\chapter{Environmentally aware}\label{env}
\section{Environments}
\LaTeX\ is environmentally aware! One of the key features
of \LaTeX\ is the way in which it uses `environments'\index{environment}:
that is, an `enclosed' structure (which may be part of
another structure), which has some kind of unity. We have
already met one environment, the \verb|document|. There
are a number of other \LaTeX\ environments, all intended
for slightly different purposes.

Environments have the common feature that they are
bounded by  a \verb|\begin{..}|\index{begin@\texttt{begin}}
\verb|\end{..}|\index{end@\texttt{end}} pair. You may only ever
terminate an environment by an environment of exactly the
same name. However, the \verb|\end{document}| will
generally close everything, perhaps with a few oblique
comments from \LaTeX.

\section{Quoting}
To illustrate an environment, look first at quoting. \LaTeX\
has two slightly different environments for quoting:
\verb|quote| and \verb|quotation|. A \texttt{quote}\index{quote@\texttt{quote}} 
is intended for short, perhaps one line, quotations. 
\begin{verbatim}
\begin{quote}
Joe Bob says check it out.
\end{quote}
\end{verbatim}
yields
\begin{quote}
Joe Bob says check it out.
\end{quote}
What do we note about this? The quote is indented to offset it
from the existing text, and there is a little white space
above and below. If we had a longer quotation, we would
be able to see whether the indentation we see was a `normal' paragraph
indentation or whether the whole quote was indented or whether the whole quote
is indented:
\begin{quote}
I am not now, and never have been, a member of the
American Communist Party -- \emph{attributed to Mao Xedung}.
\end{quote}
Obviously it applies to all lines.

The other alternative is \verb|quotation|\index{quotation@\texttt{quotation}},
which is really designed for longer quotations containing several
paragraphs: 
\begin{quotation}
I know your works, that you are neither hot or cold: I would that you were
either hot or cold.

But because you are lukewarm,  neither hot or cold, I will spit
you out.
\end{quotation}
which may be obtained from
\begin{verbatim}
\begin{quotation}
I know your works, that you are neither 
hot or cold: 
I would that you were either hot or cold.

But because you are lukewarm,  
neither hot or cold, I will spit you out.
\end{quotation}
\end{verbatim}

\begin{question}
Use the \texttt{quotation} environment to produce a piece of
quoted text. If your mind is a complete blank, you could
always try:
\begin{verbatim}
It's one hundred and six miles to Chicago;
we got a full tank of gas, half a pack of 
cigarettes, it's dark, and we're wearing 
sunglasses.

Hit it.
\end{verbatim}
\end{question}

\section{Verse or worse}
Lamport  introduces another environment
which he describes thus:
\begin{quote}
Poetry is displayed with the \texttt{verse} environment.
\end{quote}
Few poets would want to have their work described as
verse\index{vers@\texttt{verse}}, and even fewer would use the \texttt{verse}
environment to present it. However, the environment is
not without its uses.
And in fact, to be `fair' it has some rather appealing
qualities. Unlike every other situation we have seen so
far, \LaTeX\ asks that you do the line breaking\index{line breaking} yourself
in the \texttt{verse} environment. Do this by terminating a
line with \verb|\\|\index{backslash}. As we shall discover, this works in all
sorts of other places too.
%\begin{verbatim}
\begin{verse}
Bring me my Bow of burning gold:\\
Bring me my Arrows of desire:\\
Bring me my Spear: O clouds unfold!\\
Bring me my Chariot of fire!
\end{verse}
%\end{verbatim}
was produced from
\begin{verbatim}
\begin{verse}
Bring me my Bow of burning gold:\\
Bring me my Arrows of desire:\\
Bring me my Spear: O clouds unfold!\\
Bring me my Chariot of fire!
\end{verse}
\end{verbatim}

One feature may be worth emphasising. The last line of a stanza does not
have to end with \verb+\\+. Since it is immediately followed by
\verb+\end{verse}+ there is no need to introduce an additional (and
therefore redundant) instruction.
\begin{question}
Find out what happens when you end the last line with \verb+\\+. Does
\LaTeX\ handle it in an elegant and friendly way?
\end{question}



