# Citation in “dcu” bibliography style sometimes return “et al” other times full author list

I am having a problem when I \cite (or \citep) using natbib with dcu.bst.

For some citations with 3 or more authors the full list of authors is inserted in the text but in others it is shortened to "et al." (which is what I want).
The problem is not related to the dcu numnames# option as the "et al." shortening applies for most references with 3 or more authors. I noticed that this problem only happens for a few entries in the reference list only (Three examples given below).

I am not using longnamesfirst option for natbib:

\usepackage{natbib}
\bibliographystyle{dcu}


After spending a long time trying to figure out what the problem is, it seems to be (only) related to the last name of the first authors of the references where the problem appears. If I change their last name (e.g., delete or add a letter) et al. works properly (possibly due to a conflict with other references for the same first author).

Here are some bib entries that are creating unexpected citation call-outs. Two entries have Chmait, Nader as the first listed author, different co-authors, and, importantly, the same publication year. In contrast, two others entries have Insa-Cabrera, Javier as the first listed author but do not have the same publication year.

@inproceedings{InsaCabrera11,
author = {Insa-Cabrera, Javier and Dowe, David L. and {Espa{\~n}a-Cubillo}, Sergio and {Hern{\'a}ndez-Lloreda}, M. Victoria and Hern{\'a}ndez-Orallo, Jos{\'e}},
title = {Comparing Humans and {AI} Agents.},
booktitle = {Artificial General Intelligence (AGI)},
isbn = {978-3-642-22886-5},
keywords = {dblp},
pages = {122-132},
publisher = {Springer},
series = {Lecture Notes in Computer Science (LNCS)},
volume = {6830},
year = {2011}
}

@inproceedings{Insa2012,
author      = {Insa-Cabrera, Javier and Benacloch-Ayuso, Jos{\'e}-Luis  and Hern{\'a}ndez-Orallo, Jos{\'e}},
title       = {On Measuring Social Intelligence: Experiments on Competition and Cooperation},
booktitle   = {Proceedings 5th International Conference on Artificial General Intelligence (AGI)},
editor      = {Joscha Bach and  Ben Goertzel and Matthew Ikl{\'e}},
volume      = {7716},
series      = {Lecture Notes in Computer Science (LNCS)},
publisher   = {Springer Berlin Heidelberg},
year        = {2012},
pages       = {126-135}
}

@inproceedings{Chmait2016a,
title     = {Factors of Collective Intelligence: How Smart Are Agent Collectives?},
author    = {Chmait, Nader and Dowe, David L. and Li, Yuan-Fang and Green, David G. and Insa-Cabrera, Javier},
booktitle = {Proceedings of 22nd European Conference on Artificial Intelligence {ECAI}},
address   = {The Hague, The Netherlands},
ISBN      = {978-1-61499-671-2},
Series    = {Frontiers in Artificial Intelligence and Applications},
Editors   = {Gal A. Kaminka and Maria Fox and Paolo Bouquet and Eyke H{\"u}llermeier and Virginia Dignum and Frank Dignum and Frank van Harmelen},
Volume    = {285},
pages     = {542--550},
doi       = {DOI10.3233/978-1-61499-672-9-542},
year      = {2016}
}

@inproceedings{workshopChmait2016b,
author    = {Chmait, Nader and Li, Yuan-Fang and Dowe, David L. and Green, David G.},
title     = {A Dynamic Intelligence Test Framework for Evaluating {AI} Agents},
booktitle = {Proceedings of 1st International Workshop on Evaluating General-Purpose {AI (EGPAI 2016)}},
address   = {The Hague, The Netherlands},
year      = {2016},
pages     = {1--8}
}

• usually (or at least sometimes) inconsistency here is due to errors in the bib file where a list of authors is marked up so that bibtex thinks it is one author with many names, so it does not get et al processing. – David Carlisle Jan 11 '17 at 11:17
• Maybe we can help you to find the reason or mistake in the bib file, if you would add a MWE into your question. – Schweinebacke Jan 11 '17 at 11:24
• Thanks. I have edited the post and added some further comments and gave examples of the references that don't display et al. when I cite them. Is there anything I could do to debug this problem? How come editing the last name of the first author can affect the display of the citation from et al. to full author list? – NadTeX Jan 14 '17 at 7:08
• @DavidCarlisle - It turns out that the behavior reported by the OP is not due to either incorrect input or a bug in the dcu bibliography style. Instead, it's a design feature. – Mico Jan 14 '17 at 7:16

(too long for a comment, hence posted as an answer)

The dcu bibliography style -- "dcu" stands for "Design Computing Unit", Department of Architectural and Design Science, University of Sydney -- is part of the harvard citation management package. (Aside: If you use dcu with the natbib package instead of the harvard package, it's a good idea to load the har2nat package as well, especially if the hyperref package is in use.)

