Summary #e article presents some aspects of text explanation in the systemic functional model of language. In the systemic functional model text is conceived as a semantic unit created in the process of selection and realization of meaningful choices encoded in words and structures. #is implies that in order to uncover what types of meanings are being encoded in the text one has to take into account the lexico-grammatical patterns which realize it. #e article treats some of the possibilities for the application of this model in a comparative text analysis involving texts in different languages. As an illustration it presents a comparison of some aspects of the experiential function of two presidential speeches, one in English and one in Slovene. #e comparison focuses on the distribution of different process types in both texts. Povzetek Članek obravnava nekatere vidike razlage besedil v sistemsko-funkcijskem modelu jezika. V tem modelu je besedilo pojmovano kot pomenska enota, ki nastaja v procesu uresničevanja pomenskih izbir, ki so izražene z besedami in strukturami. Iz tega sledi, da je potrebno pri ugotavljanju besedilnih pomenov upoštevati leksikalne in slovnične vzorce, ki te pomene uresničujejo. Članek poskuša razkriti nekatere možnosti za uporabo tega modela v primerjalni analizi besedil v različnih jezikih. Za ilustracijo je predstavljena primerjava nekaterih vidikov predstavne funkcije v dveh besedilih, predsedniških govorih, od katerih je eno v angleščini in eno v slovenščini. Primerjava je osredotočena predvsem na distribucijo glagolskih dogodkov v obeh besedilih. DOI: 10.4312/elope.1.1-2.23-36 Scholarship on text and discourse linguistics has drawn from different theoretical sources. An important contribution to the study of text has come from functional approaches to language. One of the more influential functional schools is the theoretical framework developed by M.A.K. Halliday and his followers. #eir linguistic orientation has become known under the term systemic functional (SF) linguistics. Grammatical categories and the model of text developed within the framework of SF linguistics have been applied in a wide range of text linguistics contexts, from discourse analysis to literary stylistics and criticism, from text typology to register and genre theory. It has also become the theoretical background of much of today’s study of the relationship between language, power and ideology. In order to explain why this particular model of language is especially suitable as the theoretical underpinning of different kinds of text studies, we must first observe some common features of linguistic functionalism in general. #e core question in linguistic functionalism is why language is as it is given the functions it has to fulfill. #is also includes the question of what is natural or iconic about the structural code of language and how this code is constrained by the cultural and biological environment in which it has evolved. According to Givon (1995, xv), the lowest common denominator of all functional thought is the non-autonomy postulate - the fundamental assumption that language, including grammar, can be neither described nor explained adequately as an autonomous system. In order to understand linguistic structures one has to take into account different parameters that shape language, such as cognition and culture, change and variation, acquisition and evolution, the brain and language processing; in short, all the complexity of the bio-cultural reality of language. Different functional schools have treated these parameters with a varying degree of emphasis, but they all accept some version of the non-autonomy postulate. Since they all emphasize the relevance of the performance to our understanding of synchronic structure, they usually view themselves as some kind of opposition or complementation to Chomskyan linguistics. Halliday (1994, xxviii), for example, sees the basic opposition in grammars of the second half of the twentieth century as that between formal grammars with their roots in logic and philosophy and the functional ones with their roots in rhetoric and ethnography. #e return to preoccupation with discourse in the 1970’s has increased the importance of the anthropological point of view and of the conceptualization of language as a resource for meaning, rather than as a system of rules. Functionalism rejects a strict segregation of performance and competence and stresses the importance of the rich textual database for the study of the system. According to Halliday (ibid., xxii), the distinction between text and system means that text is conceptualized as a realization of the choices available in the system. Halliday (ibid.) also argues that the grammar is at once both a grammar of the system and a grammar of the text. Consequently, the study of text necessarily has to include the study of the system and vice-versa. Givon (1995, 5) explains that functionalism has developed to a large extent as a reaction to the dogmatic linguistic assumptions of structuralism, i.e. the legacy of structuralism and its dogmas, that is the arbitrariness, the idealization and the strict segregation of diachronic from synchronic description. Givon (ibid., 6) dismisses such formalist reductionism in linguistics by arguing that the idealization, although useful as a methodological device, dissimulates the fact that all the functional-adaptive pressures