Note also that what makes the contents of minds unacceptable as meanings is their inherent non-shareability; thus an alternative approach might be to develop a theory of mind which would take mental contents to be not inviolably private3.
In order for it to mean something, it is not enough that each of them individually makes the association, he/she must also know that the others do the same, that he/she can use the word to intelligibly express its meaning in various public circumstances etc. Language is essentially public; and as such it cannot rest on private associations.
It makes no sense to say that what we subject to rules are already pawns, bishops etc. – the pieces acquire the values via being subjected to the rules.
It makes no sense to say that what we subject to rules are already meaningful words – the words acquire the meanings via being subjected to the rules.
Hence the value of a piece and our normative attitudes to the way it is treated are two sides of the same coin - it makes no sense to say that something is, say, a king independently of the attitudes - to be a king is to enjoy this kind of attitudes.
It makes no sense to say: "What you can check is obviously a king not a mere piece of wood – hence you cannot formulate rules of chess unless you have pieces which already are kings, pawns, bishops .
Hence the meaning of a word and our normative attitudes to the way it is treated are two sides of the same coin it makes no sense to say that something is, say, a conjunction connective independently of the attitudes - to be a conjunction connective is to enjoy this kind of attitudes.
It makes no sense to say: "What you can assert is obviously a meaningful sentence not a mere meaningless sound/inscription – hence you cannot establish rules of language unless you have expressions which already are meaningful."
As Quine showed, the holistic character of language makes it impossible to distribute the relatively clearcut boundary between semanticsL and semanticsW to individual statements and expressions in any unique way; and this makes the boundary between meaningL and meaningW of an individual linguistic item rather illusory.
Thus, semantic analysis is always ultimately a matter of translating the language that is to be analyzed into another language - it makes sense if the latter is in some relevant sense more perspicuous than the former.