On Halliday’s thematic theory and Information Structure Wang Huang 1. Introduction Mathesius first advanced the thematic theory in 1939. So far, researchers at home and abroad had conducted multidimensional studies on this theory. Among which M.A.K Halliday is the most influential one. This article first reviews the development history of thematic theory, then gives a brief and comprehensive introduction to Halliday's theme/rheme theory. Based on that, the author presents different critiques on Halliday’s thematic theory, and also put forward the disagreements and weaknesses in the theory as well as its application studies. At last, directions for further studies and researches are presented. 2. The development history of Theme/Rheme Theory Since the theme/rheme theory was proposed, it has become an analytical tool with increasing popularity. This popularity can be attributed to its interesting position at the interface of grammar and discourse (Mauranen 1993:104). However, the definition of theme/rheme is very different from different linguistic schools, even within a single linguistics school, a variety of approaches can be met. Therefore, it would be critical and meaningful to have an overview of the development history of theme/rheme theory. It will also be useful to begin with the Prague School, which is the first group that interested in the study of theme/rheme theory. Most researchers hold the view that Vilem Mathesius was the first scholar who proposed the concept of ‘O tak zvanem aktualnim cleneni vetnem’, which was translated into English later as “Functional Sentence Perspective FSP” by J. Firbas. This concept is quickly recognized and accepted by researchers. Mathesius’ study of theme/rheme benefited from similar ideas of French scholar Henri Weil, who first put forward a similar concept in his book The order of words in the Ancient Language Compared with that of the Modern Language. According to Weil: Words are the signs of ideas; to treat of the order of words is, then, in a measure, to treat of the order of ideas” and “a word which is affected wither by the logical or grammatical emphasis is placed before the rest; a word which has the restrained emphasis is placed after the others. (Weil 1978:11). He further stated, ‘there is […] a point of departure, and initial notion which is equally present to him who speaks and to him who hears which forms as it were the ground upon which the two intelligences meet and another part of discourse which forms the statement properly so called. This division is found in almost all we say' (Weil 1978:17). This idea is much like 'the departure of message', which is a core principle in the definition of Theme later. Mathesius got great inspiration from Weil’s opinion and put forward the famous Functional Sentence Perspective. He proposed this concept to study the different roles of different components in sentences in language communication. He believes that in addition to the division of subject and predicate forms used in the traditional grammatical analysis, syntactic analysis can also implement practical segmentation based on communicative functions to analyze how the structure expresses the information in a sentence. He found that the majority of all sentences contain two basic content elements: a statement and an element about which the statement is made (Mathesius 1975:81). According to Mathesius, ‘the basic elements of the ‘topical’ structure are the starting point of the utterance, i.e. what is in the given situation known or at least evident and from what the speaker starts, and the nucleus, that is what the speaker says about it or with respect to the starting point’(Marketa Mala and Pavlina Saldova, 2012:50). He then further divided a sentence into three parts, namely, Theme, Transition and Rheme. Some researchers do not agree with Mathesius’s definition of Theme. They pointed out that, Mathesius actually has two definitions, firstly ‘the starting point of the utterance’ and secondly ‘what is in the given situation known’. These two definition do not totally agree with each other thus lead to confusion. For example, twenty people were killed in the earthquake. Here, ‘twenty people’ is what is being spoken about, but it is not the given information. Due to the disagreement in division standard, his theory was soon criticized by other scholars. Among which Jan Firbas is a representative. Although Firbas accepted the triad ‘theme-transition-rheme’ theory, he then passes over to the concept of ‘Communicative Dynamism’ (CD henceforth). Communicative dynamism defined as a continuum into which all sentence elements fall. Firbas states that ‘the theme is constituted by the sentence element (or elements) carrying the lowest degree(s) of CD within the sentence’ (Firbas 1964:272). The Rheme pushes the communication forward the most and may be viewed as the informative part of the sentence. Firbas asserts that the basic distribution of CD would reflect what Weil called the "movement of the mind". This