A Fallible Collection of Infallible Books? The Canon Problem for Sola Scriptura

R. C. Sproul, for half a century one of the most widely read Reformed teachers in America, once summarized the Protestant doctrine of the biblical canon in a single phrase. The canon of Scripture, he wrote, is “a fallible collection of infallible books.”⁠¹ He did not mean it as a concession. He meant it as a confession of faith, and he made it with characteristic candor: “I think it’s possible that wrong books could have been selected, but I don’t believe for a minute that that’s the case.”⁠²

I think it is the most honest sentence in the Protestant theology of the canon, and I think it gives the argument away.

When I first wrote about sola scriptura in 2020, the canon was the argument on which I leaned hardest: to affirm Scripture as the only infallible authority, one must first accept some authority outside Scripture to say which books Scripture contains. I still think that is right. But I made the argument then in a few paragraphs, and I did not take on the strongest Protestant replies to it. This post tries to make it properly, against those replies, ancient and modern, including the very good case Gavin Ortlund has recently made for Sproul’s formula. My thesis is narrow. The question is not whether a fallible Church can recognize inspired books; of course it can. The question is whether a collection held fallibly can be held the way Christians, Catholic and Protestant alike, ordinarily hold the New Testament: closed, binding, and beyond appeal. Sproul’s phrase says it cannot. Everything else follows from that.

What the Argument Is Not

An honest argument about the canon has to begin by clearing away a caricature, and Catholics are responsible for one of the most common.

The caricature runs like this: the Catholic Church made the Bible. Some council in the fourth century voted, and the books that won became Scripture. Catholic popularizers sometimes put the argument almost this way, and Protestants are right to object to it. It is also not what the Catholic Church teaches. The First Vatican Council said so in as many words:

These books the church holds to be sacred and canonical not because she subsequently approved them by her authority after they had been composed by unaided human skill, nor simply because they contain revelation without error, but because, being written under the inspiration of the holy Spirit, they have God as their author, and were as such committed to the church.⁠³

The Second Vatican Council repeated the formula: the books are canonical because, “written under the inspiration of the Holy Spirit, they have God as their author and have been handed on as such to the Church herself.”⁠⁴ The Church does not confer canonicity. She receives it.

Set that beside the Westminster Confession, and the agreement is striking. “The authority of the Holy Scripture, for which it ought to be believed, and obeyed,” the Westminster divines wrote in 1646, “dependeth not upon the testimony of any man, or church; but wholly upon God (who is truth itself) the author thereof.”⁠⁵ On the question of what makes a book Scripture, Rome and Westminster say the same thing. God is the author. A council is not. Westminster wrote that sentence against Rome, and it concerns why Scripture is to be believed, not only what makes it Scripture. But Vatican I also teaches that we believe what God has revealed “because of the authority of God himself, who makes the revelation,” so on the ground of Scripture’s authority the two still agree.⁠⁶

Michael Kruger, perhaps the most careful evangelical scholar of the canon now writing, has put the Protestant view memorably: the Church is a thermometer, not a thermostat. It responds to what God has already done in the books; it does not set the temperature. Kruger draws the contrast against Rome—“the Roman Catholic view of the church is that it creates the canon. It makes the canon what it is.”⁠⁷ But Vatican I, as we have just seen, denies exactly that. A Catholic can accept the thermometer. The disagreement lies elsewhere.

It lies, I think, in two places. The first is how the canon is known. The Second Vatican Council answers that question in a single clause: “Through the same tradition the Church’s full canon of the sacred books is known.”⁠⁸ The Latin verb is innotescit—it “becomes known.” It is a verb of recognition, not of creation, and the Catechism follows it: “It was by the apostolic Tradition that the Church discerned which writings are to be included in the list of the sacred books.”⁠⁹ The Reformed answer, as we shall see, is the inward testimony of the Holy Spirit.

The second is what kind of assent the list receives. Here the question is not psychological certainty. John Henry Newman warned long ago against confusing certitude with infallibility. “It is persons and rules that are infallible,” he wrote, and a man may be certain of a great many things without being infallible about any of them.⁠¹⁰ A Protestant can be certain, in Newman’s sense, that the Gospel of Mark belongs in the Bible. The question is whether the list as a whole can be held as a rule of faith—a boundary that binds the conscience of every Christian and that no Christian may revise—when the authority that drew the boundary is, by the Protestant’s own account, capable of having drawn it wrongly.

That is the question Sproul’s phrase answers. To see why his answer is a problem, it helps to begin with the answer that came before it.

The Reformed Answer: The Spirit and the Self-Evidencing Word

John Calvin addressed the canon question head-on in the seventh chapter of the first book of the Institutes, and he did not avoid its hardest form. He framed the objection himself:

With great insult to the Holy Spirit, it is asked, who can assure us that the Scriptures proceeded from God; who guarantee that they have come down safe and unimpaired to our times; who persuade us that this book is to be received with reverence, and that one expunged from the list, did not the Church regulate all these things with certainty?⁠¹¹

Notice the scope of the question. Calvin does not limit himself to whether Scripture as a whole is from God. He includes “this book” and “that one expunged from the list”—the boundary itself. His answer comes in two parts.

The first is that Scripture is evident in itself. “Scripture bears upon the face of it as clear evidence of its truth,” Calvin wrote, “as white and black do of their colour, sweet and bitter of their taste.”⁠¹² The second is that this evidence becomes conviction only by the work of the Spirit: “those who are inwardly taught by the Holy Spirit acquiesce implicitly in Scripture; that Scripture, carrying its own evidence along with it, deigns not to submit to proofs and arguments, but owes the full conviction with which we ought to receive it to the testimony of the Spirit.”⁠¹³

Calvin was not dismissive of the Church’s role. He granted that “the consent of the Church is not without its weight,” and that human testimonies may serve “as secondary helps to our weakness.”⁠¹⁴ He also offered a careful reading of Augustine’s famous line—“I should not believe the gospel except as moved by the authority of the Catholic Church”—arguing that Augustine was describing how an unbeliever is first led to the Gospel, not the ground of a believer’s certainty.⁠¹⁵ (I have discussed that dispute over Augustine elsewhere, in a longer reply on the Reformation’s historical premise.) But the foundation, for Calvin, was the Spirit.

