27 December 2013

Unpacking Randal Rauser's claim that "testimony is properly basic belief"

I have to confess that when I first came across Randal Rauser's claim that testimony is properly basic, my gut reaction was that it was one of the most ridiculous apologetic arguments I'd ever heard. Surely, I thought, he cannot actually be arguing what I think he is. Surely there is some sophisticated aspect to the argument I'm overlooking or failing to properly comprehend.  Recently the debate was rekindled after I pointed out the research which shows eyewitness testimony to be of fairly dubious reliability. Spurred by the debate, I decided to revisit Randal's original post and see what I might have overlooked. Is it really as stupid an argument as I think it is, or am I totally failing to comprehend an incisive philosophical argument that helps substantiate (among other things) the Resurrection of Christ and modern-day miracles?

In fairness to Randal, I'll reprint his summary of his original argument:
It is true that you can think of testimony as evidence just like you can think of sense perception and rational intuition as evidence. But so long as there are no defeaters present, we also find that each of these sources of rational belief can provide a stopping point to the regress of justification and thus each provides a means of prima facie proper basicality. Not only that, but sufficiently strong testimony can also withstand the assault of formidable defeaters. Consequently, just as rational intuition can provide us with properly basic belief that 7+5=12, and sense perception can provide us with properly basic belief that “The banana is on the counter”, so testimony can provide us with properly basic belief that “Common descent is true” or “Jesus rose from the dead.”

The terms of engagement

First, I think it's helpful to define the terms being used. The word "testimony" is generally used in a legal or religious context, but in philosophy it can simply be synonymous with an assertion or claim. From the Stanford Encyclopedia of Philosophy:
The primary speech-act of testimony is a speaker's saying, telling, or asserting something (Searle 1969). Assertion puts forth a proposition that the speaker represents as true (see the entry on assertion, on norms of assertion, Rescorla 2009)

"Properly basic belief" is a term that's not used much outside of theologians referencing Alvin Plantinga (more on that later). The term "basic belief" derives from Foundationalism and its various derivatives. A basic belief is essentially one that needs no other justification – one that's actually irrational to question.

The most obvious example of a basic belief is "I exist". The assumption of one's own existence is not only a necessary starting point for any epistemological framework, but it's irrational to be skeptical of one's own existence: the statement "I am skeptical of my own existence" assumes the existence of an "I" that is capable of skepticism!

Randal mentions sense perception as a basic belief ("The banana is on the counter"), but I'd be more specific than that. It's entirely possible that there is no banana, that his senses are being deceived – like a thirsty desert wanderer mistaking the reflection of the sun against the sand for an oasis. As optical illusions show, our senses can and do deceive us. But we must operate on the assumption that our senses are at least sometimes reliable, otherwise we've painted ourselves into an epistemological corner.

Now on a personal note, I very much dislike the term "basic belief", because the word "belief" generally refers to assumptions or convictions that we consciously hold, and which can be amended with the introduction of new evidence. Just like skepticism of one's own existence, belief in one's own existence is paradoxical – or at least redundant: "I believe in my own existence" already assumes the existence of an "I" capable of belief. So rather than the term "basic belief", I prefer terms like "foundational assumption" or "necessary assumption". We don't consciously worry about our own existence when thinking about knowledge and epistemology. But that's just a personal quirk, and I'll continue to use the term "basic belief" to be consistent with Randal's use of the term.

Why it's ridiculous

My gut reaction upon reading Randal's argument was that it's incredibly ridiculous, for reasons that seem so obvious that they shouldn't even need mentioning. 

While it's clearly irrational to be skeptical of our own existence or the entirety of our sensory perception, there are many, many reasons to be skeptical of assertions made by others. A litany of cognitive biases color our perception of events – confirmation bias, groupthink, attribution bias, clustering illusion, causal fallacies, and many, many more. Worse, our memories themselves often deceive us. It's a common error among those unfamiliar with cognitive psychology to assume that our memories work like video cameras, capturing external information and storing it just as it happened. The reality is that memory generation is an active process whereby the brain takes bits and pieces of sensory data and fills in the gaps retroactively to make the events coherent – this is known as reconstructive memory. In fact, research indicates that it may be impossible for us to remember an event without altering it in some way [1], [2]. And that's just when you're dealing with the testimony of one person. Every time testimony is passed from one person to the next, bits of information are omitted, added, or altered as each person's recall distorts the original memory. 

