No beetle? Wittgenstein’s ‘grammatical illusions’ and Dalabon emotion metaphors

Maïa Ponsonnet
Australian National University and Dynamique du Langage (CNRS/Université Lyon 2)

Apart from a few fruitful but pointed encounters, linguistics and philosophy of language often talk past each other. In this post, I try and establish a dialogue between these two disciplines, around the question of private states (or inner states) and their linguistic descriptions. I suggest a ‘translation’ of Wittgenstein’s stance on private states into more technical linguistic terms, and I show how empirical description of the way private states are described in various languages may relate to some of Wittgenstein’s philosophical questions.

Wittgenstein’s stance on private states

Most languages in the world have words to refer to ‘private’ or ‘inner’ states such as emotions (love, fear, anger…) or other affects (pain, hunger…). The ‘ontological’ status of such private states has been an issue of debate in philosophy. Behaviorists discarded them altogether; others, like Ryle (1949), considered such concepts superfluous, and an obstacle to the adequate understanding of human behavior. Wittgenstein (1953) adopted a more nuanced stance.

Wittgenstein does not question the ontological status of such private states. On the contrary, he suggests that the fact that one is tempted to think that private states should ‘have’ an ontological status at all results from a ‘grammatical confusion’, i.e. a misinterpretation of our own linguistic descriptions and of how language works. In other words, private states are not ‘things’ about which it would make sense to ask whether they exist or not. We may characterize these private states to some extent (e.g. pain can be qualified as sharp, darting…), but this does not mean that they are anything else other than purely linguistic devices. This stance on private states, which is closely tied to Wittgenstein’s ‘private language argument’,1 is encapsulated in the parable of the ‘beetle in a box’, in §293 of the Philosophical Investigations (1953), where Wittgenstein discusses pain.

§293. – Suppose everyone had a box with something in it: we call it a ‘beetle’. No one can look into anyone else’s box, and everyone says he knows what a beetle is only by looking at his beetle. – Here it would be quite possible for everyone to have something different in his box. One might even imagine such a thing constantly changing. – But suppose the word ‘beetle’ had a use in these people’s language? – If so it would not be used as the name of a thing. The thing in the box has no place in the language-game at all; not even as a something: for the box might even be empty. – No, one can ‘divide through’ by the thing in the box; it cancels out, whatever it is.

That is to say: if we construe the grammar of the expression of sensation on the model of ‘object and designation’, the object drops out of consideration as irrelevant.

The last sentence indicates how Wittgenstein believes the confusion arises: it results from analogies suggested by our language. In German and English (or other European languages), private states (affects and emotions) are commonly depicted by nouns. An unsophisticated understanding of language—the one Wittgenstein criticizes—leads one to assume that nouns normally denote things (material entities) and, therefore, to implicitly construe private states on the model of ‘things’—‘entities’ that one can see, or point to, etc.

§304. ‘And yet you again and again reach the conclusion that the sensation itself is a nothing’– Not at all. It is not a something, but not a nothing either! The conclusion was only that a nothing would serve just as well as a something about which nothing could be said. We have only rejected the grammar which tries to force itself on us here.

The paradox disappears only if we make a radical break with the idea that language always functions in one way, always serves the same purpose: to convey thoughts—which may be about houses, pains, good and evil, or anything you please.

A linguist’s perspective

Wittgenstein’s concept of ‘grammar’ differs from what linguists call ‘grammar’ and encompasses all rules of language, in particular semantic rules. Wittgenstein’s notion of ‘grammatical confusion’ leaves the linguistic mechanisms which trigger these ontological or philosophical confusions relatively under-defined. Such confusions operate on various levels. One of these levels is general and ‘metaphysical’: our mistake is about how language is supposed to relate to the world. At another level, confusions may relate to the particular architecture of what Wittgenstein calls the ‘grammar’ of our language, as suggested in §78 of his Philosophical Investigations.

§78 Compare knowing and saying.
How many feet high Mont Blanc is –
How the word “game” is used –
How a clarinet sounds –
If you are surprised that one can know something and not be able to say it, you are perhaps thinking of a case like the third.

I contend that linguistic categories and terminology often allow us to specify which analogies trigger the confusions in question. With respect to private states, for instance, Wittgenstein points out that ‘pain’ should not be construed as a ‘thing’. Such a misconception could arise on the basis of an analogy between the following sentences for instance:

(a) ‘I feel pain.’
(b) ‘I smell bread.’
(c) ‘I see a house.’
(d) ‘I bring bread.’
(e) ‘I’m painting a house.’

