Solipsism and the Self in Wittgenstein's Tractatus
This paper addresses Ludwig Wittgenstein's discussion of solipsism in the Tractatus Logico-Philosophicus. It demonstrates that, throughout his early and middle periods, Wittgenstein's work on solipsism was focused on traditional solipsism, not semantic solipsism as is standardly supposed. It furthermore argues that the Tractatus's stated support for solipsism should be taken as a straightforward endorsement of the doctrine. In establishing this, it analyzes the connection between Wittgenstein's thought on solipsism and the metaphysical subject through the subjective-objective and subject-object distinctions, arguing that this holds the key to his comments on solipsism in the Tractatus.
Wittgenstein, solipsism, subjectivity, metaphysical subject, the self
This paper aims to settle the question of what Ludwig Wittgenstein meant by "what solipsism means":
This remark provides the key to the problem, how much truth there is in solipsism.
For what solipsism means is quite correct; only it cannot be said, but makes itself manifest.
The world is my world: this is manifest in the fact that the limits of language (of that language which alone I understand) mean the limits of my world. (TLP 5.62)1
Throughout this paper, I will paraphrase "what solipsism means" simply as 'solipsism.'2 The aim of the present paper can then be restated as an attempt to specify the conception of solipsism Wittgenstein was addressing in the Tractatus [End Page 127] Logico-Philosophicus. A secondary aim is to specify the extent to which we should take Wittgenstein as having considered solipsism to be true.
These aims are narrow; but that is because this paper really intends to settle these issues. Both of them arise due to the fact that the Tractatus itself does not specify what it means by 'solipsism.' In its most general sense, solipsism is the doctrine that reality is in some sense limited or restricted in virtue of its unique relation to oneself. The nearest the Tractatus comes to defining 'solipsism' is its statement, "The world is my world" (TLP 5.62). The difficulty this presents the reader is that the book contains two conceptions of the world—the world as the totality of possible states of affairs (TLP 5.6), and the world as the totality of obtaining states of affairs, or the totality of facts (TLP 1–1.21, 2.04)—each of which corresponds to a different version of solipsism. The first version, sometimes called traditional solipsism, corresponds to the world as the totality of facts. Traditional solipsism can be defined as the doctrine that reality, or everything that actually exists at any given time, exists only through and in virtue of its relation to the solipsist. Most commonly, this reality-constituting relation is taken as experience. The traditional solipsist says that only their own experience is real, or that their present experience exhaustively constitutes reality. It is this kind of solipsism that, I will argue, is addressed in Wittgenstein's discussion of solipsism.3 The second version, semantic or linguistic solipsism, corresponds to the world as possible states of affairs.4 Semantic solipsism focuses on representation; in particular, on the idea that all representation of the world—everything that can be conceived or described in language or thought as being or possibly being—is in some way dependent on the solipsist, and a unique relation to language the solipsist claims to possess.5 The semantic solipsist presents their own use and understanding of language as constituting language per se, with the upshot that the world, as it is or can be given in language, therefore is the world only through and for the solipsist.
It is important to see that the key point of difference between these kinds of solipsism is that each relates to a different conception of the world: linguistic solipsism to the world-as-possibilities, and traditional solipsism to the world-as-facts. In asking which the text addresses, the use of labels like 'linguistic solipsism' to designate the particular way of conceiving of solipsism ascribed to Wittgenstein [End Page 128] can present its own difficulties, insofar as the expression is also open to being taken as a kind of general descriptor, broadly applicable to any solipsistic thesis that grounds its position by reference to the nature of language, or which posits solipsism as properly classified, as it were, as a purely linguistic or grammatical phenomenon. As such, it encourages conflation of the considerations from which Wittgenstein's endorsement of solipsism arises with the thesis he was actually calling 'solipsism.' Obviously, given the book's subject-matter, whichever solipsistic idea we might take the Tractatus to be endorsing, the grounds for its endorsement can be related to linguistic considerations. This being the case, however, in itself tells us nothing about what that idea of solipsism was; in particular, whether it was a form of semantic or traditional solipsism. This is the question we are interested in. In this sense, semantic and traditional solipsism are differentiated, not primarily by how each draws its solipsistic restriction—whether it involves reference to language or not—but by what each claims is being restricted: possibilities (the meaningfulness of propositions) or facts (the truth of propositions).6
Wittgenstein's endorsement of the truth of solipsism is commonly treated as being intended to relate to some form of semantic solipsism.7 This view can then prompt the question of whether or how far Wittgenstein really meant to endorse solipsism. Obviously, this is because semantic solipsism is not solipsism, at least as philosophy has traditionally understood it. So, if Wittgenstein judged "how much truth there is in solipsism" as being just that there is some truth in semantic solipsism, this could be taken to imply he also judged there to be untruth in traditional solipsism. Thus, commentators who interpret the positive assessment in 5.62 as relating to semantic solipsism tend either to portray it as one step in an overall subversion of solipsism (traditionally understood) or to describe the aspect of solipsism Wittgenstein was assessing positively as something so weak or distantly removed from the doctrine as to barely be discernible as solipsism at all. Each line of interpretation, however, must then also produce an explanation either for why Wittgenstein would say solipsism is correct when he is really rejecting it, or why he would call his doctrine solipsism when that is not what it really is.
This paper argues for a more straightforward reading of the 5.6s.8 It proposes that the argument Wittgenstein makes in its opening sections (TLP 5.6–5.621) should be seen as addressing a traditional solipsism—with its conclusion, "The world is my world," stating a traditionally solipsistic restriction of reality to Wittgenstein's experience—and as making an unqualified endorsement of it—by which I mean it intends to endorse (and not less than endorse) traditional solipsism (and not less than traditional solipsism). That is, it means we should consider the [End Page 129] Tractatus to have endorsed a traditional solipsism just as it endorsed (for example) a distinction between saying and showing.9
At core, my argument is simple: It is that throughout his early and middle periods, Wittgenstein's writings on solipsism addressed traditional solipsism; that this being the case, there is insufficient reason for us (as readers) not to take 5.62 as addressing traditional solipsism; and that, moreover, there are compelling reasons for why we should take it this way.10 It begins by demonstrating why we should expect the 5.6s to engage with traditional solipsism. It should not be overly contentious to suggest that, on its own, Wittgenstein's discussion of solipsism in the Tractatus is too brief, and too opaque, for us to think we could make an unequivocal interpretative decision regarding precisely how he conceived of solipsism based on that text alone. For this reason, the best way to recognize the kind of solipsism the Tractatus was concerned with is by first looking outside the Tractatus, that is, at Wittgenstein's pre-Tractatus notebooks and immediately post-Tractatus writings, which discuss solipsism more forthrightly, and at much greater length. Sections 1 and 2 address these periods, starting with Wittgenstein's middle period work on solipsism.11 These sections aim to illustrate that, both pre- and post-Tractatus, [End Page 130] Wittgenstein's engagements with solipsism centered on traditional conceptions of the doctrine, and to delineate the clear continuities in his thought on solipsism across his writings. In doing so, they offer an informed framework for what we should expect to find within the Tractatus.
Section 3 proceeds to address the Tractatus directly. Having established that the 5.6s should be expected to engage with traditional solipsism, and with 5.62 seemingly endorsing the doctrine, I argue that reading the Tractatus as other than endorsing traditional solipsism must therefore require justification. It proposes three reasons that might be considered to provide this, with the remaining sections looking at each in turn. Section 4 addresses the so-called "resolute" readings of the Tractatus. Section 5 discusses Wittgenstein's argument in 5.6–5.63. Section 6 concludes by defending this reading in relation to his wider discussion of solipsism and the self in the 5.6s. In doing so, I show not only that an endorsement of traditional solipsism is consistent with that discussion, but, further, that it is implied by Wittgenstein's comments on the self.
1. wittgenstein's middle period writings on solipsism
In looking at Wittgenstein's post-1929 work, it is important to distinguish the kind of solipsism under analysis from the analysis being given. The solipsism under analysis was unmistakably traditional. This is evident in the characterizations of solipsism Wittgenstein gave during this period. Wittgenstein's solipsist is a person who "says that only his experiences are real" (BB 58; AWL 22), that "only I feel real pain," and "only I really see (or hear)" (BB 60). He similarly connected solipsism with such statements as "I can only know that I have personal experiences, not that anyone else has" (BB 48), and wrote, "Sometimes, the most satisfying expression of our solipsism seems to be this: 'When anything is seen (really seen) it is always I who see it'" (BB 61). G. E. Moore, who attended Wittgenstein's lectures during this period, recalls:
As regards Solipsism and Idealism he said that he himself had been often tempted to say "'All that is real is the experience of the present moment" or "'All that is certain is the experience of the present moment"; and that anyone who is at all tempted to hold Idealism or Solipsism knows the temptation to say "The only reality is the present experience" or "The only reality is my present experience."
Wittgenstein's analysis of such statements, on the other hand, focused on their linguistic function. Rather than metaphysical descriptions, he argued they should be seen as contentions about grammar (BB 57, 59; AWL 25). This should not, however, be confused with a focus on semantic solipsism: Wittgenstein gave a linguistic analysis of traditional solipsism, not an analysis of linguistic solipsism.
That analysis is too large to be completely recounted here; in any case, it is not necessary. We can gain detail by focusing on a single part of the analysis; specifically, Wittgenstein's observation that the solipsist's goal of having reality identified uniquely with their own experience is achievable, though only in a grammatical sense, through making certain modifications to the rules of grammar (BB 63). Looking at these modifications, and how Wittgenstein thought them capable of [End Page 131] satisfying the solipsist, provides a clear idea of what he took such statements as "Only my experiences are real" to be trying to say.