That's the science, and it's pretty damning to Randal's case. If the fallibility of human cognition and memory renders testimony inaccurate and unreliable, then how can it be considered a properly basic belief? But even if one is unaware of the research showing just how depressingly unreliable our perception and memories actually are, it ought to be common sense that people's testimony can be and often is mistaken. But wait a second here... have we thrown the baby out with the bathwater? Because obviously lots of information we know about the world comes through the testimony of others. Does our unreliable cognition and memory mean that it's never rational to trust testimony?

Where was I supposed to turn, again?

In the latest discussion/debate on the matter, Randal used the example of getting directions from a stranger as a counter-argument to my skepticism:

Randal overlooks several important things here, and his example ends up making my own argument for me. Remember, the question is not whether testimony can ever be trusted at all, but whether it can be considered a properly basic belief along with things like sense perception and self-existence.

Firstly, our general trust of others comes from experiential evidence. We know, from experience, that people tend not to lie to us without some vested interest in doing so, and it's unlikely that someone is going to give us false or misleading directions just to mess with us. We acknowledge that this is a remote possibility, but our experience tells us that strangers are generally kind in such circumstances. Note, for example, that Randal describes the witnesses in the above example as "intelligent looking people". That statement alone is predicated on a litany of evidence-based assumptions (which can be and often are wrong) about which types of people are trustworthy and/or reliable, and which are not.

Secondly, whenever we get directions from someone we implicitly acknowledge that their directions may in fact be incorrect. It's not that we think they intentionally deceived us, but that we know people's recall is fallible. And the more detailed directions we require, the higher the probability of error.

Both of these facts demonstrate that it is rational to have some degree – if only a small degree – of skepticism regarding the accuracy of directions given to us by others. Note that this contrasts with true properly basic beliefs which are irrational to hold any skepticism toward at all.

Randal then gives another bad example regarding reports of a rare bird:

There's a field of pseudoscience called cryptozoology that deals with mythical creatures like bigfoot, yeti, and the chupacabra. Often, belief in the existence of such animals is buttressed by the eyewitness testimonies of indigenous peoples who claim to have encountered the beasts. But whenever independently verifiable empirical evidence of such creatures is sought, the creatures turn out to be either entirely mythical or simply rare cases of mistaken identity.

The point is that testimony is not, and should not, be taken at face value precisely because we humans often perceive and remember events inaccurately.

Multiple attestations

One of Randal's rebuttals here implies that I ought to reject scientific research as well, because my knowledge of it relies on testimonial evidence. But again, the question is not whether testimony is ever reliable, but whether it can be considered a properly basic belief that needs no other evidential justification.

Multiple eyewitnesses to a crime is generally considered to be stronger evidence than a single eyewitness. That's because we recognize that as more people are involved who can independently corroborate an event, the probability that our judgements will succumb to the errors of a single individual decreases.

Even if I know nothing at all about the science of common ancestry, it's rational for me to accept it on the basis of the fact that an overwhelming majority – virtually the entirety, in fact – of biological scientists all over the world, comprising millions of people and 150 years of research, have corroborated its veracity as a fact. Moreover, they have extensively documented their research and made it accessible to almost anyone, and if I want to contribute to the field I had better be prepared to base my hypotheses on independently verifiable empirical evidence. Clearly, "testimonial evidence" in this regard should never be conflated with a single person, or even a small group of people, who claim to have witness a "miracle" or other supernatural event. This not merely the testimony of others being taken at face value, but testimony that is corroborated with independently verifiable empirical evidence to which I, and others, have access.

Randal's biggest blunder

Remember that Randal's argument here is that someone's testimony is unto itself a rational reason to affirm the truth of an assertion: "It is true that you can think of testimony as evidence just like you can think of sense perception and rational intuition as evidence. But so long as there are no defeaters present, we also find that each of these sources of rational belief can provide a stopping point to the regress of justification and thus each provides a means of prima facie proper basicality."

Where Randal really slips into nonsensical territory is with his statements about "defeaters". In epistemology, a "defeater" is just something which causes a belief to lose some degree of positive epistemic status. A true properly basic belief cannot have a defeater, because to be skeptical of it is irrational (as mentioned earlier). But the fallibility of human cognition clearly gives us a rational reason (or rather, a set of reasons) to be skeptical of testimony.

