At Keilah, David asks God a direct question and is given a definite answer — and the rest of the story records that answer not happening.
At Keilah, David asks God a direct question and gets a definite answer — Saul will come down, the town will hand you over — then leaves, and neither happens. A close reading of 1 Samuel 23, set beside the inquiry that held (Ziklag) and the one that met only silence (Saul before Gilboa). The careful claim: of the inquiries that return a stated outcome, this is the one the narrative records as averted. It locates a hard question about foreknowledge and freedom without trying to settle it.
David asks God a yes-or-no question, receives a clear answer about what is going to happen, acts on it — and the thing God said would happen never happens. The text reports all of this plainly, in a few verses, without pausing to explain it. It is one of the quieter strange moments in the Hebrew Bible, and as far as the record goes, it may be the only one of its kind.
David is a fugitive. Saul wants him dead, and David is moving from hideout to hideout with roughly six hundred men. He hears that the Philistines are raiding the town of Keilah, inquires of God twice about whether to intervene, is told to go, and rescues the town (1 Sam 23:1–5). Then he settles inside its walls.
Saul hears that David has boxed himself into a walled town with gates and bars — and sees his chance. David, learning Saul is mobilizing, calls for the priest Abiathar to bring the ephod, and puts two questions to God:
"O LORD, the God of Israel, your servant has heard that Saul seeks to come to Keilah, to destroy the city on my account. And now, will Saul come down, as your servant has heard? O LORD, the God of Israel, I beseech you, tell your servant." The LORD said, "He will come down."
1 Samuel 23:10–11 · NRSVue
Then David said, "Will the men of Keilah surrender me and my men into the hand of Saul?" The LORD said, "They will surrender you."
1 Samuel 23:12 · NRSVue
Two answers, both definite: Saul will come; the people David just rescued will hand him over. And then:
Then David and his men, who were about six hundred, set out and left Keilah and wandered wherever they could go. When Saul was told that David had escaped from Keilah, he gave up the expedition.
1 Samuel 23:13 · NRSVue
So Saul does not come down. The men of Keilah do not surrender David. Both foretold events are recorded, a sentence later, as having not occurred. The reason is on the surface of the text: David heard the answer and left, and once he was gone the situation the answer described no longer existed. The narrator offers no commentary. He simply lets the prediction and its non-fulfilment sit next to each other.
David is not overhearing a prophet or interpreting a dream. He has called for the ephod, the priestly garment carried to him by Abiathar, who had fled to David after Saul slaughtered the priests at Nob (1 Sam 22:20–23; 23:6). Bound up with the ephod is the priestly oracle — the Urim and Thummim — the sanctioned instrument for putting a question to God and receiving an answer (Exod 28:30; Num 27:21). The verb the chapter keeps using for what David does is inquire (שָׁאַל | šāʾal | ) — to ask, to request. It is, by its nature, a tool meant to be used: you ask in order to act on the answer.1
The two key answers turn on two ordinary verbs. Saul would come down (יָרַד | yārad | ), and the men of Keilah would surrender (סָגַר | sāgar | ) him — a verb whose root sense is "to shut up, close in," used here in the sense of handing a person over. Neither word carries a hidden conditional in itself. The answers read as flat statements of fact: he will come; they will hand you over.
David's two questions and God's two answers do not line up in order. In verse 11 David asks first about the surrender and then about Saul coming down — but the answer given is only to the second question ("He will come down"). David then repeats the surrender question in verse 12, and only then receives its answer ("They will surrender you"). The doubling is a real feature of the Masoretic text, not a translation artifact; the oracle answers one question at a time, in the order it chooses.2
Scripture is full of divine predictions that do not come to pass as stated — but almost all of them belong to a recognizable pattern, and it is a different pattern from this one. A prophet announces coming judgment, the people respond, and God relents:
"Forty days more, and Nineveh shall be overthrown!" … When God saw what they did, how they turned from their evil ways, God changed his mind about the calamity that he had said he would bring upon them, and he did not do it.
Jonah 3:4, 10 · NRSVue
The same shape governs Hezekiah, told "you shall die; you shall not recover," who prays and is granted fifteen more years (2 Kings 20:1–6), and Ahab, warned of disaster on his house, who humbles himself so that the disaster is deferred (1 Kings 21:21–29). In each case a judgment is proclaimed, a human repents or pleads, and God relents. The announcement carries an implied "unless."
Keilah is not that. It is not an unsolicited proclamation of judgment but a requested answer to a direct question. No call to repent is attached. And the thing that undoes the prediction is not repentance moving God to mercy — it is the inquirer using the information to remove himself from the circumstances the answer described. David does not change God's mind. He changes his own location.
