Handbook of Typography for the

Mathematical Sciences
Steven G. Krantz
January 21, 2003
©2001 CRC Press LLC ©2001 CRC Press LLC ©2001 CRC Press LLC
©2001 CRC Press LLC
To Johann Gutenberg (n´ee Gensfleisch) and Donald E. Knuth.
©2001 CRC Press LLC ©2001 CRC Press LLC
©2001 CRC Press LLC
Table of Contents
1 Basic Principles
1.1 An Overview
1.2 Choice of Notation
1.3 Displaying Mathematics
1.4 Consistency
1.5 Overall Design
2 Typesetting Mathematics
2.0 Introductory Remarks
2.1 What is T
E
X?
2.2 Methods of Typesetting Mathematics
2.3 A Lightning Tour of T
E
X
2.4 The Guts of T
E
X
2.5 Modes of Typesetting Mathematics
2.6 Line Breaks in Displayed Mathematics
2.7 Types of Space
2.8 Technical Issues
2.9 Including Graphics in a T
E
X Document
2.9.1 Handling Graphics in the Computer Environment
2.9.2 The Inclusion of a PostScript

Graphic
2.9.3 Graphics and the L
A
T
E
X2
ε
Environment
2.9.4 The Use of PCT
E
X

2.9.5 Freeware that Will Handle Graphics
3 T
E
X and the Typesetting of Text
3.1 Other Word Processors and Typesetting Systems
3.2 Modes of Typesetting Text
3.3 Hyphens and Dashes
3.4 Alignment
3.5 Typesetting Material in Two Columns
3.6 Some Technical Textual Issues
4 Front Matter and Back Matter
4.1 The Beginning
4.2 The End
4.3 Concluding Remarks
5 Copy Editing
5.1 Traditional Methods of Copy Editing
5.2 Communicating with Your Copy Editor
5.3 Communicating with Your Typesetter
©2001 CRC Press LLC
5.4 Communicating with Your Editor
5.5 Modern Methods of Copy Editing
5.6 More on Interacting with Your Copy Editor
5.7 Manuscript Proofs, Galley Proofs, and Page Proofs
5.8 The End of the Process
6 The Production Process
6.1 Production of a Paper
6.2 Production of a Book
6.3 What Happens at the Printer’s
7 Publishing on the Web
7.1 Introductory Remarks
7.2 How to Get on the Web
7.3 Web Resources
7.4 Mathematics and the Web
7.5 Software to Go with your Book or Article; Web Sites
Appendix I: Copy Editor’s/Proofreader’s Marks
Appendix II: Use of Copy Editor’s Marks
Appendix III: Specialized Mathematics Symbols
Appendix IV: Standard Alphabets
Appendix V: Alternative Mathematical Notations
Appendix VI: T
E
X, PostScript,

Acrobat,

and Related Internet Sites
Appendix VII: Basic T
E
X Commands
Appendix VIII: A Sample of L
A
T
E
X
Glossary
References
Resources by Type
©2001 CRC Press LLC
AUTHOR
Steven G. Krantz, Ph.D., received his doctorate from Princeton Uni-
versity in 1974 and his Bachelor of Arts degree from the University of
California at Santa Cruz in 1971. Dr. Krantz was an Assistant Profes-
sor at UCLA, a visiting Associate Professor at Princeton University, and
Associate Professor and Professor at Penn State.
Dr. Krantz is currently Professor and Chairman of the Department of
Mathematics at Washington University in St. Louis; he is also a visiting
professor and lecturer at many universities around the world.
Dr. Krantz has written more than 105 research papers and many ar-
ticles and reviews. He has written or edited thirty books. He is the
founder and managing editor of the Journal of Geometric Analysis. He
is also the editor of the CRC Press series Studies in Advanced Mathe-
matics. Dr. Krantz is the holder of the UCLA Foundation Distinguished
Teaching Award, the Chauvenet Prize, the Beckenbach Prize, and the
Outstanding Academic Book Award.
©2001 CRC Press LLC ©2001 CRC Press LLC
©2001 CRC Press LLC
PREFACE
Ellen Swanson’s book Mathematics into Type is a unique and im-
portant contribution to the literature of technical typesetting. It set a
standard for how mathematics should be translated from a handwritten
manuscript to a printed book or document. While Swanson’s book was
intended primarily as a resource for technical typesetters, it was also im-
portant to mathematical and other technical authors who wanted to take
an active role in ensuring that their work reached print in an attractive
and accurate form.
The landscape has now changed considerably. With the advent and
wide availability of T
E
X,
1
most mathematicians can take a more active
role in producing typeset versions of their work. Indeed, many mathe-
maticians currently use T
E
X to write preliminary versions of their work
that are very similar (in many respects) to what will ultimately appear
in print.
While the output from T
E
X has a more typeset appearance than that
from most word processors, the T
E
X product is not automatically (with-
out human intervention) “ready to go to press.” There are still “post-
processing” typesetting issues that must be addressed before a work
actually appears in print. The style and format of running heads, sec-
tion headings and other titles, the formatting of theorems and other
enunciations, the text at the bottom of the page, page break issues, and
the fonts and spacing used in all of these go under the name of “page de-
sign”. These are often customized for a particular book or journal. The
index and table of contents must be designed and typeset. Graphics,
and sometimes new fonts, must be integrated. Additional questions of
style in the formatting of equations and superscripts and subscripts can
also arise. Most T
E
X users do not know how to handle the questions just
listed, which is why most publishers currently send T
E
X documents for
books or journal articles to a third-party T
E
X consultant. The purpose
of the present work is to serve as a touchstone for those who want to
learn to make typesetting decisions themselves.
1
T
E
X is a markup language for doing mathematical typesetting. We will talk about
T
E
X in more detail as the book develops.
©2001 CRC Press LLC
Let us set out once and for all what this book is not. It is not a text
for learning T
E
X. Indeed, A T
E
X Primer for Scientists by Sawyer and
Krantz provides a venue for the rapid assimilation and mastery of T
E
X
basics. There is no need to repeat the lessons of A T
E
X Primer here.
Instead, the present volume is (in part) a book on how to use T
E
X.
But the typesetting principles enunciated here will apply equally well
for those whose work is being typeset by a different method. The user
of Microsoft Word,

for example, will not himself implement (as would
a T
E
X user) the kerning and formatting and page design commands
which we discuss in Chapters 2 and 3, but he will communicate with
the typesetter about those commands. He will not be able to format
equations with the level of precision and detail that we describe, but
he will (after reading this book) be equipped with the vocabulary and
skills to tell the typesetter what he wants. He will not be importing an
encapsulated PostScript

figure into his document in just the manner
that we lay out, but he will learn the process and thus be able to ensure
that his book or tract comes out in the form desired.
There is no point to mince words. We believe strongly, and we are
certainly not alone in this belief, that the medium of choice for producing
a mathematical document today is T
E
X. Most mathematicians use T
E
X,
most publishers use T
E
X, and most Web sites are set up either to handle
T
E
X documents or to handle files produced from T
E
X code. T
E
X code is
more portable than the files from any word processor, and its output is
of vastly higher quality than the output from any other system available.
With suitable plug-ins, T
E
X can handle graphics beautifully. There is
no formatting problem that T
E
X cannot handle.
2
If you send your work
to a journal or a publisher in any electronic format other than T
E
X (or
one of the variants of T
E
X), then you are only inviting trouble and, in
some cases, derision. We hope that this book will serve to convince you
of the correctness of all these assertions.
Prerequisites for reading this book are a knowledge of the elements of
mathematical writing (for which see, among other sources, A Primer of
Mathematical Writing by this author) and an interest in mathematical
typesetting and graphics issues. We certainly do not assume that the
reader is an active user of T
E
X. We include a brief description of T
E
X
and its most basic commands, just because T
E
X is so much a part of
mathematical life today; and also because it is easier to describe some
2
Consider this typesetting problem: You have an expression that consists of a 3 × 3
matrix divided by an integral, and you want to typeset it in displayed fashion as
a fraction. T
E
X can perform this task beautifully and easily, with simple and sure
commands. Your word processor cannot, nor can any other document preparation
system that is available as of this writing. A similar remark applies to commutative
diagrams, to tables, and to many other high-level typesetting tasks.
©2001 CRC Press LLC
of the typesetting procedures that are essential to mathematics if one
can make reference to T
E
X. In Appendices III and VII we include a
compendium of all the T
E
X commands that are most commonly used in
mathematical writing. Appendix VIII contains a sample of T
E
X code
together with the compiled output. The reader who spends some time
with the present book will certainly come away with considerable moti-
vation for learning more about T
E
X.
Likewise, we do not assume that the reader is conversant with the tools
for providing his book or manuscript with graphics—such as PostScript
or bitmap (*.bmp) files or *.jpg files or *.gif or *.pdf files or P
I
CT
E
X.
Instead, we hope to acquaint the reader with these and some of the other
graphics options that are available to the mathematical author.
The tools that are now available for creating the index, the table of
contents, the list of figures, the bibliography, tables, diagrams, and other
writing elements are both powerful and marvelous. We wish to compile
here a resource for the author who wants to take control of these portions
of the creation of a book or document.
Finally, many a book author today will want his book to contain a
computer diskette, or a reference to software that is available on the
World Wide Web, or source code for software. We will discuss issues
attendant to this part of mathematical writing, and offer some solutions
as well.
The reader who becomes acquainted with the present work will be a
well-informed author who is equipped to deal with publishers, compos-
itors, editors and typesetters, with T
E
X consultants, with copy editors,
and with graphics designers of every sort. He may not be tempted to
perform the various typesetting and formatting and graphical tasks him-
self, but he will be prepared to communicate with those who do. It is
our hope that the result will be an author who has a better understand-
ing of the publishing process, and one who will want to and be able to
create better mathematics books.
Steven G. Krantz
St. Louis, Missouri
©2001 CRC Press LLC
ACKNOWLEDGEMENTS
Over the years, my friend and collaborator Stanley Sawyer has taught
me a great deal about T
E
X and about typesetting. He was also good
enough to read many drafts of the manuscript for this book, and to
offer innumerable suggestions. For all of his help and suggestions I am
grateful.
George Kamberov has been a great resource in helping me learn how
to include graphics in a T
E
X document. His friendship and patience are
much appreciated.
CRC Press has engaged several reviewers to help me hone this book
into the precise and accurate tool that it should be. To all of them I
express my indebtedness.
My editor, Bob Stern, has been encouraging and helpful in every as-
pect of the production of this book. He helped me interpret the review-
ers’ remarks and keep this book on track. As always, thanks.
©2001 CRC Press LLC
chapter 1
Basic Principles
1.1 An Overview
The ability to write well is not a gift from the heavens. It is a craft that
is honed and developed over time. This assertion is as true for technical
writing as it is for prose and poetry. A large part of the craft is the
ability to harness one’s thoughts and to organize them into sentences,
paragraphs, and chapters. Today’s technical writer is also involved in
the physical process of putting the words on the page. Word processing
systems like Microsoft Word and computer typesetting systems like T
E
X
put the writer in charge of the composition of the page, the choice of
fonts, the layout of section and chapter titles, the design of running
heads, and of many other aspects of the book or document as it will
finally appear.
Thus the creation of a document in the modern writing environment
involves not only the traditional process of organizing one’s thoughts,
but also planning the form of the document. There are some easy choices
that can be made. L
A
T
E
X (a dialect of T
E
X invented by Leslie Lamport)
1
allows you to choose one of several different pre-formatted document
styles (in L
A
T
E
X 2.09) or document classes (in L
A
T
E
X2
ε
). As an example,
you can choose the book document style or document class by entering
\documentstyle{book} or \documentclass{book} as the first line of
your T
E
X source code file. See the sample code in Appendix VIII. Once
you have made this choice, many typesetting and page design decisions
are automatically made for you. Your sections, subsections, theorems,
propositions, definitions, examples, equations, and so forth are all num-
bered automatically. Every reference to an equation or theorem or other
enunciation will be linked to that theorem or equation, so that all of your
1
See Chapter 2 for more information about T
E
X and L
A
T
E
X. Among dialects of T
E
X,
L
A
T
E
X is often preferred by publishers because it gives authors fewer choices, and
because it is a structured language.
©2001 CRC Press LLC
references will always be accurate. You will be able to refer to equations
and theorems and other items by nickname, and L
A
T
E
X will number them
automatically; should you move one of these items to a different part of
the document, L
A
T
E
X will still get the numbering right.
L
A
T
E
X2
ε
further allows you to customize your stylistic choice with the
\usepackage command. Some of the packages that may be invoked are
babel.sty, multicol.sty, \epic.sty, and varioref.sty. As their
titles suggest,
• babel.sty gives the choice of several languages in which the doc-
ument may be typeset.
• multicol.sty enables several different multi-column formats. It is
much more powerful than the multi-column format of L
A
T
E
X 2.09.
• epic.sty provides a high-level user interface to the L
A
T
E
X picture
environment.
• varioref.sty provides powerful cross-referencing capabilities.
There are many other packages provided with L
A
T
E
X2
ε
, and more that
you can download from the Web (see Appendix VI).
Please note that non-customized (out-of-the-box) L
A
T
E
X styles look
grey and dull to many experienced editors, technical writers, and type-
setters. Few publishers will publish a manuscript in that form, if for no
other reason than that each publisher wants his or her books to have a
distinctive look.
In addition, few publishers will simply accept what the author sends
in. A copy editor will go through the manuscript and analyze line breaks,
page breaks, kerning, spacing, displays, running heads, graphics, choice
of fonts, and many other aspects of the manuscript that you might have
thought was a finished product. He will think about your syntax, your
use of language, your use of specialized terms and jargon, and the pre-
cision of your English. If you want to interact on an informed and
intelligent level with that copy editor, and if you want your book to
come out the way that you want it, then it behooves you to learn about
these and many other features of the typesetting process.
In many cases, the publisher will send you a T
E
X or L
A
T
E
X style file
that will determine the style of the finished product. Suppose, for def-
initeness, that you are using L
A
T
E
X2
ε
and that the style file is called
publstyle.sty. This file will be “called” at the top
2
of your document
2
The “top” of your L
A
T
E
X source code file—the part that occurs af-
ter the \documentstyle or \documentclass command and before the line
\begin{document}—is called the “preamble”. Various special L
A
T
E
X commands must
be placed in the preamble. Others must occur after the preamble.
©2001 CRC Press LLC
source or root file (or the first line of each source file if you have different
chapters in different self-contained files). In L
A
T
E
X2
ε
, the first two lines
will be
\documentclass{book}
\usepackage{publstyle}
Roughly speaking, the book.cls file invoked in the first line makes broad
stylistic choices that will apply to most any book. The publstyle.sty
file invoked in the second line will introduce refined stylistic features. In
L
A
T
E
X 2.09, the \input command may be used in a style analogous to
\usepackage in L
A
T
E
X2
ε
, but it is considerably less powerful.
Typically, the publisher’s style files will specify the size of the print
block on the page, fix the margins, fix the formatting of chapter and
section and subsection titles, and select fonts for all these features. It will
format enunciations (theorems, definitions, etc.), displayed equations,
and other features in your writing. It will sometimes add graphical
elements like horizontal lines to set off titles. Where suitable, it will also
format the bibliography, index, and table of contents. A well-designed
page is a delight to behold (and by no means an easy thing to create).
Unless you want to spend a great many hours mastering the principles
of page design, I urge you to let your publisher design the pages of your
book or paper. Even if you think you know how to do it, it is quite
likely that your publisher will modify or veto whatever page style you
may create.
1.2 Choice of Notation
An important part of planning a document, and one that is special to
the writing of mathematics, is choosing your notation. Even the unsea-
soned mathematical writer will know that notation should be consistent
throughout a given work. One should not use the symbol 1 to stand
for a group in Section 2 and for a topological space in Section 7 (unless,
of course, the group and the topological space are one and the same
entity). But, from the point of view of typography and clarity, there are
other issues at play.
Strive for simplicity and elegance in your notation. This dictum some-
times runs counter to the way that a mathematician thinks. For exam-
ple, it is very natural for a mathematician to use
1
j
1
,j
2
,...,j
k
to denote an entity (namely, 1) that depends on indices ,
1
, ,
2
, . . . , ,
k
.
But it is simpler to typeset, and certainly easier to read, the expression
1(,
1
, ,
2
, . . . , ,
k
).
©2001 CRC Press LLC
The meaning is unchanged, but the conveyor of the information is sim-
pler. A more extreme, and perhaps more convincing, example is to use
the notation
1(i
1
, i
2
, . . . , i
m
; ,
1
, ,
2
, . . . , ,
k
)
in preference to
1
i
1
,i
2
,...,i
m
j
1
,j
2
,...,j
k
.
It is best to avoid stacked accents. They are difficult to typeset, dif-
ficult to center properly, and need not be used. They can cause trouble
with between-line spacing. For example, in Fourier analysis the em-
bellishment
´
is used to denote the Fourier transform. And, if ) is a
function on 1, then the symbol
¯
) is defined by
˜
)(r) = )(−r).
If one wishes to compose these two operations (as in the standard proof
of Plancherel’s theorem—see [KRA2]), then one might be tempted to
typeset (this is the default in T
E
X):
˜
´
) .
You can see that the tilde is not properly centered over the
´
). And,
if you wish, you can blame the software T
E
X for this clumsiness. But
there it is, and adjusting the T
E
X code to obtain the desired result
˜
´
)
is a tedious business. It is simpler, and more attractive, to introduce
the notation 1)(r) = )(−r) and then to write 1(
´
)) for the composi-
tion. One can go further and introduce T) to stand for
´
). Then the
composition becomes 1T). Some of the stacked accent problems have
been fixed in L
A
T
E
X2
ε
.
In a related vein, it is not always convenient, nor aesthetically pleasing,
to use
´
) to denote the Fourier transform—especially if the argument
is a large expression. Rather than typeset
¯
¹ + 1 + C + 1
(note that the
´
is far too short) one can typeset
_
¹ + 1 + C + 1
_
´
(L
A
T
E
X2
ε
has special commands for handling this formatting situation).
Perhaps even better, you could typeset
T
_
¹ + 1 + C + 1
_
.
©2001 CRC Press LLC
The experienced T
E
X user will know that the Fourier transform’s hat
and also the tilde come in two forms: there is
ˆ
) and
´
), as well as
˜
) and
¯
). These are given by the T
E
X command pairs \hat{f}, \widehat{f}
and \tilde{f}, \widetilde{f}. It should be noted that, unlike the
overbar (given by \overline), the Fourier hat and the tilde are not
infinitely extendable. There are serious typesetting problems connected
with both the shape of and the positioning of hats and tildes. A similar
comment applies to the so-called “inverse Fourier transform”, or “bird”
(the symbol
ˇ
, given by \check). The user who needs a wide hat, tilde,
or bird will either need to learn METAFONT (in order to manufacture
his own) or may be able to find a third-party font which includes these.
Sometimes we do things on the computer just because we can. It is
fun to enter the code
e^{{a \over b} + {c \over d}}
and then to see the compiled outcome
c
a
b
+
c
d
.
But it is much clearer, and less prone to error, to typeset
c
(a/b)+(c/d)
with the fractions typeset in shilling form. Better still is
exp
_
(o,/) + (c,d)
¸
(note the use of parentheses () to help parse the fractions) or
exp
_
o
/
+
c
d
_
.
It is critical when you set a fraction in shilling form to use sufficient
parentheses so that the meaning is unambiguous. An expression like
¹,1+C has two distinct possible meanings. You want either ¹,(1+C)
or (¹,1) + C.
All notation should strive for simplicity, elegance, and clarity. If 1 is
a function of the matrix argument
_
o /
c d
_
then one might be tempted to typeset
1
_
o /
c d
_
.
©2001 CRC Press LLC
But it is much clearer to write
Let
¹ =
_
o /
c d
_
and consider
1(¹) .
A final example to round out the message being promulgated here: It
may be natural to typeset an expression like
_
o + /
2
c + sind
2o/
o
2
+ /
2
.
In some displayed equations, this format will probably serve; and it is
great fun to see the huge square root sign make its appearance. But
in many contexts it is clearer, and less intrusive on the neighboring
material, to render the formula as
_
o + /
2
c + sin d
_
1/2

2o/
o
2
+ /
2
or even
_
o + /
2
c + sin d
_
1/2

_
2o/
o
2
+ /
2
_
.
See Appendix V for a treatment of various typesetting choices.
This may be the moment to note that you should never begin a sen-
tence or a phrase (following a comma, for instance) with notation. For
example, never write
)(r) is a positive function on the interval [o, /].
or
G is a group of finite order.
Such usage will cause the reader to do a double-take; it will only serve
to obfuscate your prose. Better is
Let ) be a positive function on the interval [o, /].
or
The function ) is positive and has domain the interval [o, /].
For the second example, it is preferable to say
©2001 CRC Press LLC
We use the notation G to denote a group of finite order.
or perhaps
The group G has finite order.
Observe that, in the first example, we have corrected another common
error along the way. The character ) is the name of the function and
)(r) is the value of that function. In the present context, it is correct
to use “) ” rather than “)(r)”.
1.3 Displaying Mathematics
An important decision, which the mathematical author must make fre-
quently and effectively, is whether to display a mathematical expression
or leave it inline (i.e., in text).
3
If the expression is more than one fourth
of a line long, then you should usually display it. If it is not complex and
if it occurs near the beginning of a paragraph (so that you do not have
to worry about line breaks), then you may leave it inline. If it involves
more than a few subscripts or superscripts, or if these embellishments
are compounded, then you should display it so that it does not interfere
with the lines above and below. If it involves integrals which are not
completely elementary, or any kind of matrices, then you should display
it. It is always better to err on the side of clarity, and therefore better
to have too many displays rather than too few. But try to avoid having
so many displays that the mathematical exposition becomes harder to
follow.
Mathematical writing expert N. J. Higham [HIG] tells us that we
should display a mathematical expression when
• it needs to be numbered;
• it would be hard to read if placed inline;
• it merits special attention (perhaps because it contains the first
occurrence of an important mathematical entity).
These are excellent guidelines for deciding when to display an item. In
the end these decisions must be made by the author—based on meaning,
3
Here a piece of mathematics is said to be “displayed” if it occurs by itself, separated
vertically from the text before and after, and is centered left-to-right. Often, but
not always, displayed text will be larger—or at least appear larger—than text in the
main body of text. If mathematics is not displayed, then it is typeset “inline”, which
means that it is part of the regular flow of text.
©2001 CRC Press LLC
context, and appearance. Refer also to Section 2.5, where more subtle
issues of displaying mathematics are discussed.
4
1.4 Consistency
In fact the planning to which we alluded in the first and second sections
can lead to some attractive, and very desirable, consistency. As we will
explain in Chapters 2 and 3, the natural way to execute the planning of
the form of a book or document (at least when you are using T
E
X) is to
write some macros. The creation of a macro is simply the assigning of a
name to a collection of commands. The aggregate of commands is then
executed by simply invoking the macro name. Thus one creates macros
for section heads, subsection heads, chapter headings, running heads,
formatting of theorems, formatting of displayed equations, etc. It also
makes sense to create macros for complicated mathematical expressions
that will be used repeatedly. (See [SAK] for the full picture of macro
creation. Here we merely give a couple of simple examples.) As an
instance, the line of T
E
X code
5
\def\smsqr#1#2{\sqrt{{#1}^2 + {#2}^2}
+ \frac{1}{{#1}^2 + {#2}^2}}
creates a new T
E
X command called \smsqr. This command accepts two
arguments (designated by #1 and #2). The way of T
E
X is to replace
every occurrence of \smsqr with the code
\sqrt{{#1}^2 + {#2}^2} + \frac{1}{{#1}^2 + {#2}^2}.
With this macro definition at the top of your T
E
X file, entering the line
\smsqr{x}{y} will result in the typeset expression
_
r
2
+ j
2
+
1
r
2
+ j
2
.
A second example is a macro for theorems:
\def\theorem#1{{\sc Theorem} \ #1: }
This is a macro with a single argument (designated by #1), which will
usually be the number of the theorem. If you place this line of code at
the top of your T
E
X document, then you can invoke the macro with the
code
4
It might be mentioned that the command \displaystyle allows you to format
mathematics in display size and format while you are in inline mode; the similar
command \textstyle allows inline math formatting in display mode.
5
Refer to Appendix VII for the meanings of the different T
E
X commands.
©2001 CRC Press LLC
\theorem{3.5} Let $\epsilon > 0$. Let $[a,b]
\subseteq \RR$ be an interval. There is a function
$f$ such that \dots
The result will be the typeset text
Theorem 3.5: Let c 0. Let [o, /] ⊆ 1 be an interval.
There is a function ) such that . . .
The nice thing about this macro is that now all your “theorem” declara-
tions will be set in the same font and will have the colon included; you
need not remember, nor perform, a check for consistency.
The practice of using macros not only saves time and eliminates ag-
gravation, but it also serves to promote accuracy. By using macros, you
will guarantee that size, spacing, indentation, and other features of your
composition are maintained uniformly throughout the document.
There are other aspects of consistency about which many authors
are blissfully unaware: spacing above and below a displayed equation,
spacing above and below a theorem,
6
space after a proof, the mark at
the end of a proof (QED, or the Halmos “tombstone” , for example).
Again, a good macro can be invaluable in addressing these issues; but
awareness of the problem is also a great asset.
1.5 Overall Design
Proper page design requires an artistic sense and experience that most
mathematicians and other technical writers may not have. For this
reason, page design is best left to professionals, or at least should only
be done with input from professionals. If the publisher provides a set
of style macros that you can include at the top of your T
E
X file (see
Section 1.1), then many of the design decisions will already have been
made.
Even if you are using style macros provided by the publisher, there
are issues of which you should be aware. T
E
X will not align the bottom
lines of facing pages in a book unless the style macros do this explicitly.
7
T
E
X can be told to prohibit orphans and widows,
8
but this is sometimes
6
Here, and throughout this book, we use the word “theorem” to denote any displayed
body of text, including proposition or lemma or sublemma or definition or remark or
any number of other similar constructions. It is common among typesetters to call
these “enunciations”, and we will use that terminology as well.
7
The two pages of production notes in the classic [GMS] discuss the rather clever
and technical means by which these three very skilled authors had to hand-correct
decisions made by L
A
T
E
X. Almost all of these corrections concerned page breaks.
8
Here an orphan is a brief phrase—less than a full line—that occurs at the bottom
of a page and that begins a new paragraph. A widow is a brief phrase—less than a
full line—that occurs at the top of a page and that ends the preceding paragraph.
©2001 CRC Press LLC
not done because it can cause problems elsewhere. A chapter should
not end on a page with fewer than four lines from that chapter. A
section heading should be followed by at least two lines of text of the
new section on the same page. At some stage it will be necessary for
someone to go through the manuscript by hand and correct problems of
this kind. (With many publishers, this “someone” will be the author.)
Likewise, T
E
X can be clumsy with line breaks. If T
E
X gets jammed up
trying to fit just the right number of words into a line, or if it encounters
an unfamiliar word that it cannot figure out how to hyphenate, then it
may break a line in an inauspicious place—in fact, T
E
X may leave too
much space between words or not enough. Worse, it may run part of
a word or formula out into the margin.
9
T
E
X treats certain formulae
in math mode as unbreakable units; it is easy to see that such a rigid
chunk of typeset text can also be run out into a margin—resulting in
an unsightly paragraph and page. It is best if the author is the one to
deal with a problem such as this, because it is sometimes necessary to
rephrase or rewrite a sentence or paragraph to make things come out
as they should. You certainly do not want someone else rewriting your
material just to make the typesetting come out right!
10
As we have suggested, things become a bit easier once you have a
publisher. The publisher will then provide a T
E
X style file or (the old-
fashioned approach) a hard-copy style manual. If you do not yet have a
publisher, then you should avoid the temptation to be seduced by T
E
X
(or whatever document preparation system you are using) into writing
an overly complex source file. You should either use the simplest choices
in a standard T
E
X dialect (like Plain T
E
X, L
A
T
E
X, /
/
o-T
E
X, or /
/
o-
L
A
T
E
X), or else make sure that your style macros are well-designed and
very well-documented. Otherwise, your eventual publishers may accuse
you of “freestyling”, which suggests a T
E
X file with rigid or non-uniform
stylistic choices and T
E
X source that is incomprehensible except per-
haps to you. If you are guilty of freestyling, then it can be extremely
difficult for either the publisher or yourself to translate your work into
publishable form.
At the risk of belaboring this last point, let us mention a few examples.
Some authors go to great pains to have each chapter, or each section,
begin with a dropped gothic capital letter. Or they format pages in two
or three columns. Or they equip the left-hand pages with a running
9
The habit of T
E
X is to place a black box in the margin next to a line that has not
been typeset according to T
E
X’s standards. You can adjust or weaken the standards,
and thereby eliminate some of the black boxes. Usually it is better to eliminate the
boxes by learning to be a better T
E
Xnical typesetter.
10
George Bernard Shaw tells of rewriting sentences or passages in his plays in order
to satisfy the typesetter, and then the typesetter getting things wrong anyway.
©2001 CRC Press LLC
head and the right-hand pages with a running foot. Or they use very
exotic fonts. Or they may place the page numbers (the folios) in strange
places, set in peculiar fonts. All very charming, but no publisher will
want to preserve and to propagate these whimsies. You make everyone’s
life easier if you eschew the eccentric and stick to the most basic con-
structions. This advice is valid for the Plain T
E
X user, for the L
A
T
E
X
user, for the Microsoft Word user, and for every other user of electronic
tools.
©2001 CRC Press LLC
chapter 2
Typesetting Mathematics
2.0 Introductory Remarks
When mathematical formatting works well, we are hardly aware of it.
What we see is what we expect to see: there is a smooth transition
from what is printed on the page to what the eye sees to the cognitive
processes of the brain.
When one is creating typeset mathematics, one must be aware of the
components of good typesetting, so as best to effect the attractive pro-
cess just described. In particular, one must be aware of the issues of
spacing and positioning that control the clarity and meaning of mathe-
matical expressions.
In this chapter, we discuss such issues. While we will often invoke T
E
X
in this chapter, such invocations are generally self-contained. As you
read along, you will want to refer to Appendices III, VII, and VIII, which
are designed to speed along your acquaintance with T
E
X. Certainly this
book will tell you enough about T
E
X so that you will find everything that
we discuss comprehensible, and so that it will travel well to whatever
document preparation system you are using.
2.1 What is T
E
X?
T
E
X is a computer package for producing document output of typeset
quality. T
E
X is a computer typesetting language—a high-level computer
programming language. It is used by individuals to write letters and
articles with a personal printer as well as by technical publishers to
typeset mathematics books on a Varityper 6000 or a Linotronic 630
or other electronic typesetting device. T
E
X is particularly powerful and
useful for preparing manuscripts with mathematical formulas and tables.
Since being invented and introduced by Donald Knuth in the early 1980s,
©2001 CRC Press LLC
it is now almost universally used by mathematicians to typeset mathe-
matics, and is also widely used in areas of science and engineering that
emphasize mathematics.
1
Part of the reason for T
E
X’s long life and wide use is that it implements
a “markup language” instead of creating output on a computer screen
as you enter it. The computer file that you write for a manuscript is
a text or ASCII file that contains commands that describe in more-or-
less plain English terms how you want the pages formatted and what
mathematical symbols you want to use. Viewing output as it might be
printed is a separate step. This has the advantage that your computer
source file is independent of improvements to your computer hardware.
Currently laser printers have around 1200 dots per inch or around 10,200
horizontal dots per page, while most current computer monitors go up
to at most 1200 or 1600 horizontal dots per screen. So you can see
that, from the outset, screens display material to a different tolerance
than the printer will print it. With the help of its markup commands,
T
E
X positions elements on the printed page to within 10
−6
of an inch.
In effect, the T
E
X code mandates exactly how you want the document
to look. The screen provides an approximation to this truth, usually
suitable for accuracy checking. Depending on the quality of your printing
device, the printer will give a hard-copy rendition of the ideal document
dictated by the T
E
X code. The superb degree of accuracy of which T
E
X
is capable should protect T
E
X from obsolescence for many decades to
come.
Fortunately, you do not have to decide yourself how to position each
character on the printed page to within a millionth of an inch. T
E
X
makes most of these decisions itself. You can use T
E
X commands to
customize these decisions, but this is usually not advisable unless you
have a considerable amount of experience with either typesetting or T
E
X.
However, as we described in Chapter 1, there are a number of situations
where human intervention is necessary for highest-quality output. One
of the main purposes of this book is to explain what those situations are
and how to deal with them, or else how to communicate with a copy
editor and/or typesetter so that he or she can deal with them.
Like many languages that have been around for some time, T
E
X has
a number of dialects. The most common dialects are L
A
T
E
X, Plain T
E
X,
/
/
o-T
E
X, and /
/
o-L
A
T
E
X. These are roughly as close as American
English and English English, with Plain T
E
X being slightly more di-
vergent. Computer installations that support one of these dialects will
generally support all four.
1
In fact, T
E
X has been used by law offices and even by the periodical TV Guide. The
portability—in the large and in the small—of T
E
X make it a powerful and versatile
tool.
©2001 CRC Press LLC
A note about pronunciation: T
E
X is pronounced as in “technical,” not
as in “Texas”. Many purists pronounce the final “X” like the “ch” sound
in German or Scottish, but a simple “k” sound is more than sufficient.
The most common pronunciation of L
A
T
E
X is lay

-tek, presumably in
analogy with the word “latex” in “latex gloves”. Some people, however,
enjoy saying lah

-tek.
A command or special symbol in T
E
X is typically an alphanumeric
word preceded by the symbol ¸ (the “backslash”). Some examples are
2
\alpha , \parindent , \vskip.35cm , \bigtriangleup ,
\matrix , \biggl , \kern-.2in
A more complete list of T
E
X commands appears in Appendix VII; see
also Appendices III and VIII. Ordinary words are entered as text, just as
they would be with a typewriter or a word processor. In the remainder
of this chapter we shall provide a more detailed picture of T
E
X.
2.2 Methods of Typesetting Mathematics
In today’s world, there are basically five types of mathematical typeset-
ting:
(1) Cold type typesetting: This is the most traditional form of the art,
used for centuries. Characters, diphthongs, accents, and spaces are
actualized in pieces of lead. These lead slugs are positioned in a
rack, and that rack serves as the initial printing plate.
(2) Photoelectric typesetting: The typesetter works from a keyboard,
and his output is actualized on film. From that piece of film, an
image is produced on a copper or zinc plate. The plate is etched,
and becomes the printing plate.
(3) Paper tape typesetting: This is a variant of (2), in which the
output is on a stream of paper tape (known in the industry as
“idiot tape”). The tape is then used—as in a Turing machine—to
operate either a phototypesetting or metal typesetting machine.
(4) Camera-ready copy: The manuscript is produced by typewriter,
word processor, calligraphy, or other means. It is photographed,
2
The first command typesets the Greek letter α, the second is for paragraph inden-
tation, the third is a vertical space of size .35 centimeters, and the fourth is a large
triangular symbol () like the Laplace operator. On the second line, the first com-
mand is for forming a matrix, the second controls the size of a left delimiter, and the
third moves the point of typesetting 0.2 inches to the left.
©2001 CRC Press LLC
and then the photographic image reproduced on a copper or zinc
plate. The plate is then etched, and becomes the printing plate.
(5) T
E
X: The work is produced in T
E
X computer code. The source file
is compiled and a *.dvi file produced. The *.dvi file is translated
to PostScript. The PostScript file is used to make an image of
each page on film. The film is then used to produce an image on a
copper or zinc plate. The plate is etched and becomes the printing
plate.
Of these five techniques for mathematical typesetting, (5) is becoming
the most prevalent. This book will concentrate on ideas and techniques
connected to (5).
2.3 A Lightning Tour of T
E
X
When you create a document with a word processor, you enter text
and perhaps mathematical formulas (using suitable syntax) from a com-
puter keyboard; the code then appears on a CRT or computer screen.
The word processor formats your text as entered, perhaps with right
justification (i.e., alignment at the right margin) if desired.
Most word processors have poor capabilities for typesetting math-
ematics. On the positive side, most word processors are essentially
WYSIWYG (“What you see is what you get.”). Whatever you see on
the screen is more-or-less what is printed, although individual text char-
acters or graphics may be printed at a higher pixel resolution.
T
E
X is more sophisticated. It gives you microscopic control over for-
matting. With T
E
X, as in other high-level word processors, you can print
your page in multiple columns, flow text around a graphic or chart, and
import multiple fonts and typeset them in a bewildering array of sizes
and positions. T
E
X will give you more control over positioning and sizing
than will any other document preparation system. T
E
X has more fonts
and (with METAFONT) the ability to create fonts without limit. But,
most important for the mathematician, you can integrate mathematics
3
into your work and have complete control over the outcome. You also
can integrate and manipulate graphics in T
E
X in a variety of useful and
convenient formats.
From the point of view of portability, T
E
X has a decided advantage
over word processors. First of all, T
E
X is virtually platform-independent.
3
You can download utilities from the Web to tackle a variety of very complex dia-
grams (mathematical commutative diagrams and others), as well as chemical charts,
Feynman diagrams, tables, graphs, bar charts, histograms, chess problems, musical
scores, and many other challenging typesetting tasks. See Chapter 7 and Appendix
VI for details on Web resources.
©2001 CRC Press LLC
2.3 A Lightning Tour of T
E
X
myfile.tex → myfile.dvi →
_
¸
¸
¸
¸
¸
¸
¸
¸
_
¸
¸
¸
¸
¸
¸
¸
¸
_
directly to printer
directly to computer screen
myfile.hp or other file that can
be understood by your printer
myfile.ps (a PostScript file for
various output devices)
myfile.pdf (Acrobat

file)
Figure 2.1
You can enter your T
E
X code (into an ASCII file called myfile.tex, for
example) on a PDP-11, compile it on a Macintosh,

and preview it on
a Pentium machine. But it is portable in another important sense: if
you set up a complicated display in TeX, then you can copy or transfer
that piece of T
E
X code to another part of your document or to another
document altogether and the result (in the compiled output) will be
unchanged. This sort of “translation-invariance” does not obtain with
word processors. Because a word processor has hidden formatting com-
mands, cutting and pasting often corrupts the formatting of text, and
you must waste time fiddling around and re-formatting. That will never
happen with T
E
X.
An effective T
E
X user must learn to read many T
E
X commands and
T
E
X code contexts and understand what they say and what effect they
have on the printed output. He must also learn the concepts that under-
pin the basic logic by which T
E
X operates, and the fundamental com-
mands that select a typesetting operation and then implement it.
We begin by diagraming the logic of creating a T
E
X document. Sup-
pose that you are creating a business report. You might entitle the
source code file myfile.tex. Although it is not mandatory, you will
usually find it more convenient to give your source code file the three
letter extension .tex in its name. Here myfile.tex is the file that you
write in a text editor or word processor.
4
The file myfile.tex is called the T
E
X source file, and contains T
E
X
commands and unformatted text as you have entered it. A software
engineer might view myfile.tex as analogous to the source code file for
4
This file should be a straight text or ASCII file. In Microsoft Word, the file should be
saved as either “Text Only” or “Text Only with Line Breaks”. Other word processors
might require “non-document mode” or “ASCII with carriage returns and line feeds.”
If you are using a text editor that is part of an integrated T
E
X package, then it will
write plain ASCII output files by design and you need not worry. Many dedicated
T
E
X users use one of the standard programmer’s editors: emacs or vi on a UNIX
system, or epsilon or crisp on a PC.
©2001 CRC Press LLC
a FORTRAN or C or Java computer program. Unlike a word processor,
T
E
X does not emulate the ultimate page of output; instead, T
E
X issues
formatting commands that specify where and how the page elements
will appear.
Refer to Figure 2.1. The first arrow in the diagram represents the
“compile” operation, which converts the source code file to a *.dvi file
(this latter is the binary file that carries all the machine-coded infor-
mation about your document). The binary codes in this file resemble
codes that might be sent to a printer or computer screen, but are in
a language that is independent of any specific printer or computer. In
this sense, the compiled file is “device independent” (hence the name
myfile.dvi), and is usually called a *.dvi file for short. The infor-
mation in the *.dvi file specifies positions on the printed page within
10
−6
of an inch (in particular, it thinks of each character or typesetting
element as occupying a box, and records the dimensions and position of
each such box), and also has information that determines line breaks,
page breaks, between-line and between-paragraph spacing, what fonts
should be used, and other formatting specifications.
Software is available for most all modern (personal) computers to in-
terpret T
E
X *.dvi files. In particular, you can create myfile.dvi on
a Macintosh or UNIX computer, display it on a DEC Alpha, and then
print it on an Intel Pentium machine. This platform universality is a
second reason for the long life of T
E
X. If you buy a new computer or
upgrade your computer hardware, you merely must make sure that you
have a text editor, that the T
E
X program (that converts myfile.tex to
myfile.dvi) works on your new computer, and that you have screen
and printer programs that can display *.dvi files.
5
These steps taken
together are generally easier than porting an entire integrated word-
processing package to a new machine. Such a word processor will cer-
tainly be platform-dependent, and its data files will not travel well.
The second arrow in Figure 2.1 represents some of the things that you
can do with a *.dvi file. (Printing or screen display refers to the typeset
output, not the literal T
E
X commands and text in myfile.tex.) Current
printers generally have much finer resolution than current display devices
(see Section 2.1). Most screen preview programs have magnification
features to make it easier to see fine detail. Some T
E
X preview programs
display smaller characters more coarsely to make them easier to read.
This is sometimes called “hinting,” and is not used if that part of the
5
If you are a real hacker, then you can actually obtain the C and Pascal source code
for T
E
X and compile it on any machine you like. Most of us will instead obtain a
suitable version of T
E
X, for whatever machine we may wish to use, from a commercial
vendor or from the Web. Appendix VI discusses Web sites that offer public domain
versions of T
E
X for a variety of platforms.
©2001 CRC Press LLC
screen output is sufficiently magnified. Other preview programs do not
use hinting. This makes the output more attractive and closer to what
would be printed, but also makes smaller characters harder to read.
There are some T
E
X programs that write output files directly in Adobe
PDF

format instead of *.dvi format. There are also widely-used com-
puter programs that convert *.dvi files to either Adobe PostScript or
PDF format. One advantage of PDF and PostScript files is that they
generally contain complete font information, while *.dvi files do not
(the *.dvi file only records a sequence of boxes). This has the advan-
tage that screen display and printer programs do not need access to font
information in auxiliary files. However, displayed or printed PDF or
PostScript files can be noticeably cruder if they are displayed or printed
at a different resolution than that of their embedded font information
(unless the user is smart enough to use scalable fonts).
6
2.4 The Guts of T
E
X
In this section, we discuss some of the types of commands and text that
can go into a typical T
E
X source file. This is meant as part of a brief
overview of what can go into a *.tex file. See references such as [SAK],
[KNU], [LAM], or [GMS] for more detail. Also refer to Appendices III,
VII, and VIII.
Text: Literal text that you want to appear in your printed output.
This is entered directly into the source file, just as with typewriter or
traditional word processor. For example, if you enter the text
The cat with a bat saw a rat take a hat off the mat.
then the sentence
The cat with a bat saw a rat take a hat off the mat.
will appear in your output—either on screen or on the printed page.
You begin a new paragraph by leaving a blank line in your source code.
T
E
X will understand that all the material appearing in an aggregate
clump—without vertical spacing—is a paragraph. No matter how you
type it, with any number of words on a line and any number of spaces
between words, T
E
X will collapse an positive number of spaces into a
single space and will calculate the correct number of words to put on
a line. It will format the text into paragraphs in a highly professional
way, and will also hyphenate words if need be.
6
Refer to Appendix VI for some URLs where PostScript resources can be obtained.
©2001 CRC Press LLC
Fonts: A font is a style of type. According to [GMS], a font has five
attributes:
• monospaced or proportionately spaced
• serif or sans serif
• shape
• weight
• width
Perhaps a word of explanation is in order.
On a traditional typewriter, the font is “monospaced.” This means in
effect that each character occupies the same amount of horizontal space.
If you think about how a typewriter mechanism works then monospacing
makes a lot of sense: after you strike each key, the carriage moves by
a pre-determined amount of space. So of course each character (from
“i” to “m” to “w”) must have the same width. But genuine typesetting
is not monospaced; it is kerned. This means that the way space is set
around a given character depends both on that character and on what
characters are its neighbors. Certainly one of the characteristic features
of T
E
X is that it is not monospaced; T
E
X is a full-featured typesetting
system and each character comes equipped with complete information
as to how it should be typeset against every other character.
A serif on a character is a small embellishment (a perpendicular cross)
at the end of one of the main strokes that composes the character. The
roman character “I” has two serifs: one at the top and one at the bottom.
The roman character “R” has three serifs.
7
The attributes “shape” and “width” are self-explanatory.
The “weight” of a character describes the thickness of the strokes
that make up the character. For example, a boldface b is composed of
thicker strokes (i.e., has more weight) than a roman b.
The size of a font is generally specified in points (where a typesetter’s
“point” is about
1
72
of an inch), but other units are sometimes used. The
Glossary defines a number of different typesetter’s units of measure.
In T
E
X, and in other typesetting situations, the word “font” can also
refer to the size of the type (measured in points). Some of the standard
fonts are roman, boldface, and italic. These are normally from the
Computer Modern font family in T
E
X, but some T
E
X implementations
7
A useful font for some applications is the “sans serif” style of font. As the name
suggests, the characters in such a font have no serifs. The phrase This is a sans serif
font is set in sans serif. We sometimes use sans serif for abstracts, headings, and
displayed statements. Some people like to use sans serif in business correspondence.
©2001 CRC Press LLC
use Times Roman or Lucida or other popular fonts. Other common T
E
X
fonts are big cap-small cap, slanted roman, and typewriter-like.
Today there are many hundreds of fonts available for T
E
X. These in-
clude bitmap fonts, outline fonts, PostScript fonts, METAFONT fonts,
and other types of fonts. The books [SAK] and [GMS] contain a thor-
ough discussion of fonts.
The default font in T
E
X is (Computer Modern) roman. (The basic
Computer Modern fonts were created by D. Knuth at the time that
he created T
E
X.) Most of the words in this paragraph are typeset in
roman. This is the font that T
E
X uses in text unless you enter T
E
X
commands that mandate otherwise. In L
A
T
E
X 2.09, to typeset a phrase
in (for example) boldface, isolate the phrase with braces and put the
command \bf just after the opening brace. Thus the T
E
X code
As you can see, {\bf this is boldface} text.
produces the output
As you can see, this is boldface text.
Similarly, you can use the commands \it, \sc, \sl, and \tt (inside
matched braces) to typeset (respectively) text in italics, big cap-small
cap, slanted roman, and typewriter-like fonts.
L
A
T
E
X2
ε
has a more sophisticated protocol for invoking fonts other
than the default roman. Although L
A
T
E
X2
ε
will recognize the L
A
T
E
X
2.09 commands, there are advantages to using the new command (to
obtain boldface, for example)
As you can see, \textbf{this is boldface} text.
The output is of course the same. In order to keep the logic straight,
you must keep in mind that (among other things), L
A
T
E
X2
ε
is a superset
of /
/
o-T
E
X, and /
/
o-T
E
X now has distinct commands for boldface
in text mode and boldface in math mode. The new \textbf command
works in both text and math modes, and in all versions of T
E
X. It has
other sophisticated capabilities which are best learned by reading [GMS].
Of course there are analogous commands for italic (instead of the L
A
T
E
X
2.09 command \it one uses the L
A
T
E
X2
ε
command \textit), and the
more sophisticated L
A
T
E
X2
ε
command even takes into account the special
spacing requirements that the built-in slant of italic characters requires.
The even more sophisticated L
A
T
E
X2
ε
command \bfseries calls in not
just one font but a whole series of fonts. We again refer the reader to
[GMS] for the details of these subtle and powerful commands.
A curiosity in T
E
X is that it offers “old style” numerals for specialized
applications. These are contained in the font cmmi10. A sample of such
numerals is 1:¸¸¸6¿8µo.
©2001 CRC Press LLC
Formatting Mathematics: In T
E
X, learning the commands for dif-
ferent mathematical symbols is easy. It is the formatting that is difficult.
We will treat only the most elementary formatting questions here.
Although you may never have explicitly noted the fact before, spacing
in text is done differently from spacing in mathematics. Thus T
E
X has
two environments: the text environment and the mathematics environ-
ment. The text environment is the default: just start entering ASCII
code, with no special commands, and you are in the text environment.
There are two slightly different mathematics environments: inline and
display. To enter the inline mathematics environment, type $ and then
begin your mathematics. Type a terminal $ when you are done. To en-
ter the display mathematics environment, type $$ and then begin your
mathematics. Type a terminal $$ when you are done. For instance, the
T
E
X code
Let $y = x^2 + 2 + \int_1^x t\sin t\,dt$. That says it.
produces the typeset output
Let j = r
2
+ 2 +
_
x
1
t sin t dt. That says it.
In contrast, the T
E
X code
Let $$y = x^2 + 2 + \int_1^x t\sin t\,dt.$$ That says it.
produces output with the formula in display math mode:
Let
j = r
2
+ 2 +
_
x
1
t sin t dt.
That says it.
Since T
E
X has more room (vertical and horizontal space) in display
mode, it uses a larger integral sign in the integral and a less-cramped
style for upper and lower limits on the integral.
To keep your source file from getting too cramped, it is better to enter
code for display math as
Let
$$
y = x^2 + 2 + \int_1^x t\sin t\,dt .
$$
That says it.
The effect to T
E
X is the same, but the source file will be much easier to
read.
©2001 CRC Press LLC
Many T
E
X gurus recommend an alternative method for designating
math mode. For inline math mode, the usage (instead of $ $) is
\(f(x) = x^2 + 3\) and for displayed math mode, the usage (instead
of $$ $$) is \[f(x) = x^2 + 3\]. One advantage of these notations
is that they are oriented: the user can tell the left delimiter from the
right; this fact will presumably help to reduce errors, and help to find
them when they occur.
Mathematical Symbols: Many different technical symbols are used
in mathematics. See also Appendices III, VII, and VIII. The T
E
X com-
mands for these symbols are mostly straightforward to learn. The fol-
lowing chart will get you well on your way to typesetting mathematics.
T
E
X Code Typeset Result
\alpha α
\beta β
. . . . . .
\omega ω
\Gamma Γ
\Delta ∆
. . . . . .
\Omega Ω
\int_a^b f(x) \, dx
_
b
a
)(r) dr
{d \over dx}
d
dr
A_j^m ¹
m
j
\sum_{j=1}^\infty a_j/b_j^2


j=1
o
j
,/
2
j
\pmatrix{ a & b \\ c & d}
_
o /
c d
_
{a+b \over c} \leq {E^x \over s^y}
o + /
c

1
x
:
y
\overline{\partial} u = \nabla g ∂n = ∇p
Observe that & is used as T
E
X’s alignment character. If you need the
actual symbol &, then enter the T
E
X code \&.
©2001 CRC Press LLC
Aligned Equations: Equations that are aligned at an equality or in-
equality sign are common in mathematical writing. The T
E
X commands
for typesetting aligned equations vary slightly with the T
E
X dialect. In
L
A
T
E
X, the code
\begin{eqnarray*}
y &=& (x+2)(x+5) - (x+1)(x+6) \\
&=& x^2 + 7x + 10 - (x^2 + 7x +6) \\
&=& 4 .
\end{eqnarray*}
produces the output
j = (r + 2)(r + 5) −(r + 1)(r + 6)
= r
2
+ 7r + 10 −(r
2
+ 7r + 6)
= 4.
L
A
T
E
X, by default, generates equation numbers for all equations in an
array. The * in \begin{eqnarray*} and \end{eqnarray*} suppresses
the equation numbers. If you enter instead
\begin{eqnarray}
y &=& (x+2)(x+5) - (x+1)(x+6) \\
&=& x^2 + 7x + 10 - (x^2 + 7x +6) \nonumber \\
&=& 4 .
\end{eqnarray}
then the output becomes
j = (r + 2)(r + 5) −(r + 1)(r + 6) (2.1)
= r
2
+ 7r + 10 −(r
2
+ 7r + 6)
= 4. (2.2)
In Plain T
E
X or /
/
o-T
E
X, the code
$$
\eqalign {
y &= x^2 + 2x + 17 \cr
&\le w^3 + 4 w^{A \otimes B} \cr
&\equiv z\times w \cr }
$$
is typeset as
j = r
2
+ 2r + 17
≤ n
3
+ 4n
A⊗B
≡ . n.
©2001 CRC Press LLC
For equation numbers, enter
$$
\eqalignno {
y &= x^2 + 2x + 17 &(17)\cr
&\le w^3 + 4 w^{A \otimes B} \cr
&\equiv z\times w . &(18)\cr }
$$
The output is
j = r
2
+ 2r + 17 (17)
≤ n
3
+ 4n
A⊗B
≡ . n. (18)
To put the equation numbers on the left, replace \eqalignno by
\leqalignno.
Many people like to put their equation numbers on the left. Others
like to put them on the right. The wisdom among typesetters is that the
equation number is less important than the equation itself. As a result,
the recommendation is that the label come after the equation (i.e., on
the right).
Finally, choose a system for numbering your equations that is trans-
parent, and makes it easy to find equations. The reference [KRW] dis-
cusses this matter in greater detail.
2.5 Modes of Typesetting Mathematics
There are two standard formats for mathematics on the typeset page:
inline mathematics and displayed mathematics. Inline mathematics is
part of the body of text; displayed mathematics is set off from the text.
In broad strokes, they are very much the same—except that the dis-
played math is sometimes larger overall. Let us now discuss, in some
detail, their similarities and differences.
Similarities: The default in T
E
X (for example) is to put (essentially)
no spaces between characters in either of the math modes. This point
must be clarified. If you enter the code $The cat sat in the hat.$—
putting an English sentence in math mode—then the typeset output will
be
T/ccot:otint/c/ot.
Not only is the font incorrect, but the standard between-word spacing
is gone. T
E
X is actually very smart at finding the correct spaces to put
between symbols in math formulas (for example the spaces around a +
sign, or the space just to the right of an integral
_
sign). But some spaces
©2001 CRC Press LLC
(for example just before the dr in an integral) must be hand-inserted
by the user or typesetter. And, by default, the characters are set in a
font called math italic. If you look in any mathematics monograph, you
will confirm instantly that the font in mathematical formulas is different
from the roman font used for text.
Differences: Displayed math will be generally larger than inline math.
That is, certain components, such as integral signs and summation signs
will be larger. The result is that an expression in displayed math mode
will appear altogether larger than that same expression in inline math
mode. The text itself in a math display will be the same size as ordinary
text.
A more interesting difference is in the way that sums and integrals and
other mathematical artifacts with labels above and below are typeset.
Consider the following examples:
Inline Mode Display Mode
(1)

10
j=1
o
j
vs.
10

j=1
o
j
.
(2)
_
b
a
)(r) dr vs.
_
b
a
)(r) dr .
(3) ∪
j∈A
o
j
vs.
_
j∈A
o
j
.
In each instance, the displayed version is more attractive in its layout,
and easier to read. But it is also clear that such a user-friendly display
would not work in inline mode, because the limits above and below
would interfere with the text in the adjacent lines. That is why the
inline mathematics mode has such labels adjacent to the carrier rather
than above and below it.
2.6 Line Breaks in Displayed Mathematics
It is often the case that a displayed mathematical formula, or string of
inequalities, cannot fit comfortably on one line. Compare the display
(r+3)
4
= r
4
+12r
3
+54r
2
+108r+81 = r
_
108+r
_
54+r(12+r)
¸
_
+81
©2001 CRC Press LLC
with the much more readable display
(r + 3)
4
= r
4
+ 12r
3
+ 54r
2
+ 108r + 81
= r
_
108 + r
_
54 + r(12 + r)
¸
_
+ 81.
In practice, it is best to disabuse oneself of the idea that a displayed
equation belongs on one line, and instead to think in terms of displaying
mathematics in multi-line chunks. In T
E
X, this means to think in terms
of the command \eqalign{ } and in L
A
T
E
X it means to think in terms
of \eqnarray{ } rather than the displayed math command
$$ . . . $$.
In substantive mathematical discourse, very long and complicated
equations are quite common. The issue therefore arises of how to break
a long equation across a line.
The first choice is to break at a “verb” connective. Our last example
showed breaks at = signs, and breaks at ≤ or ≥ or similar connectives
are very natural and easy to read. But it is sometimes necessary to break
at connectives like a + or − sign, or perhaps in the middle of a product
of two large expressions. We now discuss, by way of some examples, how
best to do this.
EXAMPLE I: The break
(r + 2j)
6
= r
6
+ 12r
5
j + 60r
4
j
2
+ 160r
3
j
3
+240r
2
j
4
+ 192rj
5
+ 64j
6
is more attractive than either
(r + 2j)
6
= r
6
+ 12r
5
j + 60r
4
j
2
+ 160r
3
j
3
+240r
2
j
4
+ 192rj
5
+ 64j
6
or
(r + 2j)
6
= r
6
+ 12r
5
j + 60r
4
j
2
+ 160r
3
j
3
+
240r
2
j
4
+ 192rj
5
+ 64j
6
.
First of all, it is a good idea to indent the second line, so that the reader’s
eye is signaled of a continuation. Second, it is recommended to place
the plus sign on the second line so that the reader knows that the line
is a continuation.
©2001 CRC Press LLC
¹
3
+
rj
2
+ c
x
+ 7 log 1
−4rj
3
+ .n
2
=
_
erf(: + t) −cosh t
3
j
: +
a+b
c
_

_
r
2
−4j
r + n −tan θ
ln(3t) + 2ncos ψ + arg(τ)
_
is more attractive than either
¹
3
+
rj
2
+ c
x
+ 7 log 1
−4rj
3
+ .n
2
=
_
erf(: + t) −cosht
3
j
: +
a+b
c
_

_
r
2
−4j
r + n −tanθ
ln(3t) + 2ncos ψ + arg(τ)
_
or
¹
3
+
rj
2
+ c
x
+ 7 log 1
−4rj
3
+ .n
2
=
_
erf(: + t) −cosh t
3
j
: +
a+b
c
_

_
r
2
−4j
r + n −tanθ
ln(3t) + 2ncos ψ + arg(τ)
_
or
¹
3
+
rj
2
+ c
x
+ 7 log 1
−4rj
3
+ .n
2
=
_
erf(: + t) −cosht
3
j
: +
a+b
c
_
_
r
2
−4j
r + n −tan θ
ln(3t) + 2ncos ψ + arg(τ)
_
.
First of all, it is a good idea to indent the second line, so that the
reader’s eye is signaled of a continuation. Second, it is best to include
the initial on the second line so that the reader knows that the line is
a continuation. Third, the multiplicative is hard to read and too easily
misinterpreted. Last, while juxtaposition of two expressions is commonly
interpreted in mathematics to mean multiplication, that meaning is lost
when the two expressions occur on different lines.
As mathematicians, we have the ability to use notation as a weapon.
This device is exploited with insufficient frequency. For instance, in the
last example one could write
Let
α =
erf(: + t) −cosh t
3
j
: +
a+b
c
,
β = r
2
−4j
r + n −tan θ
ln(3t) + 2ncos ψ + arg(τ)
,
©2001 CRC Press LLC
and
γ =
rj
2
+ c
x
+ 7 log 1
−4rj
3
+ .n
2
.
Then
¹
3
+ γ = α β .
EXAMPLE III: If the material in your display is grouped with paren-
theses ( ), or braces ¦ ¦, or brackets [ ], then it is best to keep matching
pairs on the same line. But if the expression between delimiters is quite
long and complicated, then it may be necessary to break it. Doing so,
you want to be certain that the left and right delimiters match (in both
size and shape), and that they are easily and quickly paired up in the
reader’s eye. Thus it is best to make them a bit larger than is perhaps
necessary. For example, the displayed expression
j + r
2
=
_
r −j
r + 2j

o −/
o + 2/

: + n
:−n
+
: + t
: −4t
_
16n
4
is much easier to read than
j + r
2
=
_
r −j
r + 2j

o −/
o + 2/

: + n
:−n
+
: + t
: −4t
_
16n
4
or
j + r
2
=
_
[r −j],[r + 2j] −¦[o −/],[o + 2/]¦ ¦[: + n],[:−n]¦
+[: + t],[: −4t]
_
16n
4
.
It is helpful to have the material on the second line shorter than the
material on the first—so that the second line looks like an appendage
(which is the effect that you want).
EXAMPLE IV: If a displayed expression is very long, then it will have
to be broken into several lines (not just two). In such a case, alignment
and judicious use of indentation can be of great assistance to clarity. For
example,
©2001 CRC Press LLC
j +
o + / + c
o −/ + 2c
=
o −/ −c
o + / −2c
+
d + / −c
o −d + 2c
+
o + ) + c
o −/ + 2p

/ + / + /
o −/ + 2/

/ + ) + :
o −n + 2.

: + / + ¡
o −/ + 2j
≤ r
2
+ 3j + 7
is much clearer than
j +
o + / + c
o −/ + 2c
=
o −/ −c
o + / −2c
+
d + / −c
o −d + 2c
+
o + ) + c
o −/ + 2p

/ + / + /
o −/ + 2/

/ + ) + :
o −n + 2.

: + / + ¡
o −/ + 2j
≤ r
2
+ 3j + 7
or
j +
o + / + c
o −/ + 2c
=
o −/ −c
o + / −2c
+
d + / −c
o −d + 2c
+
o + ) + c
o −/ + 2p

/ + / + /
o −/ + 2/

/ + ) + :
o −n + 2.

