%% tolkienfontsdoc.tex
%% Part of the tolkienfonts package
%% Copyright 2010 Ariel Barton
%
% This work may be distributed and/or modified under the
% conditions of the LaTeX Project Public License, either version 1.3
% of this license or (at your option) any later version.
% The latest version of this license is in
%   http://www.latex-project.org/lppl.txt
% and version 1.3 or later is part of all distributions of LaTeX
% version 2005/12/01 or later.
%
% This work has the LPPL maintenance status `maintained'.
% 
% The Current Maintainer of this work is Ariel Barton.
%
% This work consists of all the files listed in Appendix F of the file
% `tolkienfontsdoc.pdf'.
%
% Date: 2010/08/29
%
% Version: 2

\documentclass{amsart}

\usepackage{tipa, boustr, tolkienfonts}
\sunrune

\makeatletter\def\@cite#1#2{{\@citestyle [\citeform {#1}\if@tempswa \ #2\fi ]}}\makeatother
% This line gets rid of the comma in citations with optional arguments

\def\<{$\langle$}
\def\>{$\rangle$}

\usepackage{color}
\definecolor{blue}{rgb}{0,0,1}
\definecolor{green}{rgb}{0,0.8,0}
\definecolor{red}{rgb}{0.9,0,0}
\definecolor{orange}{rgb}{1,0.5,0}
\definecolor{purple}{rgb}{0.7,0,1}
\definecolor{gray}{gray}{0.8}

\newcommand{\ment}[1]{\textsl{#1}}

\usepackage[colorlinks=true, citecolor=blue, linkcolor=black]{hyperref}
\def\sectionautorefname{Section}
\def\appendixautorefname{Section}
\def\subsectionautorefname{Section}
\def\subsubsectionautorefname{Section}

\title{Tolkienian Scripts with Virtual Fonts}

\author{Ariel Barton}

\newcommand{\chr}{\small\char`\\char}

\newlength{\halfwidth}
\setlength{\halfwidth}{0.4\textwidth}

\begin{document}

\maketitle

\makeatletter % I don't want it to print ``Appendix'' in the TOC.
\def\tocsection#1#2#3{\indentlabel {\@ifnotempty {#2}{\ignorespaces
   #2.\quad }}#3}
\makeatother

\tableofcontents
\listoftables

\section{Introduction}

This package consists of virtual fonts (and a few macros) designed to
do what \TeX\ is meant to do: convert legible, easy-to-type ASCII text
files into attractive, properly formatted documents. It's just that
``properly formatted'' in this case means written correctly in one of
the writing systems invented by J.R.R. Tolkien.

This package supports the tengwar and cirth as described in
\cite{appE}, and Anglo-Saxon runes as used in \cite{thror}. Some
support is also provided for writing Quenya (but no other language)
with sarati and the valmaric script.\footnote{Sarati is the first of
these four scripts that Tolkien invented. It has a place in the
Middle-earth cosmology as the alphabet invented by R\'umil of Tirion.
F\"eanor thus did not invent the tengwar out of whole cloth; they're
an improvement of an earlier system.

The valmaric script was invented by Tolkien after sarati but before
tengwar, and looks quite a bit like both of them. I don't know if
Tolkien still thought of it as the alphabet of Valinor after inventing
the tengwar.}

This package uses several freely available TrueType fonts for the
Tolkienian scripts. It is designed to be used with pdf\/\TeX.

This package uses virtual fonts to convert normal text to tengwar.
This is a radically different approach from Ivan Derzhanski's
\textsf{tengtex} package, which instead uses sophisticated \TeX\
macros. Virtual fonts compile faster and there are more books and
articles on the subject, which is why I used them; however, they do
mean that I have to make several new supporting files for even the
most trivial changes in mode, which Derzhanski does not.

If you only want to write Elvish, you can stop reading this document
after \autoref{sec:lastneeded}. The remaining sections are for
writing English (much more complicated), for tengwar experts who want
to know exactly how my transcription works, or for anyone who wants
extra fonts or color effects.

\section{License}

This work is copyright Ariel Barton, 2010. It may be distributed
and/or modified under the conditions of the \LaTeX\ Project Public
License, either version 1.3 of this license or (at your option) any
later version.

 The latest version of the license is in
 \begin{quote}
    \href{http://www.latex-project.org/lppl.txt}
    {http://www.latex-project.org/lppl.txt}
  \end{quote}
 and version 1.3 or later is part of all distributions of
 \LaTeX\ version 2003/06/01 or later.

This work has the LPPL maintenance status ``maintained''. I am the
current maintainer; I may be reached at \texttt{origamist@gmail.com}.

This work is a package designed to make it easier to \emph{use} fonts
such as Tengwar Annatar. Those fonts \textbf{are not} considered a
part of this work. The font family Tengwar Annatar is copyright Johan
Winge; the fonts Sarati Eldamar, Valmaric Eldamar, Tengwar Parmaite,
and Tengwar Eldamar are copyright  M\r{a}ns Bj\"orkman; the fonts
Cirth Erebor, Tengwar Quenya, Tengwar Sindarin, and Tengwar Noldor
are copyright Daniel Smith. It is your responsibility to make sure
that your use of these fonts does not violate the creators'
copyrights. (The sunrune font, and this font alone, is considered part of the package, because I made it.)

Similarly, the tengwar, cirth, sarati and valmaric scripts were
created by J.R.R. Tolkien; it is your responsibility to ensure that
your use of this package does not violate the rights of the person,
persons, or entity who owns this intellectual property.

As of August 28, 2010, these fonts are all available for free download,
and I can use them in pdf\/\TeX\ documents. No other claims are made
by the author regarding these fonts. In particular, I cannot guarantee
and take no responsibility to ensure that they will remain available
or will work with your distribution of \TeX. Furthermore, any attempts
to download, install, or use these fonts are taken at your own risk; I
am not liable for any bad consequences of downloading, installing,
using, attempting to download, install, or use, or being unable to
download, install, or use these fonts.

\section{Installing the package}

\label{sec:install}

This package requires over a hundred supporting files, not all of which
are in its CTAN directory. None of the tengwar fonts come with the 
package; you have to get them from their creators' websites.

As of this writing, you can get these fonts from
\begin{quote}
\href{http://www.acondia.com/fonts/cirth/index.html}
{http://www.acondia.com/fonts/cirth/index.html}\\
\href{http://home.student.uu.se/jowi4905/fonts/annatar.html}
{http://home.student.uu.se/jowi4905/fonts/annatar.html}\\
\href{http://at.mansbjorkman.net/downloads.htm}
{http://at.mansbjorkman.net/downloads.htm}
\end{quote}

It is possible to use the \texttt{.pfb} versions of these files instead of the \texttt{.ttf} versions. (I find, for example, that pdf\/\TeX\ doesn't like the file \texttt{tengel\_a.ttf}, so when using Tengwar Eldamar, I have to use \texttt{tengel\_a.pfb}.) You change by editing \texttt{tolkien.map}. See the comments in that file. (\TeX\ won't use the \texttt{.afm}, \texttt{.pfm} and \texttt{.inf} files.)

The fonts and the files that come with this package should be put in
appropriate places. Your distribution of \TeX\ may be able to do some of
this for you; however, it probably cannot deal with the TrueType fonts, so you will have to place them yourself.

Most modern \TeX\ distributions have a folder, usually named \texttt{texmf}, where you can store supporting files for the packages you add yourself.\footnote{If you're using Mac\TeX,
this folder should be \texttt{Users\slash username\slash
Library\slash texmf}. If it isn't there, create it.

If you're using MiK\TeX, it is possible to designate any folder you like as the root of your local tree, that is, the place where you store supporting files. Instructions may be found at \href{http://docs.miktex.org/manual/localadditions.html\#id573803}{\texttt{http:/\slash docs.miktex.org\slash manual\slash localadditions.html\allowbreak\#id573803}}
 or through the manual which should have come with MiKTeX.

Any time you add supporting files to a local MiKTeX root, you have to refresh the file name database; see 
\href{http://docs.miktex.org/manual/configuring.html\#fndbupdate}{\texttt{http:/\slash docs.miktex.org\slash manual\slash configuring.html\allowbreak\#fndbupdate}}. } 
All supporting files should be
sorted into specific subfolders of \texttt{texmf}.
The sorting rules are:
\begin{itemize}
\item \texttt{.fd} and \texttt{.sty} files go in \verb.texmf/tex/latex.
%\item \texttt{.tex} files go in  \verb.texmf/tex/plain.
\item \texttt{.tfm} files go in \verb.texmf/fonts/tfm.
\item \texttt{.vf} files go in \verb.texmf/fonts/vf.
\item \texttt{.pfb} files go in \verb.texmf/fonts/type1.
\item \texttt{.map} files go in \verb.texmf/fonts/map.
%\item \texttt{.enc} files go in \verb.texmf/fonts/enc.
\item \texttt{.mf} files go in \verb.texmf/fonts/source.
%\item \texttt{.afm} files go in \verb.texmf/fonts/afm.
\item \texttt{.ttf} files go in \verb.texmf/fonts/truetype.
\end{itemize}
In all cases they can go in sub-subfolders; for example, \texttt{.tfm}
files may be put into \texttt{texmf/fonts\slash tfm/tengwar} and not
\texttt{texmf/fonts/tfm}.

If you're using some other distribution, you may have some entirely
different place where you can put these files. Your distribution's
documentation should tell you where.

If you really can't figure out where to put the files, if you're in a
hurry, or if you're using someone else's computer and don't want to mess
with their \texttt{texmf} folder, just dump every file you think you
might need into the same folder as the document that uses the package.

You aren't done! \TeX\ now knows everything it needs to do its job and \emph{arrange} the characters in the font, and so your document will compile, but the postprocessing software (your DVI viewer, your printer, or the PDF files that pdf\/\TeX\ produces) don't know about the fonts themselves.

There's a simple way to tell pdf\/\TeX\ about the fonts: use the command
\begin{quote}\verb|\pdfmapfile{+tolkien.map}|\end{quote}
or the lines
\begin{quote}
\verb|\pdfmapline{+CirthEXX\space <CirthErebor.ttf\space <ansinew.enc}|\\
\verb|\pdfmapline{+CirthSXX\space <sunrune.pfb}|\\
%
\verb|\pdfmapline{+SaratiE \space <sarael__.ttf\space <ansinew.enc}|\\
\verb|\pdfmapline{+SaratiB \space <saraelb_.ttf\space <ansinew.enc}|\\
\verb|\pdfmapline{+Valmaric\space <valmarel.ttf\space <ansinew.enc}|\\
%
\verb|\pdfmapline{+TengwarA\space <tngan.ttf\space <ansinew.enc}|\\
\verb|\pdfmapline{+TengwarAA\space <tngana.ttf\space <ansinew.enc}|
%
\end{quote}
These map lines can go in the respective \texttt{.fd} files,
\texttt{tolkienfonts.sty}, or the file that uses the package.

You may need to downoad the file \texttt{ansinew.enc}; it may be found through \href{http://ctan.org/search.html#byName}{http://ctan.org/}.
It goes in \verb.texmf/fonts/enc..

The advantage to this is that it involves nothing outside of the
document you are compiling. Also, the \verb.\pdfmapline. command is
part of pdf\/\TeX, and has been since 2004 (and \verb.\pdfmapfile. is
even older); any distribution of \TeX\ installed or updated in the last
six years will be able to deal with the syntax above.

On the other hand, it can be annoying to have to say that everywhere.
And this doesn't work at all if you decide you want to produce DVI files
and use a postprocessor such as \textsf{dvips}.

With Mac\TeX, I can
cause \textsf{dvips}, \textsf{dvipdfm}, and pdf\/\TeX\ to know about
these fonts by opening a Terminal window (command prompt) and typing
{\tt updmap -{}-enable Map=tolkien.map}.

With MiK\TeX, I need to say
\texttt{initexmf -{}-edit-config-file updmap}
and then add the line \texttt{Map tolkien.map} to the file which opens, then run \texttt{updmap} from the command line.

Something similar should work with most distributions; it is always wise to check and see what your distribution's documentation says about \texttt{updmap} before using it.

\section{Setup}

Using the package, once it's installed, is meant to be simple. Say
\verb.\usepackage.\linebreak[0]\verb.{tolkienfonts}. in the preamble.
This will define text commands (such as \verb.\Quenya.) and
environments (such as \verb.{quenya}.) which convert their argument or
enclosed text into another script:
\begin{quote}
\parbox{\halfwidth}{\raggedright 
\texttt{Sing \char`\\Quenya\{nam\char`\\'arie!\}}
\texttt{\char`\\par\char`\\begin\{quenya\}\\
ai! laurie lantar lassi s\char`\\'urinen!
\char92end\{quenya\}}
}\hfil
\parbox{\halfwidth}{\raggedright Sing \Quenya{nam\'arie!}
\par\begin{quenya}
ai! lauri\"e lantar lassi s\'urinen!
\end{quenya}
}
\end{quote}

If for some reason you want to produce tengwar without using
\texttt{tolkienfonts.sty} (for example, if you're using plain \TeX\ or
Con\TeX t), see \autoref{sec:plain}.

This package supports M\r ans Bj\"orkman's Sarati Eldamar and Valmaric Eldamar fonts, because they seem to be the only Sarati and Valmaric fonts available. It supports Dan Smith's Cirth Erebor font; when the first version of this package came out, this was the only font available that had all of the variant cirth seen in \cite{ori}. Even it does not have all the variant runes seen on \cite{thror}; the sunrune font was originally made just so I'd have a font that had everything. You can switch between sunrunes and Cirth Erebor with \verb|\sunrune| and \verb|\fonterebor|. Note that these commands must precede the runic commands they are to affect. \verb|\fonterebor| \verb|\Moria{balin| \verb|fundinul| \verb|\sunrune| \verb|uzbad| \verb|khazad-d\^umu}| produces \fonterebor \Moria{balin fundinul \sunrune uzbad khazad-d\^umu}; to get \fonterebor \Moria{balin fundinul} \sunrune \Moria{uzbad khazad-d\^umu}, you have to say \verb|\fonterebor| \verb|\Moria{balin| \verb|fundinul}| \verb|\sunrune| \verb|\Moria{uzbad| \verb|khazad-d\^umu}|. \sunrune

There are \emph{many} available Tengwar fonts on the internet. I have put a fair amount of time and effort into making sure my modes look good with Tengwar Annatar; tehtar placement might be suboptimal for other fonts. For this reason, \textsf{tolkienfonts} as is supports only this one tengwar font. For information on how to modify \textsf{tolkienfonts} to support a few other tengwar fonts, see \autoref{sec:fonts}.

\begin{table}[tp]
\caption{Modes, sorted by script and language} \label{table:modes}

\vbox{\offinterlineskip\halign to \hsize{\tabskip = 0pt plus 1fil
# \vrule height 10pt&& \vtop{\hsize= 0.2\textwidth\normalbaselines\tt\parindent=0pt  #\vrule width 0pt depth 5pt\relax}\cr
\vadjust{\hrule}&\rm Quenya&\rm  Sindarin&\rm  English&\rm  Dwarvish \cr
\rm Tengwar
&{Quenya\\ QuenyaA\\ Quenyagen\\ Quenyared}
&{Gondor\\ Arnor\\ Beleriand}
&{Orthmode\\ Mazarbul\\ Tehtamode\\ Phonemic}\cr
\rm Cirth
&{Eregion\\\color{red}fancyeregion}&{Daeron\\\color{red}fancydaeron}&{Erebor\\\color{red}fancyerebor}&{Moria\\\color{red}fancymoria}\cr
\rm Runic&&&\color{black}Hobbit\\\color{red}fancyhobbit&\cr
\rm Sarati&{Sarati\\SaratiA\\\color{red}barsarati\\\color{red}barsaratia}\cr
\rm Valmaric&Valmaric\\ ValmaricA\cr
}}
\end{table}

Tolkien had several ways of writing with runes, tengwar, and so on. For example, on Thror's map in \ment{The Hobbit}, \ment{the} is spelled \Hobbit{the}; on the title-page inscription for \ment{The Lord of the Rings}, \ment{the} is spelled \Erebor{dhE}. Different ways of writing with tengwar are called \ment{modes}; for simplicity I will use the word \ment{mode} to describe any Tolkien writing system. 

\textsf{tolkienfonts} supports several modes; each mode has an environment (like \texttt{\{quenya\}} above), and most have a text command (like \texttt{\char`\\Quenya} above) as well.
The available modes, arranged by language and writing system,
are shown in \autoref{table:modes}. The command names are shown; to get the environment names, simply set everything to lowercase.

Modes which are only available as environments are listed in red in
\autoref{table:modes}; these environments all involve active space
characters, which can lead to odd behavior.

You may also use most of the modes as package options (that is, say something like
\verb|\usepackage|\penalty100\verb|[quenya]{tolkienfonts}|); this will
cause the entire document to be typeset in the appropriate mode. Any
such option will define the \verb|\textlatin| command, which produces
text in Computer Modern.

\subsection{Differences between modes}

Most of the modes are based on one or a few specific Tolkienian texts. Often, there is only one (or no) mode for a given language/script combination; however, sometimes there are more.

I've only implemented sarati and valmaric modes for Quenya (there is a general phonemic sarati mode, but I thought it would be too annoying to deal with). The differences between the various sarati and valmaric modes are mostly cosmetic. 
\texttt{barsarati} (or \texttt{barsaratia}) is \texttt{sarati} drawn
with a bar: \begin{barsarati}{r\'umil tiriono}\end{barsarati} instead
of \begin{sarati}r\'umil tiriono\end{sarati}. Many cirth texts put
dots in the interword spaces: \Erebor{tHE\ lord\ ov\ tHE\ rings}. In
the normal cirth modes, you get these dotted spaces with \verb*.\ .
(backslash-space); \texttt{fancyerebor} and so forth put the dots in
the spaces automatically.

Sarati and valmaric quenya modes normally do not bother writing the
letter \ment{a} unless absolutely necessary; instead they have a special
tehta for a consonant followed by no vowel. \texttt{sarati} and
\texttt{valmaric} follow this convention (\texttt{calma} is
transliterated \Sarati{calma} or \Valmaric{calma}), but \texttt{saratia}
and \texttt{valmarica} do not (\SaratiA{calma} or \ValmaricA{calma}).

\texttt{Quenya} is the standard Quenya mode (as seen in \cite{namarie}
and described in \cite{appE}). It has several variations. \cite{appE}
says that you can omit the \ment{a}s when writing tengwar; since almost
no extant Quenya tengwar texts do this, \verb|Quenya| defaults to
writing \ment{a}s with tehtar \Quenya{calma} and \verb|QuenyaA| omits
them \QuenyaA{calma}.\footnote{If you really want \Quenya{clm}, type
\texttt{\char`\\Quenya\{clm\}}. I recommend against doing this unless
you've found some native Quenya speakers to read your document.}

\texttt{Quenyared} {is a mode identical to \texttt{Quenya} except that
(in pdf\/\TeX) the tehtar and punctuation are red (and the tengwar
are black): \Quenyared{tengwar ar tehtar}. For instructions on how to
set up a font with this kind of interesting color effects, see
\autoref{sec:color}.}

In the standard Quenya mode, it is impossible to correctly transcribe,
for example, ``Essenya Bilbo'' (my name is Bilbo); the closest you can
do is ``Essenya Mbilvo''. \texttt{Quenyagen} is a mode designed mainly
for Quenya, but in which it is possible to write Sindarin or Mannish
names such as \ment{Bilbo}.

The three Sindarin tengwar modes reflect writing styles of Beleriand, Gondor and Arnor. (\texttt{Beleriand} is based on \cite{durin} and \cite{elbereth}; \cite{durin} explicitly calls it the Mode of Beleriand. \texttt{Arnor} and \texttt{Gondor} are based on \cite{kingsletter}; I got the terms ``mode of Arnor'' and ``mode of Gondor'' from \cite{history}.) The mode of {Beleriand} has some of the same problems as the usual Quenya mode (\ment{Bilbo} can be written, but \ment{Marcho} cannot); it is possible to transcribe these Mannish names in the modes of Gondor and Arnor. (\texttt{Gondor} is a tehta mode \Gondor{gondor} and \texttt{Arnor} is a full mode \Arnor{arnor}. Otherwise they are much the same.)

