Relation is a very problematic category. Aquinas says (Summa Theologica, 1 28.2):
In the genera, apart from that of 'relation,' as in quantity and quality, even the true idea of the genus itself is derived from a respect to the subject, for quantity is called the measure of substance, and quality is the disposition of the substance. But the true idea of relation is not taken from its respect to that which it is, but from its respect to something else.I have been re-reading Indefinability by Josephine Pasternak (Boris' sister, one feels obliged to mention, just to get it out of the way). Pasternak's lovely little book seems to have "fallen stillborn," like many a significant philosophical work, despite its admiring "preface" (effectively a short blurb printed on page vii) by Dame Iris Murdoch. In it, Pasternak starts out by insisting that the notion of Relation is, you guessed it, indefinable. Exhibit one in her case is a sleek little demonstration:
Even a complex sentence implying a definition can be reduced to the syntactic skeleton of "S is P," where "S" stands for subject, "is" is a link (but "Link" is a kind of relation), and finally, "P" is a predicate.... Would it be possible to define [Relation] in the formal terms of a basic syntactic construction? Relation in it would be the subject; the link "is" (remember, a link is a kind of relation) would be a relation; the "unknown," the "sought for," would stand in the place of "P." We might as well replace the linguistic symbol "unknown" by the graphic punctuation symbol "?". ...The result, though looking monstrous, is correct: Relation (S) Relation (link "is") ? (P), that is:Pasternak's book suffers from a certain unfortunate anti-Aristotelian bias which occasionally mars its elegance with a smear of polemic, and seems to me to be based in part on a misconstrual of the Greek logos, trying to tease apart its various senses (especially "logic" and "word") in a way that feels anachronistic. However, when I read her vis-a-vis another somewhat tendentious anti-Stagirite tract, Robert Pirsig's beautiful Zen and the Art of Motorcycle Maintenance, a certain symmetry emerges.
This construction shows that the attempt to define Relation lands us in a vicious circle where the thing to be defined (namely Relation) is antecedent to its own definition. (Indefinability pp3-4)
Pirsig's book is not always taken seriously by philosophers. Partly this is because it is a novel; partly because it had the misfortune of being a very successful and popular novel; partly because there are, arguably, some mistakes in it. (The only factual mistake I know if is its claim that "Phaedrus" is the Greek for "wolf;" most of the other "mistakes" are, precisely, arguable, and tend to reduce to a certain riding-roughshod over accepted scholarship de jour.) I do take Pirsig seriously. At some point I may devote a post as to why I think he's a significantly under-rated thinker. For now, I want to underscore the resonance between him and Pasternak.
As you know if you've read ZAMM ("an inquiry into values," the subtitle announces) and maybe even if you've just read about it, its central philosophical contention is that Quality is indefinable, and comes (ontologically) before the division of experience into subjective and objective.
Quality is not a thing. It is an event. It is the event at which the subject becomes aware of the object... Quality is not just the result of the collision between subject and object. The very existence of subject and object themselves is deduced from the Quality event. The Quality event is the cause of the subjects and objects, which are then mistakenly presumed to be the cause of the Quality (ZAMM, p 304)This ontological priority leads directly to a defense of Pirsig's claim that Quality is indefinable (although Pirsig had actually formulated this claim earlier):
to take that which has caused us to create the world, and include it within the world we have created, is clearly impossible. That is why Quality cannot be defined. If we do define it we are defining something less than Quality itself.Allow, for the moment, these two citations to stand as abbreviations for a more complete argument. (I grant that Pirsig is leaving out some steps, or I am in citing just these two snippets.) What is significant in the present context is not simply that Pirsig and Pasternak are both claiming that they have an indefinable reality on their hands (or rather, we do on ours). The similarity is closer, and it depends upon, of all people, Aristotle, for whom Pirsig has almost nothing good to say (and he is frank about his partisanship.) Both Quality and Relation are, in Aristotle's terms in the Categories, "accidents." Well, yes, you'll say, but then anything in Aristotle that isn't a substance is an accident. Yes, but there is a symmetry, or rather, a polarity, between these two categories, since Quality is normally taken to be, as it were, the "strongest" accident (Quality contains four inherent species -- habit, potency, passive quality, shape and figure), and Relation, the "weakest," the most contingent. They are also, arguably, the two that have been historically found most problematic. Aristotle provides no argument for why he places the four aforementioned species under Quality, and exegetes do not agree on a rationale or even on whether he had one. (This subsection of the entry at the Stanford Encyclopedia of Philosophy is a good start on the difficulties.) As to Relation, Aristotle himself recognized that it posed problems. Whereas earlier in the Categories when he discusses Relation he says,
Those things are called relative which, either being said to be of something else or to be related to something else, are explained by reference to that other thing (Categories 6a),he later acknowledges that this account needs modification. After all, Aristotle's account is usually glossed as a "substance/accident"ontology; there are substances, and there are the accidents that adhere to them. This is an over-simplification, but the details do not seriously affect the point that since any Relation is an accident, it clearly should not apply to Substance. But, Aristotle goes on to note,
If our definition of the relative was complete, it is very difficult, if not impossible, to prove that no substance is relative. ...[But] only those things are properly called relative in the case of which relation to an external object is a necessary condition of existence.... The former definition indeed applies to all relatives; but the fact that a thing is explained with reference to something else does not make it essentially relative. (Ibid, 8a)The difficulty arises because all the categories are "relative" in a certain important sense. After all, to "explain" anything is to explain it in terms of something else. But there are, on the other hand, some things that are inherently related. "Daughter," for instance, is always daughter-of- ; the relation in this case is not just part of the explanation, but (we might say) part of the grammar of the thing to be explained.
