Non-Pronominal Coding of Active Referents The purpose of this paper is to explore the concept of English sentence structures with regard to non-pronominal coding of active referents. In order to do this, it is important to have a baseline definition of non-pronominal (NP) coding and active referents. We look to recent literature and case study of not only English but other languages examined to understand sentence structure. Upon reviewing the literature, it was found that definitions for the pronominal approach were plentiful and easy to understand. As a means of comparison an understanding the NP application, we are also exploring the pronominal approach that acts as a framework for literature. Once these definitions are established, we will look at active referents and their role in sentence structure. Available literature suggests non-pronominal coding is used for active referents. As part of this analysis, it is important to look at other languages as they tend to reintroduce referents sooner than English. These languages vary greatly from the English's structure which can provide a whole different meaning to what is written or spoken. This study looks to languages like Hebrew, German, Maltese, Russian and Wardaman for examples of this theory. In order to check this theory, a piece of literature has been selected to see if non-pronominal coding of active referents and reintroducing referents in languages other than English is a viable theory. The piece of literature chosen was Hans Christen Andersen's The Princess and the Pea, written in English. For the sake of comparison and contrast, it is important to look to real life situations to see if this theory happens there as well. These findings will be compared to the original theory in the hopes acquiring new insight into sentence structure and possible oral and written meanings.

Non-Pronominal Coding of Active Referents

Hartwell Francis et al., establish a direct relationship of the grammatical role of subject is the syntactic expression of the discourse role of the topic (1). Making this clarification allows for the referent to be identified so that a status can be established with regard to the topic of the sentence. A referent may be established in the topic role in the very act of commenting about it. Topics tend to be associated as textually evoked referents. Evoked status is prominently found in pronominal coding. It is logical that active referents would be associated with non-pronominal coding. For example, "my sister has a, she just had a baby. He's about five months old and she was worrying about going back to work" (Hartwell et al. 3). In this sentence, the baby is introduced as an indefinite referential non-pronominal in object position and then reference to the baby is continued with pronouns, beginning in the subject position as a clause topic. Lambrecht concludes, the speaker exploits the potential for easy activation of the family member referent and "conveys a request to the hearer to act as if the referent of the NP were already available" (114). Outside the familial pronoun, the use of NP functions as a reactivation of a topic for which there exist competitors in the intervening discourse. According to Baumann and Grice, "a referent is already active in the listener's mind at the time of the utterance" (1). A referent can also become active from a previous inactive state or even active from a semi-active state. A referent is accessible if by text or situation opportunities. It is contingent on the speaker or writer experience. In this respect the utterance will be different depending on focus (pronoun), pitch or accent. Understanding an expression comes down to coherence and focusing. These factors and their differences have been ignored by language interpretation. Such approached as PA and NP may work in well-formed familiar texts but do not aid in cross cultural or cross generational coherence active or non-active referents (Grosz, et al. 45)

This method is called the Pronominal Approach (PA) because of the role it assigns to paradigms of pronouns that are proportional to constituents containing lexical (i.e. non-pronominal) elements. (The relation termed proportionality is defined below.). The PA first examines the set of "pronominal sentences" and their internal relations. It thus is based on language-immanent observations such as those to which Harris's procedures are applied. We will also show how it furnishes a constructive proof2 of basic syntactic concepts, and in particular reinforces Harris's String...

Consequently, the investigation of the relation between syntactic constructions has gained interest during the last decade. This led to many interesting insights into the syntax-semantics interface and to proposals for a highly structured and hierarchical design of the lexicon (Pustejovsky 1995).
Harris exploited a pair-wise equivalence relation between constructions (e.g. Harris 1969a [1970:615]). He suggests entailment between constructions as a criterion for establishing construction groups or networks leading to a syntactico-semantic typology. Recent work in the PA combines both aspects: it establishes relations between two or more constructions of a verbal predicator, based on the observation that terms are shared in those constructions; moreover relationships are classified according to criteria such as the following: is syntactic function of terms preserved or not, is the relationship predicator-specific or not, is one construction implied by the other. This results in a network of relations between constructions, some of which may be combined. For an application to French.

In Harris's and Whiteley's approaches, construction slots are usually filled by lexical phrases. In Whiteley's approach this contains a risk: the relations of implication or entailment might be established only for particular lexical items on a more or less ad hoc basis, and therefore would not prove the existence of systematic relationships (of logical implication) between the constructions as such, independently of the lexical items filling the slots. Whiteley reduces the risk of establishing unjustified implications between constructions by using variables representing referents of the constructions (Whiteley 1960 150) . The PA takes this approach one step further by using pronominal paradigms instead of variables. In this way it applies the principles underlying the following quotation from Quine: "To be is to be in the range of reference of a pronoun. Pronouns are the basic media of reference; nouns might better have been named pro-pronouns" ( 19964, 13)

In order to obtain a workable definition of the pronominal paradigms for syntactic analysis, the notion of proportionality is essential for the establishment of construction groups. This concept is procedurally defined by establishing the ratios between the lexical (sub) syntagms of a sentence and their pronominal counterparts (cf.

Harris's String Analysis provides a series of systematic tests for the identification of syntactic elements, among which the excision or omissibility test, but as shown in Eynde (1998:147-150) this test in fact presupposes the foreknowledge of the correct structure, based upon the proportionality with minimal referentials.

Contrary to concepts as NP, VP..., to syntactic functions (subject, object,...), or to roles (agent, patient,...), pronouns are elements that belong to the language itself and pronominal sentences are directly observable: they are open to judgment of grammaticality or acceptability. Moreover, the inventory of pronouns is finite and constitutes a closed set.

These observations constitute the basis of the PA: when establishing the valency of a predicator, we exploit this proportionality in order to reduce the huge number of combinations between lexical elements to a much smaller number of combinations with pronouns. In other words, the finite nature of the list(s) of pronouns makes it possible to examine their combinations with predicators (in our dictionaries the full verbs, i.e. lexical as opposed to function verbs such as auxiliaries or modals) in a systematic and exhaustive way, without having to appeal to particular semantico-interpretative features. The possibility or impossibility for a particular pronoun to appear is indeed meaningful: the pronouns reveal the primary (i.e. basic) characteristics a predicator imposes on its dependents.

This presents the basic notions of the PA progressing from the basic elements, the pronouns, to paradigms of pronouns and relations between paradigms and predicators forming constructions. It considers relations between constructions with a view to verb classification. Examples from French will illustrate the basic notions. It will be clear that pronouns and pronominal paradigms can be used in many different contexts such as discourse analysis, prosodic analysis and text cohesion. However, we will concentrate exclusively on aspects related to verb classification.

In the first part of this section, we describe how the basic elements, the pronouns, are identified, delimited from lexical elements, and further classified. Next we present the organization of pronouns in paradigms. Furthermore, we describe how paradigms are related to predicators, thus forming constructions, and finally how relations between different constructions of the same predicator can be established.

Referents and types of referents

Degree of referential specification

The degree of specification distinguishes pronouns from lexicalized constituents, i.e. constituents containing lexical elements. The degree of lexicalization coincides with the degree of specification: the pronouns are generic elements with minimal reference; the lexicalized forms are necessarily more specific. Thus, (1)…