is an advancement in the definition of Theme, because it points out the relationship between Theme and information unit. In this sense, Theme is no longer determined only by its position, but also by its communicative importance. This problem would be further discussed in section 4. Another Czechoslovakian linguist Frantisek Danes applied the FSP into context analysis, and firstly put forward the concept of ‘Thematic Progression(TP)’. He thinks that although most studies of FSP predominantly concern the problems of Theme, it is the Rheme that represents the core of the utterance (the message proper) and pushes the communication forward. Therefore, in terms of text organization, it is the Theme that plays an crucial constructional role. The Rheme shows its significances as the conveyor of the ‘new’ - actual information, while the Theme, being informatively insignificant, will be employed as a suitable means of the construction (Danes 1974:113). He says, ‘the choice and ordering of utterance themes can show their mutual concatenation and hierarchy, as well as their relationship to the hyperthemes of the superior text units (such as the paragraph, chapter…), to the whole text, and to the situation’ (Danes 1974:114). Based on this assumption, he further analyzed five common models of Thematic Progression. Many linguists home and abroad had done great work in this aspect, for instance, Van dijk in Nederland and Shengheng Xu and Yan Huang in China. 3. Halliday’s theme/rheme theory and information structure Mathesius’s essay Functional Sentence Perspective was published in Czech in 1939 during the Second World War, it did not arouse many responses from the academic circle until the first volume of The Prague school of Linguistics and Language Teaching was published. The Prague School representative M.A.K. Halliday inherits and further improve the theme/rheme theory. It would be beneficial to first summarize Halliday’s main points of theme/rheme theory before further critiques are made. In short, Halliday's main contribution can be divided into three parts, namely, the definition of Theme, the differences between theme/rheme structure and information unit, the differences between theme/rheme theory and topic/comment theory, and the markedness of Theme and multiple Theme. 3.1 The definition of Theme According to Halliday, Theme and Rheme together constitute a message in the structural configuration of a clause. Theme is ‘the element which serves as the point of departure of the message; it is that which locates and orients the clause within its context’. The Rheme is ‘the remainder of the message, the part in which the theme is developed' (Halliday 2004:64). 3.2 The differences between theme/rheme structure and information unit 3.2.1 Different realization Theme embodies itself in the linear structure of a clause; what comes in the first position is the Theme of the clause, the rest part of the clause is Rheme. However, the information unit is realized by phonologically by ‘tonality’, the distribution of the text into tone groups (Halliday 1967:200-201). 3.2.2 Different appearing order As for Theme, constrained by its definition, it can only appear before the Rheme. The situation that Rheme goes before Theme does not exist. However, in the information unit, generally speaking, given information will appear before the new information, but this is not always the case. Sometimes, new information would come before the given information. Therefore, there is no corresponding relationship between Theme and given or Rheme and new. 3.2.3 Different relationship with context Theme/rheme structure is the inner structure of a clause that organize the message, it does not have a close relationship with the context. However, the context is closely related to information unit. Without the context, we cannot divide what is new information and what is given information. 3.2.4 Different perspective 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 of have accessible to you. Theme + Rheme is speaker-oriented, whereas Given+New is listener-oriented (Halliday 2004:93). 3.3 The difference between theme/rheme theory and topic/comment theory When analyzing sentences from the functional perspective, many linguists treat theme/rheme and topic/comment as the same. Halliday clearly opposes such confusion. He states: Some grammarians have used the terms Topic and Comment instead of Theme and Rheme. But the Topic-Comment terminology carries rather different connotation. The label 'Topic' usually refers to only one particular kind of Theme, the 'topical' Theme; and it tends to be used as a cover term for two concepts that are functionally distinct, one being that of Theme and the other being that of given. It seems preferable to retain the earlier terminology of theme/rheme (Halliday 2004:65). 