From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel

I. What nonsense might not be

Diamond distinguishes between what she calls the natural view of nonsense, which she criticizes, and the Frege-Wittgenstein view of nonsense (also called the substantial view as opposed to the austere view in later debates, for example Conant 2006). One example Diamond gives is (A) “Caesar is a prime number” (usually taken to contain a category mistake) and (B) “Scott kept a runcible at Abbotsford” (usually taken to contain a word which lacks meaning) (Diamond 1991). According to the natural view, the words (or logical elements) in (A) are combined in an “illegal” way, whereas in the second (B), one word has the wrong meaning. These “facts” explain why the sentence lacks sense. According to the natural view, “Caesar” is a proper name, and in that place of the sentence, in combination with “is a prime number” there could only be a number. Therefore, the combination is illegitimate. The result of this mistaken combination is a nonsensical sentence.

Contrary to that story, and in line with what Diamond promotes and calls the Wittgenstein-Frege view of nonsense, a word has meaning only in the context of a sentence. This is often referred to as Frege’s context principle. The question is what “Caesar” means in this sentence. Psychologically, Diamond writes, we think that “Caesar” must be the same in (A) as in “Caesar crossed the Rubicon” but it can’t be this way. Since words do not have meaning in isolation they cannot be combined in the wrong way to make a nonsensical sentence.

On the Wittgenstein-Frege view, a sentence is not nonsense because of some meaning that the words in a sentence already have, or consequently, because of the fact that some rules of combination of logical elements (such as putting a proper name in the place of a verb) are violated against. Only when a sentence makes sense, can it be seen as a combination of logical elements, as having “a structure”. This is what a logically nonsensical sentence lacks. In a logically nonsensical sentence, no parts are to be discerned, and the “sentence” has no structure. There is only one way to produce nonsense, according to Diamond – the two example sentences are, after all, nonsense in the same way.

Anything that is nonsense is so merely because some determination of sense has not been made. (Diamond 1991: 106)

One could rephrase Diamond’s view a bit: nonsense is what cannot be put into any category. Is not categorized in the same way as the items in the other boxes – it is actually not, in a sense, categorized at all, but goes into the rest bin. Another way of putting it: there is no recipe for making nonsense. Thinking that one can make a “diagnosis” of nonsense by showing that a proposition has two or more parts (that it is a combination of categories which cannot be combined) presupposes that one can discern a structure in it and in a sense understands the proposition. Then the sentence is seen as a sentence, and does say something, or we have given it some sense – and hence it is not completely nonsensical after all. That would be paradoxical: a nonsensical proposition with sense. Logically speaking, nonsense, the opposite of what makes sense, is not “sentences which lack sense”. What lacks sense is not a sentence or a proposition or the like. There is simply nothing there. This idea is articulated in the TLP. A sentence has sense, per definition, and it is used, that is, “it is a Satzzeichen in its projective relation to the world”.

A Satz, according to the TLP, has sense (Sinn), and is bipolar: it is either true or false. So called “Scheinsätze”, or pseudo-propositions, are nonsensical, and they are neither true nor false. The word “pseudo-proposition” can be taken as shorthand for a proposition-like entity, which is no proposition because it lacks sense (“Scheinsätze”,TLP 4.1272, 5.354, 5.535). A proposition-like entity, in my reading, is an expression which we are tempted to take as a sentence; perhaps something like “Caesar is a prime number” as opposed to “jsd ffdjiniobglfdsk”. “Expression” is perhaps not a good choice of word here, because of course, it is not expressed by anyone if it is not a sentence.

Diamond’s view in “What Nonsense Might Be” is not clear-cut, it may be argued. She retains elements of the natural view which she criticizes. Diamond keeps the idea of “nonsensical sentences”, even though she gives another explanation to them – otherwise she would not accept A and B as examples of nonsense.

Diamond takes three examples from Wittgenstein and writes that

I should claim that the view of nonsense expressed in those three quotations is one that was consistently held to by Wittgenstein throughout his writings, from the period before the Tractatus was written and onwards. (Diamond 1991:107)

Hertzberg (2001) has argued that this view does not hold for the Philosophical Investigations. I will take a look for support and counter-evidence to Diamond’s claim in the TLP.

II. The Tractatus and Nonsense

In the TLP, Wittgenstein uses the word “Unsinn” or “unsinnig” (nonsense/nonsensical) altogether 22 times (preface, 3.24, 4.003, 4.124, 4.1272, 4.461, 5.473, 5.5303, 5.5351, 5.5422, 5.5571, 6.51, 6.54. “Sinnlos” is used four times: 4.461, 5.312, 5.1362, 5.5351).

