Google Books Ngram Viewer: Functional Vs Formal Grammar

Projection Vs Realisation

Projection is a logical relation between the material order of experience and the semiotic order of experience — between a macrophenomenon (a senser sensing or a sayer saying) and a metaphenomenon (an idea or locution).

Realisation is both a logical relation and an experiential relation.  

As a logical relation, it is elaboration, a type of expansion.  

As an experiential relation, it is one of symbolic identity, a type of identifying relation.  

Realisation is an ideational relation between different levels of symbolic abstraction.

"Reading Images"

		Token	Process	Value
consciousness (Symboliser symbolic processing)	projects	meaning x construed by language (Value)	means	y
		pictorial re-presentation of visual experience (Token)

Why Stratify The Content Plane?

Halliday & Matthiessen (1999: 237):

Of course, what we are recognising here as two distinct constructions, the semantic and the grammatical, never had or could have had any existence the one prior to the other; they are our analytic representation of the overall semioticising of experience — how experience is construed into meaning. If the congruent form had been the only form of construal, we would probably not have needed to think of semantics and grammar as two separate strata: they would be merely two facets of the content plane, interpreted on the one hand as function and on the other as form.

Cognitively Projected Propositions: Reported Interrogatives

Halliday & Matthiessen (2004: 450-1):

In the environment of ‘mental’ projection, the contrast between statement and question is not concerned with the speech functional orientation of giving vs demanding information but rather with the status of the validity of the information.  In a statement, it is fixed with respect to the polarity and the elements of transitivity (realised by an indirect declarative clause optionally introduced by that), but in a question, it is open with respect [to] the polarity (realised by an indirect yes/no interrogative clause introduced by whether or if) or one (or more) of the elements of transitivity (realised by an indirect wh- interrogative clause introduced by who, which, when, where, etc).  Consequently, mental clauses representing an ‘undecided’ state of mind are used to project indirect questions.  These include clauses of wondering and doubting, finding out and checking, and contemplating, which tend to be characterised by special lexical verbs such as wonder, ascertain …

Halliday & Matthiessen (2004: 463):

If the reported clause is interrogative it typically shifts into the declarative; the declarative is the unmarked mood, and is used in all clauses that do not select for mood independently, including all dependent clauses. A yes/no interrogative becomes declarative, introduced by if or whether; a WH- interrogative becomes declarative with the WH- element remaining at the front.

How To Improve A Theory

Step 1: First make sure you understand the theory in its own terms.

Martin’s Cline Of Instantiation Applied To Martin’s Stratification

Martin's strata: genre, register, discourse semantics, lexicogrammar, phonology

Martin's cline of instantiation: system, genre/register, text type, text, reading.

On this model,

(1) register is simultaneously more abstract than discourse semantics (in terms of stratification) and a sub-system of discourse semantics (in terms of the latter stratum's instantiation);

(2) genre is simultaneously more abstract than register (in terms of stratification), and a sub-system of register (in terms of the latter stratum's instantiation);

(3) register is simultaneously a stratal system and a sub-system of itself (in terms of that stratum's instantiation);

(4) genre is simultaneously a stratal system and a sub-system of itself (in terms of that stratum's instantiation);

etc …

cf Halliday's original (logically coherent) model:

Instantiation / Stratification Matrix

Locating Genre in The SFL Model

Because Martin's model of genre as a level of context isn't theorised according to SFL principles — see for example arguments here here and/or here — components of the model are scattered across the architecture of SFL theory. For example:

1. to the extent that genre refers to text type, it is a point of variation on the cline of instantiation: specifically, it is register viewed from the instance pole;
2. to the extent that genre refers to the rôle of language in context, it is defined by features of the system of MODE, and so by the textual metafunction at the level of context;
3. to the extent that genre refers to the sequence in which ideational and interpersonal meanings are organised, it is within the textual metafunction at the level of semantics.

The Difference Between Material Setting, Context And Cotext

When we construe a person speaking, we construe a material order and a semiotic order of experience.

• The material order includes the person speaking and their physical circumstances; while
• the semiotic order includes the projected content of what is said: the wording that realises meaning. 

So

1. the term material setting refers to the material order of a person speaking;
2. the term context refers to the (instance of) culture being realised in the content of what is being said — i.e. the context is of the semiotic order: it is the culture construed as a semiotic system (field, tenor and mode) that is realised in language;
3. the term cotext refers to the rest of the language that accompanies an identified portion of what is being said.

It is important to understand that the term context does not refer to either the material setting or the (semiotic) cotext.

The misunderstanding that context refers to the cotext and the material setting can be found in Martin (1992: 33), as recorded here.

Semiotic Context And Material Setting

Context is of the semiotic order of experience and realised by the content that is projected from the material order through a 'sayer saying' or a 'senser sensing'. The material order realisation of context is the expression plane of language.

