• Keine Ergebnisse gefunden

1. INTRODUCTION

1.2. Description of the corpus

1.2.1. Criteria used for compiling the examples

1.2.1. Criteria used for compiling the examples

Due to the limits of the present study and with the aim of thoroughly analysing an adequate amount of examples, I had to select which constructions to include in the analysis: the preliminary analysis was done on 230 examples (115 representing initial detachments and 115 representing final detachments), from which a number of examples were extracted for the discussion. As will be explained below in two following subchapters, the final number of examples discussed and the principles used in classifying them are somewhat unequal for the two categories, but this difference was ultimately considered acceptable in a qualitative study such as the present thesis. Notwithstanding the differences in the distribution of examples in both categories (initial and final detachments), the discussion will be structured in a way that allows for general conclusions to be drawn, based on the research questions formulated at the beginning of the thesis. The process of selection remains nevertheless subjective, but the examples were selected to be as representative as possible: first of all, a general (typical) pattern was identified that allowed the creation of subcategories of initial and final detachments respectively. The examples were also intended to demonstrate the observed variability in all examples: in all categories, the most frequent types (from a formal point of view) are most represented and some less frequent examples are included on the condition that they are not unique occurrences, but seem to represent a less frequent type.

In addition to the examples from the oral corpus, 13 examples from different written sources are discussed in the study, mostly representing initial detach-ments.

1.2.1.1. Examples of initial detachments

Compared to final detachments, initial detachments form a more homogeneous category in the sense that there are fewer difficulties in identifying these constructions as such: the examples are easier to define and delimit using formal criteria; however, the classification inside the category remains controversial as the criteria to be taken into account have in all cases some

18

problematic aspects and the internal complexity of detached elements seems to be greater than in final detachments.

The choice of examples to discuss was made on the basis of 115 examples, from which 35 were selected and divided into 3 groups.

The classification principles can vary according to the perspective adopted, but in the case of initial detachments it seemed justified to first create a category of examples with a relative clause modifying the detached element (or more generally, the complex detached element), secondly to separately analyse a category of examples where the detached element seems to play a clear role in the structuring of the discourse, and thirdly, analyse the examples where typically the detached element serves to extract an item from a set or occurs as a contrast in a situation of competing referents; this also includes examples where the whole construction contributes to a specification of a referent, by repairs or other means.

Examples have been included in this analysis of initial detachments which contain a lexical element, pronoun or proper noun as a detached element, followed by the main clause, but the biggest group is formed by utterances containing a detached full NP.

At this stage, I tried to find the proper balance by excluding examples whose most important characteristics in the light of my investigation occurred only once or twice in the corpus. Of course, with a larger corpus and a quantitative analysis this distribution could be somewhat modified, but some general principles should remain (in similar types of corpora), for example the fact that lexical elements are far more frequent in a detached position as pronouns or names and that certain grammatical cases (nominative, partitive) occur more often in main clauses.

The following table shows some characteristics and statistics of the examples of initial detachments discussed in the present thesis:

Table 1

Initial detachments Total occurrences in corpus Occurrences discussed in detail Complex detached element

(relative clause) 45, of which 12 are

detached pronouns (39%) 13 Extraction of a referent

(adjustment, contrast) 58 (50%) 18

Resumption of a sequence,

other textual functions 12 (11%) 7

Total 115 (100%) 38

We considered that the so-called ‘accessibility’ or the status of the referent in the discourse cannot be used as a general criterion in order to classify the examples in a systematic way due to the variable interpretation of this

19

property – there is a continuum rather than clear-cut limits between different cases. Therefore, it was decided to use one formal criterion (the complexity of the detached element, relative clause) and textual criteria related to the reference (extraction of one element from a set or contrast between several items and resumption of a sequence by using a detachment construction).

Of the 36 examples, 11 occur in interrogative utterances.

1.2.1.2. Examples of final detachments

The choice of examples to discuss was made on the basis of 115 examples, from which 29 were selected and divided into 3 groups for a more detailed analysis.

Table 2

Final detachments Total occurrences in corpus Occurrences analysed in detail Generalisation, resumption

of a sequence 22 (19%) 7

Referent present in discourse framework

59 (51%) 14

Discourse in elaboration 34 (30%) 8

Total 115 (100%) 29

The question of representativity is somewhat different in the case of final detachments: as the interpretation of the “right periphery” is more complicated due to the “openness” of this construction (the detached element does not necessarily mark the end of the utterance), it was decided to discuss these more ambiguous examples in one group (“discourse in elaboration”) where most of the examples represent cases of different repair constructions. The total number of examples discussed is slightly smaller than in the case of initial detachments, firstly because of recurrent patterns in many unambiguous examples (i.e. the most typical examples do not show great variability, which is why the biggest category is represented with a smaller proportion of examples discussed in detail) and secondly due to a number of examples that are not in the focus of the present study. This mostly concerns examples that merit a separate analysis, with different types of borderline cases where prosodic criteria likely play a more important role; that is why the cases of “discourse in elaboration” are less represented relative to their overall number in the corpus.

Of the 29 examples, 7 occur in interrogative utterances.

20