that shape the synchronic structure of language are exerted during actual performance. For this reason, functionalists are normally text-sensitive linguists who focus on texts as authentic products of social interaction and are concerned with issues such as properties of discourse and their frequency distribution. It is thus primarily in text where they seek explanations for the facts of structure. Functionalism, however, has also been criticized for its own brand of reductionism. Criticism often comes from linguists working within the functional paradigm itself. Givon (1995), for example, criticizes the type of functionalism which denies formal morpho-syntactic structure its own reality, using circular definitions and not taking cognition seriously. It is the study of cognition that occupies a prominent position in Givon’s work (ibid., 389), which conceptualizes cognition as the “invisible process” underlying the “visible artefacts” of the observable communicative transaction. In functionalism of the Hallidayan type, cultural explanations and aspects of language have traditionally been given priority. In Hallidayan linguistics, the main stress thus lies on the socio-cultural factors that have shaped language. Halliday (Halliday and Hasan 1989, 4) is careful enough to stress that the socio-cultural perspective he adopts in his work does not exclude other modes of interpretation. However, his functional-systemic model of language tries to describe and explain language primarily in terms of the social aspect of human experience. Some more recent work in linguistic functionalism (notably Givon 1995; Dik 1989; Croft 2000; etc.) tries to include all the factors constraining the use and evolution of language, i.e. its communicative tasks and the socio-cultural as well as cognitive and biological contexts in which it has evolved. T o attain a more complete understanding of linguistic phenomena, both perceptual and cultural aspects of cognition are taken into account. Dik (1989, 37) thus speaks of the culturally and psychologically defined cognitive world of natural language users. #e systemic-functional approach initiated by M.A.K. Halliday has been one of the most influential approaches within the functional paradigm. Its applications range from research of theoretical nature to practical problem-solving tasks. Over the years it has been applied particularly by discourse analysts who are concerned with the ideological aspects of language use, i.e. in critical linguistics (e.g. Martin 1992), but also in literary theory (e.g. Fowler 1996) and educational linguistics (e.g. Halliday and Hasan 1989). It seems that the reason for the the usefulness of this model in so many different areas is its focus on language as social rather than individual phenomenon. Although criticized by some as reductionist, this orientation has a lot to offer both to research within discourse studies and to language-related studies outside linguistics proper. As it attempts to develop a grammar that can provide insights into the meanings of a text, the discourse grammar of this kind is semantic in its orientation and thus sensitive to textual considerations. It wants to provide the basis on which to relate the language and the text to the non-linguistic universe of its situational and cultural environment. For this reason it is also popular with those literary theorists and critics who want to establish the linguistic bases of different aspects of literary writing. As Toolan (1998, viii) puts it, linguistic labels used for this kind of analysis should not be “empty ones”, rather, they should have “content” and they should “represent specific insights concerning the structure of texts and the nature of language communication”. It seems that this is why the semantically-oriented categories offered by Hallidayan linguistics, while disputed by some grammarians, are particularly suitable for literary stylistics, but also for the study of how language is used in texts in any other field of experience, e.g. advertising, political discourse etc. Before looking at the systemic-functional model of the text, we should first examine some assumptions on which it is based. Halliday (e.g. 1989) stresses the social nature of acts of meaning, as they are symbolic acts that will not work unless there is a receiver, pointing out that acts of meaning are by nature social acts and that all symbolic systems are social systems. #ese symbols have evolved to serve certain functions. In fact, the term functional is used in SF linguistics in three different but related senses. In the first sense, functional is synonymous with the term use, signifying a generalized notion of the functions of language. #ese uses are extrinsic to language, referring generally to the way people use language to achieve different goals. Halliday goes further by claiming that these uses of language are clearly reflected in the structure of language and form the basis of its semantic and grammatical organization. Functions are thus regarded as the fundamental semantic property of language itself. In the third sense, the term functional refers to the clause as a functional unit with corresponding organic configurations of functions, i.e. to grammatical functions. #e hypothesis of three general functions or metafunctions as the basic principle of the organization of the semantic system occupies the central position in the functional part of