Scotus on Doing Metaphysics in statu isto
Giorgio Pini
Scotus makes the interesting claim that those who think that being is
not univocal destroy philosophy. In his Oxford Lectura, we read: “I say that
[by positing being as univocal] I do not destroy philosophy, but those who
posit the contrary [i.e. that being is not univocal] necessarily destroy
philosophy”.1 How should we interpret this claim? It is easy to understand that
Scotus is willing to attach great importance to a doctrine that struck his
contemporaries as novel and created considerable controversy. All the same,
to state that to deny univocity is to destroy philosophy sounds as exaggerated.
For one thing, one may regard Scotus’s claim as bizarre. Before Scotus, a large
amount of philosophy had already been done, but no one had ever thought of
arguing systematically for the univocity of the concept of being. Furthermore,
even somebody looking at Scotus’s innovation with sympathy may seriously
doubt whether this doctrine plays such a fundamental role in philosophy. As is
well known, the gist of Scotus’s doctrine is that when we think of God as a
being or of a creature as a being, as well as when we think of a substance as a
being or of an accident as a being, we are making use of one and the same
concept of being.2 So why should we suppose that this doctrine plays such a
fundamental role in philosophy?
When Scotus makes his claim about the role of univocity in
philosophy he is specifically thinking of the univocity of being with regard to
substances and accidents. Let us follow Scotus and leave out the part of his
doctrine that concerns God and creatures. The claim that substances and
accidents are thought of as beings in the same sense of the word ‘being’ is
1
Lect. I, d. 3, p. 1, q. 1-2 (Vat. XVI, 265): “Dico quod non destruo philosophiam, sed
ponentes contrarium necessario destruunt philosophiam.”
2
See Lect. I, d. 3, pars 1, qq. 1–2, nn. 21–34 (Vat. XVI, 232–37); Lect. I, d. 8, p. 1, q. 3,
nn. 60–88 (Vat. XVII, 20–30); Ord. I, d. 3, p. 1, q. 3, nn. 131–166 (Vat. III, 81–103); Ord. I, d. 8,
p. 1, q. 3, nn. 44–89 (Vat. IV, 171–195). The literature on Scotus’s doctrine of univocity is vast.
See S.D. Dumont, “Scotus’s Doctrine of Univocity and the Medieval Tradition of Metaphysics,” in
Was ist Philosophie im Mittelalter? Miscellanea Mediaevalia 25, eds. J.A. Aertsen and A. Speer
(Berlin and New York: de Gruyter, 1998), 193–212. On the context of Scotus’s doctrine, see S.
Donati, “La discussione sull’unità del concetto di ente nella tradizione di commento della
“Fisica”: commenti parigini degli anni 1270-1315 ca.,” in Die Logik des Transzendentale.
Festschrift für Jan A. Aertsen zum 65. Geburstag, Miscellanea Mediaevalia 30, ed. M. Pickavé
(Berlin and New York: de Gruyter, 2003), 60–139; G. Pini, “Univocity in Scotus’s Questions on
the Metaphysics: The Solution to a Riddle,” Medioevo 30 (2005): 69-110.
1
controversial enough. For this claim seems to be in stark opposition with one
of the central tenets of Aristotle’s metaphysics. Being, according to Aristotle, is
said in many ways. This is usually taken to entail that ‘to be’, for a substance,
means to be a substance; by contrast, ‘to be’, for a quality, means to be a
quality, and so on for any of the Aristotelian category. ‘Being’ is an ambiguous
term to which no single concept corresponds.3 Since the claim that being is
said in many ways seems to be at the basis of any attempt to do philosophy in
a sober Aristotelian way, it is not surprising the first criticism Scotus has to
face is that to posit being as univocal destroys all philosophy.4 It is against this
criticism that Scotus retorts that it is not he the one who destroys philosophy;
quite the opposite, it is those who deny that being is univocal who make
philosophy impossible.
Before dismissing Scotus’s claim as a rhetorical exaggeration, I think
that we should give it serious consideration from a philosophical point of
view. My point in this paper is to illustrate in some detail the argument that
lies behind Scotus’s claim. Specifically, I will do five things. First, I will
consider Scotus’s claim that those who deny that being is univocal cannot
account for our cognition of substances. Second, I will show why the question
of how we cognize substantial essences was such an important issue for
Scotus and his predecessors and contemporaries—an issue that lay at the very
foundation of philosophy. Third, I will consider in some detail Scotus’s
treatment of this topic and his evolution as it can be traced through his
Questions on the Metaphysics to his Lectura and Ordinatio. Fourth, I will
consider Scotus’s explanation of why we do not have direct cognition of
substances and consequently why the univocal concept of being is so
important for us. Fifth and finally, I will turn to some unexpected
consequences of Scotus’s position and his attempts to relocate Aristotle’s
metaphysics in the light of these consequences.
I
Let us start focusing on Scotus’s claim that those who deny that being
is univocal destroy philosophy. This claim is particularly interesting for two
reasons. First, it sheds some light on the way Scotus himself considers his
doctrine of univocity. He does not regard it as a daring innovation on
Aristotle’s metaphysics. Quite the contrary, he considers it as a necessary
3
Aristotle, Meta. VII, 1, 1028a10–13. The interpretation of Aristotle’s notion of focal
meaning or multivocity is in itself controversial. See C. Shields, Order in Multiplicity. Homonymy
in the Philosophy of Aristotle (Oxford: Clarendon Press, 1999).
4
Lect. I, d. 3, p. 1, q. 1-2, nn. 105-109 (Vat. XVI, 264–5).
2
move to rescue Aristotle from the inconsistencies that has plagued his
commentators’ and followers’ works so far. According to Scotus, at the basis
of any good way of doing philosophy there is the assumption—no matter
whether explicitly or implicitly made—that being is univocal. Scotus’s
intention is not to create a new metaphysical system, but to make that
assumption explicit and to provide a defensible version of Aristotelianism at
last free from contradictions. Accordingly, he regards his metaphysics as a
vindication of good philosophy, that is to say, in his opinion, of
Aristotelianism. Second and more specifically, Scotus’s claim is interesting
because of the argument he gives in its support. For he goes on to say:
For if being did not have a common concept, it would be impossible
for us to have the concept of substance, because substance does not have a
proper species in the possible intellect. But [we] only [have] the concept of
being obtained by way of abstraction from the species of accidents.
Therefore, if the concept of being were not one, we would have no concept
of substance, neither of substance in general nor of a substance in particular.
(Trans. mine)
5
Thus, the main reason why we should posit that being is a univocal
concept is that, if it were not, we could not account for our knowledge of
substances. Scotus specifies that, without a univocal concept of being, we
could have neither a concept of substance in general—i.e. our concept of
what underlies accidents—nor of specific substances, such as dogs, cats, and
human beings. Obviously, however, we do have both a concept of substance
in general and concepts of specific substances. It follows that our concept of
being is univocal.
The crucial step in this argument is clearly the major premise, i.e. that
we could not have the concept of substance if being were not a univocal
concept. Scotus’s demonstration of this premise will be the focus of the
second part of this paper. But before turning to this part of Scotus’s argument, I
would like to notice that Scotus’s main philosophical reason to posit that
being is a univocal concept does not have to do with the topic of the subject
matter of metaphysics (whether God or being) or with the object of our
intellect (whether material substances or being). Both these issues will be of
5
Lect. I, d. 3, p. 1, q. 1-2, n. 110 (Vat. XVI, 265): “… quia si ens non importaret
conceptum communem, impossibile esset quod haberemus conceptum substantiae, quia
substantia non habet propriam speciem in intellectu possibili, sed tantum conceptum entis
abstrahendo a speciebus accidentium. Si ergo ens non haberet unum conceptum, nullum
conceptum—nec in communi nec in particulari—haberemus de substantia.” See also Ord. I, d. 3,
p. 1, q. 1–2, nn. 139, 140, 145 and 146 (Vat. III, 86, 88, 89, 90).
3
course relevant to our discussion. I should also add that the attention of
Scotus’s scholars have been attracted almost exclusively to these issues. The
issue of our knowledge of substances, however, has been relatively neglected.
But it is this issue that Scotus indicates as the main philosophical motivation
behind the adoption of doctrine of the univocity of being.
The question that Scotus is concerned with is the problem of how we
cognize substances, both the concept of substance in general and the
concepts of specific substances. I think that here we should take Scotus as
referring to our concepts of specific essences or natures, i.e. our concepts of
what a dog is, what a cat is, what a human being is, and so on. In what
follows, we should consider three questions. First, why is this a problem at all
for Scotus? Second, why does he think that the solution to this problem is so
central to philosophy? Third, why does he think that we can solve this
problem only if we posit that being is univocal?
This is one of the cases in which a chronological consideration of
Scotus’s works sheds some light on their philosophical value. Virtually absent
in his logical commentaries, the question of how we cognize substantial
essences makes its appearance in the first draft of a question on the
Metaphysics. Afterwards, we can closely follow Scotus’s attempt to come to
grips with this issue in several long and tortured passages added to his
Questions on the Metaphysics. In the Lectura and the Ordinatio we finally find
Scotus’s solution. But it is only in a handful of passages in the Ordinatio and
in some probably late additions to the Questions on the Metaphysics that we
find Scotus’s attempt to draw all the consequences of his solution.
Unfortunately, these late elaborations do not result in a fully worked-out view.
