Go to page:

Search:   Help

Edit

LyX /

LinguistLyX

<< | Page list | >>

Using LyX for Linguistic Papers

by Maria Gouskova, Stacia Hartleben and Jürgen Spitzmüller

LyX and Linguistics

Linguists, as opposed to, say mathematicians, have diverse and specific typographic needs, such as:

• a numbered paragraph environment, in parentheses, with numbering sequential throughout the document
• table editing: double lines, dashed lines, shaded table cells (for Optimality Theory)
• various symbols, such as the pointy hand and the bomb, flower, etc. (for Optimality Theory)
• phonetic notation (IPA fonts)
• trees, syntactic and autosegmental/prosodic
• aligned glosses
• specific notation for conversation analysis
• math symbols intermingled with phonetic symbols
• switching from one language's orthography to another in the same document/paragraph/line

LyX serves these needs pretty well, even if some of the tasks remain difficult. This site documents LyX's linguistic features and shows some workarounds and feature requests for future LyX versions.

For general information about using LyX in humanities, which covers a lot of linguists' needs as well, consult the site HumanitiesLyX.

For general information on LaTeX for linguists please see Doug Arnold's LaTeX for linguists page, the UPenn LaTeX page, and Ling-TeX. LyX for Linguists

Linguistics Module [new in 1.6]

LyX 1.6 introduces the feature of modules that add specific features to your document. It ships a Linguistics module that adds some native support for numbered examples, glosses, OT tableaux and some semantic markup. You can load the module in Document→Settings→Modules.

Instructions for this module can be found throughout this site. Additionally, LyX 1.6 also includes a small manual that demonstrates the features (cf. Help→Specific Manuals→Linguistics Manual).

If you have suggestions for features that could be added to this module, don't hesitate to contact the LyX developers list.

Phonology/Phonetics

Using LyX to Display Phonetic Characters (IPA)

Beginning with Version 1.5.0, LyX has switched to Unicode. If you are used to working with Unicode fonts such as the SIL Doulos IPA font, you can principally continue entering International Phonetic Alphabet (IPA) symbols in the same way you usually do (for example, using a Keyboard Layout). You can also copy and paste Unicode IPA text from other sources into LyX or insert them via Insert→Special Characters→Symbols.... Such "direct" Unicode input is particularly advised if you're using XeTeX? or LuaTeX with "non-TeX fonts". With this specific setting, there's even no other way to enter IPA (up to LyX 2.0). That's not a problem. You just have to take care to select a font that includes the IPA glyphs, and everything should just work. However, in all other settings (i.e., with "TeX fonts", be it with XeTeX/LuaTeX or traditional LaTeX), direct input has several drawbacks. Read the following section a.), if you are interested in details and technical reasons. If you believe us and/or want to insert good-looking IPA without further ado, just hop to section b.) below.

a.) Drawbacks of direct Unicode input with "TeX fonts" and advantages of the dedicated IPA inset

First, outside XeTeX/LuaTeX, not all IPA symbols are known by LyX yet, If you enter such an unknown symbol, LyX will yield an error message. In this case, find out the UCS4 code point, look up the command name in the TIPA manual, and add a line to the file unicodesymbols that you will find in your LyX installation directory. Please send the newly added symbols to the developer list, so that they can be included in the next release. The newest version of this file can always be found at http://www.lyx.org/trac/browser/lyx-devel/trunk/lib/unicodesymbols. The syntax is explained at the beginning of this file.

Second, if you enter Unicode symbols directly (again, with "TeX fonts"), LyX will take the "normal" (ASCII) characters from your document's main font, the IPA characters from the IPA font. This will most likely look odd.

Third, native insertion of Unicode glyphs results in sub-optimally kerned words (i.e., bad spacing between letters). This is due to the way LyX resolves these glyphs to LaTeX (see below for the technical explanation).

LyX's native IPA support via the dedicated IPA inset cures these issues. Here's why.

If "TeX fonts" are used, IPA symbols are produced by means of the TIPA package, LaTeX's standard package for IPA support. This package includes two font families. One has been designed to be used with LaTeX's computer modern family of fonts, the other with times (i.e. mathptmx.sty). Please refer to the voluminous TIPA manual for details.

