Thematic Progression as a Functional Resource in Analysing Texts
Universidad Complutense de Madrid
Frantisek Danes' (1974) proposal of three main types of thematic progression constitutes a functional explanation of the ordering of information in discourse. As such, it corresponds to Nichols' (1984) fifth distinction of 'function' as 'function in context', and specifically in textual organization.
Danes' consideration of the Theme-Rheme nexus operating over sequences of text is seen to compare favourably with treatments of Theme in isolation. Application of the three TP types to texts of a more everyday nature than those previously explored for this type of organization (Dubois 1987, Nwogu and Bloor 1991) reveals that all three manifestations of TP are to be found, depending on the text type and rhetorical purpose of the writer. The study concludes with suggestions for further exploration of Danes' model within educational settings.
1. Introduction. Functional explanations and uses of
This study attempts to test the applicability of a particular functional resource - Danes's thematic progression types - to the analysis of of certain linguistic choices in texts. A functional analysis or explanation involves explaining why a given phenomenon occurs, by showing what its contribution is to the text in question or to a larger system of which it is itself a subsystem. A functionalist view, then, sees language as 'goal-oriented', although I would agree with Leech that the term 'function' is preferable to 'goal' since 'goal' evokes deliberate, conscious and personal objectives; by contrast, 'function' leaves open how far the attainment of goals is due to conscious states of the individual, or for that matter, whether the goal is an attribute of the individual, the community or the species (Leech 1983:3.2.)
One possible problem arises from the fact that in the functionalist literature, 'function' is used in a perhaps bewildering variety of ways. According to Nichols (1984), there are at least five main senses, all of them signifying the dependency of some given structural element on elements of another order, or another domain.
First, there is function in the sense of interdependence; one thing is the function of another in a way related to the mathematical sense of function.
Second, there is function in the sense of purpose. This is what speakers think they are doing or intend to carry out through language: questioning, declaring, giving an order, persuading and other such illocutionary forces. This is the use of language to enact a speech-act type. Taxonomies of the functions of language such as those of Bühler, Malinowski, Jakobson, Halliday, Popper are taxonomies of function as the purposive use of languae.
Third, there is function in relation to context. This is the indexical meaning of language: language as it reflects the context of the speech act. Nichols sees it as covering two subtypes: function in in the speech event and function in text.
As regards the speech event, functional categories index speech roles in terms of status, power, social distance and so forth of the participants in the speech event. These may be reflected in terms of politeness and deference categories.
In text, function has to do with the textual categories of discourse organisation and includes functional accounts of such questions as narrative structure, cohesion and other markers of textuality. Thematic progression, as discussed below, forms part of this functional dimension.
The fourth sense of function is that of relation, 'the relation of a structural element to, or within a higher-order structural unit, and its contribution to that unit' (Nichols 1984:100). In this sense a functional analysis is opposed to a categorial analysis. Categories or classes of unit such as noun phrase or clause can, for instance, realise the function of Subject. Conversely, a single structural element may have the category NP and conflate various functions such as Subject, Agent, Theme, Given information.
Finally, as Nichols points out, function is often used in a loose sense, roughly equivalent to meaning, especially where meaning is taken to include the pragmatics of purpose and context. Here can be included the 'metafunctions' proposed by Halliday, who suggests that languages are organised around two main kinds of meaning, the 'ideational' and the 'interpersonal'. These are visualised as functional components, together with a third, the 'textual' or 'enabling' function which 'breathes relevance into the other two' (Halliday 1994:xiii).
