A defence of the austere view of nonsense

Abstract

The austere view of nonsense says that the source of nonsense is not a violation of the rules of logical syntax, but nonsense is always due to a lack of meaning in one of the components of a sentence. In other words, the necessary and sufficient condition for nonsensicality is that no meaning has been assigned to a constituent in a sentence. The austere conception is the key ingredient of the resolute reading of Tractatus Logico-Philosophicus that presents a therapeutical interpretation of the work, and rejects a possibility of conveying some ineffable truths by means of nonsensical sentences. In this paper, I defend the austere view against the most important objections. Firstly, I discuss textual sources for ascribing to Wittgenstein the austere view that have not been sufficiently thoroughly discussed yet. Next, I indicate that the apparent occurrence of words out of the context of a sentence on lists of verbs and in definitions is not in conflict with the version of the context principle recommended by austere theorists, since this argument does not sufficiently take into account the distinction between sign and symbol. Furthermore, I argue that the austere conception and the principle of compositionality are jointly tenable, for the former view does not exclude the existence of syntactic and semantic rules, but only shows that they are recognizable conditionally. Finally, I indicate that the austere view of nonsense should be seen as a viable option by a contemporary theorist of nonsense, for it helps us to avoid making controversial assumptions about the nature of meaning and content.

1 Introduction

There are two main positions concerning what nonsense is in Ludwig Wittgenstein’s Tractatus Logico-Philosophicus. More traditional and well-known is the substantial view of nonsense. It is the view holding that the main source of nonsense is a clash in the logical categories of words. That view of nonsense has been criticized and rejected by the proponents of the austere view of nonsense. The adherents of the austere conception claim that nonsense is never a matter of category mismatch, but it always results from the absence of meaning in one of the components of a sentence. The exchange concerning the correct view of nonsense is at the heart of the debate between the standard readers of the Tractatus and the resolute readers of this work.¹ The standard reading has it that the main aim of the work was to pass on some ineffable truths concerning the nature of the world, language, logic and ethical life (Anscombe, 1965; Baker & Hacker, 2005; Geach, 1976; Glock, 1996, 2015; Hacker, 1986, 2000; Pears, 1987; Proops, 2001; White, 2011). That position would be false without the substantial view of nonsense, since nonsensical sentences consisting of categorially incompatible words are supposed to be the main bearers conveying unsayable insights. On the other hand, the resolute readers of the Tractatus reject the substantial conception of nonsense in favour of the austere view (Conant, 1998, 2002, 2004, 2011; Conant & Diamond, 2004; Diamond, 1995, 2010, 2018, 2019a, 2019b; Floyd, 1998, 2007; Goldfarb, 1997, 2011; Kremer, 1997, 2001, 2007). They claim that there are no ineffable truths, and the aim of the work was, in fact, anti-metaphysical and therapeutic. The austere conception of nonsense is the favourite object of attacks for the critics of the resolute reading. Nevertheless, I believe that the austere view is attractive both for exegetical and substantive reasons. It is appealing from the exegetical point of view because it is more coherent with the rest of the Tractatus and other Wittgenstein’s early writings than the substantial view. Furthermore, the austere view and the matter of its assessment should be interesting to the broad philosophical audience, since it is a position that can play an important role in the latest revival of interest in the nature, understanding, and scope of nonsense in philosophy and scientific discourse in general. For that reason, the last part of my paper concerns the substantive merits of the austere view that prove the need for broader interest in this view.

The paper is intended as a defence of the austere view of nonsense against its most important critics. I want to consider four main issues crucial for the acceptance of the austere view. Firstly, I discuss passages of Wittgenstein’s works witnessing that the author of the Tractatus accepted the austere conception, and I point out some previously neglected passages in Wittgenstein’s works. Both the proponents and critics of the austere view put a lot of emphasis on the rational reconstruction of Wittgenstein’s philosophy. This strategy has some advantages, since it leads to a number of interesting interpretations, and shows how to link the very particular strand of Tractatus to the more general aim of the work. However, it proves to be inconclusive and questionable on methodological grounds. For these reasons, I begin with a discussion of Wittgenstein’s texts rather than a detailed reconstruction of his views. Secondly, I defend the austere view against an argument saying that the main component of the conception, namely the context principle, should be rejected as an obvious falsehood. Thirdly, I argue that the austere view and the context principle are consistent with the principle of compositionality. The principle of compositionality is the basis of contemporary semantics, hence the incoherence between the austere reading and compositional view of language would make the Tractarian view of nonsense unpalatable for most philosophers interested in how language works. Finally, I show that the austere view is attractive beyond the restrictions of the exegetical debates, since it is a view that can be appealed to by all semantic frameworks, and allow us to avoid many controversial debates without resigning from the useful role that the notion of nonsense can play in philosophical debates. Additionally, when discussing the aforementioned points I also want to shed light on some confusions and misunderstandings occurring in the debate between the resolute and standard readers. At the same time, I do not consider some important topics that were raised by Peter Hacker (2000, 2003) and answered back by James Conant (2001; Conant & Diamond, 2004) and Cora Diamond (2005).² I am going to focus on arguments against the austere view that are crucial for the whole debate and have not been satisfactorily answered yet.

The paper is organized as follows. The next section presents the main aspects of the austere conception of nonsense and contrasts this conception with the substantial view of nonsense. In Sect. 3, I discuss textual support in favour of the austere view of nonsense that can be found in Wittgenstein’s writings. In Sect. 4, I challenge an argument against the austere conception and the context principle saying that they contradict well-known facts about meaning, such as the occurrence of meaningful words on the lists of verbs and in definitions. In Sect. 5, I examine the view that the austere conception of nonsense leads to a conflict with the principle of compositionality. In Sect. 6, I argue that a contemporary theorist of nonsense should be interested in the austere conception of nonsense because of its independence from theories of meaning and content.

2 Conceptions of nonsense in the Tractatus

As it is well-known, Wittgenstein states in the penultimate thesis of the Tractatus that some/all of his thesis are nonsensical, and for that reason they must be thrown away. In detail, his claim is as follows:

My propositions serve as elucidations in the following way: anyone who understands me eventually recognizes them as nonsensical, when he has used them — as steps — to climb up beyond them. (He must, so to speak, throw away the ladder after he has climbed up it.)

He must transcend these propositions, and then he will see the world aright (TLP, 6.54).³

How should we understand the term ‘nonsensical’ and what is the aim of repudiating the Tractarian sentences? These two questions are interconnected, and so are the answers to them. As I noted in the introduction, there are two basic conceptions of nonsense in the Tractatus, namely the substantial view and the austere view. I will start by presenting the substantial conception.

