I have yet to insert some Greek terminology denoted as [Greek] and the full citation details.
There is no interpretative consensus among Aristotelian scholars on the methodological principles Aristotle employed in arriving at his list of categories nor indeed how the categories themselves are related to these “classes” of entities. Are they primarily classes (a theory of reality) or classes of expressions (a theory of language) of a certain sort we classify under substance, quality or quantity? Are the entities classified according to a classification of the “categorematic” (Frede 1981: 2) expressions by which they are “signified” or are they classified according to the classification of the “entities” they signify? Aristotle himself “was never in a position to fix his topic . . . because his conception of what ought to be achieved by a discussion of [Greek] constantly outgrew his definitions” (Wiggins 1995: 214); Aristotle’s conception of substance changed across the Categories, Physics and Metaphysics (Driscoll 1981: 129-159). Matters are further complicated in that all that the Aristotelian tradition held to should not be laid at the door of Aristotle. Thus in what follows I keep to a perceived orthodoxy in interpretation (I do look at some competing interpretations) but and in so doing also licence myself to take a more suggestive reading rather than being constrained by a close grained exercise in scholarship.
For Aristotle the unit of thought is the proposition in which something can be affirmed or denied, stated as true or false and which he analyses into two factors:
(1) The Subject – that about which something is affirmed or denied; and
(2) The Predicate – that which is affirmed or denied of it.
Consequently his doctrine of the classification of Terms is based on a classification of Predicates, or of Propositions according to the special kind of connection between the Subject and Predicate which they affirm or deny. Two such classifications, which cannot be made to fit into one another, are the “Categories” or “Predicables”. The list of “Categories” reveals itself as an attempt to answer the question in how many different senses the words “is “a” or “are” (one needs to mark that Aristotle’s first word for substance was “being” which “obscures the etymological connection between [Greek] and being” (Gill 1989: 13)) are employed when we assert that x is y or x is a y or xs are ys. Such a statement may tell us:
(1) what x is for if I say that “x is a cat”; the predicate is then said to fall under the category of Substance;
(2) what is x like, for example when I say that “x is black, or wise” – the category of Quality;
(3) how much or how many x is – the category of Quantity;
(4) how is x related to something else, say x is to the right of y, x is the father of y – the category of Relation.
For Aristotle “certain elements of a language have key-designating roles, the full understanding of which requires that we understand the designata as falling within those classes which jointly form the set definitive of, that to which a sensible particular must be related” (Moravcsik 1968: 145). The word key is crucial here. The structure of language was not isomorphic to the structure of reality a la Wittgenstein’s Tractatus, but that there are key parts of language and reality which do have connections.
In Categories 1b25ff Aristotle introduces the list of categories as containing the kinds (quality, quantity) of items (for example, white, grammatical, three cubits long) which are signified by “things said without any combination.” In Topics 101b26-28 he writes that the key elements in a statement are property, genus, definition and accident but that none of them in of themselves comprise a statement. Does the theory that all the elements of such sentences designate an item? Does every element that has a designation fall into one of the categories? The answer offered by Moravcsik (1968: 128-135) is summarized as follows.
(1) Aristotle did not think that every word or phrase which could be part of a sentence of subject-predicate form has the function of designating an entity.
(2) Likewise, not every noun or verb designates an item in one of the categories.
(3) Some noun or verb-phrases do designate items falling under more than one category.
(4) Not all relevantly simple nouns and verbs designate items falling into only one category.
(5) Those elements of subject-predicate form, or definitions (Aristotle did not consider definitions as true or false or as produced by combination but we must assume that he thought of parts of definitions as falling under one category), which
(a) are not connectors;
(b) are not, like “being”, otherwise non-designative in nature;
(c) cannot be turned into sentences of subject-predicate form, by the mere addition of connectors;
(d) designate genuine unities, are “uncombined elements” of language, and designate items falling into one and only one category.
The designative link between simple parts of language (d) and the simple parts of reality which fall into only one category, is the key link between the structure of language and the structure of reality.
Consider an ordinary universal affirmative proposition of the form “all xs are ys.” Now if this statement is true it may also be true that “all ys are xs,” or it may not. On the first supposition we have two possible cases:
(1) the predicate may state precisely what the subject defined is; then y is the Definition of x, for example “men are mortal animals capable of discourse.” Here it is also true to say that “mortal animals capable of discourse are men,” and Aristotle regards the predicate “mortal animal capable of discourse” as expressing the inmost nature of man;
(2) The predicate may not express the inmost nature of the subject, and yet may belong only top the class denoted by the subject and to every member of that class. The predicate is then called a Proprium or property, an exclusive attribute of the class in question. Thus it was held that “all men are capable of laughter” and “all beings capable of laughter are men,” but that the capacity for laughter is no part of the inmost nature or “real essence” of humanity.