They are sufficiently clear, however, to indicate what Scotus thinks of the
nature and method of metaphysics in his maturity. In any case, a careful
consideration of his writings shows that Scotus comes to deny that we have
any cognition of what a substantial nature is (both in general and specifically
for each kind of substantial nature) apart from the fact that it is something or a
being.
To anticipate the main point of my argument, I think that Scotus’s
remarks about our cognition of substantial essences allow us to say that
possibly the driving motivation behind Scotus’s metaphysical output is to posit
Aristotle’s metaphysics on firm ground while at the same time taking into full
consideration the limits of our knowledge. Scotus tries to reach that goal by
carefully separating two questions. On the one hand, there is the question of
how things are, independently of whether and how we know them. On the
other hand, there is the question of what we can know in our present
condition. The former question is properly metaphysical, and Aristotle’s
4
Metaphysics should be read with this question in mind. The latter question
concerns cognitive psychology in our present condition (in statu isto).
Aristotle’s De anima should be read with that question in mind. As Scotus
said, when we take into account the former question, we speak
metaphysically (metaphysice). By contrast, when we take into account the
latter question we speak psychologically (animastice).6 Scotus thinks that these
two questions should be kept separated. Many mistakes of his predecessors
should be ascribed to their incapacity to keep these two questions separated,.
As a result, the metaphysical treatments of many of Scotus’s predecessors
were, in Scotus’s assessment, confused and plagued by contradictions. By
contrast, Scotus’s intention is to take into full account the cognitive limitations
that our present condition (due probably to the Fall) presents us with.
Accordingly, his attempt may be described as that of giving a solid foundation
to metaphysics in our present condition (in statu isto), that is to say, after the
Fall and before the beatific vision. His attempt to give metaphysics autonomy
from cognitive psychology may perhaps be considered cognate to his attempt
to separate logic from metaphysics and with his attempt to separate some
aspects of cognitive psychology, such as his explanation of intentionality, from
natural philosophy more in general.7
II
Let me now turn to the second part of this paper. I will try to answer
the three questions I have posed above. First, why is the way we cognize
substances (i.e. substantial essences or natures) a problem at all for Scotus?
Second, why does Scotus think that the solution to this problem is so central
to philosophy? Third and finally, why does he come to think that we can solve
that problem only if we posit that being is univocal?
Specific substantial essences or natures such as humanity (i.e. what a
human being is) or horseness (i.e. what a horse is) are the basic constituents of
reality in Scotus’s metaphysics. One of Scotus’s most famous claims is that
6
Quaest. Metaph. VII, 15, nn. 12–32 (OPh IV, 297–306).
On Scotus’s separation between logic and metaphysics, see G. Pini, “How Is Scotus’s
Logic Related to His Metaphysics? A Reply to Todd Bates,” in Medieval Commentaries on
Aristotle’s Categories, ed. L. Newton (Leiden and New York: Brill, 2008), 277–94. On Scotus’s
cognitive psychology, see R. Pasnau, “Cognition,” in The Cambridge Companion to Duns Scotus,
ed. T. Williams (Cambridge: Cambridge University Press, 2003), 285-311; P. King, “Rethinking
Representation in the Middle Ages,” in Representation and Objects of Thought in Medieval
Philosophy, ed. H. Legerlund (Aldershot: Ashgate, 2005), 83-102.
7
5
these essences have a mind-independent, less-than-numerical unity.8 It is not
necessary to deal with this doctrine in detail. Suffice it to say that Scotus
argues that an essence such as humanity has its own unity or identity and that
this unity is different from and independent of both the unity that an individual
human being has and the unity that our concept of humanity has.
As I have said, Scotus considers these essences as the basic
constituents of reality. Accordingly, he holds that metaphysicians, whose task
is to study the structure of reality, are mostly concerned with these essences. It
is these essences that metaphysicians define by their definitions. And it is
these essences that are the truth-makers of the metaphysicians’ preferred type
of propositions, namely true per se propositions such as “Human beings are
rational animals.” As Scotus says in his Ordinatio:
… In accordance with this natural priority [namely, the priority of
the essence considered by itself over the essences as it is in the individuals
and as it is cognized by way of a universal concept], the essence is the per se
object of the intellect and is per se, as such, considered by the metaphysician
and expressed by a definition. True propositions in the first per se mode are
true by virtue of the essence so taken.
9
Since accidental essences such as blackness and whiteness are in
some way dependent on substantial essences or natures such as horseness and
humanity,10 it seems to be safe to assume that here Scotus is mainly referring
to substantial essences or natures. The metaphysicians’ definitions are real
definition of those essences. So, since these substantial essences are what
metaphysicians define and what they talk about, it seems necessary to
conclude that, in order for metaphysics to be possible at all, we must cognize
those essences. The content of this cognition is precisely what our definitions
and true per se propositions spell out. If substantial essences such as horseness
8
Lect. II, d. 3, p. 1, q. 1, nn. 8–32 (Vat. XVIII, 230–7); Ord. II, d. 3, p. 1, q. 1, nn. 7–34
(Vat. VII, 394–405).
9
Ord. II, d. 3, p. 1, q. 1, n. 32, p. 403: “… secundum prioritatem naturalem est ‘quod
quid est’ per se obiectum intellectus, et per se, ut sic, consideratur a metaphysico et exprimitur
per definitionem; et propositiones ‘verae primo modo’ sunt verae ratione quiditatis sic acceptae,
quia nihil dicitur ‘per se primo modo’ de quiditate nisi quod includitur in ea essentialiter, in
quantum ipsa abstrahitur ab omnibus istis, quae sunt posteriora naturaliter ipsa.” The translation is
taken, with some little modifications, from Five Texts on the Mediaeval Problem of Universals.
Porphyry, Boethius, Abelard, Duns Scotus, Ockham. Translated and edited by P. V. Spade
(Indianapolis and Cambridge: Hackett Publishing Company, 1994), 63.
10
On accidents’ dependence on substances, see G. Pini, “Substance, Accident, and
Inherence: Scotus and the Paris Debate on the Metaphysics of the Eucharist,” in Duns Scot à Paris
1302-2002. Actes du colloque de Paris 2-4 septembre 2002, eds. O. Boulnois et al. (Turnhout:
Brepols, 2004), 273–311.
6
and humanity were “something I know not what”, to use Locke’s famous
formula, the metaphysicians’ definitions could not certainly be counted as real
definitions. They could at best be definitions of the terms we use to speak of
them.
At first sight, the idea that we cognize substantial essences does not
seem to pose any problem. Substances are not just cognizable; they are the
first things we cognize, according to Aristotle.11
So how do we cognize substances? According to an account of
cognition common in the late thirteenth and early fourteenth century, we
cognize something when we receive in our intellect its intellectual likeness,
i.e. its so-called ‘intelligible species’. Thanks to the reception of this likeness,
an extramental thing becomes present to our intellect and is the object of the
intellect’s first act of cognition, the so-called act of simple apprehension. This
likeness or species present in the intellect is abstracted by the agent intellect
from some data present in the phantasm. The phantasm is in turn the result of
imagination’s re-elaboration of the data coming from the senses in the form of
sensible likenesses, the so-called ‘sensible species’.12 This is admittedly a very
crude description of how cognition takes place and there are many points that
need clarification. But I think that this account captures the main traits of the
theory of cognition to which Scotus subscribes. The point that we should
retain from this picture is that we cognize a certain thing when we receive its
intelligible species.
Scotus adopted a sophisticated version of this account of cognition.
He argued that, in order to cognize something by an act of abstractive
cognition, we need to receive its intellectual likeness. This intellectual likeness
does not come to our intellect by any form of illumination. It is just the reelaboration carried out by our natural faculties over information coming from
the senses.13
Accordingly, the answer to the question ‘how do we cognize
substances?’ seems to be straightforward. Our senses receive some
information from the external world by way of sensible species; our faculties
elaborate on that information; finally, our agent intellect abstracts an
intelligible species that contains the same information as the sensible species
11
Aristotle, Metaph. VII, 1, 1028a31–b1.
For two treatments focused on Aquinas but useful as general introductions to these
issues, see R. Pasnau, Thomas Aquinas on Human Nature. A Philosophical Study of Summa
theologiae Ia 75–89 (Cambridge: Cambridge University Press, 2002), 267–329; E. Stump, Aquinas
(Routledge: London, 2003), 262–76.
13
Specifically on Scotus’s defense of the intelligible species, see P. King, “Scotus on
Mental Content,” in Duns Scot à Paris, 1302-2002, 65–88.
12
7
does but in a generalized form. This intelligible species is stored in the
possible intellect, and our cognition of substantial natures can take place. So
substantial essences, which by themselves are mind-independent things, are
cognized when their likeness or intelligible species becomes present in the
intellect. In this way, substantial essences are both something outside our
intellect and present in our intellect by way of their likeness or intelligible
species.14
But there is a big problem with this account. For this account assumes
that, in order to cognize something, its likeness or intelligible species must be
present in the intellect. If there is no likeness or species, there is no
cognition—at least, no direct cognition of anything. By inference or by
analogy we can indeed get some sort of glimpse of things that we cannot
sense and of which, consequently, we do not have any intelligible species.
