Adamson, Robin

2018-01-31T11:18:03Z

2018-01-31T11:18:03Z

1984

This description of the language of the war poetry of Paul Eluard is the result of a research project which had several aims. The first was stylistic: to attempt a largely objective description of the poetry which would be of interest and use to other Eluardians and to those studying literary style. A second aim grew out of the first. If the description was to be linguistic (in the sense that it was to use language as its starting point) and objective, the methods used would have to be assessed for their usefulness in producing such a description. Hence this second, but not subsidiary, aim: an exploration of method.

These first two aims, the stylistic and the methodological, were largely complementary. The methodology evolved combines computer analysis and statistical calculations with close textual study, and the final result, although often apparently far removed from the 'mechanical' procedures, owes much to them both. From the outset considerable use has been made of an index and word frequency count and of concordances produced by computer. This involvement in computing led to the decision to prepare the final manuscript in machine-readable form for printing by Lasercomp, a decision which, although it considerably lengthened the preparation time, means that both the data and the conclusions are easily accessible and readily available to others.

As it is presented here, the work retains considerable methodological speculation, as well as documented conclusions. Many questions remain unsolved, and at least some of the conclusions are tentative. In its broad outlines however the method followed is based on the thinking of various groups of linguists in the fields of computer-aided literary research, linguistic statistics, semantics and syntax. The study has been divided into three sections: Lexis, Semantics and Syntax, and in each of these sections a number of theoretical hypotheses are examined. From among these are selected those which were most useful in analysing this particular corpus. Thus, while it is hoped that the methodological conclusions drawn can be applied to other studies of a similar kind, it is possible that particular problems require individually tailored solutions.

What seems clear from this study is that the provision of an index and word frequency count as a basic tool of literary analysis must be considered highly desirable. Similarly, the possibility offered by computer analysis of producing concordances of various kinds puts the researcher in a very powerful position vis-a-vis his text, and allows the exploration of many avenues closed to ordinary research because of their very time-consuming nature. Not that computer-aided research can, by any stretch of the imagination, be described as time-suvz'/zg. For any but those lucky few who can have their texts prepared and programs run by computer staff or research assistants, the preparation and processing of texts by computer is a long and very wearing business.

The various kinds of statistical analysis presented here provide a mass of useful information whose potential is far from exhausted, and theories of semantics and syntax, in particular those of the structuralists and post-structuralists, offer a number of very useful tools which have been used in stylistic analysis.Only some of them, those which seemed most useful, have been tried here. Perhaps the most important idea in all this is the one of potentiality, for when at last the computer print-out is ready, the statistics calculated, the semantic fields and structures identified, the syntactic patterns recognised, the stylistic analysis still remains to be done. The criteria we apply from this point on are essentially subjective and aesthetic. Although the decisions about stylistic relevance are based on sound information, they are nonetheless personal, and they take us away from the comfort of what can be objectively verified into the heady world of hypothesis.

The attempt to strike a balance between the objective and the subjective, the methodological and the stylistic, gives to the chapters which follow something of the air of a Pilgrim's Progress. Considerable detail is given of the manner of setting out, and of the journey, an early stage of which, the parsing of all the words in the index, seemed rather like imprisonment in Doubting Castle at the mercy of Giant Despair. In including some of the 2 theoretical background to the analytical methods adopted, an effort was made not to wander into By-Path Meadow, and, in the application of the methods chosen, to avoid alike the Mountain of Error and Mount Caution. It has been thought preferable to retain the atmosphere of searching, of travelling hopefully, rather than simply to present conclusions, for it is doubtful whether, in such an undertaking as this, one can ever be certain of having achieved a "safe arrival at the desired countrey".

It has been a privilege to work on the poetry of Eluard, to perceive through his words, his poesie resistante, some glimpses of eternal truths. Like Faithful, Eluard and his companions at arms have achieved immortality:

http://hdl.handle.net/1842/26279

The University of Edinburgh

Annexe Thesis Digitisation Project 2017 Block 15

Already catalogued

La poesie resistante: lexis, semantics and syntax in the war poetry of Paul Eluard

Thesis or Dissertation

Doctoral

PhD Doctor of Philosophy