Bernard, Ben, and list: I am still working on the question of what, if anything, is wrong in my account of the ten sign classes (as resulting from the cross-combination of the three basic sign trichotomies) in my paper on Peirce's semiotic in the Encyclopedic Dictionary of Semiotics (commissioned and edited by Tom Sebeok and Umberto Eco), originally published in 1986 and presently available in a revised version at Arisbe:
http://members.door.net/arisbe/menu/library/aboutcsp/ransdell/eds/eds.htm >From what I have been able to figure out thus far, there is one important error in the original version which I eliminated in the revised version, so there is nothing formally wrong with the version at Arisbe as it presently stands. There is, however, one "infelicity" -- which is a euphemistic way of talking about something that might well be misleading even though not formally erroneous. My elimination of the formal error in the revision was, I think, only a fortunate accident since the corrective move I made was made only in order to eliminate something from the diagram which I adjudged to have no real role to play in anything that I said in the rest of the paper. In short, you were right in sensing something wrong there, Fortunately, it is not the catastrophic blunder I feared that it might be. Let me explain what I did -- or didn't -- do or say that was misleading and, in the original version, strongly so, though considerably less so in the revised version that has been available for the past six years or so. In the original version I did not use a tree diagram but rather a tabular form which is equivalent to and easily transformed into a tree diagram: (1) qualisigns: (i) icons: (i) rhemes (I) (2) sinsigns: (a) indexes (including symbol replicas): (i) rhemes (II) (ii) dicisigns (III) (b) iconic signs: (i) rhemes (IV) (3) legisigns-- (a) symbols: (i) rhemes (V) (ii) dicisigns (VI) (iii) arguments (VII) (b) indexical signs: (i) rhemes (VIII) (ii) dicisigns (IX) (c) iconic signs: (i) rhemes (X) Notice that if you read across the lines with Roman numerals at the end and simply collapse the table appropriately into ten corresponding lines you get a list of the ten classes of signs: qualisigns icons rhemes (I) sinsigns indices rhemes (II) sinsigns indices dicisigns (III) sinsigns icons rhemes (IV) legisigns symbols rhemes (V) legisigns symbols dicents (VI) legisigns symbols arguments (VII) legisigns indices rhemes (VIII) legisigns indices dicents (IX) legisigns icons rhemes (X) Now, this does indeed list out the ten classes according to their differing three-component combinations and is not mistaken in itself. However, when we notice the correlation with the ten roman numerals we find the important mistake, which is owing to the fact that in the passage from the Syllabus of Logic that this is based upon Peirce himself used Roman numerals to number the classes and he numbered them differently. The proper numbering, following Peirce, would rather be: qualisigns icons rhemes (I) sinsigns indices rhemes (III) sinsigns indices dicisigns (IV) sinsigns icons rhemes (II) legisigns symbols rhemes (VIII) legisigns symbols dicents (IX) legisigns symbols arguments (X) legisigns indices rhemes (VI) legisigns indices dicents (VII) legisigns icons rhemes (V) That gives us Peirce's ordering both in the diagram of the ten-box triangle at CP 2.264, where Peirce inserts the Roman numerals in the boxes, and in the several pages just prior to that where he gives paragraph-long descriptions of each of the ten classes, wherein he does not use Roman numerals but does use ordinal English numbers (first, second, etc.). Thus the ordering I suggested with my numbering in the original version was simply mistaken insofar as it suggested that Peirce ordered them numerically in that way, which he clearly did not. My numbering there was not absolutely mistaken because neither way of ordering them makes any difference as to what the ten classes actually are, as regards their differing defining elements. But still, it was clearly a mistake. That mistake was corrected in my revised version when I simply omitted the numbering of the classes after noting that there was no need to include them since I made no use of the numbers in that paper. But still, it could be misleading in case someone were to mistakenly think that the tree diagram which I used to replace the tabular diagram used in the original version ought to present them in a diagram which, read top-down. would give you Peirce's own numbering. That probably wouldn't affect many readers, but still, it is an undesirable possibility. But there is one other infelicity which I think might actually induce some confusion because I think it has confused me at one time and another, namely, that I also reversed the order of listing the three components moving horizontally. Thus where I list the constituent components of the first class as follows: qualisign icon rheme Peirce reads them in the reverse order: rheme icon qualisign That is, he does so when he names them (e.g. "rhematic iconic qualisign"). Now, I believe he reads them in that order because it is more awkward in English to say them in the other order. That is, it is natural to say, for example, "rhematic indexical legisign" but very forced and awkward to say "legisignal indexical rheme". (In fact, he would have had to abandon the neologisms "qualisign", "sinsign", and "legisign" for that reason alone, in my opinion, and that would have been unfortunate since that set of three names is actually one of his more inspired neologisms, nearly as good as his decision to use "icon", "index" and "symbol".) Or maybe there was some other reason for him to decide to name them in that order of their elements. But it should be noted that in the paragraphs in which he describes the combinations that define the classes he usually begins by speaking of the sign's mode of being and proceeds from that in explaining the categorial constraints and possibilities that follow from that. For example, in explaining the rhematic iconic qualisign, he doesn't begin from it being a rheme but rather from it as being a qualisign, explaining that this necessitates it being an icon and further necessitates it being a rheme (CP 2.254). See CP 2.254-263 for his descriptions of the classes. It was natural, then, for me to take the proper horizontal order to be from the first to the third rather than the reverse of that when I constructed the tree diagram in which the ten classes are represented in that way, and the possibility simply did not occur to me that Peirce's order insofar as that is represented in the triangle of ten boxes is actually the mirrored (hence reversed) image of that, given the numbering order he put in those boxes. I am uncertain at this point what to do about that. There is nothing in the substance of my account (as regards the present question) that I see any need to change, so it is a matter of replacing the tree graphic by one with a rectified vertical order to keep it from being misleading in that respect. But should the horizontal order be changed, too? That is not so clear to me yet. Maybe a sentence or two in explanation of how to avoid being misled by it when one turns to the way Peirce does it in the Syllabus will do it. Or maybe it is not worth explaining there, everything considered. But I am not finished with this matter yet, so this is just an interim report on what I have come up with so far. Joe Ransdell -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.394 / Virus Database: 268.9.0/367 - Release Date: 6/16/2006 --- Message from peirce-l forum to subscriber archive@mail-archive.com
