Abstract
When Wittgenstein took up philosophy again in 1929, he was convinced that the theories he had put forward in the Tractatus Logico-Philosophicus were not sustainable as they stood. He began to produce a vast series of texts which were to lead him within a few years to formulate his mature philosophy. It would be pointless to locate the formation of this mature philosophy in a specific text or a particular time. It was by no means a linear process and involved several intermediate stages. However, there is no doubt that when Wittgenstein dictated the Blue Book in 1933-34, he had gone well over half-way down the road that was to lead him to the Philosophical Investigations. Prior to the Blue Book we find a series of manuscripts and typescripts, which are only now beginning to be published and investigated in depth. If examined in chronological order, these texts undoubtedly show Wittgenstein's ideas moving further and further away from the Tractatus. The critiques and divergences are steadily sharper and more profound. It is one of these ruptures that I want to discuss in what follows.
Table of contents
When Wittgenstein took up philosophy again in 1929, he was convinced that the theories he had put forward in the Tractatus Logico-Philosophicus were not sustainable as they stood. He began to produce a vast series of texts which were to lead him within a few years to formulate his mature philosophy. It would be pointless to locate the formation of this mature philosophy in a specific text or a particular time. It was by no means a linear process and involved several intermediate stages. However, there is no doubt that when Wittgenstein dictated the Blue Book in 1933-34, he had gone well over half-way down the road that was to lead him to the Philosophical Investigations. Prior to the Blue Book we find a series of manuscripts and typescripts, which are only now beginning to be published and investigated in depth. If examined in chronological order, these texts undoubtedly show Wittgenstein's ideas moving further and further away from the Tractatus. The critiques and divergences are steadily sharper and more profound. It is one of these ruptures that I want to discuss in what follows.
In the Philosophical Remarks Wittgenstein acknowledges that language is essentially linked to time and the physical universe. Using a terminology that seems to derive from the later writings of Ramsey, he states: "Language itself belongs to the second system. If I describe a language, I describe something essentially physicalist... What we understand by the word 'language' unfolds in the homogeneous time of physics" (Wittgenstein 1989, 98). Let us try to grasp the meaning of these propositions more fully.
First, the use of the expression "physicalist language" in this context enables us to clear up a grave misunderstanding that frequently obstructs analysis of the texts Wittgenstein wrote during what is commonly referred to as his "verificationist phase". It would be a mistake to project into the Philosophical Remarks the meanings that the expressions "physicalist language" and "phenomenological language" have in contemporary writings by members of the Vienna Circle. By saying that "language is essentially physicalist", Wittgenstein does not mean that names refer (or should refer) to physical objects, much less that propositions are verified by events in the physical world as distinguished from phenomenical events. On the contrary. All the philosophy in the Philosophical Remarks is organised around the opposition between propositions as such, which describe phenomenical events, and "hypotheses", which cannot be verified and hence do not describe anything at all. For the Wittgenstein who concerns us here, "physicalist language" in the sense of language that refers to physical objects would simply be a contradiction in terms. Any description refers to phenomenological objects, since all descriptions are verified by the occurrence of phenomenical events.
What Wittgenstein means on this point is something quite different. Far from denying that language can refer to phenomena, Wittgenstein is saying that it is impossible to refer phenomenologically to phenomena. The target is always a phenomenon, but the arrow that points to it is essentially physical and temporal. The best way to understand this proposition is to analyse the argument that leads up to it. This argument takes the form of an imaginary experiment that has the force of a reductio ad absurdum. Wittgenstein tries to imagine languages from which all hypothetical elements have been eliminated. These imaginary languages are laid out in a series. With each step we take along the series we find a hypothetical residue which the next language unsuccessfully attempts to eliminate. At the end we have a language that is as close to phenomena as a language can be, but even here a physicalist residue is evidently present and it is clearly impossible to move another step towards phenomena without destroying the representative nature of this language. Let us see how this is done.
The starting-point is the possibility of constructing a "phenomenical autobiography": in other words, an exhaustive account of each of the phenomenical experiences that make up my life. For this purpose I imagine that I have been endowed with a prodigious memory and elaborate my descriptions in a language that makes direct references to phenomena. To underline the "directness" of these references, Wittgenstein limits himself to the representation of events in my visual field, imagining reference by means of partially painted plaster models. Only the portion effectively seen is painted, carefully bringing out what is blurred in everyday language, i.e. the experience we have effectively had, in opposition to the horizon of the infinite possible experiences corresponding to the idea of a "physical object".
To the description of this first "imaginary language" Wittgenstein adds his first objection: "So far, so good" he says. "But what can be said of the time I take to arrive at these representations?" (Wittgenstein 1989, 97). The objection apparently refers to the practical possibility of making such "records". It seems reasonable to suppose that even a thousand years would not be enough to construct representations capable of accounting for all the phenomena present in a single day of our lives. However, this is an "empirical" limitation, a practical problem that has no relevance to the establishment of a logical possibility. The real problem lies elsewhere. To facilitate the discussion, suppose my life lasts exactly two days. On the first day I have a number of experiences, which I store accurately in my memory. On the second day I set about recalling each of the experiences I had on the previous day and recording these experiences using my plaster models. I remember a visual experience and immediately afterwards I reproduce the scene in a plaster model. The process clearly involves three distinct phases: the original experience, remembering it, and constructing the model.
