20th World Congress of Philosophy Logo

Philosophical Anthropology

Impossible Descriptions, Superfluous Descriptions, and Mead's "I"

Ingvar Johansson
Umeå University, Sweden

bluered.gif (1041 bytes)

ABSTRACT: Some kinds of utterances which have an indicative grammatical form seem, for different reasons, to be unable to say something true of the world. Logical contradictions are only the prime example of something the author baptizes impossible descriptions. So-called performative contradictions (e.g., "I do not exist") make up another kind, but there are at least two more such kinds: negating affirmations and performatives which cannot be explained within the philosophy of language. Only philosophical anthropology can explain their feature of "impossibleness," and a distinction between unreflective and reflective consciousness is central to the explanation. Particularly important here is G. H. Mead's distinction between two aspects of the self: the "I" and the "me." Each of the four kinds of impossible descriptions distinguished has its own contrary opposite. These are, in turn, logical tautologies, performative tautologies, affirming negations, and omissive performatives. The last three types as types have not received the philosophical recognition that they deserve. All four fit a general characterization which is given as a definition of the concept of superfluous description.

bluered.gif (1041 bytes)


A logically self-contradictory utterance is not only false, it cannot possibly describe anything. Therefore, it may also be called an impossible description. A tautological utterance, on the other hand, says something true, but it supplies no new information about the world. Therefore, from a common sense point of view, it is a superfluous description. There are at least, I will show, three other kinds of utterances which adequately can be called impossible descriptions and three which can be called superfluous descriptions. Only views which belong to philosophical anthropology can explain the feature of "impossibleness" of some of these impossible descriptions. Logic and philosophy of language cannot explain all our language intuitions.

The terms 'impossible description' and 'superfluous description' will be used in accordance with the following definitions:

Impossible description =def. An utterance with an indicative grammatical form which has a propositional content which cannot (in general or in the context at hand) possibly be used to say something true.

Superfluous description =def. An utterance with an indicative grammatical form which has a propositional content which cannot (in general or in the context at hand) possibly be used to present ("logically") new information about the world.

In both definitions 'propositional content' refers to what is entertained in an utterance. It does not refer to an assertion or proposition in the traditional sense. Some interpretational problems disregarded, propositional content is what Husserl has called "act matter" and "noematic what", Austin "locutionary act", Hare "phrastic", and Stenius (when explaining Wittgenstein's Tractatus) "sentence-radical"; Searle calls it "propositional content". A propositional content can be used, among other things, to make assertions and to make different kinds of performative utterances. Such a content can be supplemented by anything which adds illocutionary force to it, but it is only assertions and performatives which are discussed in this paper.

Logical self-contradictions make up one kind of impossible description, and logical tautologies make up one kind of superfluous description. Below, I will present and discuss the three other kinds of impossible and superfluous descriptions hinted at.


One kind of utterance which fits my definition of impossible descriptions are the so-called performative contradictions. (The label was, I think, invented by K-O Apel.) Utterances like 'I do not exist' and 'I cannot talk' are necessarily false because their propositional content cannot, in ordinary utterances, possibly be used to describe anything in the world. All performative contradictions are first person sentences. If we insert 'He' instead of 'I', then both the utterances become assertions which are either true or false depending upon what the world looks like. The necessary falsity of a performative contradiction does not arise from some logical self-contradiction, but from the fact that the asserted propositional content contradicts some pragmatic presupposition for an utterance; in the examples that the speaker exists and that the speaker can talk, respectively.

A performative contradiction can be described as a logical contradiction between the asserted propositional content of the utterance ('I cannot talk') and a description of one of the pragmatic presuppositions for an utterance ('I, the speaker, can talk'). This fact, however, does not mean that performative contradictions are logical contradictions. Traditional logical contradictions belong wholly to the semantic or the syntactic dimension of a language.

The contrary opposites of 'I do not exist' and 'I cannot talk' are 'I exist' and 'I can talk'. As far as I know, no philosopher has described these assertions with the concept which is the contrary opposite of performative contradictions, namely performative tautologies. But both of them may very well be given that epithet. They can also be called superfluous descriptions. If I say 'I can talk', I say something whose asserted propositional content is true, but the assertion is "logically" superfluous because the propositional content is also conveyed by the sounds made. The same propositional content appears twice in the same speech act, once on the semantic level and once on the pragmatic level. This is so because the asserted propositional content describes a pragmatic presupposition for itself, and pragmatic presuppositions are always implicitly conveyed (cf. Grice's concept of implicature).

In my view, Descartes' famous utterance Cogito ergo sum is a performative tautology. This phrase was used by Descartes in his Discourse on the Method (1637), but in The Meditations Concerning First Philosophy (1641) he re-wrote it in a way which makes my point more easily visible. In the second meditation he wrote: "Thus, after having thought well on this matter, and after examining all things with care, I must finally conclude and maintain that this proposition: I am, I exist, is necessarily true every time that I pronounce it or conceive it in my mind."

