6

The Unity of General and Special

Metaphysics: Aristotle's Conception
of Metaphysics

If one tries to get clearer about Aristotle's conception of metaphysics, one natu-
rally turns to the treatise that by its very title promises to give us an account of
Aristotle's metaphysics. Unfortunately, the title itself does not provide us with
any clue. "Metaphysics" is not an Aristotelian term. It only gains some currency
in late antiquity. Thus, the commentary on Isaiah attributed to St. Basil (164)
speaks of those things, higher than the objects of the theory of nature, "which
some call metaphysical." The earliest catalog of Aristotle's writings, the one
preserved in Diogenes Laertius, does not yet contain the title "Metaphysics."
Hence, it is clear that our title is the title later editors gave to the treatise. It is
first attested in Nicolaus of Damascus' compendium of Aristotle's philosophy,
i.e., in the first century B.C. But even these editors presumably did not mean
to suggest any particular conception of the discipline by chosing this title. Proba-
bly, they were at a loss regarding a proper title for the treatise and just named
it after its position in the corpus of Aristotelian writings, namely, as coming af-
ter the physical writings. It would also be a mistake to assume that the title in-
directly expresses a certain conception of the discipline metaphysics by referring
to its "natural" place in the order of Aristotelian writings. The place is anything
but "natural." The order of the corpus follows the Academic, and then Stoic, di-
vision of philosophy into logic, physics, and ethics. And though there were sub-
divisions of this scheme in Hellenistic times, none made provision for a dis-
cipline metaphysics, whether called by that or another name. Not as if
Hellenistic philosophers did not do any metaphyics, but they did not regard it
as a separate discipline. Sometimes physics was divided into physics, more
properly speaking, and theology. And since there is no natural place for the
Metaphysics in Aristotle's corpus, a position after the physical writings must
have seemed least disturbing, especially since Aristotle himself in the
Metaphysics at times had identified the subject of the treatise as theology. More-
over, the treatise clearly belonged with the theoretical treatises, rather than with

81
 82 UNITY OF METAPHYSICS

the Organon or the ethical writings, and hence had to come before or after the
physical writings. But the basic order of the corpus is clearly didactic: we start
with logic, proceed to a doctrine of the sensible world, on the basis of this move
on to a doctrine of the intelligible reality underlying the sensible world, and,
finally, in the spirit of Hellenistic philosophy, move to ethics as the ultimate end
of all philosophical endeavor. In late antiquity it will become natural to identify
the intelligible with the supra-sensible and to think of the move from the physical
writings to the metaphysical treatises as the move from the doctrine of the sensi-
ble world to the doctrine of the supra-sensible world. And now the term
"metaphysical" is easily understood to refer to the doctrine of supra-sensible en-
tities, God, the ideas, the umoved movers, and or angels. But there is not su-
fficient reason to believe that this was what the Hellenistic editors of our text
had in mind. If there is a question about why they chose the title "Ta meta ta
physika," it is why did they not call the treatise "theology" or, at least, "first phi-
losophy"? The answer to this must be that in Hellenistic philosophy one did not
have much use for the notion of a first philosophy, and that the treatise did not
fit the conception of theology one had, just as it does not fit our conception of
theology. But that the matter is more complex we can see from the fact that
Nicolaus of Damascus (p. 74 Dossaart-Lulofs) still identifies the subject of the
Metaphysics as theology and calls it the "first science." Once we come to late
antiquity, the situation has changed radically. Given the dominance of Platonism
with its two-world view, its identification of the realm of forms with the Divine,
and its doctrine of the ascent from the physical to the metaphysical, it became
easy to see the Metaphysics as a theological treatise (cf. Asclepius in Met. p.
1, 18-2, 3). But this way of looking at things hardly fits into the first century
B.C. Hence, we do not have sufficient reason to suppose that the title
"Metaphysics" originally referred to anything more than the position of our trea-
tise in the corpus.
To get a notion of Aristotle's conception of metaphysics, then, we cannot rely
on the title, but have to turn to the treatise itself. Unfortunately, the treatise, too,
does not owe its present form to Aristotle. There is good reason to believe that
the treatise, as Aristotle left it, was composed only of books A, B, F, E, Z, H,
0,1, A, M, N, and that a, A, and K were added later. But even this underlying
treatise turns out not to be of one piece. The evidence for this is abundant and
well known. It will suffice here to recall that, e.g., the beginning of Metaphysics
Z suggests a certain program in the course of which we shall deal with separate
substances and the claim of ideas and mathematicals to be substances. But
though we get these discussions in Met. A and Met. M-N, respectively, it is
fairly clear that these books were not written in one piece with the beginning
of Met. Z. They, rather, seem to be revised versions of treatises Aristotle incor-
porated to temporarily fulfill the need of a discussion of this sort. Once we rea-
lize this, it is also clear that to determine Aristotle's conception of metaphysics,
 UNITY OF METAPHYSICS 83

