For the most part I second Bob’s remarks.  TG, through no fault of its own, has gotten blamed both the decline in teaching diagramming and in teaching grammar in general, and much of this blame is associated with the truly awful Paul Roberts series, which was a combination of poor TG and poor pedagogy.  The decline in the teaching of grammar is a social and philosophical phenomenon well laid out in David Mulroy’s The War against Grammar and has been the result of social forces independent of any particular theory of grammar.  Bob’s precisely on point when he argues that any approach to grammar that doesn’t make explicit what native speakers already know implicitly fails pedagogically.  Bob and I probably differ on the amount of formalism we include in our grammar courses, but probably much less in the sort of content we cover.

 

I make my plea once again for a careful traditional grammar of the sort presented in the major reference grammars like Jespersen, Quirk et al., Greenbaum, Biber et al., and Huddleston&Pullum.  These are demanding works to read, but they describe the language well and make the necessary distinctions both between form and function and within form and function.  One of the goals of a college grammar course ought to be to orient students to one of more of these works so that they have a sense of what a reference grammar is and how to use one, and so that they may even have one on their shelves as an antidote to the watered-down, often inaccurate grammar that gets presented in too many course texts.

 

 

It is tiresome to read about the shortcomings of a transformational-generative grammar approach without understanding why it remains an important perspective for many who study the nature of language as their academic discipline.  Here is an example of  what I mean.

Diagramming is especially useful for visual learners.  If it fell

out of favor, I suppose that was due to the influence of

transformational-generative grammar, but TG grammar is especially focused

on "deep" structures, not on the surface structures that get

diagrammed. 

Consider two famous sentences that are used to argue for the importance of an abstract representation for purposes of interpretation.

1) John is easy to please.
2) John is eager to please.

In (1), John is the one being pleased, but in (2) John is the one doing the pleasing. 

My understanding is that a concern for SURFACE STRUCTURE would not be able to show this.  Conventional diagramming (and I may be wrong on this) would diagram (1) and (2) the same.  

Conventional diagramming could show this difference, but that would require the addition of some "abstract" form to capture that John is the underlying object of please in (1) but is the underlying subject of please in (2).  Such an addition would destroy the "advantage" of diagramming as showing "surface" structures.

This example is important for any teacher of English (L1 or L2) to understand.  There are any number of examples of L2 learners interpreting (1) as John doing the pleasing (in other words, just like (2)), and Carol Chomsky has shown that many L1 learners of English until about the age of eight also interpret (1)  like (2). 

 There is room for both approaches. 

This may be the case, but I don't understand the need for an approach which does not accurately describe the knowledge that all competent speakers of the language know.

Bob Yates, Central Missouri State University