Lamport clearly had Blake in mind when he created the
{\tt verse} environment, since lines which are too long
for the page width wrap over nicely, with a little bit
of extra indentation:
\begin{verse}
Excess of sorrow laughs. Excess of joy weeps.\\
The roaring of lions, the howling of wolves, the raging
of the stormy sea, and the destructive sword, are
portions of eternity too great for the eye of man.\\
The fox condemns the trap, not himself
\end{verse}

There are two other features of note here. The first is that the \verb|\\|
instruction has a variant\index{variant} form, the \texttt{*}\index{asterisk@*}
form. If you write \verb|\\*|, \LaTeX\ will not allow the page to be
broken at that point. This useful feature is 
an example of a variant form for the instruction. We will find that many
\LaTeX\ instructions have a \verb|*| variant, which  modifies
the default way in which the instruction is handled.

The second feature is that the instruction may take an optional\index{options}
argument. We have met optional arguments before, and we shall meet them again and
again. Optional arguments are always enclosed in square brackets --
\mbox{\texttt{[}\thinspace\texttt{]}}. Here the optional argument allows us to
specify how much additional space to leave between lines -- a
\emph{dimension}\index{dimensions} is specified, complete with its units. In the
case of the
\verb+\\+ instruction we would have something like \verb+\\[0.5cm]+ where some
dimension, in this case 0.5\,cm, was specified in the square brackets. This is
similar to the optional arguments we have already met for
\verb|\documentclass|, except that
\verb|\documentclass| also  requires an argument in braces. Once we know
what units we can use, this  optional argument offers some useful
possibilities. 

\section{Units\index{units} -- a digression}\label{units}
We have come a long way, without any real reference to the sizes\index{size} of
things, except fonts. In general, we shouldn't be too concerned with sizes of most
things on the page, since the selection of the document style implicitly
specifies many of the relationships and sizes of parts of the document.
Nevertheless, there are times when we want to do something particular,
where we must somehow specify a size.

\begin{table*}
\begin{center}
\begin{tabular}{|cll|}
\hline
abbreviation & name & conversion\\
\hline
 pt & point & \\
 pc & pica & 1\thinspace pc=12\thinspace pt \\
 in & inch & 1\thinspace in=72.27\thinspace pt \\
 bp & big point & 72\thinspace bp=1\thinspace in \\
 cm & centimeter & 1\thinspace in=2.54\thinspace cm \\
 mm & millimeter & 10\thinspace mm=1\thinspace cm \\
 dd & did\^ot point & 1157\thinspace dd=1238\thinspace pt \\
 cc & cicero & 1\thinspace cc=12\thinspace dd \\
 sp & scaled point & 65536\thinspace sp=$2^{16}$sp=1\thinspace pt \\
 em & width of  `M' & (current font)\\ 
 ex & height of  `x' & (current font)\\
\hline
\end{tabular}
\end{center}
\caption{Units\index{units} used in \LaTeX\ to specify dimensions}\label{tunit}
\end{table*}

Table \ref{tunit} shows the units which \LaTeX\ `knows' about,
together with their conversions. To specify a dimension, say
\verb|12pt|, \verb|0.7pc| or \verb|12in|, or even \verb|12 pt|, \verb|0.7 pc|
or \verb|12 in| if you find that easier to write or read. If you inadvertently
say \verb|12inches|, \LaTeX\ will assume that the \verb|12in| is the
dimension, and that the \verb|ches| is  text to be printed. Of course
\verb|12points| wouldn't even be understood and would generate an error. 

(\La)\TeX\ uses the conversions as \emph{exact}
ratios -- in other words, even if they are not the exact and true
conversions, they are what \LaTeX\ uses (this ensures a degree of
standardisation within \LaTeX). Any value preceding one of these units
may be specified as either a whole (integer) number, or one with a decimal
point (decimal numbers would not make much sense in the case of `scaled
points'). \LaTeX\ does not support mixtures of units, so that  `two picas
and four points'  would not be acceptable, except as 2.3333333pc or 28pt.

You do not have to stick to any particular dimensions throughout your
document. Mix metric, Imperial and printers' dimensions as you wish. The
printer's point\index{points} makes some sort of sense as a base unit, since all
the fonts are described in terms of their `point' size. But outside that
particular area, any other  `size' can be described in any of the above
dimensions. If you do not include a dimension, \LaTeX\ will often assume
points are meant; it is safer to give the dimension. If the dimension is
zero, it still needs units.