: + / + ¡
o −/ + 2j
≤ r
2
+ 3j + 7 .
EXAMPLE V: If the first line of a multi-line expression is very long,
then the verb connective (= or ≤ or something similar) and the right-
hand side of the first relation will have to go on a succeeding line. In this
circumstance, the second and subsequent lines should be aligned with a
two-em quad indent from the left:
.
3
−rj.
2
+
o + / −c
d −c
2
+ )
+ sin[r
3
j −.j
2
+ rj.n]
= .
x
j
w
:
qr
:t + o / c
de
) p −/
i
,
k
|
m
n
o
j
≤ 23 −45 + 17 + 333 −4769.
2.7 Types of Space
T
E
X is very good at the spacing that is needed for a well-formatted
mathematics formula. It knows how much space to put after an integral
©2001 CRC Press LLC
sign, how to space out a matrix, and how to typeset an arrow. But the
default in T
E
X mathematics mode is to put no space between objects.
Thus, for example, the T
E
X code
$x y z^2 w$
is typeset as rj.
2
n. In situations like this, all spacing that occurs is
imposed by the user. So one must have positive cognizance of spacing.
Some of the most standard horizontal spaces in typesetting are these
• The thin space: This is a horizontal space with thickness approx-
imately 0.023 in. Some standard instances in mathematics where
the thin space occurs are
1. Before and after a connective that can be read as a “verb”:
These include
=

= → ⊂ ⊃ ≡
.
= ∈ ±
Examples are r = j, ¹ ⊃ 1, and o ∈ ¹. T
E
X supplies the
space automatically as part of the “glue” that comes with
each of the connectives.
2. In an integral:
_
b
a
)(r) dr.
There is a thin space both before and after the )(r). (In T
E
X,
the first of these is automatic and the second not.)
3. After a coefficient:
α

j=1
r
j
β
_
b
a
)(r) dr.
4. Before and after a connective that can be read as a “conjunc-
tion”. These include
+ − ÷ ∩ ∨ ⊕
Examples include r + j, o ∩ T, and \ ⊕\.
5. Before, but not after, certain binary connectives when they
are used as adjectives:
¦¦ + 3 ¦¦ −4 ¦¦ ±8
6. After function names that are set in roman type:
sinr log 3 deg α dimT exp 4.
©2001 CRC Press LLC
7. After commas in a list or sequence:
¦1, 2, 3¦ (1, 2) o, /, c, . . .
8. Before and after a single vertical bar or a colon used as a
mathematical symbol:
¹ [ 1 ¦r: r
2

9. Before a “back subscript” or “back superscript”:
r
3
1
H o
2
`
• The thick space: This is a horizontal space of thickness equal to
two thin spaces. Some standard instances in mathematics where
the thick space occurs are:
1. In congruences (before the parentheses):
9 = 3 (mod 6)
2. In conditional statements (before the parentheses) in text:
r
n
n (n = 1, 2, . . .)
• The em quad: This space is so named because, traditionally, it
was the width of the letter capital “M” in the font being used.
Nowadays the definition is variable, and depends on the font and on
other parameters as well. Some standard instances in mathematics
where the em quad occurs are:
1. Between mathematical expressions and verbal statements:
o
n
0 for most values of n
or
o ,= / by Theorem 1.3.8.
2. Around conjunctions (or other connectives):
r 0 or j < 5
or
o
n
0 if n ≥ 5.
• The two-em quad: The space is so named because it is twice the
thickness of an em quad. Some standard instances in mathematics
where the two-em quad occurs are:
©2001 CRC Press LLC
1. Between two equations:
r
2
−j
3
= 9, r
4
+ j
5
= 3 .
2. To separate an equation from an accompanying condition in
display:
o
n
0 (n = 0, 1, . . .)
or
/
n
< 0 (n odd).
• Negative space: The most commonly used negative space com-
mand in T
E
X is \! . It denotes a negative thin space; that is to
say, an amount of space equal to a thin space is removed at the
place where the command occurs.
Here is a common use of \!. If you wish to typeset two integrals
in a row, such as
__
)(r, j) drdj,
then it will not do to typeset
\int \int f(x,y) \, dx dy.
In fact, every character or expression in T
E
X comes equipped with
“glue”, which is a certain amount of spacing that is formed around
it. When two integrals are juxtaposed, their built-in spacing su-
perimposes, and the result is too great a space between the two
integral signs. Using the T
E
X code \int \int f(x,y) \, dx dy,
one obtains
_ _
)(r, j) drdj.
The improved T
E
X code \int \! \! \! \int f(x,y) \, dx dy
produces the more desirable
__
)(r, j) drdj.
Note that L
A
T
E
X2
ε
has the command \iint which is, in effect,
a macro for the formatting protocol that we have just defined.
The analogous commands \iiint and \iiiint produce triple and
quadruple integrals.
• Forced horizontal space: In T
E
X, one can mandate a horizontal
space of any size with the command \hskip#, where # denotes a
linear measure. For instance, the code
©2001 CRC Press LLC
Bob \hskip.3in and Sally or Sally \hskip-.2in and Bob
yields the compiled output
Bob and Sally or Sally and Bob
• Kerning: To kern is to move the position at which the typesetting
is taking place—either left-right or up-down. Thus if one enters
the T
E
X code
book \kern-.1in binder and book \kern.1in binder
then one obtains the output
bookbinder and book binder
This is an extreme example, just to illustrate what kerning does.
But one can plainly see that negative kerning will have the same
effect as a negative space, and positive kerning will have the same
effect as a positive space (except that the horizontal space com-
mand \hskip has a certain amount of flexible latitude, or “glue”,
associated with it while \kern does not).
2.8 Technical Issues
Here we list and discuss some technical matters that come up frequently
in mathematical typesetting. We recommend the book [SWA] for the
mathematical typesetting details and [SAK] for technicalities of T
E
X.
Here we concentrate on just a few issues. We will frequently filter these
topics through the sensibilities of T
E
X.
Abbreviations in Mathematics: Certain words, such as “sine”, “co-
sine”, “angle”, and “exponential”, are commonly abbreviated in math-
ematical formulas. Thus we see, in equations, expressions like
sinr + cos θ −exp t = 0 .
What you should observe about these expressions is that (i) the abbre-
viated word is in roman (not italic) and (ii) there is a thin space after
the abbreviated word and before the argument (unless the argument is
in parentheses or brackets or braces, in which case there is no space).
T
E
X supplies macros for the most commonly used math abbreviations.
For example, \sin in math mode gives the sine function in roman fol-
lowed by a thin space. If you fail to use the built-in T
E
X command
©2001 CRC Press LLC
for sine, and instead enter the code $sin x$, then you will get the un-
sightly result :inr. (Is this an eccentric rendition of the word “sinx” or
the product of :, i, n, r or something else?)
For some functions, T
E
X will not have the name built in and you
will have to define your own macro. As an instance, you can enter
\def\sgn{\mathop{\rm sgn}} to define a T
E
X command \sgn as an
abbreviation for the signum function.
Blackboard Bold Characters: Because it is difficult to write boldface
characters with a piece of chalk at the blackboard, the font blackb oard
b old was invented. Over time, people became accustomed to black-
board bold and it was used in print as well. When this author published
[KRA1], the publisher declined to use blackboard bold for the complex
numbers because the font needed to be imported from Japan and was
considered to be too expensive. Nowadays, however, it is a common-
place. It is standard in printed mathematics to denote the integers,
rationals, reals, and complex numbers as Z, ¸, 1, and C. Donald Knuth
does not like the blackboard bold font (see [KNU] to learn why), hence
blackboard bold characters are not a standard part of T
E
X or L
A
T
E
X.
The American Mathematical Society (AMS), however, has come to
the rescue. The AMS font msbm has these characters, but you need to
learn how to call the font. You should include the lines
8
\newfam\msbfam
\font\tenmsb=msbm10 \textfont\msbfam=\tenmsb
\font\sevenmsb=msbm7 \scriptfont\msbfam=\sevenmsb
\font\fivemsb=msbm5 \scriptscriptfont\msbfam=\fivemsb
\def\Bbb{\fam\msbfam \tenmsb}
\def\ZZ{{\Bbb Z}}
\def\QQ{{\Bbb Q}}
\def\RR{{\Bbb R}}
\def\CC{{\Bbb C}}
at the top of your *.tex file. Then typeset the characters with
$\ZZ, \QQ, \RR, \CC$. The result is Z, ¸, 1, C.
Delimiters, Size of: A delimiter is a bracketing device such as a paren-
thesis or a brace or a bracket. The book [SAK] contains a discursive dis-
cussion of delimiters—their names, their meaning, their use, and their
limitations. Delimiters parse information, group information, and orga-
nize information. It is essential, for maximum clarity in mathematical
8
Calling a text font involves a simple one-line command—see Section 3.6. Calling a
math font involves calling an entire family and is a bit complicated. See [SAK] for a
thorough and cogent discussion of these matters.
©2001 CRC Press LLC
writing, to use delimiters of the correct size, shape, and position. A
passage like
j = r + (. +
_
1
0
)(t) dt −
o
/
)
(typeset with the T
E
X code
$$
y = x + (z+\int_0^1 f(t) \, dt-\frac{a}{b} )
$$
is hard to read just because the parentheses are too small. T
E
X will size
the delimiters for you if you simply modify your code in this way:
$$
y = x + \left ( z+\int_0^1 f(t) \, dt-\frac{a}{b} \right ) .
$$
With this change, the compiled output is
j = r +
_
. +
_
1
0
)(t) dt −
o
/
_
.
The commands \left and \right generally do a good job at suiting
delimiters to the text that comes between them. T
E
X measures the
heights of the intervening boxes and adapts the delimiters to those boxes.
But sometimes you must intervene when T
E
X makes a poor choice. For
example, in the expression
_
n+1

k=1
/
/
2
+ 1
_
2
,
typeset with the code
$$
\left ( \sum_{k=1}^{n+1} \frac{k}{k^2 + 1} \right )^2
$$
the parentheses are a bit too large. Since the limits / = 1 and n +1 are
not seen by the human eye as an integral part of the summation, they
can (and should) be allowed to extrude slightly beyond the upper and
lower limits of the parentheses. In circumstances such as these, we use
explicitly sized delimiters. The commands for explicit sizing are
\bigl \bigm \bigr \big
\Bigl \Bigm \Bigr \Big
\biggl \biggm \biggr \bigg
\Biggl \Biggm \Biggr \Bigg
©2001 CRC Press LLC
Thus we would typeset
$$
\biggl ( \sum_{k=1}^{n+1} \frac{k}{k^2 + 1} \biggr )^2
$$
to obtain the somewhat more desirable output
_
n+1

k=1
/
/
2
+ 1
_
2
.
Complete details may be found in [SAK]. We sometimes also use ex-
plicitly sized delimiters when the left and right units occur on different
lines. L
A
T
E
X has alternative tools for handling that situation (see [LAM]
or [GMS]).
Ellipses: An ellipsis is a sequence of three dots used either to indicate
a break in text or the trailing end of a thought. For instance
Consider the sequences o
1
, o
2
, . . . , o
N
and /
1
, . . . , /
M
of real
numbers. Define ¹ = o
1
o
2
o
N
, 1 = /
1
+ + /
M
, and
C = /
1
/
M
. Now we ponder these numbers . . . .
The displayed text exhibits six ellipses: one on the baseline in the list of
the o
j
’s, one on the baseline in the list of the /
j
’s, one in the definition
of the product, one in the sum, one in the definition of C, and one at
the end of the passage. If the “missing material” is bracketed by binary
operators, then usually a raised (or vertically centered) ellipsis is used.
Otherwise a baseline ellipsis is appropriate. In almost all applications in
mathematics, an ellipsis is surrounded either by operators or commas.
If the ellipsis occurs at the end of a sentence, then you must include
a fourth dot (the period) to denote the end of the sentence—unless, of
course, you mean to indicate that the sentence is incomplete.
In T
E
X, it is incorrect to form an ellipsis simply by concatenating
three periods (for a baseline ellipsis) or by concatenating three \cdot
commands to adjoin three centered dots. The reason is that the period
. and the centered dot each come with glue, and they do not perform
well under concatenation. T
E
X provides us with \dots for a baseline
ellipsis and \cdots for a raised (or vertically centered) ellipsis. Please
use them.
Embellishments, Accents, and Diacritical Marks: Twenty-six al-
phabetical letters, both upper- and lower- case, are generally insufficient
mat´erial to express our mathematical thoughts. Therefore we frequently
use embellishments over single letters. Examples are
ˆ
) , ˇ c , ˙ o , ˘ c , · .
©2001 CRC Press LLC
It is best not to over-use embellished symbols, as they are harder to read
(and to remember) than un-embellished ones. An equation such as
ˇ o ` c + ` c
´ :,
ˆ
) − ¯ n ˜ n
= /

+ ˙ o + ˘ .
is a nightmare to read and to comprehend.
As noted in Section 1.2, compound embellishments are best avoided,
because they are difficult to typeset and may interfere with the line
above. Embellishments under letters are generally a bad idea, as they
can interfere with letters (like g ) that hang below the baseline.
If you are tempted to flout the suggestions in the previous paragraph,
consider instead using a boldface character, or a character from another
alphabet.
Hard vs. Soft Spaces: If you enter plain text in a *.tex source file,
then T
E
X will decide where to put line breaks. The first choice for a line
break is between words (it obviously only requires a trivial filter for T
E
X
to find these), and the second choice is at the inter-syllable hyphenation
of a word. In fact T
E
X comes equipped with a rather remarkable algo-
rithm for hyphenating words, and usually gets these breaks right. But
there are some places in your text where you definitely do not want a line
break, and other places where not having a break is definitely preferred.
As an instance of the first of these, you do not want a break in “Section
5” between the word “Section” and the number “5”. As an instance of
the second of these, you probably do not want a break in “Mr. Smith”
between the title “Mr.” and the name “Smith”. In either of these cir-
cumstances, or in situations like them, you use the T
E
X code Section~5
or Mr.~Smith to tell T
E
X that you want a space, but the space cannot
be broken at the end of a line. The books [KNU] and [HIG] provide
more information on the use of the so-called hard space ~ .
One-symbol Formulas: In a passage like
Let r be a real variable. The function ) is assumed to operate
on r in the following way:
it is essential that the one-symbol formulas r and ) be formulated in
math italic. If not, the sentence would be difficult to read. This ob-
servation simply means that you must enter math mode (using either $
signs or \( and \)) to typeset the r and the ). Thus the correct code is
Let $x$ be a real variable. The function $f$ is assumed
to operate on $x$ in the following way:
©2001 CRC Press LLC
By contrast, the incorrect code
Let x be a real variable. The function f is assumed
to operate on x in the following way:
produces the less desirable output
Let x be a real variable. The function f is assumed to operate
on x in the following way:
Page Breaks: T
E
X usually has pretty good judgement about where to
break pages, and probably you do too. But here are a few breaks to
avoid:
1. In the middle of a displayed equation or sequence of equations that
run more than one line;
2. In a series of equations that are numbered as a group;
3. After the first line of a new section (i.e., a new section title should
be followed by at least two lines of that section on the same page);
4. So as to leave a widow (see Section 1.4);
5. So as to leave an orphan (see Section 1.4);
6. So as to isolate the last three lines of a chapter.
If there is a sequence of equations separated by commas, then it is
all right to put a page break in the middle of the sequence (the logic
being that such a sequence does not make up a single assertion). An
acceptable place to break a page is before a short phrase, not the end of
a sentence, that precedes a displayed equation.
Punctuation and Formulas, Interaction of: The final word about
punctuation in formulas has not yet been enunciated, and it will not
be enunciated here. In particular, there is considerable disagreement
concerning whether punctuation should appear at the end of a displayed
formula. That is, should it be
Substituting the values for r, j, and . results in the formula
¹ = 1
2
+ C1.
or
Substituting the values for r, j, and . results in the formula
¹ = 1
2
+ C1
©2001 CRC Press LLC
One argument for the second protocol in favor of the first is that if
the formula was replaced by a figure or photograph, then you certainly
would not put a period at its lower right. An argument in favor of the
first protocol over the second is that, if you read the sentence aloud,
then the formula translates to words and demands a period.
Many scientific style manuals militate against punctuation at the end
of a displayed formula. Most mathematics books and journals do include
the punctuation. After all, if the formula was typeset in inline mode,
then you most certainly would include the punctuation. So why not
do the same with a displayed formula? On the other hand, the vertical
space after a displayed formula provides the necessary logical pause; with
that thought in mind, punctuation is redundant.
Clearly the question does not have a clear answer. In your own writing,
you should choose a paradigm for punctuating formulas and then be
consistent. Some publishers will have a preferred style, and you should
learn it and adhere to it.
Struts, Use of: Struts are a technical device that allows T
E
X to provide
a “roof” on which to rest certain symbols. For example, the T
E
X code
$$
\sqrt{a} = \sqrt{X} + \sqrt{y}
$$
typesets rather badly as

o =

A +

j
with square root signs of different heights. The situation may be reme-
died by the use of the \mathstrut command:
$$
\sqrt{\mathstrut a}=\sqrt{\mathstrut X}+\sqrt{\mathstrut y}
$$
The \mathstrut provides an invisible strut of zero width atop which
the square root rests. The output now is
_
o =
_
A +
_
j
and all the square roots have the same height.
There is also a \strut command for use in text mode, but we shall
say nothing about it here. Complete details are in [SAK].
2.9 Including Graphics in a
T
E
X Document
The original design of T
E
X made no provision for the inclusion of graph-
ics stored in external graphics files. The number of popular graphics
©2001 CRC Press LLC
formats was growing then and is probably continuing to grow as you
read this book, and any assumptions about graphics formats of external
files would have rapidly made those parts of T
E
X obsolete.
Today there are two principal types of graphics: bitmap graphics and
vector graphics. The Web site
http://www.rcc.ryerson.ca/rta/brd038/nsr/ptenk/bitvec1.htm
has a lovely discussion of the merits of these two graphic protocols.
Briefly, a bitmap graphic stores the image pixel by pixel while a vector
graphic uses a mathematical language to describe lines and curves. Vec-
tor graphic files tend to be smaller than bitmap files. Vector images have
the advantage of being easily scalable, while bitmap images are better
at providing the detail needed in high quality (photographic, for exam-
ple) images. New technologies are being developed that combine the
best features of both formats. In what follows, we discuss both vector
graphic formats such as PostScript or Acrobat (actually both of these
are much more than vector graphics systems) and also bitmap graphic
formats, such as *.gif and *.bmp.
T
E
X does have the capacity for producing graphics internally by po-
sitioning text characters at arbitrary positions within an open box. The
P
I
CT
E
X package for T
E
X and the L
A
T
E
X picture environment are based
on this approach. The T
E
X companion program METAFONT can also be
used to produce graphics. METAFONT is used to design fonts for T
E
X,
and uses a script language to specify the shapes of characters (so it is, in
effect, a vector graphics language). This script language can be used to
produce line drawings. The result would appear to the user as a new font
with a single character that might be of arbitrary size. But these pack-
ages do not have the power and flexibility of general-purpose computer
packages for producing bitmapped graphics or line/vector graphics. In
particular, scaling and rotation of graphics is generally awkward in the
METAFONT environment.
One significant advantage of the “internal” graphics packages just de-
scribed is that they travel well. If you re-position material in your docu-
ment, add or change labels to figures, scale figures, etc., then everything
works as it should—just like any other part of the T
E
X document. Third
party graphics packages pose real problems with scaling, and especially
with labeling (move a figure and the labels may stay where the figure
used to be). PSTricks is one of many packages that addresses this prob-
lem with captions and labels. This point will be discussed in more detail
below.
Even though out-of-the-box T
E
X has no provision for graphics, the
door was left open for extensions. The T
E
X command \special{...}
allows the user to send whatever information he or she wants directly
to the printer or screen display program. The printer or screen display
©2001 CRC Press LLC
program can then do whatever it likes with this information. The most
common use of \special commands is to send to the printer information
about external graphics files and about how the information in these files
should be positioned and scaled.
In recent years, PostScript has become widely used for the production
and manipulation of graphics. T
E
X printer and screen display programs
that can read and display PostScript graphics files are widely available.
The book [GRM] is devoted to this topic. We cannot include all of
that information here.
9
However, please note that you can use any
good Web browser—such as Netscape

— to view a PostScript file. In
the Windows

environment, simply click on <File and then <Open
Page and enter the path of the *.ps file. Another useful utility for
this purpose is ghostview, which can be downloaded from the Web.
Traditional (non-electronic) methods for the inclusion of graphics in
a book or paper are more laborious. The author creates line drawings,
photographs, and/or halftones on separate pieces of paper. The figures
are submitted along with the finished manuscript to the publisher, who
handles the technical details from that point. Specifically, specialists
in the production department at the publishing house will either re-
draw each figure from scratch or will scan each figure into MacDraw

or
CorelDRAW

or some other powerful graphics software and render it as
a cleanly and professionally drawn figure. This “traditional” method is
still used, and most current publishers will perform this service for an
author.
10
There are advantages and disadvantages to handling graphics yourself,
as an integral part of the electronic document, as opposed to submitting
graphics to the publisher on separate pieces of paper. Among these are:
Advantage: You have complete control over the form and the format-
ting of your document. You will be certain that each of your
graphics is exactly what you want, positioned as you want, and
scaled as you want, in just the form that you want.
11
9
Refer to Appendix VI for some URLs where Acrobat and PostScript resources can
be obtained.
10
There may, however, be a cost. The publisher may offer a lower royalty rate if his
production department must do the figures. I have heard experienced editors yearn
for the “good old days when the publishing house did everything.” But the modern
business reality is that the house wants to outsource to the author.
On the other hand, the publisher may pay the author if he is willing to handle the
graphics himself.
11
If you have never before published a document with graphics, then you may have
no idea what a struggle it is to communicate your desires to a technical artist who
is probably a few thousand miles away and who has no idea what you are trying
to depict. The publisher himself can complicate matters by complaining that the
production of your graphics is becoming too convoluted and too costly.
©2001 CRC Press LLC
Disadvantage: The method you choose for incorporating graphics may
depend on the system you are using and the software you are using.
Disadvantage: You must be sure that you are able to generate output
that your publisher can use. This means not simply that you end
up with satisfactory figures in your printed output, but also that
you end up with a computer file that the publisher can read and
manipulate. This might be a T
E
X or L
A
T
E
X file that has PostScript
input commands and an accompanying PostScript file for each fig-
ure. Or it might be a PostScript file for each chapter, which in-
cludes all the text and all the graphics for that chapter. Most
publishers can handle PostScript files. Many (but not all) publish-
ers can process a T
E
X file with suitable “include” commands for
PostScript graphics files. However, the method for inclusion may
be system-dependent.
Disadvantage: If the publisher wants you to provide “camera-ready”
output in hard copy, then you must have access to a 1200 dpi
or higher resolution printer, and you must be able to obtain the
sort of RC (resin-coated) paper that is required for camera-ready
output. If your publisher is able to transfer your data directly from
a computer disk or computer file to film,
12
then this step will not
be necessary.
Disadvantage: You must determine how to position each graphic on
the page, and decide how to attach labels to each figure. If you
typeset the labels in T
E
X, it may be possible that, each time you
revise your document, all the labels will be disturbed. You may
find it more convenient to make the labels for each graphic a part
of that graphic (i.e., to draw the labels in the fonts provided with
your graphics package, or to use PSTricks or another package that
makes the labels portable).
Disadvantage: If you want to include color graphics, then you may
need to learn additional procedures for color filtering. Tradition-
ally, a “screen” or “benday” was used to sort a color picture into
its yellow, magenta, cyan, and black source. Then four separate
printing plates were generated, and the page was run through the
printer four times.
13
With modern technology, many of these steps
12
Here the film is used to print a photographic image on a (usually copper or zinc)
lithography plate. The plate is then etched and used as the printing master in the
press.
13
These two sentences explain why color graphics traditionally drove up the price of
a book or document.
©2001 CRC Press LLC
may be eliminated. But screening, and keeping track of the pri-
mary colors, is still an important part of the process, and is the
responsibility of the person who is in charge of the graphics.
Disadvantage: If you are creating graphics with CorelDRAW or xfig
or another standard graphics package, then you will have no trou-
ble generating a PostScript file for inclusion in your document.
But there are different versions of PostScript (encapsulated, non-
encapsulated, compressed, and others), and these are not all com-
patible. The heading of the PostScript file must contain crucial
data in just the right form or else you will not be able to use it. If
you are so foolish as to introduce a blank line at the top of your
PostScript file then some systems will not read it.
Disadvantage: You can use graphics packages that are internal to T
E
X
and L
A
T
E
X. These include P
I
CT
E
X, L
A
T
E
X’s native picture en-
vironment, PSTricks, Gnuplot, METAFONT, and others. Such
an internal graphics package has the advantage of portability—its
source code becomes part of the T
E
X source and is as portable
as the *.dvi files that T
E
X produces. Also the labels and cap-
tions become, in effect, part of the figure; when the figure moves
or re-sizes, then so do these ancillaries. However, many of these
packages are limited in the types of graphics they can produce,
and some are quite tedious to use.
Thus, if you are writing a book or article that has graphics, then
you are either going to have to (i) become your own system manager,
and configure your computer and printer to produce the types of files
and hard-copy output that you require, or (ii) consult closely with the
system manager at your place of work to find out what tools are available
on your system for doing the tasks that you need to perform.
In the following, we will describe two or three common methods for
including graphics in your T
E
X document. All of these exploit commonly
used and readily available software. However, there is no guarantee that
precisely these resources will be available on your system. And our
description here is by no means exhaustive. The real hacker will have
other means at his disposal; he can use a preprocessor to decompose a
graphic into a number of fonts, each containing a small piece of the the
graphic as a pixel pattern. Alternatively, graphics can also be made up
of raw printer commands, proprietary to your particular printer, that
you send to the printer using the T
E
X \special command. We shall
not treat any of these more advanced methods here.
©2001 CRC Press LLC
2.9.1 Handling Graphics in the Computer Environment
Powerful commercial packages are available today for rendering graphic
images for almost any application. CorelDRAW, Adobe Illustrator,

Harvard Graphics,

and FreeHand

are some of these. These prod-
ucts are obtainable both for the Windows and the Macintosh platforms.
A UNIX installation will probably include xfig or idraw. The math-
ematics/computer algebra packages Mathematica,

and Maple,

and
MatLab

will generate beautiful graphs and other graphics;
14
they can-
not, however, be used as full-feature graphics packages. For example, if
you needed to draw a figure for a Euclidean geometry text, you would
find it quite difficult to use Mathematica. But a full-featured graph-
ics package like CorelDRAW would make the job easy. Most commerical
graphics packages have the feature that their output can be saved in
PostScript files.
Many of these products have their own proprietary (vector) graphics
language in which figures are encoded. In addition, most graphics pack-
ages nowadays allow you to export your graphic to any of dozens of dif-
ferent graphics formats, including PostScript and a variety of bitmapped
and line graphic formats; these include formats that at one time were
considered proprietary to Microsoft DOS,

UNIX, or Macintosh

sys-
tems. The PostScript format has the advantage of creating an ASCII or
text file, and is one for which the interface with T
E
X is highly developed.
For the purposes of the present discussion, we will assume that you have
converted your graphic to a PostScript file.
15
2.9.2 The Inclusion of a PostScript

Graphic
Suppose that your graphic is called Figure 1, and that the name of your
corresponding PostScript file is fig1.ps.
16
Now your source code file (in
L
A
T
E
X 2.09) should include the line \input{epsf} (in L
A
T
E
X2
ε
it would
14
It is actually possible to produce an entire document using Maple or Mathematica.
The graphics will be rendered in PostScript and the text will be rendered in L
A
T
E
X.
But the T
E
X capabilities are limited. The software Scientific Workplace

is a
document-creation utility that gives L
A
T
E
X output for text and has a Maple kernel
for calculations. One can actually do calculations on the fly as one is writing in
Scientific Workplace.
15
Virtually every UNIX installation, and many personal computer systems too, has
a system utility called dvips. Many versions of this utility can also be downloaded
from the Internet. Using a simple syntax, dvips will convert your *.dvi file to a
PostScript file. It will either send that file to the printer or will store the PostScript
file to disc. The utility dvips can also input external PostScript graphics files and
format them as part of your document.
16
Observe that *.ps is the default extension for a PostScript file name. Also *.eps
is the default extension for an encapsulated PostScript file name.
©2001 CRC Press LLC
be include graphics and in Plain T
E
X it would be \input epsf) in its
preamble; this calls in the package for handling PostScript files. At the
point of impact—where the figure should actually appear—there should
be the lines
\epsfysize=3in
\centerline{\epsfbox{fig1.ps}}
In point of fact, after you compile and view a file containing the code
just described, you may be unsatisfied with the results. Indeed, you may
be unable to find the figure on the page! In practice, the positioning of
graphics is one of the most difficult features of the entire process. You
can, however, cut through much of the complexity by using L
A
T
E
X’s
built-in figure environment. Instead use the code
\begin{figure}
\epsfysize=3in
\centerline{\epsfbox{fig1.ps}}
\caption{This is a picture of whatever it is a picture of.}
\end{figure}
Let us explain what each piece of this code signifies.
L
A
T
E
X’s figure environment turns the object inside it into a “float”.
A float is an object—this could be a graphic or a table or even a body of
text—into a displayed object that does not have a pre-assigned position
in the document. In fact, L
A
T
E
X will solve an optimization problem
to determine where best to place the float, so that it causes the least
difficulty with page and line breaks.
17
Combined with \centerline,
18
the figure environment will usually do a nice job of positioning and
centering the figure, of positioning and centering the caption, and also
of making peace with the text that surrounds the figure.
The \epsfysize command scales or sizes the figure. A smaller number
yields a smaller figure and a larger number yields a larger figure. There
is not necessarily any direct correlation between the numerical value of
size that you mandate (in our example, this is “three inches”or 3in)
and the actual physical size of the figure that you will obtain in the
output. This is, in part, because that size will depend on the size of
the figure with which you began, and on the overall magnification of
the T
E
X document. T
E
X will not be sizing the collection of lines and
objects that make up your figure, but rather the “minipage” on which it
17
It is not difficult to imagine that a poorly placed figure could cause a great deal of
blank space to occur on a page, or a section title that is separated from its section.
18
As you might suppose, this T
E
X command simply centers—from left to right—the
object inside it.
©2001 CRC Press LLC
lives. So you will have to learn through trial-and-error how to size your
figures.
The \caption command provides a suitably typeset caption for the
figure, and numbers it automatically (in just the same way that L
A
T
E
X
numbers chapters and sections and equations). Later, you can use L
A
T
E
X
utilities to collect all the caption names and numbers and pages and
produce (without human intervention) a List of Figures to go in the
front of your book or document.
In all, you will find that the simple five lines of code offered above
will nicely dispatch almost any PostScript graphic that you want to
include in your L
A
T
E
X document. If you are using Plain T
E
X instead of
L
A
T
E
X, then you will need to use the float commands \midinsert and
\endinsert instead of the figure environment. And you will have to
label your figures manually. The other basic ideas are the same.
In both T
E
X and L
A
T
E
X, it is quite common to use the alternative
utility (which is actually a T
E
X package or macro) \psfig for the inclu-
sion of PostScript graphics. The command \psfig{psfile=fig1.ps}
will create the appropriate amount of vertical space for the figure and,
assuming the macro has been suitably customized, will insert a T
E
X
\special command into the *.dvi file that will cause the figure to be
included. There are various options that will enable you to scale the
figure.
2.9.3 Graphics and the L
A
T
E
X2
ε
Environment
L
A
T
E
X2
ε
, the latest implementation of L
A
T
E
X, has considerably enhanced
graphics capabilities (over L
A
T
E
X 2.09, the preceding release). It has both
the graphics and the more advanced graphicx packages. To use the
first of these packages on a PC, one enters
\usepackage[dvips]{graphics}
and then
\includegraphics*[llx, lly][urx, ury]{filename}
to call in a particular graphic file.
19
The use of graphicx is similar.
On many Macintosh installations, the command for implementing
graphics will be
\usepackage[textures]{graphics}
This change has to do with the nature of the T
E
X implementation on a
Mac. Everything else about graphics is the same as for a PC.
19
Here (llx, lly) are the coordinates of the lower left corner of the image and
(urx, ury) are the coordinates of the upper right corner of the image.
©2001 CRC Press LLC
L
A
T
E
X2
ε
has the delightful feature of a uniform syntax for the inclusion
of every kind of graphics file that can be handled by the drivers. In many
respects, it is much easier to use than earlier versions of L
A
T
E
X.
2.9.4 The Use of PCT
E
X

Just to illustrate another method for handling graphics in a T
E
X environ-
ment—perhaps on a home computer—we will now discuss PCT
E
X. In
fact PCT
E
X offers
20
several methods for including graphics. In its cur-
rent implementation, PostScript graphics may be included, viewed, and
printed out. PCT
E
X can also handle *.wmf, *.jpg, and other graphics
formats. It has a full implementation of dvips.
Previous releases did not allow screen viewing of PostScript graphics;
the current implementation does, but it is buggy. Since we have already
discussed PostScript graphics in the preceding subsection anyway, we
shall instead now treat bitmap (or *.bmp) graphics for PCT
E
X.
PCT
E
X will allow you both to preview your T
E
X file with the *.bmp
graphics displayed, and also to print out the results. The commands
needed to carry out this procedure are analogous to those for PostScript
graphics that we described in the last subsection, but have modified
syntax in order to handle bitmaps. Specifically, in the preamble to the
document, include the line \input setbmp. Then, at the point of impact
(where the figure is to appear), include the line
\setbmp{dim1}{dim2}{dim3}{fig1.bmp}
A few words of explanation are in order. In fact, \setbmp is a macro for
the more discursive T
E
X command
\def\setbmp#1#2#3#4{\vskip#3\relax\noindent\hskip#1\relax
\special{bmp:#4 x=#2, y=#3}}
Even those not well-versed in T
E
X will see that the first argument (that
is, #1) corresponds to horizontal spacing, the second (#2) is width, the
third (#3) is vertical spacing, and the fourth (#4) is the name of the file.
Embedded inside the definition of \setbmp is a \special command,
which is the basic tool for inputting printer commands and graphic ma-
chine codes.
A complete example of a typesetting command in the PCT
E
X envi-
ronment for Figure 1 (with file name fig1.bmp) is
20
For specificity, we will discuss PCT
E
X 32,

version 4.0. That was the latest release
at the time of this writing.
©2001 CRC Press LLC
\setbmp{-2.55in}{1.88in}{1.8in}{fig1.bmp}
\begin{center}
Figure 1. A picture of everything.
\end{center}
As in previous discussions, the dimension parameters given here do not
correspond literally to anything that you will actually see in the printed
output. In particular, the way to get the figure positioned properly is
to fiddle. You will have to learn by trial and error what these numbers
signify. The centered material following the figure call is, of course, the
caption.
2.9.5 Freeware that Will Handle Graphics
The T
E
X implementation known as BCT
E
X (for the PC), available for
free from the World Wide Web, has powerful graphics capabilities. An
implementation of BCT
E
X may be downloaded from
http://199.26.180.160/winnt/misc/page2.html
There are certainly more modern freeware versions of T
E
X that are
available as downloads (such as MikT
E
X—see below). But BCT
E
X is
tried and true and readily available; and its graphic interface is partic-
ularly easy to describe.
To import a bitmap graphic (i.e., a *.bmp file) using BCT
E
X, the
command is
\special{isoscale fig1.bmp, 1.5in 1.7in}
You can see the explicit role of the \special command in this example.
The command isoscale is one of several choices in BCT
E
X, indicating
the scaling method (isotropic or non-isotropic) that one wishes to use.
The dimensions 1.5in and 1.7in denote the horizontal and vertical
sizes of the figure. The usual remarks about positioning the figure on
the page still apply.
Today, the software MikT
E
X (for the PC) is one of the most popu-
lar, and most powerful, freeware implementations of T
E
X that can be
downloaded for free from the Web.
21
MikT
E
X has a useful dvips utility,
flexible tools for handling graphics, and many other helpful add-ons. It
may be downloaded from the ftp site
ftp://ctan.tug.org/tex-archive/systems/win32
/miktex/1.20/index.html
21
OzT
E
X is a package available online for Macs. Created by An-
drew Trevorrow, it comes both in freeware and shareware versions. See
http://www.trevorrow.com/oztex/ozfaq.html for more information.
©2001 CRC Press LLC
More information about MikT
E
X (and downloads) can be obtained from
any of the Web sites:
http://members.tripod.com/~upem/miktex.html
http://www.miktex.de
http://wuarchive.wustl.edu/packages/TeX/systems
/win32/miktex/1.20/
(This last is a site at the author’s university.) Incidentally, the CTAN
(Comprehensive T
E
X Archive Network) site is, in general, a great place
to get T
E
X fonts, utilities, articles, and other materials related to T
E
X.
See also Appendix VI for other T
E
X Web sites.
The T
E
X Users Group makes available the T
E
X package T
E
XLive4 (for
the PC). This is a large installation (about 170 Mb), but it contains
useful packages such as Xy-pic (for commutative diagrams). However,
this package is available only for purchase or by joining the T
E
X Users
Group.
For the Macintosh, OzT
E
X is the standard freeware package. Another
is called CMacT
E
X. A Mac T
E
X Web site is
http://www.esm.psu.edu/mac-tex/versions.html
It contains fonts, add-ons, and other useful information about T
E
X for
the Mac user. See Appendix VI for other Internet sites that will be of
interest to the T
E
X user.
©2001 CRC Press LLC
chapter 3
T
E
X and the Typesetting of
Text
Although T
E
X was designed specifically for its prowess with mathemat-
ics, it is also a powerful tool in the typesetting of English prose.
Straight prose, in paragraph format, is child’s play with T
E
X. You
simply type away, as on a traditional typewriter or word processor. You
need not worry about line breaks or page breaks; T
E
X can calculate
those, and take care of right justification as well. To begin a new para-
graph, just leave a blank line. T
E
X will automatically indent the new
paragraph and provide the right amount of vertical space between para-
graphs.
What is tricky in the typesetting of English text is handling various
advanced formatting issues. Certainly the possibilities are limitless: T
E
X
can format a page in two (or more!) columns, can display an indented
block of text, can flow text around a graphic or figure, and can per-
form other standard and exotic typesetting tasks. Fortunately for you,
these issues rarely come up in the writing of a mathematical paper or
book. For those who need these powerful but complicated T
E
X func-
tions, the book [SAK] provides all necessary details. For most of us, the
information in this chapter will suffice.
3.1 Other Word Processors and Typesetting
Systems
Before T
E
X, there were word processors. Compared to the old-fashioned
typewriter, a word processor is a marvel. Using one, you can type as
fast as you like, never fearing for errors. Making changes, corrections,
and revisions is trivial. Most importantly, you can move around large
blocks of text; and you can try things. Word processors have changed
the way people write; and many of the changes are for the good.
©2001 CRC Press LLC
The word processor shows on the screen just what you will be getting on
the printed page. A printer driver is built right into the word processor,
thus guaranteeing complete compatibility between screen output and
hard-copy output. Of course you can store your documents for future
use, and call them up and modify them at will.
But the convenience of a word processor is also the source of its limi-
tations. The best Super VGA screens are only about 1200 pixels
1
across.
A high quality laser printer has a resolution of 1200 dots per inch. Thus
the accuracy, and level of detail, that the printer can achieve is an order
of magnitude higher than what is possible on the computer screen.
Much more powerful is a markup language, such as T
E
X. T
E
X is a
computer typesetting system—not a word processor. In other words,
T
E
X is a high-level computer language, and it allows you to issue com-
mands that declare exactly what will appear on your page and how it
will be formatted.
Of greatest interest for readers of this book is that a word processor
is generally quite limited in its ability to handle mathematics. Microsoft
Word, Corel WordPerfect,

and other popular word processors can, in
fact, format equations. Commercial patches are available which enhance
the mathematical capabilities of these products. But, even so, a word
processor cannot handle the variety of fonts and font sizes, and especially
the formatting and positioning issues, that are essential to the quality
typesetting of mathematics. Try creating a fraction whose numerator
is an integral and whose denominator is the determinant of a matrix.
With T
E
X the matter is a triviality; with your word processor or other
system, good luck.
If you send your work to a publisher and it is created in T
E
X, then the
publisher will tinker with it and then publish it in T
E
X (or in PostScript,
which is easily produced from the output *.dvi file). But if you send
your work to the publisher in WordPerfect then it will have to be type-
set,
2
almost as if you had sent in a typed manuscript.
3
As of this writing, T
E
X is virtually the only complete software system
for typesetting both mathematics and English. It is used universally,
and around the world. It is available in Russian, Japanese, and other
1
A pixel is a screen dot.
2
Not quite. Many publishers can utilize the ASCII code in your word processed
document. But all the mathematics will have to be typeset from scratch.
3
It is worth noting explicitly that it is impossible to create a without-human-
intervention translator to mediate between a word processor and T
E
X. That is
because a word processor and T
E
X are conceptually different: the word processor
is a device of convenience, for putting words on the page in a workable form; T
E
X,
which is a typesetting system, is much more exacting, and specifies page format to a
much higher degree of accuracy than is possible on any word processor.
©2001 CRC Press LLC
languages. It is infinitely portable, and it is stable. Every young (and
old) mathematician should learn to use T
E
X.
3.2 Modes of Typesetting Text
The typeset page has margins mandated, and these usually are fixed
once and for all throughout the entire manuscript. In T
E
X, the default
margins are 1

on all sides. For a full-sized 8.5

11

page and 10 point
type, such margins are too small. You need to select margins that are
suitable for the size of the page, the size of the type, and the purpose of
the document. A medical reference work will have large pages, possibly
two columns, very small type, and narrow margins. A children’s book
will have even larger pages, but also large type and large margins. Most
documents will have specifications somewhere in between.
In a mathematical document—either an article or a report or a book—
most text will be in paragraph mode. Usually paragraphs are indented,
and there is customarily a bit of extra vertical space between paragraphs
(i.e., more space than between successive lines in the same paragraph).
We now spend some time discussing various standard typesetting situa-
tions that deviate from the standard paragraph layout.
Bibliographic References: A great deal might be said about how to
typeset bibliographic references. Consult [HIG], [SWA], and especially
[SKI] for the complete chapter and verse. Our simple advice is that
you should pick a standard format and stick to it. If you use /
/
o-
T
E
X or BibT
E
X, for example, then the package will choose a format for
you. This software has special ways of formatting references to books,
articles, theses, preprints, and other forms of scholarly work. If you use
other typesetting environments, then you will have to exert some control
over the situation.
One standard format for a reference to a paper is
S. Schlobodkin, An incorrect proof of the Riemann hypoth-
esis, Journal of Unpublishable Results 43(1997), 23–59.
Here 43 is the volume number of the journal, 1997 is the year, and 23–59
is the range of pages.
For a book, the information required and the format are slightly dif-
ferent:
Dewyew Luffmea, The Complete Index of Hypothetical The-
orems, Voracious Vanity Press, Vancouver, 1943.
For a thesis, an acceptable format is
A. Lincoln, The slaves should be freed, Thesis, University of
Albemarle, Albemarle, 1994.
©2001 CRC Press LLC
For a preprint, use
J. Davis, I’m not so sure, but we lost the war, preprint.
Of course there are occasionally other types of sources to which you
will refer. One is the “private communication”—indicating that someone
has told you something but that it has not been published anywhere.
This could have been a personal letter, or a conversation, or perhaps an
e-mail message. The custom in this circumstance is to write
J. P. Jones, private communication.
Displayed Quotations: A quotation of length greater than one sen-
tence, or fifteen words, should be displayed. This means that it is set off
from the main body of text, with extra space above and below and wider
margins than the default. Sometimes—especially for long quotations—
the text is set in a smaller font. When a quotation is displayed then
quotation marks are not used. An example is
A rare delicacy in Korea is a gruel made of mosquito eyeballs.
There is a certain bat that eats the mosquitos, and that bat
can digest all parts of the mosquito except the eyeball. Thus
the eyeballs are harvested by processing the leavings of the
bats.
Footnotes: A footnote is designated by a special mark in the text—
sometimes a number or a dagger † or another special symbol—at the
point of impact. Then the text of the footnote appears at the bottom
of the page. This text is generally in a smaller font than the main
text body, and is often separated from the main text by a horizontal
line.
4
The footnote at the bottom of this page was typeset with the
code \footnote{This is an example of a typeset footnote.}.
Many publishers of mathematics discourage footnotes. This is an in-
teresting historical development. In the nineteenth century and early
twentieth century, bibliographic references in a mathematical work were
generally set in footnotes. Now such references are usually collected in
a list at the end of the document. In more modern times, publishers
objected to footnotes because they were extra work for the typesetter.
Also, the footnote labels could be misconstrued as superscripts. With
the advent of electronic typesetting systems—such as T
E
X—this objec-
tion has grown moot. But the custom is still to minimize the use of
footnotes.
4
This is an example of a typeset footnote.
©2001 CRC Press LLC
Marginalia: Sometimes, in a reference work, it is useful to have topic
headings or sometimes even notes to the reader in the margin. In a
textbook, one sometimes has exercise numbers hang-indented into the
margin. But, generally speaking, works of mathematics contain no ma-
terial in the margin. Again, part of the reason for this custom is that
it is extra work for the typesetter to typeset such material. Even with
electronic typesetting systems, it is time-consuming and tedious to set
up marginalia. Marginalia are best avoided.
Even so, there are situations where a “hang indent” format is desir-
able. When you are presenting a list of problems (as in the exercise
section of a textbook), it is common to hang the problem numbers in
the margin. This can be done with the T
E
X code
\medbreak\noindent\llap{{\bf 1.}\enskip}\ignorespaces
Consider a deck of 53 cards. Drop it on the floor and
step on the cards with flamenco shoes \dots
\medbreak\noindent\llap{{\bf 2.}\enskip}\ignorespaces
Three men have \$47 dollars among them. Each man hands
\$3.50 to the man on his right, taking back a quarter
when the other man isn’t looking \dots
The result is
1. Consider a deck of 53 cards. Drop it on the floor and step on the cards
with flamenco shoes . . .
2. Three men have $47 dollars among them. Each man hands $3.50 to
the man on his right, taking back a quarter when the other man isn’t
looking . . .
The key command here is \llap, and the others are of secondary im-
portance. We refer the reader to [SAK, p. 152] for a detailed discussion
of all the T
E
X commands used in this last example.
Sometimes a “hangindent” can make for an attractive display. This
typesetting concept entails having one or more initial lines flush against
the left margin, and the subsequent lines uniformly indented. An exam-
ple is obtained from the T
E
X code
\medbreak \hangindent=.3in \noindent S. G. Krantz,
{\it Function Theory of Several Complex Variables},
Second Edition, Wadsworth Publishing, Belmont,
California, 1992. {\sf This is a fine text, written
©2001 CRC Press LLC
in the classical style. The author, while young,
seems to know what he is talking about.}
which typesets as
S. G. Krantz, Function Theory of Several Complex Variables, Second
Edition, Wadsworth Publishing, Belmont, California, 1992. This is
a fine text, written in the classical style. The author, while young,
seems to know what he is talking about.
The key command here is \hangindent, and the others are of secondary
importance. We refer the reader to [SAK, p. 186] for a detailed discus-
sion of all the T
E
X commands used in this last example.
Tables or Charts: Of course a table or chart cannot appear inline;
it must be displayed—with extra space above and below and on both
sides. A table will often use multiple fonts, perhaps bold for head-
ings and italic for subheadings. L
A
T
E
X has special macros, notably the
tabular environment, for typesetting tables and the like. See also the
macros provided in [SAK]. The creation of a table (unaided) is remark-
ably difficult in Plain T
E
X.
Theorems, Definitions, and Other Displayed Material: It is cus-
tomary in mathematics to organize material into blocks of text, or enun-
ciations; theorems, definitions, examples, remarks, and other devices are
common organizational tools. Usually there is a small amount of extra
vertical space above and below each of these displayed items.
There are many different paradigms for typesetting these units. You
should pick one and use it consistently. Many times, the word “The-
orem” or “Definition” or “Remark” or “Example”, together with its
number or other designation, is set in boldface. Some authors choose to
set the word “theorem” (or other sobriquet) in all caps, or big cap-small
cap, or some other special form; the main thing is to choose a format and
be consistent. The text of the enunciation is generally typeset in roman.
However, many authors find it useful to put the text of an enunciation
in slanted roman, just so that it will stand out.
L
A
T
E
X includes a utility called \newtheorem for typesetting enuncia-
tions. For example, if you want to specify in advance how a proposition
and a lemma will look, then (in your preamble) you include the code
\newtheorem{proposition}{PROPOSITION}
\newtheorem{lemma}{Lemma}
Now when you want to typeset a lemma followed by a theorem then
you use the code
©2001 CRC Press LLC
\begin{lemma}
Let $\epsilon > 0$. For every continuous function $f \dots$
\end{lemma}
\begin{proposition}
Let $\zeta$ be Riemann’s zeta function. Let $P$ be a zero
of $\zeta$ that lies in the critical strip \dots
\end{proposition}
The typeset result is
LEMMA 3.1
Let c 0. For every continuous function ) . . .
PROPOSITION 3.1
Let ζ be Riemann’s zeta function. Let 1 be a zero of ζ that lies in the
critical strip . . .
As you can see, it is the default in L
A
T
E
X to typeset these displayed
units in italic. This author finds such a typesetting decision to be in
error; italics were invented to slow down the eye, and should be used
only for single words or short phrases. A lengthy enunciation should not
be typeset in italics. Many typesetters would agree that slanted roman
is preferable to italic in this instance. You can either invoke the alter-
native font by setting \sl by hand or else (in L
A
T
E
X2
ε
) you can use the
command \theorembodyfont.
3.3 Hyphens and Dashes
English typesetting commonly uses three types of hyphens or dashes:
the hyphen ( - ), the en-dash ( – ), and the em-dash ( — ). Each of
these three punctuation marks has a distinct and well-defined use.
The Hyphen: The hyphen is used in standard hyphenation of words:
well-defined, semi-retired, ultra-modern. Although infrequently invoked
by mathematicians, numbers between twenty and one hundred, written
out in words, are hyphenated: forty-two and eight-seven. It is common
in mathematics to refer to the Cauchy-Kowalewski theorem or the Heine-
Borel theorem. Do not put spaces on either side of the hyphen.
©2001 CRC Press LLC
The En-Dash: The en-dash is longer than the hyphen but shorter than
the em-dash. It is customarily the length of an en, which is the width
of the letter “N” in the current font. The en-dash is used to designate
a range of numbers or years: pages 3–42 or the reign of King Henry
(1833–1841). Do not put spaces on either side of the en-dash.
An interesting artifact of modern life is that some publishers (and
authors) consider Cauchy-Kowalewski and Heine-Borel to be politically
incorrect, because the use of the hyphen might suggest that Cauchy and
Kowalewski or Heine and Borel were married (and had therefore hyphen-
ated their names). Thus the publisher will typeset Cauchy–Kowalewski
and Heine–Borel (with an en-dash).
The Em-Dash: The em-dash is the longest of the commonly used
dashes. It is customarily the length of an em, which is the width of the
letter “M” in the current font. The em-dash is used for a pause or break
in a sentence:
Bob began systematically to move Mary’s possessions out of
the apartment—he was really the best man to do it.
There is general agreement that you should not put spaces on either side
of the em-dash.
5
3.4 Alignment
Alignment is a useful tool for organizing material. When making a list,
or creating a table, or designating a correspondence, you will use align-
ment to make your ideas clear.
Name Rank Number
Jones sergeant 37-99405
Smith major 43-89427
Larson lieutenant 11-48950
Masters colonel 53-48045
An example of alignment.
You mark items in a manuscript for alignment by running a vertical
(sinuous) line through the initial letters of the leading words and writing
“align” in the margin.
5
But customs in other countries may vary.
©2001 CRC Press LLC
Making a good table—in which the information is clear and easy to
retrieve—is a skill that must be honed over time. The book [HIG] has
a detailed and informative discussion of how to form clear and easy-to-
read tables. Other fine references for table-making are [CHI] and [TUR].
The book [TUF] is an authoritative source on the display of quantitative
information.
3.5 Typesetting Material in Two Columns
We sometimes encounter two-column typesetting in encyclopedias (for
example the Encyclopaedia Britannica), in medical reference books, and
in insurance contracts. It is rarely appropriate to use the two-column
format in a serious mathematical work. The presence of complex, often
long, equations is directly in conflict with the narrowness of columns. On
the other hand, the two-column format is sometimes used in statistical
writing. Consult the Author Guidelines for any publication you are
considering before you think about setting your text in two columns.
Certainly L
A
T
E
X has macros which make the two-column format a
breeze to invoke. But, using them, you will find yourself doing a lot of
handwork to correct line over-runs and awkward line and page breaks.
L
A
T
E
X2
ε
is more versatile in this regard than is L
A
T
E
X 2.09.
3.6 Some Technical Textual Issues
The formatting of text has a great many technical questions attached to
it. Fortunately, most of them will not apply to mathematical writing.
We focus here on only a few of the most common topics.
Font Calls: The lore of fonts is vast and varied. It is discussed rather
thoroughly in [SAK] and [GMS]. We give just a quick idea here of how
a text font call is done. The font cmss10 is a sans serif font. The basic
logic of a font call is that the name that you use for a font in a T
E
X
call cannot contain any digits. But we keep track of fonts by their name
and their size. So a font is typically stored on your hard drive with a
name that uses both characters and digits—such as cmss10. So you must
assign a nickname to the font; let us use the nickname sanser. Thus,
at the top of your file, you would include the code
\font\sanser=cmss10
This statement assigns the “working nickname” sanser to the font
cmss10.
Then if you want to set the phrase “Quoth the raven, ‘Nevermore’. ”
in sans serif you enter the T
E
X code
©2001 CRC Press LLC
{\sanser Quoth the raven, ‘Nevermore’.}
The compiled output is then
Quoth the raven, ‘Nevermore’.
Periods, Spaces After: One of the ingenious features of T
E
X is that
most every character comes equipped with “glue”. In a nutshell, a char-
acter’s glue tells T
E
X how much space that character wants on either
side of it. This is not a trivial issue. For example, a left parenthesis
wants more space to the left than to the right. A + sign wants an equal
amount of space, of a certain size, on both sides; an = (equal sign) will
want a slightly different amount of space. And, in order to effect attrac-
tive line breaks, T
E
X must build a certain amount of flexibility into its
spacing mandates.
A period “ . ” is often used as a device for ending a sentence. As such,
it wants no space on the left and a double space on the right. In fact,
that is the default space that T
E
X gives to the humble period. But if
instead you are typesetting “Mr. Smith” then you definitely do not want
a double space after the period. In that case you must either typeset
Mr.\ Smith, where the \ is a manual single space, or else typeset
Mr.~Smith, where we have used the hard space discussed in Section 2.8.
If the period follows a capital letter, then T
E
X assumes that the let-
ter is an initial (as in P. G. Wodehouse) and puts a single space after
the period. You may want to use the hard space anyway (typeset as
P.~G.~Wodehouse) to guarantee that there will not be an awkward line
or page break between or after the initials.
Quotation Marks, Usage of: There is reason to be grateful that this
issue does not come up very frequently in mathematical writing. Logic
dictates that one should write
Euclid called it “one of those Pythagorean things”. (∗)
The thought here is that Euclid’s statement is a proper subset of the
full sentence, therefore the closing double quotation mark should come
before the full stop (i.e., the period). Unfortunately, typesetting con-
vention runs contrary to logic. The correct mode is
Euclid called it “one of those Pythagorean things.” (∗∗)
Matters are actually more complicated than the previous paragraph
might indicate. Rules of grammar and syntax, and rules of typesetting
as well, are in a constant state of genesis and flux. The convention we
have described above is very much in force—simply go and consult a
©2001 CRC Press LLC
popular novel to verify this claim for yourself. But the inertia in current
mathematics is to follow the logical rather than the popular; that is to
say, more and more mathematical manuscripts adhere to the first custom
(∗) rather than the second custom (∗∗).
The notion that the punctuation mark should precede the closing
quotation marks applies both to commas and to periods. For colons and
semi-colons the rule is just the opposite: the punctuation mark should
follow the quotations marks. But the rules are different in Britain.
Whatever paradigm you may choose to follow, do be consistent. This
author has advocated (in the book [KRW]) that the standard rule should
be broken when the quotation marks surround a short phrase without a
verb. For example, the sentence
Bob’s favorite words were “apple”, “ludicrous”, “apostle”,
and “perquisite”.
is arguably clearer and less clumsy than
Bob’s favorite words were “apple,” “ludicrous,” “apostle,”
and “perquisite.”
Professional typesetters have told us that this view is probably correct
and sensible, but it ought not to be uttered in public. Judge for yourself.
See also the following discussion of smart quotes vs. straight quotes.
Quotes, Smart vs. Straight: The method in T
E
X for specifying quo-
tation marks is to use pairs of single quotation marks. For example,
the code George said, ‘‘Hello Bob.’’ produces the output George
said, “Hello Bob.” The clear advantage of using pairs of single quotation
marks is that it allows us to specify left quotes and right quotes explic-
itly, so that they are typeset correctly as left curly quotes and right curly
quotes. If you instead use straight quotation marks in your T
E
X code,
i.e., George said, "Hello Bob." then you will get the incorrect result
George said, ”Hello Bob.”
It is easy enough to get used to T
E
X’s method for handling quota-
tions. But in some systems there is now a system of “smart quotes”.
For example, the latest release of Microsoft Word allows you to enter
straight quotes both on the left and the right, and it correctly converts
them to left curly quotes and right curly quotes. The Smart Quotes

software of Sam Kington will automatically convert straight quotes to
(correctly chosen) curly quotes in any desktop situation. Smart Quotes
is freeware. You can become better acquainted with Smart Quotes at
the Web site http://www.illuminated.co.uk/acorn/smartquote/.
Spell Checkers: Whether you use T
E
X or a word processor, do develop
the habit of using a spell checker. You may be the world’s greatest
©2001 CRC Press LLC
speller—winner of every spelling bee from London to Newcastle—but
when typing rapidly you could set teh instead of the or enromous in-
stead of enormous. You could, in your haste, double a consonant that
should not be doubled or put an “i” before an “e” when it should be the
other way around. The spell checker will help you to catch such errors.
Please note that the spell checker will not catch mistakes like typing
homeomorphism for homomorphism or conversation for conservation
because these are all legitimate words. In point of fact, your spell checker
may flag both “homeomorphism” and “homomorphism” as errors be-
cause they are not part of the standard English lexicon. But you can,
and should, add words to your spell checker’s dictionary so that the
jargon you usually use in your work does not get flagged.
If you are a T
E
X user, then your spell checker will flag \parindent
and \eqnarray and other standard T
E
X and L
A
T
E
X commands. But
some popular spell checkers (such as MicroSpell

) have either a T
E
X
mode that will ignore T
E
X commands or a T
E
X filter that will sort out
the T
E
X commands and just check the text.
©2001 CRC Press LLC
chapter 4
Front Matter and Back
Matter
4.1 The Beginning
The front matter for a book comprises the front and back title pages,
the Table of Contents or TOC, the Preface, the Foreword (optional),
the List of Figures (optional), the Acknowledgements (optional), the
Dedication (optional), and like material. The back matter consists of
Appendices (optional), the Bibliography or References, the Glossary,
tables and other resources (optional), and the Index.
L
A
T
E
X and other standard sets of T
E
X macros enable the automatic
generation of the TOC, List of Figures, the Glossary, the Bibliography,
and the Index. A few words of explanation are relevant in this regard.
In the L
A
T
E
X book style, one begins a chapter, a section, and a sub-
section thusly:
\chapter{All about Everything}
\section{Specific Comments about Everything}
\subsection{General Comments about Very Little}
In order to create a Table of Contents, you need to enter the line
\tableofcontents in your source code file after you enter the all-impor-
tant line \begin{document}. L
A
T
E
X is what computer scientists call a
two-pass system.
1
The first time that you run latex on a source file
(call it mybook.tex), it creates an auxiliary file called mybook.toc. The
*.toc file contains information about the names and the page numbers
of the various chapters, sections, and subsections. When you then run
latex on the file a second time, it builds a Table of Contents, typeset in
T
E
X. Of course all the titles and page numbers are literal copies of what
1
For some packages (such as BibT
E
X), even more than two passes may be required.
©2001 CRC Press LLC
actually appears in the text. This is a wonderful labor-saving device,
and guarantees perfect accuracy.
The creation of the List of Figures is just the same. You insert the
command \listoffigures after the \begin{document} line. When
you run latex on your file mybook.tex, then the file mybook.lof is
created. It contains information about the caption, the number, and
the page number for each figure. When you run L
A
T
E
X a second time, a
typeset List of Figures is generated. A List of Tables may be produced
in an identical manner. The interested reader may consult [GMS], and
the various T
E
X Web sites (such as the T
E
X Users Group, whose URL
is http://www.tug.org) will contain additional macros for specialized
applications. See also Appendix VI.
4.2 The End
The back matter for a book consists of the references or bibliography,
the index, perhaps some appendices, and other auxiliary material of the
author’s choosing.
In a mathematics book (as opposed to, say, a romance novel), the
Index is of vital importance. A mathematics book is thick and dense,
and the author should do his utmost to help the reader navigate his way
around. The Index is an important tool in that quest.
In the old days, the author would painstakingly plod through the page
proofs; he would note each word to be indexed, and its page number, on
a 3

5

card. At the end of this process, the author would have a stack
of said cards, and would then alphabetize them. Then he would type out
the index; in the process, he would eliminate redundancies, amalgamate
cognate terms, classify words either as principal entries or subentries
(see the display below for a sample of a finished Index, with entries and
subentries clearly exhibited), and so forth. The entire procedure could
easily take a week of hard work.
A sample passage from an Index might look like this:
homology 37
class 38
group 40
homotopy 29
deformation 31
equivalence 30
In this sample, “homology” and “homotopy” are principal entries and
“class”, “group”, “deformation”, and “equivalence” are subentries.
With T
E
X, the entire process just described has been automated. The
author puts the command \makeindex in the T
E
X source code file before
the \begin{document} command (i.e., in the preamble to the electronic
©2001 CRC Press LLC
file). Then he inserts tagged words in the file wherever they are to be
indexed. In out-of-the-box L
A
T
E
X, if the word “homotopy” occurs in
the T
E
X file in line 1227, then the author adds \index{homotopy} to
that line adjacent to the original occurrence of that word. In L
A
T
E
X2
ε
,
one can also use an index tag of the form \index{homotopy!stable}
to indicate that “homotopy” is a principal index entry and “stable” is
a subentry. When the T
E
X file (call it myfile.tex) is compiled, an
auxiliary file myfile.idx is created. It contains each word that was
tagged, together with the number of the page on which the word occurs.
Now here is data, ready to be manipulated.
Simple operating system commands can now be used to alphabetize
the myfile.idx file. And it is straightforward, using a good text editor,
to eliminate redundancies and to sort entries as either principal entries
or subentries. When the final manipulation is complete, the myfile.idx
file is incorporated into myfile.tex, and it compiles as a finished Index.
Usually, the entire task can be handled in an evening.
Incidentally, with the use of auxiliary T
E
X packages (such as the utility
makeindex) that are available on many systems, much of the editing
described in the last paragraph can be done by machine, without any
human intervention. You simply enter the command
makeindex myfile.idx
and the file myfile.ind will be produced. This new file alphabetizes
the index, removes redundancies, and sorts out index entries and index
subentries. The reference [GMS] gives a more detailed treatment of the
use of the makeindex command.
Many of the public domain T
E
X archives, such as CTAN (see Ap-
pendix VI), contain utilities for performing the sort of sorting and filter-
ing described in the last paragraph. See also the indexing macros that
are provided with the book [SAK].
A glossary may be created using exactly the same procedures as we
have described for creating an index. Words are tagged with the T
E
X
syntax \glossary{ }. The glossary is compiled and formatted in just
the same way.
The creation of bibliographies is a world of its own. Different versions
of T
E
X each have their own macros for formatting the bibliography. In
L
A
T
E
X, for example, each style macro (book, article, report, etc.) has
its own paradigm for formatting the references. There is also BibT
E
X,
which gives the user the power to create a bibliographical database with
a filename *.bib. BibT
E
X is premised on the notion that we tend to use
the same references repeatedly in our work. So why not have them all
in a single database, and add to that database as new references come
along? Each entry in the database—be it book, or article, or thesis, or
other item—has a nickname, and any new document being created can
©2001 CRC Press LLC
invoke an item in the database simply by using its nickname. All very
slick, and you can read the details in [GMS].
As with any good piece of software, BibT
E
X will cause you to re-think
the process of doing bibliographical work. BibT
E
X will
1. allow you to re-format the entire bibliography of a paper or book
simply by selecting a different bibliography style file;
2. allow you to keep the reference lists of all your works up-to-date
(i.e., which preprints have appeared, which book manuscripts have
been published by which publisher, which “private communica-
tions” have actually become papers) simply by keeping your mas-
ter *.bib file up-to-date;
3. allow you to share your bibliographic database (by putting it on
the Web, for instance) and to access the bibliographic databases
of others.
2
Here are some sample entries from a file called mybooks.bib, a typical
database for BibT
E
X:
@ARTICLE{whole-journal,
key = "GAJ",
journal = {\mbox{G-Animal’s} Journal},
year = 1986,
volume = 41,
number = 7,
month = jul,
note = {The entire issue is devoted to gnats
and gnus (this entry is a cross-referenced
ARTICLE (journal))},
}
@INBOOK{inbook-full,
author = "Donald E. Knuth",
title = "Fundamental Algorithms",
volume = 1,
series = "The Art of Computer Programming",
publisher = "Addison-Wesley",
address = "Reading, Massachusetts",
edition = "Second",
month = "10~" # jan,
2
Many subject areas have rather complete, publicly accessible, bibliographic
databases posted on the Web. These are updated regularly.
©2001 CRC Press LLC
year = "{\noopsort{1973b}}1973",
type = "Section",
chapter = "1.2",
pages = "10--119",
note = "This is a full INBOOK entry",
}
@BOOK{book-full,
author = "Donald E. Knuth",
title = "Seminumerical Algorithms",
volume = 2,
series = "The Art of Computer Programming",
publisher = "Addison-Wesley",
address = "Reading, Massachusetts",
edition = "Second",
month = "10~" # jan,
year = "{\noopsort{1973c}}1981",
note = "This is a full BOOK entry",
}
@INPROCEEDINGS{inproceedings-crossref,
crossref = "whole-proceedings",
author = "Alfred V. Oaho and Jeffrey D. Ullman
and Mihalis Yannakakis",
title = "On Notions of Information Transfer in
{VLSI} Circuits",
organization = "",
pages = "133--139",
note = "This is a cross-referencing INPROCEEDINGS
entry",
}
@PHDTHESIS{phdthesis-minimal,
author = "F. Phidias Phony-Baloney",
title = "Fighting Fire with Fire: Festooning
{F}rench Phrases",
school = "Fanstord University",
year = 1988,
}
@TECHREPORT{techreport-full,
author = "Tom T{\’{e}}rrific",
title = "An {$O(n \log n / \! \log\log n)$}
Sorting Algorithm",
©2001 CRC Press LLC
institution = "Fanstord University",
type = "Wishful Research Result",
number = "7",
address = "Computer Science Department,
Fanstord, California",
month = oct,
year = 1988,
note = "This is a full TECHREPORT entry",
}
@UNPUBLISHED{unpublished-full,
author = "Ulrich {\"{U}}nderwood and Ned {\~N}et
and Paul {\={P}}ot",
title = "Lower Bounds for Wishful Research Results",
month = nov # ", " # dec,
year = 1988,
note = "Talk at Fanstord University (this is a
full UNPUBLISHED entry)",
}
Your T
E
X source file myfile.tex might have this form:
\documentclass{article}
\begin{document}
\section{This is the First of Many Sections}
Now is the time for all good men to come. Come, men,
come. An appropriate reference for these ideas is
\cite{whole-journal}. Further developments appear in
\cite{techreport-full}. And if that doesn’t float your
boat,\index{{\sc Bib}\TeX, sample of use}
then look in \cite{phdthesis-minimal}.
\bibliographystyle{plain}
\bibliography{mybooks}
\end{document}
Compiling this T
E
X ASCII source file will result in a document with
a full bibliography at the end and appropriate citations to the biblio-
graphic entries appearing in the text.
©2001 CRC Press LLC

