There is a story that apologists from Josh McDowell to Lee Strobel have been telling for decades. It goes something like this: hundreds of years before Jesus of Nazareth was born, the Hebrew prophets laid down a series of detailed predictions about a coming Messiah. Jesus fulfilled those predictions — sometimes down to the letter — thereby proving his divine identity through the mathematical impossibility of coincidence. It is a tidy narrative. It is a compelling narrative. And it fundamentally misrepresents how the New Testament authors actually engaged with the Hebrew scriptures.
When you sit down and examine what the New Testament writers were actually doing with the Old Testament texts they cited, you do find people saying that the prophets foretold these things. They say it plainly and often. What you do not find is people reading the prophets the way the modern argument from prophecy reads them. You find people doing something far more sophisticated, far more creative, and far more characteristic of first-century Jewish interpretive practice. You find pesher. You find midrash. You find typology. And none of those methods works the way the apologetic argument from prophecy needs it to.
That distinction matters enormously. If you misidentify the method, you misidentify the meaning. And if you misidentify the meaning, you build an entire apologetic edifice on a foundation the prophets themselves never laid.
The Interpretive World of Second Temple Judaism
To understand what the New Testament authors were doing, you have to understand the interpretive environment they inhabited. These were not modern historical-critical scholars attempting to recover the "original meaning" of a text through grammatical-historical analysis. They were participants in a vibrant, diverse, and often wildly creative tradition of Jewish scriptural interpretation that flourished in the Second Temple period — roughly 516 BCE to 70 CE. As John J. Collins shows in The Scepter and the Star, the period's messianic hopes ran along several lines. He identifies "four messianic paradigms in Judaism around the turn of the era: king, priest, prophet, and heavenly messiah or Son of Man," and adds that "these paradigms were not always distinct" (Collins 2010, ch. 9). But Collins also argues that one of them dominated: the Davidic warrior-king "constitutes the common core of Jewish messianism around the turn of the era" (ch. 3). A crucified Messiah, I would add, matched none of these figures, and least of all that one. That is precisely why an interpretive method, rather than a checklist of predictions, had to do the heavy lifting.
Donald Juel, in his landmark work Messianic Exegesis: Christological Interpretation of the Old Testament in Early Christianity (1988), argued that early Christian reading of the Hebrew scriptures was not an anomaly. It was a species of a broader genus. The interpretive moves that the earliest Christians made with texts like Psalm 2, Psalm 110, and Isaiah 53 were of a piece with the moves being made at Qumran and, later, in the rabbinic tradition. In Juel's words, Qumran pesher and New Testament exegesis "belong within a common ethos in which texts are read in particular ways" (Juel 1988, 55). The Christians were not inventing a new method. They were applying an existing method to a new set of circumstances — specifically, the circumstances of a crucified and (they believed) risen Messiah.
Pesher: "This Is That"
Let's start with pesher, because it is perhaps the most illuminating parallel. The word is Hebrew, and it means something like "interpretation" or "solution"; its Aramaic cognate, pishra, is the word Daniel uses for his interpretations of dreams. We know it best from the Dead Sea Scrolls — the sectarian documents produced by the Qumran community, a Jewish group that retreated to the Judean desert and lived in apocalyptic expectation from roughly the second century BCE to 68 CE.
The Qumran community produced a series of documents known as pesharim (the plural of pesher), in which they took prophetic texts — Habakkuk, Nahum, the Psalms — and applied them directly to the events of their own community. The Habakkuk Pesher (1QpHab) is the most famous example. The commentator takes a verse from Habakkuk, quotes it, and then says, in Vermes's rendering, "Interpreted, this concerns ..." — and what follows is an application to the Teacher of Righteousness, the Wicked Priest, or some other figure or event within the community's own experience. Anyone who wants to see this for themselves can read the pesharim in Geza Vermes's standard translation, The Dead Sea Scrolls in English; the formula is relentless and unmistakable, verse after verse re-pointed at the community's present.
Here is what is crucial about pesher: it does not care what the prophet understood himself to mean. Habakkuk was writing about the Chaldeans, the Babylonian invaders of the late seventh century BCE. The commentator reads them as the Kittim, the invaders of his own era ("Interpreted, this concerns the Kittim," 1QpHab II 12), and he states the principle outright: "God told Habakkuk to write down that which would happen to the final generation, but He did not make known to him when time would come to an end" (1QpHab VII 1-2). The meaning belongs instead to the Teacher of Righteousness, "to whom God made known all the mysteries of the words of His servants the prophets" (VII 4-5). The assumption undergirding pesher interpretation is that the prophetic text has a hidden meaning — a meaning deposited there by God, opaque to the original author, and only now revealed to the inspired interpreter. The text is a divine code, and the community possesses the key.