So if you insert a Schwa glyph, LyX will automatically convert this glyph to the respective TIPA macro (\textschwa). LaTeX then typesets the Schwa glyph from the TIPA font. If you insert a normal alphabet glyph (such as b), it will be passed to LaTeX literally, however, so LaTeX typesets the b not from the IPA font, but from the main text font, whatever that may be. LyX's dedicated IPA inset, however, assures that all text uses the IPA font. This cures the font mixing problem.

Also, with the dedicated inset, you can additionally use the handy "TIPA shortcut notation" (see below), which is in fact faster than Unicode input, once you are used to it. Moreover, the kerning problems described above do not occur with the shortcut notation. Here's the technical explanation: When LyX exports the Unicode glyphs to TIPA macros (such as \textschwa), it terminated those macros by braces (i.e., \textschwa{}). LyX does this in order to prevent LaTeX errors. Think of a Schwa followed by a b. Without the braces, this would result in \textschwab (instead of \textschwa{}b), and this would confuse LaTeX, because the command \textschwab does not exist. These braces, however, disturb the kerning. The correct solution would be a blank, i.e. \textschwa b, but this requires a much more complex algorithm (In fact, LyX 2.1 will follow this more complex algorithm and thus fix the kerning problems).

So much for technical details. The following sections describe how to use LyX's dedicated IPA inset.

b.) Entering the IPA inset via the Menu

The easiest way to start IPA mode is to use the menu entry Insert→Special Characters→Phonetic Symbols, which will bring you to the IPA mode immediately (this menu entry was introduced in LyX 1.4.3).

You can now enter text, using the shortcut characters that TIPA provides (see the appendix of the TIPA manual for a list of all symbols and the corresponding shortcuts):

That's all. If the TIPA package has been installed correctly, this text will come out as phonetic notation in the DVI or PDF output. Moreover, as long as LyX's Instant Preview feature is enabled, LyX will display the phonetic text inside LyX as soon as the cursor leaves the math inset. Voilà:

As of LyX 2.0, you can also input or paste Unicode IPA glyphs into the IPA inset. LyX will automatically export them to the correct macros in the output, and no font mixing will occur. Note though, that this has the kerning drawbacks described above.

c.) Inserting the IPA inset manually

Alternatively, you can start IPA mode as follows. First switch to math mode. This is done by the keyboard shortcut C-M (or Insert→Math→Inline formula). To enter math's text mode, hit C-M twice. This introduces the so called math_textinset:

To switch to TIPA mode, insert \textipa into this math inset and hit the space key. The \textipa will disappear and everything looks as before, but LyX is now in TIPA mode.

Now enter the text as described in the previous section.

d.) Converting from text to math mode

You can also insert it this way: write \text{\textipa{Ekspl@"neIs@n}} in LyX (outside math), select this text, and then hit C-M. It will be converted automatically to math mode.

e.) If everything else fails: Using ERT

Finally, you can use LyX's LaTeX mode (ERT) to enter any TIPA command. This might be necessary sometimes to use things that are not yet supported. Of course, you'll lose all the nice features of the native support, most notably Instant Preview.

f.) Perspectives: an improved IPA inset

Forthcoming LyX 2.1 (currently still in an early development stage) will feature a new IPA inset, which does not rely on the math editor and which will fix some of the shortcomings of the current IPA inset. Watch out for details here.

Tables and Optimality-Theoretic Tableaux

Unlike some LaTeX editors, LyX allows you to edit tables visually, though its table formatting capabilities are somewhat limited. The most important OT-specific table formatting need is dashed vertical lines--the typical way of showing that two constraints cannot be ranked with respect to each other based on available evidence. Dashed lines are created using the package arydshln. Another need, which is formally somewhat superfluous(approve sites), is table cell shading--the way to show the violations that are irrelevant to the fate of a candidate. See the LyX wiki page Tables for dashed lines and shading.

