Contrastive Analysis of German "weil" and English "because" as Discourse Markers in Connected Speech

Table of Contents

1. Introduction
1.1 General approach
1.2 Primary objective

2. Discourse markers
2.1 Characteristic features
2.2 Primary functions

3. Contrastive analysis of English because and German weil
3.1 weil
3.1.1 Propositional versus non-propositional meaning
3.1.2 Functions and examples of weil in corpus
3.1.2.1 weil as a causal subordinating conjunction
3.1.2.2 Epistemic weil
3.1.2.3 weil referring to speech acts
3.1.2.4 weil as discourse marker
3.2 because
3.2.1 Propositional versus non-propositional meaning
3.2.2 Functions and examples of because in corpus
3.2.2.1 because as a causal subordinating conjunction
3.2.2.2 Paratactic uses of because
3.2.2.3 because as discourse marker

4. Results

5. Bibliographic references
5.1 Online sources
5.2 Corpora

6. Appendix

“ Although spoken English doesn't obey the rules of written language,
a person who doesn’t know the rules thoroughly is at a great disadvantage. ”

Marilyn vos Savant^[1]

1. Introduction

Discourse markers mark discourse. Although this paraphrase may not be regarded incorrect, the subject proves to be far more intricate. Since verbal communication is regarded a highly structured social activity, connected speech as part of interactional spoken dialog requires a substantial amount of metalinguistic knowledge from all participants, speakers and hearers alike. When attempting to interpret possible speaker intentions, identifying word, phrase, and utterance borders as well as recognizing occurrences of speech repair and hesitation, the hearer relies on specific cues and signposts, respectively, assisting in structuring information distributed within discourse as well as organizing the alternation of speakership. Recent linguistic research has shown proactive interest in defining and categorizing such devices, including, inter alia, the rather controversial class of discourse markers.

The presented paper aims at investigating various uses and functions of German weil and English because as discourse markers in spoken discourse, also taking into account functional dissociations to other fields of operation.

1.1 General approach

After attempting to provide a general, cross-linguistic definition of the class of discourse markers with focus on functional aspects, both entities of interest, i.e. because and weil, will be examined in detail. Transcripts used for analysis will be taken from the “Forschungs- und Lehrkorpus für gesprochenes Deutsch” (FOLK) exemplifying German weil. Occurrences of because will be illustrated and analyzed by corpus excerpts from the “British National Corpus” (BNC). For reasons of better readability, each object of investigation will be dealt with separately, describing language-specific criteria and exemplifying such by using appropriate corpus excerpts for the respective item. Before analyzing various functions because and weil may serve as discourse markers, their propositional use as causal connectives will be described and exemplified, hoping to establish a more holistic view on the subject. The preposition because of will be excluded from analysis since its use as discourse marker appears to differ from that of because.^[2]

Due to the lack of annotations in FOLK as well as in BNC, aspects of prosody related to discourse markers, such as intonation and stress, will be largely excluded from investigation. A more detailed description of the modus operandi on which the contrastive analysis is based will follow in chapter 3.

The textual extend of the presented paper is substantiated in the complexity of the subject as well as in the amount of transcript excerpts chosen for analysis.

1.2 Primary objective

Authentic spoken discourse is dynamic, spontaneous, and therefore oftentimes unpredictable (cf. Rühlemann 293). So, how exactly do speakers and hearers alike manage to communicate successfully? In which way are discourse markers, such as because and weil, of importance in regards to the speaker’s responsibility of supporting the hearer’s comprehension processes? What functions do they serve, and how may that choice be determined by discourse-pragmatic principles? And finally, is it expedient or even possible to analyze functions of discourse markers separately?

Although the analysis will have to remain of a merely exemplary kind, it is hoped to identify different discourse-pragmatic functions weil and because serve in connected speech, expecting similarities as well as contrasts, both of which will be presented in chapter 4.

2. Discourse markers

Attempting to provide an unambiguous definition of the term ‘discourse marker’ has proven to be an enormous challenge throughout the past twenty-five years of linguistic research. There are various factors which appear to impede the possibility of reaching a consensus on the subject.

