Functional grammar, first introduced by Dik, is “a general theory of the grammatical organization of natural language based on the functional view of the
nature of language” (Siewierska, 1991, p.1). It views the organization of language from the perspective of human social interaction, emphasizing the communicative functions of grammar and the appropriate use of a particular linguistic form in a given context. In other words, because the focus of functional grammar mainly rests on communication, the relationship between the forms and functions of language becomes significant. Bloor and Bloor (1995) contended that the same words can convey different functions on the one hand. For example, “Good afternoon” (p.8) carries the function of greeting. However, when a teacher says “Good afternoon” to a student, who enters the classroom after half of the first class, the function is reproach with sarcasm, instead of greeting. Halliday (2004) also indicated that the relationship between forms and functions is complicated and based on the principle that utterances make sense in the very context where they are being said. On the other hand, the identical function can be realized in various ways, as shown in the following examples:
[2.21a] Close the door, Kate.
[2.21b] Can you close the door, Kate?
[2.21c] It seems that you still leave the door open, Kate.
Examples [2.21a] to [2.21c] express the same idea of asking Kate to “close the door”
even though it is encoded into sentences with different tones and forms.
Siewierska (1991) claimed formal grammar is structural illustration irrespective of language use. Lock (1996) argued that traditional grammar stresses the grammaticality of sentences and the relation between them instead of “their meanings or their uses in different contexts” (p.1). In contrast, functional grammar not only includes the subareas like semantics, syntax, morphology, and phonology in formal grammar but also views them as tools to achieve “communicative and interactional
functions” (Siewierska,1991, p.3).
How are active voice and passive voice considered in formal and functional grammar respectively? Lock (1996) indicated that formal grammar focuses on the formation of active and passive construction and the transformation rules, explaining how to switch the subject and object position and how to change the form of verb. In contrast, functional grammar is more concerned with questions such as “(1) how the communicative effect of the message in the sentence is different in either active or passive voice; (2) what the effect is of putting agent at the end of a clause; (3) what features of the context may have led the writer to select passive rather than active voice” (p.2). Therefore, voice from the perspective of functional grammar is viewed as a device to promote communication.
In written discourse, voice choice is closely related to thematic progression as the voice of a sentence is usually dependent on the selection of its subject and hence influences subject continuity/discontinuity (Pinkster, 1985, p.115). The use of the same subject in two or more successive sentences can develop subject continuity while the change of either or both of them can cause subject discontinuity.
In the following, we will investigate the roles voice may play in organizing information and facilitating thematic progression from the perspective of functional grammar. Specifically, we will introduce information and theme system and their relation to voice selection. The ideas of theme/rheme, given-new information, text cohesion, texture, and thematic progression are also presented. It is hoped that the discussion can shed light on voice choice in research articles.
Information System
According to Halliday (2004), the function of grammar can maintain the discourse flow by two interrelated and parallel systems: one is “system of
information” while the other is “system of clause” (p.88). The former is made up of information units. The latter, often called Theme system, views clause as a message, and analyzes it as the combination of theme and rheme. In the following, we will first explicate the discourse functions of passive voice in relation to the system of information and its pattern.
The system of information is “a form of discourse organization” (Halliday, 1976, p.175). An information unit is not exactly correspondent to any unit in syntax, but a unit of the same status as a clause. It can be operated within less than a clause or extended to more than one. Thus, one clause may comprise more than one information unit and a single information unit can stretch to more than one clause. It is proposed that information is the interaction between what has been known and unknown or what has been predictable and unpredictable (Halliday, 2004, p.89). The former is often called “given” while the latter is termed “new.” Given information is
“recoverable to the hearer” and new information, “non-recoverable” (Halliday &
Hasan, 1976, p.27). In other words, given information is what a reader or hearer can refer to in the previous text or “something that is not around at all but that the speaker wants to present as the given for rhetorical purpose” (Halliday, 2004, p.91). On the other hand, new information can be either based on which to be introduced or something not expected by audience. In addition, Bloor and Bloor (1995) noted that given information is an element to enhance the effectiveness of communication since it represents the shared and mutual knowledge as the background in a particular context. Furthermore, a default form of an information unit consists of a given and a new constituent. Halliday (2004) suggested two “natural” ways for structuring given and new information: (1) new information is the position for information focus; (2) the arrangement of information should follow the pattern of given to new. He contended that new information represents a speaker’s choice of information focus.
The element with this feature of prominence, no matter in spoken or written language, is the one that the whole sentence intends to introduce or even center.
