In Philosophical Investigations, Ludwig Wittgenstein characterized his thought as an attempt “to shew the fly the way out of the fly-bottle.” As one commentator has observed, this claim encapsulates Wittgenstein’s proclivity for dissolving rather than solving philosophical problems. Wittgenstein believed that, the relevant concepts being clarified, many “philosophical problems should completely disappear.”
Yet, one can benefit from what ultimately turns out to be a meaningless or incoherent question. St. Augustine, for example, had to clarify his conceptions of being and goodness in order to dissolve the question “unde malum.” Before he discovered the “libri platonicorum,” Augustine thought of evil as a substance and sought for one principle that might explain its nature. Once he realized that being and goodness were convertible, however, Augustine abandoned his original inquiry. As he put it, “And thus I sought for the origin of evil, although I sought it evilly. For I did not see the evil in the very inquiry itself.” Like Wittgenstein, Augustine made progress by showing that what initially appeared to be an insurmountable philosophical problem is actually incoherent, though, pace Wittgenstein, it is not therefore meaningless.
Following Augustine’s example, this paper will clarify evil’s relation to ἐπιστήμη or scientific knowledge. In particular, I will combine Aristotle’s argument in Metaphysics Eta against a science of accidents with Proclus’s and Pseudo-Dionysius’s insight that evil is an accident. Consequently, I will argue that, given their classification of evil, Proclus and Dionysius are committed to the position that κακολογία, a science of evil, is impossible.
My analysis, then, will proceed in two steps. First, I will examine Aristotle’s argument in Metaphysics Eta against a science of τὸ συμβεβηκός. Given the enormous complexity of the concepts at stake, I will not be able to discuss Aristotle’s concepts of ἐπιστήμη, causality, and matter at sufficient length. Nonetheless, these concepts are crucial for understanding Aristotle’s claim that there cannot be a science of accidents. For accidental being, like the matter responsible for it, is indefinite and inherently unintelligible. Second, I will analyze Proclus’s and Dionysius’s joint disjunctive argument for classifying evil as a quasi-being per accidens. Overall, I will argue that both are committed to the impossibility of κακολογία, given their claim that evil cannot have a per se cause. Lastly, I will conclude with a few thoughts on the significance of this ‘philosophical dissolution.’
I: Against a Science of τὸ συμβεβηκός
In Metaphysics Eta, Aristotle says “that there is no knowledge of the per accidens is manifest”. At first glance, this claim seems paradoxical. What could be more trivial than knowing that someone is tan or in the agora? Both predicates signify accidents insofar as one may or may not have them, since they do not follow necessarily from one’s essence, i.e. from humanity itself. Neither does their acquisition or loss alter one’s humanity. Nonetheless, it seems as though one can give an account of their generation. The tan man was outside in the sun. The man in the agora needed a few things from the market, etc. Given such explanations, it seems as if there is some knowledge of accidents.
Such accounts, though, overlook that scientific knowledge is oriented towards what Aristotle calls τὸ ἀεί or τὸ ὡς ἐπὶ τὸ πολύ. That is, there are several differences between knowing that Socrates is in the agora and knowing that he is a rational animal. First, the latter claim touches upon something universal. In this case, Socrates’s essence is something he shares with every other man insofar as they are all members of one species. His location, on the other hand, is particular. Second, scientific knowledge is of something necessary. While Socrates may or may not be in the agora, he must be a rational animal. Third, ἐπιστήμη must be infallible. For one cannot know what is false or what is liable to falsification. One cannot know, for instance, that Socrates is a fish, nor can one know that he might become one tomorrow.
Given the above distinctions, it is clear that scientific knowledge must be knowledge of causes. In the above case, one has grasped the formal cause of humanity as such and has, consequently, achieved ἐπισήμη. Furthermore, if scientific knowledge is universal, necessary, and infallible, the causes one grasps ought to be stable. For if rational animality were not always the formal cause of man, one would not know that ‘man is a rational animal.’ As Aristotle says, scientific knowledge must be of “eternal causes.”
What, then, is a cause for Aristotle? In the Physics he defines a cause as that which answers the question διὰ τί, “on account of what?.” A cause is first and foremost something that explains one thing’s dependency on another. It explains why some being exists by means of some necessary relation it has to some other being.