There are many English modes because Tolkien spent a lot of time experimenting with writing English with strange alphabets. \texttt{Hobbit} and \texttt{Erebor} are the runic writing systems used in \ment{The Hobbit} and \ment{The Lord of the Rings}, respectively.

Two of the tengwar modes reflect in-universe regional variations: \texttt{Orthmode} is designed to replicate Aragorn's mode in \cite{kingsletter}, and \texttt{Mazarbul} is designed to replicate Ori's mode in \cite{ori}.

Tolkien rarely wrote English with tehtar; \texttt{Tehtamode} is designed to replicate the most well-known example (\cite{lotr}).

\texttt{Phonemic} is designed to replicate a number of long inscriptions: \cite{errantry}, \cite{luthien}, and \cite{bombadil}. The main difference between \texttt{Phonemic} and the others is that it is based on English pronunciation, not English spelling: in Aragorn's mode, \ment{their} and \ment{there} are transliterated \Orthmode{tHeir} and \Orthmode{tHerE}, while in the phonemic mode, both should be transliterated as \Phonemic{tHer}.

\section{What to put inside the mode commands}

\subsection{Text}

For Hobbit runes, nothing special need be done; just say
\texttt{\char`\\Hobbit\{The record of a year's journey made by Bilbo
Baggins of Hobbiton.\}} In other English modes, you have to change the
spelling a great deal to get it to work properly. See
Sections~\ref{sec:EO} and~\ref{sec:EP}.

For a Tolkienian language, you can generally get the right results by
typing the Quenya, Sindarin or Dwarvish phrase in lowercase letters.
(The \verb.\lowercase. or \verb.\MakeLowercase. commands can be useful
here.)

In the Mode of Gondor, you have to do one other thing. A \ment{w}
after a \ment{d}, \ment{g}, or \ment{ch} is transcribed in a special
way, which occasionally interferes with preceding vowel tehtar. The
simplest way to get the correct result is to replace all \texttt{dw}s,
\texttt{gw}s and \texttt{chw}s with \texttt{Dw}s, \texttt{Gw}s, and
\texttt{Chw}s or \texttt{'dw}s, \texttt{'gw}s, and \texttt{'chw}s.

Experts might need a few other uppercase letters.

In the Mode of Arnor, \ment{r} is almost always spelled~\Arnor{r}; however, it is occasionally spelled~\Arnor{R}, and I haven't figured out any underlying rule; so if you want~\Arnor{R}, type~\texttt{R} instead of~\texttt{r}.

In the Quenya modes, \texttt{N}/\texttt{ng}, \texttt{S}/\texttt{th}
and \texttt{V}/\texttt{w} may be used in words that are pronounced
with \ment{n},~\ment{s} or~\ment{v} but are spelled as if they used
\ment{\ng},~\ment{th} or~\ment{w}. There are three different \ment{h}s
in Quenya; you get \Quenya{H} with \texttt{H} or just before
\texttt{r}/\texttt{l}, you get \Quenya{ch} with \texttt{ch} or just
before \texttt{t}, and you get \Quenya{h} with \texttt{h} otherwise.
(See \autoref{sec:Qh}.)

To the extent that it is possible, I have tried to ensure that text
which is spelled correctly in a given mode will produce appropriate
results in other modes configured for the same language. For example,
\verb.D.,~\verb.G., \verb.C., and~\verb.R. may be used instead of
\verb.d.,~\verb.g., \verb.c., and~\verb.r. in the mode of Beleriand,
even though there's no reason to.

Otherwise, uppercase letters will often produce odd results; don't use
them unless you know what they do and want that effect to occur.

Some rare Sindarin words begin with~\ment{ng}. \cite[\unskip,
p.~63]{Road} indicates that this \ment{ng} is pronounced as in
\ment{sing}. If you believe that the word you are typing instead begins
with the sound in \ment{finger}, use \texttt{'ng} instead
of~\texttt{ng}.

If you want to transcribe ``Elessar Telcontar aran Gondor ar Arnor''
(``Elfstone Strider, King of Gondor and Arnor''), you should probably
pick one mode and stick to it, even though ``Elessar Telcontar'' is in
Quenya and the titles are in Sindarin. Transcribing ``Elessar
Telcontar'' with \texttt{Quenya} and ``aran Gondor ar Arnor'' with
\texttt{Gondor} is confusing, and Tolkien didn't do it
(see \cite{kingsletter}).

This means that you often need to be able to write names in modes
designed for other languages.

If you use Sindarin names with \ment{dh}, \ment{y}, \ment{d} (except \ment{nd}, \ment{rd},~\ment{ld}), \ment{b} (except~\ment{mb}),  or  \ment{g} (except~\ment{ng}),  you need to use \texttt{Quenyagen} and not \texttt{Quenya}. Otherwise, Quenya and Sindarin names can be used in each others' modes with almost no special spelling. I've tried to deal gracefully with the fact that \ment{y} is a vowel in Sindarin and a consonant in Quenya. Thus, in these modes, \texttt{y} produces an appropriate character for a vocalic~\ment{y}; the correct character for a consonantal \ment{y} may be produced by either a \texttt{Y} or a \texttt{y} followed by a vowel.

There is one subtle point: if a Sindarin word starts with an~\ment{i} followed by another vowel, the \ment{i} is pronounced (and transcribed as) a consonantal~\ment{y}. So to mention Iorhael in a Quenya text, it is best to say \verb|\Quenyagen{yorhael}| \Quenyagen{yorhael}; if for some reason you need to start a Quenya word with \ment{io-} in a Sindarin text, type \texttt{'io} and not just~\texttt{io}.

We don't have any attested examples of Westron (or English, or Dwarvish)
names in Quenya or Sindarin texts; make something up. A (reasonable)
guess is to spell things phonetically, using the conventions
described in \cite{appE}. The symbols
\Gondor{j},~\Gondor{sh},~\Gondor{zh} for \ment{j},~\ment{sh}, \ment{zh}
(in \ment{a\textbf{z}ure}) are available in \texttt{gondor},
\texttt{arnor} and \texttt{quenyagen}. The symbol \Gondor{tj} for
English \ment{ch} (in \ment{\textbf{ch}ur\textbf{ch}}) is also
available, but you have to get it with \texttt{tj}, which was the best
letter combination available that I did not want to reserve for
Sindarin~\ment{ch} (\Gondor{ch}~or~\Beleriand{ch}).

\subsection{Accented letters}

Tolkien used three diacritical marks with the Latin alphabet: \'e, \^e
and~\"e. You can get appropriate transcriptions by using the appropriate
\TeX\ accent commands \verb.\'.,~\verb.\^. or~\verb.\".. In any mode
designed for a Tolkienian language, you can also get appropriate (if
sometimes different) transcriptions for acute or circumflex accents
by doubling the vowel: \texttt{uu}
instead of \verb|\'u|,~\verb|\^u|. Dieresis marks aren't actually necessary when
writing Elvish; they can be ignored. (Nothing bad will happen if a
\verb.\".\<\emph{vowel}\> winds up in a Quenya text by accident.)

\subsection{Punctuation}

In transcribing Elvish to the Latin alphabet, a dash indicates
compound words; an apostrophe indicates a letter that is pronounced
but not written. Neither of these are written in the tengwar text:
\ment{lumenn'} becomes \mbox{\Quenya{lumenn'}}, not
\mbox{\Quenya{lumenn''}}. However, occasionally these marks make a
difference to tengwar spelling: \mbox{\ment{Pant-hael}} is not
pronounced \ment{panth-ael}, and so should be spelled
\mbox{\Gondor{pant-hael}}, not \mbox{\Gondor{panthael}}.

In all fonts, we can force \Gondor{pant-hael} by putting
\verb.-.,~\verb.'., \verb.{}., or~\verb.\-. between the \texttt{t}
and~\texttt{h}. \verb.-.~and~\verb.'. behave slightly differently.

Most of the modes have special rules at the start or end of words. For
example, in the mode of Arnor, if a word starts with \ment{i}
followed by another vowel, the \ment{i} is pronounced as a \ment{y}
and transcribed with a \tengfont{l}; so \ment{iarwain} is transcribed
\Arnor{iarwain}, but \ment{Imladris} is transcribed as
\Arnor{imladris}.

A hyphen (but not an apostrophe) restores start-of-word and
end-of-word behavior on either side; so \verb.\Arnor{ben-ion}.
produces {\Arnor{ben-ion}}, but
\verb.\Arnor.\penalty0\verb.{ben'ion}. produces
{\Arnor{ben'ion}.}




\begin{table}[t]
\caption{Tengwar Punctuation}\label{table:punct}
\begin{tabular}[t]{|c|c|l|}
\cline{3-3}
\multicolumn{2}{c|}{}&Meaning\\\hline
\Tehtamode{?}   &\texttt{?}     &Question mark (from \cite{namarie})    
\\ \hline
\Tehtamode{!}   &\texttt{!}     &Exclamation point (from \cite{namarie})    
\\ \hline
\Tehtamode{,}   &\texttt{,}     &Short pause    \\ \hline
\Tehtamode{.}   &\texttt{.} or \texttt{;}       &Medium pause   
\\ \hline
\Tehtamode{:}   &\texttt{:}     &\\\cline{1-2}
\Tehtamode{,.}  &\texttt{,.}    &Long pause     \\\cline{1-2}
\Tehtamode{.,}  &\texttt{.,}    &   \\\hline
\Tehtamode{,.,}	&\texttt{,.,}    &Very long pause/paragraph break    
\\\cline{1-2}
\Tehtamode{..}	&\texttt{..}    &   \\\hline
\Tehtamode{(}   &\texttt{(} or \texttt{)}   &Parenthesis (from \cite{kingsletter}, \cite{thorin})   
\\ \hline
\Tehtamode{--}  &\texttt{-{}-}  &\\\cline{1-2}
\Tehtamode{---}&\texttt{-{}-{}-}    &Decoration or end of inscription       
\\\cline{1-2}
\Tehtamode{=}   &\texttt{=} &   \\\hline
\end{tabular}
\begin{tabular}[t]{|c|c|}
\hline
\Gondor{``}&\texttt{``} or \texttt{`}\\\hline
\Gondor{''}&\texttt{"} or \texttt{''}\\\hline
\Gondor{(`}&\texttt{[} or \texttt{(`}\\\hline
\Gondor{)`}&\texttt{]} or \texttt{)`}\\\hline
\Gondor{;`}&\texttt{;{}`}\\\hline
\Gondor{.`}&\texttt{.{}`}\\\hline
\Gondor{,`}&\texttt{,{}`}\\\hline
\Gondor{?`}&\texttt{?{}`}\\\hline
\Gondor{!`}&\texttt{!{}`}\\\hline
\end{tabular}
\end{table}

Visible punctuation is also available. In general, the Elves used
(varying numbers of) dots to indicate pauses in speech. So in all
fonts, the punctuation characters \texttt{.,:;} produce some number of
dots; for decorative effects, or at the end of inscriptions, you can
repeat symbols to get clusters of more dots.

In the tengwar fonts, special symbols for \texttt{?}\ and \texttt{!}\
are used (they come from \cite{namarie}). In the other fonts, these
symbols are again dots, since I didn't know what else to make them.

Again in the tengwar fonts only, there are available symbols for
``Elvish parentheses''; furthermore, you can get Latin punctuation,
including quote marks. See \autoref{table:punct}.

I've done some virtual-font tricks so that the elvish punctuation (but
not the Latin punctuation) has small spaces on the left and right, not
just on the right. When written ragged right, the spaces on both sides
are the same size; however, for technical reasons all of the stretch
and shrink has to go on the right-hand side of the punctuation signs.
(Except in the active space modes.) If you don't like this effect, see
\autoref{sec:spacing}.

\subsection{Numbers}\label{sec:lastneeded}

In Middle-Earth, numbers are written with the least significant digit
first (this is backwards relative to the way we write). Also, Elves
and Dwarves often wrote numbers in base 12 instead of base 10. (See
\cite{linnumbers}.)

We don't know how to write numbers in Valmaric or Cirth. (We do have
some digit tables for Sarati.) The numbers 3, 4, 5, 6 show up in
\cite{ori}, but it's hard to extrapolate upwards (or down to 0).

You can get numbers in Sarati or Tengwar in two ways:

\begin{itemize} \item Type them in directly: Bilbo invited
\verb.\Orthmode{441}. guests to his special party, and Thorin set out
for the Lonely Mountain with one hobbit, one wizard, and
\verb.\Orthmode{01}. dwarves, or a total of \verb.\Orthmode{21}.
companions. (The commands \verb.\ten. and \verb.\eleven. produce the
appropriate digits.) \item Use the \verb.\reversedigits. macro. This
macro does the work of converting a base-10 number written with most
significant digit first to a base-whatever number written with the
least significant digit first. (The default default base is 10. You
can get other bases up to 12 by either using an optional first
argument or by renewing the command \verb.\defaultbase..)

For example, \verb.\Orthmode{\reversedigits{16}}. produces
\Orthmode{\reversedigits{16}}, and

\medskip

\begin{quote}
\noindent\verb.\Orthmode{\reversedigits[12]{16}}.

\medskip

\noindent \verb.\renewcommand{\defaultbase}{12}.\\
\verb.\Orthmode{\reversedigits{16}}.
\end{quote}

\medskip

\noindent both produce \Orthmode{\reversedigits[12]{16}}.
\end{itemize}

\section{Extra symbols}

Long tengwar look like \Quenya{th|}. These are used on the One Ring, and
also in English mode as abbreviations for ``the'' and ``of''.
Occasionally they are used elsewhere as variants. They can be accessed
in a tengwar mode by placing a | after the code for the corresponding
tengwa from Grade~1, 2, 3, or~4. So, for example, \verb.\Gondor{p| dh|}.
produces \Gondor{p| dh|}.

{
\begin{table}[tp]
\caption{\ment{s}-curls}\label{table:curls}
\tt
\begin{tabular}[t]{|c|l|}
\hline
\rm Curl & \rm Code \\\hline
\tengalt{\color{blue}1}\Quenya{\char0}&\chr0 {\rm or} \char92 s \\
\Quenya{{\color{blue}ch}\char11}&\chr11 \\
\tengalt{\color{blue}1}\Quenya{\char12}&\chr12 \\
\tengalt{\color{blue}2}\Quenya{\char13}&\chr13 \\
\Quenya{{\color{blue}l}\char14}&\chr14 \\
\hline
\end{tabular}
\begin{tabular}[t]{|c|l|l}
\cline{1-2}
\rm Curl & \rm Code \\\cline{1-2}
\Quenya{{\color{blue}c}\char5}&\chr5 \\
\tengalt{\color{blue}1}\Quenya{\char6}&\chr6 \\
\Quenya{{\color{blue}l}\char7}&\chr7 \\
\tengalt{\color{blue}1}\Quenya{\char8}&\chr8 \\
\tengalt{\color{blue}z}\Quenya{\char9}&\chr9 \\
\cline{1-2}
\end{tabular}
\end{table}
}

Tolkien often used curls and flourishes as an abbreviation for
\ment{s}: \Quenya{p\s}, \Quenya{q\s}, \SaratiA{t\s}, \ValmaricA{t\s},
and so on. To get these in a tengwar, sarati or valmaric font, you can
use the commands \verb.\s. or \verb.\char0..

With sarati and valmaric, flourishes may be used on either side of a
letter. \verb.\s.~by default produces the flourish that attaches to
the left side of the following symbol; at the end of a word, the
flourish which attaches to the right side of the preceding symbol is
substituted automatically.

In a Sindarin mode, or English phonemic mode, that's all you have to do.
(The fonts will switch to a good \ment{s}-curl automatically. If you
don't like the curl the font picked out, see
\autoref{table:curls}.\footnote{If you say \texttt{t\char`\\char11},
the fonts should usually type out \Quenya{t\char11} instead of
\Quenya{t'\char11}. That is, some fonts have special versions of some
letters for use with curls, which usually appear automatically. If they
don't, you can use
\texttt{\char`\\tengfont}\slash\texttt{\char`\\tengalt} to get them; if
you don't want them to happen, you can say  \texttt{t'\char`\\char1}.}

In Quenya tengwar modes, sometimes you want to get things like
\Quenya{t\sa}. From \cite{appE}, it is clear that \Quenya{t\sa} means
\ment{tsa} and not \ment{tas}; that is, the \ment{s}-curl should happen
only if~\ment{s} occurs directly after another consonant. In \cite{qwl},
this happens only in the combinations \ment{ts,~ps, ls, ms, ns, rs,}
and~\ment{x}.

Since those letters are all (except \Quenya{l}~\ment{l}) from Series~I
or~II, we can use the same \ment{s}-curl for all of them. So the
commands \verb.\sa., \verb.\se.,\,\ldots,\verb.\su. after~\verb.t.,
\verb.p., \verb.m., \verb.n.,~\verb.r., and some other letters, produce
this curl and a tehta. (After~\verb.l., or many other letters, they
produce a silme nuquerna plus the appropriate tehta. If you really want
{\Quenya{la\char14}}, you have to use \verb.la\char14..)

In English orthographic modes, similarly, if a word ends with a
\emph{silent} E, and you want to attach a \ment{s}-curl, use the
command \verb.\es. instead of~\verb.\s..

In all tengwar modes, you can get the short and long carriers~\Quenya{\i} and~\Gondor{\j}
with the commands \verb.\i. and~\verb.\j., and the variant r\'omen
\Phonemic{\R} with~\texttt{\char`\\R}. There are some other symbols
present in the base fonts which are not accessible in the normal fonts;
to get them, you need to use \verb.\tengfont.. (See
\autoref{sec:DSmode}.)

\section{English orthographic modes}\label{sec:EO}

When writing Elvish, spelling always dictates pronunciation; aside
from a few leftover Quenya spelling rules, pronunciation dictates
spelling.

In English, this is not the case. There are two ways to make an
English mode in a non-Latin alphabet: by basing it on English
spelling, and by basing it on English pronunciation. The first kind
are referred to as \ment{orthographic} modes, and the second kind are
referred to as \ment{phonemic} modes.

In an ``orthographic'' mode, you can determine the Latin spelling from
the tengwar spelling,\footnote{Except for \ment{c/k} and, sometimes,
\ment{i/y} and~\ment{u/w}.} but you can't determine the tengwar
spelling from the Latin spelling. Extra information is present: for
example, Tolkien used separate letters for silent and
normal~\ment{e}s. While this would be very nice for any hypothetical
Elves who might come here and try to decipher our writing, it does
mean that the input has to be a bit more complicated than just normal
English text.

To the extent that it is possible, I have attempted to arrange things
so that the argument of \verb.orthmode., \verb.mazarbul.,
\verb.tehtamode. and \verb.erebor. can be a case change away from
normal English spelling---that is, you get \Orthmode{danCE}, the
correct transliteration of \ment{dance}, with \verb.\Orthmode{danCE}.
and not \verb.\Orthmode{dans}..

I have also tried to make sure that you can use the same input for all
four modes. Not all of the following rules actually \emph{do} things
in all the modes.

You will need to use the following special spelling rules:

\newcounter{temp} \begin{enumerate}

\item A dipthong is two vowels in a row which are run together into
one syllable. These are often, but not always, written in a special
way. For example, in the mode of Beleriand, \ment{taur} (one syllable)
is spelled \Beleriand{taur}, but \ment{Tuor} (two syllables) is
spelled \Beleriand{tuor}.

