1. Act and potency

1.1 The general theory

1.1.1 Origins of the distinction

The first of the famous twenty-four Thomistic theses[1] reads:

Potency and act are a complete division of being. Hence, whatever is must be either pure act or a unit composed of potency and act as its primary and intrinsic principles. (Wuellner 1956, p.120)

Fundamental to Scholastic philosophy in general

Differences between Scotists and Suarezians

Questions about:

· metaphysics of substance

· essence

· causation

Also for that matter to scholastic philosophy of nature, philosophical psychology, natural theology, and even ethics.

Begin an outline of the theory

· origins in Aristotle resolution between the Eleatics and Heraclitus

o (change versus permanence)

o (multiplicity versus unity)

Aristotle Against Parmenides and Zeno: Permanence

Parmenides and Zeno denied the reality of change.

(I) Change would require being to arise out of non-being or nothingness

(II) From non-being or nothingness, nothing can arise

Therefore

(III) change is impossible.

Zeno offered paradoxes to demonstrate the impossibility or motion, some of which presuppose that traversing a finite distance would require traversing an infinite number of shorter distances.

First response: retorsion

It is self-contradictory to argue the impossibility of motion.

To argue practically:

· Requires moving one’s his lips etc.

· Changing the minds of one’s listeners or readers

Nevertheless, the reality of change is not self-evident, insofar as it is not a necessary truth that any change ever actually occurs.

“But it is still evident insofar as we have to acknowledge it in order to argue for anything at all.”

But this only tells us that something has gone wrong, but not what, exactly, has gone wrong.

Aristotle diagnosed the problem.

Not the inference from (1) and (2) to (3)

Not premise (2)

Rather premise (1) is false.

According to Aristotle, it is not the case that being arises from non-being, but rather one kind of being arising from another kind.

Specifically,

1. there is being-In-act --the ways a thing actually is

2. there is being-In-potency --the ways a thing could potentially be.

For instance, a given rubber ball might "in act" be:

· spherical

· solid

· red

· sitting on the desk, etc.

But "in potency" it is (be):

· flat

· squishy (if melted)

· light pink (if left out in the sun too long)

· rolling across the ground (if dropped).

Potentialities or potencies are real features of the ball itself even if they are not actualities.

They are not nothing, even if they do not have the kind of being that the ball's roundness, solidity, smoothness, etc. currently have.

The rubber ball can become flat, squishy, and rough in a way it cannot become sentient, or eloquent, or capable of doing arithmetic.

So “Being-in-potency” is a middle ground between “being-in-act” on the one hand, and sheer “nothingness or non-being” on the other.

Change is NOT a matter of being arising from nonbeing.

Change is being-in-act arising from being-in-potency. (the actualization of a potential)

Zeno too …

Regarding Zeno: The infinite number of smaller distances in the interval between two points A and B are indeed there in potentially, rather than actually.

“Hence there is no actually infinitely large number of distances the runner must traverse, and Zeno's purported reductio fails.”

Axristotle Against Parmenides and Zeno: Unity

Parmenides denies that there can possibly be more than one being. Imagine A and B are both exist. This means they are both “being.” But if a being A and B really were distinct, there would have to be differentiated by something. But since A and B both are, by hypothesis, being, the only thing that could differentiate them would be non-being and non-being, since it is just nothingness, does not exist and thus cannot differentiate them. Therefore there can be no multiplicity and all reality is one undifferentiated being

Thus, Parmenides denies that there can possibly be more than one being. But of course, being in act can be distinguished from other being in act in terms of its potentialities. While not act, potential is not non-being either

Zeno argues along the same lines via his paradox of parts.

1. If there is more than one being, then either these multiple beings have size or they do not.

a. If they do not have any size, then, even when combined, they can never yield anything with size.

i. Therefore if they have no size, it would follow that there is nothing of any size at all.

ii. But if all that is has no size then, one thing cannot be distinguished from another (no boarders o differentiate)

iii. Therefore if being has no size, then there cannot be multiple beings of no size.

iv. Therefore if being has no size, then there cannot be more than one being.

b. If these multiple beings do have size, then they are infinitely divisible.

i. If they are infinitely divisible then they have an infinite number of parts.

ii. If they have an infinite number of parts (regardless of how small), then they must all be of infinite size.

iii. So there are multiple beings of infinite size.

iv. But there cannot be more one being of infinite size

v. Therefore is being has size there cannot be more than one being.

But Aristotle would counter that if multiple beings do have size, then they are divisible ad infinitum in potentiality. This is not mean they constitute a actual infinity of parts. Thus they need not be infinitely large and there is nothing paradoxical about a plurality of finitely large things.

Aristotle against Heraclitus: Change

Heraclitus argued that there is no being but only endless becoming. (Change and change alone is real.)

· no stability or persistence of even a temporary sort

· nothing that corresponds to Aristotle's notion of being-in-act.

Retorsion:

Reasoning itself requires fixed premises, etc.

Likewise requires enduring interlocutors (persisting subjects)

Requires that there be such thing as "the" argument for one’s conclusion --some single, stable pattern of reasoning

But perhaps even more inherent, any change is change toward some outcome, the acquisitions of new actualities.

And the changes themselves manifest repeatable patterns, thus a degree of stability that belies Heraclitus' doctrine of flux.

Aristotle against Heraclitus: Multiplicity

The Heraclitean position entails that there is only multiplicity and no unity

There is this particular thing we call "round," that one, and a third one, but no one thing, roundness, that they all instantiate; there is this perceptual experience of what we call a "ball," that one, and a third one, but no one thing, that ball itself, that these experiences are all experiences of, and no one subject, the perceiving self, which has the various perceptual experiences.

“(To be sure, Heraclitus himself adopted a kind of monism on which there is one thing, the world itself, which is the subject of endless change --a dynamic monism rather than the static monism of the Eleatics. Still, none of what J. L. Austin called the "middle-sized dry goods"* of everyday experience could count as unified subjects on this view.)

"But then we have to ask, of course, what this class comprises. We are given, as examples, 'familiar objects'--chairs, tables, pictures, books, flowers, pens, cigarettes; the expression 'material thing' is not here (or anywhere else in Ayer's text) further defined. But does the ordinary man believe that what he perceives is (always) something like furniture, or like these other 'familiar objects'—moderate-sized specimens of dry goods?"

Sense and Sensibilia page 8.

Retorsion: (a pox on both your houses)

Eleatics: How can the Eleatic so much as distinguish between:

himself and his interlocutor?

his premises and his conclusion?

reality and appearance?

Heraclitean: How can a Heraclitean ever

• validly draw a (fixed) conclusion from his (fixed) premises?

• state his (a) thesis unless there were stable, recurring patterns in terms of which to characterize change or becoming?

Application of act and potency:

non-being or nothingness is not the only candidate for a principle by which two beings A and B could be differentiated. (differentiated by reference to their potencies.)

Zeno:

the infinite number of parts are in the thing only in potentially.

Science would be impossible if either the Eleatic position or its Heraclitean opposite were true.

On the former, there would be no world of distinct, changing things and events and perceptual experiences as distinct and changing perceptual episodes, would be entirely illusory.

On the latter there could be no stable, repeatable patterns for the scientist to uncover (no laws of physics, etc.) Each position undermines the metaphysical presumptions of science.

“Yet there is no way to avoid the Eleatic and Herclitean extremes without affirming the distinction between act and potency.”

1.1.2 The relationship between act and potency

Distinction between

1. real distinctions (differences in extra-mental reality itself)

2. logical distinctions (differences in our ways of thinking about extra-mental reality)

Scotists add to this classification the notion of a

3. formal distinction (something intermediate between a real and a logical distinction)

Thomists regard the distinction between act and potency as a real distinction

Scotists and Suarezians regard it as a formal distinction.

Thomists differ with Scotists and Suarezians about whether anything other than potency limits act.

The roundness of a certain rubber ball, which is actual, but in a limited way (i.e. roundness is perfect but the ball's roundness is not perfect)

The Thomist position is that potency accounts for these limitations on a thing's actuality. (This is the second of the twenty-four Thomistic theses)

Because act is perfection, it is limited only by potency which is a capacity for perfection. Hence, a pure act in any order of being exists only as unlimited and unique; but wherever it (act) is finite and multiplied, there it unites in true composition with potency. (Wuellner 1956, p. 120)

It is matter which limits the roundness to this rather than that degree as well as this or that particular time and place.

(matter is always potentially at some other point in time and space)

Scotists and Suarezians, hold that the limitations of a thing's actuality can be accounted for by reference to the thing's cause.

The ball's roundness is imperfect because the ball's cause put, as it were, only so much roundness into it.

Thomist, however, such an extrinsic principle of limitation (the limit of the cause) is possible only if there is an intrinsic principle --something in the limited thing itself by virtue of which its cause is able to limit its actuality --and this can only be potency.

“Hence the cause of the ball can put a limited degree of roundness into it precisely because the ball has the potency to be something other than perfectly round; and it can cause the roundness to be instantiated here and now rather than some other time and place precisely because the rubber which takes on that form has the potency to be at various times and places.

Related:

· whether the distinction between act and potency is a real distinction

· whether the distinction between a thing's essence and its existence is a real distinction

For those who regard the distinction between act and potency as real, act is prior to or more fundamental than potency.

1. Any potency is always defined in relation to act.

2. A thing's potencies are grounded in its actualities. (because the ball is actually made of rubber rather)

3. A potency can be actualized only by what is already actual.

4. Act is prior to potency insofar as while there can be nothing that is pure potency

That which is absolutely pure actuality or actus purus is the core of Scholastic philosophy's conception of God.

1.1.3 Divisions of act and potency

Subdistinctions:

Subdivisions of potency:

1. pure or logical possibility on the one hand and real potency on the other.

Unicorns would be examples of the logically possible (objective potencies- qua objects of thought)

2. Real potencies are grounded in the natures of real things.

Regarded as potencies in the proper sense (subjective potencies- insofar as they are grounded in a real subject)

Further distinction of real potencies:

1. active potency: the capacity to bring about an effect

2. passive potency: capacity to be affected.

EX: Fire's capacity to melt rubber (active potency)

Rubber's capacity to be melted ( passive potency)

An active potency is a power.

A passive potency is a potentiality in the strict sense.

Further distinction WRT Active Potency: uncreated active potency and created active potency

“…active potency is, strictly speaking, a kind of act or actuality (in particular, what is called a "first actuality"); more precisely, it is a kind of act relative to the substance possessing it, though a kind of potency relative to the action it grounds (Koren 1955, p. 59). By "potency" what is usually meant is passive potency. (cf. Koren 1960, p. 122; Renard 1946, p. 29) Pure active potency or power unmixed with any passive potency or potentiality is just pure actuality, and identified by the Scholastics with God; in everything other than God active potency is mixed with passive potency. This difference is marked by the Scholastic distinction between uncreated active potency and created active potency.

Distinctions WRT passive potency.

1. Passive potency considered in relation to the thing that has it

a. passive potency considered in relation to the essence of a thing

b. passive potency considered in relation to the existence of a thing.

2. Passive potency considered in relation to the agent which brings about an effect in the thing that has it.

a. natural passive potency

i. points to an outcome that can be realized given only a thing's natural capacities

ii. can be actualized by some agent that is itself a mixture of active and passive potency

b. a supernatural or obediential passive potency

i. potency points to an outcome that cannot be realized given only a thing's natural capacities (i.e. for humans, eating, sleeping, walking, etc.)

ii. requires as an agent a purely actual divine cause. (i.e. for humans Beatific Vision)

Prime matter: pure potentiality for the reception of form

Second matter: matter which has taken on some substantial form, but is in potency relative to the reception of accidental forms.

Divisions of act or actuality:

1. pure act: (Actus puras) utterly unmixed with any potentiality: God

a. absolutely pure act (God who is in in act relative His existence.)

b. relatively pure act. (e.g. an angel - form without matter with its essence is still in potency relative to its existence)

2. mixed act: Everything else, act in some way mixed with potency.

Second distinction:

3. operative act (concerns a thing's operations or activities)

4. entitative act. (Considered as pure entity; abstracted from all circumstances- concerns what it is statically speaking)

i. a thing's essential act --its essence or nature, what it is

1. A thing's substantial form, that which makes it the kind of substance it is (sometimes called its "first act")

2. A thing’s accidental form, which modifies an already existing substance. (sometimes called its "second act")

ii. a thing’s existential act -its existence, or that it is.

