Certainty and Irrevisability

Certainty, Irrevisability, and Theological Beliefs[1]

Michael Czapkay Sudduth

Christians often claim that their theological beliefs are “absolutely certainty.” In the following epistemological discussion I clarify the senses in which I think this claim is false.

I. A Basic Distinction: Psychological and Epistemic Certainty

The term "certainty" generally divides into two different uses:

Psychological (or descriptive) certainty

Epistemic (or normative) certainty.

The former refers to the degree of confidence, conviction, belief a person has in some proposition. If a person is certain in this sense, he has a maximal (or nearly so ) subjective conviction or assurance of the truth of something.

By contrast, epistemic certainty refers to the degree of *warrant* possessed by a belief. Warrant is, roughly, a truth-indicating property, typically cashed out in terms of the extent to which one's ground, basis, reason, or source of belief are adequately truth-indicating. ("Knowledge," will be understood here as a strongly warranted true belief). An epistemically certain belief is, roughly, one that has the best possible grounds or evidence, typically the sort that preclude all possible reasons for doubting the truth of the proposition or belief in question. Epistemically certain beliefs will have maximal warrant.

What are the relations between psychological and epistemic certainty?

Psychological certainty does not entail epistemic certainty, though the converse typically is true. People are - subjectively speaking - "sure" of many things the warrant for which is less than maximal. Suppose I believe that <the 49ers will beat the Vikings> because I consulted the Magic 8-Ball, which said "It is certain" when I asked, "Will the 49ers beat the Vikings this Sunday?" I might be extremely confident that <the 49ers will beat the Vikings> on the basis of the testimony of the Magic 8-Ball, say because I'm very sure about the reliability of the Magic 8-Ball in divining the outcome of NFL games. But my belief, even if it has some degree of warrant, surely has nowhere close to maximal warrant. Many other beliefs will have considerably more by way of warrant and there is a seemingly endless number of possible grounds for doubting my belief that <the 49ers will beat the Vikings>. The belief in question is psychologically certain but not epistemically certain.

It is also possible, indeed quite common, for beliefs to have a high degree of warrant and be true, but fail to be epistemically certain. Take my belief that <it is raining outside>. Suppose that I believe very firmly that <it is raining outside>. Suppose further that I have good reasons to suppose that it is raining outside, e.g., my eye-sight is quite good, I have a clear view of my backyard, my wife has just entered the house drenched, and the local weather report I heard the night before said that it would rain today. It follows that I have warrant for my belief that <it is raining outside>, and the degree of warrant possessed by my belief is pretty high given how the scenario is described. So I have a highly warranted, firm belief that <it is raining outside>. Let's add now that it *is* raining outside. So my belief is true to boot. With the satisfaction of all these conditions (highly warranted true belief), it is sensible to suppose that I *know* it is raining outside.

Despite all of this, though, there are *possible* grounds for doubting the proposition <it is raining outside>. It is possible after all, that I am a victim of a community-wide practical joke. It is possible that I am in fact dreaming. It is possible that what I take to be rain falling is just water from my neighbor's garden hose falling into my yard. It is possible that I have ingested a drug that engenders rain hallucinations. Now these scenarios may not be very likely. I may have no reason to suppose that they are true. In fact, they may be false. But they still represent possible grounds for doubt, which don't have to be either true or likely to engender a requirement for doubt. They only have to be possible and such that *if* they were true, my warrant for believing <it is raining outside> would be ever so slightly reduced.

Now epistemic certainty is usually taken, as in the Cartesian tradition, to rule out all possibility of doubt. That is, certainty is either equivalent to or entails *indubitability* as just described above. Indubitability entails *maximal warrant* since there can be no better grounds or reasons for believing a proposition than there being no possible grounds for doubting the truth of the proposition. In this case, though, it becomes fairly clear that we must distinguish epistemic certainty from firmly believing something, firmly believing what happens to be true (though one has no or little warrant for the belief), and having a firm, highly warranted true belief. Since the latter is likely necessary and sufficient for knowledge, epistemic certainty must also be distinguished from knowledge. All beliefs that are epistemically certain are items of knowledge but not every instance of knowledge is epistemically certain.

