Wittgenstein and Peirce on inner experience
Rosa M. Calcaterra

This essay surveys and critically discusses the anti-cartesian approach of both Peirce and Wittgenstein to the problem of private experiences, bringing to light the affinity of their own usage of the notion of "outward criteria", that both introduce as alternative to the traditional logical principle of introspection. In particular, are taken in consideration the treatment of sensation language in Wittgenstein's Philosophical Investigations and Peirce's critique of introspective method as a means for knowledge of so called "inner world", in an attempt to put into focus the emphasis on pragmatic function of language they have in common.
Key words: emotion, introspection, language, self-consciousness, sensation, testimony.

1. "An 'inner process' stands in need of outward criteria": so reads No. 580 of Philosophical Investigations[1]. I propose to put into focus some of the progressions in Wittgenstein's laborious scrutiny, which come together in this statement, in order to compare them with some of Charles S. Peirce's utterances about the problem of so-called "inner facts". In their respective arguments, in my opinion, we encounter a significant affinity, both theoretical and methodological, which can, for the most part, be traced back to their shared anti-mentalist attitude, an attitude which, moreover, marks their similarity in other important statements in their philosophies.

In introducing the notion of "outward criteria", Wittgenstein certainly does not intend to deny that inner facts and processes exist: that sensation, for example, is also a private and subjective fact, or that phenomena such as memory entail a mental process. Rather, this concerns showing how explanations centered on the notion of awareness prove inadequate and even misleading. In this sense, a good part of his reflections on the language of sensations seems aimed at putting into focus the epistemic and semantic inadequacy of the introspection principle. The introspective approach, in particular, will be rejected in the end, both as a point of reference for investigating the meaning of the terms of sensation and as a fact-finding tool for the "reality" of sensorial experience. Both the linguistic utterances and the psychological processes of sensation will become the field for an articulated and complex interpretive dynamic, one which will refer back to a reconstruction of the naturalistic origins and socio-pragmatic function of this particular linguistic game, of the behavioral attitudes that accompany it, and of the rules that underpin its use and its understanding between individuals. The concept of "outward criteria", where such factors flow into each other, will provide the key to understanding the problem of private language, which is the starting point for a specific analysis of sensation language, and it will extend to the Remarks on the Philosophy of Psychology[2], constituting a model of Wittgenstein's version of the so-called third-person approach to the problem of the emotions.

In these pages I will limit myself to considering the sections of Philosophical Investigations where the inferential nature of inner experience is presented, first expanding on those that deal with the origins and meaning of the terms of sensation. I will attempt to show how these passages are analogous to some parts of Peirce's critique of the introspection principle's logical value.

The fact that the issue of the language of sensations constitutes a specific and central aspect of the discourse on private languages underscores the continuity this part of the Investigations has with Wittgenstein's critique of solipsism. Sensations, indeed, understood as absolutely private objects, represent the sole object with which the solipsist expects to deal in any way. And the condition of the solipsist is exactly that of the subject of private language. No. 243, which opens the section dedicated to this theme, is very precise on the matter:

The individual words of this language are to refer to what can only be known to the person speacking; to his immediate private sensations. So another person cannot understand the language.

Wittgenstein specifies that the case of the monologue, or the possibility that human beings accompany "their activities by talking to themselves", will not be discussed; rather, the possibility at issue is that of a language built by exclusive reference to one's own inner experiences, a language that someone could use to express in writing or orally "his inner experiences - his feelings, moods, and the rest - for his private use". As is well known, Wittgenstein will conclude by suggesting that the idea of such a language is unsustainable, not simply because an absolutely individual form of language would be unintelligible to anyone else, but also because the same individual constructing it would have a difficult time connecting his own linguistic terms to adequately defined objects.

This aspect has to do with the theme of language rules. Even if one were to admit that a private language could, on a par with any other, constitute a "system that follows a rule" - even though completely sui generis - that is, one that could produce a language and, hence, a set of meanings based on an "absolutely private rule model", it still would be a language that could not be analyzed. In fact, it would require, in this case, the ability to penetrate inside a world that involves matters dealing solely with the person making use of that rule, completely leaving aside any reference to the community to which he or she belongs, as well as any criteria that belonging to a community implies for understanding and applying the rule itself[3]. On the other hand, and perhaps more importantly, such a language would never even reach the point of taking shape: lacking any point of reference, either to the physical world or to the linguistic and behavioral practices of a community, the very possibility would be diminished whereby the meanings which the individual would attribute to the signs he himself had constructed could be kept constant[4].

