Ezekiel's Shepherd-King, Reduced to One Line

One of the most widely circulated fulfilled-prophecy lists on the internet, AccordingtotheScriptures.org's "356 Prophecies Fulfilled in Jesus Christ," has exactly two entries from Ezekiel. Entry 298 is Ezekiel 34:23-24, described as "Descendant of David" and matched to Matthew 1:1. Entry 299 is Ezekiel 37:24-25, also described as "Descendant of David" and matched to Luke 1:31-33. That is everything the list says about them. Two of the fullest royal oracles in the prophetic books each come down to three words and a cross-reference: the first line of Matthew's genealogy for one, and Gabriel's promise of "the throne of his father David" for the other.

The label is not wrong. Both passages do promise a ruler from David's house, and Christian readers since the New Testament have heard them as promises about Jesus. So the interesting question is not whether the list includes Ezekiel. It is what the list keeps from Ezekiel, and what it leaves on the floor. Set the one-line entries beside the two chapters and the gap is wide. The list keeps the lineage and, through Luke, the word "forever". It drops the reunion of Israel and Judah, the land, the statutes, the sanctuary, and the temple-state of Ezekiel 40-48, where the "prince" brings a sin offering "for himself and for all the people of the land" (45:22). What gets dropped is not incidental. It is most of what the passages say.

The list gives the same three words to Jeremiah 23:5-6 (entry 293) and Jeremiah 33:14-15 (entry 297). Four different oracles, from two prophetic books, share one identical line (Jeremiah 23:5-6 gets a second entry as well, "The Messiah would be both God and Man"). This essay puts the two Ezekiel passages back together and works them through several lenses, naming each as it goes: their literary structure, their ancient Near Eastern and philological background, their long Jewish and Christian reception, and finally the ideological question of whose reading the one-line entry serves.

The One Shepherd and the Shepherd Before Him

Ezekiel 34:23-24 reads, in the World English Bible:

The literary lens first. These verses close a chapter that moves in three steps. Verses 1-10 indict the shepherds of Israel, the kings who fed themselves instead of the flock. Verses 11-22 do something readers often skip: God takes the job himself. "Behold, I myself, even I, will search for my sheep," he says (34:11), and "I myself will be the shepherd of my sheep" (34:15); then he judges "between sheep and sheep" (34:17). Only after that does he appoint "one shepherd… my servant David." The nineteenth-century commentator Malbim (Meir Leibush Wisser) reads that order as the point. Only after God has judged and shepherded the flock himself is the Davidic kingship revealed, and even then it rises in stages, until "my servant David" climbs from the rank of "shepherd" to the rank of "prince", "who will be the King Messiah" (on 34:23; my translation). The chapter then ends not with the king but with the land: a "covenant of peace", "showers of blessing", and a people "secure in their land" (34:25-27).

Now the comparative lens. Shepherd language for kings is about as conventional as ancient Near Eastern royal rhetoric gets. In the epilogue to his law code Hammurabi says, "I am the salvation-bearing shepherd, whose staff is straight" (L. W. King's translation), and Egyptian kings carried the crook as an emblem of rule. Ezekiel is not coining an image; he is using the political vocabulary of his world. So is Jeremiah, a generation earlier, who runs the same sequence: "Woe to the shepherds who destroy and scatter the sheep of my pasture!" and then "I will raise to David a righteous Branch" (Jer 23:1, 5). Ezekiel 34 belongs to the restoration program of chapters 33-48, written in Babylonian exile after the fall of Jerusalem in 587/586 BCE.

"David" here is almost certainly a dynastic name, not a prediction that the historical David will come back. The medieval commentators saw this too. Isaac Abarbanel (1437-1508), statesman and exegete, compares it to Egyptian royal names: "just as all the kings of Egypt were called by the name Pharaoh for a long time, and afterwards by the name Ptolemy for a long time, so the King Messiah and all his seed will be called by the name David" (on 37:24; my translation). He knows the alternative, and does not dismiss it. On 34:24 he calls the dynastic reading "correct", then reports that the kabbalists, who held the doctrine of the transmigration of souls, said the King Messiah would be David himself, David's soul returned in him. Ezekiel, he adds, repeats that view at 37:25 ("David my servant will be their prince forever"), and he has defended transmigration against the philosophers' objections elsewhere, in his commentary on Deuteronomy (my translation).