4.3 Concluding Remarks
The front matter of a book will include the Preface, Table of Contents,
Acknowledgements, the front and back title pages,
3
and perhaps a Fore-
word and a brief Biography of the author. The back matter will include
the Bibliography and Index and may also include Appendices, Tables (of
notation, for example), a Glossary, and other auxiliary material. L
A
T
E
X
has dedicated macros for creating most of these book components. We
urge you to consult [GMS] or [LAM] for all the details.
3
The back title page is where the Copyright and ISBN number, and like information,
appears.
©2001 CRC Press LLC
chapter 5
Copy Editing
5.1 Traditional Metho ds of Copy Editing
Traditionally, even as recently as twenty years ago, the copy editor
worked with the hard-copy manuscript submitted by the author. He
first made a pass through the manuscript in order to make notes on
stylistic issues: formatting of section and chapter heads, use of remarks
and examples and other enunciations, numbering of theorems, use of
unusual spellings or constructions, syntax, grammar, usage, specialized
words or jargon, and so forth. Then the copy editor would go through
the manuscript again, correcting each page and bringing it into a con-
sistent form that adheres to the usual customs and paradigms of the
publishing business. All of the copy editor’s marks would be made by
hand, usually with red or blue pencil, in the margins of the manuscript.
The usual collection of copy editor’s marks (see Appendix I and Ap-
pendix II) would be exploited. The copy editor would also write queries
to the author in the margins. Appendix I, Plate 4 shows how this is
done. Then the author would be given a chance to respond to the copy
editor’s ministrations. As an author, you should learn the copy editor’s
marks (Appendix I). They will facilitate your communications with the
copy editor.
5.2 Communicating with Your Copy Editor
The copy editing process is described in detail later in the book. Your
primary method for communicating with your copy editor is by way
of his markings and your markings on the manuscript. Learn to use
the standard proofreader’s marks (see Appendix I); they are as sure
and accurate a way of communicating with your copy editor as learning
French is a sure way of communicating with a resident of Paris. A
complete list of those proofreader’s/copy editor’s marks that are most
useful to the technical writer may be found in Plates 1–4 of Appendix
©2001 CRC Press LLC
I. See also Appendix II for instances of the actual usage of these marks.
The sources [SKI, p. 71] and [SWA, pp. 13, 85] contain further details
about these marks, plus examples of their use. The reference [HIG]
contains some useful exercises.
1
You can also communicate with your copy editor using English prose,
but the chance of a misunderstanding or inaccuracy will be higher. The
fact that you are both native speakers is no guarantee that you will
understand each other.
When communications with the copy editor become bogged down or
too complex—and this will happen occasionally, despite the best in-
tentions of all parties—then you will find yourself talking to the copy
editor on the telephone. Our own experience is that this happens most
frequently in trying to understand graphics, and what they are meant
to represent. We wish to stress that this is not a confrontational trans-
action. Your copy editor wants nothing more than to make your book
or paper come out right. You do too.
Electronic mail and FAXes are also an effective means for communi-
cating with the copy editor. When a substantial (or surprising) change
is necessary, we have sometimes found it useful to re-typeset a passage
(in T
E
X, of course), compile and print it, and then FAX the result to
the copy editor. Alternatively, one could e-mail (or ftp) the T
E
X code
or the *.dvi file.
5.3 Communicating with Your Typesetter
Generally speaking, you will not communicate directly with the typeset-
ter. In many instances you will be the typesetter, so the issue is moot.
But even when the publisher outsources to a third-party typesetter, the
copy editor or production editor will serve as a mouthpiece—both to
you and to the typesetter. This is prudent, for most authors do not
have the experience and the vocabulary to communicate directly with
the typesetter. The copy editor can interpret the author’s wishes for
the typesetter and vice versa. The editor is the unique person in the
transaction who is equipped to see that the page comes out as it should.
1
Note that it will virtually never happen that you and the copy editor are sending
(modified) electronic T
E
X files back and forth. In such a scenario, it would be too
difficult to locate the passages which have been changed, and to compare the old
and the new. There is too much chance for error, and for misunderstanding. The
advantage of a hand-marked hard copy is that you have both the existing text and
the proposed change before your eyes, and you always know what is meant.
©2001 CRC Press LLC
5.4 Communicating with Your Editor
Here we speak of your developmental editor, or your acquisitions editor,
or your production editor (not your copy editor). Much communication
with the editor will be by telephone. In the best of circumstances, the
editor will interpret your wishes as expressed in proofreader’s marks (see
also the discussion in Section 5.2). If your attempts to communicate how
a graphic should appear have failed, then a telephone conversation can
often clear things up. When misunderstanding occurs, then a phone call,
or FAX, or e-mail, or even a FedEx may be necessary. Generally speak-
ing, your relationship with your editor will develop into a salubrious,
friendly, and productive one.
5.5 Mo dern Metho ds of Copy Editing
Today the process just described is frequently streamlined. First, the
copy editor is likely working with a printout of the author’s electronic
file—either T
E
X, or L
A
T
E
X, or Microsoft Word, or some other piece of
document-preparation software. The copy editor would never work di-
rectly with the author’s electronic file; doing so would create too much
potential for confusion (in part because one could not distinguish the
original material from the corrected material), and increase the likeli-
hood of miscommunication with the author.
Some things remain the same: the copy editor will still use the tra-
ditional copy editing marks (Appendix I), and will still place them in
the margins of the manuscript, using red or blue pencil. He will write
queries to the author, by hand, in the margins. As before, these will be
recognizable because each begins with “Au:” or “Qu:” or “Qy:”.
Usually the publisher will send this edited manuscript back to the
author, thus giving the author a chance to respond. Then the publisher
will generate page proofs, and the author will have another look. Other
publishers will have the copy editor work with an already edited ver-
sion of the author’s work, rendered in page proofs, and this will be the
author’s one and only chance to review corrections.
Whether you have a publisher that adheres to the traditional method-
ology or the modern one, take heart. You will have at least one last
chance to review your beloved book before it goes to press.
5.6 More on Interacting with Your Copy Editor
A copy editor is a trained technician, with considerable knowledge of
the English language and syntax. He is not a mathematician. Such a
person will have a good sense of English sentence structure and, because
he does not get lost in the meaning of the sentences he is reading, can
©2001 CRC Press LLC
be a good judge of proper phraseology and overall writing. You should
take the suggestions and corrections of the copy editor seriously, and
reply to each one with care.
But, because of lack of knowledge of mathematics, the copy editor will
make blunders. This author once had the experience of a copy editor
intervening in a discussion of the Riemann zeta function and changing
every occurrence of “critical strip” to “critical ship”. On another occa-
sion, a copy editor changed the name “Riemann mapping theorem” to
“Riemannian mapping theorem” because the name “Riemannian met-
ric” had occurred elsewhere in the manuscript. You will sometimes find
the copy editor recommending a certain construction just because it
seems to be correct English; but it will be incorrect mathematics. In-
deed, what the copy editor is proposing may be just the opposite of what
you want to say. What is one to do?
It is natural for one to become outraged. How dare the copy editor
overstep his role and try to change your carefully chosen prose? If he
is proposing this particular outrage, then what other travesties has he
committed on your beloved manuscript elsewhere?
Be of stout heart. The copy editor is a serious professional, and he
is aware of his limitations. In each instance where the copy editor has
made an inapposite proposal, you merely need to write in the margin
“This may be unfamiliar English, but it is standard mathematical jar-
gon. Please leave as set.” The standard editing argot, whose use will
endear you to the copy editor, is STET (see the description of this
acronym in Appendix I). Just write STET after your comment and your
message will be clear. And you will get no further argument.
As indicated elsewhere, the discussion of figures and other artwork in
the manuscript is going to be trickier. You may have submitted a sketch
that was perfectly clear to you. That clarity stems in part, no doubt,
from the fact that you yourself understand the mathematics behind the
picture. The artist does not. So it is agonizingly likely that the pub-
lisher’s first drafts of some of your figures will not at all represent what
you had intended. Often, in a geometric figure, certain intersections or
tangencies or containments must be depicted just so; the artist will not
know that, and will render the picture rather differently. Somehow you
must react calmly to what you’ve been sent, and communicate your true
feelings so that the next draft will be more accurate.
There are several difficulties with this process. First of all, in many
instances you will be communicating with the copy editor rather than
directly with the commercial artist. Second, it is difficult to articulate in
words what you want the picture to show; that is perhaps why you were
using a picture in the first place. Thus there really is a lot to be said for
either (i) submitting a professionally hand-drawn figure (drawn locally
under your direct supervision!) to the publisher or (ii) learning to use
©2001 CRC Press LLC
software to generate the figures for your work, or at least to generate
a figure accurate enough to communicate to the publisher’s artist what
you want.
If you choose to go the first route, and if you are unable to generate
the hand-drawn figure yourself, then at least you can hire someone to do
it for you, and you can stand right next to him and tell him what you
want. The cost at a university of getting a skilled technician to generate
your graphics is on the order of $30–$40 per figure (in the year 2000).
This is money well spent if you want your book to come out right; and
you may be able to convince your publisher to help to defray the costs.
5.7 Manuscript Proofs, Galley Proofs, and Page
Proofs
In other parts of this book we have made reference to different types of
proofs of your work. To recapitulate, the standard drafts of a work in
progress that you may see are
(1) The Manuscript. This is the original hard-copy draft that you
submit to the publisher. Today, this will generally be the printout
from some electronic document preparation system such as T
E
X,
or Microsoft Word, or some analogous piece of software. Most
publishers will still accept typewritten copy; but working from
such copy will add to the cost of producing the book or paper.
(2) The Galley Proofs. This is the typeset draft of the work, prepared
from the submitted manuscript, but now correct for formatting,
spacing, line breaks, numbering, and English usage and syntax. It
is not correct for page breaks. The galley proofs will not be spaced
for figures, tables, and other displayed matter.
(3) The Page Proofs. This is a representation of how the work will
ultimately appear in print. It preserves all of the format and syntax
formalities that were present in the galley proofs, but is now correct
for page breaks as well. In particular, the page proofs will exhibit
where all figures, tables, and other displays will appear. They will
be correct for size, shape, and position. This is the moment when
“what you see is what you get.” It is your last opportunity to
review your work before it goes to the printer.
Today, many publishers eliminate the galley proof stage; or else they
meld the galley proof and page proof stages into a single stage. Because
of electronic media, it is trivial to reformat page breaks and reposition
figures. Today the publisher is not significantly inconvenienced if, even
at the page proof stage, you need to move or to alter a figure (in the old
©2001 CRC Press LLC
days this would have been a major calamity). Therefore a lot of time
and expense in the production process can be saved, and steps can be
skipped.
5.8 The End of the Process
After you have reviewed your page proofs, you will send them into the
publisher, usually by express mail. That is the end of the editorial
process. Are there any more steps, from the point of view of the author?
There are a few. First, you may be asked to sign a statement pro-
claiming that you accept the manuscript in its current form, you give
permission for it to go to press, and you will offer no further corrections.
If you have been doing your homework all along, then this declaration
should be easy for you.
Now let us assume that you are publishing a book (rather than an
article). If your publisher is being particularly upfront and solicitous
(professional societies are particularly good about this), then you will
be asked/allowed to review the format and content of the cover. Part of
this process is for you to evaluate the artistic and design qualities of the
cover. The other part is for you to review the words on the cover. Of
course make sure that the title is correct and the author’s name spelled
accurately; but also check the text on the back cover and on the flyleaves.
Feel free to offer suggestions, corrections, and emendations. More people
will read this material than will actually read the book, so you want it
to make a good impression. In fact, many decisions on whether to buy
the book will be based on the impression that this material makes. So
be sure that it reflects well on you.
Less certain, but definitely desirable from your point of view, is to
review the advertising copy for your book. Many marketing departments
are rather territorial, and will not let even the editor see what they are
creating. Let us hope that the marketing crew at your publisher’s is not
like that. You should be allowed to see the advertising copy, and you
should be allowed to criticize it just as you did with your copy edited
manuscript. Your editor should act as your broker/attorney, and make
sure that your wishes are met. You certainly do not want the advertising
copy to be inaccurate or embarrassing.
These last two transactions—the cover and the advertising copy—
take place while the book is going into production. By the time you
have approved the advertising copy, your book will be high in the queue
at the printer’s. In today’s market, advertisements for your book will
appear before the book has actually hit the street. Check out the Web
sites:
©2001 CRC Press LLC
http://www.amazon.com
or
http://www.barnesandnoble.com
or the pages of the Notices of the AMS to begin to develop an appreci-
ation for your pear-shaped tones. In a short while, a box of books will
be arriving on your doorstep.
©2001 CRC Press LLC
chapter 6
The Production Process
6.1 Production of a Paper
These days, mathematical authors frequently operate under the delusion
that they themselves will typeset their paper or book in T
E
X or some
other markup language, send the disc into the publisher, and that is the
end of it: The author’s diskette is popped into one end of a big machine
and a box of books or reprints comes out the other end. Not true.
For a first line journal or book, that subscribes to quality typesetting
practices, there is a lot more to the process.
Many authors will begin by writing the manuscript out longhand.
This is so because, when you are writing out complicated mathematical
expressions or ideas, it is easiest to maintain control with a pen or pencil.
Then either the author or his typist renders the document in a “typed
up” form. Today, it is really best if this can be done on a computer.
If you submit a paper to a journal and do not have an ASCII file on
a diskette, then the journal will incur extra expense and aggravation
getting your work typeset. Many journals will give a paper not prepared
in electronic form a low priority; some journals have an explicitly stated
policy to this effect. Other journals simply will not consider papers
unless they are typeset in T
E
X. If you are not yourself a T
E
X user, or
if your department does not have a T
E
X typist, you may actually find
yourself hiring someone to do your T
E
Xing for you.
It is the case now, and always has been, that a journal will expect you
to produce your own artwork. That is to say, when you submit the final
draft of your paper to the journal, then you should submit repro-quality
printouts of your graphics as well. The journal may in turn decide to
render those graphics as bitmaps, or in PostScript, or in some other
graphic format. Generally speaking, however, you cannot expect the
journal to draw your pictures for you.
Let us suppose that you are preparing a paper with the intention of
publishing it in a journal. After the “typed up” document is prepared,
©2001 CRC Press LLC
the manuscript is then submitted—usually in duplicate or triplicate—to
a journal. Unless the journal explicitly encourages electronic submission,
you should submit your work in hard-copy form, accompanied by the
customary cover letter to the editor.
1
Do not submit your diskette at
this time, because there will probably be modifications to the paper—
initiated either by you or by the referee—before the paper is in final,
accepted form.
After the paper is refereed and, presumably, accepted, there may be
some revisions required. When you submit the final draft of the paper,
again send two or three hard copies (marked as revisions and dated), the
diskette with a T
E
X or other suitable electronic file, and the standard
cover letter.
In due course, the author will receive page proofs for the article. This
will not simply be a printout from the disc that the author submitted.
Most mainstream mathematics journals outsource to a T
E
X consultant
who painstakingly works over each paper for the journal, rendering it
in the journal’s style and seeing to it that formulas, and theorems, and
the other parts of each paper are all typeset to an exacting—and to
the same—standard. In an effort to achieve uniformity, some journals
will change the author’s numbering system. Many journals will modify
the author’s method of formatting the references, and particularly his
method of labeling the books and papers listed in the references. Cer-
tainly all of the page breaks, and even some of the line breaks, will be
different from those in the preprint.
Note that, twenty years ago or more, it was common for the author to
receive galley proofs—not broken into pages. These galley proofs would
not have shown placement of figures, tables, and other displayed mat-
ter. Today, because of electronic typesetting, what the author receives
really is a representation of how the article will finally appear. These
proofs will be accompanied by instructions on how to proofread the
manuscript, how to mark corrections, and there may be author queries
noted in the margins. Do follow whatever instructions may be provided
by the publisher for that particular journal; if the publisher provides no
instructions, then instead follow the guidelines provided in this book.
The page proofs will be accompanied by copyright transfer forms and
order forms for reprints and other clerical materials. It is generally re-
quested that proofs be returned within forty-eight hours. Do endeavor
to handle your proofs efficiently, accurately, and with dispatch. Also be
careful to mail your proofs to the correct address—often it will be an
address different from the one to which you originally submitted your
1
The book [KRW] contains a discursive discussion of how to deal with journal editors
and referee’s reports, and we shall not repeat that advice here.
©2001 CRC Press LLC
paper.
That is the end of the story. After you mail your page proofs back to
the publisher, the next you will hear of your work will be on receipt of a
packet of reprints and on your library’s receipt of the appropriate issue
of the journal.
6.2 Production of a Book
The steps in the production of a book are similar to those in the pro-
duction of an article, but there are more of them. We give here a sketch
of the process.
Many authors will begin by writing the book manuscript out longhand.
This is so because, when you are writing out complicated mathemati-
cal expressions or ideas, it is easiest to maintain control with a pen or
pencil. Then either the author or his typist renders the document in
a “typed up” form. Today, it is really best if this can be done on a
computer. Many publishers provide authors with a “manuscript prepa-
ration grant” ($2000 or so) to help defray the cost of getting the book
T
E
Xed. Alternatively, the publisher will offer to pay the author to do
the T
E
Xing. A warning is in order here. In many instances, if you sign a
contract agreeing to produce the T
E
Xed version of the document, then
you may very well be agreeing to produce the final version that meets
all the requirements of the copy and production editors. Thus you have
agreed to engage in a process that goes way beyond just keying in the
mathematics. You will have to wrestle with line and page breaks, spac-
ing, running heads, and other formatting issues. In other words, you
will have to master all the ideas in this book! Some authors will choose
to simply forego the extra cash and let the publisher worry about these
technicalities.
After the “typed up” document is prepared, the manuscript is then
submitted to a book publisher.
2
(Of course if you are already under
contract to a publisher and/or if a publisher has helped you to to pre-
pare the document, then you are already committed and this step is
automatic.) You should definitely submit your work in hard-copy form
only at this time. One copy will suffice, and it should be printed on just
one side of the page, with ample margins (so that editors and reviewers
can make notes in the margins). Do not submit a disc yet, as there will
certainly be changes later on. Although it may seem obvious to observe,
this really needs to be said: Your book manuscript should be as pol-
ished and complete at the time of submission as a mathematics paper
2
A journal article can only be submitted to one journal at a time. By contrast, a
book manuscript can be sent to several different publishers simultaneously.
©2001 CRC Press LLC
that you would submit to a journal. You would never send an article
into the Annals of Mathematics that said
Theorem: The zeros of the Riemann zeta function in the
critical strip all lie on the critical line.
Proof: George—include a plausible argument here. Enough
details to be convincing. Add some figures for pizzazz.
But in fact people submit book manuscripts in this form—incomplete,
with notes that the author has written to himself—all the time.
Your book manuscript should be as polished and complete as you can
make it: include the Table of Contents, Preface or Prospectus, List of
Figures, all text, all illustrations and tables and figures, all Appendices,
the Glossary, the Bibliography, and the Index. You should include a
detailed cover letter explaining just what this manuscript is, and for
what audience it is intended. You should tell the publisher just who
you are. Include your Curriculum Vitae, plus all current information
about your address, phone number, e-mail address, FAX number, and
so forth. The book [KRW] contains detailed information about dealing
with a publisher over a book manuscript; we shall not repeat that advice
here.
In an effort to evaluate your project, the publisher will send your
manuscript out to various editorial advisors and reviewers. After a time,
he will evaluate the reviewer reports and will render a decision about
whether to pursue publication of your manuscript. It is rare that a
publisher will say, “This is great and we will rush it to press.” Most
likely, you will be asked to consider some revisions. Eventually, you will
produce a finished product, complete with all the elements described
above, and be ready to submit this final draft to the publisher. In fact,
there are several ways to do so.
You certainly will want to send the publisher a final hard copy. This
hard copy represents what you think the book is supposed to look like. If
an editor has any trouble with any of your electronic code, or is not sure
what a certain formula or passage is supposed to say, then he will use
your hard copy as a reference.
3
Of course the publisher will also want
your electronic files. If the book is not too long, then a couple of 3.5

diskettes is a perfectly acceptable medium on which to submit your book.
However, if you are submitting PostScript files or graphics in electronic
format, then your files are liable to be quite large. Therefore you may
wish to use a Zip disc (which holds 100 megabytes, as opposed to the
paltry 1.44 megabytes that a 3.5

high density diskette holds).
4
It is also
3
If that fails, then the editor will simply phone you up and ask you.
4
Iomega Zip drives and discs have become a standard in the publishing industry.
©2001 CRC Press LLC
possible, with some publishers, to submit your files over the Internet by
ftp (file transfer protocol). A common procedure is to tar your files
together so that you are manipulating just one big file, compress it with
gzip or some other standard compression utility, and then ftp the file
to a Web site that the publisher has set up for this purpose.
If your files are in a pre-agreed upon format that the publisher can use
(say T
E
X and PostScript), then the technicians at the publishing house
will have no trouble decompressing, un-taring, and printing or viewing
your files.
At this point one or more copy editors will go to work on your book.
Their task is both laborious and painstaking. The copy-editing proce-
dure for a 400 page book could take several weeks. The editor will go over
each page for style, formatting, consistency, (and possibly) grammar,
syntax, and usage. You will then be sent the marked-up manuscript.
Read the publisher’s instructions carefully. If you are doing the type-
setting, then you will respond to the copy-editing by incorporating the
editor’s instructions into the T
E
X source code. If the publisher is do-
ing the typesetting, then it is up to you to read all of the copy editor’s
marks and either to accept them or to change them (by putting your
own marks on the pages) or to challenge them (this latter option will be
the exception). At the end of this process, you will either
(i) [If you are doing the typesetting] Print out a new hard copy and
send it, along with your elecronic files, into the publisher.
(ii) [If the publisher is doing the typesetting] Send the newly marked-up
manuscript back to the publisher.
The publishing industry is almost always in a “hurry up and wait” mode.
Thus it is considered de rigueur to send things in using express mail
(Federal Express or DHL or another service).
The copy editors will now go back to work. Depending on which
publishing house is handling your book, you may now engage in some
give-and-take with the editors, fine tuning the manuscript to a mutually
agreeable finale. Or you may in fact receive a (putatively) final set of
page proofs for your approval.
As with a journal article, the page proofs are your very last opportu-
nity to see how everything will appear—including positioning of figures,
all page breaks, formatting of special sections and materials, and front
and back matter. Take a few days and spend some time with your book
before it goes off to press. This is how it is going to look to your public!
©2001 CRC Press LLC
6.3 What Happ ens at the Printer’s
The reference [SKI, pp. 511–532] gives a detailed and authoritative de-
scription of all the activities that take place in a print shop. It is a long,
complicated, and charming story—fraught with much history and tra-
dition. In particular, page 518 of that book tells all the steps involved
in the production of a book by the traditional methods (that were used
almost universally up until 20 years ago). It would take us far afield to
describe all this history, and most of it is not relevant today. We cannot
take time to tell the whole story here. But we shall give a sketch.
In the old days, here is what would happen once your book was ready
to go into production. The publisher would print out “repro copy” on
special, high quality paper called RC paper. Then a technician would
“shoot” the book: take a photograph of each page. Now each page of
the book was a photographic negative. That negative would then be
photographically printed on an emulsion-coated copper or zinc printing
plate.
5
And that plate would then be engraved, and become the master
printing or lithography plate for the given page.
Today, the process just described has been streamlined. Now there is
hardware and software that will go directly from the *.dvi file to film;
more precisely, the dvi file is converted to a PostScript file and then
ghostview, or ghostscript, or another
6
utility will then transfer the
image to film. The film is then used to produce the image on the plate,
the plate is engraved, and (as before) it is mounted on the printing press.
A modern printing plant is quite a technological operation. No longer
is it the case that the printer performs individual steps of printing sixteen
pages to a sheet, folding the sheets into signatures, binding the sheets
together, and compiling them between flyleafs and covers. In fact, a
machine performs all the steps—without human intervention. At the
output end of the great, rumbling behemoth, a stack of books is ejected.
That is the good news.
The bad news is that such high-tech printing operations are few in
number, and are in great demand. Often the queue at the printer is the
great unknown in the publishing process. All the editing described above
can take two or three months. Then waiting to get the book printed can
take another two or three months (the actual printing itself might take
only an hour or so!!).
Certainly even a well-informed and high-tech author need not know
the details of how the printing process works. But if something goes
5
Depending on the quality of hard-copy output that is desired, the printer will some-
times use plastic or compressed fiber plates.
6
Refer to Appendix VI for some URLs where ghostview, ghostscript, and related
tools may be found.
©2001 CRC Press LLC
wrong, it is helpful if you can speak to your editor knowledgeably about
a reprint, or a tip-and-tear (see the Glossary), or paper weight, or types
of ink. Again, see [SKI] for the chapter and verse on these matters.
The one certain fact—and this is in large part why the present book was
written—is that you do not want to find yourself unable to communicate
your needs or concerns to your editor, and you also do not want to find
yourself unable to understand what he is telling you. You need to learn
the language!
©2001 CRC Press LLC
chapter 7
Publishing on the Web
7.1 Introductory Remarks
The Internet is a vast network of interlinked computers. Thanks to the
World Wide Web, material posted on many of these computers can be
easily accessed from any of the others with a simple, intuitive graphics
interface.
It has become desirable, for many reasons, to publish scientific mate-
rial on the Web. The World Wide Web provides an essentially cost-free,
instantaneous vehicle for universal dissemination of your ideas. Write
up a new theorem tonight, post it on the Web tomorrow, and your
colleagues in Uzbekistan can be reading it right away. No longer are
we dependent on the vagaries of the publisher, the printer, the postal
service, the university library, or other distribution systems.
There are disadvantages to Web publishing from the point of view of
archiving. As of this writing, there does not exist any permanent and
reliable protocol for archiving electronic media. The traditional method
of archiving a book is to print a thousand hard copies of the volume
and to distribute them to libraries (both public and private) around
the world. Since each of those “mirror copies” is stable, we can be
reasonably sure that a fire or a rodent attack at one location will not
affect copies at another location. By contrast, sun spots or cosmic rays
could simultaneously damage all electronic mirror images everywhere.
We encourage you to archive your work in the time-tested hard-copy
format. But, for rapid communication and dissemination, you had better
learn to use the Web.
7.2 How to Get on the Web
If you are on a UNIX system, then in a subdirectory with a name
something like ~/public_html you will create a file with the name
index.html. That file is the basis of your Web page. It may (ulti-
mately) contain hyperlinks to other files, but index.html is where it all
begins. If you are posting straight text, then you will use the univer-
©2001 CRC Press LLC
sal Web language HTML (Hypertext Markup Language). HTML is an
extremely straightforward, high-level computing language which allows
you to issue standard text formatting commands for size, font, spacing,
and so forth. Here is a typical bit of code for an HTML file, taken from
the book [GOR]:
<TITLE>The simplest HTML example</TITLE>
<H1>A level one heading</H1>
<P>Welcome to the world of HTML!
<P>Let’s have a second paragraph.
This will appear on the screen roughly as
A level one heading
Welcome to the world of HTML!
Let’s have a second paragraph.
The title “The Simplest HTML example” will appear in the title bar of
the browser. Of course what the end user actually sees will depend on
what browser he is using; so this “sample” should be taken only as a
suggestion of how HTML code is translated into typeset text.
As you can see, HTML uses tags to tell the Web browser how to format
text. Some tags, such as <TITLE>, have a beginning tag and an ending
tag; these are differentiated by the presence of a forward slash / in the
latter. The <H1> tag indicates a level one heading, which is a large, bold
font. Smaller, lower priority, headings are denoted with <H2>, etc. The
<P> tag indicates a new paragraph. Other tags are similar.
A surfer on the World Wide Web would locate your Web page by
typing in your URL (Uniform Resource Locator), or Web address. The
Web address of Steven G. Krantz is typical:
http://www.math.wustl.edu/~sk
You can see that the address consists of the node in our mathematics
department followed by an identifier for the individual person. Very
simple.
Private Internet providers will have their own protocol for creating a
Web page. If you are an America Online user, or a Compuserve user, or
an EarthLink user, then you will have to learn their particular paradigm
for creating a Web page. But it will be similar to what we have described
for UNIX.
1
1
There is quite a lot of software available for creating Web pages. America Online
and other ISPs have their own utilities. Third-party products—such as Micrografx’s
©2001 CRC Press LLC
7.3 Web Resources
The World Wide Web is growing rapidly as a source of information
about many topics. Thanks to the efforts of many mathematicians,
publishers, and computer scientists, the information on the Web about
mathematematical publishing is growing apace. It would be impossible
here to give anything but a hint of some of the Web resources available
to the mathematical author. We mention just a few favorites.
arXiv Available from www.arXiv.org, this is a preprint server for the
sciences. Initiated by Paul Ginsparg at Los Alamos, the arXiv has be-
come something of a national resource, and a model for how preprint
servers should be run. New preprints are posted without human inter-
vention (apart from the actions of the author to post his work), and can
be downloaded in *.pdf, *.dvi, *.ps, source code, and other formats.
AucT
E
X AucT
E
X is an add-on to the UNIX editor GNU emacs. It is
an extensible package that supports writing and formatting T
E
X files,
particularly those created in L
A
T
E
X and /
/
o-T
E
X. A 60+ page manual
is included in the distribution of AucT
E
X. There is also a link to an
on-line version of the manual.
You can download AucT
E
X from the Web site
http://www.math.auc.dk/~dethlef/Tips/auc.html
That site also contains links to sources for emacs, MiKT
E
X, ghostview,
Ispell, and Lacheck. (Some of these are discussed below.)
Bibweb This is a utility for automatically retrieving bibliographical in-
formation from the American Mathematical Society’s MathSciNet pro-
gram. Requires the use of BibT
E
X. Available from the Web page
http://www.math.washington.edu/~palmieri/Bibweb/
comp.text.tex This is an electronic newsgroup on electronic document
preparation. Their FAQ (list of “Frequently asked Questions”), called
tex-faq, is a terrific source of information about troff, PostScript,
T
E
X, and other typesetting ideas. The useful newsgroup also contains
the utility TeX-index, which is an archive of all public domain T
E
X
macros.
Webtricity, Sierra Home’s Complete Web Studio, IBM’s Top Page, and SoftQuad’s
HotMetal Pro 6.0—may be purchased at any computer store. This software provides
an interface, very much like a word processor, between the user and HTML.
©2001 CRC Press LLC
Front Greg Kuperberg’s contribution to the quality of mathematical
life, Front is a very user-friendly front end for arXiv. You can try out
Front for free at http://front.math.ucdavis.edu.
Ispell Also available from the Web site
http://www.math.auc.dk/~dethlef/Tips/auc.html
Ispell is a spell-checker for all types of documents. You can also down-
load Cygwin from the Web site; it is needed for running Ispell.
JSTOR The Mellon Foundation’s ambitious project to provide elec-
tronic archiving for major scholarly journals. There are now about
130 journals archived on JSTOR, 11 of these in mathematics. New
journals are continually being added. Other fields covered range from
Afro-American Studies to Economics, Philosophy, and Sociology. If a
university subscribes to JSTOR, then there is no limit to the number
of its people who can read the same article at the same time. Arti-
cles may be downloaded and printed. Learn more about JSTOR at
http://www.jstor.org.
Lacheck This is a utility for checking cosmetic errors (at a rather re-
condite technical level) in a T
E
X source code file. Downloadable from
the Web site http://www.math.auc.dk/~dethlef/Tips/auc.html.
MathSciNet Created by the American Mathematical Society, in this
author’s opinion MathSciNet is the most important new mathematics re-
source to come along in many a year. Type in the name of an author and
in a few seconds MathSciNet will return complete bibliographical refer-
ences for all his papers. Type in a full or partial title and MathSciNet
will return all papers with that text in the title. Type in key words
and MathSciNet will return a list of all papers with those words in the
title, or in the Math Reviews review. MathSciNet is also a great tool
for seeing what that “special someone” has been up to in the past few
years.
MathSciNet is not free. In fact, the American Math Society requires
that you subscribe to the hard-copy Math Reviews before you can sub-
scribe to MathSciNet (for an additional fee). But it is worth it. Learn
more about MathSciNet at http://www.ams.org/mathscinet.
MR Lookup This is a new utility for accessing the MathSciNet database
to verify and create references that can link to reviews and original
sources. If you input basic reference data into MR Lookup, then it re-
turns electronic publication-ready references with live links to reviews
©2001 CRC Press LLC
in MathSciNet and to original articles. Available from the AMS Web
site http://www.ams.org/mrlookup.
7.4 Mathematics and the Web
The interesting issue about Web publishing from the point of view of the
mathematician is that HTML cannot handle mathematics. There does
not even exist universally readable HTML code for the Greek letter π.
2
The mathematician seeking to post his ideas on the Web is best advised
to use another medium.
3
Many, if not most, mathematicians will create their documents in
some form of T
E
X. Most publishers prefer L
A
T
E
X, just because it gives
the author fewer choices and because it emphasizes logical markup over
visual markup.
4
So let us assume, for the sake of this brief discussion,
that you have created a mathematics document—perhaps a research pa-
per, or a chapter of a book, or an article for the Notices of the American
Mathematical Society—in L
A
T
E
X. Now you want to put it on the Web.
What do you do?
The simple answer is that your document needs to be rendered in
*.pdf format. Adobe Acrobat is designed to read and manipulate *.pdf
files.
5
An Acrobat file can be recognized by the extension *.pdf (where
pdf stands for “portable document format”). The utility pdftex will, in
fact, turn your T
E
X file directly into *.pdf code. You can use the utility
dvipdf to turn your *.dvi file into a *.pdf file. Alternatively, either
Adobe Acrobat distiller or ghostview will turn a PostScript file into
*.pdf. Then anyone with an Acrobat reader will be able to read your
document, with formatting and resolution quality comparable to that of
T
E
X.
You can, if you wish, use the utility latex2html to translate your
L
A
T
E
X file directly into HTML. If your system is so equipped, it will turn
2
It should be noted that Greek letters are now part of the ISO 10646 standard, and
they are implemented in the HTML specification. Not every Web browser supports
the code, however. As of this writing, HTML does not have an international standard
(like the ANSI Standard in C), and SGML is still under development.
3
Everything about publishing, and especially about publishing mathematics, on the
Web is in flux. It is constantly developing, and at a rapid rate. So whatever we say
about it here should be understood in that context.
4
Any description of “what publishers think” is going to depend on the particular
publishing house, on the particular individual expressing the opinion, and on what
that individual’s position is. Most editors are not fluently conversant with T
E
X, and
view it as a black box. They see T
E
X in terms of what end product it will produce,
and how easily. A member of the publisher’s production department will take a more
technical and empirical view of T
E
X.
5
Refer to Appendix VI for some URLs where Acrobat-related tools may be found.
©2001 CRC Press LLC
your mathematics characters and formulas into bitmaps. The result is
a display that is difficult to read, and files that are not very portable.
If your document includes graphics, then you must be sure that these
are in the proper format. The pdftex utility can handle *.png, *.jpg,
and *.pdf graphics files. Whereas—earlier—we touted encapsulated
PostScript as the way to go for incorporating graphics in a T
E
X doc-
ument, we now must tell you to translate (using ghostview or Adobe
Acrobat distiller, for example) those *.eps files to *.pdf. The util-
ity pdftex is included in the T
E
X implementations MikT
E
X and fpT
E
X,
both of which may be downloaded free from the Web.
One big issue that impacts—philosophically, practically, and perhaps
legalistically—on the person publishing on the Web is this: What the
end user sees is highly dependent on the particular Web browser that
he is using; it does not depend in any essential way on the browser that
the promulgator is using. It is quite possible that the end user’s browser
will format things differently than the author intended, or it could make
font substitutions. For casual reading these differences may not be im-
portant. But for scientific data, or for a table of vital information, or
for a mathematical formula, or a business document, a small slip may
entirely change the intended meaning. If a Web browser encounters a
piece of code that it does not understand, then it will simply pass over
that piece of code. There will be no error message! In that case, the end
user will see a document with the meaning drastically changed. Use of
Acrobat tends to minimize these differences, and to offer some guarantee
of coherence and integrity in the transmission of information.
6
Another set of legal issues concerns copyright infringements. Many
of these questions are still being argued in the courts. Let us note,
for example, that if you are writing a book for a commercial publisher
then at some point you may sign the rights over to that publisher. The
publisher will probably be happy to have you post your book on the
Web up until the time the book appears in (hard copy) print. At that
moment, however, the publisher will tell you in no uncertain terms to
take it down. The situation with research papers is similar. Everyone
wants to have their work posted on the Web. But as soon as the publisher
has the rights to the paper or book then he can exercise control over
where, and in what form, the paper can appear. So you must be careful.
These days there are a great many utilities available for putting your
T
E
X or L
A
T
E
X document on the Web. Space limitations prevent us from
discussing any of these in detail. Suffice it to say that they all have their
strengths, and they all have limitations. Some are accurate but very
slow. Some do not handle fonts well. Some cannot handle scaling. Some
6
Refer to Appendix VI for some URLs where Acrobat-related tools may be found.
©2001 CRC Press LLC
have trouble with graphics. Some do not interact well with certain Web
browsers. The monograph [GOR] gives the full picture of many of the
most popular of these tools. We briefly will mention just a few.
The IBM product techexplorer allows you to, in effect, manipulate
a T
E
X or L
A
T
E
X file as though it were an HTML document. Note that
techexplorer does not treat the *.dvi file; instead it directly treats
the *.tex file. With techexplorer you will manipulate your document
in MathML.
7
Therefore much of the “attribute customization” that can
be performed on an HTML document can also be performed on a T
E
X
or L
A
T
E
X document. The utility techexplorer does not support style
files, is touchy about user-made macros, and requires that you enter
your markup commands in a certain order. It may skip over markup
commands that occur in a place where it does not expect them.
The utility TeX4ht creates configurable hypertext versions of L
A
T
E
X
documents. With TeX4ht, you manipulate the *.dvi file rather than
the *.tex file. It requires multiple passes to make everything come out
right, and often post-processing of loose-end tasks is required. Generally
speaking, it is troublesome to handle graphics with TeX4ht.
The utility WebEQ is a popular Java applet for rendering mathematics
in a browser. It includes tools for dispalying math. With WebEQ, you
store documents in the Mathematical Markup Language MathML.
The language WebTeX is not quite T
E
X and not quite L
A
T
E
X. In fact, it
is a version of these languages adapted for the Internet. It retains many
features of T
E
X and L
A
T
E
X but discards others. If you are going to use
WebTeX, then you will have to do some retooling.
The book [GOR] gives detailed tutorials in the uses, attributes, and
limitations of all the utilities described here. It treats several others as
well, and furthermore describes the differences and relationships among
the markup languages HTML, SGML, XML, and MathML. We recom-
mend it highly to the stalwart reader who really wants to understand
how all these utilities work.
7.5 Software to Go with your Book or Article; Web
Sites
These days, it is becoming more and more common to have software
that is connected in some way to a hard-copy book or article. Some of
the ways that this connection is implemented are as follows:
7
Here MathML is a markup language to be used with XML, and XML is an extensible
markup language that improves on HTML in many ways.
©2001 CRC Press LLC
• The author has a number of specialized references, or arcane and
lengthy examples, or other ancillary material that would not be
attractive to include in the hard-copy material. He instead puts
them on a publicly accessible Web site. The URL is provided with
the hard-copy material.
• The author has written a textbook, and may want to put worked
problem solutions, or additional examples, or review material, or
color figures, or animated graphics, on a Web site.
• The author has created executable software, either in C, or Fortran,
or C++, or Java, or some other standard programming language
(for which most readers will have a compiler, or at least access to
one), that he wants the reader to be able to download and use.
He puts the *.exe file, or perhaps the source code, together with
suitable documentation, on a publicly accessible Web site.
In any of these instances, the author may wish to place the T
E
X code,
or perhaps the *.dvi or *.ps files for the entire text of the document,
on the Web site as well. There may, however, be copyright or legal
or fiduciary considerations that need to be worked out before doing so.
Consult with your publisher before you put your entire book or MS on
the Web.
This is not the place to describe in detail just how to put software
on a Web site. Suffice it to say that you will put it somewhere in your
public_html directory, and you will create a button on your home page
that will lead users to it. Along the tree that leads to your software,
you will offer options including (i) documentation, (ii) examples, (iii)
additional information, (iv) FAQs (Frequently Asked Questions), (v)
other references and Web sites, and (vi) hyperlinks to related material.
You will be doing your readers a great favor by making your Web site
as user-friendly as possible. Many Web sites consist of little more than a
list of hyperlinks, each of which leads to a download. It is up to the user
to figure out what is what, and how to download. With a little extra
effort, you can give your site an attractive and readily understandable
graphic interface and you can also provide some text that will lead the
user to the particular choice that will satisfy his needs.
If you are writing a book, you may wish to include a diskette with
the book. (For example, [SAK] has such a diskette; it contains T
E
X
macros that are discussed in the book.) In that instance, you must
decide whether to render the files in Windows format, or Macintosh
format, or some other format. Whatever choice you make will of course
exclude some readers. The really classy thing to do is to put your files
on a CD-ROM, which has plenty of storage space, and then provide the
files in several formats.
©2001 CRC Press LLC
It is a nice touch to include a few pages in the book which explain
how to use the provided disc or how to access the Web site, as the case
may be. Many people are still quite naive and inexperienced at using a
computer. If you provide them with nothing but a disc then they may
just assume that it is a self-booting disc that will run all by itself. Likely
as not, that is untrue; there may be a readme file or an executable or a
batch file. The experienced hacker will know just what to do, but the
neophyte will not. Providing a little help is your job. If the reader cannot
figure out what to do with your disc then he will become frustrated and
may set your book aside. That is not the effect that you want. After all
the hours that you will invest making the disc, invest a few more making
it easy for people to use the disc.
A final note: Sometimes it is attractive to include written-out com-
puter code in the book you are writing. If it is a book about dynamical
systems, then you may want to include the code for generating the Man-
delbrot set. If it is a book on numerical methods, then you may want
to include the code for Euler’s method or for Runge-Kutta. It is now
quite standard to use the “typewriter-like font” for computer code. For
instance, here is a fragment of code that this author wrote (many years
ago) for a utility for the pharmaceutical industry to calculate volumes
and surface areas of pills (tablets):
1573 FOR J = 0 TO 16:PRESET (550+J,35):NEXT J
1577 LINE (383,44)-(398,44)
1580 INPUT "THICKNESS OF BELLY BAND (FIG DIM BB)";BB
1590 IF BB < 0 THEN GOTO 2130
1593 FOR J = 0 TO 16:PRESET (383+J,44):NEXT J
1597 LINE (394,12)- (401,12)
1600 INPUT "LAND (FIG DIM L) ";INC
1610 IF INC < 0 THEN GOTO 2130
1615 FOR J=0 TO 8:PRESET (394+J,12):NEXT J
1617 LINE (428,12)-(458,12)
1618 INPUT "MINOR RADIUS (FIG DIM MINR) ";MINR
1620 IF MINR < CD THEN GOTO 2130
1623 IF 2*MINR*CD - CD^2 < 0 THEN GOTO 2130
1624 G = SQR(ABS(2*MINR*CD - CD^2))
1625 FOR J = 0 TO 31:PRESET (428+J,12):NEXT J
1635 S2 = (PI*(MI^2)*.25 + (MA - MI)*MI)*BB
The more fundamental question is what syntax to use. The code that
we just exhibited is Turbo BASIC, which most people would consider to
be Neantderthal. Should your code instead be in C, or in Fortran, or in
©2001 CRC Press LLC
APL, or should it be in pseudocode?
8
Speaking from my own experience,
I find true code, coming from a specific language, to be much clearer
than pseudocode. Since pseudocode does not have a generally accepted
syntax, I often do not know what the writer’s pseudocode means. In
any event, C is a nearly universal scientific programming language these
days. So is Java. I would recommend that you use one of these. In
my own work, I have found that attempting to understand an author’s
provided language-specific code gave me just the motivation I needed to
learn more about C. Perhaps it will work in the same way for you.
8
Here the jargon “pseudocode” denotes a generic, made-up code that comes from no
particular language. It will tell, by paradigm, the well-informed reader how to write
code in the language of his choice.
©2001 CRC Press LLC
©2001 CRC Press LLC
©2001 CRC Press LLC
©2001 CRC Press LLC
©2001 CRC Press LLC
9
We have used marks from our tables as well as some new marks that should be clear
©2001 CRC Press LLC
©2001 CRC Press LLC
Appendix III: Specialized Mathematics
Symbols
We present here, in table form, the standard mathematics symbols
that make up most mathematical manuscripts that you will see. These
are typeset using the computer typesetting system T
E
X. We begin with
the Greek alphabet.
Lower-Case Greek Letters
α \alpha λ \lambda φ \phi
β \beta j \mu χ \chi
γ \gamma ν \nu ψ \psi
δ \delta ξ \xi ω \omega
c \epsilon o o ε \varepsilon
ζ \zeta π \pi ϑ \vartheta
η \eta ρ \rho c \varpi
θ \theta σ \sigma ¸ \varrho
ι \iota τ \tau ς \varsigma
κ \kappa υ \upsilon ϕ \varphi
Observe that the Greek letter “omicron” is just the same as the lower-
case roman “o”. The characters ε, ϑ, c, ¸, ς, ϕ are, in effect, script ver-
sions of the lower-case Greek letters c, θ, π, ρ, σ, φ.
Capital Greek Letters
¹ A Λ \Lambda Φ \Phi
1 B ` M A X
Γ \Gamma · N Ψ \Psi
∆ \Delta Ξ \Xi Ω \Omega
1 E O O
7 Z Π \Pi
H H 1 P
Θ \Theta Σ \Sigma
1 I T T
1 K Υ \Upsilon
©2001 CRC Press LLC
One can typeset slanted upper-case Greek letters by using the com-
mand \mit. For example, the code $\mit\Phi=\Psi(\Delta,\Gamma)$
will compile to Φ = Ψ(∆, Γ). Boldface capital Greek may be typeset in
an obvious way: the code $\bf \Gamma + \Delta$ typesets to Γ +∆.
The characters
/B ( Tc T ( H1 ¸ /L/^ OT Q¹o T | 1 JA } Z
are typeset by using the command \cal inside the T
E
X math environ-
ment to invoke the calligraphic alphabet.
A standard /
/
o-T
E
X implementation will contain certain commonly
used fonts for mathematicians. These include Euler fraktur, which is
invoked by the font call \frak and has the form Euler fraktur, and
blackboard bold, which is invoked by the font call \Bbb and has the
form B'ACKB´A1| B´'|.
It is convenient, from the point of view of T
E
X, to group characters
according to their “spacing type”. With that paradigm in mind, here
are the characters of class Ord:
Characters of Class Ord
ℵ \aleph / \prime \neg
¯/ \hbar ∅ \emptyset | \|
ı \imath ∇ \nabla ¸ \backslash
, \jmath

\surd . \flat
/ \ell · \top ; \natural
℘ \wp ⊥ \bot ; \sharp
' \Re

\angle ♣ \clubsuit
· \Im ´ \triangle ♦ \diamondsuit
∂ \partial ∀ \forall ♥ \heartsuit
∞ \infty ∃ \exists ♠ \spadesuit
Large Operators of Class Op

\coprod

\bigcup

\bigoplus
_
\bigvee
_
\int

\bigodot
_
\bigwedge

\prod
_
\oint

\biguplus

\sum

\bigsqcup

\bigcap

\bigotimes ∫ \smallint
©2001 CRC Press LLC
Symbols for Footnotes
∗ * † \dag ‡ \ddag
¸ \S | \| ¶ \P
Binary Operations
\times ∩ \cap \triangleleft
÷ \div ∪ \cup > \triangleright
∓ \mp ¯ \sqcap ´ \bigtriangleup
± \pm . \sqcup _ \bigtriangledown
∧ \wedge ¬ \uplus ◦ \circ
∨ \vee • \bullet _ \bigcirc
‡ \ddagger ¸ \odot ¸ \setminus
† \dagger ¸ \oslash \cdot
H \amalg ⊗ \otimes ∗ \ast
\diamond ¸ \ominus × \star
/ \wr ⊕ \oplus
Relations
∝ \propto ≈ \approx → \to
_ \sqsubseteq ∼ \sim ← \leftarrow
_ \sqsupseteq · \simeq ↑ \uparrow
| \parallel

= \cong ↓ \downarrow
[ \mid _ \succeq ⊥ \perp
¬ \dashv _ \preceq ≡ \equiv
¬ \vdash ⊃ \supset · \asymp
,= \ne ⊂ \subset
.
= \doteq
, ∈ \notin ⊇ \supseteq \smile
, \not ⊆ \subseteq · \frown
≤ \le ∈ \in > \bowtie
≥ \ge ÷ \ni [= \models
~ \succ ¸ \gg
÷
÷ \rightleftharpoons
≺ \prec ¸ \ll ⇐⇒ \iff
©2001 CRC Press LLC
Punctuation Marks and Ellipses
. \ldotp
.
.
. \vdots
.
.
.
\ddots
\cdotp . . . \ldots \cdots
: \colon . . . \dots
Arrows
¸ \nearrow ⇔ \Leftrightarrow ←→ \longleftrightarrow
¸ \searrow ↔ \leftrightarrow ⇐⇒ \Longleftrightarrow
¸ \nwarrow =⇒ \Longrightarrow → \hookrightarrow
¸ \swarrow −→ \longrightarrow ← \hookleftarrow
→ \mapsto ←− \longleftarrow ÷ \leftharpoonup
⇐ \Leftarrow ⇐= \Longleftarrow ÷ \leftharpoondown
⇒ \Rightarrow −→ \longmapsto ÷ \rightharpoonup
⇐⇒ \iff → \to ÷ \rightharpoondown
← \leftarrow ↑ \uparrow ↓ \downarrow
Delimiters
( ( ¸ \langle ↑ \uparrow
[ [ ¸ \backslash ↓ \downarrow
¦ \{ ¸ \lfloor ⇑ \Uparrow
[ | ¸ \lceil ⇓ \Downarrow
| \|
_
_
\lgroup ¸ \Updownarrow
, /
_
_
\lmoustache
Next is a display of the two most standard AMS-T
E
X symbol fonts:
AMS Symbol Font MSAM10
0 1 2 3 4 5 6 7 8 9 A B C D E F
0. ¯ ¬ ¯ ¯ B . ♦ # ¸ ¸ = = ¬ ' ' = 0.
10. ÷ ÷ ⇔ ⇒ | _ ` ¸ , ¸ ÷ ÷ ¯ · ¨ ¨ 10.
20. ~ - ÷ + = _ _ _ ÷ ∴ ∵ = = _ _ _ 20.
30. ¸ ¸ - ` - _ ¸ ≶ \ - = = , _ ¸ ≷ 30.
40. ¸ ¸ < _ _ ± ¸ * > ¬ - - ¬ # V 40.
50. = _ ¸ _ _ Y = ÷ . Y ¯ , ∠ X < ∝ 50.
60. · · ¸ ¸ J + · ` · ¸ ¸ = · ≪ ≫ 60.
70. ' ¯ ¡ ¸ . ÷ ~ - , ¸ ± U ¡ ¸ ¸ ¸ 70.
©2001 CRC Press LLC
AMS Symbol Font MSBM10
0 1 2 3 4 5 6 7 8 9 A B C D E F
0. _ _ _ _ ≮ ≯ ⊀ , _ _ ¸ ¸ _ _ _ _ 0.
10. _ _ _ _ _ _ _ _ _ _ _ _ ~ æ , ¸ 10.
20. _ _ ¿ ¸ _ _ ¸ ¿ _ _ _ _ ∦ [ · + 20.
30. - ' = ÷ _ _ ; ; ÷ ÷ = = = ÷ ÷ ∅ 30.
40. ± A B C | E F G H " , K ' , N ´ 40.
50. P ¸ 1 S ¯ U V W X Y Z ¯ 50.
60. · , G ð - Q ג ¯ < 60.
70. . + ∼ ≈ ¸ _ _ . } κ k h / ~ 70.
We conclude with the commands for the math abbreviations (or func-
tion names) included in T
E
X.
Math Abbreviations
\arccos \cos \csc \exp \ker \limsup \min \sinh
\arcsin \cosh \deg \gcd \lg \ln \Pr \sup
\arctan \cot \det \hom \lim \log \sec \tan
\arg \coth \dim \inf \liminf \max \sin \tanh
©2001 CRC Press LLC
Appendix IV: Standard Alphabets
We now exhibit some of the most standard fonts that are used in mod-
ern typesetting. These are rendered using T
E
X.
Upper Case Alphabets
roman boldface italic slanted roman calligraphic fraktur
A A A A / A
B B B B B B
C C C C ( C
D D D D T D
E E E E c E
F F F F T F
G G G G ( G
H H H H H H
I I I I 1 I
J J J J ¸ J
K K K K / K
L L L L L L
M M M M / M
N N N N ^ N
O O O O O O
P P P P T P
Q Q Q Q Q Q
R R R R ¹ R
S S S S o S
T T T T T T
U U U U | U
V V V V 1 V
W W W W J W
X X X X A X
Y Y Y Y } Y
Z Z Z Z Z Z
©2001 CRC Press LLC
Lower Case Alphabets
roman boldface italic slanted roman calligraphic fraktur
a a a a a
b b b b b
c c c c c
d d d d d
e e e e e
f f f f f
g g g g g
h h h h h
i i i i i
j j j j j
k k k k k
l l l l l
m m m m m
n n n n n
o o o o o
p p p p p
q q q q q
r r r r r
s s s s s
t t t t t
u u u u u
v v v v v
w w w w w
x x x x x
y y y y y
z z z z z
Observe that there are no lower case calligraphic letters in T
E
X.
©2001 CRC Press LLC
Appendix V: Alternative Mathematical
Notations
Here we present various typesetting situations together with some
choices in how to address them.
Option 1 Option 2 Remarks
lim, lim limsup , liminf Option 2 more common
c
x
3
−y
2
xy+2
exp
_
(r
3
−j
2
),(rj + 2)
_
Option 2 better in text

r
n
+ j
m
(r
n
+ j
m
)
1/2
Option 1 better in text
x+2y
3
,
4
3
(r + 2j),3 , 4,3 Option 2 better in text

n
j=1
,

k
=0
n

j=1
,
k

=0
Option 1 better in text
Option 2 better for display
o
j
1
,...,j
m
o(,
1
, ,
2
, . . . , ,
m
) Option 2 easier to read
o
k
1
,...,k
p
j
1
,...,j
m
o(,
1
, . . . , ,
m
; /
1
, . . . , /
p
) Option 2 easier to read
_
e
f
_
C
e,f
Option 2 better in text
C
f
→1 ) : C →1 Option 2 easier to read
A ∪ Y Cl(A ∪ Y ) Option 2 better in text
· v Option 2 better in text
˙
¹,
¨
¹, ˜ n,
´
) , ¯ r ¹

, ¹

, n

, T()) , r
#
Option 2 better in text
log
2
y
5
_
c −
b
z
log(2,j)
[c −/,.]
1/5
Option 2 easier to read
)
_
r j
. n
_
)(¹) , ¹ =
_
r j
. n
_
Option 2 preferable
©2001 CRC Press LLC
Appendix VI: T
E
X, PostScript, Acrobat,
and Related Internet Sites
There are a number of Internet sites that archive T
E
X fonts, macros,
and other utilities that will be of interest to the T
E
X user. We record
some of the ftp addresses and Web sites here.
Always bear in mind that Web sites are transient. They are con-
stantly in flux, they change form, they move, and (sadly) sometimes
they disappear. As of this writing, all the given sites are operational.
As an alternative to logging on to one of the given sites, you can use
one of the utilities archie, gopher, or wais to search for the package
you are looking for. Or else use the altavista, or yahoo, or google
Web search engines to conduct a search.
Sites from which to ftp T
E
X Packages
ftp.math.utah.edu /pub/tex/tex-index
This is Nelson Beebe’s server, and it is particularly strong in
information about BibT
E
X. It also contains a concordance
of all articles in TUGboat, the publication of the T
E
X Users
Group.
labrea.stanford.edu
This is the “official” Stanford University repository for T
E
X,
METAFONT, dvips, and related files.
http://www.tug.org
This is the site maintained by the T
E
X Users Group. It is
a source for emTeX, publicTeX, publicMF, and many other
fonts and macros and T
E
X utilities.
e-math.ams.org
This is the ftp site of the American Mathematical Society
(AMS). The login and password are both e-math. The Soci-
ety is the creator and promulgator of /
/
o-T
E
X, and the site
makes both /
/
o-T
E
X and /
/
o-L
A
T
E
X available for down-
loading. Once you are logged on, use your <Up-<Down
keys to highlight <Sitemap and then press <Enter. Again
use your <Up-<Down keys to scroll to <TeX Resources
and again press <Enter. Now you are in the AMS’s world
of T
E
X. There are numerous fonts, macros, style files, author
©2001 CRC Press LLC
packages, manuals and other T
E
X resources available at the
e-math site.
ftp.dante.de /pub/tex/help/TeX-index
is a source of the “Mainz” packages of Mittelbach and Sch¨opf
(these resources include multicol, verbatim, theorem, NFSS,
ftnright, and array). It is a mirror site of ftp.tex.ac.uk,
and has the Web address http://www.dante.de.
ftp.tex.ac.uk /pub/archive/help/TeX-index
This is a CTAN site (a mirror of ftp.dante.de), and also
the home of the UKTeX newsletter. The corresponding Web
address is http://www.tex.ac.uk.
Other URLs for T
E
X Goodies
http://199.26.180.160/winnt/misc/page2.html
An implementation of BCT
E
X may be downloaded from this
site.
ftp://ctan.tug.org/tex-archive/systems
/win32/miktex/1.20/
MikT
E
X may be downloaded from this site.
http://www.esm.psu.edu/mac-tex/versions.html
This is a T
E
X Web site for Macintosh users.
http://www.tug.org
This is the Web site of the T
E
X Users Group. It has pointers
to several different CTAN sites.
http://www.cs.wisc.edu/~ghost/
This site is a source for the utility ghostscript, for manip-
ulating PostScript.
http://www.adobe.com
Adobe’s home page. Adobe is the the company that created
Acrobat and PostScript.
http://www.math.uakron.edu/~dpstory/acrotex.html
©2001 CRC Press LLC
A repository for PDF-based math tutorials. Makes great and
extensive use of Acrobat forms.
http://www.tug.org/applications/pdftex/
A resource of pdfT
E
X examples.
http://www.pdfzone.com/
The PDF Zone.
http://www.adobe.com/prodindex/postscript/main.html
Information about PostScript at Adobe.
ftp://ctan.tug.org/tex-archive//support/latex2html/sources/
Sources for L
A
T
E
X2HTML on the CTAN site.
http://www.5-d.com/niknak.htm
This is a commercial PostScript

-to-PDF converter.
http://www.wargaming.net/Programming
Adobe PostScript resources.
http://math.berkeley.edu/~vojta/xdvi.html
This is Paul Vojta’s (he is a math professor at U. C. Berke-
ley) X-Windows T
E
X previewer.
http://www.yandy.com/
The home page of Y&Y T
E
X.
http://www.cs.wisc.edu/~ghost/
Source for Ghostview, Ghostscript, GSview.
http://godel.ph.utexas.edu/Members/timg/gs/gs.html/
Source for Atari Ghostscript.
————————————————————–
For those not entirely conversant with the process of conducting an
ftp transaction, we now provide a little guidance. On a UNIX system,
log on and get to a system prompt. On a Macintosh machine, go to the
ftp icon. Instructions for America Online are given below.
©2001 CRC Press LLC
To go to the first ftp site given in this Appendix, ftp now to the ftp
site ftp.math.utah.edu. (On a UNIX system, this step is performed
by typing
ftp ftp.math.utah.edu
at your system prompt. On other systems, or with a browser, this will be
done by typing ftp.math.utah.edu in a suitable window and clicking on
<Go or a similar icon.) Perform the logon by entering “anonymous”
and using your e-mail address as a password. Type cd pub followed by
<Enter. Then type cd tex followed by <Enter. Now you will be
at Nelson Beebe’s famous T
E
X site. Enjoy!
Of course you can always use your Web browser to ftp. With America
Online, click on <Keyword, type ftp in the keyword window, and then
click on <Go. With Netscape, type ftp://ftp.math.utah.edu in the
URL search window and press <Enter. Other Web browsers have
similar protocols.
©2001 CRC Press LLC
App endix VI I: Basic T
E
X Commands
We provide here a concordance of the most basic commands in T
E
X.
This list will supplement, and in some instances duplicate, what appears
in the tables in Appendix III. While those tables emphasized mathemat-
ics symbols, the present list also contains formatting and text commands.
Our wish is that this will serve as a quick-and-dirty reference for the
beginner, and also a quick review for the more experienced T
E
X user.
Command Meaning or Purpose
“\ ” hard single space
\, thin space
\! negative thin space
\’ acute accent (or egout) in text (e.g., ´ a)
\‘ accent grave in text (e.g., ` o)
\~ tilde in text (e.g., ˜e)
\" dieresis or umlaut in text (e.g., ¨ a)
\^ circumflex in text (e.g.,
ˆ
f)
\. dot accent in text (e.g., ˙ o)
\= macron in text (e.g., ¯e)
!‘ ¡
?‘ ¿
$ $ inline math mode
\( \) inline math mode
$$ $$ display math mode
\[ \] display math mode
\b bar-under accent in text (e.g., e
¯
)
\c cedilla (e.g., ¸ c)
\d dot-under accent in text (e.g., o
.
)
\H long Hungarian umlaut in text (e.g., ˝ o)
\t tie-after accent in text (e.g., oo)
\u breve in text (e.g., ˘ o)
\v h´ aˇcek or check in text (e.g., ˇa)
\AA
˚
A
\aa ˚a
A^b superscript ¹
c
A_c subscript ¹
c
\AE Æ
\angle angle (