This is notoriously the case of the knowledge we have of God in this life. But
the problem is that a good Aristotelian seems to be forced to concede that we
cannot have any intelligible species of material substances either. The reason
is clear enough. All our cognition originates from the senses. This means that
all the content of our acts of cognition ultimately comes from the senses. It is
true that our faculties, among which the intellect plays a prominent role,
elaborate on that content. But our faculties can only re-work and elaborate on
what they receive from the world, they cannot add any new content to it. It
follows that we have direct cognition only of what can be sensed and of what
can be abstracted from the information conveyed by the senses. Now, the
crucial fact is that only accidents can be sensed—qualities such as colors,
sounds, smells and the like.15 Of course, the information coming from the
senses is subjected to abstraction by the intellect. The operation of abstraction,
however, is an operation of generalization. As I have said, no new content can
be added to the information coming from the senses. So, for example, from a
particular shade of white our intellect gets the concept of white, the concept
of color and the concept of quality. But it seems that neither the substance in
which that shade of white inheres nor the concept of substance in general can
be reached by means of this process of generalization. So if we just take into
account the intellect’s first act of cognition, namely the so-called act of simple
14
This was the way Aristotle’s famous claim that cognition is an identity between
knower and thing known was interpreted by the defenders of the account of cognition through
intelligible species towards the end of the thirteenth century. Specifically on Aquinas, see Stump,
Aquinas, 273–5.
15
Ord. I, d. 3, p. 1, q. 1, n. 139 (Vat. III, 87). On sensible accidents, see Aristotle, Cat.
8, 9a35-b7; De gen. et corr. II, 2, 329b19. See also Pasnau, Thomas Aquinas on Human Nature,
181.
8
apprehension, it seems that our intellect cannot reach substantial natures.
Only accidents can be cognized by abstraction and simple apprehension.
Substantial natures, if they are cognized at all, must be reached in another
way.
This problem is well known to Scotus and his contemporaries. It can
be summed up as a contrast between the requirements of Aristotle’s cognitive
psychology on the one hand and of his metaphysics and theory of science on
the other hand. On the one hand, by way of intelligible species we can
cognize only accidents, not substances. On the other hand, substantial natures
are the object of real definitions, so we must get some cognitive access to
them. This contrast can be found in Aquinas, who may have been the first to
come to terms with it, even though his attempts to solve it do not seem to be
completely successful. Aquinas recognizes that the only possible way to
acquire cognition of substances is not by direct acquaintance, but by
inference. We start from information concerning accidents, and from that
information we infer that there must be a subject behind accidents, even
though we do not have any direct acquaintance with that subject. So
substantial natures can be cognized only as the result of a discursive act—an
act of reasoning, not an act of simple apprehension and abstraction. Aquinas’s
problem was that, for several reasons, he was unwilling to give up the idea
that substantial natures are the object of the intellect’s act of simple
apprehension. As a result, there is an inner tension in his account of how we
cognize substances.16
The generation of thinkers who came after Aquinas could not avoid
dealing with this problem. Several solutions were suggested. Some thinkers
granted that substances do have intelligible species, even though of a very
peculiar kind, i.e. arrived at by an act of gathering together information that
was sometimes called ‘collatio’ and which does not seem to be clearly
distinguishable from an inferential act.17 Here I will briefly mention another
solution to this question, whose main elements can be found in the Franciscan
16
Several times Aquinas indeed stated that we do not know the substantial differences
of things and that we use accidental differences to refer to them instead a consequence. See for
example Sent. IV, d. 14, q. 1, sol. 6, ad 1; Sent. IV, d. 44, q. 2, a. 1, sol. 1, ad 1; De Ver. q. 4, a. 1,
ad 8; De Ver. q. 10, a. 1, ad 6; ST I, q. 29, a. 1, ad 3; ST I, q. 77, a. 1, ad 7.
17
See Vital du Four, Quaestiones de cognitione, q. 5, in F. Delorme, “Le cardinal Vital
du Four. Huit questions disputées sur le problème de la connaissance,”Archives d’histoire
doctrinale et littéraire du Moyen Age 2 (1927): 151–337, esp. 252–72. See also J.E. Lynch, The
Theory of Knowledge of Vital du Four (St. Bonaventure, NY: The Franciscan Institute Publications,
1972), 113–22; A. Robert, “L’universalité réduite au discours. Sur quelques théories franciscaines
de l’abstraction à la fin du XIIIe siècle,” Documenti e studi sulla tradizione filosofica medievale 18
(2007): 363-393. I also thank Timothy Noone for sharing with me the text of a still unpublished
lecture he gave on this topic at the Antonianum (Rome) in the Fall 2007.
9
Richard of Middleton, who wrote in the 1280s. According to Richard of
Middleton, we simply have to give up the idea that we can have an act of
simple apprehension of substances. Since there is no intelligible species of
substances, they cannot be cognized by a non-discursive act of cognition. But
this does not mean that substances cannot be cognized at all. They are indeed
cognized, but by an act of reasoning, i.e. by inferring their existence from the
existence of accidents and by inferring something about their nature from the
kind of accidents that inhere in them. For example, we can infer that the
substance in which color inheres is a material substance. Accordingly, we can
end up knowing very much about substances. But our cognition takes place
by a discursive act of reasoning and not by simple apprehension.18
But are we certain that we do not have intelligible species of
substances? Maybe the premise from which this conclusion is reached should
be rejected or revised. If this were the case, it would be much easier to
account for our cognition of substances. Richard of Middleton, however, has
several arguments in support of the view that we do not have intelligible
species of substances, quite independently of the Aristotelian account of
cognition that he adopts. Among those arguments, there is one that should
retain our attention, as Scotus will also use it to prove that we do not have
intelligible species of substances. The argument runs as follows. If we had an
intelligible species of a substance, we would be able to know whether a
substance is present or not. For any time a certain substance is present, we
would receive its species. By contrast, any time a certain substance is absent,
we would not receive its species. Now, in Transubstantiation the substance of
bread is present before Consecration and absent after Consecration but the
accidents of bread (its color, smell, taste, etc.) are present both before and
after Consecration. So, if we indeed had an intelligible species of bread’s
substance providing us with some direct, i.e. non-discursive, cognitive access
to such a substance, we would be able to perceive a difference in the Host
before and after Consecration, for after Consecration we would be able to
know that the essence of bread is absent. But this is not the case. Our
perception of the Host is the same before and after Consecration. Therefore,
both Richard and Scotus conclude that we do not have any intelligible species
and in general any non-discursive cognitive access to substance. We merely
infer that there is a substance from the presence of the accidents.19
18
Richard of Middleton, Super quatuor libros Sententiarum Petri Lombardi quaestiones
subtilissimae, II, d. 24, art. 2; Tomus II (Brixiae, 1591; reprint: Frankfurt am Main: Minerva, 1963,
309–11.
19
Richard of Middleton, Sent. II, d. 24, art. 2, 310; Scotus, Lect. I, d. 3, p. 1, q. 1-2, n.
111 (Vat. XVI, 266); Ord. I, d. 3, p. 1, q. 1, n. 140 (Vat. III, 88). As Stephen Dumont has pointed
10
III
So far, we have seen why the question of whether and how we
cognize substances is a problem for Scotus. We have also found out why this
is such an important issue for Scotus and his contemporaries. Since substantial
essences are the basic constituents of reality and what most of a
metaphysician’s statements are about, if we have no cognitive access to them,
it is indeed not an exaggeration to say that the entire philosophical enterprise
is at risk. So we have found an answer to the first and second question that we
posed above. It now remains the third question: why does Scotus think that
we could get some cognitive access to substantial essences only if we posit
that being is univocal?
Scotus came to see that this was the case only gradually. We can
trace Scotus’s evolution thanks to the study of several of his works, notably his
Questions on the Metaphysics, his Lectura and his Ordinatio, as well as of
several additions to his Questions on the Metaphysics. I suggest distinguishing
three main stages in Scotus’s consideration of the issue. First, Scotus still
defends the view that we do have direct (i.e. non-inferential) cognitive access
to substantial essences, even though he admits that we cognize accidents
before substances, so that substances’ priority in knowledge cannot be
interpreted temporally; all the same, when substances are indeed cognized,
they are most perfectly cognized. Second, Scotus comes to realize that that
position cannot work and struggles to find a new solution. Third, he finds a
new solution to the question of how we cognize substantial essences thanks to
his doctrine of the univocity of being.
So let us start with the first stage of Scotus’s evolving position on the
cognition of substantial natures. In this first stage, Scotus thinks that we do
have a non-inferential grasp of substantial essences. He holds this position
both in his logical commentaries and in what we should consider as the first
out to me, this argument does not take into account God’s power to create an intelligible species
of the substance of bread even when no bread is present, i.e. after Transubstantiation has
occurred. Scotus clearly contemplates such a possibility: see Ord. I, d. 3, p. 3, q. 2, nn. 477–9
(Vat. III, 285–6). So Richard’s and Scotus’s argument is conclusive only if we make the further
assumption that God does not create in us an intelligible species of bread after Transubstantiation.
The point is well taken. I think, however, that positing that that God creates in us an intelligible
species of bread after Transubstantiation would force us to assume that God positively misleads us
into believing that there is something that actually is not there, i.e. bread. This assumption seems
to me very problematic. It seems to be preferable to posit that God has no intention to mislead us
and that our incapacity to detect the presence or absence of a substance is due to some limitation
of our cognitive powers.
11
draft of his Questions on the Metaphysics. Scotus is aware that the claim that
substance is prior in knowledge cannot be given a temporal interpretation. As
a matter of fact, we cognize accidents before substances. For example, I first
receive the impression of softness and blackness before forming the concept of
a cat. All the same, Scotus still maintains that substantial essences, when they
are finally grasped, are grasped by way of a simple act of apprehension.