It is important to make a careful distinction between a visual experience and the memory of it. The purpose of the plaster model, of course, is to be a faithful reproduction of the latter, which we could call a "mnemonic experience", and not a direct reproduction of past visual experience. The relationship between memory and past experience is evidently hypothetical and presupposes the availability of physical time. At the level of assertions about the physical world we can enumerate criteria that must be met in order for a remembrance to be accurate. None of these enumerations will be exhaustive. It will always be possible to make sceptical objections to the accuracy of memory. However, Wittgenstein decides to limit the problem to the very last link in the chain: the relations between the plaster representation and the mnemonic experience. He leaves aside the question of the accuracy of memory, which in any event does not arise at the level of phenomena. In phenomena, the past is merely an internal property of experiences necessarily confined to a continuous present. From this standpoint the question is simply: to what extent can I reproduce immediately what my memory presents to me immediately? How far can language reproduce immediately the experience of memory? The time employed in constructing representations, however short, evidently becomes relevant. If remembering occurs at one moment and the elaboration of a model takes place later, then it is possible to raise the question asked in regard to the physical world about the relationship between memory and past experience. The plaster representation of memory becomes hypothetical, since it is possible to imagine sceptical objections to the accuracy of this representation. To what extent have I reproduced faithfully in the model which was given to me immediately in memory?
Wittgenstein answers this objection by imagining a second language that adheres to the phenomenon more closely than the first. Suppose, he says, that I am "capable of 'writing' - constructing representations - at the speed of my memories" (Wittgenstein 1989, 97). In other words, I write at the same time as I recall so that the accuracy of my writing is guaranteed by the mere presence in the phenomenical field of the memories I describe. As always, we may doubt the accuracy of memory when we project our experience in the physical world. That is a different matter. The important point here is that it should be logically impossible to doubt the fidelity of the representation of memory, which is a present experience. By adding this "real-time accompaniment" to our requirements, we seem to have achieved the goal of immediate representation. Or have we? Our requirements have been met only with regard to the moment at which the representation is being made. Let us now suppose, says Wittgenstein, "that I re-read the entire description. Would it not be hypothetical this time?" (Wittgenstein 1989, 97).
Indeed, who can be sure that this plaster figure now before me is faithful to the mnemonic experience I had while making it? It is no use saying in the present that the making of the representation was concurrent with the experience. While I was making the representation, the simultaneity of the mnemonic experience guaranteed identity of meaning. Now, however, all I have is the plaster figure before me. The plaster figure affirms the occurrence of a past experience but there is no way in which I can verify immediately whether this experience really occurred in the past. At most, I have the memory of a past memory while I read. But it is to the first memory that the plaster figure is intended to refer and not to the second.
In developing this argument, Wittgenstein takes to an extreme his attempt to imagine what an immediate representation of the past might be. He imagines a mechanism and a "language" associated with it. Here is his description of this imaginary experiment:
Let us imagine the following representation. The bodies I apparently see are moved by a mechanism in such a way as to give two eyes fixed in a certain place in the model the visual figure to be represented. The visual figure described is therefore determined by the location of the eye in the model and by the location and movement of the bodies. One could conceive, for example, of moving the mechanism by turning a handle and thus 'parroting' the description." (Wittgenstein 1989, 97).
It seems evident that this imaginary mechanism is designed to externalise the perceptive situation, creating the possibility of supposedly objective control over representation. The experience to be represented is the experience captured by the eye fixed at the centre of the mechanism. However, another eye is assumed to exist in the experiment. This is an eye taking notes of the coordinates corresponding to the fixed eye, the objects, and the movements of the machine. A handle enables a person "outside" the mechanism to reproduce the movements performed. This person uses the notes on the coordinates like a musical score. From the standpoint of this external person, the text is "read" mechanically ? it is "parroted", as Wittgenstein says. The sensations this external person has as he turns the handle have no necessary relation to the meaning of the signs on the score. We might imagine that the person does not even see the objects moving as he turns the handle. The objects and their movements are seen by the eye placed at the centre of the machine. From our own standpoint, what the eye now sees is an accurate representation of what it saw before. Our aim is to ensure that this eye does not perceive plaster figures, much less mathematical descriptions of the movements of bodies. We want to bring about a situation in which the representation of past visual figures is produced by present visual representations. Phenomena representing phenomena in the most immediate way possible.
It is obvious that the mechanism imagined here would give the images produced a hypothetical nature. For us on the outside, turning the handle, it would be impossible to verify the absolute accuracy of the representation. Everything would depend on the fidelity of the transcriptions made and the functioning of the machine. The criteria at our disposal for this purpose are necessarily fallible and provisional. If we want to advance still further toward immediacy we will have to assume the viewpoint of the immovable eye at the centre of the mechanism, watching the scenes unfold. In the world of this eye there are neither machines, nor handles nor transcriptions. There is just the recurrent scene, which is eternally present. In itself the scene is not a "reproduction of the past" or indeed a reproduction of anything. It is a continuous flow of necessarily differing sense impressions. For the representation to become even more immediate, we must put ourselves in the position of the fixed eye. When we assume this position, however, the representation is dissolved. What remains is the flow of life and nothing else.
This is what leads Wittgenstein to say that language belongs to the "second system". To describe a language, he says, is "to describe something essentially physicalist". Not that language cannot refer to phenomena. On the contrary. From the perspective adopted by Wittgenstein in 1929, there is nothing else to which language can refer. It is phenomena that verify the descriptions we produce; hence it is only about phenomena that we can speak. But this reference is always made in a "nonphenomenical" manner - that is, by means of an essentially precarious relationship between the linguistic sign and the expected phenomenon. This creates an evident tension between the two requirements, which if not explicitly contradictory seems at least to pull in opposite directions. It is not unreasonable to presume that this tension played an important role in leading Wittgenstein to abandon the phenomenological project that characterised his work in the early 1930s.
References
- Wittgenstein, L. (1989) Philosophische Bemerkungen. Frankfurt am Main: Suhrkamp.
Refbacks
- There are currently no refbacks.