Wednesday, April 02, 2008

Paradox In Christian Theology

Paradox in Christian Theology: An Analysis of Its Presence, Character and Epistemic Status

-Reviewed by Paul Manata-

Available here.

Introduction To A Mystery Novel

What would you say of a book that makes it part of its aim to establish that certain essential Christian doctrines (say, the doctrine of the Trinity and the doctrine of the Incarnation) present the appearance of logical contradiction? What would you think of a book that argues that no theologian in history has been able to present a statement of those doctrines that do not avoid logical tension via (some kind of) inconsistency, while simultaneously remaining faithful to Christian orthodoxy as presented in the culture-identifying creeds of the early church, or to the explicit (and implicit) statements of Scripture (from which those creeds derive their authority)? In other words, Christians can have their logical consistency, or they can have their orthodoxy, but they cannot have both. What if the author of this book believes that he established those points? In addition, what if he looked at some of the best contemporary Christian philosophers and their attempts to put forward fully consistent models (ones which do not lead to any logical headaches) and showed that they all fail as well (in the sense described above)? Indeed, what if this book argued that the Christian was without theological and philosophical defenses that save both orthodoxy and logical consistency (of the implicit kind)? You might think I am describing the latest atheological work to hit the market.

What if I also added that the author not only demonstrates the above, agrees that some of our most precious doctrines of the Christian faith resist full logical consistency in our formulations of those doctrines, but that he is also a Christian? He is an orthodox Christian who operates out of the venerable Reformed tradition. Not only that, what if I told you that the author not only believes these doctrines to appear logically inconsistent (making them paradoxical), that we also have not resolved the paradoxes, but he also believes that the Christian (almost any kind of Christian, from scholar to layman) is warranted in believing the conjunction of claims that lead to the paradox? That it is a perfectly rational thing to believe? That the presence of paradox cannot be seriously considered as an intellectual obstacle to belief in Christianity? What if I included that information? You might very well scratch your head and call it a mystery! Don’t we all love a good mystery novel? . . . The book I am about to describe is just that. It is a mystery novel. One in which the mystery is left unsolved, at that!

In his bold book, Paradox In Christian Theology: An Analysis of Its Presence, Character, and Epistemic Status,1 James Anderson sets out to show that certain doctrines of the Christian faith are paradoxical, but may be reasonably believed in spite of this feature (if not because of it). Anderson also argues that these doctrines are not actually contradictory, but merely apparent.2 However, believing this appearance could be cause for the charge of irrationality to stick. Thus, Anderson provides a model by which the Christian cannot only show that the doctrines are not actual contradictions, but how he can also be rational in affirming these apparently contradictory sets of propositions. With the rigorous mind of a philosophical theologian operating within the analytic tradition (and I believe with the heart of a pastor), Anderson successfully presents and defends this model. In so doing, he offers one of the most intriguing and ingenious responses to one of the most difficult challenges to the Christian faith.

If that were not enough, PCT will lift the Christian reader to profound new heights of reverence and awe as they contemplate a God whose ways are not our ways and whose thoughts are not our thoughts (Isa. 55:8).3 And so, if I can offer a seemingly paradoxical observation of my own: PCT will provide new insights about God and how he has structured things in our world and in our cognitive equipment, yet we will stand in more awe of God with this increased information. This is seemingly paradoxical because usually the more you learn about someone, the more the gap between the two of you decreases, but it is just the opposite with God. The more we learn about him, the more we realize how far the distance is between creature and Creator. It appears that the more we know, the less we know. (This is not paradoxical in the sense used in the book, though; but in keeping with the theme of the book under review, I figured my hyperbole might be forgiven.)

I am sure it is obvious that I think highly of this book. It is a work of philosophical theology, and Anderson writes in a clear, precise, and detailed fashion. (However, this does not stop him from slipping in a few humorous quips from time to time.) His command of the literature, the history, the intricacies of the historical debates, and the arguments on all sides is nothing short of impressive. He interacts with contemporary theologians as well as contemporary philosophers. He discusses a multitude of topics and brings them all to his reader in a clear, accessible way. The reader may have to work hard in parts, but that hard work definitely pays off in spades.

In PCT, Anderson thus establishes himself as one of the brightest, rising stars in Christian philosophy and apologetics today. The rest of this review will consist of looking at the three aspects Anderson seeks to analyze regarding paradoxes in Christian theology: (i) its presence, (ii) its character, and (iii) its epistemic status, and provide his answers. Anderson presents and defends his thesis in two parts. I will follow the structure of his book for this review. I do include some criticisms of his book throughout the review; but be careful, if you blink you may miss these comments (especially since most of them are filed away in the endnotes), as I agree substantially with Anderson and really have no major gripes or criticisms. (I do recommend that the reader read the endnotes as they contain what I take to be some valuable resource material, as well as (I hope) some valuable comments above and beyond any remarks critical of PCT.)

The Introduction (Ch.1)

Before Anderson looks at his paradigmatic doctrines, which begin in part one of the book, he lays some foundations, defines some terms, and offers in a broad-brushed way the approach he takes in the book. This is all very helpful, and is indicative of the patient way Anderson approaches his subject, always remaining careful to make sure his reader is given the necessary framework to follow Anderson’s building project. Chapter 1 lays some groundwork.

As stated above, Anderson’s goal in the book is to show that certain doctrines of the Christian faith are paradoxical, but may be reasonably believed in spite of this feature (if not because of it).

But why even write a book on this issue? Is it even significant? The significance of paradox in Christian theology “lies in the potential implications for the epistemic status of Christian beliefs.” Atheists and agnostics have appealed to their presence in support of their non- or disbelief. Within the Christian camp, however, some have been disposed to laud their presence, while others loathe their presence.

The various attitudes towards paradox can be expressed thusly:

[P1] It is always irrational to affirm a paradoxical doctrine; some central Christian doctrines are paradoxical; therefore, adherence to Christian faith is always irrational (because of paradox).

[P2] It is always irrational to affirm paradoxical doctrine; no central Christian doctrines are paradoxical; therefore, adherence to Christian faith is not always irrational (because of paradox).

[P3] It is not always irrational to affirm paradoxical doctrine; some central Christian doctrines are paradoxical; therefore, adherence to Christian faith is not always irrational (because of paradox).4

Each position has problems, notes Anderson. [P1] is problematic for those Christians who believe they have good epistemic grounds for their faith; [P2] is out of touch with the widespread notion (of both Christian and non) that there are unresolved paradoxes in the Christian faith; and [P3] has failed to offer a satisfactory account of the circumstances under which said paradoxical doctrines might rationally be believed. Anderson says that some advocates of [P3] have offered sketchy defenses of the intellectual propriety of paradox; none of them address the prior question of what constitutes rationality; i.e., (i) what is required for a belief to have that honorific title ‘rational’ bestowed upon it, and (ii) whether adherence to paradoxical doctrines can ever meet these requirements. Anderson sets out to fill this gap.

However, you cannot get very far in discussions of this nature without people asking you how ‘paradox’ is being defined. Anderson is sensitive to this, and so immediately offers his definition.5 Anderson defines ‘paradox’ thusly:

X is paradoxical = df X amounts to a set of claims which taken in conjunction appear to be logically inconsistent.

Note well the qualifier ‘apparent.’ Thus, a paradox does not entail a logical inconsistency per se, just the appearance of logical inconsistency. This definition “presupposes that a meaningful distinction can be made between apparent and real contradiction.” Anderson defends this distinction in 6.2.1 and 7.4.1.

Lastly, Anderson gives the reader a preview of what to expect in the coming pages, whetting the appetite of the reader.

Anderson notes that [P1] --> [P3] suggest the presence of two key questions concerning paradox in Christian theology: (i) Are any essential Christian doctrines genuinely paradoxical? If they are, (ii) can a person rationally believe them? Answering these two questions will determine which of the positions, [P1] --> [P3], “is closest to the mark.” Anderson’s book consists of two parts. Part one deals with (i) and part two with (ii). How Anderson fills out parts one and two make up the material of the book. Part one has three chapters (2, 3, 4) dedicated to answering (i) as well as various strategies employed to deal with the presence of paradox. Part two has four chapters (5, 6, 7, 8) dedicated to answering (ii).

Anderson answers both questions in the affirmative. As for (i) Anderson first argues (in Ch. 2) that the doctrine of the Trinity is paradoxical. He reaches this conclusion by: (a) surveying the early Trinitarian controversies that lead us to two “definitive statements of orthodoxy” (the Niceno-Constantinopolitan and Athanasian creeds), and by (b) critically examining contemporary interpretations of the doctrine of the Trinity, especially those that try to develop the doctrine with the aim of logical consistency and maintained orthodoxy. Then in Chapter 3, Anderson looks at the doctrine of the Incarnation and concludes that it is paradoxical in much the same way he proceeded in Chapter 2. Anderson chose these two doctrines because they have a putative status as paradoxical, and because of their ecumenical appeal. Thus, all stripes of Christians can profit from Anderson’s book (he also believes the same about “several doctrines distinctive to the Reformed tradition,” but does not focus on these for obvious reasons, such as “relevance” for “the wider Christian community”). Then in Chapter 4, Anderson considers a “range of responses to the paradoxicality of Christian doctrines.” He intends to show that these “coping strategies” fail on either theological or philosophical grounds (or both). This paves the way for him to present his case for an affirmative answer to (ii).

In part two Anderson supports his affirmative answer by first setting forth an account of how affirming Christian doctrine in general can be rational (Ch. 5). Chapter 5 introduces the reader to many epistemically important concepts that lay the groundwork for his account. Chapter 6 “sets forth a model for construing theological paradox” in terms of which believers that run the intellectual gamut can be rational in holding to said paradoxical doctrines. Chapter 7 defends this model against many potential objections (potential, you must remember, because Anderson’s work is highly original). Lastly, Chapter 8 highlights what Anderson takes to be the main implications and offers suggestions for further study.

Part 1-The Presence of Paradox
The Trinity & The Incarnation (Ch. 2 & 3)

To ask about whether it is rational to believe the paradoxes in the Christian faith, there must first be paradoxes in the Christian faith. Chapters 2 and 3 set out to substantiate the claim that there are. Chapter 2 discusses the Trinity, chapter 3, the Incarnation. I will go over Anderson’s discussion of the Trinity and refrain from his discussion of the Incarnation. I will give my reasons for this in my closing paragraph of this section.

All Christians agree that the Trinity is a mysterious doctrine. A brief formulation is that “there is only one God who exists in three distinct persons.” This immediately raises problems about the logical status of this essential doctrine. Anderson asks a series of questions:

Is not God (at least in the Judeo-Christian tradition) a personal being? Does this not suggest that if there is only one God, there is only one divine person? How then can this one divine person be also three divine persons? Put another way: if there are three numerically distinct persons, each fully divine, does this not imply that there are three Gods, rather than one?6
The answers to these questions bring out the paradox of it all. Anderson attempts to show that this charge of paradoxicality is well-justified. Anderson notes that most thinkers admit that interpretations of the Trinity have been offered which appear, under scrutiny, to be free of logical difficulty, yet whether they are also free of theological difficulties, at the same time, is another matter.

This brings out the criticism, framed in the form of a dilemma: Either remain orthodox and face paradox, or, banish paradox and face heterodoxy. Anderson’s assessment of this criticism is as follows: (1) He studies the historical development of the doctrine of the Trinity. The focus is not only on the words of the creeds and confessions, but the concepts associated with the words, as well as the theological concerns that conditioned the development of the doctrine. Anderson’s (stated) main aim is to secure an adequate understanding not only of the formal statement of the doctrine of the Trinity, but also the constraints placed on legitimate interpretations of that doctrine. This will allow the reader to see how paradox arose, as well as how the formulators were prepared to tolerate paradox within their theology. (2) Anderson reviews representatives of the contemporary scene with an eye towards those who have sought to avoid the charge of paradox. These explications of the doctrine are judged by two criteria: (i) their fidelity to orthodoxy in light of the result of the work done in (1), and (ii) their success in avoiding paradoxical formulations. Anderson then ends his romp with a look at those contemporary thinkers who have conceded the paradoxicality of the doctrine of the Trinity. I will touch on (1) and (2) respectively.

