This kind of grammar theead usually ends in tears.
And, no doubt, we are shortly to be thoroughly flamed for being off-topic.
I think what it points out is that, those of us who do not know enough about
grammar, should not presume to suggest that fixes to grammar are
unimportant. Bar-room gramarians are, perhaps, as unhelpful in the IETF as
bar-room lawyers, and the reason why we stoop to employ professionals is
because we are not qualified to distinguish our defining elbows from our
You might go to Strunk and White for a good and clear view of the topic
since this work is particular to the American usage that we are required to
turn out our RFCs in.
"That" is the defining, or restrictive, pronoun, "which" the nondefining, or
nonrestrictive... But it would be a convenience to all if these two pronouns
were used with precision. Careful writers go which-hunting, remove the
defining whiches, and by so doing improve their work.
If you have time, you should read Fowler's essayon the topic and extensive
examples (prefering an old edition since modern editors have tended to
remove his "ramblings" in later editions). He gives a simple sentence that
should clearly illustrate the difference...
I always buy his books that/which have influenced me greatly.
Buy them all - "which"
Buy a subset - "that"
Better yet, let us construct an example in the context of a protocol
The defined protocol error conditions which/that are carried in the Notify
message are for information only and MUST NOT cause any change in the
"which" - all defined protocol error conditions are for information only,
they are carried in the Notify message.
"that" - the protocol error conditions carried in the Notify message are for
information only, but there may be other error conditions.
PS. I thought the Chicago definition clear, and the example poor.
Ietf mailing list