Systemic Functional Grammar – Part 4 (The Textual Metafunction & Conclusion)

This post is the final part of a series on Systemic Functional Grammar. Read the first part for an overview of SFG. The second part examined the Interpersonal Metafunction and the third part considered the Experiential Metafunction.

In this final part of my exploration of SFG, I look at the Textual Metafunction, a part of SFG that is used in EFL teaching more than the other metafunctions, and often related to cohesion and coherence. Following that, I will briefly outline my views on the implications that SFG should be having on our teaching and conclude this four-part thread on SFG.

This post is based on research from the following books

An Introduction to Functional Grammar

The Functional Analysis of English

Using Functional Grammar: An Explorer’s Guide

Aptly named ‘Clause as Message’ (Halliday and Matthiessen, 2004), the textual metafunction helps organise the message within and between clauses, and is closely linked to theories of cohesion. I will only be looking at textual organisation within the clause.

Thematic Structure

The Theme is the departure point the speaker has chosen for his/her text. In English, the Theme, on which the clause depends for its orientation within the context (ibid), takes initial position in the clause. Theme^Rheme makes up the thematic structure of a clause.

Theme and Rheme

Chia bought some curry yesterday.
Theme Rheme…. 
Paul was bought some curry yesterday.
Theme Rheme…

Multiple Themes

Well, surely, Ken, grammar is your forte.
Textual Theme Interpersonal Theme Interpersonal Theme Topical Theme Rheme
Textual Adjunct Mood Adjunct Vocative Subject Finite Complement

Textual (discourse markers, conjunctives) or interpersonal (vocatives, Mood Adjuncts) Themes can combine to create multiple Themes, but it is the topical Theme, the first word carrying meaning in an experiential sense, realised by a Participant, Process or Circumstance, that is incorporated in every clause and anchors the starting point of the message (Bloor and Bloor, 2004). In an unmarked sentence, the topical Theme is the Subject of the interpersonal metafunction.

The most common marked Themes utilize adverbial groups or prepositional phrases serving as Circumstantial Adjuncts. More highly marked are Themes realized by nominal groups that are not Subjects, as seen in many informal spoken conversations e.g. ‘My reading, I’m done with it.’ Syntactical structures are highly neglected in EFL, as most grammar work focuses on tenses. Students use sentences like ‘Mario, yesterday, I gave the book to him,’ without intending to have a highly-marked Theme. It is perhaps useful to teach students basic unmarked structures, especially at lower levels.

Information structure

The Theme-and-Rheme theory was first conceived in The Prague School, where Themes were associated to the Given unit of information, and Rheme to the New (ibid). While ‘Given’ refers to the previously-mentioned, or the un-newsworthy, ‘New’ reveals new information or what is deemed newsworthy, and is often indicated by the placement of the tonic nucleus. Given+New makes up the information structure of the clause. I have chosen to use ‘+’ over ‘^’ because, for Halliday, Given units are not always thematized. In unmarked imperatives, the Theme is assumed to be ‘you’ (ibid). In marked declaratives, New could occur anywhere, especially in spoken English, where contrastive stress is shown through intonation change, e.g. ‘Derren has three brothers, not me.

Thematic and information structures occur across languages but may succumb to different rules. In Latin-based languages, the inflected nature of words allows positional flexibility, and results in interference errors such as ‘Today, happened something.’  In Japanese, Themes are signalled by the particle ‘wa’, while in Chinese, ‘ba’ is added to Complements when highlighting the Process as New (Halliday and Matthiessen, 2004). However, because we cannot say ‘I the food disposed’ in English, separable phrasal verbs have evolved to allow the Process to be the New, as in ‘I threw the food away’ (ibid). Information organisation through thematic and information structure principles can enable students to construct full texts with appropriate attention to the desired details and overall message.

Implications in Teaching

SFG attempts a ‘view from above’ (ibid), and the description of how each metafunction is realised in their systems and structures is best applied to the observation of authentic language use. Authentic texts can be analysed for structures related to the three metafunctions, encouraging students to spot patterns within the needed genres, e.g. the thematic progression of discursive essays tends to follow certain patterns (Bloor and Bloor, 2004), and EAP students or those preparing for language exams like IELTS can use this framework to develop their essay-writing skills.

