A while ago, I wrote a post explaining why I didn't think Clark's position that "words are arbitrary signs" which "tag thoughts" made sense. I was thinking about this today and believe I have found a different way to phrase my problem with it. Firstly, however, I want to say that the point of that post and this post was and is not to refute Clark so much as develop a theory of language more consistent with his broader philosophy.
For instance, I agree with Clark that propositions are the objects of knowledge. Everything we "know," philosophically speaking, is propositional. Clark stated that a proposition is the "meaning of a declarative sentence" (Logic, pg. 28). But what is meaning? Or more to the point: at root, is that which we philosophically "know" wordless? No. Language is a precondition for knowledge. Knowledge comes from communication with God. But the implications of Clark's statement that words are "mere symbols or signs" (cf. here) seems to lead to the conclusion that our knowledge is, at root, wordless.
Suppose words are, as Clark said, arbitrary signs. Now consider any declarative sentence. The words in that sentence are not, if arbitrary, intrinsically meaningful; each of those words "tag" something else. If they "tag" other words, we must remember that the words which are tagged are likewise arbitrary by Clark's definition, so those words would in turn "tag" something else. This process can either continue ad infinitum or terminate on a wordless Ding an sich. In either case, whence comes meaning? What is the proposition?
Clark rejected the empirical version of the correspondence theory of truth, where the object of knowledge is "a representation and not the reality itself" (Language and Theology, pg. 29). I agree with him. I think the reality is the object of knowledge and the physical correspondent is the representation. But then again, signs and symbols are themselves intermediaries of sorts. I believe Clark himself made the point somewhere in God's Hammer that symbols must symbolize something. What is it, then, that Clark think words arbitrarily signify or symbolize? Is it something "physical"? Then it would seem Clark's theory of language relies on the very correspondence theory of truth he opposed. Is it something "spiritual"? Then it would seem Clark's theory of language logically leads to mysticism, which he also opposed.
Thus, while it may be the case some words can be "arbitrary" - that is, they are assigned meaning by "tagging" other words - it appears that other words - the words to which the "arbitrary" words must eventually be traceable - must necessarily be intrinsically meaningful. As such. they would not be signs or symbols.
I realize this post may leave much to be desired, and that is because the philosophy of language is highly complex for me. To show my awareness, I wish to explicitly state that I do not here intend to comprehensively address 1) which words are intrinsically meaningful, 2) the relation between words which are intrinsically meaningful and the fact knowledge is proposition rather than conceptual, nor 3) the fact that despite the intrinsic meaningfulness of certain words, to know one proposition requires knowledge of other propositions. I think these are "problems" either capable of resolution or merely the result of ill-definition, for which I fully bear the responsibility.
To alleviate some of that responsibility, I would, in brief, argue that 1) isn't a problem because I don't see why I need to know which words are intrinsically meaningful, 2) isn't a problem because what I mean when I assert that some words are "intrinsically meaningful" is nothing other than that some subjects have necessary predicates (predicates which, when are subject to definition, also have necessary predicates), and 3) isn't a problem both because of my answer to 2) and the fact that not all knowledge is tautologous (thus, knowledge of predicates is to a certain extent necessary in order to know a subject). But I will give all of this more thought.