Here are some examples of tableaux:

Tableaux Floats and List of Tableaux: [new in 1.6] The Linguistics Module provides a separate float for Tableaux as well as a List of Tableaux (with a separate counter). You can find and use both the same way as table and figure floats. Note that the captions are always below the tableau by default, no matter where you put them in the LyX view. To get captions on top, put in preamble

\floatstyle{plaintop}
\restylefloat{tableau}

If your using a KOMA class and the option "tablecaptionsabove", use the following instead for better spacing:

\floatstyle{komaabove}
\restylefloat{tableau}

For LyX < 1.6, you have to create such a float yourself. The procedure is explained in this message to lyx-users.

Feature Matrices/Stuff in Brackets

For SPE-style feature representations in brackets, go to the Math Panel (Insert→Math→Math Panel) and click on the button to open the delimitor and bracket dialog. Select square brackets from the many available options, and then click on the "insert matrix" button to insert a matrix (basically, a mini-table without grid marks) inside the brackets. You can specify the number of rows and columns inside the matrix in the emerging dialog box.

If you need to use the pound sign, #, inside math mode, you must type it in as \#, otherwise LyX will not recognize it as a valid character. If you want the text inside the matrix to not appear in italics, select the line and change it to text mode (see above).

Other Special Characters and Diacritics

The Comprehensive LaTeX symbol list by Scott Pakin covers many of the symbols that can be produced with LaTeX. You can insert them in LyX by means of ERT or in math-text mode as described above. Furthermore, LyX supports the direct input of unicode characters as of version 1.5.0 and automatically translates these glyphs to the according LaTeX macro on output (almost the complete list of Pakin and more should be supported).

For example, an acute accent on an "a" can be entered

• as character á (if the keyboard supports it, maybe via a dead key or a key combination),
• via cut and paste from external sources,
• with the accent-acute lfun (M-x accent-acute a),
• via the lfun unicode-insert 0x00e1 (where 0x00e1 is the unicode code point of the glyph á),
• as ERT (Ctrl-l \'a).

[new in 1.6] LyX 1.6.0 will introduce a general symbols dialog (Insert→Special Characters→Symbols...) from which you can directly insert all supported characters (cf. http://wiki.lyx.org/LyX/NewInLyX16#symbolsdialog)

Symbols such as arrows, zero, relations of domination, relative harmony, and so on can be found in the Math Panel. If you need to place a primary/secondary stress (grave/acute accent) or an unstressed/short mark on top of a Greek letter, the easiest way is the Math Panel/Frame Decorations.

If you need a pointy hand to mark the winner in your OT tableau, there are several options available in various packages. The bbding package has a nice pointy hand with a cuff, which in ERT is entered as \HandRight. See The Comprehensive LaTeX symbol list for more. The pointy hand is supported in Instant Preview.

Using LyX to Draw Trees

Packages

One of the following packages is needed to draw trees in LyX:

1. The forest package. A tree drawing package specifically for linguistics which draws on tiKZ/PGF. Probably the most powerful and intuitive package currently available. Supports bracket (i.e. "qtree-style") input syntax.
2. The qTree package (location on CTAN). Take care that you have a recent version of qtree that supports the pict2e package.
3. The pst-qtree package is a frontend for the PS-Tricks tree package pst-tree. It uses a similar syntax as qTree. Compared to qTree, the package can do more complex things (such as traces and dotted branches); however, it does not work with pdflatex out of the box (look here for details), and you need to have PS-Tricks installed.
4. Ralf Vogel's xyling package, which is based on xypic.
5. Koaunghi Un's xytree package, based on xypic as well. This package is quite similar to xyling, but differs a bit in the syntax and in what you can do in detail.
6. For extra power, you may have to resort to xypic itself.
7. Till Tantau's tiKZ, a quite powerful vector drawing package (similar to PSTricks) included to the PGF bundle, can draw trees as well. It has a quite intuitive syntax, but is an overall complex beast.
8. tikz-qtree adds a wrapper around tiKZ and allows the use of the qTree syntax for tiKZ trees. In a similar vein as pst-qtree (which was written by the same author, David Chiang), it aims to combine the intuitiveness of qTree with the power of tiKZ.

