A Critique of the Nature of Theme in Systemic Functional Grammar
Abstract 
The notion of theme has been under much debate since it was first advanced by the German-French scholar Henry Weil in 1887. Based on Halliday’s definition of theme, this paper critically reviews the evolution of its nature in systemic functional grammar through detailed examination of the definitions proposed by scholars like Mathesius, Travnicek, Fries, Huddleston, Matthiessen, Martin and Fawcett. Firstly, theme and given information are two different notions and should be kept apart. Theme is what the speaker choose to take as the point of departure, while given information is what the listener already know about. Secondly, it is not appropriate to simply define theme as what the message is about. Theme is what locates and orients the clause within its context. Thirdly, Halliday’s notion of multiple theme is suspect of equating the initial position with theme. And Fawcett is right to distinguish between elements which always come early in a clause and those placed here as a thematic choice. 
Key words: theme  given information  systemic functional grammar  

The concept of theme in modern linguistics can be traced back to 19th century when the German-French scholar Henry Weil points out that “the syntactic march is not the march of ideas” (1887:21) in The Order of Words in the Ancient Language Compared With That of the Modern Languages. The syntax is the image of a sensible fact, while the succession of words is the progression of the thought. “Syntax relates to the exterior, to things; the succession of the words relates to the speaking subject, to the mind of man” (1887:30). And “there are in the proposition two different movements: an objective movement, which is expressed by syntactic relations; and a subjective movement, which is expressed by the order of the words (1887:43).” He believes that “the order of words corresponds to the order of ideas” (1887:36) and “there is a point of departure, an initial notion as something present and known, in order to reach out to something less present or unknown” (1887:29). 
Mathesius, representative of the Prague school, accepts and further develops Weil’s idea and defines “the starting point of the utterance as something known or at least obvious in the given situation, and from which the speaker proceeds”. Mathesius (1939) ragards theme as an element about which something is stated or the basis of utterance; and what is stated about the basis of utterance is the nucleus or rheme. Mathesius also finds out two varieties of the theme-rheme order: the objective sequence and subjective sequence. Objective sequence means that theme occurs before rheme as this sequence relates to the cognitive process of acceptance on the part of the addressee, while the subjective sequence refers to the situation when rheme precedes theme since the order is related to contextual factors. In this regard, Mathesius’ idea of “objective” and “subjective” have different meanings in contrast to Weil’s uses of the words “objective” and “subjective”. 
Travnicek (1962) criticizes Mathesius’ proposal, because the “conveyance of known information is not the essential feature of theme”. He argues that an essential feature of theme must be its ability to be truly general in character and to remain invariably the same in all cases. Therefore, he defines theme “as the sentence element that links up directly with the object of thought, proceeds from it, and opens the sentence hereby”.
theme vs given information
The salient difference between Halliday and Mathesius is that Halliday clearly distinguishes theme from given information. Although they are usually expressed by the same words, their meanings are different from his point of view. As he points out, “given” means “what you were talking about” (or “what I was talking about before”, while “theme” means “what I am talking about” (or “what I am talking about now”); and, “as a student of rhetoric knows, the two do not necessarily coincide” (1967:212). Meanwhile, he argues that the two pairs of concepts (theme/rheme and given/new information) belong to different linguistic systems and he repeats his argument in his essay Language Structure and Language Function. 
According to Halliday and Hasan (1976:325), given and new “is expressed in English by the intonation patterns, and it is therefore a feature only of spoken English”. In other words, information is tightly interconnected with tone group. The information systems specify information as structural units and relate them to the previous discourse. He further points out that the theme in a clause is what is prominent for the speaker while the new in an information unit is what is made prominent. The elements of given and new comprise parts of the information structure, while the takeoff point of the message, theme, comprises (together with rheme) part of the thematic structure.
Halliday (1985) admits that “the Theme is what I, the speaker, choose to take as my point of departure. The Given is what you, the listener, already know about or have accessible to you. Theme+Rheme is speaker-oriented, while Given+New is listener-oriented” (1985:278). Theme sometimes is the given while new is the rheme, and the opposite is also common, especially in oral English. New information, signaled by prosodic prominence, may conflate with theme. For example:
(1)	Who took her to the party? 