As you've discovered, the dcu bibliography style is programmed not to employ the et al abbreviation in citation call-outs if there are two (or more) multi-author entries which (a) start with the same first author and (b) do not have all authors in common. In this regard, dcu is similar to the agsm bibliography style, which -- not coincidentally! -- is also part of the harvard citation management package. See the posting AGSM bibliography style sometimes doesn't abbreviate to "et al." for duplicate author+year for more on this subject.

The upshot: If you can't stand this behavior, you should probably look into using a different bibliography style.

A full MWE:

\RequirePackage{filecontents}
\begin{filecontents}{mybib.bib}
@inproceedings{InsaCabrera11,
author = {Insa-Cabrera, Javier and Dowe, David L. and {Espa\~{n}a-Cubillo}, Sergio and {Hern\'{a}ndez-Lloreda}, M. Victoria and Hern\'{a}ndez-Orallo, Jos{\'e}},
title = {Comparing Humans and {AI} Agents.},
booktitle = {Artificial General Intelligence (AGI)},
isbn = {978-3-642-22886-5}, keywords = {dblp},
pages = {122-132},
publisher = {Springer},
series = {Lecture Notes in Computer Science (LNCS)},
volume = {6830},
year = {2011}
}
@inproceedings{Insa2012,
author = {Insa-Cabrera, Javier and Benacloch-Ayuso, Jos\'e-Luis and Hern\'andez-Orallo, Jos\'e},
title = {On Measuring Social Intelligence: Experiments on Competition and Cooperation},
booktitle = {Proceedings 5th International Conference on Artificial General Intelligence (AGI)},
editor = {Joscha Bach and Ben Goertzel and Matthew Ikl{\'e}},
volume = {7716},
series = {Lecture Notes in Computer Science (LNCS)},
publisher = {Springer Berlin Heidelberg},
year = {2012},
pages = {126-135}
}
@inproceedings{Chmait2016a,
title = {Factors of Collective Intelligence: How Smart Are Agent Collectives?},
author = {Chmait, Nader and Dowe, David L. and Li, Yuan-Fang and Green, David G. and Insa-Cabrera, Javier},
booktitle = {Proceedings of 22nd European Conference on Artificial Intelligence {ECAI}},
address = {The Hague, The Netherlands},
ISBN = {978-1-61499-671-2},
Series = {Frontiers in Artificial Intelligence and Applications},
Editors = {Gal A. Kaminka and Maria Fox and Paolo Bouquet and Eyke H\"ullermeier and Virginia Dignum and Frank Dignum and Frank van Harmelen},
Volume = {285},
pages = {542--550},
doi = {DOI10.3233/978-1-61499-672-9-542},
year = {2016}
}
@inproceedings{workshopChmait2016b,
author = {Chmait, Nader and Li, Yuan-Fang and Dowe, David L. and Green, David G.},
title = {A Dynamic Intelligence Test Framework for Evaluating {AI} Agents},
booktitle = {Proceedings of 1st International Workshop on Evaluating General-Purpose {AI (EGPAI 2016)}},
address = {The Hague, The Netherlands},
year = {2016},
pages = {1--8}
}
\end{filecontents}

\documentclass{article}
\usepackage{natbib,har2nat}
\bibliographystyle{dcu}

\begin{document}
\citet{InsaCabrera11}

\citet{Insa2012}

\citet{Chmait2016a}

\citet{workshopChmait2016b}

\bibliography{mybib}
\end{document}

• I'd like to add that using \citeasnoun{} only once has solved this issue. Now all multi-author (> 3 authors) citations are shortened using et al. – NadTeX Jan 16 '17 at 1:01
• @NadTeX - I'm afraid I don't follow you. The har2nat package provides the following definition: \newcommand{\citeasnoun}{\citet}; thus, \citeasnoun should behave exactly as \citet does. This is borne out if I add a couple of \citeasnoun instructions to the example code shown above. Are you maybe using a different code sample? If so, it would be helpful if you provided more information about how your code differs from mine. – Mico Jan 16 '17 at 1:14
• Apologies my mistake. In fact I could not resolve the issue using the dcu bibstyle. You are right \citeasnoun behaves exactly as \citet does. (I was confused as I have used apalike instead of dcu and then switched back to dcu but the citations were not updated after the first one or two compiles. That's why I thought the issue was resolved). Is it possible to edit the name formatting in dcu.bst (after renaming it) to ignore full citations if there are two (plus) multi-author entries which (a) start with the same first author and (b) do not have all authors in common? – NadTeX Jan 16 '17 at 3:13
• @NadTeX - I suppose that (almost) anything is possible. However, should it be done? In the case of the agsm and dcu bibliography styles, my view is, "no". I wouldn't change them, in no small part because doing so would be exceedingly tricky. Instead, I would like to suggest that you become familiar with the makebst utility -- part of the [custom-bib](Equilibrium models of exchange rates) package -- which makes it straightforward to create your own, bespoke bibliography style file. It's not difficult to learn how to use this utility -- really! – Mico Jan 16 '17 at 4:10
• Thank you. I will have a look at makebst. – NadTeX Jan 16 '17 at 4:17