Who Wrote Isaiah? And Why It Matters for Prophecy

If you've ever heard a Christian apologist cite Isaiah 53 as a prophecy of the crucifixion of Jesus, you've encountered one of the most popular arguments in the entire apologetic arsenal. The Suffering Servant — despised, rejected, pierced for our transgressions, led like a lamb to the slaughter — it reads like a script for the Passion narrative. And if this was written by the prophet Isaiah in the 8th century BCE, more than seven centuries before the crucifixion, then you'd have to admit that's at least striking.

But here's the thing. Critical scholarship has known for over two centuries that the book of Isaiah was not written by a single author. This has been the overwhelming majority view in critical scholarship since the late nineteenth century: the book of Isaiah is a composite work assembled over several centuries, with an eighth-century core in chapters 1–39, an anonymous prophet of the Babylonian exile behind most of 40–55, and later hands, probably several, in 56–66 and in parts of 1–39 itself. There is dissent. A real minority of credentialed scholars, John Oswalt and Alec Motyer among them, defend single authorship by the eighth-century prophet. Others, notably Brevard Childs and Christopher Seitz, accept that the book is not one man's work from one century, but argue that it was shaped to be read as a single work; they are canonical readers, scholars who treat the finished book as the object of interpretation. I'll come back to both groups. But the compositional thesis is not a fringe position being imposed on a divided field; it is the working consensus that the field's standard commentaries assume as their starting point. And once you understand it, the Cyrus prediction disappears, and the Isaiah 53 argument loses its foundation: it was never a free-standing forecast, but a poem about a servant, in a book that names its servant as Israel.

This is the foundation beneath the most-cited messianic prophecy in Christianity.

The Three Isaiahs

The two-part division is older than most readers realize. R. N. Whybray traces it to Johann Christoph Döderlein in 1775, "already anticipated by Ibn Ezra in the twelfth century," and reports that "all but the most conservative scholars now accept" it (Whybray 1975, 20). The three-part framework was established by the German scholar Bernhard Duhm in his 1892 commentary on Isaiah. Duhm identified three major sections of the book, each with a distinct historical setting, theological emphasis, and literary style.

The first section — chapters 1 through 39, which scholars call Proto-Isaiah — is rooted in the 8th century BCE. Isaiah ben Amoz was active from about 740 to 700 BCE, under Uzziah, Jotham, Ahaz, and Hezekiah; his call is dated to "the year that king Uzziah died" (6:1). The threats he addresses are immediate and specific: the Syro-Ephraimite crisis (c. 735–732 BCE, when Syria and the northern kingdom of Israel moved against Judah), the expansion of Assyria, the siege of Jerusalem by Sennacherib. The theology is dominated by themes of judgment against Judah's unfaithfulness, with oracles against specific nations. This is the section that can plausibly be attributed, at least in part, to the historical prophet Isaiah, and "in part" is doing real work. Chapters 1–39 contain material from well after his lifetime. The oracle against Babylon in chapter 13 names its destroyer: "Behold, I will stir up the Medes against them" (13:17), a threat to Babylon that belongs to the sixth century, not the eighth. Duhm himself dated passages in chapters 25 and 26 to the destruction of Samaria by John Hyrcanus, "between 113 and 105 B.C.," as a contemporary reviewer summarized him (Cheyne 1893). And chapters 36–39 are a narrative that runs parallel to 2 Kings 18:13–20:19.

Then everything changes at chapter 40. Deutero-Isaiah — chapters 40 through 55 — shifts the historical setting entirely. It is no longer 8th-century Judah under Assyrian threat. It is the Babylonian exile of the 6th century BCE — specifically the years before 539 BCE, when the Persian king Cyrus was advancing on Babylon and the Jewish exiles could hope for release. The tone shifts from judgment to comfort. The very first words of chapter 40 are "Comfort, comfort my people" (40:1, WEB). The theological emphasis moves to monotheism, creation theology, and the redemption of Israel from captivity. The literary style is different — more lyrical, more rhetorical, more expansive.