(2)	John did.
In this conversation, John is the theme and conveys new information. The information structure is not directly a structure of the clause, but one of the information unit signaled by tone group in spoken language. In my opinion, it is necessary to separate thematic structure from information structure. But we can not separate them completely or we would go to the other extreme. 
The second edition of An Introduction to Functional Grammar did not change much on this issue. However, the third edition of the book has some adjustments. Firstly, the content of the speech part is compressed. The original Chapter 8 is withdrawn. Two sections on given/new information have been merged into Chapter 3, which discusses themes/rheme. Secondly, there are some changes in the formulation of the discourse construction in the last section of Chapter 9. Secondly, there are some changes in the formulation of discourse construction in the last section of Chapter 9. While distinguishing structural and cohesive text resources, the author also introduces another classification. The text resources are divided into discourse transitions and discourse identity. The former includes the clause complex of the structural part and the connection of the cohesive part, and the latter includes the thematic structure and information structure of the structural part.
Fries(1992: 464) tries to give an overt signal of the location of new information in written language, arguing that “new information falls at the end of the clause”. This argument resonates with Halliday’s (1990) claim that the main item of news has an unmarked location at the end. The location of new makes it reasonable for new to conflate with rheme. Considering their frequent co-occurrence of new and rheme at the end of clause, Fries (1992) coins a new term N-Rheme, which is usually put at the unmarked end of a clause that bears some contrastiveness against that of theme, to indicate the last constituent of the clause for convenience of text analysis and states that N-Rheme typically expresses the core of the newsworthy part of the clause.  Nevertheless, the status of the rest part of a clause except N-Rheme is still problematic to identify since it is hard to define its status along a fixed standard in functional linguistics.  
theme vs what the message is about
Halliday begins to discuss the notion of theme in 1967 when he firstly clarifies his idea of it in his essay Notes on Transitivity and Theme in English. He has described theme in different ways. We should notice that there is a change in Halliday’s definition of theme. The changes of the notions imply that theme features the function of transitivity gradually. We can see that Halliday treats theme as a functional unit and the association of theme with initial position as a description of how theme can be recognized in English. Clearly an approach to theme which uses the description of its function as its definition requires a robust definition which can be easily applied to many specific cases in a non-circular manner. It is also clear that notions such as “starting-point” or “peg on which the message is hung” are metaphorical and difficult to interpret. Indeed, many discussions of themes would ignore or dismiss such wordings. The phenomenon that “the point of departure” and “what is being talked about” or “what the speaker is going to say” are juxtaposed together by the two definitions seems to indicate that they are the same. However, they don’t always overlap. One constituent might be “the point of departure” but what it refers to is probably not “what is being talked about” and vice versa. As a result, the definitions cause many disputes, though Halliday once distinguishes theme from topic clearly. 
Huddleston(1988) argues that “it is not clear that ‘point of departure’ can sustain an interpretation that is independent of syntactic sequence”. Huddleston (1991: 97) lists 11 examples from the first edition of An Introduction to Functional Grammar to prove that Halliday’s theme is closely related to “aboutness”. Fries (1983) notices the use of the first position to indicate the pattern or organization of a text. He uses the term “method of development” (Fries, 1983:130) to refer to the notion of how a text develops the ideas which it encodes, and notes that some texts seem to develop their ideas in a single, simple consistent way. So Fries (1981/1983) follows Halliday in holding to a sequentially-ordered, or position-based definition of theme, i.e. themes invariably precede rhemes. Following Fries, Martin (1992) argues that Huddleston misunderstands Halliday’s theme, and does not consider the important role played by the theme in the development of discourse. Matthiessen and Martin (1991: 49) also believe that understanding themes should not be based on the empirical semantic “aboutness” of the terms, but rather the contribution of the themes in the development of discourse. Therefore, the theme is the embodiment of the method of discourse development. This is the essence of the theme.
