Paraphrase is certainly a form of rewording. My late friend Guido Almansi, who also taught in Great Britain for a long time, once published, along with Guido Fink, an anthology of parodies.5 One of the chapters of that amusing collection of masterpieces was entitled ‘Il falso innocente’ (The innocent fake) and concerned unconscious parodies – that is, versions of great works ad usum Delphini which represent a form of interpretation by summary. One of the most preposterous examples of paraphrase was given by the Tales from Shakespeare written at the beginning of the nineteenth century by Charles and Mary Lamb. Let me quote only a few passages concerning Hamlet.
Hamlet’s madness was love. And the queen wished that the good beauties of Ophelia might be the happy cause of his wildness, for so she hoped that her virtues might happily restore him to his accustomed way again [. . .] But Hamlet’s malady lay deeper than she supposed, or than could be so cured [. . .] His very melancholy, and the dejection of spirits he had so long been in, produced an irresoluteness and wavering of purpose which kept him from proceeding to extremities.
Coming to our scene, there is the undoubtedly clarifying paraphrase:
Said the queen, ‘if you show me so little respect, I will set those to you that can speak,’ and was going to send the king or Polonius to him. But Hamlet would not let her go, now he had her alone, till he had tried if his words could not bring her to some sense of her wicked life; and, taking her by the wrist, he held her fast, and made her sit down. She, affrighted at his earnest manner, and fearful lest in his lunacy he should do her a mischief, cried out; and a voice was heard from behind the hangings, ‘Help, help, the queen!’ which Hamlet hearing, and verily thinking that it was the king himself there concealed, he drew his sword and stabbed at the place where the voice came from, as he would have stabbed a rat that ran there, till the voice ceasing, he concluded the person to be dead. But when he dragged forth the body, it was not the king, but Polonius, the old officious counsellor, that had planted himself as a spy behind the hangings. ‘Oh me!’ exclaimed the queen, ‘what a rash and bloody deed have you done!’ ‘A bloody deed, mother,’ replied Hamlet, ‘but not so bad as yours, who killed a king, and married his brother.’
Is that a good translation from Shakespeare? I suspect it is not. But let us go on. We have said that even a comment which draws inferences from a text is a form of interpretation. In the milieu of the French Oulipo it was suggested that the opening line of À la recherche by Proust, Longtemps je me suis couché de bonne heure, could be interpreted by the following inference: It took me a long time to convince my parents to let me go to bed after nine. This is certainly another excellent example of interpretation which is certainly not a good example of translation.
Now, what do all these cases of rewording have in common? Each of them tried to communicate the same content as the original, but with different words. This means that in order to render the same content the interpretation has to change the form of the expression, or the discourse – in the sense that, linguistically speaking, the definitions felis catus or animal which miaows are linguistic strings different from cat. Can we say that even translation proper does the same? From a certain point of view it does, since nobody can deny that if I translate the passage from Hamlet into Italian the linguistic strings would be different. However, why can come? un topo! be considered a correct translation of how now! a rat! when we consider a translation like How now? Any of several kinds of black, brown or grey, long-tailed rodents, resembling, but larger than, the mouse! to be a joke?
Reformulations in other semiotic systems
We must first of all consider the fact that under the headings of what Jakobson called rewording (intended as the reformulation of a given expression within the same semiotic system) one should also put cases of reformulation in a non-verbal system. In music such a procedure is usually called transcription, as when a musical piece is rewritten in a different key, or changed from major to minor or (in ancient times) from the Dorian mode to the Phrygian mode. In these cases it seems there is no change in substance (we are always using sounds, perhaps produced by the same instrument) and both the melodic line and the harmonic relationship are preserved. But there are cases in which a given composition is transcribed for a different instrument. Take for instance Bach’s Solo Cello Suites transcribed for alto recorder. The change in the substance of the expression is at this point relevant, even for an absent-minded listener. First of all, it is true that the melodic line is transported point by point from one instrument to the other, without variations, but the same does not happen with the chords. The bow of the cello can be drawn across more than one string at the same time, while with a recorder one can play no more than one note at a time. The solution is ‘to translate’ a given chord into an arpeggio: the soloist plays several notes one after the other very rapidly so to create the aural impression of performing all of them at the same instant.
Second, there is a shift from a given timbre to another, and this feature is certainly perceived even by the most inattentive listener. Both from the melodic and harmonic point of view the transcription should allow everybody to recognise the original score, but such identification is not as easy as it seems. I play the Solo Cello Suites on the alto recorder, and even though I am a very modest performer I could say I know them by heart. In spite of this, it has happened many times (while doing something else) that I have listened absently to a melody played by a cello on the radio, and have the impression of knowing it but am unable to immediately identify it. It took a certain effort to realise that it was one of those suites I knew so well. Changing the timbre meant that the effect of the piece was a different one. Thus we can say that a mutation in substance has a consistent impact on the listener.
Can we say that this does not happen with verbal languages?
Substance in rewording
In chapter 1 I wrote that at the expression level of a text there is not just one single substance at play. There is the properly linguistic substance, but it is conveyed and supported by many suprasegmental elements such as tone of voice, pitch, the rhythm and speed of the utterance and so on. All these features of an expressed substance do not have anything to do with the language as a system. Accent is for instance a substance phenomenon but, in respect of the form of a language, is considered a suprasegmental business – in other words, an Australian worker pronounces the sentence There is a pot on the table differently from an Oxford professor, but this has nothing to do with English grammar and both the uttered sentences can be paraphrased or translated in the same way. Thus in every new utterance of the same sentence there is a change in substance in the sense that the same expression uttered by two different persons displays two different tones of voice.
Differences in timbre are still linguistic features (Hjelsmlev called them connotators) but are relevant only when we want to recognise someone who is speaking out of sight. They are mainly used as clues in order to identify things or individuals. From the point of view of the sense of the expression, or of the content substance, these features are irrelevant, and the same happens with all cases of rewording such as definitions, summaries, comments, inferential developments. In these cases we are interested in grasping what an expression means, and the physical way such a meaning is conveyed does not interest us.
The same seems to happen with variations in graphic substances. Suppose I produce a written expression and then I reproduce it on this page many times:
Mothers love their children
Mothers love their children
Mothers love their children
Mothers love their children
Mothers love their children
In these five cases we have the same Linear Manifestation, and the physical variations of the five strings are irrelevant (only through a microscope could one detect infinitesimal differences in the printing). Thus I have repeated the same written sentence five times. But now let’s suppose I reproduce the same sentence in three different fonts:
A font is a particular ‘form’ but this change in the form of the graphic expression also represents a change in substance, since different fonts produce different visual reactions. Might we say that these three sentences act on our senses as if they were pronounced respectively by a cockney woman, by an Eton student and by an old Highland shepherd? I think that there is something more: in the graphic case we shold also appreciate the aesthetic aim of the printer, to such an extent that if we consider a pseudo-futurist line like
in an English translation we should also keep the graphic substance as a relevant feature of the poem, thus translating:
The same would happen if the sentence