In the English modes, you can prevent two vowels from forming a
dip\-thong by putting something between them (\verb.'., \verb.-.,
\verb.{}.). You can also prevent a dipthong by using~\verb.\".:
\verb.ro\"ac. produces \Tehtamode{ro\"ac}, not
\Tehtamode{roac}.\footnote{This sometimes---but not always---works in
the Elvish modes. In Quenya and Sindarin, every pair of vowels is either
\emph{always} a dipthong or \emph{never} a dipthong, so the
\texttt{\char`\\"} is superfluous unless something strange is going on.
It would have been complicated to make \texttt{t\char`\\"ai}
transliterate as \Quenya{ta{}i} (not a dipthong) instead of
\Quenya{t\"ai} (dipthong), so I didn't bother. \texttt{ta\char`\\"i}, or
\texttt{t\char`\\"ai} in a Sindarin mode, should still prevent a dipthong.}

If a dipthong is available, you can force it to appear by replacing
the second vowel with its capital.\footnote{This rule may or may not
take precedence over \texttt{\char92"}. Being consistent on this
particular point was not worth the effort.} (\ment{w} is considered to
be a vowel; \ment{beware} should be transcribed as \Orthmode{bewarE},
but \ment{dewy} should be transcribed as \Orthmode{deWy}. To get the
second one, use \texttt{W}.)

If you do not do any of these things, what happens depends on the pair
of vowels and the mode; experiment if you're curious.

\item Apostrophes should usually be skipped: if there's a short form
of \ment{nt}, you want to use it in \ment{can't}.

\item Vocalic \ment{y} should be written with \texttt{y}; consonantal
\ment{y}, or \ment{y} as the second letter of a dipthong, should be
written with~\texttt{Y}.

\item Silent \ment{e}s should be input as \texttt{E}, not~\texttt{e}.
(Note that the \ment{e}s in \ment{dale, water, walked} are considered
to be silent.)

\item A soft \ment{c} (as in \ment{princess}) should be input as
\texttt{C}, not~\texttt{c}. (A hard \ment{c} in \ment{cat} should be
left as~\texttt{c}.)

\item There are two different \ment{ng}s in English: the \ment{ng} in
\ment{sing} and the \ment{ng} in \ment{finger}. Since the first one is
more common, you get it with \texttt{ng}; to get the \ment{ng} in
\ment{finger}, you should type \texttt{Ng}.

\item English has two kinds of \ment{ch}: the soft \ment{ch} in
\ment{church}, and the \ment{ch} pronounced as \ment{c} which is so
common in Greek loan words like \ment{echo}. German and Sindarin loan
words contain a third kind of \ment{ch}. The first should be spelled
with \texttt{ch}.

The other two should be spelled with \Orthmode{CH}/\Erebor{\small
cH}~\texttt{CH}/\texttt{cH}/\texttt{kH}/\texttt{KH} or
\Orthmode{Ch}/\Erebor{\small Ch}~\texttt{Ch}/\texttt{Kh}. Which one is
which is up to you; Tolkien uses \Orthmode{CH} in \ment{Christmas} in
\cite{brogan}, but uses \Orthmode{Ch} in \ment{Michael} in
\cite{endorion}.

I prefer to use \Gondor{ch}/\Daeron{\small ch}~\texttt{Ch} for the
Sindarin \ment{ch}, so that you use the same symbol in the Sindarin
and English modes. This leaves \Orthmode{CH}/\Erebor{\small
CH}~\texttt{CH} for \ment{Christmas}.

Since words like \ment{backhand} are so common in English, \texttt{kh}
just produces \ment{k+h}.

\item There are two different \ment{th} sounds in English, which
Tolkien describes as ``the voiceless \ment{th} of English in
\ment{thin cloth}'' and ``the voiced (soft) \ment{th} of English
\ment{these clothes}''. You get the first one with \texttt{th}; you
get the second one with~\texttt{dh} or~\texttt{tH}.

This means that a normal \ment{dh} (in, say, \ment{childhood}) needs
to be spelled as \texttt{d'h} or \verb.d{}h.. \ment{dh}s in Elvish
names (such as \ment{Maedhros}, \ment{Caradhras}) are intended to be
pronounced as this voiced \ment{th} (it's not \ment{Carad-hras}), so
do not separate them.

\setcounter{temp}{\value{enumi}} \end{enumerate}

The following rules only apply to tengwar (not cirth):

\begin{enumerate} \setcounter{enumi}{\value{temp}}

\item A \ment{w} which is silent or pronounced in combination with
another consonant (\ment{two}, \ment{twenty}) or a silent
\ment{u} after a consonant (\ment{quiet}, and probably also \ment{build}, \ment{guest},)
should be written with a twist tehta. This should not be done in
compounds such as \ment{westward}, so you have to tell
\textsf{tolkienfonts} that you want the \ment{u} or \ment{w} to become
a twist by using the capital version.\footnote{Except after
\texttt{q}. \texttt{q} always produces the proper glyph or glyphs for \ment{qu}, since there isn't a known way to write just \ment{q} with any Tolkienian script. Any \texttt{u} after a \texttt{q} should silently vanish.} I'm pretty sure twists are used only
in the combinations \ment{gu}, \ment{bu}, \ment{tw}, \ment{dw},
\ment{sw}, \ment{thw}, \ment{gw}, and \ment{schw}, plus \ment{Chw} if
you include Sindarin names. These are the only combinations where
\texttt{W}, \texttt{U} produce twists.

In the tehta mode, you have to worry about vowel and twist tehtar
interfering.

As in the mode of Gondor, \texttt{T}, \texttt{B}, \texttt{D},
\texttt{S}, and \texttt{G} behave exactly as their lowercase
counterparts, except that they do not take vowel tehtar. Since
\texttt{C} is too busy representing soft \ment{c} to represent
\Tehtamode{c} with no tehtar, if you want a \Tehtamode{chW} or
\Tehtamode{ChW} just after a vowel, I'm afraid you'll just have to put
a \verb.'. or \verb.{}. before it.

\item Tolkien abbreviated the common words \ment{and}, \ment{of},
\ment{the}, and \ment{of the} as \Orthmode{ndE}/\Orthmode{nd},
\Orthmode{v|}, \Orthmode{d|}, and \Orthmode{vv|}, instead of using the
longer spellings \Orthmode{and}/\Tehtamode{and},
\Orthmode{of}/\Tehtamode{of}, \Orthmode{tHe}/\Tehtamode{tHe}, and
\Orthmode{of tHe}/\Tehtamode{of tHe}. You can get the abbreviations
with \verb.ndE., \verb.v|., \verb.tH|./\verb.d|., and \verb.vv|., as
well as with the commands \verb.\&., \verb.\of.,
\verb.\The./\verb.\THE., and
\verb.\ofthe..\footnote{\texttt{\char92the} and \texttt{\char92and}
are predefined \TeX\ macros, so I did not want to redefine them.}

\item No English text uses \Tehtamode{Rh}. It's anyone's guess what
Tolkien would have used it for. But a reasonable guess is \ment{rh} in
\ment{rhyme}, \ment{rhythm}, (or \ment{Rhovanion}),\footnote{The other
obvious possibilities are \ment{rr}, \ment{rd} and that it simply
isn't used. \ment{rd} and \ment{rr} are spelled \Tehtamode{rd} and
\Tehtamode{RR} in all three drafts of \cite{kingsletter}, and
\ment{rh} has to be spelled with \emph{something}.} and so you get it
with \texttt{Rh}. (Words like \ment{neighborhood}, which shouldn't be
spelled this way, are too common for me to let \texttt{rh} produce
\Orthmode{Rh}.)

\item In tehta mode, never use \texttt{N} unless the next letter is
\texttt{g} or \texttt{G}. For technical reasons,
\<\ment{vowel}\>+\texttt{Ng} and \<\ment{vowel}\>+\texttt{mg} produce
the same thing in tehta mode; in the unlikely event that you actually
want a word with \ment{mg} in it,\footnote{\cite{ispell} knows two
such words: \ment{broomgrove} and \ment{slumgullion}.} you'll want to
separate them \verb.m{}g..

\item In tehta mode, (at least, with Tengwar Annatar) you probably
want to get \Tehtamode{Wh} with \texttt{Wh} after vowels because it
will place the tehtar better. You usually want to get \Tehtamode{CH}
with \texttt{CH} (not~\texttt{c|}) for the same reason.

\end{enumerate}

If you want to go further and spell \ment{age} and \ment{his} as they
are pronounced (as \ment{aje}, \ment{hiz}), you will need to spell
them in the input as \texttt{ajE}, \texttt{hiz}. (Tolkien used
\ment{az} for \ment{as} in \cite{lotr}, but spelled \ment{his} and
\ment{bridge} as \ment{his} and \ment{bridge} (not \ment{hiz} and
\ment{bridje}) in \cite{kingsletter}.)

\section{English phonemic mode}\label{sec:EP}


\begin{table}
\def\gobble#1{}
\caption{English full phonemic mode}\label{table:EP}
\parbox{\hsize}{\parindent = 3ex
This table shows the tengwar and tehtar available in the phonemic mode.

Symbols shown in black or purple are probably used in the mode of
\cite{errantry}, \cite{bombadil} and \cite{luthien}. Symbols shown in red or orange are attested
in some other example (in most cases, \cite{thorin}, \cite{treebeard} or
\cite[39]{DTS}) but are probably not used in the mode of
\cite{errantry}. Symbols in green
are not attested in any phonemic examples that I know of. (Blue and
orange symbols seem to be variants of other symbols.
{\color{purple}\Phonemic{kh}} is unattested, and {\color{purple}\Phonemic{zh}} is attested only in \cite{thorin}.)

It is unclear whether certain letters are best transcribed as
\Phonemic{\'o},~\Phonemic{\'a} or \Phonemic{\^o},~\Phonemic{\^a}.
Therefore, both sets are available.

Note that \Phonemic{V} is not available in all fonts; \Phonemic{U} will
be substituted if necessary.

\texttt{e} produces \Phonemic{\.e} if the next tengwa is in
Series~III, as in the Mode of Beleriand; \texttt{\char`\\.e} may be used otherwise.
}

\bigskip

\begin{tabular}{|ccl|ccl|ccl|ccl|}\hline
\gobble{1}\Phonemic{t}  &\texttt{t} &\textbf{t}o &
\gobble{q}\Phonemic{p}  &\texttt{p} &\textbf{p}ass&
\gobble{a}\Phonemic{ch} &\texttt{ch}&bran\textbf{ch}  & %Bombadil II
\gobble{z}\Phonemic{k}  &\texttt{k} &\textbf{c}all
\\
\gobble{2}\Phonemic{d}  &\texttt{d} &ha\textbf{d}  &
\gobble{w}\Phonemic{b}  &\texttt{b} &\textbf{b}uilt&
\gobble{s}\Phonemic{j}  &\texttt{j} &oran\textbf{g}e &
\gobble{x}\Phonemic{g}  &\texttt{g} &\textbf{g}ilded
\\
\gobble{3}\Phonemic{th} &\texttt{th}&\textbf{th}ink  & %Bombadil II
\gobble{e}\Phonemic{f}  &\texttt{f} &\textbf{f}or&
\gobble{d}\Phonemic{sh} &\texttt{sh}&fi\textbf{sh} &%Bombadil I
\color{purple}\gobble{c}\Phonemic{kh}&\texttt{kh} &lo\textbf{ch}
\\
\gobble{4}\Phonemic{dh} &\texttt{dh,tH}&\textbf{th}ere &
\gobble{r}\Phonemic{v}  &\texttt{v} &ri\textbf{v}er&
\color{purple}\gobble{f}\Phonemic{zh}  &\texttt{zh} &occa\textbf{s}ion & %Thorin
\color{green}\gobble{v}\Phonemic{gh} &\texttt{gh} &
\\
\gobble{5}\Phonemic{n}  &\texttt{n} &i\textbf{n} &
\gobble{t}\Phonemic{m}  &\texttt{m} &\textbf{m}erry&
\color{red}\gobble{g}\Phonemic{\~n}   &\texttt{nY}    && % DTS 39
\gobble{b}\Phonemic{ng} &\texttt{ng}    &lo\textbf{ng}%Bombadil I
\\
\gobble{6}\Phonemic{r}  &\texttt{r} &fo\textbf{r}   &
\gobble{y}\Phonemic{u}  &\texttt{u} &p\textbf{u}ll, t\textbf{o}&
\gobble{h}\Phonemic{U}  &\texttt{U} &\textbf{u}nder &%Bombadil I
\gobble{n}\Phonemic{a}  &\texttt{a} &c\textbf{a}rgo
\\\hline
\gobble{7}\Phonemic{R}  &\texttt{R} &me\textbf{rr}y  &
\color{green}\gobble{u}\Phonemic{Rh} &\texttt{Rh}& &
\gobble{j}\Phonemic{l}  &\texttt{l} &{\textbf{l}oad}&
\color{green}\gobble{m}\Phonemic{L}  &\texttt{L}& 
\\
\gobble{8}\Phonemic{s}  &\texttt{s} &acro\textbf{ss}  &
\color{red}\gobble{i}\Phonemic{S} &\multicolumn{2}{l|}{\texttt{S}, \texttt{C}} & %Thorin
\gobble{k}\Phonemic{z}  &\texttt{z} &wa\textbf{s}&
\color{green}\gobble{,}\Phonemic{Z} &\texttt{Z}& 
\\
\gobble{9}\Phonemic{h}  &\texttt{h} &\textbf{h}e   &
\color{green}\gobble{o}\Phonemic{hw} &\texttt{hw}& &
\color{red}\gobble{l}\Phonemic{Y} &\texttt{Y}&& %Thorin
\gobble{.}\Phonemic{o}  &\texttt{o} &{\textbf{o}range}
\\
\gobble{\char'176}\Phonemic{y}  &\texttt{y}     &\textbf{y}ellow   &
\tengalt{\char'277}&\texttt{wh}     &\textbf{wh}at & %Bombadil II
\gobble{\char'263}\Phonemic{w}  &\texttt{w}&\textbf{w}as &
&&
\\
\color{blue}\gobble{\char'176}\Phonemic{\j}  &\texttt{\char`\\j}     &\textbf{y}ou   & % Bombadil II-consistently.
& &&
\color{blue}\gobble{\char'263}\Phonemic{W}  &\texttt{W}&\textbf{w}ith &
&&
\\
\hline\end{tabular}
\par\medskip\par 
\begin{tabular}{|ccl|ccl|ccl|ccl|ccl|}\hline
\Phonemic{A}  &\texttt{A,E,I,O}   &\multicolumn{1}{l|}{\textbf{a}, \textbf{o}f}&
\gobble{n}\Phonemic{a}  &\texttt{a} &c\textbf{a}rgo&
\gobble{.}\Phonemic{o}  &\texttt{o} &{g\textbf{o}ndola}
\\
\Phonemic{ee}    &\texttt{iy, ee}    &betw\textbf{ee}n&
\Phonemic{ai}    &\texttt{ay, ii}        &br\textbf{igh}t&%Bombadil I
\Phonemic{oy}    &\texttt{oy}        &j\textbf{oy} %Luthien
\\
&&&
\Phonemic{au}    &\texttt{au, aw}        &d\textbf{ow}n&%Bombadil I
\Phonemic{ou}    &\texttt{ow, oe}    &yell\textbf{ow}
\\
\Phonemic{i} &\texttt{i}     &\textbf{i}n, h\textbf{e}*&
\Phonemic{aa}    &\texttt{aa}    &h\textbf{a}t& %Bombadil I
&&
\\
&&&
\Phonemic{\'a} &\texttt{\char`\\'a} &f\textbf{a}ther &
\Phonemic{\'o} &\texttt{\char`\\'o} & l\textbf{aw}
\\
\hline
\Phonemic{e'}  &\texttt{e}     &m\textbf{e}rry&
\gobble{y}\Phonemic{u}  &\texttt{u} &p\textbf{u}lled, t\textbf{o}*&
\gobble{h}\Phonemic{U}  &\texttt{U} &\textbf{u}p %Bombadil I
\\
\color{blue}\Phonemic{\.e}    &\texttt{\char`\\\char`\.e}        &f\textbf{ea}ther& %Bombadil II
\Phonemic{uw}    &\texttt{uw, ue}        &bl\textbf{ue}&%Bombadil I
\color{blue}\Phonemic{V}    &\texttt{V}        &\textbf{u}nder%Bombadil I
\\
\Phonemic{ei}    &\texttt{ey, ae}        &l\textbf{ay}&
\Phonemic{uu}    &\texttt{uu}   &bl\textbf{ue}&%Bombadil II
&&
\\
\hline
\end{tabular}

\par\medskip\par

{\footnotesize *H\textbf{e} and t\textbf{o} should be spelled with \Phonemic{i} and \Phonemic{u} only when unstressed; otherwise, use \Phonemic{ee} and \Phonemic{uu}. \par}

\par\medskip\par %\let\texttt\Phonemic
Vowels and tehtar\par\medskip\par
\begin{tabular}{|c@{ }l|c@{ }l|c@{ }l|c@{ }l|c@{ }l|c@{ }l|c@{ }l|c@{ }l|}
\hline
\tengfont{n}                            & \texttt{a} &
\tengfont{]}                            & \texttt{e} &
\tengfont{`}                            & \texttt{AEIO} &
\tengfont{.}                            & \texttt{o} &
\tengfont{y}                            & \texttt{u} &
\tengfont{h}                            & \texttt{U} &
\tengfont{\char`\~}                     & \texttt{\j} 
\\
\tengfont{n\char213}                    & \texttt{ay/ai, ii} & 
\tengfont{]\char214}                    & \texttt{ey/ei, ae} &
\tengfont{`\char215}                    & \texttt{iy, ee} & 
\tengfont{.\char214}                    & \texttt{oy/oi} &
\color{green}\tengfont{y\char213}       & \texttt{uy/ui} &
\color{red}\tengfont{h\char213}         & \texttt{Uy} &
%\color{green}\tengfont{l\char214}       & \texttt{Yy} &
\color{red}\tengfont{\char`\~\char215}  & \texttt{yy} 
\\ 
\tengfont{n\char233}                    & \texttt{aw/au} &
\color{red}\tengfont{]\char234}         & \texttt{ew/eu} &
\color{red}\tengfont{`\char235}         & \texttt{iw/iu} &
\tengfont{.\char234}                    & \texttt{ow/ou, oe} & 
\tengfont{y\char233}                    & \texttt{uw, ue} &
\color{green}\tengfont{h\char233}       & \texttt{Uw} &
%\color{green}\tengfont{l\char234}       & \texttt{Yw} &
\color{red}\tengfont{\char`\~\char235}  & \texttt{yw} 
\\ 
\tengfont{nT}                           & \texttt{\char`\\\char`\.a, aa} & 
\color{blue}\tengfont{]G}                           & \texttt{\char`\\\char`\.e} & 
\tengfont{`B}                           & \texttt{\char`\\\char`\.\i, i} & 
\color{green}\tengfont{.G}              & \texttt{\char`\\\char`\.o} &
\color{green}\tengfont{yT}              & \texttt{\char`\\\char`\.u} &
\color{green}\tengfont{hT}              & \texttt{\char`\\\char`\.U} &
%\color{green}\tengfont{lG}              & \texttt{\char`\\\char`\.Y} &
\color{blue}\tengfont{\char`\~B}        & \texttt{\char`\\\char`\.\j} 
\\ 
\tengfont{nR}                           & \texttt{\char`\\\char`\'a} &
\color{red}\tengfont{]F}                & \texttt{\char`\\\char`\'e} &
\color{red}\tengfont{`V}                & \texttt{\char`\\\char`\'i} &
\tengfont{.F}                           & \texttt{\char`\\\char`\'o, oo} &
\tengfont{yR}                           & \texttt{\char`\\\char`\'u, uu} &
\color{green}\tengfont{hR}              & \texttt{\char`\\\char`\'U} &
%\color{green}\tengfont{lF}              & \texttt{\char`\\\char`\'Y} &
\color{green}\tengfont{\char`\~V}       & \texttt{\char`\\\char`\'y} 
\\
\color{blue}\tengfont{nY}               & \texttt{\char`\\\char`\^a} &
\color{orange}\tengfont{]H}             & \texttt{\char`\\\char`\^e} &
\color{orange}\tengfont{`N}             & \texttt{\char`\\\char`\^i} &
\color{blue}\tengfont{.H}               & \texttt{\char`\\\char`\^o} & 
\color{blue}\tengfont{yY}               & \texttt{\char`\\\char`\^u} &
\color{green}\tengfont{hY}              & \texttt{\char`\\\char`\^U} &
%\color{green}\tengfont{lH}              & \texttt{\char`\\\char`\^Y} &
\color{green}\tengfont{\char`\~N}       & \texttt{\char`\\\char`\^\j} 
\\ \hline
\color{red}\Phonemic{aaa}               & \texttt{aaa} & 
\color{red}\tengfont{6R}                & \texttt{\char`\\\char`\'r} &
\color{orange}\tengfont{6Y}             & \texttt{\char`\\\char`\^r}&
\color{red}\tengfont{l}                 & \texttt{Y} &
&& 
\color{blue}\tengalt{h}                 & \texttt{V} &
&\\
\hline
\end{tabular}

\end{table}

Tolkien often wrote English with Tengwar phonemically. I'm not too
fond of this mode; I find that the orthographic modes are easier to
read and easier to write. However, I've heard some people say they
prefer the phonemic modes, so I'm making one available.