we cannot look just at what we take to be the original treatise to see what kind
of project he is carrying out. For the project never seems to have been com-
pleted. Thus, the only avenue that is left to us seems to be the following: we
have to go by Aristotle's explicit remarks about the project he is engaged in, see
to what extent this project actually is carried out, and extrapolate on what the
finished project would have looked like.
Unfortunately, it turns out that Aristotle, on the face of it, does not even seem
to have a clear conception of his project himself. Different parts of the Ur-
Metaphysik seem to be written with different conceptions in mind. The most
striking example of this is the following: the first lines of Met. T (1003a 2Iff.)
introduce the discipline as a science that considers being qua being quite gener-
ally and set it off from the particular sciences, which single out a particular part
of being, particular kinds of beings, as their subject of study. And the rest of
Met. F, in particular Met. T 2, tries to show how there could be such a universal,
and yet unified, discipline. Met. E 1, on the other hand, introduces a discipline
that is concerned with a particular subject matter, namely, with the kinds of be-
ings that come first in the order of being. Hence, Aristotle calls the science "first
philosophy" (1026a 24). And, assuming for the moment that there might be di-
vine beings prior to natural objects, he also calls the discipline "theology" (1026a
19). Thus, we seem to have two radically different conceptions of the enterprise
of the Metaphysics. According to one, we deal with what traditionally has been
called "metaphysica generalis," a general study of being as such, of all there is
insofar as it is, according to the other with metaphysica specialis, a study of a
special kind of beings, supra-sensible beings.
Much modern scholarship has been devoted to the question how these two no-
tions might be related. And yet agreement seems to be so far out of reach that,
given the present state of the art, it might seem hopeless to make another attempt
to arrive at a generally acceptable interpretation. If I, nevertheless, make the at-
tempt, it is not because I think that I have a radically new answer, but because
it seems to me that basically the correct answer was given by Patzig in his "The-
ologie und Ontologie in der 'Metaphysik' des Aristoteles" (an English version of
which appeared in: J. Barnes, M. Schofield, R. Sorabji, eds.: "Articles in Aris-
totle," vol. Ill) in 1960, but this interpretation has not won the acceptance it
deserves. This is in part owing to the fact that Patzig's view needs to be revised
and elaborated in various respects.
Any interpretation has to start from the fact that it is clear from Aristotle's
own remarks that Aristotle himself does not see a conflict between the two no-
tions. Though E 1 introduces, in addition to physics and mathematics, a theoreti-
cal science of separate, unchanging substances as the object of our study, he,
right from the beginning of the chapter, also talks as if he were still concerned
with the universal discipline introduced in Met. T which studies being as such,
and not just a particular kind of being (1025b 9-10). And toward the end of
 84 UNITY OF METAPHYSICS

Chapter 1, he faces the issue squarely by pointing out that there is a problem
of whether first philosophy is universal or particular. But, obviously, for Aris-
totle this is not much of a problem. For, hardly having raised it, he settles it by
the succinct remark: "it is universal in this way because it is first" (1026a 30-31).
This is all Aristotle cares to say about the matter.
How are we going to interpret this remark? We cannot follow Natorp who
tried to deal with our problem in a manner fashionable in the late nineteenth cen-
tury, namely, by excising from the text all references to the theological interpre-
tation of the Metaphysics as later interpolations. This would not leave us with
a coherent text for E 1. The remark at the end of Chapter 1 also excludes the
possibility suggested by Jaeger (Aristoteles pp. 226ff.) that the two views reflect
different stages in Aristotle's development. It also seems to rule out the attempt,
suggested, e.g., by Merlan, to make the apparent conflict disappear by interpret-
ing the phrase "being qua being" to refer to God's being, rather than to being
in general. For the final remark of E 1 does assume that the study of being qua
being as such is general or universal, and tries to explain how it could be so,
given that it has a particular subject-matter. The problem raised in 1026e 23
would not be a problem and would not need the answer it receives in the last
sentence of the chapter, if the study of being qua being were not as such univer-
sal, but already in itself concerned with a particular kind of being. For we can
hardly attribute to Aristotle the much later view that God is just being.
The explanation I want to offer for the final remark of E 1 is the following:
(i) theology deals with beings of a certain kind, namely, separate substances. But
in doing so, it also deals with a particular kind or way of being, a way of being
peculiar to divine substances, (ii) It turns out that this way of being is the one
in terms of which all other ways of being have to be explained, i.e., it turns out
that a study of being as such resolves itself in three steps into a study of how
all the different ways of being that characterize the different kinds of beings ulti-
mately have to be explained in terms of the way of being that is characteristic
of divine substances, (iii) Since theology studies this focal way or sense of being,
it also provides the natural point to discuss how all other ways of being depend
on this primary way of being, especially since this primacy would seem to reflect
the very nature of divine substances. In developing this explanation, theology
does carry out at least the substantial core of the program of general metaphysics
and to that extent can be identified with general metaphysics. This is one way
in which theology, because of the primacy of its objects, will be universal. For,
in taking into account the primacy of the being of its objects, it will also deal
with the ways of being that are dependent on it. (iv] But general metaphysics
involves more than this kind of ontology. It also discusses certain universal prin-
ciples, like the principle of non-contradiction, and certain notions of universal
applicability like the notions of unity and identity. Again, this can be explained
in terms of the primacy of theology. For though these principles and notions are
 UNITY OF METAPHYSICS 85

