Foundationalism and Virtuous Circularity
You do not want to call our natural sciences as ‘knowledge’. I do not even mind that. (For instance, you might want to define ‘knowledge’ as ‘true and justified belief’. You might show me that our theories in natural sciences are only believed to be true and are not proved to be true. Therefore, you might want to say that our natural sciences are not ‘knowledge’ according to your definition. Fair enough.) Since you want to conduct a discussion about knowledge and if our sciences are not examples of knowledge, I need to know whether there are other examples of knowledge. If such an entity exists, our discussions make sense. If we have no examples of knowledge at all, according to some or another definition of the word, how could we discuss about that? I mean to say, if not even the logics we use in out argument is an example of knowledge, how could we ever have any discussion about the scope, range, etc. of knowledge? We cannot. In other words, if we want to have discussions about knowledge, we need to have examples of knowledge and need to use these examples in our discussions about their nature. So, you have to tell me: (a) whether we have examples of knowledge; (b) what they are; (c) how we should use them in our discussions; (d) show that their use does not pervert our discussions about knowledge.
Looking at the issue abstractly, such an argument might appear ‘circular’: we presuppose what knowledge is in order to know what knowledge is. A moments’ reflection should tell you what is wrong with this impression.
1. We have to assume that ‘circularity’ should be avoided. (Where does this piece of ‘knowledge’ come from, or what is its foundation? How do we know this is true?)
2. We have to assume that tautologies (‘p’ implies ‘p’) do not advance knowledge. (However, is a tautology true? How do we know this? What proofs do we have for this?)
3. We have to assume as true that there is no difference between ‘vicious’ and ‘virtuous’ circularities.
4. We have to assume that human knowledge is not ‘circular’. And so on.
How do we ‘know’ that these assumptions are true? What is ‘truth’ anyway? So, you see, if you want to have discussion about ‘knowledge’, you need to make assumptions about its nature. In this sense, there is no possibility for us to proceed from ‘first principles’.
There is something else too. Consider the problem that a computer confronts us with. A computer can read some or another programme if and only if the operating system interprets the language of the programme. The operating system can do that if and only if the computer hands over the control to the operating system. The computer can do that if and only if the CPU has an interpreter that allows it to interpret the language the operating system is compiled in. This interpreter must itself be written in some language or another. So, we need another interpreter to interpret the interpreter. That should also be written in some language or another and therefore … and so on. So, we can conclude this is a circular problem, which has no resolution. But computers do boot, they do load the operating system and the latter does hand over the control to the programme that is called. This process is called “boot-strapping” that solves ‘circular’ problems. This should also tell you that ‘circularity’ appears as a problem when you assume that each sub-step in a complex task requires the same kind of intelligence that the entire task demands. (It is to the credit of Artificial Intelligence to have solved some of the long-standing philosophical problems, including the problem of ‘object recognition’. To recognize an object, we need to have a ‘copy’ of the object in mind; to recognize the copy of the object, we need another copy… and so on. AI shows that we can split up an intelligent task into a series of stupid moves and generate the desired result.) Do we assume these ‘facts’ (like running MSWord on a computer) as true in our discussions about ‘knowledge’ or not? If yes, are we not being ‘circular’? If no, why not?
[Check Harold Brown’s Circular Justifications ]