According to Wittgenstein, while appearing to claim "monopoly" over experience, the solipsist's utterance "Only my experiences are real" does not exclude, but rather implies, other owners of experience, "my" and "real" only having sense if their opposition is not logically excluded (AWL 22). What the solipsist therefore needs "is not a notation in which the ego has monopoly over experience, but one in which the ego vanishes" (AWL 22). The notation Wittgenstein suggests involves two grammatical changes: First, that the solipsist (and the solipsist alone) not be required to use personal pronouns in describing their experience—indeed, they are precluded from doing so; and second, that application of the epithet 'real' be restricted to what would ordinarily be called the solipsist's experiences (BB 59). It can be "tempting" to adopt this notation, Wittgenstein says, "because the description of a sensation does not contain a reference to either a person or a sense organ":
The truth of the proposition, "The noise is approaching my right ear," does not require the existence of a physical ear; it is a description of an auditory experience, the experience being logically independent of the existence of my ears. The audible phenomenon is in an auditory space, and the subject who hears has nothing to do with the human body.(AWL 22–23)
The temptation to adopt the solipsist's notation is therefore its reflecting the grammatical insight that propositions ascribing self-experience need not refer to oneself as the person having the experience (or, more generally, that 'I' does not refer to a possessor in sentences detailing experience [AWL 21]).
How does this relate to solipsism? What Wittgenstein appears to have had in mind is that, by precluding the solipsist from using pronouns in describing their experience, their notation makes no formal distinction between sentences describing what he called "material" facts—what we generally consider objective happenings in the "external" world—and those describing subjective "personal experiences," such as pain, auditory experience etc. (BB 46). Thus, "Instead of saying 'I think' or 'I have an ache,' one might say 'It thinks' (like 'It rains')" (AWL 21). Transposed onto the world, their notation therefore does not allow for differentiating a thing's having reality only for the solipsist from its being a subject-independent fact. At the same time, this equivalency with fact attaches only to the solipsist's experience, since only in their notation are the two presented the same. Restricting "real" to the solipsist's experience furthermore excludes such statements as "Jones has a real toothache" or "Jones really sees" (where Jones is not the solipsist) meaning any account of reality—of everything really seen, really heard etc.—could only include the solipsist's experiences (BB 59).12 [End Page 132]
The solipsist's notation therefore appears designed to appease someone who wishes both to restrict the having of experience exclusively to themselves, and the having of reality exclusively to what is their experience; or who, in other words, wishes to identify their experience as reality. This is plainly a traditional solipsism. At the same time, it is significant that Wittgenstein should have considered the solipsist to want a notation "in which the ego vanishes" (AWL 22). In this, his post-Tractatus writings partake in a movement of thought that features in all three periods this paper addresses, wherein solipsism is linked with an analysis of the "I" (or what Wittgenstein variously called the "self" or "metaphysical subject") that involves the "I" being in some sense nullified. In the middle period work, we see this in Wittgenstein's repeated, inconclusive attempts to specify identifying criteria for the I, and the exclusion of "I" from the solipsist's notation (with its broader analysis of "I" not being a referring term); in the "psycho-physical parallelism" of the pre-Tractatus notebooks; and in the expulsion of the I to the "limit" of the world in the Tractatus. I discuss each in detail in this paper. It is worth looking here, however, at why this connection should recur in Wittgenstein's thought, as it offers a valuable tool for reading the notebooks and Tractatus.
The topical progression, from solipsism to the subject, that recurs in Wittgenstein's writings is often presented as essentially accidental: A case of solipsism—with its contention that reality is uniquely mine—leading naturally to the question of what "I" am here. While there is some truth in this, two further factors require consideration. First, while solipsism does raise the question of the nature of the subject of experience, it equally raises the question of the nature of the object of experience. If it "naturally" suggests investigating anything, it is therefore the subject in connection with its object, and vice versa.13 This is important for recognizing how Wittgenstein's thought on solipsism and the subject each respectively sets parameters for interpreting the other. Second, as mentioned above, Wittgenstein not only proceeds from solipsism to discuss the subject, but to discussions that in each case—and in different ways—reject the idea of the subject as an individuated entity. We should therefore prefer an explanation of why he connects these topics to go some way toward accounting for this aspect in itself.
Taking these points into consideration, I believe the best way to understand Wittgenstein's thought here is by looking at solipsism as essentially a rejection of a subjective-objective distinction—a distinction between the solipsist's experience and objective reality—with his comments on the "I" reflecting the negation of the subject-object distinction that follows as a consequence. These two distinctions, in other words, stand and fall together. The connection is that the subject cannot be individuated as a reality distinct from its object if that object is itself denied the subject-independent reality accorded under a subjective-objective distinction.
If they wish to individuate themselves as subject, the solipsist's most promising option appears to be to identify themselves with their subjectivity as such. The subject can neither be identified empirically, in or as a part of experience (by definition, an object of experience cannot be the subject of that experience), [End Page 133] nor through juxtaposition with the independent reality of objects—by arguing that as things exist independent of me, I must exist independent of them—since under solipsism, objects have no subject-independent reality (cf. PR 100). The solipsist might instead try appealing to the continuity of their experience, beyond experience of any particular object, as showing their independent reality. As Wittgenstein saw, this amounts to the solipsist defining themselves as their subjectivity, or as the having of their experience:
Now let us ask ourselves what sort of identity of personality it is we are referring to when we say "when anything is seen, it is always I who see." What is it I want all these cases of seeing to have in common? . . . When I think about it a little longer I see that what I wished to say was: 'Always when anything is seen, something is seen.' I.e., that of which I said it continued during all the experiences of seeing was not any particular entity "I," but the experience of seeing itself.(BB 63)
In having rejected objectivity, however, the reality of objects has already been defined as the solipsist's experience of them. By then also defining themselves as their experience, the solipsist would therefore be identifying their own reality and that of the objects of their experience as one and the same thing:
When I made my solipsistic statement, I pointed, but I robbed the pointing of its sense by inseparably connecting that which points and that to which it points. I constructed a clock with all its wheels, etc., and in the end fastened the dial to the pointer and made it go around with it. And in this way the solipsist's "Only this is really seen" reminds us of a tautology.(BB 71)
Wittgenstein thus diagnosed the solipsist as being led to define their reality as subject as indistinct from the reality of the objects of their experience, just as (indeed, because) they define their (subjective) experience as constituting (objective) reality. Solipsism is therefore linked with the inability to identify the I, not just because the solipsist is unable to individuate their I as an object of experience, but more importantly, because lacking a subjective-objective distinction, they are unable to individuate their I from the objects of experience.
These connections therefore help explain Wittgenstein's progression from solipsism to the impossibility of individuating the subject. At the same time, the fact that they are formally connected means that his discussions of the I provide a means for identifying the kind of solipsism he was addressing. Because solipsism prompts negation of the subject-object distinction through its denial of objective reality, and because the subject Wittgenstein discusses in connection with solipsism is the subject of that solipsism (the solipsist), this means the object he discusses in relation to the solipsist should be the reality their solipsism is denying objectivity to. By identifying the kind of object Wittgenstein discusses in relation with the subject, we can therefore identify the solipsism of which it is the subject.
It will help to describe this connection in terms of the two kinds of solipsism this paper is looking at. The semantic solipsist identifies the world with their world through rejecting a subjective-objective distinction within language: It argues language is the solipsist's language; and because language is in some sense the world, the world is the solipsist's world. If Wittgenstein were discussing this kind of solipsism, then the subject and object we should expect him to connect are the subject as linguistic representer and the world as representable in language, [End Page 134] respectively. Traditional solipsism, on the other hand, denies a distinction between subjective experience and objective reality. If Wittgenstein were discussing this solipsism, the subject and object should be the subject of experience and the world of facts as given in experience, respectively.
The important thing is that we are able to work backwards from the kind of subject and object, to the kind of solipsism. As an interpretative tool, this is not so crucial for reading Wittgenstein's post-Tractatus work—as we have seen, it straightforwardly defines the traditional solipsism it addresses. It is more useful, however, for understanding his pre-Tractatus notebooks. While containing few remarks directly on solipsism (and like the Tractatus, no definitions) they discuss the "I of solipsism" at length. We can therefore use these discussions to specify the solipsism Wittgenstein had in mind, which the text otherwise leaves largely underdetermined for us.
2. wittgenstein's pre-tractatus notebooks
Wittgenstein's Notebooks 1914–1916 show remarkable affinity with his immediately post-Tractatus thought regarding solipsism and the nature of the I: Both track the same progression from solipsism to nullification of the subject, to identification of the subject with its objects of experience. Yet equally remarkable is their divergence: For whereas the later work simply analyzed this line of thinking, the early Wittgenstein embraced it, developing his recognition of the non-appearance of the I in experience into a Schopenhauerian "psycho-physical parallelism" (NB 85).
Rather than as the subject of experience or linguistic representor, in his notebooks Wittgenstein introduces the I as a necessary ethical notion (NB 79). It is in trying to explain the nature of this ethical subject that focus shifts to the I as the subject of experience:
The I, the I is what is deeply mysterious!
The I is not an object.
I objectively confront every object. But not the I.
So there really is a sense in which there can and must be mention of the I in a non-psychological sense in philosophy.