The \verb|em|\index{em} and \verb|ex|\index{ex} deserve a little more
explanation. They are the only two `relative measures'. In
typographic terms, an em (also known as a \emph{mutton}) is a horizontal
measure (being approximately the width of a capital `M' in the \emph{current}
font), while an ex (also known as an \emph{en-nut}) is a vertical measure (the
height of an `x' in the \emph{current} font). If you change font, the values of
em and ex will also change. Although an em is intended as a horizontal
measure, it will be possible to use it in a vertical sense: similarly, the ex
may be used for horizontal amounts. \LaTeX\ is not that picky.

\subsection{The point}
In this sort of situation, in verse, it is more
general to specify a distance which is related in some way to the font in
use: a relative rather than an absolute figure:
\begin{verse}
Our plesance here is all vain glory,\\
This fause warld is but transitory;\\
The flesh is brukle, the Fiend is slee:\\*[2ex]
\emph{Timor mortis conturbat me.}
\end{verse}
The `refrain' is separated from the rest of the poem by an additional
2\,ex, obtained by terminating the  line preceding the refrain by
\verb|\\[2ex]|:
\begin{verbatim}
The flesh is brukle, the Fiend is slee:\\[2ex]
\end{verbatim} 
\begin{question}
Try that last example, but use different units. How would you leave a gap
which was equivalent to a blank line?
\end{question}
\section{Lists}
If \LaTeX\ is remembered for nothing else, it will be
remembered for the way in which Lamport elevated the
`list'\index{lists} into a common, powerful, all-pervading structure.
The more I think about it, the more I see myself
surrounded by documents which are most effectively
presented as lists, or lists within lists. As a technique
for structuring an argument, and giving relative weights
to various parts and sub-parts, it can be enlightening and fundamental.

\LaTeX\ has three basic list structures, each with
slightly different attributes and visual properties.
There are also facilities to extend these. The structures
are\index{index@\texttt{itemize}}\index{enumerate@\texttt{enumerate}}\index{description@\texttt{description}}
\begin{itemize}
\item enumerate
\item itemize
\item description
\end{itemize}
Of course, this is a list! It was generated by
\begin{verbatim}
\begin{itemize}
\item enumerate
\item itemize
\item description
\end{itemize}
\end{verbatim}
Now we have (almost) the whole syntax\footnote{You should
be able to guess that there are probably some optional
arguments lurking about somewhere. There usually are.} of
these list-making environments. 

As usual, the environment must be terminated with the
correct name. 

\begin{question}
Try doing something like this:
\begin{verbatim}
\begin{itemize}
\item just a word or two
\item and another one
\end{enumerate}
\end{verbatim}
The error message which \LaTeX\ gives in this
context is very annoying:
\begin{verbatim}
! \begin{itemize} on input line 3 ended 
  by \end{description}.
\end{verbatim}
It is informative that \LaTeX\ stops at this point, but
also helpful that it will allow you to continue. A pity it misleads.
\end{question}

If you are an \texttt{emacs} guru, or have another similarly
powerful editor\footnote{Correctly, \texttt{emacs} is a way
of life. Like (\La)\TeX, there is nothing \texttt{emacs} cannot do, 
in the right hands.}, and skill, you could ensure that your
\verb|\begin{..}| was always terminated
correctly. 

\subsubsection{Itemizing}
Using the \verb|itemize|\index{itemize@\texttt{itemize}} environment, each item
is marked in some way. Different levels of itemization are given
different symbols. This implies that you can
nest itemization instructions. In fact, you can nest the
whole concept of environments in general, and lists in
particular\footnote{Almost true; don't go too deep.}. 

\begin{question}
Find out what the symbols are for the various levels of
itemization.
\end{question}
\begin{question}
Find out just how deeply you can nest itemization.
\end{question}

\subsubsection{Enumerating}
As you may guess, the \texttt{enumerate} environment numbers each
item\index{enumerate@\texttt{enumerate}}. This implies that there must be a
counter marching away somewhere, incrementing each time \verb|\item| is
encountered. And so there is. There may be several, since we can, as
usual, nest enumeration too.
\begin{question}
Find out what happens when you nest levels of
enumeration. For example, you could  use one of your last exercises,
changing the \verb|\begin{itemize}| to \verb|\begin{enumerate}|.
Remember to change the \verb|\end| too.
\end{question}
\begin{question}
Now nest enumeration within itemization (or even
itemization within enumeration).
\end{question}

