Unlike other semantic components, logical meanings are expressed recursively through the structure of “unit complexes” (Halliday 1978). The recursive structures can be formed by paratactic or hypotactic combinations at each rank, including the clause complex. Clause complexing occurs when two or more clauses are linked together. In written language a clause complex is a sentence. But the term clause complex is used to cover both spoken and written language.
There are two sub-systems involved in the clause complex formation: the tactic system which describes the type of interdependency relationship, and the logico- semantic system which describes the specific type of meaning relationship. Of particular interest in this part of the analysis is the taxis system, as this is where the organisation of experiential meanings can be observed. In the taxis system, there are two options of interdependency relationship: parataxis for an equal, independent relationship, and hypotaxis for a dependency relationship to a main clause.
Research has shown that mature writers and less mature writers employ different strategies in terms of packing and unpacking information, which are reflected in clause combination patterns, grammatical intricacy, lexical density and
77
nominalisations. These features will be investigated as part of the analysis of the clause complex. As mentioned in Chapter 2, features such as grammatical intricacy, lexical density, and nominalisations will be dealt with in the analysis of ideational resources (instead of textual resources). These will be outlined in the following sections.
4.3.2.1 Grammatical intricacy
For the purposes of this study, a sentence comprising only one clause will be referred to as a “clause simplex”, and a sentence comprising more than one clause will be referred to as a “clause complex” (Eggins 2004, p255).
The pattern of clause combination in a clause complex, i.e. whether it is paratactic or hypotactic, will be analysed. Halliday (1985) uses Roman numbers for parataxis and Greek letters for hypotaxis as notation. In this study, however, dependency will be highlighted by means of bold-face. Table 4-6 exemplifies clause complex types in parataxis and hypotaxis:
Table 4-6: Examples of complex types in parataxis and hypotaxis (based on Halliday 1985, p83)
Parataxis
(equal, independent relationship)
Hypotaxis*
(dependency relationship to a main clause)
1 a) “and/ or”-type complexes b) “i.e./ e.g”.-type complexes c) “then/ so/ but”-type complexes
1 a) “besides/ instead of”-type complexes b) “non-defining relative” complexes c) “when/ because/ if”-type complexes 2 direct (quoted) speech complexes 2 indirect (reported) speech complexes
* Dependent clause can be finite or non-finite. Embedding is not included in hypotaxis.
In parataxis, two independent clauses relate to each other as equals. In written language, the clauses in a paratactic clause complex may be linked by adjacency and punctuation as well as by conjunctions. In hypotaxis, only one clause can stand alone as the main or head clause, with the other clause acting as a modifying or dependent clause. The dependency relationship in a hypotactic clause is signalled by either a hypotactic conjunction (e.g. while, if, when) or a relative pronoun in a non-defining relative clause (e.g. who, which, that).
78
When expanding the meaning of a clause, spoken and written varieties of language may differ in their paratactic and hypotactic clause patterns. In spoken English it is more common to link chains of independent clauses paratactically together, using the conjunctions and, then, so and but (Butt et al. 2000, p175). In written English, however, writers typically express the relationship between events in a more complex way, resulting in more crafted clause combinations of hypotaxis. For example, chains of independent clauses they are poor and they cannot live extravagantly are recombined into a complex clause made up of an independent clause and a non-finite dependent clause Being poor, they cannot live extravagantly.
When compressing the meaning of a clause, formal, written language sometimes uses embedding. By using clauses as Post-modifiers in nominal groups, complex meanings are packaged into noun groups for increasingly compact constructions. Embedding allows writers to pack more meaning into a single unit, so it requires some forethought as to how to construct a clause.
An embedded clause will not be counted as a separate clause. Because it is embedded inside a clause constituent, its original clause status becomes lower in rank, i.e. it is rank shifted: “it now functions merely as part of a clause constituent rather than as the clause constituent itself” (Butt et al. 2000, p73). The sentence below illustrates this shift. From now on examples demonstrating the linguistic analysis are taken from the two essays that will be analysed in this present study. Any errors or non-standard use of English are as in the students’ original text:
Young people [[growing up in a wealthy family]] usually go to
prestigious schools// and get higher education,// have more chances [[to get a good position with a higher salary]].///
The above sentence is a clause complex comprising three clauses (the boundary of a single clause is marked with // and the end of a clause complex with ///). Embedded clauses, shown inside double square brackets “[[ ]]”, will not be taken into account in the calculation of grammatical intricacy.
Grammatical intricacy (GI) is calculated by dividing the number of clauses in a text by the number of sentences in the same text, as follows:
79
Total number of clauses : Total number of sentences = Grammatical intricacy (GI)
In order to calculate grammatical intricacy, sentences must be divided into single clauses. What counts as a clause may differ according to grammatical theory; however, Halliday (1985) suggests that “for comparative purposes, the main requirement is consistency”. A general guideline from Halliday (1985) is that “the clause is a functional unit with a triple construction of meaning”, or “every clause embodies some pattern of selection” in transitivity, mood, and theme (p67).