)
\acute acute accent in math mode (e.g., ´ o)
\AmSTeX logo for /
/
o-T
E
X
\AmSLaTeX logo for /
/
o-L
A
T
E
X
\approx approximately equal to (≈)
\backslash backslash (¸)
©2001 CRC Press LLC
\bar macron in math mode (e.g., ¯ n)
\Bbb blackboard bold font (C, N, 1)
\bf boldface font (a, b, c, d)
\biggl left delimiter enlargement size
_
\Biggl left delimiter enlargement size
_
\biggm middle delimiter enlargement size
¸
¸
¸
¸
\Biggm middle delimiter enlargement size
¸
¸
¸
¸
¸
\biggr right delimiter enlargement size
_
\Biggr right delimiter enlargement size
_
\bigl left delimiter enlargement size
_
\Bigl left delimiter enlargement size
_
\bigm middle delimiter enlargement size
¸
¸
\Bigm middle delimiter enlargement size
¸
¸
¸
\bigr right delimiter enlargement size
_
\Bigr right delimiter enlargement size
_
\bigtriangleup large triangle, like Laplacian (´)
\bigskip large vertical space
\Box the “Halmos tombstone” ( )
\breve breve in math mode (e.g., ˘ o)
\bullet typesetter’s bullet (•)
\bye command for ending a T
E
X document
\cal the math mode calligraphic font (/, B, ()
\cap intersection sign (∩)
\cases formats a function defined by cases
\cdot small, centered dot ( )
\cdots three small, centered dots (an ellipsis )
\centerline command for centering a line of text
\check a “check” accent in math mode (
ˇ
))
\choose binomial expression (
_
a
b
_
)
\circ small, centered circle () ◦ p)
\cos cosine
\cup union sign (∪)
©2001 CRC Press LLC
\dagger typesetter’s dagger (†)
\ddagger typesetter’s double dagger (‡)
\ddot double dot math accent or dieresis (¨ r)
\def macro definer
\div division sign (÷)
\dot dot accent in math mode (e.g., ˙ o)
\dots three small, baseline dots (an ellipsis . . . )
\ell script, lower-case script ell (/)
\emptyset empty set (∅)
\eqalign utility for aligned displays
\eqalignno \eqalign with numbering
\eqno equation label at right margin
\equiv is defined to be (≡)
\exists there exists (∃)
\folio page number, as in a footline
\footline command for creating a running foot
\footnote creates a footnote
\forall for all (∀)
\frak the Euler fraktur font (f , p, /)
\ge or \geq greater than or equal to (≥)
\gg double greater than (¸)
\grave accent grave in math mode (e.g., ` o)
\hangindent command for creating hangindents
\hat a caret “hat” accent (
ˆ
))
\hbox an environment for placing text into math
\headline command for creating a running head
\hfil command for filling out a line with space
\hfill command for filling out a line with space
\hrule a solid black box ( )
\hskip horizontal skip
\int , \int_a^b integral (
_
b
a
)
\Im imaginary part of (·)
\imath the dotless “i”, for use in mathematics
\in is an element of (∈)
\indent paragraph indentation
\infty infinity (∞)
\input command for including macros
\it italic font (a, b, c, d)
\item delineates an element of a list
\itemitem delineates a sub-element of a list
\itemitemitem delineates a sub-sub-element of a list
\jmath the dotless “j”, for use in mathematics
\jot vertical space
\kern horizontal re-positioning
\l l
©2001 CRC Press LLC
\langle left inner product delimiter (¸)
\LaTeX logo for L
A
T
E
X
\le or \leq less than or equal to (≤)
\left controls size of left delimiter
\leftarrow leftward pointing arrow (←)
\Leftarrow leftward pointing double arrow (⇐)
\leqno equation number at left margin
\ll double less than (¸)
\llap command for hang indenting
\magnification controls size of font, spacing in document
\magstep specifies magnification size
\mapsto maps to (r →j)
\mathop defines a new “math operator” expression
\mathstrut an invisible box used for spacing
\matrix matrix
\medskip medium vertical space
\mid a vertical bar for use in setbuilder notation ([)
\narrower formats text with wider margins for display
\ne or \neq not equal to (,=)
\nabla nabla (∇)
\nearrow arrow point northeast (¸)
\nwarrow arrow point southwest (¸)
\ni has as an element (÷)
\noindent prevents indentation
\nopagenumbers creates a document with no page numbers
\not puts a slash ( ,) through the next symbol
\O Ø
\o ø
\OE Œ
\oe œ
\oint complex line integral (
_
)
\ominus a minus sign in a circle (¸)
\oplus a plus sign in a circle (⊕)
\otimes a times sign in a circle (⊗)
\over used in making fractions
\overbrace a brace over text (
¸ .. ¸
r + j)
\overline a line over text (r + j)
\parindent paragraph indentation
\partial a partial differentiation sign (∂))
\perp perpendicular sign (⊥)
\pm plus or minus (±)
\pmatrix matrix with parentheses
\qquad a double em quad
\quad an em quad
©2001 CRC Press LLC
\rangle right inner product delimiter ())
\Re real part of (')
\right controls size of right delimiter
\rightarrow rightward pointing arrow (→)
\Rightarrow rightward pointing double arrow (⇒)
\rm roman font
\sc big cap-small cap font
\searrow arrow point southeast (¸)
\setminus sign for set-theoretic difference (¹¸ 1)
\sim similar to (o ∼ /)
\simeq variant of \sim (o · /)
\sqrt square root (

r + ¹)
\ss ß
\supset is a (proper) superset of (¹ ⊃ 1)
\supseteq is a superset of (¹ ⊇ 1)
\swarrow arrow point southwest (¸)
\sin sine
\sl slanted roman font
\smallskip small vertical space
\subset is a (proper) subset of (⊂)
\subseteq subset of (⊆)
\tan tangent
\TeX macro for TeX (T
E
X)
\tilde tilde (
˜
¹)
\times a times or multiplication sign ()
\tt typewriter-like font
\underbrace a brace under text (r + j
. ¸¸ .
)
\underline a line under text (o + /)
\vbox a typesetting device for isolating text
\vec the vector accent (o)
\vee a vee (o ∨ T)
\vrule a solid black box ( )
\vskip vertical skip
\wedge a wedge, as in an exterior product (· ∧ n)
\widehat wide caret or “hat” in math mode (e.g.,
´
))
\widetilde wide tilde in math mode (e.g.,
¯
/)
©2001 CRC Press LLC
Appendix VIII: A Sample of L
A
T
E
X
We now present a sample of L
A
T
E
X code and then show its compiled
output. You will be able to do a fair amount of L
A
T
E
X typesetting just
by imitating what you see here. For convenience, we present the code
both in L
A
T
E
X 2.09 form and L
A
T
E
X2
ε
form.
©2001 CRC Press LLC
\documentstyle{article}
\newfam\msbfam
\font\tenmsb=msbm10 \textfont\msbfam=\tenmsb
\font\sevenmsb=msbm7 \scriptfont\msbfam=\sevenmsb
\font\fivemsb=msbm5 \scriptscriptfont\msbfam=\fivemsb
\def\Bbb{\fam\msbfam \tenmsb}
\def\RR{{\Bbb R}}
\def\HollowBox #1#2{{\dimen0=#1 \advance\dimen0 by -#2
\dimen1=#1 \advance\dimen1 by #2
\vrule height #1 depth #2 width #2
\vrule height 0pt depth #2 width #1
\llap{\vrule height #1 depth -\dimen0 width \dimen1}%
\hskip -#2
\vrule height #1 depth #2 width #2}}
\def\BoxOpTwo{\mathord{\HollowBox{6pt}{.4pt}}\;}
\begin{document}
\begin{center}
\large \bf The Fundamental Theorem of Calculus
\end{center}
Let $f$ be a continuous function on an open interval
$I \subseteq \RR$. Fix a point $a \in I$.
For any point $x \in I$, we define
$$
F(x) = \int_a^x f(t) \, dt . \eqno (\dagger)
$$
The substance of the Fundamental Theorem of Calculus is
to claim that $F$ is an anti-derivative for $f$.
More precisely, we have
\smallskip \\
\noindent {\sc Theorem 1:} The function $F$ defined above
is differentiable, and
$$
\frac{d}{dx} F(x) = f(x)
$$
for every $x \in I$.
©2001 CRC Press LLC
\smallskip \\
\noindent {\bf Proof:} We endeavor to calculate the
derivative of $F$ by forming the difference or Newton
quotient for $h \ne 0$:
\setcounter{chapter}{0}
\begin{eqnarray}
\underbrace{\frac{F(x + h) - F(x)}{h}}_{\hbox{Newton quotient}}
& = & \frac{\int_a^{x+h} f(t) \, dt
- \int_a^x f(t) \, dt}{h} \nonumber \\
& = & \frac{\int_x^{x+h} f(t) \, dt}{h} .
\end{eqnarray}
Now fix a point $x_0 \in I$. Let $\epsilon > 0$. Choose
$\delta > 0$ such that $|t - x_0| < \delta$ implies that
$|f(t) - f(x_0)| < \epsilon$. Now we may rewrite (1) as
\begin{eqnarray*}
\frac{\int_x^{x+h} f(t) \, dt}{h} & = &
\frac{\int_x^{x+h} f(x) \, dt}{h}
+ \frac{\int_x^{x+h} [f(t) - f(x)] \, dt}{h} \\
& = & f(x) + \frac{\int_x^{x+h} [f(t) - f(x)] \, dt}{h} . \\
\end{eqnarray*}
If $|h| < \delta$ then we may estimate the last fraction as
$$
\left | \frac{\int_x^{x+h} [f(t) - f(x)] \, dt}{h} \right | \leq
\frac{\int_x^{x+h} |f(t) - f(x)| \, dt}{h} \leq \epsilon .
$$
Thus, in summary, we have
$$
\frac{F(x + h) - F(x)}{h} = f(x) + \hbox{error} ,
$$
where the error is not greater than $\epsilon$. In conclusion,
$$
\lim_{h \rightarrow 0} {F(x+h) - f(x)}{h} = f(x) . \eqno \BoxOpTwo
$$
\end{document}
©2001 CRC Press LLC
\documentclass{article}
\newfam\msbfam
\font\tenmsb=msbm10 \textfont\msbfam=\tenmsb
\font\sevenmsb=msbm7 \scriptfont\msbfam=\sevenmsb
\font\fivemsb=msbm5 \scriptscriptfont\msbfam=\fivemsb
\def\Bbb{\fam\msbfam \tenmsb}
\def\RR{{\Bbb R}}
\def\HollowBox #1#2{{\dimen0=#1 \advance\dimen0 by -#2
\dimen1=#1 \advance\dimen1 by #2
\vrule height #1 depth #2 width #2
\vrule height 0pt depth #2 width #1
\llap{\vrule height #1 depth -\dimen0 width \dimen1}%
\hskip -#2
\vrule height #1 depth #2 width #2}}
\def\BoxOpTwo{\mathord{\HollowBox{6pt}{.4pt}}\;}
\begin{document}
\begin{center}
\textbf{\large The Fundamental Theorem of Calculus}
\end{center}
Let $f$ be a continuous function on an open interval
$I \subseteq \RR$. Fix a point $a \in I$.
For any point $x \in I$, we define
$$
F(x) = \int_a^x f(t) \, dt . \eqno (\dagger)
$$
The substance of the Fundamental Theorem of Calculus is
to claim that $F$ is an anti-derivative for $f$.
More precisely, we have
\smallskip \\
\noindent \textsc{Theorem 1:} The function $F$ defined above
is differentiable, and
$$
\frac{d}{dx} F(x) = f(x)
$$
©2001 CRC Press LLC
for every $x \in I$.
\smallskip \\
\noindent \textbf{Proof:} We endeavor to calculate the
derivative of $F$ by forming the difference or Newton
quotient for $h \ne 0$:
\setcounter{chapter}{0}
\begin{eqnarray}
\underbrace{\frac{F(x + h) - F(x)}{h}}_{\hbox{Newton quotient}}
& = & \frac{\int_a^{x+h} f(t) \, dt
- \int_a^x f(t) \, dt}{h} \nonumber \\
& = & \frac{\int_x^{x+h} f(t) \, dt}{h} .
\end{eqnarray}
Now fix a point $x_0 \in I$. Let $\epsilon > 0$. Choose
$\delta > 0$ such that $|t - x_0| < \delta$ implies that
$|f(t) - f(x_0)| < \epsilon$. Now we may rewrite (1) as
\begin{eqnarray*}
\frac{\int_x^{x+h} f(t) \, dt}{h} & = &
\frac{\int_x^{x+h} f(x) \, dt}{h}
+ \frac{\int_x^{x+h} [f(t) - f(x)] \, dt}{h} \\
& = & f(x) + \frac{\int_x^{x+h} [f(t) - f(x)] \, dt}{h} . \\
\end{eqnarray*}
If $|h| < \delta$ then we may estimate the last fraction as
$$
\left | \frac{\int_x^{x+h} [f(t) - f(x)] \, dt}{h} \right | \leq
\frac{\int_x^{x+h} |f(t) - f(x)| \, dt}{h} \leq \epsilon .
$$
Thus, in summary, we have
$$
\frac{F(x + h) - F(x)}{h} = f(x) + \hbox{error} ,
$$
where the error is not greater than $\epsilon$. In conclusion,
$$
\lim_{h \rightarrow 0} {F(x+h) - f(x)}{h} = f(x) .
\eqno \BoxOpTwo
$$
\end{document}
©2001 CRC Press LLC
Let ) be a continuous function on an open interval 1 ⊆ 1. Fix a point
o ∈ 1. For any point r ∈ 1, we define
1(r) =
_
x
a
)(t) dt. (†)
The substance of the Fundamental Theorem of Calculus is to claim that
1 is an anti-derivative for ). More precisely, we have
Theorem 1: The function 1 defined above is differentiable, and
d
dr
1(r) = )(r)
for every r ∈ 1.
Proof: We endeavor to calculate the derivative of 1 by forming the
difference or Newton quotient for / ,= 0:
1(r + /) −1(r)
/
. ¸¸ .
Newton quotient
=
_
x+h
a
)(t) dt −
_
x
a
)(t) dt
/
=
_
x+h
x
)(t) dt
/
. (1)
Now fix a point r
0
∈ 1. Let c 0. Choose δ 0 such that [t −r
0
[ < δ
implies that [)(t) −)(r
0
)[ < c. Now we may rewrite (1) as
_
x+h
x
)(t) dt
/
=
_
x+h
x
)(r) dt
/
+
_
x+h
x
[)(t) −)(r)] dt
/
= )(r) +
_
x+h
x
[)(t) −)(r)] dt
/
.
If [/[ < δ then we may estimate the last fraction as
¸
¸
¸
¸
¸
_
x+h
x
[)(t) −)(r)] dt
/
¸
¸
¸
¸
¸

_
x+h
x
[)(t) −)(r)[ dt
/
≤ c.
©2001 CRC Press LLC
Thus, in summary, we have
1(r + /) −1(r)
/
= )(r) + error,
where the error is not greater than c. In conclusion,
lim
h→0
1(r + /) −)(r)/ = )(r).
©2001 CRC Press LLC
Glossary
The definitions included here have been drawn from a variety of books
and articles listed in the References, but especially from the Web page
of Budgett and Johnstone.
A4 paper size The international standard size for business paper:
297mm 210mm. See international paper sizes and ISO.
AA’s (author’s alterations) Changes to the copy made by author
after the manuscript has been typeset. A publisher may charge the
author for AA’s that exceed 15%.
accent A mark on a character, usually a letter, indicating stress.
Common accents are named acute (´e), grave (`e), circumflex (ˆa), cedilla
(¸ c), tilde (˜ n).
accent grave The accent (used in French, Spanish, and other Euro-
pean languages) that has the form ` a and is given by the T
E
X code
\‘.
acquisitions editor The editor who evaluates a new manuscript from
an author, gets it reviewed, and determines whether it is to be published.
acute accent The accent (used in French, Spanish, and other Euro-
pean languages) that has the form ´ a and is given by the T
E
X code \’.
Also called “accent egout”.
Adobe Acrobat A reader for the pdf computer graphics and page
design language. Particularly well-suited for use on the Internet.
agate An old unit of measure in typesetting equal to slightly less
than 5.5 points. For years the agate was the standard for measuring
advertisements. In England this unit was called the ruby.
AMS American Mathematical Society.
A
M
S-L
A
T
E
X A version of T
E
X that incorporates many of the best
features of L
A
T
E
X and of /
/
o-T
E
X.
A
M
S-T
E
X A macro package for T
E
X that has been created by the
American Mathematical Society (specifically, by Michael Spivak) in or-
der to facilitate the creation of mathematics documents. /
/
o-T
E
X
contains special style macros for the Society’s journals, special fonts,
and also macros for creating commutative diagrams and for other math-
ematical displays.
archive A collection of software located at a particular Web address
that can be accessed through anonymous ftp. Also: a single file that
contains a collection of other files, for manipulation with the tar com-
mand.
©2001 CRC Press LLC
ascender The portion of a lower-case letter that extends above the
main body of the letter. Compare descender.
ASCII code The American Standard Code for Information Inter-
change. An enumeration of all characters, both alphanumeric and sym-
bolic, that can be typed in on a computer keyboard.
back matter Material that follows the main body of text. This
may include appendices, glossaries, tables, references, indexes, and like
material. See front matter.
bar-under accent The accent having the form c
¯
and given by the
T
E
X code \b.
baseline An imaginary line that passes through the bottoms, or bases,
of the capital letters.
bed The base on which the forme is held when printing by letterpress.
Newspapers speak of “sending an issue to bed.”
benday The screen used to create a halftone. See also halftone.
BIBT
E
X A macro for L
A
T
E
X that is designed for the preparation of
bibliographies. Bibliographic data is archived in *.bib files in a special
BibT
E
X format.
big cap-small cap font A font in which lower-case letters are replaced
by small capital letters. This sentence is set in big cap-small cap.
binary code The machine code that is read and understood by a
computer.
binding A method (of which there are many) for securing loose leaves
or sections into a book. Among these methods are perfect and saddle-
stitched.
bitmap A machine language graphic format. See also *.bmp file.
blanket cylinder The cylinder by way of which the inked lithography
plate transfers the image to the paper. Usually the cylinder is covered
with a rubber sheet which prevents wear from the plate coming into
contact with the paper.
bleed Illustrations that extend to the edge of the page, and are trimmed
off at the time of binding, are said to bleed.
block in The process of sketching in the main parts of an image prior
to making the more formal design.
block quotation An excerpt or quotation set in reduced type. Such
a quotation is often displayed. Also called an extract.
blow up An enlargement of a graphic image or a photograph.
*.bmp file A graphics file that is stored in the bitmap format.
body size The height of a type font, measured in points, as measured
from the top of the tallest ascender to the bottom of the lowest descender.
©2001 CRC Press LLC
boldface A font consisting of heavy, dark, thick letters. In some fonts
boldface is available in either upright A, B, C or slanted A, B, C
form.
bond A sized, finished writing paper of 50 grams per square meter or
heavier. Such paper can also be printed upon.
braces The delimiters
_ _
.
brackets The delimiters
_ ¸
.
breve The accent that has the form ˘ o and is given by the T
E
X code
$\breve{a}$.
broadsheet See broadside.
broadside See landscape.
browser A piece of software used for reading material on the World
Wide Web. See also search engine.
bullet A large dot placed before text to add emphasis. Invoked in
L
A
T
E
X with the command \begin{itemize}.
camera copy or camera-ready copy Proof sheets that are ready
to go to the printer. Usually printed on RC (resin-coated) paper. Also
known as “mechanicals”. See also repro copy.
cap line An imaginary line drawn through the tops of all capital
letters.
caps Capital letters.
cap size The distance from the cap line to the baseline.
caption The heading or title of a figure, table, illustration, or of a
chapter or article. As distinct from a legend.
caret The symbol ∧, used in proofreading to indicate an insertion.
Similar to, but generally larger than, the circumflex.
case bound A hardbound book made with stiff outer covers. The case
is often covered with cloth, leather, or vinyl.
cast off To estimate the number of typeset pages that will be made
from a manuscript. Alternatively, to estimate the number of pages of
text that will be made from a set of galleys. Also used as a noun: to
make a castoff.
catchline A temporary heading—provided for identification purposes—
at the top of a galley proof.
cedilla The accent (used in French and other European languages)
that has the form ¸ c and is given by the T
E
X code \c.
chalking If the ink fails to dry satisfactorily after printing, then a
resulting powdering effect on the surface is called chalking.
character count The number of characters (i.e., letters, numerals,
signs, spaces, symbols, etc.) in a piece of copy or a line or a paragraph.
Used as the first step in a cast off.
©2001 CRC Press LLC
chase A metal frame in which (cold) metal type and blocks (i.e.,
engravings) are locked into position in order to make up a page.
check The accent that has the form ˇ a and is given in text by the
T
E
X code \v{a} and in math mode by the T
E
X code $\check{a}$.
circumflex The “hat” accent that is used to denote the Fourier trans-
form, and also occurs in foreign languages such as French.
close up A proof correction mark (copy editor’s mark) that is used to
indicate reduction of the amount of space between characters or words.
cold type composition The traditional technology used for typeset-
ting. Each character or ligature is engraved on a piece of lead. Words,
sentences, and paragraphs are composed by juxtaposing the pieces of
lead in a rack.
collate To assemble the separate pages or section of a book into the
proper order for binding. See gathering.
collective sign A mathematical symbol, such as a sum, product,
union, or integral, that signifies some aggregation.
colophon The publisher’s logo or trade emblem, usually appearing on
the title page of a book. The term also refers to a passage at the end of
a book describing the type and how the book was produced.
color separation The division of a four-color or multi-colored original
into the primary process colors of yellow, magenta, cyan, and black. See
also benday, screen, halftone.
column inch In a newspaper or magazine, a portion of the type body
that is one column wide and one inch deep. The measure is used to
calculate the cost of display advertising.
commercial “at” The symbol @, commonly used in e-mail addresses.
composition The process of setting a manuscript into type. The same
as typesetting.
compositor In traditional typesetting, the person who typesets text.
consonantal digraph A combination of two consonants to express a
single sound. See also diphthong.
copy editing The process by which a manuscript is read and evaluated—
by a trained expert—with a view to correcting both the use of language,
grammar, spelling, syntax, style, consistency, and the formatting of the
page.
copy editor The person who performs copy editing functions.
copy editor’s marks See proofreader’s marks.
copyholder A person who reads manuscript aloud to the proofreader.
copyright The legal entity that gives protection to the creator of
material to prevent use without express permission or acknowledgement
of that creator.
©2001 CRC Press LLC
corner marks The angle marks printed on a sheet to denote the trim
or register marks.
cover The outside binding material of a book or journal. Usually
made of heavier material than the actual pages. The outside front cover,
inside front cover, inside back cover, and outside back cover are often
designated covers 1, 2, 3, and 4 respectively.
CPU Central processing unit. The chip that is the “brains” of the
computer.
cropping The editing away of portions of a photograph or other orig-
inal graphic that are not desired in the printing. Usually cropping is
done in order to allow the other parts of the graphic to be enlarged to
fill the indicated space.
CRT Cathode ray tube. The computer’s monitor. Nowadays some of
the new “thin” monitors use digital technology, not cathode ray tubes.
CTAN (Comprehensive T
E
X Archive Network) A Web site, with
many mirror images, that holds T
E
X macros, fonts, implementations of
T
E
X and L
A
T
E
X, and many other T
E
X resources. See Appendix VI for
more information about CTAN sites.
cursive A typeface that resembles written script.
cut flush A method of trimming a book after the cover has been
attached.
cutout A halftone in which the background has been cropped out in
order to produce a silhouette.
dagger In typesetting, the symbol † that is used to mark equations,
displays, or footnotes.
daisy wheel printer An impact printer which prints from a wheel
that has many different engraved print images. This method of print-
ing is slow, and has been replaced by ink-jet, laser, and other printing
technologies.
ddagger In typesetting, the symbol ‡ that is used to mark equations,
displays, or footnotes.
dash A short horizontal rule used for punctuation. This could be an
em-dash or an en-dash.
delimiter See fence.
descender That portion of a lower-case letter that extends below the
baseline.
developmental editor This is the editor that oversees the writing and
genesis of the book—from early draft to final form that will be sent into
production. The developmental editor will get reviews, from peers of
the author and from others, and will help the author to interpret those
©2001 CRC Press LLC
reviews. The developmental editor will help the author to render his
graphics, will guide the author through writing problems, and will help
him to assemble the final book. The role of the developmental editor is
most decisive, and pervasive, in the writing of elementary textbooks.
die A hardened steel engraved block used to print an inked image.
Used, for example, in the production of high quality letter headings.
dieresis The accent that has the form ¨ a and is given in text by the
T
E
X code \"{a} and in math mode by the T
E
X code $\ddot{a}$.
digraph A combination of two letters to express a sound. See diph-
thong.
diphthong A digraph composed of two vowels. Common diphtongs
are ae and oe . The pairing is called a “diphthong”, but the resulting
compressed symbol (e.g., æ or œ) is called a ligature. See also logotype.
discretionary hyphen See soft hyphen.
displayed mathematics Mathematics that is typeset in a size larger
than the regular text and is set off, with space above and below, from
the text body.
display type Large type, usually at least 18 point and set in bold,
that is used in a heading.
DOS The disc operating system. The one in most prevalent use today
was created by Microsoft in the early 1980s and is denoted MS-DOS.
This computer operating system for PC computers has now been su-
perceded by the Windows products: Windows95,

Windows98,

WindowsNT,

and
Windows2000.

dot accent The accent that has the form ¨ a and is given in text by
the T
E
X code \.{a} and in math mode by the T
E
X code $\dot{a}$.
dot matrix printer A printer that forms each character from a matrix
of dots. Such a printer could be an impact system (in which a wire is
fired at a ribbon), or thermal, or electro-erosion. The dot matrix printer
has now been superceded by laser, ink-jet, and other printer modern
technologies.
dot-under accent The accent that has the form a
.
and is given in
text by the T
E
X code \d{a}.
double spread Two facing pages in printed copy. Also called “double
page spread”.
DPI (dots per inch) The measurement of resolution for page printers,
phototypesetting machines, and graphics screens. The image is divided
into dots and the number of dots per linear inch is the DPI. Modern
laser printers are 1200 DPI or more.
drawn on The method of binding a paper cover to a book by drawing
on the cover and gluing it to the back of the book.
©2001 CRC Press LLC
drop cap A large initial letter at the beginning of text that drops
below the baseline and into several of the lines of text below. Often
typeset in gothic.
drop folio A page number placed at the bottom of the page (as
opposed to top left or right).
dry transfer Characters, images, drawings, etc., that can be trans-
ferred to the work by rubbing on the back of a plastic transfer sheet.
Often known as Letraset.
dunhill A popular typeface that features tall, elongated letters. This
is Dunhill.
*.dvi file A device independent file, created when a T
E
X source code
file is compiled.
dye transfer A photographic color print methodology that uses special
coated paper to produce a full-color image.
editor The publishing process involves several different types of edi-
tors. The acquisitions editor is one who accepts manuscripts for review
and who ultimately decides which of these will go into publication. Once
a manuscript has been accepted, the production editor oversees the pro-
duction process. The copy editor performs a technical review of the
manuscript, correcting for formatting, usage, spelling, and syntax. See
also text editor.
elided A page range is elided if it is written in the form 465–83 rather
than 465–483.
ellipsis Three periods or dots, as . . . or ; used to denote an
omission.
em See em quad.
emacs A popular text editor for the UNIX operating system. Also
known as GNU emacs.
embellished letter A letter that has been augmented with an accent,
bar, cedilla, dot, tilde, umlaut, or other diacritical mark.
em-dash A long dash that used for a pause or break in a sentence.
em quad A space named because, traditionally, it was the width of
the capital letter “M” in the font being used. Nowadays the definition
is variable, and depends on the font and on other parameters as well.
Also called an em.
en See en quad.
encapsulated PostScript

A form of PostScript that is especially
designed for inclusion in other documents, such as a T
E
X document.
Such files have the form *.eps. See PostScript.
en-dash A dash of medium length, used to indicate a range of pages
or numbers, and in similar contexts. Also called an en.
©2001 CRC Press LLC
endpaper A folded sheet of paper, different in texture and perhaps
in color from the main pages of a book. It is pasted in the front of a
book (half to the front cover) and in the back of a book (half to the back
cover). Compare flyleaf.
en quad A space named because, traditionally, it was the width of
the capital letter “N” in the font being used. Nowadays the definition
is variable, and depends on the font and on other parameters as well.
Also called an en.
enunciation A displayed mathematical statement such as a theorem,
lemma, corollary, definition, remark, or example.
*.eps Encapsulated PostScript format. See also PostScript.
Epson emulation or standard The industry standard control codes
for dot matrix printers, developed by Epson Corporation.
even pages See verso pages.
extract See block quotation.
face An abbreviation for “typeface”. Refers to a given font family in
a given style.
FAQ On a Web site, the “Frequently Asked Questions” button.
fence A delimiting mark, usually used in pairs, such as a parenthesis,
brace, or bracket. See also delimiter.
final proof Proof which has gone through the editing and correction
process (i.e., the galley and page proof stages) and is now in final form.
first order index A superscript or subscript that directly augments
a character on the baseline. In the expression ¹
2
1
, both the 1 and the 2
are first order indices. See second order index.
first proof See galley proof.
flexography A rotary or letterpress technology consisting of printing
from a rubber or other flexible plate and using fast-drying ink.
flush left Alignment of copy along the left margin.
flush right Alignment of copy along the right margin.
flyleaf A blank leaf or page bound into the front or back of a book.
Compare endpaper.
folio The page number. Also a sheet of a manuscript.
font A style and size of type. A font includes all upper- and lower-
case letters, all numerals, and all punctuation. Most fonts will have a
boldface variant and a slanted or italic variant.
forme Type and blocks assembled in pages and placed in a metal chase
ready for printing.
©2001 CRC Press LLC
four-color press A press that can print four colors simultaneously
(usually yellow, magenta, cyan, and black), to produce all possible colors
on one pass through the printer.
freeware Software that is available (often on the World Wide Web)
without any payment or quid pro quo.
front matter Material that precedes the main body of text. This may
include the title page, foreword, preface, acknowledgements, dedication,
table of contents, the list of figures, and like material. See back matter.
ftp The acronym for “file transfer protocol”. This is both the protocol
and the software for transferring files between different computers on
the Internet.
galley proof The typeset manuscript of a book or paper, formatted
according to customary typesetting standards but not broken for pages.
The galley proof will not show space for figures, tables, and other dis-
played material.
galleys Printing term for the long metal trays used to hold type after
it had been set and before the press run.
gathering The process of placing the pages, sections, or signatures of
a book in the proper order for binding. See collate.
*.gif file A graphics file that is in the Graphics Image Interchange
format.
glue In T
E
X, the information about spacing and adjoining to other
characters that comes “packaged” with each T
E
X character.
graphics artist A technically trained artist who can render graphics
for a book or article for the purposes (and to the standards of) publica-
tion.
gravure A printing process in which the image is etched onto a metal
plate attached to a cylinder. Also called “rotogravure” or photogravure.
grey scale A range of luminance values for evaluating shading from
white to black. Used in evaluation of scanner images. The term is
sometimes applied to certain types of CRTs or computer monitors.
gutter The inner margins of a page, which are over-sized because they
will be bound.
h´aˇcek The accent that has the form ˇa and is given in text by the
T
E
X code \v{a} and in math mode by the T
E
X code $\check{a]$.
hairline rule The thinnest rule (line) that can be printed.
hairlines The thinnest of the strokes in a typeface. See hairline rule.
half title A (possibly abbreviated) title which appears by itself on a
page preceding the more formal title page.
©2001 CRC Press LLC
halftone A process by which a black-and-white photograph is filtered
through a screen so that various shades are represented by different
densities of dots. See also benday.
halftone screen A glass plate or film, carrying a network of parallel
lines, that is placed between an original photograph and the film to be
exposed. The number of lines per inch controls the coarseness of the
final dot configuration in the resulting halftone.
handwork Composition that requires manipulation by an artist rather
than just keyboarding.
hang indent A paragraph typeset so that the first line is flush left
and subsequent lines uniformly indentd.
hanging punctuation Punctuation that is allowed to fall outside the
print margins.
hardback A case bound book with a separate stiff board cover. A
book that is not paperback.
hard copy A version of a document that is printed out on a (paper)
page; as opposed to the electronic (or “soft”) version of the document.
head The margin at the top of a page.
header Computer code at the beginning of a (T
E
X) document, giving
format and style specifications.
head margin The blank space occupying the region from the top edge
of the page to the topmost element of the print body. See also sinkage.
heads Words or phrases used to differentiate different portions of the
text (chapters, sections, etc.). Often set in boldface, or big cap-small
cap, or some other special type.
helvetica A popular sans serif typeface.
house style A publisher’s or publication’s preferred style of spelling,
punctuation, hyphenation, indentation, and display.
HTML (Hypertext Markup Language) The code that is used to
prepare Web pages and text for electronic documents on the Internet.
Especially useful for hyperlinks. See MathML, SGML, XML.
hypertext The language used to format material for the Internet.
hyphen The shortest type of dash. Used to hyphenate words.
icon A pictorial image on screen that designates a utility function, a file,
a folder, or an application. In the Windows or Macintosh environment,
the icon is activated by clicking with a mouse or other pointing device.
impression cylinder The cylinder of a printing machine that brings
the paper into contact with the printing plate.
imprint The name and address of a publisher, usually appearing on
the title page of a book. Sometimes this material will appear with the
colophon (or publisher’s logo) and the date of publication.
©2001 CRC Press LLC
index These are subscripts and superscripts which augment a character
on the baseline. Also the alphabetized list of topics at the end of a book
or manuscript. See first order index and second order index.
ink-jet printer A printing technology in which particles or droplets
of electrically charged ink are sprayed from a matrix of tiny ink jets.
inline mathematics A mathematical expression that occurs in the
flow of text, and is the same size as the regular text. As opposed to
displayed mathematics.
insert An instruction to the printer for the inclusion of additional
copy.
international paper sizes The International Standards Organization
(ISO) system of paper sizes is based on a series of three sizes: A, B,
and C. Business-size paper, in many industrial countries, is size A4.
North American countries do not subscribe to the ISO standard. The
American 8.5

11

business paper is not an ISO paper size, nor is the
American legal 8.5

14

size.
Internet The union of millions of computers, linked by modem and
ethernet, around the world. Each such computer is accessible from any
other.
ISBN (International Standard Book number) Similar to the
SBN number.
ISO The International Standards Organization for paper size. See
international paper sizes.
ISP Internet service provider.
italic A slanted version of a font, used to slow down the eye when
emphasis is desired. This sentence is typeset in italic. Mathematical
expressions are generally typeset in a form of italic called math italic:
r
2
− 2j. = n. Some expressions in mathematical logic are not typeset
in italic.
*.jpg file A graphics file that is in the Joint Photographic Experts
Group format.
justified lines A line typeset flush against one or both margins. See
left justified line and right justified line.
kern To move the point of typesetting either left-right or up-down.
keyline An outline drawn on artwork to show the size and position of
an illustration.
laid In the production of high-quality stationery, the paper with a
watermark pattern showing the wire marks used in the paper-making
process.
laminate A thin, transparent plastic coating on paper or board to
provide protection and a glossy finish.
©2001 CRC Press LLC
landscape Presentation of material on an 11

8.5

page. In other
words, the page is held sideways instead of up-and-down. Contrast with
portrait.
laser printer A high-quality printing technology—in wide use today—
in which a laser beam is used to produce an image on a photosensitive
drum. The image is then transferred to paper by a process similar to
xerography.
lateral reversal An image (positive or negative) transposed from left
to right as in a mirror reflection.
L
A
T
E
X A T
E
X macro package, originally designed by Leslie Lamport,
which provides style files for articles, reports, books, and other standard
types of documents. Also called L
A
T
E
X 2.09.
L
A
T
E
X2
ε
A newer release of L
A
T
E
X that incorporates powerful, new
macros and graphics features. L
A
T
E
X3 is under preparation at the time
of this writing.
layout A sketch of a page, in preparation for printing, which shows
positioning of text and illustrations and gives overall instructions.
leading A vertical space used in typesetting. Traditionally, the space
was effected with a blank piece of lead type. It is common to put 2 points
of space between lines of text, hence the phrase “2 point leading”. A 6
point leading is sometimes used between sections, and before and after
displayed material (i.e., mathematical formulas or enunciations).
left justified line A line of text which is typeset flush against the left
margin.
legend The explanatory material that accompanies an illustration,
figure, table, or chart. Compare caption.
Letraset See dry transfer.
letterpress The type of printing in which ink is applied directly to
the metal type and then transferred to the paper by the application of
pressure. Compare lithography or offset printing.
letterset A printing process that combines offset printing with a let-
terpress relief printing plate.
letterspacing The addition of space between the letters in a word in
order to increase line length, thereby improving the appearance of a line
of text.
ligature A symbol resulting from the juxtaposition of two letters—for
instance æ or œ. See also diphthong and logotype.
line copy/line drawing (Graphic) copy that is produced with pen
strokes, without using a screen. Compare halftone.
line cut An engraving in which there are no tonal gradations. The
graphic effect is produced with lines rather than halftones.
©2001 CRC Press LLC
line gauge A metal rule used by printers, often marked with pica
measurements.
linen tester The magnifying glass designed for checking dot images
in a halftone.
lineup table A backlighted glass table used for preparing and checking
the alignment of page layouts and paste-ups.
line weight The thickness of a line or stroke used in a line drawing.
lining figures Numerals that are aligned at the baseline and at the
top.
linotype A typesetting machine that sets solid lines of slugs of metal
type. This was the standard method of typesetting for the first half of
the twentieth century.
lithography The process of printing in which ink is applied to the
metal type, then transferred to a rubber pad by way of pressure, and
then finally transferred to the paper by way of direct contact with the
rubber pad. Most modern printing is done with lithography.
logotype A single piece of type bearing two or more usually separate
elements. See also diphthong and ligature. Often the word “logotype”
is used to refer to the piece of cold type that bears the diphthong or
ligature. Abbreviated “logo”.
long Hungarian accent or umlaut The accent that has the form ˝ a
and is given by the T
E
X code \H{a}.
loose leaf A method of binding which allows for insertion and removal
of pages, for continuous updating. Binding in a three-ring binder of
ACCO folder.
lower-case Denotes those letters which are not capitalized. Compare
upper-case.
macro A collection of computer commands that may be invoked by
using a nickname that has been assigned to them.
macron An overbar in text. The accent that has the form ¯ a and is
given by the T
E
X code \={a}.
majuscule An upper-case letter.
makeready The process by which a letterpress plate is adjusted or
aligned to compensate for irregularities in the plate.
make-up The assembling of all elements in order to form the printed
image.
managing editor The person who oversees all the other editorial staff
in the production of a book—the developmental editor, the production
editor, and the copy editor.
©2001 CRC Press LLC
manuscript The hard-copy draft of a document that you submit to
the publisher.
Maple A high-level computer language, created by the mathematicians
of Waterloo, for doing mathematics by way of symbol manipulation.
margins The non-printing areas of a page. The blank space on the
edges of a printed page.
mark up Copy prepared for a compositor setting out all the typesetting
instructions.
markup language A high-level computer language that is used to
describe the content and formatting of a page of textual material. T
E
X
is a markup language.
mask Opaque material, often masking tape, that is used to block off
an area of artwork.
masthead Detailed information about publisher, editors, and editorial
staff that are usually printed on the contents page of a periodical.
Mathematica A high-level computer language, created by Wolfram
Research, Inc., for doing mathematics by way of symbol manipulation.
math italic The italic font that T
E
X uses to typeset mathematics.
MathML Mathematical Markup Language. A markup language to be
used with XML. Especially good for mathematical and scientific content.
See HTML, SGML, XML.
mechanical binding A method of binding that consists of securing
pre-trimmed pages by the insertion of wire or plastic spirals through
holes drilled in the binding edge.
mechanicals See camera copy or repro copy.
METAFONT Donald Knuth’s language for creating fonts. Part of the
T
E
X family.
miniscule A lower-case letter.
mock up The rough visual dummy of a publication or design of a
work.
moire A checkered effect that is achieved by superimposing halftone
screens at the wrong angle.
monospace The attribute of a font of having each character occupy
the same amount of horizontal space.
monotype A particular sort of typesetting machine. Material is key-
boarded, and a paper tape produced. Also used to refer to a typesetting
style in which all characters have the same width.
montage A single image formed from the assemblage of several images.
mounting board A heavy board used for mounting and assembling
artwork.
©2001 CRC Press LLC
MS See manuscript.
mutt The em quad.
negative A photographic image in which colors are reversed, or black
and white are reversed. See positive.
nipping The compression of air after sheets of a book are sewn to-
gether.
oblique strok See shilling or solidus or virgule.
OCR See optical character recognition.
odd pages See recto pages.
offprint See reprint.
offset printing See lithography.
oldstyle type A typeface that is characterized by stressed strokes and
triangular serifs. An example of an oldstyle typeface is Garamond.
old style numerals (o1:¸¸¸6¿8µ) A style in which most numer-
als ascend or descend from the r-height of the other characters in the
typeface.
opacity The degree to which paper will show print through.
optical character recognition The technology by which a machine,
by way of scanning, can recognize characters on a page and convert them
to ASCII code. As opposed to bitmapping.
orphan A brief phrase—less than a full line—from the start of a
paragraph that occurs at the bottom of a page.
outline font A typeface in which characters are formed with only their
outline specified (rather than solid strokes). Outline fonts will scale well.
overlay A transparent sheet that shows the color breakdown—used in
the preparation of multi-color artwork.
overprinting Printing a second time over an area already printed.
The method may be used to emphasize changes or alterations.
overs Additional paper used to compensate for that which is spoiled
in the printing process. Can also refer to extra copies above the number
initially ordered.
overstrike A method used in word processing, and also with a tradi-
tional typewriter, to produce a character not in the given typeface by
superimposing two given characters. For example, a mathematician will
produce a blackboard boldface C by superimposing an 1 and a C.
ozalid A patented method for producing page proofs from paper or
film.
page description language (PDL) A programming language that
enables both text and graphics to be described in a series of mathemat-
ical statements. A page description language allows the applications
©2001 CRC Press LLC
software to be independent of the physical printing device. PostScript

is an important PDL.
page printer A non-impact printer which produces a complete page in
one action. Some examples of page printers are laser, LED, LCD shutter
xerographic, ion deposition, electro-erosion, and electro-photographic.
page proof Similar to the galley proof, but broken for pages, and with
space set for figures, tables, and other displays. This is the typeset text
in its final form.
pagination The numbering of pages in a book. There are dry transfer
pagination sheets to help with pagination.
pantone A patented name for an ink color matching system. Used,
for example, by CorelDRAW.
paperback A book with a perfect binding whose covers are made of
heavy paper.
paper plate An offset printing plate for a short print run on which
matter can be typed directly.
paragraph mark A typesetting symbol (¶) used to denote the start
of a paragraph. Can also be used to mark a footnote.
paste up The elements of a layout mounted in position—often on a
mounting board—to form camera-ready artwork.
PDF format A file format (Portable Document File format) due to
Adobe that is similar to PostScript (q.v.). Files in this format generally
have extension .pdf, as in myfile.pdf. Now a standard file format
for electronic document printing and distribution. The Adobe Acrobat
reader is designed to read pdf files.
PDL See page description language.
perfect binding A style of bookbinding in which the backs of the
signatures are cut off, the edges roughened, and the pages then glued
together with adhesive. Compare signature.
perfector A printing press which can print both sides of the paper at
one pass.
PE’s (printer’s errors) Changes or errors made in the typeset text
that are the responsibility of the typesetter or copy editor. These alter-
ations are not charged to the author.
photogravure See gravure.
phototypesetting A process of typesetting onto film.
pica A unit of distance in typesetting, equal to 12 points.
pica ruler The metal rule most commonly used by a typesetter in
page design.
©2001 CRC Press LLC
pi fonts Characters not typical of a font, but added specially. These
might include mathematical symbols and timetable symbols.
pixel A screen dot on a computer monitor or CRT. A typical SVGA
monitor is about 1200 pixels across and 1000 pixels high.
plant costs Fixed costs of the book production process that are not
connected with any one book or by the size of the edition. These include
composition, plates, electricity, staff salaries, overhead, and equipment.
point A unit of length equal to 0.013837 inches. There are about 72
points to an inch.
portrait Presentation of material on an 8.5

11

page. In other
words, the page is held up-and-down instead of sideways. Contrast with
landscape.
positive The true photographic image of the original made on paper
or film. See negative.
PostScript A page design language created by Adobe. This is a high-
level computer language that allows the formatting of both graphics and
text.
preamble That part of a L
A
T
E
X source code file that occurs before the
line \begin{document}.
preface The part of the front matter of a book that introduces the
subject and that particular presentation of it. See also prospectus.
preprint A copy of the manuscript or original hard copy of an article,
chapter, or book manuscript.
press run The number of copies (of a book or article) to be printed.
primary colors These are cyan, magenta, and yellow. When mixed
together in suitable proportions with black these will produce all other
colors.
print block The entire body of text on a page.
print body The totality of print on a page, measured from the bottom
of the top margin to the top of the bottom margin, and from the right
edge of the left margin to the left edge of the right margin.
print-on-demand The digital technology that allows the printing of
a small number of copies of a document, at the moment of demand. In
traditional typesetting, the cost of setting up the printer made it too
impractical and costly to do small print runs. With digital technology
this is now feasible and affordable.
production editor The editor who oversees the process of bringing a
book from final manuscript form to final, printed copy. The production
editor oversees the copy editor, graphics artists, and the T
E
X specialists
to whom the MS is outsourced.
©2001 CRC Press LLC
progressive proofs A series of plates, each produced in one of the four
basic colors (yellow, magenta, cyan, and black). These are superimposed
progressively and used to check color quality and alignment.
proof Copy obtained from inked type, plate, block, or screen for the
purpose of checking quality and accuracy.
proof correction marks See copy editor’s marks, proofreader’s marks.
proofreader The person who reads proof copy for errors.
proofreader’s marks A universally accepted collection of symbols
and markings that are used to annotate proof copy. See the plates in
Appendix I. See copy editor’s marks.
proofreading The process of reading proofs to check them for errors—
usually by comparing those proofs with the original copy. The process
of proofreading involves marking any errors with standard proofreader’s
or copy editor’s marks.
proofs Typeset text that is ready for checking (proofreading) and
correction. See page proof and galley proof.
proportional spacing A method of spacing in typesetting whereby
each character is spaced according to the varying widths of the elements
of the font. The method uses kerning. See also monospace.
prospectus Similar to a preface, but also includes information about
the market and about other books on the same subject. See preface.
public domain A term used to describe software that has not been
copyrighted.
quad Same as an em quad.
quadding Additional space to fill out a line of type, using em quads
or en quads.
quire One twentieth of a ream (25 sheets) of paper.
ragged Lines of set type that do not begin or end at the same horizontal
position.
ragged left Describes a body of text with an unjustified left margin.
ragged right Describes a body of text with an unjustified right margin.
rag paper High quality paper for stationery that is manufactured
from cotton rags.
Raster image processor (RIP) A hardware engine which calcu-
lates the bitmapped image of both text and graphics from a series of
mathematical instructions. This process may be effected using a page
description language.
RC paper Resin-coated paper, used for camera-ready copy. If printing
plates are to be prepared in the traditional manner, then proof sheets
©2001 CRC Press LLC
are first printed on RC paper and then “shot” (photographed) for re-
production on a copper or zinc plate.
ream A package of 500 sheets of paper.
recto pages The odd-numbered pages of a book. See also odd pages.
recto-verso printing Two-sided printing (as in a book).
reference mark A symbol used in text to direct the reader to a
footnote. This could include †, ‡, ¶, (∗).
register The printing of each impression, for each primary color, in the
correct position relative to other impressions. This to check alignment.
register marks Marks used in color printing to position the paper
correctly. These marks are usually crosses or small circles.
reprint A separate reproduction of an article or book chapter; printing
of part of a larger work. See offprint. The author of a journal article
is usually given fifty reprints free of charge. These are produced by
printing fifty unbound copies of the journal.
repro copy See camera copy.
resolution Measurement used in typesetting to determine quality of
output. The units are “dots per inch” or DPI.
rest in proportion (RIP) An instruction, when sizing artwork, that
other parts of the artwork are to be sized proportionately.
retouching A method for altering or correcting artwork or color sep-
arations.
reverse out To reproduce as a white image out of a solid background.
revise The stage at which corrections have been incorporated from
previous proofs and new, corrected proofs submitted. For example, first
revise, second revise, etc.
right justified line A line of text which is typeset flush against the
right margin.
roman The default typeface in T
E
X. This definition is typeset in
computer modern roman.
rotary press A web- or reel-fed printing press which uses a curved
printing plate mounted on a cylinder.
rough A preliminary sketch of a proposed design.
royal Printing paper of size 20

25

.
rule A line, usually horizontal, for separating bodies of text. Also a
line used for decorative purposes.
runarounds Text set in short lines so that they run around a fixed
object, such as an illustration or photograph.
©2001 CRC Press LLC
run in To merge two paragraphs or bodies of text into a continuously
flowing body.
running foot The abbreviated section title, or chapter title, or author
identification that appears across the bottom of the page, below the body
of type.
running head The abbreviated section title, or chapter title, or author
identification that appears across the top of the page, above the body
of type.
sans serif A typeface or font that is plain, without serifs.
SBN number (Standard Book Number) A nine-digit number,
assigned in a unique fashion to each published book. The first part
identifies the publisher; the second part identifies the particular edition
or title; the third part is a parity check. See also ISBN.
scale The ability of a program to reduce or to enlarge the amount of
space that an image will occupy.
scaling A means of determining the amount of enlargement or reduc-
tion required to fit a photograph or graphic to an alloted space.
scamp A sketch of a design that shows the basic concept. See also
rough.
screen A traditional technique for transforming a solid block of color,
or a continuous-tone image, into a pattern of dots. The size and density
of dots determines the intensity and darkness of the color. The resulting
image is called a halftone.
script A typeface or font that resembles handwritten letters: /, T are
examples of script.
search engine A piece of software used for conducting searches on the
World Wide Web. Some popular search engines are Netscape, Microsoft
Internet Explorer,

Excite,

Yahoo,

Google,

and Dogpile.

See also
browser.
second order index A superscript or subscript that augments a first
order index. In the expression ¹
f
3, the 3 is a second order index. See
first order index.
section mark A character (¸) used at the beginning of a new section.
serif A small decoration or embellishment—a crossed line at the end
of the main strokes—that appears as part of letters on a font. The letter
R has three serifs and the letter A has two serifs.
set off The accidental transfer of the printed image from the front of
one sheet (where the ink has been applied) to the back of another.
set size The width of the type body at a given point size.
set solid Type set without line spacing (leading) between the lines.
©2001 CRC Press LLC
SGML Standard Generalized Markup Language. This is an Interna-
tional Standard that describes a generalized markup scheme for repre-
senting the logical structure of documents in a system-independent and
platform-independent manner.
shareware Software that can be distributed for free, but for which a
modest payment is requested from satisfied users.
sheet fed A printing press which handles single sheets of paper instead
of rolls or reels of paper.
sheetwise A method of printing a section by which half the pages
are imposed and printed and then the remaining half of the pages are
printed on the other side.
shilling The forward slash , used in the shilling form of a fraction.
See also solidus or virgule.
shilling form of a fraction The format of a fraction that is o,/ rather
than
a
b
(the stacked fraction format). See also solidus.
shooting a page In traditional typesetting, the process of photograph-
ing each page of the repro copy of a book or manuscript that has been
test-printed on RC paper.
sidebar A vertical bar positioned on the side of the screen—usually
on the right.
side heading A subheading that is typeset flush into the text at the
left edge.
signature A large sheet of paper, folded and cut to produce a certain
number of pages (usually a power of 2, and often 16 or 32). Classically,
books were bound in signatures. Compare perfect binding.
sinkage The vertical distance between the very top of a printed page
and the top of the body of text (or text block).
slanted roman A variant of the roman typeface in which the letters
are slanted.
slug A 6-point or 12-point piece of lead that is used for vertical spacing
in the cold type typesetting process.
slurring A smearing of the printed image caused by slippage of the
paper during the printing process.
softback See paperback.
soft cover See softback or paperback.
soft hyphen A specially coded hyphen that only displays when the
potentially hyphenated word occurs at the end of a line.
solidus The slanted line used between the parts of a fraction written
in shilling form. See also shilling or virgule.
©2001 CRC Press LLC
source code In a high-level computing language, this is the file con-
sisting of the commands pertaining to that language.
spell check A software device for checking the spelling in a document.
Modern spell-scheckers have an extensible dictionary so that the checker
may be adapted to the specialized needs of the user.
spine The back of a book; that part which joins the front and back
covers. Also called the backbone or shelfback.
spiral binding A type of binding in which a spiral of plastic or metal
is wound through holes in the pages.
s/s (same size) Instruction to reproduce to the same size as the
original.
stacked fraction The form of a fraction in which the numerator is
placed directly above the denominator, with a horizontal line in between,
for example
a
b
. Contrast with the shilling form of a fraction.
stem The main vertical stroke making up a character in a type font.
STET Sign used in proof correction work to cancel a previous correc-
tion. From the Latin “let it stand”.
strap A subheading used above the primary headline in a newspaper
article.
style file A collection of commands, or a macro, that specifies page
style for a particular document or publisher. See also header.
style sheet A collection of tags that specify page layout style, para-
graph settings, type specifications. The style sheet can be defined by
the user and saved for other documents. The *.sty files in L
A
T
E
X serve
as style sheets. See house style.
subheads Headings, subordinate to chapter heads, that mark divisions
in a chapter. These might include section heads, subsection heads, and
so forth.
subscript A symbol, smaller than those on the baseline, that ap-
pears slightly below the baseline and is an embellishment to a baseline
character—for example , is a subscript in the expression r
j
.
superior numeral A small numeral, printed above the r-height of
the font, used to denote a reference mark.
superscript A symbol, smaller than those on the baseline, that ap-
pears slightly above the baseline and is an embellishment to a baseline
character—for example , is a superscript in the expression o
j
.
surprint Printing over a previously printed area. See overprinting.
swash letters Italic characters with extra flourishes. These are some-
times used at the beginnings of chapters.
tabloid A page half the size of a broadsheet or broadside.
©2001 CRC Press LLC
tabular setting Text set in columns.
tags The many different formats which make up a style sheet—
automatic section numbering, margin and column settings, page layouts,
paragraph settings, hyphenation and justification, widow and orphan
control.
technical report A document published by a scientific organization
or company for external circulation. Usually such a report is part of a
series, and appears with a special cover bearing a logo.
template The prototype or paradigm for a standard page layout,
including basic details of the page dimensions.
T
E
X A high-level computing language created by Donald Knuth in
the 1970s for the purpose of typesetting—especially the typesetting of
technical material.
text editor A piece of software for entering ASCII code and line
breaks/carriage returns. Unlike a word processor, a text editor does not
have (hidden) formatting information.
text type The typeface(s) used for the main text of material in a given
article or book.
text wrap See runarounds.
thick space A horizontal space of thickness equal to two thin spaces.
thin space A horizontal space of thickness about 0.023