Besides a deeper knowledge of how language is used to create meaning, teachers can gain a better understanding of how language use has affected its evolution, how English differs in its functional lexico-grammatical structures, and how to better enable students to use these structures effectively. The slot-filling approach of SFG allows students to practise recognising the types of words/groups that fit into different parts of sentences, making the task of sentence construction more manageable. Assessment of student production also becomes less arbitrary, and instead of vague feedback like ‘be clearer,’ or ‘wrong register’, teachers can analyse students’ use of structures within the three metafunctions in comparison to well-produced texts, and precisely identify how they can improve their communicative skills (ibid).

However, the use of SFG in the classroom is not without problems. Being a descriptive grammar, its extensiveness could intimidate both teachers and students, who might prefer simpler rules that are easily applicable. Thompson (2004) admits that analysis of authentic texts could prove tiring and may not be as feasible as it is useful. The huge number of technical terms needed to describe language in SFG is another obstacle. Butt et al (2000) claims that we can narrow the selection of terms to teach students, yet argues that such metalanguage is necessary in making the finer distinctions in language, and does not suggest which metalanguage to teach. Currently, many ESP coursebooks have taken to teaching cohesion and textual organisation in the spirit of SFG, but it remains to be seen if the rest of SFG would make it into EFL syllabuses.


SFG offers a view where the purpose of language is to mean, and meaning could refer to our stance regarding a proposition or proposal (interpersonal), the representation of our experience or consciousness (experiential) or the relevance of its organisation in the surrounding context (textual).  Each of these different dimensions offers choices within a system, where meaning is realised in a variety of potential structures. As language continues to evolve to cater to new meanings that need representation in communication, the application of SFG to the language classroom can help teachers and students understand the overall picture and enable them to become better communicators through an understanding of how language works.

Systemic Functional Grammar (Part 3 – The Experiential Metafunction)

This is the third part of a series summarising the basic concepts of Systemic Functional Grammar. The first post was a overview of SFG, the second post looked at the Interpersonal Metafunction, this post will be looking at the Experiential Metafunction.

In my fourth and final post, I will be examining the Textual Metafunction and considering how SFG applies to our language classroom.

This post is based on research from the following books

An Introduction to Functional Grammar

The Functional Analysis of English

Using Functional Grammar: An Explorer’s Guide

If the Interpersonal Metafunction showed the point of view of the speaker at the moment of speaking through the Mood element, one could say that the Residue carried the content of that message. And the content, in Hallidayan linguistics is categorised by different processes defined by the main verb of the clause. Here’s the 2nd Metafunction.

The Experiential Metafunction

Halliday (1976) originally purported that the experiential metafunction was one of the three main metafunctions. He later includes it within the ideational metafunction (Halliday and Matthiessen, 2004), alongside the logical metafunction of relationships between clauses and clause-complexing. Thompson (2004), however, sees the latter as a fourth metafunction. As clause-complexing is beyond my scope, the focus will be on the experiential metafunction.

Also called Clause as Representation, the clause represents the ‘content’ of our experiences, answering the question ‘Who does what to whom’. This metafunction uses the grammatical system of transitivity. Although sharing the traditional view of transitivity that the focus is on the verb group (the Process), SFG refers to the system as describing the whole clause (Thompson, 2004) and does not use the labels ‘subject’, ‘verb’ and ‘object’, seeing that ‘verb’ is a word class, while ‘Subject’ is a functional term. Instead, different functional labels are given to Participants (realised by nominal groups), Processes (realised by verbal groups) and Circumstances (realised by prepositional phrases or adverbials signifying time, place or manner) of each process type.

Material process clauses refer to experiences of the external world and describe processes of doing and happening, answering the question ‘What did he/she do?’ or ‘What happened?’ (Butt et al, 2000). The distinction between participant types (Beneficiary, Goal, etc.) can alleviate students’ confusion as to which can be turned into a prepositional phrase and shifted to the end to be highlighted as newsworthy. Also, in the experiential metafunction, functional terms define the roles the Participants play in the Process, and Goals or Beneficiaries can take Subject position.