Then the reception lens, beginning with the oldest witness. Targum Jonathan, the ancient Aramaic rendering of the Prophets, is far more restrained than the later commentators. At 34:23 it drops the name David altogether: God will raise up "a single leader" (parnas) who will lead them. At 34:24 it turns Ezekiel's "prince" into "king" (malka). What it never does, here or anywhere else in its Ezekiel, is use the word "messiah". That silence is telling, because the same Targum inserts the word freely elsewhere. At Hosea 3:5, where the Hebrew has "David their king", the Targum reads "the Messiah son of David, their king" (my translation), and at Jeremiah 23:5 and Zechariah 6:12 it names the Messiah outright.

The medieval commentators go further, though not all by the same road. Rashi's entire comment on "my servant David" at 34:23 is two Hebrew words, which A. J. Rosenberg renders "A king [who will come] from his descendants": a future Davidic king rather than David himself. David Kimhi (Radak) is explicit: "the Messiah who will arise from his seed at the time of salvation" (my translation). Malbim, as we saw, reaches "the King Messiah" in stages. And Abarbanel is blunt about history. This prophecy, he writes, "was not fulfilled in the Second Temple, but will be fulfilled in the days of the Messiah" (my translation). He gives three proofs: the exiles came back only from Babylon, not from every land; throughout the Second Temple period no king of the house of David reigned; and the promised covenant of peace never came, since the returned community was always at war with its enemies.

Maimonides, whose treatment of the messianic idea in the Mishneh Torah (Laws of Kings 11-12) is the most influential codification of it, does not cite Ezekiel there at all. He works chiefly from Deuteronomy 30, Numbers 24:17-18, Zephaniah 3:9, Malachi 3 and Isaiah 11. Yet his Messiah is an Ezekiel-shaped figure, a Davidic king who restores the kingdom, builds the Temple and gathers the dispersed of Israel (11:1). He even writes that the prophets' words, read simply, "appear to imply that the war of Gog and Magog will take place at the beginning of the Messianic age" (12:2, Touger's translation). Gog and Magog are Ezekiel 38-39, the chapters that follow the dry bones.

The steelman. Here the Christian reading has its strongest ground, and it deserves to be stated at full strength. In Ezekiel 34, God first says "I myself will be the shepherd of my sheep" (34:15), and then appoints "one shepherd… my servant David" (34:23). The chapter has two shepherds, one divine and one Davidic. John's Jesus claims both roles at once: "I am the good shepherd," who gathers the scattered so that "they will become one flock with one shepherd" (John 10:11, 16). Christian readers hear the rest of the chapter in the other Gospels too. Behind "the Son of Man came to seek and to save that which was lost" (Luke 19:10) they hear "I will seek that which was lost" (Ezek 34:16). Behind the king who separates the nations "as a shepherd separates the sheep from the goats" (Matt 25:32) they hear the God who judges "between sheep and sheep" (34:17). For a Christian reader, Ezekiel 34 is not generic royal boilerplate. It is a chapter whose two shepherds converge in one person.

That reading is real, and the echoes are real. But notice what it takes and what it leaves. John takes the pastoral image, the shepherd who knows, gathers and lays down his life. He does not take the political content that surrounds the image in Ezekiel: the covenant of peace, the rain, the people "secure in their land" who "will no more be a prey to the nations" (34:27-28). The Gospels map onto Ezekiel 34 in its pastoral imagery, and there the mapping stops. The list, for its part, does not even make John's case. It files John 10's shepherd under Psalm 23:1 (entry 94, "I am the Good Shepherd") and under Isaiah 40:11 (entry 212, "A shepherd-compassionate life-giver"), and it pairs John 10:16, the "one flock with one shepherd" verse itself, with Genesis 49:10 (entry 21, "Unto Him shall the obedience of the people be"). Ezekiel 34, the chapter that actually says "one shepherd", gets only the genealogy.

The Dry Bones and the King Who Never Came

The second passage, Ezekiel 37:24-25, closes the chapter of the valley of dry bones:

The literary lens again. The chapter has three movements. First the bones are revived (37:1-14). Then two sticks, one written for Judah and one for Joseph, are joined "into one stick" (37:15-23): "They will no longer be two nations." Last comes a Davidic king ruling the reunited people in the land, with God's sanctuary "among them forever more" (37:24-28). The text glosses its own vision: "these bones are the whole house of Israel" (37:11). In context the resurrection is a nation's, back from exile. Malbim reads the reunion in 37:22 in the same national terms: the two peoples and the two kingdoms will not be divided again. He also has the kingship pass in two stages, with a Messiah son of Joseph reigning first and the son of David after him (my summary of his comment; on that figure see Messiah ben Joseph).