Notice what pesher does not deny. It does not deny prediction. The same commentary describes a priest whom God enabled to "interpret all the words of his servants the prophets, through whom he foretold all that would happen to his people" (1QpHab II 8-10). Pesher is a prediction hermeneutic. What it denies is that the prophet understood his own prediction. The prediction is real, on this view, but only the later interpreter can see what it was about.
Now watch what happens in Acts 2. Peter stands up at Pentecost and quotes Joel 2:28-32 — the passage about God pouring out the Spirit, about young men seeing visions and old men dreaming dreams, about blood and fire and billows of smoke. Peter's interpretive move is pure pesher: "But this is what has been spoken through the prophet Joel" (Acts 2:16, World English Bible). This is that. Joel was writing about a future eschatological event, the Day of the Lord, and he gave it no date. Peter supplies one. Where Joel says "It will happen afterward" (Joel 2:28), Peter's quotation says "It will be in the last days, says God" (Acts 2:17), and the last days turn out to be this morning in Jerusalem. A few verses later he says that David, "being a prophet," was "foreseeing this" when he "spoke about the resurrection of the Christ" (Acts 2:30-31). He is reading Psalm 16, in which the psalmist speaks in the first person of his own trust in God. That is a claim of prediction, and Peter means it. What makes it pesher is where the prediction was found: not in anything Joel or David understood themselves to be saying, but in a meaning read off the text after the event, by an interpreter who already knew what the event was.
The same method appears in Acts 4:25-26, where the early believers quote Psalm 2 — "Why do the nations rage, And the peoples plot a vain thing? The kings of the earth take a stand, And the rulers take council together, Against the Lord, and against his Christ [anointed one]" — and then apply it directly to the trial and crucifixion of Jesus: "For truly, both Herod and Pontius Pilate, with the Gentiles and the people of Israel, were gathered together against your holy servant, Jesus, whom you anointed" (Acts 4:27). The identification is explicit and unapologetic. Psalm 2 is about Jesus. The rulers are Herod and Pilate. The Gentiles are the Romans. The people of Israel are the Jewish authorities who handed Jesus over.
But here is what you need to notice: Psalm 2 is a royal psalm. It was composed for — and about — the coronation of a Davidic king. The "anointed one" (mashiach) in the psalm is not a future eschatological figure; it is the reigning Israelite monarch. The "nations" raging against him are the surrounding peoples threatening the kingdom. The psalm's original context is thoroughly political and thoroughly contemporary to the Israelite monarchy. The early Christians took that psalm and reapplied it, through pesher, to the events of Jesus's passion. They did treat it as prophecy; Acts introduces it as words God spoke "by the mouth of your servant, David" (Acts 4:25). But it was a prophecy discovered by re-reading a coronation hymn after the crucifixion, not one the psalmist made.
The Christians were not alone in reading Psalm 2 this way. Among the Qumran texts is the Florilegium (4Q174), which, as Collins describes it, "strings together commentaries on 2 Sam 7:10-14; Pss 1:1 and 2:1" (Collins 2010, ch. 3). Its lines on "Why do the nations rage" are damaged. But in George Brooke's reconstruction, the interpreter identifies the raging nations with the Kittim, and the psalm's faithful with "the chosen ones of Israel in the latter days" (Brooke 1985, 93, on 4QFlor 1:18-19). A Jewish group with no connection to Christianity read the same psalm as a script for its own final crisis.
Midrash: The Art of Creative Exegesis
If pesher is "this is that," midrash is something broader and, in some ways, even more interesting. The term comes from the Hebrew root darash, meaning "to seek" or "to inquire." Midrash is the practice of searching the scriptures — turning them over and over; as the Mishnah would later put it, "Turn it over, and [again] turn it over, for all is therein" (Avot 5:22) — to draw out meanings, connections, and applications that go well beyond the surface.
Midrashic exegesis is not eisegesis in the pejorative modern sense, as though the interpreter is just making things up. It operates on the assumption that scripture is inexhaustible — that every word, every letter, every textual peculiarity has been placed there by God and is therefore significant. The interpreter's job is to find the connections, fill in the gaps, and resolve the tensions. The result is often startlingly creative. Paul, when he reads Sarah and Hagar as two covenants, even names his own procedure, and it is a fourth mode: "These things contain an allegory" (Galatians 4:24).