The Reformed confessions followed him. The Gallican Confession of 1559, prepared, according to Schaff, by Calvin and his student Antoine de Chandieu and revised by a synod at Paris, says: “We know these books to be canonical, and the sure rule of our faith, not so much by the common accord and consent of the Church, as by the testimony and inward illumination of the Holy Spirit, which enables us to distinguish them from other ecclesiastical books.”⁠¹⁶ The Belgic Confession of 1561 says nearly the same: the Reformed churches receive “all these books, and these only, as holy and canonical,” “believing, without any doubt, all things contained in them,” and do so “not so much because the Church receives and approves them as such, but more especially because the Holy Ghost witnesseth in our hearts that they are from God.”⁠¹⁷

Westminster gave the doctrine its most balanced form:

We may be moved and induced by the testimony of the church to an high and reverent esteem of the Holy Scripture. And the heavenliness of the matter, the efficacy of the doctrine, the majesty of the style, the consent of all the parts, the scope of the whole … are arguments whereby it doth abundantly evidence itself to be the Word of God: yet notwithstanding, our full persuasion and assurance of the infallible truth and divine authority thereof, is from the inward work of the Holy Spirit bearing witness by and with the Word in our hearts.⁠¹⁸

This is a serious position, and it is fairer to engage it than to mock it. It does not deny the Church a role. It denies the Church the final role. The Church introduces, commends, and testifies; the Spirit persuades.

The Modern Answer: A Fallible Collection

Sproul presented his formula as “the historic Protestant position,” and in one sense it is: Westminster itself teaches that “all synods or councils, since the Apostles’ times, whether general or particular, may err,” and the councils that listed the canon were synods.⁠¹⁹ But Sproul says openly what the confessions did not. He leaves the inward testimony of the Spirit to Scripture in place; his claim concerns the list. Where Calvin spoke of a persuasion “superior to human Judgment,” and the Gallican Confession of a Spirit who “enables us to distinguish” the canonical books “from other ecclesiastical books,” Sproul spoke of the list itself as a sound but fallible historical judgment.⁠²⁰ He credited the distinction to his teacher John Gerstner: Rome holds the Bible to be “an infallible collection of infallible books,” Protestants “a fallible collection of infallible books.”⁠²¹ The Church, on this view, did its work rightly and well, but not infallibly.

C. Michael Patton of Credo House has defended the formula with real philosophical care. “Do we have a fallible collection of infallible books?” he asks. “Yes, I believe we do. When all is said and done, all of our beliefs are fallible and therefore subject to error.” His key distinction is between possibility and probability: “the possibility of error does not necessitate the probability of error.” And he turns the argument back on Catholics: “Catholics have a fallible belief about an infallible authority; Protestants have a fallible belief about an infallible authority.”⁠²²

Michael Kruger’s Canon Revisited offers the most sophisticated modern account of how the canon is known. Kruger argues that God has created the conditions under which Christians can reliably recognize the canon: the books themselves bear what he calls three “attributes of canonicity”—divine qualities, apostolic origins, and corporate reception by the Church—and the Holy Spirit enables believers to recognize them.⁠²³ Kruger presents the model as a recovery of Calvin’s self-authenticating Scripture, not a retreat from it, and he treats the Church’s reception as the corporate fruit of the Spirit’s work: the books were received because they were canonical, not canonical because they were received. Kruger is no individualist.

Finally, and most recently, Gavin Ortlund has defended Sproul’s formula, both in an episode of his program Truth Unites titled “A Fallible List of Infallible Books?” and in a written essay drawn from his book What It Means to Be Protestant.⁠²⁴ Ortlund’s case is the best I know, and it is generous: he begins from Vatican I’s denial that the Church makes books canonical, and he grants the Church a necessary role as witness. He also keeps Calvin’s terms. Scripture, for him, is “self-authenticating,” and the Church discerned it “through the enabling direction of the Holy Spirit.” His argument has three planks. The first is logical: “it is not necessary to be infallible to discern that which is infallible.” Ortlund borrows William Whitaker’s old image of the goldsmith, who recognizes gold without making it. The second is historical, and it is the one on which he rests the most weight: the canonization of the New Testament “was bottom up, not top down,” the work of local churches and local councils, not of an infallible pope or ecumenical council. The third is biblical: Israel had no infallible teaching office, yet Jesus held the Jews of his day accountable to their Scriptures. Ortlund also presses a regress: “If you think you do have to possess infallibility to discern infallibility, you have a continual regress, because now you need infallibility to receive and interpret the infallible teachings of your church.” Rome, he adds, has its own “fallible list of infallible teachings,” since the number of ex cathedra statements has never been infallibly defined.

These are good arguments. I will try to answer them on their own terms. But it is worth noticing first what the modern answer has added to the Reformed one. Kruger and Ortlund keep the Spirit and the self-authenticating Word. What is new, in Sproul, Patton, and Ortlund, is the candor about the list.

Calvin asked, “what is to become of miserable consciences in quest of some solid assurance of eternal life, if all the promises with regard to it have no better support than man’s Judgment?”⁠²⁵ He meant it as an objection to Rome: if the Church’s judgment settles the canon, faith rests on men. Sproul’s formula, taken at its word, gives the same answer about the list that Calvin feared about the books. The books are God’s; the collection is the Church’s, and the Church may have erred. There is an answer, and it is a good one: the promises rest on the books the Spirit attests, not on the list, and no one doubts the books in which the promises of eternal life are chiefly found. But that answer gives up the boundary. It secures the center by the Spirit and leaves the edge to human judgment, and the question Calvin set himself to answer included the edge: “this book” and “that one expunged from the list.” The formula says aloud what Calvin’s answer left unsaid. (I return to this reply below.)

Where the Testimony of the Spirit Ran Out

The reason it had to be said aloud is not hard to find. The Reformers’ appeal to the Spirit, whatever its power to assure a believer that Scripture is God’s word, did not produce agreement about where Scripture ends.

Martin Luther’s German New Testament of 1522 set Hebrews, James, Jude, and Revelation apart at the end of the volume. In the preface to Hebrews, as Philip Schaff renders it, Luther wrote: “Hitherto we have had the right and genuine books of the New Testament. The four that follow have been differently esteemed in olden times.”⁠²⁶ His general preface ranked James as “an epistle of straw” compared with the books he considered the true heart of the New Testament; Schaff notes that the passage “was suppressed in later editions.”⁠²⁷ (The post on the Epistle of James treats the “straw” preface in detail.)

Luther was not alone. Ulrich Zwingli “objected only to the Apocalypse and made no doctrinal use of it, because he did not deem it an inspired book”; Johannes Oecolampadius “speaks of six Antilegomena … as holding an inferior rank, but nevertheless appeals to their testimony.”⁠²⁸ Calvin himself, more conservative on this question than Luther, received 2 Peter because “in every part of the Epistle the majesty of the Spirit of Christ appears,” while confessing that he did not recognize Peter’s language in it: “to repudiate it is what I dread, though I do not here recognize the language of Peter.”⁠²⁹ He received James, which he knew “was not formerly received by many Churches without opposition,” because “I see no just cause for rejecting it” and because “it contains nothing unworthy of an Apostle of Christ.” He received Jude because “the reading of it is useful,” it “contains nothing inconsistent with the purity of apostolic doctrine,” and it “was received as authentic formerly, by some of the best.”⁠³⁰

Look at the reasoning. For 2 Peter, Calvin appealed to the book’s majesty, and the majesty gave him the book but not its author. For James, he set aside the doubts of the ancient churches on the strength of his own judgment of its doctrine. For Jude, he joined a doctrinal test to the judgment of the ancient churches—“received as authentic formerly, by some of the best.” These are sound criteria, close to what Kruger would later call divine qualities, apostolic origins, and corporate reception. But they are a scholar’s weighing of evidence, book by book, and Luther weighed the same evidence on James and Jude differently. The self-evidence of white and black settled the question for neither man.