The tradition read more into the bones, and honesty about that cuts both ways. The Talmud records an argument over whether the dead whom Ezekiel revived literally rose. Rabbi Judah calls the vision both "truth" and "a parable", and Rabbi Nehemiah resolves the paradox: "In truth, it was a parable." But the argument does not end there. Rabbi Eliezer, son of Rabbi Yosei HaGelili, answers that the revived dead went up to the land of Israel, married and had children, and Rabbi Judah ben Beteira claims to own phylacteries handed down from them (Sanhedrin 92b, William Davidson translation). Radak holds both options open. The vision may be a parable of Israel coming out of exile, or a sign that God will "revive the dead of Israel at the time of salvation" (on 37:1; my translation). At 37:24 he allows that "David" might be "a hint at the resurrection of the dead", David himself raised. Abarbanel thinks the vision itself more likely a parable, but a parable of a real bodily resurrection near the time of the redemption. So it would be wrong to say that the Jewish tradition read Ezekiel 37 as purely political. What these readers share is narrower: the king they expect is a human ruler in the land. Maimonides puts the point crisply. One should not presume that the Messiah must "work miracles and wonders… resurrect the dead" (Kings 11:3, Touger's translation).

The philological lens. Look at the titles. Ezekiel says "king" (melek) at 37:22 and 37:24, but "prince" (nasi) at 37:25 and at 34:24. Nasi is also the title of the ruler in the temple chapters, where his role is strikingly limited. He sits in the gate "to eat bread before Yahweh" (44:3) and supplies the offerings "to make atonement for the house of Israel" (45:17). How "messianic" Ezekiel's own figure is therefore remains a live scholarly question, and it is not one I can settle here; Paul Joyce's "King and Messiah in Ezekiel" (1998) and Daniel Block's "Bringing Back David" (1995) both take it up. Whether the prince of chapters 40-48 is the same figure as the "David" of chapters 34 and 37 is itself disputed. The title, at least, is the same. The Targum moves the other way: at 37:25, as at 34:24, it upgrades the "prince" to a "king" (malka), and at 37:24 it keeps "my servant David, king over them" while rendering "one shepherd" as "one leader".

The steelman on "forever". The Christian argument here is simple, and it is not weak. No mortal king reigns forever, so a promise that "David my servant will be their prince forever" can only be kept by a king who does not die. Luke 1:33 is built on exactly that ("he will reign over the house of Jacob forever"), and Luke 1:31-33 is the list's own pairing for this entry. Two things answer it. The first is the genre. The Hebrew olam belongs to the language of dynastic grants. Moshe Weinfeld argued that the covenant with David is modelled on the ancient Near Eastern royal grant, a gift bestowed on a servant for loyal service and secured to his heirs. In that idiom, as I read it, "forever" guarantees a dynasty, not an immortal individual. The second answer is in the verse itself. The same "forever" governs the people, who will dwell in the land, "they, and their children, and their children's children, forever." Nobody reads that as a promise of immortal Israelites. It is a promise of unbroken succession, and a single king who never dies is a different promise.

The land. "The land that I have given to Jacob my servant" is a territorial promise. Rashi's one note on the verse makes it more so: the land will be given "As I gave it to Jacob, without boundaries" (on 37:25, Rosenberg's translation). Nothing in these verses describes a heavenly reign.

The steelman on the statutes. Under this king, the people "will also walk in my ordinances, and observe my statutes, and do them" (37:24). Christian readers answer that Ezekiel has already explained how that obedience comes about: "I will put my Spirit within you, and cause you to walk in my statutes" (36:27). They read it as the new covenant's law written on the heart (Jeremiah 31:33; Hebrews 8:10). Taken alone, that is a fair reading of 36:27. But Ezekiel does not leave "my statutes" abstract. Eight chapters later he spells them out: a priesthood of "the sons of Zadok" who "offer to me the fat and the blood" (44:15), a festival calendar, the Sabbaths and new moons, and a prince who supplies the burnt offerings and sin offerings (chapters 44-46). Ezekiel's Spirit produces a people who keep that. Most Christian theology holds that those sacrifices have ended in Christ. Maimonides, by contrast, sets at the centre of his messianic chapter the principle that "This Torah, its statutes and its laws, are everlasting" (Kings 11:3, Touger's translation). On this point he is following Ezekiel's direction of travel, and the Christian reading has to redirect it.