3.4 The markedness of Theme and multiple Theme 3.4.1 The markedness of Theme According to Halliday, Theme can be divided into marked Theme and unmarked Theme. In typical cases, Theme is the prominent element, these are called unmarked Theme. But in some cases, there is a standing-out or marked Theme, in which the usual relationship is reversed and the nominalization becomes the Rheme (Halliday 2004:70). This is a great improvement in definition, the analysis of Theme and Rheme are then no longer constrained to common sentences, but also to other sentences that do not frequently use in daily lives, which means the application of this theory is greatly expanded. 3.4.2 Multiple theme According to Halliday (2004), the Theme can be further divided into experiential, textual, and interpersonal themes. The Theme contains one, and only one, of experiential elements. Halliday refers to this constituent, in its textual function, as the topical Theme. There may, however, be other elements in the clause preceding the topical Theme. These are elements that are either textural or interpersonal in function. Here is an example to illustrate a multiple theme containing all the types of non-topical element in thematic position: Table 1 Well but then surely Jean wouldn’t the best idea be to join in cont stru conj modal voc Finite topical Theme Rheme (Halliday 2004:81) Therefore, we can now approximate more closely to the identification of the Theme: The Theme of a clause extends from the beginning up to, and including one, and only one, of the experiential elements – that is either participant, circumstance or process. Everything after that constitutes the Rheme (Halliday 2004:85). In conclusion, Halliday is a gifted linguist who had done great work in theme/rheme theory. We can also see from the first two parts that Halliday does inherit some central ideas from the Prague School, but he also inputs many of his own understandings about the words order and information unit. Till now, a comprehensive and sound framework had been established, but that does not mean this theory is perfect without any flaws. In the next part of the essay, some deficiencies about this theory would be discussed. 4. Critiques on Halliday’s theme/rheme theory and information unit In this part, critiques on Halliday’s theme/rheme theory based on many other researcher's works will be presented with the author’s comments and thinking. As a linguistic theory, one of the essential guidelines is that it should and can be applied to any other language so that we can understand the nature of language from a broader perspective. Therefore, some Mandarin examples are used as supporting material. 4.1 Some minor changes in the definition of Theme Halliday's definition of Theme has gone through the development of nearly 50 years, during which some minor changes happened. In 1967, when Halliday first introduces theme/rheme theory into Functional Grammar, the definition is ‘the Theme is assigned initial position in the clause, and all that follows is the Rheme. The Theme is what is being talked about, the point of departure for the clause as a message’ (Halliday 1967:212). In the second edition of An Introduction to Functional Grammar, the definition of Theme is as following: 'the theme is the element which serves as the point of departure of the message; it is that with which the clause is concerned….within that configuration, the Theme is the starting-point for the message; it is the ground from which the clause is taking off’ (Halliday 1994:38).Till now, we can therefore summarize as the following: Theme is 'the point of departure of the message’, ‘the ground’ and ‘things that are being talked about’. But if we look at Theme from the perspective of the message that is being conveyed, Theme, therefore, is not always ‘the departure of the message’, ‘the ground’ as well as ‘things that are being talked about’at the same time. For example: 1. ---(Who gave you that magazine?) ---Bill gave it to me. In this sentence, Bill is the ‘departure of the message’, but it is not the ‘ground’ from which the clause is taking off. On the contrary, gave it to me is the ground. While ‘things that are being talk about’ seems to be it rather than Bill. Halliday might explain Bill as a ‘marked theme’, but the explanation is far too simple. Xiao JunHong also critique that if Theme (including from information unit perspective) can be divided into marked and unmarked, then, why there is no ‘marked rheme’ or ‘unmarked rheme’? (Xiao 2002:23) This might be a point where the definition of Theme received doubts and critiques. Therefore, in the 3rd edition of An Introduction to Functional Grammar, Halliday made a minor change one more time. This time, the definition of Theme is ‘the element which serves as the point of departure of the message; it is that which locates and orients the clause within its context’ (Halliday 2004:64). The readers might notice that, in the newest definition, Halliday delete ‘things that are being talked about’ and ‘the ground from which the clause is taking off’. At the initial stage, ‘what is being talked about’ and ‘the point of departure’ is paralleled, which means these two