The choice of the package depends on your needs and taste. Here are some pros and cons:

• The Qtree approach (plain qtree):
• PRO: the user interface is much more intuitive as xyling's (unless you are already familiar with the xypic syntax).
• PRO: works with pdflatex out of the box
• PRO: works with Instant Preview for a WYSIWYG experience (however, for more complicated trees it is advisable to put trees in an ERT snippet).
• CON: the nodes are not nicely aligned.
• CON: rather limited possibilities
• RECOMMENDED for drawing simple trees, and if the nodes of the tree (i.e., the phrases) do not need to be aligned
• The XY approach (XYLing, XYTree):
• PRO: much more features for complex trees than plain qtree (traces and other nifty things)
• PRO: better control over the alignment of nodes
• CON: the interface is less intuitive (however, the packages come with an excellent step-by-step manual)
• CON (only XYLing): for pdflatex, the package must be modified (see comment in xyling.sty)
• CON: no Instant Preview.
• RECOMMENDED for drawing complex trees and if all nodes have to be aligned
• The Qtree emulators (pst-qtree, tikz-qtree):
• PRO: intuitive syntax (same as qtree)
• PRO: much more features for complex trees than plain qtree (traces and other nifty things)
• CON (only pst-qtree): does not work with pdflatex out of the box
• CON: the nodes are not nicely aligned.
• RECOMMENDED for drawing complex trees and if the nodes of the tree (i.e., the phrases) do not need to be aligned.
• The TiKZ approach (plain tiKZ, not tikz-qtree)
• PRO: rather intuitive syntax (not as intuitive as qtree, but still much more intuitive than XY)
• Many possibilities for nifty graphic polishment
• CON: not specifically aimed at linguistics, so maybe involves more digging.
• Forest
• PRO: combines the intuitiveness of the QTree syntax with the power of TiKZ
• Comes with "styles" for specific linguistic needs.
• Aligns nodes very nicely
• Excellent support for arrows and stuff
• Many possibilities for nifty graphic polishment
• Good documentation
• CONS: None known yet (but not yet extensively tested)

Simple Trees with qTree

Install the qTree package and then put \usepackage{qtree} into your preamble. Make sure that Instant Preview is activated. Then hit C-M twice to switch to math-text mode, and type the following text for example (the blanks are important):

 \Tree[.S [.N This ] [.V is ] [.NP [.Det a ] [.N test ] ] ]

After the cursor has left the inset, the tree should be displayed inside LyX:

Alternatively, you can again write the text outside math mode, select it, and hit C-M to convert it to math (see here).

If the branches of the tree vanish in your preview, you are facing a bug in dvipng, the converter used by Instant Preview to generate the preview images for LyX. The bug has been reported to the dvipng developer and he has fixed it for the upcoming version 2.0 of dvipng. So watch out for dvipng >= 2.0 and upgrade to that.

Roofs

Roofs (for complex phrases) can be inserted with the \qroof command. In LyX, you can insert this by typing \qroof{ at the appropriate place in the formula. As soon as you have typed the curly bracket, a new subinset is created:

There you can type the phrase inside this subinset, which is then marked as follows.

Here's the Instant Preview display after the cursor left the inset:

Glosses

To put glosses under the nodes of a tree, you have to fool the Math Editor. In qtree, this can be done with the linebreak command \\, but the math editor does not accept that. So put in the preamble:

\renewcommand{\breaknode}{\\}

Then you can do:

\Tree[.S [.N Das\breaknode This ] [.V ist\breaknode is ] [.NP [.Det ein\breaknode a ] [.N Versuch\breaknode test ] ] ]

This also works inside qroofs.

Prosodic Structure Trees/Greek Symbols in Trees

If you want to include greek letters such as sigma for syllable nodes or mu for morae, you have to enter math-math mode while keeping the tree in math-text mode. The greek symbols must then be entered in LaTeX code, which requires putting them between dollar signs, as in:

\Tree [.$\sigma$ $\mu$ $\mu$ ]

Instead of manually inserting the dollar signs, you can also switch to math-math mode before entering the Greek letters by pressing again C-M. Then you can enter just, for instance, \sigma, which will be converted to the symbol (after you hit <space>). Don't forget to return to text mode again after the Greek letter by leaving the subinset (in math mode, spaces are impossible).