The Catholic controversialists of the period saw the difficulty at once, and the ablest Protestant answer to them shows exactly where it lies. William Whitaker, Regius Professor of Divinity at Cambridge, wrote the great Elizabethan reply to Rome on Scripture in 1588. He called the question of Scripture’s authority “difficult and perplexed; nor do I know whether there is any other controversy between us of greater importance.”⁠³¹ He then reported the Catholic objection fairly: if the Spirit’s testimony identifies the books, “how comes it that churches, which have this Spirit, agree not amongst themselves? For … the Lutherans disagree with you Calvinists, because you receive some books which they reject.” Whitaker’s answer was that “there is not precisely the same equal measure of the Holy Spirit in all,” and that both sides had “the precedent of some ancient churches.”⁠³²

Both parts of that answer are revealing. The first concedes that the Spirit’s testimony does not deliver the same boundary to everyone who has the Spirit. The second retreats, again, to the precedent of the ancient churches. And a little later Whitaker drew the distinction he thought answered the objection:

We do not therefore endeavour to refute others by the secret testimony of the Spirit, since it is peculiar to the individual, private and internal; but by common arguments taken from the books themselves, and from the judgment of the church.⁠³³

Whitaker did not regard this as a concession. The Spirit’s testimony, he wrote on the page before, is “the most certain argument of all,” and it alone persuades the believer himself.⁠³⁴ But the sentence is, I think, the hinge of the whole Reformation debate on the canon. The testimony of the Spirit may assure the individual believer, but it is “private and internal.” It cannot be shown to anyone else. When the question is which books—a public question, whose answer has to be the same for every Christian—the Reformed theologian, like everyone else, argues from the books and from “the judgment of the church.”

Kruger’s model has a reply. On his account the Spirit’s testimony is not merely private: its fruit is the Church’s corporate reception, which anyone can inspect. I think this is the best Protestant answer, and it moves the argument onto the ground this post has been describing, where the public evidence for the boundary is the Church’s reception. The question then is whether the Spirit protects that reception from error. Sproul reports that “some Protestants” believe “a special work of divine providence and a special work of the Holy Spirit … protected the Canon and the sorting process from mistakes,” and says, “I don’t hold that position myself.” A Protestant who answers yes has left the fallible collection behind and owes an account of where that protection is found. A Protestant who answers no, with Sproul, is back at Whitaker’s distinction.

The Reformed confessions reflect the same fact in their structure. Westminster grounds the authority of Scripture in God and the Spirit (1.4–1.5), but it does not derive its list of sixty-six books from either. It simply states the list (1.2) and excludes the rest (1.3).⁠³⁵ No verse of Scripture supplies that table of contents. Keith Mathison, a Reformed theologian, has made the point with admirable frankness against what he calls “solo” Scriptura: the table of contents is not itself an inspired text; it is, in his phrase, “a creed of the church.”⁠³⁶

Mathison is right, and he has an answer: such creeds carry “the true but subordinate authority of the church,” real but not infallible. But that answer brings us back to Sproul. A creed of a fallible church is revisable, and this creed defines the very Scripture by which every other creed is to be judged.

What the Early Church Actually Did

If the inward testimony could not draw the boundary, what did? The answer of the early Church is consistent across more than two centuries and three continents, and it is not the answer either the Catholic caricature or the Protestant one expects. The Fathers did not vote books into Scripture, and they did not rest the question on any one reader’s sense of a book’s majesty. They weighed apostolic origin and orthodoxy, and above all they asked what the churches had received.

Irenaeus of Lyons, writing around 180, imagined the canon question before there was a settled canon to ask it about. “For how should it be if the apostles themselves had not left us writings?” he asked. “Would it not be necessary, [in that case,] to follow the course of the tradition which they handed down to those to whom they did commit the Churches?”⁠³⁷ For Irenaeus, the apostolic writings and the apostolic churches were two forms of the same deposit, and the second could stand in for the first. It was the same Irenaeus who insisted that “It is not possible that the Gospels can be either more or fewer in number than they are.”⁠³⁸ The four were fixed for him before anyone had defined them.

The Muratorian Fragment, on its traditional dating the oldest surviving list of New Testament books, speaks the same language. Of certain letters forged in Paul’s name, it says that they “cannot be received into the Catholic Church, for it is not suitable for gall to be mingled with honey.” Of the Shepherd of Hermas, written, it says, “very recently in our times,” it rules that “it also ought to be read; but it cannot be made public in the Church to the people, nor placed among the prophets, as their number is complete, nor among the apostles to the end of time.”⁠³⁹ (The fragment’s date is disputed: most specialists still place it in the late second or early third century, but a serious minority argues for the fourth. The point here does not depend on the date.) The criteria are not majesty. They are apostolic origin, orthodoxy, and the Church’s public reading.

Eusebius of Caesarea, early in the fourth century, sorted the books into those accepted, those disputed, and those rejected, and he described the accepted books as the ones that “according to ecclesiastical tradition are true and genuine and commonly accepted.”⁠⁴⁰ James, Jude, 2 Peter, and 2–3 John sat in his middle category: “disputed … which are nevertheless recognized by many.” Revelation he placed in two categories at once, because the churches had not agreed about it.

Cyril of Jerusalem, catechizing new Christians around 350, gave the most direct instruction of all: “Learn also diligently, and from the Church, what are the books of the Old Testament, and what those of the New.” His rule for anything outside the list was equally plain: “whatever books are not read in Churches, these read not even by yourself.”⁠⁴¹ It is worth adding, in fairness, that Cyril’s own New Testament list omits Revelation.⁠⁴² The edges were still moving in his day.

Athanasius’s Thirty-Ninth Festal Letter of 367 is the earliest surviving list known to give exactly the twenty-seven books of our New Testament. Athanasius did not present it as his discovery. He set out “the books included in the Canon, and handed down, and accredited as Divine,” and he closed the list with a warning: “Let no man add to these, neither let him take ought from these.”⁠⁴³

Augustine, writing On Christian Doctrine in the 390s, gave the criterion its most precise statement:

Now, in regard to the canonical Scriptures, he must follow the judgment of the greater number of Catholic churches; and among these, of course, a high place must be given to such as have been thought worthy to be the seat of an apostle and to receive epistles. Accordingly, among the canonical Scriptures he will judge according to the following standard: to prefer those that are received by all the Catholic churches to those which some do not receive.⁠⁴⁴

The African councils turned that criterion into law. The list adopted at Hippo in 393 and re-adopted at Carthage in 397 is framed as a rule about what may be read in church, and it asks that “the overseas church” be consulted to confirm it.⁠⁴⁵ When the African bishops re-enacted it in their code of 419, the reason they gave was explicit: “for these are the things which we have received from our fathers to be read in church.”⁠⁴⁶

I want to be careful here, because Catholic apologists have sometimes overstated this history. The councils at Hippo and Carthage were regional, not ecumenical. The often-repeated claim that a Roman synod under Pope Damasus issued the canon in 382 rests on the Decretum Gelasianum, a sixth-century work falsely headed as a Damasine council.⁠⁴⁷ None of these councils claimed to make a book Scripture. In all of this, Ortlund’s description of the process as “bottom up” is largely correct, though the African bishops twice asked the church across the sea to confirm their list.