And the final section — chapters 56 through 66, called Trito-Isaiah — reflects a post-exilic setting. The exile is over. The community is back in Judah, struggling with the realities of rebuilding. The concerns are about proper worship, inclusion of foreigners, internal divisions within the community. The tone is neither the fiery judgment of Proto-Isaiah nor the soaring comfort of Deutero-Isaiah but a more mixed, sometimes anxious, sometimes hopeful address to a restored but struggling community. Whether it is one voice or several is unresolved. Whybray, surveying the field, found "still no consensus of opinion concerning either the date(s) of composition of chapters 56–66 or their unity of authorship" (Whybray 1975, 41).

Duhm's three sections have outlasted his three authors. The division has been refined, debated, and developed by over a century of scholarship, which has kept the sections while loosening the idea that each is one writer's work. Joseph Blenkinsopp's magisterial Anchor Bible commentary on Isaiah 40–55 treats the composite authorship as established fact and focuses on the detailed interpretation of Deutero-Isaiah as a product of the exilic period. John Goldingay and David Payne, in their two-volume International Critical Commentary on Isaiah 40–55, likewise build their analysis on the premise that these chapters address the people of the exile and come "from the sixth century BC (or later)" (Goldingay and Payne 2006, 1:2).

The Evidence: Why This Isn't Speculation

Let me be clear about where the weight of the argument sits, because it is not where it is usually put. The strongest evidence for multiple authorship does not depend on ruling out predictive prophecy. It is internal to the book, and it begins with the audience.

Prophets speak to their own contemporaries, and chapters 40–55 speak to people on the far side of a catastrophe. At the end of chapter 39, the narrative is set firmly in the reign of Hezekiah, in the late 8th century BCE. Envoys from Babylon visit, and Isaiah warns that the royal treasures "will be carried to Babylon" and that the king's descendants will serve in its palace (39:6-7). Then chapter 40 opens, and the exile has already happened. Jerusalem has "received of Yahweh's hand double for all her sins" (40:2). The audience is being comforted after catastrophe, not warned before it. Seitz, who reads the book as a whole, makes that hinge the point: chapter 39 "reminds the reader that, however mysteriously, Babylon was already at work within God's plan for Israel and the nations, when it was still far overshadowed by the Assyrian empire" (Seitz 1993, 15). The editors built a bridge. What lies on the far side of it still presupposes a century and a half that the book never narrates.

This is not a subtle shift. It is a chasm. Chapters 40–55 presuppose the exile as a present fact. Jerusalem lies waste and is promised rebuilding (44:26; 51:3), the audience is told to "Leave Babylon!" (48:20), and Zion is addressed as a city waiting for her children to come home (49:14–21; 52:1–2). Goldingay and Payne put the point exactly: these chapters "do not tell an eighth-century audience of a day when comfort will come to its descendants (contrast 39.5-8). They actually address the people to whom that message comes" (Goldingay and Payne 2006, 1:29). For them this is the clincher, though they also call the Cyrus references "the Achilles heel of any attempt to date chapters 40-55 in any period other than the mid-sixth century" (1:29). Other arguments, "e.g. linguistic ones," point the same way, "but it is the fact that they address people for whom the fall of Jerusalem is long past that remains the conclusive indication that they come from the sixth century BC (or later)" (1:2). Where the prophet stood is a live debate. Hans Barstad and Lena-Sofia Tiemeyer have argued that the chapters were written in Judah rather than Babylon (Barstad 1997; Tiemeyer 2011), and Goldingay and Payne observe that "the fact that the material addresses both Jerusalem and the exiles relativizes the question as to where the author is" (1:35). What no one has made convincing is the eighth century. Whybray is blunt: "neither Torrey nor any other commentator has succeeded in giving a convincing interpretation of the book against any other historical background than that of the Babylonian exile" (Whybray 1975, 38).