Since new information is characterized by prominence, it becomes easier for us to tell “where the new element ends” (Halliday, 2004, p.89) in a given clause. In most cases, new information is recognized as the last element in an information unit. Hence, the usual sequence of an information unit is one in which the given precedes the new (Bock & Irwin, 1980; Chafe, 1979; Daiker, Kerek, & Morenberg, 1986; Kuno, 1980;
Prince, 1978; Quirk et al., 1972). This given-new pattern forms a standard order, also known as an unmarked form, in the information system. The following two examples illustrate the unmarked and the marked information structure respectively:
[2.22] Unmarked information structure:
My mom bought a blue sweater yesterday. The sweater is very beautiful.
Given New [2.23] Marked information structure:
My mom bought a blue sweater yesterday. Beautiful the sweater is.
New Given
Kopple (1986) further proved that sentences conforming to the given-new pattern can contribute to text readability because it does not require readers to hold the new information when they wait for the given one. Daiker, Kerek and Morenberg (1986) also claimed that the given-new pattern can contribute to the continuity of thought and
“carry the thought further” (p.197). This pattern, therefore, governs the word order of a sentence in context and, in turn, usually determines the subject of the sentence. For example,
[2.24] A scanner moves along the printed text and photographs the words. The words are then converted into electronic signals and sent into a minicomputer.
In [2.24], first note that “the words” is new information in the first sentence but becomes old information in the second. Further note that the determination of “the words,” the given information, as the subject of the second sentence affects the voice choice of the second sentence.
However, Lock (1996) also noted that the given-new pattern may not be the only influencing factor in voice choice. The underlying element is the context itself. In constructing a sentence, instead of choosing between active and passive voice, writers should first select the most appropriate element and the most important information to thematize in a given context in order to fulfill the purpose of communication (Lock, 1996). For example,
[2.25] What potential remedies exist for the acid rain issue? The experts disagree.
Some say new environment laws should be enacted to control the emission of pollutants in the atmosphere. Some say that if we had known how serious acid rain was, we would have planned better to prevent it. Nevertheless, all agree that if the consumption of fossil fuel were reduced, we would have less of a problem. Another possibility is that special scrubbers could be installed in smokestacks to remove a good portion of the pollutants before they get into the atmosphere (Saslow & Mongillo, 1985, pp.141-142).
From the first sentence of [2.25], we can know its purpose is to discuss “the potential remedies existing for the acid rain.” In other words, what emphasizes here is different solutions to solve the problem of acid rain. Thus, in the ensuing clauses, when the potential remedies are proposed, that is, “new environmental laws,” “the consumption of fossil fuel,” and “special scrubbers,” they appear at the initial positions and, in turn, affect the voice selected in each clause in order to conform to the purpose of this passage. This phenomenon reflects the idea that in order to achieve communication,
context may play an influential role when writers choose between active or passive voice.
The transformation from active to passive functions to change the given and the new constituents in an unmarked sentence. Most important of all, passive transformation may also allow the agent to become the focus because of its place near the end of a clause (Chen, 1991; Lock, 1996). For example,
[2.26] Despite this system-gridlock tendency, technological innovations do occur.
Raymond Nelson’s National Research Council Workshop suggests that the explanation is both obvious and paradoxical: technological innovation is produced not by technology but by design (Lock, 1996, p.235).
The first sentence of [2.26] introduces the new information “technological innovations.” Later, when it occurs again in the sentence “technological innovation is produced not by technology but by design,” it instead occupies the initial position as given information. Moreover, in this sentence “by technology” and “by design” not only appear as new information but also function as the most important information,
“which is appropriately placed at the end of the clause” (Lock, 1996, p.244). In other words, from the perspective of information system and readability in text, passive voice is no longer merely a grammatical category, but an effective device to promote the flow of information by rearranging the information focus in the sentence and conforming to the given–new pattern. Therefore, from the functional perspective, voice plays a significant role in organizing information flow in discourse.
Theme System
Theme system, to a certain degree, is like information system. Both of them are responsible for the maintenance of the flow of discourse and operate at the clause
level. Therefore, just like information system, theme system comprises two constituents, that is, theme and rheme. In the following, the definition of theme and rheme will be presented first. Then, we will discuss the relationship between theme system and voice.
Halliday (2004) proposed that theme is the element which functions as “the point of departure of the message; it is that which locates and orients the clause within its context” (p.64). Rheme, then, refers to the rest of the message. Since theme in English is often demonstrated by position in a clause, writers usually mark theme by placing it in the initial position. Furthermore, from the perspective of functional grammar, theme is discussed in a larger sense beyond the level of individual clause. Martin, Matthiessen and Painter (1997) suggested that since themes in clauses unite together to form the context of a text, theme system involves “the current point of departure in relation to what has come before, so that it is clear where the clause is located in the text—how its contribution fits in” (p.21). Here, “the current point of departure” refers to theme. The rest of the clause that directs where it moves is called “rheme.” Thus, theme can also be interpreted as the element to present the scope of discourse in order to predicate the following constituent (Dik, 1981, p.130).
