The word “canon” is derived from the Hebrew qâneh, which means a reed, but which may be understood as a measurement. For many people this idea is circular at best. How does one measure that which alone measures us? That may seem a pious logic, but it misunderstands the question. On the other extreme there is a kind of hasty conclusion about the historical record that ignores the logic that makes the record count in the first place.

Along these lines, Kevin Vanhoozer writes that, “History alone cannot answer the question of what the canon finally is: theology alone can do that.”1 Consider why this is. The “canon of WHAT?” which raises the matter of what Michael Kruger calls “canon semantics.”2 Defining the essence of that object that we are calling Scripture isn’t simply about asking historical questions about the when and where of mere writings. Now Kruger will take from this a different lesson than we will, but it is the right starting point. What on earth does canon even mean if its truth is not exactly what it is, independent of any subjective trump card? 

The question we are really asking is this:

Whether the church creates the canon, or the canon creates the church, and is thus received on account of a set of objective criteria.

The first three points of our outline will follow exactly that of Kruger, in his book Canon Revisited: (i) Canon as community-determined; (ii) Canon as historically-determined; (iii) Canon as self-authenticating; and then, I will follow with my own (iv) evaluation and cumulative criteria. Note that each of those “typologies” of canonicity is really a single criterion acting as a sole explanation. 

Canon as Community-Determined

The label here speaks for itself. It is the church community that decided. Questions will remain: Which community? Why the need by that community? How was that community able to do so? I ask the questions in this way because I want to plant seeds of doubt right from the start in the idea that this could be the all-explaining factor—especially at so early a time, and in a time when the church had absolutely zero power, until the very eve of when that power grab supposedly happened. Let us just say that I am skeptical of such a feat.

That being said, Kruger divides this first typology into four “models”: the Historical-Critical Model, the Roman Catholic Model, the Canonical-Criticism Model, and the Existential / Neo-Orthodox Model.

About the first—the Historical-Critical Model—this is what another author calls “the naturalization of canon,”3 championed by twentieth century scholars, Walter Bauer and James Barr, namely that the whole process of recognition of the right books and codification into a single list is a human activity. It could have been in defense against the Marcionite heresy;4 or else a political power grab by the Church; or the most normal social phenomenon called “self-description,”5 that is, defining who we are.

Hence the method is descriptive, not normative. Cynically stated, “the real concern … is not to declare which books are the ‘right’ ones, but to make sure no one else declares which books are the ‘right’ ones.”6

What is Kruger’s evaluation of this? He writes, “simply demonstrating some human involvement in the canonical process is not sufficient to demonstrate sole human involvement.”7 In other words, this model is guilty of the false dilemma. Yes, the early church is involved in the process, but it simply does not follow that this is incompatible with “intrinsic qualities”8 that demand a book’s reception. 

Secondly, there is the Roman Catholic Model. This shares a common method with the historical critical model. Their threefold model of authority (Scripture, tradition, and the Magisterium) implies that the Magisterium alone may rightly interpret the other two sources. This authority includes authentication of inspired books. Their argument against the Reformed view: “you cannot have Scripture as the ultimate authority if you have no certain way of knowing what Scripture is.”9 There needs to be a sufficient external authority: apart from some “inspired table of contents,” which we do not possess.

It should be pointed out that there are different sub-models among Roman Catholic theologians: (i) the Church as epistemological authority: the magisterium is the sole way of knowing which books;10 (ii) the Church as causal grounds for the canon.11 Kruger points out that the “inspired table of contents” objection is immaterial because it too would fail to satisfy: “On what basis do you know that this twenty-eighth book comes from God?”12

The real question is a priori: Can a document ever be self-attesting? This also raises doubt “about whether the Catholic model pays adequate attention to the intrinsic authority built into these books”13 (as in Model 1). Either such qualities are decisive or not. Then Kruger makes a four point argument from Scripture and history ruling out the possibility of a Scripture “derived” from the church: 

“(1) Although the New Testament was not completed all at once, the apostolic teaching was the substance of what would later become the New Testament … (2) The earliest Christians did have a canon, namely, the Old Testament itself (Rom. 15:4; 1 Cor. 10:6; 2 Tim. 3:15-16), which seems to have existed just fine prior to the founding of the church … (3) From the very earliest days, believers received Paul’s letters as Scripture (1 Thess. 2:13), Paul clearly intended them to be received as Scripture (Gal. 1:1-24), and even other writers thought they were Scripture (2 Pet. 3:16) … (4) It was not until the Council of Trent in 1546 that the Roman Catholic Church ever made a formal and official declaration on the canon of the Bible, particularly the Apocrypha … Are we to believe that the church had no canon for over fifteen hundred years, until the Council of Trent?”14