It almost seems contradictory trying to find a universally applicable and acknowledged definition for something which, by its very nature, changes constantly: “[…] at any time we care to look at a language […] it is variable and in a state of change.” (Milroy 2). Therefore, what seem to be plausible aspects of a definition today, might call for an expansion or even revision tomorrow.

Despite or possibly due to the fact that discourse markers have not generated special attention within the linguistic realm until quite recently, there is no consensus on exactly which lexical elements the term encompasses or which functions may be considered characteristic (cf. Longman 2007: 1086). Even the simple naming of the phenomenon seems to present a controversial subject. The German “Duden – Die Grammatik” (2009) prefers the label Operator-Skopus-Strukturen (Duden 2009: 1201-1204) implying the existence of a syntactical pattern (cf. Imo 48). Other German linguists make use of terms such as Diskursmarker (Auer/Günthner 1), Gliederungssignale (Weinrich 364), and Vor-Vorfeldausdrücke (Thim-Mabrey 52). Anglicists extend the pool of terms even further, introducing expressions such as discourse operators (Redeker 1991: 1168), discourse particles (Brinton 29, Fischer 2006), discourse connectors, topic switchers, and turn-takers (Jucker/Ziv 1). This variety of competing terms is most likely attributable to the diverse frameworks, different fields of interest, methods, and objectives linguists are attempting to analyze (cf. Schourop 228).

The paper presented is concerned with the contrastive analysis of German weil and English because in their function as discourse markers, which requires the sine qua non of agreeing on non-language-specific criteria applying to discourse markers of both languages. Language-specific aspects such as semantic content, syntactic positions, and individual functions will be addressed in chapter 3. The primary objective of this chapter is not to be interpreted as a clear-cut, static definition including all aspects of this linguistic phenomenon. It is merely supposed to serve as a basic foundation on which the analysis to follow seems representative and utilitarian. As already previously employed, the term ‘discourse marker’ will be the one favored in this paper and continued to be used throughout.

2.1 Characteristic features

One of the first linguists to suggest a more in-depth look at discourse markers (even though not mentioning this term yet) was Stephen Levinson (1983):

“[…] there are many words and phrases in English, and no doubt most languages, that indicate the relationship between an utterance and the prior discourse. Examples are utterance-initial usages of but, therefore, in conclusion, to the contrary […]. It is generally conceded that such words have at least a component of meaning that resists truth-conditional treatment […] what they seem to do is indicate […] just how the utterance that contains them is a response to, or a continuation of, some portion of the prior discourse.” (Levinson 87 et seq.)

Deborah Schiffrin (1987) was able to offer a more detailed definition by analyzing eleven lexical items, which, by her claim, indicate sequentially dependent units of discourse (cf. Schiffrin 31). Similar to Levinson’s proposal, she states segment-initial position and independence from the sentential structure as typical properties of discourse markers (cf. ibid. 32). Her definition, however, is operational describing and differentiating discourse markers by the functions they serve, a topic which will be discussed further in subchapter 2.2.

A further defining cross-linguistic characteristic of discourse markers seems to be that they are pragmatic realizations of lexical items (cf. Lenk 1998: 246 et seq., Auer/Günthner 335, Fraser 1999: 950) indicating that they are semantically bleached (cf. Sankoff et al. 197, Szczepaniak 185) and at least in part have lost their lexical or propositional meaning (cf. Hansen 236, Sankoff et. Al 197, Schourop 227, Lenk 1997: 13). This allows for the high degree of optionality discourse markers often display (cf. Schourop 242, Imo 58 et seq.). However, this assumption must not automatically lead to the conclusion that they serve no vital function or lack semantic content (cf. Schourop 242). Grice argues that what he calls ‘discourse connectives’, such as so, therefore, or on the other hand, bear implicit propositions articulated in the form of conventional implicatures (cf. Grice 362). He illustrates his claim by the following example:

(1) My brother-in-law lives on a peak in Darien; his aunt, on the other hand , was a nurse in World War I. (Grice 362)