However, the expressions "first act" and "second act" are also often used in a different way, to distinguish a power from the operation of a power. For example, the power of speech is a first act or actuality, and using this power or speaking on a particular occasion would be a second act or actuality.

So…

· A man's having the substantial form of a rational animal is a first actuality

· Having the power of speech is a second actuality relative to this first actuality.

o Having the power of speech is however itself a first actuality relative to the actual exercise of that power, which is a second actuality relative to the mere having of the power.

Similar distinctions can be made vis-a-vis potentiality.

· Someone who knows no English has the potential to speak it insofar as he might learn English.

· That is a "first potentiality" for speaking English.

· Once he does learn the language he has a kind of standing ability to speak it on particular occasions if he wishes to.

· That is a "second potentiality" for speaking English.

o This second potentiality is in turn a first actuality insofar as it is a power that can be distinguished from the exercise of the power. The actual exercise of the power to speak English would, accordingly, be in turn a second actuality.

1.2 Causal powers

1.2.1 Powers in Scholastic philosophy

Analysis of Causation:

Aristotelians distinguish four causes:

1. Formal

2. Material

3. Efficient

4. final.

“In contemporary analytic philosophy, however, terms like "cause," "causality," and "causation" are generally used to refer to efficient causality almost exclusively. Occasionally final causality is discussed (even if, usually, only to reject the notion). Formal and material causes are not treated as causes at all.

At this point treats of efficient and final causality alone.

Discusses formal and material causes in chapter 3.

Efficient Cause (also called an agent in Scholastic philosophy): is that which brings something into being or changes it in some way (actualizes a potency). It does so by exercising its own active potencies or powers.

Note a cause is not always bringing about its characteristic effect. (author is the efficient cause of the book, but not continuously) .

· Thus the power (active potency) to write must therefore be distinguished from any particular act of writing,

· The power is a precondition of the latter (and thus distinct from it).

On the side of the patient (thing being affected e.g. my computer screen)

· Its passive potency or potential for being changed is the correlate of the active potency or power of the efficient cause or agent, and stands as a precondition for the this or that actual, particular instance of change (and thus distinct from it).

Must maintain the distinctions between:

· a causal power or active potency and its actual exercise on any particular occasion

· a passive potency or potentiality and its actualization on any particular occasion

Follows from the act and potency theory of change.

If only actual particular instances of causation and actual particular effects are real (i.e. that there are no such things as powers and potentialities, active and passive potencies) then we would in effect be saying that act alone is real and potency unreal. But in this only resurrects the Parmenidian problem of change.

Act and Potency as an explanations of Limited

Other than what is pure actuality, any cause only ever has a certain specific range of effects. No active power is unlimited in what it may produce. Thus the nature power itself can explain what an agent can do and what it cannot do. (A magnet (agent) can attract metal in a way a piece of wood cannot)

A causal power qua potency just is that which limits efficacy, of itself actual and unlimited, to a specific range of outcomes. Wood has the power to generate noise under certain circumstances, but not the power to attract metal.

Addresses the common charge that the scholastic notion of powers is vacuous and explanatorily useless.

The charge: Summed up in Moliere's joke[2] about the doctor who explained why opium causes sleep by attributing to it a "dormitive power."

(Good discussion) http://ftp.colloquium.co.uk/viae5.htm

Since "dormitive power" means "a power to cause sleep," the doctor's explanation amounts to saying "Opium causes sleep because it has a power to cause sleep."

Critics suggest both that this is

1. a mere tautology

2. false

But of course is can’t be both.

If it is a tautology, is explains nothing, but it is not false.

If, one the other hand, it is false (or potentially false), then it is not a tautology.

To say "Opium causes sleep because it causes sleep" would be a tautology," but the statement in question says more than that. In attributing a sleep-inducing power to opium, it tells us that the fact that sleep tends to follow the taking of opium is not merely an accidental feature of this or that sample of opium, but belongs to the nature of opium as such. That this is not a tautology is evidenced by the fact that critics of the Scholastic notion of powers regard the attribution of a dormitive power to opium as false rather than (as they should regard it if it were a tautology) trivially true. The critics do not say: "Yes, opium has the power to cause sleep, but that is too obvious to be worth mentioning." Rather, they say: "No, opium has no such power, because there are no such things as powers in the Scholastic sense."

Stephen Mumford notes:

One might still claims that appeal to a “dormitive power” is uninformative

That only tells us that there is something about the opium that causes sleep

It does not tell us exactly what that is. [3]

Nevertheless, even if true, in and of itself this gives us no reason to doubt or deny the existence of powers.

Mumford distinguishes

1. causal relations

2. causal explanations.

Whether A provides an informative explanation for B is a separate issue from whether A is in fact causally related to B.

The latter issue is more fundamental insofar as A cannot enter into a true causal explanation of B in the first place unless A really is causally related to B.

The poverty an explanation does not by itself show that the explanans is false.

Further, that powers do play an important explanatory role is precisely why Scholastic writers affirm their existence.

The Scholastic affirmation of causal powers is not in competition with the sorts of explanations put forward in empirical science.

“To say that opium has a dormitive power, for example, is not to make an assertion that conflicts with anything we know about opium from modern chemistry, because the Scholastic metaphysician is simply not addressing the same question the chemist is. The Scholastic philosopher is addressing a deeper question than the chemist is, a question, not about opium per se, but about the necessary preconditions of there being any causality at all, whether in the case of opium and sleep or in any other case. He is claiming that in order to make sense of the facts that a cause is not constantly bringing about its characteristic effects, that its efficacy involves real change, and that its efficacy is limited in just the ways it is, we have to affirm the existence of active potencies or causal powers. How precisely does this or that particular cause --opium, say --bring about its characteristic effects? That is a question for the chemist, and the Scholastic metaphysician qua metaphysician does not claim to have an answer to it. His claim is merely that, whatever the details turn out to be, they will involve the operation of real powers.

Feser notes that Locke and Boyle do actually retain something on the notion of natures possessing inherent causal powers (but it did not much catch on with later moderns.

· "Opium causes sleep because the corpuscular constitution of opium is such that, when ingested, sleep results" (Boyle or Locke)

Some modern chemists my inadvertently do the same.

· "Opium causes sleep because the chemical structure of opium is such that, when ingested, sleep results" (a more modern rendering)

But these amount to about the same thing as saying:

· "Opium causes sleep because it has a dormitive power."

If the former statements are neither tautologies nor completely uninformative --and they are not --then neither is the latter.

They are all claiming it is such as to cause sleep. But this is an appeal to the nature of opium and its fixed “powers.”

None of these statements are intended by themselves to provide a complete explanations, but simply to make a general point about what a correct explanation will have to involve, whatever the empirical details turn out to be.

Anthony Kenny[4] notes differences between the

1. possessor of a power

2. the power itself

3. the vehicle of the power

4. the actual exercise of the power

In the case of opium:

1. Opium (substance)

2. Sleep inducing (active potency possessed by the substance)

3. chemical properties

4. power exercised (actual sleep)

Other substances with the power to cause sleep may do so via other specific chemical properties and thus different vehicles.

“The difference between the metaphysician and the chemist, then, is essentially that the former is concerned with powers and the latter with vehicles.

I would add here that the metaphysician is concerned with “powers” in the broadest sense (that they exist) not in the particular as to what particular something has what particular powers. Rather, she is interested in the general setup.

For instance, the metaphysician is careful to preserve the distinction between “powers” on the one hand, and their actual exercise on the other.

(This is a metaphysics presumption for the whole “powers” set-up to work. As such it is not the stuff of Chemical investigation. One might say it is presumed by chemistry.)

To fail to do so is to reduce potentiality to actuality.

Likewise:

Reducing powers to their vehicles. (Act alone is real.)

“But just as reducing powers to their actual exercise would make act alone real, implicitly deny the reality of potency, and thus entail that change is impossible, so too would reducing powers to their vehicles have the same implication. To affirm that change is real, then, entails affirming the reality of powers as distinct from either their exercise or the vehicles by which they operate.”

Difference among Scholastics:

Some scholastics would effectively reduce powers to their possessors.

1. Thomists take the distinction between a substance and its powers to be a real distinction rather than a merely formal or logical one

2. Some other scholastics do not.

Agreed that the possessors of powers (things) are causes in the strict sense.

Powers are accidents of substances (things), not substances in their own right.

Recall, for Aristotle a substance is a thing that can exist on its own and has an essential nature determining is actual and potential properties. Accidents are features which indeed exist (are real), but cannot exist on their own.

“It is not powers which bring about effects, but rather substances which do so, by way of their powers.”

Thus is it not events which are causes but rather the substances that enter into events that are causes.

From the Scholastic point of view, an event is the actualization of a potency.

Thus events themselves presuppose causality and therefore cannot be the fundamental kind of cause.

Think here of dispositional causality. For some purposes it might be useful to say “the glass broke because it as fragile (the fragility of the glass it was cause it to break), but this is a place holder for a more fundamental notion of cause.

Neither do Scholastics seek to analyze causality in terms of regularities or counterfactual conditions. Regularities and counterfactuals follow from causal relations (powers actualizing potentialities) between substances. They may provide, at most, useful descriptions of what is going on, but not explanations.

Note: Since substances are causes (and not events) a thorough account of the Scholastic views will require an account of their view about substance, which is the subject of chapter 3

“Similarly, it is, primarily, not events which are causes but rather the substances that enter into events that are causes. An event involves the actualization of a potency; hence while there is a sense in which an event might be said to be a cause, since events themselves presuppose causality they cannot be the fundamental kind of cause. Neither, for the Scholastic, is it correct to analyze causality in terms of regularities or counterfactual conditions. These are consequences of causal relations between substances, so that to define causal relations in terms of regularities or counterfactuals is to put the cart before the horse.

Naturally, then, to understand the Scholastic position on causality requires an account of Scholastic views about substance, which is the subject of chapter 3. There is much that can be said short of that, however, and it is best approached through a consideration of recent criticisms of the post-Humean theories of causation that developed in the wake of the early modern philosophers' rejection of Scholasticism.

1.2.2

Feser examines “post-Humean” theories of causation (i.e. ones that attempt to account for causation without recourse to “spooky” powers talk or essential features of substances)

• These developed in the wake of the early modern philosophers' rejection of Scholasticism.

• Offers a consideration of recent criticisms of such attempts

Feser suggests that recent analytic philosophy has seen a revival of interest in powers, dispositions, capacities, and related notions as a way of making sense of causation.

“Powers in recent analytic philosophy Contemporary analytic philosophy has seen a revival of interest in powers, dispositions, capacities, and related notions. That this is essentially a recapitulation of Scholastic themes usually thought passé has not gone unnoticed by commentators

He sees this as a capitulation of Scholastic themes previously regarded as passé.

This is motivated by considerations from general metaphysics, and that which is motivated by considerations from philosophy of science.

1.2.2.1 Historical background

Contemporary analytic “powers theorists” are reacted to views about causation inherited from David Hume.

· These views have dominated modern philosophical thinking about causation.

· They brought to a climax a series of developments whose immediate origins lie in the debate about causation initiated by Descartes et al. (Those who sought to overthrow Scholasticism)

· There were precursors in Scholastic writers like William of Ockham and Nicholas of Autrecourt.

Ockham's theological voluntarism

Ockham maintained, largely for theological reasons, the view that the divine will is prior to the divine intellect (Something like his Existence precedes his Essence). Thus there is nothing in the nature of things that might put limits on what God could command.

· Thus he is an anti-realism about universals - variously interpreted as either nominalist or conceptualist. (Immutable universal essences or nature, would place limits on what God could will/ do. Even God could not create a French-speaking apple tree.)

· Likewise, if there is a universal human nature determining what the natural goods are for us, even God could not will that we violate these (e.g. for us to hate him).