Moreover, he also thinks that substantial essences, when they are actually
cognized, are cognized more perfectly than accidents. For that reason,
substances are said to be prior in knowledge, even though they are not
temporally prior.20 It is in the Questions on the Metaphysics that Scotus gives
the clearest account of his position at this stage. He devotes a short but very
clear question to this issue. When discussing Aristotle’s statement that
substance is prior in knowledge, he is confronted with the problem that we
have noticed above.21 How should we interpret Aristotle’s statement? Isn’t it in
contrast with the view that accidents are the first things known, and as a
matter of fact the only things known by an act of direct cognition, namely a
non-inferential act?22
Scotus first considers several answers to this question but he is not
satisfied with any of them.23 His personal solution is that we should distinguish
between two meanings of the expression ‘more perfect object of cognition’
(perfectius cognoscibile). In the first sense, ‘more perfect’ qualifies the object
that is cognized. In the second sense, ‘more perfect’ qualifies the relationship
to its cognoscibility, namely how much of that object can be cognized. In this
second sense, the expression perfectius cognoscibile could probably be
20
Quaest. super Praed., 4, n. 53 (OPh I, 290): “… primum obiectum intellectus potest
unico actu intelligendi intelligi… dico quod primum obiectum intellectus est substantia. Quia, ut
dicit Aristoteles in principio VII Metaphysicae, “substantia est primum omnium entium
congitione”, quod non intelligitur de prioritate temporis sed naturae, scilicet quod cognitio eius
est perfectissima.” See also Quaest. super Metaph. IV, 1, nn. 86: “Ad primum pro opinione
Avicennae dico quod communissima sunt primo intellecta, et decem sunt communissima. Non
tamen ista omnia sunt primo intellecta, sed substantia, ad quam omnia reducuntur. Et substantia
est prius non praedicatione ad novem genera, sed prius perfectione et causa; nec est dare aliquod
commune decem quod primo intelligatur” (OPh III, 319); n. 87: “Ad aliud: quod maior [lege:
minor, scil., quod primum obiectum intellectus est ens] est falsa; sed ilud est substantia” (OPh III,
319). I have argued elsewhere (in Pini, “Univosity”) that these passages from question IV.1 on the
Metaphysics pertain to a first draft of the question.
21
Quaest. super Metaph., VII, 3, Utrum substantia sit primum omnium entium
cognitione (OPh IV, 115–9).
22
Quaest. super Metaph., VII, 3, n. 1: “Quod non: ‘Omnis cognitio oritur a sensu’; sed
sensibilia sunt accidentia; ergo accidentia prius cognoscuntur quam substantia.”
23
Quaest. super Metaph., VII, 3, nn. 6-15 (OPh IV, 115-118). Scotus reports three
opinions. The first opinion is at nn. 6-7 (with some objections and answers at nn. 8-12); the
second opinion is at n. 13 (with an objection at n. 14); the third opinion is at n. 15 (it basically
admits that substance is cognized discursively and therefore is not prior in knowledge).
12
translated as ‘more perfectly cognizable’. So for example the sun is a more
perfect object of cognition than an instance of white, in the first sense of the
expression ‘more perfect object of cognition’, because the sun is a better thing
than an instance of white. But an instance of white is a more perfect object of
cognition in the second sense of the expression ‘more perfect object of
cognition’, because our vision of that instance of white tells us more about
that instance of white than our vision of the sun tells us about the sun. It is a
matter of how our likeness or species of the object of cognition is
proportioned to that object.24
Scotus concludes that substance is an object of cognition more
perfect than accidents in the first sense, because substance is more perfect
than accidents, both in itself and to us. By contrast, accidents are a more
perfect object of cognition in the second sense, because we understand more
perfectly the things of which we have a species in the intellect—and we have
species of accidents but not of substances. All the same, any time we cognize
a substance, our cognition of that substance is more perfect than our cognition
of accidents (in the first sense), because substance is more perfect than
accidents. Here is how the passage reads in what I take to be its original draft,
i.e. if we take away a few sentences that (as we shall see) appear to have been
added by Scotus at a later date:
Then it must be said that substance is a perfect object of cognition,
and nevertheless something can be cognized more perfectly by us in
proportion to its cognizability, because the species of some dim whiteness
represents more perfectly that white of which it is the species, than does the
species of the sun in proportion to the visibility which an eagle has of the sun
in proportion to its visibility. And I see white better in proportion to its
visibility, than an eagle sees the sun in proportion to its visibility. Then I say
that “the more perfect object of cognition is that whose species is in the
intellect” is true in proportion to its cognizability, and nevertheless subtances
as such is cognizable more perfectly in itself and by us [here there is the
added passage]. And because the cognition that we have of substance is the
more perfect cognition, once it is attained, than is the cognition of an
accident, then even though the accident is the first cognizable thing with
respect to our intellect in the order of generation, nevertheless substance is
the first in perfection and therefore is simply the first. (Trans. Etzkorn and
Wolter’s with some modifications)
25
24
Quaest. super Metaph., VII, 3, n. 16 (OPh IV, 118).
Quaest. super Metaph., VII, 3, n. 16 (OPh IV, 118-119): “Tunc dicendum quod
substantia est perfectum cognoscibile, et tamen aliquid aliud potest perfectius cognosci a nobis
secundum proportionem cognoscibilitatis suae. Quia perfectius repraesentat species alicuius albi
remissi ipsum album cuius est species quam species solis repraesentet oculo aquilae ipsum solem.
25
13
Scotus’s solution is ingenious. He successfully manages to reconcile
two claims that seemed to contradict each other. The first claim is that there
are intelligible species only of accidents and not of substances. The second
claim is that substances are nevertheless cognizable—actually, they are prior
in knowledge, even though Scotus does not provide here the details of how
they are cognized. So, when indeed substance is cognized (cum [scil., cognitio
de substantia] attingitur), substance is a more perfect object of cognition than
accidents. Everything seems to be fine with Scotus’s solution if it were not for
a passage added to the text. There are two reasons to maintain that this
passage was added to the first draft by Scotus and that originally it was not
part of Scotus’s solution. First, there is the testimony of the first editor of
Scotus’s Questions on the Metaphysics, according to which the passage in
question is marked as an addition in some manuscripts. Second and more
importantly, two manuscripts (i.e. Oxford, merton College 292 and Paris, Bibl.
Nat. lat. 16110) report this passage not in the place where we find it in the
critical edition, but at the very end of the question26. This was the usual way in
which Scotus’s additions as found in his cedulae were copied when the scribe
could not locate the place in the original question where Scotus intended to
insert his addition.27 According to the critical apparatus, the Merton
manuscript also marks this passage with two reference letters (a-b). In the light
of this evidence, both historical and philological, I think that we can safely
conclude that the passage is indeed an addition, which was originally written
by Scotus on one of the scraps (cedulae) attached to his original manuscript
and then copied down in the original question or at the end of the original
question by the scribes who copied Scotus’s original manuscript together with
his cedulae.
At a first look, these few added words seem to be relatively innocent.
But at a second look, we realize that they actually change the original
meaning of Scotus’s solution. After the passage where he had stated that
Et magis video album secundum proportionem visibilitatis suae quam aquila solem secundum
proportionem visibilitatis suae. Tunc dico quod ‘perfectius cognoscitur cuius species est in
intellectu’ verum est secundum proportionem cognoscibilitatis suae, et tamen substantia de se est
perfectius cognoscibile, et in se et a nobis [here comes the added passage]. Quia illa cognitio
quam habemus de substantia est perfectior cognitio cum attingitur quam cognitio accidentis; tunc,
licet accidens sit primum cognoscibile respectu intellectus nostri generatione, tamen substantia
est primum perfectione, et ideo simpliciter primum.” I have modified the translation given by
Etzkorn and Wolter in Questions on the Metaphysics of Aristotle by John Duns Scotus. Trans. G.J.
Etzknor and A.B. Wolter, vol. 2 (St. Bonaventure, NY: Franciscan Institute Publications, 1998),
109–10.
26
See the critical apparatus ad lin. 20 (OPh IV, 118).
27
On this practice, see Pini, “Univocity.”
14
substance is a more perfect object of cognition both in itself and for us, Scotus
now adds: “If we could ever attain it [namely, substance].” And he proceeds
to explain that we cannot cognize substances in this life. This claim is quite
unexpected. It certainly marks a change from what Scotus had said in the first
version of his solution. In the first version, nothing of what Scotus had said
implied that we do not cognize substances in this life. Quite the contrary,
Scotus seemed to be assuming that we do cognize substances in this life, even
though our cognition of substances is temporally posterior to the cognition of
accidents. He had also said that, even though what we know of substances is
less than what we know of accidents, substance is a more perfect object of
cognition than accidents, so that all in all our cognition of substances is more
perfect than our cognition of accidents. But by adding just a few words, Scotus
now modifies the entire sense of his solution. He now denies that we cognize
substances in this life. He also adds that Aristotle’s saying that substance is
prior in knowledge must be understood as said metaphysically, not
psychologically (metaphysice dictum, non animastice). To explain what these
two expressions mean, Scotus refers to what he says elsewhere in the
Quaestiones about the cognition of the individual difference.28 In that
question, Scotus claims that the individual difference is cognizable in itself,
but we cannot cognize it in this life, because of the current condition of our
cognitive powers. Only in the next life, when our cognitive powers are
restored to a better condition, will we be able to cognize what in you makes
you different from me and any other human being. Right now, we can only
know that you differ from me and any other human being. So we can just
know what the individual difference does, not what it is.29 Surprisingly, Scotus
now makes the same claim concerning substantial natures or essences. Not
just the individuating principle, but substantial essences such as humanity and
horseness are in themselves unknown to us in this life.
