Secondary Causes

One of the many things I like Aristotle for is his clear concern for what are sometimes called “secondary” causes. As usual with Aristotle, “cause” means any kind of explanation or determining reason; explanation is in general not univocal; and things are the way they are due to the combination of many causes. Secondary causes for Aristotle play an irreducible role in the overall determination of things. This is part of what I recently called the dignity of finite beings.

The way in which secondary causes operate is pluralistic; there is no single, seamless matrix of causality in the world. Instead we have a superabundance of meaning. Determination is always grounded in actuality, but actuality is never the whole story. We get a better grasp on things by taking counterfactual potentiality into account.

Secondary causes may be either “moved” or “unmoved”. If the form of an animal’s leg joint counts as an unmoved mover, the number of unmoved movers in the world is truly vast. There are also a vast number of moved movers.

Even though there is a great deal of practically meaningful determination in the world, neither God nor physics comes anywhere near completely determining human reality. The world has both real determination and real play in it. See also What and Why; Interpretation).

The Importance of Potentiality

I think modern philosophy generally is handicapped in its thinking about the empirical world by its lack of a notion like Aristotelian potentiality. To build context, I need to first say a bit more about the role of actuality.

The modern concept of a factual, existing world is relatively close to Aristotelian actuality, but the first big difference is that it is not paired with anything. The modern concept of a factual world is something that is supposed to be complete in itself, whereas for Aristotle, actuality in the world is always complemented by some correlative potentiality. Aristotle did not consider actuality alone to be sufficient to account for the world as we experience it.

Actuality also does not exactly correspond to a state of things, but rather expresses what is effectively operative. This is semantically a bit deeper than a notion of state. At the same time, it does not have state’s strong implications of complete determination. It also does not have the monolithic unity of a state. Actuality in the world consists of many coexisting things. Further, it is not intended by itself to provide all the resources needed to account for change and what happens next. This is related to the fact that for Aristotle, the operative determination of things is not entirely univocal. (See also Equivocal Determination.)

Enter potentiality. Potentiality is exactly what is not univocal in the actual determination of things. It corresponds to multiple alternative concrete possibilities of realization already implicit in current reality. This is a far more specific notion than mere logical possibility. Potentiality is closely tied to and informed by the current actuality, in that it exactly occupies the real gaps or holes in the actuality’s incompletely univocal determination. For each aspect of things where there is not univocal determination, there are instead multiple potential alternatives. This correlates with the fact that, for Aristotle — in contrast to Poincaré’s classic formulation of modern determinism — the present does not completely determine the future.

Poincaré famously claimed that from the state of the universe at any arbitrary point in time, its entire future is completely determined. This resembles the Stoic notion of fate, transferred to a modern event-based model of causality. For both the Stoics and Poincaré, the world is completely univocally determined. Like Aristotle, they emphasized the intelligibility of the world and of change in the world, but they made the very strong assumption of complete univocal determination. Aristotle did not.

Aristotle’s notion of intelligibility was broadly semantic, whereas Poincaré’s was mathematical. With semantic interpretation, there is always a question of how far we develop the account, which in principle could be extended indefinitely. It thus naturally lends itself to an account of incomplete determination, corresponding to some stopping point. Aristotle does not claim any more determination than he can show. Poincaré’s approach, by contrast, requires that we assume there is a complete univocal determination of the world by mathematical laws, even though we can never even come close to knowing enough to show it. This assumption leaves no room for anything like potentiality. Potentiality, it seems to me, could only have a place in a semantic approach to intelligibility.

The modern factual world is usually considered as something that just is, without modal qualification, but I have increasingly begun to doubt whether for Aristotle there is any non-modal account of the world. I read actuality and potentiality both as modal concepts, and everything in the Aristotelian world as parsable into actuality and potentiality.

What’s important about this is that potentiality is not just some mysterious “metaphysical” concept that we could maybe do without. It is a distinct logical/semantic modality supporting multiple virtual alternatives for the same thing. It allows us to intelligibly account for the incomplete determination we really experience, rather than treating real-world incompleteness and ambiguity as if it were a kind of flaw. (See also Structural Causality, Choice; Values, Causality; Structure, Potentiality.)

Equivocal Determination

Even though in hindsight it is possible to find reasons why things turned out as they did, the past does not completely determine the future. In general, ahead of time, multiple outcomes are possible. There are many possible ways to meet a requirement, need, or desire, and there are chance intersections and collisions of different, mostly independent vectors of determination. This is all to say that the determination of things is in some measure equivocal (i.e., not univocal). (See also Efficient Cause; Free Will and Determinism; Kantian Freedom.)

A related point is that there is such a thing as objective ambiguity in the world. This means not just that we are unsure or conflicted about something, but that the best or most complete evidence available in a particular case may have more than one reasonable interpretation. (See also Aristotelian Identity; Things in Themselves; Copernican; The Epistemic Modesty of Plato and Aristotle.)


Aristotle is the source of our modern notion of univocity. He was also the first to point out many things of interest that overflow our attempts at univocal representation.

Something is said univocally (i.e., without equivocation) if the meaning is the same whenever the representation is the same. This is an important, desirable property in logic, modern science, and engineering. A lack of univocity is one of the major sources of logical inconsistency or incoherence in representation.

A kind of univocity also applies to a Kantian unity of apperception. The unity of a unity of apperception is more or less equivalent to the univocity of an account of something. So since there is a kind of moral imperative to achieve unity of apperception or improve upon it, there is in that way also an ethical use for univocity. But just as unity of apperception requires constant renewal, so too do our attempts at comprehensive univocal accounts of things.

Aristotle frequently points out things that are “said in many ways” (i.e., not univocally). This refers not just to practices of ordinary language, but to meant realities as well (see Equivocal Determination). Being and cause are among the things said in many ways. He also points out cases where a more fluid, pluralistic approach is appropriate — the classification of animals, for instance — or just applies such an approach. Aristotle wants to be faithful to the variety, subtlety, and complexity of the world and everything in it. This is the famous Aristotelian manysidedness, often praised by Hegel.

Univocity is never simply found; it is a possible property of our constructions and passive syntheses. In the design of representations and schemas, there may be tradeoffs between coherence and correspondence, or consistency and comprehensiveness.

If there is an ethical imperative to univocity, there is also one to manysidedness. These are really two sides of one coin (a sort of responsibility our representations have to reality, as Brandom would say). To paraphrase Whitehead, we should seek univocity and distrust it. (See also Aristotelian Semantics; Mutation of Meaning.)