Site Contents
Last updated 2008-01-16.
Introduction
LaTeX is a free typesetting program that produces high-quality manuscripts ready for printing. It is frequently used for typesetting mathematical books due to its excellent support for formatting equations, but is in fact useful for nearly any purpose. Cambridge University Press typesets many of its books on linguistics, such as The Syntax of Hungarian, with LaTeX.
LaTeX can be a powerful and economic choice for those who work with classical languages and compose texts on historical linguistics. It supports, among many other languages, Latin and Greek, and even provides the correct hyphenation of text in those languages. Specialized transcription systems, such as those commonly used for reconstructions of Proto-Indo-European, can be included with relative ease. Here I shall present a brief guide to using LaTeX for the classical philologist and historical linguist. However, I do expect the reader to be computer-savvy, able to learn on his own, and to have already become familiar with the basics of LaTeX. A fairly clear primer on LaTeX is The Not So Short Introduction to LaTeX2e (PDF), though there are many other tutorials available. When one has LaTeX installed and has learned how to format basic documents, one can now learn to typeset Greek, Latin, and other philological or linguistic text.
Unicode Support
To enable Unicode support in a modern LaTeX distribution,
one need only add \usepackage[utf8]{inputenc} to the
document preamble and use an editor that can save the file in
UTF-8 encoding.
Enabling Specific Language Support
LaTeX has support for many individual languages through the
package “babel”, which can be activated in the document
preamble with the arguments being the needed languages, i.e.
\usepackage[polutonikogreek,latin,english]{babel}
The option polutonikogreek sets
hyphenation and spelling to ancient, polytonic Greek; typing
merely greek provides only support
for the modern Greek language. Latin is simply latin, though one can also somehow
configure it for Medieval Latin. The language in which most of
one’s document will be written should be the last argument to
Babel. Note that in most LaTeX installations english configures hyphenation and
spelling for American English. If your document is in British
English, try british.
This only informs Babel which languages will be in the
document to come, and the default language is that which comes
last, but to conveniently switch languages within the document
one should declare two commands within the document preamble:
\newcommand\greektext[1]{{\localgreek{#1}}}
and \newcommand\latintext[1]{\foreignlanguage{latin}{#1}}
With these declared one can simply indicate Latin or Greek
text by surrounding it with the suitable markup, e.g. \latintext{vēnī, vīdī, vīcī} for
Latin and \greektext{καὶ ὁ πρῶτος
ἐσάλπισεν} for Greek.
Typography
Typefaces
The default Greek typeface provided by Babel is unsuitable
for classical texts. Some might like it, but its odd-looking
kappa and rho with curved descender make it rather different
than the typefaces of most books. The package “psgreek”
provides five Greek fonts, three of which are free and the
others shareware. To use any of them, one must include the
package by declaring \usepackage[FONTNAME]{psgreek} in
one’s document preamble, replacing FONTNAME with the name of the typeface
one desires, which can be found in the psgreek
documentation. I myself prefer the one called “oxonia”, the
traditional typeface of the Classical Greek publications of
Oxford University Press. However, it lacks kerning and is
suitable mostly for typesetting individual words or the
occasional short sentence. If an author is going to be placing
entire Greek paragraphs in your document, he may wish to use
the two shareware fonts provided, which do contain
kerning.
Hyphenation
Once the user has a fine typeface installed, he will
certainly want to move to the next step in making classical
text look attractive: hyphenation. For Greek, one must first
acquire the hyphenation file GRAhyph3,
install it, and then edit the file
$LOCALTEXMF/tex/generic/babel/language.dat and add the line
greek grahyph3.tex.
The hyphenation file for Latin usually ships with LaTeX,
but to make the engine aware of it one may have to edit the
file $LOCALTEXMF/tex/generic/babel/language.dat and add the
line latin lahyph.tex.
Note that the user should edit a local copy of
language.dat, not the one originally installed by
your LaTeX distribution. Once that is added, one must update
the format files. On at least the tetex distribution, this is
done with: $ fmtutil --all.
Everything would have been much simpler if I had merely read Using a new language with Babel at the TeX Frequently Asked Questions.
Typesetting Indo-European Symbols
Many of these solutions for typesetting Proto-Indo-European
reconstructions and the transcription of ancient IE languages
require the “TIPA”
package. This is a pain to install, but the effort is well
worth it. A guide
to Windows installation has been provided by Yoshinari
Fujino. Once installed, one must enter \usepackage{tipa} in the document
preamble.
TIPA has one drawback, namely that it does not include a
proper italic typeface, but rather in an italic environment
(e.g. \textit{}) it simply slants the standard
typeface. The consequence is that if one uses an individual
TIPA command in an already-italicized string, the TIPA
character looks out of place. In the example below, note the
incongruity the first letter of r̥reǵeti and the last letter of swesorm̥ share against the surrounding
characters.