Richard Hays, in his influential work Echoes of Scripture in the Letters of Paul (1989), did more than anyone to show how much Paul is doing with his quotations, and he pointedly refuses to explain it by calling it midrash. The label, he warns, "tends to bring the interpretive process to a halt, as though it had explained something, when in fact we should keep pressing for clarity" (Hays 1989, 14). Paul does not simply quote proof-texts. What Hays offers in place of the label is metalepsis, a figure he takes from the literary critic John Hollander: Paul quotes a fragment of a larger passage and expects his audience to hear the unquoted context echoing behind it.
But — and this is essential — Paul's readings are not constrained by the original author's intention. Take Isaiah 49:8 ("I have answered you in an acceptable time. I have helped you in a day of salvation"). Paul quotes it in 2 Corinthians 6:2 from the Greek version (so English translations word it a little differently there) and adds his own gloss: "Behold, now is the acceptable time." He is not arguing that Isaiah foresaw a Gentile church in first-century Corinth. His claim is that the day Isaiah spoke of has arrived. Call the move midrash or, with Hays, echo: Paul is finding in Isaiah's words a pattern, a resonance, a divine principle that he sees enacted in his own experience.
Hays is careful to note that this is not arbitrary. Paul operates within a coherent theological framework — a conviction that the God who spoke through Isaiah is the same God who raised Jesus from the dead, and that the story of Israel is reaching its climax in the events Paul is living through. But coherent is not the same as foreseen. Paul is not claiming that Isaiah saw Corinth coming. He is treating Isaiah as a living word that speaks to ever-new situations precisely because it carries within it the patterns of God's action.
Typology: God Acts in Patterns
Typology is the third major mode of early Christian Old Testament interpretation, and it is perhaps the one most often confused with prediction-fulfillment by modern apologists. Typological reading identifies patterns — types and antitypes — in which earlier events, persons, or institutions foreshadow later ones. Adam is a "type" of Christ. The Exodus is a "type" of salvation. The manna in the wilderness is a "type" of the Eucharist.
The critical distinction between typology and prediction is this: typology works retrospectively. The interpreter starts with the antitype — the later event — and then looks backward to find the earlier type that corresponds to it. The claim is not that the earlier event predicted the later one, but that God acts in consistent patterns throughout history, so that earlier events can be recognized, after the fact, as prefigurations of later ones.
The Exodus pattern runs deep in the New Testament. Rikki Watts, in Isaiah's New Exodus in Mark, argues that Mark is organized around Isaiah's promised "new exodus" — the prophet's vision of a return from exile modeled on the first Exodus — and that Mark presents Jesus as the long-delayed fulfillment of that hope. That is typology at the scale of a whole narrative: Isaiah patterned his hope on the Exodus, and Mark patterns Jesus on Isaiah. Watts, writing from within the faith, calls the result fulfillment, and he means it. My point is how much of that fulfillment is made of patterning.
Craig Evans's Matthew (New Cambridge Bible Commentary, 2012) reads Matthew's infancy citations typologically: "Matthew sees biblical history repeating itself, which is what typology is all about – the conviction that God will act in the future the way he acted in the past" (Evans 2012, 47). Matthew's formula — "that it might be fulfilled which was spoken through the prophet" — is a claim of fulfillment, and Matthew means it. The question is what kind of fulfillment, and when you look at what Matthew actually does with the texts he cites, the picture is more complex.
Take Matthew 2:15, where Matthew quotes Hosea 11:1: "Out of Egypt I called my son." Matthew applies this to the Holy Family's flight to and return from Egypt. But Hosea 11:1 is not a prediction. It is a historical retrospective. Hosea is talking about the Exodus — God calling the nation of Israel, his "son," out of Egyptian bondage. Read in its original context, the verse is about the past, not the future. Evans says so without hedging: "The passage is not predictive, nor is it messianic" (Evans 2012, 58). Matthew's application is typological: just as God called Israel out of Egypt, so God called Jesus out of Egypt. The pattern recurs. The type is fulfilled in the antitype.