universal, the first time they will be used in the hierarchy of sciences is in theol-
ogy, and so it will fall to the theologian to introduce them in an appropriate man-
ner. This will be all the more fitting, since ontology will, e.g., involve the dis-
tinction of various kinds of unity and identity; not only will some of these kinds
be the ones needed by the theologian, but the theologian will be the one most
competent to elucidate them. Thus, it turns out to be true in various ways that
theology, because of its primacy, will be universal, (v) Admittedly, this will
have the result that theology, or general metaphysics, has less internal unity than
we might have expected. But Aristotle himself seems to envisage this result.
Theology, or general metaphysics, will actually consist of a series of studies
which have only generic unity. Only one of these studies will amount to theology
in a narrow sense. But given the position of theology in the narrower sense in
the hierarchy of theoretical sciences, which is, after all, owing to the nature of
its subject-matter itself, it is embedded in a whole series of studies which, taken
together, constitute general metaphysics.
Thus, let us try to understand how it is that theology is not concerned only
with a particular kind of beings, but with a particular way of being, peculiar to
its objects, and how it addresses itself to this way of being. By distinguishing
a kind of beings and a way of being I mean to make a distinction of the following
sort. Horses are a kind of beings, and camels are a different kind of beings, but
neither horses nor camels have a distinctive way of being, peculiar to them; they
both have the way of being of natural substances, as opposed to, e.g., numbers
which have the way of magnitudes, or qualities which have a yet different way
of being. The way magnitudes can be said to be is different from the way quali-
ties or natural substances can be said to be. The claim, then, is that the way sepa-
rate substances can be said to be is peculiar to separate substances. One reason
why one has such difficulties with the identification of theology with general
metaphysics is that one thinks of theology as just like other particular sciences,
like astronomy or zoology, which deal with the nature of a particular kind of
beings, but which could not be thought of as concerning themselves with the very
way of being their objects have. For in the relevant sense, animals do not have
a way of being peculiar to themselves; their way of being is the way of being
of natural substances quite generally, and thus the zoologist presupposes, but
does not concern himself with, a notion of a natural substance and what it is to
be for such a substance. Similarly, the astronomer does not concern himself with
the way of being the objects of astronomy have. They have the being of magni-
tudes, let us say, but it is not his concern to determine this way of being. For
it is shared by the objects of many other disciplines. It would be a mistake,
though, to think that it is true without qualification that particular disciplines are
like that. It, rather, seems to be the case that we have to distinguish between
the three particular theoretical sciences Aristotle distinguishes, namely, theol-
ogy, physics, and mathematics, and the diverse particular sciences that consti-
 86 UNITY OF METAPHYSICS

tute these three theoretical sciences by forming hierarchically ordered groups of