The material setting is of the same order of experience as the 'sayer saying' and the expression plane. It does not materially realise the semiotic context — the expression plane does.

Embedded (Defining) Vs Enclosed (Non-Defining) Relative Clauses Exemplified

(1) Defining, so embedded as Qualifier

Ivy’s uncle [[who is a doctor]] lives nearby (whereas her other uncle lives miles away).

(2) Non-Defining, so ranking and hypotactically elaborating:

Ivy’s uncle <<who is a doctor >> lives nearby (so let’s phone him).

Defining Vs Non-Defining Relative Clauses

Consider the following sentence:

The house that Jack built, which is a British nursery rhyme, is a 'cumulative' tale.

There are two relative clauses:

1. that Jack built
2. which is a British nursery rhyme

The first of these is a defining relative clause, while the second is a non-defining relative clause.

On the SFL model, the defining relative clause is embedded as Qualifier in a nominal group, while a non-defining relative clause is a ranking clause within a clause complex.  In the example above, the non-defining relative interrupts the clause on which it is dependent, and is termed 'enclosed'.

The distinction between the clauses is indicated by the following notation:

The house [[that Jack built]] <<which is a British nursery rhyme>> is a 'cumulative' tale.

Let's take a trinocular perspective on these types of relative clauses.

The view from above (the meaning being realised by the wording):

In terms of the logical semantic relations between the relative clause and the item it modifies, both types involve hypotactic elaboration.

The relation is hypotactic because each relative clause is dependent on the item that it modifies:

• [[that Jack built]] is dependent on the house
• <<which is a British nursery rhyme>> is dependent on the house [[that Jack built]]

The relation is elaborating because each relative clause expands the item it modifies by further specifying or describing it:

• [[that Jack built]] further specifies/describes the house
• <<which is a British nursery rhyme>> further specifies/describes the house [[that Jack built]]

Hypotactic elaboration is what the two types of relative clause have in common.  But they also significantly differ in meaning.  A defining relative clause specifies a particular subset of a general class, whereas a non-defining relative clause 'adds a further characterisation of something that is taken to be already fully specific' (Halliday & Matthiessen 2004: 400).

The view from below (how the wording is realised):

In terms of intonational realisation, the two types of relative clause differ markedly.

In the unmarked case — unless there are textual pressures for doing otherwise — the defining relative clause is realised within the same tone group realising the nominal group in which it functions as embedded Qualifier.  That is, both are packaged as a single unit of information.  So, in the above example:

• the house and [[that Jack built]] are realised within the same tone group.

On the other hand, in the unmarked case, non-defining relative clauses are realised by their own separate tone group.  That is, the relative clause and the item that it modifies are each packaged as  distinct units of information.  So, in the above example:

• the house [[that Jack built]] and <<which is a British nursery rhyme>> are each realised as distinct tone groups, thereby construing them as distinct units of information.

The view from roundabout (systemic relations at the level of wording):

In terms of systemic possibilities, the two type of relational clause differ markedly. This is a consequence of defining relative clauses being embedded as Qualifiers in nominal groups, and non-defining being ranking clauses in their own right.

Dependent elaborating clauses, of which non-defining relative clauses are a type, can be given thematic prominence in a clause nexus.  A non-finite agnate (see Halliday & Matthiessen 2004: 404) of the example above is the following:

• A British nursery rhyme, the house that Jack built is a 'cumulative' tale.
  

(If you don't understand how this clause is agnate with the original, see http://thoughts-that-cross-my-mind.blogspot.com.au/2013/09/michl-odonnell-on-non-defining-relative.html)

However, as embedded clauses, defining relative clauses are not given thematic prominence in a clause.  For example, we do not say:

• Jack built, the house, which is a British nursery rhyme, is a 'cumulative' tale.

Moreover, the lexicogrammatical domain which is elaborated by a non-defining relative clause may be a whole clause or any of its constituents, as Halliday & Matthiessen (2004: 400) point out, whereas the domain elaborated by a defining relative clause is restricted to a nominal group.

In summary, taking a trinocular vision demonstrates that defining and non-defining relative clauses share common logical semantic features, but differ markedly in every other respect.

As Halliday (2008: 6) has pointed out:

The boundaries of any grammatical category are likely to be fuzzy […] — such indeterminacy is a general property of the grammar.  The grammarian attempts to define each category as accurately as possible, looking at it from three different angles: its systemic environment (contrast with other term or terms in the system, and the relationship of that system to other systems); its meaning (proportionality in semantic terms), and it form.  In other words, the grammarian adopts a “trinocular” perspective on the stratal hierarchy so that every category is viewed “from round about”, “from above” and “from below”.  And since the views from these different angles often conflict, assigning instances to a particular category involves some degree of compromise, where criteria will depend on the purposes of the description.

The Complementarity Of Logogenesis And Instantiation

These are distinguished by the construal of time in which each unfolds as process. 

