Excerpt-Africa

New Impressions of Africa

By RAYMOND ROUSSEL

Atlas Press

Copyright © 2004 Atlas Press
All right reserved.
ISBN: 1-900565-09-9

Introduction

Raymond Roussel's New Impressions of Africa is perhaps this strange writer's strangest work. He began work on what would be his last book of poetry in 1915, but did not submit it for publication until 1927. A further five years of revision and correction of the proofs then followed, with Roussel frequently adding large sections of verse to his already intricate structure, to the despair of his typesetter.

The basic construction of each of the work's four cantos is the same: a brief "impression of Africa" is interrupted midway by a parenthesis; this new passage is then interrupted by a further parenthesis and so on until a maximum number of five parentheses is reached and they begin to close again. We then discover the second parts of the various interrupted passages and finally the closing section of the initial sentence, which can be quite a distance away. To take the example of the long, second canto, 608 lines separate the subject of the initial sentence from its main verb. There are also occasional footnotes, some also containing their own parentheses, which increase the confusion still further since they must be read in sequence otherwise the rhyme scheme breaks down. In the case of the fourth canto, there are more lines in the notes (134) than in the main body of the text (98).

The verse is standard early twentieth-century alexandrines (the classic French twelve-syllable line, equivalent in usage to English's iambic pentameter) with reasonably regular caesuras after the sixth syllable. It is written in rhymed couplets that respect the traditional French alternation of 'masculine' and 'feminine' rhymes (the latter ending in a mute e, the former not). The only rather unusual point about these rhymes is the high proportion of rimes riches, in other words the rhyming of two homonyms, or more rarely homographs: for example, the opening lines of the first canto rhyme porte (door) with porte (3rd person singular of the verb porter) and then neuf (the ninth) with neuf (new). While this practice is far commoner in French than in English (where it tends to sound rather limp), Roussel does seem to be uncommonly fond of it.

The book's basic structure is so unusual and disconcerting to read that its underlying logic has led to much speculation, which has been further fuelled by Roussel's own pronouncements about how long it took him to write it. He told Robert Desnos: "I spent twelve years on that relatively short book. According to my calculations, each line represents about fifteen hours' work" (quoted in L'Intransigeant). While doubts have been raised about the trustworthiness of such calculations, it certainly remains true that Roussel felt that it had been extremely difficult to write. Why is this? There have been several hypotheses. In his book on Roussel, Michel Foucault suggests that the reason may be that revising any single line in such a structure inevitably affects the succeeding lines with a consequent knock-on effect on the entire structure. However, my experience in translating it showed that revising particular lines generally has only a local effect and is no more damaging to the overall scheme than would be the case in a more conventional one. I feel that the explanation lies more at a line by line level. It seems clear that this book, like most of Roussel's other poetry, was written using a series of bouts rimés, or imposed rhymes, which he altered or abandoned only when he really felt incapable of progressing. In other words, as opposed to practically all other poets, who start their poems with the first word, arrive at a rhyme word, then try to find an appropriate rhyme to go with it, Roussel began by listing rhymes down the right-hand margin of his pages, and then filled up the rest of the lines as best he could. Using imposed rhymes and remaining coherent is already quite hard, but it becomes fantastically difficult in the context of Roussel's poem for two reasons: the coherency between the beginning and end of a parenthesis must be respected and the imposed rhyme greatly limits the possibilities to do so; secondly, and more importantly, in the composition of the lists of comparisons which occur. If we take the example of the second canto's long list of "things that look similar except for their size", it must have been especially hard to find appropriate comparisons using certain imposed rhymes. This may well explain the bizarre nature of some of the examples and their often tortured syntax, a result of Roussel bending language into whatever shape was required to fit his scheme.

Concerning this and other similar sections, it is also interesting to note that some critics have seen these features as examples of Roussel's "larding", a fairly facile way to lengthen a work which an author may feel is too short by progressive and repeated insertion of new lines. Though this might have been the case, writing the lines to be inserted was no easy process for the reasons mentioned above and also because, while it is relatively easy to add fresh lines when they are complete and can be intercalated between existing rhymes, it is quite difficult to do so when the new examples do not fall into convenient metric patterns. This is precisely what happens in the longest list in the poem. The first two comparisons end at line breaks (page 81):

- the machine that Franklin found, Which harmlessly draws lightning to a pit, For plain grey thread passed through a needle's slit; - For those on a new marshal's uniform, The three white stars joined in a firework storm;

But then the correspondence between syntactic and metrical units breaks down:

- By a central line bisected, a blackboard For a priest's bib; - when temperatures have soared, A thermometer's bulb near bursting full For a round-headed pin; - leashes that pull Warm collars without dogs - and so they miss them! - For parasols' slip-cords; ...

This pattern continues over an astonishing 404 lines, in which the closest we come to correspondence is in some short passages in which the rhymes are crossed (p.89):

- Marching policemen, for swarms that appear Of sad boarders at a holiday fête; - The little o that mutes draw on their slate, For the circle that covers a blackboard;

But generally syntax and metre are completely disassociated, with enjambments in every line, until we arrive at the end of the list (p.153):

- The low chapel where our heads hit the stone For the vast Dome that looms in mid Cologne; - For the harsh iceberg, born at the North Pole, The thin block that for drinks, with due control, We in the kitchen smash to lumps of ice; - A fisherwoman's hairnet full of lice, For spider-crabs slow-ranging through a trawl;))))

It thus seems likely that the initial list contained only the first two examples and the last three; exactly why Roussel chose to expand it into its final, monstrous size remains a mystery. What is certain is that it was not from a desire to find an easy way to lengthen his book. On the contrary, it was one of the reasons why it took him so long to write it.