Everyone knows some down-to-earth instances of nonsense, for example the sentences ‘Peter it Derek’ and ‘Mark is a very abnegef person’ are devoid of sense. The first sentence is not a grammatical one, the second sentence includes the word ‘abnegef’, which lacks any meaning (at least in English). According to the substantial view of nonsense, there is a second kind of nonsense that is more philosophically interesting.⁴ The second kind of nonsense comes from the incompatibility of logical categories of words in a sentence. It includes primarily sentences that are excluded by the logical syntax of language. The two well-known examples are the following category mistakes: ‘Julius Caesar is a prime number’ (Carnap, 1931), ‘There are objects’ (TLP, 4.1272).⁵ Sentences of the second kind of nonsense are nonsensical because some of its components cannot be connected, for instance by means of predication, on the basis of logical categories of words. According to the Tractatus, existence cannot be attributed to formal concepts, but it needs a proper concept to be attached to. On the other hand, the Carnapian view of syntax holds that the abstract properties like being a prime cannot be predicated about a concrete object such as Julius Caesar. In both cases, the logical categories of words are incompatible, and the resulting sentences are nonsensical. However, sentences involving category mistakes can be used in a philosophically useful way to convey some insights. For instance, we can gain insights into the nature of formal concepts and the proper symbolism by means of sentences consisting of improper use of notions such as ‘object’, ‘fact’, ‘function’.⁶

The austere view holds that there is only one kind of nonsense, namely mere nonsense.⁷ All nonsense is a matter of privation, since the sentence is nonsensical if and only if it involves a component deprived of meaning. The proponents of the austere view reject nonsense of the substantial kind, namely nonsense having a source in contravention of the rules of logical syntax. However, this does not mean that sentences involving category mistakes are meaningful.⁸ They are not. The sentences ‘Mark is a very abnegef person’ and ‘Julius Caesar is a prime number’ are of the same kind, since they have a word that lacks a meaning. The only difference between these two sentences is that in the former sentence it is obvious which component lacks a meaning, while the latter one can be fixed in two ways.⁹ Wittgenstein’s diagnosis was that generally in the case of sentences apparently involving category mistakes we oscillate between regarding different components as devoid of sense.¹⁰

Let me discuss some important aspects of the austere view of nonsense. Firstly, it is worth emphasizing that according to this conception, a violation of a rule of logical syntax is neither a sufficient nor a necessary condition of nonsensicality, since we can always use a word in a new and innovative way, which is not in agreement with the logical category of past uses (Diamond, 2005). Famously, Gottlob Frege claims that in the sentence ‘Trieste is no Vienna’, the word ‘Vienna’ is used as a concept word (Frege, 1892, p. 189).¹¹ Usually, ‘Vienna’ functions as a proper name, but in the discussed sentence we were able to find a new logical role for it. Therefore, the sentence ‘Trieste is no Vienna’ is meaningful even though ‘Vienna’ is used in contravention of the rules of logical syntax. The lack of sense of a sentence does not come from the incompatibility of logical categories of words, but from the lack of appropriate meaning conventions.¹² If we are not able to find any logical role for a word in a sentence, then we should recognise it as nonsensical. A violation of some rules of logical syntax is irrelevant here.

Another important aspect of the austere view of nonsense is the context principle. The resolute readers maintain that the context principle enabled Wittgenstein to recognise what nonsense is (Diamond, 1978, p. 201, 1981, p. 21, see also: Conant, 2002, pp. 384–385, 401–403). To see this we should ask: how do we identify the semantic values of words? How do we recognise the meaning and syntactic category of a word?¹³ The context principle says that in order to discover the logical role of a word, we must attend to its use in the given sentence. We cannot recognise the meaning and syntactic role of an expression in isolation, i.e. without taking into account the context of a sentence. Frege showed by means of the context principle that both mental ideas (anti-psychologism well-known from Grundlagen der Arithmetik (Frege, 1884)) and our previous uses of a word do not sufficiently determine the meaning of a word.¹⁴ The aforementioned example of the sentence ‘Trieste is no Vienna’ demonstrates that we must attend to a particular use of a word to recognise its logical role. Outside the context of a sentence, we cannot ascribe to a phrase the category of, for instance, proper name or first-level concept. The word ‘Vienna’ plays a role of the proper name in such sentences as ‘Vienna is the capital of Austria’ and ‘Vienna is my favourite city’. However, ‘Vienna’ has a predicative role in ‘Trieste is no Vienna’, being something incomplete and unsaturated.

The next question that we have to consider is the following: How can we identify the logical role of component words if they appear in a nonsensical sentence? The proponents of the austere conception emphasize that if the whole sentence is nonsensical, then we cannot ascribe logical roles to its constituents (see especially Diamond, 1981, pp. 10–11, 1978, pp. 200–202). If the sentence ‘Julius Caesar is a prime number’ is devoid of meaning, then we are not able to identify meanings and the logical role of ‘Julius Caesar’, ‘is’ and ‘prime number’. Furthermore, we cannot say that the sentences ‘Julius Caesar is a prime number’ and ‘5 is a prime number’ have some components in common, although, the second sentence has elements in common with the sentence ‘7 is a prime number’.¹⁵ Signs composing a sentence are to be recognized as semantic or syntactic elements only in a meaningful sentence. According to this approach to nonsense, semantic and syntactic rules work conditionally: syntactic rules determine compositionally, i.e., based on the meanings of the component words, the meaning of a sentence. If I know a language, then I know its rules. If I know the rules of the language, I know what a proposition will be expressed by a sentence composed of such and such words. However, I cannot know that such and such components have such and such meaning unless the sentence involving these components is meaningful. In other words, I cannot know the meaning of an expression in the appropriate sentence unless I know that it functions according to rules determined by language. The recognition of this requires grasping the sense of a proposition (Diamond, 1981, pp. 10–11, see also Diamond, 1978, pp. 200–202).

The substantial view of nonsense is an essential part of the standard reading, while the austere view of nonsense is a basic tenet of the resolute reading. No wonder that a significant part of the debate between these two readings consists in the critique of opposite conceptions of nonsense. In particular, the standard readers have claimed that both the austere view and supporting it the context principle are untenable, and, in fact, do not find any support in Wittgenstein’s writings. In the next section, I consider fragments of Wittgenstein’s works that demonstrate that the author of the Tractatus accepted the austere conception.