Finally, building on the “inspired table of contents” dilemma, Kruger asks, “If the Roman Catholic Church believes that infallible authorities (like the Scriptures) require external authentication, then to what authority does the church turn to establish the grounds for its own infallible authority?”15

Thirdly, there is the Canonical-Criticism Model. This model is roughly thirty years old. It says that a text is best understood not by its earliest stage of development, which demands speculation, but in its latest: (1) the final product and (2) its function within the rest of the canon.16 Only a “final product” unifies diverse streams and the redactors did what they did so that the writing would transcend any one generation. Consequently a book is canonical if it was included in the canon (1) settled upon by the early church as (2) the basis for their understanding of the gospel.17 Although in Brevard Childs’ model, this was a process and not one sudden event.

What is Childs’ definition of the canon? It is not exclusive, final, and closed, but functional: i.e., how it is received by the community. Kruger replies that Childs’ concept of redaction and receptivity actually broadens the traditional understanding of inspiration to the whole church community and post-apostolic. Theoretically “there seems to be no reason offered by Childs for why the canonical documents would not continually be open to revision even up to the present day.”18

He also leans on Carl F. H. Henry’s critique of Childs:

“If the canon represents a judgment by the community of faith … does not the community really constitute an authority just as ultimate, and even more so, than the canon?”19 

Fourthly and finally, there is the Neo-Orthodox Model. So the theology of Karl Barth and Emil Bruner had implications for even the canon of Scripture. We might think of how Childs’ view made the books canonical insofar as the whole community recognized a “final” redaction that fit their faith, and then boil it down to the individual’s faith (hence the other name for this: Existential). In this view, the Scriptures are not the Word, but a witness to the Word, namely Christ; so revelation is not propositional but an event or encounter by the Spirit.

Historical truth is unimportant. “Christians do not experience God in the Scriptures because they are canonical; rather they are ‘canonical’ because Christians experience God in them.”20 The biggest problem here, for Kruger, is the “unfortunate separation of the authority of God and the authority of Scripture,”21 to which he adds three more problems that are more or less the same as in the other problematic views.

Canon as Historically-Determined

As with the first typology and its four models, in this second typology Kruger divides between two models: the Canon-within-the-Canon Model and the Criteria-of-Canonicity Model. Both of these reject what was common in the first typology, namely that the community is determinative for what constitutes the canon. Instead, here we attempt to trace the origins through history by some standard still external.

First, let us examine this Canon-within-the-Canon Model. Examining the “historical development” of a book, one can determine which parts are genuine. Specifically the origins of the book are sought out for “core” material. Unlike Kurt Aland’s academic approach,22 the vast majority who take this position look for that core on the basis of what our faith “should” emphasize. Thus even Luther’s resistance toward James was a species of this thinking. But the Quest for the Historical Jesus and its extreme vestiges in the Jesus Seminar is a more absurd end of it. On the other hand, this highlights another crack in the armor of Kruger’s typologies. The trouble with these criteria does not seem to be their “biblical-externality” but really their subjectivity, which is quite another thing altogether.

The kind of “object” for Aland’s more serious scholarship and the Jesus Seminar’s less serious scholarship are both a matter of historical origins; whereas for early Luther the object was orthodoxy.23 Or was it? Is it not better to call it “perceived orthodoxy”? That is, unless we have come to the place where we want to reduce orthodoxy per se to a neutral or subjective perception.

Granted, there are also “ideological objects” which can function as the criterion, as Kruger mentions feminist and liberationist theologies.24 Marcionism was the most famous ancient version of this, whereas the Jefferson Bible was a famous modern version. However, one must be careful not to easily place orthodoxy as such on this list. Early Luther may have been incorrect, but his trouble was not that orthodoxy was regarded as an objective criteria.

That said, Kruger’s pattern of evaluation is by now familiar: “The fundamental problem with the canon-within-the-cannon approach is that it subjects the Scripture to a standard outside itself, namely, whatever criteria scholars set up to evaluate its truthfulness.”25 The relevant nuance may be easy to miss.