Or take Matthew 2:18, where Matthew renders Jeremiah 31:15: "A voice was heard in Ramah, Lamentation, weeping and great mourning, Rachel weeping for her children; She wouldn't be comforted, Because they are no more." Matthew applies this to Herod's slaughter of the innocents. But Jeremiah 31:15 — which in the World English Bible reads "A voice is heard in Ramah, lamentation and bitter weeping, Rachel weeping for her children" — is about the Babylonian exile. It is Rachel, mother of Joseph and Benjamin, weeping for her exiled children at Ramah, the place Jeremiah 40:1 names as the holding point for "the captives of Jerusalem and Judah, who were carried away captive to Babylon." The next verse promises that "They will come again from the land of the enemy" (Jeremiah 31:16). Evans again: "Like Hos 11:1, Jer 31:15 is neither messianic nor eschatological" (Evans 2012, 65). It is not a prediction of infanticide in Bethlehem. It is a poetic image of national grief. Matthew sees a typological correspondence: the grief of exile is reenacted in the grief of the massacre. That is not prediction-fulfillment. That is pattern recognition applied retrospectively.
The strongest reply to all this does not come from the popular lists. It comes from G. K. Beale, an evangelical New Testament scholar, and it is the position I most need to answer. Beale states the problem up front: Hosea 11:1 looks like "a mere historical reflection, but Matthew clearly understands it as a direct prophecy that is fulfilled in Christ" (Beale 2012, 697). His answer is to read the whole chapter. Hosea 11 opens with the first exodus and closes with a promise: "They will come trembling like a bird out of Egypt" (Hosea 11:11). (Beale also reads 11:5 as threatening a return to Egypt, but he notes that translations divide on the verse, and the World English Bible has "They won't return into the land of Egypt" [p. 704 n. 22].) On Beale's reading, the chapter's "main point and goal is the end-time exodus from Egypt," so Matthew borrows the language of the first exodus for a second exodus that Hosea really did predict (p. 705). That is a serious argument, and it is right about one thing: Hosea did look forward. But what Hosea promised was the return of a nation, "out of Egypt, and like a dove out of the land of Assyria" (11:11), a homecoming from exile told in the plural. Beale has an answer to that too. Hosea 1:11 says the reunited people "will appoint themselves one head," and 3:5 has them seek "Yahweh their God, and David their king." So, he argues, the future exodus is led by an individual king, and that "enhances further why Matthew could apply the corporate national language of Hos 11:1 and apply it to an individual king, Jesus" (pp. 708–709). That is the strongest form of the case. But look at how the king is found. Hosea 11 never mentions a Davidic king; the only king in the chapter comes in 11:5, where "the Assyrian will be their king." He has to be brought in from chapters 1 and 3, and 1:11 has to be read as a return from Egypt, although it says only that they "will go up from the land." Beale argues that from allusion to other passages (p. 708 n. 38). And Beale himself puts his conclusion as a question: "Could Matthew not have had such a biblical-theological reading of Hosea?" (p. 709). To see Hosea's promise in one child's journey home from Egypt, you already have to know whose journey you are looking for. And Beale frames the result as typology: his reading is, he says, "a short step away from saying that the first exodus was seen by Hosea and, more clearly, by Matthew as a historical pattern" (p. 705). On his own account, then, the pattern is seen most clearly from Matthew's end.
The infancy citation with the best claim to forward-looking prediction in its own context is Micah 5:2, quoted in Matthew 2:5-6. Evans grants that here "the prophecy of Mic 5:2 has been fulfilled in a more conventional sense," while adding that "an element of typology remains" (Evans 2012, 54). I grant it too. Even so, Micah's ruler is the one who "will deliver us from the Assyrian" (Micah 5:6), an eighth-century hope. So the claim I am defending is narrower than "no prediction anywhere". In the cases the New Testament itself puts forward, the correspondence is found by reading back from the event.
The Contrast with the Argument from Prophecy
When apologists compile their lists of "fulfilled prophecies," they are claiming what the New Testament writers also claimed: that the prophets foretold Christ. 1 Peter says the Spirit in the prophets "predicted the sufferings of Christ" (1 Peter 1:11). Acts has Peter declare that "the things which God announced by the mouth of all his prophets, that Christ should suffer, he thus fulfilled" (Acts 3:18). Nor is the apologetic use of the claim new. Around the middle of the second century, Justin Martyr offered as proof the prophets who spoke "before these things came to pass," since Christians could see events happening "just as they were predicted." This, he told the emperor, is "the strongest and truest evidence" (First Apology 30, tr. Dods and Reith).