3 Textual arguments in favour of the austere view

Let me begin this section by highlighting the key fact for our discussion on conceptions of nonsense: the austere view of nonsense is the one and only conception of nonsense discussed in detail by both early and late Wittgenstein. Furthermore, even the main proponents of the standard view acknowledge the fact that Wittgenstein in the Tractatus never speaks of violations of logical syntax (Hacker, 2003, p. 13, see also Hacker, 2000, p. 365). In support of their conception of nonsense, they mainly cite papers from the middle period of Wittgenstein’s philosophical development (circa 1929–1933) when his position was not fully worked out and in a state of constant change (Glock, 2004, pp. 231–232; Hacker, 2003, pp. 11–12).¹⁶

The first excerpt from Wittgenstein’s work that I want to consider is the following:

Though it is nonsense to say ‘I feel his pain,’ this is different from inserting into an English sentence a meaningless word, say ‘abracadabra’... and from saying a string of nonsense words. Every word in this sentence is English, and we shall be inclined to say that the sentence has a meaning. The sentence with the nonsense word or the string of nonsense words can be discarded from our language, but if we discard from our language ‘I feel Smith’s toothache’ that is quite different. The second seems nonsense, we are tempted to say, because of some truth about the nature of things or the nature of the world. We have discovered in some way that pains and personality do not fit together in such a way that I can feel his pain. The task will be to show that there is in fact no difference between these two cases of nonsense, though there is a psychological distinction in that we are inclined to say the one and be puzzled by it and not the other. We constantly hover between regarding it as sense and nonsense, and hence the trouble arises (Wittgenstein, 1935).¹⁷

As counterintuitive as the austere view of nonsense may seem, we see from the above fragment that Wittgenstein formulated such a position. The author of the Tractatus states that his aim is to show that there is no logico-semantic difference between a mere nonsense ‘abracadabra’ and the apparently metaphysically loaded nonsense of the sentence ‘I feel Smith’s toothache’. ‘Abracadabra’ is not a part of our language, and it does not have any semantic rule assigned to it, hence it lacks meaning in the simplest possible way. On the other hand, the sentence ‘I feel Smith’s toothache’ may seem to be excluded by the very nature of the thing; it seems to be a negation of a necessary truth concerning the privateness of pain. Wittgenstein emphasizes that the reason for the nonsensicality of the sentence concerning Smith’s pain is not the fact that the true nature of things turned out to be incompatible with saying that we feel someone else’s pain. There is not a special kind of nonsense occasioned by the incompatibility of metaphysical categories of things. For this reason, ‘abracadabra’ and ‘I feel Smith's toothache’ belong to the same category of nonsense. The discernible differences between them are only of a psychological variety. That means that a sentence apparently concerning Smith’s pain can be accompanied by a number of feelings, ideas and beliefs that do not arise along with ‘abracadabra’ or other mere nonsense, but they do not influence the logico-semantic status of this concept. The main role of the context principle was to convince us that mental images have no influence on the meanings of words. Therefore, the difference between the two nonsensical expressions discussed by Wittgenstein remains in the sphere of psychology without leading to any logico-semantic consequences.¹⁸

The next paper in which Wittgenstein considers the austere view of nonsense comes from the thirties too. Its main hallmark is once again the fact that Wittgenstein explicitly states that there is only one kind of nonsense since differences between nonsensical sentences have only psychological nature.

It is queer that we should say what it is that is impossible, e.g., that the mantel piece cannot be yellow and green at the same time. In speaking of that which is impossible it seems as though we are conceiving the inconceivable. When we say a thing cannot be green and yellow at the same time we are excluding something, but what? Were we to find something which we described as green and yellow we would immediately say this was not an excluded case. We have not excluded any case at all, but rather the use of an expression. And what we exclude has no semblance of sense. Most of us think that there is nonsense which makes sense and nonsense which does not – that it is nonsense in a different way to say ‘This is green and yellow at the same time’ from saying ‘Ab sur ah’. But these are nonsense in the same sense, the only difference being in the jingle of the words (AWL, pp. 63–64).¹⁹

In Wittgenstein’s view, the sentence ‘This is green and yellow at the same time’ has only the semblance of sense. In fact, it is not semantically or logically different from any other random sequence of signs (iconographic, phonic etc.). The difference between patent nonsense (‘Ab sur ah’) and latent nonsense (‘This is green and yellow at the same time’) lies in their psychological association. There is no difference in semantics, since nonsensical signs do not possess meanings, and there is no variation in logic, since nonsensical signs do not belong to logical categories (‘[…] the only difference being in the jingle of the words’).

The further fragment points to another important aspect of nonsense. The substantial view of nonsense suggests that there are illogical thoughts, i.e. propositions consisting of expressions that do possess incompatible meanings, and for that reason, they are nonsensical. Wittgenstein warns us in Philosophical Grammar and Philosophical Investigations that such a view is misleading:

How strange that one should be able to say that such and such a state of affairs is inconceivable! If we regard thought as essentially an accompaniment going with an expression, the words in the statement that specify the inconceivable state of affairs must be unaccompanied. So what sort of sense is it to have? Unless it says these words are senseless. But it isn’t as it were their sense is senseless; they are excluded from our language like some arbitrary noise, and the reason for their explicit exclusion can only be that we are tempted to confuse them with a sentence of our language (PG, p. 130).

When a sentence is called senseless, it is not, as it were, its sense that is senseless. Rather, a combination of words is being excluded from the language, withdrawn from circulation (PI, 500).

It may occur to us that we infringe the bounds of sense by means of metaphysical statements; that way we try to express something, which cannot be expressed. We tend to think that these sentences have ‘sense that is senseless’. Sentences of the substantial kind of nonsense try to represent thoughts about which, at the same time, we claim that they are nonsensical. Their nonsensicality come from the logical categories to which they belong, since, for example, ‘There are books’ involves compatible combinations of symbols, but ‘There are objects’ consists of incompatible symbols. However, Wittgenstein states that we exclude nonsensical sentences from our language like ‘some arbitrary noise’ and the reason for that is only that ‘we are tempted to confuse them with a sentence of our language’. We eliminate nonsensical sentences masking philosophical thoughts in no other way than we rule out mere nonsense like ‘ab sur ah’. There is no reason to claim that the sentence ‘There are objects’ conveys something ineffable though it is nonsensical. There is no reason to think that there are propositions that ‘would be true if they could be said’ (Anscombe, 1965, p. 162). If a sentence is nonsensical, then it is no different in kind than an arbitrarily chosen sequence of signs devoid of meaning, since there are no illogical propositions, i.e., sentences having senseless sense.

The fragments of the works discussed so far prove that Wittgenstein really did consider the austere view of nonsense and the problems that follow from the rejection of this conception. But do we know of any works written before 1929 that lead us to the same conclusion? Did the early Wittgenstein discuss the problem of illogical thoughts? Was he interested in the existence of different kinds of nonsense? Let me start from an inconspicuous remark from Notes Dictated to G. E. Moore in Norway:

An illogical language would be one in which, e.g., you could put an event into a hole (NM, p. 107).

