Truth is coherence
This view is expressed by the coherence theory of truth.
The correspondence theory of truth is problematic because one cannot step outside their own mind and see the facts. One cannot take their beliefs on the one hand and facts on another and check if they correspond. Therefore, a worry exists that correspondence theory makes truth inaccessible to us.
For the coherence theory of truth the "no access" problem is not a problem since the truth conditions of a proposition are other propositions, namely we can check if two beliefs are consistent and whether they have explanatory connections to one another. There is no need to step outside one’s mind and check what the facts are. [1]
The Coherence Theory of Truth. Retrieved October 30, 2024. ↩︎
In practice, how do we proceed when asked what makes a given proposition true? For instance:
P) Joe Biden is President of the US.
Assume P is true. What makes P true? In everyday discourse rather than philosophical one, we would answer by citing other propositions that are connected to P. For example, one would describe the system of the US government, the outcome of the 2020 election, the person referred to by the name "Joe Biden", and so on. Thus, one would not simply “display the facts”. [1]
The Coherence Theory of Truth. Retrieved October 30, 2024. ↩︎
One of the criticisms to the coherence theory is that truth and well-elaborated fictions seems to collapse and that there is no way to distinguish the truth from a well-elaborated fiction:
A Sherlock Holmes novel could have a high degree of internal coherence, but it isn’t close to the truth. One could extend the story indefinitely, but this would not make it any closer to the truth.
To this a coherentist would respond that coherence needs to be understood as including not just propositions and beliefs but other representational states, like our perceptual states. The propositions that coherentists accept are constrained by our experiences and our perceptions. For them this is the difference between theory and fiction: a theory must answer to the tribunal of experience, there is no such requirement for fiction. The Sherlock Holmes story is not coherent with relevant perceptions: if you go to 221B Baker Street, you won’t see a detective living there.
When one writes a fiction, he states propositions that contradict representational states found in experience. Because of that, fiction can never be ideally coherent. Therefore it can’t be true. [1]
The Coherence Theory of Truth. Retrieved October 30, 2024. ↩︎
One could take a proposition P like "snow is blue" and adjust other propositions about the world so that the proposition P is part of a coherent system. If this is the case, then on coherence theory it looks like almost any proposition can be made to come out as being true.
A Sherlock Holmes novel could have a high degree of internal coherence, but it isn’t close to the truth. One could extend the story indefinitely, but this would not make it any closer to the truth. In coherence theory, the distinction between truth and a well-elaborated fiction seems to collapse.
Even worse, it looks like the coherence theorist actually has no way to distinguish the truth from a sufficiently well-elaborated fiction. [1]
The Coherence Theory of Truth. Retrieved October 22, 2024. ↩︎
A coherence theorist will be committed to some sort of relativism due to the many-systems problem. A lot of philosophers take that to be a problem. [1]
The Coherence Theory of Truth. Retrieved October 30, 2024. ↩︎
Coherentists could make the following argument against the many-systems problem. Namely, that there won't be an endless variety of different coherent systems because part of what is required from coherence is going to include our experiences and perceptions. This is going to significantly narrow the available coherent systems.
However, it seems unlikely that even this requirement is going to deliver a unique coherent system – a single coherent system. The undetermination is a significant problem in philosophy of science and other areas of philosophy. There are going to be and there are multiple theories that are consistent with our perceptions.
Another issue is that experience itself is theory-laden. There are going to be many different ways of describing what it is that is given in experience. Consider many different ways of classifying color experience. Different languages carve up color in many different ways. There are languages in which blue and green are classed as the same color. There are languages where what we call dark blue is grouped with black. If you ask "what is the color of this coup?", you are going to get different answers depending on how you are carving up the color space.
Experience doesn’t provide a fixed foundation. There are going to be multiple different ways of describing what it is that is given in experience. Therefore, the many-systems problem remains a problem. [1]
The Coherence Theory of Truth. Retrieved October 30, 2024. ↩︎
A necessary condition for coherence is consistency. A set of beliefs is consistent just in case it doesn’t entail a contradiction.
What is entailment? The standard answer:
P entails Q if and only if it is impossible for P to be true and Q to be false.
At this point, coherence theory faces a circularity problem: coherence = truth, coherence depends on entailment, entailment is defined in terms of truth. One needs to have some prior conception of truth in order to say what entailment is.
The coherence theorist might define entailment in terms of inference rules:
P entails Q if and only if Q can be derived from P via the application of one or more inference rules. For example we can think of inference rules like conjunction elimination or modus ponens. Using these rules we don’t need to be concerned about what the truth is, we can just follow these mechanical rules.
However, how can we distinguish between correct and incorrect rules? In coherence theory, an inference rule is correct only if it is part of the coherent system of beliefs…
An inference rule counts as correct, only if it is part of a coherent system, but what makes the system coherent is going to depend on the inference rules used to determine what follows within the system. This again leads to circularity.
To quote Bertrand Russell:
The laws of logic supply the skeleton or framework within which the test of coherence applies and they cannot themselves be established by this test.
Since the laws of inference need to be justified independently of coherence. From this perspective, the coherence theory is falsified. [1]
The Coherence Theory of Truth. Retrieved October 30, 2024. ↩︎