The modern form of the argument, which is the position this essay examines, puts both of my Matthew texts in exactly that frame. Josh McDowell's Evidence That Demands a Verdict lists "Descent into Egypt. Hosea 11:1" and "Massacre of Innocents. Jeremiah 31:15" under the heading "Old Testament Predictions Which Were Literally Fulfilled in Christ." When McDowell asks whether Matthew was mistaken to apply Jeremiah to Herod, he answers with Theodore Laetsch's 1953 commentary on Jeremiah: "No. Certainly not. The entire context of ch. 31, beginning ch. 30:20 and continuing to ch. 33:26, is Messianic." The logic is straightforward: the prophet predicted X, X occurred centuries later, and so God's foreknowledge is demonstrated. If the premises held, it would be genuinely impressive. The disagreement, then, is not over whether the first Christians believed in prediction. It is over how they found the predictions, and whether that way of finding them can be run in reverse as evidence.
So look at how the New Testament authors found them. They are doing pesher: "This event in our experience corresponds to that text in scripture." They are doing midrash: "This text, read in light of our christological convictions, yields meanings that speak to our situation." They are doing typology: "God's action in the past prefigures God's action in the present." In each of the cases above, the movement is from event to text, not from text to event. The experience of Jesus — his life, death, and (for the believers) resurrection — came first. The scriptural interpretation came second.
Donald Juel's thesis is that "the beginnings of Christian reflection can be traced to interpretations of Israel's Scriptures, and the major focus of that scriptural interpretation was Jesus, the crucified and risen Messiah" (Juel 1988, 1). In my paraphrase: the early Christians did not find Jesus in the scriptures and then go looking for him in the world. They encountered Jesus in the world and then went to the scriptures to make sense of what they had encountered. The scriptures provided the interpretive framework, the vocabulary, the categories. But the driving force was the experience, not the text.
This is why, when you read the New Testament's Old Testament citations carefully, you so often find that the original context of the cited passage has little or nothing to do with a future Messiah. The passages are about the Exodus, or the Babylonian exile, or the Davidic monarchy, or the psalmist's personal lament. The christological reading is imposed on them — not arbitrarily, but through the interpretive methods available in the first century — by people who were trying to understand a shattering and unprecedented experience.
Why This Matters
The apologetic argument from fulfilled prophecy depends entirely on the prediction-fulfillment model. If the prophecies are genuine predictions — if Isaiah 7:14 was intended by the prophet as a prediction of a virgin birth seven centuries in the future — then their fulfillment in Jesus is indeed remarkable and demands explanation. But if the New Testament authors were doing pesher, midrash, and typology — if they were starting with Jesus and reading backward into the texts — then the "fulfillment" is not evidence of supernatural foreknowledge. It is evidence of creative scriptural interpretation.
And creative scriptural interpretation, however theologically meaningful it may be within a faith community, is not an argument that should compel an outsider. Any sufficiently motivated interpreter, armed with the methods of pesher and midrash, can find "fulfillments" of scriptural texts in virtually any set of events. The Qumran community did it with their Teacher of Righteousness. The early Christians did it with Jesus. The method produces results because it is designed to produce results. It is not a discovery procedure; it is an application procedure. And if your method is designed to produce results no matter what, the results prove nothing about foreknowledge.
That is my account of the evidence. Others will disagree. But the underlying descriptive point is widely shared, by Juel, Hays and Evans among many others: the New Testament authors were children of their interpretive age, employing the methods available to them — methods that, as the examples above show, do not work the way the argument from prophecy presupposes. I want to be precise about where the agreement ends, though. Scholars like Hays and Evans describe the method in terms that overlap with mine (Hays speaks of "reading backwards," Evans of biblical history repeating itself) while drawing a very different theological conclusion from it. For Hays, in Reading Backwards (2014), reading Israel's scriptures in light of Christ is not a debunking observation but the very heart of a figural hermeneutic he regards as legitimate and revelatory. The deflationary conclusion I am pressing here — that this retrospective method cannot function as evidence for an outsider — is my argument, not theirs. What I am borrowing from them is the description of how the method works, not their evaluation of what it proves.
So the next time someone hands you a list of "fulfilled prophecies" and asks you to calculate the odds, ask them a simple question: did the prophet intend to make a prediction, or did a later interpreter find a correspondence? Because the answer to that question determines everything. And in the cases the New Testament itself puts forward, the honest answer is the latter.
The early Christians read the Old Testament brilliantly, creatively, and passionately. They believed the prophets had spoken of Jesus, and they said so. But the correspondences they found were made at the reading end. That gap, between a correspondence found and a prediction made, is the gap through which the entire prophecy-fulfillment argument falls.