Again, in the case where it is true that “all xs are ys” but not true that “all ys are xs”, y may be part of the definition of x or it may not. If it is part of the definition of x it will be either:
(3) a Genus or wider class of which x forms a subdivision as in “All men are animals”; or
(4) a Difference, that is, one of the distinctive marks by which the xs are distinguished from other sub-classes or species of the same genus, as in “All men are capable of discourse”; or
(5) y may be no part of the definition of x, but a characteristic which belongs both to xs and some things other than xs. The predicate is then called Accident. Thus the predicate of a universal affirmative proposition is always a definition, a proprium, a genus, a difference, or an accident.
Nature herself has made certain hard and fast divisions between kinds which it is the business of our thought to recognise and follow. Thus according to Aristotle there is a real gulf, a genuine difference in kind, between the horse and the ass, and this is illustrated by the fact that the mule, the offspring of a horse and an ass, is not capable of propagating. To say that Socrates is a man tells me what Socrates is, because the statement places Socrates in the real kind to which he actually belongs; to say he is wise, or old, or a philosopher merely tells me some of his attributes. It follows from the belief in “real: or “natural” kinds that the problem of definition acquires an enormous importance for science. For Aristotle, it was held that a true classification must not only be formally satisfactory, but must also conform to the actual lines of cleavage which Nature has established between kind and kind, the task of classificatory science becomes more difficult. Science is called on to supply not merely a definition but the definition of the classes it considers, which faithfully reflects the “lines of cleavage” in Nature. This is why the Aristotelian view is that a true definition should always be per genus et differetias. It should “place” a given class by mentioning the wider class next above it in the objective hierarchy, and then enumerating the most deep-seated distinctions by which Nature herself marks off this class from the others belonging to the same wider class. Modern evolutionary science differs from the Aristotelianism on one point of importance. It regards the difference between kinds, not as a primary fact of Nature, but as produced by a long process of accumulation of slight differences. As the intermediate links between “species” drop out because they are less thoroughly adapted to maintain themselves than the extremes between which they form links, the world produced approximates more and more to a system of species between which there are unbridgeable chasms; evolution tends more and more to the final establishment of “real kinds”, marked by the fact that there is no permanent possibility of cross-breeding between them. This makes it once more possible to distinguish between a “nominal” definition and a “real” definition.
As I intimated in the opening paragraph of this section, it is not immediately obvious whether Aristotle’s ten categories, surprisingly hetergeneous, is supposed to yield an ontological classification or an analysis of the structure of propositions. The hetergeneity of the list suggests that Aristotle is primarily concerned with the types to be designated rather than the manner of designation (Moravcsik 1968: 136).
The orthodox characterization of the categories conceives them as the “highest predicables” (in a footnote, Moravcsik cites WHB Joseph’s An Introduction to Logic wherein he writes that the Greek word for category [Greek] means “predicate”). Sensing the ambiguity of “highest” Moravcsik (1968: 137) presents two ways in which the term might be taken:
(1) predicate p/ is higher than predicate p// iff all members of the class which make up the range of application of p// are also members of the class which makes up the range of application of p/, and not vice versa. i. e. animal is higher than man.
(2) p/ is higher than p// iff p/ is an attribute of p//. i. e. a category would be a higher predicate than quality or quantity, and colour, quality, category would be constitute a hierarchy.
Aristotle could not accept a hierarchy (2) for the ontological reality of the higher strata would be approaching Platonism, and so it seems that (1) is more appropriate in that as one goes to wider and wider genera, the particulars contained in any one genera are on an ontological par with the particulars of the wider genera.
Other interpretations of the ‘highest predicables’ have been put forward by Ryle, Anscombe and Ackrill, loosely grouped as the “linguistic” view. On this view the list of categories is intended to delineate the different kinds of things that can be said about a substance. As there are infinitely many things that could be said, the list is necessarily incomplete. However, the central achievement of Aristotle’s list, according to the Ryle/Anscombe view, is that Aristotle anticipated the “concept of semantic category and that the notion of a category-mistake” (Moravcsik 1968: 138). Given the discussion of Aristotle’s “key” elements of a statement, it is understandable that many commentators have taken the “linguistic” view (Frede 1981: 22). Moravcsik gives three objections to the “linguistic” view:
(1) Aristotle’s list was designed to contain mutually exclusive categories which are jointly exhaustive of reality.