For Aristotle there are four types of causes. One knows a thing if and only if one can explain either: (1) the ‘from-which’ something comes to be, such as the bronze of a statue, (2) the intelligible-form or the paradigm, as in the ratio of a harmony, (3) the first principle of change or rest, as when a father begets a child, and (4) the end or the ‘that-for-the-sake-of-which,’, as when an intention to be healthy leads to exercise.
Since formal, efficient, and final causes generate their effects on the whole and for the most part, they seem especially conducive to scientific knowledge. Their regularity can be seen in the following example from the Metaphysics, where Aristotle distinguishes what is properly caused by the housebuilder’s art from what comes about κατὰ συμβεβηκός. He says:
For the one building the house does not make as many things as happen to come about alongside the house coming into being (for these are infinite. For nothing prevents the thing being made from being sweet to some, harmful to others, useful to others, and different, so to speak, from all beings of which the housebuilding art is productive.
Not everything that results from his art, in other words, is truly ‘caused’ by it. The art of housebuilding, properly speaking, determines the man’s intent (final), his movement of the available matter (efficient), as well as the shape of the house itself (formal). One knows that these are effects because they regularly or even necessarily accompany the art of housebuilding.
However, one might think that the material cause is also regular and scientific. Does not the builder need specific materials, like stones, wood, clay, stucco, rather than water or air? While it is true that the builder cannot work with just any material, his art determines his intent, the shape of the product, and the relevant movements more than the matter involved. He may use stones or he may not, but he must intend to build a house, he must build something with a habitable form, whether that be a teepee or a log cabin. Likewise, he must move his material in a particular way determined by intent and form. However, nature provides an enormous variety of possible materials. Furthermore, in extreme cases, one can build a home out of almost anything: bottles, tires, or even dung. Such variation is not found in his intent, movement, or the habitable form.
The prioritization of causes is supported by a passage in De Partibus Animalium. Here, Aristotle seems to rank causes based on their intelligibility or determinacy. He writes:
For it appears as though that cuase is first (πρώτη), which we call the ‘that-for-the-sake-of-which.’ For the end is the ratio, which, in turn, is the principle of motion and rest equally in the things that are according to art as in things constituted by nature. For, indeed, the form having been made definite to the mind or the senses, the doctor with respect to health or the builder with respect to the house give reasons and causes of each thing they do, and why it is necessary to do thus.
The use of the word πρώτη suggests that Aristotle is giving a hierarchy of causes. Indeed, just before this passage, he says “it is necessary to define such things, which cause is first and which second by nature.” That is, Aristotle wants to know which cause is first (πρώτη) and which is second (δευτέρα) by nature.
If this is indeed a hierarchical list, it is based on each cause’s relative intelligibility. In other words, the intelligibility of one cause entails the intelligibility of another. If this interpretation is correct, then ἡ τελική or final cause is indeed πρώτη. For knowledge of the final cause implies knowing a thing’s λόγος or formal definition. If one understands what it means to be a perfect or virtuous man, one must know the form or the ‘what-it-is-to-be’ (τὸ τί ἦν εἶναι) of man. The converse is not true. Thus, the λόγος is ἠ δευτέρα αἰτία. Moreover, synchronic knowledge of the τέλος and the λόγος implies a certain range of motion from potency to actuality. If I know man’s essence as well as his end, I have some idea of how he ought to come into being and move towards perfection. Again, the converse does not seem true, though this inversion is less clear than the previous one. Nonetheless, the above list does show that final and formal causes are first because they are bounded or determinate, ὁρισάμενος, and consequently intelligible to the mind.
Significantly, the material cause is not even mentioned in this list, suggesting that it is the least intelligible of all. Lest one should doubt this, Aristotle joins the three former causes in the Physics. He says “for the three frequently come together into one. For, on the one hand, the what-it-is and the that-for-the-sake-of-which are one. On the other, the first whence-the-motion is the same as the intelligible-shape for these things. For man begets man.” Although he presents a four-fold list of causes, the final, formal, and efficient take precedence to the material.