But notice what the bottom-up process was. It was the Church’s reception—what the churches had “received from our fathers,” what was “read in church,” what “all the Catholic churches” accepted—discerned and stated by her bishops. It is what Dei Verbum means by the canon becoming known “through the same tradition.” Ortlund would not call it private judgment either: the Church, he writes, discerned the books “through the enabling direction of the Holy Spirit.” The question is what that direction guaranteed. Kruger includes “corporate reception” among the marks of canonicity; Calvin weighed what had been “received as authentic formerly, by some of the best.” Even the Protestant historian Philip Schaff conceded that the Reformed principle, “in its exclusive form,” is “diametrically opposed” to Augustine’s maxim, and added: “the two kinds of evidence supplement each other. The human authority of tradition though not the final ground of belief, is indispensable as an historical witness of the genuineness and canonicity.”⁠⁴⁸

Luther himself, before the 1522 prefaces, had granted the point in its Augustinian form. In The Babylonian Captivity of the Church (1520) he wrote: “This power the Church certainly has—that she can distinguish the word of God from the words of men. So Augustine confesses that his motive for believing the gospel was the authority of the Church, which declared it to be the gospel.” He added at once, “Not that the Church is therefore above the gospel.”⁠⁴⁹ A Catholic can affirm both sentences.

So the Protestant, too, leans on the Church’s reception. Every serious account of the canon does. The only question is what that reception is: a human report, accurate on the whole and revisable in principle, or a witness the Holy Spirit protects.

Where “Fallible” Bites

The fallible-collection formula answers that question one way. It is worth pressing what the answer costs.

Fallible means revisable

To call a judgment fallible is to say it could be wrong. Sproul said exactly this: it is “possible that wrong books could have been selected.” Patton rightly adds that a possible error is not a probable one. Fair enough. But a judgment that could be wrong is a judgment that may, in principle, be revised—and the history of the Reformation shows that the possibility is not merely theoretical.

Luther revised it. In 1522 he did not simply weigh the antilegomena; he demoted four books the Western Church had read as Scripture for more than a thousand years. The Reformed churches revised it too, in their own way. They kept Luther’s four at full standing but, like Luther in his complete German Bible of 1534, set outside the canon the Old Testament books the Western councils had received, the same councils whose New Testament list they kept. (I have traced that history in a long post on the deuterocanonical books, and I will not repeat it here.) Whatever one thinks of the merits, these were exercises of exactly the authority the fallible-collection formula concedes to every generation: the authority to reconsider.

The best Protestant reply is that the possibility of revision touches only the edges. No one will reopen Mark or Romans. Whitaker held that disagreement over the disputed letters is not a “capital” error, and Ortlund calls the disputed books the “fuzzy edges.”⁠⁵⁰ I grant the core. But the edges are where a boundary is, and the confessions bind them as firmly as the center: the Belgic Confession lists Jude beside Romans among “these books, and these only.”

On what ground, then, does a Protestant forbid the next Luther? Not on the ground that the collection is beyond error—Sproul has given that up. Not on the ground that the Spirit testifies to the boundary—Whitaker granted that the Spirit’s testimony is private, and unequally given. Only on the ground that the historical case is strong. That is a good ground for a scholarly opinion. It is not a ground for a confession of faith.

A fallible list cannot bear the weight Protestants put on it

And Protestants do treat it as a confession of faith. The Belgic Confession receives “all these books, and these only,” and believes “without any doubt, all things contained in them.” Westminster lists the sixty-six and declares the rest “no part of the canon of the Scripture,” “of no authority in the church of God.” In confessional Presbyterian and Reformed churches, ministers and elders take vows that bind them to these confessions; in Presbyterian churches, to the Westminster Confession as containing the system of doctrine taught in Scripture. Ordinary Protestant practice treats the Bible’s table of contents as settled—not as the current state of the scholarly question, open to revision by the next good argument from a manuscript discovery or a new commentary on Jude, but as the boundary of the Word of God.

That practice is correct. Christians should treat the canon as closed. But it cannot be justified by the formula. A list that one receives as “these only,” and binds on others, is a list one treats as irrevisable in practice while calling it revisable in principle. Certainty is not the difficulty; Newman was right that a fallible man can be certain. The difficulty is binding. Sproul himself holds that “Scripture alone has the authority to bind our conscience.” Westminster gives synods authority “ministerially to determine controversies of faith,” and their decrees, “if consonant to the Word of God, are to be received with reverence and submission”; but because synods “may err,” they “are not to be made the rule of faith, or practice; but to be used as a help in both.”⁠⁵¹ Sproul’s honesty consists precisely in refusing to hide what follows. He would rather admit that the list could be wrong than claim for the Church an authority he does not believe she has. I respect the choice. But it leaves the canon in an odd place. Every other article of a Reformed confession is fallible but answerable to Scripture. The table of contents is fallible and answerable to no text above it, because it defines what Scripture is, and yet it is the one judgment of a synod that functions as the rule of faith.

Newman saw this as an Anglican

In 1838, seven years before he was received into the Catholic Church, John Henry Newman published a set of lectures as Tract 85 of the Tracts for the Times, arguing that the objections Protestants raised against the Church’s doctrines, such as the apostolic succession and the priesthood, for want of clear proof in Scripture or the Fathers, applied with equal force to the canon. He put the objector’s question in its starkest form: “Why should not we be as good judges as the Church of the fourth century, on whose authority we receive it?”⁠⁵² And he drew the conclusion:

It seems, then, that the objections which can be made to the evidence for the Church doctrines are such as also lie against the Canon of Scripture; so that if they avail against the one, they avail against both.⁠⁵³

Newman even predicted the coming fight: “a battle for the Canon of Scripture is but the next step after a battle for the Creed.”⁠⁵⁴ His point was not that the canon and the creed rest on identical evidence. It was a parity argument. The same Church of the fourth and fifth centuries received the twenty-seven books and confessed the Nicene faith (the Greek churches still hesitated over Revelation, as Newman himself notes in the Essay on Development), and a man who treats her judgment as authoritative for one and negotiable for the other is choosing, not receiving. “If men then will indulge that eclectic spirit which chooses part and rejects part of the primitive Church system,” Newman wrote, “I do not see what is to keep them from choosing part and rejecting part of the Canon of Scripture.”⁠⁵⁵

When Newman revised his Essay on the Development of Christian Doctrine as a Catholic in 1878, he made the same argument and footnoted his own Anglican tract. “On what ground, then, do we receive the Canon as it comes to us,” he asked, “but on the authority of the Church of the fourth and fifth centuries?”⁠⁵⁶ The argument had not changed. Newman had.