Halliday’s (e.g. 1976, 1994) approach to language. Metafunctions are defined as basic functions which reflect certain general uses of language. #ese are the experiential function (including logical semantic relations), and the interpersonal and the textual functions. #e experiential function is the speaker’s meaning potential as an observer. It represents the patterns of experience in terms of processes, participants and circumstances, i.e. the mental picture of the external world and the speaker’s consciousness. #e interpersonal function represents the participatory function of language, i.e. the function through which the speaker expresses her own attitudes, emotions and judgments and seeks to influence the attitudes and judgments of others. #e textual function is the enabling function, the speaker’s text-forming potential expressing the relation of language to its environment and weaving together the experiential and interpersonal meanings. #e main theoretical thrust of this concept is that metafunctional meanings are systematically reflected on the grammatical and lexical level. #is means that structures are conceptualized as configurations of grammatical functions which derive from all three universal metafunctions. #ese grammatical functions are treated as belonging to the same level in the system: on the clause level the main grammatical resource for expressing the experiential function is the system of transitivity, which specifies the different types of processes and consists of the process itself, participants in the process (e.g. actor, goal, beneficiary) and circumstances attendant on it. #e interpersonal function on the clause level is expressed through the system of mood and modality. #e category of mood directly or indirectly expresses speech functions of giving or demanding information (i.e. statements and questions) and giving and demanding goods and services (i.e. offers and commands). #e grammatical function of subject and the finite part of verbal forms are treated as part of this metafunction.#e textual function consists of the resources the language has for creating text. #is function subsumes the internal resources for structuring the clause as a message, i.e. the thematic structure based on the functions of theme and rheme and the informational structure based on the functions of given and new. In addition, textual function also includes grammatical and lexical resources (e.g. reference, lexical cohesion) which exceed the limits of clausal structure and are usually grouped under the concept of cohesion. In the systemic-functional model choices from transitivity, mood and theme are realized though a clause. Interpreting a sentence is, in Halliday’s words (1989, 18), trying to relate what we say about it to general categories that are found in the grammar of the language. #is interpretation also implies that language is structured to make three main kinds of meanings simultaneously, not separately. #e clause is the place where all the choices of meaning from these three different sets of options are fused together and function in parallel, relatively independent of each other. #e treatment of these three types of functions on the same level has also provoked criticism of functional intuitionism (with regard to the notion of theme), or even naive iconism (with regard to the concept of subject) - the practice of conferring functional-sounding labels on grammatical structures (Givon 1995, 309). However, one of the main strengths of the model seems to lie in its systematic treatment of the aspects of context and its influence on how we use language at the linguistic micro-level. It offers a coherent model of text and context by systematically relating the generalized categories of context with the three metafunctions (e.g. Halliday and Hasan 1989). T o summarize briefly, there are three categories in any context of situation that have linguistic consequences: field, mode and tenor. #e category of field denotes what is going on, i.e. the social activity type, in other words, what the language is being used to talk about. It thus typically involves the degree of technicality. Mode specifies the role of language in the interaction. It involves spatial and interpersonal difference in the relation between language and situation, i.e. the distance between language and the social process occurring in the situation and the possibility of feedback between the interactants. Consequently, the category of mode specifies the contrast between spoken and written situations of language use. #e category of tenor denotes the roles of the interactants and their relationships in the exchange, thus typically involving the degree of formality. #ese three variables form register and are in a solidary relationship with the three linguistic metafunctions: field is reflected in the experiential function, tenor in the interpersonal function and mode in the textual function. In order to provide insights into the meaning of text, grammatical categories should show that grammar is naturally related to text semantics. #e categories of grammar which systematically reflect the three dimensions are functional and semantic in orientation. As Martin (1992, 2) puts it, grammar interfaces in a responsible way with textual considerations. And conversely, text analysis should be grammatically responsible. #e opposition between “sentence-centered” and “text-centered” theories