In (1) the reader is treated to the fruits of Anderson’s labors. It is obvious he has detailed knowledge of the early Trinitarian debates and controversies. He boils down hundreds of years of debate quite nicely, hitting on those highlights essential to bringing out the paradox of the Trinity. The footnotes are copious and the reader can do follow-up work if he desires. Anderson discusses Monarchianism, Praxeas, Noetus, Sabellius, Tertullian, the ante-Nicene fathers, the Nicene fathers, Origen, Athanasius, the Arian controversy, the Cappadocian fathers, Hilary of Poiters, The Post Nicene fathers, Augustine, and many in between.

It would be beyond the purpose of this review to go over the historical progression which led to the conception of the paradoxical formulation of the doctrine of the Trinity, so I will simply point out a couple of pertinent points made along the way. Then, I will summarize Anderson’s main points that he established from the survey which show the paradoxicality of the doctrine of the Trinity.

Anderson points out that in these debates the positions later dubbed heterodox or heretical did not have their basic premises challenged. Their arguments also flowed with seemingly impeccable logic. However, the conclusions were problematic. They would tend towards tri-theism, modalism, or the less-than-full divinity of Christ, etc. What had to be affirmed was a robust monotheism, and the distinctness of the persons from each other. Anderson notes, when speaking of the Nicene council,

What the Council challenged, based on what they took to be unacceptable conclusions, was the legitimacy of Arius’s inferences from those premises. In short, the overriding concern of the Fathers was not so much to develop a scrupulously coherent theology (that was arguably the principle motivation of their opponents) but to be faithful to Scripture and tradition (emphasis original).7
So, J.N.D. Kelly described Arius and his supporters as “rationalists at heart.”

The second point noted by Anderson that I will mention is that part of the debate was over the terms used and which ones the council settled on, i.e., that the son is homoousion with the Father (of the same substance), for instance. This paved the way to debate whether the identity expressed between the persons and the divine nature was of generic or numerical identity. Anderson succeeds in showing that the latter was the preferred orthodox position on the identity. One reason is that generic identity (those relationships the Social Trinitarians favor) tends towards tri-theism. He lists the reasons for why this was thought to be so.

Anderson boils down the main points he has demonstrated from his survey of the historical debates and development of the creeds to six Trinitarian claims:

(T1) An orthodox doctrine of the Trinity must uphold biblical monotheism (ruling out polytheism).

(T2) An orthodox doctrine of the Trinity must maintain the full and equal divinity of each of the three persons (ruling out subordinationism).

(T3) An orthodox doctrine of the Trinity must posit genuine distinctions between the persons (ruling out modalism).

(T4) The conjunction of (T1) and (T2) seem to require that the consubstantiality relation between the persons be construed in terms of numerical identity rather than generic.

(T5) (T3) seems to require that each divine person is numerically distinct from each other.

(T6) Thus, any formulation of the doctrine of the Trinity that seeks to meet the requirements of (T1), (T2), and (T3) will be paradoxical, given natural intuitions about concepts employed.8

In other words, it would seem that if we apply certain deductions based on certain metaphysical truths (about identity, for example), then we have: A is identical to C, and B is identical to C, then A is identical to B. The inconsistencies appear to arise due to the metaphysical affirmations made, then. Thus, Anderson maintains that though it is not impossible to develop perfectly consistent doctrines of the Trinity (ones that avoid any hint of paradoxicality); the problem comes when adherence to orthodoxy is also attempted. Therefore, it is not enough to formulate just a coherent doctrine of the Trinity. One must also try to achieve orthodoxy. If the latter is meant to be achieved, paradoxicality results. As Anderson states, “If these conclusions are correct, then trinitarians face an awkward dilemma: an apparent choice between orthodoxy and clear logical consistency.” Thus, though few deny that there are interpretations that are free from logical difficulty, being free from that, as well as being free from theological difficulty, is another matter. More on the character of this paradox below.

Next, Anderson discusses numerous contemporary writers who have maintained that interpretations of the doctrine of the Trinity are available that are both logically consistent and also uphold orthodoxy. Anderson divides this section into four headings: (1) Modalistic interpretations, (2) Social Trinitarian interpretations, (3) Relative Identity interpretations, and (4) Paradoxical interpretations.

Theologians and philosophers discussed in this section are Karl Barth9 and Karl Rahner as representatives of (1). Cornelius Plantinga, Richard Swinburne, and David Brown are representatives of (2). A. P. Martinich, Michael Rae and Jeffery Brower are representatives of (3). Stephen Davis, David Coffey, and William Hill are representatives of (4). Just as with the historical survey, Anderson clearly has done his homework and understands the field with great comprehension. I believe he successfully shows that all attempts to avoid paradox either fail to uphold orthodoxy, or do not alleviate the logical worries. Though it would go far beyond the scope of this review to discuss all his points (and then why would you buy the book!), I will look at the attempt Anderson notes is the best to date at trying to fulfill the demands of logical consistency as well as orthodoxy: that of Brower and Rae.

The relative identity thesis recognizes that the identity spoken of in the Trinitarian doctrine leads to all the problems. So, rather than an absolute identity, they relativize the identity (e.g., the Father is the same God as the Son yet the Father is not the same person as the Son). Rae and Brower recognize the major difficulties with proposed relative identity solutions. For example, what they call the ‘pure’ relative identity thesis is too strong in that it flat-out denies any absolute identity. The impure strategy offers an incomplete solution because it remains neutral on the question of whether absolute identity exists; it thus cannot rule out polytheism. Without developing and providing a plausible metaphysical account of the identity relations within the Trinity, the relative identity theorists cannot provide a theory by which they can meaningfully say “there is only one divine being” (contra polytheism) as well as “the Father is distinct from the Son” (contra modalism).

Rae and Brower believe they can meet just such a burden. They attempt to do so by bringing up the philosophical problem of material constitution.10 Anderson summarizes the problem:

Consider a statue made of bronze; let the statue and the lump of bronze be labeled ‘Athena’ and ‘Lump,’ respectively. Intuitively, we would only want to say that there is only one object here, for Athena is just the same thing as the Lump. Thus, if one wanted to purchase the statue, one would not expect to have to pay twice: once for the statue and again for the lump of bronze. According to Leibniz’s Law, however, if A is identical to B then whatever is true of A must also be true of B; in which case, if Athena could be destroyed by melting the Lump could also be destroyed by melting. Yet common sense also tells us that after melting, it would be correct to say Athena no longer exists but that the Lump remains. Is Athena identical to the Lump or not?11
Brower and Rea take an Aristotelian approach to solving the problem. The relation that exists, on this account, is unique and irreducible. Rather than an absolute identity, they favor what they call “accidental sameness without identity.” So, Athena is not absolutely identical to the Lump, she stands in a relation of accidental sameness. This employs Aristotle’s form-matter promulgations. We have sameness because one may properly say that Athena and Lump are the same, yet this sameness is accidental because Lump might have existed without being chiseled into Athena. Thus, the form is contingently instantiated by its matter. So when one counts, much importance lies on what we take ourselves to be counting. This relativizing allows us to count accidentally same ‘objects’ as one. Brower and Rae take it that something akin to this account holds between the divine persons and the divine essence.

Anderson is quick to pointout that the relation cannot be accidental sameness itself, for then it would follow that the divine persons exist contingently! So the relation should be thought of as something like “essential sameness without identity.” This is a species of “numerical sameness without identity” (cf. above). And so Brower and Rae take it that if such a relation exists between A and B, then A and B are distinct and yet can be treated as numerically one.

Anderson admits to the ingenuity and promising nature of this account, yet finds four problems: First, the Aristotelian solution is controversial in its own right. Invoking a sui generis metaphysical distinction (numerical sameness without distinction) “smacks of explaining a mystery with a mystery.”12 Second, Anderson wonders how much discontinuity can be invoked before the model loses all plausibility. Third, what about the multi-personality of God? Perhaps one can grant that the statue can be distinct from the lump, but can three statues be distinct from the same lump and distinct from one another while being one and the same material object?13 That seems paradoxical, and so paradox has not been avoided. Fourth, what is the proper name ‘God’ or ‘Yahweh’ supposed to refer to? When Christians say “God is good,” for example, they do not mean just one of the persons, but if ‘God’ is to refer to an impersonal essence, what sense does it make to say that non-personal, immaterial “stuff” can be omni-benevolent, omnipotent, and omniscient?

Anderson spends little time on those theologians of paradox. He mentions Davis and says Davis’ defense of rationally believing the paradoxical doctrine of the Trinity is, though less developed, similar in direction to the strategy Anderson employs later. Coffey has done valuable work and ends up saying that we can state that our conception of the Trinity is not contradictory, but we are at a loss to specify how it is not contradictory. We can avoid literal contradictions by introducing distinctions; we have no grasp on how this all cashes out in metaphysical terms. Hill roughly follows similar lines, pointing out that it has been the rationalists who desire to avoid paradoxical contradiction. As Anderson quotes Martinich: “[I]f faced with the alternatives of being a heretic and asserting a contradiction, the rational person will always choose heresy and trust himself to the mercy of God.” (This quote expresses Anderson’s point about the problem of holding both consistency and orthodoxy. Martinich is expressing how heresies gain rational support in this quote, he then goes on to offer a solution to the trinitarian dilemma. As noted, Anderson does not agree with Martinich’s solution. My purpose in using this quote is that it expresses an interesting truth about the hermeneutical use of the idol of rationalism.)

Anderson ends this section with what I take to be a powerful point. He says that if the doctrine of the Trinity were placed in the dock and charged with the crime of apparent contradiction, nearly all the writers he cited could be called as witnesses for the prosecution, regardless of their views on the guilt or innocence of the defendant. By setting the two parties into groups (those who are within the bounds of orthodoxy and those who are not), “an astute council for the prosecution could play off the testimonies of the former against the testimonies of the latter in order to secure a conviction.” The first group is charged with paradox, the second with violating orthodoxy. Thus, it looks as if the twin goals of complete logical consistency and complete orthodoxy cannot be had at the same time. Anderson powerfully and conclusively demonstrates this in chapter 2.

In chapter 3, Anderson does the same thing with the doctrine of the Incarnation. He looks at the history of the debate, the development of the doctrine, the concerns on all sides, and the bounds of orthodoxy the creeds constrain formulations of the doctrine of the Incarnation to follow. He also looks at various contemporary attempts to put forth a logically unproblematic, as well as honoring to orthodoxy, doctrine of the Incarnation. This section is just as detailed and well-researched as the previous. Anderson makes his case. For many reasons I will not go over his case. For one, it leaves more interest to the reader. Another reason is in interest of making this review as detailed, while as short, as possible. The last reason is that,

If the doctrine of the Trinity is inherently paradoxical, as I argued earlier, then the doctrine of the incarnation necessarily inherits that paradoxicality. Here is the argument: if the Son assumed a human nature, and the Son is God, then God assumed a human nature; but if the Father did not assume a human nature, and the Father is God, then God did not assume a human nature; therefore, God both did and did not assume a human nature.14

And so, if Anderson has demonstrated the paradoxicality of the Trinity, and I believe he has, then I am free to skip reviewing chapter 3. Some may not agree, and so the incredulous can read chapter 3 for themselves. The paradoxicality of the Incarnation, boiled down to its essentials, has to do with this: one person having both a divine and human nature.

Thus having established the paradoxicality of the doctrine of the Trinity and the doctrine of the Incarnation,15 we are in a position to answer the first of the three aspects Anderson analyzes regarding paradox in Christian theology–its presence. Does paradox have a presence in orthodox Christian theology? Yes, without a doubt. We can now move on and look at the various responses to paradox.

Responding to Paradox (Ch. 4)

Having demonstrated the presence of paradox in Christian theology, Anderson has demonstrated that orthodox Christians hold to claims which; taken in conjunction, appear to be logically inconsistent. Is the Christian therefore irrational in believing these conjunctions? This task Anderson takes up in Part 2. He offers a particular solution to dealing with paradoxical doctrines. In order to “prime the pump” for a sympathetic reading of his take on the matter, Anderson looks at a few ways of dealing with paradox offered by others (either explicitly or what could be offered in theory). He states that he does not attempt any decisive refutations of these views, but merely shows that they are inadequate for handling paradox in Christian theology. Anderson hopes that his proposal will appear less problematic than its competitors will. With this modest goal in mind, Anderson makes some helpful comments on “logical and linguistic factors which give rise to the charge of irrationality in the first place.” This allows him to categorize the various approaches to dealing with paradox to (i) the factor(s) each approach attempts to modify or eliminate in order to resolve paradox, and (ii) how the proposed modification of elimination proceeds.