· But for Ockham, God could in principle will/command us to do anything, and if he did so these things really would be good for us. (Divine Command Theory)

More to the present point, for Ockham God could also break the causal connections that ordinarily hold between things:

“Whatever God produces by the mediation of secondary causes, he can immediately produce and conserve in the absence of such causes... Every effect that God is able to produce by the mediation of a secondary cause he is able to produce immediately by himself. (Quodlibet 6, q. 6, in William of Ockham 1991, at p. 506)

It follows from this that it cannot be demonstrated that any effect is produced by a secondary cause. For even though when fire is close to combustible material, combustion always follows, this fact is, nevertheless, consistent with fire's not being the cause of it. For God could have ordained that whenever fire is present to a close by patient, the sun would cause combustion [in the patient] ... Thus, there is no effect through which it can be proved that anyone is a human being -especially through no effect that is clear to us. For an angel can produce in a body everything that we see in a human being -e.g. eating, drinking, and the like ... Therefore, it is not surprising if it is impossible to demonstrate that anything is a cause ... (Opera Theologica V, 72-93, quoted in Adams 1987, p. 750)

This would seem to entail that causes and effects are inherently "loose and separate," as Hume would later put it. So too does this passage:

Between a cause and its effect there is an eminently essential order and dependence, and yet the simple knowledge of one of them does not entail the simple knowledge of the other. And this also is something which everybody experiences within himself: that however perfectly he may know a certain thing, he will never be able to excogitate the simple and proper notion of another thing, which he has never before perceived either by sense or by intellect. (In 1 Sent., q. 3, fol. D2, recto. F, quoted in Gilson 1999, pp. 70-71)

What we witness is only constant conjunction of agent and patient. What we do NOT have is any hint of the agent when conceiving of the patient alone, thus no necessary connection between the two.

Indeed, Ockham also says things that seem to imply a "regularity" theory of causation, as when he writes that:

That is the cause of something which, not being posited, the thing does not exist, and being posited, the thing exists. (Exposino in Libros physicorum, fo1. 123c, 203a, quoted in Weinberg 1964,p.260)

But on this view it would seem that God alone is cause, and there are no (or at least there need not be any as far as we know) and secondary causes.

Accordingly, some have attributed to Ockham a proto-Humean conception of causation.

More explicit in later Ockhamite thinkers like Autrecourt argue

“that no proposition about a causal relation between A and its purported effect B is certain, because there is no logically necessary connection between A and B; and that the reason we regard A and B as causally related is that we have found A to produce B in the past, but cannot be certain that it will do so in the future (Marenbon 2009, pp. 49-51; Weinberg 1964, pp. 272-75).

The notion of a "Secondary Cause"

The notion of a "secondary cause," to which Ockham refers what he and other Scholastics thought of as a things with derivative power/causality. While it might be the proximate cause for the effect in question, it is not nor can it be the ultimate cause since it in turn derives its causal powers from some other thing. But, a system of nothing but secondary causes is an infinite regress. Thus if there are any secondary causes, there must be, at least one, ultimate source of non-derived causation (Uncaused cause.)

“B” is an instrument used by some “C” to actualize “A,” but C is in turn an instrument used by D, and D by E and so on…

“A stock example is a stick which has the power to move a stone only insofar as it is used by someone as an instrument for moving the stone. The standard Scholastic view is that relative to God, who as pure actuality is the source of all causal power, everything else that exists is a secondary cause. But secondary causes nevertheless are true causes (According to the traditional scholastic).

• In contrast to the Scholastics, some medieval Islamic theologians took the view that “secondary causes” they are NOT true causes, and that only God ever causes anything.

• This is the "occasionalist" position.

• It holds that no purported cause A really generates its apparent effect B, but rather that God causes B on the occasion when A is present.

“Autrecourt seems to have been acquainted with the arguments for this view, which relied in part on the claim that there are no necessary connections between purported causes and effects. (It is not clear, though, that Autrecourt himself was an occasionalist.)”

Occasionalism would have an enormous influence on early modern philosophy.

René Descartes (1596 – 1650)

· According to some interpreters he was an occasionalist view vis-a.-vis the apparent causal relations between material objects

o Perhaps a consequence of his conception of matter as pure extension

o Given his view that God is the total efficient cause of motion and that he recreates the material world from moment to moment, it is hard to see what is left for material objects to do.

It worth noting that the much sited weakness in Descartes’ Mind/Body dualism (the inexplicable nature of the causal interaction between mind and body) may not, for Descartes anyway, have really been any more mysterious then causality itself.

Nicolas Malebranche (1638 - 1715)

· sought to synthesize the thought of St. Augustine and Descartes

· Attempts to demonstrate the active role of God in every aspect of the world.

· He and other followers of Descartes explicitly took an occasionalist line.

George Berkeley (1685 – 1753)

· for him, “physical objects” are really just complex ideas (sensations/ perceptions) and are entirely passive

· thus he also adopted an occasionalist position vis-a.-vis the causality of physical objects (Note they no more causally interact that the characters projects on a movie screen.)

Gottfried Wilhelm (von) Leibniz (1646 –1716)

· Not himself an occasionalist, per se, ,

· Explain what he regarded as the false appearance of causality in physical objects by attributing it to a divinely pre-established harmony between them. (Note him too, they no more causally interact that the characters projects on a movie screen.)

Lest we think the attraction of occasionalism is a consequence of dualism, idealism or the monadology of these thinkers, and their rejection of materialism, we see that early modern materialism does not present an immediate alternative to occasionalism either.

Atomistic materialism had difficulties with causation as a consequent of adopting Descartes understanding of matter.

Thomas Hobbes (1588 – 1679 )

· Presupposed efficient (mechanical) causal determinism

· Rejected formal and final causation

· Regarded causation as relevant only to motion.

“Metaphysician of Motion”

Proposed we can explain all phenomena, even psychological and sociological ones, in terms of causal relations between moving bodies.

For Hobbes, it was NOT the bodies, but the accidents of bodies that are causes.

Hobbes defined a cause (efficient causes) as "the aggregate of accidents in the agent or agents, requisite for the production of the effect" (Hobbes [1655] 1839, 9.4).

He defined an effect as "that accident, which is generated in the patient" ([1655] 1839, 9.1).

But these accidents are merely motions of parts of the body. Thus “causation” consists, ultimately, in nothing other than motion. ([1655] 1839, 9.3; 9.9). All “change” is the movement of bodies and the epiphenomena these motions project.

· All causation occurs by contact

· There is no action at a distance.

· Rejected the concept of formal or final causes (merely 'disguised' efficient causes)

· Affirmed a distinction between efficient cause and material cause.

o Material cause is the receptor of the agent's activity: "the aggregate of accidents in the patient,"

o Efficient cause is the aggregate of properties in the agent required for the production of the effect.

o Both material and efficient causes are part of the “entire cause”

“…in Hobbes's universe, everything happens by necessity: "all the effects that have been, or shall be produced, have their necessity in things antecedent" ([1655] 1839, 9.5). Moreover, given the cause, "it cannot be conceived but that the effect will follow" ([1655] 1839, 9.7; italics mine). This description of cause (involving necessity) corresponds with what Taube thought to be the definition of necessity that was current in the seventeenth century, namely: "that the opposite of which is inconceivable" (Taube 1936, 102). A connection is necessary inasmuch as it is inconceivable, or contradictory, that the connection should not obtain. However, this supposed necessity is not based on any matter of fact relation. Hobbes (and most of his seventeenth century colleagues) secured necessary connection by postulating God in the causal relations of finite things. To hold that an entity acts in a manner not determined by God was inconceivable.[5]

Pierre Gassendi[6][7] (1592 - 1655).

o If the only properties of atoms are extension (i.e. size, shape, solidity, and the like), then how can they have any force or power to move other atoms?

o Gassendi took motion to be imparted and preserved by God.

Feser suggests that early modern thought is characterized by:

1. a tendency to take (what Walter Ott has called) a "top-down" approach to understanding the order that exists in the world (Ott 2009, pp. 5-6)

2. a tendency to take (what Brian Ellis has called) a "passivist" view of matter (Ellis 2002, p. 2).

By contrast,

“Aquinas and other mainstream Scholastics attributed active causal powers to material substances, and accordingly regarded the immediate source of the order they exhibit as immanent to them, their orderly behavior arising from the "bottom-up" as it were. Rejecting Scholastic powers, the early moderns came to see matter as instead entirely passive and devoid of any inner principle of change. Laws of nature, conceived of as divine decrees imposed on matter from outside and above, were Descartes' and Malebranche's alternative source of order, and other moderns would adopt a similar approach. (cf. Osler 1996)

Curiously, John Locke (1632 –1704) and Robert Boyle (1627 – 1691)

“… are partial exceptions to this trend, attempting as they do to develop a notion of powers stripped of Aristotelian features like final causality and consistent with their empiricist epistemology (see Part III of Ott 2009).

Nevertheless, Feser maintains their efforts met with only limited success and did not “catch on.”

As Kenneth Clatterbaugh notes:

Ten propositions inherited from the Scholastics were widely accepted:

(1) There are four kinds of causation --material, efficient, formal, and final.

(2) Forms preexist in efficient causes.

(3) Causation requires that something is "communicated" from the cause to the effect.

(4) Proper explanations are deductively inferential.

(5) Cause and effect are necessarily linked.

(6) Causes and effects are substances.

(7) Some substances are active (self-moving causes).

(8) Causation may be instantaneous.

(9) Proper explanations are in terms of the true or proper causes of change.

(10) God is the total efficient cause of everything. (Clatterbaugh 1999, p.15)

(Clatterbaugh's proposition (10) needs qualification. As Clatterbaugh recognizes, while the scholastics regarded God as the ultimate source of all causal power, they did not in general follow occasionalism in denying that secondary causes are true causes.)

In the debate over the nature of causality that takes place between mid-17^th to mid-18^th century, from Rene Descartes (1596 –1650) to David Hume 1711 NS (1711 – 1776), as Clatterbaugh goes on to note:

Each of these key propositions is abandoned in the course of the debate; only proposition (9) survives by the end of the debate, but what counts as true or proper cause is significantly changed by 1739. (Clatterbaugh 1999, p. 15)

By the end of this debate, what were regarded as “true and proper causes” are those identified by empirical science/ physic. But, ironically, the impoverished understanding of causation Hume leaves us with (rejecting as it does both the "bottom-up" (Scholastic) and "top-down" (Occasionalist) approaches of his predecessors, is inadequate to account for what science reveals to us.

Specifically,

1. If matter is inherently passive, then causation seems to disappear altogether as an objective feature of the natural world.

2. If divine decree is rejected as an alternative source of the regularity that exists in nature, that regularity seems to be a brute fact, without any explanation at all.

Hume embraces both of these implications, or at least denies that we can have any real knowledge of causes or of the source of the world's regularity. Echoing Autrecourt, he holds in An Enquiry Concerning Human understanding that the "constant conjunction of two objects" in our experience is what leads us to regard them as causally related, but that objectively "all events seem entirely loose and separate" rather than being necessarily connected.

“All events seem entirely loose and separate. One event follows another; but we never can observe any tie between them. They seem conjoined, but never connected. And as we can have no idea of any thing which never appeared to our outward sense or inward sentiment, the necessary conclusion seems to be that we have no idea of connexion or power at all, and that these words are absolutely, without any meaning, when employed either in philosophical reasonings or common life.[8]

So for Hume and those Moderns and Contemporaries who follow him in this, in principle, any effect/event or none might follow from any cause/event.

Thus I might throw a brick at a plate-glass window and the bring turn into a bouquet of flowers. But to so describe the situation is to talk of events and not causes at all. Indeed it is hardly to talk of “things.” (What is the “brick” as a “thing” after all. Presumably all and only what I am experiencing at the moment without any fixed nature or essence.) Strictly speaking , brick-throwing-experience might be followed by “bouquet-of-flowers” experience, though I fully expect it to be followed by shattered-window-experience. But nowhere does “cause” enter into the experience per se.

“The efficacy we think we perceive in things is really just a projection of our expectations onto the world. (Whether Hume actually denies outright the reality of objective causal relations or is rather a "skeptical realist" about them has, of course, been a matter of debate in recent Hume scholarship, but that is not an issue that needs to be settled for our purposes. cf. Read and Richman 2007.)

Ockham began, Autrecourt furthered, and the early modern occasionalists completed the removal of real causality from the world and its relocation into God. Hume's position is essentially the result of removing God from the picture as well. Nor, where Hume agrees with his predecessors, is the resemblance accidental. (My emphasis)

As Ott writes, "Hume... [was] directly influenced by Malebranche ... to the point of all but plagiarizing from his copy of Thomas Taylor's translation of The Search After Truth" (Ott 2009, p. 3). "The old saw that Hume is occasionalism minus God is," Ott judges, "not too far off the mark" (Ott 2009, p. 195).