of language is in SF linguistics regarded as “a distortion of the nature of human language” (Hasan 1978, 228). In systemic functional theory text has a central place as the basic unit through which meaning is expressed. T ext is regarded not as some kind of super sentence but as a semantic unit which is created in the process of selection and realization of choices from the functionally organized meaning potential, which are in turn coded in lexical and grammatical patterns. #e systemic part is reflected in the idea that these choices are conceptualized as “paths or passes through the networks that constitute the linguistic system” (Halliday 1989, 11). #ese patterns themselves are related to the higher level of contextual dimensions. #is basically implies that one should start to uncover what types of meaning are being encoded in the text by looking at the lexical and grammatical patterns which encode it. Within this context grammar has a central place in text analysis. Halliday (1994, xvi) argues that “a discourse analysis that is not based on grammar is not an analysis at all, but simply a running commentary on a text”. #is approach thus emphasizes the role of language in the construction of reality. In view of all this, text analysis also comes to a certain extent within the scope of grammars designed for the study of isolated sentences one at a time. #is is so because the consistent choices of sentence grammars themselves produce cumulative data through which textual meanings are expressed. #ese choices in sentence grammar are connected by lexical chains and strings, and they can interact when one or two members of the same chain or string stand in similar transitivity relations to two or more members of another string or chain. From this it follows that by examining the lexico-grammatical patterns of the text we are in fact uncovering what types of meaning are being encoded in the text (cf. Eggins 1994, 84). #e analysis of a text in grammatical terms is regarded only as the first step, but is nevertheless considered as an essential element in text explanation and interpretation. #e described model also provides a useful analytical framework for contrastive text analysis. In Eggins’s view (ibid., 310), “contrastive analysis offers a relatively easy way to tackling text analysis, because it provides some picture of how an actual text is but one realization from a total potential”. Contrastive analysis involves texts that are similar in some respects, but different in others. Consistent linguistic choices in a text are more easily discernible if they are matched against the background of possible choices realized in other texts. Different introductions to the systemic-functional model of text usually use English as the language of illustration (e.g. Halliday 1994; Martin 1992; Eggins 1994). #e question arises whether the assumptions and the descriptions of categories they contain can be used as a basis for studying other languages and the analysis of texts in languages other than English. Although Halliday warns against ethnocentrism in modern linguistics, he does not question the “universality”, or rather “generality” of particular inherent properties of language as a semiotic system. Halliday (1994, xxxiii) thus suggests that the metafunctional hypothesis is postulated for all languages. #e content systems are organized into experiential, interpersonal and textual components in every language, but the descriptive categories derived from these components are then treated as particular. #e typological diversity of languages can present certain challenges if we want to undertake a direct comparison of English texts with texts in some other language on the basis of the categories of the SF text model. #e following paragraphs show an attempt at this kind of contrastive analysis, involving a text in English and a text in Slovene. #ey present some points of a comparative analysis (Plemenitaš 1998) which focus on certain lexical and grammatical features realized at the clause level and examine how they support generalizations concerning the subject-matter of the texts. #e first task involved in this exercise has been to postulate to what extent the descriptive categories that have already been established for the English language are also valid for the Slovene language and what adjustments are needed to establish a common ground of categories functioning as a basis for the direct comparison of the two texts. For example, if we look at how SF theory (e.g. Halliday 1994, 37-67) views the ways of achieving texture through the thematic system, we see that the systems of both languages differ quite substantially. In English the word-order is rather fixed, whereas in Slovene it is much more pragmatic and consequently more flexible. #us in Slovene the system of marked and unmarked themes in different moods is not realized by the same features as in English. An additional complication arises from the fact that in Slovene the subject, which corresponds to the unmarked theme in English declarative clauses, does not have to be explicit. Whereas this particular area holds interesting typological questions, the remaining two metafunctions seem to allow a more direct comparison of the two texts at the clause level. #us the following analysis focuses on the comparison of some features of the experiential component with the aim to establish how the interpretation of the