An utterance 'I exist' is a performative tautology and a superfluous description since the asserted propositional content describes one of the pragmatic presuppositions for itself. Therefore, 'I exist' is necessarily true every time I utter it, and its propositional content does not supply us with any "logically" new information. The information is always conveyed by the mere utterance independently of content.

The opposite of 'I exist', the utterance 'I do not exist', is necessarily false every time I utter it, and it is, as already said, a performative contradiction.

The peculiar truth of Descartes' Cogito ergo sum and other performative tautologies, as well as the falsity of performative contradictions, can be explained in a philosophy of language which takes the pragmatic dimension into account. No recourse to philosophical anthropology is needed. That, however, does not apply to the sections which follow.


The third kind of impossible description that I have found, I would like to call negating affirmations. In some Christian communities pious people are ranked very high. If, in such a community, a non-pious man says 'I am a perfectly pious man', then of course he makes a false assertion. Curiously, however, 'I am a perfectly pious man' is false even when it is asserted by a man who is regarded as perfectly pious in the whole community. It may at first seem as if the pious man is just giving a true description of himself since other members of the community truly can say 'He is perfectly pious'. Nonetheless, the pious man himself cannot say it. It would be a kind of insolent pride on his part to assert 'I am perfectly pious'. If a pious man makes such an utterance he will thereby become less pious; piousness admits degrees. Really, such a man shall not even think that he is pious. In the kind of communities referred to, it is wholly impossible to say truly 'I am perfectly pious'. If a pious man affirms what others have truly said of him, then, by this affirmation, he negates both their assertions and his own. His affirmation is self-negating; it is a negating affirmation.

Let us now look at the opposite utterance, 'I am not a pious man'. A non-pious man can truly assert it, but what happens if a pious man asserts it? If a perfectly pious man is asked 'Are you pious?', he has to say 'No'. This is not lying, since it is part of the essence of piousness not to regard oneself as pious. His assertion 'No, I am not pious' may adequately be called an affirming negation. To assert the negation is to affirm what is denied. Therefore, the asserted propositional content cannot in the context at hand possibly present any new information about the world. According to my definition, then, it is a superfluous description.

Remarks similar to those made about 'I am pious' and 'I am not pious' in very Christian communities, can also be made about 'I am humble' and 'I am not humble' in some secular contexts.

In order to explain negating affirmations and affirming negations we need a distinction between living unreflectively and living reflectively. The distinction is most easily seen in relation to children. A child may be kind or nasty, it may have self-confidence or lack it, etc.; but independently of what kind of person it is, it can be wholly unaware of the fact that it is such a kind of person. It has consciousness, but it lacks self-consciousness. All normal adults have self-consciousness. Nonetheless, it is possible for adults to lack self-consciousness in some specific respect. A completely normal and sane person can live some of his character traits unreflectively.

In the kind of Christian community envisaged, a perfectly pious man is assumed to be unreflectively pious. That is the reason why he himself cannot truly say 'I am a pious man'. When a pious man is asked about his piousness, he is automatically on the brink of becoming reflectively aware of it, which means losing it. Some sociologists claim, in my opinion rightly, that Western societies have become high-reflective societies. There are in our societies strong structural forces which make every individual very aware of what kind person he or she is in all respects. Perhaps it is soon in fact impossible to live unreflectively. Even pious communities would then have to allow pious people to characterize themselves as being pious. But that does not cancel either the possibility in principle to live unreflectively or the philosophical distinction between living reflectively and living unreflectively. It is a philosophical-anthropological truth that human beings can create both low-reflective and high-reflective communities.


My fourth kind of impossible description is well known. It consists of Austin's classical performatives. In utterances like 'I promise to come' and 'I baptize you John', a grammatical indicative is used to perform an action. Such utterances can be neither true nor false. For instance, to make a promise to come is to perform the action of promising it, not to describe a future action. The utterance 'I promise to come' is not even both a performative and an assertion. It is only an action. However, utterances like 'He promises to come' and 'I promised to come' are ordinary assertions. Performatives are always first person present tense utterances.

The remarks made about 'I promise to come' can be made about 'I baptize you John' and all other classical performatives as well. All of them fit my definition of impossible descriptions. They have a propositional content which cannot possibly be used to say something true. Such a content can of course not be used to say something false either, but that does not alter the fact that, according to the definition, performatives are impossible descriptions.

All the first three kinds of impossible descriptions have, as I have shown, their own specific kind of superfluous description as a contrary opposite, and so have performatives. Their opposite I will call omissive performatives. Let me explain.

The utterance 'I promise to come' can be negated in at least three ways. I can negate it by saying (i) 'I promise not to come', i.e. I can make the opposite promise. Also, I can say, (ii) 'I promise that I will not in the future promise to come'. Here, I give a second-order promise never to promise to come, which is not equivalent to promising not to come. However, it is a third kind of negation which I want to focus attention on, namely (iii) 'I do not (now) promise to come'. In contradistinction to the first and second negations, which are promises and performatives, the third utterance is not a promise and not an ordinary performative. It relates to a promise as an act of omission (e.g. not helping someone) relates to the corresponding action (helping the person). That is the reason why 'I do not (now) promise to come' should be called an omissive performative.

