As Far as the Eye can Reach: Complete Analysis in the Intermediate Wittgenstein
As Far as the Eye can Reach:
Complete Analysis in the Intermediate Wittgenstein


From Wittgenstein’s calculus period to the language games period, the gradations internal to the multiple forms of phenomena migrate to the criteria themselves. The applicability of the calculus metaphor weakens, as the vision metaphor gains in importance. In this context, can one still speak of complete analysis? Yes, but with some qualifications. This paper addresses these qualifications.

Table of contents

    Conceptual analysis is one of the notions undergoing dramatic changes during Wittgenstein’s intermediate period. There’s an aspect of such changes which might prove characteristic of the new directions taken by Wittgenstein’s thought after 1929: it is the notion of complete analysis. We believe that Wittgenstein’s treatment of this notion is also an interesting indication of new concerns generated by the “new method” itself, in terms of its radical openness to dialogism and its anti-dogmatism. Our aim here will be to explore a couple of hints as to why this can be so.

    In Wittgenstein’s ‘calculus’ period, the work of conceptual analysis draws its elements from systems of rules in which the transitions between cases occur in the manner of a special kind of automatism (cf. PB §82; MS 209, p. 34). The inferential necessity that is internal to one system of rules, by marking a point in a transition, marks the other ones negatively – or, in a latter idiom: when making a move in the game, one makes all the other moves at once. Somehow I state that I haven’t bought five peaches, if I declare having bought only three. When the notion of language games is introduced, the aspect of gradation of those systems is not eliminated: this aspect is transferred to open systems. We have, then, two steps. The first step is that of the pluralization of the logical form in the multiple forms of phenomena. This is the aspect through which the notion of rules is introduced in the early 30s. In this regard, professor Arley Moreno writes:

    Each elementary fact is pictured within the system, i.e., by the exclusion of all the other elementary facts – and the proposition is now conceived as a system of representation, or better, as the closed set of all elementary propositions relative to a certain phenomenon of perception bearing gradation. (…) Although the central concern [of the Tractatus] remains in place, namely the idea of logical form, this latter, however, multiplies itself in the diverse logical forms of phenomena. (2007, p.71)

    The second step is taken when the closure criterion of the systems of representation of phenomena bearing gradation is pluralized. In what sense, then, could we say that the idea of gradation is not completely abandoned? In the sense that the gradations now bear on the criteria themselves. The criteria now group themselves in families, whose members are then recognizable as such. However, the transitions have therein a less uniform regularity than in the case of the closed systems of rules.

    What a strange idea – a less uniform regularity! Nevertheless, it becomes less offensive if we think of a system in which some intermediary spaces are wider than others, without thereby ceasing to be part of the same system. Figures of organic tissues occur to the mind, rather than figures of sets of gears. The family relation between chess and checkers is maintained when we insert in the list of relationships, say, the game of racket-ball – but in this case the relationship’s intermediary links have to be a bit more stretched, so as to be based in comparative traits which are less easily surveyed. Which means that these comparative traits fit less easily in the idea of an inferential automatism. Moreover, when we look perspicuously at how systems of this kind overlap, the nexuses become increasingly vaguer, more fluid, as we distance ourselves from what we could call their ideal model, which constitutes a sort of focus of view when we examine them – a sort of caricature. This doesn’t mean that there isn’t, at some point, a passage out of the system; but the identification of the passage out of the system is made in a manner which is constitutively vague, such as in the case of the limits of the scope of concepts like ‘memory’ and ‘visual field’. In short: vague – but adequately delineable nonetheless. We can find at this point an aspect of what could well be characterized as the realm of completeness of conceptual analysis in the language games period. And at the same time it becomes clearer the reason for the calculus metaphor to weaken over the years.

    More than ‘calculus’, the operative metaphor is, here, that of vision. But it must be understood not as a new problem, but in its cooperative contrast with the notion of thinking (PU §66). The question is one of seeing what is expressive without letting oneself be seduced by the idea of what is not expressive but is supposed to be, although one doesn’t see how it could.