This will produce a simple syllable (sigma) branching into two moras (mu), viewable via Instant Preview inside LyX.

Autosegmental trees

For simple feature-sharing trees, qtree is again recommended, as it can handle multiply branching trees and does not require you to define the position of the nodes for the branches to connect them correctly. If, however, you need to write a spreading rule with a dashed line, you need something more powerful, such as xyling. Unfortunately, as noted, the package does not seem to be supported in Instant Preview, so you have to use ERT. Here are a couple of examples and the trees they produce:

\Tree {\K{[back]} \B{d}\Bdash{dr} \\ \K{V} & \K{V}}
\Tree{\K{[back]} \D\Bdash{dr}\Bdash{drr} \\ \K{V} & \K{V} & \K{V}\B{d} \\ & & \K{[-low]}}

To do autosegmental delinking, you need xypic (which is the basis for xyling). Here is an example, along with the code that produced it. See XY-pic for more.

"Right-side-up" trees, lattices, OT Hasse trees (ranking diagrams)

The package xyling can do upward-pointing and crossed branches.

Using LyX to display numbered examples with Covington

The package Covington does several useful things, but one of its most convenient functions is that it provides a sequentially numbered environment that can span several sections. Your numbered examples will look as follows:

The package provides the environments example (for single examples) and examples (for multiple, subsequent examples). These can be used by means of ERT.

Setting LyX up to include a Covington environment

If you use numbered examples a lot, you can also make the Covington numbered environment available in your document as a separate paragraph environment similar to Enumerate, Itemize, and others.

[new in 1.6] As of LyX 1.6, doing this is easy. Just load the Linguistics Module and you will find styles for Numbered Examples (multiline) (covington's example environment), Numbered Examples (consecutive) (covington's examples environment) and for Subexamples (additionally defined by LyX) in the paragraph style combo box.

For versions older than 1.6, there are two ways to achieve the same (the first one is easier):

1. Grab the file linguistics.inc, which contains two environments Example for a single example and Examples for multiple subsequent examples plus an environment Subexample).
Place this file in your .lyx/layouts directory and include it in the layout file of your choice by inserting the line
Input linguistics.inc
there.
2. Create a new LyX doc class layout yourself. First, go to Steve Litt's page to get a general idea of what this involves. Then create a .layout file with the following text in it:

#% Do not delete the line below; configure depends on this
#  \DeclareLaTeXClass[article]{article (+Numbered)}
# Article textclass definition file. Taken from initial LyX source code
# Author : Matthias Ettrich <ettrich@informatik.uni-tuebingen.de>
# Transposed by Pascal Andre <andre@via.ecp.fr>
# Heavily modifed and enhanced by several developers.

# Input general definitions
Input stdclass.inc

# There are no chapters in an article.
NoStyle Chapter
NoStyle Chapter*

MaxCounter              Counter_Section
SecNumDepth             3
TocDepth                3

Style Numbered
LatexType       Environment
LatexName       example
Align           Left
AlignPossible   Left

Font
Color           Blue
EndFont
Preamble
\usepackage{covington}
EndPreamble
End


All these methods will make every Covington-numbered example display in blue font but without a number; the document will typeset in a normal way and you can cross-reference Covington-numbered examples in the text as usual (by inserting label). This is how your screen will look for the Covington example above:

Covington's example style is aimed at typical use, which means that it needs to be adjusted for specific uses. Here are some advices how to do that in Q/A form.