The I makes its appearance in philosophy through the world's being my world.(NB 80)
This passage is worth considering because it shows the genesis of "The world is my world," the closest thing to a definition of solipsism offered in the Tractatus. Wittgenstein's immediate point is that, while he does not experience the I as an object, its reality is secured through his experience of objects. At the same time, he portrays his experience of objects as likewise giving them reality (he objectively confronts every object). This suggests his reference to "my world" should then be read as designating his experience (of objects, or just experience per se). In then calling this "the world," Wittgenstein thus seems to be suggesting that his experience of the world constitutes it. [End Page 135]
The passage therefore ties the reality of the I to the world (the world as facts and objects), and vice versa, through experience. And it is this I in its relation to the world that remains Wittgenstein's focus through the remainder of his notebooks. From the outset, his insights into the I of solipsism are informed by a solipsistic elevation of his own experience to the preeminent place in determining his conception of the world:
The I of solipsism shrinks to an extensionless point and what remains is the reality co-ordinate with it.
What has history to do with me? Mine is the first and only world!
I want to report how I found the world.
What others have told me about the world is a very small and incidental part of my experience of the world.
I have to judge the world, to measure things.(NB 82)
The immediate consequence of this devaluation of objectivity is that all reality, now identified with Wittgenstein's experience, comes to be placed on the same metaphysical level in its relation to himself:
The philosophical I is not the human being, not the human body or the human soul with the psychological properties, but the metaphysical subject, the boundary (not a part) of the world. The human body, however, my body in particular, is a part of the world among others, among animals, plants, stones etc., etc.(NB 82)
As seen above, this then leaves Wittgenstein nothing to identify his subject with except his subjectivity; or, what comes to the same without a subjective-objective distinction, the object of his experience. His investigations into the I culminate in a description of what he labels the "parallelism" that "really exists between my spirit, i.e. spirit, and the world" (NB 85). His thinking here is certainly inspired by his early readings of Schopenhauer. Schopenhauer argued that persons exist in a dual nature, with an objective aspect—their human body—and a subjective aspect, characterized as will.14 He suggested that, because we can recognize ourselves as simultaneously subject and object, and since we experience the objective aspect of the rest of the world, we can justifiably infer that the world too has a corresponding subjective aspect.15 Moreover, because all experiences of the objective aspect are of one kind—in Schopenhauer's idiom, it is always experienced as representation—so too its subjective aspect must be of one kind.16 Thus, Schopenhauer concluded that the world is pervaded by a single underlying will which objectively manifests itself as the phenomenal world. In this sense, the individual person can be considered a microcosm, embodying on a smaller scale the nature of the world itself (the macrocosm).17 [End Page 136]
While undoubtedly influenced by Schopenhauer's dual-aspect conception of the world, Wittgenstein rejected the suggestion that his relation to his body could serve as the basis for inferring universal subjectivity. Commitment to the metaphysical equality of facts entails that the human body—even his own—cannot be privileged with any "preeminent place" among the facts of the world (NB 82; see also NB 84). Answering his own question regarding why he thinks of snakes and wasps as having a certain "character," Wittgenstein initially suggests he has derived it from observation of their physical appearance and behavior.
But the question arises whether even here, my body is not on the same level with that of the wasp and the snake (and surely it is so), so that I have neither inferred from that of the wasp to mine nor mine to that of the wasp.(NB 85)
Contra Schopenhauer, because all facts are on a metaphysical par, one can no more use one's body and the nature of the spirit it manifests as the basis for inferring the nature of, for example, a wasp from its body, than the other way around. For Wittgenstein, this is to say that one's subjectivity is no more manifest in one's own body than in the body of the wasp—or, for that matter, in anything else in the world. It is then possible, Wittgenstein writes, to "speak of a will that is common to the whole world. But this will is in a higher sense my will. As my idea is the world, in the same way my will is the world-will" (NB 85).
Wittgenstein thus inverted Schopenhauer's reasoning, but came to the same conclusion. Rather than using one's unique relation to one's body as the basis for positing a subjectivity that stands in an equal relationship to the world as a whole, it is just because one does not stand in a unique relation to any single part of the world that one is able to do this.
As shown below (section 6), the Tractatus develops this universal subjectivity into the defining feature of the self, which in the Tractatus means: of Wittgenstein's self; or of Wittgenstein. This is tied to his having reached a more settled view on the fundamental correctness of solipsism. Having embraced the identification of the world with his experience, Wittgenstein is led to identify himself with that experience, and so with the world. For our immediate purposes, however, the most significant aspect of Wittgenstein's discussion of psycho-physical parallelism is how it relates the subject with the world. As mentioned earlier, if in his notebooks Wittgenstein was analyzing a semantic solipsism, their discussions of the I should be addressing the nature of the subject as representer and its relation to the world as given through their language. Yet they do not. Instead, they focus on the relation between the I and the world of facts: How the subject's experience of the world determines the world's nature, and what this in turn tells us about the nature of the subject. As the I of solipsism, it is therefore the I for a solipsism likewise directed towards the world of facts; which is to say, for a traditional, not a semantic solipsism.
3. what solipsism means, and solipsism
So far, this paper has focused on highlighting some common threads that run through Wittgenstein's pre- and post-Tractatus writings on solipsism, with the aim of providing an informed view of the kind of solipsism we would expect the Tractatus [End Page 137] to be engaging with. Of course, it need not follow, from the fact that he did so preand post-Tractatus, that Wittgenstein must have engaged with traditional solipsism in the Tractatus as well. In light of the fact that that text does not tell us what it takes solipsism to be, however, and given the continuity in Wittgenstein's thought on solipsism either side of it, traditional solipsism should be taken to offer our best starting point for identifying the kind of solipsism that features in the Tractatus.
If this is granted, then the question we must ask is: What reason could there be either for saying that the Tractatus's reference to "what solipsism means" was not meant to relate to traditional solipsism—as Wittgenstein's other early texts addressed—or that its stated endorsement of this was not simply meant as an endorsement? I propose that the following three points, taken broadly, provide the only real contenders:
1. Because 5.62 says solipsism would be nonsensical to assert, and nonsense cannot be correct.
2. Because Wittgenstein's argument in the early 5.6s cannot warrant calling traditional solipsism correct.
3. Because a commitment to traditional solipsism is precluded by other of Wittgenstein's philosophical commitments.
In what follows, I look at each of these in turn, arguing that none offers a preferable way to construe the 5.6s than as endorsing traditional solipsism.
To the three objections listed we could possibly add a fourth: That the Tractatus is simply better read some other way, either as relating to semantic solipsism, or as offering limited (or no) support to solipsism. However, I address such objections by proxy. Sections 4 and 6, respectively, look at the two pronouncements made in the 5.6s that could be taken to give the strongest indication that Wittgenstein's stated support for solipsism should not be taken as a straightforward endorsement; namely, their assertions that the truth of solipsism "cannot be said" (TLP 5.62), and that the implications of solipsism, fully thought through, show it to coincide with realism (TLP 5.64). By arguing that neither statement in fact precludes a commitment to traditional solipsism, I at the same time argue that neither should be taken as mitigating Wittgenstein's stated support for the doctrine.
The question of how to evaluate seeing the Tractatus as addressing (and endorsing) traditional solipsism, as against addressing and endorsing semantic solipsism, is more difficult. I suggest the following. Responding to objection (2) will show how Wittgenstein's argument in the early 5.6s can be taken to address traditional solipsism. While I believe this is how it should be read, I will not try to argue it is the only way it can be read. Indeed, as an argument, the early 5.6s are so compressed as to be internally consistent with a range of different interpretations. What I take this to show, however, is that the question of whether Wittgenstein's argument concerns semantic or traditional solipsism can only be answered by looking outside the section of text in which he actually makes that argument. I consider it already established that traditional solipsism accords better with his work on solipsism outside the Tractatus itself. The final section of this paper, by way of responding to objection (3), argues that traditional solipsism also accords better with Wittgenstein's comments on the metaphysical subject in the second half of the 5.6s. [End Page 138]
4. objection one: the nonsensicality of solipsism
There is a sense in which the fact that Wittgenstein considered solipsism incapable of being stated does prevent us saying that he considered it true. The Tractatus uses 'true' (and its cognates) with a particular sense: to designate propositions that represent facts. Claiming to be an insight into the world as such, not about the obtaining of a particular thing in it, solipsism cannot be stated and, accordingly, is not such as could be called true. Wittgenstein of course acknowledged this, placing solipsism among the things that while they cannot be meaningfully said, yet show themselves in what can be said. So, while it is right to say that solipsism is not true—in the sense in which the Tractatus used that term—this should not prevent us saying that Wittgenstein considered what the solipsist is trying to say is correct.
This is how Wittgenstein's endorsement of solipsism has traditionally been understood; and it is, in essentials, how I mean to present it here. However, this picture has been challenged over the past two decades by the accession of so-called "resolute" readings of the Tractatus.18 Resolute readings are distinguished from what they consider the "standard" tradition of Tractatus interpretation through the "austere" conception of nonsense they take the book to contain.19 This is the idea, reasonable in itself, that an utterance cannot be both nonsensical and yet have specific content. Nonsense is only ever mere nonsense: it is "simply unintelligible—it expresses no thought"; is just "empty"; or is "completely incomprehensible gibberish."20 Significantly, however, resolute authors take this unintelligibility to apply to the Tractatus itself. Having called its propositions "nonsensical" (TLP 6.54), they deny that Wittgenstein could have intended them to convey specific—though technically ineffable—insights to the reader. Instead, they argue the book was written to seem as if it were advocating the various theses it appears to be—such as solipsism—but only to bring the reader (through a dialectical process of trying to understand them) to recognize and reject them as philosophical confusions.
