This is the second part of a series developed through the Department of Philosophical Theology at the Institute of Lutheran Theology's Christ School of Theology, exploring major results in modern logic and their significance for philosophical and systematic theology.
The great achievements of Frege, Peirce, and Cantor vastly enlarged the expressive resources available to logic and mathematics. Quantification made it possible to speak formally about all or some objects satisfying a condition; the logic of relations permitted systematic treatment of structures involving two or more objects; Cantor showed that infinity itself could be mathematically articulated and that infinite collections could differ in cardinality. Yet precisely this increase in expressive power produced a new problem. Once logic could speak about collections defined by conditions, what prevented us from forming a collection corresponding to any condition whatsoever?
The answer, discovered with particular force by Bertrand Russell, was contradiction.
Consider the apparently innocent idea of the collection of all collections that are not members of themselves. Let us call this collection R. We can describe it this way:
R = {x : x ∉ x}.
Read: R is the collection of all objects x such that x is not a member of itself.
Now ask whether R itself belongs to R. By the very condition defining R, we obtain:
R ∈ R if and only if R ∉ R.
Read: R is a member of itself if and only if R is not a member of itself.
Either answer produces its opposite. If R belongs to itself, then by definition it must not belong to itself. But if R does not belong to itself, then it satisfies the condition for membership in R, and therefore it does belong to itself.
This was not merely an amusing puzzle. Russell communicated the paradox to Frege in 1902 while the second volume of Frege’s Grundgesetze der Arithmetik was in press. Frege immediately recognized the seriousness of the difficulty. His project had aimed to show that arithmetic could be derived from logical principles, but the paradox exposed a defect in the assumptions governing the formation of extensions or classes. The new logical machinery had become powerful enough to undermine one of its own foundational constructions.
The philosophical lesson is easy to state but profound in its consequences: Not every condition that can be expressed determines an object corresponding to that condition.
Suppose we can specify a condition F and say intelligibly what it would be for something to satisfy F. It does not follow merely from this that there exists an object consisting of all and only the things satisfying F. The movement from a condition to a corresponding totality requires justification. This distinction became one of the central lessons of twentieth-century logic.
One response came through Ernst Zermelo and the subsequent development of axiomatic set theory. Instead of assuming a general principle according to which every condition determines a set, Zermelo restricted the circumstances under which sets may be formed. One does not simply move from a condition F to “the set of all Fs.” Rather, sets are generated according to specified axioms.
The dangerous principle can be expressed simply as:
For every condition F, form the set {x : Fx}.
Read: for any condition F whatsoever, there exists a set containing exactly those objects that satisfy F.
Russell’s paradox showed that this principle cannot be accepted without restriction.
Zermelo’s alternative was much more cautious. Begin with a set already given, and then select from it those members satisfying a specified condition. Thus one might have a set A and form:
{x ∈ A : Fx}.
Read: the set of those members x of A that satisfy the condition F.
The difference is crucial. We are no longer permitted to range freely over absolutely everything and collect into a set whatever satisfies an arbitrary condition. Set formation occurs within an already available domain.
Russell pursued a different strategy through the theory of types. The underlying intuition was that certain forms of self-reference arise because expressions belonging to different logical levels have been allowed to interact indiscriminately. Individuals occupy one level; predicates of individuals another; predicates of predicates another still. A predicate should not normally be allowed to take itself as an argument.
The slogan is crude but useful: Things of one logical type cannot simply be treated as things of every logical type. Russell’s solution therefore imposes hierarchy where unrestricted logical construction had permitted self-application.
These developments suggest a distinction that remains philosophically important: expressibility is not the same thing as admissible predication, and admissible predication is not the same thing as objecthood. The fact that we can describe something does not yet show that the description determines a genuine object. Logical grammar does not automatically settle ontology.
Why this matters for theology
The theological relevance is greater than it may first appear. Theology frequently speaks in terms that invite totalization: all truths, all possibilities, everything God knows, everything God can do, the totality of creation, or even everything that is not God. Russell’s paradox does not show that such theological expressions are illegitimate. It does show that theologians must distinguish carefully between quantifying over things and reifying the domain of quantification into another thing.
Suppose, for example, that we say:
For every x, if x is a creature, then God knows x.
Symbolically:
∀x (Cx → Kgx).
The symbols are simply an abbreviated way of saying: for every object x, if x is a creature, then God knows x.
Nothing in that assertion by itself requires there to be an additional object called the set of all creatures. The quantifier ranges over creatures; it need not package them into a single further entity. This becomes especially important when theologians speak of divine omniscience. One may say:
For every truth p, God knows p.
That claim does not automatically commit us to the existence of a further object called the set of all truths. We may quantify over truths without assuming that they together constitute one additional entity. Similarly, one may say:
For every creature x, x depends upon God.
Again, this does not by itself require an object called the totality of everything other than God. Quantification alone does not force reification. The lesson is methodological:
Quantification should not be confused with reification.
There is an even deeper theological resonance. Christian theology has often had to distinguish between what can be said of God and the ontological assumptions that our language appears to carry. Russell, Zermelo, and type theory remind us that grammatical or logical form can tempt us into constructing objects that our theory neither requires nor can consistently sustain.
The suspect inference has this form: We can specify what it is to be an F. Therefore, there exists one object consisting of all Fs.
The first statement does not by itself entail the second. To use the language of logic more carefully: Being able to say of each object whether it is F does not entail the existence of a set containing all and only the Fs. That is precisely the gap that Russell’s paradox forces us to notice.
The great foundational crisis thus yielded a surprisingly constructive philosophical lesson. Modern logic had acquired enormous power through Fregean quantification and Cantorian set theory. Russell’s paradox demonstrated that this power required discipline. Zermelo supplied axiomatic restrictions upon set formation; Russell supplied logical hierarchy through types. Both responses forced philosophers to distinguish more carefully between language, predication, collection, and existence.
For theology, that distinction is invaluable. Theology regularly attempts to speak about the ultimate, the universal, and the all-encompassing. Russell’s paradox teaches that whenever we speak about all, we should immediately ask another question: Have we merely quantified over everything in some domain, or have we quietly turned that domain into one more thing? That is a question worth asking whenever theology attempts to speak of God, creation, possibility, truth, or totality.
Bibliographical Note
The classic primary source for the paradox is Bertrand Russell’s 1902 letter to Gottlob Frege, together with Frege’s discussion in the appendix to the second volume of Grundgesetze der Arithmetik (1903). Russell developed the theory of types most fully with Alfred North Whitehead in Principia Mathematica (1910–1913). Ernst Zermelo’s “Untersuchungen über die Grundlagen der Mengenlehre I” (1908) provided the first major axiomatization of set theory designed in part to avoid the paradoxes of unrestricted set formation. For historical orientation, Jean van Heijenoort’s From Frege to Gödel: A Source Book in Mathematical Logic, 1879–1931 remains invaluable.