The resemblances and differences between the respective status of ‘pain’, ‘bread’ and ‘house’ in the above sentences can be straightforwardly spelt out in linguistic terms. The three lexemes are nouns, albeit from different classes (mass/count nouns)—this distinction, a covert one, may easily be overlooked by speakers, resulting in misguided associations. In each of the sentences above, these nouns are treated as syntactic objects. Here ‘object’ denotes a syntactic status or grammatical relation.2 Grammatical relations should not to be mistaken for the semantic roles, i.e. with the role of the referent in the event described. A syntactic object (syntactic function) is not a patient (semantic role):3 syntactic objects can encode other roles than the patient, and patients can be encoded by other syntactic functions than the object. In (a-c), the nouns are the syntactic object of a perception verb: in terms of semantic roles, they are stimuli. In (d), the syntactic object is a theme; in (e), it is a patient. The semantic roles of patients and themes are typically fulfilled by material entities. These semantic roles are also the prototypical roles served by syntactic objects (Taylor 1989/1995:197ff). Because ‘pain’ is treated syntactically as an object in (a), it is associated linguistically with patients and themes, and thus with material entities that can be perceived and physically affected — consistent with the fact that ‘pain’ (like ‘love’, ‘fear’, ‘anger’…) is a noun.

This description spells out the source of what Wittgenstein calls a ‘grammatical confusion’ with further precision. In English, like in German and many languages in the world, emotions and affects are expressed by the same linguistic categories as ‘things’ which can be perceived, modified etc. We may wish to call the linguistic association of a private state with a material entity a ‘generic’ or ‘ontological’ metaphor (Lakoff & Johnson 1980:25; Kövecses 2002:33-36).4 Such metaphors, embedded in the syntax of our language, are less obvious to speakers than figurative metaphors (such as ‘the burden of pain’). Therefore, speakers may fail to perceive the interplay between semantics and syntax that has just been described above. Instead, they associate pain and things because these are treated alike linguistically. As a result, speakers may implicitly construe private states on a par with ‘things’. Linguistic terminology serves Wittgenstein’s own purpose: as Wittgenstein himself argues, precise descriptions bring clarity and may thus undermine grammatical confusions.

Forms of life, language games and ethnolinguistics

Descriptive linguistics may also relate to Wittgenstein’s considerations on private states by offering a ground for comparison with other languages. Wittgenstein often compares our forms of life and our language games with those of ‘fictional tribes’. Here I suggest that real languages may also shed light on Wittgenstein’s philosophical investigations. As for private states — emotions in particular — the Dalabon language, a severely threatened language of Northern Australia (Gunwinyguan, non-Pama-Nyungan), provides an interesting example.

Dalabon is remarkable in that it mostly represents emotions as states of the person, and very rarely as entities independent of the person (Ponsonnet 2013). Emotion nouns are relatively rare. The Dalabon emotion lexicon consists mostly of predicates (verbs and adjectives, as exemplified below, together numbering more than 160 lexemes). There are only two well established emotion nouns. One of these two nouns, yolh-no ‘pep, feelings’, belongs to the class of inalienable parts and attributes of the person: feelings stand on a par with body-parts.

(1) Yolh-no, mak ka-yolh-mon ka-bo-niyan.
feelings-3sgposs neg 3sg-feelings-good 3sg-go-fut
‘His inclinations, he doesn’t feel like going. He’s not up to it.’

The other noun, yirru ‘conflict, anger’, usually refers to conflict (i.e. an observable event, not an inner state) when it is used alone. It denotes anger in metaphorical compounds which, again, depict the emotion as a state of the person rather than an autonomous entity.

(2) Bah korre nga-h-dja-yirru-ngal-minj.
but before 1sg-r-foc-conflict/anger-appear-pp
‘But I was already angry before [I had already anger-appeared].’

As illustrated in these examples, in Dalabon emotions are mostly described as states, by predicates. Most emotions predicates have a syntactic subject which in terms of syntactic role is an experiencer (the person feeling the emotion). Emotions very rarely stand as syntactic objects (and never as subjects). Therefore, the ‘grammar’ of Dalabon does not suggest analogies between emotions and material entities (‘things’). If Wittgenstein is correct about the role of grammar in our conception of private states, and if the misconceptions he has in mind have any reality, we should be able to observe differences in the way speakers of Dalabon conceive of emotions and other private states. Such differences may result, for instance, in different attitudes towards emotions.

Language games and forms of life: correlations?

I have indeed observed some differences between Dalabon speakers and their descendants’ emotion management strategies, and the Western emotion management strategies I am familiar with.

Among my Western social network, a common strategy to deal with negative emotions seeks to eradicate them by identifying their source (and reciprocally for the source of positive emotions). Emotions are construed as foreign ingredients which invade our ‘insides’ and should be eradicated or pushed out. Understanding why the foreign ingredient has come into existence is an important part of our attempts to manage negative or overwhelming emotions. There exist bodies of theories — psychoanalytic theories (Freud 1909/2008, 1930/2002) — that have devised methods to this effect.