A consideration of Popper's (1979:116) hierarchy of functions can illuminate the interconnection between function as purpose and function in textual organisation. This hierarchy, which however does not specifically include the conative or interpersonal functions nor the poetic, magical or textual functions of other proposals, is used by Popper to justify the existence of the world of 'objective knowledge', or 'knowledge without a knowing subject'. As set out by Leech (1983:49), who offers a detailed exposition of Popper's theory, stratifying functions which Popper had not placed in any strict order, there is a progression from lower to higher functions in the evolution of human knowledge. Consequently, reading from the bottom upwards we have:
D. Argumentative function (using language to present and evaluate arguments and explanations)
C. Descriptive function (using language to describe things in the external world)
B. Signalling function (using language to communicate information about internal states to other individuals)
A. Expressive function (using language expressing internal states of the individual
Fig.1 Popper's hierarchy of language functions (as in Leech 1983:49)
The two most important higher functions of human languages, according to Popper, are the descriptive and the argumentative functions; the two lower functions, the signalling and the expressive functions, are common to both human and animal languages, as well as to all linguistic phenomena. 'The argumentative function presupposes the descriptive function: arguments are, fundamentally about descriptions: they criticize descriptions from the point of view of the regulative ideas of truth, content and verisimilitude' (1979:120). The functions therefore form a hierarchy, in that a higher function must coexist with all functions lower than itself, whereas a lower function does not necessarily imply the existence of a higher function. Both phylogenetically and ontogenetically, an essential part of this thesis is that language itself has been the channel 'whereby the biological level of evolution became the basis for a more rapid and powerful kind of evolution, the evolution of knowledge (Leech, 1983:49). Popper's thesis is that 'it is to this development of the higher functions of language that we owe our humanity, our reason. For our powers of reasoning are nothing but powers of critical argument'. Popper goes so far as to emphasise the 'futility of all theories of human language that focus on expression and communication' since the human organism 'depends in its structure very largely upon the emergence of the two higher functions of language' (1979:121).
Popper attributes the accelerated evolution of knowledge to 'the tremendous biological advance of a descriptive and argumentative language'. So, according to this view, out of a sophisticated descriptive function capable of dealing not only with physical objects (e.g. houses, means to get food), mental objects and societal objects (e.g. ownership, rights, obligations, truth and falsity), which involve a high degree of abstraction, there evolves what Popper calls 'objective knowledge' or knowledge which exists independently of individuals. This kind of knowledge is typified by mathematical knowledge, knowledge stored in libraries, scientific knowledge in general. With some system of writing, the recording and storing of such knowledge is hampered.
2. The organizing principles of information in discourse
Evolution to objective knowledge therefore presupposes a written medium. And once a sophisticated written medium has evolved, a major concern is to uncover the organising principles in language that account for the ordering of information in discourse beyond the level of the sentence.
One such principle is the concept of thematic progression (Danes 1974), part of the theory of language known as Functional Sentence Perspective (FSP) propounded by scholars of the Prague School. This theory 'deals with how the semantic and syntactic structures of the sentence function in fulfilling the communicative purpose intended for the sentence' (Firbas 1986). Apart from 'functional sentence perspective', various terms used in discussions in this area include Given-New information and Theme-Rheme, both of which go back to Mathesius, writing in the nineteen thirties.
Mathesius defines Theme under two concepts: one, as 'the starting-point of the utterance, that which is known or at least obvious in the given situation, and from which the speaker proceeds', and two, as 'the foundation of the utterance, as 'something that is being spoken about in the sentence.' Rheme will be 'what the speaker says about, or in regard to the starting point of the utterance.' Theme then is 'something that can be gathered from the previous context, while Rheme expresses something new, something unknown from the previous context (translated by Firbas in Firbas 1976.11).
Following one or other of these strands, linguists have adopted what Fries terms 'the combining approach', equating Theme with Given, or the 'separating approach' disentangling the two (1983). The former include van Dijk, Firbas, Gundel and Kuno, whereas the latter is propounded by Halliday and the systemic school in general, who separate information structure (Given-New) from thematic structure (Theme-Rheme). The separation of the two appears to be useful, necessary even, for English at least. This is because, although Theme typically coincides with Given while New tends to fall within the Rheme as in (1a), the opposite is also common, especially in spoken English: New information, signalled by prosodic prominence, may conflate with Theme as in (1b):
(1a) A. (How did you get here so early?)
B. I came with TOM today.
(1b) A. (How did you get here so early?)
B. TOM brought me.
Fig. 1 Correspondences between Theme-Rheme and Given-New in English
Fries himself, a 'separater', claims that 'the information that is contained within the themes of the various sentences of a passage correlates with the method of development of the passage' (1983:116). While recognising the usefulness of the concept of 'method of development' of the text, I have suggested elsewhere that the separaters have not gone far enough, and that it is also necessary to disentangle Theme, best interpreted as a positional category such as Dik's P1 (1989) and literally the point of departure of the message, from Topic - what the message is about. These two textual categories, together with that of Given, and the syntactic category of Subject, may and often do conflate, but need not, for English at least (Downing 1991? and Downing 1992?).
