HistoricalChristian.Faith

"He descended into hell" — is the creed clause original, and should it be retained?

Type: supporting argument (off-timeline) · Assessment: contested — as a matter of text-history the clause is a late, non-original addition (absent from the oldest form of the creed, first surfacing in an Arian-leaning council creed, taken by Rufinus to mean only "buried," and not in the received Western text until the eighth century), so the historical half of the case to drop it holds; but whether to retain it is genuinely divided, because the descent doctrine the clause summarizes is older than the clause and itself sound · argued by Wayne Grudem (omit, JETS 1991) — joined on exegetical grounds by John Piper (omit, 2012) — against David Scaer and the evangelical retrieval movement (retain, JETS 1992)

Related arguments: the descent — see also descent to the dead · the raising of the saints · relocation of Paradise · content of the proclamation · excised descent-prophecy · forerunner in Hades

This argument concerns the creed clause's pedigree and retention — where the words "he descended into hell" came from and whether they belong — and is kept deliberately distinct from the companion descent-to-the-dead argument, which weighs the doctrine the clause points to (did Christ's soul go to the realm of the human dead and free the OT righteous?). One can hold the doctrine sound and the clause a latecomer at once; the two questions must be answered separately.

The claim

In this article my purpose is to give reasons why it seems best to consider the troublesome phrase "he descended into hell" a late intruder into the Apostles' Creed that really never belonged there in the first place and that, on historical and Scriptural grounds, deserves to be removed.

— Wayne Grudem, He Did Not Descend into Hell: A Plea for Following Scripture Instead of the Apostles' Creed, JETS 34/1 (1991), p. 103, waynegrudem.com.¹ The contrary claim, that the clause should stand:

I can respect Wayne Grudem's decision to omit "he descended into hell" from his recitation of the Apostles' Creed within the congregation … More importantly, there are good Biblical, historical and theological reasons for retaining the phrase.

— David P. Scaer, He Did Descend to Hell: In Defense of the Apostles' Creed, JETS 35/1 (1992), p. 91, etsjets.org.² Two questions are tangled here and must be untangled: (a) the textual-historical question — is the descent clause an original/apostolic article of the creed or a late interpolation, and does its first firm attestation being an Arian-leaning creed matter? — and (b) the normative question — given the answer to (a), should the clause be retained, dropped, or reworded?

Who advances it

The omission case is put most fully by Wayne Grudem (Reformed/evangelical systematician), who recites the creed in church without the line and argued in JETS (1991) that it should be dropped "once for all." The same conclusion is reached on different ground by John Piper, who would also drop the clause — not from its pedigree but from its exegesis (weighed in the assessment below). The retention case is put by the Lutheran dogmatician David Scaer in a direct JETS reply (1992), and, more recently, by an evangelical retrieval movement — Matthew Emerson's "He Descended to the Dead": An Evangelical Theology of Holy Saturday (2019) and writers at The Gospel Coalition — who urge keeping the article (often reworded "he descended to the dead") rather than excising it.³ The underlying textual history is not in dispute between them; the sources below are agreed by both sides, who differ only over what follows from them. Within the timeline, the doctrine the clause encodes is restated by Doug Wilson, whose own descent material is weighed on the descent-to-the-dead page.

The textual-historical record

The pedigree of the clause is well documented, and on the bare facts the "late addition" reading is essentially mainstream:

1. It is absent from the oldest form of the creed. The earliest extant form of the Roman baptismal creed — the Old Roman Symbol, preserved in Greek by Marcellus of Ancyra (c. 340) and in Latin by Rufinus — runs "… crucified under Pontius Pilate and buried; the third day he rose again from the dead …" with no descent clause between burial and resurrection.⁴
2. Its earliest appearance in any creed is the "Dated Creed" of 359. The Fourth Creed of Sirmium (the "Dated Creed," 22 May 359) is the first creedal text to carry a descent clause: Christ "was crucified, and died and descended into the parts beneath the earth, and regulated the things there, Whom the gate-keepers of hell saw (Job 38:17) and shuddered." But this is a Homoian (Arian-leaning) council creed — its heterodoxy lies in its avoidance of ousia/"substance" language, not in the descent line — and it is an Eastern conciliar creed, not a textual ancestor of the Western Apostles' Creed. It survives only because Athanasius (De Synodis 8) and Socrates Scholasticus (Church History II.30) quote it to refute it; neither composed the wording.⁵
3. Within the Apostles'-Creed family it is first attested at Aquileia, and Rufinus flags it as local. Rufinus of Aquileia (c. 404), expounding the creed of his own church, expressly notes that the clause is missing from the Roman and Eastern creeds and that its sense is already carried by "buried": "it should be known that the clause, He descended into Hell, is not added in the Creed of the Roman Church, neither is it in that of the Oriental Churches. It seems to be implied, however, when it is said that He was buried."⁶
4. It enters the received Western text only in the eighth century. The textus receptus — the complete, now-standard Latin form including "descendit ad inferna" — is first attested whole in the Scarapsus of Pirminius (c. 710–750); as Schaff puts it, "the earliest creed actually and in all respects identical with [the received form], that of Pirminius, does not occur till the eighth century."⁷