.
tie-after accent The accent in text that has the form oo and is given
by the T
E
X code \t{oo}.
tied letters See ligature.
*.tif file A graphics file that is in the Tag Image File format.
tilde The accent that has the form ˜a and is given in text by the T
E
X
code \~{a} and in math mode by the T
E
X code $\tilde{a}$.
tip and tear To remove a page from a bound book (perhaps one with
an error on it) and to tip in a corrected page.
tip in To paste an extra sheet into a bound book.
TLA Three-letter acronym.
tone line process The method of producing line art from a continuous
tone original.
transparency A full color, photographically produced image on trans-
parent film.
trash can The icon for the deleting of files or other (computer) objects.
See particularly the Windows desktop.
trim Cutting of the finished, printed product to the correct size.
trim marks See register marks.
©2001 CRC Press LLC
trim size The final size of the entire page of a book or journal, after
trimming.
troff An early computer typesetting system and text formatting util-
ity; a precursor of T
E
X. Still in use on UNIX systems as part of the
Documenter’s WorkBench (DWB) system.
*.ttf The True Type Font format.
two-em quad A space in typesetting that is equal to two standard
em quads.
typeface The raised or engraved surface carrying the image of a char-
acter in metal. Also used to refer to a complete set of characters of a
given family or style.
typescript A typed manuscript.
typesetting See composition.
type (text) page The total area of the print block on a page.
typewriter-like font A font that simulates the characters that
are produced by a typewriter.
typo Abbreviation for typographical error. An error in the typeset
copy or proof copy.
typographer A specialist in the design of printed matter, page design,
and the art of typography.
typography The design and planning of printed material using type.
UCC (Universal Copyright Convention) Gives protection to au-
thors, editors, and originators of text, photographs, or illustration; pre-
vents use without permission or acknowledgement. See copyright.
umlaut The accent that has the form ¨ a and is given in text by the
T
E
X code \"{a} and in math mode by the T
E
X code $\ddot{a}$. See
also dieresis.
UNIX A computer operating system developed at AT&T Laboratories
and at the University of California at Berkeley. This operating system is
public domain, and is widely used on work stations and super computers.
upper-case Denotes capital letters.
URL Universal Resource Locator. This is the address of a particular
page on a particular computer on the Internet.
vellum The chemically treated skin of a calf used as a writing material.
Also used to describe any thick, creamy book paper.
verso pages The even-numbered pages of a book. See also even pages.
vertical justification The adjusting of interline spacing (leading) and
manipulation of text in fine increments to make columns and (facing)
pages end at the same place.
©2001 CRC Press LLC
vi The primary text editor of the UNIX operating system.
vignette A small illustration in a book that is not enclosed in a border.
virgule See shilling or solidus.
watermark An impression or image incorporated in the process of
papermaking which shows the name of the paper or a company logo.
Web browser A piece of software designed for accessing locations on
the World Wide Web.
weight The degree of thickness or boldness of a character or a font.
wf Abbreviation for “wrong font”. Used as a copy editor’s mark to
indicate that a character is set in the wrong font.
widow A brief phrase—less than a full line—from the end of a para-
graph that occurs at the top of a page.
Windows The operating system created by Microsoft. Windows
comes in four flavors: Windows95, Windows98, WindowsNT, and Win-
dows2000.
word break The division of a word—usually between syllables, and
designated by a hyphen—at the end of a line.
word processor A piece of software for creating documents. Con-
tains (hidden) formatting information, and displays on the screen (in a
WYSIWYG manner) what will be printed on the page.
word wrap In a word-processing environment, the automatic right-
justification of lines and adjustment of the number of words on a line.
work and tumble A method of printing where pages are imposed
together. The sheet is printed on one side and then turned or tumbled
from front to rear to print the opposite side.
work and turn A technique of printing whereby pages are imposed
in one forme or assembled on one film. One side is printed and then the
sheet turned over and printed from the other edge using the same forme.
World Wide Web A subset of the Internet consisting of locations
that can be accessed with a standard Web browser.
WYSIWYG “What you see is what you get”.
xerography A photocopying or printing process in which the image
is formed using the electrostatic charge principle. In this process, ink is
replaced by tone.
x-height The height of lower-case characters, such as the x in a font.
XML Extensible Markup Language. An extensible markup language
that improves on HTML in many ways. A subset of SGML. See HTML,
MathML, SGML.
©2001 CRC Press LLC
References
[BUJ] H. Budgett and J. K. Johnstone, Typesetting and Publishing
Glossory, http://www.sos.com.au/files/glosray.html.
[BUT] J. Butcher, Copy-Editing: The Cambridge Handbook for Ed-
itors, Authors, and Publishers, 3
rd
ed., Cambridge University Press,
Cambridge, 1992.
[CBB] T. W. Chaundy, P. R. Barrett, and C. Batey, The Printing of
Mathematics, Oxford University Press, London, 1954.
[CHI] The Chicago Manual of Style, 14
th
revised ed., University of
Chicago Press, Chicago and London, 1993.
[GMS] M. Goossens, F. Mittelbach, and A. Samarin, The L
A
T
E
X Com-
panion, Addison-Wesley, Reading, Massachusetts, 1994.
[GRM] M. Goossens, S. Rahtz, and F. Mittelbach, The L
A
T
E
X Graphics
Companion, Addison-Wesley, Reading, Massachusetts, 1997.
[GOR] M. Goossens and S. Rahtz, The L
A
T
E
X Web Companion, Addison-
Wesley, Reading, Massachusetts, 1999.
[HIG] N. J. Higham, Handbook of Writing for the Mathematical Sci-
ences, Society for Industrial and Applied Mathematics, Philadelphia,
1993.
[KNU] D. Knuth, The T
E
Xbook, Addison-Wesley, Reading, Massachusetts,
1984.
[KRW] S. G. Krantz, A Primer of Mathematical Writing, American
Mathematical Society, Providence, 1997.
[KRA1] S. G. Krantz, Function Theory of Several Complex Variables,
1
st
ed., John Wiley & Sons, New York, 1982.
[KRA2] S. G. Krantz, A Panorama of Harmonic Analysis, The Math-
ematical Association of America, Washington, D.C., 1999.
[LAM] L. Lamport, L
A
T
E
X: A Document Preparation System, 2
nd
re-
vised ed., Addison-Wesley, Reading, Massachusetts, 1994.
[OCO] M. O’Connor, How to Copyedit Scientific Books and Journals,
ISI Press, Philadelphia, 1986.
[PEM] J. E. Pemberton, How to Find Out in Mathematics, 2
nd
ed.,
Pergamon Press, London, 1969.
[SAK] S. A. Sawyer and S. G. Krantz, A T
E
X Primer for Scientists,
CRC Press, Boca Raton, Florida, 1995.
©2001 CRC Press LLC
[SKI] M. Skillin, R. Gay, et al, Words into Type, Prentice Hall, En-
glewood Cliffs, New Jersey, 1974.
[SWA] E. Swanson, Mathematics into Type, Updated Edition, Amer-
ican Mathematical Society, Providence, 1999.
[TUF] E. Tufte, The Visual Display of Quantitative Information, Graph-
ics Press, Cheshire, Connecticut, 1983.
[TUR] K. Turabian, A Manual for Writers of Term Papers, Theses,
and Dissertations, 5
th
ed., The University of Chicago Press, Chicago
and London, 1987.
©2001 CRC Press LLC
Resources by Type
Sources on Writing
1. A Manual of Style, 14
th
ed., Chicago: The University of Chicago
Press, 1993.
2. S. Baker, The Practical Stylist, 6
th
ed., Harper and Row, New
York, 1985.
3. R. Barrass, Scientists Must Write: A Guide to Better Writing for
Scientists, Engineers, and Students, Chapman and Hall, London,
1978.
4. T. Barrass, Students Must Write: A Guide to Better Writing in
Course Work and Examinations, Methuen, London, 1982.
5. D. F. Beer, ed., Writing and Speaking in the Technology Profes-
sions: A Practical Guide, IEEE Press, New York, 1992.
6. T. M. Bernstein, The Careful Writer: A Modern Guide to English
Usage, Atheneum, New York, 1965.
7. T. M. Bernstein, Miss Thistlebottom’s Hobgoblins: The Careful
Writer’s Guide to the Taboos, Bugbears and Outmoded Rules of
English Usage, Farrar, Straus and Giroux, New York, 1971.
8. A. Bierce, Write it Right: A Little Blacklist of Literary Faults, The
Union Library Association, New York, 1937.
9. L. Bj¨ orck, M. Knight, and E. Wikborg, The Writing Process:
Composition Writing for University Students, 2
nd
ed., Studentlit-
teratur, Lund, Sweden and Chartwell Bratt, Bromley, Kent, Eng-
land, 1990.
10. V. Booth, Communicating in Science: Writing and Speaking, Cam-
bridge University Press, Cambridge, 1984.
11. B. Bryson, The Penguin Dictionary of Troublesome Words, 2
nd
ed., Penguin, London, 1987.
12. B. Bryson, Mother Tongue: English and How It Got That Way,
Avon Books, New York, 1990.
©2001 CRC Press LLC
13. R. Burchfield, Unlocking the English Language, Faber and Faber,
London, 1989.
14. L. L. Cherry, Writing Tools, IEEE Trans. Comm., COM-30(1982),
100–105.
15. Collins Cobuild English Language Dictionary, London, 1987.
16. B. M. Cooper, Writing Technical Reports, Penguin, London, 1964.
17. D. Crystal, The English Language, Penguin, London, 1988.
18. P. J. Davis, Fidelity in mathematical discourse: Is one and one
really two?, Am. Math. Monthly 79(1972), 252–263.
19. R. A. Day, How to Write and Publish a Scientific Paper, 3
rd
ed., Cambridge University Press, Cambridge, and Oryx Press,
Phoenix, Arizona, 1988.
20. R. A. Day, Scientific English: A Guide for Scientists and Other
Professionals, Oryx Press, Phoenix, Arizona, 1992.
21. L. Dupr´e, Bugs in Writing, Addison-Wesley, Reading, Massachu-
setts, 1995.
22. H. F. Ebel, C. Bliefert, and W. E. Russey, The Art of Scien-
tific Writing: From Student Reports to Professional Publications in
Chemistry and Related Fields, VCH Publishers, New York, 1987.
23. A. Eisenberg, Guide to Technical Editing: Discussion, Dictionary,
and Exercises, Oxford University Press, New York, 1992.
24. J. R. Ewer and G. Latorre, A Course in Basic Scientific English,
Longman, Harlow, Essex, 1969.
25. R. Flesch, The Art of Plain Talk, Harper and Brothers, New York,
1946.
26. H. W. Fowler, Modern English Usage, Oxford University Press,
Oxford, 1965.
27. M. Frank, Modern English: A Practical Reference Guide, 2
nd
edi-
tion, Regents/Prentice-Hall, Englewood Cliffs, New Jersey, 1993.
28. L. Gillman, Writing Mathematics Well: A Manual for Authors,
The Mathematical Association of America, Washington, D.C., 1987.
29. E. Gowers, The Complete Plain Words, 3
rd
ed., Penguin, London,
1986.
©2001 CRC Press LLC
30. P. R. Halmos, How to write mathematics, Enseign. Math. 16(1970),
123–152.
31. J. B. Heaton and N. D. Turton, Longman Dictionary of Common
Errors, Longman, Harlow, Essex, 1987.
32. A. J. Herbert, The Structure of Technical English, Longman, Har-
low, Essex, 1965.
33. N. J. Higham, Handbook of Writing for the Mathematical Sciences,
SIAM, Philadelphia, Pennsylvania, 1993.
34. P. J. Hills, ed., Publish or Perish, Peter Francis Publishers, Berry-
croft, Cambs., 1987.
35. J. Kirkman, Good Style: Writing for Science and Technology,
E & FN Spon (Chapman and Hall), London, 1992.
36. D. E. Knuth, T. Larrabee, and P. M. Roberts, Mathematical Writ-
ing, Mathematical Notes No. 14, The Mathematical Association of
America, Washington, D.C., 1989.
37. B. Luey, Handbook for Academic Authors, Cambridge University
Press, Cambridge, 1987.
38. A. J. MacGregor, Graphics simplified: Charts and graphs, Schol-
arly Publishing 8(1977), 151–164.
39. A. J. MacGregor, Graphics simplified: Preparing charts and graphs,
Scholarly Publishing 8(1977), 257–274.
40. Merriam-Webster’s Dictionary of English Usage, Merriam-Webster,
Inc., Springfield, Massachusetts, 1994.
41. J. P. Mitchell, The New Writer: Techniques for Writing Well with
a Computer, Microsoft Press, Redmond, Washington, 1987.
42. E. Newman, Strictly Speaking, Warner Books, New York, 1974.
43. M. O’Connor, Writing Successfully in Science, Chapman and Hall,
London, 1991.
44. Oxford University Press, Oxford Advanced Learner’s Dictionary of
Current English, 4
th
ed., London, 1989.
45. E. Partridge, Usage and Abusage: A Guide to Good English, Pen-
guin, London, 1973.
46. C. R. Perry, The Fine Art of Technical Writing, Blue Heron Pub-
lishing, Hillsboro, Oregon, 1991.
©2001 CRC Press LLC
47. G. Piranian, Say it better, Math. Intelligencer 4(1982), 17–19.
48. R. Quirk and G. Stein, English in Use, Longman, Harlow, Essex,
1990.
49. W. Safire, A Lighthearted Guide to Grammar and Good Usage,
Dell Publishing, New York, 1990.
50. D. L. Schwartz, How to be a published mathematician without
trying harder than necessary, in The Journal of Irreproducible Re-
sults: Selected Papers, 3
rd
ed., George H. Scherr, ed., 1986, p.
205.
51. N. Steenrod, et al., How to Write Mathematics, the AMS, Provi-
dence, Rhode Island, 1973.
52. W. Strunk and W. White, The Elements of Style, 3
rd
Edition,
Macmillan, New York, 1979.
53. J. Swales, Writing Scientific English, Thomas Nelson, London,
1971.
54. M. Swan, Practical English Usage, Oxford University Press, Ox-
ford, 1980.
55. E. Tufte, The Visual Display of Quantitative Information, Graph-
ics Press, Cheshire, Connecticut, 1983.
56. E. Tufte, Envisioning Information, Graphics Press, Cheshire, Con-
necticut, 1990.
57. C. Turk and J. Kirkman, Effective Writing: Improving Scientific,
Technical and Business Communication, 2
nd
ed., E & FN Spon
(Chapman and Hall), London, 1989.
58. B. Turner, Effective Technical Writing and Speaking, 2
nd
ed., Busi-
ness Books, London, 1974.
59. M. C. van Leunen, Handbook for Scholars, revised ed., Oxford
University Press, New York, 1992.
©2001 CRC Press LLC
60. F. K. Baskette, J. Z. Sissors, and B. S. Brooks, The Art of Editing,
5
th
ed., Macmillan, New York, 1992.
61. J. Butcher, Copy-Editing: The Cambridge Handbook for Editors,
Authors, and Publishers, 3
rd
ed., Cambridge University Press,
Cambridge, 1992.
62. T. W. Chaundy, P. R. Barrett, and C. Batey, The Printing of
Mathematics, Oxford University Press, London, 1954.
63. H. Hart, Hart’s Rules for Compositors and Readers at the Uni-
versity Press Oxford, 39
th
ed., Oxford University Press, Oxford,
1983.
64. D. E. Knuth, Mathematical typography, Bulletin of the AMS (new
series) 1(1979), 337-372.
65. L. Lamport, Document production: Visual or logical?, Notices
Am. Math. Soc. 34(1987), 621–624.
66. M. O’Connor, How to Copyedit Scientific Books and Journals, ISI
Press, Philadelphia, 1986.
67. O. Patashnik, BIBT
E
Xing, Documentation for general BibT
E
X
users. File btxdoc.tex in the standard distribution, February 8,
1988.
68. J. E. Pemberton, How to Find Out in Mathematics, 2
nd
ed., Perg-
amon Press, London, 1969.
69. G. C. Reid, Thinking in PostScript, Addison-Wesley, Reading,
Massachusetts, 1990.
70. S. E. Roth, ed., Real World PostScript. Techniques from PostScript
Professionals, Addison-Wesley, Reading, Massachusetts, 1988.
71. R. Rubinstein, Digital Typography—An Introduction to Type and
Composition for Computer System Design, Addison-Wesley, Read-
ing, Massachusetts, 1988.
72. M. E. Skillin, R. M. Gay, et al, Words into Type, Prentice-Hall,
Englewood Cliffs, New Jersey, 1994.
73. R. Smith, Learning PostScript—A Visual Approach, Peachpit Press,
1085 Keith Avenue, Berkeley, CA 94708, 1990.
©2001 CRC Press LLC
74. E. M. Stainton, A bag for editors, Scholarly Publishing 11(1977),
111–119.
75. E. Swanson, Mathematics into Type, revised edition, the AMS,
Providence, Rhode Island, 1979.
76. J. A. Tarutz, Technical Editing: The Practical Guide for Editors
and Writers, Addison-Wesley, Reading, Massachusetts, 1992.
Sources on T
E
X
77. G. Gr¨ atzer, Math into T
E
X: A Simple Introduction to /
/
o-L
A
T
E
X,
Birkh¨ auser, Boston, 1993.
78. D. Knuth, T
E
X and METAFONT, New Directions in Typesetting,
The American Mathematical Society and Digital Press, Bedford,
Massachusetts, 1979.
79. D. Knuth, The T
E
Xbook, Addison-Wesley, Reading, Massachusetts,
1986.
80. D. Knuth, The METAFONT Book, Addison-Wesley, Reading, Mas-
sachusetts, 1986.
81. L. Lamport, L
A
T
E
X: A Document Preparation System. User’s
Guide and Manual., 2
nd
Ed., Addison-Wesley, Reading, Massachu-
setts, 1994.
82. S. Sawyer and S. G. Krantz, A T
E
X Primer for Scientists, CRC
Press, Boca Raton, Florida, 1995.
83. E. M. Stainton, The Fine Art of Copyediting, Columbia University
Press, New York, 1991.
84. R. Wonneberger and F. Mittelbach, BibT
E
X reconsidered, TUG-
boat 12(1991), 111–124.
©2001 CRC Press LLC

To Johann Gutenberg (n´e Gensfleisch) and Donald E. Knuth. e

©2001 CRC Press LLC

Table of Contents
1 Basic Principles 1.1 An Overview 1.2 Choice of Notation 1.3 Displaying Mathematics 1.4 Consistency 1.5 Overall Design 2 Typesetting Mathematics 2.0 Introductory Remarks 2.1 What is TEX? 2.2 Methods of Typesetting Mathematics 2.3 A Lightning Tour of TEX 2.4 The Guts of TEX 2.5 Modes of Typesetting Mathematics 2.6 Line Breaks in Displayed Mathematics 2.7 Types of Space 2.8 Technical Issues 2.9 Including Graphics in a TEX Document 2.9.1 Handling Graphics in the Computer Environment 2.9.2 The Inclusion of a PostScript Graphic A 2.9.3 Graphics and the L TEX2ε Environment 2.9.4 The Use of PCTEX 2.9.5 Freeware that Will Handle Graphics 3 TEX and the Typesetting of Text 3.1 Other Word Processors and Typesetting Systems 3.2 Modes of Typesetting Text 3.3 Hyphens and Dashes 3.4 Alignment 3.5 Typesetting Material in Two Columns 3.6 Some Technical Textual Issues 4 Front Matter and Back Matter 4.1 The Beginning 4.2 The End 4.3 Concluding Remarks 5 Copy Editing 5.1 Traditional Methods of Copy Editing 5.2 Communicating with Your Copy Editor 5.3 Communicating with Your Typesetter

©2001 CRC Press LLC

4 5.2 How to Get on the Web 7. PostScript.8 6 The 6.1 6.3 Communicating with Your Editor Modern Methods of Copy Editing More on Interacting with Your Copy Editor Manuscript Proofs. and Page Proofs The End of the Process Production Process Production of a Paper Production of a Book What Happens at the Printer’s 7 Publishing on the Web 7.1 Introductory Remarks 7.5 Software to Go with your Book or Article.3 Web Resources 7.4 Mathematics and the Web 7.2 6. Appendix VII: Basic TEX Commands A Appendix VIII: A Sample of L TEX Glossary References Resources by Type ©2001 CRC Press LLC . Web Sites Appendix I: Copy Editor’s/Proofreader’s Marks Appendix II: Use of Copy Editor’s Marks Appendix III: Specialized Mathematics Symbols Appendix IV: Standard Alphabets Appendix V: Alternative Mathematical Notations Appendix VI: TEX.5 5.7 5.5. and Related Internet Sites Acrobat.6 5. Galley Proofs.

. and Associate Professor and Professor at Penn State. Dr. he is also a visiting professor and lecturer at many universities around the world. ©2001 CRC Press LLC . and the Outstanding Academic Book Award. Krantz is currently Professor and Chairman of the Department of Mathematics at Washington University in St. the Beckenbach Prize.D. the Chauvenet Prize. received his doctorate from Princeton University in 1974 and his Bachelor of Arts degree from the University of California at Santa Cruz in 1971. Krantz. Louis. a visiting Associate Professor at Princeton University. Krantz has written more than 105 research papers and many articles and reviews. He is the founder and managing editor of the Journal of Geometric Analysis. Krantz is the holder of the UCLA Foundation Distinguished Teaching Award. He is also the editor of the CRC Press series Studies in Advanced Mathematics. Dr. Ph.AUTHOR Steven G. Dr. He has written or edited thirty books. Krantz was an Assistant Professor at UCLA. Dr.

PREFACE

Ellen Swanson’s book Mathematics into Type is a unique and important contribution to the literature of technical typesetting. It set a standard for how mathematics should be translated from a handwritten manuscript to a printed book or document. While Swanson’s book was intended primarily as a resource for technical typesetters, it was also important to mathematical and other technical authors who wanted to take an active role in ensuring that their work reached print in an attractive and accurate form. The landscape has now changed considerably. With the advent and wide availability of TEX,1 most mathematicians can take a more active role in producing typeset versions of their work. Indeed, many mathematicians currently use TEX to write preliminary versions of their work that are very similar (in many respects) to what will ultimately appear in print. While the output from TEX has a more typeset appearance than that from most word processors, the TEX product is not automatically (without human intervention) “ready to go to press.” There are still “postprocessing” typesetting issues that must be addressed before a work actually appears in print. The style and format of running heads, section headings and other titles, the formatting of theorems and other enunciations, the text at the bottom of the page, page break issues, and the fonts and spacing used in all of these go under the name of “page design”. These are often customized for a particular book or journal. The index and table of contents must be designed and typeset. Graphics, and sometimes new fonts, must be integrated. Additional questions of style in the formatting of equations and superscripts and subscripts can also arise. Most TEX users do not know how to handle the questions just listed, which is why most publishers currently send TEX documents for books or journal articles to a third-party TEX consultant. The purpose of the present work is to serve as a touchstone for those who want to learn to make typesetting decisions themselves.
1 T X is a markup language for doing mathematical typesetting. We will talk about E TEX in more detail as the book develops.

©2001 CRC Press LLC

Let us set out once and for all what this book is not. It is not a text for learning TEX. Indeed, A TEX Primer for Scientists by Sawyer and Krantz provides a venue for the rapid assimilation and mastery of TEX basics. There is no need to repeat the lessons of A TEX Primer here. Instead, the present volume is (in part) a book on how to use TEX. But the typesetting principles enunciated here will apply equally well for those whose work is being typeset by a different method. The user of Microsoft Word, for example, will not himself implement (as would a TEX user) the kerning and formatting and page design commands which we discuss in Chapters 2 and 3, but he will communicate with the typesetter about those commands. He will not be able to format equations with the level of precision and detail that we describe, but he will (after reading this book) be equipped with the vocabulary and skills to tell the typesetter what he wants. He will not be importing an encapsulated PostScript figure into his document in just the manner that we lay out, but he will learn the process and thus be able to ensure that his book or tract comes out in the form desired. There is no point to mince words. We believe strongly, and we are certainly not alone in this belief, that the medium of choice for producing a mathematical document today is TEX. Most mathematicians use TEX, most publishers use TEX, and most Web sites are set up either to handle TEX documents or to handle files produced from TEX code. TEX code is more portable than the files from any word processor, and its output is of vastly higher quality than the output from any other system available. With suitable plug-ins, TEX can handle graphics beautifully. There is no formatting problem that TEX cannot handle.2 If you send your work to a journal or a publisher in any electronic format other than TEX (or one of the variants of TEX), then you are only inviting trouble and, in some cases, derision. We hope that this book will serve to convince you of the correctness of all these assertions. Prerequisites for reading this book are a knowledge of the elements of mathematical writing (for which see, among other sources, A Primer of Mathematical Writing by this author) and an interest in mathematical typesetting and graphics issues. We certainly do not assume that the reader is an active user of TEX. We include a brief description of TEX and its most basic commands, just because TEX is so much a part of mathematical life today; and also because it is easier to describe some
2 Consider

this typesetting problem: You have an expression that consists of a 3 × 3 matrix divided by an integral, and you want to typeset it in displayed fashion as a fraction. TEX can perform this task beautifully and easily, with simple and sure commands. Your word processor cannot, nor can any other document preparation system that is available as of this writing. A similar remark applies to commutative diagrams, to tables, and to many other high-level typesetting tasks.

©2001 CRC Press LLC

of the typesetting procedures that are essential to mathematics if one can make reference to TEX. In Appendices III and VII we include a compendium of all the TEX commands that are most commonly used in mathematical writing. Appendix VIII contains a sample of TEX code together with the compiled output. The reader who spends some time with the present book will certainly come away with considerable motivation for learning more about TEX. Likewise, we do not assume that the reader is conversant with the tools for providing his book or manuscript with graphics—such as PostScript or bitmap (*.bmp) files or *.jpg files or *.gif or *.pdf files or PICTEX. Instead, we hope to acquaint the reader with these and some of the other graphics options that are available to the mathematical author. The tools that are now available for creating the index, the table of contents, the list of figures, the bibliography, tables, diagrams, and other writing elements are both powerful and marvelous. We wish to compile here a resource for the author who wants to take control of these portions of the creation of a book or document. Finally, many a book author today will want his book to contain a computer diskette, or a reference to software that is available on the World Wide Web, or source code for software. We will discuss issues attendant to this part of mathematical writing, and offer some solutions as well. The reader who becomes acquainted with the present work will be a well-informed author who is equipped to deal with publishers, compositors, editors and typesetters, with TEX consultants, with copy editors, and with graphics designers of every sort. He may not be tempted to perform the various typesetting and formatting and graphical tasks himself, but he will be prepared to communicate with those who do. It is our hope that the result will be an author who has a better understanding of the publishing process, and one who will want to and be able to create better mathematics books. Steven G. Krantz St. Louis, Missouri

©2001 CRC Press LLC

thanks. His friendship and patience are much appreciated. my friend and collaborator Stanley Sawyer has taught me a great deal about TEX and about typesetting. has been encouraging and helpful in every aspect of the production of this book. and to offer innumerable suggestions. He helped me interpret the reviewers’ remarks and keep this book on track. Bob Stern. CRC Press has engaged several reviewers to help me hone this book into the precise and accurate tool that it should be. George Kamberov has been a great resource in helping me learn how to include graphics in a TEX document. To all of them I express my indebtedness.ACKNOWLEDGEMENTS Over the years. For all of his help and suggestions I am grateful. As always. ©2001 CRC Press LLC . My editor. He was also good enough to read many drafts of the manuscript for this book.

definitions. and so forth are all numbered automatically. See the sample code in Appendix VIII.09) or document classes (in L TEX2ε). Once you have made this choice. A large part of the craft is the ability to harness one’s thoughts and to organize them into sentences. paragraphs. and of many other aspects of the book or document as it will finally appear.1 An Overview The ability to write well is not a gift from the heavens. but also planning the form of the document. propositions. subsections. ©2001 CRC Press LLC .chapter 1 Basic Principles 1. Your sections. Today’s technical writer is also involved in the physical process of putting the words on the page. Every reference to an equation or theorem or other enunciation will be linked to that theorem or equation. Among dialects of TEX. and chapters. As an example. A L TEX is often preferred by publishers because it gives authors fewer choices. This assertion is as true for technical writing as it is for prose and poetry. the layout of section and chapter titles. many typesetting and page design decisions are automatically made for you. examples. Word processing systems like Microsoft Word and computer typesetting systems like TEX put the writer in charge of the composition of the page. theorems. It is a craft that is honed and developed over time. L TEX (a dialect of TEX invented by Leslie Lamport)1 allows you to choose one of several different pre-formatted document A A styles (in L TEX 2. so that all of your 1 See A Chapter 2 for more information about TEX and L TEX. There are some easy choices A that can be made. you can choose the book document style or document class by entering \documentstyle{book} or \documentclass{book} as the first line of your TEX source code file. equations. Thus the creation of a document in the modern writing environment involves not only the traditional process of organizing one’s thoughts. the design of running heads. and because it is a structured language. the choice of fonts.

the publisher will send you a TEX or L TEX style file that will determine the style of the finished product. This file will be “called” at the top2 of your document 2 The A “top” of your L TEX source code file—the part that occurs after the \documentstyle or \documentclass command and before the line A \begin{document}—is called the “preamble”. In addition. your use of specialized terms and jargon. displays. your use of language. You will be able to refer to equations A and theorems and other items by nickname. ©2001 CRC Press LLC . If you want to interact on an informed and intelligent level with that copy editor. and varioref.references will always be accurate. few publishers will simply accept what the author sends in. and L TEX will number them automatically. A There are many other packages provided with L TEX2ε.sty gives the choice of several languages in which the document may be typeset. kerning.sty provides a high-level user interface to the L TEX picture environment. that you are using L TEX2ε and that the style file is called publstyle. graphics. Few publishers will publish a manuscript in that form. Others must occur after the preamble. A copy editor will go through the manuscript and analyze line breaks. for defA initeness. and more that you can download from the Web (see Appendix VI). As their titles suggest. • multicol.sty. \epic.sty. Some of the packages that may be invoked are babel. Suppose.sty. and typesetters. L TEX will still get the numbering right.sty provides powerful cross-referencing capabilities. and the precision of your English. should you move one of these items to a different part of A the document. choice of fonts.sty. It is A much more powerful than the multi-column format of L TEX 2. • babel. page breaks.sty enables several different multi-column formats. and if you want your book to come out the way that you want it. technical writers. A In many cases. Various special L TEX commands must be placed in the preamble. then it behooves you to learn about these and many other features of the typesetting process. A Please note that non-customized (out-of-the-box) L TEX styles look grey and dull to many experienced editors. A • epic. multicol. running heads. and many other aspects of the manuscript that you might have thought was a finished product. He will think about your syntax. if for no other reason than that each publisher wants his or her books to have a distinctive look. • varioref. A L TEX2ε further allows you to customize your stylistic choice with the \usepackage command.09. spacing.sty.

The publstyle.). I urge you to let your publisher design the pages of your book or paper. But it is simpler to typeset. the \input command may be used in a style analogous to A \usepackage in L TEX2ε.09. there are other issues at play. and one that is special to the writing of mathematics. . . from the point of view of typography and clarity. j2 . and certainly easier to read. . the first two lines will be \documentclass{book} \usepackage{publstyle} Roughly speaking. In A L TEX 2. and table of contents.. the publisher’s style files will specify the size of the print block on the page. . It will format enunciations (theorems. displayed equations. But. Typically. it is quite likely that your publisher will modify or veto whatever page style you may create. the group and the topological space are one and the same entity). it will also format the bibliography. Unless you want to spend a great many hours mastering the principles of page design.sty file invoked in the second line will introduce refined stylistic features. In L TEX2ε. it is very natural for a mathematician to use Ej1 . etc. jk .cls file invoked in the first line makes broad stylistic choices that will apply to most any book..source or root file (or the first line of each source file if you have different A chapters in different self-contained files). Even if you think you know how to do it. Even the unseasoned mathematical writer will know that notation should be consistent throughout a given work. but it is considerably less powerful. ©2001 CRC Press LLC . This dictum sometimes runs counter to the way that a mathematician thinks. A well-designed page is a delight to behold (and by no means an easy thing to create). definitions. . and select fonts for all these features. the book. 1. Where suitable. Strive for simplicity and elegance in your notation. the expression E(j1 .. j2 . is choosing your notation.2 Choice of Notation An important part of planning a document. . E) that depends on indices j1 . . One should not use the symbol E to stand for a group in Section 2 and for a topological space in Section 7 (unless. jk ). . index. fix the formatting of chapter and section and subsection titles. fix the margins. of course.jk to denote an entity (namely.j2 . and other features in your writing. For example. It will sometimes add graphical elements like horizontal lines to set off titles..

im .i Ej1 . They can cause trouble with between-line spacing. Some of the stacked accent problems have A been fixed in L TEX2ε.j2 . In a related vein. And. Rather than typeset A+B+C +D (note that the is far too short) one can typeset A+B+C +D A (L TEX2ε has special commands for handling this formatting situation). . in Fourier analysis the embellishment is used to denote the Fourier transform. if you wish. i2 . jk ) in preference to i1 ....The meaning is unchanged. you could typeset F A+B+C +D . you can blame the software TEX for this clumsiness. You can see that the tilde is not properly centered over the f . im . and perhaps more convincing. . . and adjusting the TEX code to obtain the desired result ˜ f is a tedious business. ©2001 CRC Press LLC . but the conveyor of the information is simpler.. Perhaps even better. 2 .jk It is best to avoid stacked accents.. . and more attractive. example is to use the notation E(i1 . to use f to denote the Fourier transform—especially if the argument is a large expression. it is not always convenient.. . But there it is. If one wishes to compose these two operations (as in the standard proof of Plancherel’s theorem—see [KRA2]). j2 . And. A more extreme. Then the composition becomes IFf . .. . It is simpler. to introduce the notation If (x) = f (−x) and then to write I(f ) for the composition. j1 . and need not be used. For example. One can go further and introduce Ff to stand for f .. nor aesthetically pleasing. then the symbol f is defined by ˜ f (x) = f (−x). then one might be tempted to typeset (this is the default in TEX): ˜ f. if f is a function on R. They are difficult to typeset. difficult to center properly. .

But it is much clearer. or bird will either need to learn METAFONT (in order to manufacture his own) or may be able to find a third-party font which includes these. as well as f and f . You want either A/(B +C) or (A/B) + C. a c It is critical when you set a fraction in shilling form to use sufficient parentheses so that the meaning is unambiguous. unlike the overbar (given by \overline). ©2001 CRC Press LLC . It should be noted that. If F is a function of the matrix argument ab cd then one might be tempted to typeset F ab cd . elegance. and clarity. An expression like A/B +C has two distinct possible meanings. These are given by the TEX command pairs \hat{f}. There are serious typesetting problems connected with both the shape of and the positioning of hats and tildes. Sometimes we do things on the computer just because we can. A similar comment applies to the so-called “inverse Fourier transform”. \widehat{f} and \tilde{f}. to typeset e(a/b)+(c/d) with the fractions typeset in shilling form. the Fourier hat and the tilde are not infinitely extendable. Better still is exp (a/b) + (c/d) (note the use of parentheses () to help parse the fractions) or exp c a + b d . \widetilde{f}. All notation should strive for simplicity. and less prone to error. .The experienced TEX user will know that the Fourier transform’s hat ˆ ˜ and also the tilde come in two forms: there is f and f . tilde. It is fun to enter the code e^{{a \over b} + {c \over d}} and then to see the compiled outcome eb +d . or “bird” (the symbol ˇ given by \check). The user who needs a wide hat.

But it is much clearer to write Let A= and consider F (A) . For example. never write f (x) is a positive function on the interval [a. or The function f is positive and has domain the interval [a. b]. For the second example. for instance) with notation. See Appendix V for a treatment of various typesetting choices. A final example to round out the message being promulgated here: It may be natural to typeset an expression like 2ab a + b2 . and it is great fun to see the huge square root sign make its appearance. b]. Such usage will cause the reader to do a double-take. or G is a group of finite order. it is preferable to say ©2001 CRC Press LLC . This may be the moment to note that you should never begin a sentence or a phrase (following a comma. c + sin d a2 + b2 In some displayed equations. this format will probably serve. it will only serve to obfuscate your prose. to render the formula as a + b2 c + sin d or even a + b2 c + sin d 1/2 ab cd · 2ab a2 + b2 2ab + b2 1/2 · a2 . Better is Let f be a positive function on the interval [a. and less intrusive on the neighboring material. But in many contexts it is clearer. b].

Observe that. in text). a piece of mathematics is said to be “displayed” if it occurs by itself. Often. 3 Here ©2001 CRC Press LLC .3 If the expression is more than one fourth of a line long. which means that it is part of the regular flow of text. If it is not complex and if it occurs near the beginning of a paragraph (so that you do not have to worry about line breaks). is whether to display a mathematical expression or leave it inline (i. • it would be hard to read if placed inline. or perhaps The group G has finite order. then you should display it so that it does not interfere with the lines above and below. which the mathematical author must make frequently and effectively. in the first example. and is centered left-to-right. separated vertically from the text before and after.We use the notation G to denote a group of finite order. we have corrected another common error along the way.. or if these embellishments are compounded. If it involves more than a few subscripts or superscripts. displayed text will be larger—or at least appear larger—than text in the main body of text. In the end these decisions must be made by the author—based on meaning. then it is typeset “inline”. Higham [HIG] tells us that we should display a mathematical expression when • it needs to be numbered.e. and therefore better to have too many displays rather than too few. • it merits special attention (perhaps because it contains the first occurrence of an important mathematical entity). The character f is the name of the function and f (x) is the value of that function. In the present context. it is correct to use “f ” rather than “f (x)”. then you may leave it inline. If it involves integrals which are not completely elementary. 1.3 Displaying Mathematics An important decision. But try to avoid having so many displays that the mathematical exposition becomes harder to follow. If mathematics is not displayed. These are excellent guidelines for deciding when to display an item. then you should usually display it. It is always better to err on the side of clarity. then you should display it. or any kind of matrices. Mathematical writing expert N. J. but not always.

then you can invoke the macro with the code might be mentioned that the command \displaystyle allows you to format mathematics in display size and format while you are in inline mode. where more subtle issues of displaying mathematics are discussed. consistency. the natural way to execute the planning of the form of a book or document (at least when you are using TEX) is to write some macros. This command accepts two arguments (designated by #1 and #2). the similar command \textstyle allows inline math formatting in display mode.4 Consistency In fact the planning to which we alluded in the first and second sections can lead to some attractive. E 4 It ©2001 CRC Press LLC . Here we merely give a couple of simple examples. The way of TEX is to replace every occurrence of \smsqr with the code \sqrt{{#1}^2 + {#2}^2} + \frac{1}{{#1}^2 + {#2}^2}. It also makes sense to create macros for complicated mathematical expressions that will be used repeatedly. The aggregate of commands is then executed by simply invoking the macro name. x2 + y 2 A second example is a macro for theorems: \def\theorem#1{{\sc Theorem} \ #1: } This is a macro with a single argument (designated by #1). (See [SAK] for the full picture of macro creation. running heads. and very desirable. and appearance. the line of TEX code5 \def\smsqr#1#2{\sqrt{{#1}^2 + {#2}^2} + \frac{1}{{#1}^2 + {#2}^2}} creates a new TEX command called \smsqr.context. If you place this line of code at the top of your TEX document. which will usually be the number of the theorem. formatting of theorems. etc.5. entering the line \smsqr{x}{y} will result in the typeset expression x2 + y 2 + 1 . chapter headings. Refer also to Section 2.) As an instance. subsection heads.4 1. As we will explain in Chapters 2 and 3. formatting of displayed equations. The creation of a macro is simply the assigning of a name to a collection of commands. Thus one creates macros for section heads. 5 Refer to Appendix VII for the meanings of the different T X commands. With this macro definition at the top of your TEX file.

There is a function f such that . indentation. but awareness of the problem is also a great asset. then many of the design decisions will already have been made. spacing.5: Let > 0. or the Halmos “tombstone” Again. a good macro can be invaluable in addressing these issues. Let [a. The nice thing about this macro is that now all your “theorem” declarations will be set in the same font and will have the colon included. 8 Here an orphan is a brief phrase—less than a full line—that occurs at the bottom of a page and that begins a new paragraph. There are other aspects of consistency about which many authors are blissfully unaware: spacing above and below a displayed equation. 1. a check for consistency. and we will use that terminology as well.5} Let $\epsilon > 0$. we use the word “theorem” to denote any displayed body of text. Almost all of these corrections concerned page breaks. including proposition or lemma or sublemma or definition or remark or any number of other similar constructions. b] ⊆ R be an interval. you need not remember. There is a function $f$ such that \dots The result will be the typeset text Theorem 3. for example). or at least should only be done with input from professionals. It is common among typesetters to call these “enunciations”. there are issues of which you should be aware. Even if you are using style macros provided by the publisher. If the publisher provides a set of style macros that you can include at the top of your TEX file (see Section 1.b] \subseteq \RR$ be an interval. . By using macros. The practice of using macros not only saves time and eliminates aggravation. nor perform. spacing above and below a theorem. and other features of your composition are maintained uniformly throughout the document.5 Overall Design Proper page design requires an artistic sense and experience that most mathematicians and other technical writers may not have.8 but this is sometimes 6 Here. For this reason. TEX will not align the bottom lines of facing pages in a book unless the style macros do this explicitly.6 space after a proof.7 TEX can be told to prohibit orphans and widows. the mark at . the end of a proof (QED. page design is best left to professionals. Let $[a.1). but it also serves to promote accuracy. . and throughout this book. ©2001 CRC Press LLC .\theorem{3. A widow is a brief phrase—less than a full line—that occurs at the top of a page and that ends the preceding paragraph. you will guarantee that size. 7 The two pages of production notes in the classic [GMS] discuss the rather clever and technical means by which these three very skilled authors had to hand-correct A decisions made by L TEX.

Or they format pages in two or three columns. Worse. this “someone” will be the author. If you are guilty of freestyling. Usually it is better to eliminate the boxes by learning to be a better TEXnical typesetter. TEX may leave too much space between words or not enough.) Likewise. The publisher will then provide a TEX style file or (the oldfashioned approach) a hard-copy style manual. It is best if the author is the one to deal with a problem such as this. or else make sure that your style macros are well-designed and very well-documented. or AMSA L TEX). and then the typesetter getting things wrong anyway. TEX can be clumsy with line breaks. it may run part of a word or formula out into the margin. At the risk of belaboring this last point. At some stage it will be necessary for someone to go through the manuscript by hand and correct problems of this kind. your eventual publishers may accuse you of “freestyling”. Otherwise. AMS-TEX. then it can be extremely difficult for either the publisher or yourself to translate your work into publishable form.not done because it can cause problems elsewhere. which suggests a TEX file with rigid or non-uniform stylistic choices and TEX source that is incomprehensible except perhaps to you. If TEX gets jammed up trying to fit just the right number of words into a line. 10 George Bernard Shaw tells of rewriting sentences or passages in his plays in order to satisfy the typesetter. You certainly do not want someone else rewriting your material just to make the typesetting come out right!10 As we have suggested. it is easy to see that such a rigid chunk of typeset text can also be run out into a margin—resulting in an unsightly paragraph and page. L TEX. let us mention a few examples. A chapter should not end on a page with fewer than four lines from that chapter. begin with a dropped gothic capital letter. things become a bit easier once you have a publisher. Some authors go to great pains to have each chapter. or each section. You should either use the simplest choices A in a standard TEX dialect (like Plain TEX. Or they equip the left-hand pages with a running 9 The habit of TEX is to place a black box in the margin next to a line that has not been typeset according to TEX’s standards. and thereby eliminate some of the black boxes. ©2001 CRC Press LLC . then you should avoid the temptation to be seduced by TEX (or whatever document preparation system you are using) into writing an overly complex source file. A section heading should be followed by at least two lines of text of the new section on the same page.9 TEX treats certain formulae in math mode as unbreakable units. then it may break a line in an inauspicious place—in fact. because it is sometimes necessary to rephrase or rewrite a sentence or paragraph to make things come out as they should. or if it encounters an unfamiliar word that it cannot figure out how to hyphenate. You can adjust or weaken the standards. If you do not yet have a publisher. (With many publishers.

Or they use very exotic fonts.head and the right-hand pages with a running foot. and for every other user of electronic tools. but no publisher will want to preserve and to propagate these whimsies. This advice is valid for the Plain TEX user. You make everyone’s life easier if you eschew the eccentric and stick to the most basic conA structions. All very charming. Or they may place the page numbers (the folios) in strange places. for the Microsoft Word user. ©2001 CRC Press LLC . set in peculiar fonts. for the L TEX user.

It is used by individuals to write letters and articles with a personal printer as well as by technical publishers to typeset mathematics books on a Varityper 6000 or a Linotronic 630 or other electronic typesetting device. and so that it will travel well to whatever document preparation system you are using.chapter 2 Typesetting Mathematics 2. As you read along. and VIII. What we see is what we expect to see: there is a smooth transition from what is printed on the page to what the eye sees to the cognitive processes of the brain. we discuss such issues. which are designed to speed along your acquaintance with TEX. Since being invented and introduced by Donald Knuth in the early 1980s.1 What is TEX? TEX is a computer package for producing document output of typeset quality. When one is creating typeset mathematics. In this chapter. TEX is particularly powerful and useful for preparing manuscripts with mathematical formulas and tables. so as best to effect the attractive process just described. ©2001 CRC Press LLC . While we will often invoke TEX in this chapter.0 Introductory Remarks When mathematical formatting works well. you will want to refer to Appendices III. one must be aware of the issues of spacing and positioning that control the clarity and meaning of mathematical expressions. In particular. one must be aware of the components of good typesetting. VII. Certainly this book will tell you enough about TEX so that you will find everything that we discuss comprehensible. 2. such invocations are generally self-contained. we are hardly aware of it. TEX is a computer typesetting language—a high-level computer programming language.

the TEX code mandates exactly how you want the document to look. The screen provides an approximation to this truth. In effect. With the help of its markup commands. Viewing output as it might be printed is a separate step. as we described in Chapter 1. or else how to communicate with a copy editor and/or typesetter so that he or she can deal with them. TEX has A a number of dialects. T X has been used by law offices and even by the periodical TV Guide. The E portability—in the large and in the small—of TEX make it a powerful and versatile tool. Currently laser printers have around 1200 dots per inch or around 10. ©2001 CRC Press LLC . TEX makes most of these decisions itself. The most common dialects are L TEX. TEX positions elements on the printed page to within 10−6 of an inch. This has the advantage that your computer source file is independent of improvements to your computer hardware. Plain TEX. The superb degree of accuracy of which TEX is capable should protect TEX from obsolescence for many decades to come. Fortunately. 1 In fact. but this is usually not advisable unless you have a considerable amount of experience with either typesetting or TEX. You can use TEX commands to customize these decisions.1 Part of the reason for TEX’s long life and wide use is that it implements a “markup language” instead of creating output on a computer screen as you enter it. the printer will give a hard-copy rendition of the ideal document dictated by the TEX code. The computer file that you write for a manuscript is a text or ASCII file that contains commands that describe in more-orless plain English terms how you want the pages formatted and what mathematical symbols you want to use.it is now almost universally used by mathematicians to typeset mathematics. Computer installations that support one of these dialects will generally support all four. A AMS-TEX. So you can see that. screens display material to a different tolerance than the printer will print it. One of the main purposes of this book is to explain what those situations are and how to deal with them. with Plain TEX being slightly more divergent. usually suitable for accuracy checking. However. you do not have to decide yourself how to position each character on the printed page to within a millionth of an inch. These are roughly as close as American English and English English. while most current computer monitors go up to at most 1200 or 1600 horizontal dots per screen. and AMS-L TEX. and is also widely used in areas of science and engineering that emphasize mathematics. there are a number of situations where human intervention is necessary for highest-quality output. Like many languages that have been around for some time. from the outset.200 horizontal dots per page. Depending on the quality of your printing device.

just as they would be with a typewriter or a word processor. (2) Photoelectric typesetting: The typesetter works from a keyboard. Some people. presumably in analogy with the word “latex” in “latex gloves”. and that rack serves as the initial printing plate. \kern-. The tape is then used—as in a Turing machine—to operate either a phototypesetting or metal typesetting machine. enjoy saying lah -tek. however.2in A more complete list of TEX commands appears in Appendix VII. or other means.35cm . the first command is for forming a matrix. the third is a vertical space of size . \bigtriangleup . see also Appendices III and VIII. in which the output is on a stream of paper tape (known in the industry as “idiot tape”).35 centimeters. Some examples are2 \alpha . and spaces are actualized in pieces of lead. diphthongs. calligraphy. Characters. the second is for paragraph indentation. These lead slugs are positioned in a rack. \parindent . It is photographed. and his output is actualized on film. (4) Camera-ready copy: The manuscript is produced by typewriter. there are basically five types of mathematical typesetting: (1) Cold type typesetting: This is the most traditional form of the art. Ordinary words are entered as text. A The most common pronunciation of L TEX is lay -tek. used for centuries. From that piece of film. and the fourth is a large triangular symbol ( ) like the Laplace operator.2 inches to the left. On the second line. and the third moves the point of typesetting 0. and becomes the printing plate. \vskip. In the remainder of this chapter we shall provide a more detailed picture of TEX. accents. Many purists pronounce the final “X” like the “ch” sound in German or Scottish. an image is produced on a copper or zinc plate. word processor.A note about pronunciation: TEX is pronounced as in “technical. 2 The ©2001 CRC Press LLC . The plate is etched. \matrix . first command typesets the Greek letter α. 2. the second controls the size of a left delimiter.” not as in “Texas”.2 Methods of Typesetting Mathematics In today’s world. but a simple “k” sound is more than sufficient. (3) Paper tape typesetting: This is a variant of (2). A command or special symbol in TEX is typically an alphanumeric word preceded by the symbol \ (the “backslash”). \biggl .

3 A Lightning Tour of TEX When you create a document with a word processor. histograms. First of all. and becomes the printing plate. chess problems. tables. TEX has a decided advantage over word processors. The source file is compiled and a *. and import multiple fonts and typeset them in a bewildering array of sizes and positions.dvi file is translated to PostScript. With TEX. See Chapter 7 and Appendix VI for details on Web resources. alignment at the right margin) if desired. The plate is then etched. TEX will give you more control over positioning and sizing than will any other document preparation system. you can integrate mathematics3 into your work and have complete control over the outcome. most important for the mathematician. perhaps with right justification (i. (5) is becoming the most prevalent. musical scores. most word processors are essentially WYSIWYG (“What you see is what you get. and many other challenging typesetting tasks. From the point of view of portability.dvi file produced.”). This book will concentrate on ideas and techniques connected to (5). It gives you microscopic control over formatting. graphs. The film is then used to produce an image on a copper or zinc plate. TEX is more sophisticated. You also can integrate and manipulate graphics in TEX in a variety of useful and convenient formats. The plate is etched and becomes the printing plate. 3 You ©2001 CRC Press LLC .and then the photographic image reproduced on a copper or zinc plate. TEX is virtually platform-independent. flow text around a graphic or chart. On the positive side.. the code then appears on a CRT or computer screen. But. The PostScript file is used to make an image of each page on film. Of these five techniques for mathematical typesetting. as in other high-level word processors. as well as chemical charts. although individual text characters or graphics may be printed at a higher pixel resolution. The word processor formats your text as entered. Most word processors have poor capabilities for typesetting mathematics. (5) TEX: The work is produced in TEX computer code. you enter text and perhaps mathematical formulas (using suitable syntax) from a computer keyboard. 2.e. The *. can download utilities from the Web to tackle a variety of very complex diagrams (mathematical commutative diagrams and others). Whatever you see on the screen is more-or-less what is printed. Feynman diagrams. you can print your page in multiple columns. bar charts. TEX has more fonts and (with METAFONT) the ability to create fonts without limit.

He must also learn the concepts that underpin the basic logic by which TEX operates. you will usually find it more convenient to give your source code file the three letter extension .1 A Lightning Tour of TEX You can enter your TEX code (into an ASCII file called myfile.tex is called the TEX source file.tex → myfile. You might entitle the source code file myfile.3   directly to printer    directly to computer screen     myfile. Many dedicated TEX users use one of the standard programmer’s editors: emacs or vi on a UNIX system.2. compile it on a Macintosh. or epsilon or crisp on a PC.pdf (Acrobat file) Figure 2.tex as analogous to the source code file for 4 This file should be a straight text or ASCII file. the file should be saved as either “Text Only” or “Text Only with Line Breaks”.4 The file myfile.tex. But it is portable in another important sense: if you set up a complicated display in TeX. We begin by diagraming the logic of creating a TEX document. An effective TEX user must learn to read many TEX commands and TEX code contexts and understand what they say and what effect they have on the printed output. Because a word processor has hidden formatting commands. A software engineer might view myfile.ps (a PostScript file for     various output devices)    myfile.tex in its name. and contains TEX commands and unformatted text as you have entered it. then it will write plain ASCII output files by design and you need not worry.tex is the file that you write in a text editor or word processor. and the fundamental commands that select a typesetting operation and then implement it. Although it is not mandatory. ©2001 CRC Press LLC .tex.dvi →   myfile. cutting and pasting often corrupts the formatting of text. In Microsoft Word.hp or other file that can  be understood by your printer myfile. Other word processors might require “non-document mode” or “ASCII with carriage returns and line feeds. for example) on a PDP-11.” If you are using a text editor that is part of an integrated TEX package. This sort of “translation-invariance” does not obtain with word processors. and you must waste time fiddling around and re-formatting. and preview it on a Pentium machine. Here myfile. That will never happen with TEX. Suppose that you are creating a business report. then you can copy or transfer that piece of TEX code to another part of your document or to another document altogether and the result (in the compiled output) will be unchanged.

and other formatting specifications.1). Some TEX preview programs display smaller characters more coarsely to make them easier to read.” and is not used if that part of the you are a real hacker. The first arrow in the diagram represents the “compile” operation.5 These steps taken together are generally easier than porting an entire integrated wordprocessing package to a new machine.dvi file specifies positions on the printed page within 10−6 of an inch (in particular.dvi file. display it on a DEC Alpha.1. page breaks. Unlike a word processor.) Current printers generally have much finer resolution than current display devices (see Section 2.dvi) works on your new computer. TEX does not emulate the ultimate page of output. it thinks of each character or typesetting element as occupying a box. and is usually called a *. then you can actually obtain the C and Pascal source code for TEX and compile it on any machine you like. The second arrow in Figure 2. The binary codes in this file resemble codes that might be sent to a printer or computer screen.dvi file (this latter is the binary file that carries all the machine-coded information about your document). This is sometimes called “hinting.tex to myfile.dvi file for short. instead. but are in a language that is independent of any specific printer or computer. which converts the source code file to a *.tex. (Printing or screen display refers to the typeset output. from a commercial vendor or from the Web. In particular. that the TEX program (that converts myfile. you merely must make sure that you have a text editor. The information in the *. and that you have screen and printer programs that can display *.a FORTRAN or C or Java computer program. what fonts should be used. TEX issues formatting commands that specify where and how the page elements will appear. Such a word processor will certainly be platform-dependent. Software is available for most all modern (personal) computers to interpret TEX *. and records the dimensions and position of each such box). for whatever machine we may wish to use. 5 If ©2001 CRC Press LLC . Most of us will instead obtain a suitable version of TEX. Appendix VI discusses Web sites that offer public domain versions of TEX for a variety of platforms. and then print it on an Intel Pentium machine.dvi). In this sense. If you buy a new computer or upgrade your computer hardware.dvi files. you can create myfile. Most screen preview programs have magnification features to make it easier to see fine detail. the compiled file is “device independent” (hence the name myfile. This platform universality is a second reason for the long life of TEX. between-line and between-paragraph spacing.dvi files. and its data files will not travel well. and also has information that determines line breaks. Refer to Figure 2. not the literal TEX commands and text in myfile.1 represents some of the things that you can do with a *.dvi on a Macintosh or UNIX computer.

You begin a new paragraph by leaving a blank line in your source code. displayed or printed PDF or PostScript files can be noticeably cruder if they are displayed or printed at a different resolution than that of their embedded font information (unless the user is smart enough to use scalable fonts). There are some TEX programs that write output files directly in Adobe PDF format instead of *. with any number of words on a line and any number of spaces between words.dvi file only records a sequence of boxes). See references such as [SAK]. This makes the output more attractive and closer to what would be printed. [KNU]. 6 Refer to Appendix VI for some URLs where PostScript resources can be obtained. while *. It will format the text into paragraphs in a highly professional way. This has the advantage that screen display and printer programs do not need access to font information in auxiliary files. but also makes smaller characters harder to read. This is meant as part of a brief overview of what can go into a *. TEX will collapse an positive number of spaces into a single space and will calculate the correct number of words to put on a line. TEX will understand that all the material appearing in an aggregate clump—without vertical spacing—is a paragraph. However. then the sentence The cat with a bat saw a rat take a hat off the mat. will appear in your output—either on screen or on the printed page. just as with typewriter or traditional word processor. Text: Literal text that you want to appear in your printed output. For example. VII.dvi files to either Adobe PostScript or PDF format.6 2. There are also widely-used computer programs that convert *. Other preview programs do not use hinting.dvi files do not (the *. This is entered directly into the source file. No matter how you type it. if you enter the text The cat with a bat saw a rat take a hat off the mat. Also refer to Appendices III.dvi format.4 The Guts of TEX In this section. and will also hyphenate words if need be. or [GMS] for more detail.tex file. ©2001 CRC Press LLC . One advantage of PDF and PostScript files is that they generally contain complete font information. [LAM]. we discuss some of the types of commands and text that can go into a typical TEX source file. and VIII.screen output is sufficiently magnified.

As the name suggests. but some TEX implementations 7 A useful font for some applications is the “sans serif” style of font.Fonts: A font is a style of type. On a traditional typewriter. but other units are sometimes used. Some of the standard fonts are roman. For example. The Glossary defines a number of different typesetter’s units of measure. In TEX. The “weight” of a character describes the thickness of the strokes that make up the character. TEX is a full-featured typesetting system and each character comes equipped with complete information as to how it should be typeset against every other character. the font is “monospaced. The roman character “I” has two serifs: one at the top and one at the bottom. has more weight) than a roman b. These are normally from the Computer Modern font family in TEX. the carriage moves by a pre-determined amount of space. The roman character “R” has three serifs. Certainly one of the characteristic features of TEX is that it is not monospaced. Some people like to use sans serif in business correspondence. If you think about how a typewriter mechanism works then monospacing makes a lot of sense: after you strike each key. A serif on a character is a small embellishment (a perpendicular cross) at the end of one of the main strokes that composes the character. it is kerned.. The phrase This is a sans serif font is set in sans serif. We sometimes use sans serif for abstracts. a boldface b is composed of thicker strokes (i. and italic.e. and in other typesetting situations.” This means in effect that each character occupies the same amount of horizontal space. the word “font” can also refer to the size of the type (measured in points). The size of a font is generally specified in points (where a typesetter’s 1 “point” is about 72 of an inch). and displayed statements.7 The attributes “shape” and “width” are self-explanatory. the characters in such a font have no serifs. ©2001 CRC Press LLC . According to [GMS]. But genuine typesetting is not monospaced. headings. This means that the way space is set around a given character depends both on that character and on what characters are its neighbors. So of course each character (from “i” to “m” to “w”) must have the same width. a font has five attributes: • monospaced or proportionately spaced • serif or sans serif • shape • weight • width Perhaps a word of explanation is in order. boldface.

you can use the commands \it. to typeset a phrase in (for example) boldface.09. outline fonts. slanted roman. Knuth at the time that he created TEX. The default font in TEX is (Computer Modern) roman. A Of course there are analogous commands for italic (instead of the L TEX A 2. Other common TEX fonts are big cap-small cap. PostScript fonts. and typewriter-like fonts. We again refer the reader to [GMS] for the details of these subtle and powerful commands. \sl. METAFONT fonts. this is boldface text. Today there are many hundreds of fonts available for TEX. and typewriter-like.09 commands. and \tt (inside matched braces) to typeset (respectively) text in italics. \textbf{this is boldface} text. \sc. and in all versions of TEX. big cap-small cap. slanted roman. isolate the phrase with braces and put the command \bf just after the opening brace. The books [SAK] and [GMS] contain a thorough discussion of fonts. there are advantages to using the new command (to obtain boldface. for example) As you can see. These are contained in the font cmmi10. The output is of course the same. Thus the TEX code As you can see. {\bf this is boldface} text.) Most of the words in this paragraph are typeset in roman.use Times Roman or Lucida or other popular fonts. The new \textbf command works in both text and math modes. Although L TEX2ε will recognize the L TEX 2. ©2001 CRC Press LLC . and other types of fonts. A sample of such numerals is . and AMS-TEX now has distinct commands for boldface in text mode and boldface in math mode. A you must keep in mind that (among other things). (The basic Computer Modern fonts were created by D. A L TEX2ε has a more sophisticated protocol for invoking fonts other A A than the default roman. This is the font that TEX uses in text unless you enter TEX A commands that mandate otherwise. A The even more sophisticated L TEX2ε command \bfseries calls in not just one font but a whole series of fonts. In L TEX 2. produces the output As you can see. Similarly. and the A more sophisticated L TEX2ε command even takes into account the special spacing requirements that the built-in slant of italic characters requires.09 command \it one uses the L TEX2ε command \textit). In order to keep the logic straight. It has other sophisticated capabilities which are best learned by reading [GMS]. L TEX2ε is a superset of AMS-TEX. These include bitmap fonts. A curiosity in TEX is that it offers “old style” numerals for specialized applications.

It is the formatting that is difficult. Type a terminal $ when you are done. $$ That says it. There are two slightly different mathematics environments: inline and display. ©2001 CRC Press LLC . type $$ and then begin your mathematics. The effect to TEX is the same. For instance. learning the commands for different mathematical symbols is easy. We will treat only the most elementary formatting questions here. To enter the display mathematics environment.dt$. That says it. produces output with the formula in display math mode: Let y = x2 + 2 + 1 x t sin t dt. it uses a larger integral sign in the integral and a less-cramped style for upper and lower limits on the integral. the TEX code Let $y = x^2 + 2 + \int_1^x t\sin t\. Since TEX has more room (vertical and horizontal space) in display mode. t sin t dt. the TEX code Let $$y = x^2 + 2 + \int_1^x t\sin t\. with no special commands. it is better to enter code for display math as Let $$ y = x^2 + 2 + \int_1^x t\sin t\. type $ and then begin your mathematics. That says it. The text environment is the default: just start entering ASCII code. produces the typeset output Let y = x2 + 2 + In contrast. Although you may never have explicitly noted the fact before. spacing in text is done differently from spacing in mathematics. To enter the inline mathematics environment.dt. and you are in the text environment. x 1 That says it. Thus TEX has two environments: the text environment and the mathematics environment. To keep your source file from getting too cramped. but the source file will be much easier to read.$$ That says it. Type a terminal $$ when you are done.dt .Formatting Mathematics: In TEX.

the usage (instead of $$ $$) is \[f(x) = x^2 + 3\]. ω Γ ∆ .. then enter the TEX code \&.Many TEX gurus recommend an alternative method for designating math mode. VII. The TEX commands for these symbols are mostly straightforward to learn. the usage (instead of $ $) is \(f(x) = x^2 + 3\) and for displayed math mode.. One advantage of these notations is that they are oriented: the user can tell the left delimiter from the right. If you need the actual symbol &. The following chart will get you well on your way to typesetting mathematics. dx {d \over dx} A_j^m \sum_{j=1}^\infty a_j/b_j^2 \pmatrix{ a & b \\ c & d} {a+b \over c} \leq {E^x \over s^y} \overline{\partial} u = \nabla g d dx Am j ∞ j=1 aj /b2 j a b c d Ex a+b ≤ y c s ∂u = ∇g Observe that & is used as TEX’s alignment character.. α β . \Omega b Ω f (x) dx a \int_a^b f(x) \.. and help to find them when they occur. For inline math mode. See also Appendices III... and VIII. \omega \Gamma \Delta . this fact will presumably help to reduce errors.. Mathematical Symbols: Many different technical symbols are used in mathematics. TEX Code Typeset Result \alpha \beta .. ©2001 CRC Press LLC .

the code \begin{eqnarray*} y &=& (x+2)(x+5) . A L TEX. In Plain TEX or AMS-TEX.2) ©2001 CRC Press LLC . \end{eqnarray} then the output becomes y = (x + 2)(x + 5) − (x + 1)(x + 6) = x2 + 7x + 10 − (x2 + 7x + 6) = 4. \end{eqnarray*} produces the output y = (x + 2)(x + 5) − (x + 1)(x + 6) = x2 + 7x + 10 − (x2 + 7x + 6) = 4.1) (2. In A L TEX. generates equation numbers for all equations in an array.Aligned Equations: Equations that are aligned at an equality or inequality sign are common in mathematical writing. The * in \begin{eqnarray*} and \end{eqnarray*} suppresses the equation numbers. the code $$ \eqalign { y &= x^2 + 2x + 17 \cr &\le w^3 + 4 w^{A \otimes B} \cr &\equiv z\times w \cr } $$ is typeset as y = x2 + 2x + 17 ≤ w3 + 4wA⊗B ≡ z × w.(x+1)(x+6) \\ &=& x^2 + 7x + 10 . If you enter instead \begin{eqnarray} y &=& (x+2)(x+5) . The TEX commands for typesetting aligned equations vary slightly with the TEX dialect. by default.(x^2 + 7x +6) \\ &=& 4 .(x+1)(x+6) \\ &=& x^2 + 7x + 10 .(x^2 + 7x +6) \nonumber \\ &=& 4 . (2.

2. But some spaces ©2001 CRC Press LLC . and makes it easy to find equations. displayed mathematics is set off from the text. Similarities: The default in TEX (for example) is to put (essentially) no spaces between characters in either of the math modes. they are very much the same—except that the displayed math is sometimes larger overall. Inline mathematics is part of the body of text. the recommendation is that the label come after the equation (i. Not only is the font incorrect. but the standard between-word spacing is gone. The reference [KRW] discusses this matter in greater detail. In broad strokes. enter $$ \eqalignno { y &= x^2 + 2x + 17 &(17)\cr &\le w^3 + 4 w^{A \otimes B} \cr &\equiv z\times w ..5 Modes of Typesetting Mathematics There are two standard formats for mathematics on the typeset page: inline mathematics and displayed mathematics.e. on the right). choose a system for numbering your equations that is transparent. Let us now discuss. replace \eqalignno by \leqalignno. (17) (18) To put the equation numbers on the left. This point must be clarified. As a result. The wisdom among typesetters is that the equation number is less important than the equation itself. their similarities and differences. &(18)\cr } $$ The output is y = x2 + 2x + 17 ≤ w3 + 4wA⊗B ≡ z × w. or the space just to the right of an integral sign).$— putting an English sentence in math mode—then the typeset output will be T hecatsatinthehat. TEX is actually very smart at finding the correct spaces to put between symbols in math formulas (for example the spaces around a + sign. If you enter the code $The cat sat in the hat. Finally. in some detail. Others like to put them on the right. Many people like to put their equation numbers on the left.For equation numbers.

a f (x) dx . certain components. That is why the inline mathematics mode has such labels adjacent to the carrier rather than above and below it. But it is also clear that such a user-friendly display would not work in inline mode. If you look in any mathematics monograph. 2. and easier to read. because the limits above and below would interfere with the text in the adjacent lines. Differences: Displayed math will be generally larger than inline math. That is.6 Line Breaks in Displayed Mathematics It is often the case that a displayed mathematical formula. And. such as integral signs and summation signs will be larger.(for example just before the dx in an integral) must be hand-inserted by the user or typesetter. In each instance. (3) ∪j∈A Sj vs. The result is that an expression in displayed math mode will appear altogether larger than that same expression in inline math mode. the displayed version is more attractive in its layout. cannot fit comfortably on one line. you will confirm instantly that the font in mathematical formulas is different from the roman font used for text. Compare the display (x+3)4 = x4 +12x3 +54x2 +108x+81 = x 108+x 54+x(12+x) +81 ©2001 CRC Press LLC . the characters are set in a font called math italic. j=1 b aj . j∈A Sj . or string of inequalities. A more interesting difference is in the way that sums and integrals and other mathematical artifacts with labels above and below are typeset. (2) b a f (x) dx vs. The text itself in a math display will be the same size as ordinary text. by default. Consider the following examples: Inline Mode (1) 10 j=1 Display Mode 10 aj vs.

Second. We now discuss. and instead to think in terms of displaying mathematics in multi-line chunks. it is recommended to place the plus sign on the second line so that the reader knows that the line is a continuation. . and breaks at ≤ or ≥ or similar connectives are very natural and easy to read. it is best to disabuse oneself of the idea that a displayed equation belongs on one line. But it is sometimes necessary to break at connectives like a + or − sign. by way of some examples. In practice. how best to do this. EXAMPLE I: The break (x + 2y)6 = x6 + 12x5 y + 60x4 y 2 + 160x3 y 3 +240x2 y 4 + 192xy 5 + 64y 6 is more attractive than either (x + 2y)6 = x6 + 12x5 y + 60x4 y 2 + 160x3 y 3 +240x2 y 4 + 192xy 5 + 64y 6 or (x + 2y)6 = x6 + 12x5 y + 60x4 y 2 + 160x3 y 3 + 240x2 y 4 + 192xy 5 + 64y 6 . this means to think in terms A of the command \eqalign{ } and in L TEX it means to think in terms of \eqnarray{ } rather than the displayed math command $$ . The issue therefore arises of how to break a long equation across a line. . Our last example showed breaks at = signs. First of all. or perhaps in the middle of a product of two large expressions. very long and complicated equations are quite common. $$. so that the reader’s eye is signaled of a continuation. In TEX. In substantive mathematical discourse. The first choice is to break at a “verb” connective. ©2001 CRC Press LLC . it is a good idea to indent the second line.with the much more readable display (x + 3)4 = x4 + 12x3 + 54x2 + 108x + 81 = x 108 + x 54 + x(12 + x) + 81.