Anderson asks about which “factors give rise to a genuine contradiction?”. A popular candidate for what entails a ‘contradiction’ is that some proposition is both true and false, or some proposition and its negation are both true. If someone S asserts this, then S is guilty of irrationality. Anderson does not think the propositions (sets of statements that make up the doctrines) treated in chapters 2 and 3 are guilty of so blatant a contradiction as, for example: “There is only one God and there is not only one God.”

So what kind of apparent contradiction is Anderson talking about here? He proposes to distinguish three important types: (i) an explicit contradiction arises when one both affirms a proposition and its logical negation. (ii) A formal contradiction arises when one affirms a set of propositions are not explicitly contradictory, but from which an explicit contradiction may be deduced. Finally, (iii) an implicit contradiction occurs if, and only if, one adds one or more necessary truths to a set of statements to derive a contradiction.

Anderson claims that, “the type of contradiction, apparent or otherwise, involved in paradoxical Christian doctrines is best characterized as an implicit contradiction. That is, certain statements in the doctrines seem to imply claims that explicitly contradict other statements of Christian doctrine.” Hence, one of the factors that invite a charge of irrationality is logic. Laws of logic are considered canons of rationality. To assert something, which violates those laws, is to delve into irrationality. However, not only does the presence of contradiction depend on logic alone, it also depends on the meaning of the statements from which the alleged contradiction is derived. One meaning may invite a contradiction “whilst”(!) another may not.

Therefore, from the above “paradox-generating” factors, we can see two approaches to dealing with paradox: (i) revisions of the laws of logic, and (ii) dealing with the meaning of the doctrinal statements. There are two sub-groups for both (i) and (ii): (ia) strategies trying to avoid contradiction, and (ib) strategies aiming to allow contradiction. With (ii) we have: (iia) strategies aiming to revise doctrine, and (iib) strategies aiming at retaining doctrine while claiming that the doctrines don’t contain anything that necessarily implies contradiction.

This is one of the tougher chapters in the book. A little familiarity with the philosophy of logic, and some of the contemporary debates in that field, makes for better comprehension, though this is not necessary for reading and grasping the chapter (or for understanding Anderson's later model (Ch.6), for that matter). Anderson does a good job at distilling all the intricacies and boiling down the important aspects which bear on the nature of his thesis.

Before discussing the above four strategies, Anderson looks at theological anti-realism. This is the thesis that our theological talk does not refer to anything that actually exists in the talk-independent world. That is, any entity existing independently of our thought or language. Anderson successfully rebuts this view in so far as it might be thought useful for dealing with paradox.

After dealing with theological anti-realism, Anderson deals with (i) and (ii) in their various subsets. He deals with anti-deductivism and dialetheism in (ia) and (ib) respectively. He deals with doctrinal revisionism with respect to (iia) and semantic minimalism and complementarianism with respect to (iib). Time does not permit a discussion of this interesting subject matter, or Anderson’s critiques of all these strategies. Having said that, I will mention a few points in passing.

With respect to (i), both sub-groups seem to grate against some of our most powerful intuitions such that to reject the universality of some of the logical laws they require us to deny some pretty basic beliefs we hold.16 Phenomenologically speaking, if we can reject beliefs of this status, what can’t we reject? Propositions with the same phenomenologically strong credentials could also be cast into doubt. Moreover, with respect to dialetheism specifically, whatever its strengths are, and however good the arguments in favor of it are, and however difficult it is to refute, this approach does not help us in resolving paradoxes. If real contradictions could be true, then the desire to preserve orthodox interpretations is gone. Indeed, one could no longer object to heterodox statements. Thus, dialetheism makes preserving orthodoxy irrelevant.

Anderson critiques (iia) from many angles, including an argument that the creeds are what identify the Christian community. To revise such basic identifying doctrines results in a kind of identity crisis. If we can do this, should doctrine play any role at all in identifying what it is to be a Christian? Anderson also argues that the creeds enjoy a certain authority, which is derived, on Protestant lines, from Scripture. To undermine the creeds is to undermine their authority, and hence their derivative authority also. However, Anderson notes, this strategy may be appealing for those concerned more with logical orthodoxy than theological orthodoxy.

Under (iib) semantic minimalism is the view that creedal doctrinal statements are excessively meaningful. One need not go to the excess. If we just stick with the minimum required, we can escape contradiction. Anderson argues, among other things, that semantic minimalism is not minimal enough to avoid paradox. Lastly, complementarianism takes its cue from one popular way of resolving scientific paradoxes (e.g., the light particle-wave paradox made famous by the double-slit experiment). This view does not advocate logical revision or doctrinal revision. It focuses on the meaning of the statements. This view offers the greatest promise and is somewhat along the lines Anderson will proceed in later chapters, but it ultimately has difficulties, which make it less than a satisfactory resolution. This section is more detailed and subtle and so I will refrain from discussing it further in this review. Suffice it to say, Anderson’s model will neither advocate logical revision or doctrinal revision, but focuses on meaning. It does so in a way that accommodates the doctrines of the Trinity and the Incarnation (as well as some other doctrines traditionally thought to be paradoxical) without sacrificing the orthodoxy or rationality of these traditional Christian beliefs.

Overall, I think Anderson succeeds in this chapter. Anderson shows that, for those with an eye towards maintaining rationality and orthodoxy, the popular strategies for dealing with theological paradox are unsatisfactory to that aim.

Part 2-The Propriety of Paradox
Warranted Christian Doctrines (Ch. 5)

Similar to chapter 4, chapter 5 requires some familiarity with some contemporary epistemological issues; and specifically, the work of Alvin Plantinga (particularly his Warrant trilogy), in order to achieve maximum comprehension.17 But the same point above applies here: Anderson does a good job distilling those parts necessary to make his case and presents them to the reader in clear and accessible fashion. Thus, with some effort by the reader insufficiently familiar with Plantinga for maximum comprehension of this chapter, Anderson’s main points can be grasped and appreciated.

Anderson’s main goal in this chapter is to “consider whether any Christian doctrine, paradoxical or otherwise, invite or enjoy rational assent in the first place-and if so in what way and to what degree.” So the first question must be whether the component claims that constitute paradoxical doctrines, are individually worthy of assent. And just what is this positive quality Christians want to ascribe to their doctrines? It is an epistemic property, as opposed to psychological, prudential, or other non-epistemic properties, that invite belief. Anderson now points out that the epistemic property he will focus on is warrant.18 He gives three reasons for this: (i) his conclusions will be more interesting by claiming that they are warranted as opposed to some weaker epistemological properties (e.g., plausibility, rational entitlement, etc.,), and even some stronger ones (i.e., justification); (ii) there are important types of rationality necessary for warrant (thus if a belief is warranted, it is rational in these ways, and hence if belief in paradoxical doctrines were warranted, they would be rational too); and (iii) warrant seems to have been the quality most Christians have wanted to ascribe to their beliefs, from the earliest of times.

Thus by concentrating on warranted Christian doctrines, Anderson will be focusing on whether some Christian doctrines rise to the level of knowledge. Answering this question sets up the answer to the main question of the book: whether paradoxical doctrines can be warranted, and thus rationally believed. Is the paradoxicality of a doctrine a strike against its warrantability? Does it make its warrantability less probable? These are some of the issues and questions Anderson pursues.

Anderson wishes to place his theory of warrant in substantial agreement with Plantinga’s model. He begins by setting up Plantinga’s model by discussing some epistemological history and surveying various key positions in this debate (e.g., externalism/internalism, deontologist justification, coherentism, foundationalism, Gettier counter-examples, etc.). Rather than discuss Anderson’s analysis of the situation, I will simply quote a relevant part from his review of Plantinga’s book, Warranted Christian Belief:

Ten years later, Plantinga turned his analytical skills to an analysis of knowledge in general in the first two volumes of his Warrant trilogy. Defining ‘warrant’ as that which (in sufficient measure) distinguishes knowledge from mere true belief, he argued that none of the extant contemporary theories of knowledge — varieties of classical deontologism, internalism, coherentism, and reliabilism — offered a satisfactory analysis of warrant. On the basis of various imaginative counterexamples, Plantinga maintained that in each case the conception of ‘justification’ or ‘warrant’ propounded was either not necessary for knowledge, or not sufficient for knowledge, or both. Plantinga proceeded to argue that these counterexamples (as well as the classic ‘Gettier’ cases) teach us that what is lacking in current analyses of knowledge is the notion of proper function, i.e. of beliefs being formed by noetic processes functioning in the manner in which they were ‘designed’ (whether by God or by evolution) to function. In Warrant and Proper Function, Plantinga fleshed out in more detail his basic contention that “a belief has warrant if and only if it is produced by cognitive faculties functionally properly in a congenial epistemic environment according to a design plan successfully aimed at the production of true belief” by addressing various objections, making some important refinements, and suggesting how his analysis of warrant might cash out in terms of the various types of knowledge we possess (a priori, perceptual, inductive, etc.).19

So, a belief is warranted if, and only if, it is produced by cognitive faculties functioning properly in a congenial epistemic environment according to a design plan successfully aimed at the production of true belief. We can also add this constraint on warrant: “that a belief B is warranted for [someone] S only if on adequate reflection S would not be aware of any sufficient reason to reject B,” adds Anderson in PCT. Call this the negative internal constraint.

So, basically, what we have here is an account of what is necessary and sufficient for a belief to be warranted. Why should your belief be warranted if the belief is produced by cognitive malfunction (say, a brain lesion producing the belief that you have a brain lesion, this belief may be reliably produced, even true, but it is due to malfunction/dysfunction, luck, or accident)? Why should it be warranted if your epistemic environment is malfunctioning (say cosmic rays emanating from Alpha Centauri causing everything to appear red)? Why should your beliefs be warranted if the design plan (the purpose or function of your cognitive faculties) is not aimed at producing true beliefs (say, your beliefs are only aimed at evolutionary survival, not truth20)? And why should they be warranted if you were, on adequate reflection, aware of the plan of the Alpha Centaurians to make all of earth’s objects appear red by means of directing their red-illuminating rays towards earth? Therefore, in my estimation, Anderson presents an accurate account of what it is for a belief to be warranted.

At this point, Anderson moves on to the question of warranted theistic belief, and then warranted Christian belief. (Again, I refer the reader to his review of Plantinga on this section for a fuller account of how, on Plantinga’s account of things, a Christian’s beliefs may have warrant (see n.9). There are three previous books (roughly totaling 1,000 pages) that provide the background for this account (as well as hundreds of pages of articles) from which Anderson is drawing here.) Anderson, following Plantinga, presents a model for warranted theistic belief, and then an extended model for warranted Christian belief. I agree substantially with Anderson (and thus Plantinga!) here, but it would be far beyond the scope of this review to trace all the ins and outs of Anderson’s discussion.

Suffice it to say, on the Christian story, God has made us in his image and designed us to know truths about the world, and truths about Him and His requirements of us in relation to Him and our fellow man. He also designed the world, and makes sure it is a properly functioning environment. Thus the Christian, on this story, can be externally and internally rational, and warranted, in her theistic beliefs. It should be plain to see, then, that objections to the rationality of Christian belief (de jure objections) only land by assuming the fact of the matter (de facto) is not the case. That is, the objection to the rationality and warrant of the Christian belief only hits if the Christian view of reality is false. Therefore, you cannot object to the rationality of the Christian’s beliefs without assuming that God does not exist, and that the Christian story of things is false. In his review of Warranted Christian Belief, Anderson sums this up well:

Plantinga happily notes in agreement that if Christian theism is false, then Christian belief is most probably unwarranted. But on the basis of the [Aquinas/Calvin] model, he further notes that if Christian theism is true, then Christian belief is most probably warranted. Plantinga takes it that the failure of the Freudian and Marxian objections are merely illustrations of a more general principle, one of the most significant theses of WCB: that the de jure question regarding Christian belief cannot be answered independently of the de facto question. In other words, stances such as the following are no longer tenable:

‘Well, I certainly don’t know whether theistic belief is true — who could know a thing like that? — but I do know this: it is irrational, or unjustified, or not rationally justified, or contrary to reason or intellectually irresponsible or . . .’ (WCB, p. 191)

Thus, the Christian’s theistic beliefs arise from cognitive faculties functionally properly in a congenial epistemic environment according to a design plan successfully aimed at the production of true belief. (Plantinga discusses specific defeaters, and Anderson discusses defeaters pertinent to his thesis later). These beliefs are then warranted given the truth of Christian theism. This cognitive faculty is called the sensus divinitatis.22 But given disagreement about God, examples of unbelief, the worry is that perhaps the sensus divinitatis does not lead to warranted beliefs with a degree of warrant sufficient for knowledge. But Anderson notes, following Plantinga, that the Christian can bring sin into the picture. The noetic effects of sin can account for the above discrepancies. Thus, the Christian story has more than enough muscle to do all the heavy lifting required of it.