Obviously Halliday is familiar with these discussions, so he revises the second clause in his definition of theme as “with which the clause is concerned” in the second edition of An Introduction to Functional Grammar (Halliday, 1994). What’s more, the definition of theme is further changed into “that which locates and orients the clause within its context” in the third edition of the book (Halliday & Matthiessen, 2014:64). These changes mark that theme is not merely “what is being talked about”. The change in wording before and after shows that Halliday also tries to avoid misunderstanding of the “aboutness” of the theme by others, while clearly emphasizing the role of the theme in discourse development. The third edition of IFG especially added two texts, Text 3-2 and Text 3-3, to illustrate the theme.
A second difficulty with Halliday’s description is that the wording “what the clause is going to be about” has regularly been linked with topic although he tries to avoid the term “topic” because this term has been “associated with only one particular type of theme”. Fries also shows that theme is not always topic. Fries and Francis (1992) argue that such expressions are metaphorical and easily fall into the context of circular argument, especially when the analysis object is a sentence that is out of context. Fries attempts to provide evidence that theme is a meaningful concept by showing that theme is relevant to the description of texts and he rewords Halliday’s description of theme in a bit less metaphorical way. “The theme of a T-unit provides a framework within which the rheme of that T-unit can be interpreted.” 
The T-unit, in short, refers to a clause complex with clauses. Therefore, the units used by Halliday and Fries when defining themes are different. Compared to Halliday’s clause unit, Fries’ T-unit is obviously larger. In my opinion, whether it is Fries’ framework or Halliday’s “locate” or ‘orient”, the definition of the theme is functional, highlighting the theme’s role in text development.
theme vs initial position
    At the early years of his statement, Halliday seems to mix the concepts of “the point of departure” with “what is being talked about” and “the element which is put in first position”. In 1970, Halliday regards theme as “the element which is put in first position” (1970:161). And in 1974, he claims that theme is the element which realizes its function by being put in the first position from the functional sentence perspective. Then in 1985, he deems it necessary to distinguish theme from initial position. 
Although Halliday tries to make it clear, some of Halliday’s arguments are still doubtful. And some of his ideas are self-contradictory. For example, according to him, all the words before topical themes are regarded as multiple themes. It seems that he still takes the initial position as the parameter to decide a theme. Huddleston (1988:158) questioned whether there are other explanations of “the point of departure” besides “the initial position of a sentence”. In fact he means that there is no doubt that “the initial position” is the standard to define a theme if the conjunctions and interrogatives appearing in the first positions of the sentences are categorized as themes. In this sense, it is not a functional definition. Thus in the following exchange, Huddleston points out that the themes in the question and the answer are different according to Halliday’s analysis. “The answer will be analyzed as being about the new boss, but the question won’t——its theme is what”. But actually they are about the same subject, that is, the new boss (1988:159) .  
(3)	What’s the new boss like?
(4)	She seems OK.
Huddleston also mentions yes/no interrogatives, for example: 
(5)	Well but then Ann, surely, wouldn’t the best idea (be to join the group)?
According to Halliday, the theme is Well but then Ann, surely, wouldn’t the best idea and wouldn’t is the interpersonal theme and the best idea is the topical theme. But wouldn’t is part of the discontinuous process element wouldn’t...be. Why can’t it be a topical theme? Then why wouldn’t in (5) and isn’t in (6) are different themes?  
(6)Isn’t (the best idea to join the group)?
In Matthiessen and Martin’s response to Huddleston’s review of Halliday’s introduction to functional grammar, they firstly ascertain that there are definitely extra explanations of “point of departure” besides what appears in the first position. They claim that themes are marked by particles instead of positions in Tagalog. The unmarked themes of the language are at the end of sentences, not the beginnings (1991:42). As for the themes in WH-questions, they say the theme in (3) is what the speak posits the listener can provide and the new participant is the rheme, while in (4), the participant is the theme and the real answer is the rheme. And with regard to yes/no interrogatives, they’re convinced that Huddleston’s interpretation is wrong. According to Halliday’s analysis, the themes in (5) and (6) are the same. Both wouldn’t and isn’t are interpersonal themes, and the best idea is the topical theme. 