patterns found bears on the subject-matter of the analyzed texts and what kind of interaction there is between experiential and interpersonal meanings. #e texts chosen for analysis belong to the genre of speeches. #e English text is the inaugural speech by the US president Bill Clinton given at his first inauguration in 1993, whereas the Slovene text is the speech by the Slovene president Milan Kučan on the occasion of the reconciliation festivities at Kočevski rog in 1990. #e experiential metafunction involves different process types encoded by the grammar of the clause as representation (cf. Halliday 1994; Eggins 1994). #is is achieved by the system which models experience as a manageable set of process types, i.e. the system of transitivity. A process consists of three semantic components: the process itself (realized in the verbal group of the clause), participants in the process (typically realized in the nominal groups), and circumstances associated with the process (typically expressed through adverbial groups and prepositional phrases). #e analysis of the experiential function presented here is primarily concentrated on the frequency distribution of different process types. Each process type determines different participant roles, and these can appear in different configurations. #e participants more closely connected with the process itself are nuclear or obligatory participants, whereas other participants are optional (opt.). #e analysis is based on the following overview of the process type system as it has been suggested by Halliday (e.g. 1994). Due to space limitation it shows only the general categories of processes and their participants (for a detailed survey and discussion of criteria for identification see Eggins 1994). Each English label is followed by a proposed Slovene translation in brackets. #e circumstantial system is not included here, as it was not part of the analysis. !e process type system: - material process (materialni glagolski dogodek): Actor (vršilec), opt. Goal (cilj), Range (območje), Beneficiary (koristnik), which can be Recipient (prejemnik) or Client (uporabnik), Attribute (prilastek); - mental process (duševni glagolski dogodek): Senser (doživljalec), Phenomenon (pojav); - verbal process (govorni glagolski dogodek): Sayer (govorec); opt. Receiver (prejemnik), Verbiage (govorjeno), Target (tarča); - behavioral process (vedenjski glagolski dogodek): Behaver (obnašalec), opt. Behavior (obnašanje), which can be Phenomenon (pojav) or Verbiage (govorjeno); - existential process (obstojanski glagolski dogodek): Existent (obstoječe); - relational process (odnosni glagolski dogodek), which is further divided into identifying (istovetnostni) with participants Token (element) and Value (vrednost), and attributive (prilastni) with participants Carrier (nosilec) and Attribute (lastnost), opt. Beneficiary (koristnik). #e three main types of process in English are material, mental and relational. Material types are processes of doing, representing the processes of the external world, whereas mental types are processes of sensing, belonging to the world of consciousness. Relational types are processes of being, construing relations of classifying and identifying. #e remaining three categories are located at the boundaries of the main types, i.e. behavioral processes on the borderline between material and mental types, verbal processes on the borderline between mental and relational, and existential on the borderline between material and relational types. Behavioral processes are processes of typically human physiological and psychological behavior, whereas verbal processes are processes of saying. Existential processes represent simply that something exists. #e fact that the system of language allows fuzzy areas at the margins of categories has functional reasons, since learning and diachronic extension of categories can proceed only through shaded graduality. But in the context of text analysis this often means alternative explanations of indeterminacies, places where one has to weigh one factor against another. All this makes a certain degree of interpretation necessary. For example, processes denoting typically human behavior, especially social behavior, often cannot be distinguished from material types . #us the clause from the inaugural speech “T oday, we celebrate the mystery of American renewal” is interpreted as a material process. Another example involves the interpretation of certain types which can be interpreted either as material or relational. According to Halliday (Halliday and Martin 1993, 65) verbs expressing the causing of a specific effect, e.g. speed up, encourage, improve (‘make faster, more likely, better’) can be interpreted either as relational or as material processes. In this analysis such clauses are interpreted as relational. For example, the following two clauses, one for each text, are interpreted as identifying relational processes with the causative relationship expressed through a causative circumstantial verb: Yes, you, my fellow Americans, have forced the spring; Omogočimo si življenje, spravljeno z mrtvimi. In view of the strong semantic orientation of the functional interpretation of experiential structure, the existing typological grammatical differences between the two languages do not represent any major difficulty for the identification of process types in Slovene by analogy and similarity