This move clears the way for Anderson to discuss Plantinga’s discussion of warranted Christian belief. This is more than a belief in God, and some of His attributes and demands. This is belief in some of the revealed propositions in the Holy Bible. Beliefs about Jesus, soteriology, and other essentials called “the great things of the gospel” (Plantinga’s terminology). If the above model is called the Aquinas/Calvin (A/C) model, then what we need here is an extended A/C model. Thus, we keep all the above comments about being made in God’s image, having a cognitive faculty that produces knowledge of God in a basic way, etc., and we now add other elements to it to account for some specific Christian beliefs known only by way of revelation.23

Since the sensus divinitatis has been marred or damaged by the fall, God has graciously provided us a remedy. The remedy is made through the person and work of Jesus Christ. God has also graciously given us a revelation, and His Holy Spirit to produce faith in us. This faith is a cognitive affair, but also more than that. It involves the will and affections, according to Plantinga. This “internal invitation” (though a Calvinist might want to use stronger terminology than Plantinga does in saying, “invitation”!) of the Holy Spirit is a source of belief for us, a cognitive process that produces beliefs. This meets the conditions for warrant laid out above. This whole process is part of a properly functioning design plan for the renewed Christian. The epistemic environment is certainly congenial, and the aim of the faculty is the production of true beliefs, says Plantinga.

Now Anderson wants to look at all of this. Here is where things get more interesting. Anderson cites Plantinga’s three main elements involved in a person’s coming to know “the great things of the gospel.” Plantinga means to follow Jonathan Edwards here: (i) God gives us the Scriptures to supply the content of belief; (ii) God gives us the Holy Spirit to produce belief, and (iii) this outcome of the operation of the Spirit results in the production of faith. Notice that (i) and (ii) are external to the believer while (iii) is internal. Anderson points out that on Plantinga’s model the Scriptures are needed only to provide the propositional content of belief and the occasion of coming to believe that content. So upon reading the Bible, a person may entertain one of the propositions (the “great things” propositions), and then the Holy Spirit affects the mind to induce belief in that proposition. But Anderson notes that this is inadequate since many Christians have taken the Bible not only to serve as a mere conduit for true propositions, but it is also divine testimony to their truth (cf. Jn.5:39-40, 21:24; Rom. 3:21; Isa. 8:20, etc.). So on Plantinga’s model, even if Scripture is testimony from God, it plays “no role”24 in the formation of Christian belief.

Anderson supplies evidence from Plantinga’s writing to this effect.25 Anderson then notes that Plantinga’s above position is interesting because he seems to switch gears later, affirming that the Bible is indeed testimony, as if this were an important feature.26 Anderson claims that the picture Plantinga later paints is “substantially different” from his early account of the role of Scripture. Scripture provides more than mere propositional content. This testimony is not enough by itself, though. The Spirit is required to encourage or enable belief. The main point here is that Scripture does play some role in one coming to know the truth. So how should this ambiguity within Plantinga be resolved?

First things first. Anderson takes issue with Plantinga’s explication of the operation of the Spirit as a process. Plantinga does this because he does not want to say that the Holy Spirit is a function; that is, the Spirit is not part of the cognitive apparatus of the Christian. The internal invitation (alternatively, instigation) of the Holy Spirit cannot be thought of as either a faculty or the operation of a faculty. So, to Plantinga, it seems to make more sense to speak of the operation as a process of belief-formation. At this point Anderson brings up a problem. Though it makes sense to speak of a faculty as “‘functioning properly,’ it is rather less clear that it makes sense to speak of a process as ‘functioning properly.’” Therefore, processes are not things that we can say function properly. If this is true, then they also cannot malfunction. Moreover, it looks like the conditions of warrant do not apply here.

Anderson offers a couple of suggestions as to how Plantinga might resolve this problem.27 The first is to treat both the cognitive apparatus of the believer and the work of the Holy Spirit as a system of belief-formation. This system is designed to bring about the requisite belief-formation needed in order for the believer to have warranted beliefs. A system may not be a faculty, but it can be something that functions properly. Anderson thinks this can address the worry of those like Richard Gale, though it does not address the ambiguity on Plantinga’s account of the role of Scripture and “tends to reinforce the notion that the testimonial value of Scripture does not contribute to the warrant of Christian beliefs.”28 Thus Anderson has another, more preferable, solution in mind. According to Anderson, the model should involve those cognitive faculties designed for the formation of beliefs by testimony. Earlier Plantinga had denied that the beliefs in the great things of the gospel did not come by way of the ordinary function of cognitive faculties originally created; they come by way of the Holy Spirit. But, Anderson brings back in this natural element. Since Plantinga thinks there is a place for testimony, it is natural to involve those cognitive faculties designed for the formation of belief based on testimony.

At this juncture Anderson dovetails back into full agreement with Plantinga and his defense of knowledge by testimony given in Warrant and Proper Function. Following Thomas Reid, the principle of credulity in the context of testimony from our fellows is a “gift of nature.” Reid finds that if he had doubted his parents’ testimony from an early age, he should not be here today (yesterday, for us!). The majority of our beliefs come by way of testimony. It is an extremely important part of our cognitive equipment. It is a key feature of our design plan that we form beliefs based on the testimony of others. “[T]hus according to a proper function account, testimonial beliefs can be warranted to a degree sufficient for knowledge.”29

How would these faculties play a role here in Anderson’s (extended) extended A/C model? He puts forward various options. For example, the Holy Spirit could work to strengthen the beliefs so they have warrant to a degree sufficient for knowledge, or counteracting the effects of sin in the life of the believer. Renewing their minds. Taking away unwarranted incredulity, etc. The Holy Spirit could aid in bringing about a more congenial epistemic environment. Third, the Holy Spirit could re-start, in a sense, the original design plan which included taking God’s testimony in a credulous way. (In fact, this was the natural path for man, much like it is very natural for our children to form warranted beliefs based off their parents’ testimony.)30 This faculty would be repaired. On the other hand, the Holy Spirit need not repair or re-start anything. He could adapt or modify already existing faculties, thus the believer would operate according to a new design plan. He does not settle on any one of these, but says perhaps that a combination of all three would be best. Nevertheless, Anderson clearly shows that these three are plausible, and that if it (or something close) were true, we would meet the conditions for warrant.

Lastly, Anderson asks why such an elaborate scheme should be needed for warrant. Could not the Christian’s beliefs be warranted through normal historical investigation into the reliability of the biblical documents? Anderson follows Plantinga’s response (as well as many others; Alston, for example). Such methods are “beyond the grasp of the average believer.” In addition, “a predominately inferential strategy could never deliver the goods in any case.” This is due to the “principle of dwindling probabilities.” Thus, a model where beliefs are formed in a basic way is to be preferred.31

Anderson notes next that Plantinga’s extended A/C model does not cover many specific Christian beliefs, two of which are crucial for Anderson’s purposes—the Trinity and the Incarnation. If we mean the “Nicene-Constantinopolitan trinitarian statements and the Chalcedonian Christological statements” as our standards of orthodoxy, then Plantinga’s model does not cover them. This is because those doctrines (and others) are not explicitly stated in the Bible. “If the text of Scripture were such that belief in the doctrines of the Trinity and the Incarnation could arise just as the extended A/C model suggests, there would have been no need for such creeds to be formulated in the first place.” Thus, Plantinga’s position does not go far enough. It does not tell us how Christian doctrines can also be warranted. So Anderson attempts to fill this lacuna.

He begins by looking at four major views as to the epistemological foundations of doctrine. One is selected as providing a basis for a model for warranted Christian doctrine. Though there are many interesting discussions one could get into regarding views on the nature of doctrine, Anderson’s goal is to look at the epistemic question. He seeks to answer two such questions: (i) what is the source of the propositions expressed in the doctrines, and (ii) how could anyone come to know those propositions are true? Anderson points out that the majorities have viewed doctrine as grounded in divine revelation. However, many have diverged after this basic agreement. “[T]he disagreement concerns the precise character, location, and appropriation of this divine revelation.” The four positions Anderson looks at are (1) the ‘Reformed’ perspective; (2) the ‘Catholic’ perspective; (3) the ‘Neo-Orthodox’ perspective; and (4) the ‘Liberal perspective.’ After giving his reasons why, Anderson sides with the ‘Reformed’ perspective. But he is quick to point out that it is still possible that alternative models for warranted Christian beliefs could be developed on the ‘Catholic’ perspective. However, the Catholic perspective is less simple than the ‘Reformed’ perspective, and their views on private interpretation would not sit well with Plantinga’s model of warranted beliefs formed in a basic way on reading the text alone independently of ecclesiastical teaching.

Anderson wants to defend the ‘Reformed’ perspective. On this perspective, “Christian doctrines are only warranted insofar as they are grounded on God’s special revelation through the Holy Spirit speaking through the Scriptures alone.” How would this perspective cash out in terms of warranted belief in Christian doctrine? It should be recognized, first off, that if doctrines such as the Trinity and the Incarnation are warranted by being affirmed (explicitly or implicitly) by Scripture, “then a person must be warranted in believing that whatever is affirmed by Scripture is true in order to be warranted in believing the doctrines themselves.” This belief could be warranted in numerous ways.

One way is by ordinary teaching and testimony. But warranted beliefs based on testimony are derivative, thus it cannot be that all warranted Christian beliefs arise by way of (at least) human testimony. Other ways could come into play too. Another way is the internal instigation of the Holy Spirit. These beliefs would come, perhaps over time, by reading the Bible and its various tota divinitus inspirata passages. Another view would be that a person comes to believe in the total inspiration of the Bible by inductive inference. This, coupled with a design plan aimed at producing true beliefs based on inductive inferences, could be warranted to a high degree. Anderson notes a few other possibilities too. Spiritual perception with attendant phenomenology arising from reading the text of Scripture producing beliefs that are warranted in a basic way, and thus would not be an inference from that phenomenology, is another possible view. Anderson’s basic point is not to argue which one of these is true, it may be a mixture. The main point is that there seems to be no reason to doubt that, given the truth of Christian theism, the belief that Scripture is God’s word could be warranted on an extended-extended A/C model, given Plantinga’s claims about warrant and proper function, et al.

Therefore, scriptural teaching can function as a basis for warranted belief in Christian doctrine. Now Anderson wants to explore how individual Christians can be warranted in believing the statements of the Trinity and the Incarnation found in the respective orthodox creeds. Anderson proposes four paradigmatic ways in which doctrinal beliefs can be warranted. The boundaries are not sharp, and some Christians may exhibit elements of more than one way:

[1] The first way provides the epistemic basis for the other three. A Christian studies the text of Scripture, accurately interprets the texts, and comes to the warranted conclusion that some set of propositions is in harmony with some Christian doctrine. The Holy Spirit is no doubt involved in this process. The believer also makes use of good exegetical tools, e.g., background study of the culture, history, writings, and other exegetical desiderata, makes use of commentaries, logic textbooks, etc., in order to better interpret the text of Scripture. On this first case, a Christian is warranted in believing Christian doctrine because he has directly studied the text of Scripture in a scholarly way, and he can explain and defend his reasoning to others.