The speaker here utters two explicit propositions, first that his brother lives on a peak in Darien (explicit proposition p1), and second that his brother-in-law’s aunt was a nurse in World War I (explicit proposition p2). Assuming on the other hand (implicit proposition pi) carries the meaning of contrasting (p1) with (p2), it can still not be concluded that (pi) contributes to the truth-conditions^[3] of A and B since by erasing(pi) from the entire utterance, the speaker of (p1) and (p2) – assuming that he is speaking the truth – cannot be accused of being a liar (cf. ibid.), which may lead to the conclusion that “a linguistic expression which is non-truth-conditional might nonetheless encode conceptual rather than procedural information” (Wilson/Sperber 13 et seq.).

Other linguists regard discourse markers as lexical items holding a semantic core meaning which may or may not vary from their lexical “twin(s)” in other word classes (cf. Fraser 2006: 197, Lenk 1998: 253, Fischer 2006a: 431). Whether or not this applies to the discourse markers to be analyzed will be illustrated in chapter 3.

In sum, linguists, German and English alike, seem to agree on the following properties to be cross-linguistic characteristics:

Discourse markers occur mostly, but not exclusively, in spoken language (cf. Lenk 1998: 246, Auer/Günthner 335, Gohl/Günthner 60, Brinton 33, Schourup 234), are mostly placed as disjuncts in segment-initial position, oftentimes outside of the syntactic structure of a sentence (cf. Gohl/Günthner 59, Traugott 1995: 6, Brinton 34, Schiffrin 31 et seq., 328, Fraser 2006: 189, Auer/Günthner 336, Longman 2007: 1086), and are often claimed to be a quite heterogeneous syntactic class (cf. Schourup 234) since they constitute “a class of lexical expressions drawn primarily from […] onjuncttions, adverbs, and prepositional phrases.” (Fraser 1999: 931) and are therefore homophonic with their lexical “twins” (cf. Auer/Günthner 335). As a result of their typically reduced semantic content (cf. Gohl/Günthner 60, Sankoff 197, Szczepaniak 185) and increased pragmatic meaning (cf. Lenk 1998: 246 et seq., Auer/Günthner 335, Fraser 1999: 950, Lenk 1997: 13) the scope in which they operate expands (cf. Auer/Günthner 337, Fraser 1999: 943). Even though their sphere of activity increases, discourse markers are strongly characterized by their non-truth conditionality favored by their syntactic position, or paraphrasing: Discourse markers do not contribute anything to the truth-conditions of a proposition expressed by a speaker and may therefore be regarded non-obligatory (cf. Schourup 232, Traugott 1995: 5, Imo 58, Fraser 1990: 390, Sankoff 197, Hansen 236, Levinson 87 et seq.).

When investigating transcribed conversations, discourse markers are often displayed orthographically separate from the “main” utterance implying a prosodically isolated realization (cf. Imo 71). This seemingly distinct audible and, when in writing, visible clue however cannot be regarded a credible identifier for discourse markers in general since a vast amount of discourse markers in use lack a notable connection of prosodic realization and specific functional aspects (cf. Günthner/Imo 14 et seq.). As a result, a more in-depth look at prosodic marking as a possible feature of discourse markers will be refrained from in the course of the presented paper.

Surely, there are more properties applicable in language- or variety-specific contexts, and the attempted definition is by no means exhaustive. Its aim is to simply summarize the most important features of discourse markers in general, fostering a common ground on which to base the following analysis and helping to identify contrasts more efficiently.

When looking at the defining properties of discourse marker noted above it becomes apparent that they seem to be characterized mostly by negative terminology such as ‘non-propositional’, ‘little or no semantic value’, ‘situated outside the syntactic structure of a speech unit’, and ‘non-truth conditionality’, which consequently raises questions of why discourse markers are used, which functions they serve, and why speakers prefer using them as implicit signals of relations rather than uttering their intentions more explicitly.

