The Transcendental Subject: Three Good Tractarian Reasons to Accept It
The Transcendental Subject: Three Good Tractarian Reasons to Accept It


The transcendental subject has a logical function. It is "transcendental" in the two philosophical understandings of the term. It conditions the possibility of meaning, being outside the reach of it. In the order of reasons, its ethical dimension comes as a kind of side-effect. Given its semantical functions, it must be the bearer of a non-psychological kind of will. Accordingly, it must be the subject of an absolute ethical stand.

Table of contents

    The most common approach to the mystical doctrine of the Tractatus has been to trace it back to this or that author in the history of philosophy. Wittgenstein was certainly not the first philosopher to talk about the existence of a "transcendental subject". But this can be a good place to apply a mild version of the so-called "principle of charity". If strong evidences are not produced, we should not take Wittgenstein for a plagiary. He probably did something more interesting at the final remarks of the Tractatus than just transcribing a curious piece of someone else's philosophy. It is just fine to look for similarities between his philosophy and any other, but first of all we must have good reasons to use the possessive pronoun in this context. That's what I will try to achieve in the minimalist way imposed by time constraints. I will try to show that, from the Tractarian viewpoint, the transcendental subject is a logical necessity. I think it is "transcendental" in the two philosophical acceptions of the term. It conditions the possibility of meaning, being outside the reach of it. Moreover, it has a kind of side-effect. Given its semantical functions, it must be the bearer of a non-psychological kind of will. Accordingly, it must be the subject of an absolute ethical stand. Let us see how.

    The word "intentionality" does not appear in the Tractatus, but the idea is pervasive. Wittgenstein chose a somewhat misleading expression to nickname it – "abbildende Beziehung", or "depicting relation". This relation is formally defined as "the coordinations of the elements of the picture and the things" (2.1514), and more pictorially described as "the feelers of its elements with which the picture touches reality" (2.1515). We may think of it as the relation through which a logically proper name acquires its meaning by being assigned to a logically simple object.

    The relation must be conceived as an asymmetrical one. The name must name the object, and not the opposite. The relation must "go" to the object, as the feeler of an insect goes to the things it tests by touch. But that is not enough. To qualify as an authentic naming relation, this feeler must be outside the world. Being a necesary condition of any description, it cannot be described. It must be what the Tractatus calls an "internal relation". For suppose that this object is taken as the name of that one. If you try to say that this names that, you will be forced to use the name you are trying to speak about, and your proposition will have to be true in order to have a meaning. But it is obvious that in the logical order of precedence meaning must come first. It is possible to say what colour an object has, but not which name was assigned to him. This is something that language cannot say, but only show. Strange as it may seem, the naming relation cannot have a name. It cannot be the object of a meaningful discourse.

    But the most interesting mark of this relation is yet to be mentioned. Besides being asymmetrical and ineffable, it must be instituted, established. It is not "ready-made", it is not "engraved" in the objects forming the facts we use as pictures of other facts. Nothing is by its own nature the name of anything. It becomes so. It must be made so. For imagine that a and b are two objects belonging to the same logical category. Would language lose something if we exchaged their names throughout? Is it not possible to use another tool taken from the same categorial box? The relation between name and object is internal, as we saw some moments ago. But at the same time it must be established. It is not "resident" in the object that is being used as a name. The relation must be established, so to speak, from outside on. Differently said, no object has a projective relation inscribed in it. No object is a logical feeler "by its own nature". It must be taken as a name, associated to an object, projected onto reality. It has an internal relation to an object, it is true, but this relation is supervenient. Not supervenient in time, for sure, since what must "happen" cannot be a fact – it cannot have any place inside the world probed by the feelers of our thought.

    This is the "mystery of meaning". Naming is akin to transubstancialization. When the object becomes a name of something, it does not suffer any change in its "external properties", but acquires a new internal one. An action is wanted here. An ineffable action performed outside the bounds of sense – a trascendental intervention of a transcendental subject. When he introduced this element at the end of the Tractatus, Wittgenstein was not mesmerized by this or that chapter of the history of philosophy. Without the transcendental subject, the whole logical engine of the Tractatus simply does not work.