with English ones, and consequently for the direct comparison of the experiential components in the two texts. #e same identification criteria as in English were also applied in the classification of process types in Slovene, with some adjustments where necessary. #e analysis of the Slovene text had to account for some structural configurations which are not analogous with the English ones, or have been given a different interpretation in Slovene grammar. For example, a special interpretation is required for Slovene structures with the ‘logical subject’, i.e. the actor or doer that is not in the nominative case, but whose oblique case is not the consequence of the clause negation or a premodifying quantifier. #is participant is also termed the dative or accusative of the logical subject (cf. T oporišič 1984, 205), e.g. Z njim je šlo na slabše. Functional grammar questions the ‘logical’ terminology of participant roles and postulates that different functional configurations reflect different interpretations of reality. #erefore such examples have been analysed as attributive relational, with no recognized actor or carrier, but with a recipient and an attribute, e.g. “T esno (attribute) bi nam (recipient) bilo”. In order to avoid the problem of the access of datives to subjecthood in Slovene, subjects of such clauses are simply treated as implicit in the verb. On the other hand, the participants occurring in the genitive case because of the negation of the clause have been given the same interpretation as participants in the nominative case in related affirmative clauses. In addition, there are types of Slovene relational clauses which in Slovene linguistics have received a slightly different interpretation than in the SF model. #e Slovene text abounds with relational clauses such as T ega dejanja ni mogoče ponoviti. In Slovene grammar the attribute in such a clause is classified as belonging to a word class called “naklonski nepregibni povedkovniki” (modal words used as subject complements) and the infinitive following the attribute as “nedoločniško predmetnopomensko določilo”, (infinitival complement) (T oporišič 1992, 120). #e analysis here, however, is based on the analysis of the interpersonal meaning of English clauses of the type It is not possible to do this, where the infinitive is treated as a postponed subject, and the pronoun it as anticipatory pronoun in the normal subject position. In terms of experiential meaning this would mean that possible is attribute and to do this carrier in an attributive relational clause. Accordingly, the above Slovene clause can be analyzed as a relational process with two participants, a discontinued carrier ((T)ega dejanja.....ponoviti) and an attribute (mogoče). Such analysis of this type of processes also has implications for the evaluation of the interpersonal function, in particular for the role of the subject. #ere is also a difference in the identification of existential processes. In English, processes of this type usually contain the structural there, whereas in Slovene they were identified on the basis of being introduced by a verb of existence (e.g. Je kraj življenja). Another criterion for the existential interpretation of such Slovene clauses is if their translations yield English existential clauses. It should also be noted that the unit of analysis has been limited to ranking clauses, i.e. clauses which are direct constituents of the sentence (or the clause complex in functional terminology), thus excluding clauses which function as elements of the nominal group. #e tables below show the frequency distribution of the main process types in both texts. Process type no. % Material 52 39.7 Relational 46 35.2 Existential 18 13.7 Mental 8 6.1 Verbal 7 5.3 Behavioral - - Process type no. % Material 85 54.8 Relational 32 20.6 Mental 19 12.3 Verbal 11 7.1 Behavioral 4 2.6 Existential 4 2.6 #e comparison of the use of process types shows the dominance of material process in both texts. However, the proportion of material processes in the inaugural text is considerably higher (54.8 %) in comparison with that of the reconciliation text (39.7%). In both texts, second place is occupied by relational processes, but the proportion of relational processes in the inaugural text is considerably lower (20.6 %) compared to that of the reconciliation text (35.2 %). Mental processes figure more prominently in the inauguration text, where they occupy third place (18, 11.6 %), whereas in the reconciliation text they are in fourth place (8, 6.1%). In the inaugural text, mental processes are followed by verbal processes in fourth place (11, 7.1 %), while in the reconciliation text these form the smallest group (7, 5.3 %), closely following the category of mental processes. Behavioral processes, which together with existential processes constitute the smallest proportion in the inaugural text (4, 2.6 %), have not been identified in the reconciliation text. However, the texts differ significantly in the use of existential processes. In the inaugural text they form the smallest group (4, 2.6 %), but in the reconciliation text they figure prominently constituting the third biggest category (18, 13.7%). #e following step in the analysis was to interpret the transitivity patterns in the inaugural and reconciliation texts as the expression of the contextual variable of field. Here we take into account that both texts are largely