\subsubsection{Describing}
In the \verb|description|\index{description@\texttt{description}} environment,
you
\emph{must} provide an optional (\emph{sic}) argument on the \verb|\item|.
This argument will be the `key' to the list. In the
previous list environments, \LaTeX\ generated something
(a symbol, or a number\slash letter) to mark each \verb|item|.
This time, you have to provide the `mark':
\begin{verbatim}
\begin{description}
\item[aardvark] mainly harmless -- 
as in ``aardvark never hurt anyone''.
\item[Betty Boop] a pleasure and a 
delight. One of \emph{the} great artistes.
\item[crow] family of birds including the 
magpie (\textsl{pica pica})
\end{description}
\end{verbatim}
which comes out looking like
\begin{description}
\item[aardvark] mainly harmless -- as in ``aardvark never
hurt anyone''.
\item[Betty Boop] a pleasure and a delight. One of \emph{the}
great artistes.
\item[crow] family of birds including the magpie (\textsl{pica pica})
\end{description}
\begin{question}
Perhaps I exaggerate when I say that this environment requires an optional
argument. Demonstrate what happens if you omit the optional argument on the
\verb+\item+s in \verb+description+. Or whether the other list environments
may take optional arguments.
\end{question}
\begin{question}
It has always seemed to me that these three basic list environments are not
well named -- with two verbs and a noun. Devise a better and more consistent
set of names. \textsc{Html} uses the rather terse \texttt{ol} (ordered list),
\texttt{ul} (unordered list), and \texttt{dl} (data list).
\end{question}
%%%%%%%%%%%inserted section 3-e.tex
\section{Read my lips}
Another two pairs of environments may be useful. As these
notes show, it is often useful to be able to express chunks
of text in typrewriter form -- that is `verbatim', where
every symbol is recorded as it was typed in. This may not be 
a frequent requirement. \LaTeX\ has a \verb+verbatim+ environment 
which will handle this. Obviously it has to turn off lots of \LaTeX's normal
processing, and as a result there are some limitations. When the environment is
ended, the \verb+\end{verabtim}+ must occur on a line by itself. A variant is
the *-form, where any blanks (spaces) within the material will be printed out
as \verb*+ +. Besides needing to be able to present chunks of text literally,
it may also be necessary to do this for odd  symbols or words. The command
\verb+\verb+ is provided for this. For example, if we write
\begin{verbatim}
The backslash, \verb+\+, introduces every instruction, with a few
exceptions, including \verb=~=.
\end{verbatim}
then the result will be 
\begin{quote}
The backslash, \verb+\+, introduces every instruction, with a few
exceptions, including \verb=~=.
\end{quote}
Note how \verb+\verb+ works. The material which is to be expressed verbatim is
`enclosed' within a pair of repeated arbitrary symbols. The only symbols which
may not be used are \verb+*+, since there is a variant, \verb+\verb*+, where
any enclosed blanks are again represented by \verb*+ +. For example
\begin{verbatim}
The tag \verb*+< html>+ will not be 
recognized, since the space (\verb*+ +) 
after the \verb+<+ should not be present
\end{verbatim}
would give us
\begin{quote}
The tag \verb*+< html>+ will not be recognized, since
the space (\verb*+ +) after the \verb+<+ should not be present
\end{quote}
Although line endings are respected in the \verb+verbatim+ environment,
 you should not have a line end in the body of a \verb+\verb+. This is seldom a
problem, but if you use an editor which wraps (like \Textures), you may
inadvertantly break a \verb+\verb+ over a line. \LaTeX\ will note the error.
Another important factor is that neither of these instructions may appear in
the argument of another instruction.


The sister environment to \verb+verbatim+ is \verb+alltt+. This is identical,
except that \verb+\+, \verb+{+ and \verb+}+ all have their `normal' meanings.
We may therefore have
\begin{verbatim}
\begin{alltt}
Even in text we need to  \textsl{slant}
\end{alltt}
\end{verbatim}
to yield
\begin{quote}
\begin{alltt}
Even in text we need to  \textsl{slant}
\end{alltt}
\end{quote}
If we wish to use this feature we should first ensure that the \verb+alltt+
package is loaded. We'll cover this aspect a little later.

