Roland's Razor

In an online discussion group we stumbled on something that just seems like common sense but which none of us have come across anywhere else and which has come to be dubbed "Roland's Razor" by Paul.  I think he just liked the alliteration but thanks anyway.

If Proponent_a uses an argument to come to Conclusion_a and Proponent_b uses the same argument to come to Conclusion_b and Conclusion_a and Conclusion_b are incompatible or mutually exclusive the the argument itself is not able to be used to distinguish between Conclusions_a and Conclusion_b. It is not that the argument is invalid, it can be a perfectly valid argument, but if it can be used to come to mutually incompatible conclusions then the argument can't be used to distinguish between the conclusions.

By way of example, suppose that a party is trying to decide between god_a and god_b both of whom make the claim that they are the one and only god. So Proponant_a makes the argument that "She prayed for someone to be healed, and she was healed therefore god_a exists" while Proponant_b makes the argument that "She prayed for someone to be healed, and she was healed therefore god_b exists". For the party trying to decide between god_a and god_b the argument "She prayed for someone to be healed and she was healed" can't be used to decide between god_a and god_b.

Roland's Flock

Lately, I have been coming across references to Gettier and his article "Is Justified True Belief Knowledge?" and what this has to say about a proposition being absolutely true.  The example below isn't the one's involving Smith and Jones in Gettiers article but, I think, it illustrates the same point.

An Observer_a views a paddock and sees a sheep and makes a statement to the effect that "There is a sheep in the paddock" given the evidence that she is able to see that there is a sheep in the paddock. Unbeknownst to her what she is looking at is a target that looks like a sheep but there happens to really be a sheep in the paddock, just not where she is looking. In this case, she has expressed a justified, evidence courtesy of sight, true i.e. there really is a sheep in the paddock, belief. She is not wrong in her statement but the reasons she uses to justify the statement are divorced from the truth of the statement.

For me, this calls into question how we can use evidence, sight in this case, to make pronouncements on what is true or not; are we left in a quandary where we can't know anything because we have to distrust any evidence provided as a reason for justified belief?

I would like to extend the example a little at a time:

Another Observer_b views the same paddock and also sees a sheep and makes a statement to the effect that "There is a sheep in the paddock" given the evidence that she is able to see that there is a sheep in the paddock. Unbeknownst to her what she is looking at is a wolf in sheep's clothing but there happens to really be a sheep in the paddock, just not where she is looking. In this case, she has expressed a justified, evidence courtesy of sight, true i.e there really is a sheep in the paddock, belief. She is also not wrong in her statement but the reasons she uses to justify the statement are again divorced from the truth of the statement.

And so on but no matter how much evidence piles up from different observers and different perspectives the justification of the statement that "There is a sheep in the paddock" still doesn't lead to proof that there actually is a sheep in the paddock as it is all co-incidentally true rather than causally true.

There is an implied third observer that exists in the examples above which is the Observer_n that is making the statement "There really is a sheep in the paddock but not where the other Observers are looking". So how do we know that the third Observer_n themselves aren't mistaken and, in reality, there may or may not even be a sheep in the paddock.

Now we have called into question not only evidence as a means of determining if something is true but whether it is even possible to know if something is true in an absolute sense.

My take on this is that we have no way of determining what is absolutely true because it seems to me that in order to do so one would need to view the paddock from all possible perspectives and through the lens of all possible evidence(s) and that, as humans, I don't think is possible. Perspectives and evidence that are impossible at the current time to access could very well become possible in the future but we are only able to work with what is currently possible and available to us at the moment.

As uncomfortable as that may make us I don't think that all is lost and that the example points a way to, while not being certain, at least being confident, that there is or is not a sheep in the paddock by combining our perspectives and evidences so that we can state with confidence that "There is a sheep in the paddock" even though we can't be absolutely certain that there is one there.

As new evidence or perspectives become available to us we can revisit the question and potentially, the answer may change. So in the example Observer_a,b,n get together and walk around to their respective perspectives to see how things look. When gathered at Observer_a's perspective the target will look like a sheep, the actual sheep looks like a sheep but the wolf in sheep's clothing becomes clear for what it is. When gathered at Observer_b's perspective the wolf will look like a sheep, the actual sheep looks like a sheep but the target becomes clear for what it is. Finally, when gathered at Observer_n's perspective the sheep looks like a sheep but the target and the wolf in sheep's clothing are clear for what they are.

In the example, Observer_n's perspective isn't in some way a privileged perspective but it is the one that what is observed remains a sheep from both Observer_a's, Observer_b's and Observer_n's perspectives but the introduction of Observer_c's perspective could potentially show them to all be wrong but not that there really is a sheep in the paddock! 

So as counter-intuitive as it feels, it seems that we are able to "know an absolute truth" i.e. "There is a sheep in the paddock" but unable to know whether "There is a sheep in the paddock" is an absolute truth or not. 

Comments 

 A friend pointed out that "It seems to me that you're arguing for corroboration as a form of validating what we take to be true" which isn't quite what I was intending but I think I need to explore the point a little bit.

On the "Yes" side of corroboration as a form of validation I was pointing out that having Observer_a, Observer_b leave behind their perspectives to take on Observer_n's does strengthen Observer_n's claim that there is a sheep in the paddock and that it can be a justified belief albeit not a proven belief.

On the "No" side, the competition is between perspectives and not about the number of people sharing the same perspective but, that being said, perspectives are only held by people so, in the end, numbers do seem to count by virtue of holding a particular perspective and not another. Numbers are useful in determining the dominant or consensus perspective but not when it comes to determining which perspective, if any, is absolutely true.

Come to think about it, I think that we mostly base our beliefs on consensus perspectives simply because it is not possible for us to each be a cosmologist, a philosopher, a biologist, a mathematician, a physicist etc, etc so we use a heuristic of consensus on a perspective without going any deeper into the topic.