Answering the Best Objections

Ortlund’s objections deserve direct answers. I take them in turn.

“A necessary witness need not be an infallible witness”

This is true, and in one sense it settles nothing. A fallible witness can tell me something true. A fallible jeweler can correctly identify gold. William Whitaker made the classic form of the argument in 1588: the goldsmith “does not make gold, even in respect of us, but only indicates what is gold, so that we the more easily trust it.”⁠⁵⁷

But consider how the analogy works. The goldsmith’s verdict helps me because I could, in principle, verify it myself—with the same touchstone, the same scales, the same knowledge of metals. Whitaker’s own illustration makes the point: “it is on account of the matter and the form impressed upon the coin that I perceive it to be sterling and royal money.” The witness is useful because the hearer can, in the end, see for himself.

With the canon, the hearer cannot do so alone. The Spirit’s testimony, Whitaker granted, is “private and internal”; it cannot be produced as evidence. The public evidence is the historical reception of the churches. And at that point there are only two possibilities. Either the reader can assay the books himself, in which case Luther’s assay of James, which found straw, and Calvin’s, which found gold, are rival private verdicts with no one authorized to choose between them. Or he cannot, in which case the reliability of the Church’s witness is not one consideration among several but the whole matter. The jeweler analogy works for gold, where trained assayers get the same result. It does not work for a boundary on which the best assayers have disagreed since the second century.

“You cannot escape the regress”

Ortlund presses a regress: Catholics also use fallible judgment to identify the Church, to identify which of her statements are infallible, and to interpret them. There is no infallible list of infallible papal statements. So the Catholic, too, has a fallible list.

The first half of the objection is true and Catholics should simply grant it. The act of faith is preceded by judgments of credibility that are not themselves infallible. Vatican I calls the Church herself “a kind of great and perpetual motive of credibility,” recognizable by public marks—her propagation, her holiness, her unity, her stability.⁠⁵⁸ Those marks are weighed by fallible minds. No one is asked to be infallible before he believes.

But the regress does not run the same length on both sides. The Catholic uses fallible judgment to recognize a living authority, one that can then teach, correct, and settle what he could not settle alone—an authority that, in the Second Vatican Council’s words, “is not above the word of God, but serves it.”⁠⁵⁹ The Protestant uses fallible judgment for the list and for its meaning, and every link remains, by his own account, revisable. The Reformed answer is that the chain ends in “the Holy Spirit speaking in the Scripture.”⁠⁶⁰ But that is the testimony Whitaker called “private and internal,” and it does not tell anyone else which books Scripture contains. A layman can make a fallible judgment that a court has jurisdiction over his case; that does not make the court’s judgment as fallible as his own guess about the law.

More importantly, the objection misidentifies where Catholics locate the canon’s certainty. The Catholic claim is not that the canon became certain in 1546, or that it waits on an ex cathedra list. Vatican I teaches that what is “contained in the word of God as found in scripture and tradition” is to be believed with divine and Catholic faith when the Church proposes it “whether by her solemn judgment or in her ordinary and universal magisterium.”⁠⁶¹ Vatican II teaches that “the entire body of the faithful, anointed as they are by the Holy One, cannot err in matters of belief” when, “from the Bishops down to the last of the lay faithful,” they “show universal agreement in matters of faith and morals,” a discernment “exercised under the guidance of the sacred teaching authority.”⁠⁶² On Catholic premises, then, the bottom-up reception Ortlund describes so well need not have lacked an infallible witness. Catholic theologians commonly hold that the core of the canon was held infallibly by the whole Church, bishops included, long before any council defined it.⁠⁶³ Where reception was not yet universal—Revelation in the Greek East, Hebrews in the early Latin West, the deuterocanonical books for Jerome—the bishops judged, Florence enumerated, and Trent defined.⁠⁶⁴ Joe Heschmeyer of Catholic Answers made this point in a 2025 response to Ortlund: the question is whether that reception was protected, not whether a pope signed it.⁠⁶⁵

“Jesus held Israel to its Scriptures without an infallible magisterium”

This is the most interesting of the three objections, and the answer has three parts.

First, Israel’s canon in Jesus’ day had a recognized core and disputed edges, as Ortlund himself grants: he notes disagreements over Ecclesiastes and Esther, and the Samaritans’ Torah-only canon. Jesus’ own summary in Luke 24:44 names “the law of Moses and the prophets and the psalms,” without naming a closed third collection.⁠⁶⁶ (The deuterocanonical post treats the evidence in detail.)

Second, Israel was never left with a book and no voice. It had prophets, priests, and judges, and in Jesus’ own day it had a teaching authority he explicitly told his disciples to obey: “The scribes and the Pharisees sit on Moses’ seat; so practice and observe whatever they tell you, but not what they do.”⁠⁶⁷ I do not claim that the Pharisees were infallible. I claim that God’s people were not governed by a text without an authorized interpreter. That does not refute Ortlund’s claim, and I should say so. His claim is only that God’s people could recognize Scripture and be held to it without an infallible decree, and I granted that at the outset. What his example does not show is a closed and binding boundary held fallibly. Israel had a recognized core, disputed edges, and living voices to settle what needed settling, and at last the Lord himself. The Church, unlike the synagogue, was promised the Spirit who would “guide you into all the truth.”⁠⁶⁸

Third, the analogy proves too little. The Christian Old Testament was not received from the synagogue’s consensus but from the Church’s. If Jewish reception were the rule, the Reformed would owe us an account of why first-century Jewish reception governs the Old Testament while fourth-century Christian reception governs the New. Whatever the answer, it is not given by Scripture.

“Rome had a fallible canon until Florence and Trent”

Ortlund presses the point that Rome itself had a “fallible canon” for most of her history, since, on Catholic terms, the canon “was not infallibly defined until the Councils of Florence and Trent toward the end of the Middle Ages.” Others add that respected Catholic scholars questioned parts of the received canon right up to the eve of Trent. The facts are true. Cardinal Cajetan, Luther’s opponent at Augsburg, followed Jerome in setting the deuterocanonical books outside the canon in the strict sense and doubted the authorship of several New Testament letters; in the preface to his commentary on Hebrews he wrote that “we have received Jerome as our rule that we may not err in the separation of the Canonical books.”⁠⁶⁹ Erasmus raised doubts of his own. A Catholic should concede all of this plainly. What does not follow is that an undefined canon was a fallible one. On Catholic premises, as I argued above, the core of the canon was held without error long before it was defined. The edges were what remained open until Trent.