For Kruger (signalled by the logically inclusive word “namely”), a “standard outside” of the Scripture is logically equivalent to (or “namely”) “whatever criteria scholars set up to evaluate its truthfulnesss.”

But is that a fact? What we are beginning to see is that the classicalist versus presuppositionalist debate features in the question of canon as well. And that will feature more explicitly in Kruger’s critique of the next model.

For the Criteria-of-Canonicity Model, objective criteria like “apostolicity, orthodoxy, usage, etc.”26 function as the criteria for determining whether or not a book is canonical. In other words, this view is just the standard evidentialist apologetic for the canon, according to Kruger, and he sees B. B. Warfield as the standard-bearer of this position among modern theologians. Once these criteria are assumed, historical investigation then shows how our New Testament books “meet these criteria.”

He even cites Paul Helm’s use of the term “externalism,” which says that “external data are required to validate the Scriptures as the Word of God.”27 We may want to ask what all is being loaded into the terms “required” and “validate.” It may surprise conservative Evangelical readers, looking for the most authoritative contemporary work on this subject of canon, to find it at odds with Warfield and the idea of objective criteria. 

What then are Kruger’s criticisms? There are three that are basic: 1. “Neutral” historical investigation, namely based upon the “accepted canons” of present research, is problematic: “Are scholars really able to check their worldviews at the door so easily?”28 And some methodologies are based on explicitly non-Christian assumptions. 2. The criteria themselves unduly inflates the church’s intentional role, as if the early church went out, like a “search committee.”29 3. He also calls attention to deficiencies in the criteria, but as he describes the function of each, he does so as if any one of them were expected to stand on their own,30 a view which no classicalist or evidentialist would recognize as their view. 

At any rate, “The criteria of canon are best understood as the way in which the Scripture sets the terms for how its own origins are to be investigated and explored.”31

He offers as an alternative term “attributes of canonicity,” rather than criteria for canonicity. Recall how he had used the word intrinsic rather than objective. We have no problem with the word intrinsic—unless one reduces that which is ‘intrinsic to Scripture’ to what cannot be objective and knowable in the world outside of our minds. 

Notice that, against both of these models, that which is objective as a criteria is problematic because whatever those objects are become, via human reasoning, the real standard over Scripture itself. As Kruger said elsewhere, “All highest authorities are self-authenticating … or it’s not the highest authority.”32 So if all the other views “ground the authority of the canon in something outside the canon itself,”33 let us see if Kruger can succeed in limiting his criteria in the canon to the canon alone.

Canon as Self-Authenticating

The third typology is the one Kruger himself endorses—the one that he takes to be the traditional Reformed view, consistent with the Westminster Standards, but which I say is only Van Tillianism applied to the question of canon and not in step with those Westminster Standards. 

To say “that the canon is self-authenticating is simply to recognize that one cannot authenticate the canon without appealing to the canon.”34 To be fair, he is quick to add that external data is not incompatible with self-authentication. The key is that external evidence cannot be autonomous. But how exactly would we know if evidence is being used in an “autonomous” manner?

The function of the evidence must be “warranted by Scripture and interpreted by Scripture.”35 Moreover, his attributes of canonicity—which he shorthands as 1. divine qualities, 2. corporate reception, 3. apostolic origin36—are combined with the internal testimony of the Spirit to overcome the noetic effects of sin, so that the mind can properly assess those divine characteristics. Of course on this point there is agreement among all who call themselves Reformed.

The key distinction, then, lies in what Kruger means by evidence that is “warranted and interpreted” by Scripture. To be warranted by Scripture may only mean that the use of evidence is consistent with Scripture. However, the use of the word “interpreted” here is sharper. This suggests that the use of evidence must (in some way) be more stringently set forth by Scripture.

Interestingly, Kruger self-consciously acknowledges that he is using the Reformed concept of self-authentication in a “broader fashion”37 than the classical Reformed tradition did.

