BibLaTeX: How to keep first given name only?

Question

In my Bibliography, I would like to keep the first “given” name only (besides, obviously, the “family” name): No further given names, no initials. This has been surprisingly difficult to achieve. Consider this MWE:

\documentclass{scrreprt}
\usepackage[shorthands=off,american]{babel}
\usepackage{csquotes}
\begin{filecontents}{\jobname.bib}
@article{Becker.2012,
  title = {Male gender bias in autism and pediatric autoimmunity},
  author = {Becker, Kevin G and John Q. Public},
  date = {2012},
  journaltitle = {Autism research},
  volume = {5},
  number = {2},
  pages = {77--83},
  langid = {english},

}
@article{Bejerot.Eriksson.2014,
  title = {Sexuality and gender role in autism spectrum disorder},
  author = {Bejerot, Susanne Barbara Camelia and Eriksson, Jonna M},
  date = {2014},
  journaltitle = {PloS one},
  volume = {9},
  number = {1},
  pages = {111--141},
  langid = {english},

}
@book{Cardon.Matson.2016,
  title = {Technology and the Treatment of Children with Autism Spectrum Disorder},
  editor = {Cardon, Teresa and Matson, Johnny X.},
  date = {2016},
  publisher = {Springer},
  location = {Heidelberg and New York and Dordrecht and London},
  langid = {english},
 }
\end{filecontents}
\usepackage[bibstyle=verbose, language=auto, autolang=other, dashed=true,
 maxcitenames=2, sorting=nyt, alldates=year]{biblatex}
\addbibresource{\jobname.bib}
% FAMILYNAME in small caps:
\renewcommand*{\mkbibnamefamily}[1]{\textsc{#1}}
\begin{document}
  \nocite{*}
  \printbibliography
\end{document}

How can I get rid of the “superfluous” names, as it were, short of removing them from my .bib file, which is frequently recompiled from sources that I have little control over?

Answer 1

As mentioned in the comments, biblatex (and indeed BibTeX) has no concept of a middle name. Everything that is not family name, "von" part (prefix), "junior" part (suffix) is understood as given name.

It would be possible to make biblatex distinguish between first given names and middle names if we use the extended name format and add a new name part. But this requires a slightly unnatural input (cf. Bibtex/Biber: how to cite an author using Ethiopian conventions?) and in particular would not change the handling of names specified in the normal <prefix> <family>, <suffix>, <given> or <given> <prefix> <family> order in the .bib file. This solution has the advantage that Biber knows about the different name parts and can act accordingly for sorting, name uniqueness etc.

A simpler solution (with the drawback that Biber does not know about it) is to have the middle names identified and dropped on the LaTeX side after the names have been parsed. l3regex makes that fairly easy if you know about the special format biblatex uses for names (name parts are separated with \bibnamedelima, \bibnamedelimb, \bibnamedelimi and not just spaces).

\documentclass{scrreprt}
\usepackage[shorthands=off,american]{babel}
\usepackage{csquotes} 
\usepackage[bibstyle=verbose, language=auto, autolang=other, dashed=true,
 maxcitenames=2, sorting=nyt, alldates=year]{biblatex}
\renewcommand{\mkbibnamefamily}{\textsc}
\renewcommand{\mkbibnamegiven}{\getfirstgivenname}
\ExplSyntaxOn
\seq_new:N \l_ingmar_givennames_seq
\regex_const:Nn \c_ingmar_splitnames_regex { \c{bibnamedelim(a|b|i)}|\s+ }
\cs_new_protected_nopar:Npn __ingmar_getfirstgiven:n #1
  {
    \regex_split:NnN \c_ingmar_splitnames_regex { #1 } \l_ingmar_givennames_seq
    \seq_item:Nn \l_ingmar_givennames_seq {1}
  }
\cs_generate_variant:Nn __ingmar_getfirstgiven:n { o }
\NewDocumentCommand \getfirstgivenname {m}
  {
    __ingmar_getfirstgiven:o {#1}
  }
\ExplSyntaxOff
\begin{filecontents}{\jobname.bib}
@article{Becker.2012,
  title        = {Male gender bias in autism and pediatric autoimmunity},
  author       = {Becker, Kevin G. and John Q. Public},
  date         = {2012},
  journaltitle = {Autism research},
  volume       = {5},
  number       = {2},
  pages        = {77--83},
  langid       = {english},