• How can I change the label of the subexamples?
Place [(whatever)] in TeX mode at the very beginning of the subexample.
• How can I adjust the example(s) indendation?
Covington has a fixed indendation that fits well if the example number is not too large. However, with more than 99 examples or if you use a book class, where covington numbers chapterwise [e.g. (1.10)], the indendation looks odd especially for multi-line examples.
You can adjust the indendation with the following preamble code (adjust the values if needed):
\let\oldexamples\examples
\let\endoldexamples\endexamples
\renewenvironment{examples}{%
\setlength\leftmargini{3.05em}
\begin{oldexamples}%
\settowidth{\labelwidth}{(0.00)}}
{\end{oldexamples}}

\renewenvironment{example}{%
\setlength\leftmargini{3.05em}
\begin{examples}%
\settowidth{\labelwidth}{(0.00)}%
\item}
{\end{examples}}

• How can I make Covington number my examples consequently, not per chapter?
Covington (mis)uses LaTeX's equation counter, so you need to adjust this counter in your preamble as follows:
% use package remreset to omit counter resetting on new chapter
\usepackage{remreset}
\@removefromreset{equation}{chapter}
% redefine equation counter (do not prepend chapter number)
\renewcommand{\theequation}{\arabic{equation}}

If you just want to omit the chapter prefix and still want to have the examples start from (1) at each chapter start, just use the last two lines of the code.

A Note on Example Number Alignment

If you are using a table (e.g. an OT tableau) as the content of an example, the default is for the example number to align vertically with the center of the table if no title/comment is included before it within the example. To adjust this so that the example number is by the top of the table, enter the "Table Settings" dialogue and, in the "Table-wide settings" box, change the "Vertical alignment" setting to "Top".

Aligned Glosses (Covington)

Often linguists need to have a list of examples which are glossed word by word. A great package for this is, again, Covington.

[new in 1.6] The Linguistics Module shipped with LyX 1.6 provides some native support for covington glosses, even though you still need some LaTeX syntax. After you have loaded the module, you will find two insets "Glosse" and "Tri-Glosse" in Insert→Custom Insets. The following screenshot shows how a sentence and its glossed translation are input into the inset, just as two consecutive lines (for Tri-Glosse, that would obviously be three). The last line is the (mandatory) (grammatically well-formed) translation which must be preceded by \glt (not in ERT, just verbatim):

The PDF output of the example looks as follows:

In LyX 1.5 and earlier, you will have to stick to ERT snippets. The above example would have to be inserted in TeX mode as follows:

\begin{examples}
\item This is an example in English.
\item \gll Dit is een voorbeeldje     in het Nederlands.
This is a {little example} in {}  Dutch.
\glt This is a little example in Dutch.'
\glend
\end{examples}


If you have set up LyX to support covington examples natively (as described here), you do not need the \begin{examples} and \item parts. Simply insert an ERT with everything between \gll and \glend to an Example paragraph:

Also note that up to LyX version 1.3 you will have to separate the lines inside the ERT with a linebreak (C-Return). Using the return key will lead to wrong output. As of LyX 1.4, you will not have to bother about this anymore. Just use Return.

Semantics

Markup

[new in 1.6] The Linguistics Module that ships with LyX 1.6 provides some semantic markup for "expression" (emphasized by default), "concept" (small caps by default) and "meaning" (enquoted by single quotes by default). The definitions can be changed in the preamble.

Logical Semantics

Fortunately, the logical semantics notation is pretty much covered by LyX's math mode. So if you want to insert logical semantic notation, do as if you would want to insert a math formula. You'll find most of the required symbols in the Math Panel (Insert→Math→Math Panel). If you are a bit familiar with the LaTeX notation of that symbols, you can enter them directly into the formula. LyX will convert them for you. With Instant Preview enabled, you'll even get a WYSIWYG display in LyX.

Other stuff

There are packages available for other semantic needs.

• Discourse Representation Structures (DRS) can be noted with Covington, the sdrt or the drs package.

Conversation Analysis

In Conversation Analysis, turns need to be aligned to mark overlaps, turn takings, etc. LyX does not allow multiple blanks and does not have a tabulator, but with LaTeX's tabbing environment, the alignment can be managed quite well. Unfortunately, LyX does not yet support tabbing natively, so we are stuck with ERT.