The retention debate

Grudem (omit). Grudem builds the normative case on the pedigree: the phrase "was not found in any of the early versions of the Creed … until it appeared in one of two versions from Rufinus in A.D. 390. Then it is not included again in any version of the Creed until 650," and even Rufinus "did not think that it meant that Christ descended into hell but understood the phrase simply to mean that Christ was 'buried.'" His conclusion: the clause "has no clear warrant from Scripture … no claim to being apostolic … was not in the earliest versions of the Creed and later was included … because of an apparent misunderstanding," so "there would be all gain and no loss if it were dropped from the Creed once for all."¹

Scaer and the retrieval (retain). Scaer grants Grudem's history but denies the inference. The clause encodes a biblical motif older than the clause itself: "1 Peter 3:18–22 has an outline strongly resembling our creed: Christ's death and glorification; his preaching in the prison … his resurrection; and his ascension," so "our Apostles' Creed with the descent clause is maintaining a fundamental Biblical motif." To excise it costs something: "By eliminating the descent from the creed we lose the one reference to the victory of Christ and the defeat of God's major opponent. In following the Apostles' Creed we are following the Scriptures."² The modern retrieval makes the same move — the descent is an ancient, near-universal patristic confession (woven through the timeline above and adjudicated sound, on its narrowest form, on the descent-to-the-dead page), and a burial already implies a descent, so the right remedy is to confess the clause well, not to delete it.³

The confessions split the difference. Present and unchanged in the Church of England's 1662 Book of Common Prayer ("… crucified, dead, and buried: He descended into hell; The third day he rose again …") and in the Lutheran Church–Missouri Synod's liturgy; reworded to "he descended to the dead" in many recent ecumenical texts; and quietly dropped in the traditional United Methodist form (American Methodists removed it in 1792), with the 1989 United Methodist Hymnal restoring "he descended to the dead" only in its ecumenical version and relegating "into hell" to a footnote.⁸ The Lutheran tradition does more than retain the clause: the Formula of Concord (1577) builds an entire confessional article on it, directing the churches to Luther's 1533 Torgau sermon and reading the descent as Christ's victory — "the entire person, God and man, after the burial descended into hell, conquered the devil, destroyed the power of hell" — a reading distinct alike from Calvin's (the soul's torment on the cross) and from the Catholic emptying of the limbus patrum, and a pointed counter to the omission case.¹¹

Adversarial assessment

What holds. The textual-historical spine of Grudem's case is sound and not seriously contested by Scaer: the clause is demonstrably not original to the Apostles' Creed. It is absent from the Old Roman Symbol, local to Aquileia when Rufinus writes (who says so plainly), and only universal in the West from the eighth century. On the narrow question "is the clause an original/apostolic article?" the answer is no.

What is overstated or imprecise.⁹

1. Grudem's first-appearance timeline is imprecise. He dates the clause's debut to "Rufinus in A.D. 390"; but Rufinus is conventionally dated c. 404, and a descent clause appears in a creed earlier still — the Dated Creed of 359. That refines rather than overturns his point (the Sirmium creed is a separate Eastern document, not the source of the Roman clause), but the clean "390 → 650 gap" should be handled with care.
2. The Arian provenance of the first attestation does not taint the clause. That the earliest creedal descent clause sits in an Arian-leaning creed is a genealogical curiosity, not a doctrinal stain: the heterodoxy of the Dated Creed is its homoian dodge on ousia, not its descent line, and the Apostles'-Creed clause is not textually descended from Sirmium anyway. The descent doctrine itself is pre-Nicene and confessed across the orthodox tradition (Irenaeus, Clement, Origen, Cyril of Jerusalem), so the clause's lateness as words does not make the doctrine late.
3. "Late words" does not entail "drop it." This is the load-bearing weakness in the omission case. Creeds legitimately develop; the test of an article is not whether the exact phrase is apostolic but whether it confesses apostolic teaching. Rufinus's "it means He was buried," which Grudem reads as evidence of confusion, is just as naturally read as evidence of continuity — the clause was felt to add nothing new because the descent was already believed. The retrieval side's reframing (keep it, confess it rightly, reword to "to the dead") meets Grudem's real worry — that "into hell" wrongly suggests Christ suffered in gehenna — without excising the article.
4. Both sides over-read 1 Peter 3, in opposite directions. Scaer rests retention partly on 1 Peter 3:18–22 as a descent text; but that passage is itself genuinely disputed (the "spirits in prison" may be the imprisoned angels, not the human dead — see the 1 Peter 3:19 argument). So the clause's scriptural warrant is firmer from Acts 2:27/31 and the broad patristic descensus than from 1 Peter alone, and Grudem is right to deny that 1 Peter 3 by itself settles it — while overreaching when he treats the doctrine as nearly unscriptural.

A second omission voice on exegetical grounds (Piper)

Grudem's is not the only modern evangelical case for dropping the clause. John Piper reaches the same verdict by the opposite route — not the clause's pedigree but its exegesis. He grants no New-Testament warrant for a preaching-in-Hades: he reads 1 Peter 3:19 of Christ preaching "through the voice of Noah" to Noah's own generation, "whose spirits are now 'in prison,'" and 1 Peter 4:6 of those who "have since died," not of any postmortem proclamation; and he points to the penitent thief, "with me in paradise" that very day, as evidence that the righteous dead need no holding-place from which to be later delivered:

With regard to 1 Peter 3:19, I take these words to mean that Christ, through the voice of Noah, went and preached to that generation, whose spirits are now "in prison," that is, in hell. In other words, Peter does not say that Christ preached to them while they were in prison. He says he preached to them once, during the days of Noah, and now they are in prison.

I would say, therefore, that there is no textual basis in the New Testament for claiming that between Good Friday and Easter Christ was preaching to souls imprisoned in hell or Hades … For these and other reasons, it seems best to me to omit from the Apostles Creed the clause, "he descended into hell," rather than giving it other meanings that are more defensible, the way Calvin does.¹⁰

Piper's route is, if anything, more vulnerable than Grudem's. Where Grudem's textual-historical spine is secure, Piper's omission rests on two contested exegetical judgments — that the "spirits in prison" are Noah's generation evangelized in their own day (one option among several on the 1 Peter 3:19 identity argument) and that the descent proclaimed nothing to the dead (the content-of-the-proclamation argument). And his own positive datum tells only against the strong holding-place form, not against the bare descensus, which is creedal and grounded in Acts 2:27/31 quite apart from 1 Peter. So Piper reinforces the case that the clause is dispensable only by leaning on readings that are themselves live arguments — which is part of why "omit" stays a minority verdict even among those who grant the clause's late pedigree.

Assessment

Contested. Two questions, two different answers. On the textual-historical question the verdict is firm and favors Grudem: the clause is a late, non-original addition — absent from the Old Roman Symbol, first appearing in a creed at Arian-leaning Sirmium (359), local to Aquileia when Rufinus writes (c. 404) and taken by him to mean only "buried," and not in the received Western text until the eighth-century Scarapsus of Pirminius. Anyone claiming the words are apostolic or original is simply mistaken. But the normative question — should the clause be retained? — is genuinely divided, and that is where the dispute really lives. The omission case overreaches in three places: it treats the Arian provenance of the first attestation as if it discredited the clause (it does not), it slides from "the words are late" to "the article should go" (creeds develop, and a late phrase can confess early doctrine), and it leans on the contested reading of 1 Peter 3 cutting only its way. The retention case is stronger than its pedigree looks because the doctrine the clause summarizes is ancient and, on its narrowest form, sound (see the descent-to-the-dead argument). Net: the clause is rightly called a latecomer (sound history) but wrongly called dispensable on that ground alone (the retention debate is live, with the better of the theology on the retrieval side and the better of the philology on Grudem's). Strong by one discipline's lights, weaker by another's — the textbook shape of a contested argument. (John Piper's parallel case for omission is weaker still: it rests not on the firm pedigree but on contested readings of 1 Peter — see above.) The fairest practical resolution, and the one most of the tradition is converging on, is to keep the article while rewording it "he descended to the dead," which preserves the apostolic doctrine and removes the misleading suggestion that Christ suffered in gehenna that worried Grudem in the first place.

Sources