Hume's arguments in contrast with Scholastic reasoning:

Note:

Contemporary (non-Scholastic) accounts of causation depend on regularity theories, counterfactual theories. These begin from the Humean starting point. Much of the debate takes place within the boundaries set by Hume.

Ironically, the motivations for the Humean account were theological assumptions that neither the mainstream Scholastics nor secular Humeians would accept. More than just ironic, this undercuts the claim that the burden of proof lies with those reject Hume’s construct of the problem of causation.

According to Feser:

“There is no objective reason to regard Hume's assumptions as the default ones. And as contemporary philosophers with no Scholastic ax to grind have argued, there is good reason to question them.

1.2.2.2 Considerations from metaphysics

Hume offers the following definition:

"We may define a cause to be an object, followed by another, and where all the objects similar to the first, are followed by objects similar to the second" (section VII).

Regularity theories of causation

Regularity theories of causation attempt to captured the causal relation in terms of the regular correlations that exists between events. They resist any reference to a “power” in A by which it generates B. Nor do they seek to establish any “necessary connection” between A and B, presuming as they do that nothing necessary is ever given IN experience. Thus, causation reduces to a Humean, purely contingent "constant conjunction" of inherently "loose and separate" items.

But:

1. We certainly WANT to say more than A is has been (inexplicably) conjoined to B in the past, even if this is all we get via experience.

2. We NEED more than this is science is going to be a source of explanations and predictions.

Further:

1. There is an asymmetry between causes and effects. Correlation supports the claim A causes B just as well as the claim B causes A.

a. We might stipulate A must temporally precede its effect B to be its cause, but some causes and effects are simultaneous.

2. Furthermore, the regularity approach cannot distinguish relevant and irrelevant temporal antecedents in a casual chain. (e.g. The sound of the splashing was regularly precedes but is causally irrelevant to the movement of the leaf on the surface of the pond.)

3. Third, the regularity approach cannot distinguish relevant temporally prior irrelevant “causes” in cases of causal over determination. (e.g. a second stone thrown.) (Eating arsenic and then getting hit by a bus.)

Counterfactual Account of Causation

Another alternative to the regularity account is the “B if and only if A”,

· Still Humean in spirit

o "Or in other words, where, if the first object had not been, the second never had existed."

· Here Hume clearly goes beyond regularity adding a counterfactual condition.

o Prominent defender: David Lewis (1973).

“Such theories hold that causality is essentially a matter of counterfactual dependence. It's not just that whenever A occurs, B also occurs; what makes for a causal connection between A and B is that if A had not occurred, B would not have occurred either. (See Paul 2009 for an overview and Collins, Hall, and Paul 2004a for a collection of key essays.)

Summarizes Lewis's account:

1. If A had not occurred, B would not have occurred.

2. If A had occurred, B would have occurred.

3. A and B both occurred.

(~A -> ~B) & (A -> B)

· Condition (1) states that A is a necessary condition

o at least in the actual circumstances, if not absolutely

· Condition (2) that it is a sufficient condition, for the occurrence of B.

1. Attempt to handle the asymmetry problem

2. Seems to weed out irrelevant temporal predecessors.

Essentially Humean insofar as

"hold[sJ that causal facts are to be explained in terms of--or more ambitiously, shown to reduce to --facts about what happens, together with facts about the fundamental laws that govern what happens" (Collins, Hall, and Paul 2004b, emphasis in original).

Thus the analysis is entirely in terms of actualities rather than potentialities or powers.

To say that the relevant actualities are related by "facts about ... fundamental laws" is to say that they have no intrinsic or necessary connection to one another but are related only extrinsically and contingently. It is to say that causes are (as Ellis would put it) "passive" rather than having any active tendency to bring about their effects, and that (as Ott would put it) the connection between causes and effects is therefore imposed "top-down" via laws that could have been other than they are. Inherently, causes and effects are "loose and separate" and the "constant conjunction" enshrined in the laws that connect them reflects mere nomological[9] necessity rather than metaphysical necessity. As with Ockham, Autrecourt, and the Islamic and early modern occasionalists, the order we find in the natural world is in no way inherent to it but is entirely imposed from outside. The difference is that this external source of order is to be identified with a set of contingent laws rather than with God.

Problems with the Counterfactual: Finks and Masks

Finks:

C. B. Martin's "electro-fink" Example

Consider a live wire, which if touched by a conductor will cause electricity to flow into it. On the counterfactual account:

If the wire is touched by a conductor, then electrical current flows from the wire to the conductor.

Now suppose the wire is attached to an “electro-fink,” (renders a dead wire live when it touches a conductor and renders a live wire dead when it touches a conductor).

The conditional above will not be true of a wire when it is live.

The conditional will be true of a wire when it is dead.

Hence “the conditional fails to give necessary conditions for the wire's being live, since a wire could be live even when it is not true that it will transmit current to a conductor. And when the wire is dead, current will still flow from it to the conductor, because it will be made live by the electro-fink; hence the conditional fails to give sufficient conditions for the wire's being live, since a wire could in fact be dead even when it will transmit current to a conductor. The proper way to characterize the wire, in Martin's view, is to say that it has a power when it is live which is prevented from operating by the electro-fink, and lacks such a power when it is dead but is then given this power by the electro-fink --a power the wire's having of which cannot, as Martin's example shows, simply be reduced to the obtaining of certain counterfactual conditions. An ontology of real powers, in short, captures a crucial aspect of the causal situation that the counterfactual analysis cannot capture.

Lewis attempts to solve the difficulty posed by "finks" (as examples like Martin's have come to be known in the literature) by proposing the following "Reformed conditional Analysis" (Lewis 1997):

(RCA) Something x is disposed at time t to give response r to stimulus s if, for some intrinsic property b that x has at t, for some time t' after t, if x were to undergo stimulus s at time t and retain property b until t', sand x's having of b would jointly be an x-complete cause of x's giving response y, where an x-complete cause of y is one that includes all the intrinsic properties of x which contribute causally to y's occurrence.

This implies that the problem is a certain brief time lag between the conductor's touching the wire and the current's flowing to the conductor. During this lag the flow is blocked by the electro-fink. But if the wire has some property b that would at some time t' have caused the current to flow when the stimulus of the conductor is present if the electro-fink hadn't operated as quickly as it does, then the RCA will still be true.

Thus the notion of "cause" involved in the RCA can be analyzed in terms of a further counterfactual statement. The resulting "double counterfactual" analysis explicate “powers.”

Masks:

However, finks are not the only problem cases facing the counterfactual analysis, and examples of another sort --called "antidotes" or "masks" in the literature --have been raised against the RCA. To borrow an example from Alexander Bird (2007, pp. 27-29), consider a fatal poison for which someone who has ingested it has also taken an antidote. Suppose the antidote works by changing the body's physiology so that the poison does not have its typical effect. This is different from "finkish" cases insofar as the poison (unlike the wire in the electro-fink example) is not changed; it is rather the environment in which it operates that changes. And the example is such that while the antecedent of the RCA is true of it, the consequent is not.

That is to say, it is true that the poison retains at t' the intrinsic properties that give it a disposition to kill the one ingesting it, but it is nevertheless false that at t' those properties result in the death of the one ingesting it. Hence we have a counterexample to the RCA.

As further counterexamples, George Molnar cites "intrinsic maskers," powers the operation of which constitute an antidote for or mask the operation of other powers (Molnar 2003, p. 93). For instance, King Midas had the power to nourish himself, but this power was masked by his power of turning everything he touched into gold. The RCA does not capture his having the first power, since though Midas always retains the intrinsic properties by virtue of which he could nourish himself, his turning the food he touches into gold prevents them from ever causing him to be nourished. Molnar also points out that conditional analyses cannot capture powers that operate continuously and unconditionally.

He writes:

Rest mass is such a power according to General Relativity. Massive objects are spontaneously manifesting their gravitational power in continuous interaction with space-time.

Note that this line of criticism does not depend on reference to actual cases of unconditionally manifesting powers. The mere possibility of the existence of spontaneous manifestations is enough to refute relational analyses of powers in which the relation is conditionalized on some triggering event. (2003, p. 87)

Recent analytic powers theorists share the view of Scholastics, that we must be cause to distinguish between the possessing of a power and the manifestation of a paper.

…(P)owers must be distinct from their manifestations insofar as it can be possible for a thing to produce a certain outcome even though it never in fact produces it. For instance, it is possible for the phosphorus in the head of a match to generate flame and heat if the match is struck, and this is true even if the match is never in fact struck and is destroyed without ever having been used. The power of the phosphorus to generate flame and heat is what grounds this possibility.

Some might object that there would be, in the case of the match at least, one could articulate a “verification” test, and so, an observable situation in which the manifestation would actually occur.

However, D. H. Mellor (1974) gives the example of a nuclear reactor which has the power to cause an explosion, but never does so precisely because the safety mechanism which monitors for possible explosions shuts the reactor down before one can (actually) occur. Yet had the reactor not had the power to cause an explosion, the safety device would not have been needed in the first place.

A second argument

· manifestations of a power are eventS,

· possession of a power is a kind of state

Mumford adds a third argument:

Some powers do not persist through their manifestations. (e.g. A match loses its power to generate flame and heat once it is struck and actually generates it.)

Dispositions vs. Powers

Some writers use the terms interchangeably (e.g. Mumford and Anjum 2011, p. 4).

Not others (Bird 2013; cf. Oderberg 2007, pp. 131-32).

Bird suggests that someone could regard a dispositional analysis of a particular phenomenon as correct but nevertheless reducible (to, say, a counterfactual analysis), thus denying the metaphysical reality of powers.

“Powers would (if I understand Bird correctly) be dispositions which are not reducible via an analysis that makes no reference to dispositions.”

The attraction of a “powers theory” is that it provides a ground for causality that other, non-powers approaches cannot provide.

Causation is just the manifestation of a power.

It also seems better able to handle “polygenic” cases where an effect is the result of a combination of the manifestations of several powers operating in tandem. (e.g. two horses pulling a barge from either side of a canal).

‘Each horse pulls the barge in a direction at an angle from the canal, but the effect of this combination of manifested powers is that the barge moves straight ahead down the canal. This example also illustrates how powers are "pleiotropic" in the sense of making contributions to different kinds of effect. The pulling action of horse A, taken in tandem with that of horse B, may produce motion in a northerly direction, but the very same pulling action will result in motion in a different direction in other contexts.

This analysis of causation implies that there is indeed a relationship between causation and necessity, but it is complicated, perhaps more complicated than is often recognized.

1. There is a necessary connection between a power and its manifestation.

A stock example would be how solubility necessarily has dissolving as its manifestation.

2. Powers are pleiotropic (producing more than one effect) and an effect is typically polygenic (having more than one cause). Therefore, there are bound to be examples of cause and effect relations that are not necessary. We often say things like "Throwing the stone caused the window to shatter," but of course shattered windows don't always and necessarily result from thrown stones.

“…Even if there is a necessary connection between a power and its manifestation, effects are typically the result of several active (and passive) potencies operating in tandem --in this case, the solidity of a particular stone, the brittleness of a particular pane of glass, the strength of a particular person's arm, etc. are all relevant to the actual outcome --and if some of these potencies are absent (or if a "fink" or "mask" is present), an effect that would follow when a certain power is operating in their presence (or when the fink or mask is absent) would not in this case follow.

Mumford and Rani Lill Anjum (2011) characterize powers as "vectors" which combine in various ways to produce divergent outcomes. (Though Mumford and Anjum do not regard even the relationship between powers and their manifestations as either necessary or contingent, but as a tending towards which in their view constitutes an irreducible modality intermediate between contingency and necessity.)

This account can also explain the plausibility of counterfactual accpuonts.

Generally certain powers regularly generate certain predictable effects and some counterfactual description of the causal situation will naturally seem correct (at least initially).

But given the polygeny of effects, counterfactual accounts fail in cases of "finkish" and "masking" counterexamples. We cannot reduce causality to counterfactual dependence.

“The problem with Humean analyses is not so much that regularity and counterfactual dependence are not real aspects of causation, but that they are the consequences of causal relationships rather than being constitutive of causal relationships. What is constitutive is what can only be captured in the language of powers and their manifestations.

