I’ve long had an issue with the idea that beliefs can be justified simply by virtue of them fitting into an overall coherent worldview.

This can be seen in practice by the number of coherent yet mutually exclusive philosophies, theory-laden political ideologies, and worldviews that surround us. Taken alone, they back themselves up. Held up to the light of reality, they have no foundational premise on which to deserve the label “True”.

In a recent post on Fake Naus, Michael Huemer, Professor of Philosophy at the University of Colorado, explores this idea in more detail, taking “the Coherentist view to be that coherence alone suffices to justify our beliefs, without the need for any degree of foundational justification”, pointing out that this endorses circular reasoning [it is also the basis for most conspiracy theories, and endless efforts of Coherence-making from organized religion].

The Probabilistic Defense

A probabilistic argument can be put forward to defend Coherentism. A very abridged version of the argument goes along the lines of: the more people that claim X, the more likely it is that X is true. “This is true even if you antecedently had no opinion about how reliable the witnesses were”. On a parallel: the more of our beliefs that happen to coherently fit together, the more we can assume they resemble truth.

Huemer’s objection to this line of reasoning is that we need at least some foundational justification for, say, the reliability of the believers, or witnesses. Without some foundational bedrock, a Coherent worldview may as well be a work of fiction.

A large number of witnesses who tell the same story simply isn’t enough. In probabilistic terms, why should the likelihood of those witnesses telling the truth be higher than the likelihood of them telling a lie? How can we take those witnesses as a reliable source of knowledge? As Karl Popper argues: Induction can never yield certainty, because there is always the chance for an observation to be made that refutes the testimony of a large number of people.

This relates to Stewart Cohen’s “KR principle”; A potential knowledge source K can yield knowledge for S, only if S knows K is reliable.

In other words Coherentism cannot be reliably separated from Foundationalism without falling victim to circular reasoning.

Where This Leaves Us

My issue here is that a foundational belief must also be justified by something other than faith. Therefore, although Coherentism cannot be a replacement for Foundationalism, its failure is not an endorsement of Foundationalism.

I do not yet have a strong opinion on where this leaves us. I am drawn to Karl Popper’s idea of falsification. This is the idea that no matter how many times an argument is said to be true, just a single case of falsification is enough to label that theory wrong, which should prompt a new theory to pop up in its place, a la Bayesian thinking.

With this method, we can gradually become less wrong and more coherent, without ever knowing we have fully arrived at truth. We cannot prove we’re right, but we can prove we’re wrong. That’s something to work with.