Note that I am not trained at all in linguistics; it's likely that I
got some of the stuff in \autoref{table:EP} quite wrong. Generally,
your best bet is to figure out for yourself what tengwa you should
use, and then look up in the table what letters you should use with
it.

Here are some notes:

Accented (long) vowels are usually treated as simply stressed vowels in
this mode (there's no special way of writing them); in \cite{luthien},
\ment{L\'uthien} is spelled \Phonemic{luuthi'en}, with the same letter
as in \Phonemic{yuus} \ment{use}, \Phonemic{huu} \ment{who}.

The abbreviations \Tehtamode{v|} and \Tehtamode{vv|} are not used in
this mode; \ment{of} is spelled \Phonemic{Ov}, \ment{of the} is
spelled \Phonemic{Ov tH|}. \Tehtamode{\THE} is still used for
\ment{the}. \Phonemic{End} is not an abbreviation; that's the correct
spelling of \ment{and} in this mode.

You can get a schwa using any of the letters \texttt{A}, \texttt{E},
\texttt{I}, or \texttt{O}. A schwa can be indicated either with an
undotted carrier, or an underdot under the \emph{following} tengwa. I
haven't been able to figure out a systematic rule for when a dot is
used and when a carrier is used; therefore, \texttt{E} and \texttt{A}
turn into underdots if they are followed by a reasonable letter and
\texttt{I} and \texttt{O} do not.

A \texttt{y} after most consonants turns into double overdots (as in
\Phonemic{pIrfyum} \ment{perfume}); a \texttt{W} after some consonants
or consonant groups (\texttt{t}, \texttt{d}, \texttt{g}, \texttt{s},
\texttt{th}, and~\texttt{sh}) becomes an overtwist.

Note that some pronunciation symbols show up twice; for example,
both~\Phonemic{i} and~\Phonemic{ee} have IPA symbol~i. This is not a
coincidince; Tolkien sometimes used one, sometimes the other. (He used
\emph{each of them} in the word \ment{beard}. This might be a mistake;
or it might be because both spellings are ``correct''. Usually,
\Phonemic{i} is used at the ends of words and \Phonemic{ee} is used
inside words. The difference between \Phonemic{u} and \Phonemic{uu}
seems to be mostly a matter of stress.)

Finally, note that all this analysis is based on Tolkien's examples,
which are based on the way English was pronounced by educated
Englishmen in the early twentieth century. If you want to spell things
differently because you pronounce things differently because you are a
twenty-first-century American, that's probably all right.

\section{Writing direction}

According to \cite{sarati}, sarati was often written right-to-left,
boustrophedon (alternating left-to-right), or vertically.

A small package called \texttt{boustr.sty} should have come with
\textsf{tolkienfonts}. It contains some environments for writing
direction:

\bigskip

\parbox{\halfwidth}{\raggedright
\texttt{\char`\\begin\{rtl\} Right to left \char`\\end\{rtl\}}
}\hfil
\parbox{\halfwidth}{
\begin{rtl} Right to left \end{rtl}}

\bigskip

\parbox{\halfwidth}{\raggedright
\texttt{\char`\\begin\{boustrophedon\} Boustrophedon\ldots\ 
\char`\\end\{boustrophedon\}}
}\hfil
\parbox{\halfwidth}{
\begin{boustrophedon} Boustrophedon, starting on the right and
switching to the left. \end{boustrophedon}}

\bigskip

\parbox[t]{\halfwidth}{\raggedright
\texttt{\char`\\boustrightfalse \char`\\begin\{boustrophedon\} 
Boustrophedon\ldots\ \char`\\end\{boustrophedon\}}
}\hfil
\parbox[t]{\halfwidth}{\boustrightfalse
\begin{boustrophedon} Boustrophedon, starting on the left and
switching to the right. \end{boustrophedon}}

\bigskip

\parbox{\halfwidth}{\raggedright
\texttt{\char`\\begin\{sidewaysflip\}\{1in\} 
Flipped vertical text \char`\\end\{sidewaysflip\}}
}\hfil
\parbox{\halfwidth}{\begin{sidewaysflip}{0.7in}\raggedright 
Flipped vertical text \end{sidewaysflip}}

(I didn't see the need to make a plain sideways environment; the
\textsf{rotating} package already has one.)

Be warned that \texttt{rtl} and \texttt{boustrophedon} are delicate.
They do bad things to displayed math, and stop working if you put any of
a number of odd things (such as \verb.\clearpage.) inside; don't do
anything too weird inside these environments. In particular, don't make
any zero-point vertical skips inside the environment; if for some reason
you really want to say
\verb.\setlength.\penalty0\verb.{\parskip}.\penalty0\verb.{0pt}., make
it \verb.\setlength.\penalty0\verb.{\parskip}.\penalty0\verb.{1sp}.
instead.

\texttt{rtl} could perhaps have avoided this, but I wanted it to be
capable of breaking across pages.

\section{Raw fonts}\label{sec:DSmode}

If you examine the documentation that came with your tengwar fonts,
you'll see that the keymapping is, frankly, bizarre. To get
\Quenya{nam\'arie} out of these fonts, you have to type
\texttt{5\#t\char`\~C7T`V}; being able to get it by typing
\texttt{nam\char`\\'arie} instead is the entire point of this package.

Nonetheless, fonts in their raw form have their uses. Occasionally,
you may find that you want to use, say, \tengfont{\char'246}. But
Tolkien only used this symbol when writing Old English, so it's not
available in any standard mode. Alternatively, you may have already
determined that in the raw font, \verb.7D~Vj%. \verb.j%k$w#3G.
\verb.w6E15^. produces your name, and want to copy and paste. You can
get the raw encoding with the commands \verb.\tengfont. and
\verb.\tengalt..

In the raw font, many important glyphs are placed over special
characters. The easiest way to get \tengfont{\char`\\},
\tengfont{\char`\~} and {\tengfont{2\char`\^}} is with
\verb.\tengfont{\char`\\}., \verb.\tengfont{\char`\~}. and
\verb,\tengfont{2\char`\^},.\footnote{\texttt{\char`\\tengfont\char`\{2\char`\\\char`\^\char`\{\char`\}\char`\}} will \emph{not} produce \tengfont{2\char`\^}.} Several of the other characters (\#, \$,
\%, \&, \_, \{, and~\}) must be preceded by a backslash \char92\ to
come out properly.

The unusual characters can be referenced with the \verb.\char.
command. For example, the Tengwar Annatar keyboard map says that
\tengfont{\char190} is character 190; to get it in \TeX, use
\verb.\tengfont{\char190}..

The commands \verb|\cirthfont|, \verb|\saratifont|, and
\verb|\valmarfont| are defined similarly.

\appendix

\section{Doing without \rm\texttt{tolkienfonts.sty}}\label{sec:plain}

Most of the work of \textsf{tolkienfonts} is done by the virtual
fonts; the file \texttt{tolkien\-fonts.sty} provides macros,
environments, package options, etc., but can be dispensed with if
necessary.

\begin{table}[tp]
\caption{Font family names and modes}\label{table:fontnames}

When using font names or font family names, beware of capitalization issues. Sometimes you have to say \texttt{TengwarASB}; other times you have to say \texttt{tengwarasb}.

\bigskip

\begin{tabular}{|l|l|}
\multicolumn{2}{c}{Scripts and fonts}\\\hline
\texttt{TengwarA}& Tengwar Annatar\\
\texttt{CirthE} & Cirth Erebor\\
\texttt{CirthS} & Sunrunes\\
\texttt{SaratiE} & Sarati Eldamar\\
\texttt{SaratiB} & Sarati Eldamar with bar\\
\texttt{Valmaric}& Valmaric Eldamar\\\hline
\end{tabular}

\bigskip

\begin{tabular}{|l|l|}
\multicolumn{2}{c}{Languages and modes}\\\hline
\texttt{} & Raw tengwar, sarati, or valmaric font\\
\texttt{A} & Raw tengwar alt font\\
\texttt{XX} & Raw cirth font\\
\hline\multicolumn{2}{|c|}{Tengwar}\\
\texttt{QQ} & Quenya default mode\\
\texttt{QR} & Quenya red mode\\
\texttt{QG} & Quenya generic mode\\
\texttt{QA} & Quenya mode without \ment{a}s\\
\texttt{SA} & Sindarin mode of Arnor\\
\texttt{SG} & Sindarin mode of Gondor\\
\texttt{SB} & Sindarin mode of Beleriand\\
\texttt{EO} & English full orthographic mode\\
\texttt{EM} & English mode of Mazarbul\\
\texttt{ET} & English tehta orthographic mode\\
\texttt{EP} & English full phonemic mode\\
\hline\multicolumn{2}{|c|}{Runes}\\
\texttt{QE} & Angerthas Daeron/Eregion, arranged for Quenya\\
\texttt{SD} & Angerthas Daeron/Eregion, arranged for Sindarin\\
\texttt{DM} & Angerthas Moria, arranged for Dwarvish\\
\texttt{EE} & Angerthas Erebor, arranged for English\\
\texttt{EH} & Runes from \ment{The Hobbit}, arranged for English\\
\hline\multicolumn{2}{|c|}{Sarati/Valmaric}\\
\texttt{QQ} & Quenya default mode\\
\texttt{QA} & Quenya mode with \ment{a}s\\\hline
\end{tabular}

\end{table}

The (real and virtual) fonts used by \textsf{tolkienfonts} do not
follow the Berry naming scheme. Instead, the font names are
constructed by starting with the script name (usually
\texttt{Tengwar}) and appending three capital letters to indicate
font, language, and mode, plus (for Tengwar Annatar) a few lowercase
letters to indicate font shape.

So the virtual font for \texttt{beleriand} is \texttt{TengwarASB}. See
\autoref{table:fontnames} for the complete font/mode list.

The names of the \LaTeX\ font families are constructed by taking the
names of the (normal-shaped) fonts and translating them to lowercase.

In \LaTeX, you can use these fonts without \textsf{tolkienfonts} in several ways:

\begin{itemize}
\item You can create a text-mode command\\
\indent\verb|\DeclareTextFontCommand{\Beleriand}|\\
\indent
\verb|{\fontencoding{T1}\fontfamily{tengwarasb}\selectfont}|\\
which you can use like \verb|\textbf|:\\
\indent\verb|\Beleriand{tin\'uviel}|
\item Alternatively, you can change fonts directly inside the 
document:\\
\indent\verb|\usefont{T1}{tengwarasb}{m}{n} tin\'uviel|
\item If you want the entire document in tengwar, you can use\\
\indent\verb|\renewcommand{\rmdefault}{tengwarasb}|\\
\indent\verb|\begin{document}|\\
\indent\verb|tin\'uviel|
\item If you're having trouble getting font families to work, you can use \verb|\newfont|:\\
\indent\verb|\newfont{\Beleriand}{TengwarASB at 10pt}|\\
\indent\verb|{\Beleriand tin\'uviel}|\\
However, \verb|\newfont| does not play nicely with \LaTeX's size-changing mechanisms, so if you can get any of the above techniques to work, it's probably better to use them.
\end{itemize}

All of these produce {\usefont{T1}{tengwarasb}{m}{n}tin\'uviel}.


In plain \TeX, you use direct font definitions
\begin{quote}
\verb|\font \beleriand = TengwarASB|
\end{quote}
as usual.

In either case, you should be aware of a few things. First, to adjust 
the spacing, you need to adjust some spacefactor codes:
\begin{quote}
\verb|\sfcode`,=2000| \verb|\sfcode`;=2000|
\end{quote}
for all fonts, and
\begin{quote} \raggedright
\verb|\sfcode`)=1000| \verb|\sfcode`]=1000|
\verb|\sfcode`?=1999| \verb|\sfcode`!=1999|
\end{quote}
for tengwar.

If you're likely to use a capital letter just before a punctuation mark, 
you might want to say \verb|\sfcode`R=1000| to fix the spacing.

Also, the commands
\begin{quote} \tt  \raggedright
\char`\\ten
\char`\\eleven
\char`\\twelve
\char`\\ldots
\char`\\of
\char`\\ofthe
\char`\\The
\char`\\THE
\char`\\\char`\&\
\char`\\r
\char`\\R
\char`\\s
\char`\\es
\char`\\sa
\char`\\se
\char`\\si
\char`\\so
\char`\\su
\char`\\sy
\end{quote}
and \verb*|\ | (runic modes) are
redefined by \texttt{tolkien\-fonts.sty}; don't use them with tengwar
without it. You can get ten and eleven with \texttt{<} and~\texttt{>}.

You can do without \verb|\"| (by using quote marks or dashes); you can
get long vowels with a left quote \texttt{a`}. Alternatively, if you
want \verb|\'| (and \verb|\"|,~\verb|\^|) to work properly, in
\LaTeX, you can use the \textsf{fontenc} package: \begin{quote}
\verb|\usepackage[T1]{fontenc}|\end{quote} In plain \TeX, you can use
\textsf{plainenc}: \begin{quote} \verb|\input plainenc|\\
\verb|\fontencoding{T1}| \end{quote} If you do this, \verb|\^| will occasionally produce a different result from~\verb|\'|; they will both be legitimate transliterations of long vowels, but possibly different ones.

\section{Tengwar discussion: why did I make the mode choices I did?}

Given the paucity of examples, there are a lot of decisions I had to
make regarding the fine points of mode operation. They are discussed
below.

Note: \ment{kh} refers to the Scottish, German or Sindarin hard
\ment{ch}, not the English soft \ment{ch}; it sounds more like a cat
hissing than anything in English. It is usually transliterated
\ment{ch}; in this appendix, I am going to call it \ment{kh} to
distinguish the two. If you don't speak German, \ment{h} is probably the
best English substitute when speaking aloud.

\subsection{Quenya modes}

The general mode was based on \cite[~38, 42, 43, 44, 46,~64]{DTS}. It's
been designed to produce reasonable things if Sindarin words, as well as
Quenya words, are entered. Therefore, many of the Sindarin special rules
(e.g.\ \Quenyagen{ng} at the ends as well as the beginnings of words)
work in the general mode \texttt{Quenyagen} but not in \texttt{Quenya}.

A notable feature of this mode is that dipthongs are always written with
two tehtar. All the other Elvish tehta modes write dipthongs with a
tehta and a special tengwa; however, \Quenyagen{ai} for \ment{ai} is
well attested. This is also a good idea in that a generic mode needs to
support the Sindarin \ment{e}-glide dipthongs \ment{ae}, \ment{oe} as
well as the Quenya \ment{i}-glide dipthongs \ment{ai}, \ment{oi},
\ment{ui}. You can force dipthongs to be written with special tengwar by using \texttt{E},~\texttt{I} or~\texttt{U} as the second letter.

The Quenya tengwar mode is based on descriptions in \cite{appE}, the
example \cite{namarie}, and also on the analysis in \cite{textbook}
and \cite{linquenya}. (If you don't know anything about the Quenya
mode and you would like to, \cite{linquenya} is a good place to
start.)

To construct \verb.\Eregion. and the sarati and valmaric fonts, I used
\cite{appE}, \cite{sarati} and \cite{valmar} to determine which glyph
meant which sound. Additional rules were then taken from the tengwar
mode. \verb.\Eregion. and \verb.\Daeron. were carefully constructed to
be as similar as possible; there are a few minor differences.

There was one decision I had to make: what to do with~\texttt{h}.

\subsubsection{The \ment{h} question}\label{sec:Qh}

So: does \ment{aha} have an aha in it?

It would not be unprecedented for the answer to be no. If we follow
the usual \ment{r}-rule, \ment{\'ore} is spelled \Quenya{\'ore}, with
a r\'omen---no \'ore is present.

Here's what we know.

Originally, Quenya had a \ment{kh} sound. It was written with
\Quenya{ch}; the Noldor named this letter \ment{kharma}. Quenya also
had an \ment{h} sound (as in English \ment{house};) it was written
with \Quenya{H}. Finally, it had a \ment{hy} sound, written with
\Quenya{h}.

At some point, the Noldor decided that words which \emph{began} with
\ment{kh} should begin with \ment{h} instead. Somewhat unusually
(possibly because they were keeping \ment{kh}s that occured inside
words), they decided to alter the spellings of these words. At or
around this time, they started to use \Quenya{h} for the \ment{h}
sound, and \Quenya{hy} for \ment{hy}.\footnote{They kept \Quenya{H}
around to indicate voiceless \ment{r} and \ment{l}:
\Quenya{hR}, \Quenya{hl}. These are the same sounds which
Sindarin writes with \Gondor{rh} and \Gondor{lh}; in Quenya, they are transcribed as \ment{hr} and \ment{hl} instead of \ment{rh}
and \ment{lh}.}

So now \ment{harma} did not have a \Quenya{ch} in it. The Noldor began
to call \Quenya{ch} \ment{aha}; presumably, at that time, \ment{aha}
was spelled \Quenya{acha}. However, according to \cite{appE},
\emph{any} \ment{h} standing on its own with no nearby consonants is
pronounced as in \ment{house}. (\ment{ht}~is for some strange reason
pronounced as~\ment{kht}.)

There are two possible explanations for this situation.

\begin{itemize}

\item When they started calling \Quenya{ch} \ment{aha}, it was
pronounced \ment{aha}, not \ment{akha}. This doesn't make sense to me.

\item When they started calling \Quenya{ch} \ment{aha}, it was
pronounced \ment{akha}, not \ment{aha}. Quenya \emph{later} lost all
of its remaining \ment{kh}s except those just before \ment{t}s. For
whatever reason, the Noldor didn't change the name of \Quenya{ch}
again.

It should be noted that Tolkien mentions \ment{kh} as ``remaining
medially'', but that \cite{qwl} has no \ment{ch}s or \ment{kh}s in it;
as far as we know, in Third Age Quenya, \ment{kh} only happens just
before \ment{t}. It is possible that \ment{kh} originally only
happened at the start of words or before \ment{t}, but it seems
unlikely.

It should also be noted that the Sindarin words \ment{acharn} and
\ment{achas} mean \ment{vengeance} and \ment{dread}; it seems likely
that they and \ment{aha} come from a common proto-Eldarin root with a
\ment{kh} in it.

\end{itemize}

In the first case, all medial \ment{h}s should be spelled with
\Quenya{ch}. In the second case, \ment{h}s before \ment{t}s should
probably be spelled with \Quenya{ch}. Medial \ment{h}s that used to be
\ment{kh}s might \emph{also} still be spelled with \Quenya{ch} (which
would explain why \ment{tehta} is not spelled \ment{techta});
unfortunately, we don't actually know which \ment{h}s used to be
\ment{kh}s. You can guess based on Sindarin cognates; you can assume
that it is all of them; or you can just write them with \Quenya{h}s
instead.

