A Note on the Late Wittgenstein’s Use of the Picture Concept
A Note on the Late Wittgenstein’s Use of the Picture Concept


In his post-Tractatus writings, Wittgenstein continued to employ the analogy between the picture and the sentence. In 1936 he wrote: “The sentence as picture. And that is very pleasing, but [[d]er Satz als Bild. Und das ist recht schön, aber] there are still lifes, portraits, landscapes, depictions of mythology [Y] maps, diagrams etc., etc.” (MS 116, 223). This paper focusses on the uses of the picture concept in the late Wittgenstein and will emphasise in particular its new application in the description of religious language use.

Table of contents


    In the following I wish to draw your attention to two related ideas that occur in Wittgenstein’s later writings. In making this emphasis I am at the same time claiming a certain continuity in Wittgenstein’s thought – a continuity of a quite particular kind. The argument that I shall present in the following can be summarised under three points: 1. in both his early and his late writings, Wittgenstein makes a natural-historical claim that, as humans, we are picture-creating and picture-using creatures; 2. the crucial analogy between the picture and the sentence that appears in the Tractatus Logico-Philosophicus is retained in Wittgenstein’s later descriptions of language; and 3. the use of this analogy serves two diametrically opposed purposes when considered in relation to religious language, whereby the earlier use determines the propositions of natural science and delimits these from religious propositions, and the later use of the analogy provides the impetus for a grammatical investigation of religious language and religious beliefs.

    1. Man – a picture-using creature

    In neither the Notebooks 1916-1918 nor in the Tractatus does Wittgenstein offer any argument for, or investigation of, the particular quality or ability that enables human beings to perform the experimental putting together (TLP 4.031) or constructing (TLP 4.023) of names in accordance with the criteria for logical description implied by the theory of representation. Instead, he simply affirms, in November 1914, that „[w]e can make ourselves a picture of it [a situation] [Wir können uns ein Bild von ihm machen]“ (NB, 24). The wording of this remark allows us to regard it as a crucial premise underlying the theory of representation in the early philosophy. In essence, this premise is that humans are by definition picture-using creatures. This implies in turn that we are naturally inclined to create language, whereby the essential quality of language is assumed to be that it depicts the things that exist and the ways in which these are mutually related to one another. It is assumed that this kind of representation is achieved in virtue of the human ability to create such pictures in accordance with the criteria for logical representation. This premise is put forward anew four years later in the introductory remarks that summarise the representational theory of the picture (TLP 2.1-2.12). – „We make to ourselves pictures [[w]ir machen uns Bilder] of facts“ (TLP 2.1). Here as well the premise has the form of a natural-historical assertion to the effect that, in anthropological terms, humans are picture-creating creatures. Somewhat earlier, in October 1913, when also framing linguistic philosophical reflections within a natural-historical perspective, he wrote: „Man possesses an innate capacity for constructing symbols“ (NB, 100; TLP 4.002). The use of the verb „make/machen“ in the two assertions just quoted encapsulates the account of sentence formation as an experimental putting together or constructing of names. The verb emphasises the human activity of creating pictures. With this emphasis in mind, it is worth noting that the corresponding summary of the picture theory in the Prototractatus implies a broader natural-historical field, for in that context Wittgenstein also asserts that humans perceive, understand and grasp propositions and facts in the world as (logical) pictures; or, as he puts it: „Facts are grasped by us [begreifen wir] in pictures“ (PT 2.1). This also means that, in more general terms we should speak of humans as picture-perceiving creatures. It is these natural-historical facts that Wittgenstein presupposes despite his failure to investigate the question in depth: what is it that enables us to form pictures (thought as logical picture (TLP 3); proposition as (logical) picture (TLP 4.021)) of states of affairs?