How should we interpret this surprising claim? And what happened
between the time Scotus first drafted the solution to this question, in which he
assumed that we cognize substances in this life, and the time he added the
few words that changed the sense of his original answer?
28
Quaest. super Metaph., VII, 3, n. 16 (OPh IV, 118.19–119.3): “… si possemus ad
illam pertingere; sed non in vita ista, ut habetur in II, 3 quaestione. Et tunc ad dictum Philosophi
intelligitur metaphysice dictum, non animastice. Distinctio haec habetur in quaestione ‘De
singulari’.”
29
Quaest. super Metaph., VII, 15, nn. 12-32 (OPh IV, 297-306). It should be noticed
that this question is almost certainly a late question, since it is supposed to replace the previous
question, VII, 14, as the editors remark (see OPh IV, 281, note 1).
15
I think that we can find the answer to these problems if we turn to
another place of Scotus’s Questions on the Metaphysics. In Bk. II, q. 2-3
Scotus considers this interesting problem: when we say that something is
difficult or even impossible to know, are we talking about an intrinsic
complexity of the world or about our cognitive limitations? Is there an
objective difficulty due to the very nature of the thing to know or just a
weakness of our senses and intellect? In other words, is it the world that is
difficult to know or are our cognitive powers that are weak?30 Scotus’s
question consists largely of successive layers added one after the other. This
makes Scotus’s question very difficult to interpret but also extremely
interesting. For in these additions we can witness Scotus’s struggle to find a
solution to this question, which is indeed central to his metaphysics. It is in
these additions that I identify the second stage of Scotus’s position on the
cognition of substantial essences.
Scotus starts with an attentive analysis of the ways in which our
cognitive powers, i.e. the senses and the intellect, cognize things. He lists four
kinds of cognition31. The first kind of cognition is intuitive cognition. Here
Scotus does not give a definition of intuitive cognition, but it seems that we
may safely refer to his usual definition of intuitive cognition as the cognition of
something present and existent as present and existent.32 Scotus here affirms
that, in this life, intuitive cognition is only sensitive cognition. For example,
this is the way sight cognizes color when it sees color. The second kind of
cognition is cognition by way of species. Scotus says that this is the kind of
cognition by which, at the sensory level, phantasia imagines color. In the
absence of a certain color, say red, we may evoke a sensible image of that
color. But this kind of cognition can also be found in the intellect. As we may
expect, however, Scotus says that we have intellective cognition by way of
30
Quaest. super Metaph. II, 2, Utrum difficultas cognoscendi sit ex parte intellectus vel
ex parte rerum cognoscibilium. Scotus answers this question together with the following question,
i.e. Utrum substantiae immateriales possint intelligi a nobis pro statu isto.
31
Six in Quaest. super Metaph., II, 2-3, n. 80 (OPh III, 224), four at n. 114 (OPh IV,
232). The second passage, which leaves out the privative modes, is an addition, so it probably
pertains to what should have been the second draft of the question. I follow the second passage.
32
See example Quodl., q. 13, n. 8 (Vivès XXV, p. 321): “Aliqua ergo cognitio est per se
existentis, sicut quae attingit obiectum in sua propria existentia actuali. Exemplum de visione
coloris, et communiter in sensatione sensus exterioris. Aliqua etiam est cognitio obiecti, non ut
existentis in se, sed vel obiectum non existit, vel saltem illa cognitio non est eius, ut actualiter
existentis. Exemplum, imaginatio coloris, quia contingit imaginari rem, quando non existit, sicut
quando existit.” On intuitive cognition, see S. Day, Intuitive Cognition: A Key to the Later
Scholastics (St. Bonaventure, NY: Franciscan Institute Publications, 1947), 114–23; C. Bérubé, La
connaissance de l’individuel au Moyen Age (Montréal-Paris: Presses universitaires de France,
1964), 134–224; S.D. Dumont, “The Scientific Character of Theology and the Origin of Duns
Scotus’ Distinction between Intuitive and Abstractive Cognition,” Speculum 64 (1989), 579–99.
16
species only of sensible accidents, not of substances, because substances do
not have intelligible species, as demonstrated by the argument about
Transubstantiation. The third kind of cognition is the cognition that results
from the composition of different species. In this way, our phantasia can
compound different sensible species, say of gold and of a mountain, and form
an image of something that it has not sensed, say a golden mountain.
Similarly, our intellect can compound different intelligible species to produce
the intellectual likeness of something that has not been grasped by any simple
act of apprehension. Scotus hypothesizes that the cognition we have of
separate substances is of this kind. Finally, the fourth kind of cognition is
cognition per accidens. It is in this way that our senses cognize a human being
when we sense his or her shape and color. A human being is not sensible per
se, only shape and colors are. As far as human beings are concerned, being
human applies accidentally to shape and color. Consequently, our senses
cognize substances accidentally. Similarly, Scotus remarks that this is the way
our intellect cognizes substantial essences, for our intellect cognizes by
abstraction only accidents, and it is therefore not directly acquainted with
substantial essences, which it reaches in a different way.33
So far, Scotus is just rehearsing in some detail what we already know.
But then, he focuses on the way our intellect cognizes substance both in
general and specifically. By working out in some detail the way our intellect
cognizes substances through accidents, he now sees clearly that any time we
cognize a specific substance or even the concept of substance in general, a
key role is played by our concept of being. So for example—Scotus says—let
us posit that we sense a quality and that we also sense a quantity, say an
instance of the color black and a certain weight. From these two sensed
accidents, we get by abstraction the concepts of quality and quantity. Then,
we notice that these two accidents occur together, let us say in a cat, which
both is black and weights 10 pounds. So we ask: what accounts for their
occurring together? We reason that neither the quality nor the quantity can
account for their occurring together, for other times we notice that the color
black occurs without that particular weight and the other way around.
Consequently, their union cannot depend on either of them, i.e. what the
color black is or what weighing ten pounds is. Nothing in the nature or
essence of a quality and a quantity accounts for their regularly occurring
together. We conclude that there must be a third thing or being, which
accounts for their union. That particular instance of black and that particular
weight occur together because they both inhere in a third thing or being. This
33
Quaest. super Metaph., II, 2-3, n. 115 (OPh III, 233).
17
third thing cannot in turn be an accident, otherwise we would never stop in
our search for something that justifies the common occurrence of that quality
and that quantity. Also, we reason that this third thing or being must be
different from other things that function as subjects of other accidents
precisely because different kinds of accidents inhere in those other subjects.
This third thing is what we call a substance, say a cat. In this way we arrive
both at the cognition of substance in general and at the cognition of a specific
kind of substance. We start from accidents and then we infer that there must
be a underlying thing to account for their conjoined occurrence.34
Scotus’s detailed consideration of how we arrive at the cognition of
substances allows him to notice that this account can works only if we have
the concept of something or, as he says, of some being (ens). Otherwise, we
could not take the crucial step in our inference, i.e. we could not posit that
there must be some underlying thing or being apart from the accidents that we
are acquainted with. To draw this conclusion, we must have the concept of a
thing or being. But at this point, Scotus faces a formidable problem. We have
seen that all concepts are just re-elaborations of information coming from the
senses via sensible and then intelligible species. So how do we acquire this
concept of thing or being? Perhaps there is an intelligible species of being?
This possibility would clearly be problematic for Scotus. So far, all his
reasoning has been based on the premise that there are intelligible species
only of sensible qualities. Now being is not a sensible quality. But then, how
is the concept of being acquired? At his stage, Scotus does not have a solution.
So he concludes this long addition to his question with the word with which
he usually concludes his most tormented passages: Stude, “Study this”. No
definitive conclusion is reached and the problem is still open.35
Scotus’s problem may be summarized as follows. All the information
we have comes from the senses. But only sensible accidents can be sensed. So
34
Scotus describes this inferential process from accidents to substances in a different
way in Quaest. super Metaph. II, 2-3, n. 115 (OPh III, 233). The description I have given is
actually based on Ord. I, d. 22, q. unica, n. 7 (Vat. V, 344–5): “Nam concipiuntur ab aliquo multa
accidentia, concurrentia in eodem, puta talis quantitas et talis qualitas,—et probatur neutrum
illorum esse alterum, quia utrumque illorum manet sine altero; probatur etiam utrique illorum
aliquid aliud esse subiectum commune, quia utrumque illorum potest destrui altero non destructo:
ergo aliquid concluditur esse subiectum utrique, ut qualitati et quantitati,—illud autem quod
subest, non concipitur in conceptu quiditativo nisi entis, vel ‘huius entis’. Et cum frequenter
contingat quod talis quantitas et talis qualitas coniunguntur in aliquo et alibi non coniunguntur, et
hoc non est ex natura qualitatis et quantitatis, ut praeostensum est,—concluditur quod hoc est ex
natura illius tertii, in quo fundantur ambo ista; non autem coniunguntur talia in isto toto, qualia in
illo: ex quo enim diversimode coniunguntur in diversis, concluditur substratum istis esse diversum
a substrato illis, et ex hoc concluditur hoc esse aliud ab alio tertio.”