Furthermore, it is not only Themes and Topics which deserve attention in a consideration of the clause as message. It is also necessary to investigate how Theme-Rheme combinations develop and pattern over text. A concentration on Theme among systemicsts has frequently tended to involve a relative neglect of the Rheme, except in its static conception as harbouring unmarked Focus, the 'newsworthy' element of the message. Thus Francis (1989:201), concentrating on the realizations of Theme, considers that 'essentially, method of development can be taken to mean something like typical thematic selection - the point of departure most often selected within a text or part of a text' (author's emphasis). Fries (1994:234) adopts a similar criterion in his treatment of what he calls N-Rhemes (the last clause constituent, typically assocated with newsworthiness). Analyzing a fund-raising written text, he notes the number of occasions in which N-Rhemes are filled by evaluative items which serve to contribute to the writer's communicative goals. Building on Firbas's insights (that Themes will be the 'main constructing elements of the text', while Rhemes will 'push the message forward'), these studies are illustrative of how the functions of Themes and Rhemes have tended to be treated separately. In both, the frequency of occurrence of either thematic of rhematic realizations of certain types is what is considered significant. Neither deals with the Theme-Rheme nexus across sentence boundaries. This concentration on either Theme or Rheme or Rheme, rather than the relationship between the two over stretches of text perhaps derives from Halliday's early statement that 'thematization is independent of what has gone before' (1967:17), and which is challenged by Danes (1974:109).
3. Exploring the textual function: F. Danes's 'thematic progression'
A more dynamic view is that of Danes's concept of thematic progression. He claims that the organization of information in texts is determined by the progression in the ordering of utterance themes and their Rhemes. His spelling out of the relationship between successive themes and their Rhemes would appear to provide a more satisfactory account of the 'method of development' of texts.
Danes makes two very telling observations about the property of being New:
First, New may mean two things:
(i) not mentioned before in the preceding context.
(ii) related as Rheme to a Theme to which it has not yet been related.
In the former case, the property 'new' is assigned to the expression itself. In the latter, it is the T-R nexus that appears as new.
b) Second, 'it is not Rheme alone, but its connection with the given T that is communicatively relevant. Firbas points out that it is (typically) the Rheme that represents the core of the utterance (the message proper) and 'pushes the communication forward'; however, from the point of view of text organisation it is the Theme that plays an important constructing role. It is this latter notion that Fries has renamed 'method of development of the text'.
The thematic organisation of the text is closely connected with discourse coherence or text connexity. A text is defined as text largely in terms of its semantic coherence; however, it is useful to remember, as Danes points out (quoting the Prague School scholars Hausenblas and Trost), that texts are not always perfect; they not only display coherence to an uneven degree, but some may be characterised as 'discontinuous'. Text connexity, in this light, is not a yes/no property but rather a more-or-less property. This resistance of coherence to formulation as rules is perhaps one of the causes of difficulty in the teaching of prose writing, academic or otherwise.
The reason for less than optimal coherence may be that the speaker/writer is simply not controlling the mass of new information that is successively accumulated as the text unfolds. This mass of information is mostly so extensive that the speaker/writer necessarily makes a choice, and this choice, Danes claims, is determined , directly or indirectly by the selection of utterance theme.
4. Danes' concept of thematic progression: an application
Danes's important contribution is to extend the concept of theme as point of departure of a single utterance (sentence) to that of explaining the inner connexity of texts.
His basic assumption is that text connexity is represented, among other things, by thematic progression. By this he means 'the choice and ordering of utterance themes, their mutual concatenation and hierarchy, as well as their relation to the hyperthemes of the superior text unit (such as paragraph, chapter, etc.), to the whole text, and to the situation. Thematic progression might be viewed as the skeleton of the plot.' (1974:114).
Danes postulates three main type of Thematic progression:
(1) Simple linear progression: Each R becomes the T of the next utterance.