(2) The basis of individuation of the categories was not linguistic. Aristotle explicitly states (11a15-20; 6a26) their differentiating characteristics – for example, quality is that in virtue of which things can be said to be similar; quantity is that which can be said to be equal or unequal.
(3) Were Aristotle to have semantic anomaly in mind, “he would have missed the glaring fact that to describe a shape as red or blue is sematically odd, even though both shape and colour belong to the same category – i. e. quality” (Moravcsik 1968: 139).
Ackrill (1963: 78-81) advances the view that Aristotle arrived at his list of categories in two ways.
(1) Through the sorting out of the different types of question that can be asked about substances.
(2) Through the asking what any given thing is and to continue by repeating that question with reference to whatever the previous answer disclosed (Topics 103b22ff).
The process of (1) and (2) is concluded when irreducible genera are reached, presumably because nothing more can be asked. Again, this account has its weakness. The emphasis on questions, or classes of properties concerning substances seem to run on one example, man. Moravcsik (1968: 140-141) quite rightly points out that it would be very surprising if the two approaches yielded the same list: “Why should the classification of the aspects of one kind of entity, e. g. substance, coincide with an exhaustive classification of the essences of all entities that make up reality?” Ackrill’s view also suffers from the same criticism levelled against the Ryle/Anscombe view in that there is an indeterminacy in trying to classify things by repeatedly asking a question.
Ackrill’s second approach runs on a passage in the Topics. It states, among other things, that the essence of anything will be found in one of the categories, but only that the categories make up such an exhaustive classification of reality that no real essence will overlap any one category. This is consistent with the claim that there are irreducible genera within the categories and that the number of categories could be reduced. Most importantly, the statement is only a necessary condition for the correct list of categories; it does not provide a procedure for arriving at or a principle of unity for the list.
In light of the above interpretations, Moravcsik (1968: 142) gives three necessary conditions for an adequate interpretation drawing on the Topics, Physics and Metaphysics, all pointing to Aristotle’s using the list in his analyses of key concepts such as being and change, but also in his intention to show substance as ontologically prior to all else. The three necessary conditions are as follows.
(1) the list must be exhaustive of all that Aristotle takes to exist;
(2) no reduction in the number of categories should be possible without violating the principle that the essence of anything will be found in one of the categories and that no real essence will overlap any of the categories. In other words, each category must be mutually exclusive;
(3) no further subdivision of the categories should be possible without violating the constitutive principle.
These conditions go some way to resolving the debate whether there are particulars and universals in each category, or only universals in all but the first. The conditions do not entail that each category must contain both universals and particulars but rather that they jointly contain all the universals and particulars thus allowing for some categories to be comprised solely of universals, or particulars or both.
For Moravcsik (1968: 143-144) an interpretation that goes some way to providing a more adequate understanding is that of Bonitz (Àber die Kategorien des Aristotles, Vienna 1853). Bonitz’ suggestion is that Aristotle’s list “yields a survey of what is given in sense experience, and that each entity thus given must be related to some item in each of the categories.” On this view the constitutive principle of the list of categories is that they constitute those classes of items to each of which any sensible particular, substantial or otherwise, must be related. A sensible particular includes not only substance, but sound, event etc. and must be related to some substance, it must display some quality and quantity, it must have relational properties, have spatio-temporal existence, subject to the categories of affecting and being affected.
Aristotle’s list was first and foremost an ontological inventory: that is, first determining an ultimate class of entities consisting of all qualities and then introducing a category of quality which is characterised by the fact that the item predicated belongs to the antecedently determined class of qualities. Frede (1981: 22) thinks that the reverse is true: he takes Aristotle in the Categories to rely on the already understood knowledge of what it is to say of something what it is like. And qualities are just those items which we attribute to something when we say what it is like, quantities just those items which we refer to in saying what amount it is, and so on.
I concede that Aristotle did arrive at his list of categories with both logical and grammatical considerations in mind – after all logic and meaning is vital to metaphysical discourse – but I don’t hold to the view of Frede’s (1981: 23) that “the real interest of the Aristotelian notion of a category seems to lie in the fact that we are supposed to have one set of notions, or perhaps, two or three very closely related sets of notions, which do indeed work in logic, in grammar, and in metaphysics.” The reason for attributing an ontological character to Aristotle’s enterprise is because of his emphasis on substance, the preeminent category, having ontological primacy over all other categories predicable of substance, the investigation of which lays the ground for scientific explanation.
