Convergence

       Now, at long last, I am ready to begin explaining why faith does not work as means of obtaining knowledge. I apologize for the lengthy detour, but I did not want to skip over the foundations and risk confusing you. The way in which I will argue, and the points I will make, are very different from what you are likely to have heard before. This is partly because anyone who wasn’t convinced by an argument the first time is unlikely to be convinced the second, third, or even hundredth time. But it is also partly because I am going to be avoiding hazardous terms, words with very strong connotations but not much agreed upon meaning.

       Trying to argue that the conclusions of science are true and those of faith are typically false would require me to pin down what I mean by true and false, as both are hazardous terms when discussing religion.  This would probably result in definitions that appear to beg the question, and responses along the lines of “that’s not what ‘true’ means!”  So I’m going to avoid all of that.  I’m going to argue that science is a better conclusion-seeking method than faith, and I will do so without using words like true, false, or reality.  Instead, I’m going to reference a number of desirable properties that science and its conclusions have and that faith and its conclusions lack. This will make my arguments more resistant to such verbal shenanigans as re-defining truth or moving the definitional goalposts.

I will begin with a desirable property of conclusion-seeking methods I call convergence.  I’m going to first define convergence in a way that will sound like it’s coming from a mathematics textbook, in order to be precise.  Then I will present a diagram and describe in more accessible terms what that definition generally means.

Convergent: A finite* method M is convergent if it meets the following:  Given a question Q and any two potential beliefs B and C regarding Q, applying M at B will reach result R if and only if applying M at C will reach result R.

The basic idea is that when you use a convergent method to answer a question, you will obtain one particular answer regardless of where you started.  In other words, your conclusion does not depend on what you believed when you began investigating the question.

Keep in mind, convergence has nothing to do with how fast the conclusion is reached.  It may be that Charlie with belief C will reach R in a matter of seconds, while Bob with belief B will take ten years.  But whatever the time difference, the property of convergence guarantees that Bob and Charlie will eventually agree.

One side-effect of convergent methods is that two people using a convergent method cannot “agree to disagree.”  If at any point they disagree, a simple application of their shared method will (eventually) result in an agreement.  We see this quite often in science.  When two scientists disagree about something they study, they do not simply say “well you have your opinion and I have mine.”  Instead, they apply the scientific method in a way that resolves the disagreement.  And after this application, the two scientists will agree.  This is how we have physicists from all over the world, with hundreds of different cultural and political backgrounds, all reaching agreement over the validity of the theory of relativity or the violation of Belle’s inequality.  Science is a convergent method, so scientists tend to reach agreements over the questions they pursue.

Compare this to what has historically happened with disagreement in religions.  Instead of resolving the disagreement and presenting a united front, many disagreements eventually result in a splitting.  This is where we get reformed versions of religions while still having the old versions around.  It’s why there are Catholics and Protestants, or even Christians and Jews and Muslims.  Faith isn’t convergent, so different people applying faith from different starting points will reach different conclusions, giving us a plethora of disparate religions that disagree with one another.  So while physicists and chemists both agree about the validity of physics and chemistry, Christians and Muslims largely disagree about the validity of Christianity and Islam.  It’s one of the reasons we have Muslims calling for the deaths of Christians but we don’t have chemists calling for the deaths of physicists.

Scientists don’t settle their theories over Jihads.  They don’t talk about tolerating each other’s different scientific views or accepting other people’s opinions on scientific matters.  They don’t insist on respecting all scientific beliefs.  They don’t do any of these things because they have a method that resolves their conflicts by converging to a shared conclusion.  So while religious disagreements can result in centuries of animosity or even outright war, most disagreements in science result in one scientist saying “Hey guys, turns out I was wrong.  Glad we figured that out.  Let’s move on.”

Convergence also generates (and perhaps is equivalent to) objectivity.  I’ve heard a lot of (usually religious) people complaining about scientists being arrogant enough to present their conclusions as objective.  But arrogance has nothing to do with it.  Scientists say that their conclusions are objective because their method is convergent.  You can do the experiments in India or China or England or Mexico and you’ll get the same conclusion.  You can hire a team of Germans or Australians or Hungarians or Brazilians and you’ll get the same conclusion.

In fact, science has a series of built-in checks to ensure convergence.  Independent confirmation is a big thing in science.  When you do an experiment, it’s expected that other groups will be able to perform the experiment too.  All over the world we have groups attempting to verify other groups’ results.  In addition to serving as a check against honest mistakes in procedure, independent confirmation serves to ensure convergence and objectivity.  Thus convergent behavior is built into the scientific method in a way you’re unlikely to ever see in any type of religion.




*For infinite or infinitely recursive methods, you have to use the concept of a limit (or an epsilon-neighborhood, which even fewer people know).  Basically, the limit as you apply the method indefinitely from B is the same as the limit as you apply the method indefinitely from C.