Further, requiring temporal sequence between events is not even prima facie plausible for causal examples such as

“two books leaning against one another and keeping each other from falling over, and a refrigerator magnet sitting motionless in place (Mumford 2009, pp. 275-76).

These cases of causation are not well described as sequential events, but rather as cases where "in a sense, nothing is happening." Rather it is a continuous state.

Further, some talk of causal event can be understood as a single “event” described differently:

The stone's pushing through the glass and the glass's giving way to the stone are distinct aspects of the causal situation, but they are not "loose and separate" events. They are rather two aspects of a single event. And the aspects in question are most plausibly just the ones a powers analysis would identify: the stone's active potency or power to shatter glass, and the glass's passive potency or power to be shattered. Similarly, the books, the magnet, the sugar, and the person tasting the sugar all plausibly manifest various powers even if the examples in question do not plausibly involve distinct and temporally separated events.

So then:

1. Arguments from contemporary analytic metaphysics support the thesis that causal powers are real features of the world

2. The standard Humean alternatives are inadequate

3. The “powers analysis” captures what is correct in those alternatives (regularities and counterfactuals) and also explains why they are subject to counterexamples of the finkish and masking sort.

4. The powers account can handle instances of causation for which the Humean approach is not even a prima facie plausible analysis in the first place (simultaneous causation).

5. The powers account explains why it is possible for a cause to generate a certain effect even if it never in fact does so.

6. The powers account can handle the complexity of actual causal situations (the pleiotropic and "vector" -like nature of causes and the polygenic nature of effects capture) in a way Humean analyses cannot.

To be sure, there are further issues that have arisen in the contemporary debate over powers, such as how powers ever get manifested, the relationship between "dispositional" properties and "categorical" ones, and the role a power's purported "directedness" toward a manifestation plays in explaining its necessary connection to the latter. We will address these issues below. Before doing so, let us look at how considerations from the philosophy of science have also contributed to a revival of the notion of causal powers.

1.2.2.3 Considerations from philosophy of science

Nancy Cartwright argues that an ontology of powers (or "capacities," as she usually calls them) makes better sense of the analytic method employed in sciences like physics than Humean approaches can (1989; 1992, reprinted in a slightly shortened form as chapter 4 of Cartwright 1999). Controlled experiments aim to determine what effect a factor will have in idealized circumstances, acting alone in a way it does not in the ordinary course of things. For instance:

Consider Coulomb's law of electrostatic attraction and repulsion. Coulomb's law says that the force between two objects of charge ql and qz is equal to qlq2/~' Yet, this is not the force the bodies experience; they are also subject to the law of gravity .... Coulomb's is not the force that actually occurs; rather, it is a hypothetical power hidden away in the actual force ....

Coulomb's law tells not what force charged particles experience but rather what is in their nature, qua charged, to experience. Natures are something like powers. To say it is in their nature to experience a force of qlq2/~ is to say at least that they can experience this force if only the right conditions occur for the power to exercise itself; for instance, if they have very small masses so that gravitational effects are negligible. (Cartwright 1992, p. 48)

A Humean counterfactual analysis would hold that Coulomb's law tells us the force two bodies would experience if their masses were equal to zero. But there are, as Cartwright notes, several problems with this suggestion. First, the antecedent of this conditional can never in fact be instantiated, which doesn't sit well with the Humean's insistence on analyzing causation in terms of actualities or "facts about what happens." Second, the appeal to what would happen if the masses were equal to zero suggests an interest in "what the total force would be, were there no other forces at work" (Cartwright 1992, p. 49). The counterfactual analysis itself thus seems implicitly to assume that there are powers whose operations can affect and be affected by each other. Third, that the focus is, specifically, on what would happen in circumstances where no other forces are at work as opposed to all the other circumstances in which a charged body might operate --suggests a commitment to there being a specific behavior that charged bodies will by nature exhibit on their own and try to exhibit even when impeded. We have, that is to say, an implicit recognition that a power makes a unique contribution to an overall outcome --what Molnar calls the "pleiotropic" character of powers, and what Mumford and Anjum call their "vector"-like operation.

The causal regularities the Humean would make fundamental are, in Cartwright's view, in fact an artifact of what she calls "nomological machines," relatively stable arrangements of components whose capacities or powers in combination give rise to relatively stable patterns of behavior (Cartwright 1999, chapter 3; Cartwright and Pemberton 2013). Even the fundamental laws of physics, Cartwright holds, only operate in a ceteris paribus way. Newton's law of inertia holds only in circumstances where no forces act on a body, circumstances which never actually obtain. Kepler tells us that planets move in ellipses, but this is only approximately true insofar as planets are always acted upon by the gravitational pull of other bodies. Kepler's law holds to the extent that it does only because the solar system constitutes a kind of nomological machine, whose components and their powers are arranged in a stable enough way that they give rise to behavior that approximates the law. Most nomological machines are, unlike the solar system, artificial, the product of experimental conditions.

Within the domain of scientific evidence for causal claims, Cartwright and John Pemberton distinguish between what they call "what-evidence," which concerns the things that enter into causal relations, and their arrangements (as in nomological machines); "how-evidence," or information about the processes these things and their arrangements are involved in; and "that-evidence," which concerns the regularities that result (Cartwright and Pemberton 2013). The trouble with Humean approaches to causation is that they deal only with "that-evidence" and cannot plausibly account (as the powers approach can) for the underlying "what-evidence" and "how-evidence." Empirical science involves activities like: identifying arrangements of things in the world into nomological machines and predicting their future states; constructing arrangements so as to control future events (as when setting up experiments or making artifacts); intervening in preexisting arrangements so as to alter the usual outcomes; building up knowledge of the markers of the presence of certain powers (such as a thing's having a certain microstructure); knowledge of the particular contributions such powers make to various outcomes; and knowledge of how these contributions combine in the processes found in various nomological machines (Cartwright and Pemberton 2013, p. 104).

The Humean has to insist that all of this can somehow be captured in a set of laws connecting certain stating features of a causal situation with certain effects. Even for a simple context like the flushing of a toilet, where the powers theorist would make reference to the way the causal powers of the various component parts combine or are impeded given the circumstances and the arrangements of the parts, the Humean has to posit a complex network of laws connecting (say) the exact shape of this specific part, the exact shape of that specific part, the exact arrangement they happen to be in, the vibrations caused by nearby passing objects, etc., with exactly the sort of outcome that occurs in such-and-such a particular case. But the number and complexity of such laws that would have to be postulated is immense; the suggestion that they can be reduced to some smaller set of laws is an unbacked promissory note; and appeal to such laws is neither necessary nor what actually characterizes our practice (Cartwright and Pemberton 2013, pp. 106-108).

Stathis Psillos notes that the Humean could object that we need to appeal to regularities or Humean laws in order to identify what capacities a thing has in the first place (Psillos 2008; cf. Psillos 2002, pp.190-96). He might say, for instance, that we can attribute to aspirin the capacity to make a headache go away only after we have established a regular association between taking aspirin and headaches going away. But then (so the argument goes), pace Cartwright, capacities are not more fundamental than Humean law-like regularities.

Cartwright's response (2008) is that the "laws" that enter into identifying capacities are not of the Humean sort, viz. regular associations between occurrent properties (that is, actualities or "facts about what happens," as I referred to them earlier). Rather, they will be laws which themselves make reference to capacities. An example of such a law might be: If an object of mass m manifests its capacity to attract an object of mass M a distance r away and nothing interferes, the second object will have an acceleration Gm/r² -- where the capacity is ascribed to a property we have other ways to identify and where we have a claim about what behavior occurs when the capacity is manifested. Even if there is a sense in which "a given capacity is what it is because of the laws it participates in" (Cartwright 2008, p. 195), it is not a sense that vindicates the Humean position.

Anjan Chakravartty argues that an ontology of powers (he uses the term "dispositions") is especially useful in defending scientific realism (2013; cf. Chakravartty 2007). Scientific realism is the view that our best scientific theories correctly describe mind-independent reality (as opposed, say, to being merely useful instruments for making predictions). The main consideration in its favor is, as Hilary Putnam famously put it, that "it is the only philosophy that doesn't make the success of science a miracle" (1975, p. 73). But scientific realism comes in different varieties. Entity realism holds that the theoretical entities posited by our best scientific theories really exist; structural realism, by contrast, holds that it is the structure of relations between the entities posited by such theories, rather than the entities themselves, which really exists. These are alternative way of dealing with the problem that many scientific theories of the past have turned out to be mistaken, and currently accepted scientific theories may turn out to be mistaken too. The entity realist accommodates this fact by affirming only the existence of certain entities posited by our best scientific theories, and not necessarily the other aspects of the theories. The structural realist holds instead that it is only the relations between the entities posited by the theories that the realist need affirm, while allowing that the other aspects may be false.

One virtue of a powers ontology, in Chakravartty's view, is that it allows the scientific realist to combine insights from both of these versions of realism. The strength of entity realism is its emphasis on the idea that causal knowledge of a putative entity that allows us to manipulate it gives us grounds for believing that it is real. The strength of structural realism is its emphasis on the idea that the relational features of a theory are the ones most likely to survive theory change. Now as Chakravartty writes:

The behaviours that entities manifest in virtue of the dispositions [or powers] they possess are generally described by scientific theories in terms of relations, often in the form of mathematical equations relating variables whose values are determinate magnitudes of the properties in question. (2013, p. 117)

To attribute powers to a thing, then, is both to identify its causal features and to do so precisely by reference to its relations. This unifies what would otherwise seem competing elements of the two versions of scientific realism in question.

A second unifying job a powers ontology performs, in Chakravartty's view, concerns the relationship between causation, laws, and natural kinds. All three notions commonly play a role in defenses of scientific realism, and all three are controversial. Defending them is easier when they can be shown to be tightly integrated, as they are on a powers ontology. For to attribute powers to a thing is precisely to attribute to it certain causal properties; these properties are commonly regarded as typical of the kind to which it belongs; and laws of nature can be understood as descriptions of the behavioral regularities that follow upon the manifestation of the causal powers a thing has by virtue of being the kind of thing it is.

Finally, a powers ontology affords, in Chakravartty's view, a way of dealing with a skeptical objection to scientific realism, to the effect that realism cannot account for the way that explanatory models that are equally successful but incompatible can apply to the same systems. For instance, in studies of fluid flow, it is sometimes useful to model a fluid as a continuous medium, and sometimes as a collection of discrete particles in motion. Since it cannot be both continuous and a collection of discrete particles, this might seem to pose a problem for realism. But the problem is avoided, Chakravartty argues, if we think in terms of attributing certain powers or dispositions to fluids. For a power manifests itself in different ways in different circumstances. (Recall Molnar's point about the "pleiotropic" character of powers and the "polygenic" nature of effects, and Mumford and Anjum's treatment of powers as "vectors.") We should not be surprised, then, that a fluid will by virtue of its powers behave in some circumstances in ways that makes it useful to describe it as if it were continuous, and in other circumstances in ways that make it useful to describe it in terms of discrete particles.

Other writers approaching our topic from a philosophy of science perspective have revived the Scholastic distinction between active and passive potencies --characterizing it instead as a distinction between powers and liabilities, or between active causal powers and passive causal powers (Harre and Madden 1975; Swinburne 1979, pp. 42-44; Bhaskar 2008, p. 87; Ellis 2001, p. 110) --and have argued that it is implicit in what science tells us the world is like. Brian Ellis writes:

Scientists today certainly talk about inanimate things as though they believed they had such powers. Negatively charged particles have the power to attract positively charged ones. Electrostatic fields have the power to modify spectral lines. Sulfuric acid has the power to dissolve copper. (Ellis 2001, p. 109)

Of course, the Humean will insist that such talk can be cashed out in terms of laws of nature or the like. But the writers in question respond that this has things precisely backwards --that laws of nature themselves must be explained in terms of powers and liabilities. Powers are what Ellis calls the "truth-makers" for laws of nature (Ellis 2001, pp. 112 and 222; Cf. Bhaskar pp. 45-56).