The unsustainability of the idea of an absolutely private language is connected to the critique of the epistemic and semantic value of the principle of introspection, which, in the Investigations, coincides with his critique of the descriptive notion of the sensation terms. Wittgenstein's polemic target here does not seem to be the capability of introspection as such, but rather the conviction that it represents a privileged mode of access to the "data" of the subjective consciousness, locatable upstream of language, of behavior, of relations with the external world and with others. This idea is, in fact, implicit in the presupposition that there is a direct relation between an "inner" sensorial datum and a linguistic utterance, that is, a relation of objective reference, one on which the descriptive concept of sensation terms is based. No. 244 clearly announced both the methodological program and the theoretical perspective that Wittgenstein will use to articulate his argument. Trying to understand how words "refer" to sensations is the equivalent of investigating an ordinary linguistic practice, namely, the fact that "we talk about sensations every day, and give them names". It is a matter, therefore, of asking how, in this case, a connection between the name and the named is established, and this question - says Wittgenstein - "is identical with the other: how does a man learn the meaning of the names of sensations?" Thus he reprises the methodological style he had already adopted in The Blue and Brown Books[5] for his critique of the reduction of language to a solely descriptive function and, in particular, of traditional name theory, based on the object-designation model[6]. The significant results he obtained by moving from ordinary language to the question of modes of learning now undergo new developments and find additional confirmation. The explanation in which the names of sensations constitute a group of linguistic conventions whose use is the result of mechanical learning and whose function is to describe a thing, is supplanted by the idea that the language of sensations does not describe, but rather "replaces", the sensorial experience. In other words, it is "a new behavior" which the child gradually learns to interpolate between his or her own "immediate" organic and natural experiences and the surrounding social environment.

In Nos. 256 and 257, the "outward signs" of sensorial experience are, in fact, presented as a necessary condition for learning and for the "intentional" function of the language of sensations. They constitute, in other words, the concrete occasion through which adults teach the child to give a name to his or her own natural reactions to a certain sensation, and they are, furthermore, the condition which will make it possible to elicit in others attitudes that are suitable and consistent with his or her own state (see also Nos. 246, 261, and 273).

The descriptive criterion for terms of sensation seems to rest, among others, on the conviction that concentrating the attention on inner experience serves to "impress" on one's mind the connextion between the sign and sensorial datum:

But "I impress it on myself" can only mean: this process brings it about that I remember the connextion right in the future. But in the present case I have no criterion of correctness. One would like to say: whatever is going to seems right to me is right. And that only means that here we can't talk about "right" (No. 258).

In reality, we can never gather our sensible impressions exactly, in other words, we can never know their essence: We understand only their language (No. 348), while their "deep aspect readily eludes us" (No. 387). For this reason, the descriptive concept is not only misleading, but also decidedly without any ground. This is especially so to the extent that it carries with it an opposition between a "datum" of consciousness and language, or between the "inner" and the "outer", an opposition which Wittgenstein instead endeavors to eliminate by showing how these two aspects are indissolubly joined precisely in the specific linguistic game of the terms of sensation. Actually, it is, at the same time, both public and private language. That is, on the one hand, it expresses and communicates private experiences; on the other, it uses terms that belong to the common language which - flattened out under the descriptive function - ends up disguising its origin and its specific function (Nos. 246, 256, 257, 261, and 273).

Of particular interest for our argument is a group of observations that indicate how the notion of consciousness, or "interiority", only appears and is embodied under specific conditions, or at the reflexive levels of communicative praxes. Wittgenstein notes that to speak of a sensation by specifying that it is also an inner fact is merely a useless "ceremony". Put otherwise, "The proposition 'Sensations are private' is comparable to: 'One plays patience by oneself'" (No. 248). In ordinary use, terms of sensation are primarily linked to the organism or to the properties of external objects. For example, the word "pain" expresses, above all, an experience that is localized in a given part of the body, and not, therefore, something inner and inaccessible to others, just as the term "red" designates "something confronting us all ", instead of the inner, private sensation of the color red (No. 273). Only someone who is not expressing himself spontaneously speaks by pointing to himself, and referring to what is private to express a sensation is something typical of "a particular experience in doing philosophy" (Nos. 274, 275).