The steelman on the temple. The obvious Christian reply to all this is that the New Testament does take up Ezekiel's temple, only transposed. Revelation's new Jerusalem is built out of Ezekiel 40-48. John's guide measures the city with "a golden reed" (Rev 21:15), as Ezekiel's guide carried "a measuring reed" (Ezek 40:3). The city has "twelve gates" bearing "the names of the twelve tribes of the children of Israel" (Rev 21:12), as Ezekiel's city has gates "named after the tribes of Israel" (Ezek 48:31). "A river of water of life" flows "out of the throne of God and of the Lamb", lined with trees whose "leaves… were for the healing of the nations" (Rev 22:1-2), as Ezekiel's river flows from the sanctuary and its trees bear leaves "for healing" (Ezek 47:1, 12). Then comes the decisive line: "I saw no temple in it, for the Lord God, the Almighty, and the Lamb, are its temple" (Rev 21:22). That is a serious reading, and no late invention; it is in the canon. But it is a reinterpretation. It keeps Ezekiel's imagery and drops his altar, his priests, his prince and his tribal allotments. Ezekiel wrote a building code. Revelation turned it into a vision without a building.

Some Christian readers, dispensationalists above all, refuse that move and take Ezekiel 40-48 literally, as a temple still to come in a future reign. The Scofield Reference Bible (1917) says of the temple's sacrifices, "Doubtless these offerings will be memorial, looking back to the cross" (note on Ezekiel 43:19). On that view the list is not dropping Ezekiel's land, statutes and sanctuary, only deferring them: the lineage was fulfilled at the first coming, and the rest awaits the second. That is a coherent position, and it avoids the reinterpretation. But it concedes the point that matters here. By its own account most of what these two passages promise has not happened yet, and a list of fulfilled prophecies that shows the one line that has happened, without saying that it is one line, still tells its reader a fraction of what Ezekiel said.

What the One Line Leaves Out

The ideological lens, finally: whose reading does the one-line entry serve? Here are the two passages set beside what the list does with them. The list has no other Ezekiel entries, so every "No" below means that element of Ezekiel's promise is not on the list.

The pattern is not special to Ezekiel. Nathan's oracle in 2 Samuel 7 is on the list four times, and two of its entries pull out "His Kingdom is everlasting" and "The Son of God" as separate fulfilments (see the overlapping prophecies). Where a passage carries a detail that can be matched to a New Testament verse, the list splits it into several items. Where the rest of a passage is about land, statutes and sanctuary, the rest does not appear. Ezekiel's two oracles are kept, but only as the genealogy line.

That is the extraction method in its plainest form. A prophecy list works by lifting specific details out of their literary and historical setting and lining them up with New Testament episodes, and a lineage here, a birthplace there, a manner of death somewhere else, soon adds up to an impressive-looking pattern. Ezekiel 34 and 37 show what the method has to do to a passage whose meaning cannot be separated from its setting. They are core texts of exile-era Davidic hope. They describe a king who rules in the land, governs by Ezekiel's statutes, reunites the two kingdoms, and presides over a sanctuary with a priesthood and a sacrificial calendar. The list can include them only by reducing each to "Descendant of David", the one element that fits.

Christian readers have answers to each dropped element, and I have tried to put them at their strongest: the pastoral shepherd of John 10, the Spirit of Ezekiel 36:27, the everlasting king of Luke 1:33, and the temple that becomes God and the Lamb in Revelation 21:22. Each is a real reading. But each is a reinterpretation of what Ezekiel said, not a fulfilment of it, and a list that counts fulfilments hides the difference. The list does not catalog what the Hebrew Bible says about the messiah. It catalogs what can be made to look like it maps onto the Gospels. Those are two very different things, and entries 298 and 299 show the difference as clearly as anything on the list.

That is how I weigh the evidence. Others will weigh it differently. But the choice the two entries pose is a real one. Either a prophecy means what its whole passage says, in which case Ezekiel's shepherd-king is a ruler in the land who governs by the statutes and presides over a rebuilt sanctuary, and Jesus has not yet fulfilled that prophecy (which the most literal Christian readers themselves concede, by placing most of it in a future still to come). Or a prophecy can be cut down to one line, "Descendant of David", and then it proves very little, because almost any claimant of David's line satisfies it. The list needs the first kind of prophecy for its argument, and uses the second kind for its entries.