35
Quaest. super Metaph., II, 2-3, n. 122 (OPh III, 235): “Item, quomodo intelligitur ens?
Numquid per speciem propriam? Stude.”
18
how do we get cognitive access to substances? Not by abstraction from what
we get through the senses, but by inference, as we have seem. We form a
complex concept or description of a substance in general—as well as of
specific substances—as something or a being that underlies accidents. So it is
clear that the concept of being plays a key role in our cognition of substances
through accidents, for we cognize a substance as “something” or “a being”
that accounts for the common occurrence of accidents. But if ‘being’ is an
ambiguous term and the concept of being we use when we cognize
substances as beings is different from the concept of being that we use when
we cognize accidents as beings, we are still facing the same problem in
different terms. Granted, we cognize substances through accidents. And to
pass from accidents to substances, we need the concept of being. But this
concept is different when said of substance and when said of accident. So we
are back to our problem: how do we know the concept of being that is proper
to substance?
Scotus can answer this question when he finally comes to think that
being is a univocal concept. This is the third stage that I would like to single
out in Scotus’s development on the issue of our cognition of substances.
When we say that a substance is a being and an accident is a being, we use
the word ‘being’ in one and the same meaning. Scotus reaches this
conclusion probably for theological reasons in his Lectura. We find it first
expressed in his discussion of the concept we have of God. But in the Lectura
itself Scotus applies this conclusion to the issue of the cognition we have of
substances. He finally realizes that, if we did not have a univocal concept of
being, we could not have any cognition of substances. It is the univocal
concept of being that functions as a sort of bridge between the accidents,
which we cognize directly, and substances, which we reach by inference.
Here is Scotus’s argument as he gives it in his Ordinatio:
What was assumed about substance, namely that it does not
produce an immediate change on the intellect so that the intellect may have
an act about it, is demonstrated in this way. The intellect can naturally know
that a certain thing is absent any time it is not changed by it if that thing
produces a change on the intellect when it is present. This is clear from what
Aristotle says in the second chapter of the De anima, namely that sight
perceives darkness when light is not present, and therefore sight is not
changed then [i.e., when light is not present]. Therefore, if the intellect were
naturally changed by a substance in an immediate way, so that the intellect
would have an act about that substance, it would follow that, any time that
substance is not present, the intellect would be able to know naturally that
that substance is not present. And thus, the intellect would be able to know
19
naturally that there is no substance of bread in the consecrated Host. But this
is manifestly false. (Translation mine)
36
Scotus’s mature account of how we cognize substances can be
reconstructed as follows. The only things we are directly acquainted with are
accidents—more specifically, the infimae species of sensible accidents, such
as a certain shade of white or a certain specific sound. These are the objects of
which we have likenesses in our senses (i.e. sensible species) and in our
intellect (i.e. intelligible species). These are the things that we cognize noninferentially. These accidents, however, can be cognized in two ways. First,
they are cognized in what Scotus calls ‘a confused way’ (confuse). Scotus
defines the cognition of something in a confused way as the sort of cognition
that we have when we can name something, i.e. when we can pick it out
among other things and refer to it but we are not able to provide any
description of it. In this order of cognition, the infimae species of sensible
qualities are the first things cognized. Second, these same sensible qualities
are cognized in what Scotus calls ‘a distinct way’ (distincte). Something is
cognized in a distinct way when it is possible to provide a definition of it, i.e.
a description that captures its essential features. When we cognize something
in a distinct way we analyze the information that we have received from the
senses and that we store in the intellect into its simple components. So for
example we analyze a certain shade of white into the concept of color able to
refract the light in a certain way. We can go on in our conceptual analysis
until we reach a definition of the kind “something or a being such that etc.”
Accordingly, the most fundamental concept in this order of cognition is the
concept of being or something. This concept cannot be further analyzed into
anything simpler. It is presupposed by any other, more complex concept, but
in turn it does not presuppose any other concept. It is for this reason that
Scotus claims that the concept of being is the first or most fundamental
concept in the order of knowing something distinctly, i.e. when we carry out
the task of defining things. Now, the concept of being or something that we
have arrived at in our analysis of the accidents we are directly acquainted
with is so simple and fundamental that it can form the core of a description of
36
Ord. I, d. 3, p. 1, q. 1, n. 140 (Vat. III, 87-88): “Quod autem est suppositum de
substantia, quod non immutat intellectum nostrum immediate ad actum circa se, hoc probatur,
quia quidquid praesens immutat intellectum, illius absentia potest naturaliter cognosci ab
intellectu quando non immutatur, sicut apparet II De anima, quod visus est tenebrae perceptivus,
quando scilicet lux non est praesens, et ideo tunc visus non immutatur. Igitur si intellectus
naturaliter immutatur a substantia immediate ad actum circa ipsam, sequeretur quod quando
substantia non esset prasens, posset naturaliter cognosci, non esse praesens—et ita naturaliter
posset cognosci in hostia altaris non esse substantiam panis, quod est manifeste falsum.”
20
something different, i.e. what underlies those accidents. So we can use the
same concept of being or something that we have acquired by analyzing the
accidents we are acquainted with, in order to form a description of what we
infer must be underlying those accidents. In this way, we come to cognize
substances—as what lies behind the accidents we are acquainted with. But we
could not have formed our description either of substance in general or of
specific substances if the concept of being that we have acquired by analyzing
the accidents we are acquainted with had not been univocal, i.e. so simple
that it could be used to describe also things belonging to a different
ontological kind such as substances.37
So we can finally answer the third question that I posed above, i.e.
why Scotus thinks that we can account for the way we cognize substances
only if we posit that being is univocal. Only if the concept of being is the
same when used to describe accidents and substances can we use it to infer
that there must be something underlying the accidents we are acquainted
with, in the same sense of ‘something’ in which we say that accidents are
something. Since this concept is obtained through analysis of the cognitive
content conveyed by the intelligible species abstracted from accidents, we can
say that the concept of something or being is obtained by abstraction from
accidents, as Scotus says.38 This does not mean that there is an intelligible
species of being impressed in our intellect. No new species is required in
addition to the species of sensible qualities. All further concepts are obtained
by a process of analysis that our intellect carries out over the content
conveyed by the senses. So from one and the same intelligible species of, say,
a certain shade of white, we obtain several further concepts, such as the
concepts of color and, more fundamentally, of being. Scotus clearly considers
this process of analysis as part of the act of abstraction, even thought it should
be separated from the reception of an intelligible species.39 The reception of
an intelligible species is only the first stage in a complex process. Scotus
regards cognition as such a complex process from knowing something in a
confused way to knowing something in a distinct way. In this respect, Scotus
37
Ord. I, d. 3, p. 1, q. 1-2, nn. 71–3, 80 (Vat. III, 50–5). See also Lect. I, d. 3, p. 1, q. 12, nn. 70 and 75 (Vat. XVI, 252–3).
38
Lect. I, d. 3, p. 1, q. 1, n. 112 (Vat. XVI, 266).
39
Scotus distinguished between the abstraction of an intelligible species from a sensible
quality, which results in a specific concept, and the abstraction of a generic concept, which he
calls “a bigger and more difficult abstraction” in Ord. I, d. 3, p. 1, q. 1-2, n. 86 (Vat. III, 58). The
latter abstraction seems to be identical with what Scotus usually calls ‘resolutio’, i.e. a process of
conceptual analysis.
21
is just like Aquinas.40 But Scotus differs from Aquinas in thinking that only a
univocal concept of being can allow us to pass from the cognition of
accidents to the cognition of substances.
IV
At this point, Scotus’s mature position should be clear. He is aware of
a problem lying at the heart of late medieval Aristotelianism. On the one
hand, we obviously have cognitive access to substantial essences, if
philosophy is to be taken as mapping out the real structure of the world. On
the other hand, it seems that no coherent account of how we may get this
access can be given within the Aristotelian framework if we do not have a
univocal concept of being, i.e. if we are not able to describe both substances
and accidents as beings in the very same sense of the word ‘being’. Scotus’s
claim that those who deny that being is univocal are the true destroyers of
philosophy seems to be, after all, justified. In other words, Scotus now realizes
that the doctrine of the univocity of the concept of being—which he had
probably first entertained as a theological position necessary to account for
our cognition of God—is also necessary in philosophy if the Aristotelian
account of how we get to know substances must be vindicated and freed from
the inconsistencies that had plagued it in the hands of Scotus’s predecessors.
Also, Scotus’s evolution should be clear. First he just assumed that we
do have cognitive access to substances, presumably by an act of simple
apprehension. Afterwards, he came to realize that this is not compatible with
the Aristotelian claim that all our cognition comes from the senses—for only
sensible accidents can be sensed. Accordingly, he struggles to reconcile the
Aristotelian position concerning sensible qualities with Aristotle’s own claim
that substances are prior in knowledge. He finally arrives at a satisfactory
solution thanks to his doctrine of the univocity of the concept of being. Our
inferences from sensible qualities to substantial natures are warranted because
we have a univocal concept of being, which we obtain by analyzing the
cognitive content conveyed by the intelligible species abstracted from sensible
qualities.