A3 + xy 2 + ex + 7 log B = −4xy 3 + zw2 erf(s + t) − cosh t3 p s + a+b c × x2 − 4y · x + w − tan θ ln(3t) + 2w cos ψ + arg(τ ) is more attractive than either A3 + xy 2 + ex + 7 log B = −4xy 3 + zw2 erf(s + t) − cosh t3 p s + a+b c × x2 − 4y · or A3 + xy 2 + ex + 7 log B = −4xy 3 + zw2 erf(s + t) − cosh t3 p s + a+b c · x2 − 4y · or A3 + xy 2 + ex + 7 log B = −4xy 3 + zw2 erf(s + t) − cosh t3 p s + a+b c x2 − 4y · x + w − tan θ ln(3t) + 2w cos ψ + arg(τ ) . ln(3t) + 2w cos ψ + arg(τ ) β = x2 − 4y · ©2001 CRC Press LLC . s + a+b c x + w − tan θ . Third. in the last example one could write Let α= erf(s + t) − cosh t3 p . x + w − tan θ ln(3t) + 2w cos ψ + arg(τ ) x + w − tan θ ln(3t) + 2w cos ψ + arg(τ ) First of all. For instance. so that the reader’s eye is signaled of a continuation. Second. that meaning is lost when the two expressions occur on different lines. Last. we have the ability to use notation as a weapon. the multiplicative · is hard to read and too easily misinterpreted. As mathematicians. This device is exploited with insufficient frequency. it is best to include the initial × on the second line so that the reader knows that the line is a continuation. it is a good idea to indent the second line. while juxtaposition of two expressions is commonly interpreted in mathematics to mean multiplication.

and that they are easily and quickly paired up in the reader’s eye. then it may be necessary to break it. −4xy 3 + zw2 A3 + γ = α · β . then it will have to be broken into several lines (not just two). you want to be certain that the left and right delimiters match (in both size and shape). ©2001 CRC Press LLC . Thus it is best to make them a bit larger than is perhaps necessary. Doing so. or brackets [ ]. EXAMPLE III: If the material in your display is grouped with parentheses ( ). the displayed expression y + x2 = x−y a−b m+n − · x + 2y a + 2b m − n s+t + · 16w4 s − 4t is much easier to read than y + x2 = a−b m+n x−y − · x + 2y a + 2b m − n s+t + · 16w4 s − 4t or y + x2 = [x − y]/[x + 2y] − {[a − b]/[a + 2b]} · {[m + n]/[m − n]} +[s + t]/[s − 4t] · 16w4 . In such a case. But if the expression between delimiters is quite long and complicated. It is helpful to have the material on the second line shorter than the material on the first—so that the second line looks like an appendage (which is the effect that you want). For example. or braces { }.and γ= Then xy 2 + ex + 7 log B . For example. EXAMPLE IV: If a displayed expression is very long. then it is best to keep matching pairs on the same line. alignment and judicious use of indentation can be of great assistance to clarity.

7 Types of Space TEX is very good at the spacing that is needed for a well-formatted mathematics formula.y+ a−b−c d+b−c a+b+c = + a − b + 2c a + b − 2c a − d + 2c a+f +c h+b+k + − a − b + 2g a − k + 2 +f +m r+b+q − − a − n + 2z a − k + 2p ≤ x2 + 3y + 7 is much clearer than a−b−c d+b−c a+b+c = + y+ a − b + 2c a + b − 2c a − d + 2c a+f +c h+b+k + − a − b + 2g a − k + 2 +f +m r+b+q − − a − n + 2z a − k + 2p ≤ x2 + 3y + 7 or y+ a−b−c d+b−c a+b+c = + a − b + 2c a + b − 2c a − d + 2c a+f +c h+b+k + − a − b + 2g a − k + 2 +f +m r+b+q − − a − n + 2z a − k + 2p ≤ x2 + 3y + 7 . In this circumstance. then the verb connective (= or ≤ or something similar) and the righthand side of the first relation will have to go on a succeeding line. EXAMPLE V: If the first line of a multi-line expression is very long. 2. It knows how much space to put after an integral ©2001 CRC Press LLC . the second and subsequent lines should be aligned with a two-em quad indent from the left: z 3 −xyz 2 + a+b−c + sin[x3 y − zy 2 + xyzw] d − e2 + f = z x yw mqr st + a · b · cde · f · g − hi jk lm no p ≤ 23 − 45 + 17 + 333 − 4769.

2. In an integral: b f (x) dx. and V ⊕ W . 5. A ⊃ B. 4. TEX supplies the space automatically as part of the “glue” that comes with each of the connectives. but not after. a There is a thin space both before and after the f (x). S ∩ T .023 in.) 3. and how to typeset an arrow. for example. all spacing that occurs is imposed by the user. Thus.sign. and a ∈ A. Before and after a connective that can be read as a “conjunction”. the first of these is automatic and the second not. Before. Before and after a connective that can be read as a “verb”: These include = ∼ = → ⊂ ⊃ ≡ . the TEX code $x y z^2 w$ is typeset as xyz 2 w. Some of the most standard horizontal spaces in typesetting are these • The thin space: This is a horizontal space with thickness approximately 0. Some standard instances in mathematics where the thin space occurs are 1. In situations like this. After function names that are set in roman type: sin x log 3 deg α dim T exp 4z ©2001 CRC Press LLC . After a coefficient: ∞ b α j=1 xj β a f (x) dx. These include + − ÷ ∩ ∨ ⊕ Examples include x + y. how to space out a matrix. = ∈ ± Examples are x = y. certain binary connectives when they are used as adjectives: {} + 3 {} − 4 {} ± 8 6. So one must have positive cognizance of spacing. (In TEX. But the default in TEX mathematics mode is to put no space between objects.

traditionally.) • The em quad: This space is so named because. Some standard instances in mathematics where the thick space occurs are: 1. Some standard instances in mathematics where the em quad occurs are: 1. .3. Nowadays the definition is variable. . Around conjunctions (or other connectives): x>0 or an > 0 if n ≥ 5.8. 2) a. it was the width of the letter capital “M” in the font being used. 8. Some standard instances in mathematics where the two-em quad occurs are: or y<5 ©2001 CRC Press LLC . Before and after a single vertical bar or a colon used as a mathematical symbol: A|B {x : x2 > 0} 9. 2.7. After commas in a list or sequence: {1. c. and depends on the font and on other parameters as well. . Before a “back subscript” or “back superscript”: x 3H 1 a 2M • The thick space: This is a horizontal space of thickness equal to two thin spaces. . . Between mathematical expressions and verbal statements: an > 0 for most values of n or a=b by Theorem 1. • The two-em quad: The space is so named because it is twice the thickness of an em quad. In congruences (before the parentheses): 9 = 3 (mod 6) 2. 2. In conditional statements (before the parentheses) in text: xn > n (n = 1. 2. 3} (1. b. .

dx dy. 1. y) dxdy. A Note that L TEX2ε has the command \iint which is. in effect. one obtains f (x. When two integrals are juxtaposed. To separate an equation from an accompanying condition in display: an > 0 (n = 0. and the result is too great a space between the two integral signs. y) dxdy. x4 + y 5 = 3 . one can mandate a horizontal space of any size with the command \hskip#. . The analogous commands \iiint and \iiiint produce triple and quadruple integrals. Using the TEX code \int \int f(x. a macro for the formatting protocol that we have just defined. For instance. dx dy produces the more desirable f (x. Between two equations: x2 − y 3 = 9. If you wish to typeset two integrals in a row. • Negative space: The most commonly used negative space command in TEX is \! . • Forced horizontal space: In TEX. such as f (x. Here is a common use of \!. the code ©2001 CRC Press LLC . every character or expression in TEX comes equipped with “glue”. that is to say. .y) \. y) dxdy. which is a certain amount of spacing that is formed around it. then it will not do to typeset \int \int f(x. In fact. an amount of space equal to a thin space is removed at the place where the command occurs. their built-in spacing superimposes. It denotes a negative thin space.y) \. 2. .1.y) \. where # denotes a linear measure. The improved TEX code \int \! \! \! \int f(x. dx dy.) or bn < 0 (n odd).

are commonly abbreviated in mathematical formulas. just to illustrate what kerning does. \sin in math mode gives the sine function in roman followed by a thin space. Thus if one enters the TEX code book \kern-. in equations. expressions like sin x + cos θ − exp t = 0 . and positive kerning will have the same effect as a positive space (except that the horizontal space command \hskip has a certain amount of flexible latitude.3in and Sally or Sally \hskip-. We will frequently filter these topics through the sensibilities of TEX. or “glue”. associated with it while \kern does not). “angle”.1in binder and book \kern. such as “sine”. in which case there is no space). Abbreviations in Mathematics: Certain words. What you should observe about these expressions is that (i) the abbreviated word is in roman (not italic) and (ii) there is a thin space after the abbreviated word and before the argument (unless the argument is in parentheses or brackets or braces. If you fail to use the built-in TEX command ©2001 CRC Press LLC .8 Technical Issues Here we list and discuss some technical matters that come up frequently in mathematical typesetting. TEX supplies macros for the most commonly used math abbreviations. 2. Thus we see.1in binder then one obtains the output book binder and book binder This is an extreme example. But one can plainly see that negative kerning will have the same effect as a negative space. and “exponential”. Here we concentrate on just a few issues. For example.2in and Bob yields the compiled output Bob and Sally or Sally Bob and • Kerning: To kern is to move the position at which the typesetting is taking place—either left-right or up-down. We recommend the book [SWA] for the mathematical typesetting details and [SAK] for technicalities of TEX. “cosine”.Bob \hskip.

(Is this an eccentric rendition of the word “sinx” or the product of s. Blackboard Bold Characters: Because it is difficult to write boldface characters with a piece of chalk at the blackboard. and instead enter the code $sin x$. It is standard in printed mathematics to denote the integers. it is a commonplace. Then typeset the characters with $\ZZ.6. You should include the lines8 \newfam\msbfam \font\tenmsb=msbm10 \font\sevenmsb=msbm7 \font\fivemsb=msbm5 \def\Bbb{\fam\msbfam \def\ZZ{{\Bbb \def\QQ{{\Bbb \def\RR{{\Bbb \def\CC{{\Bbb Z}} Q}} R}} C}} \textfont\msbfam=\tenmsb \scriptfont\msbfam=\sevenmsb \scriptscriptfont\msbfam=\fivemsb \tenmsb} at the top of your *. group information. reals. for maximum clarity in mathematical 8 Calling a text font involves a simple one-line command—see Section 3. When this author published [KRA1]. Q. Donald Knuth does not like the blackboard bold font (see [KNU] to learn why). and C. then you will get the unsightly result sinx. Delimiters. As an instance. and their limitations. Calling a math font involves calling an entire family and is a bit complicated. their use. TEX will not have the name built in and you will have to define your own macro.tex file. ©2001 CRC Press LLC .for sine. The book [SAK] contains a discursive discussion of delimiters—their names. Size of: A delimiter is a bracketing device such as a parenthesis or a brace or a bracket. It is essential. i. people became accustomed to blackboard bold and it was used in print as well. n. R. The American Mathematical Society (AMS). however. the font blackboard bold was invented. has come to the rescue. you can enter \def\sgn{\mathop{\rm sgn}} to define a TEX command \sgn as an abbreviation for the signum function. See [SAK] for a thorough and cogent discussion of these matters. and organize information. rationals. Nowadays. The AMS font msbm has these characters. however. R. \QQ. Delimiters parse information. The result is Z. \CC$. Over time. their meaning. \RR. and complex numbers as Z. C. but you need to learn how to call the font. Q. the publisher declined to use blackboard bold for the complex numbers because the font needed to be imported from Japan and was considered to be too expensive. x or something else?) For some functions. hence A blackboard bold characters are not a standard part of TEX or L TEX.

In circumstances such as these. to use delimiters of the correct size. they can (and should) be allowed to extrude slightly beyond the upper and lower limits of the parentheses. For example. and position. The commands for explicit sizing are \bigl \Bigl \biggl \Biggl \bigm \Bigm \biggm \Biggm \bigr \Bigr \biggr \Biggr \big \Big \bigg \Bigg ©2001 CRC Press LLC . shape. dt-\frac{a}{b} \right ) . dt-\frac{a}{b} ) $$ is hard to read just because the parentheses are too small. we use explicitly sized delimiters. TEX measures the heights of the intervening boxes and adapts the delimiters to those boxes. $$ With this change.writing. in the expression n+1 k=1 k k2 + 1 2 . But sometimes you must intervene when TEX makes a poor choice. A passage like 1 a y = x + (z + f (t) dt − ) b 0 (typeset with the TEX code $$ y = x + (z+\int_0^1 f(t) \. Since the limits k = 1 and n + 1 are not seen by the human eye as an integral part of the summation. TEX will size the delimiters for you if you simply modify your code in this way: $$ y = x + \left ( z+\int_0^1 f(t) \. The commands \left and \right generally do a good job at suiting delimiters to the text that comes between them. typeset with the code $$ \left ( \sum_{k=1}^{n+1} \frac{k}{k^2 + 1} \right )^2 $$ the parentheses are a bit too large. the compiled output is 1 y =x+ z+ 0 f (t) dt − a b .

Ellipses: An ellipsis is a sequence of three dots used either to indicate a break in text or the trailing end of a thought. .case. Complete details may be found in [SAK]. then usually a raised (or vertically centered) ellipsis is used. For instance Consider the sequences a1 . L TEX has alternative tools for handling that situation (see [LAM] or [GMS]). then you must include a fourth dot (the period) to denote the end of the sentence—unless. an ellipsis is surrounded either by operators or commas. If the “missing material” is bracketed by binary operators. . and Diacritical Marks: Twenty-six alphabetical letters. of course. v. one in the definition of C. and one at the end of the passage. Please use them. . We sometimes also use explicitly sized delimiters when the left and right units occur on different A lines. both upper.e . and C = b1 × · · · × bM . bM of real numbers. you mean to indicate that the sentence is incomplete. a . aN and b1 . . Otherwise a baseline ellipsis is appropriate. In TEX. one in the definition of the product. . one in the sum. . . . .and lower. Embellishments. . Accents. Now we ponder these numbers . are generally insufficient mat´rial to express our mathematical thoughts. one on the baseline in the list of the bj ’s. a2 . and the centered dot · each come with glue. TEX provides us with \dots for a baseline ellipsis and \cdots for a raised (or vertically centered) ellipsis. . B = b1 + · · · + bM . Therefore we frequently e use embellishments over single letters. Examples are ˆ ˇ ˙ ˘ f . Define A = a1 · a2 · · · aN .Thus we would typeset $$ \biggl ( \sum_{k=1}^{n+1} \frac{k}{k^2 + 1} \biggr )^2 $$ to obtain the somewhat more desirable output n+1 k=1 k 2+1 k 2 . In almost all applications in mathematics. it is incorrect to form an ellipsis simply by concatenating three periods (for a baseline ellipsis) or by concatenating three \cdot commands to adjoin three centered dots. ©2001 CRC Press LLC . and they do not perform well under concatenation. The reason is that the period . If the ellipsis occurs at the end of a sentence. c . The displayed text exhibits six ellipses: one on the baseline in the list of the aj ’s.

If you are tempted to flout the suggestions in the previous paragraph. as they are harder to read (and to remember) than un-embellished ones. as they can interfere with letters (like g ) that hang below the baseline. you probably do not want a break in “Mr. The function $f$ is assumed to operate on $x$ in the following way: ©2001 CRC Press LLC . consider instead using a boldface character. The first choice for a line break is between words (it obviously only requires a trivial filter for TEX to find these). In either of these circumstances. Hard vs. you use the TEX code Section~5 or Mr. and the second choice is at the inter-syllable hyphenation of a word. the sentence would be difficult to read. As an instance of the first of these. This observation simply means that you must enter math mode (using either $ signs or \( and \)) to typeset the x and the f . In fact TEX comes equipped with a rather remarkable algorithm for hyphenating words. The function f is assumed to operate on x in the following way: it is essential that the one-symbol formulas x and f be formulated in math italic. But there are some places in your text where you definitely do not want a line break. The books [KNU] and [HIG] provide more information on the use of the so-called hard space ~ . If not. and usually gets these breaks right. Soft Spaces: If you enter plain text in a *. Smith” between the title “Mr. As an instance of the second of these. because they are difficult to typeset and may interfere with the line above.tex source file. then TEX will decide where to put line breaks. you do not want a break in “Section 5” between the word “Section” and the number “5”. or a character from another alphabet. compound embellishments are best avoided. Thus the correct code is Let $x$ be a real variable. As noted in Section 1. and other places where not having a break is definitely preferred.It is best not to over-use embellished symbols. One-symbol Formulas: In a passage like Let x be a real variable.~Smith to tell TEX that you want a space.” and the name “Smith”. but the space cannot be broken at the end of a line. or in situations like them. Embellishments under letters are generally a bad idea. An equation such as a·c+e ˇ ` ` m/f − w · n ´ ˆ ¯ ˜ =h +a+z ˙ ˘ is a nightmare to read and to comprehend.2.

In particular. there is considerable disagreement concerning whether punctuation should appear at the end of a displayed formula. So as to leave a widow (see Section 1. 5.e. not the end of a sentence. that precedes a displayed equation. y. the incorrect code Let x be a real variable. should it be Substituting the values for x. But here are a few breaks to avoid: 1. An acceptable place to break a page is before a short phrase. a new section title should be followed by at least two lines of that section on the same page). So as to leave an orphan (see Section 1. The function f is assumed to operate on x in the following way: produces the less desirable output Let x be a real variable.4). 6. y. The function f is assumed to operate on x in the following way: Page Breaks: TEX usually has pretty good judgement about where to break pages.. or Substituting the values for x. So as to isolate the last three lines of a chapter. Punctuation and Formulas. Interaction of: The final word about punctuation in formulas has not yet been enunciated. and z results in the formula A = B 2 + CD . In the middle of a displayed equation or sequence of equations that run more than one line. 2. and z results in the formula A = B 2 + CD ©2001 CRC Press LLC . and probably you do too.4). That is.By contrast. then it is all right to put a page break in the middle of the sequence (the logic being that such a sequence does not make up a single assertion). After the first line of a new section (i. 4. 3. In a series of equations that are numbered as a group. and it will not be enunciated here. If there is a sequence of equations separated by commas.

and you should learn it and adhere to it. For example. then you most certainly would include the punctuation.9 Including Graphics in a TEX Document The original design of TEX made no provision for the inclusion of graphics stored in external graphics files. After all. The situation may be remedied by the use of the \mathstrut command: $$ \sqrt{\mathstrut a}=\sqrt{\mathstrut X}+\sqrt{\mathstrut y} $$ The \mathstrut provides an invisible strut of zero width atop which the square root rests. The number of popular graphics ©2001 CRC Press LLC . An argument in favor of the first protocol over the second is that. The output now is a= X+ y and all the square roots have the same height. you should choose a paradigm for punctuating formulas and then be consistent. then the formula translates to words and demands a period. Complete details are in [SAK]. Some publishers will have a preferred style. Clearly the question does not have a clear answer.One argument for the second protocol in favor of the first is that if the formula was replaced by a figure or photograph. but we shall say nothing about it here. punctuation is redundant. then you certainly would not put a period at its lower right. the TEX code $$ \sqrt{a} = \sqrt{X} + \sqrt{y} $$ typesets rather badly as √ √ √ a= X+ y with square root signs of different heights. In your own writing. Many scientific style manuals militate against punctuation at the end of a displayed formula. Struts. Use of: Struts are a technical device that allows TEX to provide a “roof” on which to rest certain symbols. the vertical space after a displayed formula provides the necessary logical pause. if the formula was typeset in inline mode. 2. with that thought in mind. if you read the sentence aloud. There is also a \strut command for use in text mode. So why not do the same with a displayed formula? On the other hand. Most mathematics books and journals do include the punctuation.

The A PICTEX package for TEX and the L TEX picture environment are based on this approach. and especially with labeling (move a figure and the labels may stay where the figure used to be).rcc.ca/rta/brd038/nsr/ptenk/bitvec1. Vector images have the advantage of being easily scalable. add or change labels to figures. and any assumptions about graphics formats of external files would have rapidly made those parts of TEX obsolete. METAFONT is used to design fonts for TEX. for example) images. Briefly. such as *. Vector graphic files tend to be smaller than bitmap files. New technologies are being developed that combine the best features of both formats.bmp.htm has a lovely discussion of the merits of these two graphic protocols.formats was growing then and is probably continuing to grow as you read this book.. Today there are two principal types of graphics: bitmap graphics and vector graphics. The TEX companion program METAFONT can also be used to produce graphics. a bitmap graphic stores the image pixel by pixel while a vector graphic uses a mathematical language to describe lines and curves.. scale figures. But these packages do not have the power and flexibility of general-purpose computer packages for producing bitmapped graphics or line/vector graphics. etc. a vector graphics language). then everything works as it should—just like any other part of the TEX document. in effect. PSTricks is one of many packages that addresses this problem with captions and labels. One significant advantage of the “internal” graphics packages just described is that they travel well. Third party graphics packages pose real problems with scaling. TEX does have the capacity for producing graphics internally by positioning text characters at arbitrary positions within an open box. This point will be discussed in more detail below.. Even though out-of-the-box TEX has no provision for graphics.gif and *. and uses a script language to specify the shapes of characters (so it is. The result would appear to the user as a new font with a single character that might be of arbitrary size. scaling and rotation of graphics is generally awkward in the METAFONT environment. In particular. If you re-position material in your document.} allows the user to send whatever information he or she wants directly to the printer or screen display program. The printer or screen display ©2001 CRC Press LLC . the door was left open for extensions. This script language can be used to produce line drawings. The Web site http://www.ryerson. while bitmap images are better at providing the detail needed in high quality (photographic. we discuss both vector graphic formats such as PostScript or Acrobat (actually both of these are much more than vector graphics systems) and also bitmap graphic formats. The TEX command \special{. In what follows.

please note that you can use any good Web browser—such as Netscape — to view a PostScript file. Another useful utility for this purpose is ghostview. I have heard experienced editors yearn for the “good old days when the publishing house did everything. and scaled as you want.9 However. The publisher may offer a lower royalty rate if his production department must do the figures. positioned as you want. in just the form that you want. the publisher may pay the author if he is willing to handle the graphics himself. be a cost. which can be downloaded from the Web. On the other hand. specialists in the production department at the publishing house will either redraw each figure from scratch or will scan each figure into MacDraw or CorelDRAW or some other powerful graphics software and render it as a cleanly and professionally drawn figure. The most common use of \special commands is to send to the printer information about external graphics files and about how the information in these files should be positioned and scaled. and/or halftones on separate pieces of paper. 10 There may.program can then do whatever it likes with this information. The figures are submitted along with the finished manuscript to the publisher.ps file. You will be certain that each of your graphics is exactly what you want. as opposed to submitting graphics to the publisher on separate pieces of paper. We cannot include all of that information here. PostScript has become widely used for the production and manipulation of graphics. 11 If you have never before published a document with graphics. then you may have no idea what a struggle it is to communicate your desires to a technical artist who is probably a few thousand miles away and who has no idea what you are trying to depict. Specifically. The publisher himself can complicate matters by complaining that the production of your graphics is becoming too convoluted and too costly.11 9 Refer to Appendix VI for some URLs where Acrobat and PostScript resources can be obtained. photographs.10 There are advantages and disadvantages to handling graphics yourself. and most current publishers will perform this service for an author. Among these are: Advantage: You have complete control over the form and the formatting of your document. Traditional (non-electronic) methods for the inclusion of graphics in a book or paper are more laborious. This “traditional” method is still used. as an integral part of the electronic document. however.” But the modern business reality is that the house wants to outsource to the author. The book [GRM] is devoted to this topic. who handles the technical details from that point. In recent years. In the Windows environment. simply click on <File> and then <Open Page> and enter the path of the *. ©2001 CRC Press LLC . The author creates line drawings. TEX printer and screen display programs that can read and display PostScript graphics files are widely available.

each time you revise your document. or to use PSTricks or another package that makes the labels portable). and the page was run through the printer four times.13 With modern technology. magenta. ©2001 CRC Press LLC .e. to draw the labels in the fonts provided with your graphics package. many of these steps 12 Here the film is used to print a photographic image on a (usually copper or zinc) lithography plate. However. Disadvantage: If you want to include color graphics. 13 These two sentences explain why color graphics traditionally drove up the price of a book or document. all the labels will be disturbed. Many (but not all) publishers can process a TEX file with suitable “include” commands for PostScript graphics files. the method for inclusion may be system-dependent. If you typeset the labels in TEX. Or it might be a PostScript file for each chapter. Traditionally. Disadvantage: You must be sure that you are able to generate output that your publisher can use. cyan. You may find it more convenient to make the labels for each graphic a part of that graphic (i. This means not simply that you end up with satisfactory figures in your printed output. Disadvantage: If the publisher wants you to provide “camera-ready” output in hard copy. it may be possible that. a “screen” or “benday” was used to sort a color picture into its yellow. and black source.Disadvantage: The method you choose for incorporating graphics may depend on the system you are using and the software you are using. The plate is then etched and used as the printing master in the press. which includes all the text and all the graphics for that chapter. then you must have access to a 1200 dpi or higher resolution printer.. If your publisher is able to transfer your data directly from a computer disk or computer file to film. Then four separate printing plates were generated. Disadvantage: You must determine how to position each graphic on the page. Most publishers can handle PostScript files. and decide how to attach labels to each figure. then you may need to learn additional procedures for color filtering.12 then this step will not be necessary. This might be a TEX or L TEX file that has PostScript input commands and an accompanying PostScript file for each figure. but also that you end up with a computer file that the publisher can read and A manipulate. and you must be able to obtain the sort of RC (resin-coated) paper that is required for camera-ready output.

graphics can also be made up of raw printer commands. These include PICTEX. L TEX’s native picture environment. and others. then you will have no trouble generating a PostScript file for inclusion in your document. Such an internal graphics package has the advantage of portability—its source code becomes part of the TEX source and is as portable as the *. and keeping track of the primary colors. However. that you send to the printer using the TEX \special command. And our description here is by no means exhaustive. he can use a preprocessor to decompose a graphic into a number of fonts. then you are either going to have to (i) become your own system manager. part of the figure. compressed. All of these exploit commonly used and readily available software. and others). METAFONT. is still an important part of the process. Alternatively. Disadvantage: You can use graphics packages that are internal to TEX A A and L TEX. Disadvantage: If you are creating graphics with CorelDRAW or xfig or another standard graphics package. and these are not all compatible. In the following. proprietary to your particular printer. then so do these ancillaries.may be eliminated. each containing a small piece of the the graphic as a pixel pattern. But screening. if you are writing a book or article that has graphics. we will describe two or three common methods for including graphics in your TEX document. PSTricks. The real hacker will have other means at his disposal. We shall not treat any of these more advanced methods here. in effect. nonencapsulated. Also the labels and captions become. and is the responsibility of the person who is in charge of the graphics.dvi files that TEX produces. Gnuplot. many of these packages are limited in the types of graphics they can produce. there is no guarantee that precisely these resources will be available on your system. when the figure moves or re-sizes. The heading of the PostScript file must contain crucial data in just the right form or else you will not be able to use it. If you are so foolish as to introduce a blank line at the top of your PostScript file then some systems will not read it. and some are quite tedious to use. However. Thus. and configure your computer and printer to produce the types of files and hard-copy output that you require. or (ii) consult closely with the system manager at your place of work to find out what tools are available on your system for doing the tasks that you need to perform. But there are different versions of PostScript (encapsulated. ©2001 CRC Press LLC .

most graphics packages nowadays allow you to export your graphic to any of dozens of different graphics formats. 15 Virtually every UNIX installation. Many versions of this utility can also be downloaded from the Internet.9. Adobe Illustrator.ps is the default extension for a PostScript file name.14 they cannot. including PostScript and a variety of bitmapped and line graphic formats.ps. CorelDRAW. has a system utility called dvips.2.dvi file to a PostScript file. dvips will convert your *. Many of these products have their own proprietary (vector) graphics language in which figures are encoded. and FreeHand are some of these.eps is the default extension for an encapsulated PostScript file name. you would find it quite difficult to use Mathematica. these include formats that at one time were considered proprietary to Microsoft DOS.15 2. These products are obtainable both for the Windows and the Macintosh platforms.09) should include the line \input{epsf} (in L TEX2ε it would 14 It is actually possible to produce an entire document using Maple or Mathematica. Also *. ©2001 CRC Press LLC . and is one for which the interface with TEX is highly developed. and MatLab will generate beautiful graphs and other graphics.2 The Inclusion of a PostScript Graphic Suppose that your graphic is called Figure 1. and that the name of your corresponding PostScript file is fig1. or Macintosh systems. But the TEX capabilities are limited. Using a simple syntax. For example. A The graphics will be rendered in PostScript and the text will be rendered in L TEX. Most commerical graphics packages have the feature that their output can be saved in PostScript files. But a full-featured graphics package like CorelDRAW would make the job easy. The utility dvips can also input external PostScript graphics files and format them as part of your document. In addition. Harvard Graphics. 16 Observe that *. One can actually do calculations on the fly as one is writing in Scientific Workplace.16 Now your source code file (in A A L TEX 2. and Maple. we will assume that you have converted your graphic to a PostScript file. The PostScript format has the advantage of creating an ASCII or text file. UNIX.1 Handling Graphics in the Computer Environment Powerful commercial packages are available today for rendering graphic images for almost any application. The software Scientific Workplace is a A document-creation utility that gives L TEX output for text and has a Maple kernel for calculations. if you needed to draw a figure for a Euclidean geometry text. The mathematics/computer algebra packages Mathematica. A UNIX installation will probably include xfig or idraw. It will either send that file to the printer or will store the PostScript file to disc. however. be used as full-feature graphics packages.9. For the purposes of the present discussion. and many personal computer systems too.

This is. 18 As you might suppose. You A can. and also of making peace with the text that surrounds the figure. you may be unsatisfied with the results. after you compile and view a file containing the code just described. Instead use the code \begin{figure} \epsfysize=3in \centerline{\epsfbox{fig1. In fact. this is “three inches”or 3in) and the actual physical size of the figure that you will obtain in the output. A smaller number yields a smaller figure and a larger number yields a larger figure. of positioning and centering the caption. The \epsfysize command scales or sizes the figure. this calls in the package for handling PostScript files. the positioning of graphics is one of the most difficult features of the entire process. cut through much of the complexity by using L TEX’s built-in figure environment. L TEX will solve an optimization problem to determine where best to place the float. There is not necessarily any direct correlation between the numerical value of size that you mandate (in our example. and on the overall magnification of the TEX document. or a section title that is separated from its section.be include graphics and in Plain TEX it would be \input epsf) in its preamble. however.ps}} \caption{This is a picture of whatever it is a picture of. Indeed. A float is an object—this could be a graphic or a table or even a body of text—into a displayed object that does not have a pre-assigned position A in the document. TEX will not be sizing the collection of lines and objects that make up your figure. but rather the “minipage” on which it 17 It is not difficult to imagine that a poorly placed figure could cause a great deal of blank space to occur on a page.18 the figure environment will usually do a nice job of positioning and centering the figure. because that size will depend on the size of the figure with which you began.ps}} In point of fact. At the point of impact—where the figure should actually appear—there should be the lines \epsfysize=3in \centerline{\epsfbox{fig1.} \end{figure} Let us explain what each piece of this code signifies. in part. A L TEX’s figure environment turns the object inside it into a “float”. so that it causes the least difficulty with page and line breaks. this T X command simply centers—from left to right—the E object inside it. ©2001 CRC Press LLC .17 Combined with \centerline. you may be unable to find the figure on the page! In practice.

To use the first of these packages on a PC. 2. ©2001 CRC Press LLC . ury) are the coordinates of the upper right corner of the image. the latest implementation of L TEX.ps} will create the appropriate amount of vertical space for the figure and. A In both TEX and L TEX.09. It has both graphics capabilities (over L TE the graphics and the more advanced graphicx packages. lly][urx. There are various options that will enable you to scale the figure.19 The use of graphicx is similar.lives. the command for implementing graphics will be \usepackage[textures]{graphics} This change has to do with the nature of the TEX implementation on a Mac. And you will have to label your figures manually. Everything else about graphics is the same as for a PC. If you are using Plain TEX instead of A L TEX. In all. Later. it is quite common to use the alternative utility (which is actually a TEX package or macro) \psfig for the inclusion of PostScript graphics. 19 Here (llx. On many Macintosh installations. ury]{filename} to call in a particular graphic file.9. and numbers it automatically (in just the same way that L TEX A numbers chapters and sections and equations). lly) are the coordinates of the lower left corner of the image and (urx. has considerably enhanced A X 2. The other basic ideas are the same. one enters \usepackage[dvips]{graphics} and then \includegraphics*[llx. The command \psfig{psfile=fig1. then you will need to use the float commands \midinsert and \endinsert instead of the figure environment. The \caption command provides a suitably typeset caption for the A figure. the preceding release).dvi file that will cause the figure to be included. will insert a TEX \special command into the *. you can use L TEX utilities to collect all the caption names and numbers and pages and produce (without human intervention) a List of Figures to go in the front of your book or document. So you will have to learn through trial-and-error how to size your figures.3 A Graphics and the LTEX2ε Environment A A L TEX2ε. assuming the macro has been suitably customized. you will find that the simple five lines of code offered above will nicely dispatch almost any PostScript graphic that you want to A include in your L TEX document.

and also to print out the results.wmf. Embedded inside the definition of \setbmp is a \special command. PostScript graphics may be included. viewed. In its current implementation. the third (#3) is vertical spacing. 2. The commands needed to carry out this procedure are analogous to those for PostScript graphics that we described in the last subsection. it is much easier to use than earlier versions of L TEX. at the point of impact (where the figure is to appear). include the line \input setbmp.9.bmp) graphics for PCTEX. which is the basic tool for inputting printer commands and graphic machine codes. *. A complete example of a typesetting command in the PCTEX environment for Figure 1 (with file name fig1. and the fourth (#4) is the name of the file. at the time of this writing.0. y=#3}} Even those not well-versed in TEX will see that the first argument (that is. but have modified syntax in order to handle bitmaps. That was the latest release ©2001 CRC Press LLC . in the preamble to the document. \setbmp is a macro for the more discursive TEX command \def\setbmp#1#2#3#4{\vskip#3\relax\noindent\hskip#1\relax \special{bmp:#4 x=#2. PCTEX will allow you both to preview your TEX file with the *. In many A respects. #1) corresponds to horizontal spacing. the second (#2) is width. Then. we will discuss PCTEX 32. we shall instead now treat bitmap (or *. and other graphics formats. the current implementation does.bmp) is 20 For specificity. It has a full implementation of dvips.bmp} A few words of explanation are in order. include the line \setbmp{dim1}{dim2}{dim3}{fig1. version 4.jpg.4 The Use of PCTEX Just to illustrate another method for handling graphics in a TEX environment—perhaps on a home computer—we will now discuss PCTEX. Since we have already discussed PostScript graphics in the preceding subsection anyway. Specifically. but it is buggy. In fact PCTEX offers20 several methods for including graphics.bmp graphics displayed.A L TEX2ε has the delightful feature of a uniform syntax for the inclusion of every kind of graphics file that can be handled by the drivers. In fact. PCTEX can also handle *. and printed out. Previous releases did not allow screen viewing of PostScript graphics.

html 21 OzT Created by EX is a package available online for Macs.org/tex-archive/systems/win32 /miktex/1. and its graphic interface is particularly easy to describe. You will have to learn by trial and error what these numbers signify.7in} You can see the explicit role of the \special command in this example. A picture of everything. Today. 2. In particular.5 Freeware that Will Handle Graphics The TEX implementation known as BCTEX (for the PC). the way to get the figure positioned properly is to fiddle.160/winnt/misc/page2.26.. of course. flexible tools for handling graphics.20/index. and most powerful.html There are certainly more modern freeware versions of TEX that are available as downloads (such as MikTEX—see below). The dimensions 1.com/oztex/ozfaq. the command is \special{isoscale fig1.21 MikTEX has a useful dvips utility.9.8in}{fig1.html for more information. the software MikTEX (for the PC) is one of the most popular. a *. An implementation of BCTEX may be downloaded from http://199.bmp file) using BCTEX. To import a bitmap graphic (i.bmp.5in and 1.55in}{1. it comes both in freeware and shareware versions.7in denote the horizontal and vertical sizes of the figure.e.88in}{1. the dimension parameters given here do not correspond literally to anything that you will actually see in the printed output. But BCTEX is tried and true and readily available. available for free from the World Wide Web. http://www. indicating the scaling method (isotropic or non-isotropic) that one wishes to use.bmp} \begin{center} Figure 1. and many other helpful add-ons. drew Trevorrow.5in 1. 1. the caption. freeware implementations of TEX that can be downloaded for free from the Web.180. has powerful graphics capabilities.tug. AnSee ©2001 CRC Press LLC . The usual remarks about positioning the figure on the page still apply. The centered material following the figure call is.trevorrow. \end{center} As in previous discussions. It may be downloaded from the ftp site ftp://ctan.\setbmp{-2. The command isoscale is one of several choices in BCTEX.

a great place to get TEX fonts. However.edu/packages/TeX/systems /win32/miktex/1.wustl. but it contains useful packages such as Xy-pic (for commutative diagrams). This is a large installation (about 170 Mb).20/ (This last is a site at the author’s university.) Incidentally.html http://www.miktex. articles. the CTAN (Comprehensive TEX Archive Network) site is.html It contains fonts.de http://wuarchive.esm. this package is available only for purchase or by joining the TEX Users Group.com/~upem/miktex. For the Macintosh. and other materials related to TEX.psu.edu/mac-tex/versions. The TEX Users Group makes available the TEX package TEXLive4 (for the PC). OzTEX is the standard freeware package. utilities. See Appendix VI for other Internet sites that will be of interest to the TEX user. and other useful information about TEX for the Mac user. A Mac TEX Web site is http://www.tripod. ©2001 CRC Press LLC . See also Appendix VI for other TEX Web sites. add-ons.More information about MikTEX (and downloads) can be obtained from any of the Web sites: http://members. Another is called CMacTEX. in general.

1 Other Word Processors and Typesetting Systems Before TEX. you can move around large blocks of text. Certainly the possibilities are limitless: TEX can format a page in two (or more!) columns. Most importantly. What is tricky in the typesetting of English text is handling various advanced formatting issues. and can perform other standard and exotic typesetting tasks. and revisions is trivial. these issues rarely come up in the writing of a mathematical paper or book. For most of us. you can type as fast as you like. Making changes. and you can try things. as on a traditional typewriter or word processor. there were word processors. TEX can calculate those. is child’s play with TEX. in paragraph format. TEX will automatically indent the new paragraph and provide the right amount of vertical space between paragraphs. 3. it is also a powerful tool in the typesetting of English prose. just leave a blank line. and take care of right justification as well. corrections. the information in this chapter will suffice. never fearing for errors. can display an indented block of text. You need not worry about line breaks or page breaks. To begin a new paragraph. Using one. Word processors have changed the way people write.chapter 3 TEX and the Typesetting of Text Although TEX was designed specifically for its prowess with mathematics. Fortunately for you. a word processor is a marvel. For those who need these powerful but complicated TEX functions. Compared to the old-fashioned typewriter. Straight prose. ©2001 CRC Press LLC . can flow text around a graphic or figure. and many of the changes are for the good. You simply type away. the book [SAK] provides all necessary details.

is much more exacting. even so. Commercial patches are available which enhance the mathematical capabilities of these products. and especially the formatting and positioning issues. thus guaranteeing complete compatibility between screen output and hard-copy output. with your word processor or other system. a word processor cannot handle the variety of fonts and font sizes. that the printer can achieve is an order of magnitude higher than what is possible on the computer screen. which is a typesetting system. 2 Not 1A ©2001 CRC Press LLC .2 almost as if you had sent in a typed manuscript. Corel WordPerfect. that are essential to the quality typesetting of mathematics. 3 It is worth noting explicitly that it is impossible to create a without-humanintervention translator to mediate between a word processor and TEX. TEX is virtually the only complete software system for typesetting both mathematics and English. But the convenience of a word processor is also the source of its limitations. TEX. With TEX the matter is a triviality. which is easily produced from the output *. and around the world. Of course you can store your documents for future use. good luck. and call them up and modify them at will. Thus the accuracy. If you send your work to a publisher and it is created in TEX. and specifies page format to a much higher degree of accuracy than is possible on any word processor. It is used universally. That is because a word processor and TEX are conceptually different: the word processor is a device of convenience. A printer driver is built right into the word processor. and other pixel is a screen dot. Microsoft Word. then the publisher will tinker with it and then publish it in TEX (or in PostScript. But all the mathematics will have to be typeset from scratch. But. Japanese. such as TEX. Many publishers can utilize the ASCII code in your word processed document. It is available in Russian. Much more powerful is a markup language. Try creating a fraction whose numerator is an integral and whose denominator is the determinant of a matrix. TEX is a high-level computer language. But if you send your work to the publisher in WordPerfect then it will have to be typeset. and other popular word processors can. for putting words on the page in a workable form. TEX is a computer typesetting system—not a word processor. The best Super VGA screens are only about 1200 pixels1 across. A high quality laser printer has a resolution of 1200 dots per inch. format equations.3 As of this writing. and level of detail. and it allows you to issue commands that declare exactly what will appear on your page and how it will be formatted.The word processor shows on the screen just what you will be getting on the printed page. in fact. In other words. Of greatest interest for readers of this book is that a word processor is generally quite limited in its ability to handle mathematics.dvi file). quite.

very small type. Thesis. A children’s book will have even larger pages. and narrow margins. For a book. The Complete Index of Hypothetical Theorems. We now spend some time discussing various standard typesetting situations that deviate from the standard paragraph layout. and especially [SKI] for the complete chapter and verse. If you use AMSTEX or BibTEX. Consult [HIG]. If you use other typesetting environments. Usually paragraphs are indented.languages. an acceptable format is A. For a thesis. Voracious Vanity Press. then you will have to exert some control over the situation. and the purpose of the document. the default margins are 1 on all sides. You need to select margins that are suitable for the size of the page. Every young (and old) mathematician should learn to use TEX. Albemarle. In TEX. In a mathematical document—either an article or a report or a book— most text will be in paragraph mode.5 × 11 page and 10 point type. articles. Schlobodkin. This software has special ways of formatting references to books.2 Modes of Typesetting Text The typeset page has margins mandated. Our simple advice is that you should pick a standard format and stick to it. A medical reference work will have large pages. Vancouver. [SWA]. It is infinitely portable. Bibliographic References: A great deal might be said about how to typeset bibliographic references. possibly two columns. Journal of Unpublishable Results 43(1997). For a full-sized 8. for example. and 23–59 is the range of pages. theses. University of Albemarle. 1943. 23–59. ©2001 CRC Press LLC . and other forms of scholarly work. 1997 is the year. 3.. 1994. and there is customarily a bit of extra vertical space between paragraphs (i. the size of the type.e. Here 43 is the volume number of the journal. more space than between successive lines in the same paragraph). preprints. and it is stable. such margins are too small. the information required and the format are slightly different: Dewyew Luffmea. then the package will choose a format for you. The slaves should be freed. An incorrect proof of the Riemann hypothesis. but also large type and large margins. Lincoln. Most documents will have specifications somewhere in between. One standard format for a reference to a paper is S. and these usually are fixed once and for all throughout the entire manuscript.

Now such references are usually collected in a list at the end of the document. But the custom is still to minimize the use of footnotes. P. Many publishers of mathematics discourage footnotes. with extra space above and below and wider margins than the default. In the nineteenth century and early twentieth century. Of course there are occasionally other types of sources to which you will refer. or fifteen words. With the advent of electronic typesetting systems—such as TEX—this objection has grown moot. In more modern times. Davis. should be displayed. private communication. This could have been a personal letter. This is an interesting historical development. Displayed Quotations: A quotation of length greater than one sentence. When a quotation is displayed then quotation marks are not used. I’m not so sure. ©2001 CRC Press LLC . use J. This means that it is set off from the main body of text. Also. and that bat can digest all parts of the mosquito except the eyeball. Jones. Sometimes—especially for long quotations— the text is set in a smaller font. but we lost the war. One is the “private communication”—indicating that someone has told you something but that it has not been published anywhere. and is often separated from the main text by a horizontal line. There is a certain bat that eats the mosquitos. Thus the eyeballs are harvested by processing the leavings of the bats. bibliographic references in a mathematical work were generally set in footnotes. The custom in this circumstance is to write J. or a conversation. or perhaps an e-mail message. This text is generally in a smaller font than the main text body.For a preprint. Then the text of the footnote appears at the bottom of the page.4 The footnote at the bottom of this page was typeset with the code \footnote{This is an example of a typeset footnote. Footnotes: A footnote is designated by a special mark in the text— sometimes a number or a dagger † or another special symbol—at the point of impact. 4 This is an example of a typeset footnote. preprint. publishers objected to footnotes because they were extra work for the typesetter.}. An example is A rare delicacy in Korea is a gruel made of mosquito eyeballs. the footnote labels could be misconstrued as superscripts.

Drop it on the floor and step on the cards with flamenco shoes .50 to the man on his right. {\sf This is a fine text. . The key command here is \llap. it is common to hang the problem numbers in the margin.3in \noindent S.Marginalia: Sometimes. Each man hands \$3. 1992. and the others are of secondary importance. p. Sometimes a “hangindent” can make for an attractive display. .}\enskip}\ignorespaces Three men have \$47 dollars among them. Drop it on the floor and step on the cards with flamenco shoes \dots \medbreak\noindent\llap{{\bf 2. Even with electronic typesetting systems.50 to the man on his right. works of mathematics contain no material in the margin. Each man hands $3. in a reference work. and the subsequent lines uniformly indented. written ©2001 CRC Press LLC . there are situations where a “hang indent” format is desirable. Belmont. 2. it is useful to have topic headings or sometimes even notes to the reader in the margin. Consider a deck of 53 cards.}\enskip}\ignorespaces Consider a deck of 53 cards. Wadsworth Publishing. taking back a quarter when the other man isn’t looking \dots The result is 1. This can be done with the TEX code \medbreak\noindent\llap{{\bf 1. G. Krantz. generally speaking. taking back a quarter when the other man isn’t looking . . Again. An example is obtained from the TEX code \medbreak \hangindent=. But. 152] for a detailed discussion of all the TEX commands used in this last example. {\it Function Theory of Several Complex Variables}. This typesetting concept entails having one or more initial lines flush against the left margin. Three men have $47 dollars among them. When you are presenting a list of problems (as in the exercise section of a textbook). one sometimes has exercise numbers hang-indented into the margin. In a textbook. Even so. Second Edition. it is time-consuming and tedious to set up marginalia. part of the reason for this custom is that it is extra work for the typesetter to typeset such material. . We refer the reader to [SAK. California. Marginalia are best avoided.

for typesetting tables and the like. while young. the word “Theorem” or “Definition” or “Remark” or “Example”. See also the macros provided in [SAK]. Definitions. Wadsworth Publishing. We refer the reader to [SAK. the main thing is to choose a format and be consistent. and the others are of secondary importance. remarks. and Other Displayed Material: It is customary in mathematics to organize material into blocks of text. or big cap-small cap.in the classical style. A L TEX includes a utility called \newtheorem for typesetting enunciations. 1992. Function Theory of Several Complex Variables. G. The author. However. p. together with its number or other designation. The text of the enunciation is generally typeset in roman. examples. California. This is a fine text. it must be displayed—with extra space above and below and on both sides. Theorems. definitions. seems to know what he is talking about. notably the tabular environment.} which typesets as S. You should pick one and use it consistently. For example. Second Edition. while young. Belmont. Usually there is a small amount of extra vertical space above and below each of these displayed items. The author. L TEX has special macros. or enunciations. A table will often use multiple fonts. if you want to specify in advance how a proposition and a lemma will look. 186] for a detailed discussion of all the TEX commands used in this last example. Tables or Charts: Of course a table or chart cannot appear inline. is set in boldface. written in the classical style. many authors find it useful to put the text of an enunciation in slanted roman. theorems. or some other special form. There are many different paradigms for typesetting these units. and other devices are common organizational tools. just so that it will stand out. perhaps bold for headA ings and italic for subheadings. Some authors choose to set the word “theorem” (or other sobriquet) in all caps. Many times. The key command here is \hangindent. Krantz. The creation of a table (unaided) is remarkably difficult in Plain TEX. seems to know what he is talking about. then (in your preamble) you include the code \newtheorem{proposition}{PROPOSITION} \newtheorem{lemma}{Lemma} Now when you want to typeset a lemma followed by a theorem then you use the code ©2001 CRC Press LLC .

Many typesetters would agree that slanted roman is preferable to italic in this instance. . You can either invoke the alterA native font by setting \sl by hand or else (in L TEX2ε) you can use the command \theorembodyfont.). Let P be a zero of ζ that lies in the critical strip . The Hyphen: The hyphen is used in standard hyphenation of words: well-defined. A lengthy enunciation should not be typeset in italics. semi-retired.1 Let ζ be Riemann’s zeta function. are hyphenated: forty-two and eight-seven.1 Let > 0. ©2001 CRC Press LLC . Each of these three punctuation marks has a distinct and well-defined use. For every continuous function f . ultra-modern. A As you can see. italics were invented to slow down the eye. the en-dash ( – ). . it is the default in L TEX to typeset these displayed units in italic. written out in words. numbers between twenty and one hundred. This author finds such a typesetting decision to be in error. Let $P$ be a zero of $\zeta$ that lies in the critical strip \dots \end{proposition} The typeset result is LEMMA 3.\begin{lemma} Let $\epsilon > 0$. \end{lemma} For every continuous function $f \dots$ \begin{proposition} Let $\zeta$ be Riemann’s zeta function. Although infrequently invoked by mathematicians. . PROPOSITION 3. and the em-dash ( — ).3 Hyphens and Dashes English typesetting commonly uses three types of hyphens or dashes: the hyphen ( . Do not put spaces on either side of the hyphen. and should be used only for single words or short phrases. . 3. It is common in mathematics to refer to the Cauchy-Kowalewski theorem or the HeineBorel theorem.

5 But customs in other countries may vary. or creating a table. It is customarily the length of an em. You mark items in a manuscript for alignment by running a vertical (sinuous) line through the initial letters of the leading words and writing “align” in the margin. The em-dash is used for a pause or break in a sentence: Bob began systematically to move Mary’s possessions out of the apartment—he was really the best man to do it. or designating a correspondence. When making a list. Do not put spaces on either side of the en-dash. ©2001 CRC Press LLC . There is general agreement that you should not put spaces on either side of the em-dash. which is the width of the letter “M” in the current font. An interesting artifact of modern life is that some publishers (and authors) consider Cauchy-Kowalewski and Heine-Borel to be politically incorrect. which is the width of the letter “N” in the current font. It is customarily the length of an en. Name Jones Smith Larson Masters Rank sergeant major lieutenant colonel Number 37-99405 43-89427 11-48950 53-48045 An example of alignment. Thus the publisher will typeset Cauchy–Kowalewski and Heine–Borel (with an en-dash). because the use of the hyphen might suggest that Cauchy and Kowalewski or Heine and Borel were married (and had therefore hyphenated their names). you will use alignment to make your ideas clear.4 Alignment Alignment is a useful tool for organizing material. The en-dash is used to designate a range of numbers or years: pages 3–42 or the reign of King Henry (1833–1841).The En-Dash: The en-dash is longer than the hyphen but shorter than the em-dash. The Em-Dash: The em-dash is the longest of the commonly used dashes.5 3.

” in sans serif you enter the TEX code ©2001 CRC Press LLC .5 Typesetting Material in Two Columns We sometimes encounter two-column typesetting in encyclopedias (for example the Encyclopaedia Britannica). The book [HIG] has a detailed and informative discussion of how to form clear and easy-toread tables. We give just a quick idea here of how a text font call is done. The presence of complex. let us use the nickname sanser. 3. Then if you want to set the phrase “Quoth the raven. you would include the code \font\sanser=cmss10 This statement assigns the “working nickname” sanser to the font cmss10. Thus.6 Some Technical Textual Issues The formatting of text has a great many technical questions attached to it. the two-column format is sometimes used in statistical writing. Font Calls: The lore of fonts is vast and varied. at the top of your file. A Certainly L TEX has macros which make the two-column format a breeze to invoke. So a font is typically stored on your hard drive with a name that uses both characters and digits—such as cmss10. Consult the Author Guidelines for any publication you are considering before you think about setting your text in two columns.Making a good table—in which the information is clear and easy to retrieve—is a skill that must be honed over time. most of them will not apply to mathematical writing. Fortunately. 3. The basic logic of a font call is that the name that you use for a font in a TEX call cannot contain any digits. But we keep track of fonts by their name and their size. in medical reference books. On the other hand. equations is directly in conflict with the narrowness of columns. using them. The book [TUF] is an authoritative source on the display of quantitative information. A A L TEX2ε is more versatile in this regard than is L TEX 2. But. So you must assign a nickname to the font. often long. and in insurance contracts.09. Other fine references for table-making are [CHI] and [TUR]. ‘Nevermore’. We focus here on only a few of the most common topics. It is rarely appropriate to use the two-column format in a serious mathematical work. you will find yourself doing a lot of handwork to correct line over-runs and awkward line and page breaks. It is discussed rather thoroughly in [SAK] and [GMS]. The font cmss10 is a sans serif font.

The convention we have described above is very much in force—simply go and consult a ©2001 CRC Press LLC . This is not a trivial issue. ‘Nevermore’. Wodehouse) and puts a single space after the period. a left parenthesis wants more space to the left than to the right. Periods. A + sign wants an equal amount of space. a character’s glue tells TEX how much space that character wants on either side of it. TEX must build a certain amount of flexibility into its spacing mandates. then TEX assumes that the letter is an initial (as in P. ‘Nevermore’. And. But if instead you are typesetting “Mr.8. Unfortunately. A period “ .~G.e. In fact.\ Smith. As such.} The compiled output is then Quoth the raven.” (∗∗) Matters are actually more complicated than the previous paragraph might indicate. therefore the closing double quotation mark should come before the full stop (i. an = (equal sign) will want a slightly different amount of space. where the \ is a manual single space. In a nutshell. In that case you must either typeset Mr. typesetting convention runs contrary to logic. are in a constant state of genesis and flux.~Smith. G. where we have used the hard space discussed in Section 2.~Wodehouse) to guarantee that there will not be an awkward line or page break between or after the initials. Spaces After: One of the ingenious features of TEX is that most every character comes equipped with “glue”. If the period follows a capital letter. Smith” then you definitely do not want a double space after the period. the period).{\sanser Quoth the raven. ” is often used as a device for ending a sentence. Logic dictates that one should write Euclid called it “one of those Pythagorean things”. that is the default space that TEX gives to the humble period. Rules of grammar and syntax. or else typeset Mr. in order to effect attractive line breaks. and rules of typesetting as well. Usage of: There is reason to be grateful that this issue does not come up very frequently in mathematical writing. The correct mode is Euclid called it “one of those Pythagorean things. Quotation Marks. (∗) The thought here is that Euclid’s statement is a proper subset of the full sentence.. it wants no space on the left and a double space on the right. You may want to use the hard space anyway (typeset as P. of a certain size. on both sides. For example.

and it correctly converts them to left curly quotes and right curly quotes. and “perquisite”. that is to say. Spell Checkers: Whether you use TEX or a word processor. "Hello Bob.co. Judge for yourself.” “ludicrous. You may be the world’s greatest ©2001 CRC Press LLC . is arguably clearer and less clumsy than Bob’s favorite words were “apple. But in some systems there is now a system of “smart quotes”. i. But the rules are different in Britain. Smart Quotes is freeware. “Hello Bob. For colons and semi-colons the rule is just the opposite: the punctuation mark should follow the quotations marks. ”Hello Bob. But the inertia in current mathematics is to follow the logical rather than the popular. George said. See also the following discussion of smart quotes vs. For example.’’ produces the output George said. The Smart Quotes software of Sam Kington will automatically convert straight quotes to (correctly chosen) curly quotes in any desktop situation. ‘‘Hello Bob.” and “perquisite.e.” “apostle. This author has advocated (in the book [KRW]) that the standard rule should be broken when the quotation marks surround a short phrase without a verb. “ludicrous”. straight quotes.illuminated. The notion that the punctuation mark should precede the closing quotation marks applies both to commas and to periods. You can become better acquainted with Smart Quotes at the Web site http://www. the code George said. For example. do develop the habit of using a spell checker.popular novel to verify this claim for yourself. so that they are typeset correctly as left curly quotes and right curly quotes.” The clear advantage of using pairs of single quotation marks is that it allows us to specify left quotes and right quotes explicitly. Straight: The method in TEX for specifying quotation marks is to use pairs of single quotation marks..” It is easy enough to get used to TEX’s method for handling quotations." then you will get the incorrect result George said. Whatever paradigm you may choose to follow. “apostle”. Quotes. the latest release of Microsoft Word allows you to enter straight quotes both on the left and the right. If you instead use straight quotation marks in your TEX code.” Professional typesetters have told us that this view is probably correct and sensible. Smart vs. For example. do be consistent. the sentence Bob’s favorite words were “apple”. but it ought not to be uttered in public. more and more mathematical manuscripts adhere to the first custom (∗) rather than the second custom (∗∗).uk/acorn/smartquote/.

in your haste. The spell checker will help you to catch such errors. add words to your spell checker’s dictionary so that the jargon you usually use in your work does not get flagged. In point of fact. your spell checker may flag both “homeomorphism” and “homomorphism” as errors because they are not part of the standard English lexicon. But some popular spell checkers (such as MicroSpell ) have either a TEX mode that will ignore TEX commands or a TEX filter that will sort out the TEX commands and just check the text. Please note that the spell checker will not catch mistakes like typing homeomorphism for homomorphism or conversation for conservation because these are all legitimate words. ©2001 CRC Press LLC . But you can. You could. double a consonant that should not be doubled or put an “i” before an “e” when it should be the other way around. then your spell checker will flag \parindent A and \eqnarray and other standard TEX and L TEX commands. If you are a TEX user. and should.speller—winner of every spelling bee from London to Newcastle—but when typing rapidly you could set teh instead of the or enromous instead of enormous.

(call it mybook. the Table of Contents or TOC. the Glossary. When you then run latex on the file a second time. A few words of explanation are relevant in this regard. typeset in TEX. tables and other resources (optional). it builds a Table of Contents. the List of Figures (optional).toc. the Foreword (optional). the Dedication (optional). List of Figures. and the Index. A L TEX and other standard sets of TEX macros enable the automatic generation of the TOC. The back matter consists of Appendices (optional). you need to enter the line \tableofcontents in your source code file after you enter the all-imporA tant line \begin{document}. L TEX is what computer scientists call a 1 The first time that you run latex on a source file two-pass system.toc file contains information about the names and the page numbers of the various chapters. the Bibliography. a section. and the Index. and like material. even more than two passes may be required. one begins a chapter.tex). sections. the Preface. ©2001 CRC Press LLC . The *. A In the L TEX book style. it creates an auxiliary file called mybook. the Glossary. the Bibliography or References. and a subsection thusly: \chapter{All about Everything} \section{Specific Comments about Everything} \subsection{General Comments about Very Little} In order to create a Table of Contents. Of course all the titles and page numbers are literal copies of what 1 For some packages (such as BibTEX).chapter 4 Front Matter and Back Matter 4. the Acknowledgements (optional).1 The Beginning The front matter for a book comprises the front and back title pages. and subsections.

and so forth. and the author should do his utmost to help the reader navigate his way around.. The Index is an important tool in that quest. The entire procedure could easily take a week of hard work. and its page number. When you run latex on your file mybook. See also Appendix VI. “deformation”. the author would have a stack of said cards. amalgamate cognate terms. a romance novel). “group”.actually appears in the text. a typeset List of Figures is generated. the number. the index. “homology” and “homotopy” are principal entries and “class”. This is a wonderful labor-saving device. A mathematics book is thick and dense. then the file mybook.2 The End The back matter for a book consists of the references or bibliography. A sample passage from an Index might look like this: homology 37 class 38 group 40 homotopy 29 deformation 31 equivalence 30 In this sample. In the old days. With TEX. Then he would type out the index. and guarantees perfect accuracy. The author puts the command \makeindex in the TEX source code file before the \begin{document} command (i. the entire process just described has been automated. When you run L TEX a second time. the Index is of vital importance.e. In a mathematics book (as opposed to. whose URL is http://www. perhaps some appendices.tex. with entries and subentries clearly exhibited). he would note each word to be indexed. on a 3 × 5 card. It contains information about the caption. the author would painstakingly plod through the page proofs. 4.lof is created. At the end of this process. and A the page number for each figure. The creation of the List of Figures is just the same.tug. classify words either as principal entries or subentries (see the display below for a sample of a finished Index. and other auxiliary material of the author’s choosing. and “equivalence” are subentries. A List of Tables may be produced in an identical manner.org) will contain additional macros for specialized applications. say. and would then alphabetize them. The interested reader may consult [GMS]. You insert the command \listoffigures after the \begin{document} line. and the various TEX Web sites (such as the TEX Users Group. he would eliminate redundancies. in the process. in the preamble to the electronic ©2001 CRC Press LLC .

for example. without any human intervention.file). article. report. and any new document being created can ©2001 CRC Press LLC . or thesis. Simple operating system commands can now be used to alphabetize the myfile. or other item—has a nickname. removes redundancies. A glossary may be created using exactly the same procedures as we have described for creating an index. In A L TEX.idx is created.tex) is compiled. Many of the public domain TEX archives. You simply enter the command makeindex myfile. with the use of auxiliary TEX packages (such as the utility makeindex) that are available on many systems. Now here is data. to eliminate redundancies and to sort entries as either principal entries or subentries.idx file. When the final manipulation is complete. the myfile. ready to be manipulated. and it compiles as a finished Index. In out-of-the-box L TEX.bib. contain utilities for performing the sort of sorting and filtering described in the last paragraph. much of the editing described in the last paragraph can be done by machine.) has its own paradigm for formatting the references. It contains each word that was tagged. such as CTAN (see Appendix VI). So why not have them all in a single database. The reference [GMS] gives a more detailed treatment of the use of the makeindex command. the entire task can be handled in an evening. using a good text editor. See also the indexing macros that are provided with the book [SAK]. The creation of bibliographies is a world of its own. or article. and sorts out index entries and index subentries. Usually. each style macro (book. together with the number of the page on which the word occurs. which gives the user the power to create a bibliographical database with a filename *.tex. Then he inserts tagged words in the file wherever they are to be A indexed. BibTEX is premised on the notion that we tend to use the same references repeatedly in our work. The glossary is compiled and formatted in just the same way. This new file alphabetizes the index.idx file is incorporated into myfile. if the word “homotopy” occurs in the TEX file in line 1227. then the author adds \index{homotopy} to A that line adjacent to the original occurrence of that word. There is also BibTEX. etc.idx and the file myfile. Incidentally. In L TEX2ε.ind will be produced. an auxiliary file myfile. and add to that database as new references come along? Each entry in the database—be it book. And it is straightforward. Words are tagged with the TEX syntax \glossary{ }. Different versions of TEX each have their own macros for formatting the bibliography. one can also use an index tag of the form \index{homotopy!stable} to indicate that “homotopy” is a principal index entry and “stable” is a subentry. When the TEX file (call it myfile.

key = "GAJ". BibTEX will cause you to re-think the process of doing bibliographical work. which preprints have appeared. allow you to keep the reference lists of all your works up-to-date (i. number = 7. publisher = "Addison-Wesley". bibliographic databases posted on the Web. journal = {\mbox{G-Animal’s} Journal}. for instance) and to access the bibliographic databases of others. and you can read the details in [GMS]. allow you to re-format the entire bibliography of a paper or book simply by selecting a different bibliography style file. 2 Many subject areas have rather complete.2 Here are some sample entries from a file called mybooks. 3. year = 1986. address = "Reading. 2. which “private communications” have actually become papers) simply by keeping your master *. BibTEX will 1. allow you to share your bibliographic database (by putting it on the Web. } @INBOOK{inbook-full. Knuth". publicly accessible. month = jul. volume = 1. All very slick. volume = 41. series = "The Art of Computer Programming". which book manuscripts have been published by which publisher.bib file up-to-date.bib.. month = "10~" # jan. title = "Fundamental Algorithms". note = {The entire issue is devoted to gnats and gnus (this entry is a cross-referenced ARTICLE (journal))}.invoke an item in the database simply by using its nickname. As with any good piece of software. edition = "Second". author = "Donald E. These are updated regularly.e. Massachusetts". a typical database for BibTEX: @ARTICLE{whole-journal. ©2001 CRC Press LLC .

year = "{\noopsort{1973b}}1973". publisher = "Addison-Wesley". title = "Seminumerical Algorithms". pages = "133--139". organization = "". year = "{\noopsort{1973c}}1981". note = "This is a cross-referencing INPROCEEDINGS entry". } @BOOK{book-full. author = "F. volume = 2. year = 1988. Massachusetts". Oaho and Jeffrey D. title = "An {$O(n \log n / \! \log\log n)$} Sorting Algorithm". } @INPROCEEDINGS{inproceedings-crossref. title = "Fighting Fire with Fire: Festooning {F}rench Phrases". crossref = "whole-proceedings". title = "On Notions of Information Transfer in {VLSI} Circuits". Knuth". edition = "Second". Phidias Phony-Baloney". series = "The Art of Computer Programming". school = "Fanstord University". note = "This is a full INBOOK entry".2". } @PHDTHESIS{phdthesis-minimal. address = "Reading. type = "Section". pages = "10--119". } @TECHREPORT{techreport-full. author = "Donald E. month = "10~" # jan. Ullman and Mihalis Yannakakis". author = "Tom T{\’{e}}rrific". ©2001 CRC Press LLC . chapter = "1. note = "This is a full BOOK entry". author = "Alfred V.