T1 ------ R1
T2 (=R1)------ R2
T3 (=R2) ------ R3
(2) TP with a constant (continuous) theme:Themes are derived from a 'hypertheme'
T1 -------- R1
T2 -------- R2
T3 -------- R3
(3) TP with derived themes
These three types may be employed in various combinations, combinations of Types 1 and 2 being particularly frequent. Certain combinations may constitute TP-types of a higher order, 'representing a formal frame for the employment of the basic types.' The most important of such frames is that of 'a split Rheme'. (Split Rheme is often quoted as 'Type 4', but it is in fact a combination of Types 1 and 2. )
Danes goes on to insist that these three types are to be considered as abstract principles, models or constructs. The implementation or manifestation of these models in particular languages depends on the properties of the given language, especially on the different means available for expressing FSP' (1974:121). In applying this notion to a pro-drop language such as Spanish, for instance, it is clear that the manifestation of theme would have to be interpreted differently from the corresponding manifestation in English or French. Danes also points out that languages have at their disposal special means for the purposes of TP. as for instance, both....and; on the one hand.... on the other, the distribution of which builds up a 'network of orientation' of the text. He remarks on signalling devices in technical texts, scientific or technical prose. He also illustrates possible 'complicated' utterances obtained by coordination, apposition, nominalizations and relative clauses.
Further applications of thematic progression types to scientific texts have been carried out (Dubois 1987, Nwogu and Bloor 1991), both studies based on naturally occurring data. While both found Types 1 and 2 to be frequent, their findings differ as regards the manifestation of Type 3 (derived themes). Whereas Nwogu and Bloor found derived themes represented in their research article data but not in their more popular medical texts, Dubois, working with biomedical research texts, found one single instance of this type. She consequently reduces Danes' three types to two, one themic the other rhemic, each of which may be simple or multuple, contiguous or gapped. In this reformulation, Type 3 disppears as a separate type, subsumed under the themic type.
The present study is based on data extracted from more popular text types, all encountered within the 'leisure' sections of newspapers and magazines: first, a feature article by a wildlife expert, and second, extracts from the 'travel' sections. In one variant of the second of these text types it turns out that Type 3 is well represented.
My first text is a paragraph from an article by Ralph Whitlock, nature correspondent of the Guardian Weekly. As a genre, this type of journalistic text does not come within the 'field' of 'national daily news', 'field', in recent genre/register studies referring to the nature of the social action that is taking place, the ideational dimension of register. Its purpose is entertainment and as a text type it frequently combines narrative with reflective modes, with a 'personal' writer addressing the reader (McCarthy & Carter 1994). But like other journalists, Whitlock is no doubt writing to a strict time schedule, although not so severe as a reporter working for instance on a war front.
Birds and animals recognise us not merely as human
beings but as individuals (1). When my wife and I arrive
home from our morning walk (2) we are observed by our
resident pair of collared doves, perched on a convenient
tree, cable or roof-top (3). They recognise not only us but
our car (4). Strangers and unfamiliar cars are viewed with
suspicion (5), but of ourselves they are interested
spectators (6). They have decided we are harmless (7), and
so they enjoy watching us (8).
Ralph Whitlock in The Guardian Weekly
Summary and discussion of TP types
• Clause 1 establishes a Theme (Birds and animals) with the rest as Rheme.
• Clauses 1 to 2: Simple linear (Type 1)
• Clauses 2 to 3: Constant (Type 2)
• Clauses 3 to 4: Simple linear (Type 1)
• Clauses 4 to 5: Simple linear by contrast (pivot,
marked Theme, topic switch) (Type 1)
• Clauses 5 to 6: constant by means of contrast (Type 2)
• Clauses 6 to 7: simple linear (Type 1)
• Clauses 7 to 8: Constant (Type 2)
• From the Rheme of clause 1, an element 'us...as individuals' provides the Theme of Clause 2, a subordinate clause: 'When my wife and I...' The thematic progression is simple linear (Type 1).
• Clause 3 is a main clause 'we are observed by our resident pair of collared doves', in which a the Theme we has the same referent as that of the subordinate clause in 2 (my wife and I). TP from 2-3 is therefore of Type 2 (continuous/constant).
The thematic progression types of the first three clauses in the text are diagrammed below:
T1 - R1
Birds and animals ...us......
T2 - R2
my wife and I...
T2 - R3
we .... collared doves
They - us
Fig. 2 Thematic progression across clauses 1 to 3
• The Rheme of Clause 3 is long , extending from 'are observed' upto and including 'roof-top', and containing what Danes calls a 'condensed unit' represented by the heavy nominal followed by a clause, all describing the collared doves. The latter provide the Theme for Clause 4, 'they', the progression type from clause 3 to clause 4 being simple linear (Type 1).
• The Rheme of cl. 4 contains us and our car.