}
@article{Bejerot.Eriksson.2014,
  title        = {Sexuality and gender role in autism spectrum disorder},
  author       = {Bejerot, Susanne Barbara Camelia and Eriksson, Jonna M.},
  date         = {2014},
  journaltitle = {PloS one},
  volume       = {9},
  number       = {1},
  pages        = {111--141},
  langid       = {english},

}
@book{Cardon.Matson.2016,
  title     = {Technology and the Treatment of Children with Autism Spectrum Disorder},
  editor    = {Cardon, Teresa and Matson, Johnny X.},
  date      = {2016},
  publisher = {Springer},
  location  = {Heidelberg and New York and Dordrecht and London},
  langid    = {english},
}
\end{filecontents}
\addbibresource{\jobname.bib}
\begin{document}
  \nocite{*}
  \printbibliography
\end{document}

You can see how that goes wrong in

\documentclass{article}
\usepackage[shorthands=off,american]{babel}
\usepackage{csquotes} 
\usepackage[style=authoryear]{biblatex}
\renewcommand{\mkbibnamefamily}{\textsc}
\renewcommand{\mkbibnamegiven}{\getfirstgivenname}
\ExplSyntaxOn
\seq_new:N \l_ingmar_givennames_seq
\regex_const:Nn \c_ingmar_splitnames_regex { \c{bibnamedelim(a|b|i)}|\s+ }
\cs_new_protected_nopar:Npn __ingmar_getfirstgiven:n #1
  {
    \regex_split:NnN \c_ingmar_splitnames_regex { #1 } \l_ingmar_givennames_seq
    \seq_item:Nn \l_ingmar_givennames_seq {1}
  }
\cs_generate_variant:Nn __ingmar_getfirstgiven:n { o }
\NewDocumentCommand \getfirstgivenname {m}
  {
    __ingmar_getfirstgiven:o {#1}
  }
\ExplSyntaxOff
\begin{filecontents}{\jobname.bib}
@book{doe:a,
  title        = {Book A},
  author       = {Jane Emilia Doe},
  date         = {2020},
}
@book{doe:b,
  title        = {Book B},
  author       = {Jane Edith Doe},
  date         = {2020},
}
\end{filecontents}
\addbibresource{\jobname.bib}
\begin{document}
  \cite{doe:a,doe:b}
  \printbibliography
\end{document}

The two different works appear to have the same citation label and doe:b is seemingly incorrectly sorted before doe:a.

---

A fusion of the two approaches would be to use a Biber sourcemap to drop the undesired middle name initials directly when the data is read. Unfortunately, such a solution would have to replicate parts of the name parsing itself, because it would have to deal with all of Jane Emilia Doe, Doe, Jane Edith, Georgina Anne de la Name Nother before they are split into their name parts.

Answer 2

(This is sort of a non-answer, but too long for a comment.)

You can't safely make an acceptable name form automatically from a full name like that. If you have "John Ronald Reuel Tolkien" in your bib file it would be just bad to have a program erroneously make him into "John Tolkien" in the bibliography. Similarly with "Alan Milne", "Thomas Eliot" and many more.

Also bad to have "Johan August Strindberg" in your bib file and have a program output that as "Johan Strindberg". That is just wrong. In this case you should have just "Strindberg, August" in the bib, since that is that author's name as used, on title pages and elsewhere. The bibliography is not the place for the full legal name, so don't put it in the bib file to begin with.

If an author uses several given names on title pages etc., like for example Philip José Farmer it is not up to you to decide that just "Philip Farmer" is enough. The same with initials. It may be the case that a name appears in print in various forms. There are for example book covers where Robert A. Heinlein is called just "Robert Heinlein" but an automatic system doing what you want won't know that, and if let loose will create name forms that just aren't acceptable.