The resolute reading therefore challenges my position through disputing the hermeneutic framework on which it is based. While a comprehensive defense of the claim that the "standard" framework (at least as it relates to Wittgenstein's ideas of showing and nonsense) more closely reflects his thought in the Tractatus is beyond the scope of this paper, a few comments can be made on my basic disagreement with the resolute alternative in its particular application to the 5.6s. The fundamental flaw, which appears inherent in the resolute hermeneutic, is the arbitrariness it displays in ascribing content only to certain passages of the Tractatus. This is particularly evident in its claim that the text can be divided into [End Page 139] a so-called "body" of nonsensical passages, which perform an elucidatory function by being recognized as such, and a meaningful "frame" that explains how that function is performed.21 Now, a frame/body distinction is essential to the resolute reading. Without one, the passages of the Tractatus would have to be considered either as all devoid of content—a position that could not be justified by reference to anything Wittgenstein actually said, since the claim would be just that he is not saying anything—or as all having content, which is simply a traditional reading. As I have argued elsewhere, however, there is no way to justify drawing this division.22 Wittgenstein does not say the book is split into austerely nonsensical and ordinarily meaningful passages (his failure to state explicitly in 6.54 that every proposition in the book is nonsensical aside);23 without such a statement, we as readers cannot assign lack of sense for Wittgenstein to any of its passages (since according to the austere conception of nonsense, it is the author of a sentence, not a reader, who arbitrates its meaningfulness for that author);24 nor is a frame/body distinction necessary. It can appear necessary only if we think that Wittgenstein considered all nonsense "austere."25 However, and this is the important point, any attempt to establish this, if not arbitrarily selective, is ultimately self-defeating because it has to be justified by reference to what Wittgenstein actually said about nonsense. Acknowledging this, resolute readings assign content/frame status not only to 6.54 but also to certain other passages in which Wittgenstein talks about nonsense, on the grounds that we must understand how Wittgenstein uses the term in those passages if we are to understand its use in 6.54.26 (Indeed, James Conant, a leading proponent of the resolute interpretation, writes that it is "a criterion of adequacy that must be met by any textually faithful account of what Wittgenstein means" by 'nonsense' that it must accord with what Wittgenstein says in those other passages.)27 Of course, those passages also employ concepts that, likewise, must be understood in order to understand the passages in which they feature, and thereby justify the claim that Wittgenstein meant nonsense austerely in 6.54. Appearing again to recognize this, resolute authors have shown willingness to extend the frame to a further rank of passages, which feature and explain concepts involved in the previous rank.28 [End Page 140]
Yet if it is justifiable to extend content in this way—and indeed it is necessary, to substantiate interpretation of one passages through what is said in others—there seems to be nothing that could justify bringing the process to an end, short, that is, of ascribing content to the Tractatus in entirety. Concepts that feature in the third rank of frame passages are likewise linked to a fourth rank, and those to a fifth. The Tractatus forms a network, each passage linked to others that address common ideas, and through those passages to others again. Consistently applying the reasoning resolute authors use (and need) to justify their claim that Wittgenstein meant nonsense austerely in 6.54 would therefore result in content being extended throughout the entire book, thus demonstrating that this could not have been what Wittgenstein meant at all.
As I have said, this is only a summary assessment of the resolute program. What it shows, however, is that in being dependent on not following its interpretative process through, the resolute position is essentially arbitrary.29 The idea that the book's nonsensicality could preclude the Tractatus from having meant to advocate solipsism—or indeed, any of the other theses expounded in its text—therefore involves attributing to Wittgenstein a position on nonsense we cannot consistently judge him to have held. At the same time, the fact that our ability to recognize the nonsensicality of the Tractatus not only allows but actually requires that its passages, despite their nonsensicality, still be taken to convey content, suggests the specific inability to state the truth of solipsism likewise should neither be taken to exclude any particular way of understanding that truth, nor to mitigate the book's endorsement of it.
5. objection two: wittgenstein's argument in the 5.6s
This section presents a reading of Wittgenstein's argument in the early 5.6s, according to which the argument was intended to demonstrate the correctness of a traditional solipsism. I discuss the question of whether this is consistent with Wittgenstein's wider philosophical commitments in the following section.
The commonly-held view, that the Tractatus meant to endorse only a form of semantic solipsism, is encouraged by a difficulty the reader faces in approaching this section of the book. In evaluating philosophical arguments, we generally expect to be able to understand the position being argued for independently of understanding how it is being argued for, usually through definition of the terms involved. We also expect to understand the conclusion prior to evaluating its argument, since what we are evaluating is the success of the argument in establishing that conclusion. When it comes to Wittgenstein's argument in the early 5.6s, however, because it does not define its terms, and because its conclusion—"The world is my world"—is so equivocal in its meaning, the reader of the Tractatus [End Page 141] seems compelled to reverse this usual order: To try and work through how the argument proceeds first, and from this extract definitions for the terms of its conclusion. In other words, it seems that the only option available to the reader is to seek out a (more or less solipsistic) position that Wittgenstein's argument could justify calling correct, and then simply infer that it must therefore be the position that his argument meant to be calling solipsism.
It is this exegetical situation that encourages reading semantic solipsism into the early 5.6s. Because the opening premise of its argument—"The limits of my language mean the limits of my world"—is also the premise in which Wittgenstein says the truth of solipsism "is manifest" (TLP 5.62), 5.6 clearly concerns the limits of the world-as-possibilities. Insofar as we consider ourselves compelled to treat this as the starting point for deciding which idea of the world Wittgenstein's solipsism relates to, it quite naturally suggests thinking that the world-as-possibilities must remain the focus of his entire discussion.
If the only way to establish what Wittgenstein meant by 'solipsism' were by extracting this ex post facto from one's reading of his argument in the 5.6s, it is then easy to see why these passages should so commonly be taken as endorsing a form of semantic solipsism. As I have argued, however, we are, in fact, not in this position. Looking first outside the 5.6s, at what Wittgenstein called solipsism in his other early writings, means we can approach that section of text with at least some idea of what its subject-matter meant for him. And if we do this, it opens the possibility for a different way of viewing its progression. Rather than being led into thinking, by the fact that 5.6, as its opening premise, addresses the world-as-possibilities, it must be doing so in an argument whose conclusion addresses the world-as-possibilities, the informed expectation that the Tractatus would engage with traditional solipsism suggests we should instead be looking at 5.6 to see how it might be intended to relate to that doctrine. Our prior appreciation for what Wittgenstein called solipsism, in other words, raises the possibility of looking at his argument as intended to progress from an insight into our relation to the world-as-possibilities, to a conclusion about our relation to the world-as-facts.
If we do approach the 5.6s from this point of view, the argument Wittgenstein puts forward becomes reasonably direct. Before looking at this argument, however, we should begin by noting that, independent of the 5.6s, the Tractatus presents a generally solipsistic worldview. Its early sections develop a model of the world in abstract. The world is constituted by the totality of facts (TLP 1.1), with a fact being a state of affairs that obtains (TLP 2). A proposition is true when it represents such a state of affairs (TLP 2.222), with the totality of true propositions therefore constituting a "complete description of the world" (TLP 4.26). Truth or falsity cannot be determined a priori (TLP 2.225). Rather, propositions are established as true when they have been compared with reality and found to represent it correctly (and, vice versa, falsely) (TLP 2.223). A complete description of the world is therefore given by the totality of propositions that have been compared with reality and found to be true.
These doctrines begin pointing towards solipsism when we consider what such a description would actually consist in. Wittgenstein writes: "The world is completely described by giving all elementary propositions, and adding which of them are true [End Page 142] and which false" (TLP 4.26). Now, suppose someone set out to do this; namely, to list all the propositions, and catalogue which are true and which false. It should also be supposed that they catalogue only propositions known to be true: First, because otherwise they should have no reason for calling their list a description of the world (or this world), distinct from other possible worlds they could similarly construct in language; and, second, because, according to the Tractatus, propositions are not true independent of being known to be true. "It is impossible to tell from the picture alone whether it is true or false. There are no pictures that are true a priori" (TLP 2.224–2.225). Note that propositions not only cannot be known to be true or false a priori—that is, prior to having been compared with reality (TLP 3.05)—but, further, that they are not true or false a priori.30 This is because it must be possible to represent wrongly. In order to explain how propositions can be false, while yet remaining meaningful (and equally, how it is possible to say what is not the case), Wittgenstein argued that propositions must represent not states of affairs as such, but possibilities of states of affairs (TLP 2.15–2.1511; 2.2–2.225; 3.24; 4.06–4.061). If propositions meant states of affairs, then a state of affairs that did not obtain both could not be said not to obtain—since there would be no state of affairs for a proposition to mean—nor falsely be said to obtain; for the same reason, such a proposition would not be false but meaningless (TLP 3.24). For this reason, Wittgenstein argued that propositions must represent only the possibility that a situation obtains, with the logical arrangement of names in a proposition expressing the possibility that objects corresponding to those names—and sharing their logical form—stand in a logically-isomorphic arrangement. Taken a priori, propositions themselves therefore cannot be considered true or false; rather, they are bipolar, like the possibilities they represent, gaining truth-value only through being compared with reality and being found to depict it correctly or incorrectly. [End Page 143]
As Wittgenstein suggests in 2.22–2.225, one consequence of the bipolarity of representation is that a proposition's being true becomes inseparable from its being known to be true. This is because bipolarity entails that representation can be made only of the possibility that a situation obtains, not of the fact that a situation obtains: "A proposition shows how things stand if it is true. And it says that they do so stand" (TLP 4.022). We are therefore unable to represent that a proposition is true (or false). An attempt to state a proposition's truth-value a priori could only be made through using a second proposition, which, as such, must therefore be itself bipolar. This second proposition would then simply be representing the possibility that the situation represented by the first proposition obtains, and so be representing the same thing as the proposition of which it tries to state the truth. For the same reason, propositions also cannot be represented as having any truth-value (a truth-value per se) a priori. We might try to argue that, irrespective of whether it has yet been determined to obtain, a given state of affairs must still either obtain or not, such that a proposition representing it really does have a truth-value a priori, but simply a truth-value we have not yet determined. Yet bipolarity leaves this empty because, again, in stating the argument, we could only be representing something possible. If we tried to say of propositions in general that they must have truth-value a priori, we would be describing a possibility no experience could possibly show to be true.31 And if we tried to say of a particular proposition that, even a priori, it must either be true or false, we could only be saying that the situation it represents must either obtain or not, and so, again, be representing the same thing as the proposition we are trying to talk about. The sense in which a represented state of affairs can be considered to necessarily obtain or not (and so the representation of it to necessarily have some truth-value) a priori, is that these possibilities are inherent in the fact of its being represented; which is to say, in the fact of a representation of it being bipolar. But for that reason, we cannot represent that one of those possibilities either is or must be the case.
