Paul Ricoeur on Metaphor

This time I will not translate my notes but try to put what I remember into words. The exam's in two days, tick tock.

Metaphors are interesting for linguistics and language philosophy alike because they are riddles that need to be solved, they have a twist. According to cognitive linguistics a metaphor is a mapping of a structure from one cognitive model to another. Philosophy has viewed it somewhat differently. In Ancient Greece, for example, metaphors were viewed as rhetorical devices that could be replaced, hence the substitution theory. Later, however, it was decided that not all metaphors could be removed and literal expressions used in their stead. Think of time in terms of space: on Monday, in June, in five minutes etc: they are all metaphorical expressions. The substitution theory lost its meaning.

In the twentieth century philosophers treated the metaphor in three different ways. Firstly, it was claimed to be a clash of ideas. If we call Jane a cow, it is obvious that Jane as a person cannot be a cow. What we do then, prompted by some sort of pressure to understand this absurd proposition, is solve the riddle by means of guessing which features of a cow (in our culture) can be projected onto Jane, her looks or perhaps behaviour. This brings us to the second view, according to which there is not as much of a clash between the word and its superficially absurd new meaning as may be thought, since we use metaphors only when certain things are similar to our construal of the world. Every metaphor is thus a reduced simile. 'Jane is a cow' means 'Jane is (in some way) like a cow'. This does not work everywhere, I must add, as 'Let's meet on Monday' cannot be turned simply into 'Monday is like an object we can be on so let's meet on it.' The same is true of the third approach, which characterises the metaphor as a miniature poem. In a poem the most important function is that focusing on the message, claimed the famous scholar Roman Jakobson, which is why Ricoeur and other hermeneuts prefer to analyse innovative metaphors that have to be guessed by the reader/interpreter and leave 'dead' metaphors such as 'Time is money' or 'Ideas are parcels' and their various popular expressions to linguists.

In poems we usually deal with metaphors where X is Y, that is one person/thing is called another person/thing. Edgar Allan Poe called the female figure in his poem 'an island in the sea (...) a fountain and a shrine'. Needless to say, Poe's metaphors quoted here are not part and parcel of everyday English and need to be interpreted rather than understood without effort just as we understand notices saying 'First floor' or 'No smoking'. According to Ricoeur, whenever we encounter new metaphors and manage to reveal their meaning, we can observe the surplus of meaning, something that simple substitution cannot give the text. Here it is also important for the text to be written. When we talk to someone and fail to understand the speaker's figurative language, we can always ask the person to explain what they mean. In writing, however, there is distance that allows for no substitution on the speaker's/writer's part and therefore needs effort on the reader's part. Enter the hermeneutic circle. We have a naive understanding of the text as a whole, then solve the little riddles disguised as metaphors, allegories and symbols, the latter two including metaphorical expressions, analyse the structure of the text and go back to understanding of the entire text, this time a better one. The circle may continue, depending on the depth of figurative language and how saturated it is with symbols, for example. Alternatively, we may not want to interpret it any further.

 The difference between the metaphor and the symbol is that while the former belongs entirely to language and can be interpreted without reference to tradition or culture at large, the latter cannot. Ricoeur was interested in symbols and wrote his most famous work on the symbolism of evil but later he decided that symbols were too difficult to interpret by means of language alone and decided to deal with text, discourse and metaphor instead. The fragment of his book about symbols seems the most difficult to me, in particular when he describes Eliade's views on hierophanies and their (omni)presence in probably all aspects of human existence. For me it's one of the places where philosophy leaves the steady ground of empiricism and says things that can be neither proved nor disproved.

Finally, I may attempt to characterise the structure of the metaphor. It was traditionally thought of as lying in a word. However, as the Ancient Greeks noted, the essence of a word's meaning lies not in it itself but in how it is employed in context. Ricoeur prefers to use the term 'metaphorical expression' in order to show that the metaphorical lies not in identification (onoma) but in predication (rhema) and they together form a metaphor. Thus in 'Jane is a cow' it is not just 'cow' that is metaphorical but the entire expression, the absurd placed in the auxiliary verb - after all we could replace the 'is' with 'likes' and the entire sentence would change into a plain assertion of a fact or opinion. Thus 'X is Y' is the structure in question. And yet we have such expressions as 'arrive at a conclusion', 'get at something' (all phrasal verbs are idioms) and 'Let's meet on Monday'. This is something that hermeneutics appears to leave to linguists.