disciplines. Once we make this distinction, it is clear that not just theology, but
also physics and mathematics, i.e., all three of Aristotle's theoretical sciences,
do concern themselves not just with a particular kind of beings, but also with
a particular way of being peculiar to their objects. Though physics deals only
with a particular part of reality, namely, natural substances, its objects do have
a distinctive way of being, namely, the way of natural substances. And physics,
as we can see from Aristotle's Physics, does address itself to this particular way
of being by asking the question what it is to be a natural substance, i.e., by ask-
ing what it is to have a nature in the sense that distinguishes natural substances,
whether there is such a thing as a nature, and what is involved in assuming natu-
ral substances to have such a nature. Even in the Metaphysics (Z 11, 1037a
14-16) Aristotle tells us that in a way the, as we would say "metaphysical," in-
quiry into sensible substance is part of physics or second philosophy. And in E
1 he is quite explicit that were it not for the assumed fact that there are sub-
stances prior to natural substances, metaphysics would be part of physics and
hence physics would be first philosophy (1026a 27-29). Obviously, Aristotle is
of a divided mind concerning the natural place of a theory of the way of being
of natural substances. And this accounts for the strangely metaphysical character
of Aristotle's treatise called "Physics." But there is no doubt that he does assume
that the objects of physics do have a distinctive way of being, peculiar to them,
and that the science of physics at least presupposes a theory of this way of being,
if it does not itself involve it. For physics as an axiomatic science would have
as one of its principles an assumption of what it is to be for a natural substance
and of the existence of such things.
The situation in the case of mathematics is less clear. But, again, it seems that
mathematical entities not only are a separate kind of entities, but also have their
separate way of being, peculiar to them, namely, the being of magnitudes.
Moreover, mathematics, as Aristotle conceives of it, does have a place at which
the mathematician naturally would address himself to the peculiar kind of being
mathematical entities have, or at which he would at least presuppose an answer
to this question, namely, in general mathematics, a subdiscipline of mathematics
Aristotle refers to in E 1 (1026a 27). General mathematics deal with the nature
and the properties of magnitudes in general, or as such, and thus at least presup-
poses a notion of what it is to be for a magnitude and that there are such entities
with this kind of being. Thus, theology is unlike the particular sciences that fall
under physics or mathematics, in that it is concerned with beings which have
a way of being peculiar to them and in that it somehow has to address itself to
this way of being by at least making an assumption about the nature of its objects
and their existence, but it does not differ in this respect from physics or
mathematics as a whole. This matter will need closer consideration at a later
point. For the moment it suffices to recognize (i) that theology deals with beings
 UNITY OF METAPHYSICS 87

that qua beings differ from all other beings and (ii) that theology somehow has
to address itself to their peculiar way of being, if only to assume what it is to
be for this kind of being and to assume that there are beings that have this pecu-
liar way of being.
Now, this way of being, peculiar to divine substances, I want to suggest, is
the focal way or sense of being in terms of which all other ways of being have
to be explained. This explanation comes in three steps. The reasons why Aristot-
le thinks that all other ways of being presuppose, and have to be explained in
terms of, the being of substances are well known and do not have to be rehearsed
here. Aristotle thinks, e.g., that the being of qualities can be understood only
in terms of the being of substances that are qualified in some way or other. But
it would be a mistake to assume, as it often is, that Aristotle thinks that his task
has been completed by showing how the various ways of being depend on the
way of being of substances. For as soon as we start to pursue the question what
is it to be for a substance, it turns out that this question has a single answer as
little as the question of what it is to be for a being does. Even in the case of sensi-
ble, perishable substances, it has at least three answers, one for matter, one for
substantial forms, and one for the composite of matter and form. What is more,
these three ways of being a substance stand in a certain relation of priority and
posteriority to each other. The being of a substance primarily belongs to the sub-
stantial form, only secondarily to the concrete physical substance in virtue of its
having a substantial form, and in a third way to matter, insofar as it potentially
is a composite substance. Thus, the focal way of being a substance, for sensible
substances, turns out to be the being a substance of substantial forms. And since
the ways of being of the entities in all the other categories depend on the way
of being of sensible substances, the way of being of substantial forms turns out
to be the focus for all non-substantial entities. Thus, in a second step, all ways
of being are shown to be dependent on the way of being of substantial forms.
But there are not just sensible, perishable substances. Aristotle at times distin-
guishes as many as three different kinds of substances: sensible, perishable sub-
stances; the imperishable heavenly bodies; and immaterial, nonperceptible sub-
stances (cf. A 6, 1071b 2ff.) And these are not just different kinds of substances
in the sense in which horses and donkeys are different kinds of substances. They
differ from each other qua substances. This is most easily seen if we just distin-
guish between sensible and nonsensible substances, as Aristotle himself some-
times does. The substantiality of sensible substances has to be explained in terms
of the substantiality of their substantial forms, whereas nonsensible substances
are just substantial forms. What is more, the substantial forms of sensible sub-
stances do not have the same way of being as the substantial forms that are sepa-
rate substances. The substantial forms of sensible substances, in order to be at
all, have to be realized in a composite substance that has various non-substantial
characteristics, size, weight, shape, color, etc. Separate substances, on the other
 88 UNITY OF METAPHYSICS