\index{{\sc Bib}\TeX.tex might have this form: \documentclass{article} \begin{document} \section{This is the First of Many Sections} Now is the time for all good men to come. month = oct. " # dec. \bibliographystyle{plain} \bibliography{mybooks} \end{document} Compiling this TEX ASCII source file will result in a document with a full bibliography at the end and appropriate citations to the bibliographic entries appearing in the text. year = 1988. } @UNPUBLISHED{unpublished-full.institution = "Fanstord University". Further developments appear in \cite{techreport-full}. address = "Computer Science Department. author = "Ulrich {\"{U}}nderwood and Ned {\~N}et and Paul {\={P}}ot". sample of use} then look in \cite{phdthesis-minimal}. men. An appropriate reference for these ideas is \cite{whole-journal}. ©2001 CRC Press LLC . number = "7". title = "Lower Bounds for Wishful Research Results". note = "This is a full TECHREPORT entry". California". type = "Wishful Research Result". come. month = nov # ". note = "Talk at Fanstord University (this is a full UNPUBLISHED entry)". Come. And if that doesn’t float your boat. year = 1988. Fanstord. } Your TEX source file myfile.

4.3 Concluding Remarks
The front matter of a book will include the Preface, Table of Contents, Acknowledgements, the front and back title pages,3 and perhaps a Foreword and a brief Biography of the author. The back matter will include the Bibliography and Index and may also include Appendices, Tables (of A notation, for example), a Glossary, and other auxiliary material. L TEX has dedicated macros for creating most of these book components. We urge you to consult [GMS] or [LAM] for all the details.

3 The back title page is where the Copyright and ISBN number, and like information, appears.

©2001 CRC Press LLC

chapter 5

Copy Editing
5.1 Traditional Methods of Copy Editing
Traditionally, even as recently as twenty years ago, the copy editor worked with the hard-copy manuscript submitted by the author. He first made a pass through the manuscript in order to make notes on stylistic issues: formatting of section and chapter heads, use of remarks and examples and other enunciations, numbering of theorems, use of unusual spellings or constructions, syntax, grammar, usage, specialized words or jargon, and so forth. Then the copy editor would go through the manuscript again, correcting each page and bringing it into a consistent form that adheres to the usual customs and paradigms of the publishing business. All of the copy editor’s marks would be made by hand, usually with red or blue pencil, in the margins of the manuscript. The usual collection of copy editor’s marks (see Appendix I and Appendix II) would be exploited. The copy editor would also write queries to the author in the margins. Appendix I, Plate 4 shows how this is done. Then the author would be given a chance to respond to the copy editor’s ministrations. As an author, you should learn the copy editor’s marks (Appendix I). They will facilitate your communications with the copy editor.

5.2 Communicating with Your Copy Editor
The copy editing process is described in detail later in the book. Your primary method for communicating with your copy editor is by way of his markings and your markings on the manuscript. Learn to use the standard proofreader’s marks (see Appendix I); they are as sure and accurate a way of communicating with your copy editor as learning French is a sure way of communicating with a resident of Paris. A complete list of those proofreader’s/copy editor’s marks that are most useful to the technical writer may be found in Plates 1–4 of Appendix

©2001 CRC Press LLC

I. See also Appendix II for instances of the actual usage of these marks. The sources [SKI, p. 71] and [SWA, pp. 13, 85] contain further details about these marks, plus examples of their use. The reference [HIG] contains some useful exercises.1 You can also communicate with your copy editor using English prose, but the chance of a misunderstanding or inaccuracy will be higher. The fact that you are both native speakers is no guarantee that you will understand each other. When communications with the copy editor become bogged down or too complex—and this will happen occasionally, despite the best intentions of all parties—then you will find yourself talking to the copy editor on the telephone. Our own experience is that this happens most frequently in trying to understand graphics, and what they are meant to represent. We wish to stress that this is not a confrontational transaction. Your copy editor wants nothing more than to make your book or paper come out right. You do too. Electronic mail and FAXes are also an effective means for communicating with the copy editor. When a substantial (or surprising) change is necessary, we have sometimes found it useful to re-typeset a passage (in TEX, of course), compile and print it, and then FAX the result to the copy editor. Alternatively, one could e-mail (or ftp) the TEX code or the *.dvi file.

5.3 Communicating with Your Typesetter
Generally speaking, you will not communicate directly with the typesetter. In many instances you will be the typesetter, so the issue is moot. But even when the publisher outsources to a third-party typesetter, the copy editor or production editor will serve as a mouthpiece—both to you and to the typesetter. This is prudent, for most authors do not have the experience and the vocabulary to communicate directly with the typesetter. The copy editor can interpret the author’s wishes for the typesetter and vice versa. The editor is the unique person in the transaction who is equipped to see that the page comes out as it should.

1 Note that it will virtually never happen that you and the copy editor are sending (modified) electronic TEX files back and forth. In such a scenario, it would be too difficult to locate the passages which have been changed, and to compare the old and the new. There is too much chance for error, and for misunderstanding. The advantage of a hand-marked hard copy is that you have both the existing text and the proposed change before your eyes, and you always know what is meant.

©2001 CRC Press LLC

and this will be the author’s one and only chance to review corrections. the editor will interpret your wishes as expressed in proofreader’s marks (see also the discussion in Section 5. First. or your acquisitions editor. He is not a mathematician. the copy editor is likely working with a printout of the author’s electronic A file—either TEX. and increase the likelihood of miscommunication with the author. and will still place them in the margins of the manuscript. because he does not get lost in the meaning of the sentences he is reading. If your attempts to communicate how a graphic should appear have failed. or some other piece of document-preparation software. rendered in page proofs. Much communication with the editor will be by telephone.2). doing so would create too much potential for confusion (in part because one could not distinguish the original material from the corrected material). He will write queries to the author. can ©2001 CRC Press LLC . then a phone call. Some things remain the same: the copy editor will still use the traditional copy editing marks (Appendix I). When misunderstanding occurs.4 Communicating with Your Editor Here we speak of your developmental editor. The copy editor would never work directly with the author’s electronic file. Usually the publisher will send this edited manuscript back to the author. with considerable knowledge of the English language and syntax. friendly.5 Modern Methods of Copy Editing Today the process just described is frequently streamlined. and productive one.6 More on Interacting with Your Copy Editor A copy editor is a trained technician. thus giving the author a chance to respond. 5. or FAX. As before. Whether you have a publisher that adheres to the traditional methodology or the modern one. or even a FedEx may be necessary. In the best of circumstances. Such a person will have a good sense of English sentence structure and. in the margins. 5. take heart. your relationship with your editor will develop into a salubrious. Generally speaking. by hand. these will be recognizable because each begins with “Au:” or “Qu:” or “Qy:”. or e-mail.5. Other publishers will have the copy editor work with an already edited version of the author’s work. and the author will have another look. or Microsoft Word. You will have at least one last chance to review your beloved book before it goes to press. then a telephone conversation can often clear things up. using red or blue pencil. Then the publisher will generate page proofs. or L TEX. or your production editor (not your copy editor).

You should take the suggestions and corrections of the copy editor seriously. certain intersections or tangencies or containments must be depicted just so. In each instance where the copy editor has made an inapposite proposal. The copy editor is a serious professional. What is one to do? It is natural for one to become outraged. in a geometric figure. and communicate your true feelings so that the next draft will be more accurate. the copy editor will make blunders. This author once had the experience of a copy editor intervening in a discussion of the Riemann zeta function and changing every occurrence of “critical strip” to “critical ship”. and reply to each one with care. That clarity stems in part. in many instances you will be communicating with the copy editor rather than directly with the commercial artist. then what other travesties has he committed on your beloved manuscript elsewhere? Be of stout heart. Somehow you must react calmly to what you’ve been sent. and he is aware of his limitations. First of all. is STET (see the description of this acronym in Appendix I). There are several difficulties with this process. You may have submitted a sketch that was perfectly clear to you. the discussion of figures and other artwork in the manuscript is going to be trickier. what the copy editor is proposing may be just the opposite of what you want to say. it is difficult to articulate in words what you want the picture to show. that is perhaps why you were using a picture in the first place. but it is standard mathematical jargon. because of lack of knowledge of mathematics. Second. you merely need to write in the margin “This may be unfamiliar English. On another occasion. How dare the copy editor overstep his role and try to change your carefully chosen prose? If he is proposing this particular outrage. Often. from the fact that you yourself understand the mathematics behind the picture. But. no doubt. Indeed.” The standard editing argot. So it is agonizingly likely that the publisher’s first drafts of some of your figures will not at all represent what you had intended. Thus there really is a lot to be said for either (i) submitting a professionally hand-drawn figure (drawn locally under your direct supervision!) to the publisher or (ii) learning to use ©2001 CRC Press LLC . the artist will not know that. Just write STET after your comment and your message will be clear. You will sometimes find the copy editor recommending a certain construction just because it seems to be correct English. a copy editor changed the name “Riemann mapping theorem” to “Riemannian mapping theorem” because the name “Riemannian metric” had occurred elsewhere in the manuscript. And you will get no further argument. but it will be incorrect mathematics. Please leave as set. whose use will endear you to the copy editor. and will render the picture rather differently.be a good judge of proper phraseology and overall writing. As indicated elsewhere. The artist does not.

This is money well spent if you want your book to come out right. and other displayed matter. spacing. or some analogous piece of software.software to generate the figures for your work. line breaks. Today. The cost at a university of getting a skilled technician to generate your graphics is on the order of $30–$40 per figure (in the year 2000). you need to move or to alter a figure (in the old ©2001 CRC Press LLC . or else they meld the galley proof and page proof stages into a single stage. If you choose to go the first route. This is the moment when “what you see is what you get. but is now correct for page breaks as well. This is the original hard-copy draft that you submit to the publisher. To recapitulate. and you can stand right next to him and tell him what you want. then at least you can hire someone to do it for you. Because of electronic media. many publishers eliminate the galley proof stage. (2) The Galley Proofs. it is trivial to reformat page breaks and reposition figures. It preserves all of the format and syntax formalities that were present in the galley proofs. but working from such copy will add to the cost of producing the book or paper. Galley Proofs. or Microsoft Word. prepared from the submitted manuscript. This is a representation of how the work will ultimately appear in print. and if you are unable to generate the hand-drawn figure yourself. the page proofs will exhibit where all figures. It is not correct for page breaks. The galley proofs will not be spaced for figures. the standard drafts of a work in progress that you may see are (1) The Manuscript. shape.” It is your last opportunity to review your work before it goes to the printer. and English usage and syntax.7 Manuscript Proofs. this will generally be the printout from some electronic document preparation system such as TEX. tables. (3) The Page Proofs. and you may be able to convince your publisher to help to defray the costs. Today. and other displays will appear. This is the typeset draft of the work. 5. They will be correct for size. even at the page proof stage. numbering. Today the publisher is not significantly inconvenienced if. In particular. but now correct for formatting. and Page Proofs In other parts of this book we have made reference to different types of proofs of your work. and position. or at least to generate a figure accurate enough to communicate to the publisher’s artist what you want. Most publishers will still accept typewritten copy. tables.

and you should be allowed to criticize it just as you did with your copy edited manuscript. but also check the text on the back cover and on the flyleaves. Of course make sure that the title is correct and the author’s name spelled accurately. Are there any more steps. Many marketing departments are rather territorial. Therefore a lot of time and expense in the production process can be saved. advertisements for your book will appear before the book has actually hit the street. In today’s market. If your publisher is being particularly upfront and solicitous (professional societies are particularly good about this). you may be asked to sign a statement proclaiming that you accept the manuscript in its current form. The other part is for you to review the words on the cover. and emendations. That is the end of the editorial process. many decisions on whether to buy the book will be based on the impression that this material makes. Now let us assume that you are publishing a book (rather than an article). So be sure that it reflects well on you. usually by express mail. You certainly do not want the advertising copy to be inaccurate or embarrassing. Check out the Web sites: ©2001 CRC Press LLC . First. you will send them into the publisher. then you will be asked/allowed to review the format and content of the cover. If you have been doing your homework all along. and you will offer no further corrections. you give permission for it to go to press. is to review the advertising copy for your book. Feel free to offer suggestions. and steps can be skipped. In fact. and make sure that your wishes are met. so you want it to make a good impression. Part of this process is for you to evaluate the artistic and design qualities of the cover. and will not let even the editor see what they are creating. These last two transactions—the cover and the advertising copy— take place while the book is going into production. 5. then this declaration should be easy for you. but definitely desirable from your point of view. You should be allowed to see the advertising copy. from the point of view of the author? There are a few. Your editor should act as your broker/attorney. your book will be high in the queue at the printer’s.8 The End of the Process After you have reviewed your page proofs.days this would have been a major calamity). By the time you have approved the advertising copy. More people will read this material than will actually read the book. corrections. Let us hope that the marketing crew at your publisher’s is not like that. Less certain.

amazon.com or the pages of the Notices of the AMS to begin to develop an appreciation for your pear-shaped tones.com or http://www. In a short while.barnesandnoble.http://www. a box of books will be arriving on your doorstep. ©2001 CRC Press LLC .

©2001 CRC Press LLC . Many authors will begin by writing the manuscript out longhand. mathematical authors frequently operate under the delusion that they themselves will typeset their paper or book in TEX or some other markup language. or if your department does not have a TEX typist. there is a lot more to the process. Then either the author or his typist renders the document in a “typed up” form. however. Other journals simply will not consider papers unless they are typeset in TEX. send the disc into the publisher. when you submit the final draft of your paper to the journal. Let us suppose that you are preparing a paper with the intention of publishing it in a journal. then the journal will incur extra expense and aggravation getting your work typeset. and that is the end of it: The author’s diskette is popped into one end of a big machine and a box of books or reprints comes out the other end.chapter 6 The Production Process 6.1 Production of a Paper These days. This is so because. Not true. that a journal will expect you to produce your own artwork. If you are not yourself a TEX user. when you are writing out complicated mathematical expressions or ideas. or in some other graphic format. some journals have an explicitly stated policy to this effect. The journal may in turn decide to render those graphics as bitmaps. After the “typed up” document is prepared. Generally speaking. Many journals will give a paper not prepared in electronic form a low priority. For a first line journal or book. If you submit a paper to a journal and do not have an ASCII file on a diskette. it is really best if this can be done on a computer. that subscribes to quality typesetting practices. then you should submit repro-quality printouts of your graphics as well. or in PostScript. and always has been. It is the case now. you may actually find yourself hiring someone to do your TEXing for you. it is easiest to maintain control with a pen or pencil. That is to say. you cannot expect the journal to draw your pictures for you. Today.

if the publisher provides no instructions. These galley proofs would not have shown placement of figures. Today. twenty years ago or more. because there will probably be modifications to the paper— initiated either by you or by the referee—before the paper is in final. what the author receives really is a representation of how the article will finally appear. rendering it in the journal’s style and seeing to it that formulas. you should submit your work in hard-copy form. and there may be author queries noted in the margins. the diskette with a TEX or other suitable electronic file. presumably. it was common for the author to receive galley proofs—not broken into pages. Certainly all of the page breaks. Many journals will modify the author’s method of formatting the references. Note that. Also be careful to mail your proofs to the correct address—often it will be an address different from the one to which you originally submitted your 1 The book [KRW] contains a discursive discussion of how to deal with journal editors and referee’s reports. then instead follow the guidelines provided in this book. the author will receive page proofs for the article. and even some of the line breaks. accepted form. In due course. It is generally requested that proofs be returned within forty-eight hours. will be different from those in the preprint. This will not simply be a printout from the disc that the author submitted. accepted. and we shall not repeat that advice here. accurately. Do follow whatever instructions may be provided by the publisher for that particular journal. ©2001 CRC Press LLC . and particularly his method of labeling the books and papers listed in the references.1 Do not submit your diskette at this time. In an effort to achieve uniformity. there may be some revisions required. When you submit the final draft of the paper. and the other parts of each paper are all typeset to an exacting—and to the same—standard. how to mark corrections. tables. and other displayed matter.the manuscript is then submitted—usually in duplicate or triplicate—to a journal. Unless the journal explicitly encourages electronic submission. some journals will change the author’s numbering system. again send two or three hard copies (marked as revisions and dated). accompanied by the customary cover letter to the editor. and with dispatch. After the paper is refereed and. Do endeavor to handle your proofs efficiently. The page proofs will be accompanied by copyright transfer forms and order forms for reprints and other clerical materials. and the standard cover letter. because of electronic typesetting. Most mainstream mathematics journals outsource to a TEX consultant who painstakingly works over each paper for the journal. and theorems. These proofs will be accompanied by instructions on how to proofread the manuscript.

but there are more of them. this really needs to be said: Your book manuscript should be as polished and complete at the time of submission as a mathematics paper 2 A journal article can only be submitted to one journal at a time. then you may very well be agreeing to produce the final version that meets all the requirements of the copy and production editors. 6.) You should definitely submit your work in hard-copy form only at this time. In many instances.2 Production of a Book The steps in the production of a book are similar to those in the production of an article. After you mail your page proofs back to the publisher. ©2001 CRC Press LLC . You will have to wrestle with line and page breaks. as there will certainly be changes later on. it is easiest to maintain control with a pen or pencil. the next you will hear of your work will be on receipt of a packet of reprints and on your library’s receipt of the appropriate issue of the journal. you will have to master all the ideas in this book! Some authors will choose to simply forego the extra cash and let the publisher worry about these technicalities.2 (Of course if you are already under contract to a publisher and/or if a publisher has helped you to to prepare the document. Alternatively. Although it may seem obvious to observe. the publisher will offer to pay the author to do the TEXing. After the “typed up” document is prepared. and other formatting issues. Then either the author or his typist renders the document in a “typed up” form. Many publishers provide authors with a “manuscript preparation grant” ($2000 or so) to help defray the cost of getting the book TEXed. Today. and it should be printed on just one side of the page. We give here a sketch of the process. That is the end of the story. a book manuscript can be sent to several different publishers simultaneously. spacing. Thus you have agreed to engage in a process that goes way beyond just keying in the mathematics. running heads. One copy will suffice. if you sign a contract agreeing to produce the TEXed version of the document. with ample margins (so that editors and reviewers can make notes in the margins).paper. A warning is in order here. when you are writing out complicated mathematical expressions or ideas. In other words. By contrast. it is really best if this can be done on a computer. then you are already committed and this step is automatic. Many authors will begin by writing the book manuscript out longhand. Do not submit a disc yet. This is so because. the manuscript is then submitted to a book publisher.

Add some figures for pizzazz. the Bibliography. Include your Curriculum Vitae.4 It is also 3 If 4 Iomega that fails. You should include a detailed cover letter explaining just what this manuscript is. If the book is not too long. Preface or Prospectus. the Glossary. then the editor will simply phone you up and ask you. phone number.3 Of course the publisher will also want your electronic files. all Appendices. Therefore you may wish to use a Zip disc (which holds 100 megabytes. In fact. ©2001 CRC Press LLC . If an editor has any trouble with any of your electronic code. List of Figures. It is rare that a publisher will say. and the Index. there are several ways to do so. However.44 megabytes that a 3. and for what audience it is intended. Eventually.that you would submit to a journal. and so forth. The book [KRW] contains detailed information about dealing with a publisher over a book manuscript. Proof: George—include a plausible argument here. Zip drives and discs have become a standard in the publishing industry. e-mail address. You would never send an article into the Annals of Mathematics that said Theorem: The zeros of the Riemann zeta function in the critical strip all lie on the critical line. Enough details to be convincing. you will produce a finished product. “This is great and we will rush it to press. you will be asked to consider some revisions. then he will use your hard copy as a reference. or is not sure what a certain formula or passage is supposed to say. You should tell the publisher just who you are. Your book manuscript should be as polished and complete as you can make it: include the Table of Contents. all text. he will evaluate the reviewer reports and will render a decision about whether to pursue publication of your manuscript.5 diskettes is a perfectly acceptable medium on which to submit your book. with notes that the author has written to himself—all the time. But in fact people submit book manuscripts in this form—incomplete. plus all current information about your address.” Most likely. the publisher will send your manuscript out to various editorial advisors and reviewers. as opposed to the paltry 1. complete with all the elements described above. In an effort to evaluate your project. and be ready to submit this final draft to the publisher. then your files are liable to be quite large. After a time. You certainly will want to send the publisher a final hard copy.5 high density diskette holds). FAX number. then a couple of 3. we shall not repeat that advice here. if you are submitting PostScript files or graphics in electronic format. This hard copy represents what you think the book is supposed to look like. all illustrations and tables and figures.

The publishing industry is almost always in a “hurry up and wait” mode. to submit your files over the Internet by ftp (file transfer protocol). the page proofs are your very last opportunity to see how everything will appear—including positioning of figures. If you are doing the typesetting.possible. This is how it is going to look to your public! ©2001 CRC Press LLC . compress it with gzip or some other standard compression utility. un-taring. then the technicians at the publishing house will have no trouble decompressing. If the publisher is doing the typesetting. Thus it is considered de rigueur to send things in using express mail (Federal Express or DHL or another service). with some publishers. and front and back matter. As with a journal article. (ii) [If the publisher is doing the typesetting] Send the newly marked-up manuscript back to the publisher. along with your elecronic files. (and possibly) grammar. formatting of special sections and materials. you will either (i) [If you are doing the typesetting] Print out a new hard copy and send it. Take a few days and spend some time with your book before it goes off to press. At this point one or more copy editors will go to work on your book. and then ftp the file to a Web site that the publisher has set up for this purpose. and printing or viewing your files. You will then be sent the marked-up manuscript. The copy-editing procedure for a 400 page book could take several weeks. The editor will go over each page for style. and usage. Depending on which publishing house is handling your book. you may now engage in some give-and-take with the editors. then you will respond to the copy-editing by incorporating the editor’s instructions into the TEX source code. all page breaks. Their task is both laborious and painstaking. The copy editors will now go back to work. syntax. At the end of this process. Or you may in fact receive a (putatively) final set of page proofs for your approval. consistency. Read the publisher’s instructions carefully. then it is up to you to read all of the copy editor’s marks and either to accept them or to change them (by putting your own marks on the pages) or to challenge them (this latter option will be the exception). fine tuning the manuscript to a mutually agreeable finale. formatting. A common procedure is to tar your files together so that you are manipulating just one big file. into the publisher. If your files are in a pre-agreed upon format that the publisher can use (say TEX and PostScript).

the printer will sometimes use plastic or compressed fiber plates.3 What Happens at the Printer’s The reference [SKI. Then a technician would “shoot” the book: take a photograph of each page. 6 Refer to Appendix VI for some URLs where ghostview. The film is then used to produce the image on the plate. and charming story—fraught with much history and tradition. and are in great demand. the process just described has been streamlined.6.dvi file to film. folding the sheets into signatures. and become the master printing or lithography plate for the given page.5 And that plate would then be engraved. page 518 of that book tells all the steps involved in the production of a book by the traditional methods (that were used almost universally up until 20 years ago). and most of it is not relevant today. pp. or another6 utility will then transfer the image to film. the plate is engraved. That is the good news. But if something goes 5 Depending on the quality of hard-copy output that is desired. In particular. Then waiting to get the book printed can take another two or three months (the actual printing itself might take only an hour or so!!). We cannot take time to tell the whole story here. Now each page of the book was a photographic negative. high quality paper called RC paper. In fact. a stack of books is ejected. a machine performs all the steps—without human intervention. The bad news is that such high-tech printing operations are few in number. It is a long. Today. The publisher would print out “repro copy” on special. All the editing described above can take two or three months. At the output end of the great. ghostscript. and related tools may be found. and (as before) it is mounted on the printing press. No longer is it the case that the printer performs individual steps of printing sixteen pages to a sheet. more precisely. It would take us far afield to describe all this history. rumbling behemoth. binding the sheets together. or ghostscript. Certainly even a well-informed and high-tech author need not know the details of how the printing process works. Often the queue at the printer is the great unknown in the publishing process. ©2001 CRC Press LLC . Now there is hardware and software that will go directly from the *. But we shall give a sketch. and compiling them between flyleafs and covers. In the old days. A modern printing plant is quite a technological operation. 511–532] gives a detailed and authoritative description of all the activities that take place in a print shop. complicated. the dvi file is converted to a PostScript file and then ghostview. That negative would then be photographically printed on an emulsion-coated copper or zinc printing plate. here is what would happen once your book was ready to go into production.

wrong. or paper weight. it is helpful if you can speak to your editor knowledgeably about a reprint. The one certain fact—and this is in large part why the present book was written—is that you do not want to find yourself unable to communicate your needs or concerns to your editor. see [SKI] for the chapter and verse on these matters. or types of ink. and you also do not want to find yourself unable to understand what he is telling you. Again. or a tip-and-tear (see the Glossary). You need to learn the language! ©2001 CRC Press LLC .

the printer.chapter 7 Publishing on the Web 7. for many reasons. to publish scientific material on the Web. sun spots or cosmic rays could simultaneously damage all electronic mirror images everywhere. or other distribution systems.html. No longer are we dependent on the vagaries of the publisher. the university library. intuitive graphics interface. It has become desirable. then you will use the univer- ©2001 CRC Press LLC .2 How to Get on the Web If you are on a UNIX system. But. Since each of those “mirror copies” is stable. 7. then in a subdirectory with a name something like ~/public_html you will create a file with the name index. That file is the basis of your Web page. We encourage you to archive your work in the time-tested hard-copy format. Thanks to the World Wide Web.1 Introductory Remarks The Internet is a vast network of interlinked computers. for rapid communication and dissemination. post it on the Web tomorrow. The traditional method of archiving a book is to print a thousand hard copies of the volume and to distribute them to libraries (both public and private) around the world. we can be reasonably sure that a fire or a rodent attack at one location will not affect copies at another location. The World Wide Web provides an essentially cost-free. As of this writing. It may (ultimately) contain hyperlinks to other files. instantaneous vehicle for universal dissemination of your ideas. Write up a new theorem tonight. there does not exist any permanent and reliable protocol for archiving electronic media. If you are posting straight text. but index. the postal service. There are disadvantages to Web publishing from the point of view of archiving.html is where it all begins. material posted on many of these computers can be easily accessed from any of the others with a simple. you had better learn to use the Web. By contrast. and your colleagues in Uzbekistan can be reading it right away.

headings are denoted with <H2>.wustl. This will appear on the screen roughly as A level one heading Welcome to the world of HTML! Let’s have a second paragraph. A surfer on the World Wide Web would locate your Web page by typing in your URL (Uniform Resource Locator). As you can see. America Online and other ISPs have their own utilities. bold font. The <P> tag indicates a new paragraph.sal Web language HTML (Hypertext Markup Language). these are differentiated by the presence of a forward slash / in the latter. Other tags are similar. HTML is an extremely straightforward.edu/~sk You can see that the address consists of the node in our mathematics department followed by an identifier for the individual person. If you are an America Online user. Private Internet providers will have their own protocol for creating a Web page. The Web address of Steven G. Some tags. so this “sample” should be taken only as a suggestion of how HTML code is translated into typeset text. have a beginning tag and an ending tag. or an EarthLink user. such as <TITLE>. high-level computing language which allows you to issue standard text formatting commands for size. spacing. Of course what the end user actually sees will depend on what browser he is using. or Web address. The <H1> tag indicates a level one heading. HTML uses tags to tell the Web browser how to format text. or a Compuserve user. But it will be similar to what we have described for UNIX. which is a large. then you will have to learn their particular paradigm for creating a Web page. Third-party products—such as Micrografx’s ©2001 CRC Press LLC . Very simple. etc. Here is a typical bit of code for an HTML file.math. taken from the book [GOR]: <TITLE>The simplest HTML example</TITLE> <H1>A level one heading</H1> <P>Welcome to the world of HTML! <P>Let’s have a second paragraph. font. and so forth. Smaller. The title “The Simplest HTML example” will appear in the title bar of the browser. Krantz is typical: http://www.1 1 There is quite a lot of software available for creating Web pages. lower priority.

auc. AucTEX AucTEX is an add-on to the UNIX editor GNU emacs. this is a preprint server for the sciences.ps. publishers. MiKTEX.7.tex This is an electronic newsgroup on electronic document preparation. It is an extensible package that supports writing and formatting TEX files. and other typesetting ideas. New preprints are posted without human intervention (apart from the actions of the author to post his work). and a model for how preprint servers should be run.math.arXiv. the information on the Web about mathematematical publishing is growing apace. is a terrific source of information about troff. It would be impossible here to give anything but a hint of some of the Web resources available to the mathematical author. Available from the Web page http://www.text. and can be downloaded in *. and computer scientists. Webtricity.pdf. TEX. You can download AucTEX from the Web site http://www. Ispell.washington. *. IBM’s Top Page.) Bibweb This is a utility for automatically retrieving bibliographical information from the American Mathematical Society’s MathSciNet program. (Some of these are discussed below.0—may be purchased at any computer store. *. Requires the use of BibTEX.math. called tex-faq. The useful newsgroup also contains the utility TeX-index. PostScript. Their FAQ (list of “Frequently asked Questions”). Thanks to the efforts of many mathematicians. Sierra Home’s Complete Web Studio. We mention just a few favorites.edu/~palmieri/Bibweb/ comp. which is an archive of all public domain TEX macros.html That site also contains links to sources for emacs. and Lacheck. and SoftQuad’s HotMetal Pro 6.dk/~dethlef/Tips/auc. This software provides an interface. ©2001 CRC Press LLC . arXiv Available from www.3 Web Resources The World Wide Web is growing rapidly as a source of information about many topics. A 60+ page manual is included in the distribution of AucTEX. and other formats. the arXiv has become something of a national resource.org. very much like a word processor. ghostview. A particularly those created in L TEX and AMS-TEX. between the user and HTML. Initiated by Paul Ginsparg at Los Alamos.dvi. source code. There is also a link to an on-line version of the manual.

New journals are continually being added.math. and Sociology. it is needed for running Ispell.org/mathscinet. then there is no limit to the number of its people who can read the same article at the same time. Articles may be downloaded and printed. Learn more about JSTOR at http://www. Type in key words and MathSciNet will return a list of all papers with those words in the title. Learn more about MathSciNet at http://www. In fact. Type in the name of an author and in a few seconds MathSciNet will return complete bibliographical references for all his papers. MathSciNet is not free. If a university subscribes to JSTOR. JSTOR The Mellon Foundation’s ambitious project to provide electronic archiving for major scholarly journals.html Ispell is a spell-checker for all types of documents. Philosophy. Front is a very user-friendly front end for arXiv.ams.auc.Front Greg Kuperberg’s contribution to the quality of mathematical life. But it is worth it. the American Math Society requires that you subscribe to the hard-copy Math Reviews before you can subscribe to MathSciNet (for an additional fee). or in the Math Reviews review. If you input basic reference data into MR Lookup. Type in a full or partial title and MathSciNet will return all papers with that text in the title. Other fields covered range from Afro-American Studies to Economics. MathSciNet is also a great tool for seeing what that “special someone” has been up to in the past few years. You can also download Cygwin from the Web site. MR Lookup This is a new utility for accessing the MathSciNet database to verify and create references that can link to reviews and original sources. Downloadable from the Web site http://www.math. in this author’s opinion MathSciNet is the most important new mathematics resource to come along in many a year. There are now about 130 journals archived on JSTOR.auc.edu.dk/~dethlef/Tips/auc.html. then it returns electronic publication-ready references with live links to reviews ©2001 CRC Press LLC .ucdavis. MathSciNet Created by the American Mathematical Society. 11 of these in mathematics. You can try out Front for free at http://front. Ispell Also available from the Web site http://www.dk/~dethlef/Tips/auc.math.org. Lacheck This is a utility for checking cosmetic errors (at a rather recondite technical level) in a TEX source code file.jstor.

and especially about publishing mathematics. that you have created a mathematics document—perhaps a research paper. or a chapter of a book.pdf files. mathematicians will create their documents in A some form of TEX. and at a rapid rate.pdf (where pdf stands for “portable document format”). Not every Web browser supports the code. 4 Any description of “what publishers think” is going to depend on the particular publishing house.ams.4 So let us assume. Now you want to put it on the Web. So whatever we say about it here should be understood in that context. and how easily. just because it gives the author fewer choices and because it emphasizes logical markup over visual markup.pdf file.org/mrlookup. use the utility latex2html to translate your A L TEX file directly into HTML. it will turn 2 It should be noted that Greek letters are now part of the ISO 10646 standard. Alternatively. Available from the AMS Web site http://www. and view it as a black box. What do you do? The simple answer is that your document needs to be rendered in *. There does not even exist universally readable HTML code for the Greek letter π.in MathSciNet and to original articles. and on what that individual’s position is. Then anyone with an Acrobat reader will be able to read your document. HTML does not have an international standard (like the ANSI Standard in C). A member of the publisher’s production department will take a more technical and empirical view of TEX.dvi file into a *. You can. or an article for the Notices of the American A Mathematical Society—in L TEX. If your system is so equipped. Most editors are not fluently conversant with TEX. turn your TEX file directly into *.pdf code.5 An Acrobat file can be recognized by the extension *.2 The mathematician seeking to post his ideas on the Web is best advised to use another medium. ©2001 CRC Press LLC . and they are implemented in the HTML specification.pdf format. Adobe Acrobat is designed to read and manipulate *. on the particular individual expressing the opinion. and SGML is still under development. however. You can use the utility dvipdf to turn your *. As of this writing. They see TEX in terms of what end product it will produce. 3 Everything about publishing. The utility pdftex will. 5 Refer to Appendix VI for some URLs where Acrobat-related tools may be found. if not most. with formatting and resolution quality comparable to that of TEX.4 Mathematics and the Web The interesting issue about Web publishing from the point of view of the mathematician is that HTML cannot handle mathematics. Most publishers prefer L TEX. if you wish. in fact.3 Many. on the Web is in flux. for the sake of this brief discussion. either Adobe Acrobat distiller or ghostview will turn a PostScript file into *. 7.pdf. It is constantly developing.

At that moment.pdf.pdf graphics files. ©2001 CRC Press LLC . Whereas—earlier—we touted encapsulated PostScript as the way to go for incorporating graphics in a TEX document. the end user will see a document with the meaning drastically changed. and *. or a business document.png. But for scientific data. practically. These days there are a great many utilities available for putting your A TEX or L TEX document on the Web. The pdftex utility can handle *. The utility pdftex is included in the TEX implementations MikTEX and fpTEX. a small slip may entirely change the intended meaning. we now must tell you to translate (using ghostview or Adobe Acrobat distiller. There will be no error message! In that case. The result is a display that is difficult to read. for example) those *. It is quite possible that the end user’s browser will format things differently than the author intended. or for a table of vital information. the paper can appear. or it could make font substitutions.jpg. Everyone wants to have their work posted on the Web. and in what form. Some do not handle fonts well. Some cannot handle scaling. and perhaps legalistically—on the person publishing on the Web is this: What the end user sees is highly dependent on the particular Web browser that he is using. If a Web browser encounters a piece of code that it does not understand. the publisher will tell you in no uncertain terms to take it down. But as soon as the publisher has the rights to the paper or book then he can exercise control over where. and to offer some guarantee of coherence and integrity in the transmission of information. both of which may be downloaded free from the Web. then you must be sure that these are in the proper format. for example. If your document includes graphics. however. *.eps files to *. For casual reading these differences may not be important. Let us note. it does not depend in any essential way on the browser that the promulgator is using. Many of these questions are still being argued in the courts. Suffice it to say that they all have their strengths.6 Another set of legal issues concerns copyright infringements. and they all have limitations. Space limitations prevent us from discussing any of these in detail. or for a mathematical formula. that if you are writing a book for a commercial publisher then at some point you may sign the rights over to that publisher. Use of Acrobat tends to minimize these differences. Some 6 Refer to Appendix VI for some URLs where Acrobat-related tools may be found.your mathematics characters and formulas into bitmaps. The publisher will probably be happy to have you post your book on the Web up until the time the book appears in (hard copy) print. One big issue that impacts—philosophically. then it will simply pass over that piece of code. and files that are not very portable. So you must be careful. The situation with research papers is similar. Some are accurate but very slow.

you store documents in the Mathematical Markup Language MathML. attributes. We briefly will mention just a few. It includes tools for dispalying math. It may skip over markup commands that occur in a place where it does not expect them. A The language WebTeX is not quite TEX and not quite L TEX. It treats several others as well. Web Sites These days. It requires multiple passes to make everything come out right. you manipulate the *. 7. it is a version of these languages adapted for the Internet. it is becoming more and more common to have software that is connected in some way to a hard-copy book or article. Some do not interact well with certain Web browsers. The monograph [GOR] gives the full picture of many of the most popular of these tools.tex file. Some of the ways that this connection is implemented are as follows: 7 Here MathML is a markup language to be used with XML. in effect.tex file. The IBM product techexplorer allows you to. XML. The book [GOR] gives detailed tutorials in the uses. If you are going to use WebTeX. and XML is an extensible markup language that improves on HTML in many ways. Note that techexplorer does not treat the *. SGML. is touchy about user-made macros. The utility WebEQ is a popular Java applet for rendering mathematics in a browser.7 Therefore much of the “attribute customization” that can be performed on an HTML document can also be performed on a TEX A or L TEX document. With TeX4ht. and requires that you enter your markup commands in a certain order. instead it directly treats the *. it is troublesome to handle graphics with TeX4ht. and MathML. With WebEQ. manipulate A a TEX or L TEX file as though it were an HTML document. and limitations of all the utilities described here.dvi file. then you will have to do some retooling.dvi file rather than the *. ©2001 CRC Press LLC . It retains many A features of TEX and L TEX but discards others. With techexplorer you will manipulate your document in MathML.5 Software to Go with your Book or Article. In fact. We recommend it highly to the stalwart reader who really wants to understand how all these utilities work. The utility techexplorer does not support style files. and often post-processing of loose-end tasks is required. Generally speaking. A The utility TeX4ht creates configurable hypertext versions of L TEX documents. and furthermore describes the differences and relationships among the markup languages HTML.have trouble with graphics.

This is not the place to describe in detail just how to put software on a Web site. There may. (iii) additional information. each of which leads to a download. He puts the *. or Macintosh format. or perhaps the source code. The URL is provided with the hard-copy material. or other ancillary material that would not be attractive to include in the hard-copy material. or additional examples. you can give your site an attractive and readily understandable graphic interface and you can also provide some text that will lead the user to the particular choice that will satisfy his needs. and then provide the files in several formats. The really classy thing to do is to put your files on a CD-ROM. on the Web site as well. or arcane and lengthy examples. (ii) examples. With a little extra effort. ©2001 CRC Press LLC . either in C. Along the tree that leads to your software. the author may wish to place the TEX code. Consult with your publisher before you put your entire book or MS on the Web.) In that instance. If you are writing a book.ps files for the entire text of the document. (For example.dvi or *. or C++. you must decide whether to render the files in Windows format. and how to download. and you will create a button on your home page that will lead users to it. (iv) FAQs (Frequently Asked Questions). or some other format. however. it contains TEX macros that are discussed in the book. [SAK] has such a diskette. or color figures. which has plenty of storage space. Whatever choice you make will of course exclude some readers. Suffice it to say that you will put it somewhere in your public_html directory. He instead puts them on a publicly accessible Web site. • The author has created executable software. It is up to the user to figure out what is what. or Fortran.exe file. that he wants the reader to be able to download and use. and (vi) hyperlinks to related material. • The author has written a textbook. You will be doing your readers a great favor by making your Web site as user-friendly as possible. be copyright or legal or fiduciary considerations that need to be worked out before doing so. or animated graphics. or at least access to one). on a Web site. and may want to put worked problem solutions. or Java. In any of these instances. you may wish to include a diskette with the book. Many Web sites consist of little more than a list of hyperlinks. (v) other references and Web sites. or review material. or some other standard programming language (for which most readers will have a compiler. you will offer options including (i) documentation. or perhaps the *. together with suitable documentation.• The author has a number of specialized references. on a publicly accessible Web site.

44)-(398.MI)*MI)*BB The more fundamental question is what syntax to use.12). Many people are still quite naive and inexperienced at using a computer.44):NEXT J LINE (394. Providing a little help is your job. then you may want to include the code for Euler’s method or for Runge-Kutta. here is a fragment of code that this author wrote (many years ago) for a utility for the pharmaceutical industry to calculate volumes and surface areas of pills (tablets): 1573 1577 1580 1590 1593 1597 1600 1610 1615 1617 1618 1620 1623 1624 1625 1635 FOR J = 0 TO 16:PRESET (550+J. It is now quite standard to use the “typewriter-like font” for computer code. If the reader cannot figure out what to do with your disc then he will become frustrated and may set your book aside. The code that we just exhibited is Turbo BASIC. If it is a book about dynamical systems. then you may want to include the code for generating the Mandelbrot set. That is not the effect that you want. invest a few more making it easy for people to use the disc. The experienced hacker will know just what to do.CD^2)) FOR J = 0 TO 31:PRESET (428+J. that is untrue.25 + (MA . If it is a book on numerical methods.CD^2 < 0 THEN GOTO 2130 G = SQR(ABS(2*MINR*CD . Should your code instead be in C.35):NEXT J LINE (383. For instance.MINR IF MINR < CD THEN GOTO 2130 IF 2*MINR*CD .12) INPUT "MINOR RADIUS (FIG DIM MINR) ".It is a nice touch to include a few pages in the book which explain how to use the provided disc or how to access the Web site. If you provide them with nothing but a disc then they may just assume that it is a self-booting disc that will run all by itself. Likely as not. but the neophyte will not.12)-(458.INC IF INC < 0 THEN GOTO 2130 FOR J=0 TO 8:PRESET (394+J. A final note: Sometimes it is attractive to include written-out computer code in the book you are writing.44) INPUT "THICKNESS OF BELLY BAND (FIG DIM BB)".BB IF BB < 0 THEN GOTO 2130 FOR J = 0 TO 16:PRESET (383+J.(401.12) INPUT "LAND (FIG DIM L) ". which most people would consider to be Neantderthal. there may be a readme file or an executable or a batch file. as the case may be. or in Fortran.12):NEXT J LINE (428. or in ©2001 CRC Press LLC . After all the hours that you will invest making the disc.12):NEXT J S2 = (PI*(MI^2)*.

coming from a specific language. ©2001 CRC Press LLC . by paradigm. In any event. C is a nearly universal scientific programming language these days.APL. It will tell. to be much clearer than pseudocode. I find true code. So is Java. or should it be in pseudocode?8 Speaking from my own experience. In my own work. Perhaps it will work in the same way for you. made-up code that comes from no particular language. the well-informed reader how to write code in the language of his choice. 8 Here the jargon “pseudocode” denotes a generic. Since pseudocode does not have a generally accepted syntax. I would recommend that you use one of these. I have found that attempting to understand an author’s provided language-specific code gave me just the motivation I needed to learn more about C. I often do not know what the writer’s pseudocode means.

©2001 CRC Press LLC .

©2001 CRC Press LLC .

©2001 CRC Press LLC .

©2001 CRC Press LLC .

9 We have used marks from our tables as well as some new marks that should be clear ©2001 CRC Press LLC .

©2001 CRC Press LLC .

Lower-Case Greek Letters α β γ δ ζ η θ ι κ \alpha \beta \gamma \delta \epsilon \zeta \eta \theta \iota \kappa λ µ ν ξ o π ρ σ τ υ \lambda \mu \nu \xi o \pi \rho \sigma \tau \upsilon φ χ ψ ω ε ϑ ς ϕ \phi \chi \psi \omega \varepsilon \vartheta \varpi \varrho \varsigma \varphi Observe that the Greek letter “omicron” is just the same as the lowercase roman “o”. in table form. the standard mathematics symbols that make up most mathematical manuscripts that you will see. . The characters ε. Capital Greek Letters A B Γ ∆ E Z H Θ I K A B \Gamma \Delta E Z H \Theta I K Λ M N Ξ O Π P Σ T Υ \Lambda M N \Xi O \Pi P \Sigma T \Upsilon Φ X Ψ Ω \Phi X \Psi \Omega ©2001 CRC Press LLC . We begin with the Greek alphabet. ϕ are. ϑ. These are typeset using the computer typesetting system TEX. in effect. θ. . φ. π. ρ. ς.Appendix III: Specialized Mathematics Symbols We present here. σ. script versions of the lower-case Greek letters .

and blackboard bold.One can typeset slanted upper-case Greek letters by using the command \mit. It is convenient. Boldface capital Greek may be typeset in an obvious way: the code $\bf \Gamma + \Delta$ typesets to Γ + ∆. here are the characters of class Ord: Characters of Class Ord \aleph \hbar \imath \jmath \ell ℘ \wp \Re \Im ∂ \partial ∞ \infty ℵ h ¯ ı  ∅ ∇ √ ⊥ ∀ ∃ \prime \emptyset \nabla \surd \top \bot \angle \triangle \forall \exists ¬ \ ♣ ♦ ♥ ♠ \neg \| \backslash \flat \natural \sharp \clubsuit \diamondsuit \heartsuit \spadesuit Large Operators of Class Op \coprod \bigvee \bigwedge \biguplus \bigcap \bigcup \int \prod \sum \bigotimes ∫ \bigoplus \bigodot \oint \bigsqcup \smallint ©2001 CRC Press LLC . which is invoked by the font call \frak and has the form Euler fraktur. For example. from the point of view of TEX. These include Euler fraktur. the code $\mit\Phi=\Psi(\Delta. With that paradigm in mind. Γ ). to group characters according to their “spacing type”. A standard AMS-TEX implementation will contain certain commonly used fonts for mathematicians. The characters ABC DE F G HI J KLMN OP QRS T U V W X Y Z are typeset by using the command \cal inside the TEX math environment to invoke the calligraphic alphabet.\Gamma)$ will compile to Φ = Ψ (∆. which is invoked by the font call \Bbb and has the form BLACKBOARD BOLD.

Symbols for Footnotes ∗ * § \S Binary Operations † \dag \| ‡ \ddag ¶ \P × ÷ ∓ ± ∧ ∨ ‡ † \times \div \mp \pm \wedge \vee \ddagger \dagger \amalg \diamond \wr \cap \cup \sqcap \sqcup \uplus • \bullet \odot \oslash ⊗ \otimes \ominus ⊕ \oplus ∩ ∪ ◦ \ · ∗ \triangleleft \triangleright \bigtriangleup \bigtriangledown \circ \bigcirc \setminus \cdot \ast \star Relations ∝ \propto \sqsubseteq \sqsupseteq \parallel | \mid \dashv \vdash = \ne ∈ \notin / \not ≤ \le ≥ \ge \succ ≺ \prec ≈ ∼ ∼ = ⊃ ⊂ ⊇ ⊆ ∈ \approx \sim \simeq \cong \succeq \preceq \supset \subset \supseteq \subseteq \in \ni \gg \ll → ← ↑ ↓ ⊥ ≡ . = |= ⇐⇒ \to \leftarrow \uparrow \downarrow \perp \equiv \asymp \doteq \smile \frown \bowtie \models \rightleftharpoons \iff ©2001 CRC Press LLC .

30.Punctuation Marks and Ellipses . ∝ 50. \vdots \ldots \dots . ≷ 30. 70. 20... ≪ ≫ 60.. 10. 20... 70. . . 40. ≶ ©2001 CRC Press LLC . 50. ··· \ddots \cdots \nearrow \searrow \nwarrow \swarrow → \mapsto ⇐ \Leftarrow ⇒ \Rightarrow ⇐⇒ \iff ← \leftarrow Delimiters \Leftrightarrow \leftrightarrow \Longrightarrow \longrightarrow \longleftarrow \Longleftarrow \longmapsto \to \uparrow ←→ ⇐⇒ → ← ↓ \longleftrightarrow \Longleftrightarrow \hookrightarrow \hookleftarrow \leftharpoonup \leftharpoondown \rightharpoonup \rightharpoondown \downarrow ( [ { | / ( [ \{ | \| / \     \langle \backslash \lfloor \lceil \lgroup \lmoustache ↑ ↓ ⇑ ⇓ \uparrow \downarrow \Uparrow \Downarrow \Updownarrow Next is a display of the two most standard AMS-TEX symbol fonts: AMS Symbol Font MSAM10 0 0. 40. 1 2 ⇔ 3 ⇒ 4 5 6 ♦ 7 8 9 ∴ A ∵ ∠ B C D E F 0. 60. 10. \ldotp · \cdotp : \colon Arrows ⇔ ↔ =⇒ −→ ←− ⇐= −→ → ↑ . . . .

70. 60. 40. 50. Math Abbreviations \arccos \arcsin \arctan \arg \cos \cosh \cot \coth \csc \deg \det \dim \exp \gcd \hom \inf \ker \lg \lim \liminf \limsup \ln \log \max \min \Pr \sec \sin \sinh \sup \tan \tanh ©2001 CRC Press LLC . 60. 50. 20. 30.AMS Symbol Font MSBM10 0 0. 1 2 3 4 ≮ 5 ≯ 6 ⊀ 7 8 9 A B C ∦ D E F 0. P A Q B R C S ∼ D T ≈ E U F V G W ð H X I Y J Z ‫ג‬ K κ L k J N ∅ O We conclude with the commands for the math abbreviations (or function names) included in TEX. 70. 30. 20. 40. 10. 10.

Upper Case Alphabets roman A B C D E F G H I J K L M N O P Q R S T U V W X Y Z boldface A B C D E F G H I J K L M N O P Q R S T U V W X Y Z italic slanted roman A A B B C C D D E E F F G G H H I I J J K K L L M M N N O O P P Q Q R R S S T T U U V V W W X X Y Y Z Z calligraphic fraktur A A B B C C D D E E F F G G H H I I J J K K L L M M N N O O P P Q Q R R S S T T U U V V W W X X Y Y Z Z ©2001 CRC Press LLC . These are rendered using TEX.Appendix IV: Standard Alphabets We now exhibit some of the most standard fonts that are used in modern typesetting.

Lower Case Alphabets roman a b c d e f g h i j k l m n o p q r s t u v w x y z boldface a b c d e f g h i j k l m n o p q r s t u v w x y z italic slanted roman a a b b c c d d e e f f g g h h i i j j k k l l m m n n o o p p q q r r s s t t u u v v w w x x y y z z calligraphic fraktur a b c d e f g h i j k l m n o p q r s t u v w x y z Observe that there are no lower case calligraphic letters in TEX. ©2001 CRC Press LLC .

....f k . f . . .... . jm . k1 . w∗ . .jm a(j1 .. j2 ... lim inf exp (x3 − y 2 )/(xy + 2) Remarks Option 2 more common Option 2 better in text Option 1 better in text Option 2 better in text Option 1 better in text e √ n x + ym x+2y 3 n j=1 x3 −y 2 xy+2 (xn + y m )1/2 (x + 2y)/3 . w .. .Appendix V: Alternative Mathematical Notations Here we present various typesetting situations together with some choices in how to address them. . A . A = x y zw f x y zw Option 2 preferable ©2001 CRC Press LLC . . F(f ) . lim Option 2 lim sup . 4 3 k =0 .. x A . jm ) a(j1 ..k e f C→D X ∪Y v f f :C→D Cl(X ∪ Y ) v ˙ ¨ ˜ A . 4/3 n k . . . x# ¯ log 5 2 y b z c− log(2/y) [c − b/z]1/5 f (A) . . A . .jm 1 p aj1 . Option 1 lim . kp ) Ce. j=1 =0 Option 2 better for display Option 2 easier to read Option 2 easier to read Option 2 better in text Option 2 easier to read Option 2 better in text Option 2 better in text Option 2 better in text Option 2 easier to read aj1 . .

Appendix VI: TEX, PostScript, Acrobat, and Related Internet Sites
There are a number of Internet sites that archive TEX fonts, macros, and other utilities that will be of interest to the TEX user. We record some of the ftp addresses and Web sites here. Always bear in mind that Web sites are transient. They are constantly in flux, they change form, they move, and (sadly) sometimes they disappear. As of this writing, all the given sites are operational. As an alternative to logging on to one of the given sites, you can use one of the utilities archie, gopher, or wais to search for the package you are looking for. Or else use the altavista, or yahoo, or google Web search engines to conduct a search.

Sites from which to ftp TEX Packages
ftp.math.utah.edu /pub/tex/tex-index This is Nelson Beebe’s server, and it is particularly strong in information about BibTEX. It also contains a concordance of all articles in TUGboat, the publication of the TEX Users Group. labrea.stanford.edu This is the “official” Stanford University repository for TEX, METAFONT, dvips, and related files. http://www.tug.org This is the site maintained by the TEX Users Group. It is a source for emTeX, publicTeX, publicMF, and many other fonts and macros and TEX utilities. e-math.ams.org This is the ftp site of the American Mathematical Society (AMS). The login and password are both e-math. The Society is the creator and promulgator of AMS-TEX, and the site A makes both AMS-TEX and AMS-L TEX available for downloading. Once you are logged on, use your <Up>-<Down> keys to highlight <Sitemap> and then press <Enter>. Again use your <Up>-<Down> keys to scroll to <TeX Resources> and again press <Enter>. Now you are in the AMS’s world of TEX. There are numerous fonts, macros, style files, author

©2001 CRC Press LLC

packages, manuals and other TEX resources available at the e-math site. ftp.dante.de /pub/tex/help/TeX-index

is a source of the “Mainz” packages of Mittelbach and Sch¨pf o (these resources include multicol, verbatim, theorem, NFSS, ftnright, and array). It is a mirror site of ftp.tex.ac.uk, and has the Web address http://www.dante.de.

ftp.tex.ac.uk /pub/archive/help/TeX-index This is a CTAN site (a mirror of ftp.dante.de), and also the home of the UKTeX newsletter. The corresponding Web address is http://www.tex.ac.uk.

Other URLs for TEX Goodies
http://199.26.180.160/winnt/misc/page2.html An implementation of BCTEX may be downloaded from this site. ftp://ctan.tug.org/tex-archive/systems /win32/miktex/1.20/ MikTEX may be downloaded from this site. http://www.esm.psu.edu/mac-tex/versions.html This is a TEX Web site for Macintosh users. http://www.tug.org This is the Web site of the TEX Users Group. It has pointers to several different CTAN sites. http://www.cs.wisc.edu/~ghost/ This site is a source for the utility ghostscript, for manipulating PostScript. http://www.adobe.com Adobe’s home page. Adobe is the the company that created Acrobat and PostScript. http://www.math.uakron.edu/~dpstory/acrotex.html

©2001 CRC Press LLC

A repository for PDF-based math tutorials. Makes great and extensive use of Acrobat forms. http://www.tug.org/applications/pdftex/ A resource of pdfTEX examples. http://www.pdfzone.com/ The PDF Zone. http://www.adobe.com/prodindex/postscript/main.html Information about PostScript at Adobe. ftp://ctan.tug.org/tex-archive//support/latex2html/sources/
A Sources for L TEX2HTML on the CTAN site.

http://www.5-d.com/niknak.htm This is a commercial PostScript -to-PDF converter. http://www.wargaming.net/Programming Adobe PostScript resources. http://math.berkeley.edu/~vojta/xdvi.html This is Paul Vojta’s (he is a math professor at U. C. Berkeley) X-Windows TEX previewer. http://www.yandy.com/ The home page of Y&Y TEX. http://www.cs.wisc.edu/~ghost/ Source for Ghostview, Ghostscript, GSview. http://godel.ph.utexas.edu/Members/timg/gs/gs.html/ Source for Atari Ghostscript. ————————————————————– For those not entirely conversant with the process of conducting an ftp transaction, we now provide a little guidance. On a UNIX system, log on and get to a system prompt. On a Macintosh machine, go to the ftp icon. Instructions for America Online are given below.

©2001 CRC Press LLC

To go to the first ftp site given in this Appendix. On other systems. Type cd pub followed by <Enter>. this will be done by typing ftp. or with a browser. click on <Keyword>. and then click on <Go>.math. ©2001 CRC Press LLC .) Perform the logon by entering “anonymous” and using your e-mail address as a password.edu in a suitable window and clicking on <Go> or a similar icon. Then type cd tex followed by <Enter>.utah.math. Other Web browsers have similar protocols. Now you will be at Nelson Beebe’s famous TEX site. (On a UNIX system.utah. ftp now to the ftp site ftp.edu at your system prompt. type ftp://ftp.utah.utah.edu.math.edu in the URL search window and press <Enter>. With America Online.math. With Netscape. Enjoy! Of course you can always use your Web browser to ftp. this step is performed by typing ftp ftp. type ftp in the keyword window.

ˆ f) \.g.g.... This list will supplement. Command “\ ” Meaning or Purpose hard single space \.g. Our wish is that this will serve as a quick-and-dirty reference for the beginner.g.g... and also a quick review for the more experienced TEX user.g. ¯) e !‘ ¡ ?‘ ¿ $ $ inline math mode \( \) inline math mode $$ $$ display math mode \[ \] display math mode \b bar-under accent in text (e.g.App endix VI I: Basic TEX Commands We provide here a concordance of the most basic commands in TEX. a) ´ \AmSTeX logo for AMS-TEX A \AmSLaTeX logo for AMS-L TEX \approx approximately equal to (≈) \backslash backslash (\) ©2001 CRC Press LLC . \H long Hungarian umlaut in text (e.. and in some instances duplicate. a) ´ \‘ accent grave in text (e.. thin space \! negative thin space \’ acute accent (or egout) in text (e...g. o) ˙ \= macron in text (e.g.. oo) \u breve in text (e.g.g. o) ˝ \t tie-after accent in text (e. o) ˘ \v h´ˇek or check in text (e. dot accent in text (e. the present list also contains formatting and text commands.g.... While those tables emphasized mathematics symbols. e) ¯ \c cedilla (e.g.. a) ¨ \^ circumflex in text (e.g. c) ¸ \d dot-under accent in text (e. o) ` \~ tilde in text (e.g.. ˇ) ac a ˚ \AA A \aa ˚ a A^b superscript Ac A_c subscript Ac \AE Æ \angle angle ( ) \acute acute accent in math mode (e. o) . what appears in the tables in Appendix III. ˜) e \" dieresis or umlaut in text (e.

like Laplacian ( ) large vertical space the “Halmos tombstone” ( ) breve in math mode (e. C ) intersection sign (∩) formats a function defined by cases small. d) left delimiter enlargement size left delimiter enlargement size middle delimiter enlargement size middle delimiter enlargement size right delimiter enlargement size right delimiter enlargement size left delimiter enlargement size left delimiter enlargement size middle delimiter enlargement size middle delimiter enlargement size right delimiter enlargement size right delimiter enlargement size large triangle. centered dots (an ellipsis · · ·) command for centering a line of text ˇ a “check” accent in math mode (f ) binomial expression ( a ) b small. N.\bar \Bbb \bf \biggl \Biggl \biggm \Biggm \biggr \Biggr \bigl \Bigl \bigm \Bigm \bigr \Bigr \bigtriangleup \bigskip \Box \breve \bullet \bye \cal \cap \cases \cdot \cdots \centerline \check \choose \circ \cos \cup macron in math mode (e..g.. u) ¯ blackboard bold font (C. centered circle (f ◦ g ) cosine union sign (∪) ©2001 CRC Press LLC . B. R) boldface font (a. centered dot ( · ) three small.g. b. a) ˘ typesetter’s bullet (•) command for ending a TEX document the math mode calligraphic font (A. c.

c. as in a footline command for creating a running foot creates a footnote for all (∀) the Euler fraktur font (f . a) ` command for creating hangindents ˆ a caret “hat” accent (f ) an environment for placing text into math command for creating a running head command for filling out a line with space command for filling out a line with space a solid black box ( ) horizontal skip integral ( a ) imaginary part of ( ) the dotless “i”.\dagger \ddagger \ddot \def \div \dot \dots \ell \emptyset \eqalign \eqalignno \eqno \equiv \exists \folio \footline \footnote \forall \frak \ge or \geq \gg \grave \hangindent \hat \hbox \headline \hfil \hfill \hrule \hskip \int ... h) greater than or equal to (≥) double greater than ( ) accent grave in math mode (e. for use in mathematics vertical space horizontal re-positioning l b ©2001 CRC Press LLC . lower-case script ell ( ) empty set (∅) utility for aligned displays \eqalign with numbering equation label at right margin is defined to be (≡) there exists (∃) page number.g. . .g. \int_a^b \Im \imath \in \indent \infty \input \it \item \itemitem \itemitemitem \jmath \jot \kern \l typesetter’s dagger (†) typesetter’s double dagger (‡) double dot math accent or dieresis (x) ¨ macro definer division sign (÷) dot accent in math mode (e. d) delineates an element of a list delineates a sub-element of a list delineates a sub-sub-element of a list the dotless “j”. a) ˙ three small. baseline dots (an ellipsis . g. for use in mathematics is an element of (∈) paragraph indentation infinity (∞) command for including macros italic font (a. ) script. b.

spacing in document specifies magnification size maps to (x → y ) defines a new “math operator” expression an invisible box used for spacing matrix medium vertical space a vertical bar for use in setbuilder notation (|) formats text with wider margins for display not equal to (=) nabla (∇) arrow point northeast ( ) arrow point southwest ( ) has as an element ( ) prevents indentation creates a document with no page numbers puts a slash ( ) through the next symbol Ø ø Œ œ complex line integral ( ) a minus sign in a circle ( ) a plus sign in a circle (⊕) a times sign in a circle (⊗) used in making fractions a brace over text (x + y ) a line over text (x + y ) paragraph indentation a partial differentiation sign (∂f ) perpendicular sign (⊥) plus or minus (±) matrix with parentheses a double em quad an em quad ©2001 CRC Press LLC .\langle \LaTeX \le or \leq \left \leftarrow \Leftarrow \leqno \ll \llap \magnification \magstep \mapsto \mathop \mathstrut \matrix \medskip \mid \narrower \ne or \neq \nabla \nearrow \nwarrow \ni \noindent \nopagenumbers \not \O \o \OE \oe \oint \ominus \oplus \otimes \over \overbrace \overline \parindent \partial \perp \pm \pmatrix \qquad \quad left inner product delimiter ( ) A logo for L TEX less than or equal to (≤) controls size of left delimiter leftward pointing arrow (←) leftward pointing double arrow (⇐) equation number at left margin double less than ( ) command for hang indenting controls size of font.

