On an Infinite Regress of Causes

The general structure of cosmological arguments is well known: Starting from general features about the world (e.g., that there is movement), these arguments proceed by pointing out that these general features must have a cause, and this cause must have a cause, and since if there were no first cause there would not be any subsequent causes, there must indeed be a first cause.

Much has been written over the years about cosmological arguments, and there are, in truth, many different types of such arguments. One must distinguish in esse arguments (arguments in the order of ontological dependency) from in fieri arguments (arguments in the order of temporal becoming). One must distinguish arguments that proceed from the fact of contingency from arguments proceeding from movement or causality. One must distinguish arguments that use the principle of sufficient reason from arguments that do not make explicit use of this principle. To enter into any discussion of the diverse array of such arguments shall not be our concern here. What I want to deal with in this post is an interesting attack by James Sadowsky on the notion that there can be an infinite regress of causes ("Can there be an Endless Regress of Cause," International Philosophical Quarterly, 20-4, 1980).

Sadowky points out that the operative principle in the cosmological argument is that "if each cause of A were itself in need of a cause, then no cause of A could exist and hence A itself could not exist." From this the argument proceeds easily: A (let us say there is motion in the world) exists and thus all of A's causes are not in need of a cause, that is, there is some cause that is itself not in need of a cause. [One thinks here perhaps of Schopenhauer's quote that the law of universal causation is like "a hired cab that one dismisses when one reaches one's destination."]

Critics of cosmological arguments oftentimes point to the obvious fact that in order for A to be, there must be some B which causes A, and in order for B to be there must be some C that causes B, and that this series can run back to infinity. Think for a moment about the infinite series of integers. For every integer I, there is some integer 'I - 1' such that I is generated from 'I - 1' by adding '1'. Any integer can be "caused" by taking the preceding integer and adding one. There is no problem with this series running back to infinity, of course. If it did not, we would have a pretty truncated mathematics.

But proponents of cosmological arguments often make claims about how an actual infinite is not possible - - after all, Aristotle said so - - and that the analogy between an infinite causal series in the world and the infinite series of integers is not great. For the infinite causal series, the operative principle specified previously holds, which does not in the generation of infinite mathematical series: If each cause of A were itself in need of a cause, then no cause of A could exist and hence A itself could not exist.

Sadowsky asks us to compare the statement of the cosmological argument that no causation can take place because each act of causation requires a previous act of causation with the following: no permission can be asked for because each asking of permission requires a prior asking of permission. Consider this statement:

1) No one may do anything (including asking for permission) without asking for permission.

Is (1) true? It seems not, for how could it be that the condition for asking for permission is itself the asking of permission. It seems that permission asking in order to do every X cannot run back to infinity, because X includes the asking of permission. The activity of asking for permission cannot run back to infinity because there would be no first asking of permission and thus no subsequent series of permission asking.

Sadowsky asks us now to consider Ryle's demolition of the so-called "Intellectual Legend": Never do anything (including thinking) without first thinking about it. Consider then (2):

2) No one ought to anything (including thinking) without first thinking about it.

Is (2) true? It seems not, for how could it be that the condition for thinking is itself based upon thinking? It seems that an infinite series of intellectual reflection based upon intellectual reflection is impossible, for how can it be that one's reflection on something (call it X) must result from X?

Although Sadowsky does not explicitly say so, he supposes that (1) and (2) are unsatisfiable, that is, there cannot be a state of affairs of every act of intellectual thinking being dependent upon anterior acts of intellectual thinking. Why? Because if there is real contingency in intellectual thinking - - if it is possible to consider propositions either shrewdly (intellectually) or stupidly - - and the condition for considering propositions shrewdly (intellectually) is a prior condition of having considered propositions shrewdly (intellectually) and not stupidly, then in order for there to be subsequent acts of intellectual consideration there must have been a first act of intellectual consideration. In other words, there is no possible world in which there can be an infinite regress in the order of prior intellectual operations as a prerequisite of subsequent intellectual operations. There must be a first intellectual operation that grounds subsequent intellectual operations, or there would have been no subsequent intellectual operations. Similarly, there must be a first permission that grounds subsequent acts of granting permission. There can be no possible world in which one cannot do anything without first asking permission, if it is true that "doing anything" includes the seeking of permission.

In (2) it is impossible to break into a series of intellectual considerations without there being an intellectual consideration not grounded in anterior intellectual considerations. In (1) there cannot be a breaking into the series of permissions without there being a first permission granting that needs not anterior permission. We have here the claim that there must be intellectual consideration that is not the result of an intellectual consideration, and a permission seeking that is not the result of a permission seeking. Now the question is simply this: is an infinite regress in the order of causes analogous to these two cases? Is it true that (3) is unsatisfiable?

3) For each and every cause, there must be a cause of that cause.

Is the denial of (3) somehow contraditory? Is it contradictory to have an uncaused causer? Or, put differently, if there must be an an unpermitted permitter, and a nonintellectualized, intellectualizer, why not an uncaused causer? Why should causality be regarded differently?

It seems that the answer to this might lie in the different contexts in which intellectual considerations, permission seeking, and causing inhabit. It strikes me that intellectual considerations and permission-seekings are teleological activities. Take, for instance, the notion of an infinite series of purposes. It seems like an infinite order in the series of final causes is indeed unthinkable. If everything that occurs, occurs for the sake of something else, is it not true that there must be finally something for which all things occur. (Heidegger traces this back to Dasein, of course.) No infinite regress in the order of teleological "reasons for" is possible, for it seems, that in order for there to be subsequent "reasons for" there must be a first "that upon which all reasons are ultimately reasons for."

Most of the time, however, we regard the order of causes as a nonteleological context: A causes B which causes C, etc. In a universe without meaning or purpose, why would an infinite series of causes not be allowed? Of course, there is not a first cause on the basis of which subsequent causes are! That is the point of thinking about an order of causes purely extensionally. There is nothing unsatisfiable about (3), though there might be about (3') below:

3') For each and every reason, there must be a reason for that reason.

I think many people would dispute (3') being satisfiable on the basis of there being finally a 'brute reason or purpose' on the basis of which other reasons find their positions. (Heidegger would agree here.) We often trace human reasoning back to a human telos generally. Why did Bob do x? He had such and such reasons for doing x. But why did he have these reasons? Because he ultimately desired that some y come about, and he reasoned in ways that would eventuate in y. But why did he desire that some y come about? Reasons must stop somewhere, and one might just say that his desire for y just is. Is it reasonable? Perhaps, but it is not reasonable based upon other reasons. It is an unreasoned reason.

Sadowsky has forced us to see more clearly into what we often mean by an infinite regress in the order of causes. We mean something that is quite without meaning. It seems in an unthinking universe without value and purpose there could be an infinite series of causes. Whether a thing is or is not is not the same kind of question as whether a proposition is reasonable or not. While the second concerns a teleological context where an infinite regress is impossible, this is not so of the first. Or at least that is what one might reasonably say.