as a near-proper name by the Chronicler. By the New Testament, Satan is fully personalized. The concurrent-causation reading does not deny the development; it argues the underlying theology of divine and adversarial agency was stable across the period, and the vocabulary changed faster than the conceptual content. The two-traditions reading takes the timeline as suggestive but not decisive, noting that any given text can sit at a different point on the trajectory depending on its source materials.

What the chapter is doing inside each book

2 Samuel 24 closes the book. The chapter sits after the so-called appendix (2 Sam 21-24), a non-chronological collection of David traditions that the editors of Samuel placed at the end. The chapter has structural counterparts in chapter 21 (the famine in Saul's house) and chapter 22 (David's song), and the book closes on a darker note than it began. The census is the king's last narrated act before the succession narrative resumes in 1 Kings 1. Read inside Samuel, the chapter functions as the final shadow over David's reign, and the threshing-floor purchase becomes a quiet moment of restoration that does not undo what the census cost.

1 Chronicles 21 is doing something different. The Chronicler has cut large stretches of David's life that appear in Samuel (Bathsheba, Absalom's revolt, the Sheba revolt) and replaced them with an extended account of David's preparations for the Temple. The census chapter is rewritten so that its conclusion is the identification of the future Temple site (1 Chr 22:1). The plague is not the closing shadow on David's reign but the inciting moment for the Temple project. Chapters 22-29 then narrate David's preparations in detail. The Chronicler's substitution at 21:1 fits a chapter whose function in the book is to begin the Temple narrative, not to close the David narrative.

Both framings affect how the opening verse reads. In Samuel, attributing the incitement to the LORD's anger fits a chapter that ends on a note of unresolved cost. In Chronicles, attributing the incitement to an adversary fits a chapter whose endpoint is the Temple site and the divine approval of David's altar. The narrative work each chapter is doing inside its book pushes in the direction of the verse-level change the chapter opens with.

Why a census was the wrong act

One question the chapter does not directly answer is why counting the people was an offense worthy of plague. The Pentateuch authorizes a census in Numbers 1 and Numbers 26 without sanction, and Exodus 30:12 prescribes a half-shekel ransom payment 'when thou takest the sum of the children of Israel' so that 'there be no plague among them when thou numberest them.' The Exodus rule is the clearest hint at why the David census was the wrong act. A census without the ransom payment was understood to expose the people to plague.

Joab's protest in both accounts (2 Sam 24:3; 1 Chr 21:3) does not invoke the Exodus rule but treats the census as a moral problem of pride and royal overreach. Joab's reading is the line most often picked up by traditional commentators: David is counting the army to measure his own strength rather than trusting the LORD. The combination of the Exodus rule (the technical liability of a census) and Joab's protest (the moral dimension of military pride) gives both Samuel and Chronicles the working framework for why the plague follows. The two accounts then diverge on who set the chain in motion.

Where the positions actually disagree

Stripped to the spine, the three positions disagree on three questions. First, what is the literary relationship between Samuel and Chronicles. The development reading treats Chronicles as a deliberate theological updating of Samuel. The two-traditions reading allows for partial independence, with the Chronicler working from variant source material. The concurrent-causation reading does not require a strong position on the literary relationship; it can be paired with either of the other two.

Second, what counts as the same event when narrated by two writers with different theological vocabularies. The development reading treats the verses as describing the same event with different language. The concurrent-causation reading treats them as describing different layers of the same event. The two-traditions reading allows for the possibility that the writers had partially different accounts of what occurred.

Third, what the satan figure is doing by the late Persian period. The development reading puts 1 Chr 21:1 at a transitional point in a long career. The concurrent-causation reading treats the figure as a long-standing role in Israelite theology (Job, Zechariah) that Chronicles is now describing more directly. The two-traditions reading takes the trajectory seriously but argues the trajectory was not linear and that texts at the same date may have used the vocabulary differently.