constitutive of their fields. Closer observation of individual process type patterns shows that in the inaugural text the main theme is change, or the need for change. #e most prominent patterns of process types are created by material and relational processes. Material processes reinforce the theme of the need for change and for activities demanded from participants in order to achieve it (e.g. we will work to shape change; we must revitalize our democracy; while America rebuilds at home; we must do the work), also representing characteristic processes in the present (powerful forces are shaking and remaking our world) and negative processes which have to be changed (e.g. the cost of health care devastates families; drifting has eroded our resources). Relational clauses mainly depict the state and characteristics of participants in the past and present (e.g. investment is global; technology is almost magical). Identifying relational clauses define entities which are important for the American self-image (e.g. our strength is the power of our ideas). 4 of 7 relational clauses display a circumstantial feature of causation (e.g. we force the spring; Communism’s collapse has called forth old animosities and new dangers), which reinforces the theme of bringing about change. #e text also contains mental process of all the 3 subtypes. Cognitive mental clauses express the knowledge or beliefs of participants about the proper way to act (e.g. we know what we have to do; they knew that America ....would have to change), or decisions to act properly ((L)et us resolve to reform our politics), while mental processes of perception put the participants metaphorically in touch with the world of concrete senses (e.g. see the promise; feel the pain; see ourselves in the light of posterity). Mental processes of affection depict emotional states of participants triggered by the mentioned processes and the state of affairs (e.g. worry endlessly about who is in and who is out), or state the need for emotional ties among the participants (e.g. we must care for each other). Verbal processes express mainly rhetorical processes (e.g. [W]hen George Washington first took the oath; they threaten to bankrupt our enterprises), some bordering on causative relational clauses (e.g. you have summoned the change). Behavioral clauses, which play only a minor role, are mostly used for the expression of celebrating the moment and paying tribute (e.g. I thank millions of men and women; I salute my predecessor). Existential processes form the smallest group. In all of them, the existent is expressed by a lengthy word group with a clause postmodifier. It seems they are mainly used for the purpose of moving the participant to the focal position at the end of the clause (e.g. [T]here is nothing wrong with America that cannot be cured by what is right with America). #us the meaning expressed by the existential process type is clearly in the background. In the Slovene (reconciliation) text the main theme is the need for the reconciliation of the Slovene people due to the fratricidal conflict during and after the Second World War. In this text, too, the dominant patterns are formed by material and the relational process types, but the emphasis is different. #e proportion of material clauses in the reconciliation text is smaller, and they mainly express happenings in the past that led to the need for reconciliation (e.g. [S]lovenski narod so pribili na križ; so zgorela premnoga slovenska življenja; tu smo se pobijali – the Slovene nation had been nailed to the cross; many Slovenian lives were burnt down; here we killed each other ), and symbolic reconciliatory gestures (e.g. pokopljimo mrtve; končajmo narodno diasporo; obrnimo se k skupni prihodnosti – let’s bury the dead; let’s end the national diaspora; let’s turn ahead to the common future). Identifying and attributive relationals form the second largest group. #e identifying relationals express the definition of reconciliation or reconciliatory acts (e.g. sprava...je preizkušnja narodove zrelosti; ...je odpoved lažnemu upanju; ...je zgodovinska nujnost – reconciliation… is the test of the nation’s maturity; …is the renunciation of false hope;… is a historical necessity). Attributive relationals also constitute a large proportion, particularly those with modalized attributes expressing possibility or impossiblity to act (e.g. [M]ogoče nam je prepoznati okoliščine, dejstva; [N]aše preteklosti ni mogoče predrugačiti ne zamolčati - It is possible to recognize the circumstances, facts; our past cannot be changed or kept untold). #e subcategory of possessive relational clauses is mainly used to express generic statements, including quotations from the Old T estament (e.g. [V]se ima svoj čas, vsako opravilo ima svojo uro pod nebom – to everything there is a season, and a time to every purpose under the heaven). #e proportion of existential clauses in this text is considerably higher than in the inaugural text. #e existential clauses, which represent that something exists, form a kind of meaningful complementation to the modalities expressed in the attributive relationals and emphasize the feeling of determinism. #is is also explicitly expressed in the following sentences, taken from Marx: “Človek sam dela zgodovino. T oda dela jo v razmerah, ki si jih ne izbira v celoti sam” (Man creates history by himself. But he creates it in circumstances which are not