But a definition is not the beginning of a belief. Florence enumerated the same books in 1442, in its bull of union with the Copts: the Church “accepts and venerates their books, whose titles are as follows.”⁠⁷⁰ That bull pronounced no anathema on anyone who rejected a book on its list. (Its anathema, a sentence later, falls on the Manichees.) Whether it was itself a dogmatic definition is debated, and Cajetan’s later dissent shows that it did not end the discussion. Trent’s decree settled it for Catholics, and Trent’s anathema rests on prior reception, not on a fresh decision: it condemns anyone who does not receive the books “as they have been used to be read in the Catholic Church.”⁠⁷¹ At Trent, and arguably already at Florence, the Church defined what she had read for more than a thousand years.

This is how doctrine has ordinarily been defined. Christians worshipped Christ as God before Nicaea defined the homoousios in 325. The definition did not create the faith; it closed a dispute. Newman, in the Essay on Development, put the canon of the New Testament and the homoousion side by side as instances of exactly this pattern.⁠⁷² The difference between Catholics and Protestants is not that Catholics had a defined canon early and Protestants late. It is that Catholics have an authority competent to close the dispute, and the fallible-collection formula, by its own account, does not.

Why Sproul Was Right

I began by calling Sproul’s formula the most honest sentence in the Protestant theology of the canon, and I meant it as praise. It refuses two evasions. It does not pretend that the Bible’s table of contents is taught by the Bible. And it does not pretend that the inward testimony of the Spirit, however real, has delivered the same table of contents to every Christian who has the Spirit. Having refused both, it says what is left: the Church’s judgment, which a Protestant must regard as human and therefore fallible.

The Catholic does not disagree with the first two refusals. He disagrees only with the last step. He holds that the Church’s judgment about the canon was not merely human—that the same Spirit who inspired the books guided the community that received them, so that what the churches “received from our fathers to be read in church” was not a guess, however educated, but a witness. That is a large claim, and it requires an argument about the Church that this post has not tried to make. But it is the claim that best matches what Catholics and Protestants alike actually do with the Bible. We treat the list as closed. We bind it on the conscience. We do not wait for the next good argument about Jude.

Every Bible’s table of contents is a creed. The question is only whether we believe the Church that first confessed it.

Frequently Asked Questions

What does “a fallible collection of infallible books” mean?

It is R. C. Sproul’s summary of what he called “the historic Protestant position,” a formula he credited to John Gerstner. Each book of the Bible is inspired and therefore infallible, but the Church’s judgment about which books belong in the Bible was a human judgment and could in principle have been wrong. Sproul did not believe it was wrong; he believed the Church did its work well.

Does the Catholic Church claim it created the Bible?

No. The First Vatican Council taught that the Church holds the books to be sacred and canonical “not because she subsequently approved them by her authority,” but because they were written under the inspiration of the Holy Spirit and have God as their author. The Catholic claim concerns how the canon is known—through the Church’s tradition—not what makes a book Scripture.

Didn’t the early Church get along without an infallible canon list?

The early Church received the core of the New Testament long before any council listed it, and the councils that did list it, such as Hippo (393) and Carthage (397), were regional. Catholics agree. The Catholic claim is that the Church’s reception of those books, settled by her bishops where the churches had disagreed, was guided and protected by the Holy Spirit. That is why the African bishops, re-enacting the list in 419, appealed to what they had “received from our fathers to be read in church” rather than to their own authority.

Isn’t the Catholic also relying on fallible judgment to accept the Church?

Yes, at the start. The judgments that lead a person to faith are not themselves infallible. The difference is where those judgments lead: the Catholic’s lead to a living authority that can settle questions the individual cannot, while the fallible-collection view leaves the boundary of Scripture resting on a historical judgment that remains, by its own account, open to revision.

Did the Reformers agree among themselves about the New Testament canon?

Not entirely. Luther set Hebrews, James, Jude, and Revelation apart in his 1522 New Testament; Zwingli made no doctrinal use of Revelation; Oecolampadius ranked six disputed books lower than the rest. Calvin received the disputed letters on which he commented, though he doubted that Peter personally wrote 2 Peter. Most Reformed confessions later fixed the list by enumeration.

Further Reading

For the Reformed view at its source, read John Calvin, Institutes of the Christian Religion, book 1, chapters 7–8, and William Whitaker, A Disputation on Holy Scripture (1588; Parker Society, 1849), which remains the most thorough early Protestant answer to the Catholic argument. Michael J. Kruger’s Canon Revisited (Crossway, 2012) is the best modern Protestant account of how Christians can know the canon, and Gavin Ortlund’s What It Means to Be Protestant (Zondervan, 2024) makes the case for Sproul’s formula in a broader defense of Protestant theology. For the history of the canon itself, Bruce Metzger’s The Canon of the New Testament (Oxford, 1987) is still the standard survey. On the Catholic side, John Henry Newman’s lectures “Holy Scripture in Its Relation to the Catholic Creed” (Tract 85, 1838, reprinted in Discussions and Arguments) are the indispensable starting point, along with chapter 4 of his Essay on the Development of Christian Doctrine.

Footnotes

1. 1. R. C. Sproul, Now, That’s a Good Question! (Wheaton, IL: Tyndale House, 1996), as excerpted in “Tough Questions with R. C. Sproul,” BibleGateway, May 21, 2024. Sproul’s full sentence describes “the historic Protestant position shared by Lutherans, Methodists, Episcopalians, Presbyterians, and so on.”

2. 2. Sproul, Now, That’s a Good Question!, in the same excerpt. Sproul adds that some Protestants hold that “a special work of divine providence and a special work of the Holy Spirit … protected the Canon and the sorting process from mistakes,” and says, “I don’t hold that position myself.”

3. 3. First Vatican Council, Dei Filius (1870), ch. 2, trans. Norman P. Tanner, as reproduced at papalencyclicals.net. The Latin reads: Eos vero Ecclesia pro sacris et canonicis habet, non ideo quod sola humana industria concinnati, sua deinde auctoritate sint approbati … sed propterea quod Spiritu Sancto inspirante conscripti Deum habent auctorem, atque ut tales ipsi Ecclesiae traditi sunt (vatican.va).

4. 4. Second Vatican Council, Dei Verbum §11 (1965).

5. 5. Westminster Confession of Faith (1646), 1.4, in the text published by the Orthodox Presbyterian Church.

6. 6. Dei Filius, ch. 3 (Tanner).

7. 7. Michael J. Kruger, “Did the Church Decide Which Books Made It into the Bible?” video clip and transcript from his series The New Testament Canon, Ligonier Ministries, May 22, 2019. In the same clip Kruger says the Protestant view does not claim “that the church creates the canon or constitutes the canon.”

8. 8. Dei Verbum §8; Latin: Per eandem Traditionem integer Sacrorum Librorum canon Ecclesiae innotescit (vatican.va).