\section{One paragraph at a time}
The other major environments are connected with the setting of paragraphs. In
a later  chapter we will look at how \LaTeX\ actually handles the
setting of text. In the mean time we will note that sometimes it does not
manage to align the right margin and complains of `overfull boxes'. There are
many solutions to this. One is to provide \LaTeX\ with more flexibility in
setting the paragraph. This can be translated as `being more sloppy'. The two
environments \verb+sloppypar+ and \verb+fussypar+ are provided to handle this.
The default is \verb+fussypar+, and therefore it is unclear just why you might
want to switch it on, when a `sloppy' set of paragraphs can be delimited
easily. Note this is a paragraph matter. Since \LaTeX\ does its
typesetting on a paragraph by paragraph basis, there is no point trying to set
part of a paragraph `sloppy' and part `fussy'. The concept just does not exist.

An alternative to this \verb+sloppypar+ environment is to use the instruction
\verb+\sloppy+. There is a corresponding
\verb+\fussy+ instruction too. One of \LaTeX's quirks, a consequence
of treating text on a paragraph basis, is that the inclusion of
one of these instructions within a paragraph, whether at the end or the
beginning, turns the whole paragraph into that style; and of course every
following paragraph too, unless you issue new instructions, or are careful to
limit the action in some other way.



It is also possible to exercise control over the way \LaTeX\ sets lines within
paragraphs. Normally \LaTeX\ spreads out the space between words to have both
the right and left margins aligned vertically -- set `flush left' and 
`flush right'
(or \emph{justified}). If, instead of adjusting the space, 
it is made constant, the left margin remains fixed and a ragged
right margin would appear. The \verb+flushleft+ environment has this effect.
Its mirror-twin, \verb+flushright+ will make the left margin ragged. And a third
variant may be more useful, \verb+center+, where the text is effectively
centred around the middle axis. Taken with \verb+\\+ to break lines, this is
sometimes a useful way of creating title information. All three of these
environments have corresponding instructions: respectively they are
\verb+\raggedright+, \verb+\raggedleft+ and \verb+\centering+. Again, recall that
these instructions and environments will apply to whole paragraphs.

Observation will demonstrate that the \verb+verse+ environment is set 
ragged right, presumably to minimise
hyphenation, while \verb+quotation+ is justified.

\subsection{More to come}
This is was just a glimpse of the environments available in \LaTeX.
A few more will crop up. It makes little sense to present
 \emph{all} the environments at once, especially
when they each have quite different functions. The key
things to note are 
\begin{itemize}
\item that they must be enclosed by the same
identifier or label (e.g.~\texttt{quote} or \texttt{itemize}), 
\item that they \emph{may} have
optional arguments, 
\item and that they may not overlap.
\end{itemize}

\section{Sections\index{sections} and the like}
Many (most) documents can be thought of as highly
hierarchical, with a subdivision into successively
smaller units, down the scale through `parts',
`chapters', `section', `subsections', right down through
`paragraphs'\index{paragraphs} to sentences and even words. \LaTeX\ recognizes
this and provides a set of `sectioning' instructions, shown in
Table~\ref{tsect}
\begin{figure*}
\begin{center}
\begin{tabbing}
\makebox[1.5in]{}\=\verb|\part|\\
\>\quad\=\verb|\chapter|\\
\>\>\quad\=\verb|\section|\\
\>\>\>\quad\=\verb|\subsection|\\
\>\>\>\>\quad\=\verb|\subsubsection|\\
\>\>\>\>\>\quad\=\verb|\paragraph|\\
\>\>\>\>\>\>\quad\=\verb|\subparagraph|
\end{tabbing}
\end{center}
\caption{The sectional hierarchy}\label{tsect}
\end{figure*}

Now, note that we do not in fact go down right to the
sentence or word level. And also note that Lamport's
\verb|\paragraph|\index{paragraph@\texttt{paragraph}} and \verb|\subparagraph|
are not really paragraphs at all. Lamport merely says
\begin{quote}
``The names \dots\ are unfortunate, since they usually
denote units that are usually composed of several
paragraphs; they have been retained for historical
reasons.''
\end{quote}
Make of that what you will.

Each sectioning instruction requires an `argument' -- its own
title. Since this is mandatory, it appears in braces.
(Remember that only optional arguments appear in square
brackets.)