This is what \textsf{tolkienfonts} does: it uses \Quenya{ch} before a
\ment{t}, \Quenya{H} before a \ment{r} or \ment{l}, and \Quenya{h} for
all other \ment{h}s which are not part of combinations. If you want a
medial \ment{h} to be written with \Quenya{ch}, you have to use
\texttt{ch} or \texttt{kh} instead.

\subsection{Sindarin modes}\label{sec:S}

The mode of Beleriand is based on \cite{elbereth} and \cite{durin}, as
well as on the analysis in \cite{appE}, \cite{linsindarin} and
\cite{textbook}. The Third Age Sindarin modes (\verb.Arnor. and
\verb.Gondor.) are based mainly on \cite{kingsletter}, but also on
\cite{lotr} and \cite{appE}. 

Like \texttt{Quenyagen}, the Sindarin modes have some special rules
borrowed from the other Elvish language to allow you to easily write
isolated Quenya words. You can get \Gondor{rh} and~\Gondor{lh} with
either the Sindarin initial \texttt{rh},~\texttt{lh} or the Quenya
\texttt{hr},~\texttt{hl}. A \texttt{Y}, or a \texttt{y} followed by a vowel, produces the same tengwa (\Beleriand{Y}~or~\Gondor{Y}) that is used at the start of \ment{ionnath}, that is, in the only place where Sindarin has a consonantal \ment{y} sound.

The Quenya dipthong \ment{oi} does not occur in \cite{swl}, and
\ment{iu} occurs only at the start of words. \ment{Eu} does occur (in
\ment{leutha}); the author seems to think that it is an ``unusual
dipthong'' rather than separated vowels. Consequently, these three
combinations are dipthongs like \ment{ai},~\ment{au} in all three modes
(except for initial \texttt{iu}, which produces \Gondor{iu}/\Arnor{iu}
as usual).

\verb.\Daeron., like \verb.\Eregion., is based on the Angerthas Daeron
in \cite{appE}, with some special rules that parallel rules in the
tengwar mode.

\ment{Mh} is \Gondor{mh} in the modes of Gondor and Arnor, and so is
presumed to be \Beleriand{mh} in the mode of Beleriand.

In \cite{appE}, final \ment{ng} is said to be pronounced as \ment{\ng}
(as in \ment{sing}). Initial \ment{ng} is described as being
pronounced \ment{\ng} in (archaic) Quenya; in \cite[\unskip, p.~63]{Road}, it is also mentioned as being pronounced \ment{\ng} in Sindarin. Therefore, initial \ment{ng} is transcribed with \Beleriand{ng} or~\Gondor{ng}.

However, \cite{swl} has some
words that begin with~\ment{ng} that it claims should be pronounced \ment{\ng g}. These \ment{ng}s should probably be transcribed with \Beleriand{'ng'}/\Gondor{'ng'}; to get them initially, use~\texttt{'ng}.

Most of the uncertainty in the Sindarin tengwar modes lies in how to
deal with \ment{u} after a vowel or \ment{w} in any situation.

In the tehta mode, I follow \cite{textbook} and \cite{linsindarin} and
use \Gondor{au} for \ment{au}.

\cite{kingsletter} uses overtwists in Sindarin for \ment{w} after a
consonant (labialization). It also uses overtwists for English
\ment{u}-glide dipthongs. Therefore, I use overtwists for
labialization and \ment{u}-glide dipthongs in both of the Third Age
modes.

It seems reasonable to assume that the mode of Beleriand is meant to
be easy for native Quenya readers to deal with, and that the Sindarin
modes of Gondor and Arnor are meant to be easy for native Westron
readers to deal with. The modes of Beleriand and Gondor were used
thousands of years apart, by people of different races; while they are
both Sindarin modes, there's no real reason to expect them to be
similar.

Analogy with Quenya\footnote{Also see \cite{appE}: ``\ment{hwesta
sindarinwa} [\Gondor{hw}] or `Grey-elven \ment{hw}' was so called
because in Quenya [\Quenya{chw}] had the sound of~\ment{hw}, and
distinct signs for \ment{chw} and~\ment{hw} were not required.''} 
leads me to conclude that in the mode of Beleriand, Series~IV probably
represents labialization of Series~III. So \ment{gw} and~\ment{chw}
are \Beleriand{gw} and \Beleriand{chw}, not~\Gondor{gw} and
\Gondor{chw}. (\ment{dw}~is written with \Beleriand{d} plus a tehta,
since \Beleriand{q},~\Beleriand{ghw} or~\tengfont{b} would be
confusing.)

\cite{appE} says that ``the sign for following~\ment{w} (required for
the expression of \ment{au},~\ment{aw}) was in this mode the u-curl or
a modification of it''. This could be interpreted to mean an
overtwist~\Arnor{au}, as in the Third Age modes; however, the
overtwist does not seem to be used in writing Quenya, and so the mode
of Beleriand, as used in Beleriand, probably used simply the
\ment{u}-curl~\Beleriand{au}. (Maybe the mode of Eregion used the
overtwist.) \ment{dw} is written with the same tehta.

Quenya and the mode of Beleriand both use~\Quenya{w} for~\ment{w}.
However, \cite{appE} says that \Gondor{w} was ``often used for
\ment{w}'', and~\Gondor{w} is used in \ment{war} and \ment{Westmarch}
in the title-page inscription for \ment{The Lord of the Rings}, which
is English written ``as a man of Gondor might write it''. Therefore,
Westron tehta mode\footnote{\cite{history} presents charts for a
Westron tehta mode (which does use \Gondor{w} for~\ment{w}) and
Westron full mode which are virtually identical to the charts for
English tehta and orthographic modes. The full Westron mode there,
like English orthographic mode, uses \Orthmode{w} for~\ment{w},
\Orthmode{u} for~\ment{u} and \Orthmode{a} for~\ment{a}. We know that
the mode of Arnor uses \Arnor{u} for~\ment{u}. It could theoretically
use either \Arnor{w} or~\Arnor{W} for~\ment{w}; it would, however, be
less confusing to Westron writers if \ment{u} and~\ment{w} just
switched.} probably used~\Gondor{w}, and so the modes of Gondor and
Arnor are assumed to do the same.

\subsection{English and Dwarvish}

The full orthographic English mode supplied by \verb.orthmode. is based
on \cite{kingsletter}; the \verb.mazarbul. variant is based on
\cite{ori}. (The use of \Mazarbul{wh} as \ment{wh} comes from
\cite{lowdham}.) The cirth mode \texttt{erebor} is based on \cite{ori}
as well (it has some pages written with cirth and some pages written
with tengwar).

The mode \texttt{moria} is based almost entirely on \cite{appE}. The
only extant text written in this mode (in fact, the only extant text
in Dwarvish at all) is \cite{balin}. If Dwarvish has any special
rules, I don't know what they are.

We have very few tehta mode examples. (There is \cite{lotr},
\cite{ttt}, \cite{brogan}, and \cite{endorion}. That's about it.)
What's more, the examples we do have are inconsistent. \cite{ttt} uses
\Tehtamode{u} for \ment{o}; \cite{brogan} uses~\Tehtamode{i}
for~\ment{e} and~\Tehtamode{e} for~\ment{i}; etc.

Therefore, the English tehta mode supported by \textsf{tolkienfonts}
derives many of its rules from the English orthographic mode and the
mode of Gondor.

These two modes agree on everything, except: \begin{itemize}

\item \ment{ch, j, sh, x, z,} silent \ment{e}, and the \ment{c+h} in
\ment{echo} occur in English. Thus, \Orthmode{z} cannot be \ment{ss},
and the tengwar for the others are actually used.

\item The letter \Orthmode{C} is used for soft \emph{c} in the
orthographic full mode examples. Soft \emph{c} does not come up in any
of the tehta mode examples; Tolkien never uses \Orthmode{C} for
\ment{s}, even when there are tehtar present and using a \Orthmode{C}
would be convenient.

\item In the mode of Gondor, \Gondor{E} is used for \ment{e} only when
it is the second letter of a dipthong (that is, only for
``\ment{e}-glide''); in  English orthographic mode, it is used for all
non-silent~\ment{e}s. (The Gondorians also use it for an \ment{i}
followed by a vowel at the start of a word, which is pronounced as a
consonantal~\ment{y}.)

\item English full modes use \Orthmode{i} and \Orthmode{y} for
\ment{i} and \ment{y}. Obviously, these symbols are not available in
any tehta mode. In English tehta mode, tehtar are available to
represent \ment{i} and vocalic \ment{y}; some letter needs to be
chosen for consonantal~\ment{y}.

\item The letters \Orthmode{u} and \Orthmode{o} mean \ment{u} and
\ment{o} in full English orthographic mode, but mean \ment {w} and
\ment{i}-glide in the Mode of Gondor.

\item The letters \Orthmode{a} and \Orthmode{w} mean \ment{a} and
\ment{w} in full English orthographic mode. One of these probably
means \ment{u}-glide in the Mode of Gondor, and the other is probably
not used. \cite{textbook} and \cite{linsindarin} use \Gondor{U} for
\ment{u}-glide, but I don't believe that we have any attested
examples.

\end{itemize}

So to construct a tehta mode, we need to decide the following:
\begin{itemize}

\item What tehta do we use for vocalic \ment{y}? (I use \Tehtamode{y},
since both \cite{lotr} and \cite{brogan} do. The mode of Gondor
uses~\Gondor{y}.)

\item What tengwa do we use for consonantal \ment{y}? (\cite{brogan}
uses \Tehtamode{Y}, so that's what I use.)

\item Which tengwa do we want to use for \ment{w}? (I follow
\cite{brogan} and \cite{lotr} and use~\Tehtamode{w}.)

\item Do we use \Tehtamode{e} for \ment{e} and \Tehtamode{i} for
\ment{i}, or vice versa? The first one is consistent with both Quenya
mode and the mode of Gondor; the second one uses some form of dots for
both kinds of \ment{e}. This is particularly useful if we want to
imitate Christopher Tolkien and use other undertehtar. (\cite{brogan}
uses \Tehtamode{i} for \ment{e}; \cite{lotr} and \cite{endorion} use
\Tehtamode{e} for \ment{e}. I don't like undertehtar, and I find that
I can remember that underdot and overaccent mean the same letter
fairly easily, whereas if I use \Tehtamode{i} for \ment{e}, I get
confused whenever I try to read an inscription in Quenya mode or the
mode of Gondor.)

\item How do we write dipthongs? This is really the big one. We have
three standard tengwar left: \tengfont{n}, \tengfont{l},
and~\tengfont{.}. (We also have \tengfont{]} and
\tengfont{\char"B3},~\tengfont{\char"BE}.) We could just write out the
two tehtar, with a carrier or two. Tolkien does this for \ment{ea} and
\ment{io} in \cite{brogan}, but had switched from \Tehtamode{ia} to
\Tehtamode{ea} by the time he wrote \cite{endorion}; I prefer to write
as he did in \cite{endorion}.

\begin{itemize}

\item \cite{endorion} uses \tengfont{]} for its \ment{a}-glide
dipthong.

\item \cite{endorion} uses \tengfont{l} for its \ment{e}-glide
dipthong. So does the mode of Gondor.

\item Since we're using \Tehtamode{Y} for the consonantal \ment{y} in
\ment{you}, it makes sense to use it for \ment{y}-glide dipthongs as
well. And given the relationship between \ment{i} and \ment{y} (e.g.\
\ment{they/their}, \ment{liar/lying}), it makes sense to use the same
character for \ment{i}-glide and \ment{y}-glide, if we have to have
any overlaps.

\item \cite{brogan} makes \Tehtamode{w} do double duty as \ment{w} and
\ment{u}-glide. \cite{lotr} and \cite{endorion} have no
\ment{u}-glides, so they cannot contradict this. On the other hand,
this is ambiguous, and it makes no sense to use \tengfont{.} or
\tengfont{n} as \ment{y}; they are available to be \ment{u}-glide and
\ment{o}-glide.

I have a strong personal preference for unambiguous transliteration, so I really do not like overloading \Tehtamode{w} for both \ment{w} and \ment{u}-glide. Also, I prefer to arrange things so that, if you happen to find yourself mentioning \ment{Aul\"e} or
\ment{Sauron} in your tengwar texts, you will spell those names more
or less as they were spelled in Middle-Earth. Thus, \texttt{Tehtamode} defaults to using \Tehtamode{U} for \ment{u}-glide. If this bothers you, use \texttt{w} to force a~\Tehtamode{w}.

\item \ment{o}-glide dipthongs (as in \ment{people}) are rare, unless
you include words ending in \ment{tion}, \ment{cion}, \ment{sion},
which I don't like to since the \ment{i} in those words goes to
modifying the preceding consonant---it's not part of the vowel.
Unfortunately, we have no attested examples of \ment{people}, and only
one example of \ment{tion}, which is in \cite{brogan}. There are no
\ment{o}-glide dipthongs in any Elvish language, and the full Elvish
and English orthographic modes all use \Beleriand{o} for \ment{o},
which is taken. English phonemic inscriptions use \Beleriand{u} for
\ment{o}, which I want to use for~\ment{u}. I use \Tehtamode{O} for
\ment{o}-glide just because it's the only thing left.\footnote{You
could also make \Tehtamode{w} do double duty, as in \cite{brogan},
which would free up \tengfont{.} to be \ment{o}-glide. Or you could
just write \ment{eo} as \Tehtamode{eo} no matter how it's pronounced.}

\end{itemize}

\end{itemize}

\section{Virtual Fonts and the {\textsf{fontinst}}
package}\label{sec:fontinst}

The way \textsf{tolkienfonts} is constructed, I need a separate
virtual font for each mode\slash base font\slash color combination.
For full color effects as seen in \cite{namarie}, you'd want both a
red-tehtar-black-tengwar and a black-tehtar-red-tengwar font (plus a
plain font for normal use). I've got this package set up to allow for
nine different tengwar fonts (counting the four flavors of Tengwar
Annatar).

Rather than having this package come with hundreds of supporting
files, I've decided to include one font (Tengwar Annatar normal) and
one colored mode as examples, and give what you need to make the
supporting files for colored tehtar and other fonts yourself.

The correct Quenya spelling of ``Arda'' is \Quenya{{arda}}. In most
modern tengwar fonts, you get \Quenya{{arda}} by typing \texttt{`CuF}.
The purpose of this package is to let you get
\Quenyared{\textbf{arda}} by typing \verb.\Quenyared{\textbf{arda}}..
Here's how \LaTeX\ does it: \begin{itemize}

\item The \texttt{.sty} file defines the command \verb|\Quenyared|,
which says, ``\LaTeX! Switch to the font encoding T1 and the font
family \textsf{tengwaraqr}!''

\item \LaTeX\ finds the file \texttt{t1tengwaraqr.fd}. It has a bunch
of font-family commands, one of which says, ``\LaTeX! If you're using
this family, then tell \TeX\ to use the font \textsf{TengwarAQRb} for
bold-width, normal-shaped text!''

\item \TeX\ looks for the files \texttt{TengwarAQRb.tfm} and
\texttt{TengwarAQRb.vf}. The \texttt{.tfm} file says, ``\TeX! You
should allocate \setbox0=\hbox{\Quenyared{a}}\the\wd0\ for that first
\texttt{a}, you should turn all \texttt{rd}s into the character in
Slot~245, and you should turn all \texttt{a}s that come after that
character into the character in Slot~205!''

\definecolor{darkred}{rgb}{0.8,0,0}

\item The \texttt{.vf} file says, ``\TeX, you're not done yet! I need
you to replace all the \texttt{a}s with the characters \texttt{`C} from
the font \textsf{TengwarAb} and replace the character in Slot~245 with
the character \texttt{u} from \textsf{TengwarAb}. Oh, and while you're
at it, make the \texttt{\color{darkred}C} red.''

\item \TeX\ goes and looks up the width of the \texttt{`} in the file
\texttt{TengwarAb.tfm}, so that it knows where to put the~\texttt{C}.

\item \TeX\ checks to make sure there's no file \texttt{TengwarAb.vf},
and, satisfied, finishes its job. We now know that we want the
sequence of glyphs \texttt{`{\color{darkred}C}u\color{darkred}F} from
the font \textsf{TengwarAb}.

\item Now, it's some other program's job to figure out whether
\textsf{TengwarAb} means \texttt{tnganb.ttf} or \texttt{tnganb.pfb} or
\texttt{TengwarAb.mf} or what, actually grab the appropriate glyphs,
and place them into the PDF or PostScript file (or display them on the
screen). pdf\/\TeX\ combines this step with the previous steps;
\textsf{dvips} and so on keep them separated.

\end{itemize}

So, if you want to use Tengwar Parmaite (for example) instead of
Tengwar Annatar, you need to make all those files.

\subsection{Adding new fonts}\label{sec:fonts}

I've included the files you'll need to set up for the following fonts:
\begin{itemize}
\item The other three Tengwar Annatar fonts
\item Tengwar Parmaite
\item Tengwar Eldamar
\item Tengwar Noldor
\item Tengwar Quenya
\item Tengwar Sindarin

%\item Tengwar Formal
%\item Tengwar Elfica
%\item Tengwar Gothika
%\item Tengwar Beleriand
%\item Tengwar Teleri
%\item Tengwar Galvorn
%\item Tengwar Mornedhel
%\item Tengwar Cursive
%\item Tengwar Naive
\end{itemize}

You can switch between fonts with the commands \verb.\annatar.,
\verb.\eldamar., \verb.\noldor., \verb|\parmaite|, \verb.\fontquenya.,
and \verb.\fontsindarin.. Like \verb|\sunrune| and \verb|\fonterebor|,
these commands need to be given outside of the font commands they are to
affect.

To set up these fonts, start by going to the \texttt{tolkieninst}
folder. Find the file \texttt{tolkien\-inst.tex} and open it.

Find the lines
\begin{quote}
\verb|%%%%% Tengwar Parmaite %%%%%|

\bigskip

\noindent\verb|\installfonts|\\
\verb|%\installfontas{TengwarP}{T1}{tengwarp}{m}{n}{}|
$$\vdots\quad\vdots\quad\vdots$$
\verb|\endinstallfonts|
\end{quote}

Uncomment every line between the
\verb|\installfonts| and the \verb|\endinstallfonts|. If
for some reason you only want a few modes, uncomment just the lines for that mode. (The \verb.\installfontas. lines
make the \texttt{.fd} files for \verb.\tengfont. and \verb.\tengalt..)

All the \verb.\installfont. and \verb.\installfontas. lines for Tengwar
Annatar, Cirth Erebor and sunrunes, Sarati and Valmaric are uncommented,
since this is the file I used to make the files that came with this
package; you probably want to comment them out.

Now, run \texttt{tolkieninst.tex} through plain \TeX\ (not \LaTeX).
This will generate a bunch of \texttt{.vpl} files. Open up a command
line. The following commands will convert multiple VPL files to VF/TFM files:
\begin{quote}
    \verb|for file in *.vpl; do vptovf $file; done| (Mac OS X or UNIX)\\
	\verb|for %A in (*.vpl) do vptovf %A %~nA.vf %~nA.tfm| (Windows)
\end{quote}
If this doesn't work, you'll have to type something like
\begin{quote}
\verb|vptovf TengwarPQQ.vpl TengwarPQQ.vf TengwarPQQ.tfm|
\end{quote}
for every generated \texttt{.vpl}.

This should give you a bunch of \texttt{.fd} files, a bunch of
\texttt{.vf} files, and a bunch of \texttt{.tfm} files, which should
be distributed to appropriate places, as described in \autoref{sec:install}. The \texttt{.vpl} files may be deleted.

Quenya, Noldor, and Sindarin can be dealt with in the same way. If you
want to include multiple shapes of Tengwar Annatar, it's strongly
recommended that you do all four shapes at once. This will
automatically write all four shapes to the \texttt{.fd}s, which in
turn will let you pick them out with the \LaTeX\ commands \verb|\emph|
and \verb|\textbf|.