The Cyrus problem corroborates all this, and it is worth being candid about what kind of evidence it is. In Isaiah 44:28, the text says: "Who says of Cyrus, 'He is my shepherd, and shall perform all my pleasure,' even saying of Jerusalem, 'She will be built;' and of the temple, 'Your foundation will be laid.'" In Isaiah 45:1, the text says: "Yahweh says to his anointed, to Cyrus, whose right hand I have held, to subdue nations before him."

Cyrus is named. By name. Cyrus the Great, king of Persia, who conquered Babylon in 539 BCE and issued the edict allowing the Jewish exiles to return home. The text doesn't vaguely gesture at a future deliverer. It identifies him by name and assigns him a program: conquering nations, and ordering Jerusalem and its temple rebuilt (44:28; 45:13, "He shall build my city, and he shall let my exiles go free").

Now, if Isaiah ben Amoz wrote this around 700 BCE, he named a king who would not take Babylon for another century and a half. That's the apologetic claim. And yes, if you already believe in predictive prophecy as a supernatural phenomenon, you can accommodate this datum. I should state the premise openly rather than hide it: the Cyrus argument is decisive only for a reader who does not grant that kind of prediction. For that reader, the parsimonious explanation, the one that requires the fewest assumptions, is that the author was writing when Cyrus was a known figure on the world stage. Whybray narrows it further. Cyrus "is always represented as having already established a reputation as an invincible conqueror and consequently become a potential menace to Babylon, but as having not yet attacked that city" (Whybray 1975, 22), which on his reckoning puts the prophet's career between 550 and 538 BCE (23). Whether excluding prediction is a bias or a method is the real dispute between the two sides, and I take it up below.

And then there are the vocabulary and style differences. Scholars have catalogued extensive differences in vocabulary, syntax, and literary technique between Proto-Isaiah and Deutero-Isaiah. The phrase "the Holy One of Israel" appears in both, which is often cited as evidence of unity — but it is a title, not a stylistic signature, and its presence in both sections is more easily explained by the later author's deliberate adoption of Isaianic tradition than by single authorship. The count bears this out. In the WEB, 25 of the 31 Old Testament verses that use the phrase are in Isaiah, spread across all three sections (12 in 1–39, 11 in 40–55, 2 in 56–66), and H. G. M. Williamson has argued that Deutero-Isaiah deliberately took up and edited the earlier prophet's work (Williamson 1994). Meanwhile, Deutero-Isaiah develops forms that are rare or absent in Proto-Isaiah: the formal salvation oracle to Israel ("Don't be afraid, for I have redeemed you," 43:1; compare 41:10, 44:2), the creation-redemption theology linking God's primordial creative acts to present salvation, and the trial speeches where Yahweh challenges the gods of the nations to prove their reality. The words "don't be afraid" do occur in 1–39 (7:4, 10:24, 37:6), but the salvation-oracle form, like the trial speech, is concentrated in chapters 40–55.

Finally, there are the Servant Songs. Duhm's most influential contribution was his identification of four "Servant Songs" within Deutero-Isaiah: Isaiah 42:1–4, 49:1–6, 50:4–9, and 52:13–53:12. These poems describe a figure called "the servant of the Lord" who has a mission to Israel and the nations, who suffers unjustly, and who is ultimately vindicated. Duhm argued that these songs were composed separately and inserted into the larger Deutero-Isaianic corpus. Whether or not they were originally independent (scholars disagree on this), they are products of the exilic or early post-exilic period. Duhm himself, as the same reviewer summarized him, thought the songs "were of later origin than the work of the Second Isaiah, in which they have been inserted" (Cheyne 1893). They address the condition of exile. They process the suffering of the community or its representative leader. They look forward to restoration.

Isaiah 53, the most famous of the Servant Songs, describes a figure who is despised and rejected, who bears the sins of others, who is led like a lamb to slaughter, and who is "cut off out of the land of the living." Christians read this as a prediction of Jesus' crucifixion. But in its original context, a poem of the exile or its immediate aftermath, the servant is most naturally understood as a personification of Israel itself, or as a reference to a specific historical figure within the exilic community. Blenkinsopp, in his Anchor Bible commentary, surveys the options: the servant as collective Israel, the servant as an individual prophetic figure, the servant as an idealized Israel representing what the nation should be. Whybray takes the individual option firmly. The servant of chapter 53 is "the same as the 'servant' of the three earlier 'Servant Songs' … that is, the prophet Deutero-Isaiah himself," and the "we" who speak are "a group of the prophet's fellow-exiles" (Whybray 1975, on 53:1–12). None of these options require a reference to a figure who wouldn't appear for another five centuries.