functions are equally important. However, we can notice that later Halliday put more focus on ‘the point of departure’ and do not mention about ‘things that are being talked about’. However, Halliday does not explain in depth what does he mean by ‘the departure of the message’. The definition of ‘the departure of the message’ is very vague, which will cause more confusion when we consider the multiple Theme. This will be further elaborated in section 4.2. What is more, when putting this theory into practice, the most important guideline to divide Theme and Rheme seems to be ‘the initial position’ of a clause. As Halliday puts it, ‘as a general guide, the theme can be identified as that element which comes in first position in the clause’ (Halliday 1994:37). Therefore, we can almost conclude that sentence-initialness is the most crucial principle when dividing Theme and Rheme. Although Halliday explained that ‘this is not how the category of theme is defined, the definition is functional’ (Halliday 1994:37). We can tell from his book that the division between Theme and Rheme is based on linearity, but linearity is not the way we understand language. Some sentences can be understood in its linear arrangement, but this is only a coincident; we cannot, therefore, conclude that all the sentence can be understood in this way. There are also questions to do with the status of the subject. Under the theme/rheme system, the subject can be regarded as unmarked topical Theme in a declarative clause. If the subject does not come in the first position, then anything in the initial position of the clause is marked topical Theme. In this way, the subject element of structure is more independently existed in the theme component of the grammar than in that of mood. As Eirian C Davies (2014) puts it, ‘how does a part of an element of one kind (the subject as part of the mood element) combine with the whole of an element of a different kind, such as theme or actor?’. 4.2 On multiple Theme 4.2.1 On the Definition of Multiple Theme Halliday divides Theme into single Theme and multiple Theme. Some sentence patterns hold lexical meanings as well as grammatical meaning, some sentence (for example yes/no interrogative), the Theme only contains grammatical meaning, that is why Halliday extends its Theme to the following elements and proposes the concept of multiple Theme. This is an advancement, but ‘it still constrains itself in the “form” rather than focus on the message’ (Xiao 2001:38). Multiple Theme including topical, textual and interpersonal Theme, however, the information conveyed by the three themes are entirely different in its nature, then how could those three themes combined be ‘the departure of the message’ and ‘locates and orients the clause within its context’ at the same time? Halliday further explains his definition of multiple themes as follows: Continuatives and conjunctions are inherently thematic…conjunctive adjunct, vocative, modal comment adjunct and finite verbal operators are called characteristically thematic…since continuative and conjunctions are thematic by default, when one of them is present it does not take up the full thematic potential of the clause in which it occurs. What follows it will also have thematic status (Halliday 2004:83). If we do deep thinking along with this logic, then how do we explain the WH-word in WH-interrogative, the verb in a declarative sentence, and the interjection in an exclamatory sentence? Those elements are also element 'come at the beginning'; therefore, they are 'thematic by default', then why they can ‘take up the full thematic potential of the clause’ and what follows them do not have ‘thematic status’? Such confusion might be caused by the vague definition of topical Theme, which will be discussed in section 4.2.2. 4.2.2 On the definition of topical Theme Among the three types of multiple Theme, Halliday thinks that 'the theme contains one, and only one, of these experiential elements and we refer to this constituent as the topical theme' (Halliday 2004:79). How does 'theme' and 'topic' combined into 'topical theme'? Some linguists think that a clause is consists of topic and comment, ‘the speaker announces a topic and them says something about it’ (Gilian Brown et al 1983:70). Halliday does not agree with such a division. He thinks what those linguists refer to as topic is only one particular kind of Theme; that is the 'topical theme'. As he puts it, '(topic) tends to be used as a cover term for two concepts that are functionally distinct, one being that of theme and the other being that of given' (Halliday 2004:65). Therefore, for him, the Theme extends from the beginning up to, and including the first element that has an experiential function. But this definition ignores the different functions carried by different constituents as the Theme. For example, does participant as theme function the same as the circumstance as Theme? Halliday regards Theme as 'the point of departure of the message […] with