\subsection{Adding color}\label{sec:color}

The easiest way to get odd color effects is to build them into the
virtual font.

Open the file \texttt{tolkieninst.tex}. Figure out which font you want
to edit.

Suppose that you decide you want \verb.\Beleriand. after
\verb.\noldor. to produce green tehtar and punctuation. You want the
line \verb.\installfont.\penalty0\verb.{TengwarNSB}.: \texttt{N}~for
the Noldor font, \texttt{S} because \verb.\Beleriand. is a Sindarin
mode, and \texttt{B} for Beleriand. (See \autoref{table:fontnames}.)
Uncomment that line.

The important line starts with:

\medskip

\noindent\verb.\installfont{TengwarNSB}. 
\verb.{fontNoldor,effectPunctspace,.\ldots

\medskip

Add \texttt{effectColor} just before \texttt{effectPunctspace}:

\medskip

\noindent\verb.\installfont{TengwarNSB}.
\verb.{fontNoldor,effectColor,effectPunctspace,.\ldots

\medskip

Now, open the file \texttt{effectColor.mtx}. Find the lines
\begin{quote}\parindent=0pt
\verb|\setcommand\tehta#1{%   Red|\\
\verb|   \ifisglyph{#1}\then|\\
\verb|      \setglyph{#1black}\glyph{#1}{1000}\endsetglyph|\\
\verb|      \setleftrightkerning{#1black}{#1}{1000}|\\
\verb|      \resetglyph{#1}|\\
\verb|         \glyphspecial{pdf: 0.8 0 0 rg 0.8 0 0 RG} %pdfTeX|\\
\verb|         %\glyphspecial{color push rgb 0.8 0 0} % dvips|\\
\verb|         \glyph{#1}{1000}|\\
\verb|         %\glyphspecial{color pop} % dvips|\\
\verb|         \glyphspecial{pdf: 0 g 0 G} %pdfTeX|\\
\verb|      \endresetglyph|\\
\verb|   \fi|\\
\verb|}|
\end{quote}

It's the \verb.\glyphspecial. lines that change color. Next we have to
fix the color code to be green instead of red.

If I say
\begin{center}
\verb|\definecolor{mygreen}{rgb}{0,0.6,0} {\color{mygreen} test}|
\end{center}
I get a nice green; so I change the \verb|0.8 0 0 rg 0.8 0 0 RG| to
\verb|0 0.6 0 rg 0 0.6 0 RG|. (It's a lot faster to fiddle with colors
in \LaTeX\ before you start changing them in \texttt{color.mtx}.)

Now, we're ready. Run the file \texttt{tolkieninst.tex} through plain
\TeX\ (not \LaTeX). The result should be a few generated files. Go to
a command line and run \texttt{vptovf TengwarNSB}; among your newly
generated files should be the file \texttt{TengwarNSB.vf}. Replace the
old \texttt{TengwarNSB.vf} with this one, and you're done.

If you want to make entirely new fonts rather than replacing old ones, 
then you should copy the line
\par{\raggedright\tt \hangindent = 2em
\char`\\installfont\{TengwarNSB\} 
\{fontNoldor,\penalty0effectPunctspace,\penalty0scripttengwar\} 
\{modeBeleriand\} \{T1\}\{tengwarnsb\} \{m\}\{n\}\{\}
\par}
\noindent and change it to
\par{\raggedright\tt \hangindent = 2em
\char`\\installfont\{TengwarNSBG\} 
\{fontNoldor,\penalty0effectColor,\penalty0effectPunctspace,%
\penalty0scripttengwar\} \{modeBeleriand\} \{T1\}\{tengwarnsbg\}
\{m\}\{n\}\{\}
\par}
This will generate a new font \textsf{TengwarNSBG} and a new font
family \textsf{tengwarnsbg}. See \autoref{sec:plain} to see how to
use them.

\subsection{Adjusting the spacing}\label{sec:spacing}

Getting rid of the special tengwar spacing rules is even easier. Just 
delete the \texttt{effectPunctspace} in the \verb.\installfont. lines:

\medskip

\noindent\verb.\installfont{TengwarNSB}.%
\texttt{\char`\{fontNoldor,effectPunctspace,scripttengwar\ldots}

\medskip

\noindent becomes

\medskip

\noindent\verb.\installfont{TengwarNSB}.\texttt{\char`\{fontNoldor,scripttengwar\ldots}

\medskip

Then you run \texttt{tex tolkieninst} and compile your \texttt{.vpl}s as 
usual.

\clearpage

\section{Font tables}\label{sec:tables}

Asterisks indicate letters or letter combinations whose behavior
depends on neighboring characters.

Single *s indicate word boundaries: \mbox{\Quenya{nw} *nw} means that
\texttt{nw} produces \Quenya{nw} \emph{only} when they are the first
two letters in the word. If for some reason you really want
\Quenya{'nw'} at the start of a word, use an apostrophe.

Double **s indicate that the characters that trigger the special
behavior are all vowels: \mbox{\Gondor{S} **s} means that \texttt{s}
produces \Gondor{S} to accomodate tehtar, \mbox{\Quenya{I} **i} means
that \texttt{i} produces \Quenya{I} when used in appropriate
dipthongs, and \mbox{\Quenya{R} r**} means that \texttt{r} produces
\Quenya{R} whenever the next letter is a vowel.

\begin{table}[hp]
\caption{Quenya mode}\label{table:quenya}
\texttt{\begin{tabular}{|ll|ll|ll|ll|}\hline
\Quenya{t}          &t
&\Quenya{p}         &p
&\Quenya{k}         &c, k
&\Quenya{\color{blue}q}         &q, qu\\
\Quenya{d}          &d, nd
&\Quenya{b}         &b, mb
&\Quenya{\color{blue}g}     &g, ng
&\Quenya{\color{blue}gw}    &gw, ngw\\
\Quenya{\color{red}S}           &th, S
&\Quenya{f'}         &f, ph
&\Quenya{\color{red}ch} & h***, ch, kh
&\Quenya{\color{red}hw} &hw, chw, khw\\
\Quenya{nt} &nt
&\Quenya{\color{blue}mp}    &mp
&\Quenya{\color{blue}nc}&nc
&\Quenya{nq}    &nq, nqu\\
\Quenya{n}          &n
&\Quenya{m}         &m
&\Quenya{\color{red}ng}         &*ng, N
&\Quenya{\color{red}ngw} &*nw, *ngw, Nw\\
\Quenya{r}          &r
&\Quenya{v}         &v, f*
&\Quenya{\color{green}Y}            &Y
&\Quenya{w}         &w, V\\\hline
\Quenya{R}          & r**, R
&\Quenya{rd}    	&rd
&\Quenya{l}         &l
&\Quenya{L}         &ld, L\\
\Quenya{s}          &s
&\Quenya{C}         & s**, C
&\Quenya{z}         &ss, z
&\Quenya{Z}         &ss**, Z\\
\Quenya{h}          &h
&\Quenya{\color{green}Hw}   & Hw
&\Quenya{I}         &**i, I
&\Quenya{U}         &**u, U\\\hline
\Quenya{\color{blue}H}      &h***, H
&\Quenya{\color{blue}x}         &x, cs, ks
&\Quenya{\color{green}A}         &A
&\Quenya{y}         &y\\\hline
\Quenya{a}          &a
&\Quenya{o}         &o  
&\Quenya{ai}        &ai
&\Quenya{au}        &au
\\
\Quenya{e}         &e
&\Quenya{u}          &u
&\Quenya{oi}        &oi 
&\Quenya{\color{red}eu} &eu
\\
\Quenya{i}         &i
&&
&\Quenya{ui}         &ui
&\Quenya{\color{red}iu}     &iu \\\hline
\end{tabular}}

\bigskip

\parbox{\textwidth}{\parindent = 3ex The letters shown in black are
used in \cite{namarie}; the letters shown in blue are not used in
\cite{namarie}, but are described explicitly in \cite{appE}; the
letters shown in green are probably not used in this mode; the letters
and combinations shown in red are deduced or have values which changed
from Age to Age.

\texttt{h} produces \Quenya{h} normally, \Quenya{H} before
\texttt{r}/\texttt{l}, and \Quenya{ch} before~\texttt{t}.}

\bigskip

{General mode}

\bigskip

\texttt{\begin{tabular}{|ll|ll|ll|ll|ll|}\hline
\Quenyagen{t   }       &t
&\Quenyagen{p  }       &p
&\Quenyagen{tj}   &tj
&\Quenyagen{c  }       &c, k    
&\Quenyagen{q  }       &q, qu\\
\Quenyagen{d   }       &d,
&\Quenyagen{b  }       &b
&\Quenyagen{j }       &j
&\Quenyagen{g  }       &g   
&\Quenyagen{gw  }       &gw   \\
\Quenyagen{th}     &th
&\Quenyagen{f' }       &ph, f
&\Quenyagen{sh}   &       sh
&\Quenyagen{kh}    &ch, kh 
&\Quenyagen{khw}    &chw, khw \\
\Quenyagen{dh} &dh
&\Quenyagen{v  }       &v, f*
&\Quenyagen{zh}   &           zh, ssh
&\Quenyagen{gh} &gh 
&\Quenyagen{ghw} &ghw \\
\Quenyagen{n   }       &n
&\Quenyagen{m  }       &m
&\Quenyagen{nY}      &nY, ny**
&\Quenyagen{ng}         &ng*, *ng   
&\Quenyagen{ngw}         &*ngw   \\
\Quenyagen{r    }       &r
&\Quenyagen{w   }       &w
&\Quenyagen{Y   }       &Y, y**
&\Quenyagen{@}        &@  
&&\\\hline
\Quenyagen{ R   }       &R
&\Quenyagen{rd  }       &rd
&\Quenyagen{l  }       &l
&\Quenyagen{ld}          &ld, L \\
\Quenyagen{s   }       &s
&\Quenyagen{C   }       &C, s**
&\Quenyagen{z  }       &z, ss
&\Quenyagen{Z}      &Z, ss**   \\
\Quenyagen{h   }       &h
&\Quenyagen{hw  }       &hw
&\Quenyagen{E}          &E
&\Quenyagen{U}  &U\\ \hline
\Quenyagen{H}         &H, h***
&\Quenyagen{x}        &x
&\Quenyagen{{\color{gray}tE}\tengfont{\char202}}      &E
&\Quenyagen{A}     &A
&\Quenyagen{dw}        &dw
\\ \hline
\end{tabular}}

\bigskip

Gray characters are included for illustration; E produces just the underdot.
\end{table}



\begin{table}[tp]
\caption{Mode of Beleriand}\label{table:beleriand}
\texttt{\begin{tabular}{|ll|ll|ll|ll|}\hline
\Beleriand{t    }       &t
&\Beleriand{p   }       &p
&\Beleriand{c   }       &c, k, C
&\Beleriand{\color{red}q    }       &q, qu  \\
\Beleriand{d    }       &d, D
&\Beleriand{b   }       &b
&\Beleriand{g   }       &g, G
&\Beleriand{\color{red}gw}      &gw, Gw\\
\Beleriand{th}  &th
&\Beleriand{f'}     &f, ph, F
&\Beleriand{ch} &ch, kh, Ch
&\Beleriand{\color{red}Chw}&chw, khw, Chw\\
\Beleriand{dh}  &dh
&\Beleriand{v   }       &v, f*
&\Beleriand{\color{red}gh}  &gh
&\Beleriand{\color{red}ghw} &ghw    \\
\Beleriand{N}           &nn, N
&\Beleriand{mm}     &mm, M
&\Beleriand{\color{red}'Ng'}  &ng*, *ng, Ng
&\Beleriand{\color{green}\~n}   &\char`\\\char`\~n, *ngw, *nw  \\
\Beleriand{n    }       &n
&\Beleriand{m   }       &m
&\Beleriand{o   }       &o
&\Beleriand{w   }       &w  \\\hline
\Beleriand{r    }       &r, R
&\Beleriand{\color{red} rh}     &*rh
&\Beleriand{l   }       &l
&\Beleriand{\color{red}lh}      &*lh, L \\
\Beleriand{s    }       &s
&\Beleriand{y   }       &y
&\Beleriand{\color{red}ss}      &ss, z
&\Beleriand{\color{green}Z  }       &Z  \\
\Beleriand{h    }       &h
&\Beleriand{\color{red}hw}  &hw
&\Beleriand{e   }       &e
&\Beleriand{u   }       &u  \\\hline
\Beleriand{a},  \tengfont{]G}       &a
&\Beleriand{i   }       &i
&\Beleriand{\color{red}Y}       &I, Y, y**, *i**
&\Beleriand{\color{red}mh}  &mh \\
\Beleriand{\color{red}dw}           &dw, Dw
&\Beleriand{\color{green}oi  }       &oi
&\Beleriand{\color{green}eu  }       &eu
&\Beleriand{\color{green}'iu  }       &iu
\\
\Beleriand{ai}      &ai
&\Beleriand{ei}     &ei
&\Beleriand{ui}     &ui
&\Beleriand{\color{red}au}      &au, aw*    \\\hline
\end{tabular}}

\bigskip

\parbox{\textwidth}{The letters shown in black are used in
\cite{elbereth} or \cite{durin}; the letters shown in green are
probably not used in this mode; the letters and combinations shown in
red are deduced based on \cite{appE}, analogy with Quenya, or analogy
with the mode of Arnor or English full modes.} \end{table}

\begin{table}[tp]
\caption{Third Age Sindarin modes}\label{table:sindarin}
\texttt{\begin{tabular}{|ll|ll|ll|ll|}\hline
\Gondor{t   }       &t
&\Gondor{p  }       &p
&\Gondor{\color{green}tj}   &tj
&\Gondor{c  }       &c, k, C    \\
\Gondor{d   }       &d, D
&\Gondor{b  }       &b
&\Gondor{\color{green}j }       &j
&\Gondor{g  }       &g, G   \\
\Gondor{th}     &th
&\Gondor{f' }       &ph, f, F
&\Gondor{\color{green}sh}   &       sh
&\Gondor{kh}    &ch, kh \\
\Gondor{dh} &dh
&\Gondor{v  }       &v, f*
&\Gondor{\color{green}zh}   &           zh, ssh
&\Gondor{\color{red}gh} &gh \\
\Gondor{n   }       &n
&\Gondor{m  }       &m
&\Gondor{\color{green}nY}      &nY
&\Gondor{\color{red}'Ng'}         &ng*, *ng, Ng  \\
\Gondor{    }       &
&\Gondor{\color{red}w   }       &w
&\Gondor{   }       &
&\Gondor{\color{green}@}        &@  \\\hline
\Gondor{    }       &
&\Gondor{\color{red}rh  }       &*rh, hr
&\Gondor{l  }       &l
&\Gondor{\color{red}L}          &*lh, hl, L \\
\Gondor{s   }       &s
&\Gondor{   }       &
&\Gondor{z  }       &z, ss
&\Gondor{}      &   \\
\Gondor{h   }       &h
&\Gondor{\color{red}hw  }       &hw
&\Gondor{}          &
&\Gondor{}  &\\\hline
\Gondor{mh}         &mh
&\Gondor{dw}        &dw, Dw
&\Gondor{gw}        &gw, Gw
&\Gondor{\color{red}chw }       &chw, Chw\\ \hline
\end{tabular}}

\bigskip

Mode of Arnor

\bigskip

\texttt{\begin{tabular}{|ll|ll|ll|ll|}\hline
\Arnor{r    }       &r
&\Arnor{y   }       &y
&\Arnor{o   }       &o
&\Arnor{\color{green}Z  }       &Z  \\
\Arnor{R    }       &R
&&
&\Arnor{e   }       &e
&\Arnor{u   }   &u
\\\hline
\Arnor{a'},\Arnor{a}&   a
&\Arnor{ae  }       &ae
&\Arnor{ai}     &ai
&\Arnor{\color{green}oi}     &oi
\\
\Arnor{i}          &       i
&\Arnor{\color{red}oe   }       &oe
&\Arnor{ei}     &ei 
&\Arnor{\color{green}eu}           &   eu
\\
\Arnor{I}          &   I, Y, y**, *i**
&\Arnor{\color{red}au}      &au, aw*
&\Arnor{\color{red}ui}      &ui
&\Arnor{\color{green}'iu}    &iu
\\\hline
\end{tabular}}

\bigskip

Mode of Gondor

\bigskip

\texttt{\begin{tabular}{|ll|ll|ll|ll|ll|}\hline
\Gondor{r   }       &R, r*
&\Gondor{S  }       &S, **s
&\Gondor{I  }       &I, **i
&\Gondor{\color{red}Z   }       &Z, **ss    
&&\\
\Gondor{r'  }       &r
&\Gondor{}      &
&\Gondor{E}         &E, Y, **e, y**, *i**
&\Gondor{\color{red}U   }   &U, **u
&\Gondor{\color{green}A}	&A
\\\hline
\Gondor{a}          &a
&\Gondor{o}         &o
&\Gondor{ae }       &ae
&\Gondor{ai}        &ai
&\color{green}\Gondor{oi} &oi
\\
\Gondor{e}          &e
&\Gondor{u}         &u
&\Gondor{\color{red}oe}     &oe
&\Gondor{ei}        &ei
&\color{green}\Gondor{eu} &eu
\\
\Gondor{i}          &i
&\Gondor{y}         &y
&\Gondor{\color{red}au}     &au, aw*
&\Gondor{ui}            &ui
&\color{green}\Gondor{'iu} &iu
\\
\hline
\end{tabular}}

\bigskip

\parbox{\textwidth}{The characters which are shown in black occur in
\cite{kingsletter}; the letters shown in green are probably not used
in this mode; the letters or combinations shown in red have their
values deduced from \cite{appE}, \cite{lotr}, similar letters or
combinations, or other modes.} \end{table}

\begin{table}[tp]
\caption{English orthographic modes}\label{table:EO}
\texttt{\begin{tabular}{|cc|cc|cc|cc|}\hline
\Orthmode{t}        &{t, T}
&\Orthmode{p}       &{p}
&\Orthmode{ch}  &{ch} %Letter, not Mazarbul
&\Orthmode{k}       &{k, c}
\\
\Orthmode{d}        &{d, D}
&\Orthmode{b}       &{b, B}
&\Orthmode{\color{blue}j}       &{j}
&\Orthmode{g}       &{g, G}
\\
\Orthmode{th}   &{th, Th}
&\Orthmode{f}       &{f}
&\Orthmode{sh}  &{sh}
&\Orthmode{\color{red}Ch}   &Ch, Kh
\\
\Orthmode{dh}   &tH, dh%
&\Orthmode{\color{blue}v}       &{v} %Mazarbul, not letter
&\Orthmode{\color{green}zh}     &zh
&\Orthmode{gh}  &{gh} %Letter, not Mazarbul
\\
\Orthmode{n}        &{n}
&\Orthmode{m}       &{m}
&\Orthmode{\color{green}\~n}    &\char`\\\char`\~n \textrm{(\~n)}
&\Orthmode{ng}  &{ng}
\\
\Orthmode{r}        &{r}
&\Orthmode{}        &
&\Orthmode{}        &
&\Orthmode{}        &
\\\hline
\Orthmode{R}        &{r**, R}
&\Orthmode{\color{red}Rh}   &{Rh}
&\Orthmode{l}       &{l}
&\Orthmode{L}       &ll, L
\\
\Orthmode{s}        &{s, S\phantom{*}}
&\Orthmode{C}       &{C} %Letter, not Mazarbul
&\Orthmode{\color{blue}z}       &{z}     %Letter, not Mazarbul
&\Orthmode{Z}       &Z, **z      %LOTR only
\\
\Orthmode{h}        &{h}
&\Orthmode{}    &
&\Orthmode{}        &
&\Orthmode{}        &
\\\hline
\Orthmode{tW}       &{tW, TW}
&\Orthmode{\color{red}dW}       &{dW, DW}
&\Orthmode{\color{red}bU}       &{bU, BU}
&\Orthmode{\color{purple}SW}        &{sW, SW}
\\
\Orthmode{\color{red}q}         &{q, qu}
&\Orthmode{\color{red}thW}      &{thW, ThW}
&\Orthmode{\color{red}chW}      &{chW}
&\Orthmode{\color{red}ChW}      &{ChW}
\\
\Orthmode{\color{purple}Ng}     &{Ng}
&\Orthmode{\color{red}ngU}      &{ngU, ngW}
&\Orthmode{\color{red}GW}       &{gU, gW}
&\Orthmode{\color{red}NGW}      &{NgU, NgW}
\\
\Orthmode{\color{blue}CH}       &{CH, KH, cH, kH}%, \color{purple}K
&\Orthmode{\color{red}ph}       &ph%, \color{purple}P
&\Orthmode{\color{blue}x}       &{x}
&&
\\
\hline
\end{tabular}}