hand, exist without matter and without accidents. The unmoved mover, e.g.,
and quite generally divine substances, are such separate forms.
One may ask, though, what reason we have to believe that this difference in
the way of being of substantial forms is relevant for Aristotle's account of what
it is to be a substance and, hence, of what it is to be a being.
There are two passages in Met. Z which suggest that Aristotle himself thinks
that the difference is relevant. In Met. Z 11 (1037q llff.) he tells us
"whether . . . we have to look for a different kind of substance (i.e., im-
material substance), like numbers or something of this sort, we will have to see
later. For it is because of this that we try to get clear also about sensible sub-
stances. For in a way it is the task of physics and second philosophy to consider
sensible substances." Aristotle here clearly assumes that separate substances and
physical substances differ qua substances, and that in a way we in Met. Z only
discuss physical substance to get clearer about separate substance. We find a
similar thought in Z 3, 1029b 3ff. There we are told that we will start with a
consideration of sensible substances that are generally agreed to be substances.
Thus, it is said, we shall proceed from what is better known to us to what is
better known by nature, i.e., from what we are familiar with to what in the order
of nature and hence scientific knowledge is prior such that ultimately what we
are familiar with has to be explained in terms of it. Since both remarks are made
in the context of a discussion of the question "what is substance?," these texts
seem to suggest the following: (i) our discussion of the substantiality of sensible
substances is preliminary to a discussion of the substantiality of nonsensible sub-
stances; (ii) nonsensible substances qua substances are prior to sensible sub-
stances; and thus (iii) we shall achieve a full understanding of the substantiality
of sensible substances only when we have understood the substantiality of non-
sensible substances.
Though Aristotle, unfortunately, does not explain this relation, one can still
vaguely see what he must have in mind. Before we try to get clearer about this,
though, it is important to distinguish two kinds of priority and dependence. It
will be readily granted that, according to Aristotle, all other beings depend for
their being on the being of separate substances, in particular the prime mover.
And, hence, Patzig, e.g., originally thought that the relevant kind of dependence
consisted in the fact that a complete account of the being of anything will have
to make reference to the unmoved mover. But it seems that the kind of depen-
dence we need for our account is not primarily this quasi-causal dependence, but
a different kind of dependence. The kind of dependence we need for our account,
rather, is of the following kind: sensible substances are dependent on separate
substances qua substances, and this in the sense that their way of being a sub-
stance, and hence their way of being, has to be explained in terms of the way
separate substances are substances.
But how could this be? Aristotle in Met. A not only assumes that the first un-
 UNITY OF METAPHYSICS 89

moved mover is the primary being, but also that it is the primary intelligible ob-
ject (1072a 26ff.), thus giving rise to the medieval debate whether it is God or
being that is the first object of the intellect, and hence the primary subject of
metaphysics. Aristotle seems to assume not only that ultimately everything de-
pends on God for its being, but also that ultimately nothing is intelligible unless
it is understood in its dependence on God. And this in various ways. It is not
just that everything depends on God as its first cause. There is also the notion,
reflected in Aristotle in various ways, that lower forms of being somehow imi-
tate higher forms of being. Animals procreate; this is their way of sharing in
the eternal. The heaven eternally rotates to imitate, as well as it can, the un-
changing nature of the unmoved mover. This suggests a scale of perfection in
which the less perfect is to be understood in terms of the more perfect and ulti-
mately the unmoved mover, as if everything was like him in the limited way it
could be. But the central books of the Metaphysics seem to rely on a much more
precise notion, though this never comes out explicitly.
To understand what it is to be a substance, and hence what it is to be a being,
one has to understand, as Met. Z argues, what it is to be a substantial form or
essence. For even if we consider a composite physical substance and look for
what it is independently of its ever changing characteristics, look for what it is
that remains the same throughout its life-span, while its matter is changing, it
seems that it is the form that provides the object with its identity. But from the
thought that it is substantial forms that are the substances, there are two lines
of argument leading to the conclusion that it is separate forms that qua sub-
stances are prior to everything else.
The forms of sensible substances and separate substances are both substantial
forms and they are both actualities, i.e., the reality of the object they are is con-
stituted by their reality. But whereas separate substances turn out to be substan-
tial forms and actualities without qualification, the substantial forms of sensible
objects have to be understood as substantial forms and actualities of a certain
limited kind. Thus, to understand them properly one first of all has to understand
what it is to be a substantial form and an actuality without qualification, and then
to understand the qualifications with which the substantial forms of sensible sub-
stances are substantial forms and actualities.
The forms of sensible substances involve potentiality in two ways, and hence
are not pure actualities, though it is of the essence of a form to be an actuality.
They need matter to be realized in, and thus are the forms of objects subject to
change. But, what is more, when we turn to the paradigms of sensible sub-
stances, living beings, it turns out that their forms themselves essentially contain
an element of potentiality. When Aristotle in De anima II, 1 defines the soul as
the "first actuality" of a certain kind of body, this very language reflects the fact
that the soul in a way is constituted by the various abilities to exercise the life-
functions characteristic of the kind of living being in question, but that not all
 90 UNITY OF METAPHYSICS