Consider the following simple example:

\begin{tabbing}
1 \=Peter: \=Do you know [=LyX=] \=the the document\\
2 \>Mary:\>\>yeah I know
\end{tabbing}


With \= you save a position, at which \> will push the text. Note that <Return> does not work inside tabbing; you have to use linebreaks (\\). To adjust the vertical space, you can also use the optional argument of the linebreak, e.g. \\[\baselineskip].

Caveats:

• Inside tabbing, the lines are not broken automatically. You have to do it manually to prevent the lines go into the margin.
• The tabbing environment redefines the LaTeX's accent commands. If you use accents, you'd better try the package Tabbing which does not touch the accent commands and provides another tabbing syntax instead.

If you need arrows and other symbols for your notation, you'll most likely find them in the Math Panel. If not, read this.

Switching languages

To switch languages, e.g. between English and Greek, go to Edit→Text Style→Customized...→Language and select the respective language (e.g. "Greek"). This will apply to all following text (or the selected text). If you select "reset", the language will be reset to the document's main language (as set in Document→Settings→Language).

Foreign languages (i.e. those differing from the document's main language) are marked by a blue underline.

If you use "aspell (library)" as spell checking backend and if the spell checking dictionaries are installed, LyX will spellcheck all marked passages in the allocated language.

If you need to switch languages a lot, opening the text style dialog every time might be too clumsy. To speed up the workflow, you can define keyboard shortcuts for your common languages:

• go to Tools→Preferences→Edit→Shortcuts
• Hit the "New" button to add a new shortcut
• enter "language <language name>" to the function command, e.g. "language english" or "language ngerman" (see the file languages in the LyX system directory for a list of valid language names)
• press the grey shortcut field
• Enter the shortcut of your choice (e.g. Alt+l+e for English, Alt+l+g for German)
• Hit OK
• Hit Apply
• Hit Save

Now you can switch to English by just pressing Alt+l+e (or whatever keyboard shortcut you have defined). Note that the language function acts as a toggle. I.e., if you mark a foreign German passage of text and press Alt+l+g (for German), the language of the passage will be reset to the document's main language.

Note that in older (< 1.5) versions, only one kind of writing system will be displayed inside LyX (i.e. Latin letters instead of Greek ones if the main language is English). In the PDF output, however, the correct writing systems (Roman alphabet for English, Greek alphabet for Greek) are used. As of version 1.5, LyX uses the unicode encoding internally, so it can to display Latin and Greek (and Cyrillic ...) letters in parallel. However, this is only done if the letter is keyed in as a Greek one (e.g. via a character picker, drag-and-drop from a unicode encoded source or a matching keyboard mapping).

To switch between English and Russian, go to Document→Settings→Customized...→Language and set the document's language to Russian. Switch to the Cyrillic keyboard layout of your OS when you need to type in Russian, and switch back when you need English. Both Russian and English will be displayed in LyX.

Bibliographies

LyX has native BibTeX support for bibliographies. By default, LyX is configured to use the bibliography style plain, which is used in other sciences and which cites references in text by number rather than by name and date. Most linguists will probably not find it useful. If you need to use a bibliography style similar to that of Linguistic Inquiry or Language, you should use the natbib or the jurabib package (see this page for information on the differences of those two packages). To find out more about using BibTeX in LyX, see the BibTeX Group on this wiki and LaTeX for Linguists/Bibliographies (the main page for LaTeX for Linguists at that page is wrong and should be http://www.essex.ac.uk/linguistics/external/clmt/latex4ling/). A bare-bones guide which should suit linguists' basic needs is included in the LyX for Humanities page.

A bibliography style popular among linguists is linquiry2.bst. It's designed for natbib.

List of Publications

You can generate a list of your own publications semi-automatically, using BibTeX. See this page for details.

Acknowledgment footnote

It is common in linguistics articles to label the acknowledgment footnote (usually the first footnote) with an asterisk rather than with a regular number. To accomplish this in LaTeX, put the following in ERT: \thanks{text of your acknowledgment footnote}`. Note that in the title environment, LyX automatically outputs footnote insets as \thanks, so you can just insert a footnote as usual.