does this reflect Voice reality?

I am trying to come up with a kit sack of homely examples, as close as I
can come to what people would feel they already know, to illustrate how the
field structure of dialogs changes between media or interaction modes, such
as between screen and voice.

An example that I am currently running around quoting is

"'City, State' is two independent fields in a standard Web form.  'City'
will be a free text entry and 'State' a select-from-enumeration.  But in
voice this is one input field, and the recognition is coupled."

Is it OK for me to say that?  Is it a fact that voice dialogs will
generally take "City, State" input in one gulp (even 'though the grammar
matches it as two or more tokens)?  

Al