1.2.2.4 Powers and laws of nature

Here the concerns of the metaphysicians and the philosophers of science dovetail. Here we also come full circle, back around to the key notion with which the early moderns, who began the long intellectual trajectory against which recent analytic powers theorists are rebelling, sought to replace the Scholastic notion of causal powers. As noted already, because the early moderns came to regard matter as essentially passive, some of them relocated the source of activity in the world in divine decrees. Laws of nature were descriptions of how the world operated given these decrees. The idea of laws of nature was, then, originally theological. of course, most contemporary philosophers and scientists who appeal to laws of nature don't think of them in theological terms, but it is at least an open question whether laws can be made sense of apart from God. At least one contemporary philosopher with no theological ax to grind thinks not (Cartwright 2005).

Be that as it may, it is certainly difficult to see how laws of nature, understood non-theologically, can plausibly replace causal powers. For what is a law of nature if it is not a divine decree? There are four main candidate answers. Empiricists maintain that a law is a regularity to be found in nature. (There are different accounts of what sort of regularity counts as a law, but that is a complication we can ignore for present purposes.) There are several objections that can be raised against this sort of view (Cf. Mumford 2004), one of them being Cartwright's point that the ceteris paribus character of regularities is more naturally interpreted in terms of the operation of powers rather than laws. But the point to emphasize here is that if a law is just a regularity, then it doesn't explain anything. For what we need to know is why there are just the regularities that exist in nature, rather than some other regularities or no regularities at all. We might regard some given level of regularities as a special case of deeper regularities, but this will still leave the deepest regularities unaccounted for. Calling these regularities "laws" would merely be to re-describe them rather than to explain them. The powers theorist, by contrast, has an explanation of these regularities, and of why they hold in a ceteris paribus way: they are the "pleiotropic" or "vector" like manifestations of the powers things have by virtue of their essences, of course, the Humean may shrug his shoulders and say the basic regularities just exist without any explanation, but that is hardly to give a reason for preferring laws of nature to causal powers.

Another approach would be to interpret laws instrumentally rather than realistically. Laws are just useful tools for making predictions, developing technologies, and the like. But this faces the Putnamesque objection that it makes a miracle of the success of science's use of the notion of a law of nature. We need an explanation of why laws are such useful instruments if they are not real.

One realist alternative is to regard laws as relations between universals, with universals conceived of in terms of either Platonic realism or Aristotelian realism (Dretske 1977, Tooley 1977, Armstrong 1983). It might seem that this sort of view can explain the regularities that exist in nature, without resort to powers. Things in the world are related in the regular ways that they are because they are instances of universals, which are related in parallel ways. But if we interpret this approach in a platonic way, then we need an explanation of how laws conceived of as abstract entities existing outside the natural world come to have any influence on it, which merely pushes the problem back a stage. (Cf. Cartwright 2005) Yet if, following David Armstrong, we interpret it in an Aristotelian way, then the laws will depend for their existence on their instances, in which case they cannot be the explanation of those instances. (Cf. Mumford 2004, pp. 101-3). One might suggest, in the platonic case, that the laws operate by virtue of God's using them as a blueprint when creating the world; or, in line with the standard scholastic development of Aristotelian realism, that they pre-exist their instantiation in individual things as ideas in the divine intellect. But of course, in either case we will have brought God back into the picture, when the point was to find a nontheological account of laws.

A further objection to Armstrong's version of this position is put forward by Alexander Bird (2007). Armstrong takes universals to be related by "nomic necessitation." This is a move away from the Humean conception of things as entirely "loose and separate," but is still "semi-Humean" insofar as it relates things in such a way that they are only contingently necessary. Given the laws of nature which necessarily connect being an F with being a G, every individual F will be a G; but being an F and being a G could have been related by different laws instead. So, where N is nomic necessitation, given that N(F,G), it will be true that Vx (Fx -Gx).

But, asks Bird, is it necessary (in a fully anti-Humean sense that goes beyond mere nomic necessitation) that if N(F,G), then Vx (Fx -Gx)? If not, then the relationship between N(F,G) and Vx (Fx -Gx) is either accidental, in which case we do not have an explanation after all; or it is a relation of nomic necessitation, in which case we have a vicious regress. On the other hand, ifN(F,G) and Vx (Fx -Gx) are related in a strongly necessary, anti-Humean way, then N is essentially like the relationship between powers (Bird uses the term "potencies") and their manifestations --and thus not really a true alternative to the powers account at all. (See chapter 3 of Mumford 2007 for a useful survey of the debate over Armstrong's position.)

This brings us, finally, to the account of laws that some recent powers theorists have adopted. For Ellis (2001, 2002), a law is just a matter of a natural kind's having an essential property; and a causal law is just a matter of a natural kind's essentially having a certain dispositional property or causal power. On this view, laws of nature are necessary in the strong, metaphysical, anti-Humean sense that Armstrong's position shrinks from. Other powers theorists (such as Mumford) have opted to abandon the notion of laws as unnecessary once one rejects the passivist, anti-Aristotelian conception of nature that made the early moderns see a need for them. But even some Scholastic writers have refrained from going that far, one of them defining a law of nature or physical law in essentially Aristotelian terms as follows:

physical law,

1. an intrinsic tendency in a natural body or other nature to produce definite effects proper to its nature in a definite uniform way and measure or by determinate means ...

2. the scientific or mathematical expression of this constant way in which a natural body or other nature acts ... (wuellner 1956a, p. 70; cf. Bittle 1941, p. 422, and Smith 1950, pp. 97-99)

More recently, David Oderberg (whose influences are no less Scholastic than analytic) has endorsed something like Ellis's view, holding that the laws of nature are the laws of the natures of things, the ways things will behave give their essences (2007, pp. 143-51).

Naturally, laws thus understood can hardly replace causal powers and the rest of the anti-Humean metaphysical apparatus, since thus understood they presuppose the latter. Nor does there seem to be any principled reason for affirming laws of nature if they are not to be understood in either this Aristotelian, "bottom-up" way (to borrow Ott's terminology) or the theological, "top-down" way. As Ott notes, laws understood in neither of these ways seem to be "brute facts" (2009, p. 7) --and, he suspects, "either vacuous or incoherent" (p.249). Certainly it is difficult to see any motivation for them, other than their provision of an ad hoc way of avoiding a commitment to either Aristotelian causal powers or theism.

As this lengthy excursus on contemporary analytic metaphysics and philosophy of science shows, despite the long dominance of Humeanism, an essentially Scholastic notion of causal powers is very much alive, and supported by a wide range of arguments. We will see that the same thing is true ofthe other main elements of the Scholastic approach to causation.

1.3 Real distinctions?

We noted above that there is disagreement among Scholastics about whether the distinctions between act and potency, and between a substance and its powers, are real distinctions. These disputes are paralleled by recent debate within analytic philosophy over the relationship between categorical and dispositional properties (where "categorical" properties correspond roughly to actualities and "dispositional" ones roughly to potencies). Properly to understand the issues requires making a number of distinctions between kinds of distinctions. (Cf. Bittle 1939, chapter XII; Coffey 1970, pp. 104-7 and 139-57; De Raeymaeker 1954, pp. 62-69; Harper 1940, volume I, pp.342-60; Koren 1955, pp. 70-74)

1.3.1 The Scholastic theory of distinctions

Scholastics define a real distinction as one that reflects a difference in extra-mental reality and a logical distinction (or "distinction of reason") as one that reflects only a difference in ways of thinking about extra-mental reality. A logical distinction can be either purely logical or virtual. It is purely logical when it is merely verbal, without any foundation in reality. The distinction between "human being" and "rational animal" is (given the Aristotelian definition of a human being) a distinction of this sort. It is virtual when it has some foundation in reality. For example, a man's nature as a rational animal is (given the Thomistic account of essence, to be discussed in chapter 4) in reality one thing, not two. But we can view it either under the aspect of rationality or under the aspect of animality, for we know of instances when animality exists apart from rationality. Hence there is a virtual distinction between the two aspects. A virtual distinction can in turn be either major or minor (or perfect or imperfect). It is major or perfect when the concepts expressing the different aspects do not include one another, as is the case with animality and rationality since (again) there are cases when the one exists without the other even though they are united in human beings. It is minor or imperfect when the concepts do include one another implicitly, as in the case of "being" and "substance," since "being" covers everything that exists, including substances, and a "substance" is a kind of being.

A real distinction, which holds entirely apart from the way the intellect conceives of a thing, can also be either major or minor (or absolute versus modal). A major or absolute real distinction is a distinction between entities, though the entities may be of different types. Most obviously, individual objects like people, dogs, trees, and stones are really distinct. Also really distinct are parts of an individual object, such as two halves of a stone, an apple and the tree it hangs from, and the paw and leg of a dog. A third instance of a major real distinction would be that between a substance and its positive accidents --for instance, between a stone and its color. A fourth would be the distinction between accidents themselves, such as the distinction between quantity and quality. A minor or modal real distinction would be a distinction not between things but between a thing and its modes, understood as features that have no being apart from the thing. An example would the distinction between a material object on the one hand and its location or state of rest or motion on the other.

Among the marks of a real distinction, the clearest is separability. Hence we regard two dogs, or a dog and its leg, as really distinct because they can exist apart from each other. We regard an object and its location as really distinct because the former continues to exist even when the latter changes. But separability is not the only mark of a real distinction. Another is contrariety of the concepts under which things fall, i.e. an incompatibility between some of the elements of these concepts. For example, being material and being immaterial obviously exclude one another, so that there must be a real distinction between a material thing and an immaterial thing. A third mark sometimes suggested is efficient causality --the idea being that if A is the efficient cause of B, then A and B must be really distinct -though one writer objects that such a causal claim arguably presupposes, and thus cannot ground a claim about, a real distinction between A and B (Coffey 1970, p. 148).

A major or perfect virtual distinction may appear at first glance hard to distinguish from a real distinction. But the key to understanding the difference between any logical distinction and a real one is this: If the intellect's activity is essential to making sense of a distinction, it is logical; if not, it is real. Consider again the example of man's nature as a rational animal, or an animal's nature as a sentient corporeal substance. On the one hand, a man and an animal are each one thing. A particular animal's sentience is not really distinct from its corporeality, nor is either really distinct from its substancehood. It is a single substance which is at once corporeal and sentient. A particular man's rationality is not really distinct from his animality; nor, for that matter, are the sentience, corporeality, and substancehood he has by virtue of being an animal really distinct. He is a single substance which is at once corporeal, sentient, and rational. All the same, there are animals that lack rationality, corporeal substances that lack sentience, and (more controversially) substances that lack corporeality. An intellect that knows all this can therefore distinguish a man's rationality from his animality, an animal's sentience from its corporeality, and a corporeal thing's corporeality from its substancehood. Because there are animals that are not rational, corporeal things that are not sentient, etc., these distinctions have a foundation in reality. But because these things are not really distinct in men and animals themselves, and the distinction arises only when the intellect notes that there are animals without rationality, etc., the distinction is a logical one (specifically, a major virtual one) rather than a real one.

Here we come to some matters famously in dispute among Scholastics. If separability is not the only mark of a real distinction, is it nevertheless a necessary condition? Is a distinction between A and B real only if A and B are separable? Thomists answer in the negative, and thus draw a further distinction between a real physical distinction (which entails separability of the really distinct aspects) and a real metaphysical distinction (which does not entail separability). But Scotus and Suarez answer in the affirmative, maintaining that a distinction is real only when it entails separability. (Or at least this is so in created things; the Persons of the Trinity are held to be distinct but inseparable. Cf. Cross 2005, p. 109.)

Scotus also adds to the distinction between real and logical distinctions a third and intermediate kind, the formal distinction. (Cf. in addition to the literature on the theory of distinctions cited above: Ingham and Dreyer 2004, pp. 33-38; King 2003, pp. 22-25) Consider yet again a man's rationality and animality. Scotus agrees that there is no real distinction between them. However, the animality of a man is the same thing as the animality of a dog or any other non-human animal, and the animality of a dog is distinct from rationality (since, of course, it exists entirely apart from rationality). So the animality of a man must be distinct from his rationality. But though this distinction is not a real one, neither is it a logical one, since it reflects a difference that exists even apart from the intellect's consideration. Scotists call it a distinction between "formalities" --the formality of animality and the formality of rationality --and the distinction, purportedly neither real nor logical, is accordingly labeled a formal distinction.