Martin and Rose (2003:178) change a bit about the notion of multiple themes. They maintain that the last topical theme should be the subject in the experiential function. The topical themes functioning as adverbials can only be marked themes. Only the themes which are also subjects are unmarked themes. For example, in the following examples, the underlined parts are marked themes, while he in (7) and I in (8) are unmarked themes.   
(7)Instead of resting at night, he would wander from window to window. 
(8)In the early hours of the morning between two and half past two, I jolt awake from his rushed breathing.
In the third edition of An Introduction to Functional Grammar, to better clarify the relationship between theme and the initial position, the authors add an annotation which explains that different languages vary greatly. 
Fawcett’s theory is most different from Halliday’s. In his book The Many Types of Theme in English: Their Syntax Semantics and Discourse Functions, Fawcett claims that a clause can have many experiential themes. Besides marked themes, subjects can also be unmarked themes, which are called as subject themes. What’s more, Fawcett thinks it is not necessary to divide themes based on the three metafunctions and some themes are not real themes. From the basic rule of systemic functional grammar: meaning implies choice. He emphasizes that only the elements that are put in the first position of the sentences out of thematic choice are real themes. The elements which are put at the beginning due to other reasons, like the conjunctions and, but, if, because or the interrogatives like what, when and where, can not be regarded as themes. In other words, the elements that mark mood can not be considered as themes. 
There are two kinds of elements in Halliday’s mood system: subject and finite. Halliday and Matthiessen both make it clear that “the type of meaning that is realized by the occurrence and sequence of the subject and the finite (or operator, in Cardiff Grammar terms) in these positions is interpersonal” (Fawcett: 123). The sequence of them decides whether a sentence is a statement or a question. When an finite is put in the sentence-initial position, Fawcett disagrees with Halliday’s opinion that it is an interpersonal theme.
Fawcett thinks that the reason an finite is put in a sentence-initial position is because the mood is interrogative. Someone might say Who is the theme on the grounds that we may also say Fred hit who? But Fawcett claims that the meanings of the two sentences are different. “A clause such as Who did Fred hit? is a straightforward ‘new content seeker’, while Fred hit who? is a ‘content challenging check’ (an ‘echo-question’ in the terms of traditional grammar) (Fawcett: 123).” In other words, it is a choice within the mood network. However, Fawcett admits that there are two situations where wh-elements can be the themes e.g. Who hit you? vs. Who were you hit by? And when there are two wh-elements in the same sentence, like Who kissed who? Or Who did who kiss? Fawcett clarified that in terms of form, a common feature of the different kinds of themes is that they all occur in the beginning parts of sentences. But this doesn’t mean that all the beginning parts are themes.
Eggins (1994) and Halliday (1985, 1994) believe that the theme is divided to the end of the first topical theme and includes the first topical theme. The subject matter is either a participant in the experimental elements, a circumstance, or a process. However, some scholars such as Martin & Rose (2003) believe that the theme should  include the subject, even if there is a marked theme before the subject.
In this respect, I agree with Fawcett’s opinions. On the one hand if we insist theme a functional notion instead of a form, while on the other hand we regard all the elements appearing in the initial positions as themes. Actually we do equate “the point of departure” with “the initial position”. Thus it makes no sense. Fawcett tries to distinguish the reasons the elements occur early in the sentences and only admit those who occur as themes on the basis that theme is a functional concept. 
There are internal and external reasons for the theme’s controversial definition and identification, this is closely related to the “non-constituent nature” of the theme itself. Gómez-González (2001) points out that the theme is not a clause component, but a range of clauses, which is the “thematic zone”. Matthiessen’s (1992) phenomenon of “the weakening of the theme wave” indicates that as the clause extends, the weakening of the theme wave will be replaced by the information wave. Therefore, thematic boundaries are inherently inaccurate. Martin believes that if the wave-like implementation of discourse metafunction is considered, the segmentation of theme and rheme remains to be discussed, because the determination of the boundary between the peaks and troughs must have a certain degree of subjectivity and arbitrariness. Nevertheless, actual discourse analysis still needs to clearly define the demarcation point of the theme, because this is the starting point of discourse meaning analysis. And researchers often determine different thematic demarcation points for different research purposes and contextual factors.