This is indeed Scotus’s mature and reasoned view. Still, we should
notice that Scotus formulates it in a remarkably cautious way. Even in his
40
See N. Kretzmann, “Infallibility, Error, and Ignorance,” in Aristotle and His Medieval
Interpreters (Canadian Journal of Philosophy, suppl. vol. 17), eds. R. Bosley and M. Tweedale
(Calgary: University of Calgary Press, 1991), 159–94; Pasnau, Thomas Aquinas on Human Nature,
324–9.
22
Ordinatio and in the Collationes parisienses, he puts forward this position as a
sort of afterthought—something that is necessarily implied by his assumptions
but that he still embraces with some hesitation.41 He usually refers to our
cognition of quiddities and to the presence of species in the intellect in a
generic way, without specifying whether the quiddities and the intelligible
species are of substances or of accidents42. Only when considering the issue
more in detail does Scotus specify that the quiddities we cognize by direct
acquaintance, i.e. non-inferentially, are actually the quiddities of accidents.
Similarly, the species present in our intellect are actually the species of
accidents. Even these clarifications, however, are always made with a note of
cautiousness.
I think that there are two possible explanations for Scotus’s caution.
The first explanation is that Scotus never had the chance to give a
definitive formulation of his philosophical and theological views. Even his last
writings give us quite often merely a glimpse of what his considered opinion
might have been. Accordingly, Scotus may have gradually come to the
conclusion that a coherent defense of his opinion on how we cognize
substances would have required the reformulation and re-writing of vast parts
of his work in order to harmonize them with his insights concerning
substances. For example, his defense of the role of the intelligible species in
cognition can be retained but should be reformulated as concerning
exclusively the only intelligible species we have, i.e. the intelligible species of
sensible accidents. Unfortunately, Scotus died just when he realized that such
an adjustment was necessary and he never managed to carry it out. So what
we have right now in Scotus’s writings is a mixture of his old and new views,
where Scotus has not yet drawn all the consequences deriving from his
opinion on our cognition of substances. Accordingly, the details of this
opinion are still formulated with some hesitation.
The second explanation for Scotus’s caution in putting forward his
views concerning our cognition of substances touches on an important aspect
41
See in particular Collationes parisienses, collat. 4 (cod. Oxon. Coll. Merton. 194, f.
66va), as reported in Vat. III, 225–6: “Quiditas accidentis verius ‘esse’ habet in intellectu per
speciem quam quiditas substantiae, quia forte substantia non intelligitur per speciem propriam, eo
quod per speciem propriam non facit in intellectu, sicut nec in sensu; similiter, quiditas
substantiae materialis verius ‘esse’ habet in specie intelligibili quam quiditas substantiae
immaterialis, quae non habet proprium phantasma, nec etiam phantasma accidentis.” Scotus
refers to this passage in an addition to the Ordinatio where he discusses the presence of species in
the intellect. See Ord. I, d. 3, p. 3, q. 1, n. 370 (Vat. III, 225): “Cuiuslibet quod per se et primo
intelligitur, est propria species – Collationes.”
42
On the role of essences in metaphysics, see Ord.. II, d. 3, p. 1, q. 1, nn. 29–34 (Vat.
VII, 402–5). On the role of the intelligible species in cognition, see Lect. I, d. 3, p. 3, q. 1 (Vat.
XVI, 325–48); Ord. I, d. 3, p. 3, q. 1 (Vat. III, 200–44).
23
of Scotus’s thought. Scotus’s cautious tone may be due to the fact that he
thinks that our incapacity to cognize substances directly, i.e. in a noninferential way, is a contingent fact. At the heart of Scotus’s metaphysics there
lies a contrast between, on the one hand, the way we currently are and know
things (in statu isto) and, on the other hand, the way we were meant to be and
to know things—which is also the way we will be and will know things in the
next life, when the original order between soul and body is going to be
restored. According to Scotus, the current dependence of our souls on our
bodies is a contingent fact. Specifically concerning cognition, it is not part of
our nature of human beings that we acquire all cognitive information from the
senses. What for Aquinas was an essential feature of human nature, for Scotus
is a contingent situation.43 Scotus lists two possible causes of this contingent
situation. The first possible cause of the current situation is the Fall. That we
now acquire all information through the senses may be a punishment for the
Original Sin. After our catastrophic fall, the soul was made subject to the
body, and a consequence of this new subjection may have been our intellect’s
current cognitive dependence on the senses. The second possible cause of the
current situation is God’s willingness to harmonize our body with our soul in
the present state. This harmonization, however, is contingent. As a matter of
fact, it seems to be nothing else than the intellect’s subjection to the body. The
intellect is currently dependent on the body for all the information it has about
the extramental world. This dependence will disappear in the next life, when
the original and genuine harmony between soul and body is going to be
restored and consequently the intellect is not going to have access to the
world exclusively through the senses.44
43
Ord. I, d. 3, p. 1, q. 3, n. 186 (Vat. III, 112–3): “Obiectum primum potentiae
assignatur illud quod adaequatum est potentiae ex ratione potentiae, non autem quod adaequatur
potentiae in aliquo statu: quemdamodum primum obiectum visus non ponitur illud quod
adaequatur visui exisistenti in medio illuminato a candela, praecise, sed quod natum est
adaequari visui ex se, quantum est ex natura visus. Nunc autem ut probatum est prius – contra
primam opinionem de primo obiecto intellectus, hoc est adaequato, quae ponit quiditatem rei
materialis primum obiectum – nihil potest adaequari intellectui nostro ex natura potentiae in
ratione primi obiecti nisi communissimum; tamen ei pro statu isto adaequatur in ratione motiva
quiditas rei sensibilis, et ideo pro isto statu non naturaliter intelliget alia quae non continentur sub
isto primo motivo.”
44
Ord. I, d. 3, p. 1, q. 3, n. 187 (Vat. III, 113–4): “Sed quae est ratio huius status? –
Respondeo. ‘Status’ non videtur esse nisi ‘stabilis permanentia’, fermata legibus sapientiae.
Firmatum est autem illis legibus, quod intellectus noster non intelligat pro statu isto nisi illa
quorum species relucent in phantasmate, et hoc sive propter poenam peccati originalis, sive
propter naturalem concordantiam potentiarum animae in operando, secundum quod videmus
quod potentia superior operatur circa idem circa quod inferior, si utraque habebit operationem
perfectam. Et de facto ita est in nobis, quod quodcumque universale intelligimus, eius singulare
actu phanasiamur. Ista tamen concordantia, quae est de facto pro statu isto, non est de natura
intellectus unde intellectus est, - nec etiam unde in corpore, quia tunc in corpore glorioso
24
So both explanations of the current situation stress that the intellect’s
current dependence on the senses is contingent. We were originally meant to
cognize substantial essences, and we will again be able to do so when our
nature is restored to its original condition in the next life. Now, however, we
are unable to grasp substantial essences by a non-inferential act. The
intellect’s current dependence on the senses implies that the only cognition
we can now have of substances is inferential and by way of description.
Fortunately, we do have a univocal concept of being that, even in the current
situation, enables us to reconstruct the structure of reality notwithstanding our
cognitive limitations. Accordingly, Scotus’s hesitations in his formulating his
views on the cognition of substances may depend on the point of view he
chooses. If he speaks about what we are meant to cognize, he emphasizes the
role of substances as first things known and our capacity to grasp them by a
direct, i.e. non-inferential act of cognition. By contrast, any time he focuses on
our current, contingent condition, Scotus stresses our incapacity to grasp
substances by a non-inferential act, for we currently have intelligible species
only of sensible accidents. It is also interesting to notice that Scotus’s mention
of the role of the Fall to explain our current cognitive limitations is present in
the Ordinatio but is absent in his previous Lectura. So it seems that only at a
relatively late date did Scotus come to interpret the contrast between the
perfect cognizability of substances and our incapacity to grasp them in a noninferential way as a contrast between what we were meant to be (and
hopefully will be again in the next life) and what we currently are after the
Fall.
It is interesting to notice that Scotus explains what was originally a
tension internal to the Aristotelian framework as a distinction between the preand post-lapsarian state. In the pre-lapsarian state, substances can be grasped
by way of a direct act of cognition. In the post-lasparian state, this is not
possible anymore and our intellect is made dependent on the species reiceved
from the senses. Scotus’s explanation of this tension is ultimately dependent
on his theological anthropology.
necessario haberet similem concordantiam, quod falsum est. Undecumque ergo sit iste status, sive
ex mera voluntate Dei, sive ex iustitia puniente (quam causam innuit Augustinus XV De Trinitate
cap. Ultimo: “Quae causa” – inquit – cur ipsam lucem acie fixa videre non possis, nisi utique
infirmitas? et quis eam tibi fecit, nisi utique iniquitas?”) sive—inquam— haec sit tota causa, sive
aliqua alia, saltem non est primum obiectum intellectus unde potentia est et natura, nisi aliquid
commune ad omnia intelligibilis, licet primum obiectum, adaequatum sibi in movendo, pro statu
isto ist quiditas rei sensibilis.”
25
V
There is a last difficulty. Let us consider again our cognition of
substantial natures. What do we know about them? We know that they are
beings, in the same sense in which accidents are beings. Apart from that, all
our knowledge concerning substances is actually knowledge of the accidents
that inhere in substances. So, if we distinguish the cognition of what
something is from the cognition of how it is qualified, we must conclude that
the only cognition that we have of what substances are is that they are beings.