Material Process

Chia bought some curry yesterday.
Actor Process: material Goal circumstance
Nominal group Verbal group Nominal group Adverbial group

Material process with Beneficiary

Chia bought Paul some curry yesterday.
Actor Process: material Beneficiary Goal circumstance
Nominal group Verbal group Nominal Group Nominal group Adverbial group

Material process with Beneficiary shifted to the end

Chia bought some curry for Paul yesterday.
Actor Process: material Goal Beneficiary circumstance
Nominal group Verbal group Nominal group Nominal Group Adverbial group

Agentless passive structure with Beneficiary as Subject

Paul was        bought some curry.
Beneficiary Process: material Goal
Subject Finite Predicator Nominal group

Agentless passive structure with Goal as Subject

Some curry was         bought for Charles.
Goal Process: material Beneficiary
Subject Finite Predicator Nominal group

Relational process types serve to identify and characterize, and are further subdivided into processes of ‘being’ (intensive or circumstantial) and ‘having’ (possessive). (Halliday and Matthiessen, 2004).  Attribution specifies the class the Carrier belongs to, while identification narrows the class down to one.

Relational Process of intensive attribution

Emma is pretty.
Carrier Process: Relational: intensive atrribution Attribute
Nominal group Verbal group Nominal group with adjective as Head.

Relational Process of intensive identification with Value as Subject

Emma is the prettiest.
Value Process: Relational: intensive identification Token

Relational Process of intensive identification with Token as Subject

Emma is the leader.
Token Process: Relational: intensive identification Value

In ‘Emma is the prettiest one’, ‘Emma’ is the Value identified by the Token ‘the prettiest one’, since Emma is represented by the prettiest one. However, in ‘Emma is the leader’, ‘Emma’ represents the leader and is now the ‘Token’, identified by ‘the leader’ as the Value. Because of this structural distinction, we cannot combine the two and say, ‘Emma is the prettiest one and the leader’. (Halliday and Matthiessen, 2004)

Below, are more examples of attribution and identification within other types of relational processes.

Relational Process of possessive attribution

Emma has a Wii console .
Carrier Process: Relational: possessive attribution Attribute

Relational Process of possessive identification

That Wii console is Emma’s.
Token/Possessed Process: Relational: possessive identification Value/Possessor

Relational Process of circumstantial attribution

The deadline is on Tuesday.
Carrier Process: Relational: circumstantial attribution Attribute

Relational Process of circumstantial identification

Tuesday is the deadline for the blogpost.
Token Process: Relational: circumstantial identification Value

Serving to construe processes of sensing, happenings within our consciousness, mental processes are subdivided into processes of emotion, perception, cognition, and desideration (ibid).

Mental process of emotion with nominal group as Phenomenon

David liked the headphones.
Sensor Process: Mental: Emotion Phenomenon

Mental process of perception with embedded clause as Phenomenon

David saw what happened.
Sensor Process: Mental: Perception Phenomenon

Mental process of cognition with projected clause

David knew he was getting headphones for Christmas.
Sensor Process: Mental: Cognition Projected Clause

Mental process of cognition with projected clause

David hoped that he would get headphones for Christmas.
Sensor Process: Mental: Desideration Projected clause 

Notice that in mental processes of emotion and perception, what is loved or hated, seen or heard, is labelled Phenomenon, even when the fact is realised as an embedded clause. However, mental processes of cognition and desideration often bring wishes and ideas into existence by projecting a separate clause (Thompson, 2004).

Pairs such as ‘like/please’, which show a different direction in Sensor-Phenomenon relationships, are often unaccounted for in traditional grammar, but commonly occur in cognitive and emotive mental processes (ibid).

David liked the headphones.
Sensor Process: Mental: Emotive Phenomenon
The headphones pleased David
Phenomenon Process: Mental: Emotive Sensor 

In ‘David liked the headphones’, ‘David’ is the Sensor of this emotive mental process which is denoted by the verb ‘liked’. ‘The headphones’ is the Phenomenon which summarises what is thought, perceived, or liked/disliked. Contrast that with ‘The headphones pleased David’, where the Subject is now the Phenomenon, and the Sensor, which is the conscious being, fills the interpersonal slot of Complement.  The example below shows that ‘realise’ is a ‘like’ type verb, while ‘occur to’ is similar to ‘please’. (Halliday and Matthiessen, 2004)

David realised the fact that he was wrong.
Sensor Process: Mental: Cognitive Phenomenon
The fact that he was wrong occurred to David.
Phenomenon Process: Mental: Cognitive Sensor

Other process types

Between material and mental are behavioural processes, while existential processes are between relational and material. Verbal processes share the ability of mental processes to project what is said or thought in a separate clause. Some indirect-speech verbs, e.g. ‘urge’, ‘force’, which take  to-infinitives when projecting,  and direct-speech verbs like ‘whispered’, ‘sneered’ can convey illocutionary force (Bloor and Bloor, 2004).