What would it mean to put an event into a hole? What would be a language that allows for putting an event into a hole? What is a language in which there is no such thing as that? Let me take as an example the sentence s₁ ‘The event: the execution of Charles I in 1649, is in the crater of Vesuvius’.²⁰ An illogical language would be one in which such a sentence would say of the event (the execution of Charles I) that it is in the crater. The situation would take place if and only if it were possible to make a category mistake in a way described by the substantial view of nonsense. Only the sentence in which illogical combinations are possible would really have the structure that—we think—we can perceive in that sentence. Nevertheless, according to Wittgenstein, there is no such thing as saying that an event is in a crater, since there is no illogical language. There is no such thought that the execution of Charles I is in the crater; there is only a string of words imitating an expression of thought. This example illustrates a confusion characteristic for the substantial view of nonsense. The substantial conception holds that s₁ would consist of logically incompatible symbols. An illogical language is a language in which you can put an event into a hole. So it is a system of symbols in which you can really do what the substantial conception of nonsense proposes in the case of category mistakes, namely to combine symbols in an illicit manner. To put it another way, illogical languages would allow us to genuinely combine logically incompatible symbols.

The proponents of the austere view maintain that there is no such thing as a combination of incompatible symbols resulting in a nonsensical sentence. Wittgenstein’s view does not say that it is impossible to put an event into a hole. There is simply no such thought: there is no thought that an event is into a hole; there is only a string of signs imitating the thought. This aspect of the austere view is challenged by Hacker (2003, pp. 7–8). He emphasizes that Wittgenstein accepts the possibility of putting an expression in an illicit sentential position.²¹ According to Hacker, the acceptance of illicit combinations of symbols does not lead to the confused idea of an illogical language. Now, how would Wittgenstein react to this opinion? He comments on his new way of sentence analysis in a letter to Bertrand Russell (16.01.1913) in the following manner:

But if I analyse [it] (as I do now) into Socrates and (∃x)x is mortal or generally into x and (∃x)ϕ(x) it becomes impossible to substitute the wrong way round, because the two symbols are now of a different kind themselves. What I am most certain of is not however the correctness of my present way of analysis, but of the fact that all theory of types must be done away with by a theory of symbolism showing that what seem to be different kinds of things are symbolised by different kinds of symbols which cannot possibly be substituted in one another’s places (WiC, p. 38).

Wittgenstein claims in his letter to Russell that in the theory of symbolism it is impossible to combine symbols in the wrong way. It is impossible to substitute something in the wrong way. The fact that there is no substitution of one word by another leading to an illogical language was part of Wittgenstein’s view that ‘logic must take care of itself’.²² The correct view of logic must show that an inappropriate substitution of logical categories for one another is not possible (NB, pp. 11–12; TLP, 5.473).

The above considerations lead to the following idea: there is only one kind of nonsense and nonsense results from the absence of meaning of one of the components of a sentence. In the Notes Dictated to G. E. Moore in Norway, one can read:

The reason why ¬ x is meaningless, is simply that we have given no meaning to the symbol ¬ε (…). The reason why, e.g., it seems as if ‘Plato Socrates’ might have a meaning, while ‘Abracadabra Socrates’ will never be suspected to have one, is because we know that ‘Plato’ has one, and do not observe that in order that the whole phrase should have one, what is necessary is not that ‘Plato’ should have one, but that the fact that ‘Plato’ is to the left of a name should (NM, p. 115).

The expressions ‘¬ x’ and ‘Plato Socrates’ are nonsensical due to the lack of appropriate meaning conventions. In the former case, we have given no meaning to ‘x’ as to the possible argument of operation of negation. In the latter case, there is no meaning convention saying that we can predicate another name about ‘Plato’. Alternatively, we can just say: ‘Socrates’ has no meaning as a predicate.²³ Wittgenstein suggests that ‘Plato Socrates’ and ‘Abracadabra Socrates’ are no different in terms of the kind of nonsense they are. In both cases, one of the components of a sentence lacks a meaning, so the whole sentence is devoid of sense. The author of the Tractatus does not say that these combinations are prohibited or violate any other semantic condition. Nonsense is explained there as the absence of an appropriate meaning convention that could be adopted in other circumstances.²⁴

Let me note that in the last quoted passage Wittgenstein talking about nonsense has used the same phrase ‘abracadabra’ as nearly twenty years later discussing the austere view in the notes taken by Margaret Macdonald.²⁵ Moreover, I think that the remarks from the Tractatus that are routinely cited in support of the austere view of nonsense will seem clearer and less confusing now. In the Tractatus, Wittgenstein devoted a number of theses to his conception of nonsense:

Logic must look after itself.

If a sign is possible, then it is also capable of signifying. Whatever is possible in logic is also permitted. (The reason why ‘Socrates is identical’ means nothing is that there is no property called ‘identical’. The proposition is nonsensical because we have failed to make an arbitrary determination, and not because the symbol, in itself, would be illegitimate.)

In a certain sense, we cannot make mistakes in logic (TLP, 5.473).

We cannot give a sign the wrong sense (TLP, 5.4732).

Frege says that any legitimately constructed proposition must have a sense. And I say that any possible proposition is legitimately constructed, and, if it has no sense, that can only be because we have failed to give a meaning to some of its constituents.

(Even if we think that we have done so.)

Thus the reason why ‘Socrates is identical’ says nothing is that we have not given any adjectival meaning to the word ‘identical’. For when it appears as a sign for identity, it symbolizes in an entirely different way — the signifying relation is a different one —therefore the symbols also are entirely different in the two cases: the two symbols have only the sign in common, and that is an accident (TLP, 5.4733).

In TLP, 5.473, Wittgenstein explicitly states that the only reason why an expression is nonsensical is the absence of a meaning convention. No sign is logically incoherent or illicit, but a sentence is nonsensical if its component lacks a meaning. In TLP, 5.4732, Wittgenstein alludes to the idea that nonsensical sentences do not possess senseless sense, since their components are devoid of meaning. This is the essence of the austere view of nonsense, and this is also the main message expressed in TLP, 5.4733. In this remark, Wittgenstein reiterates the view that there are no illegitimate combinations of symbols. The sentence ‘Socrates is identical’ is nonsensical since there is no adjectival use of the expression ‘identical’ in our language. A fate of ‘identical’ is similar to the previously discussed case of ‘Plato’ that has no meaning as a subject of attribution of other names. According to the Tractatus, both ‘identical’ and ‘Plato’ are only signs in these contexts, but not symbols.

I believe that the above excerpts from Wittgenstein’s works are evidence strong enough to ascribe the austere conception of nonsense to the early Wittgenstein. In the remainder of my article, I will discuss some important objections to the austere view. In the next section, I will defend the austere conception of nonsense against the criticism that it involves a version of the context principle which is obviously false, for it excludes meaningful uses of words outside the context of a sentence on the lists of verbs and in definitions.

4 The restrictive context principle is defensible

It is hard to question that Wittgenstein believes in a form of the context principle, since he seems to embrace it explicitly:

Only propositions have sense; only in the context of a proposition does a name have meaning (TLP, 3.3).²⁶

An expression has meaning only in a proposition (…) (TLP, 3.314).

(…) If I can imagine objects combined in states of affairs, I cannot imagine them excluded from the possibility of such combinations (TLP, 2.0121).