    Wittgenstein returns to the picture concept again after 1929, when it is integrated into and modified to suit his new account of language. One of the more prominent formulations in this regard is that, among the countless and diverse uses of language, there are certain forms of expression that can be described or classified as pictures. This claim is worded in a variety of ways that could in fact be confused with the earlier formulations of the assumed natural-historical premise that underlies the picture theory. In MSS 116 and 120, for example, he says that many of the difficulties (Schwierigkeiten) that we encounter and have to contend with in philosophical work result from the situation „that we make pictures for ourselves [dass wir uns Bilder machen]“ (MS 116, 162; MS 120, 26); these pictures often tempt or encourage us to expect that words have some other use than they commonly have in everyday life. In the first half of MS 115 he says more generally: „To make pictures for ourselves is part of our life [uns Bilder zu machen ist Teil unseres Lebens]“ (MS 115, 93).

    Judging by this kind of formulation, it would appear that Wittgenstein adhered to his natural-historical premise, that we are picture-creating, or picture-using (Z '381) creatures. The reason for saying it would appear so is not least that the premise is only put forward in a weakened form; on the one hand, the forms of expression that are described as pictures cannot be equated with the pictures of representational logic of the Tractatus, and on the other, since the thesis only applies to certain or particular forms of expression, it cannot be regarded as extensively applicable to all language uses, as was the case in the early philosophy. Thus the premise is retained but only in a limited sense, insofar as it is posited in relation to a characteristic of certain elements or features of language use. – „The various positions [[d]ie verschiedenen Stellen] that a picture can assume in a theory (or in any kind of language game)“ (MS 116, 226). – Features that are closely associated with other forms of language, such as physical gestures and reactions. In several contexts Wittgenstein is more precise in specifying the principal elements or features of language that the picture concept relates to, and these happen to be a relatively limited range of expressive forms together with their ideas and conceptions. I shall mention a few examples: „It is quite clear that the way of talking about [die Ausdrucksweise vom] memory as a picture is merely a picture“ (MS 108, 33-34); „[A] certain picture, a certain idea [[E]in gewisses Bild, eine gewisse Idee]“ (MS 115, 205); „A new arithmetical technique should provide us with a new picture, a new form of expression [ein neues Bild liefern, eine neue Ausdrucksweise]“ (MS 121, 129); „You have a picture. (A form of expression.) [Du hast ein Bild. (Eine Ausdrucksweise.)]“ (MS 121, 3).

    2. The analogy – picture and sentence

    With this focus in mind it becomes relevant to ask, what is it about such forms of expression that permits us to regard them as pictures? Or, what is the basis for retaining the analogy between sentence and picture in the late works?

    At no point does Wittgenstein argue in any depth for reintroducing the picture concept, and in a sense it might seem surprising that he should retain such a central element of his early philosophy, since many of the ideas that are presupposed by this reuse of the concept stand in conflict to the earlier ideas about logical representation. No justification for the reintroduction of the concept is given in Philosophical Investigations, although one is sketched in the Big Typescript. In a brief section of Chapter 21 entitled „Ähnlichkeit von Satz und Bild“ he says: „In what sense can I say, the sentence is a picture [[i]n welchem Sinne kann ich sagen, der Satz sei ein Bild]? When I think about it I wish to say: it must be a picture if it is to show me what I should do, if it is to let me act in accordance with it. But then what you want to say is that you act in accordance with it in the same sense in which you act in accordance with a picture. [Y] To say that the sentence is a picture highlights certain features in the grammar of the word ‘sentence’. [Zu sagen, dass der Satz ein Bild ist, hebt gewisse Züge in der Grammatik des Wortes „Satz“ hervor.]“ (TS 213, 83).

    Apart from the point that the reintroduction of the concept serves a heuristic purpose, Wittgenstein claims that the basis for retaining the picture concept is the insight that what is entailed in understanding, acting upon or acting on the background of a sentence is closely related to what is entailed in understanding, acting upon or acting on the background of a picture. To put it another way, our attitude towards and our use of certain forms of expression in language is in many respects analogous to our attitude towards and use of pictures. But can we say more precisely what aspects of approach and use (of sentence and picture) it is that are being equated? Only a few are suggested. Two are mentioned in the Big Typescript: (a) a semantic aspect: in many contexts the semantic content of a picture is like that of a linguistic description – „I can translate the description of the garden into a picture [ein gemaltes Bild], and the picture into a description“ (TS 213, 83); and (b) a functional aspect: in many contexts a picture can prompt the same reactions as a linguistic description – „When the ugliness of a person affects us as repellent, it can do so in a picture, just as it can in a description, in words“ (TS 213, 86). A third point of comparison is found in Zettel: (c) a discursive aspect: a picture, like a sentence, allows certain inferences or conclusions to be drawn – „Of course a picture in which the face smiles does not shew how it looks when weeping. But it does permit inferences“ (Z '514). And finally there is (d) an explicative aspect: in many contexts a picture can play the same role as a linguistic explanation – „Take ‘God created Man’. Pictures of Michelangelo showing the creation of the world. In general, there is nothing which explains the meanings of words as well as pictures“ (LCA, 63).