If causes are ranked by intelligibility and if science is of determinate causes, then both the ἄπειρα and the ἑτέρα are necessarily imcompatible with ἐπιστήμη. “The limitless qua limitless is unknowable.” In the housebuilding example, Aristotle uses these terms to exclude everything that merely “happens to come about” (συμβαίνει) from the builder’s art. He says that, even if the housebuilder’s art or his product result in pleasure, pain, or even profit, such things are not, strictly speaking, caused by his art. The list of such ‘effects’ would truly be limitless (ἄπειρα), other (ἑτέρα), and thus outside the scope of his art.
This indeterminacy is due to matter, the least intelligible cause of all, as Aristotle argues in Metaphysics Eta. One could interpret his argument as being ‘transcendental’ in the Kantian sense. For Aristotle argues from an obvious experience, that of causal irregularity, then asks what principles one must presuppose in order to account for it. The argument begins by pointing out that not everything follows necessarily from causes. For the builder’s art sometimes brings about pleasure, othertimes pain. Thus, causes do not always have the same outcomes. Yet, causes, at least formal, final, and efficient causes, make things like themselves. Such causes necessarily make their effects “sibi simile,” as the builder’s art showed. How, then, is divergence from causal regularity possible? Aristotle argues that we must posit a principle of variation and alterity. He writes:
Since not everything comes to be by necessity and exists always, whether they be beings or things that have come to be, but rather for the most part, it is necessary that there exist that which is per accidens, such as when the white man is musical, which is neither always nor for the most part. Since it comes to be sometimes it will be per accidens. If not, then everything will be by necessity, such that matter will be the cause admitting things outside of or different from that which is for the most part.
Therefore, unless one holds that effects follow uniformly from causes—a claim contradicted by experience—one must assume the existence of some principle of τὀ ἕτερον or τὸ ἄλλως.
For Aristotle, one should posit ὕλη as this principle. Matter receives many forms into one substance, thereby joining εἴδη that would not ordinarily exist together. By being devoid of all form, it is unbounded and ‘other’ with respect to every form. As the conclusion of a transcendental argument, one could say that matter’s absolute indeterminacy is the necessary precondition for the possibility of combinations of unrelated εἴδη within a primary substance. This nonuniform bundling of forms is, in turn, a necessary precondition for accidental being.
From this argument, it follows that there cannot be a principle of the per accidens. If matter is the principle of the variation and alterity necessary for per accidens being, and if what follows from a principle is only as intelligible as its principle, then accidents will be unintelligible in themselves. That which owes its being to the ἕτερον or the ἄπειρον will be opposed to the determinate intelligibility that characterizes per se causality and thus scientific knowledge. Consequently, there cannot be a science of accidents given that they follow from indeterminate matter. Indeed, As Aristotle says, the per accidens, due to its indeterminacy is not really a being at all. It is ἐγγύς τι τοῦ μὴ ὄντος. The term συμβεβηκός does not even signify something, but exists ὥσπερ…ὄνομά τι μόνον (as if it were a name only)
II: Against κακολογία
In order to rule out a science of evil, we must turn to Proclus and Dionysius. In particular, we must examine their claim that evil is a quasi per accidens being. In De Malorum Subsistentia and De Divinis Nominibus, Proclus and Dionysius make a similar disjunctive argument justifying their classification of evil.
Some reconstruction, though, is necessary to formalize their reasoning. First, both assume that evil must either be per se or per accidens, since even quasi or analogous being must fall into this universal division of being-in-general. Second, they assume that per se being follows from per se causes, while per accidens being follows from per accidens causes. Third, both assume the standard Aristotelian list of per se causes—though they add the paradigmatic cause. Fourth, they reject every possible per se cause for evil, substituting an ersatz cause in its place. This last premise is the hinge of the entire argument, since it shows that evil does not have a per se cause, entailing that it is not a per se being but rather per accidens.
How, then, do Proclus and Dionysius rule out the possibility of a per se cause of evil? Their methodology is peculiar, since they do not simply rule out the list of per se causes. Rather, they posit what can be called a counter-list of ψευδοαιτίαι or, as one commentator put it, “negative causes,” capturing the idea that such ‘causes’ work through dissonance and absence of form rather than by means of presence and order.