In this way, bipolarity effectively precludes the representation of truth-value. This means a distinction between what truth-value a proposition might be determined to have a posteriori, and the truth-value we might want to say the proposition "had" a priori is incapable of being formulated. Knowledge of truth-value thus becomes indistinguishable from truth-value itself. Wittgenstein's point, in 2.22–2.225, is that we therefore cannot represent (or think) our way to truth. The bipolarity inherent in representation can be transcended only through the added involvement of something other than representation; specifically, through the involvement of experience, and the direct comparison of a proposition with [End Page 144] what it purports to represent.32 What this means, however, is that a proposition can be considered true, or a state of affairs said to obtain, only if or while the individual using or evaluating it is able to make this comparison; and that means, while the element of reality it purports to represent is present to them in experience. At all other times, bipolarity means that it simply cannot be attributed a truth-value. Even propositions previously found to be true cannot be said to be true now if they cannot be compared with reality now, since that would ignore the possibility of their now being false.
Due to these constraints, a complete description of the world of the kind outlined above would only seem able to contain propositions that the person giving the description could verify as true at the time of giving it. In this way, their complete listing of the facts would be just a complete description of their present experience.33
In short, the Tractatus appears to provide no basis for classifying any of the inherently possible states of affairs a person can represent in language as facts (or for saying that such-and-such a proposition is true or that such-and-such is known) other than the present availability of that fact in experience. Having recognized them earlier in the text, these same ideas can then be seen to underlie Wittgenstein's argument in the early 5.6s:
5.6 The limits of my language mean the limits of my world.
5.61 Logic pervades the world: the limits of the world are also its limits.
So we cannot say in logic, "The world has this in it, and this, but not that."
For that would appear to presuppose that we were excluding certain possibilities, and this cannot be the case, since it would require that logic should go beyond the limits of the world; for only in that way could it view those limits from the other side as well. [End Page 145]
We cannot think what we cannot think; so what we cannot think we cannot say either.
5.62 This remark provides the key to the problem, how much truth there is in solipsism.
For what solipsism means is quite correct; only it cannot be said, but makes itself manifest.
The world is my world: this is manifest in the fact that the limits of language (of that language which alone I understand) mean the limits of my world.(TLP 5.6–5.62)
Wittgenstein's argument begins by reaffirming the intrinsic co-extensionality of language and the world: Every possible state of affairs can be represented in language, and every proposition represents a possible state of affairs. In this sense, the limits of language and the limits of the world (as possibilities) are the same. Yet it is the next passage, 5.61, that performs the bulk of the argumentative work. While often missed, it is crucial to note that, with its statement, "we cannot say in logic, 'The world has this in it, and this, but not that,'" Wittgenstein's focus shifts from possibilities back to the world as the totality of facts. We can recognize this by entertaining the counterfactual, and rewriting the sentence in such a way that its reference to what the world has in it would be about the world-as-possibilities. So instead of, "The world has this in it, and this, but not that," a world-as-possibilities reading would say, "Among the possibilities there is this possibility, and this possibility, but that is not a possibility"; or, "p and q are possible, but r is not possible."34 Now, on this reading, only the final clause could conceivably be taken to exclude a possibility (as is the basis for Wittgenstein's objection to the sentence) since the first clause, "p and q are possible," offers two possibilities.35 But then, what are we to make of the final clause? For "r is not possible" to exclude a possibility, evidently r must be a meaningful proposition. This would indeed present Wittgenstein's reason for objecting to the sentence, insofar as it suggests that language can be used otherwise than to represent possibilities. More generally, it seems to imply that the world can be delimited; that we are able to classify possibilities according to their belonging to our world, or not, and thereby to identify both where the world's limit lies, and what stands beyond it.
Yet this reading of the sentence, while it would offer Wittgenstein reason for objection, does not actually fit the specific objection he makes in 5.61. What 5.61 objects to is the idea that we could say what the world has in it "in logic," which here means a priori, or as determinable without input from particular experience.36 [End Page 146] Its contention is that, because all states of affairs are inherently consistent with logic ("logic fills the world"), we cannot say without looking at the world what it has or does not have in it. The equivalency of logic, language, and the world-as-possibilities, in other words, is only being argued to rule out one way of identifying the contents of the world. The apparent implication is that "outside" of logic—by involving input from experience—we can identify what the world has in it. Yet, this would not make sense if Wittgenstein were talking about the world-as-possibilities. On a world-as-possibilities reading, the only part of the sentence that Wittgenstein's objection could apply to is its attempt to say what is not in the world that, for it to apply, needs to be taken as trying to say that a certain possibility is not possible. And obviously, we cannot say in logic that such-and-such a possible state of affairs actually is not possible; but then, we cannot say this a posteriori either. Nor could we appeal to "showing" to achieve the desired contrast, since this also is not something that could show itself.
When read as addressing the world-as-facts, on the other hand, why Wittgenstein should object only to saying "in logic" what the world has in it becomes clear. He is objecting to the idea that we could identify a priori which states of affairs obtain. All possibilities accord equally with logic, so there can be no logical basis for prioritizing any as being the facts. The same is true of language. The limits of Wittgenstein's language are the limits of his world—it inherently corresponds to what is possible—but for that reason language cannot tell him anything about what actually exists.
Of course, it is possible to determine what exists. Wittgenstein's point in 5.61 is simply that doing so requires incorporating something more than language; namely, as we have seen, experience. Yet having thus drawn attention to experience, and its role in identifying reality, Wittgenstein leaves the reader to draw the link with solipsism for themselves. At the same time, we have also seen how the concepts he is invoking regarding truth and experience already point in this direction. Because representation is inherently bipolar, the capacity for experience to fix truth-value can only work on a case by case basis, and only while the experience that fixes it is being had. The ability to ascribe truth and falsity, or facthood, or existence, is therefore only momentary; a function of the present experience of the person describing their world. It is this that brings Wittgenstein himself to the fore. When restricted to the level of possibilities, language is impersonal: All languages share the same logic, and can represent the same possibilities; so there is nothing to mark a possibility as "belonging" to one person more than another. When it comes to describing the facts, however, individuation emerges, because propositions can now be marked, through being true or false, on a basis essentially connected to the person giving the description.37 The world of facts, according to [End Page 147] the Tractatus, therefore does belong to one person: It belongs to the person who, through their description of the world, endows truth-value on propositions and facthood on states of affairs; the person whose experience differentiates what is the case from all that possibly could be. For Wittgenstein, whose description of the world we are reading in the Tractatus, the world is therefore his, coextensive with his experience, and the idea of facts obtaining "beyond" it is quite literally something that cannot be imagined.
6. objection three: realism, solipsism and the self
In this way, Wittgenstein's argument in the early 5.6s can be seen to derive a traditionally solipsistic conclusion, relating to the world-as-facts, from his account of the correspondence of language and the world-as-possibilities. The world is Wittgenstein's world, because the nature of his language renders experience his only means for classifying anything as belonging to it. Like the solipsist's notation, the model of representation developed in the Tractatus does not provide the means for drawing a distinction between descriptions of subjective experience and objective fact. And, as in the solipsist's notation, such epithets as 'real' and 'fact,' if they had meaning at all, could only be applicable to the contents of their speaker's experience. Of course, a description of the world using such formal concepts could not have meaning, so the truth of solipsism cannot be said; but it makes itself manifest, in that the limits of truth and falsity always corresponding to the limits of experience.
There is therefore nothing preventing the early 5.6s from being considered to address traditional solipsism; and indeed, ample reason to do so. Our final objection to seeing the Tractatus as genuinely endorsing solipsism, as stated in section 3, was that it is precluded by other of Wittgenstein's philosophical commitments. This section focuses on two ideas that, through Wittgenstein's having endorsed them, might be taken to suggest most strongly that he could not also have meant to endorse solipsism. These are: first, his emphasis on the inability to empirically individuate the metaphysical subject; and second, the Tractatus's supposed commitment to realism.