    The linguistic forms of expression – or of certain combinations of linguistic forms (PG, 170) – which Wittgenstein refers to as pictures can more generally be divided into two main groups: (A) particular linguistic forms of expression (pictures): here we are concerned with the afore-mentioned combinations of linguistic concepts, whose scope of application is limited to rather special or narrow fields of use; in this context, the picture (concept combination) refers to, for example, (Ai) a certain imaginative conception, such as there being a logical connection between a person’s handwriting and that person’s character (CE, 410-411; LFM, 239), and (Aii) a comprehensive or synoptic overview, such as a historic summary of developments in mathematics, which – as he puts it in the Investigations – can give „a rough picture“ (PI '23) of the conceptual changes.

    But following this, we find another main group, where the analogy between the sentence and the picture is used in conjunction with (B) constitutive or organisational linguistic forms of expression (pictures). Here we are dealing with combinations of linguistic terms that determine others, or that determine the formation of new linguistic combinations; in this case, the linguistic combination (the picture) is, as it were, decisive for the forms and categories of thought and is therefore to be regarded as a conceptual frame, which determines partly the way in which the world is perceived and comprehended (the picture establishes certain habits of thought), and partly the space within which criteria are determined for the meaningfulness of a term’s application. Thus we are concerned here with those combinations of terms (pictures) that broadly constitute, on the one hand, a (normative) conceptual perspective („eine Denkform“ (TS 220, 72)), and on the other, the ultimate (internal) authority for that conceptual perspective (justificatory authority). Our language and inherited habits of thought (Denkgewohnheiten) should therefore also be regarded as having developed in conjunction „with the oldest pictures [mit den ältesten Bildern]“ (TS 211, 398). As examples of such constitutive or organisational pictures – „a picture which is at the root of all our thinking [ein Bild am Grunde alles Denkens]“ (CV, 83) – we are given Augustine’s account of language (PI '1), the definitions of mathematics (LFM, 112) and its proofs (LFM, 55), and natural scientist’s or physicists theories about the most basic constituents of the world. In relation to the latter, Wittgenstein writes: „You are taking for granted the atomistic picture. What does this come to? We are so used to this picture that it’s as thought we had all seen atoms. Every educated eight-year old child knows that things are made of atoms. We would think it lack of education if a person didn’t think of a rod as being made of atoms“ (LCA, 17).

    3. A picture – religious language use

    In his „Lectures on Religious Belief“ Wittgenstein links his grammatical description of religious belief to the last of the concepts of picture mentioned above. He emphasises that the representational or expressive form of religious language can be regarded as pictorial. Thus he describes as pictures linguistic combinations such as „the Last Judgement“ (LCA, 53), „God’s eye sees everything“ (LCA, 71) and „we might see one another after death“ (LCA, 70-71).

    These lectures would seem to constitute an attempt to rethink the distinction between the religious and the scientific use of concepts that formerly appeared in the Tractatus and in „A Lecture on Ethics“. And as before, he points out the irreconcilability not just of the attitudes, but also of the linguistic practices and the orientations towards reality that apply in the two forms of discourse. The religious approach and sense of conviction are not founded on scientific method (in an arrangement of hypotheses and verifications, in the use of induction and of terms such as rational, assumption, probability, knowledge, facts), but rather in unshakeable faith, which manifests itself in very different linguistic practices. But in what way? And what are those linguistic practices? Wittgenstein says: „Suppose somebody made this guidance for this life: believing in the Last Judgement. Whenever he does anything, this is before his mind. [Y] [H]e has what you might call an unshakeable belief. It will show, not by reasoning or by appeal to ordinary grounds for belief, but rather by regulating for all in his life. This is a very much stronger fact – foregoing pleasures, always appealing to this picture“ (LCA, 53-54). In short, religious faith manifests itself in certain practices that appeal to and are regulated by concepts (linguistic combinations) that can be described as pictures.