Scotus draws explicitly this consequence in one of the passages from the
Questions on the Metaphysics I have already mentioned above:
But beyond the concept of being nothing more special is known of
the quiddity of any substance, not only of a separate substance but of a
material substance [as well]… to being itself we conjoin positive or privative
accidents that we know from the sense, and we make from being and many
such accidents a single description, the whole of which is never found except
in such a species. And the concept of such a description is the more perfect
concept which we have of such a species. (Trans. Etzkorn and Wolter)
45
We have complex concepts of substances, or, as Scotus says,
descriptions. These complex concepts or descriptions are made up of the
concept of being and of several concepts of accidents. So what we know of
substantial natures, apart from the accidents that inhere in them and that are
different from them, is only that those natures are beings. Far from being the
first object of our cognition, substantial natures are in themselves unknown to
us. Scotus himself notices that this has a curious consequence. Suppose that
God shows to your intellect a substantial essence, say the essence of a human
being. Suppose also that in doing this, God does not provide you with any
extra illumination; He only makes that essence present to your intellect.
Suppose, finally, that this is all that God shows you. Specifically, He does not
show you all the accidental features that are usually associated with that
substantial essence and from which you usually get your inferential cognition
of that essence. In that case, Scotus claims, you cannot know whether what
God shows you is that essence, say the essence a human being, or not. For the
only way you can identify a certain substantial essence is by way of the
45
Quaest. super Metaph., II, 2-3, n. 115 (OPh III, 233): “Sed ultra conceptum entis nihil
specialius intelligitur de quiditate alicuius substantiae; nec separatae, nec materialis… Sed ipsi
enti coniungimus accidentia positiva vel privativa, quae cognoscimus ex sensu, et facimus ex ente
et multis talibus unam descriptionem, quae tota numquam invenitur nisi in tali specie. Et
conceptus talis descriptionis est perfectior conceptus quem habemus de tali species.” The English
translation is by Etzkorn and Wolter, vol. 1, 199.
26
accidental features that are usually associated with it and which are the sole
object of your direct cognition in this life. Once these accidental features are
removed, you are not in a position to identify that substantial essence
anymore, because the only thing that you know about that essence, apart from
those accidental features, is that it is something or a being. But this is not
enough to identify anything, since everything is something or a being. 46 Our
cognition of substances amounts to ignorance:
Therefore I say that our intellect first cognizes the accidents, from
which it abstracts the concept of being—which is in turn predicated
essentially of substance just as it is predicated of an accident. And this is all
we know of substance in a nondiscursive way, and nothing more. As I said,
this is what everybody can verify in their own experience, namely they only
know about any substance that it is a being. All the rest that we know about
any substance are properties and accidents pertaining to that substance. It is
by way of those properties that we to know the features that are essential to a
substance. (Trans. mine)
47
But if this is the case, Scotus faces a formidable objection.
Quite simply, the claim that we are basically ignorant of what
substantial essences are in themselves is incompatible with metaphysics as it
is commonly practiced by Aristotelians. Scotus himself maintains that
metaphysicians make use of real definitions. Also, metaphysicians make use of
propositions per se primo modo, such as “Human beings are rational
animals”.48 But these propositions, as Scotus concedes, are true in virtue of
substantial essences. So, contrary to what Scotus claims, it seems that we do
know a lot about substantial essences. We can define them and we can make
true statements about them.49 How can Scotus account for this?
46
Lect. I, d. 3, p. 1, q. 1-2, n. 266 (Vat. XVI, 266): “Unde si per impossibile Deus
ostenderet intellectui tuo essentiam hominis—vel per possible—et non daret tibi aliud lumen, sed
tantum eam faceret intellectui tuo esse praesentem, et non ostenderet tibi descriptiones
accidentium quae intelligis vel quibus cognoscis substantiam, nescires utrum esse essentia
hominis vel non.”
47
Lect. I, d. 3, p. 1, q. 1-2, n. 112 (Vat. XVI, 266): “Unde dico quod intellectus noster
primo cognoscit accidentia, a quibus abstrahit intentionem entis, quod praedicat essentiam
substantiae sicut accidentis; et tantum intuitive cognoscimus de substantia, et non plus. Hoc, sicut
dixi, experitur quilibet in se, quod non cognoscit plus de natura substantiae nisi quod sit ens.
Totum autem aliud quod conoscimus de substantia, sunt proprietates et accidentia propria tali
substantiae, per quas proprietates intuemur ea quae sunt essentialia substantiae.” In this passage,
the term ‘intuitive’ is clearly not to be contrasted with ‘abstractive’; rather, it should be taken as
meaning ‘nondiscursive’, as is commonly the case in Henry of Ghent.
48
Ord. II, d. 3, p. 1, q. 1, n. 32 (Vat. VII, 403), quoted above, note 9.
49
Quaest. super Metaph., II, 2-3, n. 117 (OPh III, 234): “Sed contra praedicta arguitur
quod substantia a nobis per se intelligitur, quia ita definimus substantias sicut accidentia, per
genera et differentias proprias.”
27
Interestingly, Scotus does not recoil from this objection. In order to
answer it, he advances one of his most enigmatic views. He says that we have
what he calls a habitus vocalis of substances—which we may translate as a
‘naming disposition’.50 It is not immediately clear what this means. But some
light may be gained from what Scotus says in the Ordinatio about our capacity
to name substantial essences. The only concept that we have of a substantial
essence is that of being. All the same, Scotus holds that our names of
substances refer to those substances with precision, because we can name
things more distinctly than we can conceive them.51 So we lack a distinct
concept of what a substance is (both in general and in particular), but we have
a distinct name for it. For example, we lack a distinct concept of what a horse
is and of what a human being is. All the same, we can name those essences
with sufficient precision. So our names latch onto substantial essences
whereas our concepts do not. The same should be said of our definitions. The
parts of a definition latch onto real entities of which we do not have any
proper concept. For example ‘rationality’ and ‘animality’ refer to some
features in reality (the so-called ‘formalities’) even though we do not have any
distinct concept of what these features are. We can only name them; we
cannot think of them in any distinct way. Scotus makes a striking comparison
to capture the way we do metaphysics in statu isto. We only have names and
no proper concept of substances, just as a born blind person has only names
of colors but no corresponding sensation.
What is the cause of this curious mismatch between our cognitive
capacity and the way things are? As we have seen, Scotus does not give a
definite answer. Our cognitive limitation may be due to God’s decision to
harmonize in some way soul and body in the current life. Or it may just be a
consequence of the Fall. In any case, this is a contingent situation, something
that is not part of our very nature of human beings but that characterizes the
current situation in this world.
What is perhaps most remarkable about Scotus’s position is that, all
things considered, our cognitive limitations do not prevent us from doing good
metaphysics and from speaking with surprising precision about the world. For
Scotus’s comparison between what we know (or better, what we do not know)
50
Quaest. super Metaph., II, 2-3, n. 119 (OPh III, 234): “Ad primum horum dicendum
quod de substantiis habemus habitum vocalem, sicut caecus natus syllogizat de coloribus, quia
nec ipsa genera intellligimus, nisi ens.” The example of the born blind person who makes
deductions about colors comes from Aristotle, Phys. II, 1, 193a6-8. See also Aquinas’s
commentary, In octo libros de Physico audito sive Physicorum Aristotelis Commentaria, eds. F.
Angeli and M. Pirotta (Neaples: M. D’Auria, 1953), II, lect. 1, n. 306.
51
Ord. I, d. 22, q. unica, nn.4–5 (Vat. V, 343–4).
28
about substances and what a born blind person knows about colors has also a
positive side. Admittedly, the blind person knows neither what color in
general is nor what any specific color is, just as we do not know what a
substantial essence is in general or what any specific substantial essences is in
itself. All the same, a born blind person is able to make valid deductions about
the colors (caecus natus syllogizat de coloribus). So just as the blind person’s
deductions can be valid even though she is ignorant of what she is talking
about, in the same way a metaphysician’s arguments can be valid even
though she does not know what the substantial essences she is talking about
are in themselves. As to the vagueness of our concepts, it seems that the
precision with which we can name essences puts some sort of remedy to it.
Our names of substances function as signposts for things that we are ignorant
of, even though we know what they do. Accordingly, we can describe those
essences thanks to their external features and we can place them in a larger
framework thanks to those descriptions.
So it seems that our ignorance of substantial essences is not fatal to
the success of our metaphysical enquiry. Take for example Scotus’s famous
doctrine of the mind-independent, less than numerical unity of essences. It
seems that nothing of what Scotus says about that unity depends on whether
we actually cognize substantial natures in themselves or not. We may
concede that, strictly speaking, we only know that substantial essences are
beings. Nevertheless, we can still argue that those essences have a mindindependent, less than numerical unity. Stressing our ignorance of substantial
essences does not mean that metaphysics may be less daring in its exploration
of the structure of reality.
Admittedly, it may be difficult to hide some uneasiness when
considering this aspect of Scotus’s thought. The picture that emerges is in stark
contrast with the traditional view of Scotus as a metaphysician who detailed
every single aspect of reality down to the last formality. Nevertheless, Scotus’s
remarks about our cognition of substances occupy a central place in his
metaphysics, even though the overall picture may be rather sketchy. (For
example, Scotus’s reference to habitus vocalis is little more than a hint.) But
the philosophical interest of Scotus’s attempt is clear. He boldly tried to give a
solid foundation to metaphysics while taking full account of the current
limitations of our cognitive powers.
29