Behavioural process

Joe sang the song.
Behaver Process: Behavioural

Existential process

There was a boy.
Process: Existential Existent

Verbal process projecting direct speech as separate clause

Alan said, “You should read.”
Sayer Process: Verbal
Quoting Quoted
Actor Process: material

Verbal process projecting indirect speech as separate clause

Alan said you should read.
Sayer Process: Verbal
Reporting Reported
Actor Process: material 

Different process types have different tenses as their basic, unmarked, forms, e.g.  while the unmarked present tense for material processes is the present-in-present (the present progressive), that for relational and mental processes is the present simple. Halliday and Matthiessen (2004) criticise EFL syllabuses for over-simplifying the teaching of the present simple as habitual behaviour without considering that this largely depends on process types. Another oversimplification is the famous pedagogic rule that state verbs cannot take the progressive tense. However, many verbs like ‘have’ can be either states or actions, which can be confusing for students. SFG’s separation into different processes (‘have a shower’ is material, while ‘have a pen’ is possessive attributive), which are governed by different sets of rules, explains the phenomenon to students more clearly, e.g. the present-in-present for relational and mental processes implies a highly-marked narrowing of the present, which tends to signify temporality e.g. ‘I hate burgers but I’m loving this one!’ (ibid).

Furthermore, different genres have the tendency of using certain process types more than others. A written recipe contains material processes, while a chef on a cooking programme might use a combination of material and relational processes (Thompson, 2004). Existential processes are often used in narratives to introduce new characters or scenes. A good understanding of the corresponding grammatical features of processes can help students use them in expressing their experiences of the world in the appropriate register, and define the syllabuses for students learning English for Specific Purposes (ESP) or Academic Purposes (EAP).

(Bibliography in Part 1)

Systemic Functional Grammar (Part 2 – The Interpersonal Metafunction)

In my last post, we looked at an overview of what Systemic Functional Grammar is. Halliday divides the way we use language into different metafunctions. This post will explore the Interpersonal Metafunction, and in Part 3, we shall look at the Experiential Metafunction, and in Part 4, we will be examining the Textual Metafunctions and I will also be suggesting some ideas as to how we could apply SFG to our language classrooms.

This post is based on research from the following books

An Introduction to Functional Grammar

The Functional Analysis of English

Using Functional Grammar: An Explorer’s Guide

We use language to enact our personal and social relationships (the interpersonal metafunction), to construe our experience of the world and our consciousness (the experiential metafunction), and to organise discourse and create continuity and flow in
our texts (the textual metafunction) (Halliday and Matthiessen, 2004).

Although the term ‘function’ was used in Halliday’s earlier work (1976) and other books on SFG (Butt et al, 2000), the term ‘metafunction’ is now preferred, to avoid potential confusion with ‘communicative functions’ in relation to Searle’s (1965) Speech Acts, and grammatical functions of words or groups (Bloor and Bloor, 2004). Each metafunction has its own systems of choices, each choice resulting in a structure. However, realisations of these 3 metafunctions occur simultaneously, allowing language to create different meanings at the same time (Eggins, 2004).

          I  must      finish      this blogpost
   Subject  Finite Predicator Complement
InterpersonalMetafunction   Mood   Residue
 ExperientialMetafunction   Actor  Process:material Goal
 Textual Metafunction   Theme                   Rheme
  Given                   …………………..  New

The Interpersonal Metafunction

Language involves interactions where we initiate or respond to the act of giving or demanding for goods-and-services or information. Thus, Halliday and Mathiessen (2004) regard this function as one of exchange. The principle grammatical system here is the MOOD network, within which is a choice between imperative and indicative. If indicative is chosen, there is a choice between declarative and interrogative. These choices are
realised by manipulating the Mood element.