Another original aspect of the structure has also led to what is, I think, the wrong explanation. The role of the footnotes works in exactly the same way as the parentheses, in other words they interrupt the linear reading of the book, and it is not always easy to see why Roussel decided to choose one or other solution at a given moment, since some footnotes could easily be new parentheses and vice versa. A commonly held hypothesis is that he did not wish to confuse the reader by using more than five parentheses at the same time. This misapprehension seems to have arisen because of the new Pauvert edition. It has large pages, containing far more lines than in the original Lemerre edition, every other page bears text and Zo's illustrations have been relegated to the end. This completely alters the experience of reading it. To take a simple example, the second footnote in the first canto is forty-five lines long. In the Pauvert edition, the reader turns over the page, reads the note to the end, then picks up his reading again at the top of the same page. In the Lemerre edition, the reader has to turn no fewer than five pages to reach the end of the footnote, then turn five pages back to find his place once more in the main text. It is a well-known fact that Roussel was obsessive about the presentation of his books and he must have realised that, far from making the reader's life easier, his use of footnotes often complicates it.

It should also be noted that the longer footnotes mean that the reader often sees one of the illustrations three times: once when reading the footnote, then again when turning back to where the poem was left off, and once more when continuing reading the main body of the poem. This, too, has an interesting effect which must have been foreseen by Roussel.

The presence of Zo's illustrations has also generally been explained as another ruse to make the book longer while they have also come in for a certain amount of criticism (hence their relegation in the Pauvert edition). The story behind their commissioning is well known. Instead of approaching Zo himself (which he could have done), or of choosing another "better" artist, Roussel asked a private detective agency to act as an intermediary. The artist was presented with a list of 59 brief descriptions (e.g. "Saint Louis in prison in Damietta") to be used to draw the illustrations. The first time Zo saw the actual text was when Roussel sent him his complimentary copy. He was horrified to discover who he had been working for and wrote in a letter: "Allow me to tell you that I bitterly regret the fact that you wove such an impenetrable mystery around our collaboration. These are not the drawings I would have produced had I known that I was illustrating Raymond Roussel!" And that is precisely why Roussel did not want him to know. What he clearly had in mind was just this plain, unadorned, impersonal style, produced without direct reference to the text. Again, given Roussel's interest in the appearance of his books, it would seem unlikely that he would have used Zo's illustrations if he had not been happy with the result. The attentive reader will notice that their positioning has been carefully chosen to add an often strange or even comic effect to the poem. They are part and parcel of the text and of the experience of reading New Impressions of Africa. Salvador Dalí was surely right when he said: "the choice of illustrations once more displays Raymond Roussel's genius".

One point about this overall structure that has not been properly discussed by other critics is the dizzying speed at which the poet can take us from Saint Louis's imprisonment in Damietta to, for example, the instructions for operating a lift and the behaviour of winterers in Nice, in a completely logical way, and then take us back in an equally logical way via the behaviour of a jockey when being photographed, until we reach the final dolmen. What must have fascinated Roussel was this possibility of being able to go just about anywhere from any given starting point and then return to that starting point with an almost mathematical precision. As Raymond Queneau famously put it, in his perceptive review of New Impressions: "[Roussel has] an imagination which joins the mathematicians' delirium to the poets' logic".

It is also interesting to note that the four starting points of the four cantos seem likely to have been real "impressions" gleaned by Roussel during his visit to Egypt with his mother. The je at the start of the fourth canto would then be the first literal use of the first person singular in one of Roussel's books before the title of his posthumous work, How I Wrote Certain of my Books. Furthermore, if these parts are indeed autobiographical, they stand as exceptions to Roussel's oft-quoted claim: "I have travelled a great deal ... [yet] from all these travels I never took anything for my books. It seems to me that this is worth mentioning, since it clearly shows that for me imagination is everything". Perhaps one of the points of the structure of New Impressions is to show how a truly imaginative writer can use a fleeting (and intrinsically not very interesting) experience to take his readers anywhere he wants and then take them back again. No one other than Roussel has explored the rich potentiality of this technique.

As regards the uncut pages, they too contain a mystery and further illustrate the conscious artistry that went into the conception of this book as an object. What immediately strikes the reader is that it is not necessary to cut them in order to be able to read the book, because the text appears only on every fourth page. Furthermore, because the pages are uncut along the top edge they hide the illustrations to a certain degree, but not completely, as would have been the case if the fore-edges were uncut. There must be a reason for this, since leaving the top uncut instead of the fore-edge is more difficult technically and so more costly. It would seem that Roussel wanted the illustrations to be seen as though through curtains being opened. There even seems to be an "instruction for use" in illustration number 28: "A man sitting by a table on which a book is positioned vertically; he is prising apart two uncut pages to read a passage". What if this man were reading New Impressions of Africa? The reader would then be opening the pages to look at a man, who is opening the pages to look at a man, who is ... A perspective of infinite regression opens out which, to a certain degree, mirrors the structure of the book, the multiplying parentheses of which tend towards infinity inside a short sentence, rather than progressing in a linear way. This illustration occurs during the long list of "things that look similar except for their size" and the comparison drawn with the man opening the pages is also suggestive: a mother is opening the curtains of a crib, just as her own mother may have done before her and her baby may do in a few decades' time, thus allowing another infinite succession of generations into the future, and back into the past. A final indication of what Roussel was trying to do comes in the last illustration (number 59): "A stretch of starry sky without any landscape, apparently seen from a point in outer space and giving an impression of the infinite" (my italics). From what we know about Roussel, the choice of words here can hardly be anodyne. So, should the reader cut the pages or not? All I can say is that, when I receive my copies, I'll be leaving them uncut.

Continues...

Excerpted from New Impressions of Africa by RAYMOND ROUSSEL Copyright © 2004 by Atlas Press. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.