The proponents of the standard reading do not reject the context principle completely, but only its version that leads to the austere view. I will devote this section to some of Glock’s (2004, 2015) and Baker & Hacker, (2005, p. 159) arguments against the ‘restrictive’ reading of the context principle without delving into the question of the alternative renderings of the context principle.²⁷

Glock notes that the restrictive context principle plays a crucial role in one of the main arguments in favour of the austere conception of nonsense. He formulates the argument in the following way (Glock, 2004, pp. 225–226, 2015, pp. 122–123):

P₁ A word (name) has meaning only in the context of a proposition

P₂ A proposition is a sentence with a sense

C No component of a sequence of signs that lacks a sense can have a meaning.

The above argument implies that none of the elements of the sentence ‘Socrates is identical’ possess any meaning, since this sentence is not meaningful. On the basis of the second premise, if the sentence is not meaningful, then it isn’t a proposition.²⁸ The first premise, in turn, says that only parts of a proposition have meaning, i.e. components of a sentence without a sense are devoid of meaning themselves. Therefore, none of the components of a nonsensical sentence have sense.

Glock believes that the second premise is obviously true, since it is implied by the following Tractarian remarks: 3.1, 3.31, 3.5 and 4. It is very difficult to challenge this premise, because it is a slight modification of the mentioned thesis 4 ‘The thought is the significant proposition’. Glock does not question it, and I do not want to question it either. He rejects the first premise, which amounts to the restrictive view of the context principle. According to Glock, it should be obvious that this premise is false. He supports his claim by two arguments. Firstly, consider the following two columns (the first list is derived from the list of auxiliary verbs, the second one is from the list of irregular verbs):

$$\begin{array}{l@{\qquad}l} {\text{to\,be}}\quad& {\text{to\,abide}} \\ {\text{to\,have}} \quad &{\text{to\,arise}} \\ {\text{to\,do}}\quad & {\text{to\,awake}}\\ \end{array}$$

In Glock’s view, there are two indisputable things about these words. Firstly, they are not part of a proposition. Secondly, they do possess a meaning. They are neither nonsensical nor meaningless. Hence, the (restrictive) context principle must be false (Glock, 2004, p. 226, 2015, p. 122).

Glock’s second argument is as follows. Take this definition of nonsense from the Merriam Webster’s Collegiate Dictionary:

nonsense n 1 a: words or language having no meaning or conveying no intelligible ideas b (1): language, conduct, or an idea that is absurd or contrary to good sense (2): an instance of absurd action 2 a: things of no importance or value: trifles b: affected or impudent conduct.

According to Glock, it would be ‘absurd’ to maintain that the word ‘nonsense’ on the left side of the definition is devoid of meaning. He points to the fact that the definiens specifies the meaning of the definiendum. If this specification of the meaning is going to work, the definiendum must have some meaning in the first place. Moreover, Glock rejects the invocation of the type/token distinction, and emphasizes the difference between having a meaning and being used to say something. Therefore, the only possible conclusion is that the (restrictive) context principle is wrong (Glock, 2004, p. 226, 2015, pp. 122–123).

What do we have to think about this argument? Let me start from the lists of irregular and auxiliary verbs. Should a backer of the austere view believe that ‘to be’, ‘to have’ etc. figuring on this list are devoid of meaning? I do not think so. At least, it is not clear that they should say that they are devoid of meaning until we determine whether we are talking about signs or symbols. In each case, we have a sign that can be used as a different symbol. As I already explained in Sect. 2, the context principle tells us that we have to examine the whole sentence in which an expression occurs to identify its logical role. The identification of the logical role of an expression allows us to ascribe to it syntactic and semantic rules existing in our language. So, this procedure enables us to discern the symbol in the sign, and in the wake of it we gain a well-defined semantico-syntactic item that was previously just an orthographic unit. That means that without attending to the logical role of an expression in a sentence, we are not able to determine whether, for instance, the word ‘is’ has a role of the symbol of identity, of coextensionality or of copula in the particular sentence. Without examining the use of a sentence, we can only recognise a sign having multiple potential rules and roles in our language. In fact, we have to do with such a situation in Glock’s example of the lists of verbs. For instance, the verb ‘to be’ has different logical roles in each of these three sentences: (a) ‘Green is green’; (b) ‘Green is Green’; (c) ‘The colour green is the colour green’. Without taking into account the context of a sentence, we do not even know whether some signs are independent items or whether they are just part of other expressions. This is the difference between ‘Smith has Parkinson’s hat’ and ‘Smith has Parkinson’s disease’ (Diamond, 1981, p. 8). The word ‘Parkinson’ has an independent meaning in the first sentence, but it is not a working part of the second one. To take a different example, ‘have a good day!’, ‘have dinner’, ‘have a smoke’ involve different uses of ‘have’. In the first sentence, ‘have’ is not probably a self-subsistent expression, but an ineliminable part of the meaningful unit. In the last two phrases, we encounter different symbols having only common signs. Hence, there is a need for the identification of the logical role of expressions in the case of verbs too.

Wittgenstein’s use of context principle is epistemological rather than ontological.²⁹ There are signs, symbols, meaning conventions and syntactic rules. Language possesses well-defined semantics and syntax. The context principle concerns our recognition of these rules and meaning conventions in particular cases, and leads to some conclusions on the general working of a language. It points to the fact that, according to our language’s lexicon, the same sign can have different semantic and syntactic rules assigned to it, so that we cannot know which rules govern it until we examine it in the context of a sentence. Glock’s lists of auxiliary and irregular verbs consist of well-defined signs associated with many rules of our language, but they have to be used in a sentence to enable us to identify the symbols behind them.³⁰

The case of definitions is quite similar to the example of columns of verbs. Moreover, it further reinforces the observation that the number of expressions that are not perspicuous in terms of symbols is not limited only to logically confusing words, such as ‘to be’. According to the definition of nonsense cited by Glock, we can attribute the property of being nonsensical to such different types of entities as (1) words; (2) language; (3) ideas; (4) actions. Words are syntactic entities, language consists of the semantic content of sentences, ideas are mental entities, actions are performed by some subjects, and hence—in contrast to other categories—are used in an adjectival way. In fact, we attribute the property of being nonsensical to completely different ontological categories in each case. I think it is fair to say that each of these uses of the sign ‘nonsense’ symbolises in a different way, and is therefore a different symbol. This is supported by the fact that it has traditionally been assumed that logical syntax ranks categories according to their possibilities of combination (Pap, 1960; Russell, 1908; Ryle, 1938, 1949). Mental properties cannot be attributed to abstract objects, so they belong to different logical categories, and so on. The word ‘nonsense’ in Glock’s definition is a sign that can be used according to different syntactic and semantic rules; it is not a well-defined symbol. On the right side of the definition, in the definiens, we have given a set of semantic and syntactic rules that reports how we can use the phrase ‘nonsense’ in our language. These sets of rules determine different logical roles of symbols. Furthermore, it is possible to use the word ‘nonsense’ in a new, unknown way that does not automatically make it nonsensical.