In addition to the similarity between the two components (that is, given and new, and theme and rheme) in information and theme system respectively, theme system, like information system, is also divided into unmarked and marked pattern. Since the typical location of theme is at the initial position, grammatical subject and theme often coincide. Thus, when an element functions as both grammatical subject and theme, the clause is said to have an unmarked theme while any clause with a theme not in the initial position is said to have a marked theme. Unmarked and marked themes can be illustrated in the following two examples.
[2.27]Unmarked theme:
His father was born in 1943.
Theme
[2.28]Marked theme:
In 1943, his father was born.
Theme
In [2.27], since the first element (theme), “his father,” functions as the subject of the clause “His father was born in 1943,” it is an unmarked theme. In contrast, although the subject of [2.28] is also “his father” while its theme is “in 1943” because it occupies the initial position of the clause “In 1943, his father was born.” Thus, [2.28] forms a clause with a marked theme. Following the same concept of marked form in information system, marked theme should also be avoided when writers construct a clause. Researchers (Halliday, 1976; Lock, 1996) have indicated that the transformation from active to passive voice is a device to avoid marked theme. For example,
[2.29] The house, he built in 1983.
[2.30] The house was built in 1983.
As indicated earlier, clauses often take “subject” as “theme” in unmarked theme pattern. Nevertheless, clause [2.29] presents the object “the house” in the initial position and functions as theme, which results in the occurrence of a marked form while clause [2.30] constructs an unmarked theme with the inversion of subject and object and application of passive voice. In other words, with the switch of active to passive voice, writers are provided a chance to thematize clause participants such as
“Goals, Recipients and phenomena (i.e., participants mapped on the direct and
indirect objects in active voice clauses)” (Lock, 1996, p.233) and avoid a marked theme in a given clause.
Aside from the function to avoid marked theme, we can also observe another significant relationship between theme system and voice. For example (Bloor & Bloor, 1995, p.74),
[2.31] Guitar was played by Ziggy.
Theme
[2.32] Ziggy played guitar.
Theme
In [2.32], it can be inferred that “Ziggy” functions as the point of departure to introduce what Ziggy did—“played guitar.” In other words, the aim of [2.32] is to describe Ziggy’s ability or talent in playing the piano. On the contrary, [2.31] frames
“guitar” as theme and seems to present the lineup of a band, just as “Piano was played by John, drums by Ali” (Bloor & Bloor, 1995, p.74). As a result, we can see that although two clauses are constructed with the same constituents, voice choice is related to the selection of theme and rheme.
In research articles, passive voice is generally perceived as a primary device to reflect the objective nature of scientific investigation and the replicability of research irrespective of human agents (Huckin & Olsen, 1983; Quirk et al., 1972; Royds-Irmak, 1975). A number of studies (Cooray, 1965; Duskova & Urbanova, 1967; Fernalld, 1977; George, 1963; Givón, 1993; Heslot, 1982; Master, 1991; Robinett, 1980;
Swales, 1976) have further demonstrated the idea by examining its occurrences in scientific research articles. Wingard (1981) proposed that passive voice is used in reporting the sequence of procedure in author’s own research and in depicting procedures usually used by other researchers by analyzing six medical texts. Tarone,
et al. (1998) also investigated the use of voice as well as tense in two astrophysics journal papers and concluded that passive voice occurs in the following three circumstances: (1) describing standard procedurals in experiment; (2) contrasting the author’s own research with previous research; (3) indicating further research the author suggests. Nonetheless, only few studies (Gramley & Pätzold, 1992; Shaw, 1992) have related the choice of passive voice to information and theme system in research articles. Gramley and Pätzold (1992) claimed that one of the reasons for using passive in scientific research papers is to front the theme focus, that is, direct or indirect object, by putting it in the initial position in a clause. Shaw (1992) also proposed the relation between voice and word order as follows:
Rhetorical function and/or cohesion determine theme-rheme structure, which determines word order. Word-order determines what shall be the subject and hence determines voice….Voice [is] therefore not chosen to indicate function, but in most cases follow naturally from higher-level decisions (p.312).
We can conclude that, in research articles, the selection of voice concerns more than grammaticality and style. The choice between active and passive voice is not merely a question concerning if agents are either known or important. It is closely related to which information should be the shared knowledge or background and which participant is the most appropriate one to be thematized in a given context. As a result, the use of either active or passive voice in research articles should be examined at the discourse level, particularly the information and theme-rheme system and context.
In the next section, therefore, we will relate passive voice to text cohesion and thematic progression in order to gain more insight on voice selection in research articles.