The reference to the conscious is instead linked to particular pragmatic needs. This is a case of one wanting to tell someone who thinks he has fainted that he is, instead, "conscious", or that - perhaps even to tell his own doctor - he is once again in possession of a certain sensorial faculty that he had momentarily lost. On the other hand, humans "agree in saying that they see, hear, feel, and so on (even tought some are blind and some are deaf). So they are their own witness that they have consciousness" (No. 416). Moreover, Wittgenstein seems to answer the question, "Is my having consciousness a fact of experience?" (No. 418), by stating that it is a "fact" that has no need of particular justification and whose manifestation does not actually contribute to endowing meaning on the ordinary experience of language and of relations between individuals, constituting, instead, its inherent and essential condition. Hence, imagining "that the people around me are automata, lack consciousness, even though they behave in the same way as usual" not only can arouse sinister feelings, but it also is an idea that cannot be sustained "in the mist of ordinary intercourse with others (No. 420; see also Nos. 281, 283, 284). In conclusion, the reference to the conscious represents, in any case, a secondary stage in the use of sensation language, one that seems to be linked rather to "regulated" practices: namely, on one hand, to the linguistic conventions of a social community and to the rules that govern the application of certain verbal expressions; and on the other, to the whole of natural perceptions and reactions which are inscribed in the pre-established relation between man and nature in general.

From this perspective, the "outward criteria" put forward to counter the introspective criterion do not replace its epistemic claims. The dimensions of the language and behavior to which they refer represent precisely the coordinates within which inner experience finds its expression, but they do not reproduce its "truth" or essence. They are points of reference to which is entrusted the recognition between individuals of a certain inner experience, but they do not offer cognitive guarantees or univocal possibilities of definition, just as behavior and language never have univocal meanings and functions.

In this regard, some pages from Observations on the Philosophy of Psychology will prove illuminating, since the notion of "outward criteria" is openly rendered a problem: above all with respect to those objectively observable manifestations - such as facial expression, tone of voice, and gestures - which Wittgenstein seems to consider useful only for a first, hypothetical approximation of one's own psychological state and those of others[7]. Actually, the problem of knowledge of the sensations, and more generally of the so-called inner world, does not lie at the heart of an analysis of the language of the sensations. Wittgenstein sets this problem aside programmatically when he states - as has already been mentioned - that we cannot know what the sensations really are since their deep aspect readily eludes us" and "we understand only their language". And, in fact, the overall intent of his investigation seems to be, not to establish the conditions of truth for statements concerning sensorial experiences, or "inner facts", but rather to delineate the conditions under which they can be asserted, by describing the function, meaning, and correct application of certain expressions[8].

2. Wittgenstein's approach to "inner processes" brings to mind some passages in Peirce's essay, Question Concerning Some Faculties Claimed for Man[9], dedicated to a critique of the logical value of immediate intuition - whether by the senses or by the intellect. The specific goal of these pages is to demonstrate the inferential nature of all our cognitions, and, consequently, to dispose of any search for an absolute, primary source of knowledge as a pseudo problem. In this light, he confronts a series of issues that usually are resolved in intuitionist terms: in particular, the concepts of self-consciousness, the distinction between perceptions and dreams, between beliefs and ideas, the knowledge of the inner world and the very notion of thought--all explained as inferences drawn from "observation of external facts".

The matter of self-awareness is of immediate interest to us. Peirce quickly delimits the problem by specifying that, by "self-consciousness ", he means the recognition of one's own "private self"[10] and that the question is whether it "can be explained by the action of known faculties under conditions known to exist" rather than by the uncertain principle of an intuitive faculty of self-consciousness. His argumentation, starting on a psychological note, will show how self-consciousness constitutes, not the fruit of an immediate intuition, but the result of a particular phase of the individual's psychic development, that of language learning and of the interpersonal comparison that accompanies it.

