Mick O'Donnell On The Stifling Of Work In SFL

Mick O'Donnell replied to David Rose on sys-func on 14/7/22 at 9:16:

… But in moving forward, we cannot apriori know which aspects of the established model are ones that need to change.

We cannot assume SFL as developed by MAKH or by Ruqaiya Hasan or by Jim Martin is God's Truth, immutable for all time. History has shown that models are always superseded as we apply the model to reality.

To me, too much work in SFL has been stifled by the need to stay within the bounds of the establishment. We have been trying to grow the theory outwards, or in complexity, without touching the Hallidayan core.

Jim Martin has always impressed me as someone who, while respecting the whole, is willing to change core assumptions, when the needs of linguistic modeling require it. In the late 1970s, trying to handle genres within the established model, and failing, he and others broke out functional tenor as a stratum above, which I for the most part buy as necessary. But whether you accept this modification or not is not important, what is important is that Jim has been (and continues to be) willing to throw out established ideas (even his own) if they don't fit new data. And anyone who is not willing to do similarly is kidding themselves if they think they are doing linguistic science.

Blogger Comments:

[1] This is misleading, because it is untrue. The internal self-consistency of a model can be assessed purely by deduction, without reference to empirical data, and any model that is internally inconsistent, such as Martin's (evidence here)or Fawcett's (evidence here), is one that needs to change.

[2] On the one hand, this is deployment of the 'straw man' logical fallacy, since it postulates an absurd position as the one that is to be invalidated. On the other hand, this is O'Donnell once again invoking a religious stance that he attributes to others.

[3] To be clear, from the perspective of SFL Theory, models are semiotic systems, and as such, evolve as the environment in which they function changes, with semiotic systems being one source of that environmental change.

[4] To be clear, SFL Theory was devised by Halliday, and everyone else is using his theory. What gives the "Hallidayan core" its value, is, inter alia, its internal consistency as a theory, and its explanatory potential in modelling language. It "stifles" work in SFL in just the same way that Quantum Theory and General Relativity "stifle" work in physics.

[5] This is misleading, because it is untrue. Ignoring the motivation of personal ambition, Martin's models do not arise because they better fit the data, but because Martin misunderstands Halliday's model; evidence here.

[6] To be clear, this would be true, if this was, in fact, what Martin has actually done.