II. essence and accident
At the heart of the essence and accident distinction lies the notion of change and permanence: if x really changes then it cannot literally be x, the same object that underwent the change; but if we still call x x after a change, thus in some sense retaining its identity, then could it have really changed?
Aristotle thus distinguishes two senses of change, alterations on the one hand, and on the other hand, coming-to-be and passing-aways (On Generation and Corruption 319b8-24). The first, taking the example of a man, one readily acknowledges that a man may undergo any number of changes associated with growing up, illness, ageing, loss of limb etc., yet substantially or essentially be what he was before the changes. This corresponds to the first notion of the change.
Another sense of change corresponding to the second sense above is, taking the example of some artifact and inflicting some change on it. One might say as in the case of a man, a table can retain its identity were it to be painted. However, were it incinerated, the identity conditions for that table would not be possible for its essential property had been destroyed.
Thus drawing the distinction between substantial or essential change and other types or degree of change is coextensive with the distinction between essential attributes or essences, and other types of attribute, collectively termed as accidental.
Some might say that there exists a tension in Aristotle’s distinction. For if there is some underlying permanencies that persist through even the most drastic of changes, then nothing can be said ever stop or begin existing. But Aristotle seems to have anticipated this objection by positing the much stronger claim of the existence of matter.
What then is the principle for distinguishing between essence and accident? It has already been indicated that one ontological principle might be that:
if a change is perceptible, then we can say that a given property P1 is essential to an object o1 iff its loss would result in the cessation of that object whereas a property P1 is a mere accident iff the object o1 would remain identifiably and substantially the same in its absence.
An epistemic reading of the distinction between essence and accident might be that knowledge of the essence of a thing, i. e. scientific knowledge, dissolves the problem of accidental properties, for “there is no knowledge of accident as all, only of essences” (Copi 1968: 152). And scientific knowledge contrasts with two types of pseudo-definition – the definition of compounds, and the nominal definition of simples. The former are constructions of items from different categories (substance plus some accidental property); are arbitrary as no natural kinds can correspond to them, and therefore have no ontological status. Taking Aristotle’s example of a musician: no individual is constituted by the essence of a musician, since the musician is a man, but a man cannot cease to be a man and to have the essence of a man without ceasing to exist (Ayers 1991: Vol II: 23).
Ayers (1991: Vol II: 21) gives an illustration that makes this idea clearer. We know that crows are always black (cf. Hempel’s paradox of the ravens) inductively so, yet we could conceive of a crow retaining its nature as a crow without being black.
The implication of Aristotle’s drawing a distinction between essence and accident should be quite evident. By positing accidental properties, the implication is somehow we can strip them away and reveal the “real” unchanging (essence) nature of something, and of course Aristotle needed a principle of immutability to run a truly scientific enterprise.
III. matter and form
It should be noted that Aristotle begins his investigation into the structure of reality and the causes by which it is produced by starting from the existence of individual things belonging to the physical realm and which are sensible. About any such thing we may ask two questions:
(1) into what constituent factors can it be logically analysed?
(2) how has it come to exhibit the character which our analysis shows it to have?
The answer to these questions will appear from a consideration of two standing antitheses which run through Aristotle’s philosophy, the contrast between Matter and Form, between Potential and Actual, and his doctrine of the Four Causes. The interest for the present lies with the first of these.
Take any two objects made of the same metal, candlesticks and a bowl, alike in the sense that they are both made of the same metal yet unlike in that they have different shapes – the shape of the candlestick renders it suitable to be used as a candlestick, the shape of the bowl renders it suitable to be used as a bowl and so on. Now a bio-chemist can specify that the constituent tissue and chemical elements of an oak or horse are of the same kind as those of ash or an ox, but that the oak differs from the ash or the horse differs from the ox in characteristic structure. Thus in any individual thing we can distinguish two components, the stuff of which it consists – which may be identical in kind with the stuff of which things of a very different kind consist – and the structural law of formation or the arrangement which is peculiar to the special kind of thing under consideration. In each concrete individual thing these two are inseparable. They do not exist alongside each other in the way two atoms of hydrogen and one of oxygen do in water; the constitution is manifested in and through the metal or tissue. the individual is the matter as organised in accord with a determinate principle of structure, the form. Aristotle extends the analysis of the “stuff” or timber that constitutes a ship into Matter and Form by analogy beyond the range of individual substances to everything in which we can distinguish a relatively indeterminate “somewhat” and a law or type of order and arrangement giving it determination. So if we consider the relatively fixed or “formed” character of an adult man, we may look upon this character as produced out of the ‘raw material’ of tendencies and dispositions. This licences us to speak of native disposition as the matter or stuff of which character is made. Matter in the Aristotelian sense must not be confused with body; the relatively undetermined factor which receives completer determination by the structural law or Form is Matter. This is brought out in the ontological interpretation put on the logical process of definition by genus and difference – the genus may be said to stand to the ‘differences’ as Matter, the relatively indeterminate, to the Form which gives it its structure.