It is interesting to note that Peirce makes the emergence of self-consciousness coincide with a shift in our way of understanding sensorial experience. In a way analogous to that of Wittgenstein, he maintains that it is not primarily lived as an "inner" experience. The child experiences it, first of all, as a demonstration of the actual relationship between objects and their properties: "No one questions that, when a sound is heard by a child, he thinks, not of himself as hearing, but of the bell or other object as sounding"; or as the relationship between the objects and its own body: for example, the property of movement that the child experiences as belonging to a certain object will be related to its ability to interfere physically with it, and certainly not to something internal, like the desire to move it (CP 5.230,231). The decisive moment for achieving a conception of sensorial experience as inner experience as well is represented by language learning, and especially by the interpersonal communication that will lead a child to distinguish "facts" from "appearances":

A child hears it said that the stove is hot. But it is not, he says (...); but he touches it, and finds the testimony confirmed in a striking way. Thus, he becomes aware of ignorance, and it is necessary to suppose a self in which this ignorance can inhere. So testimony gives the first dawing of self-consciuosness (CP 5.233).

Moreover, there is a whole series of the child's statements - consisting of its tendency to transfer the functions of its own body to objects - that are destined not to find any confirmation either in the testimony of others or by experiential verification:

Thus, he adds to the conception of appearance as the actualization of fact, the conception of it as something private and valid for one body. In short, error appears, and it can be explained only by supposing a self which is fallible (CP 5.234).

The connection between the experience of error and self-awareness gives origin to a key principle in Peirce's philosophy: the sense and value of expressions of the subjective consciousness are defined only in relation to what is "external" to it, that is, in relation to the meanings that are determined by inter-subjectivity and to the "objective facts" that the community recognizes and develops. We cannot adequately discuss these topics here; we would only recall that the reference to "exteriority" represents in Peirce more a methodological norm than a self-sufficient epistemic criterion.

The argument of self-awareness is, in that sense, especially significant, above all for what concerns the inter-subjectivity criterion and the possibility for knowledge of the inner world it can offer. Although it is, in fact, the condition that gives rise to consciousness of one's own self and of its inherent fallibility, the testimony of others seems to take on a priority that, however, proves to be simply psychological in nature. Peirce says that it becomes "a stronger mark of fact than the facts themselves", specifying that "this remains so through life", but he also adds: "testimony will convince a man that he himself is mad" (CP 5.233). He thus underscores that, besides the inescapable nature of inter-subjective agreement, its establishment, in other words, as an essential and constant point of reference for the individual's cognitive experience, there is a risk of non-truthfulness as well, since it carries with it a strong psychological tendency to conform to the opinions of others.

Another interesting aspect, for purposes of a comparison with Wittgenstein, is the statement that any "internal fact", though belonging to consciousness, still imply a relation with something external, and it come to be known precisely as "predicate of something external". Under this heading, Peirce lumps together both sensorial perceptions and emotions. It is true - he says - that every perception is also a fact determined by "internal conditions" and, for example, "the sensation of redness is as it is, owing to the constitution of mind; and in this sense it is a sensation of something internal". But the cognition we have of it is de facto "an inference from redness as a predicate of something external" (CP 5.245).

These meager remarks summarize some of the findings in the text, A New List of Categories[11], which dates from not much before the text under consideration. There sensorial experience is seen as an event lacking per se any cognitive value, one that takes on logical importance only when, by means of the processes of differentiation and abstraction, it is translated into perceptive judgments, that is, in the predication of certain "qualities" in the objects tried. Moreover, Peirce soon will specify, in his essay, Some Consequences of Four Incapacities[12], that the "constitution of our mind" is a datum that logical processes cannot gather, an "arbitrary" datum, which means exactly "non-rational" in its double sense of the non-analyzable limit and originating condition for every logical movement (CP 5.291). From this point of view, the "inner" aspect of sensorial experience is none other than its naturalistic substratum.