these life-functions are exercised all the time. What is more, some of the abilities
that characterize the soul, like virtue or knowledge, are only acquired. Thus,
the forms of sensible substances are not pure actualities; they in part are con-
stituted by unrealized possibilities and in that sense are not fully real. The form
that is the unmoved mover, on the other hand, is pure actuality. It neither needs
matter to be realized nor does it involve any abilities that might or might not
be realized or exercised. The unmoved mover is just eternally thinking the same
thought. Thus, separate substances, in particular the unmoved mover, are pure
actualities, and thus forms, and thus substances, and thus beings in a paradig-
matic way in that they are perfectly real.
But, there is another line of argument which suggests that separate substances
are paradigmatic as substances. Perhaps the most important characteristic of
substances is that they exist in their own right, that they do not depend for their
existence on something else, or, as Aristotle puts it, are separate. Now this re-
quirement notoriously admits of various interpretations. But it seems that, on
any plausible interpretation of it, it is only separate forms that satisfy this re-
quirement straightforwardly. They do not in any sense need matter, or non-
substantial characteristics, i.e., qualities, quantities, places, etc., or anything
else to be realized. The forms of sensible substances are separate, too, but only
qualifiedly so, namely, separate in account; the account of a form is self-
contained in that it does not involve a reference to any other item in the ontology.
Still, a material form needs some matter, and the composite substance needs
nonessential properties, though these properties and their matter do not form
part of the account of the form. The second most important condition of sub-
stances is that they should be particular or individual. This, again, is a require-
ment satisfied straightforwardly only by separate forms. The individuality of
material forms, on the other hand, raises enormous problems, so much so that
one may wonder whether, or even think that, as tradition indeed did, the forms
of natural substances are universal. After all, the account of the form for all
things of the same kind is exactly the same. Definition seems to be of the univer-
sal, and form or essence seems to be exactly what is given by a definition. To
get particular forms at all we have to make up a complicated story about the way
they are individuated by their history. Similarly, there is the notorious problem
how the form of sensible substances, rather than the composite substances them-
selves or their matter, could come out as ultimate subjects of predication. Aris-
totle explicitly (Z 3, 1029a 2-3) commits himself to the view that there is a way
to construe the relation of being the subject of something in such a way that this
comes out as true. But he also takes the view that the more natural way to con-
strue the relation is such that it will be matter or the composite that comes out
as ultimate subject. Thus, this requirement is met by the forms of sensible sub-
stances only by a somewhat artificial construction of the relation. Separate
forms, on the other hand, are ultimate subjects of predication quite straightfor-
 UNITY OF METAPHYSICS 91

wardly. Hence, separate forms satisfy all three requirements of substancehood

mentioned in Z 3 straightforwardly, whereas the forms of sensible substances
meet them only in some indirect or qualified way.
Moreover, Aristotle characterizes the difference between the objects of the
three theoretical sciences by saying that the objects of theology are separate and
unchanging, the objects of physics separate and changing, the objects of
mathematics nonseparate and unchanging. Given this characterization of sepa-
rate forms, it would seem that the forms of natural substances somehow are an
inferior kind of forms in yet another way. For they are separate only qualifiedly,
namely, in account; and they are unchanging, but only qualifiedly. For though
they do not come into being or pass away, they, unlike separate forms, do not
exist eternally, but go in and out of existence instantaneously. And though they
do not suffer change, they really are different at different times, as one can see
in the case of human souls.
Moreover, only in the case of separate substances are the form and the es-
sence straightforwardly identical. For though Aristotle thinks that forms and es-
sences quite generally are identical, he at times also talks as if the specification
of the essence of a sensible substance in addition to a reference to the form had
to include a reference to the matter. Thus, there is much reason to think that one
will understand what it is to be a form, and thus what it is to be a substance,
only if one has understood how separate forms are substances, and then under-
stand how material forms, by a weakening of the conditions, count as forms and
substances. And if this should be so, then the focal way of being a substance,
and hence of being a being, is the way in which separate forms, or divine sub-
stances, are substances. Thus, general metaphysics would have as its core a
study of the way of being of divine substances.
But why would it follow from this that theology would be the natural place
to study the various ways of being, their systematical connections and thus being
qua being? There are at least two other possibilities. There could be a universal
discipline prior to theology, physics, and mathematics that studies the various
ways of being and hence being qua being. And there is the other possibility that
we leave it to the three sciences to study the ways of being peculiar to their
objects.
To take the last possibility first, it deserves to be pointed out that it is a real
possibility, one Aristotle himself seems to consider. For in the passage in Z 11
we looked at earlier, he does suggest that first philosophy is primarily concerned
with the substantiality of nonsensible substances, whereas the substantiality of
sensible substances in a way is the concern of physics. And since the way of be-
ing of nonsubstantial items like qualities and quantities depends on the way of
being of natural substances, since only they have accidents, the natural philoso-
pher would take care of all that, too.
But the shortcomings of this approach are obvious, too. Perhaps the most
 92 UNITY OF METAPHYSICS