entirely of his own choice.) #e feeling of the forces of fate is invoked already at the beginning of the text by a string of existential clauses ((J)e kraj življenja in je kraj smrti; (J)e čas sovraštva in je čas strpnosti - there is a time of life and a place of death; there is a time of hate and a time of tolerance), followed by others of the same type later in the text ([J]e čas vojne in čas ubijanja; [J]e čas odločitve in čas žrtvovanja; [B]ila je vojna, bila je okupacija - there is a time of war and a time of killing, there is a time of decision and a time of sacrifice; there was war, there was occupation). Existental processes, in two cases even wholly stripped of the existent and circumstances ( bilo je – it was), emphasize the need to come to terms with what cannot be changed. In this way the text avoids explicitly blaming either side, and rather stresses the tragic nature of human existence as such. Mental clauses are mainly of the cognitive type and are associated with the events in the past (e.g. ne smemo pozabiti tistih, ki smo se jih spominjali - we mustn’t forget those we have remembered). Verbal processes are less frequent, generally projecting desired commitment to reconciliation (recimo: nikoli več, povejmo si: tu, kjer so posute kosti – let us say it: never again; let us say it: this place strewn with bones…) or rhetorical relations among participants (in znova bomo grozili drug drugemu - and again we will threaten each other). A particular feature of the reconciliation text is frequent fragmentation of processes, which results in independent units consisting of word groups, and parts of hypotaxis or parataxis (e.g. Smo na kraju smrti. Pripravljeni in odločeni - We are at the place of death. Ready and determined…) #e analysis of the experiential component in the inaugural and reconciliation texts also takes into consideration the fact that the extension model of the system of transitivity possesses features of causation. #is kind of interpretation is becoming increasingly important in English. #e key function in causative structure is called the medium, i.e. the entity through which the process comes into existence. If there is also an external cause, it is mapped onto the function called agent (Halliday 1994, 161-73). For example, the function of agent is mapped onto that of actor in material processes with a goal. In causative material and relational processes the role of agent is distinct from other functions (e.g. they (agent) made her go, they (agent) elected him President). In the analysis presented here, the function of agent has been given a slightly different interpretation, following Dik (1989, 96). Here agent is interpreted as the function which typically has the power to determine whether or not the process will obtain, i.e. it is the controller of the process. External agency with no control is called the force. #e evaluation of the function of agent is of importance for the concept of experiential responsibility and for the analysis of the interaction between the representational and interpersonal meanings. Taking this into consideration, the analysis has established that the texts differ substantially according to the distribution of controlling participants, i.e. agents. #us the proportion of process types containing agents is 55.5% in the inaugural text compared with only 42.7% in the reconciliation text. #e analysis of the interaction between experiential and interpersonal meanings also shows significant differences. In systemic functional linguistics, subject is interpreted as the modally responsible participant, i.e. the element which is responsible for the realization of an offer or command or for the validity of the information in a statement or question. Modal responsibility differs from experiential responsibility. Experiential responsibility is assumed by agents as the real instigators or controllers of the process. #e analysis shows that the texts differ substantially according to the proportion of subjects which are also agents and thus assume experiential responsibility. In the inaugural text the proportion of subjects which are also agents constitutes 88%, whereas in the reconciliation text only 55%. #is distribution is thus consistent with the thematic interpretation of the texts based on the frequency of process types. We conclude from the above analysis that the semantic consistencies that occur as a consequence of transitivity patterns represent generalizations that influence which aspects of reality are emphasized in both texts. On a global level, the reconciliation text stresses the influence of circumstances and the participants’ need to come to terms with their history, whereas the inaugural text is about the need for the participants’ active involvement in the forthcoming changes. #e question remains whether analysis of this kind proves that discourse patterns can reinforce or even modify the reader’s habits of modelling an external world. But it undoubtedly shows that the explanation of texts based on various grammatical consistencies offers a useful analytical tool in the comparative study of textual meanings. It helps to establish differences and similarities between texts by identifying consistent choices from the functionally organized linguistic resources. Furthermore, analysis of this type can also be extended to the level of context, for example as a basis for the study of textual genres.