[2] The second paradigm case is this: A person with a warranted belief in the inspiration of the Bible is presented with the Scriptural basis for a doctrine by someone falling under heading [1].This person reflects on the teaching and has a grasp of the reasoning involved.32

[3] This case is simpler still, and is still dependent on [1]. A Christian with a warranted belief in the inspiration of Scripture is warranted based on trustworthy testimony by a parent, professor, or church leader that a certain doctrine is taught in Scripture, and so infers that this doctrine is true. Knowledge by testimony is transitive. That is, if the testifier here knows what he testifies (say that his knowledge transferred over from someone in [1] (or [2]), and [1] came from studying the testimony of an infallible God), and then the testifiee knows what has been testified to. The warrant transfers. The Holy Spirit could also play a role, removing sinful tendencies to doubt, and strengthening beliefs.33

[4] This is the simplest case. One Christian accepts a doctrine as true purely based on reliable testimony from another Christian with a warranted belief in that same doctrine. This belief is not held by inference from other beliefs, but is warranted nonetheless. Children learning from their parents fit this bill.34

On these four ways taken together, we can account for the principle way in which warranted beliefs in Christian doctrine are formed, on the assumption that belief in inspiration is also warranted (which Anderson has argued that it is). This view also includes the warrant strengthening activity of the Holy Spirit. Thus, there really is not anything novel in this approach, says Anderson. Thus, if fundamental Christian beliefs can be warranted, so can beliefs in Christian doctrine based on this extended-extended A/C model.

The last topic Anderson discusses in this chapter is that of epistemic defeaters. Recall that the notion of defeat plays into Plantinga’s account of warrant. Some objections to Plantinga’s warrant model revolved around this concept. So, granting the model, perhaps some Christians could be warranted in their beliefs, but most probably would not be. Given all the objections to Christian belief, how could the average believer hope to maintain his belief’s status as warranted? Anderson notes that this can transfer over to warranted belief in Christian doctrine. And, given his claims about the paradoxicality of some of those doctrines, it might be precisely this kind of claim that offers a defeater for the warrant a belief in those doctrines might have. Anderson says that this is precisely how the issue should be framed:

[E]ven granting that some Christian doctrines can be rational or warranted, how can paradoxical doctrines be rational or warranted given their logical difficulties seem to function as obvious defeaters? As Plantinga himself notes, Frege might have been warranted to believe the conclusions of his Foundations of Arithmetic prior to Russell’s letter pointing out the contradiction at the heat of his axiomatic system, but once appraised of this fatal flaw the rational course of action was for him to acknowledge the defeater and abandon some of his prior beliefs.35

Anderson attempts to deal with this question in chapters 6 and 7, but before he does, he finds it necessary to say some things about different kinds of defeaters, and to establish which kind of defeater paradox is alleged to be.

The first kind of defeater is a warrant defeater. This defeater removes, precludes, or substantially reduces the warrant of a belief. Defeaters sometimes arise as other beliefs, but not always. For example, the environment can be a defeater for a belief, and this is external to a person. A dog that looks like a sheep can be a defeater for a person‘s belief that he saw an actual sheep. Anderson offers other types of examples and thought experiments to show the existence of non-doxastic defeaters. If paradox were a warrant defeater, paradox would not be a defeater in these ways. The defeat paradox offers would be doxastic in nature. Following Plantinga, beliefs that defeat the warrant of other beliefs are rationality defeaters. Thus Plantinga,

[G]iven belief in the defeating proposition, you can retain belief in the defeated proposition only at the cost of irrationality. . . . A defeater for a belief b, then, is another belief d such that, given my noetic structure, I cannot rationally hold b, given that I believe d”.36

Now, rationality defeaters can either be rebutting or undercutting. A rebutting defeater is a belief that is inconsistent with another belief, giving a reason to think the latter belief is false. An undercutting defeater reveals the grounds or reasons supporting some other belief to be inadequate. In situations where these kinds of beliefs arise, rationality demands the person should give up some of their beliefs. In some cases, one can reject the defeater instead of the defeatee.

Our design plan, our proper function account of warrant, will adjudicate this according to those relevant parts of the design plan dealing with belief production and revision (in conjunction with the other parts, cognitive environment, etc.), provided this defeater system is aimed at the production or maintenance of true belief (rather than for prudential, or non-alethic reasons). What proper function requires in most cases is intuitively clear and uncontroversial. The defeater-system of the design plan “places a premium on those sorts of qualities, such as coherence and simplicity, which normally accompany true beliefs.” In some cases, what the design plan requires is not obvious. Defeaters may be irrational beliefs, but they would not be warranted since they may be internally rational, they would not be externally rational.37 Warrant requires both external and internal rationality. Anderson then states that the kind of defeater paradox (allegedly) it involves a rationality defeater. It would be categorized as a rebutting defeater rather than an undercutting one. The defeater suggests that the doctrines are false, not simply that the grounds for believing them have been removed, yet the paradoxical doctrines might still be true.

Anderson states that the objection from paradox can be stated thus:

Even if the component claims of a paradoxical doctrine can be individually warranted for S according to a proper function epistemology, once S grants that the doctrines seem to involve a logical contradiction, even an implicit one, proper function rationality will require that S not believe that doctrine–indeed, S ought to thereafter consider it (at least partly) false.38

Anderson points out that if this defeater is to be of any epistemological interest, the potential defeater must itself be held rationally. Anderson, recall, has argued that the doctrines of the Trinity and the Incarnation, interpreted in an orthodox manner, are indeed paradoxical. Anderson will now take up the challenge of presenting a model for understanding theological paradox where the beliefs in the paradoxical doctrines remain warranted, and fail to provide a rationality defeater for belief in those doctrines themselves.

In conclusion, I would say that Anderson succeeds in presenting a model whereby a Christian can be warranted in her beliefs. It is especially refreshing to see him make use of the fertile ground of the epistemology of testimony in this chapter. Anderson is a non-reductionist about knowledge by testimony, and this position is involved in much debate over the epistemology of testimony. I believe this is a fruitful field to cultivate, but Christian epistemology is still in a somewhat infant stage here. No doubt Anderson’s argument will be challenged (by some) on precisely this ground. Anderson leaves a few problems for his readers to work on and develop on their own; I would only add that this is one such area that needs further development in light of how this position (i.e., the epistemology of testimony) has blown up recently, and therefore has been receiving more criticism as of late. Overall, I am excited at this turn of events in Christian epistemology towards knowledge by testimony. After all, this is what God’s Word to man is, the testimony of a divine person to his creation. I also believe that Christianity, with its doctrine of God, can answer many of the prima facie challenges to testimonial knowledge. Naysayers may not like some of these answers, but they would be involved having to deal with Plantinga’s de jure/de facto argument.39

The Model for the Rational Affirmation of Paradoxical Theology (Ch. 6)

Chapter 6 is really the heart of the book. In this chapter, we find our answers to the other two things Anderson analyzes with respect to paradox in Christian theology: its character and epistemic status. It has been kind of a dogma, says Anderson, that it is always wrong anytime and for anyone to believe apparently contradictory propositions. Anderson’s goal in this chapter is to challenge this view by presenting a model in terms of which Christians can be rational in believing apparently contradictory doctrines. In presenting his model, we will be able to answer what the character and epistemic status is of some Christian doctrines. It is a “model for the rational affirmation of paradoxical theology (hereafter, the RAPT model)” that Anderson presents. Anderson follows Plantinga in the definition of a model.

Basically, to give a model for some proposition P or some state of affairs S is simply to show how it could be that P or S is true or actual. The model itself will be another P or S, and it will be one that is clearly possible, and one that if it is true, so is the target proposition. These two points make the target proposition possible. For Anderson’s model, the target proposition or state of affairs is something like: “Christians who affirm paradoxical doctrines (such as the Trinity and the Incarnation) are normally warranted and rational in so doing.” There are four parallel claims Anderson makes about this model: (i) it is epistemically possible;40 (ii)there are no cogent objections that are not also objections to Christian doctrines as such; (iii) the model (or something close) actually describes how things actually stand with regard to paradoxical Christian doctrine; and (iv) the model is a member of a family of similar models, one member of which will be true if the central claims of Christianity are true and some essential Christian doctrines are indeed paradoxical.

An important point Anderson wishes to make is similar to Plantinga’s de jure/de facto distinction such that there are no objections to the warrant of Christian doctrines (objections based on their paradoxicality) that are independent of their truth. Anderson’s desiderata for the model are that it avoid denying or revising the law of non-contradiction or other classical rules of logic. It renders unnecessary the need to abandon orthodox formulations of the doctrine. It does not conflict with other traditional Christian doctrines, and even derives support from them. It spells out the situations in which it would be rational to affirm paradoxical doctrine. Finally, it upholds a robust distinction between heterodoxy and orthodoxy.

In chapter 4 recall that Anderson mentioned two main strategies for dealing with paradox. The broad strategies were revisions in laws of logic or focusing on the meanings of terms. Anderson is more in line with the latter. He has most in common with the complementarian approach, but recall he did not find it ultimately satisfactory. I also did not explain that view in this review, so one definitely needs to read his book in order to grasp and comprehend the full nature of his arguments and his model.

At this point Anderson makes a valuable distinction. It is the distinction between apparent and real contradiction. The distinction is not trivial since appearance does not entail reality or actuality. Since a genuine distinction can be made between apparent and real contradictions, then so can one between apparent—and—real contradiction and apparent—but—not—real contradiction. Anderson refers to the latter as a merely apparent contradiction (MAC). Thus if there are paradoxical Christian doctrines (and there are), and if the idea that they involve real contradictions is logically and theologically anathema (and it is), then the only acceptable conclusion to draw is that these paradoxes are to be construed as MACs.

Doctrinal MACs are to be construed as being apparently contradictory not for any temporal reasons, but in regards to semantics. One should try to smooth out any semantic problems if faced with apparent contradictions, if one can do so. But that a distinction could be articulated makes for a relevant distinction between MACs and genuine contradictions. MACS of this nature are accounted for by the presence of unarticulated equivocation. Thus, Anderson refers to paradoxical doctrines as MACRUEs (merely apparent contradiction resulting from unarticulated equivocation). This is the character of paradox in Christian theology, they are MACRUEs.

Anderson proceeds by means of various examples to establish that the notion of a MACRUE is unobjectionable in principle. I believe he succeeds in showing this. His examples make many subtle points required for the intelligibility for construing doctrinal paradoxes as MACRUEs. In many of the cases, Anderson’s subjects S are rational in taking some set of apparently contradictory claims as MACRUEs because the one who spoke them (probably) would not have made such a blatant contradiction, so we can rationally believe his statements were MACRUEs. Or, S may not even understand or be aware of phenomena that remove apparent contradictions, yet S can still be rational and warranted in believing the two apparently inconsistent claims, even implicit ones. MACRUEs need not be unintelligible or meaningless for S, and the MACRUEs arise for S in some cases because S’s concepts are not sufficiently discriminating to enable S to resolve the apparent contradictions. S may be operating with “coarse” or “unrefined” concepts, but these limitations do not affect S in her interaction with her family, friends, or anyone else in the course of her everyday life. Thus, Anderson demonstrates that MACRUEs are perfectly cogent notions and apply to many plausible scenarios.

Anderson suggests that all genuinely paradoxical doctrines be construed as MACRUEs. Anderson does not believe the doctrines, or the biblical text that supplies the data for the doctrines, are explicitly apparently contradictory; but in some cases may be expressed so as to express formal contradictions, and, generally speaking, the perceived contradiction will be implicit.41 We infer, from the Bible, the “neat, succinct set of claims which serves as formal statements of orthodox belief . . .” These inferences are not made in an epistemic vacuum. The theologian, or exegete, draws from a considerable amount of extra-biblical background knowledge and prior experience and categories of thought (e.g., natural intuitions about conceptual entailment, metaphysical necessities, etc.) in order to come up with these inferred doctrines.

Anderson points out some implications for construing paradoxical doctrines as MACRUEs. For example, suppose these following statements are included or implied from the doctrine of the Trinity:42

(T1) God is one divine being.
(T2) God is three divine beings.

The RAPT model says these two statements must involve an equivocation on one or more of the terms involved. Thus, the contradiction is merely apparent. It follows, then, that equivalent but formally consistent expressions of those statements can be constructed by articulating distinctions on one or more term. Here is how Anderson maps that out:

(T1A) God is1 one divine being.
(T2a) God is2 three divine beings.


(T1B) God is one divine1 being.
(T2B) God is three divine2 beings.


(T1C) God is one divine being1.
(T2C) God is three divine beings.2.