    Nevertheless, we seem to have a tendency to mistrust grammar, in the sense of thinking and not just seeing. There we recognize the attraction of knowledge based in a systematic answer to the skeptic, in the way of the hyperbolic doubt. “Is no demon deceiving us at present?”, asks Wittgenstein in the Remarks on the Foundations of Mathematics”. “Well, if he is, it doesn’t matter. What the eye doesn’t see the heart doesn’t grieve over” (BGM, II, §78). To be enthralled in an infinite answer to the skeptic is something that stands in the way of the vision of certain consequences of the private language argument (which is crucial to the philosophical use of the metaphor of vision), as Baker & Hacker have shown (1984). 1. It does not make sense to interpose, between recognition and judgment (or assertion), a Scheinwesen, a mysterious stage of private recognition of an object or property, for which a new rule would then be needed, etc. 2. The connection between rule and case is made by way of a training and also a habit.

    It is the private language argument that allows to accommodate a notion of complete analysis to the expansion of the realm of conceptual analysis, by means of a radicalization of the role of rules in the grammatical regulation of experience. This radicalization of the role of rules, by the way, allows to accommodate within conceptual analysis not only the cases whose criterial limits are vague, but even those in which no rules are identified by the philosopher. This stems from the fact that, now, a line of kinship between games can be established up to a point where the very boundaries between what is meaningful and what is not become extremely vague – which is something intolerable within the framework of the calculus period.

    Doesn’t the analogy of language to a game enlighten us here? We can very well imagine that people in a meadow might entertain themselves by playing with a ball; and that they might do this in such a way that they’d begin various existing games one after the other without finishing them, and in between would even throw, catch, drop the ball, etc., aimlessly. Now someone might say: These people are playing a ball game the entire time, and that is why at each throw they comply with specific rules. (…) – We can say: Let’s investigate language with regard to its rules. If here and there it doesn’t have any rules, then that is the result of our investigation. (TS213, p.254r, in N)

    There’s a threefold therapeutic process of this passage, the source of Investigations §83. First, the investigator assumes an unhaunted view in regard to the possibility of a domain which is supposedly without rules. The gain here is, once again, to elude theoretic hypostases populating such dreadfully empty domain: where there is a simple acting-like-that, we postulate an experience, and the postulation is then projected onto practice like a fog covering a landscape, allowing for glimpses of isolated, uncharacteristic areas. Second, the investigator releases his view to the determinations effectively showing up in the games but just up to the point in which they show up to us, the point where we can see them. Why? Because the investigator is no longer focused in the search of some automatism of the transitions or inferences. He is no longer excessively focused on refuting the agostinian referentialist, opposing to him the autonomous field of precise contours of the closed rules systems.

    Finally, the negative results of the lineament of a rule-less field amidst experience are cheered. Reflecting latter on method, Wittgenstein states an approbation of a similar idea:

    It seems to me I am still a long way from understanding these things; from the point, that is, at which I know what I have to talk about, & what I don’t need to talk about. I still keep getting entangled in detail without knowing whether I ought to be talking about such things at all; & I have the impression that I may be inspecting a large area, simply to exclude it eventually from consideration. But even in this case these reflections would not be worthless; as long, that is, as they are not just going round in a circle. (VB, p.74; MS136, p.37a, 25.12.1947)

    The risk of taking paths that lead to nowhere is inherent to the anti-dogmatism of the new method. But not as a defect of it – in the same sense in which the sign of a dead-end street is not useless to the orientation of a lost person. On the other hand, it is also inherent to the anti-dogmatism of the new method the recognition that “not everything that can be said was said”, that “not everything that can be said was circumscribed” (Salles 2006, p.52). Sometimes, the analyst is unable to recognize if something that can be said was said or not – or, in the last analysis, even if something was said tout court. A supposedly random gesture can have a meaning, inasmuch as a view is abandoned that demands from it a meaning it must have. There is a
    subtle aspect of an attitude stated in the Philosophical Grammar that is relinquished here: that of considering that nothing is possible (i.e., no philosophical investigation) before the distinction between sense and nonsense (PG, VI, §81). In a way, a closer look at the use of symbolism in the clearly established context of continuous, or open systems of rules renders the precise identification of that distinction a more delicate matter, to say the least. It is not a question of diminishing the importance of that distinction, on the contrary: it is a question of turning it into an internal part of the work (instead of a starting point of the work, a sort of philosophy degree zero) precisely because it is important. An investigation on the limits of sense that ceases to dogmatically postulate the establishment of those limits now assimilates, in its continuous practice, that very establishment.