Such strategies do not have currency among Dalabon people. Instead, managing negative emotions is a matter of monitoring potential emotional triggers, so as to control emotional levels. Mourning practices are a good example. Immediately after losing someone, before and during the funeral, close relatives are expected to display spectacular emotional behaviour: loud crying and lament, hurting oneself with stones etc. While there is a social function to these behaviors (proving one’s attachment to the deceased), another effect is that these demonstrations exacerbate emotions, thus ensuring a massive negative discharge immediately after the loss. After the funeral, emotional triggers are thoroughly avoided. Among other things, avoidance takes the form of institutionalized taboos on the name of the deceased, which cannot be pronounced. There are also more personal strategies: avoiding certain places, certain topics of conversations; diverting one’s thoughts to different, happier matters (loving and cheerful company, being active…). Later on, thoroughly controlled emotional triggers are intentionally sought, for instance visiting a place related to the deceased, at a favourable time and in adequate company. Throughout the mourning process, emotional inputs are monitored, sometimes pursued, sometimes avoided, in order, eventually, to reach emotional balance and overcome grief. Similar practices apply when managing other negative emotions. In general, negative triggers are avoided, and positive triggers are optimized, so as to maintain a flow of positive emotions. Considering the source and rationale for an emotional state plays little or no role in emotional management. Rather, Dalabon speakers manage their emotions like I would manage my fitness and physical energy, by exercising and resting at appropriate times.

It is easy to hypothesize correlations between these strategies and the respective linguistic treatment of emotions in English (and other European languages) on the one hand, and Dalabon on the other hand. It seems that where a language encourages speakers to construe emotions as things (English), they seek to identify the source of this foreign element to eradicate it (or promote it). Where a language depicts emotions as states (Dalabon), speakers ensure that states are adequately fuelled so as to maintain a certain balance. While these correlations between linguistic representations and emotion management strategies seem plausible, they are also highly speculative. However, further systematic cross-linguistic comparison may either confirm or disconfirm them: studying a broader sample of languages and their speakers’ attitude towards emotions may unveil regular patterns of correlation.

Wrapping up

I have tried to highlight several possible lines of dialogue between Wittgenstein’s philosophy of language and linguistics. In the first paragraphs I began by suggesting a translation (or clarification) of Wittgenstein’s philosophical considerations on language and private states in linguistic terms. In the second part I showed how Wittgenstein’s intuitions on what he calls ‘grammar’ helps us formulate hypotheses on the relations between language, concepts and practice (or forms of life), hypotheses which may then be explored thanks to empirical linguistic and ethnographic description.

To some extent, Wittgenstein’s views may be proved right or wrong, depending on which correlations can be empirically identified. Another aspect of Wittgenstein’s suggestions is more prescriptive, and seems harder to test. Wittgenstein contends that our conceptions about private states of emotions are misguided and confused (since they result from grammatical confusions). Are others conceptions better informed? For instance, since Dalabon does not align private states with physical things linguistically, does this entail that Dalabon speakers have developed more appropriate views on emotions? Is it plausible to imagine that some languages yield more favourable conceptions, and thus more effective emotion management strategies? At any rate, it seems that several strategies are better than only one, and thus if different languages yield different strategies, the study of linguistic diversity may be helpful in this respect.


1  For an introduction to this philosophical issue, see

2 For an introduction on grammatical relations (or ‘arguments), see for instance

3 For an introduction on semantic or thematic roles, see for instance

4 I do not endorse the whole set of cognitive assumptions about metaphors (Lakoff & Johnson 1980).

5 Interlinear glosses. foc: focus prefix; fut: future tense; neg: negation; poss: possessive enclitic; pp: past perfective; r: realis mood; sg: singular.


Candlish, Stewart and Wrisley, George, 2012. ‘Private Language’, The Stanford Encyclopedia of Philosophy (Summer 2012 Edition), edited by Edward N. Zalta.

Freud, Sigmund, 1909/2008. Five lectures on psychoanalysis. Thousand Oaks: BN Publishing.

Freud, Sigmund, 1930/2002. Civilization and its discontents. London: Penguin.

Kövecses, Zoltán, 2002. Metaphor: A practical introduction. New York: Oxford University Press.

Lakoff, George & Mark Johnson, 1980. Metaphors we live by. Chicago: The University Of Chicago Press.

Ponsonnet, Maïa, 2013. The language of emotions in Dalabon (Northern Australia). PhD, Australian National University.

Ryle, Gilbert, 1949. The Concept of mind. London: Hutchinson.

Santorini, Beatrice & Anthony Kroch, 2007. The syntax of natural language: An online introduction using the Trees program. (21 Aug 2013).

Taylor, John R, 1989/1995. Linguistic categorization: prototypes in linguistic theory. London/New York: Clarendon Press/Oxford University Press.

Wittgenstein, Ludwig, 1953. Philosophische Untersuchungen/Philosophical Investigations, translated by G.E.M. Anscombe, P.M.S. Hacker, and J. Schulte. Malden: Wiley-Blackwell.

How to cite this post

Ponsonnet, Maïa. 2013. ‘No beetle? Wittgenstein’s “grammatical illusions” and Dalabon emotion metaphors’. History and Philosophy of the Language Sciences.

Tagged with: , , , , , , , ,
Posted in Australia, Linguistics, Philosophy, Semantics

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: