HistoricalChristian.Faith

The patristic delay of baptism: does the widespread deferral of baptism show infant baptism was not universal or apostolic?

Type: critical argument (off-timeline) · Assessment: contested — the datum is real and underrated: for centuries baptism was routinely delayed, often to adulthood, even in devout Christian homes. Tertullian counsels delay (esp. of “little children”); the Shepherd of Hermas’s one-repentance-after-baptism doctrine supplies the motive; the three Cappadocians and Ambrose, Jerome, Rufinus, and Chrysostom were all baptized only as adults though reared Christian; Augustine’s devout mother deferred his baptism when he sickened as a boy; and the fourth-century vogue for deathbed baptism was widespread enough that a ninth-century paedobaptist (Walafrid Strabo) could still recall that “in ancient times” baptism was given to the matured. This genuinely complicates a strict always-universal-from-the-apostles reading — which is why Jeremias had to treat the delay as a temporary decline from an earlier norm and Aland pressed it as evidence of development. But it does not carry the further step to non-apostolic or not-practised: the delay was an elective preference within churches that had infant baptism (emergency infant baptism was always available, and prescribed whenever life was at risk), it was driven by post-baptismal-sin anxiety rather than by any denial that infants are fit subjects, the first writer to mention infant baptism (Tertullian) argues against a custom that plainly already existed, and the same tradition resisted the delay — Cyprian’s council rejected even an eight-day postponement. So the thesis wounds the universality case without proving absence, and lands in the same Jeremias–Aland standoff as its companion · advanced by the credobaptist historians (Kurt Aland; in modern form Thomas Schreiner, Philip Schaff), answered by Joachim Jeremias and the paedobaptist reading of the same fathers

Related arguments: the historical case — see also dating the earliest mentions · antiquity and universality · Irenaeus’s “reborn to God” · Polycarp’s eighty-six years

The claim

This is the mirror image of the paedobaptist antiquity-and-universality argument, run for the other side. Where that case reasons from the early, uncontested spread of infant baptism to its apostolic origin, this one reasons from the early, widespread deferral of baptism to the opposite conclusion: if devout Christian families across the empire routinely postponed baptism — often to adulthood — then infant baptism cannot have been the universal, unquestioned, apostolic default the paedobaptist claims. The evidence is a chain of witnesses to delay:

• The counsel to delay. Tertullian, the earliest writer to discuss the question (c. 203), judges “the delay of baptism… preferable; principally, however, in the case of little children,” and would have even the unmarried and the widowed wait: let children “become Christians when they have become able to know Christ.”¹ Gregory of Nazianzus (c. 381), absent danger, counsels waiting for a healthy child “till the end of the third year, or a little more or less,” when it can “answer something about the Sacrament.”²
• The motive. The Shepherd of Hermas (2nd c.) taught that only one repentance is granted after baptism — a doctrine that made post-baptismal sin so grave a hazard that many judged it prudent to defer the washing until the worst temptations of youth had passed, and, at the extreme, to the deathbed.³
• The practice, even among the devout. The three Cappadocians are the marquee case: Basil the Great (from a famously devout Christian family) and his brother Gregory of Nyssa were baptized only as adults, as were Ambrose, Jerome, Rufinus, and Chrysostom. Augustine, son of the devout Monica, records that when he fell gravely ill as a boy and begged for baptism his mother made ready in haste — but on his sudden recovery his “cleansing was deferred,” on the reasoning that a survivor would only sin more and so incur graver post-baptismal guilt.⁴ Eusebius relays (from Clement of Alexandria) a presbyter who took a promising youth and “reared, kept, cherished, and finally baptized him” — a convert catechized and only later washed.⁵
• The long memory of it. The delay was common enough to leave a mark on later writers: Walafrid Strabo (c. 840), himself no opponent of infant baptism, wrote that “in ancient times” baptism was administered to those who had already matured.⁶

The conclusion drawn: a church in which baptism is routinely put off to adulthood, even by its most pious families, is not a church for which the baptism of newborns is the settled, apostolic norm.

Who advances it

The delay data is old — the fathers report it of themselves — but its use as an argument against paedobaptist universality is modern and credobaptist. Its ablest scholarly form is Kurt Aland’s reply to Joachim Jeremias, which read the widespread fourth-century deferral as evidence that general infant baptism was a development, not an apostolic institution; more recent credobaptist surveys (Thomas Schreiner on the post-Nicene evidence, Philip Schaff’s catalogue of the delaying fathers) press the same point.⁷ It is the natural counter-move to the paedobaptist no-founding-controversy argument: if the historical record is to settle the question, then the record of delay is part of the record.

The evidence for

Taken as a historical (not demonstrative) argument, the case has genuine, and often underrated, weight:

• The delay is real, wide, and reaches the most devout. This is the argument’s strongest point. These are not marginal figures: Basil, Gregory of Nyssa, Ambrose, Chrysostom, Jerome, Rufinus — bishops, doctors, sons of saints — baptized as adults though reared in the faith from the cradle. Monica, a canonized model of Christian motherhood, deferred her own son’s baptism. If infant baptism were the plainly-apostolic, non-negotiable default the strong paedobaptist claim requires, this pattern is hard to explain: one would expect the saints, of all people, to have been washed as newborns.
• It shows infant baptism was not the universal default even in the fourth century. The modest form of the thesis is simply true, and it is a real correction to the loose paedobaptist claim that the early church always and everywhere baptized its infants. It did not: through the fourth century a credobaptist-style deferral — faith and instruction first, the font later — was mainstream, common, and untroubled, practised by the very fathers the paedobaptist elsewhere cites.⁸
• It forces a concession from the paedobaptist’s own best historian. Jeremias, the modern champion of apostolic continuity, could not deny the fourth-century delay; he had to explain it as a temporary decline from an earlier, more general infant-baptism custom — conceding that the surface of the fourth-century evidence points the other way, and that his continuity thesis must reconstruct an earlier norm behind it. That the datum needs explaining-away is itself a measure of its force.