What emerges is something like a coherence theory of truth (often compared to a web) applied to bibliology. In this case, the attributes of canonicity are only mutually reinforcing as the Holy Spirit both creates and confirms the indicia. Again, the Reformed as a whole confess that the Spirit confirms the indicia. But it is typical of a Van Tillian recasting of this doctrine that only that which is expressly “revelational” is of the Spirit in the relevant sense. So the Spirit also creates the signs that He confirms. It is crucial for Kruger to emphasize that the attributes “are not three independent and disconnected qualities that canonical books happen to possess, but each attribute implies and involves the other two.”38 He calls it a “web of mutually reinforcing beliefs.”39 

Evaluation and Cumulative Criteria

Let me add a few more words of critique against the more unorthodox views. First an assumption that seems to be shared needs to be scrutinized: Even demonstrating maximum human involvement does not fulfill the critic’s burden. The share of involvement premise assumes that the divine and human are competing to begin with. But on what ground should we accept that premise? A second problem with the more critical models is that they are circular. For example, Barr objected that Jesus, “in his teaching,” never sanctioned a New Testament canon. But what is meant by “his teaching” apart from a coherent New Testament? 

As to the Roman Catholic model, it also contradicts the New Testament teaching that the word creates the church. Even viewing “the apostolic teaching” in any pre-canon form, this word is the substance of what the canon could ever be. Thus to reverse the relationship between church and canon is to, by implication, reverse the relationship between church and word. We can imagine Rome offering three kinds of answers to the difficulties brought up by Kruger; we can also quickly see why they fail:

First, the Church’s infallible authority is derived from Scripture. There are two problems with this: (a) circular reasoning—a proof that the Church proves Scripture’s word by Scripture proving the Church’s word! (b) There is no such biblical teaching; Second, the Church’s infallible authority is attested by the history of the Church – i.e., its origin, nature, progress. However, (a) the church is not an unbroken succession; and (b) the church is not pure in either doctrine or morals; Third, the Church’s infallible authority is self-attesting. If so, then (a) on what grounds do they reject the Protestant doctrine? (b) This completes the Roman tendency to set the church over the Scriptures.

All of this strikes at the heart of an objective view of truth. J. I. Packer put it in this way: “The Church no more gave us the New Testament than Sir Isaac Newton gave us the force of gravity. God gave us gravity … Newton did not create gravity but recognized it.”40 

To Kruger’s own view, two things must be said.

First, it is simply not the case that to be a classicalist or an evidentialist—that is, to see objective signs as ground for understanding a book to be canonical—requires that any of those criteria function independently of the others. Second, do standard methods and rules really require “checking one’s worldview at the door” to begin with? The burden of proof is on Kruger to show how any of the ordinary criteria are “non-Christian” or “anti-Christian” or “unbelieving” standards. The question is not whether or not the unregenerate will distort these. Of course, they will distort both them and those cited by Kruger. Epistemological criteria holds out no promise to regenerate. We agree that only the Spirit can do that.

Neither does the subjective conception of such criteria equal the objective criteria. As to apostolic origin, there is no tertium quid—either Paul wrote the Pastoral Letters or he did not. If he did, they possess that mark. As to whether lists existed before Nicaea, either they did or they did not. If Irenaeus’s list and the Muratorian Canon match up, then those books possess those marks as well. Yet neither of those are “in the Bible.'“ Neither is the objective left-to-right nature of the timeline by which one draw that conclusion “in Scripture,” nor, while we are at it, the logically antecedent knowledge that this is how time is signified, and so on, “in Scripture.” And that is just one aspect of one of the criteria. To concede such external criteria to a “neutral” and “autonomous” zone of secularity is, well, to do just that. It is to concede objective truth to the unregenerate, the very thing the presuppositionalists inform us that only their method avoids.   

Such criteria are not “validating” Scripture to begin with. They are signs to us, validating or invalidating, if you like, our own recognition of these books as the word of God.

To summarize the view I will take on the criteria of canonical books, these may be divided under four heads: (1) authorship of an Apostle, or one in the “apostolic circle,” (2) the book’s spiritually transformative character, (3) its orthodoxy in terms of consistency with the rest of the canon, and (4) that it was received by the universal church.

It is important to repeat that such are objective qualities. That is to say, their canonical status was recognized by the early church and thus received, not invested into the books by the church. It is also important to note that these criteria have a cumulative effect and it is not necessary that a biblical book’s author be known in all cases. The Old Testament book of Job and the New Testament book of Hebrews, while differing on a spectrum of what we know of their author, at least share this in common.  

Concluding Thoughts

I make much of Kruger’s book here not to marginalize it. Obviously I disagree with the exact typologies that he has and the rationale he gives when it comes to “evidentialism”. But this very exercise highlights a missing component of too much of our apologetics and our doctrine of Scripture. That makes his typology of canonicity a very valuable contribution. Everyone who presumes to speak about the canon of Scripture needs to be exposed to the fact that they are theologizing. More than that, they need to realize that they are engaging in circular reasoning in making so much of the actions of Nicaea, or the Church authority of that general time period.

