This  message was  originally submitted  by Edith Wollin
([log in to unmask])  to the ATEG list:
 
 
I don't see anything in Leuschner's response that helps explain why the
whom in the first sentence cannot be omitted while the one in the second
sentence can.  Whereas, if we say that if it functions as a noun, it is a
noun, we get an explanation.
 
 
> This message  was originally submitted  by
> [log in to unmask] to the ATEG list
>
> >"I don't know who(m) I saw last night".
> >"I like the girl who(m) i saw last night.
> >
> >ok, the above two, to my feeble understanding, both contain relative
> >clauses which are,
> >1-defining
> >2-and both have the relative pronoun as object
>
> The thing is not to mix up function and form. 'Relative clause' is more
on
> the function side, so we'd better not call the clause in the first
sentence
> a 'relative clause' (I know that some grammars do). To call it 'noun
> clause' (as other grammars do), is not very logical either. The term
refers
> neither to the form of the clause nor to its function. It refers to the
> fact that occasionally a noun can be found in the same place: 'I don't
know
> JACK.' Not very helpful, and it has confused me through all my student
> life.
>
> Why not apply regular logical thinking to language. It makes things so
much
> easier.
>
> Looking at texts we find that there are a great number of different
> sentence types.
>
> E.g.
>      Many Americans can speak Spanish. This is the most basic type,
usually
> called statement.
>      Can  many Americans speak Spanish?  Usually called question.
>      Many Americans can not speak Spanish. Negative sentence.
>      What many Americans can speak. This sentence type usually has no
> single name, but a different name in each function.
>
> I suggest the term 'wh-sentence'. A term which refers to the form. A
> wh-sentence is one where one (or more) functional slots are filled by a
> wh-pronoun, which goes to the front of the sentence (many Americans can
> speak what --> what many Americans can speak).
>
> Wh-sentences come in statement form (What many Americans can speak.) or
in
> question form (What can many Americans speak?)
>
> Wh-statements can be used widely, in many different functions.
>
> They can be used on their own, e.g. in headlines (How Lincoln grew a
> beard).
>
> They can fill the last slot in a paragraph (a paragraph about dogs, say,
is
> summarized by 'Which is why I hate dogs.')
>
> They can be put in any sentence slot, i.e. they can function as subjects,
> subject complements, objects, object complements, adverbials - did I
forget
> a function? In 'What you see  IS  what you get' the first wh-sentence
> functions as subject, the second as subject complement. In 'I don't know
> ...' above it is the object slot which the who-sentence fills.
>
> They can fill a functional slot in sentence parts - that is they can be
> noun attributes, in this case they are traditionally called '(defining)
> relative clauses'.
>
> They can function as appositives (My wife, who comes from H., is asleep
at
> the moment). In this function they are traditionally called 'non-defining
> relative clauses'.
>
> There are still other functions in which wh-statements can be used. But
> this should do for illustration.
>
> To distinguish clearly between form and function and use appropriate
> terminology makes grammar fascinating, even to students, while the
> traditional absurdly illogical chaos makes it a hated subject. Which
> certainly does not do justice to the importance that language has for us
> humans.
>
> B. Leuschner
> --------------------------------------------------------------------
> Burkhard Leuschner - Paedagogische Hochschule Schwaebisch Gmuend, Germany
> E-mail: [log in to unmask]    [h]    Fax: +49 7383 2212
>