These are simple illustrations. More sophisticated and informative strategies, notes Anderson, are also available for augmenting our terminology so as to render the hidden equivocation more explicit. Anderson notes that we could reconfigure our notions of identity so as to accommodate more “exceptional” cases raised by the metaphysics of divine personhood. Anderson then works through what this might look like, showing that we could make some slight alterations or qualifications to our notion of identity, yet we would still be able to say all that we wanted to about “regular” identity statements. His comments on identity, and the “trinitarian calculus” he offers, are to be thought more as refinements of first-order logic with identity, tailor-made to handle the sensitivities of trinitarian metaphysics, rather than alternatives to it.43 Alternatively, we could, “perhaps more profitably,” reconfigure our notion of numerical oneness so as to deal with the oddities of trinitarian theology. This move is “inspired by supervaluationist solutions to paradoxes of vagueness . . .” Whatever procedure we take, the main point is simply that if we take the doctrine of the Trinity to be a MACRUE, then a formally consistent statement of the doctrine of the Trinity exists. This seems perfectly acceptable to me, and it should be a point that many are willing to grant.

But on which terms do we draw the appropriate distinctions? Which of the above interpretations captures the truth of God’s triunity? Anderson correctly points out that if we were warranted in taking the doctrines to be MACRUEs, then at least one of the formally consistent statements would be correct. This point is important because it alone is sufficient to deflect the anti-trinitarian charge of falsity due to logical inconsistency. “[I]f the doctrine is a MACRUE, then it must in the nature of the case be susceptible to formally consistent expressions.” And Anderson notes that we are not stuck just saying that one interpretation is possible, either. If there are philosophical or exegetical grounds for preferring one over the other(s), then we are free to move there.

Anderson shows how this model of demarcating terms might work in an uncontroversial way. Say you have a group of people; call them “flatlanders.” Flatlanders are dimensionally impoverished with respect to their conceptual scheme. They only see in two dimensions. But, say that the Flatlanders get a revelation from the trustworthy, honest, and ultra smart “Spacelanders.” Spacelanders see things three-dimensionally. The revelation the Flatlanders get is:

(S1) The object O is shaped triangularly.
(S2) The Object O is not shaped triangularly.

Say the object is a three dimensional cone (Anderson includes pictures at this point in order to better illustrate his point). (S1) and (S2) can be rendered formally consistent in at least two ways:

(S1A) The object O is shaped1 triangularly.
(S2A) The object O is shaped2 triangularly.


(S1B) The object O is shaped triangularly1.
(S2B) The object O is shaped triangularly2.

The meanings of the terms are refined so as to remove inconsistency, e.g., “shaped1” conveys something like “horizontally-shaped;” “shaped2” means something like “vertically-shaped” (see pictures, PCT, p.231), same with “triangularly” 1 and 2. Thus, (S1A) and (S2A) effectively capture the same facts as the conjunction of (S1B) and (S2B). The dimensionally impoverished Flatlander, relying on trustworthy testimony from a superior source, can reasonably conclude that (S1) and (S2) are merely apparently contradictory statements about some “transcendent” object. Anderson then goes on to make the same kind of points regarding the Incarnation statements that taken together are paradoxical. He also says we can use this analysis for other paradoxical doctrines. For example, Reformed doctrines of providence and sovereignty coupled with biblically warranted convictions about moral responsibility can be handled similarly by “drawing appropriate distinctions within our intuitive notion of causal determinism. So, we would have:

(P1) God determined1 that people would conspire to have Jesus killed.
(P2) God did not determine2 that people would conspire to have Jesus killed.

In a footnote, he claims that (P2) would be a “deduction from the fact that those who conspired against Jesus freely chose to do so and were morally culpable for their actions.” This is an area where I disagree with Anderson.44

At this point Anderson draws on the doctrine of analogy as used in discussions of religious language; or, God-talk. He favors Ross’ approach to the doctrine of analogy. This all seems like heading in the right direction to me. Anderson would allow that language is looser, rather than rigid. So, we do not have mere equivocation, but analogy going on in the terms we use, and God used(!), to formulate (say) the doctrine of the Trinity. And is not language like this? Christians can make use of some of the insights found in, say, C.S. Lewis’s paper on language Bluspels and Flalansferes, or Lakoff’s Metaphors We Live By, for good arguments in favor of the notion that language is more analogical or metaphorical than we usually think. These insights, and others like them, could be used to bolster viewing apparent contradictions as species of MACRUEs. Viewing language this way, especially language about God, has been a tradition Christians have embraced more than eschewed. As Anderson states,

If analogy can be explicated as meaning adaptation controlled by linguistic forces exerted through the context and domain of discourse, then the model of theological paradox presented here fits hand—in—glove with at least one contemporary exposition of the Christian doctrine of analogy in religious language.45

After presenting and defending the notion of a MACRUE account of paradox in Christian theology, as well as discussing why Christians should view paradoxical doctrines this way. The three reasons he gives are: (i) it allows the Christian to avoid the irrationalism of denying classical laws of logic; (ii) it provides the groundwork for a defense of those doctrines; and (iii) it comports with the doctrine of analogy.

Another question Anderson asks is, what might account for the presence of such paradoxes in Christian theology? Is there anything in Christian theology that would lead us to expect such a phenomenon? Anderson suggests that there is just this sort of thing present: the doctrine of divine incomprehensibility. Anderson then launches into a rather magisterial exposition of “a modest” portrayal of the doctrine of divine incomprehensibility, following in the footsteps of many orthodox theologians (at least “magisterial” for Christians, and especially in light of the entire discussion; indeed, God‘s thoughts and ways are not our thoughts and ways). He mentions that the limitations of being able to comprehend the divine nature fully and exhaustively are both quantitative as well as qualitative. It is reasonable to assume that we altogether lack some of the concepts required to comprehend God’s essence perfectly. This doctrine “suggests that our minds are limited in terms of both epistemic capacity and conceptual accuracy: considered both qualitatively and quantitatively, our cognitive apparatus is simply not on par with God’s.”46
Further implications arise from this doctrine such that God has accommodated His revelation to us, using human language, employing the concepts and categories we use. None of this entails that His revelation about himself is not true; it is just not the whole truth. It is adequate for our needs, lacking precision. Calvin said God speaks baby talk to us. Anderson also invokes the Creator/creature distinction, and all that that entails for our language about God. This all points us back to the doctrine of analogy, exhibiting the internal coherence of the model. Given all this, why think our language is univocal with respect to God’s nature? This fits hand—inVglove with Anderson’s MACRUE account. Not only is this section good for the ivory tower, it is also of spiritual value. One is forced to see the majesty and transcendence of our God. Our language does not capture the totality of His essence. He is definitely not just a “bigger” and quantitatively “smarter” person than us.

This “modest model” of divine incomprehensibility contributes to the RAPT model by at least leading us to anticipate paradox in some of our theological knowledge. This doctrine (e.g., the incomprehensibility doctrine) informs us that our understanding of God is limited by man’s finitude and the imprecision of human conceptual faculties. Since our concepts and categories by which we acquire knowledge of God are approximations of the perfect concepts and categories needed to accommodate comprehensive and maximally precise knowledge of God, it is likely that at certain points in our theological reasoning the concepts we employ will be sufficiently refined to support the distinctions required to obviate any and all appearance of contradiction. We can formally articulate the existence of the distinctions, but we still lack the conceptual precision needed to be able to grasp clearly the content of those distinctions. We cannot see “just how God can be F in one respect but not-F in another respect.” Previous examples of MACRUEs make this point. However, the “modest model” does not give us reason to believe that most of our theorizing about God is paradoxical; neither does it entail paradox in our theology. For the most part, we do not need to resort to paradox. We should “treat matters on a case—by—case basis.” Anderson also gives a list of criteria that must be met in order for someone to claim paradoxicality for a doctrine. This list makes it impossible for one to claim that just any doctrine is paradoxical.47

Coming to the heart of the RAPT model, Anderson invokes the venerable tradition of “mystery.” Mystery does not threaten the rationality of Christian belief; it actually helps to explain their rationality. “A central claim of the RAPT model is that the Christian’s affirmation of paradoxical doctrine can be warranted via an appeal to mystery.” Anderson points out, through a fascinating discussion with Dale Tuggy48 regarding appeals to mystery, that the doctrines of the Trinity and Incarnation are mysterious because they appear to be logically inconsistent, but the MACRUE phenomenon is a symptom of a deeper mystery—the mystery residing in the incomprehensibility of God and our imprecise conceptual faculties. Thus, Anderson defines ‘mystery’ as: “a metaphysical state of affairs the revelation of which appears implicitly contradictory to us on account of present limitations in our cognitive apparatus and thus resists systematic description in a perspicuously consistent manner.”49

In order to bring to bear how the RAPT model allows for the rational, warranted belief in paradoxical doctrines, Anderson discusses further the notion of defeaters. This time he expands his previous discussion and looks at the notion of defeater-defeaters and defeater-insulators.50 Anderson finds agreement with Dale Tuggy and Stephen Davis regarding what criteria should be met if a paradoxical claim is to be rationally believed. First, there must be good reason to believe the contradiction to be a MAC; second, there must be strong, independent reasons for believing their component statements. Anderson agrees but also wants to add that one of these conditions may be fulfilled (in part, at least) by the other. That is, if Anderson has strong grounds for believing each component claim of the set of claims that seem inconsistent, he therefore has good reason to suspect that the inconsistency is a MAC. This was supported by Anderson’s examples of MACRUEs earlier in the chapter. In addition, in chapter 5, Anderson made the argument that belief in the component claims of central Christian doctrine may be warranted in numerous ways. If these hold, then the doctrinal beliefs will be insulated from defeat (cf. PCT, or papers referred to in n. 30 and 31 of this review).

Anderson notes that it is quite plausible to suppose that warranted beliefs in the doctrines of the Trinity and the Incarnation are intrinsically insulated from defeat. However, appearance of contradiction might be just the sort of thing that no belief could be insulated from. But Anderson points out that it would be hard to make this case. After all, whether or not defeat is the rational outcome will be determined by the defeater-system of the design plan. In addition, God designed our cognitive faculties. If Christianity is true, then the doctrines of the Trinity and Incarnation are true, and so it would be hard to see why God would have given us a defeater-system that would admit defeat for doctrines that God would want us to take as rational and warranted, despite the appearance of contradiction. Thus, one cannot prove that the doctrines of the Trinity and Incarnation are irrational without assuming that those doctrines are false. God designed our cognitive faculties to be aimed at the production of true beliefs. Therefore, Anderson sees a “close relative” to Plantinga’s thesis that the de jure question is not independent of the de facto question.

Furthermore, even if the above objections did not work (for sake of argument) other factors can be appealed to. As Tuggy and Davis point out, we also need reason for thinking the doctrines are MACRUEs. Given all that Anderson argued here, we do have warrant for believing them so. Anderson offers a few examples where all this cashes out. They show that mystery can serve as a defeater-defeater, or even better, a defeater-insulator, in the face of intellectual challenges posed by the paradoxicality of some Christian doctrines.

He also draws on the reasoning supplied by recent responses to the problem of evil. Anderson finds similarity between his RAPT model and some of these defenses.51 We might not be able to see how everything works together for good, but we trust God and take Him at His word. He far surpasses our wisdom, and we are warranted, based on His testimony (which is knowledge by testimony), that He does indeed have good reason for what He plans and allows. Likewise, we are warranted in believing the individual propositions of the paradoxical doctrines based on His testimony (which is knowledge by testimony, especially aided by the Holy Spirit to produce the proper credulity that is natural for man, given the original intent of the design plan), and given our cognitive situation, the doctrine of analogy, God’s incomprehensibility, etc., why shouldn’t we take God at His word, even if we can’t precisely point out were the unarticulated equivocates lie? “Where is the irrationality in this?”, asks Anderson. Our situation is similar to the one Flatlander finds himself in. We saw that situation to be uncontroversial. The Flatlander is not irrational for holding to (S1) and (S2) based on the conceptual authority of the honest Spacelander. He may not, due to his limitations, and the metaphysical nature of the object in question, which is so much different than the two-dimensional objects in his world (but there are analogies, they have hints at the truth, just not the entire truth), be able to understand precisely how an object can fit Spacelander’s revelation, but he can rationally take it that there is a MACRUE involved. He is warranted, based on the testimony of Spacelander, in believing the component propositions. He is also warranted in believing the conjunction, especially given his situation, then.

So according to the RAPT model, paradoxical Christian doctrines are MACs resulting from unarticulated equivocation and they originate in genuine theological mystery. Given a proper function account of warrant, with the expansions Anderson built in, the RAPT model indicates how Christians can be warranted in belief in Christian doctrine in general. The component claims of the paradoxical doctrines are well supported by the biblical teaching, and thus a Christian belief in them can be warranted. For many Christians, these claims would be warranted to a degree that they would be intrinsically insulated from defeat by the belief that they appear contradictory. Not only that, holding a “modest” belief in the doctrine of divine incomprehensibility will possess a defeater-defeater or a defeater-insulator against defeat by the appearance of contradiction. What counts as defeat will be determined by the defeater-system of the design plan.