The Apologetic Counter — And Why It Fails

I promised to come back to the dissenters, and they deserve their case in its strongest form and in their own words. The best version is not "God could have done it." It is the text's own argument. Deutero-Isaiah stakes Yahweh's claim to deity on having announced events before they happened: "I declare the end from the beginning" (46:10); "before it came to pass I showed it to you; lest you should say, 'My idol has done them'" (48:5; compare 41:21–23). John Oswalt presses the point in his commentary on 41:21–29: "If the only kinds of prophecy that existed were those that modern critics are willing to admit — those with no supernatural intervention necessary — then the ground is cut out from under the central argument for God's uniqueness." On the critical reading, he goes on, "the supposed writers are not gifted theologians, but either ignorant poets or crafty charlatans" (Oswalt 1998, on 41:21–29). Alec Motyer adds a point of method. The Old Testament is our only source for its own predictive prophecy, so "we are (to say the least) much more under its judgment on this point than it under ours. We may say that we do not believe it, but we are not at liberty to deny what it affirms." He also points to a second named prediction, "Josiah by name" in 1 Kings 13:2 (Motyer 1993, on 44:28). And he rejects the objection from distance. "It is commonplace to ask if a prophet could or, more importantly, would predict an event a century and a half after his own time," he writes, and he answers that Isaiah's own warning of exile to Babylon in 39:6-7 gave him every reason to look that far ahead (Motyer 1993, introduction).

A defender can add external evidence. The Great Isaiah Scroll from Qumran (1QIsaa) runs all sixty-six chapters with no break at 39/40. Ben Sira, early in the second century BCE, credits Isaiah with having "comforted those who mourned in Zion" (Sirach 48:24). And John 12:38–41 puts Isaiah 53:1 in the mouth of "Isaiah the prophet."

Each of these has an answer. Take the "former things" first. The appeal is to announcements the audience has already seen come true, not to forecasts sealed for a later century: "I have declared the former things from of old" (48:3). Seitz, reading the book as a whole, identifies what those former things include. The theme of victory over Babylon "is also one of the former things announced of old (45:21); in Second Isaiah this involves God's calling of Cyrus vis-à-vis Babylon (cf. 13:17)" (Seitz 1993, 6), and he reads God's stirring up of the Medes in 13:17 as the counterpart of the one raised up "from the north" in 41:25, namely Cyrus (130). That is a strong reading of the finished book. It turns the argument from fulfilled prophecy into an argument internal to the book, with the later chapters pointing back to oracles already on record under Isaiah's name. It does not show that the man who spoke those oracles also wrote the later chapters. The scroll shows how the book circulated, not who wrote it. 1QIsaa was copied in the late second century BCE, some six centuries after Isaiah, and its own major division falls not between chapters 39 and 40 but after column 27, which ends with chapter 33: there is "a three-line lacuna at the bottom of column 27, and there is also a change of sheet" (Popović, Dhali, and Schomaker 2021; on the second half, see Longacre 2013). Ben Sira and John show when the attribution to Isaiah was settled, which is the very point at issue; they are not independent evidence on it. Motyer's precedent in 1 Kings 13:2 raises the same dating question about another book rather than settling this one.

The underlying argument, though, has a serious flaw. As usually deployed, it is unfalsifiable, meaning there is no possible evidence that could prove it wrong, because any specificity that would normally indicate contemporary composition gets reclassified as miraculous foreknowledge. If a text names a future figure and describes his actions, the critic says it was written while Cyrus was already on the march, and the believer says it was predicted in advance.