• This leads to the Theme of clause 5 'Strangers and unfamiliar cars, which constitutes a new Theme, but by contrast rather than by synonymy (to us in the previous clause); the contrast is marked in reading by prosodic prominence on 'unfamiliar. TP from 4 to 5 is therefore simple linear.
• Similarly, TP from 5 to 6 is also simple linear, based on contrast, with ourselves (i.e. not strangers) as Theme of 6.
• The Rheme of 6 contains 'they are interested....', which provides the Theme of 7 'they have decided...'.again simple linear.
• Finally, the Theme of 8 'they enjoy...' maintains that of 7 (Type 2, constant theme).
According to this interpretation, the author has chosen as skeleton of the text a slight preponderance of simple linear progressions (4 occurrences) over constant theme (3 occurrences). No instances of derived themes are identified.
This approximate alternation between TP 1 and TP2 enables the author to maintain an equally approximate balance between his two sets of participants, birds and animals on the one hand, and ouselves the humans, represented by his wife and himself on the other.
On a macrostructural level, the text seems to develop towards a central point, which I have called a pivot, situated in clauses 5 and 6. This pivot introduces an alien element: strangers, that is, not us, and unfamiliar cars , that is, not ours. After this central point, the text finishes as it started, comparing birds and animals, represented by the doves, and ourselves, represented by the writer and his wife.
This alternation is achieved by syntactic choices, and it is worth remembering that they are choices, and could have been different. For instance, clause 2 starts with my wife and I as Theme of a temporal, scene-setting clause; clause 3 maintains the constant theme, but only because the passive has been chosen 'we are observed'. An active verb would have brought the collared doves to Subject and Theme, with a very different effect.
Similarly, at the two points of this short text at which a strong contrast is made, namely, clause l5 with strangers as point of departure, and clause 6 with of ourselves, the contrast is achieved in one case by the passive, and in the other by a very marked thematization (of ourselves). Unmarked choices of Theme would have maintained the collared doves as permanent point of departure throughout, and 'ourselves' as constantly part of the Rheme. As it stands, the thematic progression of successive utterances maintains a nice balance between the two sets of participants.
Yet we still get the impression that the author's ideology as a wildlife expert leads him to give the leading role to the animals, represented by the pair of collared doves. Let us attempt to see how this is done. A clue may be provided by the number of clauses in which humans are syntactic Subject: only 3 out of the total of 8. A further clue is provided by Givón's (1990, 1993) claim that Agents tend to have greater topicality than Patients in active clauses, where Agent will conflate with Subject. Passivization is used, bringing patient to Subject, when topicality is reversed.
I would suggest that this is where a semantic dimension comes in. The ideational function of language enables us to express patterns of experience, conceptualised as situation types, processes or states of affairs. The central part is what Halliday calls the 'process', typically represented by a verb. Three basic types of process are 1) material, representing actions and events, accompanied by an Agent and a Goal; 2) mental processes of cognition, perception and affectivity, in which the participants are Experiencer and Phenomenon; and 3) relational processes of being, seeming, becoming and owning, which have a Carrier and an Attribute.
If we examine this text for process type, we find a great preponderance of mental process verbs: recognise (twice, in 1 and 4), observe in 3, view in 5, decide in 7, and enjoy in 8. There are also other lexical items associated with mental processes: suspicion in 5, interested and spectators in 6. Now typically, Experiencer participants are human, especially in processes involving cognition or voluntary perception, as these items do. Here in this text, however, all these mental processes have birds and animals, particularly the collared doves, as Experiencer. We the humans are merely the Phenomenon, that which is experienced. And, moreover, humans represented by 'we' and 'my wife and I ' are the Phenomenon, not the Experiencer, no matter which syntactic position they occupy (Subject or Object), with the single exception of the subordinate clause 2, 'arrive home...', which is background information to the event expressed in the main clause. And again, we the humans are the Phenomenon no matter whether we are textually the Theme as in 3, 5 and 6, or part of the Rheme as in 4, 7 and 8. It is likely that it is the author's ideological stance which has, unconsciously no doubt, guided him towards the selections made, and which confer on this text its characteristic 'texture'.