\bigskip

Aragorn's mode (\texttt{\char92orthmode})

\bigskip

\texttt{\begin{tabular}{|cc|cc|cc|cc|}\hline
&
&\Orthmode{u}       &u
&\Orthmode{o}       &o
&\Orthmode{A}       &a, A
\\
&
&\Orthmode{\color{red}wh}   &{wh, Wh}
&\Orthmode{e}       &e
&\Orthmode{w}       &w
\\\hline
\Orthmode{\color{green}@}       &@
&\Orthmode{i}       &i
&\Orthmode{y}       &y
&\Orthmode{H}&H
\\
\Orthmode{ai}       &ai, ay
&\Orthmode{ei}      &ei, ey
&\Orthmode{\color{red}oi}       &oi, oy
&\Orthmode{uy}      &uI, uy
\\
\Orthmode{au}       &au, aW
&\Orthmode{\color{red}eu}       &eu, eW
&\Orthmode{\color{blue}oU}      &ou, ow
&&
\\
\Orthmode{\color{red}ae}        &ae
&&
&\Orthmode{\color{red}oe}       &oe
&\Orthmode{\color{red}ue}       &ue
\\
\Orthmode{\color{red}aa}        &aa
&\Orthmode{ee}          &ee
&\Orthmode{\color{blue}oo}      &oo
&\Orthmode{\color{red}uu}       &uu
\\
\hline
\end{tabular}}

\bigskip

{Mode of Mazarbul: changes from \texttt{\char92orthmode}}

\bigskip

\texttt{\begin{tabular}{|cc|cc|cc|cc|}\hline
\Mazarbul{w}&   w, W
&\Mazarbul{\color{red}wh}       &wh
&\Mazarbul{\color{red}Wh}       &Wh
&\Mazarbul{\color{green}wH}     &wH, WH
%
\\\hline
\end{tabular}}

\bigskip

\parbox{\hsize}{The symbols shown in black are used in
\cite{kingsletter}; the symbols shown in green are probably not used
in this mode; the symbols shown in blue are attested in \cite{brogan};
the symbols shown in purple are borrowed from the phonemic mode
(\cite{luthien} or \cite{bombadil}). In all cases, you can use
\texttt{G} instead of \texttt{g}.}

\bigskip

{English orthographic tehta mode}

\bigskip

\texttt{\begin{tabular}{|cc|cc|cc|cc|cc|cc|}\hline
\multicolumn{1}{|c}{\Tehtamode{}}&\multicolumn{2}{c|}{}&
\multicolumn{1}{c}{\Tehtamode{w}}&\multicolumn{2}{c|}{w}&
\multicolumn{1}{c}{\Tehtamode{Y}}&\multicolumn{2}{c|}{Y, I}&
\multicolumn{1}{c}{\color{orange}\Tehtamode{O}}&\multicolumn{2}{c|}{O}
\\
\multicolumn{1}{|c}{\Tehtamode{}}&\multicolumn{2}{c|}{}&
\multicolumn{1}{c}{\Tehtamode{wh}}&\multicolumn{2}{c|}{wh, Wh}&
\multicolumn{1}{c}{\Tehtamode{E}}&\multicolumn{2}{c|}{E}&
\multicolumn{1}{c}{\color{red}\Tehtamode{U}}&\multicolumn{2}{c|}{U}
\\
\hline
\Tehtamode{a}       &a
&&
&\Tehtamode{A}      &   A
&\Tehtamode{ae}     &ae
&\Tehtamode{au}     &au
&\Tehtamode{ai}     &ai, ay
\\
\Tehtamode{e}       &e
&\Tehtamode{ee}     &ee
&\Tehtamode{ea}     &ea
&&
&\Tehtamode{eu}     &eu
&\Tehtamode{ei}     &ei, ey
\\
 \Tehtamode{i}      &i
&\Tehtamode{ii}     &ii
&\Tehtamode{}       &
&\Tehtamode{ie}     &ie
&&
&\Tehtamode{iy}     &iy
\\
 \Tehtamode{o}      &o
&\Tehtamode{oo}     &oo
&\Tehtamode{oa}     &oa
&\Tehtamode{oe}     &oe
&\Tehtamode{ou}     &ou
&\Tehtamode{oi}     &oi, oy
\\
 \Tehtamode{u}      &u
&\Tehtamode{uu}     &uu
&\Tehtamode{}       &
&\Tehtamode{ue}     &ue
&&
&\Tehtamode{uy}     &uy
\\
 \Tehtamode{y}      &y
 &&&&
&\Tehtamode{ye}     &ye
&&&&
\\
\hline
\end{tabular}}

\bigskip

\parbox{\hsize}{There aren't really enough examples to make it worth
listing everything that's attested and everything that's not. The
\Tehtamode{\color{red}U} is debatable, and the
\Tehtamode{\color{orange}O} is \emph{spectacularly} wild guesswork.}
\end{table}

\begin{table}[tp]
\caption{Sarati Quenya}\label{table:sarati}
\texttt{\begin{tabular}{|cc|cc|cc|cc|cc|}
\hline
\SaratiA{0} &0&
\SaratiA{1} &1&
\SaratiA{2} &2&
\SaratiA{3} &3&
\SaratiA{4} &4\\
%
\SaratiA{5} &5&
\SaratiA{6} &6&
\SaratiA{7} &7&
\SaratiA{8} &8&
\SaratiA{9} &9\\
%
\SaratiA{<} &<&
\SaratiA{>} &>&
&&\Sarati{.}&.&\Sarati{:}&:\\
\hline
%
\SaratiA{p} &p&
\SaratiA{t} &t&
\SaratiA{ty}    &ty&
\SaratiA{c} &c, k&
\SaratiA{q} &q, cw, kw\\
%
\SaratiA{b} &b, mb&
\SaratiA{d} &d, nd&
\SaratiA{dy}    &dy, ndy&
\SaratiA{g} &g, ng&
\SaratiA{gw}    &gw, ngw\\
%
\SaratiA{f} &f, ph&
\SaratiA{s} &s&
\SaratiA{hy}    &hy&
\SaratiA{h} &h&
\SaratiA{hw}    &hw\\
%
\SaratiA{v} &v&
\SaratiA{r} &r&
\SaratiA{y} &y, **i&
\SaratiA{l} &l&
\SaratiA{w} &w, **u\\
%
\SaratiA{m} &m&
\SaratiA{n} &n&
\SaratiA{ny}    &ny&
\SaratiA{N} &*ng, N&
\SaratiA{Nw}    &*nw, Nw\\
%
\SaratiA{mp}    &mp&
\SaratiA{nt}    &nt&
\SaratiA{nty}   &nty&
\SaratiA{nk}    &nk, nc&
\SaratiA{nq}    &nq\\
%
\SaratiA{ts}    &ts&
\SaratiA{st}    &st&
\SaratiA{sty}   &sty&
\SaratiA{x} &x, cs, ks&
\SaratiA{ss}    &ss\\
%
&&
\SaratiA{ht}    &ht&
\SaratiA{hty}   &hty&
&&
&\\
\hline
\SaratiA{a} &a&
\SaratiA{e} &e&
\SaratiA{i} &i&
\SaratiA{o} &o&
\SaratiA{u} &u\\
%
\SaratiA{aa}    &aa&
\SaratiA{ee}    &ee&
\SaratiA{ii}    &ii&
\SaratiA{oo}    &oo&
\SaratiA{uu}    &uu\\
\hline
\color{red}\SaratiA{hl} &hl&
\color{red}\SaratiA{hr} &hr&
\color{red}\SaratiA{ly} &ly&
\color{red}\SaratiA{ry} &ry&&\\
%
%\color{red}\SaratiA{ps} &ps&
%\color{red}\SaratiA{sc} &sc&
%&&&&&\\
\color{red}\SaratiA{th} &th, S&
\color{red}\SaratiA{ch} &ch, kh&
\color{red}\SaratiA{khw}    &chw, khw&
\color{red}\SaratiA{z}  &z&
&\\
\color{red}\SaratiA{mf} &mf, mph&
\color{red}\SaratiA{dh} &dh&
&&&&&\\
\hline
\end{tabular}}

\bigskip

\parbox{\hsize}{\parindent=3ex Symbols in red are taken from the
phonemic valuation; otherwise this valuation is based on \cite{sarati}.}

\bigskip

\caption{Valmaric Quenya}\label{table:valmaric}
\texttt{\begin{tabular}{|cc|cc|cc|cc|cc|}\hline
\ValmaricA{p}   &p&
\ValmaricA{t}   &t&
\ValmaricA{c}   &c, k&
\ValmaricA{ty}  &ty&
\ValmaricA{q}   &q\\
%
\ValmaricA{f}   &f&
\ValmaricA{s}   &s, S, th&
\ValmaricA{h}   &h, H&
&&
\ValmaricA{hw}  &hw\\
%
\ValmaricA{v}   &v&
\ValmaricA{r}   &r, R&
&&
\ValmaricA{y}   &y, **i&
\ValmaricA{w}   &w, **u\\
%
\ValmaricA{mp}  &mp&
\ValmaricA{nt}  &nt&
\ValmaricA{nc}  &nc, nk&
\ValmaricA{nty} &nty&
\ValmaricA{nq}  &nq\\
%
\ValmaricA{b}   &b, mb&
\ValmaricA{d}   &d, nd&
\ValmaricA{g}   &g, ng&
&&
\ValmaricA{gw}  &gw, ngw\\
%
\ValmaricA{m}   &m&
\ValmaricA{n}   &n, N&
&&
\ValmaricA{ny}  &ny&
\ValmaricA{Nw}  &*nw, Nw\\
%
\ValmaricA{mn}  &mn&
\ValmaricA{l}   &l&
\ValmaricA{rd}  &rd&
\ValmaricA{z}   &z, ss&
\ValmaricA{ld}  &ld\\
%
\ValmaricA{rm}  &rm&
\ValmaricA{lt}  &lt&
\ValmaricA{ht}  &ht&
\ValmaricA{ry}  &ry&
\ValmaricA{lq}  &lq\\
\hline
\ValmaricA{a}   &a&
\ValmaricA{e}   &e&
\ValmaricA{i}   &i&
\ValmaricA{o}   &o&
\ValmaricA{u}   &u\\
%
\ValmaricA{aa}  &aa&
\ValmaricA{ee}  &ee&
\ValmaricA{ii}  &ii&
\ValmaricA{oo}  &oo&
\ValmaricA{uu}  &uu\\
\hline
\ValmaricA{.}   &.&
\ValmaricA{--}  &--&
\ValmaricA{---} &---\\
\cline{1-6}
\end{tabular}}
\bigskip

\parbox{\hsize}{\parindent=3ex This valuation is based entirely on
\cite{valmar}.}

\end{table}

\newcommand{\cirthline}[5]{\Erebor{\ignorespaces
%#5}&\Daeron{#1}&\Eregion{#2}&\Moria{#3}&\Erebor{#4}\\}
#5}&{#1}&{#2}&{#3}&{#4}\\}

\begin{table}[hp]

\caption{Cirth}\label{table:cirth}\def\'{}\def\-{*}
\tt
\begin{tabular}[t]{|c|c@{ }c@{ }c@{ }c|}
\hline
&\rm{Daeron}&\rm{Eregion}&\rm{Moria}&\rm Erebor\\
\hline
\cirthline{p    }{p }{p }{p }{p }
\cirthline{b    }{b }{b }{b }{b }
\cirthline{f\'  }{f }{f }{f }{f }
\cirthline{v    }{v }{v }{v }{v }
\cirthline{hw   }{hw    }{hw    }{hw    }{hw    }
\cirthline{m    }{m }{m }{m }{m }
\cirthline{mh   }{mh    }{mb    }{mb    }{mb    }
\cirthline{t    }{t }{t }{t }{t }
\cirthline{d    }{d }{d }{d }{d }
\cirthline{th   }{th    }{th    }{th    }{th    }
\cirthline{dh   }{dh    }{dh    }{dh,\,tH   }{tH}
\cirthline{n    }{n }{r }{r }{r }
\cirthline{tj    }{tj }{ch    }{ch    }{ch    }
\cirthline{j    }{j }{- }{j }{j }
\cirthline{sh   }{sh    }{sh    }{sh    }{sh    }
\cirthline{zh   }{zh    }{- }{zh    }{zh    }
\cirthline{nj   }{nj    }{z }{x }{x }
\cirthline{k    }{k }{k }{c, k  }{c }
\cirthline{g    }{g }{g }{G }{G }
\cirthline{ch,kh  }{ch,kh,h*   }{kh    }{Ch, Kh    }{Ch    }
\cirthline{gh   }{gh    }{gh    }{gh    }{gh    }
\cirthline{\-ng\-,Ng   }{\-ng\-,N    }{n }{n }{n }
\cirthline{q    }{q }{q }{q }{q }
\cirthline{gw   }{gw    }{gw    }{gw    }{gw    }
\cirthline{khw  }{khw   }{khw   }{Khw   }{Khw   }
\cirthline{ghw  }{ghw   }{ghw   }{ghw   }{ghw   }
\cirthline{ngw  }{\'ngw }{ngw   }{Ngw   }{Ngw   }
\cirthline{\-ngw\',Ngw }{\-nw\',Nw    }{nw    }{nw    }{nw    }
\cirthline{r    }{r }{j }{g }{g }
\cirthline{\-rh,hr   }{\-rh,hr   }{zh    }{Gh    }{Gh    }
\hline
\end{tabular}
\begin{tabular}[t]{|c|c@{ }c@{ }c@{ }c|}
\hline
&\rm{Daeron}&\rm{Eregion}&\rm{Moria}&\rm Erebor\\
\hline
\cirthline{l    }{l }{l }{l }{l }
\cirthline{lh,hl    }{lh,hl }{lh,hl }{L }{L }
\cirthline{\'ng }{ng\'  }{nd    }{nd    }{nd    }
\cirthline{s    }{s }{s }{s }{s }
\cirthline{S    }{S }{S }{S }{S }
\cirthline{ss,z   }{ss,z    }{Ng    }{ng    }{ng    }
\cirthline{Z    }{Z }{ng    }{Ng    }{Ng    }
\cirthline{nd   }{nd    }{nj    }{oU    }{oU    }
\cirthline{nD,Nd   }{nD,Nd    }{Nj    }{ou    }{ou    }
\cirthline{i    }{i }{i }{i }{i }
\cirthline{-    }{- }{y }{y }{y }
\cirthline{-    }{- }{hy    }{hy    }{hy    }
\cirthline{u    }{u }{u }{u }{u }
\cirthline{uu   }{uu    }{uu    }{z }{z }
\cirthline{w    }{w }{w }{w }{w }
\cirthline{\"u}{\"u}{\"u}{\"u}{\"u  }
\cirthline{e    }{e }{e }{e }{e }
\cirthline{ee   }{ee    }{ee    }{ee    }{ee    }
\cirthline{a    }{a }{a }{a }{a }
\cirthline{aa   }{aa    }{aa    }{aa    }{aa    }
\cirthline{o    }{o }{o }{o }{o }
\cirthline{oo   }{oo    }{oo    }{oo    }{oo    }
\cirthline{\"o}{\"o}{\"o}{\"o}{\"o }
\cirthline{-    }{- }{N }{N }{N }
\cirthline{h    }{h }{h }{h }{h }
\cirthline{-    }{- }{E }{E }{E }
\cirthline{-    }{- }{U }{U }{U }
%
\cirthline{H    }{H }{H }{H }{H }
\cirthline{\&   }{\&    }{\&    }{\&    }{\&    }
\hline
\end{tabular}

\bigskip

\begin{tabular}{|cccccccc|}
\hline
\multicolumn{8}{|c|}{\rm Mode of Erebor only}\\\hline
\Erebor{ps}&\texttt{ps}&
\Erebor{au}&\texttt{au}&
\Erebor{ea}&\texttt{ea}&
\Erebor{eu}&\texttt{eu}\\
\Erebor{ts}&\texttt{ts}&
\Erebor{ai}&\texttt{ai}&
\Erebor{oa}&\texttt{oa}&
\Erebor{ll}&\texttt{ll}\\
\hline
\end{tabular}

\bigskip

\parbox{\hsize}{\rm\parindent=3ex You may substitute \texttt{\char`\\'a} 
or \texttt{\char`\\\char`\^a} for \texttt{aa} in all modes.}
\end{table}

\begin{table} \sunrune

\caption{Hobbit runes}\label{table:hobbit}
\ttfamily

\begin{tabular}[t]{|c@{}cc|}
\hline
\rule[2pt]{0pt}{9pt}& \Hobbit{a}& a \\
\rule[2pt]{0pt}{9pt}& \Hobbit{b}& b \\
\rule[2pt]{0pt}{9pt}& \Hobbit{c}& c \\
\rule[2pt]{0pt}{9pt}& \Hobbit{d}& d \\
\rule[2pt]{0pt}{9pt}& \Hobbit{e}& e \\
\rule[2pt]{0pt}{9pt}& \Hobbit{f}& f \\
\rule[2pt]{0pt}{9pt}& \Hobbit{g}& g \\
\rule[2pt]{0pt}{9pt}& \Hobbit{h}& h \\
\rule[2pt]{0pt}{9pt}& \Hobbit{i}& i \\
\rule[2pt]{0pt}{9pt}& \Hobbit{j}& j \\
\rule[2pt]{0pt}{9pt}& \Hobbit{k}& k \\
\rule[2pt]{0pt}{9pt}& \Hobbit{l}& l \\
\rule[2pt]{0pt}{9pt}& \Hobbit{m}& m \\
\hline\end{tabular}\begin{tabular}[t]{|c@{}cc|}\hline
\rule[2pt]{0pt}{9pt}& \Hobbit{n}& n \\
\rule[2pt]{0pt}{9pt}& \Hobbit{o}& o \\
\rule[2pt]{0pt}{9pt}& \Hobbit{p}& p \\
\rule[2pt]{0pt}{9pt}& \Hobbit{q}& q \\
\rule[2pt]{0pt}{9pt}& \Hobbit{r}& r \\
\rule[2pt]{0pt}{9pt}& \Hobbit{s}& s \\
\rule[2pt]{0pt}{9pt}& \Hobbit{t}& t \\
\rule[2pt]{0pt}{9pt}& \Hobbit{u}& u \\
\rule[2pt]{0pt}{9pt}& \Hobbit{v}& v \\
\rule[2pt]{0pt}{9pt}& \Hobbit{w}& w \\
\rule[2pt]{0pt}{9pt}& \Hobbit{x}& x \\
\rule[2pt]{0pt}{9pt}& \Hobbit{y}& y \\
\rule[2pt]{0pt}{9pt}& \Hobbit{z}& z \\
\hline\end{tabular}\begin{tabular}[t]{|c@{}cc|}\hline
\rule[2pt]{0pt}{9pt}& \Hobbit{th}& th \\
\rule[2pt]{0pt}{9pt}& \Hobbit{ng}& ng \\
\rule[2pt]{0pt}{9pt}& \Hobbit{oo}& oo \\
\rule[2pt]{0pt}{9pt}& \Hobbit{ea}& ea \\
\rule[2pt]{0pt}{9pt}& \Hobbit{ee}& ee \\
\rule[2pt]{0pt}{9pt}& \Hobbit{eo}& eo \\
\rule[2pt]{0pt}{9pt}& \Hobbit{oa}& oa \\
\rule[2pt]{0pt}{9pt}& \Hobbit{sT}& sT \\
\rule[2pt]{0pt}{9pt}& \Hobbit{\p}& \char`\\p \\
\rule[2pt]{0pt}{9pt}& \Hobbit{.}& . \\
\rule[2pt]{0pt}{9pt}& \Hobbit{,}& , \\
%
\hline
\end{tabular}
\rm

\bigskip

\parbox{\hsize}{\parindent=3ex
Most of the letters in this mode are
from \cite{thror}, except for \Hobbit{x},~\Hobbit{z}, \Hobbit{ea}, \Hobbit{ST},
\Hobbit{j}~for~\ment{j}, \Hobbit{q}~for~\ment{q}, which are from
\cite{foreword}, and \Hobbit{oo}, \Hobbit{oa}, which are from \cite{farrer},
and \Hobbit{eo}, which is from \cite{unwin}.