Reading the chapter with the question open

Most readers do not need to choose a position to read the chapter. The shared narrative arc (census, plague, threshing floor, altar, stayed angel) is the spine, and it is identical in both books. The opening-verse divergence is the seam where the reader meets the question of how Israelite writers across two centuries described the same event with shifting theological vocabulary. Reading 2 Samuel 24 and 1 Chronicles 21 in parallel surfaces a moment in the Bible where the canon preserves two voices on the same event without resolving them. The Targum and the Septuagint show that ancient readers also felt the seam and tried to smooth it. The three positions above are different ways of holding the seam, with different costs and different commitments about how the satan became Satan.

Sources

Primary sources
  • 2 Samuel 24:1-25 (MT; Leningrad Codex B19a)
  • 1 Chronicles 21:1-22:1 (MT; Leningrad Codex B19a)
  • Septuagint 2 Kingdoms 24 and 1 Paralipomenon 21 (Rahlfs)
  • Targum Jonathan on 2 Samuel 24 (Sperber, Brill 1959-1973)
  • Targum on 1 Chronicles (McIvor, Aramaic Bible 19; Liturgical Press, 1994)
  • Job 1:6-12; 2:1-7 (MT; LXX)
  • Zechariah 3:1-2 (MT; LXX)
  • 1 Samuel 29:4; 2 Samuel 19:22; 1 Kings 11:14, 23, 25 (uses of satan as common noun)
  • Exodus 30:12 (the census ransom)
  • Numbers 1; Numbers 26 (authorized censuses)
  • Jubilees 10:7-11; 17:15-18:13 (Mastema), in Charlesworth, OTP vol. 2
  • 1 Enoch (Book of the Watchers, chs. 6-16), in Charlesworth, OTP vol. 1
  • 1QM (War Scroll), 4Q280, 4Q286 (Belial references; DJD 11, 7)
  • 4Q176 (Tanhumim), DJD V
  • 4QSamuel-a (4Q51), 4QSamuel-b (4Q52), 4QSamuel-c (4Q53), DJD XVII
  • Babylonian Talmud, Berakhot 62b (on the census)
  • Josephus, Antiquities 7.13.1-4 (on David's census), Loeb Classical Library
Modern scholarship cited
  • John Calvin, Institutes of the Christian Religion II.4.1-4 (1559)
  • Matthew Henry, Exposition of the Old and New Testament (1710)
  • D. A. Carson, Divine Sovereignty and Human Responsibility (Marshall, 1981)
  • P. Kyle McCarter Jr., II Samuel (Anchor Bible; Doubleday, 1984)
  • Peggy L. Day, An Adversary in Heaven: Satan in the Hebrew Bible (HSM 43; Scholars Press, 1988)
  • Sara Japhet, I and II Chronicles (OTL; Westminster John Knox, 1993)
  • Iain W. Provan, 1 and 2 Kings (NIBC; Hendrickson, 1995)
  • Henry Ansgar Kelly, Satan: A Biography (Cambridge, 2006)
  • John Goldingay, Old Testament Theology, vol. 2 (IVP, 2006)
  • Bruce K. Waltke, An Old Testament Theology (Zondervan, 2007)
  • Henrike Frey-Anthes, Unheilsmächte und Schutzgenien (OBO 227; Vandenhoeck and Ruprecht, 2007)
  • Ryan E. Stokes, 'The Devil Made David Do It... or Did He?' JBL 128 (2009): 91-106
  • Gary N. Knoppers, 1 Chronicles 10-29 (Anchor Bible; Doubleday, 2004)
  • Ralph W. Klein, 1 Chronicles (Hermeneia; Fortress, 2006)
  • Steven L. McKenzie, 1-2 Chronicles (AOTC; Abingdon, 2004)
  • A. Graeme Auld, I and II Samuel (OTL; Westminster John Knox, 2011)
  • John Van Seters, The Biblical Saga of King David (Eisenbrauns, 2009)
  • T. J. Wray and Gregory Mobley, The Birth of Satan (Palgrave, 2005)
  • Elaine Pagels, The Origin of Satan (Random House, 1995)
  • Ryan E. Stokes, The Satan: How God's Executioner Became the Enemy (Eerdmans, 2019)