which the clause is concerned'. This definition seems more suitable for topical Theme. If textural Theme, interpersonal Theme, and topical Theme appear in one clause, then 'the departure of the message' is realized by which Theme? Nie (2002) points out that such confusion is caused by Halliday's failure to distinguish the difference between participant as Theme and circumstance as Theme. Xu(1996) even thinks that whether circumstance at the beginning of a clause can be regarded as topical Theme still need further study. Geoff Thompson (1996:119) holds the same view. For example: 2. For centuries, yellow canaries have been used to ‘test’ the air in mining. In this clause, although 'yellow canaries' does not appear in the initial position of the clause, it is still what the clause concerns. All those discussions show that the 'aboutness' of the Theme only works when 'participant' takes the theme position, however the circumstance only works as background information. 4.3 On theme/rheme and information structure 4.3.1 Theme/rheme and given/new Information Generally, linguists regard Halliday as the first researcher to apply the concept of ‘given information’ and ‘new information’ into language information unit. Halliday (1967) put forward the Given/New information in language analysis in his essay Notes on transitivity and Theme in English. Many linguists follow his step to further study the information structure in sentences. Among several definitions proposes by different linguists, Wallace Chafe’s seems to be the one receives most attention. ‘Given (or old) information is that knowledge which the speaker assumes to be in the consciousness of the addressee at the time of the utterance. So-called new information is what the speaker assumes he is introducing into the addressee’s consciousness by what he says’ (Chafe 1976:28). KeDing Zhang and Jing Zhang think that Chafe’s definition is inspiring and enlightening, because ‘it is important to differentiate the idea of Given/New information in the communication area from that in the cognitive area’. They add one more characteristic to Given information based on Chafe’s definition, that is, ‘given information is that knowledge which the speaker thinks or assumes to be in the consciousness of the addresses at the time of utterance, including all the information that could be related to that knowledge’(Zhang 2004:6). And they think it is important to describe the relationship between theme/rheme and given/new information. Therefore, they use given/new information to define theme/rheme. Although this definition might need further study, it is an improvement in terms of the relationship between theme/rheme system and the information structure system (2004: 5-7). Zhang makes the two systems closely related to each other rather than regards them as two independent systems. In contrast, Halliday focuses more on their differences, as for their similarities and relationship, he did not explain fully and explicitly. A possible explanation for this might be his Halliday's theory mainly uses English sentences as examples, but for some other languages, Mandarin, for instance, the information structure can influence the sentence structure, thus influence the Theme and Rheme in a clause. This will be further discussed in section 4.2. 4.3.2 On the relationship between theme/rheme and information structure In this part, we want to study the relationship between the sentence structure and the information structure. According to Halliday, information unit is the grammatical reflex of the phonological tone group (Halliday 2004:31), and information structure constitutes of given and new information which we have already discussed. Halliday thinks that there is a close semantic relationship between information structure and thematic structure, but he also emphasizes that given/new and theme/rheme are not the same thing (Halliday 2004:93). In his opinion, Halliday thinks that the information structure is a component of the grammatical structure of the sentence, and many other linguists hold the same view, including Lambrecht (1994) and Shi ZhouCheng (2016). However, there are also different opinions, among which Lu JianMing (2017) is a representative. He thinks that information structure is not a part of sentence structure; instead, these are two parallel structures with influence on each other. If we take Mandarin as the study object rather than English, we will find that information structure does have an influence on sentence structure and therefore influence the theme/rheme system. 3. 那窗户玻璃 打破了，打破 那窗户玻璃 的人 是小王。 * That windowpane broken broken the windowpane who Xiaowang That windowpane was broken, the person who broken it is Xiaowang. If we analyze this sentence according to the information flow, we will find that 'the windowpane' is the topic or the Theme, which locates at the beginning of the sentence.; 'broken' is the most important new information, or the information focus; 'Xiaowang’ is also new information, but the ‘unknown information quantity’ is less than that of ‘broken’. Therefore it cannot locate after the ‘broken’. To convey this message in Mandarin, the linear arrangement of the sentence would follow the information structure perfectly —the 'windowpane' as topic locates in the initial position of the sentence, the information focus 'broken' appears in the last, and 'Xiaowang' locates in the center. The right and typical Mandarin sentence would be: 4. 那 窗户 玻璃 （被／给） 小王 打破了。 *The window glass was Xaiowang broken The windowpane was broken by Xiaowang. Therefore, we can see that in Mandarin, the sentence structure is exactly arranged according to the information structure. Linguists must attach great importance to such a phenomenon because this indicates that sentence structure and information structure are not two independent systems; there are relationships and bonds between them. However, such bonds and relationships do not receive enough attention while we study the deep structure of the sentence. And this might be an area where we can further explore and solve some problems to do with theme/rheme and their application. 5. Conclusion Since the theme/rheme theory was proposed, due to its unique position at the interface of syntax and grammar, countless linguists have tried to analyze it from various angles. Halliday is undoubtedly one of the outstanding figures. However, facts have also proved that Halliday's thematic theory still has many problems in practical application. Therefore, it is necessary to sort out and analyze the thematic theory from its definition. Based on the development history of the thematic theory, this essay collates and compares the definitions of theme/rheme put forward by various linguistic school. Furthermore, taking Halliday's thematic theory as an example, the article summarizes his contribution to the theory and then analyzes the shortcomings based on other linguists' ideas. Regarding the thematic theory, there are still many places worthy of in-depth study. For example, the research perspective is still too single, and the research methodologies are not rich enough; the depth and breadth of related research are not enough, and theoretical research is much more than applied research judging from the current research results. There are many other research topics related to thematic theory, and hoping this article will attract more linguists to discuss the theme/rheme theory further. Reference Anna Mauranen, Cultural Differences in Academic Rhetoric (Pieterlen: Peter Lang,1993). ChengZhou Shi, Exploration of Information Structure Theory (Jinan: Shandong University Press, 2016) (In Chinese). Eirian C Davies, A retrospective view of Systemic Functional Linguistics, with notes from a parallel perspective, Functional Linguistics, (1:4)2014,1-11. F. Danes, ‘Functional Sentence Perspective and the Organization of the Text’ in Papers on Functional Sentence Perspective, ed. by F. Danes, (Prague: Academia, 1974), pp.106-28. Geoff Thompson, Introducing Functional Grammar, (London: Arnold, 1996). Gillian Brown and George Yule, Discourse Analysis (Cambridge: Cambridge University Press,1983). Henri Weil, The Order of Words in the Ancient Language Compared with that of the Modern Languages, trans. by Charles W. Super (Amsterdam: John Benjamins B.V.,1978). JianMing Lu, ‘Valuing the study of Language information structure and Extending the New Horizons of language Studies’, Contemporary Rhetoric,4(2017),1-17 J. Firbas, ‘On Defining the Theme in Functional Sentence Perspective’, Travaux Linguistiques de Prague, (1)1964, 267-80. Jin Zhang and Keding Zhang, A Comparative Study of Information Structure in English and Chinese, (Kaifeng: Henan University Press, 2004) (In Chinese). JunHong Xiao, A Comparison of Theme and Rheme Theories in Halliday (1994) and Firbas(1992) , Journal of PLA University of Foreign Languages, (24)2001,37-40, (In Chinese). Knud Lambrecht, Information Structure and Sentence Form: Topic, Focus and the Mental Representations of Discourse Referents (Cambridge: Cambridge University Press,1994). Long Nie, ‘On Halliday’s topical theme’, Foreign Language Research, ser.73, 3(2002), 22-26.(In Chinese). M.A.K Halliday, An introduction to Functional Grammar, 3rd edn (London: Arnold,2004). --‘Notes on transitivity and theme in English: Part 2’, Journal of Linguistics, (3)1967,199-244. -- An introduction to Functional Grammar, 2nd edn (London: Arnold,1994). A Centenary of English studies at Charles university: from Mathesius to Present-day linguistics, ed. by Marketa Mala, Pavlina Saldova, (Prague: Charles University Press,2012). Vilem Mathesius, A Functional Analysis of Present Day English on a General Linguistic Basis (Hague: Mouton, 1975). Wallace Chafe, ‘Giveness, contrastivenss, definiteness, subjects, topics and point of view’, in Subject and Topic, ed. by C. Li (New York: Academic Press), pp.25-55. YuLong Xu, ‘The Identification of Topical Theme in Chinese and English Discourse’, Foreign Language, (6)1996,3-9,(In Chinese).