2.2 Primary functions

When contemplating on the various attempts at defining what exactly discourse markers are, it is obvious that a satisfactory solution seems more probable when focusing on what functions they perform. Several renowned linguists such as Deborah Schiffrin (1987), Karin Aijmer (2002), Uta Lenk (1997), and Eva-Maria Willkop (1988) utilize forms of categorization by function, substantiating a more operational definition. Surely denying somewhat the heterogeneity and depth of the subject, this chapter is to be interpreted as an overview for the sake of providing a general base.

Schiffrin (1987) defines discourse markers quite nebulously as “sequentially dependent elements which bracket units of talk” (Schiffrin 31). She does point out, however, that a more concise definition is only possible when focusing on operational and functional aspects (cf. ibid. 36). She raises three main questions whose answers supposedly create a firmer ground on which to base a more detailed definition: “What do discourse markers add to coherence? […] Do markers have meanings? […] Do markers have functions?” (ibid. 314). In an attempt to shed some light on the subject, she introduces a model of discourse entailing five different planes in which discourse markers operate. Units of talk within the “exchange structure” can be interpreted as turns taken in adjacency-pair-parts, i.e. in inquiry-response-cycles. Their use is based on the decision “by which speakers alternate sequential roles and define those alternations in relation to each other” (ibid. 24). In other words, they operate as signposts alternately indicating speaker-hearer-turn-taking. The “action structure” is proposed as the second plane and describes the patterned, organized occurrence of communicative actions, so-called speech acts (cf. ibid. 25), which is somewhat analog to what Goffman (1981) calls “ritual constraints” (Goffman 21) including appropriate communicative and social skills and behavior (cf Goffman 21, Schiffrin 25). Both, action and exchange structures function, according to Schiffrin, on a pragmatic level by supporting the organization of a communicative structure and influencing the negotiation of such. Thus they can be regarded as linguistic realizations of acts and turns (cf. Schiffrin 25 et seq.).

The third plane constitutes the “ideational structure”, which entails not pragmatic but semantic linguistic units (cf. ibid.). Within this plane the various relationships and interactions between propositions of a particular discourse are reflected, including cohesive relations, topic relations, and functional relations, all being part of structured propositions (cf. Schiffrin 26, Fraser 1999: 934). The “participation framework” seems self-explanatory and constitutes the fourth plane. It includes different ways in which the participants of a conversation are able to be related to their utterances and to one another (cf. Schiffrin 27 et seq.). The final plane Schiffrin calls the “information state” in which the cognitive management and organization of propositions uttered within discourse is being reflected, constantly altered, and assimilated to whatever state seems applicable and rational (cf. ibid. 28).

In all five planes discourse markers function according to Schiffrin as tools to create, or at least support different types of coherence within the domains of structuring dialogue sequentially, turn-taking organization, general speech-management, interpersonal management, topic structuring, and participation frameworks, stressing thereby that all planes and the units they entail are connected to and influenced by one another (cf. Schiffrin 24 et seq., Fischer 2006: 9, Fraser 1999: 934). But due to the fact that this categorization is based on the analysis of only eleven expressions, that it continues to be mostly theoretical, and that its groupings remain hard to identify in natural discourse, her proposal has been subject to criticism throughout research attempts following (Fraser 1999: 934, Jucker/Ziv 2, Lenk 1998: 247, Redeker 1991: 1164).

While Schiffrin (1987) focuses on the multiple roles markers of discourse may play in managing conversational situations with reference to specific domains of discourse, keeping her definition purposely quite vague (cf. Schiffrin 35), a more current approach on defining discourse markers by categorizing them by function is proposed by Uta Lenk (1997). Her model presents an outline which describes discourse markers as a subcategory of pragmatic particles, possibly attempting to provide a more narrow and intrinsic operational definition than the one Schiffrin proposed.

[...]

^[1] Provided by an online source listed in bibliographic references.

^[2] Fraser (1999) makes a clear distinction between the use of because and because of as discourse markers. He justifies his categorization with the different canonical forms they occur in (cf. Fraser 1999: 939 et seq.).

^[3] Propositions constitute the literal meaning of sentences. Truth-conditions determine whether the propositional content of an utterance is regarded true or false. A sentence S with the propositions p and q is true if both propositions are true (cf. Meibauer 41 et seq., 46).