Both begin their accounts by discussing paradigmatic causality. Normally paradigms are similar to or even identical with divine ideas, whether in νοῦς for Proclus or in God’s mind for Dionysius. Large parts of De Malorum Subsistentia and De Divinis Nominibus are devoted to rejecting the possibility that there is a paradigm of evil in God or in immaterial being generally. They argue that evil is not found in the gods/angels, the demons (at least by nature), nor in immaterial souls.
These claims follow from their shared assumption that the First Principle, the One or God, is the infinite source of goodness in beings. Given that, ex hypothesi, the maxim omne agens agit sibi simile applies even to God, it follows that nothing emanating from God can be evil. For it does not belong to the Good to create evil, as heat does not refrigerate. As Proclus puts it, even if there is limping in material things, there is no “form of lameness” in νοῦς.
In one section of De Malorum Subsistentia, Proclus imagines a so-called paradigmatic cause, though it is more of a parody of such causality. He writes:
If, though, you wish to have an exemplar cause of evils… is that which Socrates shows in the Theaetetus, a form of evil that roams around this mortal nature from necessity. Indeed, souls assimilate to evil beings, and they transmute that assimilation which is better for the life of such things. For the soul looks at the paradigms of the good things when it turns towards itself, to beings better than itself, and to the summits of beings located on their holy seat exist. But when it looks back towards that which is outside of itself, at that which is beneath itself, at those things that are singular and external to themselves, disordered and indeterminate and fluctuos by their own nature.
The contrast of “turning” “ad se” or “extra se” reveals the counterfeit nature of this paradigmatic cause. Rather than turning to the fullness of one’s form, one imitates what is “fluctuosa sui ipsorum natura.” This ‘cause,’ then, relies on absence of and opposition to form.
Yet one might think that evil results from final, formal, or efficient causes. Beginning with the last, Proclus and Dionysius show that the “factivae causae malōrum” or τὰ ποιητικά τῶν κακῶν are actually caricatures of efficient causality. Since alterity exists among material substances, somethings are inevitably “adverse to one another.” This adversion, in turn, creates “space for the coming to be of that which is contrary to nature.” For example, a dog and a fire differ according to λόγοι. Their natures, moreover, do not always interact well. A dog might warm itself by the fire, but it can also burn itself. This harm is not a per se effect of either the dog’s or the fire’s nature. Rather, their alterity makes possible a “space” for “unmeasured mixing and communion” among opposites. Hence, absence of form and disharmony among forms are the ‘efficient causes’ of evil.
Likewise, both argue that evil stems from an ersatz formal cause. True formal causation takes place when a whole substance determines itself and its parts per se, as the ratio causes harmony and rational animality causes humanity. By contrast, evil’s ‘formal cause’ occurs when a substance determines itself through some defect or deficiency. As Proclus puts it, “The form and nature of evils is deficiency, unboundedness, and deprivation.” Rather than having one complete nature that causes its parts, evils are caused by incomplete wholes with numerous privations and deficiencies. For instance, an ignorant man lacking the appropriate virtue of wisdom makes bad choices on account of his deficient humanity, i.e., his incomplete form.
Negative formal causes, moreover, are possible because things are different and unbounded in nature. This makes it possible for some creatures, acting as self-moving or self-determining principles, to choose something different, inferior, and contrary to their form, thereby bringing about corruption and evil.
Perhaps most significant of all, Proclus and Dionysius argue that evil has a faux final cause. Ignorance and desire for what is not, τὰ μἠ ὄντα, are ‘causes’ in this sense. Both thinkers follow Socrates’ claim in the Meno that people only choose evil through ignorance and by mistaking an apparent good for a real one. As both Proclus and Dionysius argue:
The principle and end of all evils will be the good. For all things come to be for the sake of the good. For we do these things desiring the good, and no one does what he does while looking at what is evil.
As with formal causation, the precondition for the possibility of ignorant choice is matter and alterity. One can only choose what is different from and hostile to form if and only if alterity exists in the first place. This diversity in being, though, does not necessarily cause evil. Responsibility for ignorance rests on the one choosing and improperly desiring the apparent good. Ignorance and malformed ἔρως for the apparent good are the ‘final causes’ of evil rather than the differences among goods themselves. For these, taken by themselves, are good.