Many paragraphs support Diamond’s view. In 5.473, Wittgenstein gives an example of a nonsensical sentence: “Socrates is identical”, and explains that there is no property called “identical”. “The proposition is unsinnig because we have not made some arbitrary determination, not because the symbol in itself is unpermissible. In a certain sense we cannot make mistakes in logic.” 5.4732: ”We cannot give a sign the wrong sense.”

5.47321, on Occam’s rule, is interesting. Unnecessary elements in a symbolism “mean nothing”, and signs which serve no purpose are logically bedutungslos (not “unsinnig”). A very frequently quoted remark in the debate on nonsense (cf. Conant 2004) is 5.4733:

Frege says: Every legitimately constructed proposition must have a sense; and I say: Every possible proposition is legitimately constructed, and if it has no sense this can only be because we have given no meaning to some of its constituent parts.

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

Thus "Socrates is identical" says nothing, because we have given no meaning to the word "identical" as adjective. For when it occurs as the sign of equality it symbolizes in an entirely different way -- the symbolizing relation is another -- therefore the symbol is in the two cases entirely different; the two symbols have the sign in common with one another only by accident.

The paragraphs above support the idea that nonsense is a lack of meaning; that one has failed to give meaning. Also, Wittgenstein writes that we cannot give a sign the wrong sense. The thought that the sign may symbolise in many ways also supports Diamond’s comments on for example the word Caesar, and “is a prime number”.

However, in the first part of this paragraph a tension looms up: “if [the legitimately constructed proposition] has no sense, this can only be because we have not given meaning to any of its constituent parts”. But is it possible to choose not to give meaning [Bedeutung] to a proposition? Can a proposition be without sense and still be called a proposition? And if it is without sense, does it have parts?

There is some support of Diamond’s view in the TLP. Now let’s see look at the paragraphs containing words for nonsense which seem to go against it.

III. Counter-evidence

At times, Wittgenstein allows for the possibility of using words so to say in the wrong way. He uses terms like “pseudo-concepts like object” and writes that when this word is used as a real concept word “nonsensical pseudo-propositions” arise. Expressions such as “1 is a number” (“and the like”) are nonsensical. And it is nonsensical to say “There is only one 1” (4.1272). Wittgenstein seems to think that concept words can only be used as concept words – i.e. he accepts that there be a form of ‘category mistake’ which produces nonsense. That formal terms can only be used as formal terms also entails that they actually do take their category with them into whatever context, and produce nonsense by being used in the wrong way. To Wittgenstein, words of logic can be toxic to sentences. That goes against Diamond’s view.

Note that in this paragraph Wittgenstein passes over from “one cannot say” and “it is impossible to say” to “nonsensical” when he talks about expressions like “”There are 100 objects” and “1 is a number”. He writes that “all similar expressions are nonsensical” and seems not to respect Diamond’s version of Frege’s context principle, but rules out both expressions and certain words beforehand. “Nonsensical pseudo-propositions arise” sounds as if there is after all a recipe for making nonsense, some way to produce pseudo-propositions, by combining words or signs of the wrong kind. The paragraph is part of an argument to show that it does not work to express the same as what is already apparent or internal to the symbolism. Something to be said about a concept script need not be said in it, it is obvious from the sign for it (i.e. from “1” you it is clear that it is a number). In this case “nonsensical” can be read as “superfluous”.

There are also “unstable” remarks in the TLP. For instance 5.473 pulls both in Diamond’s direction and in the direction of a substantial conception; the sentence is nonsensical because “wir eine willkürliche Bestimmung nicht getroffen haben”, but Wittgenstein also gives the reason why the Satz is nonsensical – but a Satz, as mentioned, cannot be nonsensical – it would then be a Scheinsatz. In 5.5352 Wittgenstein hints that Russell’s formalisation of “There are no things” is not a proposition.

Wittgenstein should not be ascribed a “conception” of nonsense, because he writes “Es ist schon darum Unsinn...“ (5.5351). Does he not mean that it is simply unnecessary or even stupid? In this case, the Unsinnigkeit may amount to no more than a plain rejection. It is another talk, a non-technical talk of nonsense, which maybe should not be ignored. (The Pears-McGuinness translation of this remark harbours inconsistencies; Unsinn and unsinnig are translated into both meaningless and senseless.) My examination of the instances of the use of the words Unsinn, unsinnig andsinnlos and terms connected to them (such as Satz) reveals many other internal tensions in Wittgenstein’s use of words.

IV. A conception of nonsense at all?

At this point, it is clear that the matter is much more complex than it seemed at the outset. Wittgenstein does not entertain only the conception of nonsense that Diamond claims, nor clearly another competing consistent conception. What then are we, as interpreters, warranted to say about Wittgenstein’s conception of nonsense, and of the Satz, in the TLP, at all?

