Johanna Rubba wrote:

> This makes it sound like the only people we are teaching to when we teach
> grammar is speakers of nonstandard dialects. I don't think Bob intends
> this to be inferred.

That is a most ungenerous reading of what I wrote.  I am assuming that
we teach grammar to all native speakers of English to help them make
conscious decisions about what is appropriate, defining appropriate as
being standard or textually felicitous.

> And I do see a connection to text structure. I've claimed before that
> choices about sentence structure are governed by the need to manage
> information flow in texts.

This statement is right.  However, the central question is whether
knowledge of
argument structure is necessary to resolve issues of information flow in
a text.
>
> Where different phrases get placed, and which grammatical forms get chosen
> to code them, can also vary depending on text-level needs. Yet the
> thematic role of a constituent doesn't change no matter how it's 'clothed'
> or placed.

So, how does argument structure help with this?
>
> Karen broke the window with a hammer.
> A hammer was used to break the window.
> A hammer broke the window.
> The window was broken with a hammer.

Notice you can also have: The window broke.
>
> In all of these sentences, 'a hammer' retains its instrument thematic
> role, although it is 'packaged' into different syntactic roles: object of
> preposition, subject of passive clause, subject of active clause. It may
> seem like the subject of the active clause is agent, but remember that
> thematic roles depend on real-world status (or our conceptualization of
> same), not syntactic status in a given sentence. Languages differ in what
> kind of thematic roles can become subjects of passive sentences, and some
> languages will not allow an instrument to be coded as subject of a
> passive.

Right.  And, in English we are talking about verbs which allow this.

Karen painted the house with a sprayer.
The house was painted with a sprayer.
A sprayer was used to paint the house.
? A sprayer painted the house.  (where sprayer is a tool and not a
person)
*The house painted.

I know of NO dialect of English which would immediately allow the last
two sentences. All of these facts are interesting and raise really
interesting questions about our knowledge of language, but . . . do
public school teachers, who see themselves teaching native speakers,
need to know the difference between paint and break?

Do I need to know about thematic roles to make decisions about
information flow?  I don't have Kolln's Rhetorical Grammar in front of
me, but I don't remember a chapter which deals with this topic.  Do
native speaker have difficulty in information flow in a text because
they are not aware which thematic roles can be moved around with or
without passive morphology?  I don't think so.

All of this is interesting to dispel the notion that the "subject" of a
sentence/clause is a "doer."  Ok, so we don't want to define the
grammatical subject of a sentence as a "doer."

> > On the other hand, thematic roles are a problem for non-native
> > speakers.

> This is true, and it makes thematic roles more immediately useful in
> teaching non-native speakers. But that's not our primary concern in ATEG,
> as we have been reminded in the past.

My entire post was responding to this question:

> > Are thematic roles of any use in grammar teaching?

Excuse me, but the ONLY way to answer that question is to ask who the
students are.  I can think of any number grammatical structures I have
to be prepared to teach to non-native speakers that I never have to
teach to native speakers.  For example, native speakers never have to be
taught about how the article system in English works.  The entire
distinction of count/non-count nouns is never something native speakers
have to be taught. etc.  These distinctions have important implications
for certain grammatical decisions that are influenced by information
flow.   I don't know of any study which reveals that native speakers
have any difficult with these structures.

So, is knowledge about the article system in English or the
count/non-count distinction of any use in grammar teaching?  It depends
on the students.

Bob Yates