• Logogenesis unfolds in linear time (as construed grammatically by tense), whereas 
• instantiation unfolds as a translation from the virtual into the actual (as construed grammatically by aspect).

Reconstruing Experiential As Logical

transitive : extension ::

ergative : enhancement (cause)

Reconstruing Biological Evolution Through The Interpersonal Metafunction

variation = chance (Jacques Monod) = probability = modalisation

selection = necessity (Jacques Monod) = obligation = modulation

Reconstruing Determinism Through Metalanguage

The notion of determinism conflates two angles: one ideational, one interpersonal. 

Ideationally, determinism can be seen to involve the logical relation of cause; interpersonally, it can be seen to involve modality. 

For propositions about cause, the congruent modality is modalisation: usuality and probability. 

Statements of cause and effect are concerned with usuality and probability, as demonstrated by quantum physics and non-linear systems. 

Interpretations of determinism couched in terms of obligation (eg necessity) are therefore incongruent.

Reconstruing The Laws Of Physics Through Metalanguage

Ideationally, the 'Laws of Physics' are construals of experience. 

Interpersonally, they are propositions, not proposals; they are statements not commands. 

So, the congruent modality for them is modalisation not modulation; 

that is, usuality and probability, not obligation and inclination. 

So, contrary to the usual wording of physicists, 

the Universe does not obey the 'Laws of Physics'. 

The 'Laws of Physics' are probabilistic statements about the Universe.

Adjuncts: Circumstantial vs Modal vs Conjunctive

Halliday & Matthiessen (2004: 123):

An Adjunct is an element that has not got the potential of being Subject; that is, it cannot be elevated to the interpersonal status of modal responsibility. This means that arguments cannot be constructed around those elements that serve as Adjuncts; in experiential terms, they cannot be constructed around circumstances, but they can be constructed around participants, either actually, as Subject, or potentially, as Complement … .

Halliday & Matthiessen (2004: 132-3):

The two types of Adjuncts are also similar both in their own composition (as adverbial groups and prepositional phrases) and in how they may be differentiated from circumstantial Adjuncts. Whereas circumstantial Adjuncts fall most naturally at the end of the clause, where they carry the unmarked tonic (intonational) prominence, modal and conjunctive Adjuncts occur finally only as Afterthought and can never carry the only tonic prominence in the clause. … And while they all can occur thematically, only the circumstantial Adjuncts can normally occur as predicated Theme …

Halliday & Matthiessen (2004: 133):

What is common to the modal and conjunctive Adjuncts, as distinct from the circumstantials, is that they are both constructing a context for the clause. Thus even though the same semantic feature may be involved, for example time, it has a different significance in each case. A modal Adjunct of time, such as just, yet, already, relates closely to the primary tense, which is the ‘shared time’ of speaker and listener; a conjunctive Adjunct of time, such as next, meanwhile, locates the clause in time with respect to the preceding textual environment; and both are different from time as circumstance, such as in the afternoon. And the same item may function sometimes circumstantially and sometimes conjunctively; for example, then, at that moment, later on, again.

Why Intensive Attributes Are Complements

Halliday & Matthiessen (2004: 122-3):

A Complement is an element within the Residue that has the potential of being Subject but is not; in other words, it is an element that has the potential for being given the interpersonally elevated status of modal responsibility — something that can be the nub of the argument. It is typically realised by a nominal group. … Any nominal group not functioning as Subject will be a Complement; and this includes nominal groups of one type which could not function as Subject as they stand, namely those with adjective as Head … There is an explanation of this ‘from above’ in terms of function in transitivity: nominal groups with adjective as Head can function in the clause only as Attributes, and the Attribute cannot be mapped onto the interpersonal rôle of Subject. This is because only participants in the clause can take modal responsibility, and the Attribute is only marginally, if at all, a participant.

That is, the reason that a Complement that is an Attribute cannot be raised to Subject is experiential, not interpersonal.

Reconstruing Platonic Forms Through Metalanguage

Viewed through the lens of SFL theory, (non-material abstract) Platonic Forms (or ideas) correspond to the system pole on the cline of instantiation. 

That is, the Forms are the (categorial) potential of which perceived material phenomena are actual instances; the Forms are the types of which perceived material phenomena are the tokens.

Reconstruing Cosmogenesis Through Metalanguage

The grammar construes material processes as 

• either creative or transformative, and 
• either doing (transitive) or happening (intransitive).

Models of the creation of the universe construe it as either a doing or happening.

Mythological models construe the creation as 'other-engendered': a doing in which the universe is the Goal/Medium of the process and a supernatural being is the Actor/Agent.

God	created	the Universe
Actor	Process: material: creative	Goal
Agent		Medium

Scientific models construe the creation as 'self-engendered': a happening in which the universe is the Actor/Medium of a process that has neither Goal nor Agent.

the Universe	emerged
Actor	Process: material: creative
Medium