We also observe that Matter and Form are strictly correlative. The matter is called so relatively to the form which gives it further determination. Form is taken to mean the last determination by which a thing acquires its completed character, and the Matter is that which has yet to receive this last determination. Thus in the case of the silver bowl, the spherical configuration is said to be its form, the silver its material. insofar as living organisms are concerned, the Matter is the tissue, muscles, bones, skin, etc. But each of these things which are counted as belonging to the matter of the bowl or the human body has, according to Aristotle, a development behind it. Silver is not an “element” but a specific combination of “elements” and like wise with the tissue of a living body. Thus what is matter relatively to the bowl or the living organism is Matter already determined by Form if we consider it relatively to its own constituents. Thus at every stage of a process of manufacture or growth a fresh Form is superinduced on, or developed within, a Matter which is already itself a combination of Matter and Form relatively to the process by which it has itself been originated. The implication of such a view would lead to the conclusion that in the end the simple ultimate Matter of all individual things is one and the same throughout the universe, and has absolutely no definite structure at all.
For Aristotle, at least in the Categories, anything that qualifies as a substance does so in virtue of the substance underlying all else. In the Metaphysics he begins to draw out the implications for what is actually to count as a substance – is a statue identical to its clay or to the form of its clay? Frede (1987: 64-65) suggests that an adequate answer to the substance/form distinction will need to satisfy at least three conditions:
(1) the substance must be the sort of thing that will allow one to understand why the object, whose substance it has, has the properties it has. Recall earlier, we talked of an artifact changing its properties but still remaining the same – the history of the changes might well be viewed as equivalent to the history of the object. The object might have had a different history yet have been the same object;
(2) the substance must be an individual, since we are looking for the real individuals in the category of substance which are to explain the individuality of ordinary individual objects;
(3) there must be some kind of asymmetry between substances and properties on the basis of which one can specify properties and all that exists as dependent on substances for their existence and not vice versa.
In Metaphysics Aristotle says that physical objects and the essences of objects, universals and ultimate subjects, were parallel candidates for the one title of substance (1042a3-15).
The primary substances of the Categories are individual concrete objects or things, such as a particular dog or a particular houise. In the Metaphysics, dogs and houses are now viewed as combinations of matter and substantial form and not themselves primary substances – the status of primary substance now being designated to substantial forms. Frede (1987: 74-75) sets out Aristotle’s reason for this shift.
A statement like “Socrates is healthy” introduces two entities, Socrates and health. What then is the subject of health, if health is a distinct entity from its subject? “What in the bundle or cluster of entities that constitutes Socrates is the thing itself as opposed to the properties like health which it underlies?” The two main reasons (Frede 1987: 79) why the concrete, particular substances of the Categories got replaced in the Metaphysics by substantial forms as the primary substance are:
(1) Aristotle is now concerned with the question what is the real subject in itself as opposed to its properties;
(2) Aristotle now not only has developed his own theory of forms, but also has come to assume separate substantial forms which are paradigms of substances, but which are not substances in the same way as the composites or the concrete particular objects are.
The relationship between matter and form in Aristotle the hylomorphic theory which “conceives of a thing’s form as the way in which its proximate matter has to be organized or arranged in order for a thing of that kind, made of that matter, to exist. This a piece of bronze has to be shaped in a certain way in order for a statue, made of that bronze, to exist” (Lowe 1998: 195). In what sense, if any, is a particular bronze statue a ‘combination’ of matter and form? The piece of bronze comprising a statue is not a part of it.
If a living thing is functioning, it will exhibit and behave in a characteristic way; and conversely to behave in this way is its function. Further, this characteristic behaviour depends on the orgasnisation, structure and disposition to causally effect such behaviour. For Aristotle, the form is this disposition or organization, while the matter or substance is what it is thus disposed or organized.
The question then, how could the form so conceived, satisfy the three requirements for being a substance as set out above? Substance was to explain why, despite all the changes an object had undergone, it still was the same object. If we take the example of Theseus’ ship whereby an old ship is gradually repaired with the old planks constituting a second ship, perhaps what makes for the identity of the repaired ship with the original ship is a continuity not of matter or properties, but the continuity of the organisation which functions as a ship. Puzzles such as this will be returned to in another installment.