博客

This article was originally published on the ShareLaTeX blog and is reproduced here for archival purposes.

Introduction

Since the late 1990s I had known about LaTeX; I had read a couple of manuals typeset with LaTeX and came to know the typical signature of LaTeX typeset documents. At that time, however, I did not think anything of it. Why would I? I grew up with WordPerfect 4.1 before migrating to Microsoft Word 6.0. I was bored. I read the official paper manual. I became very proficient in using this WYSIWYG editor.

Meanwhile, I started an engineering degree in computer science and I was continuously confronted with LaTeX-typeset documents. I added learning LaTeX to my to-do list. Not to the top of the list, mind, as I never found a real need to use LaTeX except for my growing curiosity.

The curriculum was heavy on mathematics and I did my mathematics homework efficiently with pen and paper. It was not until I enrolled in a course on process algebra that learning LaTeX became top priority. The derivations I had to do became longer and longer, and contained increasingly more odd symbols. I made too many errors writing down the derivations by hand. My handwriting has never been good. However, using the formula editor of Microsoft Word was not a realistic option either. I had to use something better suited for entering and typeset mathematics, I had to use LaTeX.

Ever since, I am using LaTeX for all my papers, articles, and text books.

Writing history

When I started studying history, there was no doubt in my mind that I would use LaTeX to write my papers. Historical papers are simple documents. They contain text and maybe a couple of pictures or tables, depending on the historical topic. Besides the lack of mathematics of any kind, the most notable differences with mathematical or computer science papers are the enormous amount of notes and the extensive literature references in those notes. As an example, have a look at a conference paper I wrote about the history of the ALGOL effort, ALGOL, more than just ALGOL.

In this short paper there are 53 notes, referencing 13 primary and secondary sources. The first time a source is referred to, its full reference is given. After that a short reference suffices. If a source is referred to multiple times in succession, the first time it will get a full or short reference and for the other times it is sufficient to refer to it with ibid., meaning 'the same'. Often, but not always, all works referred to are listed again with a full bibliographic entry at the end of the paper. The example paper, for example, could be referred to with:

full reference

• HT de Beer. 'ALGOL, more than just ALGOL.' In MEDICHI 2007 – Methodic and Didactic Challenges of the History of Informatics, ed. L. Böszörmenyi (Oesterreichische Computer Gesellschaft, 2007), p. 102.

short reference

• de Beer, 'ALGOL, more than just ALGOL', p. 102.

bibliographic entry

• de Beer, Huub 'ALGOL, more than just ALGOL', in: L. Böszörmenyi (Ed.) MEDICHI 2007 – Methodic and Didactic Challenges of the History of Informatics, (Oesterreichische Computer Gesellschaft, 2007) 100—111.

Modern references like these are quite straightforward. However, add translations, sources cited in other sources, multiple different version, medieval or older sources with different possible interpretations, and all kinds of non-textual media to the mix and referencing becomes one complex mess. On top of that, there are strict rules on how to use references in historical papers. A historical paper is a text based on interpretations of older texts and other sources. These references and interpretations are the core of historical craftsmanship. It is most important that the reader, i.e., another historian, can find the exact source used to evaluate its relevance and its interpretation. Referencing literature cannot go wrong.

Novice: semi-automatic literature references

Being a novice LaTeX user I did not feel confident to use BibTeX at first. The default referencing schemes of BibTeX did not match the one I had to use in my historical papers. Writing my own complex bibliographic style file seems to be overkill for the short papers I was writing at that time. I solved the problem by introducing two new commands for every source I referred to: one for the full reference and one for the short one. Furthermore, I also introduced a command for the 'ibid.' to make all references light up in my text editor with syntax highlighting. At the end of the paper I would add all bibliographic entries by hand as well.

So, if I would refer to the example paper in LaTeX I would introduce the following two commands:

\newcommand{\BeerF}[1]{ %
  \footnote{ %
    de Beer, HT (2007) 'ALGOL, more than just ALGOL', in:
    L. Böszörmenye (Ed.) \textit{MEDICHI 2007 — Methodic and Didactic Challenges
    of the History of Informatics}, p. #1.
  }
}
\newcommand{\BeerS}[1]{ %
  \footnote{ %
    de Beer, 'ALGOL, more than just ALGOL', p. #1.
  }
}

Automatic literature references with BibTeX and Jurabib

This novice strategy to literature referencing did not scale very well. When I started writing longer historical papers with a lot more references, I had a second look at BibTeX. Meanwhile I had learned to look at CTAN first when searching for an pre-existing solution to my LaTeX woes. I found JuraBib. With this package I was able to fine-tune automatically generated literature references suited for the historical craftsman.

Now all I had to do was to include the following in the preamble:

% The bibliographic style is jurabib-oxford, for creating historical style
% literature references.
\usepackage[%
  titleformat=italic,% Titles in italic 
  titleformat=commasep,% A comma between athors and title 
  titleformat=all,% Always show a title (or a short title)
  commabeforerest,% A comma after title 
  ibidem=strict,% 
  citefull=first,% The first citing in full form 
  oxford,% The oxford style
  super,% Footnotes 
  dotafter=true,% 
  see,% An extra optional argument as a prenote 
  idem
]{jurabib} 
\bibliographystyle{jox}% Use the Oxford style.

I had also to point to the bibliographic database (called bibliography.bib) in the document with \bibliography{bibliography}. And I could cite a source with a simply \cite[102]{deBeer2007a}, assuming that deBeer2007a was a key referencing to an item in the bibliographic database, of course. JuraBib and BibTex figured out if to use a full reference, a short reference, or ibid. And it automatically added a correctly formattted bibliography to the paper as well.

Once I sorted out how to use BibTeX, my life as a historian became a whole lot easier. Before, finishing a paper was a painstakingly process of checking and rechecking all references. Now I could trust LaTeX and BibTeX to do my references for me, do them consistently, and without error.

Automatic literature references with BibLaTeX

BibTex has been deprecated in favor for the more flexible BibLaTeX. The package for historians to use is biblatex-historian. This package can easily be configured to fit the literature reference style you need to use; you read all about that in the package's manual. However, most default settings are well-chosen. Now, with BibLaTeX, all I have to add to the preamble is:

\usepackage[
  style=historian,
  autocite=footnote,
  backend=biber
]{biblatex}
\bibliography{bibliography}

the line backend=biber tells LaTeX to use the biber program instead of the BibTeX program to generate the references and is not history specific.

Epilogue

For me, using LaTeX as a historian enabled me to focus on my writing. It allowed me to better plan my writing as I did not need to worry about formatting and references. At the same time, it added a new set of troubles with respect to collaboration with other historians and submissions to journals and editors. In the end I almost always have to convert my nicely typeset LaTeX documents into Microsoft Word documents. If I am the lone author, I am fine with that: I translate the document to HTML and import that into Microsoft Word or LibreOffice Writer to fix up some conversion errors and save it as a Word document.

Writing together, however, is a pain. Not only are most historians unwilling or unable to use LaTeX whatsoever, but most I worked with are not ICT savvy, at all. It still is common to do all references by hand, to write the bibliography by hand, to painstakingly check and double check all of the many, many footnotes and references. And there are almost always errors and inconsistencies. It seems as if most lack the understanding of the power and possibilities of automation and basic information processing. They waste a lot of time and effort on tasks that already have been automated. LaTeX allowed me to spend my time on what I loved most: doing the actual historical research and writing it up.