Returning again to the dimension of genre, a possible avenue of research is the relationship between genres or sub-genres and the progression types most frequently occurring. Familiarity with a variety of genres might lead to predictions that particular text-types would display preferred sequences of choices as regards thematic progression. One recent contrastive study of procedural texts in English, German and Italian found differences in the preferred options, notably in the presence of Type 3 (derived themes) in the English texts and their absence from the others. These differences were accounted for by the language-specific text-structuring of the administrative forms, and also by factors such as social distance, which differ across cultures (Downing, Lavid, Belmonte & Taboada 1995).
Danes's own illustration of Type 3 is a short description of New Jersey. Intuitively, we would feel that texts about places tend to occur in various written genres (discounting for the moment spoken genres such as casual conversation): encycopaedia entries, tourist guidebooks, travel brochures, and as settings to narratives. We might be able to hypothesize, on the basis of linguistic evidence regarding the realizations of Themes and Rhemes, which type of thematic progression might be prevalent in this particular genre.
Descriptions of places, as of persons and things, abound in everyday printed matter. What follows is a short description of St Vincent, in the holiday section of a Sunday newspaper: (Themes are underlined):
St. Vincent is small; 18 miles long and 11 wide, mountainous
and lush (1). Banana plantations cling to steep volcanic
hills (2) and coconut palms sway in the brisk trade winds
which lash the Atlantic coast (3), stirring up its black
sand (4). The people are warm, friendly and poor (5).
Unemployment is between 30 and 40 per cent (6) but few go
hungry in such lush surroundings (7).
(From The Observer)
As in Danes's illustration, the thematic progression type adopted in the St. Vincent text consists entirely of derived themes; the only exception is that of clauses 3-4 which is simple linear, with 'trade winds'in clause 3 picked up as zero in clause 4 ('stirring.....). Semantically, however, there is greater variety. Whereas the New Jersey text Themes were restricted to geographical features, in the Saint Vincent text they include not only those of the landscape but also the people. Here too, process types are not limited to the relation type of being, realised simply by be, since there are some action processes (cling, sway, lash); and even in the relational processes, attributes are lexically richer than in the New Jersey text. Rhetorically, the writer of this text seems to have made choices aimed at an educated, though conservative English-speaking tourist reader with conventional ideas about tropical countries. Banana plantations are considered exciting, as are cocnut palms, and are given dynamic processes with appropriately descriptive accompanying circumstances. The rhetorical purpose of such texts is that of persuading the potential holidaymaker, in contrast with the purely informative purpose of the encyclopaedia entry.
Even greater variety is found in another description of a place, aimed at a university readership with a strong preference for Mediterranean holidays with classical associations:
Three hours by boat from the Ionian island of Kerkyra
(Corfu) lies Paxos (1). As a holiday destination, it
was discovered by the Italians a few years ago, (2)
following in the steps of the Venetians (3). Thus far,
the rugged islanders have succeeded in defending the
tranquil simplicity of the island, of its small harbour
town and fishing villages (4). A short motor boat trip
from the harbour town of Gaios is Antipaxos ( 5), with
turquise waters and beaches softer than silk (6).
From Oxford Today
1-2 Simple linear (Paxos - it)
2-3 Simple linear (condensed) Italians - following
3-4 Derived (It (Paxos) - rugged islanders
4-5 Simple linear (harbour town - harbour town)
5-6 constant (condensed)
Even a cursory glance reveals that this text is at once more varied and more complex. On the negative side, readers may consider it more contrived. The long theme of (1) represents the means of reaching Paxos, and presupposes that the tourist will know all about Corfu under either of its names and will be willing to go there first. TP from (1) to (2) seems to be Type 1 (simple linear) (Paxos -it), with topic continuity maintained by means of the passive. The Rhemes in (2) and (3) appear to have the purpose of introducing something of the island's history. TP from (2) to (3) is simple linear Theme (Type 2), if we make the non-finite clause into a finite, as suggested by Danes (1974:115). Clause (4) starts with a derived theme (the rugged islanders) while the Rhemes of (4) also represent features of the island which in another write-up might have been placed as a series of derived themes. TP between (4) and (5) can be assigned to Type 1, though semantically complex, since it also includes the means of reaching the second island; while (5) to (6) represents constant TP of the condensed type.
This text is aimed at a readership which has a somewhat romantic attitude to all things Greek. The writer has managed to pack a great deal of information into a small space, including conventional evaluative phrases (rugged islanders, tranquil simplicity), designed to capture the educated but less wealthy holidaymaker in search of simplicity combined with exclusiveness.
