Thereafter, I showed how the greatest part of the matter of this chaos must, in accordance with these laws, dispose and arrange itself in such a way as to present the appearance of heavens; how in the meantime some of its parts must compose an earth and some planets and comets, and others a sun and fixed stars.
Nevertheless, because all participants of the games are themselves describers, these tools are also themselves, in a sense, part of the game.
Part V I would here willingly have proceeded to exhibit the whole chain of truths which I deduced from these primary but as with a view to this it would have been necessary now to treat of many questions in dispute among the earned, with whom I do not wish to be embroiled, I believe that it will be better for me to refrain from this exposition, and only mention in general what these truths are, that the more judicious may be able to determine whether a more special account of them would conduce to the public advantage.
Anyway, it seems to be quite clear that what is in the province of a linguist or a philosopher of language is meaningL, not meaningW: the project of discovering who is the present king of France, required in order to determine the meaningW of the expression the king of France and hence belonging to the project of semanticsW, is clearly not a part of the semantic theory of English.
Meaning, in the ordinary sense of the word, is a matter of semanticsL - knowing meaning is a part of knowing language, not of knowing facts about the extralinguistic world.
To assess the adequacy and reasonability of a diagram used to pursue semantic analysis we thus should not try to probe the speaker's and hearer's minds to find out whether we glimpse something which could be pictured by the diagram, but we should rather consider the following two points: (A) Is the analysandum adequate to the analysatum, does the inferential role of the former within the analyzing language 'reasonably approximate' that of the latter within the analyzed one?; and (B) is the inferential role of the analysatum, as a part of the analyzing formal language, in some sense explicit?
The approaches we classify as semiotic rest on the assumption that the significance or meaning of an expression does not depend on whether the expression is a part of language or of some other system of signs.
The basic principle of the structural approach to language, on the other hand, consists in the conviction that any significance, or any meaning of an 71 expression comes to it from its being part of the system of language.
This is the case of Quine's (1960) illustrious example of a rabbit and its undetached part.
If we accepted the semiotic view, then we could not see any problem in determining whether 'N' refers to A or B even if A and B had precisely the same properties; 'N' would simply unequivocally refer to that object which took part in the relevant process of semiosis.