Having analyzed and rejected the per se causes, it is important to stress three points. First, negative, ersatz causes are made possible by matter. This claim seems implicit in both Proclus’s and Dionysius’s analyses. As we saw, alterity and variation make possible one’s movement away from divine ideas, disharmonious associations, privations of form, and malformed desire. These evils presuppose some randomness within substance, differences among beings, and non-uniformity between causes and effects.
Second, Proclus and Dionysius rejected per se causes of evil by offering an account of ψευδοαιτίαι in their place. Evil’s ‘causes’ are perversions and caricatures of actual causes, for they act through absence and disorder rather than by presence and order.
Third, their arguments imply a response to Plotinus, who thought that matter was itself the principle of evil, τὸ πρῶτον κακόν. For Proclus and Dionysius, though, one must distinguish between necessity and evil. Matter is necessary to explain diversity and variation within the chain of causes and effects. But matter does not make evil necessary, only possible. It is the necessary precondition for the possibility of evil rather than evil itself. Likewise, both argue that matter is necessary insofar as it makes the full range of created being possible. Created being does not end with henads, angels, demons, heroes, or any other immaterial substance, but with things characterized by alterity. This last rank of material creatures, though not evil by nature, have the possibility of becoming evil.
If the above analysis is correct, it seems that the initial disjunctive argument obtains. Evil does not have a per se cause, as shown above. All per se beings have per se causes, as assumed in premise three. Therefore, evil is not a per se being. Moreover, everything that exists uni-vocally or even analogously can be divided by per se and per accidens being, as assumed in premise one. Thus, although evil does not exist in the same way as man or even whiteness, it must fall under the category of the per accidens. This conclusion, when taken together with Aristotle’s argument in Metaphysics Eta, entails that a science of evil is impossible. If a science of accidents is impossible and evil is an accident, then a science of evil must also be impossible.
Hence, if one asks unde malum thereby committing oneself to a search for a per se cause, one’s investigation will never cease. For the question itself is incoherent, like asking what color is the fastest or wisest. One could debate such a question ad infinitum and make absolutely no progress. What one needs to do is clarify the terms of the question itself. Having done this, one will see that evil cannot in principle have a per se cause, forcing one to rephrase the initial question. If this question can be rephrased so as to avoid this confusion, then one will finally be able to show the fly the way out of the fly bottle.
If the analysis of Proclus’s and Dionysius’s argument is correct and their argument itself sound, as I think it is, then one is entitled to dissolve certain questions about evil. This result is significant for ordinary language and life. For people frequently ask questions like, ‘why do bad things happen to good people?,’ ‘why is there so much suffering and evil in the world?,’ and ‘why does evil exist at all?.’ These questions have the appearance of profundity and, to be sure, they are experienced as such by those who suffer. Nonetheless, these questions are also profoundly misleading. They are bewitchments of language, as Wittgenstein says. They tempt us to search for a per se cause of evil rather than a contingent set of causes. Their apparent intelligibility leads us to search for an ultimate answer when none is available and to feel dissatisfied either when we fail to explain evil or arrive at something purely contingent.
Nonetheless, the dissolution of the original question does not imply that talk of evil is simply meaningless or incoherent. It is as intelligible as our talk of accidents. ‘The tsunami struck the town because of an earthquake’ is no less intelligible than ‘I happened upon a friend in the marketplace today.’ The fact that these statements lack some ultimate principle or cause for their being true hardly makes them meaningless. However, it does chasten our language and keep it within the realm of what is intelligible. Indeed, recognizing this fact should exorcize us of our fascination with evil and its works, revealing it for the fraud and parasite that it is. Evil’s radical unintelligibility should turn our minds towards that which is actually intelligible or even super-intelligible—the Good. Then, along with St. Augustine, we will say “every particular thing is good and at the same time the entirety is indeed good, for our God has made everything good indeed” (et singula bona sunt, et simul omnia valde bona, quoniam fecit deus noster omnia bona valde).