    In the „Lectures on Religious Belief“ he presents neither arguments nor even any basic justification for introducing the picture concept, nor any detailed account of it; but the mention of the role and function that the picture assumes allows us to infer that he is thinking here of the above-mentioned constitutive or organisational aspect of linguistic pictures. It is linguistic combinations that underlie the thoughts and actions of the religious believer. When we study the picture concept in these lectures more closely, it appears to possess three essential characteristics. (i) Religious concepts and statements (pictures) do not refer like empirical sentences (the pictures constitute referential linguistic actions that are different in kind from ostensive definitions). If we were to show Moore some pictures of a tropical plant, at least there is in this case „a technique of comparison between picture and plant“ (LCA, 63) that we could resort to. But there is no technique of comparison when it comes to Michelangelo’s picture of God creating Adam. (ii) Combinations of religious terms are pictures that cannot be justified. In MS 121 he writes: „You have a picture. (A form of expression.) But you cannot justify it. In trying to grasp something behind the form of expression, you grasp at thin air. [Wie Du hinter die Ausdrucksweise zurückgreifen willst, greifst Du in’s Leere.]“ (MS 121, 32). And (iii) religious terms and statements (pictures) cannot be reduced to or reformulated as emotive or theoretical forms of expression, since „[t]he whole weight may be in the picture“ (LCA, 72). The semantics of pictures in the religious context cannot be replaced by or transformed into other forms of expression. The pictures are untranslatable. Wittgenstein suggests these three characteristics – the picture’s lack of empirical reference, its lack of epistemological foundations, and its irreducible nature – when he writes: „Religious similes can be said to move on the edge of an abyss [am Rande des Abgrundes]. [Y] Religion says: Do this!Think like that! – but it cannot justify this“ (CV, 29). „The similes of the NT [Y] are without foundation [[d]ie Gleichnisse des N.T. [Y] sind ohne Boden [var.: ohne einen Boden]]“ (MS 162b, 128).

    But haven’t we seen these characteristics somewhere before? Indeed we have – in the description in the Tractatus of the domain from which the picture concept – as it appears in that work – is bounded off, namely the mystical, that which lies beyond all empirical happening and being-so (TLP 6.41), and which eludes any form either of meaningful conceptualisation or description, or of empirical or theoretical justification (TLP 6.522). So, to put it in a nutshell, the picture concept is a pivotal element both in the Tractatus’ logical description of natural-scientific sentences (that stand in contrast to non-empirical religious statements), and in the late Wittgenstein’s grammatical investigation of religious sentences (that stand in contrast to empirical natural-scientific statements). The analogy between the picture and the sentence is used in the Tractatus and in the late works (post 1929) in two distinct linguistic domains; the late use of the analogy in the „Lectures on Religious Belief“ simply lays claim to the characteristics from which the picture concept in the Tractatus was delimited!

    This is one of the many shifts in his Denkbewegungen that Wittgenstein has in mind when in 1944 or later he noted: „My account will be hard to follow: because it says something new but still has egg-shells from the old view sticking to it“ (CV, 44).*

    Main references

    1. Wittgenstein, L. 1953 Philosophical Investigations, Oxford: Blackwell. (PI)
    2. Wittgenstein, L. 1958 Tractatus Logico-Philosophicus, London: Routledge. (TLP)
    3. Wittgenstein, L. 1966 Lectures and Conversations on Aesthetics, Psychology and Religious Belief, Oxford: Blackwell. (LCA)
    4. Wittgenstein, L. 1980 Culture and Value, Oxford: Blackwell. (CV)
    5. Wittgenstein, L. Nachlass (the Cornell copy version & the Bergen machine-readable version). (MS & TS)
    My thanks to Peter Cripps for his translation of this paper.
    Peter K. Westergaard. 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.