1.3.3. Determining Reference in Discourse Processing

When discourses are produced, they do not repeatedly use the same word to refer to a particular object or person. A character may be introduced as ‘Jane,’ and in the next sentence be referred to as ‘she,’ and later in the discourse be referred to as ‘the woman with the brown hair.’ In order to maintain the coherence of the discourse, the comprehender must realize that ‘Jane,’ ‘she,’ and ‘the woman with the brown hair’ all refer to the same person.

The most well-studied case of reference in discourse processing is the comprehension of anaphoric reference, or reference to a person or object that has previously been mentioned (as in the above example). One factor that influences the comprehension of anaphoric references is the nature of the information currently active in memory (and foregrounded in the discourse). If there is one female character currently in the foreground of the model (Jane), then pronouns such as ‘she’ will be mapped onto Jane. If there is more than one female character in the foreground of the model, the pronoun will be mapped to the one whose representation is most active. In some cases, further knowledge is needed in order to determine the referent of an anaphor. When Jane is referred to as ‘the woman with brown hair,’ the comprehender needs to know that Jane has brown hair in order to easily understand the reference as intended.

The choice of reference for an entity in the discourse depends largely on the circumstances in which the entity is being mentioned. When a person or object is initially mentioned, it is usually marked with the article a or an, and the description provided is typically somewhat detailed. Subsequent references are less detailed, and may be made by pronouns or other ‘shorthand’ referring expressions. The more the entity has been mentioned in the discourse, the shorter the referring terms tend to be. The change in referring terms is related to the given/new distinction in language use. Information that is given (i.e., that is already established in the discourse) may be referred to in a shorthand manner because it is assumed that the comprehender already knows what is being talked about. Information that is new (i.e., information that is assumed to be unknown to the comprehender) is described in more detail to ensure that the comprehender has the relevant information to process subsequent discourse.

II. Elements Supporting Discourse: Discourse Markers

Discourse markers can be defined as those elements, such as you know, I mean, well, oh, m, you see, look, listen, that have a distinct prosodic entity, tend not to have a specific semantic meaning, and contribute to scaffold the pragmatic coherence of interaction. Broadly speaking, discourse markers have often been described by their absence of traditional linguistic properties, since neither their use nor their meaning seem to resemble any of the traditional linguistic categories. In fact, these elements have been typically discarded from rigorous linguistic studies, and in different languages have been assigned the general vague name of ‘particelle’ (Italian), ‘fillers’ (English), or ‘muletillas’ (Spanish) – both the English and the Spanish terms evidence these markers’ function of ‘filling’ or ‘supporting’ discourse. In some sociolinguistic domains, discourse markers were marginalized because they were traditionally attributed to incompetent and incoherent speakers of a language (Watts, 1989).

Schenkein (1972) and Jefferson (1978) were among the first that became convinced of the need to study these vague elements that seem to distort syntax and have a multiplicity of undefined meanings. As an instance, consider how Dik characterized these elements as ‘extra-clausal constituents’; their role in natural language was described as follows: ‘‘any natural language text can be exhaustively divided into clauses and extra-clausal constituents . . . which are neither clauses nor part of clauses.’’ In other words, discourse markers, for this author, belong to the realm of spoken language and need to be studied only with reference to the parameters of spoken language.

There are three main approaches to the study of discourse markers: the conversational, the grammatico-syntactic, and the discourse-cognitive.