The trouble with the notion of a formal distinction is that it is hard to see how it can avoid collapsing into either a real distinction or a virtual (and thus logical) distinction. For either the intellect plays some role in the distinction or it does not. If a man's rationality and animality are distinct entirely apart from the consideration of the intellect, then what we have is just a real distinction. Whereas if they are distinct because the intellect separates out the animality and the rationality on the basis of the existence of dogs and the like, then we have a logical distinction with a foundation in reality, namely a virtual distinction. There just doesn't seem to be some third, "formal" kind of distinction. However, some Scotists would argue that Scotus's formal distinction is in fact essentially the same as a virtual distinction, the difference with Aquinas being one of emphasis. A virtual distinction requires the operation of the intellect, but has a foundation in reality. Thomists emphasize the first element, thus labeling the distinction "logical" Scotus, on this interpretation, is merely concerned to emphasize the second element, the fact that virtual distinctions are grounded in mind-independent "formalities."

It is also worth noting that the motivation for drawing a purportedly intermediate formal distinction seems to disappear if we acknowledge, with Aquinas, that a real distinction need not entail separability. Since Scotus takes A and B to be really distinct only if they are separable, any two aspects of a thing that are not separable but which are evidently distinct even apart from the intellect's consideration of them will seem to be neither really distinct nor merely logically distinct, but something intermediate. This suggests an argument in favor of Aquinas's position on separability and against that of Scotus and Suarez: If every real distinction entailed separability, then there would have to be some intermediate, "formal" distinction between a real distinction and a virtual distinction; but there is no such distinction, since the formal distinction collapses on analysis into either a real distinction or a virtual distinction; so not every real distinction entails separability.

As Oderberg suggests, the claim that a real distinction entails separability is also subject to counterexamples. He writes:

Consider a circle. It has both a radius and a circumference. There is obviously a real distinction between the properties having a radius and having a circumference. This is not because, when confining ourselves to circles, having a radius can ever exist apart from having a circumference...

The radius of a circle is really distinct from its circumference, as proved by the fact that the latter is twice the former multiplied by pi. Since the radius is part of the property having a radius and the circumference is part of the property having a circumference, the properties themselves are really distinct though inseparable... [T]he same is true for triangularity and trilaterality. (2009, p. 677)

As Oderberg points out, what explains inseparability in cases like these is not identity or the absence of a real distinction, but rather the essence or nature either of the really distinct things A and B or (where A and B are qualities of a thing) of the thing whose qualities they are. "When it comes to circles (and triangles) there are mathematical laws, expressing their natures, that ensure inseparability" (p. 678).

1.3.2 Aquinas versus Scotus and Suarez

As we noted earlier in the chapter, while Aquinas regards the distinction between act and potency as a real distinction, Scotus considers it a formal distinction and Suarez a virtual distinction. Aquinas also takes the distinction between a substance and its causal powers to be a real distinction, while Scotus takes it too to be a formal distinction, and other scholastics a virtual distinction.

A natural way to think about these disputes is as follows. Potency, all Scholastics agree, cannot exist on its own but is grounded in a thing's actualities. A rubber ball has the passive potency to be melted at a certain temperature because it is actually made of rubber; a hammer has the active potency or power to shatter glass because it is actually made of steel. Now suppose we assume, with Scotus and Suarez, that a real distinction entails separability. Then for potency to be really distinct from act, it would have to be separable from act at least in principle. But, it is generally agreed, it is not separable. Therefore, the distinction between them must not be real, but only formal or virtual. And since a causal power is a kind of potency and the substance of which it is the power is a kind of act, the distinction between them must also be formal or virtual rather than real.

The Thomist, however, can reply to this as follows. First, for the reasons already given, the notion of a formal distinction intermediate between a real and a virtual distinction is dubious; the formal distinction collapses either into a real distinction or a virtual one. If it is real, then there is no genuine disagreement with Aquinas about the nature of the distinctions between act and potency or substances and their powers. If it is virtual, then there is a genuine disagreement. If the distinctions between act and potency and a substance and its powers are not real but only virtual, however, then since change and causation involve the actualization of potency, it seems to follow that change and causation are not real features of the world. That leaves us with an essentially Parmenidean view of reality. (cf. Coffey 1970, p. 303; phillips 1950, p. 182) But for the reasons given earlier, the Parmenidean view is incoherent. Therefore we must conclude that since change is real, the distinction between act and potency is real; and into the bargain, we have a further argument for the conclusion that a real distinction does not entail separability.

Further arguments for the real distinction between act and potency are as follows. (cf. Gardeil1967, pp. 197-98) First, an act or actuality involves completeness or perfection, while a potency is a mere capacity for completeness or perfection. But clearly a perfection and a mere capacity for that perfection are really distinct. For example, being spherical (as a child's rubber ball might be) is clearly really distinct from having the mere capacity to become spherical (as a parcel of molten rubber in a toy factory might have). Hence act and potency are really distinct. A second, related argument is that if a thing already has a potency (for having a spherical shape, say), but requires a cause distinct from it in order for it to come to have the corresponding actuality, then the potency and the actuality must be really distinct.

One argument for the real distinction between a substance and its powers goes as follows. Certain powers possessed by the same substance are clearly really distinct from each other. For example, the power of seeing is really distinct from the power of hearing, as is evident from the fact that an animal can exercise its power of sight without exercising its power of hearing, and vice versa. But if these powers weren't really distinct from the substance whose powers they are, then they couldn't be really distinct from each other either. Hence they must be really distinct from the substance which possesses them. (cf. Coffey 1970, pp. 304-305; Hart 1959, pp. 227-28)

Two further arguments go as follows. (cf. Aquinas, Summa Theologiae I.77.1; Koren 1955, p. 57; Koren 1962, p. 158) First, if a substance and its powers are not really distinct, then the latter will be actualized whenever the former is. Now a substance is actual as long as it exists, but its active potencies or powers are not necessarily actualized as long as it exists. For example, the phosphorus in the head of a match obviously exists even when its active potency or power of generating flame is not being actualized, and indeed even if it is never actualized. Hence the active potency or power must be really distinct from the substance. The second, related argument is that to deny a real distinction between a power and the substance that possesses it is essentially to commit a category mistake. A power is a kind of accident, and accident is just a different category from substance (cf. the discussion to come in chapter 3). Hence a substance and its powers must be really distinct.

1.3.3 Categorical versus dispositional properties in analytic metaphysics

This dispute among Scholastic metaphysicians illuminates and is illuminated by the debate over the relationship between categorical and dispositional properties in recent analytic philosophy. We noted above that while "disposition" and "power" are sometimes used interchangeably, there is another usage on which some philosophers would accept that there are dispositions but not that there are powers. The idea is this. We can think of a dispositional property as one that a thing has when a certain conditional statement is true of it, viz. the statement that if a certain stimulus is present to it, then a certain manifestation will follow. Stock examples would be fragility, which something has when, given that it is struck by a hard object, it will shatter; or solubility, which a thing has when, given that it is submerged in water, it will dissolve. Now a philosopher who thinks there are powers or potencies in the sense operative in this chapter obviously thinks there are dispositional properties. But a philosopher could deny that there are such powers or potencies and still affirm that there are dispositions, so long as he took them to be reducible to something that was not a power or potency. He could accept that conditionals of the sort in question are true, but argue that they are made true by non-dispositional or categorical properties --properties that a thing simply has, unconditionally as it were.

Stock examples of categorical properties would be shape, or having a certain structure, or spatiotemporal properties. The view that all dispositional properties can be reduced to categorical ones is called categoricalism. For a glass to be fragile, on this view, would just be (say) for the particles that compose it to bear a certain structural relationship to one another and for there to hold certain laws of nature governing particles bearing such a relationship. Opposing this position in recent analytic metaphysics is the property dualist view that there are irreducibly dispositional properties alongside the irreducibly categorical ones. And then there is pan-dispositionalism, which holds that all properties are dispositional and that there are no irreducibly categorical ones. Finally, there are monistic views which hold that there is only one kind of property but that it can be described either as categorical or dispositional. When this sort of view regards the categorical and dispositional aspects as being really there in this one kind of property, it might be labeled a two-sided brand of monism (also known as the limit view insofar as it sees the categorical and the dispositional as limits on opposite sides of a single continuum). When it regards this one fundamental property as at bottom neither categorical nor dispositional --its categorical and dispositional aspects being just different ways we might describe it --then it might be labeled neutral monism. (See Mumford 2007, chapter 5, for a useful overview of the debate between these views. cf. Armstrong, Martin, and place 1996; Damschen, Schnepf, and StUber 2009; Groff and Greco 2013; Handfield 2009; Kistler and Gnassounou 2007; Marmodoro 2010; and Mumford 1998)

Armstrong, whose views on laws of nature we considered above, is a chief proponent of the view that dispositions are real but also reducible to categorical properties. He agrees that dispositions cannot be given a purely conditional analysis. There must be some "truthmaker" which accounts for why a conditional of the sort in question holds. But the truthmaker can in his view be identified with the categorical properties of a thing together with the laws of nature governing those properties. Given the relationship of "nomic necessitation" holding between salt's molecular structure together with the circumstance of being immersed in water, on the one hand, with dissolving on the other, it follows that if salt is put in water, it will dissolve. That is all there is to salt's having the dispositional property of solubility: categorical properties plus laws of nature. No irreducible powers or potencies need be posited. (cf. Armstrong 1996a)

Of course, one objection to this account is that the view about laws of nature that it rests on is itself seriously problematic. In particular, and as we saw Bird object, Armstrong's conception of laws either makes of them non-explanatory brute facts, or leads to a vicious regress, or implicitly presupposes a power-like relationship between properties of precisely the sort Armstrong was trying to avoid. Another objection that has been raised against Armstrong is that categorical properties are essentially epiphenomenal, making no causal difference to the world. Their causal features are entirely extrinsic, depending on the laws that Armstrong takes to govern them only contingently; had the laws been different, the very same properties would have been associated with entirely different dispositions. What is the point, then, of positing such categorical properties if they don't do anything?

This consideration provides a motivation for pandispositionalism, which is defended by Mumford and Anjum (Mumford 2013; Mumford and Anjum 2011). If dispositional properties alone ever do anything, then perhaps they are the only kinds that exist in the first place. Nor, in Mumford and Anjum's view, are alleged examples of purely categorical properties compelling. Consider shape. superficially, having a certain shape might seem to confer no dispositions on a thing, but in fact it does. It is because of the difference in shape between a knife and a ball that the former can cut things and the latter cannot. Being round, the latter has a disposition to roll that a cube does not have. And so forth. A purported counterexample Mumford and Anjum attribute to E.J. Lowe would be a soap bubble, which is round but, it is suggested, does not have a disposition to roll; another alleged counterexample is Peter Unger's case of a soft sphere which squashes flat instead of rolling (Unger 2006, p. 269). But these, Mumford and Anjum insist, are not true counterexamples at all. The soft sphere fails to roll precisely because it loses its shape, and the bubble fails to roll because while it has a disposition to do so, it also has a disposition to stick to surfaces, which counteracts the first disposition. Yet as Lowe points out, that sphericity confers a power or disposition still doesn't entail that it is a power or disposition (2006, p. 138).

An objection to pandispositionalism raised by Armstrong is that its account of causality seems to lead either in a circle or a vicious regress. A disposition is the disposition it is only by reference to its characteristic manifestation. But if all properties are dispositions, then a manifestation will itself be a further disposition. The disposition to produce A will just be the disposition to produce the disposition to produce B, which will in turn be the disposition to produce the disposition to produce C, and so on.