In the next section, I will argue that we can embrace both compositionality and the austere view, since the restrictive context principle is not in tension with the standard features ascribed to compositional languages.

5 The austere view is compatible with the principle of compositionality

One of the most important and acute problems in the resolute/standard reading debate is an issue of the compositionality of language. Routinely, the principle of compositionality maintains that the meaning of a complex expression is determined by its syntactic structure and the meanings of its constituents. Hence, the meaning of a sentence is a function of the meaning of its parts together with the method by which those parts are combined (see Pelletier, 1994; Szabo, 2022; Westerståhl, 2015). The principle of compositionality is at the bottom of contemporary philosophy of language (Lycan, 2002, pp. 3–4). Furthermore, a number of remarks from the Tractatus suggest that Wittgenstein believed in the compositionality of language.

A proposition is not a blend of words.—(Just as a theme in music is not a blend of notes.) A proposition is articulate (TLP, 3.141).

Like Frege and Russell I construe a proposition as a function of the expressions contained in it (TLP, 3.318).

(…) It is understood by anyone who understands its constituents (TLP, 4.024).

When translating one language into another, we do not proceed by translating each proposition of the one into a proposition of the other, but merely by translating the constituents of propositions (…) (TLP, 4.025).

It belongs to the essence of a proposition that it should be able to communicate a new sense to us (TLP, 4.027).

A proposition must use old expressions to communicate a new sense (TLP, 4.03).

The principle of compositionality seems to be an important part both of the contemporary vision of language and the one proposed by Wittgenstein in the Tractatus. However, it is claimed that the austere view of nonsense is at odds with the compositionality of language (Glock, 2004, pp. 226–227, 228; Liptow, 2018, pp. 231–232). The critics of the austere view blame the restrictive context principle for the conflict. They believe that the compositional view of language is substantively and (in the case of the early Wittgenstein) exegetically justified³¹.The critics maintain that the context principle in its restrictive version is incompatible with compositionality; therefore, they conclude that we should reject the restrictive context principle, and—in the wake of it—the austere conception of nonsense.

To explain the conflict of the context principle with the compositionality of language, Glocks points to the fact that we ascribe meanings primarily to words. The paradigm of explanations of meanings are dictionary definitions, and not sentences. Sentences are not a primary vehicle of meaning or a unit of significance, since sentences are complex signs themselves the meaning of which depends on the meaning of constituents. Furthermore, understanding the components and the mode of composition of a sentence is a necessary condition for the understanding of the whole sentence (Glock, 2004, pp. 226–227). One understands the meaning of a sentence by means of the understanding of its components and syntax. We have to understand the simple signs first to communicate with sentences. Similarly, in translations we do not translate the whole proposition, but only its constituents. Words have primacy in the order of understanding. According to Glock (2004, p. 228), this suggests that the early Wittgenstein does not interpret the context principle in the restrictive way. Liptow sees the main problem in the primary unit of understanding as well. In his formulation, the principle of compositionality says that ‘the only way we can ever understand a sentence is by understanding its parts and the way they are put together’ (2018, p. 232). Liptow claims that the conditional interpretation of semantic and syntactic rules assumed by the austere reading implies the possibility of grasping the sense of a proposition without the understanding of its components. Understanding of a sentence must be prior to an interpretation of its components, but this contradicts the principle of compositionality. Hence, Liptow concludes that the restrictive reading of the context principle challenges the compositionality of language, and we should reject the restrictive reading along with the austere view (Liptow, 2018, p. 232).

Now, is the restrictive reading of the context principle incompatible with the compositionality of language? There are—at least—two different aspects of this issue worthy of consideration. To begin with, Diamond’s intention, in her original papers (1978, pp. 199–201, 1981, pp. 19–20), was to comply with the principle of compositionality. She speaks about semantic and syntactic rules on which language is founded. Prima facie, there is not much in them which is potentially incompatible with the principle of compositionality. In fact, the identification of these rules is based on the compositional interpretation of meaning. The idea behind this interpretation is as follows: a symbol in a sentence is that element that makes an appropriate compositional contribution to the meaning of that (actually applied) sentence. The sign ‘is’ plays the logical role of copula in the sentence ‘Mark is ill’, since it (actually) determines the thought that Mark is ill, but it does not determine (now) the thought Mark is Mark. If the contribution of a sign had been different, then the symbol represented by it would have changed as well. The semantic rules are fully compositional since the meanings of words (symbols) determine the meaning of a sentence (a propositional symbol). Diamond’s position is not much different from Glock’s and Liptow’s accounts in this regard. They agree that the meaning of an expression is that semantic item that has an appropriate compositional contribution to a sentence. Glock and Liptow believe that it is sufficient for a word to only have a given potential meaning. Diamond, on the other hand, maintains that this contribution must be actual, and not potential, to determine a meaning. Both conceptions postulate compositional determination of meaning, but they differ with respect to the methods of the individuation of meaning. Once more, Glock and Liptow support the individuation of symbols in terms of a potential use, while Diamond backs attending to an actual use.

As I indicated in the second section, Diamond postulates conditional understanding of semantic and syntactic rules. That means that we can use the general rules of our language to characterize the meaning and structure of a sentence, but only conditionally. If a sentence is nonsensical, then there is no way to identify its elements. Consider the sentence ‘Venus is more massive than Mercury’ (Diamond, 1981, p. 19). This sentence can be taken to consist of a two-term relation expression completed by the proper names ‘Venus’ and ‘Mercury’ if and only if the thought expressed by the whole sentence is meaningful. If I know the rules of language, I know that the meaning of the sentence ‘Venus is more massive than Mercury’ is determined by the meanings of ‘Venus’, ‘Mercury’ and the relational expression ‘x is more massive than y’. However, we can identify the meaning of ‘Venus’ as Venus (the planet, not the ancient goddess or a song), ‘Mercury’ as Mercury (the planet, not the chemical element) and ‘x is more massive than y’ as the relational property of being more massive only in a meaningful sentence. The restrictive context principle tells us how to identify the meaning of an expression, and its identification is not possible in a nonsensical sentence. This conditional dimension of semantic and syntactic rules seems to stay partly behind the worry that the context principle is inconsistent with compositionality.³² The context principle dictates conditional understanding of rules that require the primacy of understanding of a sentence. On the other hand, the principle of compositionality requires the primacy of understanding of words. Now, the question is what is first: understanding of a sentence or understanding of its components? The conflict seems inevitable.