Anderson then ends by showing how if a believer acquires belief in the component claims of the paradoxical doctrine by way of the four paradigm cases of how a believer could be warranted in their general theological and doctrinal beliefs mentioned in chapter 5, and the believer takes the claims taken together to only be MACs, and the believer holds to the “modest” proposal of incomprehensibility (and all that went with that), then believer can be rational in holding to paradoxical doctrines. Overall, I think he fully succeeds in achieving his goal. He has presented a “model” whereby the Christian can hold beliefs in paradoxical claims all the while being rational and warranted in so doing. Thus, we now have our answer to the question of the “epistemic status” of paradoxical doctrines. The epistemological status for belief in paradoxical Christian doctrine is that the belief is warranted. So, there are paradoxes in Christian doctrine (derived from Scripture), they are characterized as MACRUEs, and belief in the paradoxical doctrines is warranted. Anderson’s case is much more detailed and robust than presented here. And no doubt, given the nature of this review, I have left many things out. However, I think I have shown that a prima facie case has been made for the rationality of believing in paradoxical Christian doctrine. The book should be consulted for the full case.

The Model Defended (Ch. 7)

At this point many might (probably do!) have objections to Anderson’s case. Chapter 7 is dedicated to defending the model against an array of criticisms. Anderson divides the field into three main branches: (i) Biblical Concerns (regarding objections stemming from whether the fount of paradox is located in creeds or Scripture; the use of logic as a hermeneutical tool; and paradox as a defeater for inspiration); (ii) Theological Concerns (regarding objections stemming from the practice of systematic theology; how orthodoxy and heterodoxy are defined; objections regarding alternative design plans; and the apologetic mirror problem); and (iii) Philosophical Concerns (regarding the crucial distinction between contradictions, real and apparent; consistency as an intellectual virtue; alternative notions of rationality; intuitional inertia; and comprehension and conception). I will not go through all of these objections and Anderson’s answers to them. Suffice it to say, not surprisingly, I think he succeeds here as well.

Not to leave the reader completely unsatisfied, I will discuss one objection: the apologetic mirror problem. The objection runs thus: So, say Anderson has succeeded. Isn’t this a double-edged sword? It would seem that other religions (or philosophical systems) could make use of this defense. This might affect the myriad apologetical arguments Christians have given based on internal contradictions between the teachings of various non-Christian religions and philosophies.

Anderson notes that this objection has some bite. But one thing it does not do is entail that Christians are irrational in believing paradoxical Christian doctrine. The most one would concede is that adherents of these other systems are also rational in affirming their apparently contradictory doctrines. In addition, the rationality of those non-Christian beliefs would depend upon the truth of those systems, just as the Christian RAPT model does. Furthermore, these systems would need to come up with and develop their own RAPT model; Anderson’s would not be congenial to the vast majority of non-Christian systems. They would also need to come up with their own criterion. Anderson’s do not allow just any set of seemingly contradictory beliefs to be labeled a MACRUE. Anderson’s model depends upon divine revelation also. Thus, only a small minority of religions could even hope to ride Anderson’s coattails here. All the rest would need to develop their own system, from scratch. As I am sure James Anderson would attest, this is no easy feat! At the end of the day, then, Anderson says that for the few religions that could make use of his approach, this is simply the price we must pay for reconciling orthodoxy with rationality. In his estimation, “the price is worth paying.”

I would go further by suggesting that the best apologetic arguments against Islam and Judaism (two of the best candidates for riding Anderson’s coattails) based on internal contradictions are not things they could appeal to paradox for a resolution. For example, Islam claims that the Gospels are inspired by Allah. The Gospels teach that Jesus is God.52 The Koran says he is not. Therefore, the contradiction is that Jesus is and is not God. Is the Muslim really going to accept both these claims about Jesus:

(J1) Jesus is very God of very God.
(J2) Jesus is not very God of very God.


(J1A) Jesus was crucified and died for the sins of man.
(J2A) Jesus was not crucified and did not die for the sins of man.

Or, inconsistencies between their own apologetic practices and what their own holy book teaches:

(A1) The Bible is not reliable as a witness to Jesus.
(A2) The Bible is reliable as a witness to Jesus.

I do not see how.

Further arguments made from God’s just and righteous requirements as revealed in the Torah (which both religions accept as divinely inspired) as inconsistent with what both Islam and modern day Judaism teach about how man can be right before God, are likewise claims I find hard to believe they would be able to appeal to paradox in order to resolve. Not only this, if they do not accept Anderson’s constraints, then they cannot fully ride his coattails. They must spend the time coming up with their own account of paradox and what criteria must be met for a set of theological claims to be paradoxical, and provide details as to how it can be warranted. This may take a while, and so there is no immediate apologetic mirror problem. If they do accept his constraints, I do not see how appealing to a MACRUE helps them out with the majority of the best our apologetic arsenal has to offer. Lastly, if they develop such an account, then they must grant us the ability to appeal to paradox as well. Therefore, they would also lose some of their apologetical arguments based on internal contradiction So it might be a six in one, a half dozen in the other, at best.53 Therefore, I do not find much to worry about in regards to this objection at all.

Conclusion: The Prospects of Paradox

Anderson ends by discussing various implications of his thesis (e.g., apologetical benefit, etc.,); some areas for further research (e.g., further study into the doctrines of the Trinity and Incarnation, systematic theology, etc.,), and whether paradox is a blessing in disguise. This short couple of paragraphs makes for one of the more interesting directions one could go with all Anderson has shown: might we not only be warranted in believing paradoxical doctrines but also have at our disposal an argument that confers some positive epistemic benefits on the Christian faith? Well, Christianity has garnered more attention because of them (there is no such thing as bad publicity, Anderson says), but might paradox offer some evidential value to the Christian faith? Given two religions, R1 and R2, and R1 involves striking claims that are not explicitly contradictory but defy all attempts to express them in a matter that satisfies human intuitions about what is possible, and R2 involves no such claims, which would bear the mark of transcendent origin? This is definitely an area to pursue. When I think about the complexities involved in the Trinity and the Incarnation (for example), I can only conclude that a being maximally more knowledgeable, intelligent, and wise than I, could come up with such things. No mere man would have come up with these doctrines. In fact, this is the case with the entire Christian story. God’s plan of salvation, the very plan, speaks of divine origin. It is, as Aslan said, “a deeper magic.” Speaking of C.S. Lewis, his comments are apropos:

Reality, in fact, is usually something you could not have guessed. That is one of the reasons I believe Christianity. It is a religion you could not have guessed. If it offered us just the kind of universe we had always expected, I should feel we were making it up. But, in fact, it is not the sort of thing anyone would have made up. It has just that queer twist about it that real things have. So let us leave behind all these boys' philosophies--these over simple answers. The problem is not simple and the answer is not going to be simple either.

This concludes my review. I think that Anderson’s work is one of the freshest and most important theses to hit the area of philosophical theology in some time. He is to be commended for his efforts. And, even if you disagree with him, since there is no good objection against the possibility of his model, I think that he has successfully removed, once and for all, the logical problem of the Trinity and the Incarnation (much like Plantinga’s disposal of the logical argument from evil).54 Just like Plantinga’s answer to the logical problem of evil does not require that you believe it is true, only that it is possible, so too Anderson’s thesis. Thus, it is nothing short of monumental to have the objections to Christianity based on the illogicality of the doctrines of the Trinity and Incarnation shown to be rendered a non-starter. So, the apologist can disagree with Anderson’s model but still make use of it (as long as you grant the possibility of the model, which shouldn‘t be hard to do), I just happen to think you should also agree with him.


1 James Anderson, Paradox In Christian Theology: An Analysis of Its Presence, Character, and Epistemic Status, (Paternoster, 2007); hereafter, PCT.

2 Anderson discusses the distinction between apparent and actual contradiction, this distinction will be mentioned below.

3 This is all very implicit and so the atheologian need not worry that he will be being ‘preached at’ throughout the book. However, I know that it would be the author’s hopes that the non-religious readers recognize how great a God we serve in light of the arguments and conclusions of PCT.

4 [P1] through [P3] are found on p.5 of PCT.

5 This is good since to call something a “paradox” is about as vague as saying that some movie, book, or art piece is “interesting.” Copi and Cohen note in their Introduction to Logic textbook that the word “interesting” is one of the most vague terms in the English language. One could successfully debate their correctness on this with the counter-example of “paradox.” For example, in his excellent introduction to the life of Alexander the Great, historian Paul Cartledge writes of Alexander’s battle against the Persian King’s navy that, “Between 334 and 331 [Alexander] defeated the Persian Great King’s Mediterranean navy— paradoxically, unpredictably and perhaps undeservedly-by land.” Paul Cartledge, Alexander the Great, Vintage, 2004, p.41 (emphasis supplied). Used in this way, many things in the Bible would be “paradoxical.” As Anderson’s defines and sets limits on the term, Cartledge’s use won’t do. Cartledge’s use may be more apropos when speaking about biblical themes such as a “servant-King,” or “the shepherd-as-the-lamb,” or “when I am weak, then I am strong.” I call these types of “paradoxes,” the “foolishness of God.” This is more in the vein of C. S. Lewis’ “deeper magic” that Aslan used.

6 PCT, 11.

7 PCT, 18.

8 PCT, 30-31.

9 Anderson readily admits, from the outset, that Barth is no modalist in the original, Sabellian sense, and he rejects modalism as heresy. The main reason Anderson says Barth has attracted the label of ‘modalist’ is due to his dissatisfaction with the ‘person’ applied to the hypostases. The term ‘person’ should only be ascribed to the essence of God, and so God is one person existing in ‘three modes or ways of being’ (PCT, 32-33).

10 Anderson notes that there is no agreement on how to resolve this “vexing” philosophical conundrum. Indeed, it is interesting to note that there are paradoxes with the notion of material constitution! See:

11 PCT, 51.

12 William Lane Craig critiques Brower’s and Rae’s Aristotelian approach here:, but he opts for a Social Trinitarian model in the end (cf. chapter 29 of J. P. Moreland and William Lane Craig, Philosophical Foundations for a Christian Worldview(Inter-Varsity Press, 2003). Anderson’s critiques were developed prior to the publication of Craig’s paper.

13 This is a good question!

14 PCT, 79-80 (emphasis original).

15 Anderson chose these two doctrines because of their ubiquity across denominational boundaries.

16 Anderson notes that the dialetheist have put up some good arguments on their behalf, but he does not acknowledge this with respect to arguments against the universality of modus ponens. He mentions Graham Priest for the dialetheists, but he does not mention anyone on the MP side. Douglas Walton has given a very good argument against the universal validity of modus ponens; see It remains to be seen, though, if Walton’s approach could work for the anti-deductivists Anderson is dealing with (do they view the conditional is viewed as a “hook” or not, do they take the generalization “if-then“ premise as defensible in certain instances, etc). But, Walton does seem to provide the very discriminating criteria Anderson requests of the anti-deductivist, i.e., how to distinguish between valid and invalid uses of modus ponens. I would appreciate a fuller interaction with the anti-deductivists on Anderson’s end, perhaps something in line with his responses to the dialetheist. However, perhaps this is not needed since I am not familiar enough with the various ins and outs, subtleties, etc., of the anti-deductivist position on the matter; so I could simply be making things more complicated.

17 For readers of this review not sufficiently familiar with Plantinga’s basic approach, and without the time (or desire!) to read the Trilogy, James Anderson presents a nice overview of Plantinga’s approach in his review of Warranted Christian Belief; see

18 That property which, in sufficient degree, turns true belief into knowledge.


20 See Plantinga’s Evolutionary Argument Against Naturalism for arguments to this effect. You can find this argument in his books Warrant and Proper Function and Warranted Christian Belief, his paper Naturalism Defeated?, the book edited by James Beilby, Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism, and Troy Nunley’s doctoral dissertation, A Defense of Alvin Plantinga’s Evolutionary Argument Against Naturalism.