Writes Armstrong:

All serious distinction between powers and the manifestation of powers gets lost ... Causality becomes the mere passing around of powers from particulars to further particulars. To put it scholastically, the world never passes from potency to act ... nothing ever happens... There may not be a contradiction here, but it is position that I find unbelievable. (Armstrong 2005)

Mumford replies that a "passing around" is a kind of event, so that something plausibly is happening on the pandispositionalist analysis. Armstrong's chief complaint, in Mumford's view, is really that he finds irreducibly dispositional properties "mysterious" (Armstrong 1996b, p. 91). Writes Mumford:

Indeed, it appears that he thinks that the things that are passed around are not real at all. Pure powers... are thought of by Armstrong as mere potencies: potential rather than actual.. [But the] realist about dispositions or causal powers will accept such powers to be real enough ... [Powers] are certainly assumed as actual in their own right, whether or not they are manifested... When I ascribe a disposition I ascribe it actually and unconditionally. Passing round of powers would be for the realist, therefore, the passing round of something actual. (Mumford 2007, p. 88; cf. Mumford and Anjum 2011, p. 6)

As Lowe points out, though, the trouble with the threat of regress or circularity is not merely Armstrong's concern about whether anything can ever happen on such an account. It is that no property can get its identity fixed on a pandispositionalist account (2006, p. 138; Cf. Robinson 1982, pp. 114-15). The nature of a property A will be determined by reference to a property B, whose nature will be determined by reference to a property C, whose nature will be determined by reference to D, and so on either ad infinitum or in a way that leads us back to A. So what is the nature of A? If we say that the series goes on to infinity, then we never actually give the nature of A but just keep deferring the question forever; if we say that the series loops around back to A, then we give the nature of A by reference to the nature of A, which is no answer at all. Bird, who also defends pandispositionalism, suggests that the problem can be solved by appealing to the mathematical field of graph theory (2007, chapter 6). A power or dispositional property can be uniquely identified by the position it occupies in an asymmetric graph. As Oderberg argues in reply, though, even when a node in such a graph can be given a unique definition, the definition will still be circular, so that the problem is not really solved at all (20l2a; Cf. Oderberg 2011 and 2012b).

The "two-sided" or "limit view" version of monism has been defended by Martin (Martin 1996), and the "neutral monist" version by, at one point, Mumford (1998). Mumford has, under the influence of Molnar (2003), since given the latter view up as insufficiently realist about dispositions (Mumford 2013). And Armstrong poses a dilemma for Martin's version. Is the relationship between the categorical and dispositional "sides" of properties contingent or necessary? If it is contingent, then the categorical "side" could have been associated instead with different dispositions or even with no disposition. But in that case, what does it amount to to call the disposition a "side" of this property? If, on the other hand, the disposition is necessarily connected to the categorical side, then whatever causal work is supposed to be done by the dispositional "side" will necessarily flow from the categorical "side" to which it is connected. And in that case we might as well "cut out the middleman" and take the categorical to be what produces effects (Armstrong 1996b, pp. 95-96; cf. Mumford 2007, p. 85).

This leaves the property dualist view that categorical and dispositional properties are distinct and equally fundamental. It has been defended by Ellis (2001), Molnar, (2003), and U. T. place (1996). Against the categoricalist claim that the causal powers of a thing can be accounted for in terms of categorical properties like structure, Ellis and Caroline Lierse object:

[T]he causal powers of things cannot be explained, except with reference to things that themselves have causal powers. Structures are not casual powers, so no causal powers can be explained just by reference to structures. For example, the existence of planes in a crystal structure does not by itself explain its brittleness, unless these planes are cleavage planes --that is, regions of structural weakness along which the crystal is disposed to crack. But the property of having such a structural weakness is a dispositional property that depends on the fact that the bonding forces between the crystal faces at this plane are less than those that act elsewhere to hold the crystal together. Therefore the dispositional property of brittleness in a crystal depends not only on the crystal's structure, but also on the cohesive powers of its atomic or molecular constituents. However, cohesive powers are causal powers. (Ellis 2001, pp. 115-16; adapted from Ellis and Lierse 1994).

Here the considerations raised by Armstrong and Mumford against the other views might seem to pose a dilemma for the dualist. Are categorical properties themselves efficacious or not? If so, then it is hard to see why distinct dispositional properties are needed in order to account for causality, in which case we might as well opt for Armstrong's categoricalism. But if they are not, then they are epiphenomenal and do no explanatory work, in which case we might as well opt for Mumford and Anjum's pan-dispositionalism. (cf. Mumford 2007, p. 83) But Ellis and Lierse hold that though spatial, temporal, and other categorical properties are not causal powers, we know they are there because they enter into the laws that describe the operation of causal powers. For instance, spatial separation will be relevant to the strength of gravitational attraction or electrical repulsion. (Cf. Ellis 2001, pp. 137-38; cf. Ellis 2002, pp. 171-76) Molnar too argues that the operation of powers is "locationsensitive" --citing, like Ellis and Lierse, the role distance plays in the operation of a force --so that for a categorical property like location to be causally inert is not for it to be causally irrelevant (2003, pp. 162-65). place (1996) argues that there is a sense in which a categorical property causes a disposition the structure of a crystal is, after all, what makes it brittle. A manifestation might be seen, then, as the direct effect of a disposition and the indirect effect of the underlying categorical basis of the disposition (which causes both the disposition and, through it, the manifestation). In these different ways, the property dualist can defend the claim that both categorical and dispositional properties do real explanatory work.

Naturally, Scholastic philosophers will tend to sympathize less with categoricalism than with those views which affirm irreducible dispositions, as marking a welcome departure from Humean orthodoxy and a rediscovery of potency as a real feature of the world. They might also favor property dualism over the various monistic brands of dispositionalism, as closer to the Scholastic insistence on the reality of both act and potency. However, they are also bound to regard the recent debate, however salutary, as still too beholden to Humean metaphysical assumptions and insufficiently nuanced in the distinctions it presupposes.

For example, as the passage quoted above indicates, Mumford essentially agrees with Armstrong that everything real must be actual. Armstrong's view is that since irreducible dispositions or powers are not actual, they are therefore not real; while Mumford argues that since they are real, they are actual. Yet the whole point of the Aristotelian theory of act and potency is that, contrary to Parmenides' assumption, actuality does not exhaust reality --that being-inpotency is a middle ground between being-in-actuality and sheer nothingness or non-being. Similarly, Molnar claims that a power is an "actual property" rather than an "unrealized possibility" (2003, p. 126), while Martin says:

Dispositions are actual though their manifestations may not be. It is a common but elementary confusion to think of unmanifesting dispositions as unactualised possibilia; though that may characterize unmanifested manifestations. Armstrong appears to be guilty of this confusion in his reference to 'potential being'... (Martin 1996b, p. 176)

This too is a false dichotomy and misses the Scholastic philosopher's point. As we have seen, scholastics distinguish between logical or objective potencies on the one hand and real or subjective potencies on the other. Unrealized possibilities or possibilia would fall into the former category, but causal powers fall, not into the class of actualities, but rather into the class of real or subjective potencies potencies that are in a real, concrete subject rather than being mere abstract possibilities. Galen Straws on (2008), who argues for the identification of the dispositional and the categorical, presupposes that there can be no real distinction between A and B unless A and B can exist apart. But this, of course, simply begs the question against the Thomistic view that a real distinction does not entail separability. (Cf. Oderberg 2009)

Ellis seems at least to hint at the needed distinctions and at a more thoroughgoing challenge to prevailing suppositions when he says that dispositions need be grounded only in "occurrent" rather than categorical properties, and notes that the idea that the fundamental occurrent properties must be causally impotent and thus categorical rather than dispositional reflects a Humean set of assumptions (2001, pp. 116-17; on "occurrent," cf. Lowe 2006, p. 139 and Oderberg 2007, p. 132). As this indicates, the notion of the "categorical" is not exactly the same as the scholastic notion of "actuality," since Scholastics by no means regard actualities as per se causally impotent. Nor, as we have seen, is the notion of the "dispositional" exactly the same as the Scholastics' notion of potency, insofar as some dispositionalists take dispositions to be actualities. "Categorical" and "dispositional" properties are also often spoken of as if they could at least in principle exist apart from one another. As Oderberg points out, from the point of view of the theory of act and potency, this just gets things fundamentally wrong. There is no such thing as potency without act, and (apart from God, who is pure act) no such thing as act without potency (Oderberg 2007, p. 138). Potency always presupposes some actuality that shapes or circumscribes it. A power is a power to generate this particular manifestation rather than that one, and reflects the form of the substance having the power, a form which actualizes its otherwise indeterminate prime matter. (See chapter 3.) Act, in any finite and changing substance, always presupposes some potency as the principle which limits it and accounts for its changeability. Potency and act are both really distinct and inseparable. Certainly, merely to suppose otherwise is to beg the question against the scholastic position rather than to refute it. Or at least, it is to beg the question against the Thomistic version of the Scholastic position.

In particular, the categoricalist and the pan-dispositionalist essentially presuppose, with Scotus, Suarez, and Descartes, that a real distinction between what they call categorical and dispositional properties would entail separability. The categoricalist starts with the idea that dispositional properties cannot exist apart from the categorical properties in which they are grounded, and concludes that they must be reducible to categorical properties. The pandispositionalist starts with the idea that we should take purported categorical properties seriously only insofar as they have causal power, and concludes that if they have it then they are really just powers or dispositional properties. The "two-sided" and "neutral monist" views allow for both the categorical and dispositional only insofar as they are really just aspects of the same property. But if a real distinction does not entail separability, then we need not infer from the dependence of the "dispositional" on the "categorical" that the former is reducible to the latter, or from the efficacy of the "categorical" that it must really be "dispositional," or from the reality of both the "dispositional" and the "categorical" that they must really in some sense be the same property. Nor, from the real distinction between the "dispositional" and the "categorical," would we need to infer to a form of property dualism on which they could exist apart from one another.

Given the baggage associated with "categorical" and "dispositional" in contemporary philosophy, the Scholastic will in any event prefer to stick to the traditional jargon of act and potency. He might also be forgiven for thinking that while each side of the current debate has grasped an important part of the truth that the theory of act and potency seeks to capture --the categoricalist, the insight that actuality is fundamental to reality; the various brands of dispositionalism, the insight that we cannot make sense of causation without potency --both sides have also missed the larger picture, the set of problems that spawned the development of the theory of act and potency in the first place. Making all of reality "categorical" or actual entails a return to Parmenidean static monism; making all of reality "dispositional" or potential threatens a return to Heraclitean dynamic monism. But neither the Parmenidean nor the Heraclitean extremes are ultimately coherent, and the only way to avoid them is to affirm both act and potency as really distinct, even if inseparable, aspects of reality.

---

[1] http://www.u.arizona.edu/~aversa/scholastic/24Thomisticpart2.htm

Pope Pius X issued Doctoris Angelici "So far as studies are concerned, it is Our will and We hereby explicitly ordain that the Scholastic philosophy be considered as the basis of sacred studies. . . . And what is of capital importance in prescribing that Scholastic philosophy is to be followed, We have in mind particularly the philosophy which has been transmitted to us by St. Thomas Aquinas. It is Our desire that all the enactments of Our Predecessor in respect thereto be maintained in full force; and, where need be, We renew and confirm them and order them to be strictly observed by all concerned. Let Bishops urge and compel their observance in future in any Seminary in which they may have been neglected. The same injunction applies also to Superiors of Religious Orders."

[2] Jean-Baptiste Poquelin, known by his stage name Molière; 1622–1673), was a French playwright and actor who is considered to be one of the greatest masters of comedy in Western literature. Coins the phrase “virtus dormitiva,” in The Imaginary Invalid, his play in which lampoons a group of physicians providing an “explanation” of the sleep-inducing properties of opium as stemming from its "virtus dormitiva".

[3] (Mumford 1998, pp. 136-41)

[4] (Kenny 1989, pp. 73-74).

[5] Hulswit , Menno, “A Short History of ‘Causation’” http://see.library.utoronto.ca/SEED/Vol4-3/Hulswit.htm#_edn1

This is an abridged version of the first chapter of From Cause to Causation. A Peircean Perspective. Dordrecht, Kluwer Publishers, 2002.

[6] Wrote The Exercitationes which argues for the negative effects of the blind acceptance of the Aristotelian dicta on physical and philosophical study;

[7] In Syntagma, he offers a view or physics and metaphysics but with glaring contradictions, specifically between his embrace of Epicurean/Democritus physics and atomism with is support for an immaterial view of God who is the author of the universe, an immaterial rational soul, and freewill. He suggests that a calor vitalis (vital heat), a species of anima mundi (world-soul) as a physical explanation of physical phenomena. He also posits a theory of the weight essential to atoms as being due to an inner force impelling them to motion. But this too seems irreconcilable with his general doctrine of mechanical causes.

[8] David Hume (1711–76). An Enquiry Concerning Human Understanding. The Harvard Classics. 1909–14. Of the Idea of necessary Connexion Part II

[9] Def. relating to or denoting certain principles, such as laws of nature, that are neither logically necessary nor theoretically explicable, but are simply taken as true.