important difficulty with this approach would be that it would divide the account
of being on three different subjects and thus would make it episodic, when in
fact there is a continuous story to be told and when it is important for an under-
standing of the different parts of the story that they are just different parts of one
account. It is for this reason that it would seem so much more attractive to give
a continuous account of being as such, prior to the different accounts of the
different kinds of beings, i.e., prior to theology, physics, and mathematics. And
some remarks Aristotle makes might suggeest that this is the way he is inclined
to deal with the problem. In E 1, 1025b lOff., e.g., he says that the different
particular disciplines do not account for the essence and existence of their
subject-matter, as if this was left to some prior discipline, i.e., general
metaphysics. Since theology, physics, and mathematics are particular disciplines
with a particular subject-matter, this might suggest that there is some discipline
prior to all three of them which somehow accounts for their subject-matter. But
the drawbacks of this approach are obvious, too. For there clearly is a sense in
which it is the theologian who knows best about the way of being of divine sub-
stances, the physicist who knows best about the way of being of natural sub-
stances, and the mathematician who knows best about the way of being of mag-
nitudes. If one insists that a full understanding of the way of being of magnitudes
presupposes an understanding of the way of being of physical substances, and
that this, in turn, presupposes an understanding of the way of being of separate
substances, it will at least be true of the theologian that he knows best about the
way of being of separate substances. For on the basis of what could somebody
else know more about the way separate substances are? This could only be the
case, if there were principles prior to separate substances in terms of which sepa-
rate substances have to be understood, and these principles were the subject of
some further discipline. But there are no principles prior to separate substances.
Moreover, in the case of divine substances knowledge of them to a large extent,
if not entirely, amounts to no more than a knowledge of their way of being. For
assume that there is just one separate substance. In this case one might try to
argue that whatever was true of it was true of it just in virtue of its being a sepa-
rate substance, i.e., in virtue of its peculiar way of being. But even if there
should be several separate substances, it still would be the case that much, if not
most, of what is true of them would be true in virtue of their being separate sub-
stances. Thus, a study of the way of being of separate substances outside theol-
ogy would, to a large extent, just reduplicate the study of the theologian. More-
over, as I suggested earlier, it would seem to be part of the essence of separate
substances that all other ways of being depend on their way of being, just as it
seems to be essential to natural substances that the way of being of accidents de-
pend on their way of being.
It is for these reasons, I take it, that Aristotle decides, on the one hand, to
 UNITY OF METAPHYSICS 93

give a continuous account of being as such, but, on the other, to do this within
the framework of "theology" or "first philosophy."
But it has to be kept in mind that, though this kind of ontology forms the core
of general metaphysics, there is more to general metaphysics than this. The
metaphysician, according to Aristotle, does consider certain principles and no-
tions of universal applicability, like the principle of noncontradiction and the no-
tions of unity and identity and their various forms. Why would it be the task of
the theologian to consider these matters? There are two possible explanations.
The first is that since these matters have to be discussed somewhere, and since
they are most naturally discussed in the context of ontology, it will be the task
of the theologian to deal with them, since it is his task to do ontology. But there
is another possible explanation which ties this fact to Aristotle's claim that first
philosophy is universal, because it is first. Since it is first, and since these princi-
ples and notions are universal, first philosophy will be the first place where they
are used. Hence, it will be the task of the theologian to introduce them.
Now, the only way in which theology could accommodate a general study of
being as such as part of itself is by lacking the kind of unity we might expect
of an Aristotelian science on the basis of what Aristotle says in the Posterior
Analytics. But we have to keep in mind that, as we noticed earlier, physics and
mathematics, too, are one science only in the sense that there is one subject-
matter that is studied by a series of systematically connected disciplines. More-
over, Aristotle in F 1003b 22 explicitly warns us that the study of being will be
one science only generically, that it will be constituted by a series of studies that
specifically deal with the various kinds of being. Thus, there is no reason why
theology, too, should not be one only generically, why it should not actually
consist of a series of studies. There is one disanalogy, though. Theology would
not have one subject-matter in quite the straightforward way in which physics
and mathematics have one subject-matter. It not only deals with the primary kind
of beings, and thus the primary way of being, but also with being in general.
Yet, there is also a sense in which these are not really two different subject mat-
ters. For to say of the primary way of being that it is primary, is to say that all
other ways of being depend on it. And thus Aristotle can say that theology, being
first, also is universal.
If this is correct, it also throws some light on the question whether
Aristotelian metaphysics is a science. It has often been claimed that Aristotelian
metaphysics is a dialectical enterprise, rather than a science in something like
the sense of the Posterior Analytics. But it should be clear from what has been
said that Aristotelian metaphysics, either as a whole, or at least in good part,
is scientific. For otherwise, Aristotle could not in E 1 treat it as parallel to
physics and mathematics. Moreover, it does seem that there is no reason why
theology, at least in the narrower sense, should not be a science strictly speak-
 94 UNITY OF METAPHYSICS