I just hope I'll be able to remember that much during the exam. :)

Paul Ricoeur on Discourse

I'm studying for a philosophy exam these days and I think it's a good idea to combine studying with blogging, in particular since I encountered hermeneutics and modern philosophy of language only very recently. For the exam I'm supposed to become familiar with Part 3 of Hans-Georg Gadamer's Truth and Method, some of his essays on language and its role in human life, as well as quite a few chapters of Language, Text, Interpretation by Paul Ricoeur. Gadamer's writings contain many references to Plato, Herder and Humboldt, let alone Heidegger, while Ricoeur cites de Saussure and language philosophers such as Beardsley and Black. The exam and the reading list items are in Polish, so I might not always use the same terms as the ones widely accepted by Anglo-Saxon hermeneutics. If you detect any errors please leave a comment.

Paul Ricoeur in his Interpretation Theory: Discourse and the Surplus of Meaning criticises Ferdinand de Saussure's structuralism for dividing language into langue and parole, thus forgetting about discourse. If Ancient Greeks believed in the inseparability of language and thought (logos) and Aristotle stated that a noun and a verb (or onoma and rhema, as far as I can tell) could only together make an assertion, then there is no reason, says Ricoeur, why discourse should be left out. He suggests two categories of indivisible units: the sign and the sentence. The sign is potential in nature whereas the sentence is an actual event which cannot be simply reduced to a sum of its parts, be they syntactic, morphological, lexical or phonological. The analytical approach belongs to semiology and Ricoeur is not interested in that as a philosopher. What he focuses on instead is semantics understood as concerned with the meaning of discourse.

Discourse is both an event and (a) meaning. It occurs at a given moment (in particular a spoken utterance) and is therefore an event just like parole, which is the only realisation of the fully potential system of a given language. However, discourse is more than that since it can be understood again, repeated verbatim or by means of other words. The same or almost the same meaning can also be expressed in another language or translated. And yet a given discursive act retains its content. Hence discourse is also something resulting from the event-meaning relation.

Discourse is also assertion or predication. Every sentence needs to have a predicate and a predicate cannot be in opposition to other units of the same kind as is the case with phonemes, for instance. While phonemes or morphemes can be put in superordinate categories, the sentence and its predicate cannot be said to be part of anything larger. Besides, according to Benveniste the functions of predicates in language cannot be compared to those of predicates in logic. Finally, it can be stated that while a subject identifies something and thus singles something out, a predicate describes an action, state, event or relation and is universal in nature. Consequently, discourse cannot be treated as a set of subjects and predicates since it is not analytical. It can only be considered as synthetic in terms of structure since neither of its parts alone can function as a piece of discourse.

Discourse depends on the dialectic of event and meaning in a given sentence. An utterance is not to be treated predominantly as an event if a meaning analysis is undertaken. What we want to understand is the meaning of a sentence or text, something constant rather than transient but not the same as langue. Language is meinen, or intending, thanks to the act of self-deletion of a transient event (Aufhebung). In this way we can perceive an utterance as something said once and 'gone' (unless we consider audio or video recording) and yet something understood and recognised as part of our (native or otherwise well known) language. As far as recording goes, however, it is only in a given situation that Dasein, 'being there', is the same for both or all of the interlocutors. Recorded by whatever means, a given situational context can no longer put the viewer and/or listener back in time.

Ricoeur also writes about the self-reference of discourse, speech acts as well as sense and reference but I think that's too much to put in one post. If I have enough time I'll also write about Ricoeur's ideas concerning metaphors, which are different from what cognitive linguistics has to say on the issue and yet make perfect sense from the philosophical viewpoint.

Reference: Ricoeur, Paul. 1989. Język, tekst, interpretacja. Transl. by Piotr Graff and Katarzyna Rosner. Warsaw: PIW.