Note that \Hobbit{\p}, \Hobbit{ST} occur in \cite{foreword} but not in \cite{thror}, and
so they are a bit harder to get than \Hobbit{p},~\Hobbit{st}.

Uppercase letters may be used instead of lowercase letters; they are
identical except that \texttt{NG}/\texttt{Ng} produces \Hobbit{NG},
and \texttt{St}/\texttt{st} produces \Hobbit{st}.

Fonts designed for \emph{Cirth} do not always contain all
the glyphs one might like in a \emph{runic} font. Cirth Erebor does
far better than most Cirth fonts, but even it does not have
\Hobbit{p},~\Hobbit{eo}, \Hobbit{oa}, or~\Hobbit{ST}. Getting all these glyphs was the original motivation for creating the sunrune font.
}

\end{table}

\clearpage

\section{Revisions}

In Version~2, I've made some significant changes. \texttt{quenyagen} now
uses the quesse\-t\'ema instead of the calma\-t\'ema for its
\ment{k}-series; this means that it is possible to render English words
with it, at the expense of adding some extra tehtar for \ment{qu}
and~\ment{ngw}.

In general, more support has been added for writing isolated Quenya
words in a nominally Sindarin mode or vice versa. This has almost (but
not completely) eliminated the difference between \texttt{daeron} and
\texttt{eregion}.

The fonts have been modified to make accents work better in general. For
technical reasons, this means that some of the \ment{s}-curls have been
moved.

\texttt{Phonemic} has been edited to have better support for writing in
the modes of \cite{thorin} and \cite{treebeard}; these texts can now be
copied without using \verb.\tengfont. (although some of the things you
do have to type are nonintuitive).

Full support for the Sunrune font has been added.

A number of miscellaneous bug fixes and minor font improvements have
also been made.

\section{Files in this package}\label{sec:manifest}

This is a list of all files that are considered part of the
\textsf{tolkienfonts} package. They should be organized as follows:

{\tt

\begin{itemize}
\item CHANGES
\item tolkienfontsdoc.pdf
\item tolkienfontsdoc.tex
\item example.tex
\item example.pdf
\item README
\end{itemize}

\noindent tolkieninst/

\begin{itemize}
\item effectColor.mtx
\item effectPunctspace.mtx
\item effectSuppressKerns.mtx
\item effectUnderdots.mtx
\item fancykernsDM.mtx
\item fancykernsEE.mtx
\item fancykernsEH.mtx
\item fancykernsSD.mtx
\item fancykernssetup.mtx
\item fontAnnatar.mtx
\item fontAnnatarb.mtx
\item fontAnnatarbi.mtx
\item fontAnnatari.mtx
\item fontEldamar.mtx
\item fontErebor.mtx
\item fontNoldor.mtx
\item fontParmaite.mtx
\item fontQuenya.mtx
\item fontSaratiE.mtx
\item fontSaratiEBar.mtx
\item fontSindarin.mtx
\item fontSunrune.mtx
\item fontValmaricE.mtx
\item modeArnor.etx
\item modeBeleriand.etx
\item modeCirthDaeron.etx
\item modeCirthErebor.etx
\item modeCirthMoria.etx
\item modeEmazarbul.etx
\item modeEphonemic.etx
\item modeEtehta.etx
\item modeGondor.etx
\item modeQuenya.etx
\item modeQuenyaGeneral.etx
\item modeRuneHobbit.etx
\item modeSaratiQuenya.etx
\item modeValmaricQuenya.etx
\item scriptcirth.mtx
\item scriptsarati.mtx
\item scripttengwar.mtx
\item scriptvalmaric.mtx
\item TengwarModeStartingPoint.etx
\item tolkieninst.tex
\item tolkienmetrics.mtx
\end{itemize}

\noindent tex/latex/

\begin{itemize}
\item tolkienfonts.sty
\item boustr.sty
\end{itemize}

\noindent tex/latex/fd/cirth/

\begin{itemize}
\item t1cirthedm.fd
\item t1cirtheee.fd
\item t1cirtheeh.fd
\item t1cirtheqe.fd
\item t1cirthesd.fd
\item t1cirthexx.fd
\item t1cirthsdm.fd
\item t1cirthsee.fd
\item t1cirthseh.fd
\item t1cirthsqe.fd
\item t1cirthssd.fd
\item t1cirthsxx.fd
\end{itemize}

\noindent tex/latex/fd/tengwar/

\begin{itemize}
\item t1tengwara.fd
\item t1tengwaraa.fd
\item t1tengwaraem.fd
\item t1tengwaraeo.fd
\item t1tengwaraep.fd
\item t1tengwaraet.fd
\item t1tengwaraqa.fd
\item t1tengwaraqg.fd
\item t1tengwaraqq.fd
\item t1tengwaraqr.fd
\item t1tengwarasa.fd
\item t1tengwarasb.fd
\item t1tengwarasg.fd
\end{itemize}

\noindent tex/latex/fd/valmar/

\begin{itemize}
\item t1saratib.fd
\item t1saratibqa.fd
\item t1saratibqq.fd
\item t1saratie.fd
\item t1saratieqa.fd
\item t1saratieqq.fd
\item t1valmaric.fd
\item t1valmariceqa.fd
\item t1valmariceqq.fd
\end{itemize}

\noindent fonts/afm/
\begin{itemize}
\item sunrune.afm
\end{itemize}

\noindent fonts/map
\begin{itemize}
\item tolkien.map
\end{itemize}

\noindent fonts/source/
\begin{itemize}
\item sunrune.mf
\end{itemize}

\noindent fonts/tfm/raw/

\begin{itemize}
\item CirthEXX.tfm
\item CirthSXX.tfm
\item SaratiB.tfm
\item SaratiE.tfm
\item TengwarA.tfm
\item TengwarAA.tfm
\item TengwarAAb.tfm
\item TengwarAAbi.tfm
\item TengwarAAi.tfm
\item TengwarAb.tfm
\item TengwarAbi.tfm
\item TengwarAi.tfm
\item TengwarE.tfm
\item TengwarEA.tfm
\item TengwarN.tfm
\item TengwarNA.tfm
\item TengwarP.tfm
\item TengwarPA.tfm
\item TengwarQ.tfm
\item TengwarQA.tfm
\item TengwarS.tfm
\item TengwarSA.tfm
\item Valmaric.tfm
\end{itemize}

\noindent fonts/tfm/cirth/

\begin{itemize}
\item CirthEDM.tfm
\item CirthEEE.tfm
\item CirthEEH.tfm
\item CirthEQE.tfm
\item CirthESD.tfm
\item CirthSDM.tfm
\item CirthSEE.tfm
\item CirthSEH.tfm
\item CirthSQE.tfm
\item CirthSSD.tfm
\end{itemize}

\noindent fonts/tfm/tengwar/

\begin{itemize}
\item TengwarAEM.tfm
\item TengwarAEO.tfm
\item TengwarAEP.tfm
\item TengwarAET.tfm
\item TengwarAQA.tfm
\item TengwarAQG.tfm
\item TengwarAQQ.tfm
\item TengwarAQR.tfm
\item TengwarASA.tfm
\item TengwarASB.tfm
\item TengwarASG.tfm
\end{itemize}

\noindent fonts/tfm/valmar/

\begin{itemize}
\item SaratiBQA.tfm
\item SaratiBQQ.tfm
\item SaratiEQA.tfm
\item SaratiEQQ.tfm
\item ValmaricEQA.tfm
\item ValmaricEQQ.tfm
\end{itemize}

\noindent fonts/type1/
\begin{itemize}
\item sunrune.pfb
\end{itemize}

\noindent fonts/vf/cirth/

\begin{itemize}
\item CirthEDM.vf
\item CirthEEE.vf
\item CirthEEH.vf
\item CirthEQE.vf
\item CirthESD.vf
\item CirthSDM.vf
\item CirthSEE.vf
\item CirthSEH.vf
\item CirthSQE.vf
\item CirthSSD.vf
\end{itemize}

\noindent fonts/vf/tengwar/

\begin{itemize}
\item TengwarAEM.vf
\item TengwarAEO.vf
\item TengwarAEP.vf
\item TengwarAET.vf
\item TengwarAQA.vf
\item TengwarAQG.vf
\item TengwarAQQ.vf
\item TengwarAQR.vf
\item TengwarASA.vf
\item TengwarASB.vf
\item TengwarASG.vf
\end{itemize}

\noindent fonts/vf/valmar/

\begin{itemize}
\item SaratiBQA.vf
\item SaratiBQQ.vf
\item SaratiEQA.vf
\item SaratiEQQ.vf
\item ValmaricEQA.vf
\item ValmaricEQQ.vf
\end{itemize}

}

\begin{thebibliography}{Appendix Q}

\bibitem[DTS]{DTS} The Mellonath Daeron index of tengwar specimata,\\
\href{http://www.forodrim.org/daeron/mdtci.html}
{http://www.forodrim.org/daeron/mdtci.html}

\bibitem[DCS]{DCS} The Mellonath Daeron index of cirth specimata,\\
\href{http://www.forodrim.org/daeron/mdics.html}
{http://www.forodrim.org/daeron/mdics.html}

\bibitem[Endorion]{endorion} DTS 62:
A dedication to Michael Endorion, which reads
\begin{quote}
\parbox{\hsize}{%
\Tehtamode{for miChael, son v| middlE-earth.}\\
\Quenya{elen s\rlap{\hskip -0.5pt\tengfont{T\%}}l\tengfont{\char220}
luumen omentielmo.}\\
\Tehtamode{j, R, R, t}}
\end{quote}
This is mentioned on the \textsf{elfscript} message boards, at\\
\href{http://movies.groups.yahoo.com/group/elfscript/message/2424}
{http://movies.groups.yahoo.com/group/elfscript/message/2424}\\
\href{http://movies.groups.yahoo.com/group/elfscript/message/2498}
{http://movies.groups.yahoo.com/group/elfscript/message/2498}\\
A photo is at
\href{http://movies.groups.yahoo.com/group/elfscript/files/}
{http://movies.groups.yahoo.com/group/elfscript/files/}\\
English tehta mode and Quenya mode.

\bibitem[LOTR]{lotr} DTS 5: The inscription on the title page of \ment{The
Lord of the Rings}. English tehta mode.

\bibitem[TTT]{ttt} DTS 37: The inscription on the cover of some editions
of \ment{The Two Towers}, also published in \ment{J.R.R. Tolkien: Artist
and Illustrator}. English tehta mode.

\bibitem[Brogan]{brogan} DTS 10: Letter to Hugh Brogan, \ment{The Letters of
J.R.R.\ Tolkien}~\textbf{118}, containing about two sentences apiece
written in English tehta mode, full English orthographic mode, and
Dwarvish runes (mode of Erebor).

\bibitem[Letter]{kingsletter} DTS 45, 48, and 49: Aragorn's letter to Samwise Gamgee in
the rejected epilogue to \ment{The Lord of the Rings}. There are three
drafts; the first and third are in \ment{Sauron Defeated,}
Illustration~1, and the second draft is in \ment{Artist and
Illustrator}~\textbf{199}. Sindarin mode of Gondor or Arnor (depending
on the draft) and full English orthographic mode.

\bibitem[Mazarbul]{ori} DTS 13, DCS 11--13:
The Book of Mazarbul, published in
\ment{Pictures by J.R.R. Tolkien}. Tengwar full English orthographic
mode (subtly different from that in \cite{kingsletter}) and the cirth
mode of Erebor.

\bibitem[Lowdham]{lowdham} DTS 50--51, Edwin Lowdham's manuscript, published in \ment{Sauron Defeated}. Old English mode (not supported by \textsf{tolkienfonts}).

\bibitem[Errantry]{errantry} DTS 16: The beginning of the poem
``Errantry'', published in \ment{Pictures by J.R.R. Tolkien.} English
phonemic mode.

\bibitem[Bombadil]{bombadil} DTS 17--18: Two drafts of the beginning of
the poem ``The adventures of Tom Bombadil'', published in \ment{Pictures
by J.R.R. Tolkien.} English phonemic mode.

\bibitem[Leithien]{luthien} DTS 23: The tengwar passage beginning
\tengfont{8.\char'352\ jy\char'352 3`B]5}, published in \ment{The Lays
of Beleriand}. English phonemic mode.

\bibitem[Thorin]{thorin} DTS 71:
A transcription of Thorin's letter to Bilbo
in \ment{The Hobbit}, published as the frontispiece of \ment{The
History of The Hobbit, Vol.~II: Return to Bag-End}, by John
D.~Rateliff. English phonemic mode (different from that of
\cite{bombadil}, \cite{errantry} and \cite{luthien}).

\bibitem[Treebeard]{treebeard} DTS 24:
A page describing Tolkien's original conception of Treebeard, published
in \ment{Return of the Shadow}. English phonemic mode (different from
that of \cite{errantry} or \cite{thorin}).

\bibitem[Nam\'ari\"e]{namarie} DTS 20: Transcription of Galadriel's
``Nam\'ari\"e'' lament in L\'orien, \ment{The Road Goes Ever On}.
Quenya mode.

\bibitem[Elbereth]{elbereth} DTS 21: Transcription of the elvish hymn to
Elbereth Gilthoniel, \ment{The Road Goes Ever On}. Mode of Beleriand.

\bibitem[Durin]{durin} DTS 8: The illustration of the Doors of Durin, in
\ment{The Lord of the Rings}, Chapter~II.4. Mode of Beleriand.

\bibitem[Balin]{balin} DCS 6: 
Balin's tomb inscription from \ment{The Lord of
the Rings}, II.4\@. The text is in both Dwarvish and English; the
Dwarvish text (\ment{Balin Fundinul uzbad Khazad-d\^umu}) is
definitely in the mode of Moria, but the English text (\ment{Balin son
of Fundin lord of Moria}) could be either the mode of Moria or the
mode of Erebor.

\bibitem[Thror]{thror} Thror's map, published in \ment{The Hobbit}.
Anglo-Saxon runes.

\bibitem[Jacket]{unwin} The dust jacket to the original edition of
\ment{The Hobbit}, also \ment{Artist and Illustrator}~\textbf{144}.
Anglo-Saxon runes.

\bibitem[Farrer]{farrer} Letter to Katherine Farrer, \ment{The Letters
of J.R.R.\ Tolkien}~\textbf{112}, containing a few paragraphs written
with Anglo-Saxon runes.

\bibitem[Appendix E]{appE} ``Writing and Spelling'', Appendix~E to
\ment{The Lord of the Rings}.

\bibitem[Foreword]{foreword} Foreword entitled \Hobbit{\tiny The
Hobbit or there and back again}, present in most newer hardcover
editions of \ment{The Hobbit}.

\bibitem[Shibboleth]{shibboleth} ``The Shibboleth of F\"eanor'',
published in \ment{The Peoples of Middle-Earth}.

\bibitem[Road]{Road} \ment{The Road Goes Ever On}.

\bibitem[Quenyanna]{qwl} Helge K\r{a}re Fauskanger, \ment{Quettaparma
Quenyanna} (the \ment{Ardalambion} Quenya wordlist),\\
\href{http://www.uib.no/people/hnohf/wordlists.htm}
{http://www.uib.no/people/hnohf/wordlists.htm}

\bibitem[Parviphith]{swl} \ment{Parviphith Edhellen} (the \ment{Ardalambion} Sindarin wordlist),\\
\href{http://www.uib.no/people/hnohf/parviphith.doc}
{http://www.uib.no/people/hnohf/parviphith.doc}

\bibitem[History]{history} Gildor Inglorion, {\Quenya{quenta
eldatencelion:}} \ment{A brief history of Elvish writing},\\
\href{http://www.elvish.org/gwaith/pdf/history_of_writing.pdf}
{http://www.elvish.org/gwaith/pdf/history\_of\_writing.pdf}

\bibitem[Textbook]{textbook} Chris McKay, ``Tengwar Textbook''
(4th~ed.),\\
\href{http://web.archive.org/*/http://www.geocities.com/tengwar2001/}
{http://www.geocities.com/tengwar2001/}

\bibitem[Numbers]{linnumbers} Per Lindberg, \ment{Writing Numbers with
Tengwar},\\ \href{http://www.forodrim.org/daeron/md_teng_primers.html}
{http://www.forodrim.org/daeron/md\_teng\_primers.html} %, 2002-03-24

\bibitem[Sindarin]{linsindarin} Per Lindberg, {\Gondor{t\'iw i lam
thindrim:} \ment{Writing Sindarin with Tengwar}},\\
\href{http://www.forodrim.org/daeron/md_teng_primers.html}{http://www.
forodrim.org/daeron/md\_teng\_primers.html} %, 2002-03-28

\bibitem[Quenya]{linquenya} Per Lindberg, {\Quenya{quenya tengwainen:}
\ment{Writing Quenya with Tengwar}},\\
\href{http://www.forodrim.org/daeron/md_teng_primers.html}{http://www.
forodrim.org/daeron/md\_teng\_primers.html} %, 2003-04-01

\bibitem[Runes]{linrunes} Per Lindberg, {Tolkien's English Runes},\\
\href{http://www.forodrim.org/daeron/md_teng_primers.html}{http://www.
forodrim.org/daeron/md\_teng\_primers.html} %, 2007-09-16

\bibitem[FTMME]{FTMME} Ronald Kyrmse, {\ment{J.R.R.~Tolkien's Full
Tengwar Modes for Modern English: an Analysis}}, formerly available at\\
{http://www.
geocities.com/otsoandor/FTMME.htm}

\bibitem[Sarati]{sarati} M\r ans Bj\"orkman, {\ment{The
Sarati---Quenya Usage}},\\
\href{http://at.mansbjorkman.net/sarati_quenya.htm}{http://at.
mansbjorkman.net/sarati\_quenya.htm}

\bibitem[Valmaric]{valmar} M\r ans Bj\"orkman, {\ment{Valmaric Eldamar
version 1.0}},\\
\href{http://at.mansbjorkman.net/valmariceldamar.htm}{http://at.
mansbjorkman.net/valmariceldamar.htm}

\bibitem[Ispell]{ispell} The English ispell wordlist,\\
\href{http://lasr.cs.ucla.edu/geoff/ispell-dictionaries.html}
{http://lasr.cs.ucla.edu/geoff/ispell-dictionaries.html}

\bibitem[Knuth]{Knuth} Donald Knuth, \ment{Virtual Fonts: More Fun for
Grand Wizards}, TUGboat Volume~11 (1990), No.~1\\
\href{http://www.tug.org/TUGboat/Articles/tb11-1/}
{http://www.tug.org/TUGboat/Articles/tb11-1/}

\bibitem[fontinst]{fontinst} The \textsf{fontinst} package and
documentation,\\ \href{http://www.tug.org/applications/fontinst/}
{http://www.tug.org/applications/fontinst/}

\end{thebibliography}

\end{document}