Adversarial assessment

What holds. The modest claim — that infant baptism was not the unquestioned, universal default in the patristic church, and that mainstream, devout families routinely delayed baptism into adulthood — is well-evidenced and correctly punctures the overstated paedobaptist “always and everywhere” rhetoric. Anyone who claims the early church uniformly baptized its newborns has to reckon with Basil, the two Gregories, Ambrose, Chrysostom, Jerome, and Monica’s son.

What is disputed. The step from that modest claim to the strong thesis — that delay shows infant baptism was not practised or not apostolic — overreaches on four counts.

• Delay is a preference within a church that has the rite, not evidence of its absence. This is the decisive objection. In every witness adduced, elective delay coexists with an available infant baptism that is used the moment it is needed. Gregory of Nazianzus, in the very passage that counsels waiting to age three, prescribes immediate infant baptism “if any danger presses” — better “unconsciously sanctified than… unsealed and uninitiated.” The infant is a fit subject throughout; only the timing, for a healthy child, is deferred. A parent who delays a baptism he could obtain at once, and would rush to obtain if his child sickened, testifies to the rite’s availability, not its absence.⁹
• The motive was post-baptismal-sin anxiety, not a denial that infants may be baptized. Not one of the delaying fathers says an infant is an unfit subject; they say baptism’s grace is too precious to spend before the sins of youth are past. That is the Shepherd of Hermas’s one-repentance logic, Tertullian’s worry about the sponsors’ risk, and the reasoning Augustine reports of Monica — a fear of forfeiting baptismal grace, which presupposes that the grace was on offer to the child. The delay is a mistaken prudence about infant baptism, not a rejection of it — and the credobaptist cannot enlist a father who deferred baptism to guard its grace as a witness that the father denied infants the font.
• The first explicit witness argues against the practice — which means it already existed. Tertullian is quoted here for delay; but his delay-counsel is aimed at an infant baptism already common enough to need answering (the point of the companion dating argument). His protest is evidence for the practice’s existence by c. 200, so the same passage that heads the delay case also undercuts its strong conclusion.
• The tradition itself resisted and then reversed the delay. Cyprian’s council of sixty-six bishops (c. 253) rejected even an eight-day postponement as a wrong done to the infant; Basil’s and Nazianzus’s own sermons hammer adult procrastinators to stop deferring; and the deathbed-baptism vogue crested in the fourth century and receded. A practice the church argued down and then abandoned is a passing prudential fashion, not a witness to the original shape of the rite — the same “universal custom can still be a development” blade that cuts the paedobaptist here cuts the credobaptist too.¹⁰

The credobaptist rejoinder keeps the question open. It replies that “emergency” infant baptism is not the same thing as infant baptism as the norm; that a church whose saints were overwhelmingly baptized as adults looks far more like a believers’-baptism church that made exceptions for the dying than like a paedobaptist one that made exceptions for the healthy; and that Jeremias’s need to posit an earlier general practice behind the fourth-century delay is exactly the kind of silent-reversal the paedobaptist charges the credobaptist with inventing. On this rejoinder the delay evidence does real work: it shifts the burden back onto the claim that infant baptism was ever the universal default.

Assessment

Contested. The delay of baptism is a genuine and frequently-underweighted piece of the historical record, and it decisively refutes the loosest paedobaptist claim — that the early church always and everywhere baptized its newborns. On the modest reading (infant baptism was not the universal, unquestioned default; devout families widely delayed) the argument is sound, and it rightly puts Jeremias on the defensive, forcing him to treat the fourth-century deferral as a decline from an earlier norm. But the argument the page is titled for is the strong one — that the delay shows infant baptism was not practised or not apostolic — and that it cannot carry. The delay is everywhere an elective preference within a church that has, values, and on need immediately administers infant baptism; its motive is a fear of squandering baptismal grace that presupposes the grace is offered to the child; the very first witness to delay (Tertullian) is arguing against an infant baptism that already exists; and the tradition resisted the delay (Cyprian rejecting even an eight-day wait) and then let it lapse. So the datum wounds the paedobaptist universality case without establishing the credobaptist non-apostolicity case — it lands in exactly the Jeremias–Aland standoff that leaves its companion antiquity-and-universality argument contested, and for the same reason: the historical record is genuinely double-edged, strong evidence against the extreme forms of both theses and decisive proof of neither. A caution attaches to the way each side deploys it. The paedobaptist who claims uniform apostolic infant baptism overreads the record and should concede the delay; the credobaptist who reads the delay as proof that infants were not baptized overreads it the other way, and should concede that delay is a witness to an available infant baptism, not its absence. The narrower and firmer historical questions — when the practice is first attested, and whether its antiquity reaches the apostles — are weighed on the companion dating and antiquity-and-universality pages, with which this forms the historical-case set.

Sources