Conclusion 
Theme is an important concept in the discourse metafunction of systematic functional linguistics. It can be said that the long history of thematic research has been accompanied by fierce controversy, which has led to the development and improvement of the thematic research. This article discusses the theme’s controversy and development from the definition and identification. Starting from the notion of theme proposed by the Prague school, I review how linguists deepen their understandings of themes gradually. I argue that in this process the relations have become clearer between theme on the one hand and given information, what the message is about, and the initial position of a clause on the other. 
Comparing the definition of theme by the Prague school and Halliday, it is not difficult to find that the former thinks that the theme represents known information, and its definition of the theme actually combines two different concepts: given information and the starting point of information; the latter distinguishes the two terms and encode them in different structures: information structure and theme structure. Moreover, Halliday believes that the theme generally indicates given information, but it can also indicate new information. Halliday admits that in unmarked situations theme and given information are often realized by the same element. There used to be a hot debate about the relationship between theme and what the message is about. Theme is defined as “ that which locates and orients the clause within its context” in the third edition of An Introduction to Functional Grammar (Halliday & Matthiessen, 2004:64), which indicates that Halliday has distinguished the two notions. But there is still no agreement between the relation between theme and the initial positions. Halliday insists the existence of multiple theme, which seems to equate the initial position with theme, and Fawcett is right to distinguish between elements which always come early in a clause and those placed here as a thematic choice. He starts from the core idea of functional grammar: “meaning is choice”, and emphasizes that only those elements that are placed at the beginning of a sentence because of choice can be regarded as the true theme.
 Although scholars have different viewpoints concerning theme, they have general agreement on both the functions and importance of theme. Development takes place only after controversy, and the history of thematic research illustrates this. We look forward to a brighter tomorrow for thematic research.

References

[1] Fries, P. H. 1994. On Theme, Rheme and Discourse Goals. In Coulthard, M. (ed). Advances in written text analysis. Routledge.
[2] Halliday, M. A. 1970. Language Structure and Language Function. New horizons in linguistics, 1, 140-165.
[3] Halliday, M. A. K, Hasan R. 1976. Cohesion in English. London: Longman.
[4] Halliday, M. A. K. 1985. An Introduction to Functional Grammar. London: Edward Arnold.
[5] Halliday, M. A. K 1994. An Introduction to Functional Grammar (2nd Edition). London: Edward Arnold.
[6] Halliday, M. A. K, Christian MIM Matthiessen. 2004. An Introduction to Functional Grammar (3rd Edition). London: Edward Arnold.
[7] Halliday, M. A. K. 1967. Notes on Transitivity and Theme in English: Part 2. Journal of linguistics, 3(2): 199-244.
[8] Huddleston, R. 1988. Constituency, Multi-functionality and Grammaticalization in Halliday’s Functional Grammar. Journal of Linguistics, 24(1), 137-174.
[9] Martin, J. R., & Rose, D. 2003. Working with discourse: Meaning beyond the clause. Bloomsbury Publishing.
[10] Mathesius, V. 1939. Functional Sentence Perspective. Prague: Academia.
[11] Matthiessen, C. M. I. M., & Martin, J. R. 1991. A Response to Huddleston’s Review of Halliday’s Introduction to Functional Grammar. Occasional Papers in Systemic Linguistics, 5, 5-74.
[12] Trávníček F. 1962. On So-called Functional Sentence Perspective. Slovo a slovesnost, 22: 163-71.
[13] Weil, Henri. 1978. The Order of Words in the Ancient Languages Compared with That of the Modern Languages. Vol. 14. John Benjamins Publishing.