But unfalsifiability is a vice, not a virtue. Historical criticism operates by weighing evidence and inferring the most probable explanation. And here the text offers a test. Deutero-Isaiah knows the 540s intimately: it names Cyrus, tracks his victories (41:2-3, 41:25), addresses the psychology of exiles, and argues with Babylonian religion. And it gets the ending wrong in the way a contemporary would. It expects Babylon's defences to be smashed ("I will break the doors of bronze in pieces and cut apart the bars of iron," 45:2), its gods to be hauled away ("Bel bows down. Nebo stoops. … they have gone into captivity," 46:1-2), and ruin to fall on the city ("Desolation will come on you suddenly," 47:11). The Babylonian chronicle of Nabonidus's reign records something else. After a battle at Opis and the capture of Sippar "without a battle," "the army of Cyrus, without battle they entered Babylon," and in Esagil, Marduk's temple, the rites went on without interruption (ABC 7 iii 12–18, in Grayson's translation as revised at Livius.org). Whybray draws the conclusion about the gods: "this is a further example of a prophecy which was not fulfilled: in the event Cyrus proclaimed himself a follower of Marduk and actively promoted the worship of Babylon's gods" (Whybray 1975, 114). He also notes that the prophet, "in some of his oracles, regarded the capture of Babylon as imminent" (23). That is the signature of a prophet writing shortly before 539. It is not the signature of an omniscient forecaster, and not of someone writing after the event.

Defenders have a reply here too. Motyer asks "whether the passage has room for the evidence of the Nabonidus Chronicle that 'the troops of Cyrus entered Babylon without battle'," and answers, "we need look no further than verse 1," where "the gates shall not be shut" (Motyer 1993, on 45:1). That is a fair reading of 45:1, but it saves one line at the cost of the others: broken bronze doors, idols carried into captivity and sudden desolation do not describe a city that opened its gates and kept its festivals. There is also a genuine critical caution. Goldingay and Payne warn that the Cyrus material "does not mean that historically the passage dates from just before the fall of Babylon. It could be earlier, or even later" (1:8). I think the unfulfilled details make a date just before 539 the best fit, but that is an inference about probability, and critical commentators do not all draw it.

Moreover, the "God revealed it in advance" argument proves too much. If we accept this principle, then we have no basis for dating any ancient text based on its contents. Any text could have been written at any time, with God supplying the anachronistic details. This makes historical scholarship impossible. As a general method it cannot work: it is a move that, whatever its intent, immunizes a prior commitment from disconfirmation. Oswalt would answer that excluding prediction is a prior commitment too, and that is fair as far as it goes. The difference is that the critical reading can be checked against the text, as the Babylon oracles show, while the predictive reading absorbs any result.

Why It Matters

I can already hear someone saying, "So what if there were multiple authors? The text is still inspired. The prophecy is still true."

Here's the answer. The messianic prophecy argument depends on temporal distance. The whole persuasive force of "Isaiah predicted Jesus" comes from the claim that a prophet writing around 740 BCE foresaw events that occurred around 30 CE. That's a gap of about 770 years. That's impressive. That demands explanation.

Move Isaiah 53 to the late sixth or early fifth century BCE and the gap shrinks by two centuries or more, to something over five hundred years. That is still a long time, and a careful defender will say so: redating alone does not turn a prediction into a description. What does that work is what the poem says and where it sits. Its suffering is narrated largely in the past tense ("He was despised and rejected by men," 53:3; "he was pierced for our transgressions," 53:5), and it is told by a "we" who watched it happen and misjudged it: "Who has believed our message?" (53:1); "yet we considered him plagued" (53:4). And it sits in a book that names the servant as Israel: "But you, Israel, my servant" (41:8); "You are my servant, Israel, in whom I will be glorified" (49:3). The individual reading has its own anchor two verses later, where the servant's task is "to bring Jacob again to him, and to gather Israel to him" (49:5), which is hard to say of Israel itself, and it is part of why Whybray reads the servant as the prophet. On either reading the servant is a figure of the poet's own world, and the "we" are his contemporaries. The poem is describing the contemporary experience of suffering and exile. The suffering is described, not predicted; what the poem looks forward to is the servant's vindication ("he will see his offspring," 53:10; compare 53:11–12). The early Christians, centuries later, read this description and saw parallels to the death of Jesus. Of course they did — the language of unjust suffering, innocent victimhood, and redemptive death is universal enough to map onto many situations. But that's pattern-matching, not prophecy fulfillment.

