Post by David KastrupPost by Erik QuaeghebeurPost by David KastrupPost by c***@gmail.comJackson's \citeyearpar{Jackson1980} argument was that..
Which seems somehow inelegant. If there's a natbib command which
explicitly produces this result, I'd be really happy to hear about it.
\cite{Jackson1980} argues that ...
The argument of \cite{Jackson1980} was that ...
Really, there are are a number of ways to phrase this without having
to go to weird typesetting.
I thought I wouldn't jump into this debate, but here goes (and we're now
pretty much off topic.) The fact is that linguistically in English, the
possessive construction simply doesn't have an exact paraphrase, and so
as far as I can tell, you are suggesting that one avoids writing
something simply because it produces "weird" typesetting.
Post by David KastrupIt inserts grammatical changes in the _middle_ of a string used for
referencing things.
Post by Erik Quaeghebeur1. The feature is useful: I use the "\citeauthor{}'s \citeyearpar{}" trick
from time to time because it is yet another way of formulating things,
which gives authors more flexibility to write less repetitive texts.
I don't see how this would make the practice unweird. I can also
write a sentence backwards in order to be less repetitive, but that
does not make it a good idea.
Sure, but English of part not is sentence backward the, so this is
comparing apples with oranges.
The English possessive construction has a number of properties that make
it useful to use in exactly the situations we are discussing.
In a structure like "Smith's 1999 analysis" you are able to refer to
Smith using a pronoun in subsequent sentences as in (1a) below. This
doesn't work for other ways of citing (ungrammatical sentences marked
with an asterisk). (These are assumed to have no prior mention of Smith.)
(1) a. Smith's 1999 analysis showed that he had understood ...
b. *Smith 1999 showed that he had understood ...
c. *The analysis of Smith 1999 showed that he had understood ...
This same property also allows for use in sentences relating directly to
the person and not the article. For example: "Smith's 1999
insight/discovery/breakthrough"
What appears to be the closest equivalent of "X's Y", namely "the Y of
X" are simply not equivalent in English. Some examples:
(2) a. John's brother
b. *The brother of John
c. His brother
d. *The brother of him
Post by David KastrupPost by Erik Quaeghebeur2. The latex syntax part can be taken over by a newcommand such as
\newcommand{\citetpossessive}[1]{\citeauthor{#1}'s \citeyearpar{#1}},
which only leaves the optional argument part unused (could be handled,
I suppose).
What has the possibility for an implementation to do with whether
something is to be considered weird typesetting?
But if it's not the implementation that's weird, what exactly is weird,
English itself? The construction is commonly used in English. Shouldn't
LaTeX be able to typeset it?
By now we've obviously scared the OP away but hopefully he'll be happy
enough with my and Erik's suggestions.
Post by David KastrupPost by Erik QuaeghebeurI think it would be a nice feature to have included in natbib
(including correct hyperref interaction). Does (anyone know if)
Patrick Daly accept feature requests?
He does. (But doesn't have much time to address them.)
Alan