Disclaimer: The post below refers the implementation of SFL in NT studies, which may or may not necessarily reflect Halliday’s intentions. My goal is to address the methodological and theoretical flaws of the approach to SFL popularized by Stanley Porter. There is reason to believe that it represents a departure from Hallidayan linguistics (10/2/09).
This is my last post on Systemic Functional Linguistics (SFL). If I had a nickle for every time I was asked why SFL was not my preferred method, I could spend much more time salmon fishing. This series of posts is my answer to that question. For the kind of work that I am doing, SFL is ill-suited in its present form. Here are links to the other posts in the series.
The following is an excerpt from section 9.4 of my forthcoming Discourse Grammar of the Greek New Testament. I included this discussion primarily because of the inordinate amount of attention that SFL has received in NT studies. One would think this method had all but taken over the broader field of linguistics, based on its prevalence in Koine. In reality, there is some very profitable reworking of the theory that has been going on, as I understand it. Many of the shortcomings that I discuss below are being or have been addressed by scholars from the Cardiff school of SFL, lead by Dr. Robin Fawcett. So the comments that I make below are addressed to the implementation of the classic Hallidayan-SFL syntactic framework for the study of Koine.
In my grammar, I apply a cognitive-functional framework to analyze word order within the clause in Koine, based on the work of Simon Dik and Knud Lambrecht. Generally speaking, a clause is made up of both old/established information and new/non-established information. The latter is the communicative point of the clause, its reason for being. The former provides the grounding and framing within which the new information is processed and integrated. The two work together. Both kinds of information can be placed in marked positions in Koine to accomplish specific pragmatic purposes. The fronting of new information results in emphasis, taking what was already the most important part of the clause and attracting extra attention to it. Fronting the old/established information is typically referred to “topicalization,” whereby attention is drawn to changes in time, place, participants, etc. They create an explicit frame of reference for the clause that follows. For more on this, see the section on information structure in the glossary.
One more piece of background. English uses intonation or “prosody” to mark the pragmatic role of information within a clause. The new information receives primary intonation, the established information receives little to no intonation within the clause. The exception to this is topicalized “established information”–it receives secondary intonation to mark it is such.
Because of the rigid constraints on the ordering of clauses in English, there is little flexibility to move clause components. The virtual absense of inflection means that pragmatic information about information status is almost entirely communicated via intonation. There are things like cleft constructions that can mark emphasis, but they are limited in number and distribution.
Because Koine is so highly inflected, and thus has such freedom in the ordering of clause components, pragmatic information about information status is communicated by pragmatically changing the order of the clause. The intonation of the clause probably follows the principles observed in modern spoken language, with primary intonation in a clause being placed on the new information. However, since Koine is no longer spoken by native speakers, we have no way of knowing. In claims made in this area would be highly speculative at best, with little hope of independent verification. Now for a basic overview of Halliday’s approach to information structure analysis as applied to Koine. The quote begins here.
“The most prevalent linguistic approach found in NT studies today is Systemic Functional Linguistics (SFG), originated by M. A. K. Halliday. Halliday’s account of information structure derives also from the Prague school, yet with some significant differences. Halliday’s SFG sub-divides the Prague School notion of given-new/theme-rheme into new sub categories, rather than using them as synonymous terms. The given-new distinction he assigns to the ‘tone group’, which does not correlate to any specific grammatical unit, like a clause or sentence. In other words, given/new correspond to the primary device that English uses for information structuring: intonation or prosody. Theme and rheme are viewed as corresponding to constituents of the clause, i.e., to the syntax instead of prosody. This differentiation into tone units versus clause units leads to significant differences in analysis between SFG compared to FG [Simon Dik’s Functional Grammar], RRG [Foley and Van Valin’s Role and Reference Grammar], or the other functional approaches discussed above.
Most of Halliday’s theoretical work to develop his framework was conducted on English, and its application has focused on rigidly-ordered languages. Halliday’s definition of ‘focus’ is inextricably tied to stress and intonation, just as in English. His theory is well suited for English since prosody is the primary means of marking focus in rigidly ordered languages. However, SFL has received criticism from outside the SFL school—as well as from within—regarding its documented deficiencies in the area of syntax. Robin Fawcett’s A Theory of Syntax for Systemic Functional Linguistics (Philadelphia, Penn.: John Benjamins, 2000) has been praised for addressing the rather flagrant flaws in Halliday’s conceptions of syntax. It has also resulted in something of a split between the ‘Cardiff school’ of SFL lead by Fawcett, and Halliday’s ‘Sydney school’, resulting from Halliday’s unwillingness to address the criticisms lodged against his theory.
SFG [Systemic Function Grammar] postulates that the initial element in a clause, be it a conjunction, subject, or fronted focal constituent, is always the ‘theme’. This idea works fairly well in a rigidly configurational language like English, but proves inadequate for non-configurational languages like Greek and Hebrew. This pitfall is a natural consequence of formulating the theory almost exclusively using English-like languages. The choice to link theme-rheme only to intonation and not to syntax in any way has significant consequences when applied more broadly to flexibly-ordered languages.
I have little concern about the broader development of SFL in linguistics proper, though I hope to read Fawcett’s Syntax at some point to see how things turned out. My primary concern is to address the application of SFL in the specific field of Greek grammar. SFL as a method has much to offer for studying things like coherence and cohesion. However, its failure to build upon a typologically-informed foundation has resulted in misguided claims being made in its application of such a non-configurational languages like Koine Greek. In the grammar, I document a basic consensus among the major approaches to syntax regarding basic information-structuring principles. They differ primarily in why they are looking at information structure. The generative folks are seeking the underlying grammar that explains the shifts, the cognitive and relevance theory folks are looking to understand the processing and conceptualization of discourse.There is agreement on the core elements of why languages like Greek do what they do. Halliday’s claims are somewhat anamalous in that regard.
So to clarify, my problem with SFL is primarily with its application to areas it is ill-suited to handle, not with the method itself. My area of interest is discourse pragmatics, not just syntax, verbal aspect or semantics. Those implementing SFL need to critically examine its presuppositional elements and address the fundamental problems that have undermined its application to things like Greek syntax.

Systemic functional linguistics and word order-SFL IV