The same applies to the other Servant Songs. Isaiah 42 describes the servant's mission in terms appropriate to the exilic context — bringing justice, being a light to the nations. Isaiah 49 has the servant speaking of being called from the womb and being given a mission to restore Israel. Isaiah 50 describes the servant being beaten and spat upon. All of these passages process the experience of exile through the figure of the servant, whether Israel or a prophet who stood for it. All of them were subsequently mapped onto Jesus by early Christian interpreters. But the mapping is retrospective. It is an act of creative theological reading, not evidence of supernatural prediction.

The distinction is not academic. If you are evaluating Christianity's truth claims — if you are asking whether the evidence supports the conclusion that Jesus was the fulfillment of ancient prophecies — then you need to know what those ancient texts actually said, when they were written, and what they were about. And when the answer is that Isaiah 53 was written in or just after the Babylonian exile, about a servant of that world, by an anonymous prophet or one of his circle, then citing it as a prediction of Jesus' crucifixion is a category error.

The Composite Book and the Composite Argument

One final point. The book of Isaiah as we have it is a composite work that was assembled over centuries by editors who collected, arranged, and supplemented prophetic material from different periods. This is not unusual in the Hebrew Bible — most prophetic books show signs of editorial expansion and supplementation. But it means that the "book of Isaiah" is not a single coherent work by a single author with a single message. It is an anthology of prophetic traditions spanning at least 300 years, from the 8th century to the 5th century BCE.

Now, there is a serious scholarly conversation about the unity of the final book — and it deserves a fair hearing rather than a brush-off. Seitz, whose reading of the "former things" I took up above, argues that later editors shaped the book into a coherent whole, so that chapters 1–39 are meant to be read in light of what follows; he speaks of "former things that were revealed long ago to Israel and are now on the point of fulfillment" (Seitz 1993, 155). This is a genuine literary insight, and it is not the same thing as the apologetic claim. Notice what Seitz is and is not saying. He is saying the redactors, the editors who assembled the book, gave it a unified theological shape. He is not saying one 8th-century prophet predicted Cyrus by name. The final-form unity is a product of editorial work spanning centuries — which is to say, it presupposes exactly the multiple authorship and post-exilic redaction the apologetic reading needs to deny.

This doesn't mean the book is worthless, or that it can't be religiously meaningful. It means that reading it as a unified set of predictions by one prophet of the eighth century is historically indefensible. And building your apologetic case on that reading is building on sand.

That is, of course, my assessment of the evidence. The multiple-author conclusion underneath it is not mine alone: it is the working assumption of Blenkinsopp, Goldingay and Payne, Duhm, Westermann, Whybray, and nearly every critical commentator since Duhm, a confessionally mixed list. What follows from it for the prophecy argument is my own inference. Even Brevard Childs — who reads the final form of Isaiah as a deliberate canonical unity and is no friend of letting authorial-layer reconstruction become the last word on meaning — explicitly granted that this unity "cannot be formulated in terms of single authorship," and in the same breath distinguished himself from "the traditionally conservative approach represented by E. J. Young, Oswalt, and Motyer" (Childs 2001, 3). That's the point worth underlining: you can think the book hangs together as a coherent theological whole, as Childs did, and still take the multiple-author compositional history as established. What you cannot do, and keep faith with the evidence, is read the book as a set of literal predictions issued by Isaiah ben Amoz. The question is not whether Isaiah was written by multiple authors. The question is whether you are willing to let that established conclusion inform your evaluation of the prophecy argument. If you are, the argument cannot survive. If you aren't — if you insist on single authorship despite the evidence — then you have left the realm of historical inquiry and entered the realm of faith commitment.

And that's fine. But don't dress it up as evidence.