Randal seems to think that unless a belief he holds can be shown to be false, he is justified in believing it. But that is, as the saying goes, back-asswards. Beliefs do not fall into neatly divided categories of demonstrably valid and demonstrably invalid. They may also be indeterminate.

So, let's imagine that someone tells us that they've witnessed a good old-fashioned Christian miracle. If we say, "I don't believe you", that is not the same thing as saying "I can demonstrate your assertion to be false"; rather, it's simply saying, "Your assertion alone is not sufficient reason for me to affirm your claim – it's possible that you may be mistaken".

If we take the position that we are justified in believing something unless it can be shown to be false, invalid, or unreliable, then we've given ourselves epistemic license to believe just about anything we want – because there are an infinite number of assertions or beliefs which are not demonstrably false. Let's just re-word the last few lines of Randal's original argument to drive the point home. Note that even as the assertions themselves can be changed into virtually any arbitrary claim, the argument itself remains unchanged:
Consequently, just as rational intuition can provide us with properly basic belief that 7+5=12, and sense perception can provide us with properly basic belief that “The banana is on the counter”, so testimony can provide us with properly basic belief that “Buddha performed eight miracles” or “Superintelligent aliens exist beyond the horizon of the observable universe”.

Consequently, just as rational intuition can provide us with properly basic belief that 7+5=12, and sense perception can provide us with properly basic belief that “The banana is on the counter”, so testimony can provide us with properly basic belief that "Satan is our overlord” or “Jesus is a mythical figure”.
[On a side note, I think Randal's wrong about rational intuition being the basis for mathematics – mathematical axioms are based on set theory, which is derived from observational evidence (the existence of discrete objects that can be grouped into sets). I highly recommend Where Mathematics Come From by Lakoff and Nunez.]

All is not lost here, for Randal's part, as he actually inserts a little something that's sensible:
Think about it. Sense perception has also shown to be “horribly, horribly unreliable”. But the answer is not to blind your eyes and deafen your ears. Rather, it is to proceed with a skeptical caution and take the deliverances of sense perception as fallible and provisional.
He's right, of course, but the nuance here is that he erred in his original post on the topic in claiming that sensory perception itself is properly basic, rather than the assumption that senses are reliable to some degree. A true properly basic belief, such as the assumption of my own existence or that my senses are sometimes reliable, is neither fallible nor provisional; Randal makes the mistake of confusing axioms demonstrated to be reliable through induction with foundational assumptions that are required in order for me to develop any further epistemological framework.

So despite Randal's protests, I'm forced to conclude that his argument that testimony is properly basic is just as absurd as I had originally considered it to be. To call testimony "properly basic" is to dilute the term so radically as to void it of all relevance, and it's surely not a principle Randal himself consistently applies unless, like Edward Current, he's converting to every religion.

Addendum: A mainstream view?

In threads that have since been deleted, Randal accused me of being ignorant of philosophy in part because of my apparent lack of awareness that the view of testimony as properly basic was a pretty common view among academic philosophers. Not content to simply take Randal's testimony at face value (see what I did there?), I did a few searches across the interwebs. I searched for "properly basic belief" at the Stanford Encyclopedia of Philosophy, and got this:

Hmmm. Okay. Well, a search for "testimony AND 'basic belief'" gave up a single entry on epistemology, which has a subsection on testimony. Nothing therein refers to it as a "basic belief".

A similar search through the Internet Encyclopedia of Philosophy was slightly more fruitful....
... but on clicking the links and searching through the articles, not even the articles on religious epistemology or the epistemology of testimony had anything about testimony as properly basic.

Finally, a search through Google Scholar proved the most fruitful. But virtually all the links were to papers by theologians, not academic philosophers in general. Many, of course, referenced (and criticized) Alvin Plantinga and his reformed epistemology.  Nowhere could I find any evidence that the notion of testimony as properly basic was a widely held view.

Finally, since the very concept of "basic beliefs" derives from the philosophy of Foundationalism, I decided to peruse the PhilPaper surveys and see what percentage of philosophers subscribed to the view. Well... it wasn't even mentioned in either the survey or the metasurvey. The closest and most relevant to the epistemology of testimony was internalism versus externalism, and there was nothing resembling a consensus on the matter:

Now, I'm willing to accept that there may be some evidence I'm overlooking. But based on the evidence I've been able to ascertain so far, the notion of testimony as properly basic seems to be extremely fringe and not even particularly common among theologians, much less mainstream philosophy.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.