\rangle \Re \right \rightarrow \Rightarrow \rm \sc \searrow \setminus \sim \simeq \sqrt \ss \supset \supseteq \swarrow \sin \sl \smallskip \subset \subseteq \tan \TeX \tilde \times \tt \underbrace \underline \vbox \vec \vee \vrule \vskip \wedge \widehat \widetilde

right inner product delimiter ( ) real part of ( ) controls size of right delimiter rightward pointing arrow (→) rightward pointing double arrow (⇒) roman font big cap-small cap font arrow point southeast ( ) sign for set-theoretic difference (A \ B ) similar to (a ∼ b) variant of \sim (a b) √ square root ( x + A) ß is a (proper) superset of (A ⊃ B ) is a superset of (A ⊇ B ) arrow point southwest ( ) sine slanted roman font small vertical space is a (proper) subset of (⊂) subset of (⊆) tangent macro for TeX (TEX) ˜ tilde (A) a times or multiplication sign (×) typewriter-like font a brace under text (x + y ) a line under text (a + b) a typesetting device for isolating text the vector accent (a) a vee (S ∨ T ) a solid black box ( ) vertical skip a wedge, as in an exterior product (v ∧ w) wide caret or “hat” in math mode (e.g., f ) wide tilde in math mode (e.g., h)

©2001 CRC Press LLC

A Appendix VIII: A Sample of LTEX
A We now present a sample of L TEX code and then show its compiled A output. You will be able to do a fair amount of L TEX typesetting just by imitating what you see here. For convenience, we present the code A A both in L TEX 2.09 form and L TEX2ε form.

©2001 CRC Press LLC

\documentstyle{article} \newfam\msbfam \font\tenmsb=msbm10 \textfont\msbfam=\tenmsb \font\sevenmsb=msbm7 \scriptfont\msbfam=\sevenmsb \font\fivemsb=msbm5 \scriptscriptfont\msbfam=\fivemsb \def\Bbb{\fam\msbfam \tenmsb} \def\RR{{\Bbb R}} \def\HollowBox #1#2{{\dimen0=#1 \advance\dimen0 by -#2 \dimen1=#1 \advance\dimen1 by #2 \vrule height #1 depth #2 width #2 \vrule height 0pt depth #2 width #1 \llap{\vrule height #1 depth -\dimen0 width \dimen1}% \hskip -#2 \vrule height #1 depth #2 width #2}} \def\BoxOpTwo{\mathord{\HollowBox{6pt}{.4pt}}\;} \begin{document} \begin{center} \large \bf The Fundamental Theorem of Calculus \end{center} Let $f$ be a continuous function on an open interval $I \subseteq \RR$. Fix a point $a \in I$. For any point $x \in I$, we define $$ F(x) = \int_a^x f(t) \, dt . \eqno (\dagger) $$ The substance of the Fundamental Theorem of Calculus is to claim that $F$ is an anti-derivative for $f$. More precisely, we have \smallskip \\ \noindent {\sc Theorem 1:} is differentiable, and $$ \frac{d}{dx} F(x) = f(x) $$ for every $x \in I$. The function $F$ defined above

©2001 CRC Press LLC

dt}{h} . \end{eqnarray} Now fix a point $x_0 \in I$. Choose $\delta > 0$ such that $|t . dt}{h} \leq \epsilon .f(x)}{h} = f(x) .F(x)}{h} = f(x) + \hbox{error} . dt}{h} .f(x)] \. \\ \end{eqnarray*} If $|h| < \delta$ then we may estimate the last fraction as $$ \left | \frac{\int_x^{x+h} [f(t) . Now we may rewrite (1) as \begin{eqnarray*} \frac{\int_x^{x+h} f(t) \. $$ \lim_{h \rightarrow 0} {F(x+h) . \eqno \BoxOpTwo $$ \end{document} ©2001 CRC Press LLC . $$ where the error is not greater than $\epsilon$.x_0| < \delta$ implies that $|f(t) . we have $$ \frac{F(x + h) . in summary. dt}{h} \nonumber \\ & = & \frac{\int_x^{x+h} f(t) \.f(x)| \. dt}{h} \right | \leq \frac{\int_x^{x+h} |f(t) . dt}{h} & = & \frac{\int_x^{x+h} f(x) \. $$ Thus.F(x)}{h}}_{\hbox{Newton quotient}} & = & \frac{\int_a^{x+h} f(t) \. In conclusion.\int_a^x f(t) \.\smallskip \\ \noindent {\bf Proof:} We endeavor to calculate the derivative of $F$ by forming the difference or Newton quotient for $h \ne 0$: \setcounter{chapter}{0} \begin{eqnarray} \underbrace{\frac{F(x + h) . dt}{h} \\ & = & f(x) + \frac{\int_x^{x+h} [f(t) .f(x)] \. dt .f(x_0)| < \epsilon$. Let $\epsilon > 0$.f(x)] \. dt}{h} + \frac{\int_x^{x+h} [f(t) .

\eqno (\dagger) $$ The substance of the Fundamental Theorem of Calculus is to claim that $F$ is an anti-derivative for $f$. More precisely. and $$ \frac{d}{dx} F(x) = f(x) $$ The function $F$ defined above ©2001 CRC Press LLC .} \begin{document} \begin{center} \textbf{\large The Fundamental Theorem of Calculus} \end{center} Let $f$ be a continuous function on an open interval $I \subseteq \RR$. we have \smallskip \\ \noindent \textsc{Theorem 1:} is differentiable. we define $$ F(x) = \int_a^x f(t) \. For any point $x \in I$. dt .4pt}}\. Fix a point $a \in I$.\documentclass{article} \newfam\msbfam \font\tenmsb=msbm10 \textfont\msbfam=\tenmsb \font\sevenmsb=msbm7 \scriptfont\msbfam=\sevenmsb \font\fivemsb=msbm5 \scriptscriptfont\msbfam=\fivemsb \def\Bbb{\fam\msbfam \tenmsb} \def\RR{{\Bbb R}} \def\HollowBox #1#2{{\dimen0=#1 \advance\dimen0 by -#2 \dimen1=#1 \advance\dimen1 by #2 \vrule height #1 depth #2 width #2 \vrule height 0pt depth #2 width #1 \llap{\vrule height #1 depth -\dimen0 width \dimen1}% \hskip -#2 \vrule height #1 depth #2 width #2}} \def\BoxOpTwo{\mathord{\HollowBox{6pt}{.

F(x)}{h}}_{\hbox{Newton quotient}} & = & \frac{\int_a^{x+h} f(t) \. dt}{h} . we have $$ \frac{F(x + h) .f(x)}{h} = f(x) . In conclusion. dt}{h} \leq \epsilon . \end{eqnarray} Now fix a point $x_0 \in I$.x_0| < \delta$ implies that $|f(t) .f(x)] \. \smallskip \\ \noindent \textbf{Proof:} We endeavor to calculate the derivative of $F$ by forming the difference or Newton quotient for $h \ne 0$: \setcounter{chapter}{0} \begin{eqnarray} \underbrace{\frac{F(x + h) . Let $\epsilon > 0$. $$ where the error is not greater than $\epsilon$. \\ \end{eqnarray*} If $|h| < \delta$ then we may estimate the last fraction as $$ \left | \frac{\int_x^{x+h} [f(t) . in summary. dt}{h} \right | \leq \frac{\int_x^{x+h} |f(t) .f(x)] \. \eqno \BoxOpTwo $$ \end{document} ©2001 CRC Press LLC . dt}{h} \nonumber \\ & = & \frac{\int_x^{x+h} f(t) \. $$ \lim_{h \rightarrow 0} {F(x+h) . dt}{h} + \frac{\int_x^{x+h} [f(t) . dt}{h} .\int_a^x f(t) \.F(x)}{h} = f(x) + \hbox{error} . dt}{h} \\ & = & f(x) + \frac{\int_x^{x+h} [f(t) .f(x)| \. dt}{h} & = & \frac{\int_x^{x+h} f(x) \. Now we may rewrite (1) as \begin{eqnarray*} \frac{\int_x^{x+h} f(t) \.for every $x \in I$.f(x_0)| < \epsilon$. Choose $\delta > 0$ such that $|t . $$ Thus. dt .f(x)] \.

For any point x ∈ I. More precisely. = f (x) + x h x+h x x+h If |h| < δ then we may estimate the last fraction as x+h [f (t) x − f (x)] dt ≤ h x+h x |f (t) − f (x)| dt ≤ . and d F (x) = f (x) dx for every x ∈ I. h ©2001 CRC Press LLC . Choose δ > 0 such that |t − x0 | < δ implies that |f (t) − f (x0 )| < . we have Theorem 1: The function F defined above is differentiable. we define x F (x) = a f (t) dt. Proof: We endeavor to calculate the derivative of F by forming the difference or Newton quotient for h = 0: F (x + h) − F (x) = h Newton quotient = x+h x x+h a f (t) dt − h f (t) dt . (†) The substance of the Fundamental Theorem of Calculus is to claim that F is an anti-derivative for f . Now we may rewrite (1) as x+h x f (t) dt = h f (x) dt [f (t) − f (x)] dt + x h h x+h [f (t) − f (x)] dt . Let > 0. Fix a point a ∈ I.Let f be a continuous function on an open interval I ⊆ R. h x a f (t) dt (1) Now fix a point x0 ∈ I.

h where the error is not greater than . we have F (x + h) − F (x) = f (x) + error.Thus. In conclusion. h→0 lim F (x + h) − f (x)h = f (x). ©2001 CRC Press LLC . in summary.

Particularly well-suited for use on the Internet. acute accent The accent (used in French. In England this unit was called the ruby. cedilla e e a (¸). special fonts. acquisitions editor The editor who evaluates a new manuscript from an author. accent A mark on a character. Spanish. gets it reviewed. AA’s (author’s alterations) Changes to the copy made by author after the manuscript has been typeset. AMS-TEX A macro package for TEX that has been created by the American Mathematical Society (specifically. Adobe Acrobat A reader for the pdf computer graphics and page design language. For years the agate was the standard for measuring advertisements. c n accent grave The accent (used in French. ©2001 CRC Press LLC . by Michael Spivak) in order to facilitate the creation of mathematics documents. circumflex (ˆ). A AMS-L TEX A version of TEX that incorporates many of the best A features of L TEX and of AMS-TEX. archive A collection of software located at a particular Web address that can be accessed through anonymous ftp. and determines whether it is to be published. and other European languages) that has the form a and is given by the TEX code ` \‘. agate An old unit of measure in typesetting equal to slightly less than 5. for manipulation with the tar command. Also: a single file that contains a collection of other files. Spanish.5 points. and other European languages) that has the form a and is given by the TEX code \’. Common accents are named acute (´). ´ Also called “accent egout”. See international paper sizes and ISO. tilde (˜). A publisher may charge the author for AA’s that exceed 15%. indicating stress. usually a letter. but especially from the Web page of Budgett and Johnstone. AMS-TEX contains special style macros for the Society’s journals. AMS American Mathematical Society. grave (`). A4 paper size The international standard size for business paper: 297mm × 210mm.Glossary The definitions included here have been drawn from a variety of books and articles listed in the References. and also macros for creating commutative diagrams and for other mathematical displays.

bar-under accent The accent having the form c and given by the ¯ TEX code \b.bmp file. glossaries. binding A method (of which there are many) for securing loose leaves or sections into a book. Usually the cylinder is covered with a rubber sheet which prevents wear from the plate coming into contact with the paper. ASCII code The American Standard Code for Information Interchange. blow up An enlargement of a graphic image or a photograph. This may include appendices. A BIBTEX A macro for L TEX that is designed for the preparation of bibliographies. back matter Material that follows the main body of text. big cap-small cap font A font in which lower-case letters are replaced by small capital letters. binary code The machine code that is read and understood by a computer. as measured from the top of the tallest ascender to the bottom of the lowest descender.bmp file A graphics file that is stored in the bitmap format.ascender The portion of a lower-case letter that extends above the main body of the letter. block quotation An excerpt or quotation set in reduced type. bleed Illustrations that extend to the edge of the page. of the capital letters. or bases. ©2001 CRC Press LLC . Among these methods are perfect and saddlestitched. This sentence is set in big cap-small cap. See front matter. baseline An imaginary line that passes through the bottoms. Also called an extract. bed The base on which the forme is held when printing by letterpress.bib files in a special BibTEX format. See also halftone. measured in points. Such a quotation is often displayed. Bibliographic data is archived in *. indexes. and are trimmed off at the time of binding. both alphanumeric and symbolic. Compare descender. blanket cylinder The cylinder by way of which the inked lithography plate transfers the image to the paper. tables. *. references. that can be typed in on a computer keyboard. block in The process of sketching in the main parts of an image prior to making the more formal design.” benday The screen used to create a halftone. bitmap A machine language graphic format. See also *. body size The height of a type font. An enumeration of all characters. and like material. are said to bleed. Newspapers speak of “sending an issue to bed.

Also known as “mechanicals”. caps Capital letters. or of a chapter or article. to estimate the number of pages of text that will be made from a set of galleys. symbols. B. C form. Usually printed on RC (resin-coated) paper. bond A sized. but generally larger than. brackets The delimiters . the circumflex. As distinct from a legend. signs. Invoked in A L TEX with the command \begin{itemize}. ©2001 CRC Press LLC . spaces. thick letters. illustration. camera copy or camera-ready copy Proof sheets that are ready to go to the printer. then a resulting powdering effect on the surface is called chalking. cast off To estimate the number of typeset pages that will be made from a manuscript. Such paper can also be printed upon. leather. See also repro copy.. Used as the first step in a cast off. case bound A hardbound book made with stiff outer covers. ¸ chalking If the ink fails to dry satisfactorily after printing. finished writing paper of 50 grams per square meter or heavier. etc. used in proofreading to indicate an insertion. table. B. cap line An imaginary line drawn through the tops of all capital letters.e. caret The symbol ∧. Also used as a noun: to make a castoff. or vinyl. letters. braces The delimiters . dark. character count The number of characters (i. Similar to. bullet A large dot placed before text to add emphasis. Alternatively. cap size The distance from the cap line to the baseline. C or slanted A.boldface A font consisting of heavy. browser A piece of software used for reading material on the World Wide Web. breve The accent that has the form a and is given by the TEX code ˘ $\breve{a}$. numerals.) in a piece of copy or a line or a paragraph. broadside See landscape. See also search engine. catchline A temporary heading—provided for identification purposes— at the top of a galley proof. caption The heading or title of a figure. In some fonts boldface is available in either upright A. broadsheet See broadside. cedilla The accent (used in French and other European languages) that has the form c and is given by the TEX code \c. The case is often covered with cloth.

See gathering. style. color separation The division of a four-color or multi-colored original into the primary process colors of yellow. copy editing The process by which a manuscript is read and evaluated— by a trained expert—with a view to correcting both the use of language. copy editor The person who performs copy editing functions. collective sign A mathematical symbol. a portion of the type body that is one column wide and one inch deep. halftone. circumflex The “hat” accent that is used to denote the Fourier transform.. Each character or ligature is engraved on a piece of lead. The measure is used to calculate the cost of display advertising. ©2001 CRC Press LLC . Words. cyan. and the formatting of the page. such as a sum.chase A metal frame in which (cold) metal type and blocks (i. syntax. check The accent that has the form a and is given in text by the ˇ TEX code \v{a} and in math mode by the TEX code $\check{a}$. copyholder A person who reads manuscript aloud to the proofreader. collate To assemble the separate pages or section of a book into the proper order for binding. See also benday. cold type composition The traditional technology used for typesetting. and black. compositor In traditional typesetting. union. grammar. commercial “at” The symbol @. commonly used in e-mail addresses. copyright The legal entity that gives protection to the creator of material to prevent use without express permission or acknowledgement of that creator. The term also refers to a passage at the end of a book describing the type and how the book was produced. spelling. that signifies some aggregation. See also diphthong. consistency. close up A proof correction mark (copy editor’s mark) that is used to indicate reduction of the amount of space between characters or words. the person who typesets text. column inch In a newspaper or magazine.e. copy editor’s marks See proofreader’s marks. and paragraphs are composed by juxtaposing the pieces of lead in a rack. engravings) are locked into position in order to make up a page. usually appearing on the title page of a book. magenta. colophon The publisher’s logo or trade emblem. or integral. consonantal digraph A combination of two consonants to express a single sound. composition The process of setting a manuscript into type. The same as typesetting. product. screen. sentences. and also occurs in foreign languages such as French.

the symbol ‡ that is used to mark equations. and other printing technologies. with many mirror images. cover The outside binding material of a book or journal. or footnotes.corner marks The angle marks printed on a sheet to denote the trim or register marks. 2. The computer’s monitor. inside front cover. cursive A typeface that resembles written script. This could be an em-dash or an en-dash. implementations of A TEX and L TEX. laser. developmental editor This is the editor that oversees the writing and genesis of the book—from early draft to final form that will be sent into production. Usually made of heavier material than the actual pages. CRT Cathode ray tube. The developmental editor will get reviews. The chip that is the “brains” of the computer. Usually cropping is done in order to allow the other parts of the graphic to be enlarged to fill the indicated space. and many other TEX resources. dash A short horizontal rule used for punctuation. 3. The outside front cover. cut flush attached. See Appendix VI for more information about CTAN sites. from peers of the author and from others. and has been replaced by ink-jet. descender That portion of a lower-case letter that extends below the baseline. and 4 respectively. and will help the author to interpret those ©2001 CRC Press LLC . A method of trimming a book after the cover has been cutout A halftone in which the background has been cropped out in order to produce a silhouette. inside back cover. the symbol † that is used to mark equations. and outside back cover are often designated covers 1. that holds TEX macros. ddagger In typesetting. CPU Central processing unit. displays. CTAN (Comprehensive TEX Archive Network) A Web site. or footnotes. not cathode ray tubes. Nowadays some of the new “thin” monitors use digital technology. cropping The editing away of portions of a photograph or other original graphic that are not desired in the printing. fonts. displays. delimiter See fence. This method of printing is slow. daisy wheel printer An impact printer which prints from a wheel that has many different engraved print images. dagger In typesetting.

die A hardened steel engraved block used to print an inked image. and graphics screens. will guide the author through writing problems. WindowsNT.g. Such a printer could be an impact system (in which a wire is fired at a ribbon). in the writing of elementary textbooks. See diphthong. The pairing is called a “diphthong”. The role of the developmental editor is most decisive. DOS The disc operating system. This computer operating system for PC computers has now been superceded by the Windows products: Windows95. Used. usually at least 18 point and set in bold. ink-jet. or electro-erosion. discretionary hyphen See soft hyphen. The one in most prevalent use today was created by Microsoft in the early 1980s and is denoted MS-DOS. text by the TEX code \d{a}. See also logotype. from the text body. and will help him to assemble the final book. in the production of high quality letter headings. dot matrix printer A printer that forms each character from a matrix of dots. and Windows2000. double spread Two facing pages in printed copy. diphthong A digraph composed of two vowels. Also called “double page spread”. dot accent The accent that has the form a and is given in text by ¨ the TEX code \. or thermal.. dot-under accent The accent that has the form a and is given in . The image is divided into dots and the number of dots per linear inch is the DPI. and other printer modern technologies. display type Large type. Windows98.reviews. dieresis The accent that has the form a and is given in text by the ¨ TEX code \"{a} and in math mode by the TEX code $\ddot{a}$. The developmental editor will help the author to render his graphics. Modern laser printers are 1200 DPI or more. displayed mathematics Mathematics that is typeset in a size larger than the regular text and is set off. phototypesetting machines. Common diphtongs are ae and oe . æ or œ) is called a ligature. but the resulting compressed symbol (e. ©2001 CRC Press LLC . that is used in a heading. drawn on The method of binding a paper cover to a book by drawing on the cover and gluing it to the back of the book. for example. digraph A combination of two letters to express a sound. and pervasive. DPI (dots per inch) The measurement of resolution for page printers.{a} and in math mode by the TEX code $\dot{a}$. with space above and below. The dot matrix printer has now been superceded by laser.

en-dash A dash of medium length.dvi file A device independent file. elongated letters. the production editor oversees the production process. *. used to indicate a range of pages or numbers. umlaut. dot. encapsulated PostScript A form of PostScript that is especially designed for inclusion in other documents. Often typeset in gothic. Also called an en. . Also called an em. created when a TEX source code file is compiled. Such files have the form *. The acquisitions editor is one who accepts manuscripts for review and who ultimately decides which of these will go into publication. it was the width of the capital letter “M” in the font being used. en See en quad. The copy editor performs a technical review of the manuscript. and depends on the font and on other parameters as well. dye transfer A photographic color print methodology that uses special coated paper to produce a full-color image. as . Often known as Letraset. elided A page range is elided if it is written in the form 465–83 rather than 465–483. editor The publishing process involves several different types of editors. ellipsis Three periods or dots.eps. usage. dry transfer Characters. This is Dunhill. embellished letter A letter that has been augmented with an accent.. tilde. ©2001 CRC Press LLC . used to denote an omission. etc. cedilla. drop folio A page number placed at the bottom of the page (as opposed to top left or right). dunhill A popular typeface that features tall. that can be transferred to the work by rubbing on the back of a plastic transfer sheet.drop cap A large initial letter at the beginning of text that drops below the baseline and into several of the lines of text below. em quad A space named because. such as a TEX document. See PostScript. traditionally. em-dash A long dash that used for a pause or break in a sentence. bar. em See em quad. drawings. images. or other diacritical mark. emacs A popular text editor for the UNIX operating system. . Once a manuscript has been accepted. Nowadays the definition is variable. and syntax. Also known as GNU emacs. or · · · . correcting for formatting. and in similar contexts. spelling. See also text editor.

e. enunciation A displayed mathematical statement such as a theorem. forme Type and blocks assembled in pages and placed in a metal chase ready for printing.. In the expression A2 . en quad the capital is variable. remark. A font includes all upper. face An abbreviation for “typeface”. Epson emulation or standard The industry standard control codes for dot matrix printers. first proof See galley proof. flexography A rotary or letterpress technology consisting of printing from a rubber or other flexible plate and using fast-drying ink.and lowercase letters. or example. corollary. extract See block quotation. both the 1 and the 2 1 are first order indices. final proof Proof which has gone through the editing and correction process (i. developed by Epson Corporation. such as a parenthesis. *. first order index A superscript or subscript that directly augments a character on the baseline. Compare endpaper. Also a sheet of a manuscript. an en. lemma. See also delimiter. See second order index. and all punctuation. fence A delimiting mark. it was the width of letter “N” in the font being used.endpaper A folded sheet of paper. Compare flyleaf. Refers to a given font family in a given style. or bracket. It is pasted in the front of a book (half to the front cover) and in the back of a book (half to the back cover). flush right Alignment of copy along the right margin. Most fonts will have a boldface variant and a slanted or italic variant. ©2001 CRC Press LLC . Nowadays the definition and depends on the font and on other parameters as well. See also PostScript. the “Frequently Asked Questions” button. brace. definition. folio The page number. the galley and page proof stages) and is now in final form. all numerals. different in texture and perhaps in color from the main pages of a book. font A style and size of type.eps Encapsulated PostScript format. traditionally. even pages See verso pages. FAQ On a Web site. flyleaf A blank leaf or page bound into the front or back of a book. Also called A space named because. usually used in pairs. flush left Alignment of copy along the left margin.

galley proof The typeset manuscript of a book or paper. gathering The process of placing the pages. freeware Software that is available (often on the World Wide Web) without any payment or quid pro quo. galleys Printing term for the long metal trays used to hold type after it had been set and before the press run. gravure A printing process in which the image is etched onto a metal plate attached to a cylinder. and other displayed material. hairlines The thinnest of the strokes in a typeface. ftp The acronym for “file transfer protocol”. A graphics file that is in the Graphics Image Interchange glue In TEX. See back matter. The galley proof will not show space for figures. hairline rule The thinnest rule (line) that can be printed. See hairline rule. to produce all possible colors on one pass through the printer. sections. *. which are over-sized because they will be bound. gutter The inner margins of a page. grey scale A range of luminance values for evaluating shading from white to black. This is both the protocol and the software for transferring files between different computers on the Internet. The term is sometimes applied to certain types of CRTs or computer monitors. graphics artist A technically trained artist who can render graphics for a book or article for the purposes (and to the standards of) publication. and black). cyan. the information about spacing and adjoining to other characters that comes “packaged” with each TEX character. Used in evaluation of scanner images. magenta. preface. This may include the title page. h´ˇek The accent that has the form ˇ and is given in text by the ac a TEX code \v{a} and in math mode by the TEX code $\check{a]$. foreword. the list of figures. acknowledgements. tables. Also called “rotogravure” or photogravure. and like material. front matter Material that precedes the main body of text. table of contents. formatted according to customary typesetting standards but not broken for pages. ©2001 CRC Press LLC . See collate. or signatures of a book in the proper order for binding.gif file format. dedication.four-color press A press that can print four colors simultaneously (usually yellow. half title A (possibly abbreviated) title which appears by itself on a page preceding the more formal title page.

punctuation. helvetica A popular sans serif typeface. hard copy A version of a document that is printed out on a (paper) page. XML.). head margin The blank space occupying the region from the top edge of the page to the topmost element of the print body. imprint The name and address of a publisher. Often set in boldface. that is placed between an original photograph and the film to be exposed. hypertext The language used to format material for the Internet. indentation. See also sinkage. The number of lines per inch controls the coarseness of the final dot configuration in the resulting halftone. giving format and style specifications. In the Windows or Macintosh environment. hyphenation. head The margin at the top of a page. halftone screen A glass plate or film. See also benday. usually appearing on the title page of a book. and display. heads Words or phrases used to differentiate different portions of the text (chapters. etc. Especially useful for hyperlinks. hang indent A paragraph typeset so that the first line is flush left and subsequent lines uniformly indentd. Sometimes this material will appear with the colophon (or publisher’s logo) and the date of publication. a folder. or some other special type. hanging punctuation Punctuation that is allowed to fall outside the print margins. or big cap-small cap. header Computer code at the beginning of a (TEX) document. A book that is not paperback. hardback A case bound book with a separate stiff board cover. a file. house style A publisher’s or publication’s preferred style of spelling. as opposed to the electronic (or “soft”) version of the document. SGML. hyphen The shortest type of dash. the icon is activated by clicking with a mouse or other pointing device. sections. See MathML. impression cylinder The cylinder of a printing machine that brings the paper into contact with the printing plate. ©2001 CRC Press LLC . Used to hyphenate words. icon A pictorial image on screen that designates a utility function. HTML (Hypertext Markup Language) The code that is used to prepare Web pages and text for electronic documents on the Internet. handwork Composition that requires manipulation by an artist rather than just keyboarding.halftone A process by which a black-and-white photograph is filtered through a screen so that various shades are represented by different densities of dots. carrying a network of parallel lines. or an application.

See first order index and second order index. linked by modem and ethernet. As opposed to displayed mathematics. insert An instruction to the printer for the inclusion of additional copy. in many industrial countries. justified lines A line typeset flush against one or both margins. B. kern To move the point of typesetting either left-right or up-down. The American 8. ISBN (International Standard Book number) Similar to the SBN number. the paper with a watermark pattern showing the wire marks used in the paper-making process. ©2001 CRC Press LLC . italic A slanted version of a font. Each such computer is accessible from any other. keyline An outline drawn on artwork to show the size and position of an illustration. This sentence is typeset in italic.jpg file A graphics file that is in the Joint Photographic Experts Group format. Some expressions in mathematical logic are not typeset in italic. Also the alphabetized list of topics at the end of a book or manuscript.5 × 11 business paper is not an ISO paper size.5 × 14 size. Internet The union of millions of computers. transparent plastic coating on paper or board to provide protection and a glossy finish. Business-size paper. See international paper sizes. ISP Internet service provider. is size A4. and C. ISO The International Standards Organization for paper size. nor is the American legal 8. North American countries do not subscribe to the ISO standard. laid In the production of high-quality stationery. around the world.index These are subscripts and superscripts which augment a character on the baseline. See left justified line and right justified line. international paper sizes The International Standards Organization (ISO) system of paper sizes is based on a series of three sizes: A. laminate A thin. *. and is the same size as the regular text. inline mathematics A mathematical expression that occurs in the flow of text. used to slow down the eye when emphasis is desired. ink-jet printer A printing technology in which particles or droplets of electrically charged ink are sprayed from a matrix of tiny ink jets. Mathematical expressions are generally typeset in a form of italic called math italic: x2 − 2yz = w.

Compare halftone. See also diphthong and logotype. table. L TEX3 is under preparation at the time of this writing. leading A vertical space used in typesetting. and before and after displayed material (i. thereby improving the appearance of a line of text. ©2001 CRC Press LLC . without using a screen. in preparation for printing. Contrast with portrait. The image is then transferred to paper by a process similar to xerography. Traditionally. A X that incorporates powerful. letterpress The type of printing in which ink is applied directly to the metal type and then transferred to the paper by the application of pressure. The graphic effect is produced with lines rather than halftones. ligature A symbol resulting from the juxtaposition of two letters—for instance æ or œ. left justified line A line of text which is typeset flush against the left margin. Compare lithography or offset printing. letterset A printing process that combines offset printing with a letterpress relief printing plate. layout A sketch of a page. letterspacing The addition of space between the letters in a word in order to increase line length. It is common to put 2 points of space between lines of text. line copy/line drawing (Graphic) copy that is produced with pen strokes. and other standard A types of documents. which provides style files for articles. which shows positioning of text and illustrations and gives overall instructions. mathematical formulas or enunciations). hence the phrase “2 point leading”. line cut An engraving in which there are no tonal gradations. books. originally designed by Leslie Lamport.5 page. the page is held sideways instead of up-and-down. Letraset See dry transfer.09. reports. A L TEX A TEX macro package. A 6 point leading is sometimes used between sections. the space was effected with a blank piece of lead type. lateral reversal An image (positive or negative) transposed from left to right as in a mirror reflection.. laser printer A high-quality printing technology—in wide use today— in which a laser beam is used to produce an image on a photosensitive drum. legend The explanatory material that accompanies an illustration. or chart. In other words. Compare caption.e. figure.landscape Presentation of material on an 11 × 8. Also called L TEX 2. new A X2ε A newer release of L TE L TE A macros and graphics features.

Often the word “logotype” is used to refer to the piece of cold type that bears the diphthong or ligature. linen tester in a halftone. ©2001 CRC Press LLC . macro A collection of computer commands that may be invoked by using a nickname that has been assigned to them.line gauge A metal rule used by printers. The accent that has the form a and is ¯ given by the TEX code \={a}. macron An overbar in text. This was the standard method of typesetting for the first half of the twentieth century. the production editor. linotype A typesetting machine that sets solid lines of slugs of metal type. for continuous updating. Numerals that are aligned at the baseline and at the lining figures top. long Hungarian accent or umlaut The accent that has the form a ˝ and is given by the TEX code \H{a}. make-up The assembling of all elements in order to form the printed image. makeready The process by which a letterpress plate is adjusted or aligned to compensate for irregularities in the plate. line weight The thickness of a line or stroke used in a line drawing. Abbreviated “logo”. lithography The process of printing in which ink is applied to the metal type. The magnifying glass designed for checking dot images lineup table A backlighted glass table used for preparing and checking the alignment of page layouts and paste-ups. often marked with pica measurements. majuscule An upper-case letter. Most modern printing is done with lithography. managing editor The person who oversees all the other editorial staff in the production of a book—the developmental editor. See also diphthong and ligature. and then finally transferred to the paper by way of direct contact with the rubber pad. and the copy editor. Binding in a three-ring binder of ACCO folder. Compare upper-case. lower-case Denotes those letters which are not capitalized. then transferred to a rubber pad by way of pressure. loose leaf A method of binding which allows for insertion and removal of pages. logotype A single piece of type bearing two or more usually separate elements.

MathML Mathematical Markup Language. See HTML. mock up The rough visual dummy of a publication or design of a work. A markup language to be used with XML. mask Opaque material.. miniscule A lower-case letter. that is used to block off an area of artwork. for doing mathematics by way of symbol manipulation.manuscript The hard-copy draft of a document that you submit to the publisher. mechanicals See camera copy or repro copy. math italic The italic font that TEX uses to typeset mathematics. SGML. Material is keyboarded. and a paper tape produced. masthead Detailed information about publisher. monotype A particular sort of typesetting machine. Mathematica A high-level computer language. mark up Copy prepared for a compositor setting out all the typesetting instructions. Part of the TEX family. created by Wolfram Research. for doing mathematics by way of symbol manipulation. and editorial staff that are usually printed on the contents page of a periodical. markup language A high-level computer language that is used to describe the content and formatting of a page of textual material. XML. often masking tape. montage A single image formed from the assemblage of several images. mounting board A heavy board used for mounting and assembling artwork. editors. monospace The attribute of a font of having each character occupy the same amount of horizontal space. Maple A high-level computer language. created by the mathematicians of Waterloo. TEX is a markup language. margins The non-printing areas of a page. mechanical binding A method of binding that consists of securing pre-trimmed pages by the insertion of wire or plastic spirals through holes drilled in the binding edge. Also used to refer to a typesetting style in which all characters have the same width. METAFONT Donald Knuth’s language for creating fonts. moire A checkered effect that is achieved by superimposing halftone screens at the wrong angle. The blank space on the edges of a printed page. Especially good for mathematical and scientific content. Inc. ©2001 CRC Press LLC .

overs Additional paper used to compensate for that which is spoiled in the printing process. by way of scanning. odd pages See recto pages. overstrike A method used in word processing. A page description language allows the applications ©2001 CRC Press LLC . mutt The em quad. and also with a traditional typewriter. opacity The degree to which paper will show print through. to produce a character not in the given typeface by superimposing two given characters. or black and white are reversed. Can also refer to extra copies above the number initially ordered. a mathematician will produce a blackboard boldface C by superimposing an I and a C.MS See manuscript. The method may be used to emphasize changes or alterations. offset printing See lithography. outline font A typeface in which characters are formed with only their outline specified (rather than solid strokes). old style numerals () A style in which most numerals ascend or descend from the x-height of the other characters in the typeface. oblique strok See shilling or solidus or virgule. OCR See optical character recognition. optical character recognition The technology by which a machine. For example. offprint See reprint. can recognize characters on a page and convert them to ASCII code. As opposed to bitmapping. ozalid A patented method for producing page proofs from paper or film. oldstyle type A typeface that is characterized by stressed strokes and triangular serifs. negative A photographic image in which colors are reversed. An example of an oldstyle typeface is Garamond. overlay A transparent sheet that shows the color breakdown—used in the preparation of multi-color artwork. orphan A brief phrase—less than a full line—from the start of a paragraph that occurs at the bottom of a page. nipping The compression of air after sheets of a book are sewn together. page description language (PDL) A programming language that enables both text and graphics to be described in a series of mathematical statements. Outline fonts will scale well. See positive. overprinting Printing a second time over an area already printed.

page printer A non-impact printer which produces a complete page in one action. PE’s (printer’s errors) Changes or errors made in the typeset text that are the responsibility of the typesetter or copy editor. PDF format A file format (Portable Document File format) due to Adobe that is similar to PostScript (q. There are dry transfer pagination sheets to help with pagination. and electro-photographic. perfector A printing press which can print both sides of the paper at one pass.v. perfect binding A style of bookbinding in which the backs of the signatures are cut off. pica ruler The metal rule most commonly used by a typesetter in page design. ©2001 CRC Press LLC .software to be independent of the physical printing device. LED. Now a standard file format for electronic document printing and distribution. and other displays.pdf. as in myfile. phototypesetting pica A unit of distance in typesetting. This is the typeset text in its final form. A process of typesetting onto film.). electro-erosion. paste up The elements of a layout mounted in position—often on a mounting board—to form camera-ready artwork. pantone A patented name for an ink color matching system. Used. PostScript is an important PDL. pagination The numbering of pages in a book. paper plate An offset printing plate for a short print run on which matter can be typed directly. paragraph mark A typesetting symbol (¶) used to denote the start of a paragraph. Can also be used to mark a footnote. ion deposition. Compare signature. The Adobe Acrobat reader is designed to read pdf files. for example. and the pages then glued together with adhesive. and with space set for figures. tables. paperback A book with a perfect binding whose covers are made of heavy paper. photogravure See gravure. equal to 12 points. page proof Similar to the galley proof. Some examples of page printers are laser. Files in this format generally have extension .pdf. by CorelDRAW. These alterations are not charged to the author. PDL See page description language. the edges roughened. but broken for pages. LCD shutter xerographic.

5 × 11 page. portrait Presentation of material on an 8. overhead. the cost of setting up the printer made it too impractical and costly to do small print runs. PostScript A page design language created by Adobe. graphics artists. point A unit of length equal to 0. When mixed together in suitable proportions with black these will produce all other colors. but added specially. preface The part of the front matter of a book that introduces the subject and that particular presentation of it. A preamble That part of a L TEX source code file that occurs before the line \begin{document}. staff salaries. print-on-demand The digital technology that allows the printing of a small number of copies of a document. print block The entire body of text on a page. These might include mathematical symbols and timetable symbols. preprint A copy of the manuscript or original hard copy of an article. measured from the bottom of the top margin to the top of the bottom margin. Contrast with landscape. See also prospectus. press run The number of copies (of a book or article) to be printed. With digital technology this is now feasible and affordable. pixel A screen dot on a computer monitor or CRT. print body The totality of print on a page. In traditional typesetting. printed copy. and the TEX specialists to whom the MS is outsourced.013837 inches. production editor The editor who oversees the process of bringing a book from final manuscript form to final. and yellow. These include composition. The production editor oversees the copy editor. magenta. primary colors These are cyan. In other words. chapter. See negative. or book manuscript. A typical SVGA monitor is about 1200 pixels across and 1000 pixels high. the page is held up-and-down instead of sideways. electricity. positive The true photographic image of the original made on paper or film. This is a highlevel computer language that allows the formatting of both graphics and text.pi fonts Characters not typical of a font. ©2001 CRC Press LLC . and equipment. There are about 72 points to an inch. and from the right edge of the left margin to the left edge of the right margin. plant costs Fixed costs of the book production process that are not connected with any one book or by the size of the edition. plates. at the moment of demand.

quire One twentieth of a ream (25 sheets) of paper. quadding Additional space to fill out a line of type. used for camera-ready copy. See preface. proof Copy obtained from inked type. or screen for the purpose of checking quality and accuracy. RC paper Resin-coated paper.progressive proofs A series of plates. public domain copyrighted. proofreader’s marks. each produced in one of the four basic colors (yellow. ragged left Describes a body of text with an unjustified left margin. proofs Typeset text that is ready for checking (proofreading) and correction. rag paper High quality paper for stationery that is manufactured from cotton rags. and black). prospectus Similar to a preface. The process of proofreading involves marking any errors with standard proofreader’s or copy editor’s marks. See also monospace. then proof sheets ©2001 CRC Press LLC . block. ragged Lines of set type that do not begin or end at the same horizontal position. If printing plates are to be prepared in the traditional manner. proportional spacing A method of spacing in typesetting whereby each character is spaced according to the varying widths of the elements of the font. cyan. Raster image processor (RIP) A hardware engine which calculates the bitmapped image of both text and graphics from a series of mathematical instructions. See the plates in Appendix I. A term used to describe software that has not been quad Same as an em quad. magenta. These are superimposed progressively and used to check color quality and alignment. This process may be effected using a page description language. using em quads or en quads. proofreader’s marks A universally accepted collection of symbols and markings that are used to annotate proof copy. but also includes information about the market and about other books on the same subject. plate. proofreading The process of reading proofs to check them for errors— usually by comparing those proofs with the original copy. proofreader The person who reads proof copy for errors. proof correction marks See copy editor’s marks. ragged right Describes a body of text with an unjustified right margin. See page proof and galley proof. See copy editor’s marks. The method uses kerning.

royal Printing paper of size 20 × 25 . Also a line used for decorative purposes. ‡. in the correct position relative to other impressions. such as an illustration or photograph.are first printed on RC paper and then “shot” (photographed) for reproduction on a copper or zinc plate. printing of part of a larger work. when sizing artwork. The author of a journal article is usually given fifty reprints free of charge. first revise. usually horizontal. recto pages recto-verso printing Two-sided printing (as in a book). revise The stage at which corrections have been incorporated from previous proofs and new. resolution Measurement used in typesetting to determine quality of output. reverse out To reproduce as a white image out of a solid background. This to check alignment. corrected proofs submitted. See also odd pages. second revise. ©2001 CRC Press LLC . The odd-numbered pages of a book. This definition is typeset in computer modern roman. (∗). The units are “dots per inch” or DPI. These are produced by printing fifty unbound copies of the journal. retouching A method for altering or correcting artwork or color separations. runarounds Text set in short lines so that they run around a fixed object. rule A line. rest in proportion (RIP) An instruction. ream A package of 500 sheets of paper. register marks Marks used in color printing to position the paper correctly. For example. This could include †. rough A preliminary sketch of a proposed design. that other parts of the artwork are to be sized proportionately. ¶. repro copy See camera copy. for separating bodies of text. etc. See offprint. reference mark A symbol used in text to direct the reader to a footnote.or reel-fed printing press which uses a curved printing plate mounted on a cylinder. These marks are usually crosses or small circles. register The printing of each impression. for each primary color. A line of text which is typeset flush against the roman The default typeface in TEX. right justified line right margin. reprint A separate reproduction of an article or book chapter. rotary press A web.

set solid Type set without line spacing (leading) between the lines. search engine A piece of software used for conducting searches on the World Wide Web. assigned in a unique fashion to each published book. The first part identifies the publisher. running foot The abbreviated section title. without serifs.run in To merge two paragraphs or bodies of text into a continuously flowing body. In the expression Af 3 . scaling A means of determining the amount of enlargement or reduction required to fit a photograph or graphic to an alloted space. Yahoo. See also ISBN. above the body of type. script A typeface or font that resembles handwritten letters: A. sans serif A typeface or font that is plain. the 3 is a second order index. See first order index. The resulting image is called a halftone. below the body of type. or chapter title. second order index A superscript or subscript that augments a first order index. the second part identifies the particular edition or title. F are examples of script. See also browser. serif A small decoration or embellishment—a crossed line at the end of the main strokes—that appears as part of letters on a font. scale The ability of a program to reduce or to enlarge the amount of space that an image will occupy. Excite. section mark A character (§) used at the beginning of a new section. or author identification that appears across the bottom of the page. ©2001 CRC Press LLC . or chapter title. Microsoft Internet Explorer. or a continuous-tone image. See also screen A traditional technique for transforming a solid block of color. Google. Some popular search engines are Netscape. SBN number (Standard Book Number) A nine-digit number. set off The accidental transfer of the printed image from the front of one sheet (where the ink has been applied) to the back of another. The size and density of dots determines the intensity and darkness of the color. The letter R has three serifs and the letter A has two serifs. A sketch of a design that shows the basic concept. and Dogpile. running head The abbreviated section title. into a pattern of dots. scamp rough. or author identification that appears across the top of the page. the third part is a parity check. set size The width of the type body at a given point size.

shareware Software that can be distributed for free. sidebar A vertical bar positioned on the side of the screen—usually on the right. ©2001 CRC Press LLC . See also solidus. and often 16 or 32). A subheading that is typeset flush into the text at the signature A large sheet of paper. shilling The forward slash / used in the shilling form of a fraction. Classically. soft cover See softback or paperback. See also solidus or virgule. This is an International Standard that describes a generalized markup scheme for representing the logical structure of documents in a system-independent and platform-independent manner. books were bound in signatures. slanted roman are slanted. sinkage The vertical distance between the very top of a printed page and the top of the body of text (or text block). soft hyphen A specially coded hyphen that only displays when the potentially hyphenated word occurs at the end of a line. solidus The slanted line used between the parts of a fraction written in shilling form. shilling form of a fraction The format of a fraction that is a/b rather than a (the stacked fraction format). Compare perfect binding. side heading left edge. sheetwise A method of printing a section by which half the pages are imposed and printed and then the remaining half of the pages are printed on the other side. folded and cut to produce a certain number of pages (usually a power of 2. but for which a modest payment is requested from satisfied users. slurring A smearing of the printed image caused by slippage of the paper during the printing process.SGML Standard Generalized Markup Language. A variant of the roman typeface in which the letters slug A 6-point or 12-point piece of lead that is used for vertical spacing in the cold type typesetting process. See also shilling or virgule. b shooting a page In traditional typesetting. softback See paperback. sheet fed A printing press which handles single sheets of paper instead of rolls or reels of paper. the process of photographing each page of the repro copy of a book or manuscript that has been test-printed on RC paper.

swash letters Italic characters with extra flourishes. spine The back of a book. From the Latin “let it stand”. These are sometimes used at the beginnings of chapters. that appears slightly below the baseline and is an embellishment to a baseline character—for example j is a subscript in the expression xj . These might include section heads. smaller than those on the baseline. tabloid A page half the size of a broadsheet or broadside. and so forth. ©2001 CRC Press LLC .sty files in L TEX serve as style sheets. spell check A software device for checking the spelling in a document. that part which joins the front and back covers. Also called the backbone or shelfback. style sheet A collection of tags that specify page layout style. superior numeral A small numeral. b stem The main vertical stroke making up a character in a type font. See overprinting. STET Sign used in proof correction work to cancel a previous correction. paragraph settings. The *. Contrast with the shilling form of a fraction. superscript A symbol. that mark divisions in a chapter. spiral binding A type of binding in which a spiral of plastic or metal is wound through holes in the pages. that specifies page style for a particular document or publisher. subheads Headings. subscript A symbol. or a macro. stacked fraction The form of a fraction in which the numerator is placed directly above the denominator. Modern spell-scheckers have an extensible dictionary so that the checker may be adapted to the specialized needs of the user. See also header. style file A collection of commands.source code In a high-level computing language. The style sheet can be defined by A the user and saved for other documents. subsection heads. with a horizontal line in between. surprint Printing over a previously printed area. printed above the x-height of the font. strap A subheading used above the primary headline in a newspaper article. See house style. this is the file consisting of the commands pertaining to that language. subordinate to chapter heads. smaller than those on the baseline. s/s (same size) Instruction to reproduce to the same size as the original. that appears slightly above the baseline and is an embellishment to a baseline character—for example j is a superscript in the expression aj . for example a . type specifications. used to denote a reference mark.

transparency A full color. text type The typeface(s) used for the main text of material in a given article or book. template The prototype or paradigm for a standard page layout. See particularly the Windows desktop. text editor A piece of software for entering ASCII code and line breaks/carriage returns. photographically produced image on transparent film.tif file A graphics file that is in the Tag Image File format. ©2001 CRC Press LLC . TLA Three-letter acronym. tone line process The method of producing line art from a continuous tone original. TEX A high-level computing language created by Donald Knuth in the 1970s for the purpose of typesetting—especially the typesetting of technical material. Usually such a report is part of a series. including basic details of the page dimensions. widow and orphan control. tip in To paste an extra sheet into a bound book. paragraph settings. Unlike a word processor. tip and tear To remove a page from a bound book (perhaps one with an error on it) and to tip in a corrected page. and appears with a special cover bearing a logo. tied letters See ligature. trim Cutting of the finished. text wrap See runarounds. *. trash can The icon for the deleting of files or other (computer) objects. thick space A horizontal space of thickness equal to two thin spaces. tags The many different formats which make up a style sheet— automatic section numbering. technical report A document published by a scientific organization or company for external circulation. thin space A horizontal space of thickness about 0. hyphenation and justification. trim marks See register marks. printed product to the correct size.023 .tabular setting Text set in columns. margin and column settings. tie-after accent The accent in text that has the form oo and is given by the TEX code \t{oo}. page layouts. a text editor does not have (hidden) formatting information. tilde The accent that has the form ˜ and is given in text by the TEX a code \~{a} and in math mode by the TEX code $\tilde{a}$.

*. verso pages The even-numbered pages of a book. typography The design and planning of printed material using type. This operating system is public domain. after troff An early computer typesetting system and text formatting utility. upper-case Denotes capital letters. This is the address of a particular page on a particular computer on the Internet. and the art of typography. and originators of text. typographer A specialist in the design of printed matter. The final size of the entire page of a book or journal. creamy book paper. typescript typesetting A typed manuscript. URL Universal Resource Locator. ©2001 CRC Press LLC . editors. type (text) page typewriter-like font A font that simulates the characters that are produced by a typewriter. vellum The chemically treated skin of a calf used as a writing material. photographs. a precursor of TEX. page design. Also used to describe any thick. Still in use on UNIX systems as part of the Documenter’s WorkBench (DWB) system.trim size trimming. vertical justification The adjusting of interline spacing (leading) and manipulation of text in fine increments to make columns and (facing) pages end at the same place.ttf The True Type Font format. See composition. See also even pages. prevents use without permission or acknowledgement. UCC (Universal Copyright Convention) Gives protection to authors. typo Abbreviation for typographical error. UNIX A computer operating system developed at AT&T Laboratories and at the University of California at Berkeley. umlaut The accent that has the form a and is given in text by the ¨ TEX code \"{a} and in math mode by the TEX code $\ddot{a}$. and is widely used on work stations and super computers. An error in the typeset copy or proof copy. A space in typesetting that is equal to two standard two-em quad em quads. or illustration. See also dieresis. See copyright. The total area of the print block on a page. typeface The raised or engraved surface carrying the image of a character in metal. Also used to refer to a complete set of characters of a given family or style.

word break The division of a word—usually between syllables. Contains (hidden) formatting information. Web browser A piece of software designed for accessing locations on the World Wide Web. Used as a copy editor’s mark to indicate that a character is set in the wrong font. work and turn A technique of printing whereby pages are imposed in one forme or assembled on one film.vi The primary text editor of the UNIX operating system. The sheet is printed on one side and then turned or tumbled from front to rear to print the opposite side. WindowsNT. One side is printed and then the sheet turned over and printed from the other edge using the same forme. Windows98. In this process. word processor A piece of software for creating documents. Windows comes in four flavors: Windows95. MathML. and Windows2000. A subset of SGML. weight The degree of thickness or boldness of a character or a font. SGML. WYSIWYG “What you see is what you get”. the automatic rightjustification of lines and adjustment of the number of words on a line. vignette A small illustration in a book that is not enclosed in a border. wf Abbreviation for “wrong font”. and designated by a hyphen—at the end of a line. x-height The height of lower-case characters. ink is replaced by tone. World Wide Web A subset of the Internet consisting of locations that can be accessed with a standard Web browser. XML Extensible Markup Language. widow A brief phrase—less than a full line—from the end of a paragraph that occurs at the top of a page. Windows The operating system created by Microsoft. watermark An impression or image incorporated in the process of papermaking which shows the name of the paper or a company logo. and displays on the screen (in a WYSIWYG manner) what will be printed on the page. such as the x in a font. xerography A photocopying or printing process in which the image is formed using the electrostatic charge principle. word wrap In a word-processing environment. An extensible markup language that improves on HTML in many ways. virgule See shilling or solidus. work and tumble A method of printing where pages are imposed together. See HTML. ©2001 CRC Press LLC .

Copy-Editing: The Cambridge Handbook for Editors. [KNU] D. 1984. Sawyer and S. Krantz. [KRA2] S. A Panorama of Harmonic Analysis. and A.. 1999. University of Chicago Press. [KRA1] S. S. Johnstone. Authors. A. 1994. [SAK] S. Handbook of Writing for the Mathematical Sciences. A Primer of Mathematical Writing. Barrett. Reading. 3rd ed. Goossens. Butcher. [HIG] N. 1954. E. The Mathematical Association of America.sos. The TEXbook. A [GMS] M. The L TEX Graphics Companion. AddisonWesley. The L TEX Companion.. 1969. Providence... and Publishers. How to Find Out in Mathematics. Reading. Krantz. Batey. 1999. 2nd ed. P. Goossens. [OCO] M.References [BUJ] H. Chicago and London. 1995. Knuth. Krantz. Philadelphia. G. and C. O’Connor.com. Reading. Budgett and J. D. Philadelphia. Massachusetts. J. K. Mittelbach. Massachusetts. Function Theory of Several Complex Variables. W. Addison-Wesley. 1986. Chaundy. and F. F. [PEM] J. Addison-Wesley. 1st ed.au/files/glosray. Massachusetts. Reading. Goossens and S. 2nd revised ed.html.. Addison-Wesley. 1994. Florida. Oxford University Press. London.. [KRW] S. [CBB] T. How to Copyedit Scientific Books and Journals. Reading. Rahtz. Boca Raton. Pemberton. John Wiley & Sons. 1993. American Mathematical Society.C. 1997. ISI Press. G. A [GRM] M. The Printing of Mathematics. http://www. 1997. Massachusetts. 1992. 1982. CRC Press. New York. Mittelbach. 1993. Samarin. Addison-Wesley. Washington. G. Cambridge University Press. The L TEX Web Companion. L TEX: A Document Preparation System. Krantz. Cambridge. G. 14th revised ed. Typesetting and Publishing Glossory. A TEX Primer for Scientists. Society for Industrial and Applied Mathematics. Massachusetts. London. A [GOR] M. Pergamon Press. R. Lamport. [BUT] J. A [LAM] L. Rahtz. [CHI] The Chicago Manual of Style. ©2001 CRC Press LLC . Higham.

1987. [TUF] E. R.[SKI] M. Turabian. Providence. Skillin. A Manual for Writers of Term Papers. Connecticut. ©2001 CRC Press LLC . and Dissertations.. Words into Type. et al. Gay. Swanson. Theses. American Mathematical Society. Prentice Hall. Chicago and London. Tufte. Mathematics into Type. [TUR] K. Englewood Cliffs. [SWA] E. New Jersey. The Visual Display of Quantitative Information. 5th ed. Cheshire. The University of Chicago Press. Updated Edition. 1983. 1974. Graphics Press. 1999.

The Practical Stylist. New York.. Bernstein. 1990. 1984. 9. Students Must Write: A Guide to Better Writing in Course Work and Examinations. New York. 1990. 4. 1937. 1965. 6th ed. Scientists Must Write: A Guide to Better Writing for Scientists. Barrass. IEEE Press. D. 12. F. Bryson. The Writing Process: o Composition Writing for University Students. Lund. New York. Bierce. 5. Straus and Giroux. Cambridge. 1971. London. 8.. Chicago: The University of Chicago Press. Communicating in Science: Writing and Speaking. 2nd ed. and Students. 1993. Sweden and Chartwell Bratt. Beer.Resources by Type Sources on Writing 1. 6. Kent. Wikborg. New York. Cambridge University Press. 7. ©2001 CRC Press LLC . London. 1978. Harper and Row. New York. V. S. Mother Tongue: English and How It Got That Way.. Miss Thistlebottom’s Hobgoblins: The Careful Writer’s Guide to the Taboos. T.. 14th ed. 1985. The Penguin Dictionary of Troublesome Words. Atheneum. Engineers. and E. Booth. Knight. England. L. 2. Writing and Speaking in the Technology Professions: A Practical Guide. Methuen. New York. The Union Library Association. Avon Books. Write it Right: A Little Blacklist of Literary Faults. A Manual of Style. 1982. M. A. T. Bernstein. M. Penguin.. 10. 3. Bryson. R. Bugbears and Outmoded Rules of English Usage. Studentlitteratur. Bromley. M. T. 1992. Bj¨rck. B. Farrar. ed. Barrass. The Careful Writer: A Modern Guide to English Usage. 11. 1987. Chapman and Hall. Baker. London. B. 2nd ed.

16. Penguin. 1995. Massachue setts. London.. R. A.13. H. The Mathematical Association of America. E. 1987. Arizona. Ebel. Bugs in Writing. Phoenix. 24. Math. Davis. Scientific English: A Guide for Scientists and Other Professionals. The Art of Plain Talk. 1992. Cambridge University Press. Harlow. Gillman. 1993. Modern English: A Practical Reference Guide. Faber and Faber. New Jersey. A Course in Basic Scientific English. Oryx Press. 26. Cherry. Day. 1989. The English Language. Writing Mathematics Well: A Manual for Authors. Cambridge. Oxford University Press. 20. Reading. 15. Phoenix. Cooper. 1969. Russey. London. Oxford University Press. 27. Monthly 79(1972). 1988. Dupr´. 1992. 14. Guide to Technical Editing: Discussion. D. Burchfield. Regents/Prentice-Hall. Gowers. Eisenberg. Unlocking the English Language. COM-30(1982). 18. Dictionary. Washington. New York. 21. R. Englewood Cliffs. F. Fidelity in mathematical discourse: Is one and one really two?. W.C. Flesch. and W. 1987. The Art of Scientific Writing: From Student Reports to Professional Publications in Chemistry and Related Fields. 23. A. 25. Comm. R. Addison-Wesley. Ewer and G. How to Write and Publish a Scientific Paper. 1987. P. M. London. 1965. and Oryx Press. Longman. C. Modern English Usage. L. 1986. 3rd ed. Arizona. New York. London. and Exercises.. A. 100–105.. M. IEEE Trans. R. London. Oxford. L. Am. 17. ©2001 CRC Press LLC . Collins Cobuild English Language Dictionary. J. The Complete Plain Words. Writing Tools. Harper and Brothers. 2nd edition. VCH Publishers. 1964. Day. 22. H. 28. Essex. 3rd ed. Penguin. B. Penguin. L. Bliefert. 252–263. Latorre. J. Fowler. 29. D. E. 1988. 19.. Frank. 1946. New York. L. Crystal. R. Writing Technical Reports.

Scholarly Publishing 8(1977). Writing Successfully in Science. Graphics simplified: Charts and graphs. The Structure of Technical English. Longman. ©2001 CRC Press LLC . J. E. Knuth. Mathematical Notes No. Philadelphia. Essex. Partridge. Oxford Advanced Learner’s Dictionary of Current English. 1973. Halmos. Merriam-Webster. 1987. 1965. 32. Math. J. Penguin. J. E & FN Spon (Chapman and Hall). E. Longman. D. R. P. Berrycroft. Handbook of Writing for the Mathematical Sciences. M. Merriam-Webster’s Dictionary of English Usage. D. Mathematical Writing. J. 257–274. M. 14. New York. 34. 123–152. P. Washington. Kirkman. D. London.C. Warner Books. 36. Handbook for Academic Authors. 1987. Turton. Larrabee. 38. J. 1991. Publish or Perish. Redmond. 16(1970). P. 42. 46. Chapman and Hall. How to write mathematics. 1987. A. 35. 31. MacGregor. London. Newman. Longman Dictionary of Common Errors. J. Harlow. A. Washington. 41. 1989. Enseign... Oxford University Press. and P.. 45. ed. 1974. 1991. Good Style: Writing for Science and Technology. B. 1989. The Fine Art of Technical Writing. Essex. 4th ed. Cambridge University Press. 43. Inc. Hillsboro. Luey. Springfield. Herbert. 39. Usage and Abusage: A Guide to Good English. 151–164. Scholarly Publishing 8(1977). Cambridge. London. Mitchell. 40. Heaton and N. J.30. Higham. 1993. E.. London. Microsoft Press.. N. 1994. Hills. 44. C. Blue Heron Publishing. A. Oregon. T. Cambs. O’Connor. J. Harlow. MacGregor. Peter Francis Publishers. R. The Mathematical Association of America. SIAM. Perry. 33. Pennsylvania. 37. Graphics simplified: Preparing charts and graphs. 1987. Strictly Speaking. Roberts. The New Writer: Techniques for Writing Well with a Computer. Massachusetts. B. 1992.

Math. E & FN Spon (Chapman and Hall). Cheshire.. R. ©2001 CRC Press LLC . Intelligencer 4(1982). Longman. 57.. 52. Technical and Business Communication. New York. Macmillan. Graphics Press. Oxford University Press. 56. Rhode Island. Tufte. Strunk and W.. London. 1989. 3rd Edition. D. Connecticut. 51. C.. J. 1979. 1974. 17–19. Dell Publishing.. Effective Technical Writing and Speaking. in The Journal of Irreproducible Results: Selected Papers. White. 1990. N. 1986. 1990. London. Swales. George H. English in Use. The Visual Display of Quantitative Information. ed. Piranian. Cheshire. 3rd ed. 50. Practical English Usage. 1971. 205. Steenrod. p.. Say it better. the AMS. 58. C. M. 53. Schwartz. W. 1973. 1980. Graphics Press. 59. Providence. Tufte. L. Kirkman. M. Handbook for Scholars. Thomas Nelson. 55. How to be a published mathematician without trying harder than necessary. Scherr. New York. 2nd ed. Business Books. The Elements of Style. London. Harlow. 2nd ed. A Lighthearted Guide to Grammar and Good Usage. E. 1992. Essex. Swan. New York. Quirk and G. How to Write Mathematics. revised ed. G. Effective Writing: Improving Scientific. et al. Connecticut. van Leunen. 1983. 54. 49. Stein. Turk and J. Oxford University Press. Turner. W. 48. Oxford. Envisioning Information. 1990. Writing Scientific English.47. B. Safire. E.

Documentation for general BibTEX users. Oxford University Press. Sissors. Addison-Wesley. Gay. E. Hart’s Rules for Compositors and Readers at the University Press Oxford. Baskette. Peachpit Press. 71. ed. and B. Patashnik. O’Connor. Digital Typography—An Introduction to Type and Composition for Computer System Design. O. New Jersey. Englewood Cliffs. ©2001 CRC Press LLC . 1990. CA 94708. R. 3rd ed. Reading. Addison-Wesley. R. The Art of Editing. 63. F. 69. Oxford University Press. Real World PostScript. Berkeley. 64. 67. M. How to Find Out in Mathematics. 1988. The Printing of Mathematics. File btxdoc. Knuth. Document production: Visual or logical?. Philadelphia. 68. Cambridge. Learning PostScript—A Visual Approach. 72.. 5th ed. 34(1987). 1954. Barrett. S. G. C. S. Pemberton. Cambridge University Press. BIB TEXing. 1986. 61. and C.. Soc. J. H. Batey. Smith. February 8.60. Oxford.. ISI Press.. Copy-Editing: The Cambridge Handbook for Editors. M. Roth. Z. Authors. Thinking in PostScript. Prentice-Hall. Reading. Reid. 65. 70. 66. Mathematical typography. L. Techniques from PostScript Professionals. W. 62. New York. D. Brooks. Massachusetts. R.tex in the standard distribution. E. 1969. K. London. 1992. Butcher. M. Macmillan. 1988. 1085 Keith Avenue. E. 2nd ed. Skillin. E. Notices Am. Words into Type. 1983. 1988. J. Massachusetts. 1994.. Hart. J. Reading. Chaundy. Pergamon Press. 1992. and Publishers. R. Addison-Wesley. T. 337-372. Rubinstein. How to Copyedit Scientific Books and Journals. 1990. et al. Math. Bulletin of the AMS (new series) 1(1979). 73. London. P. 621–624. Massachusetts. 39th ed. Lamport.

Massachusetts. Addison-Wesley. 1991. D. CRC Press. E. a 78. Sawyer and S. D. 1979. Lamport. E. Bedford. 1993. Columbia University Press. A bag for editors. Boston. Scholarly Publishing 11(1977). Massachusetts. Addison-Wesley. A. Math into TEX: A Simple Introduction to AMS-L TEX. Stainton. Mathematics into Type. a Birkh¨user. The METAFONT Book. Reading. 1986. Reading. New York. G. Knuth. Addison-Wesley. A TEX Primer for Scientists. 79. D. Addison-Wesley. Gr¨tzer. R. TEX and METAFONT. S. 1995. the AMS. Wonneberger and F. A 81. 1986. 83. Massachusetts. revised edition. Florida. The American Mathematical Society and Digital Press. Providence. Mittelbach. BibTEX reconsidered. Technical Editing: The Practical Guide for Editors and Writers. Stainton.. L. M. Knuth. The Fine Art of Copyediting. 84. 80. L TEX: A Document Preparation System. Reading. TUGboat 12(1991). Krantz. Massachusetts. M. 75. Massachusetts. New Directions in Typesetting. Sources on TEX A 77. User’s Guide and Manual. 111–119. E.74. 76. Knuth. Swanson. The TEXbook. J. Rhode Island.. G. Reading. 2nd Ed. Boca Raton. 111–124. 1994. 82. 1979. 1992. ©2001 CRC Press LLC . Tarutz.

Sign up to vote on this title
UsefulNot useful