When I say 'I do not (now) promise to come' in a situation where nobody falsely thinks that I have promised to come, then I say something superfluous. Silence can make the same point. The utterance has a propositional content which cannot possibly be used to present new information about the world. It fits my definition of superfluous descriptions. Although it looks like an ordinary performative, it is an omission. Similarly, 'I do not (now) baptize' and other contrary opposites to the classical performatives are omissions or omissive performatives.

All omissive performative utterances are superfluous descriptions. This fact shows clearly that ordinary performatives cannot be regarded as being also assertions. Every ordinary real assertion can be negated, and its negation is a new assertion (not silence). The opposite of a classical performative, however, may well be silence, and that is because such an utterance is only an action and not an assertion. Actions have as their opposites the corresponding non-actions. As the opposite of helping is not helping, the opposite of promising can be silence.

Ordinary performatives are impossible descriptions, but why? The answer is to be found neither in logic nor in pragmatics. It relates to the distinction between living reflectively and living unreflectively which was used in the explanation of negating affirmations. But this time a deeper analysis of this distinction is needed.


Now and then most people loose themselves in some action, be it manual or intellectual. One may be so absorbed in, say, playing soccer or playing the piano that one is not really aware of oneself performing the action. The same applies to reading. Sometimes, after a really good book read under fortunate circumstances, it is as if one wakes up when one becomes aware of the fact that one has been reading for a while. It is as if the self has been somewhere else. The action in question was performed unreflectively. There was consciousness but no self-consciousness.

In some ancient teachings like Zen Buddhism, there is a very high evaluation of such 'no-self' states, both from a religious and a practical-effective point of view. A samurai, for instance, should learn to loose himself in fighting in order to become a better fighter. This strand of thought has staged a come-back in modern sports psychology. Also, it has entered discussions about what the really good life looks like. Many people claim that the best moments of their lives, their 'optimal experiences', were moments or actions where the self was absent. The distinction between being conscious and being self-conscious is of interest far outside the aims of this paper. In my view, the best philosophical analysis of the distinction so far, is to be found in G.H. Mead's classic Mind, Self, and Society.

According to Mead (especially §§18, 21, and 22), we cannot identify the self with consciousness. The defining feature of a self is that it can be both subject and object, or that it can be an object to itself. This, argues Mead, is only possible if the self has two aspects which are separated but belong together. He calls these necessary aspects of a self the "I" and the "me", respectively. Our "I" is what makes us act; it gives us our sense of freedom; it is the principle of action and impulse. Our "me" is the organized set of attitudes of others which we ourselves assume; it is passive. The movement into the future is a step by the "I" not by the "me". When we are self-conscious our "I" looks at our "me" without becoming identical with it. When a self is self-conscious, one aspect of the self looks at another aspect of the same self. What is "I" in a particular moment can be part of the "me" of a later moment, but in one and the same moment the "I" and the "me" are always distinct. Mead writes (§22):

I talk to myself, and I remember what I said and perhaps the emotional content that went with it. The "I" of this moment is present in the "me" of the next moment. There again I cannot turn around quick enough to catch myself. --- As given, it is a "me," but it is a "me" which was the "I" at the earlier time. If you ask, then, where directly in your own experience the "I" comes in, the answer is that it comes in as a historical figure.

When the "I" of a self is directed at the "me", it is necessarily directed towards the past of its self. However, when the "I" is not so directed, it can of course be directed at lot of things both in the present and in the future. I have here taken the liberty of interpreting Mead as if the "I", whatever it is, always has intentionality, i.e. directedness. If the "I" is ascribed intentionality, it becomes understandable why the "I" never can "turn around quick enough to catch /it/self". Intentional acts have a from-to structure. They are like arrows. An arrow points from itself and its apex to something other than the apex. It can, in thought at least, be turned back in such a way that the apex points to some other part of the arrow, but it can never be so turned around that the apex is directed towards the whole arrow and the apex itself. It is with the apex in relation to the whole arrow as with our eyes in relation to the rest of our body. As Mead himself notices (§18), we can see many parts of our bodies, but we can never see the whole body.

With Mead's "I" as a background, let us take a look at performatives again. A performative is an action performed by the "I" aspect of a self. Consequently, it cannot be directed at itself when it is performed. Since it is a speech act, this means that it cannot describe itself when it is performed. Afterwards, in another speech act, it can be described, but at that moment it belongs to the "me". Austin's views on performatives fits in perfectly with Mead's philosophical anthropology. Even more, this anthropology affords us an explanation why a performative utterance cannot simultaneously both perform and describe the same action, or, in other words, why performatives are impossible descriptions.

bluered.gif (1041 bytes)


Back to the Top

20th World Congress of Philosophy Logo

Paideia logo design by Janet L. Olson.
All Rights Reserved


Back to the WCP Homepage