Benchmarks

As I mentioned last week, being a skeptic is really quite simple.  It’s a one-step method, and a crucial part of a good, self-correcting epistemology.  Yet there is one glaring problem.  To see this problem, imagine that the entire world was completely skeptical in every area.  Imagine that everyone followed the basic principle of asking a question whenever they were confused or doubtful.  Not imagine how many questions that entails.  Billions at least, maybe even trillions depending on just how curious people get.  Now who’s going to answer all those questions? 

And more importantly, how?

The problem is that skepticism is not enough.  Skepticism is all about figuring out the right questions to pursue, and it achieves this function quite well.  But once these crucial questions are identified, skepticism leaves you in the dark.  It’s like giving you an address without a GPS.  Skepticism will tell you where you need to go, but not how to get there.  It’s a necessary first step, but it’s not the whole story.  Once you’ve become a skeptic, the next step is to figure out just how to answer all those questions.  That’s where the benchmarks come into play.

Benchmarks are the standards by which we evaluate things, and different people often have different benchmarks.  I won’t yet delve too deeply into the methods for determining which benchmarks we should use, as that subject will take several articles.  In this article I will assume that you already have some benchmarks (and you do, you just might not have articulated them).  Once you have these benchmarks and you’ve identified the important questions through skepticism, the next step is to apply your benchmarks to these questions.  This process (which I shall expand upon below with the help of some examples) is easiest to understand when your question is simply “Is Claim X true?”  In this case, you will obtain one of six results.

Pass

Claim X passes your benchmarks.  This means that your benchmarks indicate that Claim X must be true.  Of course, there’s still the possibility that you have lousy benchmarks.  But assuming a perfect set of benchmarks, a Pass result means that X is undeniably true.

Good

Claim X is supported by your benchmarks.  This means that your benchmarks indicate that Claim X is probably true.  While you still won’t be completely sure that X is true, if you are confident in your benchmarks, a Good result indicates that X is likely.

Shrug

Claim X is ignored by your benchmarks.  This means that your benchmarks can’t even tell you whether X is likely (or they indicate a 50/50 chance of true/false).  A perfect set of benchmarks will never Shrug, but a Shrug does not mean the benchmarks are useless.  Often a Shrug simply indicates that your benchmarks are incomplete.

Bad

Claim X is doubted by your benchmarks.  This means that your benchmarks indicate that Claim X is probably false.  While you still won’t be completely sure that X is false, if you are confident in your benchmarks, a Good result indicates that X is unlikely.

Fail

Claim X fails your benchmarks.  This means that your benchmarks indicate that Claim X cannot be true. Of course, there’s still the possibility that you have lousy benchmarks.  But assuming a perfect set of benchmarks, a Fail result means that X is undeniably false.

What?

Claim X doesn’t make sense.  This is not quite a Fail result in that, strictly speaking, it does not mean that X is false.  This result can indicate a problem with your benchmarks, but it may also indicate that X is poorly formatted, based on false assumptions, or simply too vague.

In short, your benchmarks determine when your doubts are justified, and when they are not.  Moreover, much confusion is simply the result of a statement obtaining the What result on a subconscious level.  As I said before, evaluating the benchmarks themselves is the subject of another note.  For now, I’m going to list a few common benchmarks, as well as their pros and cons.

Logic: Logic forms the backbone of philosophy and mathematics.  Most people recognize logic as a very good benchmark to use in most scenarios, though many will also claim that it doesn’t work when dealing with emotional matters like love.  Logic is highly formal and best expressed symbolically.

Pros:    Logic is very self-consistent.  When seen to be applicable, it is often recognized as the most reliable of benchmarks.  It is also very good at determining conclusions from a given set of known facts.  If you have a set of facts to start with, logic is very useful for locating new facts.

Cons:   Except in pure math, logic is almost useless by itself.  This is because logic is virtually incapable of figuring out where to start.  Unless you’re willing to assume some basic axioms, pure logic is unlikely to get you anywhere.  Many people throughout the ages have found that trying to dismiss the credibility of everything not strictly logical is a quick way to build a non-functional epistemology.

Empiricism: Empiricism is famous for its use in science.  While logicians may sit in a windowless room and argue whether or not leaves are green, an empiricist will go out and look at a leaf.  It makes a lot of intuitive sense, which may leave you wondering why anyone would bother with logic.  As it turns out, pure empiricism is about as useless as pure logic, but in the opposite way.

Pros:    Empiricism is consistent with observations, being built from them.  It is also widely recognized as a strong benchmark.  Whenever you try to convince someone of something by showing it to them, you are appealing to empiricism.  It also turns out that empiricism is very self-consistent, which seems to indicate that the universe is self-consistent. 

Cons:   While seeing is believing, pure empiricism results in a set of largely unconnected facts.  For instance, while walking through a field, empiricism will tell me that leaves and grass are both green.  What is not readily available through empiricism is why they are green, or whether the fact that both are green is significant in some way.  Furthermore, pure empiricism is largely incapable of making predictions.

Intuition: I am going to use the term intuition as a catch-all for any subconscious process which arrives at conclusions about truth.  In this sense, intuition also covers instinct, dreams, muscle memory, and any subconscious psychic ability people may claim to have.

Pros:    Intuition is fast.  You don’t have to think on a conscious level to use intuition.  This makes it incredibly useful when pressed for time.  It also works very well as a sort of guide to use when other methods have failed.  It makes a useful backup benchmark to use when you need speed, or when your other benchmarks Shrug.

Cons:   Intuition is typically built up through experience, so when you have no experience, your intuition is largely useless.  This is especially true in the more abstract and remote aspects of inquiry.  For instance, topology and quantum mechanics are often cited as fields where a newcomer’s intuition can lead them wildly astray.

Faith: By faith, I mean the kind of thing that religions typically champion.  I’ll be devoting an entire article to articulating what precisely I mean by this.  For the present purposes it is sufficient to say that belief without evidence/reason only one of the criteria used to identify religious faith.  This sort of belief is often called faith, but there are other aspects of faith as used by religion which make it fundamentally different from just belief without evidence/reason.  Lastly, it’s important to note that faith itself is not actually a benchmark.  You can’t just have faith.  You have to have faith in something.  So really, the benchmark is “Faith in X.”

Pros:    Faith can often be every bit as fast as intuition.  When someone tells you something and you decide to have faith in it, you don’t have to apply any more effort to assess the statement.  There are also several theologians who have used faith (usually faith in God) to resolve the “Can we ever really know?” questions that keep cropping up in philosophy, which may let people feel more comfortable about such things.

Cons:   Faith in X is only as good as X.  Thus in order for faith to give you the truth, you have to have faith in the truth, which in turn requires you to already have the truth.  In this sense, Faith isn’t actually an epistemology because it doesn’t allow you to start from a lack of truth and arrive at the truth.  In other words, it’s not really a method for getting at the truth.  The crux of the problem can be seen by asking “Why X and not Y?”

There are a few more common benchmarks that may be more suitably called meta-benchmarks, as they are quite useful in assessing the value of the above benchmarks.  Since this article is already longer than any of its predecessors, I will stop here.  I’ll return to the meta-benchmarks when I address the issue of evaluating your benchmarks.