    So a complete logical analysis is possible and desirable. But it is not complete in some dogmatic sense, i.e., in some sense of closure and ultimity; such would be the case if it rendered the complete grammar of the concept under exam. This sense of completeness is not given by the philosophical work, nor does the work intend to. When we look without thinking to the working with concepts, we do not know yet exactly in what way, and at what point, an analysis reaches its end. But we know that it must find this point in a logical now, so as to fulfill its function, in a localized manner, but whole nonetheless. This can be seen as an observation on method: the logical analysis in the context of the language games must be seen as an antidote against the bewilderment of the reflection on language – but stopping there, not taking the remedy too far (by extending, for instance, the reach of a clarifying comparison beyond the point of its utility). Another way of saying this is to say, with Arley Moreno (1995), that the therapy of the images does not aim at dissolving them as images, but purports to modify our attitude in relation to those images whose very necessity we consider necessary. It is to that end that the conceptual analysis strives to show the non-necessary of necessity in a complete manner.

    The “new method” is not iconoclastic: “(All that philosophy can do is to destroy idols. And that means not creating a new one – say in the “absence of an idol”)” (TS 213, p. 413, in N). The description of our language games does not purport to improve them or amend them, in a virtually infinite work, but to exhibit its internal determinations. The perspicuous presentation exhibits this determinations in the context of a suggested comparison. Such operation must be completed in each case. For clarity as an intrinsic value is the great hope of the “new method” to bring about the complete dissolution of a disquieting of the mind that paralyzes it, preventing thought from moving on in the mode of a und so weiter. But this sense of completeness is not made into a new idol. It is just the right tool for the appointed job of the philosophical business. As far as the eye can reach.


    1. Baker, Gordon & Hacker, P.M.S. 1984 Scepticism, Rules and Language, Oxford: Blackwell.
    2. Moreno, Arley R. 1995 Wittgenstein: através das imagens, 2nd ed., Campinas (Brazil): Editora da Unicamp. [Wittgenstein through the images]
    3. ––– 2007 “Pensamento e realidade: em direção a uma pragmática filosófica”, in Moreno, A.R. (org.), Wittgenstein – Aspectos pragmáticos, Campinas (Brazil): Unicamp/CLE (Coleção CLE v.49), pp. 55-94. [“Thought and Reality: Towards a Philosophical Pragmatics”]
    4. Salles, João Carlos 2006 “A caixa de gordura”, in O retrato do vermelho e outros ensaios, Salvador (Brazil): Quarteto, pp. 31-56. [The Portrait of Red and other essays]
    5. Wittgenstein, Ludwig (BGM) Remarks on the Foundations of Mathematics / Bemerkungen über die Grundlagen der Mathematik, ed. bil., ed. G.H. von Wright, R. Rhees e G.E.M. Anscombe, tr. G.E.M. Anscombe, Cambridge: The M.I.T. Press, 1967.
    6. ––– (LC) Wittgenstein’s Lectures - Cambridge, 1932-1935, ed. Alice Ambrose, NY: Prometheus Books, 2001.
    7. ––– (N) Wittgenstein’s Nachlass – The Bergen Electronic Edition, Oxford U.P./Univ. of Bergen, 2000.
    8. ––– (PB) Philosophische Bemerkungen, ed. Rush Rhees, Frankfurt: Suhrkamp, 1970.
    9. ––– (PG) Philosophische Grammatik, ed. Rush Rhees, Frankfurt: Suhrkamp, 1973.
    10. ––– (PO) Philosophical Occasions 1912-1951, ed. J. Klagge e A. Nordmann, Indianapolis: Hackett, 1993.
    11. ––– (PU) Philosophical Investigations / Philosophische Untersuchungen, 3rd. bil. ed., ed. R. Rhees e G.E.M. Anscombe, tr. G.E.M. Anscombe, Blackwell, 2001.
    12. ––– (TLP) Tractatus Logico-Philosophicus, 2nd. bil. ed., tr. and presented by L.H.L. dos Santos, São Paulo: Edusp, 1994.
    13. ––– (VB) Culture and Value / Vermischte Bemerkungen, 2nd. bil. ed., ed. G.H. von Wright, rev. Alois Pichler, tr. Peter Winch, Blackwell, 1998.
    Rafael Lopes Azize. Date: XML TEI markup by WAB (Rune J. Falch, Heinz W. Krüger, Alois Pichler, Deirdre C.P. Smith) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.