This distinction led to an elaboration in medieval scholasticism, a distinction between Relatio secundum dici, "following [the manner of] speaking", and "Categorical" relation, Relatio secundum esse, "following essence". "Daughter," the example above, is an instance of the latter. The sharp-eyed will have spotted a certain isomophism with another scholastic notion, for just as daughter is always daughter-of- , so too in the classic account of intentionality, transplanted by Brentano into modern phenomenology, any state of apprehending always has just such an -of- structure; imagination is imagination-of- , fear is fear-of- , hearing is hearing-of- , and so on.
My understanding of the connection between this phenomenological sense of intentionality and its medieval antecedent is strongly influenced by John Deely's work, especially Four Ages of Understanding and Intentionality and Semiotics. Deely argues strongly that Brentano's is a deformation of the medieval concept, but that is a detail for specialists. What is important in this context is the way intentionality involves a directedness, a towards-which- . (This is in fact very close to the Greek for Relation in Aristotle's Categories, where Relation is πρός τι.) I will come back to the scholastic notion of Relation later, in connection with Deely's exposition of the great late-medieval thinker John Poinsot, also known as "John of St. Thomas" for his extremely scrupulous manner of following Aquinas.
Like "Relation," "Quality," too, has a certain ambiguity to it. It can be a term of valuation (as in Pirsig, generally), or a term of description (it need imply no value-judgment to say the apple has the quality of redness). We might write off this ambiguity as a lexicographic accident, but note that Pirsig in claiming that Quality stands before subject and object has already stepped back from the valuation-sense (with which he started). There is a certain incipient Platonism here -- a whisper of a Good beyond good and evil. Quality in the sense of characteristics -- and, note, these are what Locke will come to call "secondary qualities," in the sense revalorized by Meillasoux -- are also qualia, those things that philosophers like Dennett deny even exist. Qualia are simply the phenomena per se. And Pasternak, too, allows us to elide this distinction for our purposes, for she contends that phenomena are indefinable:
Indefinable Phenomenals are manifestations of indefinable Relations. Unknowns exposed by unknowns: we experience their manifestations but cannot account for their occurance. In a Kantian way we can attach to them a label only, the "As Such," which is a kind of hint (like the waving of a hand indicating direction) implying the impossibility of expressing them in existing language.Pasternak invokes Kant, but this is, let it be noted, a kind of reversed Kantianism, for what is here "indefinable" is not a mysterious noumenon receding at the wrong end of the telescope, but phenomena per se. And this is just what we should expect if we follow the aforementioned phenomenological account of intentionality.
The Kantian notion (which I've argued is also universalized in Harman's ontology) is actually a mutation of a medieval distinction between things (which exist whether or not we know) and objects (which are always objects of sensory or intellectual or imaginative apprehension). This distinction is mediated, in John Poinsot's metaphysics, by the sign; for things to become objects requires intentional states, but such intentional states are relational. The fallout of this, says Deely, is that Poinsot makes an
identification of signs with pure relations as such, [with which] medieval semiotics reaches its highest point of development....the sign as such, consisting in the relation between sign vehicle and object signified, is something suprasubjective and invisible to sense. Those "things" or perceived objects that we call signs—such as traffic lights, flags, and words—are not technically speaking signs but vehicles of signification. The actual signification itself consists in the relation between the vehicles and the knowability of their objective content. Similarly, those psychological states, such as images or concepts...are also not technically speaking signs but vehicles of signification. At this stage, a new definition of signs may be said to be implicit: a sign is that which any object presupposes. (Four Ages of Understanding, p 434)If we push these identities between Quality, Relation, and Sign very far, it should be clear, plenty of ramifications follow for Speculative Realism of either its relationalist or its object-oriented chapters; and even for the possibility of a revalorization of that much-maligned bugaboo, "Correlationism." There are, for instance, strong albeit inexact resonances between Pirsig's later work (in Lila, the (less-satisfying, in my opinion) sequel to ZAMM, in which he delineates four different levels of "static" Quality -- inorganic, biological, social, and intellectual) and Meillassoux's "four worlds" (matter, life, thought, and justice). Some fuller exploration of this would need, too, to tease out the relation between what Meillassoux means by a "meaningless sign," in his paper "Iteration, Reiteration, Repetition."
Deely's observation of the ongoing play of signification is not too far, weirdly enough, from something Derrida might have said, though Deely refers instead to the more staid founder of Pragmaticism:
The objects known, in their turn, become signs of one another as new relations among them are imagined or discovered. And so, in the end, the universe as a whole, in terms of medieval semiotic theory, comes to be "perfused with signs, if it is not composed exclusively of signs," exactly as Peirce later projected. (Deely, op cit)The citation of C.S. Peirce is from his "Issues of Pragmaticism":
the entire universe -- not merely the universe of existents, but all that wider universe, embracing the universe of existents as a part, the universe which we are all accustomed to refer to as "the truth" -- ...all this universe is perfused with signs, if it is not composed exclusively of signs. (in Collected Papers vols 5-6, p 302)This is a strange plural monism or monistic pluralism; but the details of what it might mean would depend, in part, on the way one understands meaning itself.