To conclude, all of these texts are instances of professional writing whose rhetorical purpose ranges from entertainment to persuasion. Danes' model appears to be fully applicable in helping to reveal the internal organization of the text. An awareness of the types of thematic progression as expounded by Danes also raises interesting questions for language teaching pedagogy which cannot be dealt with here. For instance, whether student compositions might be improved as regards what is subjectively termed the 'flow' of the text, by consciously selecting certain thematic progressions. Within the context of learning to write, it would also be profitable to establish what limitations tend to be imposed by writers on an otherwise indefinite recursion of a particular TP type. For instance, an excessive consecutive use of Type 1 or Type 2, without variation or combining, might well lead to diffuseness and what is often termed 'wandering from the topic'.
Alternatively, this type of analysis can also be applied to the comparison of professional writing in tabloids with that of broadsheet journalism, with the aim of ascertaining whether there is any connexion between perceived coherence and thematic progression. Finally, in spoken genres, such as lectures, conversations, including interactive story-telling and jokes, an examination of thematic progression choices may yield interesting results.
Danes, F. 1974. Functional sentence perspective and the organization of the text. In F. Danes, ed. Papers on Functional Sentence Persepctive (106-128). Prague: Academia /The Hague: Mouton.
Downing, A. 1990. Sobre el tema tópico en inglés. Revista Española de Lingüística Aplicada. Anejo 1. Nuevas corrientes lingüísticas. Aplicación a la descripción del inglés. Editado por María Teresa Turrell (119-128).
Downing, A. 1991. An alternative approach to theme: a systemic- functional perspective. WORD vol. 42, number 2 (119-143).
Downing, A., Lavid, J., Belmonte, I. & Taboada, M. 1995. Socio- cultural factors and information progression strategies in multilingual administrative forms. Paper read at the VII International Systemic-Functional Workshop, Valencia, July 1995.
Dubois, B. L. 1987. A reformulation of thematic progression typology. Text 7 (2) (89-116).
Firbas, J. 1976. A study in the functional sentence perspective of the English and Slavonic interrogative sentence. Brno Studies in English 12: 9-57.
Firbas, J. 1986. On the dynamics of written communication in the light of the theory of Functional Sentence Perspective. In Cooper, C. and S. Greenbaum eds. Studying Writing: Linguistic Approaches. Beverley Hills: Sage (40-71).
Francis, G. 1994. Labelling discourse: an aspect of nominal group lexical cohesion. In Coulthard, M. ed. Advances in Written Text Analysis. London: Routledge (83-101).
Fries. C. 1983. On the status of theme in English. In Petöfi, J. and E. Sözer eds. Micro- and macro-connexity of texts. Hamburg: Buske Verlag (116-152).
Fries, C. 1994. On Theme, Rheme and discourse goals. In Coulthard, M. ed. Advances in Written Text Analysis. London: Routledge (229-249).
Givón, T. 1990. Syntax. A Functional-Typological Introduction. Amsterdam/Philadelphia: Benjamins.
Givón, T. 1993. English Grammar, I and II. Amsterdam / Philadelphia: Benjamins.
Halliday, M.A.K. 1967. Notes on transitivity and theme in English, Part 2. Journal of Linguistics 3 (177-274).
Halliday, M.A.K. 1994 (second edition). An Introduction to Functional Grammar. London: Edward Arnold.
Leech, G. 1983. The Principles of Pragmatics. London: Longman.
McCarthy, M. & Carter, R. 1994. Language as Discourse. Perspectives for Language Teaching. Cambridge: Cambridge University Press.
Nichols, J. 1984. Functional theories of grammar. Annual Review of Anthropology 13 (97-117).
Nwogu, K. and Bloor, T. 1991. Thematic progression in professional and popular medical texts. In Ventola, E. ed. Functional and Systemic Linguistics: Approaches and Uses. Berlin: Mouton de Gruyter (369-384).
Popper, K. R. 1979 (second edition). Objective Knowledge. An Evolutionary Approach. Oxford: Oxford University Press.
© Angela Downing. Círculo de Lingüística Aplicada a la Comunicación/ Circle of Linguistics Applied to Communication 5, February 2001. ISSN 1576-4737. Published in M.T. Caneda and J. Pérez (eds.), Os estudios ingleses no contexto das novas tendencias, Vigo 1996, Universidade de Vigo, 23-42.
Back to contents 5