9. 9. CCC 120.

10. 10. John Henry Newman, An Essay in Aid of a Grammar of Assent, ch. 7, §2, at newmanreader.org. Newman’s fuller sentence: “It is very common, doubtless, especially in religious controversy, to confuse infallibility with certitude, and to argue that, since we have not the one, we have not the other … but the two words stand for things quite distinct from each other.”

11. 11. John Calvin, Institutes of the Christian Religion 1.7.1, trans. Henry Beveridge (Edinburgh: Calvin Translation Society, 1845), at CCEL. Quotations of the Institutes in this post are from Beveridge’s translation, which is in the public domain.

12. 12. Calvin, Institutes 1.7.2 (Beveridge). In the same section Calvin grants that the Church “gives it the stamp of her authority,” but insists that “she does not make that authentic which was otherwise doubtful or controverted.”

13. 13. Calvin, Institutes 1.7.5 (Beveridge). Beveridge’s “carrying its own evidence along with it” renders the idea that later translations express with the word “self-authenticated.”

14. 14. Calvin, Institutes 1.8.12–13 (Beveridge), at CCEL.

15. 15. Augustine, Contra epistulam Manichaei quam vocant Fundamenti 5.6 (NPNF¹ 4), at New Advent; Calvin’s reading is in Institutes 1.7.3, where he concludes that Augustine “merely says that unbelievers would have no certainty of the gospel, so as thereby to win Christ, were they not influenced by the consent of the Church.”

16. 16. Gallican (French) Confession (1559), art. 4, English trans. Emily O. Butler, in Philip Schaff, ed., The Creeds of Christendom, vol. 3 (New York: Harper & Brothers, 1877), at CCEL. Schaff describes the confession as “prepared by Calvin and his pupil, De Chandieu.”

17. 17. Belgic Confession (1561), art. 5, in Schaff, Creeds of Christendom, vol. 3 (the English text of the Reformed Church in America, made from the Latin). The French reads non pas tant parce que l’Eglise les reçoit et approuve tels.

18. 18. Westminster Confession of Faith 1.5.

19. 19. Westminster Confession of Faith 31.3 in the OPC text (31.4 in the original 1646 numbering): “All synods or councils, since the Apostles’ times, whether general or particular, may err; and many have erred.”

20. 20. Calvin, Institutes 1.7.5 (Beveridge); Gallican Confession, art. 4.

21. 21. Sproul credits the formula to John Gerstner in “The Establishment of Scripture,” in Don Kistler, ed., Sola Scriptura! The Protestant Position on the Bible (Morgan, PA: Soli Deo Gloria, 1995). The chapter is cited here as quoted by James Swan at Beggars All (2011). The 1996 excerpt cited above uses the same pair of formulas without naming Gerstner.

22. 22. C. Michael Patton, “Why I Believe the Canon Is Fallible . . . And Am Fine with It!” Credo House, January 25, 2010.

23. 23. Michael J. Kruger, Canon Revisited: Establishing the Origins and Authority of the New Testament Books (Wheaton, IL: Crossway, 2012). Kruger’s model also requires what he calls “providential exposure”—the Church’s actual possession of the books. For a summary, see Jonathan Huggins’s review in Themelios 38, no. 3.

24. 24. Gavin Ortlund, “A Fallible List of Infallible Books?” Truth Unites, September 2, 2024 (video; transcript at Podscripts); Gavin Ortlund, “Answering 2 Objections to Sola Scriptura,” Challies.com, August 20, 2024, a sponsored excerpt-essay tied to his What It Means to Be Protestant: The Case for an Always-Reforming Church (Grand Rapids: Zondervan, 2024). The quotations in this post are from the written essay. Because the only text of the video is an automatic transcript, I paraphrase it.

25. 25. Calvin, Institutes 1.7.1 (Beveridge).

26. 26. Philip Schaff, History of the Christian Church, vol. 7, Modern Christianity: The German Reformation, §9, “The Reformation and Rationalism,” at CCEL.

27. 27. Schaff, History, 7:§9. Schaff summarizes that Luther “called it an epistle of straw as compared with the genuine apostolic writings,” and says of the passage at the close of the 1522 preface that “it was suppressed in later editions.”

28. 28. Schaff, History, 7:§9. Schaff also reports that Wolfgang Musculus “mentions the seven Antilegomena, but includes them in the general catalogue.”

29. 29. John Calvin, Commentaries on the Catholic Epistles, trans. John Owen (Edinburgh: Calvin Translation Society, 1855), “The Argument” to 2 Peter, at CCEL. Calvin concludes that if the letter is to be credited, it came from Peter, “not that he himself wrote it, but that some one of his disciples set forth in writing, by his command,” what the times required.

30. 30. Calvin, Commentaries on the Catholic Epistles, “The Argument” to James and to Jude (Owen).

31. 31. William Whitaker, A Disputation on Holy Scripture, against the Papists, Especially Bellarmine and Stapleton (1588), trans. William Fitzgerald (Cambridge: Parker Society, 1849), Controversy I, Question 3, ch. 1, p. 275, at Internet Archive.

32. 32. Whitaker, Disputation, Controversy I, Question 3, ch. 3, pp. 295–96. Whitaker attributes the objection to Thomas Stapleton and notes that it was “urged also by Campian and by others.”

33. 33. Whitaker, Disputation, Controversy I, Question 3, ch. 7, p. 318.

34. 34. Whitaker, Disputation, Controversy I, Question 3, ch. 7, p. 317: “the most certain argument of all is the testimony of the Spirit, without which it is in vain that all other proofs are applied.”

35. 35. Westminster Confession of Faith 1.2–1.3.

36. 36. Keith A. Mathison, “Solo Scriptura: The Difference a Vowel Makes,” Modern Reformation 16, no. 2 (March/April 2007): 25–29, reprinted at bible-researcher.com. Mathison’s full sentence: “It is, in a very real sense, a creed of the church declaring what the church believes to be the content of Scripture.” He makes the point against what he calls “solo” Scriptura and holds that such creeds carry real, though not infallible, authority.

37. 37. Irenaeus, Against Heresies 3.4.1 (ANF 1), at CCEL.

38. 38. Irenaeus, Against Heresies 3.11.8 (ANF 1), at CCEL.

39. 39. Muratorian Fragment (ANF 5), at CCEL. For the traditional late-second-century date, see Charles E. Hill, “The Debate over the Muratorian Fragment and the Development of the Canon,” Westminster Theological Journal 57 (1995): 437–52; for a fourth-century date, Albert C. Sundberg Jr., “Canon Muratori: A Fourth-Century List,” Harvard Theological Review 66 (1973): 1–41, and Geoffrey Mark Hahneman, The Muratorian Fragment and the Development of the Canon (Oxford: Clarendon, 1992).