II. Certainty and Irrevisability

Epistemic certainty, taken as indubitability in the above sense, should also be distinguished from irrevisability and incorrigibility - the idea that a belief is beyond revision.

(i) A belief that is irrevisable is, at least on one view, a belief that no one can show to be false. A belief might pass this test and fail to be epistemically certain, for even if it is not possible for anyone to show that some person's belief is false, there might nonetheless be possible (or even good) grounds for doubting the truth of the person's belief. I can easily have reasons for doubting that the doctrine of the universal, invisible church is true, even if I lacked reasons for supposing that it is false. Of course, a belief's being epistemically certain *does* entail that it is irrevisable, for if it is possible to show that a belief is false, there are possible grounds for doubting it.

(ii) On a slightly different account of irrevisability, a belief will be irrevisable just if no future conditions could conceivably warrant the retraction of the belief. Showing that a belief is false is a sufficient condition for warranting the retraction of a belief, so (i) entails (ii), but the retraction of a belief might be warranted by good reasons for no longer thinking that the belief is true, not necessarily reasons for thinking the belief is false. So (ii) does not entail (i).

What is required under (ii) is that it is possible that future evidence, at least equally warranted as the evidence one currently has for one's belief that p, should warrant the retraction of the belief that p. As with (i), an indubitable or epistemically certain belief would be irrevisable in the sense of (ii), but the converse would not be true. That fact that there is or can be no evidence at least equally warranted for a person that demands the retraction of the belief that p does not entail that there is or can be no other belief that has more warrant for the person than the belief that p. It may be that no conceivable or possible future evidence would make it rational or warranted for me to retract my belief that I was born in California, but surely many beliefs have more warrant for me than this belief (e.g, that I am thinking right now, that I feel sick to my stomach, that my head hurts). Similarly, it may be impossible to show that a belief is false or to acquire evidence that warrants retracting a belief, even if there are possible grounds for doubting the truth of the belief.

III. Epistemic Certainty and Theological Beliefs

A fairly crucial distinction emerges from all this. Not only is knowledge distinct from epistemic certainty, but epistemic certainty (as either maximal warrant or indubitability) is distinct from irrevisability. It is fairly important to keep these distinct, especially in the current thread.

Among other things, while showing that certain theological beliefs are epistemically certain would suffice for showing that they are irrevisable, theological beliefs might be irrevisable even if they are not epistemically certain. So one needs to decide what the real issue is here. Is it certainty? Or is it irrevisability? The thread started, I think, on the question of whether beliefs associated with the Second London Baptist Confession were beyond revision. As I recall, David Charles, Pastor Shane, and Keith S, inserted the notion of certainty. But again, it's best to keep these two notions distinct.

The crucial question, then, is whether there are any theological beliefs putatively derived from Scripture that, though they fail to be indubitable, are nonetheless irrevisable. What's the answer? I'm inclined to think that there aren't any.

My main reason for supposing that there aren't any irrevisable scripturally based theological beliefs is as follows. To the extent that such beliefs are allegedly derived from Scripture, we must be open to the possibility that future *Biblical* evidence *could* warrant the retraction of *any* member of our current set of putatively scripturally derived beliefs. Otherwise stated, since our "biblically derived" theological beliefs (or at least their warrant) depend on our interpretive principles and their particular application, our theological beliefs will be irrevisable only if our interpretive principles and our judgments about how they are applied in a particular context are irrevisable. But I can't see why we should think that the latter is so. Consequently, I have my doubts about the former.