On a par with the sensations, so, too, the emotions, which at first sight seem to refer only to the self, present a relation with something external: for instance, if a man is angry, "it can hardly be questioned that there is some relative character in the outward thing which makes him angry, and a little reflection will serve to show that his anger consists in saying to himself 'this thing is vile, abominable, etc.," and that it is rather a mark of returning reason to say 'I am angry'" ( C P .5. 247, see also 5. 292). Furthermore, as predicates of some external object, the emotions do not differ from intellectual judgments, if not to the extent that they prove in the end to be "relative to the particular circumstances and dispositions of a particular man at a particular time" (CP 5. 247; see also 5.292)[13]. The emotions thus enter into the cognitive experience circuit: the recognition of their predicative nature - of which the individual was unaware at the time - is equivalent to moving from a purely emotive state - where the object of awareness is only an immediate feeling - to the level of thought and of meaning. Only on this level will it be possible to identify the specific object of an emotion, which - as will be clarified in another context - is, in itself, a particular form of feeling: a simple "immediate presence". As such, it cannot be grasped cognitively. Knowledge and thought are carried out "over time" and imply reflective processes that yield immediate experiences only in the shape of the meaning that their representations carry (CP 5.258; CP 5.288-89). And this is the level at which Peirce's concluding statement must be considered: "There is no reason for supposing a power of introspection: the only way of investigating a psychological question is by inference from external facts" (CP 5.249).

It is not by chance that one speaks here of "investigating", because, in reality, this is not a matter of pointing out a sure way to knowledge, but one of referring to more concrete criteria than the undemonstrable principle of immediate introspection and, above all, of also acknowledging the inherent fallibility of any logical-inferential process. If one keeps in mind that, in Peirce's philosophy, representations and meanings have logical validity mainly to the extent that they are recognized inter-subjectively, we can see in this expression precisely this latter aspect. The understanding of the relationship between emotion and external object coincides, in fact, as we have seen, with the reactivation of the "telling" that the emotional state has interrupted, hence with the return to language activity, which is precisely a "return to reason". And the moment of language, by definition, cannot consist in a purely subjectivist logic; rather, it implies a relation with the "external" and, in the case at hand, a "public confirmation", a "recognition" by others of the explanations we give from time to time for our emotional states. But the fact that all this does not guarantee any absolute certainty is implicit in the limitation of the epistemic value of the testimony emerging from the distinction between "facts" and "appearances" indicated earlier[14]. In essence, as for Wittgenstein, knowledge of our inner world is, for Peirce, not only indirect, that is to say, not guaranteed by any immediate criterion, but it also is always approximate, because of the impossibility of our gathering their "profound" aspects, the "true" and "essential" ones that constitute them.

I think it useful to recall, in closing these observations, that this line of vision regarding "internal facts" is not, in any event, the sole case where the two philosophers seem to be significantly close to each other. Numerous comparisons can be made at various levels of their thought. Pronounced affinities, for example, can be traced in the epistemic distinction between believing and knowing, and concerning the subject of doubt. Let it suffice here to recall the passages in On Certainty where Wittgenstein, examining the relationship between knowledge and belief, maintains, among other things, that "knowing" is not a mental state that can be distinguished from "believing" on the basis of presumed guarantees of "clear and distinct" convictions, since, in reality, the state of mind of conviction can be the same, whether one knows or believes falsely[15]. On his part, Peirce attacked the introspective method of Descartes, noticing that "the distinction between an idea seeming clear and really being so, never occurred to him" (CP 5.391) and then he decisively rejected any distinction between true and false that appeals to a feeling of conviction, also because, according to him, each belief, by definition, implies a mental state of satisfaction, whether it is true or false (CP 5. 372). As for doubt, precisely like Wittgenstein, Peirce also maintains that it requires specific motivations ("There must be - he says - a real and living doubt, and without this all discussions are idle"); moreover, it does not make sense to assume doubt as a universal methodological rule; and finally doubt necessarily carries some consequences as well as presupposes some certainties[16]. These and other correspondences, certainly less explicit, seem somewhat surprising if one considers that we have no record of any contact between Wittgenstein and Peirce's wrigtings. Moreover, despite the importance that exquisitely pragmatic notions take on in the texts following the Tractatus, Wittgenstein apparently had no more or less direct dialogue with other pragmatists, with the exception of W. James. A comparison with their positions and, in particular, with that of Peirce would perhaps merit more space than it has obtained up to now, if for no other reason than to contextualize Wittgenstein's philosophy in the spirit of an epoch of great cultural transformations, of which it certainly is an eminent expression, but of which Peirce himself was anything but a secondary representative.