I believe that the conflict in the order of understanding between the context principle and the principle of compositionality is apparent, since the latter principle is not concerned with the way people understand sentences. Glock’s and Liptow’s argument is based on the misunderstanding of what compositionality is. I must delve into the character of the principle of compositionality a little bit deeper to demonstrate why that is. The compositionality of language is an empirical hypothesis postulated to explain two important features of natural languages, namely their productivity and systematicity.³³ Roughly, productivity is a property of language that comes down to the possibility of immediate understanding of a sentence that we have never heard before. This possibility can be explained by a thesis that the meaning of a sentence is determined by the meanings of its components (and their mode of combination) that were known to us beforehand. The thesis, of course, amounts to the principle of compositionality. Systematicity is far more contested (Johnson, 2004; Robbins, 2005). It is based on the observation that there are definite and predictable patterns among the sentences in natural language. For instance, if the sentences ‘The cat is awake’ and ‘The dog is asleep’ are meaningful, then a systematic change of the same grammatical categories should have led to new meaningful combinations such as ‘The dog is awake’ and ‘The cat is asleep’. The most popular proposal to explain the regularity of these meaningful recombinations is once again the principle of compositionality. The principle of compositionality is a thesis concerning determination. Glock and Liptow seem to confuse the principle of compositionality with the following principle: we understand a complex expression by understanding its structure and its constituents.³⁴ However, the arguments from productivity and systematicity only justify the standard view of compositionality.³⁵ The principle of compositionality is not a claim about understanding. Szabo notes in his paper (2010, p. 257) that the fact that we understand complex expressions we have never heard before is a reason to think that sometimes we do understand complex expressions in the way described by the thesis about understanding, but it does not lead to the conclusion that we always do so. In my view, in the case of complex expressions, which are used very frequently and regularly in our community, it is possible that we understand them automatically rather than interpreting each expression separately.³⁶ In this way, their understanding is similar to the understanding of idioms. Therefore, understanding of complex expression is not always based on the understanding of their components.³⁷

Glock and Liptow’s confusion of the principle of compositionality with the thesis about understanding has important consequences. On the one hand, the notion of understanding suggests that either the sentence or the constituents of the sentence (exclusive disjunction) must have priority in this relation. The relation of priority is an asymmetric relation. If x is prior to y, then y is not prior to x. However, the standard version of compositionality concerns determination, and this relation is not asymmetric. That means that if x determines y, y may determine x.³⁸ The meaning of words determines the meaning of a sentence, but a change of the meaning of the sentence may change the meaning of words as well. This shows that the meaning of a sentence sometimes determines the meaning of words. In my view, the context-dependence of some sentences and lexical items needs to be explained in this way.³⁹ For example, the sentence ‘this tree has green leaves’ means that tree t₁ has naturally green leaves when uttered in context c₁, but it means that tree t₂ has painted green leaves when uttered in context c₂.⁴⁰ This example does not contradict the principle of compositionality since the change of the meaning of the sentence is accompanied by the change of the meaning of its constituents.⁴¹ Moreover, once we distinguish between the principle of compositionality and its psychologically oriented cognate, it is possible to postulate independently that a sentence or its components must be prior in the order of understanding. The principle of compositionality is a thesis only about determination, so that it can be squared with different views on the order of understanding. Glock and Liptow’s confusion of the principle of compositionality with the thesis about understanding disguised this possibility. That said, in my view, the resolute readers do not interpret the context principle as a view holding that a sentence is prior in the order of understanding. I think that a hearer grasps the meaning of a sentence and its constituents simultaneously and these two phenomena are unified. However, as I said, this issue is independent from the very question of the principle of compositionality. Finally, the fact that neither sentences nor words are prior to each other, and that the context principle and the compositionality of language are therefore compatible, has also been recognised by Silver Bronzo (2011). He argues that the conflict between the two principles arises from the identification of the context principle with contextualism and the principle of compositionality with compositionalism. I agree with Bronzo that the conflict is superficial. However, in my view, his formulations of contextualism and compositionalism are too strong, and no Wittgensteinian scholar would accept them. Bronzo does not specify the proponents of compositionalism, and he identifies Glock as a moderate proponent of the principle of compositionality who avoids the paradoxical consequences of compositionalism. Furthermore, Bronzo rejects standard arguments for the principle of compositionality. I find this problematic, since it is impossible to determine the content of the principle of compositionality without standard empirical arguments.⁴²

As we saw in the last two sections, Glock’s, Hacker’s and Liptow’s exegetical arguments against the austere view of nonsense are not effective. They do not provide sufficient grounds to think that the early Wittgenstein rejected this view. In the next section, on the other hand, I will consider some reasons why the austere conception of nonsense can be attractive by today’s standards of theory of nonsense.

6 The austere conception of nonsense is substantially plausible

The topic of nonsense was in the mainstream of philosophy of language and analytic philosophy in the first half of the twentieth century. Thereafter, there was an observable decline in interest in this issue. It is hard to find any serious consideration of nonsense in the main strands of philosophy of language in the Post-Kripkean era. Recently, Herman Cappelen (2013) has appealed for a revival of interest in this notion, since he believes that nonsense possesses high theoretical significance, and can be useful in the debates concerning, for instance, the role of intuitions in philosophy and the nature of verbal disputes.⁴³ Hence, there arises a natural question, what is the most useful conception of nonsense? In this section I will argue that the austere conception has some properties that may be appealing to contemporary theorists of nonsense. This issue matters because the critics of the austere view maintain that this conception is inherently unattractive and unpromising (Glock, 2004, 2015; Hacker, 2003; Liptow, 2018).⁴⁴ On the one hand, they stress (apparent) problems with the principle of compositionality and with words on the lists of verbs that I have already discussed. On the other hand, Glock and Liptow suggest that the substantial conception is generally more fruitful than the austere view. In contrast to this opinion, this section will prove that Wittgenstein’s view of nonsense can be of interest not only for historical reasons. Furthermore, the following discussion will show how the austere view can be reconciled with the background of more contemporary views (e.g., a direct reference theory) and problems. I will begin with a discussion of Cappelen’s (2013) work to point out some interesting properties of the austere conception.