Why should the inability to individuate the subject be inconsistent with solipsism? The suggestion is that it denies solipsism the reference point needed to be genuinely restrictive.38 The solipsist claims that only a determinate range of things (facts, meanings etc.) exist. Where this purports to be restrictive is in [End Page 148] maintaining that the things that do exist can be differentiated from other things that do not. The specifically solipsistic aspect of the solipsist's claim is that it is a thing's relation to the solipsist that is supposed to determine which group it belongs to. Establishing solipsism's restrictiveness therefore appears to require being able to display coordination between the solipsist (as subject) and the determinate range of things that constitute their restricted world, and thereby the solipsist's non-coordination with the other things that do not. If, the objection goes, the solipsist is incapable of individuating themselves as subject (as Wittgenstein appears to suggest) then this coordination cannot be shown, and the restrictive mechanism cannot be put into place. The implication is that Wittgenstein was aware of this; and indeed, that by presenting the subject this way, his aim was precisely to show that solipsism cannot be formulated as a genuinely restrictive thesis. Hence the idea of the subject put forward in the 5.6s might be taken to show that Wittgenstein did not mean truly to endorse solipsism.
David Pears was arguably the most prominent advocate of our second suggested conflict; namely, that Wittgenstein's intention in the 5.6s was to promote realism over solipsism.39 The argument Pears described is reasonably straightforward, and indeed is built upon the previous objection.40 According to Pears, the 5.6s offer an alternative between two worldviews: A restrictive solipsism and an unrestrictive realism; the first portraying the world as limited through its relation to a particular subject (the solipsist); the second as limited only internally, through the totality of objects (or whichever constituents we take to constituent the world) simply being the totality. By showing that solipsism cannot be formulated in a genuinely restrictive way, Pears argues Wittgenstein's point was to exclude it as an alternative; to show how "we can altogether abandon it and return to realism."41
One need not formulate this argument exactly as Pears does; the suggestion that Wittgenstein saw solipsism as correct only in a sense that takes it to be fundamentally unrestrictive, and that in being unrestrictive solipsism emerges into realism, is in any case a familiar one.42 In order to see that neither Wittgenstein's comments on the metaphysical subject, nor on realism, should be taken as objections to solipsism, we can begin by noting two points. Both are brought out by juxtaposing the Tractatus's central pronouncement on the connection between solipsism and realism (TLP 5.64) with a similar statement from his notebooks:
Here it can be seen that solipsism, when its implications are followed out strictly, coincides with pure realism. The self of solipsism shrinks to a point without extension, and there remains the reality coordinated with it.(TLP 5.64)
This is the way I have travelled: Idealism singles men out from the world as unique, solipsism singles me alone out, and at last I see that I too belong with the rest of the world, and so on the one side nothing is left over, and on the other side, as unique, the world. In this way idealism leads to realism if it is strictly thought out.(NB 85) [End Page 149]
The first point revealed by comparing these passages concerns the context in which each occurs. The formulation from Wittgenstein's notebooks comes as the culmination of his discussion of psycho-physical parallelism, and the idea that his "spirit" or "will" is equally present in everything that makes up his world. Similarly, according to the numbering system of the Tractatus, 5.64 should be read as following on from 5.63, where Wittgenstein states: "I am my world. (The microcosm.)." Thus, in both cases, it is the insight that the self and the world are identical that prompts Wittgenstein to see solipsism and realism as coinciding.
This context is significant, because it clarifies what Wittgenstein was calling realism. Crucially, it highlights Wittgenstein did not have in mind any such doctrine as that the nature of reality is independent of any individual conception or perception of it. Instead, what Wittgenstein called 'realism' is principally the idea that reality is entirely constituted by a single kind of thing—facts—which are on a single metaphysical level. This is most apparent in the Notebook passage. The contrast it draws between realism, on one hand, and idealism and solipsism, on the other, centers on the idea that, whereas idealism and solipsism bestow a "unique" status on certain facts as those inhabited by subjectivity ("men" and "me" respectively), realism accords a unique status to the world, or facts, as such. With the adoption of psycho-physicalism, and its suggestion that his "spirit" is equally present in everything he experiences, Wittgenstein has therefore reached a view he considered equivalent with realism. This same idea of realism then features in 5.64. If the subject existed in the world it would have a unique status: The world's being their world makes the subject a precondition of its existence, such that its existence alone would constitute a necessarily-obtaining fact. By instead positing the self as the world's ubiquitous "limit," Wittgenstein nullifies this privilege. When the self shrinks to a point without extension, all that remains, as the monistic claim of "pure realism" asserts, is "reality," solely constituted by the contingent facts with which the self is coordinated.
Juxtaposing these passages therefore helps clarify what Wittgenstein meant by 'realism.' The second thing it reveals centers on a small but significant difference in wording between the early and later formulations. Whereas in his notebooks Wittgenstein says solipsism "leads to" realism—apparently suggesting a rational person might progress through solipsism, but eventually jettison it in favor of realism—the Tractatus says the two positions "coincide." Thus, the passage does not in itself present grounds for concluding, with Pears, that Wittgenstein "is, of course, a realist, as he implies at 5.64."43 What Wittgenstein says in 5.64 is that solipsism and realism make the same claims.44
What do these points tell us about our objections? The fact that the Tractatus's sole statement on solipsism and realism concerns the sense in which they are the same, not in which they conflict, tells us, first, that we should not expect [End Page 150] Wittgenstein's intention to involve the affirmation of one over the other. More significantly, the shift in wording between the different versions of 5.64 tells us the author of the Tractatus did not consider the restrictivity of solipsism to depend on the capacity to individuate the metaphysical subject. In the notebooks, solipsism is portrayed as involving a worldly ego, on which it bestows a unique status among the facts of the world. Upon embracing the psycho-physical conception of subjectivity—rendering no part privileged through relation to his ego—Wittgenstein therefore appears to deny solipsism in favor of the universal equality he equates with realism. The Tractatus, on the other hand, portrays solipsism and realism as compatible. This change suggests Wittgenstein no longer saw solipsism as involving an empirical ego, nor as necessitating bestowing privilege upon any particular facts as the locus of subjectivity.
What Wittgenstein appears to have realized is that the logical system he was describing in the Tractatus leaves him both unable to posit facts outside his experience and incapable of individuating himself as subject. Given his conviction regarding the fundamental correctness of the text, it is therefore to be expected that he should seek to find a way of reconciling these truths. He effects this reconciliation through what we earlier saw as the apogee of solipsism's attempt to identify the I; namely, its identification with its experience: "I am my world. (The microcosm.)" (TLP 5.63). This remarkable assertion, which begins Wittgenstein's discussion of the I in the Tractatus, is, like 5.62's affirmation of solipsism, rarely taken as meaning just what it appears to be saying. Most commonly, it is simply ignored. Yet by taking 5.62 as relating to traditional solipsism, and identifying the world with Wittgenstein's experience, it is not difficult to see that his subsequent identification of the world with himself should then be interpreted literally. Wittgenstein's expulsion of the subject from the world is sometimes taken as a rejection of the subject.45 Yet this is a mistake. While it "shrinks to a point without extension" (TLP 5.64), for Wittgenstein the subject does not lack reality ("The subject—we want to say—does not here drop out of experience but is so much involved in it that it cannot be described" [PG 156]). Instead, just as he did in his pre-Tractatus notebooks, and as he analyzed the solipsist as doing post-Tractatus, what the Tractatus was rejecting was the subject-object distinction. Being unable to identify his subject with anything in the world, he instead identifies it with his subjectivity as such. But, having already denied a subjective-objective distinction in his endorsement of solipsism, by identifying himself with his subjectivity, Wittgenstein is at the same time identifying himself with the object of his subjectivity, which is to say, with his world.
While sounding esoteric, Wittgenstein's thought in 5.63 is not particularly difficult to understand. If the solipsist as subject is nothing in itself, nothing beyond their experience, then their reality can be said to be their experience, and if their experience is all that is their world, then they can also be said to be their world. The important thing is that by carrying through the non-appearance of the I in experience into an identification of the I with its world, Wittgenstein seemingly [End Page 151] achieved a conception of the subject that allows solipsism to be genuinely restrictive without requiring an empirically identifiable subject. As the Tractatus portrays it, the realm of truths is delimited by classifying propositions as true, false or possible, on the basis of their agreeing, not agreeing, or not having been compared with reality, respectively. This makes truths an inherently restricted subset of the total range of propositions. All that is therefore needed to achieve a restriction in this realm is simply to describe reality. In describing our world, we affirm a certain range of propositions as representing facts, and thereby distinguish these from the wider range of possibilities which, while our language allows us to represent them as potentially obtaining, we are unable to verify. The fact we can still represent them, however, means our listing of the facts involves a genuine restriction.
Defining his self as his world also provides Wittgenstein grounds for considering the restricted set of truths given by his description as being essentially his; that is, grounds for calling his belief that there is such a restriction solipsism. In describing their experience, the solipsist draws a line around a set of propositions as true, and thereby individuates their world from the wider range of possible worlds they are able to construct in language. At the same time, this individuates and defines the subject as that same set of facts. This experience, this reality, is my reality, which I can distinguish from all the other sets of facts that possibly could have made up my being. This makes it plain that Wittgenstein does not need a further empirical subject to justify calling the facts he encounters in experience his. The facts are Wittgenstein's because, in essence, the facts are Wittgenstein.