The solution is to keep all transcriptions in TIPA’s
slanted style instead of using LaTeX's usual italic
environment. Place the following code in your document
preamble: \newcommand\ie[1]{\textipa{\slshape{#1}}}. Now
one can acheive a harmonious appearance by giving all
transcriptions to the \ie{} command, e.g. \ie{*\r*nnewos}.
Whatever slight roughness may now appear is a result of some PDF readers having problems displaying bitmap fonts. However, the document will print with a perfect appearance.
Proto-Indo-European Reconstructions
PIE Vowels
The long vowels ā (Unicode character u+0101 latin small letter a with macron), ē (u+0113 latin small letter e with macron), ī (u+012B latin small letter i with macron), ō (u+014D latin small letter o with macron), or ū (u+016B latin small letter u with macron) can be entered directly.
ə (u+0259
latin small letter schwa) may be entered
directly. Using the Unicode character and adding superscripts,
schwa primum can be entered as ə$^1$ and schwa secundum as ə$^1$.
Glottalic stops
The usual transcription for postulated glottalic stops must
be typeset simply by typing the consonant followed by
apostrophe, e.g. t' p' k'. I
hope to find a solution by which the consonant can be typed
followed by Unicode character u+02BC
modifier letter apostrophe, which is intended to mark
glottalisation, but this does not yet work.
Palatal velars
For the transcription k̑ and g̑ (k and g with inverted breve) one must
issue the commands \textroundcap{k} and \textroundcap{g} respectively.
For the transcription with acute accent over the consonant, ḱ (u+1E31 latin small letter k with acute) and ǵ (u+01F5 latin small letter g with acute) may be entered directly.
Labiovelars
For transcription with velar followed by superscript w, one can directly enter kʷ or gʷ (k or g plus the Unicode character u+02B7 modifier letter small w).
For transcription with velar followed by superscript u̯ (u with breve below), one may enter k\super{\textsubarch{u}} or k\super{\u*u}, and g\super{\textsubarch{u}} or g\super{\u*u}.
Syllabic resonants
For the short syllabic resonants l̥, r̥, m̥, or n̥ (l, r, m,
and n with ring below) one must enter \r*l, \r*r, \r*m, or \r*n respectively.
For the long syllabic resonants l̥̄, r̥̄, m̥̄, or n̥̄ (l, r,
m, and n with macron and ring below) one must enter \r*{\=l}, \r*{\=r}, \r*{\=m}, or \r*{\=n} respectively.
Glides
The transcription of the palatal glide i̯ (i with inverted breve below) can be
entered as \u*i or \textsubarch{u}. The labial glide
u̯ (u with inverted breve below) can
be entered as \u*u or \textsubarch{u}.
Laryngeals
The first laryngeal h1
can be typeset as h$_{1}$, the
second h2 by h$_{2}$, and the third h3 by h$_{3}$.
The transcription of an unspecified laryngeal resonant with
H̥ (capital H with ring below) must be
entered as \textsubring{\*H}.
Indo-European Languages
Avestan
Long vowels with macrons, ə (u+0259 latin small letter schwa), ą (u+0105 latin small letter a with ogonek), may be typed directly (see ).
ə̄ (schwa with macron) must be
typed with the command: \={ə}
Delta must be entered as $\delta$.
Theta θ (u+03b8 greek small letter theta)
may be entered directly, but a variant form of the letter can
be had with $\vartheta$.
ġ (u+0121 latin small letter g with dot above) may be typed directly.
m̨ (m with ogonek) is typed as
\textpolhook{m}.
ń (u+0144 latin small letter n with acute), ŋ (u+014b latin small letter eng) and ṯ (u+1e6f latin small letter t with line below) may be typed directly.
X with acute accent may be typed as \'{x}.
Hittite
For the Hittite laryngeal one can directly enter ḫ (u+1E2B latin small letter h with breve below).
Old Church Slavonic
Using the utf8x input method, most of the common symbols used in Old Church Slavonic transliteration, such as č, š, and ž (c, s, and z with caron), and ĭ and ŭ (i and u with breve) for the yers can be entered directly.
It is common for scholars to use the Latin alphabet but
retain the Cyrillic symbols for the front yer and back yer (as
in Schmalstieg’s Introduction
to Old Church Slavic and Lunt’s Old
Church Slavonic Grammar). For this the TIPA package
provides the commands \textsoftsign and \texthardsign for the soft yer and
hard yer respectively.
For information on typesetting Old Church Slavonic in the Cyrillic alphabet, one may consult my guide Typesetting Old Church Slavonic With LaTeX.
Sanskrit
Long vowels with macrons may be typed directly (see ).
ḍ (u+1E0D latin small letter d with dot below), ṃ (u+1E43 latin small letter m with dot below), ñ (u+00F1 latin small letter n with tilde) ḷ (u+1E37 latin small letter l with dot below), ṛ (u+1E5B latin small letter r with dot below), and ṣ (u+1E63 latin small letter s with dot below) may be entered directly.