It is only on their definition of canon as being “Nicaea-produced” or “Orthodoxy-produced,” or whatever else of the like, that one may conclusively deny the same canon existing beforehand. Without that definition, nothing prevents those very books from being the rule already, and the evidence of quotations from the second and third centuries lends that rule a high degree of probability.

We can put one more finer point on it. The inconceivability of there being no unanimous canons (at least regionally) before Nicaea is seldom considered. Kruger says,

“In order to recognize a certain book as Scripture in the first place, an early Christian would have needed to be able to say another book in his library was not Scripture.”41

In short, the existence of Scripture per se proves a canon per se: some canon at least. 

_______________________

1. Kevin Vanhoozer, The Drama of Doctrine (Louisville: Westminster John Knox Press, 2005), 146.

2. Michael Kruger, Canon Revisited: Establishing the Origins and Authority of the New Testament Books (Wheaton, IL: Crossway, 2012), 27

3. John Webster, quoted in Kruger, Canon Revisited, 30.

4. Adolf von Harnack cited in Kruger, Canon Revisited, 31. Harnack argued this in History of Dogma, II.62, which may be why Kruger does not factor into his own explanations for the later acceptance of certain books: e.g., 1-2 Peter.

5. Kruger, Canon Revisited, 31. Kruger cites both Paul Ricouer’s “The ‘Sacred’ Text and the Community” and D. H. Kelsey’s The Uses of Scripture in Recent Theology for this concept; but the same is said about the broader concept of orthodoxy in Alistair McGrath’s book on Heresy (2009), and Kruger does mention Bauer’s more critical thesis in this same direction: Orthodoxy and Heresy in Earliest Christianity (1934).

6. Kruger, Canon Revisited, 33.

7. Kruger, Canon Revisited, 35.

8. Kruger, Canon Revisited, 35. Note that Kruger’s choice of the term “intrinsic” over “objective” will parallel his later preference (which he states) of the term “attributes” over “criteria,” all following from the perspectival element of his commitment to Van Tillianism.

9. Kruger, Canon Revisited, 40.

10. Both the Council of Trent and Vatican I speak like this.

11. This is echoed by Karl Rahner, Hans Küng, and Peter Kreeft.

12. Kruger, Canon Revisited, 43.

13. Kruger, Canon Revisited, 43 – Note, again, Kruger’s use of “intrinsic” qualities, which are essentially “authority built in” and which oppose “external validation,” a typical conflation among Van Tillians of the logically antecedent and the ultimately authoritative.

14. Kruger, Canon Revisited, 44, 45.

15. Kruger, Canon Revisited, 46.

16. Kruger, Canon Revisited, 49.

17. Kruger, Canon Revisited, 51.

18. Kruger, Canon Revisited, 54.

19. Carl F. H. Henry, quoted in Kruger, Canon Revisited, 55.

20. Kruger, Canon Revisited, 62.

21. Kruger, Canon Revisited, 64.

22. cf. Kurt Aland, The Problem of the New Testament Canon (1962), cited in Kruger, Canon Revisited, 68.

23. Kruger, Canon Revisited, 70.

24. Kruger, Canon Revisited, 69.

25. Kruger, Canon Revisited, 71.

26. Kruger, Canon Revisited, 74.

27. Paul Helm “Faith, Evidence, and the Scriptures,” in Scripture and Truth, ed. Carson and John D. Woodbridge (Grand Rapids: Zondervan, 1983), cited in Kruger, Canon Revisited, 74.

28. Kruger, Canon Revisited, 77.

29. Kruger, Canon Revisited, 81.

30. Kruger, Canon Revisited, 82.

31. Kruger, Canon Revisited, 85.

32. Kruger, Hebrews to Revelation (RTS) Lecture 22.

33. Kruger, Canon Revisited, 88.

34. Kruger, Canon Revisited, 91.

35. Kruger, Canon Revisited, 90.

36. Kruger, Canon Revisited, 97, 103, 108.

37. Kruger, Canon Revisited, 91.

38. Kruger, Canon Revisited, 115.

39. Kruger, Canon Revisited, 115.

40. J. I. Packer, God Has Spoken: Revelation and the Bible (Grand Rapids: Baker, 1993), 109.

41. Kruger, Canon Revisited, 36.