If this or something like it is what Wittgenstein had in mind in calling himself his world—and it seems the most straightforward way of taking that statement—it allows us to recognize four things. First, the fact that its subject is not empirically individuable does not negate the restrictivity of solipsism. Hence its concept of the subject does not preclude the Tractatus from also endorsing a traditional solipsism.
Second, because its restrictivity does not require an empirically-identifiable subject, solipsism does not necessitate attributing "unique" status to certain facts with which to identify that subject. Solipsism can ascribe the same contingent status to all facts, and therefore be consistent with what Wittgenstein considered the central claim of realism. Hence any commitment the Tractatus had to realism would not preclude the book from holding a similar commitment to solipsism.
Third, Wittgenstein's identification of the self with its world shows that the 5.6s were addressing traditional solipsism. As discussed earlier, his recurrent joining of solipsism with investigation into the subject is best understood in terms of solipsism (through denying a subjective-objective distinction) leaving a subject-object distinction untenable, thus forcing the solipsist to reassess the nature of their subjectivity. This meant that the content of the two analyses should correspond: The object that the subject is connected with should be that world to which the solipsist is denying objectivity. The same general argument that was made, regarding the way in which their discussion of psycho-physical parallelism show the notebooks to have been addressing traditional solipsism, can then also be made for the Tractatus. If the 5.6s were concerned with a "solipsistic" restriction of linguistic meaning, we should expect the subject to be analyzed in connection with linguistic meaning, or linguistic representation of the world as its object. [End Page 152] But it is not. The objects Wittgenstein discusses the subject in relation to are the world (TLP 5.63), expanded as "The World as I found it," or the world of facts as encountered in experience (TLP 5.631), and the visual field (5.633–6.634).46 Taking Wittgenstein's identification of the subject with this world literally therefore allows us to recognize it as the subject of a solipsism that takes the world of facts as its restricted object, which is to say, the subject of a traditional solipsism.
Finally, we can see in Wittgenstein's advocacy of this idea the final confirmation of his advocacy of solipsism. The Tractatus follows the same path charted in Wittgenstein's later work. Adopting a solipsistic denial of a subjective-objective distinction leaves Wittgenstein unable to individuate himself as subject from the objects of his experience, and prompts him to seek resolution through identifying himself as the objects of experience. We can understand this progression only by seeing that Wittgenstein's endorsement of solipsism at 5.62 relates to the same traditional solipsism he addresses in the later work, and is endorsed without qualification.
We can therefore see that none of the suggestions in section 3 offer a valid reason for not taking the 5.6s as addressing traditional solipsism, or Wittgenstein's statement that what solipsism means is quite correct as a genuine endorsement of the doctrine. As I argued, the expectation for continuity with Wittgenstein's pre- and post-Tractatus writings prompts the suggestion that traditional solipsism would, as in those texts, be the kind under analysis in the Tractatus. This suggestion is indeed borne out, both by the direct argument made in the early 5.6s and by their consequent investigations into the metaphysical subject.
Cameron Hessell studied philosophy at the University of Canterbury
bibliography and abbreviations
1. This article employs in-text abbreviations to cite Wittgenstein's works as listed in the bibliography. References to the Tractatus are to the decimally-numbered passages in that text. References to all other works are to page number (e.g. NB 85 is to Notebooks 1914–1916, page 85).
2. I retain C. K. Ogden's correct translation, in 5.62, of "der Solipsismus" as "solipsism," rather than "the solipsist" as per the David Pears and Brian McGuiness translation. On the whole, however, this paper tends to follow Pears and McGuiness, and all references using the 'TLP ' acronym are to it (my preferred translation of "der Solipsismus" notwithstanding).
3. The Tractatus employs a decimal numbering system to designate the relative ordering of its passages. It consists of seven primary "propositions" (numbered 1–7) with the remaining passages either a comment on one of these seven, or a comment on one of those comments, and so on. Wittgenstein's discussion of solipsism and the self, which begins at passage 5.6, therefore constitutes the sixth set of passages on the fifth proposition. It runs from 5.6–5.641 (the entirety of the 5.6 subset) and will commonly be referred to in this paper simply as 'the 5.6s.'
4. The label 'semantic solipsism,' which I favor in this paper, is adopted from Merrill B. Hintikka and Jaakko Hintikka (Investigating Wittgenstein, 66). 'Linguistic solipsism' and 'traditional solipsism' are taken primarily from Pears ("Wittgenstein's Treatment of Solipsism," 58).
5. Readings that take Wittgenstein to have been addressing semantic solipsism tend to focus on one of two senses in which he (or the solipsist) could be claiming to have a "unique" relation to language; namely, either his uniquely possessing something that constitutes language, or possessing something that uniquely constitutes language. This difference is explored in Peter Sullivan, "The 'Truth' in Solipsism," 195–96. So divided, the respective interpretations of Peter Hacker (Insight and Illusion, 100) and Pears (The False Prison, 172) belong to the first category; those of G. E. M. Anscombe (Introduction, 167) and Michael Hodges (Transcendence and Wittgenstein's Tractatus, 83) to the second.
6. See n. 10.
7. See, for example, Peter Carruthers, The Metaphysics of the Tractatus, 80; Pasquale Frascolla, Understanding Wittgenstein's Tractatus, 206; Hintikka, "On Wittgenstein's 'Solipsism,'" 88–91; Hodges, Transcendence and Wittgenstein's Tractatus, 83; McGuiness, Approaches to Wittgenstein, 136; H. O. Mounce, Wittgenstein's Tractatus, 92; and Pears, "Wittgenstein's Treatment of Solipsism," 58, and The False Prison, 172–73. Anscombe (Introduction, 166–67) does not appear to distinguish semantic from traditional solipsism, defining solipsism in traditional terms (an identification of reality with the solipsist's experience) but identifying the world Wittgenstein calls his with the world-as-possibilities. Hacker (Insight and Illusion, 100), on the other hand, sees the book as defending both traditional and semantic solipsism.
8. See n. 3.
9. The Tractatus argues for a distinction between things that can be meaningfully represented in language, and things that, while they cannot be represented, are shown by (and so can be recognized in) meaningful representations. As I discuss below (sect. 4), not all commentators agree that the Tractatus does contain a saying/showing distinction.
10. There is a sense in which semantic and traditional solipsism might be thought not to float entirely free of one another. This is because it is possible to state each in terms of, and as implied by, the other. If the possibilities are essentially one's own, then insofar as some of those possibilities are facts, there is a sense in which those facts could be considered one's own, too. Similarly, if the only facts are those given by my immediate experience, there is a sense in which my experience can be said to determine the limits of language (and so of the possibilities), since the construction of propositions has an internal relation to the objects given by experience (see below, n. 38). It might then be asked why, in calling the world his, Wittgenstein could not be referring to both conceptions of the world, and thus be endorsing both semantic and traditional solipsism. The answer is simply that this would be incongruous: because, as this paper aims to show, Wittgenstein did not use 'solipsism' to mean 'semantic solipsism.' It would furthermore be unnecessary. It is important to distinguish the question of whether Wittgenstein endorsed any of the ideas commentators relate to (what they present as Wittgenstein's) semantic solipsism, from that of whether these were what Wittgenstein was endorsing as solipsism. It would of course be false to say, for example, that Wittgenstein was not interested in the correspondence between the limits of language and the world (the world-as-possibilities) and the implications these have for his ability to imagine or describe his world qua linguistic subject, or that these ideas played no role in his endorsement of solipsism. My argument is rather that, in saying these (and other) ideas show "how much truth there is in solipsism," he meant how much truth there is in traditional solipsism. So, while it is possible to derive a solipsism relating to the world-as-possibilities from a solipsism relating to the world-as-facts, and vice versa, it does not follow that, in endorsing one as solipsism, Wittgenstein must also have been endorsing the other as solipsism.
11. Historically, Wittgenstein's philosophical career has been divided into two periods—termed the early and the later—each characterized by the thought presented in the Tractatus and Philosophical Investigations, respectively. Commentators have increasingly also come to recognize a period transitionary between the two—the middle Wittgenstein—stretching roughly from Wittgenstein's return to Cambridge in 1929 until his abandonment of the manuscript that would be published posthumously as Philosophical Grammar, in November 1936. (Some also recognize a post-Investigations "Third Wittgenstein" [see Daniele Moyal-Sharrock, "The Third Wittgenstein"].) The extent of continuity or divergence in Wittgenstein's thought across these periods is a matter for debate (which is touched on, to a limited degree, in section 1). In order to avoid this issue as far as possible (that is, the issue of whether and how far Wittgenstein's post-Tractatus work can be used to draw lessons from the Tractatus) I will restrict my focus to Wittgenstein's early and middle period writings on solipsism.
12. Wittgenstein's contemporaneously-written typescript 213 (posthumously published as the Big Typescript) contains parallel discussions on the description of experience, and the role of pronouns therein. That text focuses more particularly on these topics in relation to ideas regarding the supposed privacy of sense data. The typescript also offers a deeper critique of the confusions that lead to solipsism (or idealism, or sense datum theories)—in particular, the mistaken application of the grammar of physical space to phenomenal space—than I am able to go into here (see BT 319–69; esp. 356–62).
13. We might even consider solipsism the "and vice versa" here: the investigation of the object of experience as given through its connection with its subject.
18. For a good summary of the resolute program by two of its leading proponents see, James Conant and Cora Diamond, "On Reading the Tractatus Resolutely." For specific discussions of solipsism see, Conant, "Mild Mono-Wittgensteinianism," 58; Conant and Diamond, "On Reading the Tractatus Resolutely," 79–85; and Michael Kremer, "To What Extend is Solipsism a Truth?"