40. 40. Eusebius, Church History 3.25.1–6 (NPNF² 1, trans. Arthur Cushman McGiffert), at New Advent.

41. 41. Cyril of Jerusalem, Catechetical Lectures 4.33, 4.36 (NPNF² 7), at New Advent.

42. 42. Cyril, Catechetical Lectures 4.36, which lists the four Gospels, Acts, the seven Catholic Epistles, and fourteen letters of Paul, and nothing more.

43. 43. Athanasius, Festal Letter 39.3, 39.6 (NPNF² 4), at New Advent. Athanasius goes on to name a second tier of books “not indeed included in the Canon, but appointed by the Fathers to be read by those who newly join us” (39.7).

44. 44. Augustine, On Christian Doctrine 2.8.12 (NPNF¹ 2), at New Advent.

45. 45. Breviarium Hipponense, c. 36: Ita ut de confirmando isto canone transmarina ecclesia consulatur. Latin and translation in Roger Pearse, “Canons 29–36 of the Breviarium of the Council of Hippo (393)” (2021). The acts of Hippo survive through the summary (breviarium) adopted at Carthage in 397. The standard edition is Charles Munier, ed., Concilia Africae, CCSL 149 (Turnhout: Brepols, 1974).

46. 46. Code of Canons of the African Church (419), canon 24 (NPNF² 14), at CCEL. The 419 text names the bishop of Rome: “Let this be sent to our brother and fellow bishop, Boniface, and to the other bishops of those parts, that they may confirm this canon.”

47. 47. Ernst von Dobschütz, Das Decretum Gelasianum de libris recipiendis et non recipiendis, TU 38.4 (Leipzig: Hinrichs, 1912); see F. C. Burkitt’s review in Journal of Theological Studies 14 (1913): 469–71, at tertullian.org, which accepts von Dobschütz’s conclusion that the work is “a pseudonymous literary production of the first half of the sixth century (between 519 and 553).” The Catechism still cites the Damasine heading among its references for the canon (CCC 120 n. 91, “DS 179”). A few scholars defend the attribution.

48. 48. Schaff, History, 7:§9.

49. 49. Martin Luther, The Babylonish Captivity of the Church (1520), in Henry Wace and C. A. Buchheim, eds., First Principles of the Reformation (London: John Murray, 1883), 228, at Internet Archive. Luther goes on to attribute to the Church “such a perception … by the illumination of the Spirit, in judging and approving of doctrines; a perception which she cannot demonstrate, but which she holds as most sure.”

50. 50. Whitaker, Disputation, Controversy I, Question 3, ch. 3, p. 296 (“all errors are not capital”); Ortlund, “Answering 2 Objections to Sola Scriptura” (“The only dispute was around the fuzzy edges”).

51. 51. Sproul, Now, That’s a Good Question!, in the excerpt cited in n. 1; Westminster Confession of Faith 31.2–3 (OPC text).

52. 52. John Henry Newman, “Holy Scripture in Its Relation to the Catholic Creed,” Lecture 6, §2, in Discussions and Arguments on Various Subjects (London: Longmans, Green, 1907), 203, at newmanreader.org. The lectures first appeared as Tracts for the Times no. 85 (1838). Newman is voicing the objection, not endorsing it.

53. 53. Newman, “Holy Scripture,” Lecture 6, §4, in Discussions and Arguments, 213. The lecture ends: “It is far more probable that her witness should be true, whether about the Canon or the Creed, than that God should have left us without any witness at all” (215).

54. 54. Newman, “Holy Scripture,” Lecture 6, introduction, in Discussions and Arguments, 198.

55. 55. Newman, “Holy Scripture,” Lecture 6, §2, in Discussions and Arguments, 209.

56. 56. John Henry Newman, An Essay on the Development of Christian Doctrine, rev. ed. (1878), ch. 4, §1, ¶3, at newmanreader.org. Newman adds that the Church “decided,—not merely bore testimony, but passed a judgment on former testimony.” That phrase, read alone, can sound like the Church conferring canonicity, which Vatican I denies; in context Newman means the Church’s authoritative judgment about which books had been received.

57. 57. Whitaker, Disputation, Controversy I, Question 3, ch. 2, p. 284. Ortlund draws on the same passage.

58. 58. Dei Filius, ch. 3 (Tanner), at papalencyclicals.net.

59. 59. Dei Verbum §10.

60. 60. Westminster Confession of Faith 1.10.

61. 61. Dei Filius, ch. 3 (Tanner).

62. 62. Second Vatican Council, Lumen Gentium §12 (1964).

63. 63. On the bishops, see Lumen Gentium §25: they “proclaim Christ’s doctrine infallibly whenever, even though dispersed through the world, but still maintaining the bond of communion among themselves and with the successor of Peter, and authentically teaching matters of faith and morals, they are in agreement on one position as definitively to be held.”

64. 64. On Hebrews at Rome, see Eusebius, Church History 3.3.5: “some have rejected the Epistle to the Hebrews, saying that it is disputed by the church of Rome, on the ground that it was not written by Paul” (New Advent). On Revelation in the East, Newman reports Jerome’s statement that around 400 “the Greek Church rejected the Apocalypse” (Essay on Development, ch. 4, §1).

65. 65. Joe Heschmeyer, “The Canon Question Isn’t about Papal Infallibility. Here’s Why,” Shameless Popery, Catholic Answers, January 9, 2025, responding to Ortlund and to Cameron Bertuzzi. Heschmeyer appeals to CCC 91–92 and proposes that “instead of talking about papal infallibility, let’s talk about people infallibility.”

66. 66. Luke 24:44 (RSV); the Revised Standard Version is used for Scripture quotations in this post.

67. 67. Matt 23:2–3 (RSV), at BibleGateway.

68. 68. John 16:13 (RSV).

69. 69. Cajetan, preface to his commentary on Hebrews, as translated in B. F. Westcott, A General Survey of the History of the Canon of the New Testament, 7th ed. (London: Macmillan, 1896), 483, at Internet Archive. Westcott notes that Cajetan’s commentary on the historical books of the Old Testament was dedicated to Clement VII. On Erasmus and Cajetan together, see the Catholic Encyclopedia, “Canon of the New Testament” (1908): “Even a few Catholic scholars of the Renaissance type, notably Erasmus and Cajetan, had thrown some doubts on the canonicity of the above-mentioned Antilegomena.”

70. 70. Council of Florence, Bull of Union with the Copts (Cantate Domino), Session 11, 4 February 1442, trans. Tanner, at papalencyclicals.net.

71. 71. Council of Trent, Session 4 (8 April 1546), Decree concerning the Canonical Scriptures, in J. Waterworth, trans., The Canons and Decrees of the Sacred and Oecumenical Council of Trent (London: Dolman, 1848), 17–19, at Hanover College.

72. 72. Newman, Essay on Development, ch. 4, §1, where the “Canon of the New Testament” is the first of his illustrative instances and “The Homoüsion” the fifth.