1. Mood

The Mood carries the interpersonal functions of the clause and consists of Subject+Finite. The Subject is realised by a nominal group that the speaker gives responsibility to for the validity of the clause (ibid), while the Finite is realised by the first of the verbal group. The rest of the verbal group is the Predicator, which forms part of the Residue. A clause thus
consists of Mood+Residue. The Mood element can be identified in Mood tags (pedagogically, question tags).

Josh can speak English.
Subject Finite Predicator Complement
Mood Residue
Josh can speak English, can’t he?
Subject Finite Predicator Complement Finite Subject
 Mood    Residue   Mood Tag

and is also used in short answers, the Finite being the core that is bandied about in exchanges because it carries the validity of the proposition (Thompson, 2004).

Notice how the finite is used to argue the validity of the proposition in this childish exchange:

A:      You                                didn’t                           read                                my blogpost!





    Mood    Residue

B:            Yes,                                   I                                       did.

Mood Adjunct Subject Finite

A:          No,                                 you                               didn’t!

Mood Adjunct Subject Finite

B:              Did!


A:              Didn’t!



The giving of goods-and-services is labelled an offer, usually realised by Finite^Subject signalling an interrogative, but can also be non-linguistic (I present you biscuits). A command demanding goods-and-services takes the imperative, where the Mood is non-existent, although the assumed Subject ‘you’ appears in a marked imperative (see below).  Goods-and-services are tangible commodities or activities, and responses to proposals (offers and commands) can be non-linguistic and limited to either accepting or
refusing.  Language merely facilitates the success of the exchange.

An offer realised as an interrogative

Would you like some biscuits?
Finite Subject Predicator Complement
Mood Residue

 A command realised as an unmarked imperative

    Pass the biscuit.
No Subject No Finite Predicator Complement
No Mood Residue

A command realised as a marked imperative

You,   pass the biscuit!
 Subject No Finite Predicator Complement
No Mood Residue

 A statement realised as a declarative

I           made   those biscuits.
Subject ‘past’Finite ‘make’Predicator Complement
         Mood            Residue

A question realised as an interrogative

Did you make those biscuits?
Finite Subject Predicator Complement
Mood Residue

The exchange of information involves an intangible, verbal commodity and language is the end in itself. The giving of information often takes the form of a statement, a declarative denoted by Subject^Finite. The demanding of information is expressed by a question realised by an interrogative. Statements and questions (propositions) can be argued with, denied, adjusted, etc., and the response is varied and has to be linguistic, unlike proposals. The position and existence of both Subject and Finite therefore indicates whether a clause is declarative (statement), interrogative (question, offer) or imperative (command) (see above examples with ‘biscuit’).

However, declaratives andinterrogatives could also be polite requests for goods-and-services since basic commands might be considered Face Threatening Acts, and thus highly impolite (Brown and Levinson, 1987). Modals are also often used to disguise demanding proposals or soften propositions (Bloor and Bloor, 2004), but it is important that EFL students initially learn the most straightforward grammatical realisations of the
interpersonal metafunction, before shifting towards increasing interpersonal distance through less straightforward structures (Butt et al, 2000).

In ‘I made those biscuits’, the Finite appears to be missing, but is in fact fused with the Predicator ‘make’ (made = Finite: ‘did’ + ‘Predicator: ‘make’). This could help EFL beginners understand why the so-called ‘dummy’ auxiliary ‘do/does’ magically appears in some interrogatives and negatives, while not in others that contain a separate Finite. Through the Finite, the speaker can signal the primary tense, polarity (positive or negative) and modality (the extent of validity) of the clause, seen from his/her standpoint. Teachers can help students anchor such viewpoints within the Mood.

Included in the Mood is the word ‘not’, attached to the Finite to signal negative polarity. However, according to Halliday and Matthiessen (2004), this is not always the case. Consider the two possible meanings in ‘You may not go to the party.’

If we take that sentence to mean ‘you are not allowed to go’, ‘not’ would be included as part of Mood. In the second possible meaning ‘you are allowed not to go’, ‘not’ is part of the Residue. This is a useful distinction, but interestingly, if we followed the above logic, then the ‘not’ in ‘you must not go’ ought to belong to the Residue. Halliday (ibid)
lists the above ‘not’ as part of the Finite because of the existence of the contraction ‘mustn’t’. Arguably, recognising ‘not’ as the Residue in this case might be helpful to EFL students, who are often confounded with the difference between ‘You don’t have to go’ (‘not’ in Mood, therefore ‘don’t’ negates the validity of the Residue ‘have to go’)
and ‘You mustn’t go’ (in my opinion, the Finite ‘must’ validating the Residue ‘not go’).