19. On the austere conception of nonsense, see, Conant, "Elucidation and Nonsense," 176–77, "The Method of the Tractatus," 421, and "What 'Ethics' in the Tractatus is Not," 49; Diamond, "Ethics, Imagination and Method," 149; and Kremer, "The Purpose of Tractarian Nonsense," 41.
21. See Diamond, "Ethics, Imagination and Method," 149; Conant, "The Search for Logically Alien Thought," 159; and Conant and Diamond, "On Reading the Tractatus Resolutely," 68. As I argue below, a frame/body distinction is a necessary presupposition of resolute readings as such.
23. This point is sometimes emphasized in resolute readings. See, for example, Conant, "Elucidation and Nonsense," 216n102, "The Method of the Tractatus," 458n136, and "Wittgenstein's Later Criticisms of the Tractatus," 179; and Read and Deans, "'Nothing is Shown,'" 253.
25. As suggested above, resolute authors must ascribe content to 6.54 to say Wittgenstein considered the Tractatus to contain nonsensical passages. If Tractarian nonsense is devoid of content, 6.54 could not say of itself that it is nonsensical. Hence 6.54 (at least) must remain meaningful; and since it says that other sections of the Tractatus contain nonsense, a frame/body distinction would follow.
26. See, for example, Conant, "Elucidation and Nonsense," 178. Conant extends the frame to 5.4733 and 4.112, on the grounds that these passages tell us how 6.54 employs the concepts of nonsense and elucidation, respectively. Note that whenever a resolute author uses content from the Tractatus to justify some claim, they must implicitly assign the passage from which it was derived to the frame of the text.
28. Conant, "Elucidation and Nonsense," 178, for example, features a lengthy explication of Wittgenstein's distinction between signs and symbols, primarily focused on 3.3–3.326. This distinction is integral to Wittgenstein's explanation in 5.4733 of why the pseudo-proposition "Socrates is identical" is nonsensical. (See my n. 27.)
29. The choice, between standard and resolute readings, is thus ultimately between a consistent interpretive process, which if followed suggests the book (in implying nonsense can have content) is arguably incoherent, and an inconsistent process which (given its supposedly implying that process) presents the book as equally incoherent. The fact that the standard reading is consistent suggests that the source of any incoherency it finds in the Tractatus lies with its author, while the inconsistency of the resolute reading suggests its incoherency lies with the reading itself.
30. It could be argued the text surrounding 2.225 indicates the passage should be taken in epistemic terms, as saying only that propositional truth-value cannot be known a priori, and not (as it literally says) that propositions do not have truth-value a priori. Most conspicuously, the two passages immediately preceding 2.225 (2.223 and 2.224) plainly are comments on knowledge of truth-value. Both furthermore describe truth-value as something we "recognize" (or "tell" [erkennen]) through comparing propositions with reality, seemingly suggesting truth-value must already be in propositions, waiting to be recognized prior to our doing so. However, while these points may help explain why 2.225 often is taken for an epistemic claim, they fail to show that it ought to be. While describing truth-value as recognized, Wittgenstein's point in 2.223 and 2.224 is precisely that it cannot be recognized a priori—that is, that there is no truth-value in the proposition to be recognized a priori. The only circumstance in which we can recognize truth-value is when we have the fact a proposition represents (or represents falsely) presented to us. These are the same circumstances that, according to my reading, must be in place for a proposition to have a truth-value. This may then be said to be recognized—without, however, implying the truth-value must have been "in" the proposition (unrecognizably) prior to the comparison.
The epistemic focus of 2.223 and 2.224, on the other hand, cuts both ways. While the fact that these passages are explicitly directed towards epistemic concerns might be taken to indicate 2.225 would also be so directed, this same fact—that the latter passage does not continue the epistemic language of the former—could equally be taken to show that 2.225 therefore must mean to be making a different kind of point. That these ideas should be linked so seamlessly—proceeding from the observation that we cannot know truth-value a priori, to the observation that propositions cannot be said to have truth-value a priori—actually supports my reading, since it suggests a proposition's having truth-value and our having knowledge of it should indeed be thought of as two sides of the same coin.
31. It might be thought a distinction between truth and knowledge of truth is evident elsewhere in the Tractatus; in particular, in a parallel distinction between necessary truth of tautologies and the epistemic certainty we can have in their truth. Support for this is drawn from Wittgenstein's description, in 4.464, of a tautology's truth as "certain" (gewiss). Given Wittgenstein also considered tautologies necessarily true, taking 'certain' epistemically in 4.464 appears to give two concepts—epistemic certainty and logical necessity—which might be extended to non-logical propositions.
However, this argument has two problems. First, it is not clear 4.464 is using 'certain' epistemically. Beyond describing the truth of tautologies as "certain," and propositions as "possible," it also describes the truth of contradictions as "impossible," which is difficult to read epistemically. It is not epistemically impossible to know the truth of contradictions; it is logically impossible (contradictions contain nothing epistemically impossible to know the truth of (TLP 4.461)). This suggests 'certain' should be taken in the logical sense of assured, or given (which gewiss can also denote), in the same way the concept is used in 5.25.
Second, any distinction between the truth of tautologies and our knowledge of that truth cannot be extended to non-logical propositions. We can draw this distinction for tautologies, because it is possible to imagine cases where we do not know a tautology is true; say, because we fail to recognize a tautology as such. Here we can say the tautology was true, even when we did not know it. However, we can say this only because a tautology's truth is logically certain. Tautologies are true in all circumstances, and therefore at all times. This is not the case for non-logical propositions. We cannot infer, from the fact a non-logical proposition has been determined to be true, that therefore it must also have been true before that determination. So, we cannot use any divergence between our epistemic confidence in the truth of a proposition and its verified truth or falsity, as a basis for distinguishing the a priori truth-value of propositions from a posteriori knowledge.
32. The question naturally suggests itself as to whose experience is supposed to determine a proposition's truth-value, to which the answer can only be: whoever is using the proposition. Within the Tractatus, truth-value is a function of the context of a proposition's use, not something attaching to propositions inherently. Propositions therefore do not retain truth-value across individual uses in individual contexts. So, it is the user of a proposition, whoever that is, whose experience determines whether it is in that instance representing something true, false, or possible.
33. This connection between solipsism and description of momentary phenomenal space helps explain Wittgenstein's repeatedly linking solipsism to staring. He writes, for example, that "The phenomenon of staring is closely bound up with the whole puzzle of solipsism" (NLP 309); and similarly: "Thus we may be tempted to say 'Only this is really seen' when we stare at unchanging surroundings, whereas we may not at all be tempted to say this when we look about us while walking" (BB 66).
34. For the Tractatus, "p and q are possible, but r is not possible," is not, of course, a valid proposition (TLP 4.1272; 5.525). It is useful to render the sentence this way to emphasis what, on a world-as-possibilities reading, it must be trying to say for this sentence to contravene the basis for Wittgenstein's objection to it.
35. Note that we must translate the final clause "r is not possible," and not "r is not the case," for the same reason: "r is not the case" negates a possibility—and so states a possibility itself—but it does not exclude a possibility.
36. 'Logic' is likewise used to mean a priori in 5.55–5.5521. The 5.55s and 5.6s, however, seem to feature slightly different ideas of the a priori. Within the 5.55s, 'a priori' is used to designate a preexperiential and, as it were, pre-representational state; it is what can be determined independent of all experience. Thus, names cannot be catalogued a priori, since names get meaning through coordination with objects, and the range of objects can be determined only through experience (TLP 5.55–5.552). In 5.61, on the other hand, Wittgenstein's objection to saying in logic what the world has in it—that this would illegitimately exclude possibilities—requires that what is being said designates a possibility, i.e. be a proposition that, under other circumstances, would have sense. In order for the objection to work, a level of general experience, of the kind excluded from logic in the 5.55s, must therefore be presupposed. In this sense, the idea of the a priori featuring in the 5.6s is closer to that employed in 2.225, designating that which can be determined without input from particular experience, rather than without input from experience per se.
37. Furthermore, as discussed in the following section, Wittgenstein's subsequent identification of himself with his world/experience (TLP 5.63) means any description of the world is at the same time a description of himself. It might be noted here that I accept the sense of Pears and McGuiness's translation of the parenthetical clause in 5.62 over Ogden's translation. In this, I have deferred to a number of German colleagues, who insist that the German, "der Sprache, die allein ich verstehe," makes it clear Wittgenstein is referring to a language that only he understands (not, as in Ogden's translation, the only language that he understands). At the same time, the difference seems to me not particularly significant. What Wittgenstein means to emphasize (and does, I consider it, in either translation) is simply that it is his particular relation to language—his understanding or use of it—that shows the world to be his. As described above, in using his language to depict the world, Wittgenstein both individuates his language as his (through the propositions that are true in it being those representing his experience), and, through the same mechanism, identifies his experience as the world.
44. In the German, Wittgenstein's use of the verb zusammenfallen in conjunction with the preposition mit gives the symmetrical sense of to coincide with. Without this construction, zusammenfallen can also convey the less symmetrical sense of one thing collapsing into another, but not the other way around. The same construction features in the Foreword, where Wittgenstein expresses his disinterest in the degree to which his efforts in the Tractatus "coincide" with those of other philosophers.
46. As Michael O'Sullivan has argued, the Tractatus presents the visual field as consisting of visible facts ("The Visual Field in Russell and Wittgenstein," 26–32).