    But naming is not the only place where the services of this subject are needed. Logical operations want it as well. According to the Tractatus, language needs two kinds of logical operations at work. Each of them will give us a new reason to postulate a transcendental subject acting outside the world. First of all, we have the truth-operations used to build new propositions out of simpler ones. He takes simultaneous negation as a paradigm, but a closer look at simple negation will be enough to make my point. According to Wittgenstein, an operation is "that which must happen [geschehen] to a proposition in order to make another out of it" (5.23). In order to be logical, an operation must be grounded on a formal relation between basis and result. There must be a definite logical relation between the sense of the proposition used as a basis, and the sense of the proposition obtained as a result. In writing "p" or "not-p", our hand is dealing with signs, with marks of ink. This action can be perfectly described in our language. But negating a proposition is not simply drawing graphic patterns on a sheet of paper. We only deny a proposition by reversing its sense – that is what negation "does". And the reversal of sense is not a fact at all. It is not comparable to turning my socks inside out. Again, it is something that must be done, but that cannot be done inside the logical walls of the world. It must performed by a transcendental agent dwelling into the bounds of sense.

    There is another kind of logical operation which is not used to build, but to select. Direct enumeration is the simplest selection method. We use it, for instance, when we want to construct a logical conjunction. Once more, what is "given" as an input to the conjunction is not a pair of graphic entities, but two logically determined senses. What must be "gathered" cannot be described, since it is an internal property of a pair of symbols. Accordingly, we do not have here a factual "gathering", but a transcendental one.

    The same is true of more complex gatherings, involving not two, but an infinite number of propositional senses. Quantification is the first and more salient example. In order to say that there is someone in this room, for instance, I have to consider all statements having the form "x is not a person in this room", and negate them all. What was true of two propositions keeps being true of this infinite set captured by a propositional function. What is captured is neither an infinity of real scribbles, nor an infinity of possible ones. We must gather an infinity of logically related senses, and that is something that an empirical subject could never do, even if he had an infinite span of time.

    Finally, we have the formal series, where numbers make their entrance. I can say that there is an infinite number of persons in this room by denying that there is no one, that there is exactly one, exactly two, three, four, and so on, to infinity. The Tractatus has no problem at all with infinities, provided they are formally determined. From Wittgenstein's point of view, the problem with the axiom of infinity is not the infinite, but the attempt to give the number of objects in a proposition, instead of showing it as the end result of logical analysis. By the same token, the problem with the axiom of choice would not be the mention of an object selected within an infinite set, but the arbitrariness of the selection, the fact that it is not governed by a logical rule. For Wittgenstein, logic does not play with dice. Anyway, who makes the selection, be it a finite or an infinite one, is always a subject which cannot be identified with my body, my bodily motions, mental states, bodily or mental dispositions, and what not. There is an "I" behind the sense, but it can never be identical with anything which is true about myself.

    So the transcendental subject has a logical reason to exist. More exactly, there are at least three good logical reasons for it to be postulated. Without counting with such a subject, Wittgenstein could never have spoken of names, logical operations and logical selections the way he did. In all these cases a metaphysical agent must be presupposed behind the logical tasks which must be performed. Although Wittgenstein does not speak in these terms, it is quite apropriate to think of this subject as a pure intentionality projecting names into the world, reversing the intention of propositions, and directing a non-empirical attention to certain formally defined groups of propositional sense. In more abstract terms, it is a non-empirical will – a will to mean. It is a will directed not to facts, as empirical will, but to the a-temporal world of objects. It is not a will of some contingent fact, of something that may happen or not. It is a will of what necessarily is – the a-temporal order of objects and possible states of affairs.

    This mystical contact with eternity is inevitably present where any meaning or understanding is involved, but only in a sort of "instrumental" way. When we speak we have to direct our transcendental intentionality to the realm of objects for the sake of producing propositional senses. I think Wittgenstein envisaged another possibility at this point. The Tractarian doctrine of sense lefts open the possibility for this transcendental subject to project its intentional feelers onto eternity, not for the sake of speaking meaningfully, but for its own sake. When we speak, we say how we think the world is – which objects are combined in which way. In order to form propositions, we have to project names over the realm of objects. Objects carry with them their combinatorial possibilities. To intend objects is to intend all the net of logical possibilities inscribed in the "logical space". So there is no talk about how the world is without a passage through the sempiternal order of the logical space. That is the very locus of sense. We can choose between an instrumental passage at the service of a chattering distracted reason, and the direct intuition promised by the mystical tradition. The old mystical rule of silence finds in the Tractatus its most perfect completion. The Tractarian silence is an immediate and non-instrumental contact with the a-temporal order from which every sense must come from.

    Joao Vergilio Gallerani Cuter. Date: XML TEI markup by WAB (Rune J. Falch, Heinz W. Krüger, Alois Pichler, Deirdre C.P. Smith) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.