ing. But one can also see how, given the suggested conception of Aristotle's
metaphysics, a much larger part of first philosophy could be considered as being
strictly scientific. Both physics and mathematics, in addition to studies of the
various kinds of natural substances and mathematical entities, involve a univer-
sal discipline that deals with natural substances, and magnitudes respectively, as
such. In the case of mathematics this is obvious, since Aristotle explicitly refers
to universal or general mathematics. But this is hardly less clear in the case of
physics. For our treatise called "Physics" is the dialectical counterpart to such
a science of natural substances quite generally. Thus, we may assume that theol-
ogy, too, involves a discipline that we may call general or universal theology
which will deal with separate substances as such, and, in doing so, with the way
of being peculiar to separate substances. And there is no reason why general the-
ology should not be strictly scientific, too. Immediately following general the-
ology we shall have an exposition of the various other ways of being and how
they depend on the way of being of separate substances. There is no reason why
this exposition should not take a strictly scientific form with, e.g., definitions
of qualities, quantities, relations, etc., and theorems concerning them. Having
dealt with whatever needs to be dealt with in a study of being as such, we would
move on to special theology which, again in a scientific way, would deal with
the various particular kinds of separate substances, in case there should be more
than one. Thus, it would seem that, if not the whole, then at least a very substan-
tial part of metaphysics is scientific. But what reason is there to think that any
part of metaphysics will not be scientific? The view that Aristotelian metaphysics
is essentially a dialectical enterprise rests primarily on an interpretation of Met.
T, and in particular the way Aristotle deals with the principle of noncontradic-
tion. Aristotle says that principles like this are the concern of the metaphysician,
but also says that they do not admit of proof in the strict sense, to then embark
on a nonscientific discussion of the principle, which is regarded as an actual
piece of doing metaphysics. This is the only evidence T. Irwin, e.g., refers to
in order to arrive at the conclusion that first philosophy is not demonstrative,
that the questions discussed by first philosophy are beyond demonstration
(Aristotle's Discovery of Metaphysics, Rev. of Nov. 31, 1977, 78, p. 218). But
this inference seem to me to be radically mistaken. For consider the following:
another concern of the metaphysician must be God, his essence and existence,
since God is a first principle. Moreover, Aristotle is committed to the view that
strictly speaking there is no proof of the essence and existence of God. There
will be a real definition of him as an axiom of special theology. And on the basis
of this, there will be a deduction of theological theorems. Now imagine that we
only had a text in which Aristotle says that the theologian is concerned with the
existence of God, but that his existence does not admit of any proof in the strict
sense, and then proceeded to give the kind of a posteriori proofs on the basis
of the existence of motion which he in fact does give in the Physics and in the
 UNITY OF METAPHYSICS 95

Metaphysics. We could not possibly infer from this that Aristotle thinks that the-
ology is not a science; here, we clearly could not argue that the proof constituted
a piece of actual Aristotelain theology, but obviously was not, and could not by
Aristotle be thought to be, demonstrative. For we would know that the assump-
tion of God's existence was just a starting-point for scientific theology from
which it would then deduce whatever there was to be deduced. Or consider the
existence of nature. As we can see from the Physics, Aristotle thinks that it can-
not be proved, strictly speaking, that nature exists. Nevertheless, he tries to es-
tablish it dialectically. But nobody is tempted to infer from this that Aristotle
thinks that physics is not a science; nobody is inclined to argue that the discus-
sion of the first book of the Physics is a piece of actual Aristotelian physics, and
that, hence, Aristotelian physics is dialectical. But if these inferences would be
misguided, it is equally misguided to infer from Met. F, or the Metaphysics in
general, that Aristotelian metaphysics is not a real science. It is demonstrative,
but, as with all other sciences, one arrives at its starting-points dialectically. And
just as all other writings of Aristotle's do not pretend to present us with an actual
piece of Aristotelian science, but, rather, show some of the work involved in
arriving at the proper scientific axioms, so the Metaphysics, including Met. F,
show us how we might arrive at the principles of the theory of being as such.
In one important respect, though, metaphysics differs from all other sciences.
Since it is the first in the hierarchy of sciences, it will also be the first to make
use of the principles that all sciences will rely on, like, e.g., the principle of non-
contradiction. And, hence, it will also be the task of the metaphysician to arrive
at these universal principles and to provide them with whatever backing may be
appropriate. Thus, the metaphysician, more than any other scientist, will be en-
gaged in dialectical reasoning. But this goes no way to show that metaphysics
is not a demonstrative science.
Thus, it seems to me that a proper understanding of Aristotle's conception of
metaphysics, and in particular his remark that first philosophy is universal be-
cause it is first, will restore metaphysics to its proper place, that of the first of
the demonstrative sciences.