What do we get out of sectioning? It is easiest to see by
trying it out. As you will have realised, this document
is produced with \LaTeX. We have already used
\verb|\chapter|\index{chapter@\texttt{chapter}} at the beginning of this
`section', using the statement
\begin{verbatim}
\chapter{Environmentally aware}
\end{verbatim}
Somehow, \LaTeX\ managed to work out that this is the
chapter~\ref{env}, and arranged for the details to be
printed out. So one of the things  going on in the
background is to keep track of the level of the
sections. Similarly, this particular section started with a 
\begin{verbatim}
\section{Sections and the like}
\end{verbatim}
instruction. Again, \LaTeX\ kept track of the levels. Notice
too that it presented the information in different ways.
There are no sub-sections here, but we could invent a few:
\begin{verbatim}
\section{Sections}
In article and report, section is the 
  highest level.
\subsection{Subsections}
After the section comes the subsection.
\subsubsection{Subsubsection}
After  that comes the subsubsection.
\paragraph{Paragraph}
Then we have paragraphs  which run on 
  immediately after the paragraph label.
\subparagraph{Subparagraph}
And at the very lowest level, the 
  subparagraph, which, like the paragraph 
  runs on.
\end{verbatim}
 yields Figure~\ref{sectioning} where the slight reduction in size is
deliberately introduced to  highlight these sub-divisions.
Each of these sections must contain something, if only a lower level unit.
At the lowest unit, there must be some text.
 

\begin{figure*}
\begin{center}
\begin{minipage}{0.75\textwidth}
\small
\textbf{\large 3\quad Sections}\\[0.4\baselineskip]
In article and report, section is the highest level.\\[0.75\baselineskip]
\textbf{\normalsize 3.2\quad Subsections}\\[0.4\baselineskip]
After the section comes the subsection.\\[0.75\baselineskip]
\textbf{3.2.1\quad Subsubsection}\\[0.4\baselineskip]
After  that comes the subsubsection.\\[0.75\baselineskip]
\textbf{Paragraph}\quad
Then we have paragraphs which run on immediately after the paragraph
label.\\[0.75\baselineskip]
\textbf{Subparagraph}\quad
And at the very lowest level, the subparagraph, which, like the paragraph runs
on.
\end{minipage}
\end{center}
\caption{Sub-divisions from Section to Subparagraph\label{sectioning}}
\end{figure*}




\begin{question}
This book makes no real use of \verb|\paragraph| and \verb|\subparagraph|
since its author doesn't think in those terms. Find out how \LaTeX\
actually handles the text which follows these sectioning instructions. The
results may not be entirely anticipated.
\end{question}

 Naturally, in order to begin
(say) a subsection, you will already have begun a section. And equally
naturally, you may expect to have several subdivisions within a division.
On the other hand, as may be obvious here, the
\verb|book| style does not require the presence of
\verb|\part|. It can start with a \verb|\chapter|.
Similarly, the \verb|\article| style does not use
\verb|\chapter| at all. This does mean you can string all
your articles together into a book, all too easily, just
by a few easy edits.

\begin{question}
Demonstrate what happens when you  nest the sectioning (subsectioning,
subsubsectioning\dots) instructions inconsistently.
\end{question}

What else have we gained? The table of contents\index{table of
contents}\index{contents} at the beginning is also generated automatically as a
result of creating sections. Now, examine the table of contents
carefully to determine which levels are actually
reflected in the table of contents which is created. The section
`argument' (the bit in braces) goes into the table of contents ({\em
toc}) in its entirety. The entry may also be used  to change the
running head\index{running head}\index{head} (at the top of the
page)\footnote{not here, because the \BV\ house style over-rides this.}.