4.3.2.2 Lexical density
To calculate lexical density, words must first be separated into two groups: grammatical (“function”) items, and lexical (“content”) items. Grammatical items are those that function in closed systems: determiners (including articles), pronouns, most prepositions, conjunctions and some classes of adverb. Lexical items are content- carrying items; these include nouns, the main part of the verb, most adverbs and adjectives. In contrast to a function item, a content item enters into an open system. Its class membership, or the set of items that it is related to, is limitless. For example, the word income can be related to salary, money and fringe benefit, but also to travel allowance and holidays, and even to tax and family status – the lists continue. To demonstrate the classification of grammatical and lexical items, consider this sentence: For a long time, socio-economic elements have influenced our families’ sizes and status in both developing and developed countries. In this sentence, there are 19 total words. Of these, 11 (underlined) are content-carrying lexical items, and 8 (non-underlined) are grammatical items.
Although many items clearly belong to one kind or the other, some items are intermediate cases. Modal adverbs, for example, are on the border between the two categories and can be assigned to either kind. Again, in such cases, the main issue for comparative purposes, as suggested by Halliday (1985), is consistency.
To apply the above methods of analysis to the students’ essays, after classifying the words in each sentence in their texts into the two categories, the total number of the students’ original words in each paragraph is then calculated using the Count Toolbar in MS Word. The researcher’s added corrections (as text manipulation
80
mentioned in Section 4.2.6), the students’ direct quotations and references will not be counted because these are not of the students’ creation. So, for example, in According to a study by Bloom, “health improvements accounted for as much as one-third of East Asia’s economic miracle” (2001, p5), only According to a study by Bloom will be considered for lexical items and word count.
Finally, lexical density (LD) is calculated by dividing the number of lexical items in a text by the total number of all the words in the same text (Halliday 1985). For ease of comparison, this proportion is multiplied by 100 to find the percentage. The calculation is as follows:
Number of lexical items : Total number of words x 100 = Lexical density (LD)
The calculation above assumes the same contribution from both high-frequency and low-frequency lexical items. In fact, however, these two categories do not contribute the same weight towards lexical density as some high-frequency lexical items almost function as grammatical items, for example, thing in I like that thing; make in the child makes me tired. These do not contribute much to lexical density. In reality, it is low-frequency lexical items that play a greater role in making text dense. But for immediate practical purposes, all lexical items are treated alike.
4.3.2.3 Nominalisation
It was stated above that the world of experience is represented in the transitivity system, with the process as the core of experiential meaning. The typical forms realising process, participant and circumstantial elements were then introduced. However, speaking of “typical forms” of course suggests that they are not the only forms available for meaning making. The typical or unmarked form is said to be the “congruent” realisation of a meaning (Halliday 1994, p343), with the opposite being “incongruent”. The congruent relationships of meaning to form are listed in Table 4-7 (Martin 2001b, p225).
81
Table 4-7: The congruent relationship between meaning and form (Martin 2001b, p225)
SEMANTICS (Meaning) GRAMMAR (Form)
Process verbal group
participant nominal group
quality adjective logical relation conjunction
assessment modal verb
Nominalisation is a grammatical feature of a mature writing. The term nominalisation is being used here to refer to a form of incongruent language, where there is a mismatch between meaning and form. For example, congruently a process meaning is realised by a verbal group in the grammar. Incongruently, a process meaning is being construed as if it were an entity and is realised in a nominal group. Halliday (1994, 2001) refers to the phenomenon of shifting from the congruent to the incongruent grammatical structure as “ideational grammatical metaphor”. Such incongruent realisations predominate in written language (Halliday 1985) – particularly that written by adults in academic contexts.
As mentioned above, nominalisation is the process of turning elements that are not normally realised as nominal groups into nominal groups. However, a few special notes for analysing nominalisation are needed here. First of all, as the essay prompt is Socio-economic factors govern success in life. Do you agree?, the nominal groups socio-economic factors and success in life will not be analysed as nominalisations. Instead, these terms will be viewed as a “pre-packaged” metaphor given to the students. The students do not need to construct these nominalisations by themselves. What is more important in this study is to see how the students demonstrate their ability to create nominalisations in text. Second, transcategorisation (making a noun out of a verb, e.g. by adding the suffix -ion or the suffix -er) is not the same as nominalisation. Although they “tend to draw on the same resources of derivational morphology” (Derewianka 1995, p229) and may appear to have the same form, they should be analysed as either transcategorisation or nominalisation according to their context of use. Derewianka (1995) suggests: “Where both a congruent and a metaphorical option is viable, then we could say that the instance is metaphorical. Where it cannot be “unpacked” to a more congruent meaning, it would be regarded as
82
simply a case of transcategorization” (p229). Thus, in Success in life is the result of the socio-economic factors, the nominal group the result of the socio-economic factors will be analysed as a nominalisation: its more congruent version can be “The socio-economic factors result in success in life”.
Furthermore, established technicality, where a term can be “taken for granted” within a particular field, will not be analysed as nominalisation because “it is no longer the case that both the congruent and metaphorical options are equally “at risk”, and the term has become a “faded metaphor”” (Derewianka 1995, p230). Again, there are cases “where the same term might be used metaphorically in one instance and technically in another” (Derewianka 1995, p230). For example, in they are able to receive reasonable income from their jobs, the word income is a faded metaphor. In contrast, in people in high income families have opportunities to…, the wording high income families will be analysed as a nominalisation in this context as high income changes from its more congruent form as Qualifier (families with a lot of income / having high income / which have high income) or even as an independent clause (Some families earn a lot of money), to Classifier (high income families).