First of all, Cappelen notes that there are two main rationale behind the decline in interest in the notion of nonsense. The first one is a rejection of the normative aims of the philosophy of language. Carnap, Frege, Russell and Wittgenstein perceived the construction of concept-script as a way of distancing from some misleading aspects of natural languages and improving our cognitive conditions through a clarification of language. On the other hand, the Post-Kripkean philosophy of language was programmatically descriptive. The normative aims were regarded either as a symptom of excessive philosophical ambitions or as a methodologically flawed idea. The second reason for abandoning the study of nonsense was the very strict connection between a particular theory of meaning, namely the verification theory, and theory of nonsense proposed by logical positivists. Rudolf Carnap developed his theory of nonsense by relying on the verification principle. Then, the verification theory has been broadly rejected and theory of nonsense has been discarded as well, since the latter seemed closely related with the conception of meaning preferred by the logical positivists.⁴⁵

Cappelen believes that both rationale for disregarding nonsense were misguided, especially as the connection between the theory of nonsense and the verification principle is not essential. He proposes a theory of nonsense that programmatically abstracts away from theories of meaning. This theory introduces the notion of nonsense that can be appealed to by all semantic frameworks and all theories of what content is (Cappelen, 2013, p. 22). Furthermore, Cappelen maintains that nonsense should be regarded as neutral towards various subject matters. According to this view, nonsensicality is not a domain of the specific discourse (e.g. ethics, religion, aesthetics or metaphysics), but the sources of nonsense are defective and irresponsible ways of language use. A diagnostic of nonsense requires detailed empirical knowledge of how language is and has been used by the particular individual, and it cannot be done without a detailed investigation of the history of its uses within a community (Cappelen, 2013, p. 34). Before going into the details of Cappelen’s proposal, I want to point to the fact that should be obvious to the attentive reader: Cappelen’s requirements for a correct theory of nonsense are fully fulfilled by the austere conception of nonsense. To begin with, the austere view of nonsense discards the association of the theory of nonsense with any theory of meaning. It rejects the view that nonsense has to be traced back to transgressing rules of logical syntax or any particular theory of meaning (Conant & Diamond, 2004, pp. 42–43).⁴⁶ The austere conception of nonsense says that the only source of nonsense is a lack of meaning in a component of a sentence. It does not have to point to any specific theory that further substantiates the claim that such-and-such component is devoid of meaning.⁴⁷ The reason for nonsensicality of an expression is an absence of a meaning convention. In other words, the austere conception is primarily the negative view that treats nonsense as a matter of privation. It rejects ‘positive nonsense’ resulting from illegitimate combinations of meaningful expressions. ‘Negative nonsense’—accepted by the austere theorists—comes from our not having assigned a meaning to the expression in a certain context. The austere view arises out of the rejection of the very idea of a theory of sense that might give rise to more substantial kinds of nonsense being over-and-above our ordinary conception.⁴⁸

The second common feature of Cappelen’s proposal and the austere view is the need for a detailed investigation of someone’s uses of language to diagnose nonsense. According to the austere conception of nonsense, we cannot diagnose a domain of discourse as devoid of meaning in one fell swoop, but we have to keep track of particular uses of an expression.⁴⁹ This account of nonsense requires a case-by-case approach since the rationale for nonsensicality are not meant to be derived from a theory of meaning. The dealing with sense and nonsense calls for a detailed examination of the discourse that should reveal whether we use an expression in a meaningful way. As I already explained in the Sect. 2, every expression can be used in a new and innovative way. It does not suffice to point to a contravention of logical rules to prove that a sentence is nonsensical. It does not suffice to refer to some kind of theory of sense to declare the whole field of thinking (e.g. ethics, religion, aesthetics or metaphysics) as nonsensical. We really have to take into account the use of a sentence and how its components work. Hence, this kind of thinking about nonsense requires a serious empirical and detailed investigation of language use.

In fact, Cappelen in his work accepts the austere view of nonsense or something of the kind. He observes that Carnap (1931) operates with two kinds of nonsense and choose to work with ‘the view of nonsense as sentences that have parts that lack meaning’ (2014, p. 24).⁵⁰ Cappelen (2013, p. 26) sees nonsense as a matter of privation, since he claims that nonsense is a failure of some type (failure to have a proper semantic content, failure to assert a content, or failure to have a thought). Such a failure can take place, because we are fallible with respect to grounding facts of content. That means that we believe that an expression is not devoid of sense, but it turns out that for some reason the meaning has not been given to it.⁵¹ Cappelen (2013, p. 35) indicates two possible rationale for failure. Firstly, the term T can be introduced in a defective way.⁵² Secondly, the transmission of T through a chain of communication may be defective. Both scenarios are not hard to imagine. For instance, we can imagine that some explorers discover a new island, and call it, let’s say, ‘Smogyrofsky’. We belong to a chain of communication, and we intend to use the newly introduced term to refer to the island. Sometime later, however, it turned out that the matter was more problematic. Our fellow explorers adopted the term ‘Smogyrofsky’ from the natives, who invented it to mock the arrogant conquerors. The introduction of the term ‘Smogyrofsky’ was defective: it does not refer to anything; therefore, in our scenario, we use a sentence consisting of a component that has no meaning. The sentences that we have used to communicate about the newly discovered island turned out to be nonsense.⁵³ Diagnosis of this sort requires a detailed knowledge of empirical facts concerning a linguistic practice. However, there is nothing in our discourse that excludes the possibility of defective uses of the kind that lead to nonsense.

To sum up this section, I believe that the austere conception can be attractive for contemporary theorists of nonsense. As the comparison with Cappelen’s views shows, it has a number of useful and interesting properties. Firstly, the austere view can be appealed to by all semantic frameworks. This property allows us to avoid controversial claims concerning theories of content and theories of meaning. Secondly, the austere view enables us to pursue the debate on nonsense without delving into the difficult terrain of semantic positions, arguments and disputes. The questions ‘what is nonsense’ and ‘what is meaning’ can be examined separately and independently from each other. Notice that the substantial view of nonsense obliges the supporters of the Tractatus to argue for the picture theory of meaning and the particular view of logical syntax; on the other hand, a friend of the substantial view of nonsense and of Philosophical Investigations must justify the use theory of meaning and the specific view of philosophical grammar. It should not surprise anyone how difficult it is to pursue a debate on nonsense under such circumstances. Thirdly, the austere conception of nonsense helps us to reject a diehard view that some philosophical domains are inevitably nonsensical. If metaphysics or ethics is full of nonsense, then we should demonstrate the correctness of this view in specific cases by discovering confusions staying behind uses of some sentences. This cannot be done without a detailed study of someone else’s past and present use of language. According to Wittgenstein, at least, a diagnosis of nonsense requires attention to detail rather than the development of new philosophical theories. This attitude encourages careful examination and philosophical modesty.

7 Conclusion

There are good reasons to believe that the early Wittgenstein accepted the austere view of nonsense, since he discussed and embraced this conception in a number of writings written throughout many years. Moreover, the most important challenges to the austere view of nonsense are not tenable. In my article, I explained why the restrictive version of the context principle is a crucial component of the austere conception. Then, I showed that this version of the context principle does not contradict the well-known facts about meaning (e.g. the uses of language in the dictionary entries, and the functioning of words on the lists of verbs). Next, I demonstrated that the context principle is consistent with the compositional view of language. In the last section of the work, I indicated that the austere view of nonsense should be regarded as an interesting proposal by contemporary theorists of nonsense. Its main advantage is independence from any specific theory of meaning or semantic framework. This feature can help us avoid making controversial assumptions about the nature of meaning and content while preserving the theoretical significance of the notion of nonsense.