\LaTeX\ allows you to modify the behaviour of the
sectioning\index{sections} instructions in two main ways:
\begin{enumerate}
\item we may wish a different entry
to appear in the table of contents from the `body'; or
\item we may not wish the entry to appear in the table of contents, or
the running head (and by implication, not to be numbered).
\end{enumerate}

Both these requirements may be fairly readily
accommodated.
Changing the table of contents `\texttt{toc}' entry is  readily
accomplished by adding an optional argument:
\begin{verbatim}
\subsection[What I really 
want to appear]{Sub-sub-section}
\end{verbatim}
This optional argument will now appear in the table of contents
in place of the entry which introduces the subsection.
Limiting the action, and producing a `simple' entry
is possible through the \verb|*| form\index{variant}. This is another of the
instructions which have a variant form,  followed immediately by the~\verb|*|. 
In this case, it would appear as
\begin{verbatim}
\subsubsection*{An almost anonymous entry}
\end{verbatim}
This also suppresses the increments in the 
counter for this particular subsection.
%\end{itemize}


\section{Table of contents}
Having mentioned the presence of a table of contents (or
a \emph{potential} table of contents), we have to say a
little more about it. This is not generated unless you do
actually ask for it. How do you ask for it?
Say\index{tableofcontents@\texttt{tableofcontents}}
\begin{verbatim}
\tableofcontents
\end{verbatim}
and \LaTeX\ will insert the table of contents at that
point. Now, if the table of contents appears somewhere at the beginning
of the document, how does \LaTeX\ `anticipate' what the section
descriptions are to be?
Since it cannot know, what actually happens is
that you run your \LaTeX\ file several times. The first
time round \LaTeX\ generates a new file of the same name,
but with a \verb|.toc|\index{toc@\texttt{.toc}} extension. It is this file which
is read the second time around. Note that it might still
be out of step, if you have edited the file in between.
Essentially what has to happen is that the very final,
unchanged file has to be run \emph{at least} twice to resolve the table
of content entries (recall that it includes the page
numbers too).

\section{Other background activity}
Although we have not touched on figures and tables (yet),
there are similar instructions which allow list of figures and tables to be
generated. A similar sort of process goes on. A special
file is generated which contains information about the
list of figures, the list of tables, or even a glossary
or an index. The instructions for lists of figures and
tables are:
\begin{verbatim}
\listoffigures
\listoftables
\end{verbatim}
and the files they write are given the extensions
\verb|.lof|\index{lof@\texttt{.lof}} and 
\verb|.lot|\index{lot@\texttt{.lot}}. In
fact, you will find that
\LaTeX\ can be altogether profligate, creating all sorts
of other files, usually with the main file's name, but
different extensions. Every \LaTeX\ job creates a
\verb|.log|\index{log@\texttt{.log}} and an
\verb|.aux|\index{aux@\texttt{.aux}} file. The log file can give
a lot of information about what
is going on. Sometimes some of it is useful. Typically
a log is produced as you run the file through \LaTeX, but
more information will be inserted into the \verb+.log+ file, which may
be consulted later. The `auxiliary' file is more specific and 
contains some useful information for subsequent \LaTeX\ runs. It is not
intended for reading by people, just \LaTeX.

The truth is simply that we very
rarely get everything finished, or right, the first time
around -- it is so tempting to run a bit  through \LaTeX\
and see if it looks any good. In passing we could
demolish the myth that electronic document preparation
saves any time or money, or that office productivity tools have anything to
do with increasing productivity.  When you can make the corrections yourself,
you do, using your time, which is  costly. And you seek perfection, because you
know it is attainable. Before, you were just grateful to get something --
anything.

Most of these extra `information' files are there solely for
\LaTeX, and are of little direct use to us (except in terms of what they will
allow us to do). They are just about human readable, but why bother? Perhaps
the key thing to note though is that if you send off your \LaTeX\ file to
someone else, or receive one from someone else, you
really should expect to have it run through your document at least twice before
it looks as it should. This is a bit unexpected for the naive
or  inexperienced.



\begin{question}
Now have a look and see what extra files \LaTeX\ appears
to have generated for you. Look too at an old \verb|.log|
file to see the information which \LaTeX\ will have
issued about these extra files. As usual, it is often
difficult to distinguish the useful material from the
rather less useful material (`garbage'). We really need a
\verb|.junk| file too.
\end{question}


%\mbox{\verb|[|\thinspace\verb|]|}. Here the optional argument allows us to

\begin{question}
One of the things that puzzle me about
\LaTeX\ is why sections are not environments, or even
lists. \LaTeX\ will permit you to write something like
\begin{verbatim}
\begin{itemize}
\section{...
\begin{quote}
\section{.....
\end{quote}
.
.
\end{itemize}
\end{verbatim}
While I can see that the section may be part of an
environment, I am surprised that it can span environments
in this way.

Any explanations?
\end{question}