IV. Cartesianism and Calvinism

As for the connection with Descartes, a fairly crucial element of Descartes' epistemology is the assumption (shared by the Skeptics of his day) that knowledge requires epistemic certainty. Hence, if we can engender doubt about whether anything is epistemically certain for us, then we can engender doubt about whether we have knowledge. Descartes sought to show that there are truths about which we can have epistemic certainty, beginning with the certainty of one's own existence (cogito, ergo sum). Such truths can pass the most rigorous skeptical tests. Perhaps so. The problem is that little of the rest of what we believe can. While Descartes thought he could escape this consequence by proving with certainty the existence of God, it looks as if either his proofs of God's existence don't work or they work but don't adequately bridge the gap between the certainty of our knowledge of self and the certainty of our knowledge of all other things such as the external world. It looks like Descartes is ultimately unable to defend "empirical knowledge" claims against skeptical attacks. Some of us maintain that failure is intrinsically connected to the assumption that knowledge requires epistemic certainty. That is where the big D went wrong, and where all theories of knowledge will also go wrong if they share Descartes' assumption.

For more on Descartes' epistemology and its problems, see my handouts on Descartes:

http://www.homestead.com/mscourses/files/philosophy357handout1.htm

http://www.homestead.com/mscourses/files/philosophy357handout2.htm

One of my philosophical concerns about the state of contemporary Calvinism is that it leans heavily on the Skeptical and Cartesian assumption that knowledge requires some sort of epistemic certainty. This comes up in a variety of different contexts.

For example, I raised this concern on the RB list back in May 2001

http://www.rblist.org/archive/msg18151.html

http://www.rblist.org/archive/msg18170.html

http://www.rblist.org/archive/msg17878.html

I have also more recently discussed it in relation to Robert Reymond:

http://www.rblist.org/archive/msg26638.html

http://www.rblist.org/archive/msg26649.html

For further evidence of Cartesian Calvinism, see Sean Choi's protracted battle with Cartesian Van Tilians and others on the Van Til Applied list earlier this year:

http://groups.yahoo.com/group/vantil-applied/message/1527

http://groups.yahoo.com/group/vantil-applied/message/1573

http://groups.yahoo.com/group/vantil-applied/message/1585

http://groups.yahoo.com/group/vantil-applied/message/1586

http://groups.yahoo.com/group/vantil-applied/message/1588

http://groups.yahoo.com/group/vantil-applied/message/1595

http://groups.yahoo.com/group/vantil-applied/message/1624

http://groups.yahoo.com/group/vantil-applied/message/1625

http://groups.yahoo.com/group/vantil-applied/message/1626

http://groups.yahoo.com/group/vantil-applied/message/1657

http://groups.yahoo.com/group/vantil-applied/message/1673

http://groups.yahoo.com/group/vantil-applied/message/1684

Granted, not all Calvinists who emphasize the necessity of certainty speak with sufficient philosophical clarity about the nature of "certainty." But there is at least an alarming resemblance to Cartesianism in the emphasis placed on the *need* to hhave theological beliefs that satisfy very stringent, howbeit inchoate, epistemic requirements. Some Calvinists speak with greater clarity on the issue and the connection to Descartes is more evident.

It is, of course, possible that some Calvinists have redefined certainty to fit their own purposes or agenda. But the lack of clarity at this juncture only further complicates and confuses matters. If one is going to use philosophical terminology, it is crucial to understand what meaning these terms carry in philosophy. And if one is going to redefine these terms, then *re*define them, otherwise important claims will be misunderstood. If someone says, "'quarks' don't exist," I must presume he intends to raise doubts about quantum mechanics. Sure, it's possible that he is really raising doubts about the existence of fairies. But then he shouldn't be surprised if physicists miss his point and label him an intellectual crackpot.

[1] Excerpted from an e-mail originally posted to the Reformed Baptist discussion group on September 26, 2003. For a more technical account of the varieties of certainty in connection with theistic belief, see chapter 10 of my book (in progress) The Reformed Objection to Natural Theology.