That combination is what makes the passage unusual. Among the inquiries of God in the Hebrew Bible — the times someone formally asks and receives an answer through the priestly oracle or by "inquiring of the LORD" — this appears to be the only one whose specifically foretold outcome the narrative then records as not happening.3
Is it really unique? The honest answer is that "only" is a claim about an absence, and absences are hard to prove. Other inquiries return commands rather than predictions ("Go up" — Judg 20:18), and a command followed by defeat is not a failed prediction. Cases where God declines to answer at all (1 Sam 28:6) are a different thing again. Within those limits the pattern holds, but the safe formulation is the modest one: of the inquiries that yield a stated outcome, Keilah is the one the text shows being averted — not "the only verse where God says something that doesn't happen," which the relenting passages already disprove. A deeper objection runs underneath: were these even predictions that failed, or contingent answers — "if you stay, Saul comes and the town hands you over" — whose condition David simply removed by leaving? The text does not label them conditional, yet the averting works precisely because the conditions changed, so a reader can fairly hear the oracle as contingent counsel rather than a foretelling that fell through. The observation survives either way: a definite answer about the future, recorded as not occurring once the inquirer acted on it.
The strangeness of Keilah comes into focus when it is set beside the other inquiries in the very same book. First Samuel gives us three outcomes for the same act — asking God a question through the oracle — and lets them stand side by side without smoothing them together.
The answer that holds
Raiders have burned Ziklag and taken the families. David inquires through the ephod: "Pursue; for you shall surely overtake and shall surely rescue." He pursues, and recovers everything (30:18–19). A settled promise, fulfilled to the letter.
The answer that is averted
"Saul will come down… they will surrender you." David leaves the town, and neither happens. An answer about a future that the answer itself moved David to prevent.
The answer that never comes
On the eve of battle Saul inquires, "but the LORD did not answer him, not by dreams, or by Urim, or by prophets." The same instrument, the same question — and silence.
Read together, these are not three contradictions to be reconciled but three things the narrator is content to report: a word that lands, a word that is overtaken by events, and no word at all. There is even a thread of irony running through it. The king who receives only silence is named Saul (שָׁאוּל | šāʾûl | ) — "asked for," from the very root that means to inquire — while the fugitive who keeps asking keeps being answered.4
A passage like this has a long history of being pressed into service for large arguments about God's knowledge of the future and the freedom of human choices. Those debates are real, and thoughtful readers across the traditions have read Keilah in support of quite different positions.5 This article does not enter them, and the reason is not timidity but accuracy: the passage does less than the arguments need it to do.
What the text plainly requires is modest. David asked a real question and received a real answer — accurate, usable guidance about what would happen if he stayed in Keilah. He did not stay. The conditions the answer described did not come about, so the outcome it named did not occur. Everything beyond that — whether this proves the future was genuinely open, or that God knows what people would freely do under conditions that never arrive, or that an eternal God simply sees all things at once — is a framework brought to the passage, not read out of it.
The story does not tell us what kind of knowledge the answer was. It tells us David asked, was told, and acted — and that the telling was true enough to be worth acting on.
1 Samuel 23:1–13That is the honest stopping point. The passage locates a hard question with unusual clarity; it does not answer it. What it shows, plainly, is a God who will give a straight answer to a fugitive's question — an answer meant to be used — and a man who used it, and walked out of the town before the door could close.
Sources
P. Kyle McCarter Jr. (1980). I Samuel (Anchor Bible 8). Doubleday.
Position: Historical-critical — text-critical and source analysis; weighs MT against the Greek witnesses.
Robert Alter (1999). The David Story: A Translation with Commentary of 1 and 2 Samuel. W. W. Norton.
Position: Historical-critical; literary — attentive to the narrator's craft, including the Saul/David inquiry contrast.
Walter Brueggemann (1990). First and Second Samuel (Interpretation). John Knox Press.
Position: Mainline Protestant — theological reading; treats the silence to Saul and the answering of David as relational.
Cornelius Van Dam (1997). The Urim and Thummim: A Means of Revelation in Ancient Israel. Eisenbrauns.
Position: Conservative evangelical — the fullest monograph on the priestly oracle; defends the Urim and Thummim as functioning instruments of inquiry.
Blue Letter Bible / BibleHub — interlinear and Strong's data for 1 Samuel 23:11–12 (sāgar H5462, yārad H3381, šāʾal H7592, šāʾûl H7586).
Position: Reference — lexical and morphological verification of the Hebrew (independent of the site's own materials).
Cold blind grade by a fresh, input-blinded instance (2026-06-15): A+ 19/20; Fidelity 10/12 (F1 3 — the cited scholars are background, not pinned to the uniqueness claim; F2 4; F3 3). The one-point deduction is a scope-drift between the title/lead ("didn't come true") and the body's more careful reading ("the conditions the answer described ceased"); the "Pressing the claim" box now also raises the contingent-answer-vs-failed-prediction objection. Same-model check — a different-model or human read remains the real gate.