2.    Modality

With propositions, the positive and negative poles in the Mood assert or deny what is stated in the Residue, e.g. ‘It is,’ or ‘It isn’t’. In between these certainties are degrees of probability and usuality signalled by modalization (ibid).
Modalization is expressed through the Finite showing epistemic modality (O’Halloran, 2006) e.g. ‘It might be’, or through a Modal Adjunct like ‘It probably/usually is’. Modal adjuncts are included in the Mood, and can be categorised into

(i) Mood adjuncts, e.g. ‘probably’, which occur close to the finite,

(ii) Comment adjuncts, e.g. ‘unfortunately’, which occur at any
boundary between information units.

With proposals, the positive and negative poles prescribe or proscribe e.g. ‘Do,’ or ‘Don’t’, and modulation (ibid) happens in between, showing degrees of obligation e.g. ‘You should’ for commands, and degrees of inclination e.g. ‘I should’ for offers. We can use Finites to show deontic modality (O’Halloran, 2006), or expansions of the Predicator by passive verbs like ‘I’m supposed to’, or adjectives ‘I’m anxious to’. As the assumed Subject in proposals is ‘you’, when modulated clauses implicate a third person, e.g. ‘He should know’, the proposal becomes a proposition (Halliday and Matthiessen, 2004). The distinction between proposition and proposal becomes blurred and the distinction between modalization and modulation becomes context-dependent.

Bibliography : See SFG (Part 1)

What is Systemic Functional Grammar? (Part 1)

Systemic Functional Grammar (Part 2 – The Interpersonal Metafunction)Several people I know have expressed an interest in finding out more about Systemic Functional Linguistics. Some have attempted to read Michael Halliday’s Introduction to Functional Grammar and have ‘feedbacked’ to me about it being dense and not the most accessible…So here is a summary of my research on the subject.

This post is based on research from the following books

An Introduction to Functional Grammar

The Functional Analysis of English

Using Functional Grammar: An Explorer’s Guide

I’ve found it almost impossible to do this in one blog, so I’ll be dividing my summary into 5 parts. This first part is mainly an introduction into the subject, and in Part 2, I will look at the Interpersonal Metafunction, in Part 3, the Experiential Metafunction, and in Part 4, the Textual Metafunction. The final part would also contain some applications of SFG to EFL teaching (although I’ll try to pepper the next three partts with as many EFL perspectives as possible) and some of my conclusions.

I hope I will be able to do SFG some justice with my limited knowledge and research. If you are an EFG expert, feel free to correct me or add on to what I have written. If you are an EFL professional, I’d very much appreciate it if you feel free to share your thoughts on how you use or could use SFG in your teaching.

1.    What is Systemic Functional Grammar (SFG)?

Language is used to express meanings and perform various functions in different contexts and situations of our daily lives. If grammar is ‘the way in which a language is organised’ (Butt et al, 2000), SFG attempts to explain and describe the organisation of the ‘meaning-making resources’ (Halliday and Matthiessen, 2004) we use to achieve such goals. Every linguistic choice we make is systematic, and the reason we say something in a certain way is the result of a choice, albeit unconscious. Such choices are made from a set of systems containing structures, allowing us unlimited ways of creating meaning (Bloor
and Bloor, 2004), while our experiences of the world, of text types and socially- and culturally-bounded situations, help build up our schemata of these systems. Van Djik’s (1977, in Brown and Yule 1983) Assumed Normality of the World could serve to explain how these experiences enable us to distinguish between different genres of texts (spoken or written) by their patterns of linguistic choices, and to notice when choices are inappropriate. SFG is, thus, a study of meaning construction through systems of lexicogrammatical choices that serve functions within social and cultural contexts.

To understand SFG, it is important to look at other theories of language and how they compare, after which I will briefly outline Halliday’s three basic metafunctions and how the systems of choices within each component are realised in the lexico-grammatical structures within clauses and how this could benefit teachers and students of English as a
Foreign Language. Due to space constraints, I will not be looking at theories on group- and clause-complexing, cohesion, grammatical metaphors, appraisals, or theories on genre and register.