22 In PCT Anderson does note that there is disagreement as to whether the sensus divinitatis is a cognitive faculty or a mere disposition toward true belief, or of actual knowledge of God. How to interpret Calvin is debated here. Anderson sides with such thinkers as K. Scott Oliphint, as well as Michael Sudduth in thinking the latter interpretation is to be preferred. Nevertheless, Anderson also notes that this is a minor squabble and Plantinga’s position would only need some tweaking here and there; the basics would remain the same, though.

23 A point for atheologians to consider. Christians believe that the knowledge of God in nature is insufficient to lead to salvation. In order to know salvific truths, man needs special revelation, not general (which is also known as “natural” revelation).

24 Mascord says Plantinga renders ordinary cognitive processes “useless (or secondary).” Keith Mascord, Alvin Plantinga and Christian Apologetics (Paternoster, 2006, p. 160).

25 Mascord agrees, ibid.

26 So too Mascord, ibid.

27 Anderson cites others who have pointed out this problem. Richard Gale and Andrew Dole are among those who Anderson is in agreement with on this objection.

28 Anderson believes that the notion of a “system” does not make incoherent the idea of intentional agents as components of the system. For example, the system of passport producing has intentional and non-intentional elements involved. I wonder why this cannot also be so of testimony. Educational systems are systems that provide education. There are many elements involved here, and surely testimony plays a role. Students are taught by their teachers in order to become more educated. This system, among other things, produces warranted beliefs in the students. So I am unclear as to why Anderson thinks that the notion of a system automatically “tends to reinforce” the notion that testimony does not contribute to warrant.

29 Some might balk here at Anderson’s short defense of knowledge by testimony, especially as he bases it on agreement with Plantinga in Warrant and Proper Function (WPF). One might be inclined to point out that Plantinga’s very position on knowledge by testimony has been shown to be subject to Gettier-style counter examples. Evan Fales tries to do just such a thing in his review of WPF (see here: page 2). Says Fales,

A system which is highly reliable can fail; if it can fail, it can also succeed where, but for luck, it would have failed. An example: a game of "telephone" in which A is to whisper something he knows in B's ear, B whispers it to C, and so on to the last person, Z, who accepts what he hears because he takes it to be what A said (and correctly takes A to know whereof he speaks). Everyone's hearing, speech, and cognitive systems are working sufficiently well, and the environment is sufficiently ideal, to meet Plantinga's requirements for warrant—warrant so high that, when combined with truth, it yields knowledge. Then given the way warrant is transmitted by testimony, Z's belief that A said "Plato was Greek", and hence, Z's belief that Plato was Greek, both count as knowledge on Plantinga's account, provided that both these beliefs are true. But now suppose that (improbably), some slippage occurs in the transmission of the message: halfway to Z, it has become "Pluto is green". By sheer luck, however, the slippage is reversed: by the time Y whispers it in Z's ear, "Pluto is green" has transmuted into "Plato was Greek". Surely, here, Z does not know what A said, or know that it is true merely because A said it.

The attentive reader will have made note of one of the features Anderson says must be included for a cognizer to have warrant for his belief. Thus Anderson, “that a belief B is warranted for S only if on adequate reflection S would not be aware of any sufficient reason to reject B.” This account is not included in Fales’ critique. And why is this system “reliable?” Indeed, isn’t this very argument used by atheists like Fales to conclude that the Bible has been changed or altered? The criticism goes like this: “You’ve all played telephone before, and you all know how unreliable that game is, well, how much more then is the Bible, having transmitted its information down a long line for thousands of years, to be viewed as unreliable?” Is this process reliable or not!? Furthermore, Fales focuses on the reintroduction of the correct phrase rather than the initial introduction of the incorrect phrase. How did “Pluto is green” arise? How did “Pluto is green” come to uttered by Johnny? These are questions we should look into. Perhaps Johnny has a bad ear and misheard the person who whispered the phrase into his ear? Perhaps he was, as is common with Johnny, distracted by the cute girl with pigtails in the front row and thus did not pay sufficient attention. Perhaps the person before Johnny just got a tongue ring piercing and, though she said, “Plato is Greek” it sounded like “Pluto is green.” Perhaps Johnny is a prankster and likes to mess up telephone games. These would be dysfunctions in the cognitive environment. Thus, we would not have warrant sufficient for knowledge. Furthermore, Z must believe the proposition with a sufficient degree of confidence I it. Would Z? Does Z know about Johnny and his frequent distractions? His mischievousness? Furthermore, Z’s belief is warranted only if the testifiers before him were warranted in their belief. Clearly they would not have been (if we dismiss, for argument’s sake, my above concerns), so Z would not have been warranted on Plantinga’s account. Lastly, since there was some kind of external malfunction or dysfunction up the line from Z, then the belief would not have external rationality! This is a necessary requirement for a belief to have warrant, though. Thus, the belief would not have been warranted. For these reasons, it is not clear to me, at all, that Fales has produced a Gettier -style counter example to Plantinga’s account of warrant.

30 See John Greco’s paper, Discrimination and Testimonial Knowledge for some helpful points regarding children and the formation of warranted beliefs based off testimony, here: Also relevant is the paper, Trust In Testimony: How Children Learn About Religion And Science by Harris and Koenig, see here:

31 Mascord staunchly disagrees. He critiques precisely the arguments Anderson uses, which are based off Plantinga (see Mascord, pp. 168-183). I am inclined to agree with Anderson, but an investigation into Mascord’s many arguments is beyond the scope of this review. After all, it is a review of PCT, not APCA!

32 Many who read Anderson’s book may be warranted in his conclusions based on something like [2]!

33 Many who read my review might be warranted in believing Anderson’s conclusions based on something like [3]!

34 PCT, 204-208. (As an aside, the last model should allow for parents to grant that their children really do believe, and are warranted in beliefs like: “I am saved because I believe Jesus died for me,” rather than brushing this aside as some kind of cute “copy cating.” This may allow for a smaller gap between when, say, Presbyterians baptize their children and when Baptists do. A Baptist who denies that their children are warranted in believing that Jesus died for them, based on their belief in their parents’ testimony that whoever believes in Jesus will be saved, might well have a defeater for holding to paradigm case [4]. This would not prove paedobaptism in the slightest; it would simply whittle away at the age gap that seems to be prevalent between these two camps.)

35 PCT, 210-211 (emphasis original).

36 PCT, 212.

37 Internal rationality has more to do with your mental states, your other beliefs, how they fit together, those things “downstream of experience,” etc., and external rationality has to do with your cognitive environment, those things “upstream of experience.”

38 PCT, 215.

39 Some material helpful for the Christian in working on the epistemology of testimony are: Testimony: A Philosophical Study, C. A. J. Coady, OUP, 1995; The Epistemology of Testimony, ed. Jennifer Lackey and Ernest Sosa, OUP, 2006; Is There Meaning in this Text?, Kevin J. Vanhoozer, Zondervan, 1998; The Trustworthiness of God: Perspectives on the Nature of Scripture, ed. Paul Helm and Carl R. Trueman, Wm. B. Eerdmans, 2002; Alvin Plantinga and Christian Apologetics, Keith Mascord, Paternoster, 2006; Thomas Reid and the Story of Epistemology, Nicholas Wolterstorff, Cambridge, 2002; Warrant and Proper Function, Alvin Plantinga, OUP, 1993; Intellectual Trust in Oneself and Others, Richard Foley, Cambridge, 2007; Stephen Braude, The Golden Leaf Lady and Other Parapsychological Investigations (U of Chicago Press 2007); and Stephen Braude, The Limits of Influence (Routledge & Kregal Paul 1986).

40 Epistemic possibility just means that some proposition or state of affairs is consistent with what we know, where “know” means something like propositions or states of affairs all parties in the discussion agree on.

41 Recall Anderson’s explication of these different types of contradictions in Ch.4.

42 Recall Anderson demonstrates this in Ch. 2, and showed (Ch. 4) that the most popular, and best, ways of resolving the problems were insufficient for the believer who wants to maintain rationality and orthodoxy.

43 And why should this be thought automatically problematic? Peter Unger (no theist) mentions possible refinements or inadequacies with the notion of identity as it bears on the mind/body debate. See Unger, All the Power in the World, Oxford, 2006, pgs. 106-10, as well as Ch. 5.

44 See PCT, 231-232. I do not see why we would have to say that God did not determine that people would conspire to have Jesus killed because we also have to account for the fact that “those who conspired against Jesus freely chose to do so and were morally culpable.” Why is that inconsistent with “determined?” The only reason I see is because one assumes this premise: (P1a) “determinism” is incompatible with “freedom and moral culpability.” But I do not see that taught in the Bible. This would only arise from a philosophical assumption of libertarian agency. Not only that, but (P1) and (P2) are not stated in the creeds, nor do I think they can be shown to be “implicitly” in there. There is no explicit contradiction, and a formal contradiction can only be drawn by postulating a premise committed to libertarian agency, which the Reformed confessions do not have included amongst their propositions. Rather than postulate (P2) why not have:

(P1*) God determined that the people would conspire to have Jesus killed.
(P2*) God did not force the people against their will to conspire to have Jesus killed.
(P3) The people did what they wanted to do and had reasons.
(P4) God holds them morally culpable.

(P1*), (P2*), (P3) and (P4) seem to be able to be inferred from the biblical data. They can furthermore be defended by (semi-) compatibilist philosophical arguments, showing that there is no inconsistency between moral responsibility and determinism, and Frankfurt arguments can be invoked to show that PAPs are not necessary to have “freedom.” This move is needed to rebut the challenges of the libertarians who, though they cannot challenge on exegetical grounds, nevertheless try to challenge on philosophical grounds. In discussions with Anderson he has expressed disagreement with me on this point. I would add that even if the above argument I’m making were not cogent, I also do not see how this “paradox” fits in which his constraints on what allows for some doctrine to be paradoxical. At best, if there is a “paradox” here, it would seem to result more from our view of the metaphysics of human free will and not from the nature of God. Anderson claims that paradox must result from God’s nature (this fits in with the Creator/creature distinction, the doctrine of incomprehensibility, etc). It would seem on his own terms there would not be a paradox here. But, I am open to correction on this point.

45 PCT, 36.

46 For more on the doctrines of incomprehensibility and analogy, especially as to its pedigree in Reformed circles, see John R. Muether, Cornelius Van Til: Reformed Apologist and Churchman, P&R, 2008, pp. 103-10, 109-112. See also R. Scott Clark's Janus, the Well-Meant Offer of the Gospel and Westminster Theology,” in David VanDrunen, ed., The Pattern of Sound Words: A Festschrift for Robert B. Strimple (P&R, 2004), 149-80.

47 Pace what some Van Tillians have said on this point.

48 See Anderson’s reply to Dale Tuggy, In Defense of Mystery: A Reply To Dale Tuggy, Religious Studies 41:2 (2005), 145-163, here: Much of this paper was used in chapter 6.

49 See also Kenneth Boa’s, God I don’t Understand: Answers to Biblical Questions of the Faith (Victor, 2007) for similar comments. Boa argues that human wisdom is subordinate to revelation. Says Boa, "it would be the height of egotism for a person to say that because an idea in the Bible does not make sense... it cannot be true." In addition, "when a person insists on trying to subject the two contradictory elements contained in a biblical mystery to human comprehension, he will inevitably...move to one extreme or to the other....We need to understand both elements in each mystery....To maintain proper balance we should accept the tension by supporting both ideas involved equally."

50 See Michael Sudduth’s paper The Internalist Character and Evidentialist Implications of Plantingian Defeaters, here: For another profitable discussion on defeaters, especially some of the thought experiments Anderson uses from Plantinga, see Plantinga’s Naturalism Defeated, here:

51 See Greg Welty’s paper for excellent verification of what Anderson claims here:

52 See Putting Jesus in His Place: The Case For The Divinity of Christ, by Robert M. Bowman, Jr. and J. Ed Komoszewski (Kregel, 2007), for an excellent case for the deity of Christ.

53 At best because it is not even clear that their arguments from internal contradictions are even accurate. For example, the Muslims are famous for taking us to be teaching tri-theism (cf. Sura 5:116, for example).

54 I say this for argument sake. I would agree with Greg Welty’s analysis of the situation. According to Welty, Plantinga does not succeed in refuting the logical argument from evil. See here: Nevertheless, Welty does “agree with the near-universal consensus that the logical argument from evil is a "late lamented deductive cousin" of the evidential argument . . .”

1 comment:

  1. Due to the length of this post, and the possibility that it will be printed off and read, I have dedicated another place for any comments to be posted.