My method here was “negative cherry-picking”: I specifically picked out paragraphs which go against a conception that has been ascribed to Wittgenstein. This is a very common, although certainly not unproblematic, way to treat Wittgenstein’s texts.

He is not consistent in his use of the words ‘sinnlos’, ‘unsinnig’ in relation to for example “nichts sagen” (contrast 5.5303 to 5.473) either in the TLP or in the Philosophical Investigations. Apart from “nichts sagen” (5.5303, 4.733) in the TLP, there are related expressions like “ihm entspricht nichts” (cf. 4.063), “heißt nichts“ (4.73), “keinen Gedanken ausdrücken“ (6.21) as well as “bedeutungslos” and “keinen Sinn haben”.

“Unsinn” and “unsinnig” are not always used in a technical or special philosophical way, but are sometimes outright rejections of a claim. Therefore, when these words are used it is not always the expression of “a conception”. One example of this is that “Unsinn” is used to say that something is pointless (5.351).

Even if these inconsistencies are only “verbal”, they present us with a genuine difficulty to determine whether a point of interpretation of the work is correct or not. Even if I could show that there is a lack of consistency on some account, it does not mean that I have proven that there is a genuine inconsistency. In other words, Wittgenstein could have meant one thing and nevertheless failed to express it completely clearly, or he could have been careless. That there are counterexamples or residues, that is paragraphs unaccounted for in some reading, then, does not suffice to conclude either that Wittgenstein had no definite views of what nonsense is or should be, or that the reading proposed is simply to be dismissed. Now the difficulties do not end here: even if there are inconsistencies “only verbally”, that nevertheless could be taken to show that readings of the TLP in which everything explicitly or implicitly is expected to be systematic, or readings, in which single remarks are brought to bear a heavy interpretative weight, are fundamentally problematic – as readings of Wittgenstein.

V. How should these inconsistencies be dealt with?

• 1. One could do some “positive cherry-picking” and simply ignore all inconsistencies and leave them out of one’s account. This would lead to a one-sided interpretation.
• 2. One could try to grind down the faults by for example arguing that other parts of the work are more important or that Wittgenstein did not mean that anyway.
• 3. I think one should go a little “psychological” or “philological” here.

Perhaps Wittgenstein’s ideas are still under development? This would weaken the idea that he would have “a conception of nonsense”. We have to accept that he is not as aware and deliberate on this point as many a reader has thought. Had Wittgenstein been completely systematic, he could have looked through the work to straighten out the remaining “verbal inconsistencies” and the range of words used for similar ideas; had he been structured and deliberate, had it been important.

Now does the expectancy of consistency, and the lack of it, imply an interpretative inclination on my part? Inconsistencies could play in the hands of a resolute reader: they could be taken to support the idea that Wittgenstein did not care much about consistency because it was all to be rejected, recognized as nonsense in the end. A traditional reader again, could benefit from inconsistencies, they perhaps allow for more “support” in favour of positive theses. A psychological take on the slips would not be bad either – whatever remark that does not fit into the doctrine could be explained away as that.

Instead of trying to pull in either direction, I will take another route: The inconsistencies in the Tractatus show that when it comes to nonsense, Wittgenstein is not as deliberate as many think. If the word “nonsense” does not only have a technical use, but is used by Wittgenstein in the TLP with the variation of meaning it has in everyday circumstances, as “gollywop” and “pointless” and what not, then what would look like an inconsistency to an avid interpreter, looking for the ultimate conception, is not really a genuine inconsistency – rather it shows that we are trying to find a deliberate pattern where there does not have to be one. Our flinch at these “inconsistencies” reveals our expectations. The text itself reveals that even if we may treat the TLP as a rigorous work and Wittgenstein as a rigorous thinker, this rigorousness does not mean “complete consistency in choice of words”.

In the face of this fact, it is less obvious what it is to follow the principle of charity, which in ordinary cases is to try, as much as one can, to find the text consistent and plausible. But in interpreting a text one is not allowed to help either, to improve the text to make it fit where it does not. We are only to try to make sense of it as it is, and we are not allowed to just disregard what does not fit in.

Some interpretative emphasis could be moved away from what one could call “Wittgenstein’s conception of nonsense in the TLP”. Where one might want to look for it, there may only be inconsistent splinters of uses; technical nonsense, something like category mistakes, rejections (as in everyday language). The craving for coherent “conceptions” in Wittgenstein’s work is not always a successful application of the principle of charity, but may reveal expectations to the work which may be out of contact with a potentially human writer. In the case of nonsense, both traditional readings and resolute readings will end up sweeping remarks under the carpet.

Nevertheless, nothing should stop anyone from doing philosophy inspired by Wittgenstein’s work, but that, one should not confuse with interpretation.