2.       Views on Language and Language Acquisition

Often referred to as the father of modern linguistics, De Saussure (1916, in Coffin et al, 2004)  made a distinction between  paradigmatic and syntagmatic relations in
language, which closely corresponds to Halliday’s system and structure. The
paradigmatic phenomenon is one of choice.
For example, in my last sentence, I chose to make it declarative instead of interrogative, and positive instead of negative. The syntagmatic dimension refers to prescribed sequences in which elements (phonemes, morphemes, words, groups) are combined e.g. Subject^Finite^Predicator (^ means ‘followed by’). In SFG, this chain of elements is a result of paradigmatic choice (Bloor and Bloor, 2004), structures being realisations of systems.

Alongside paradigmatic and syntagmatic distinctions, De Saussure distinguished between two aspects of language: langue, the language system, and parole, the use of this system (ibid). Chomsky later developed this view, focusing on the language system (competence) and not the use of language (performance), and strived for a ‘Minimalist Programme’ in order to formulate a limited set of principles, i.e. transformational-generative rules, which govern the formation of an unlimited number of grammatically correct sentences (ibid). Although he does not deny that meaning in language is a social phenomenon, Chomsky regards grammar as autonomous from meaning. He further asserts that such transformative-generative capacity is ‘hard-wired’ into the human brian, thus explaining the universal phenomenon of language acquisition and the biological determination of how we use language (ibid). Chomsky was interested in a model of such mental grammars,
where the basic use of language is for thought and not communication (Thompson, 2004).

Halliday, conversely, took Malinowski’s view that ontogeny recapitulates phylogeny, i.e. by explaining natural adult language and language acquisition in terms of social functions, one can understand the evolutionary origins of language (Halliday, 1976). Like Malinowski and Firth, Halliday saw language acquisition as an interplay between nature and nurture (Bloor and Bloor, 2004), suggesting that the language acquisition process of a child and the linguistic structures he/she masters reflect the functions required to serve his/her life. Learning a language was thus ‘learning how to mean’ (Halliday, 1975). Unlike Chomsky, Halliday did not believe in a finite system of rules, and preferred a descriptive approach of examining sentences as being appropriate or inappropriate to the prescriptive approach of
labelling them ‘correct’ or ‘incorrect’ (Eggins, 2004). In the tradition of Malinowski, Whorf and Firth, he believed language is moulded by culture, and the world is seen through the language we speak (Kress, 1976). Meanings are determined by the texts’ relationship with the context of culture (genre) and the context of situation (register) (Eggins, 2004), and the study of sentences should, therefore, be inseparable from its social, cultural and
situational contexts, and not done in isolation. Hence, the use of authentic texts and corpus data has become the norm in the study of Systemic Functional Linguistics. SFG has been called an ‘extravagant’ grammar (Bloor and Bloor, 2004) as it is a huge area that attempts a multi-dimensional description of how language is organised, and its linguistic viewpoints are best seen through the main issues underlying Halliday’s three metafunctions of language.


Bloor, T., and M. Bloor. (2004) The Functional Analysis of English: A Hallidayan Approach, (2nd Edition), London: Arnold

Brown, P., and S. Levinson. (1987) Politeness: some universals in language use, Cambridge: Cambridge University Press.

Butt, D., R. Fahey, S. Feez, S. Spinks, C. Yallop. (2000) Using Functional Grammar: An Explorer’s Guide, (2nd Edition), Sydney: Macquarie University.

Coffin, C. (ed.) (2006) English Grammar in Context Book 3: Getting Practical – Evaluating everyday texts (2nd Edition), Milton Keynes: The Open University.

Eggins, S. (2004) An Introduction to Systemic Functional Linguistics (2nd Edition),
London, New York: Continuum.

Halliday, M. (1975) Learning How to Mean: Explorations in the Development of Language, London: Edward Arnold.

Halliday, M. (1976) ‘The form of a functional grammar’, in Kress, G. (ed.):7-25.

Halliday, M., and C. Matthiessen. (2004) An Introduction to Functional Grammar (3rd
Edition), London: Arnold.

Kress, G. (ed.) (1976) Halliday: System and Function in Language, Oxford: Oxford University Press.

Thompson, G. (2004) Introducting Functional Grammar (2nd Edition), London: Hodder Education

%d bloggers like this: