Front matter, back matter, and apparatus

Do I need a bibliography, an index, a glossary, a timeline, a family tree, or a map?

The book's own reading demands decide most of this, and each element answers a different failure. A reader who cannot find a passage again needs an index. A reader who cannot keep a cast straight needs a family tree or a character list. A reader who cannot place events in order needs a chronology, which Chicago lists among the recognized parts of back matter.¹

None of them rescues a structural problem. A glossary added because the prose never defined its terms leaves the prose undefined, and a timeline supplied because the chapters are hard to follow leaves the chapters hard to follow.

Cost is a real input to the decision. Each element has to be written, checked, kept consistent with the text through every revision, and typeset. An editor can say which elements the manuscript appears to need and what each will cost in consistency work. Whether the book carries them is the author's call.

Will there be front matter and back matter, and who writes them?

The standard order comes from Chicago's first chapter. Front matter runs through the half title, series title, frontispiece, title page, copyright page, dedication, epigraph, table of contents, list of illustrations, foreword, preface, acknowledgments, and introduction.¹ Back matter covers acknowledgments, appendixes, chronology, glossary, endnotes, bibliography, illustration credits, index, and errata among others.¹

Authorship divides along a clear line. The author writes the dedication, preface, acknowledgments, introduction, appendixes, and glossary. The publisher produces the title page, copyright page, and cover copy. A foreword is written by somebody else, which is what distinguishes it from a preface.

The part most often left too late is the acknowledgments, because it cannot be finished until the book is. An author who starts a list at the beginning of the project is spared trying to remember two years of help in an afternoon.

Is an index needed, and who pays for it?

Indexes belong to nonfiction that readers consult rather than read straight through, and they are made at a specific moment. The American Society for Indexing states that page proofs sent to an indexer should be complete, legible, and final, which places indexing after typesetting rather than alongside editing.²

On payment, the ASI's own wording is that authors are often glad to have the work done by a professional even if it means contributing to the cost themselves.² That contribution is the usual arrangement in trade and academic publishing, whether taken from the advance, charged against royalties, or paid directly.

The EFA's 2026 rate chart puts book indexing at 2.0 to 2.5 cents a word, or $55.00 to $60.00 an hour, or $5.00 to $6.25 a page.³ Indexers charge per page, per entry, or by the hour, and the specification matters: whether the index covers only the main text, or also footnotes, figures, tables, and appendixes, changes the size of the job.²

Are there footnotes, endnotes, or citations, and does that change the edit?

It changes the edit considerably, and it changes the quote. Notes are prose and they are read at the same rate as prose. They also carry more that can be wrong, including page numbers, dates, publisher names, spellings of authors' names, and the agreement between each note and its bibliography entry.

Editors Canada includes checking mechanical consistency and factual accuracy under copy editing, and the apparatus is where most of that work lands in an academic book.⁴ A manuscript with 900 notes is not a manuscript of its main-text word count.

The practical consequence is that a word count offered to an editor should include the notes. A quote given on the text alone will move once the notes arrive, and the movement will not be small.

Are there images, captions, tables, or figures to account for?

Each of them is a separate object with its own numbering, its own caption, its own credit line, and its own place in a list at the front of the book.¹ Chicago treats lists of illustrations as front matter and illustration credits as back matter, which means one set of pictures generates entries in three different places.

What an editor checks is agreement rather than the images themselves. Figure numbers in sequence, every figure mentioned in the text, captions consistent in style and capitalization, and credit lines matching what the permission actually grants are all checkable.

Images also carry a schedule. They need resolution suitable for print, they need permissions, and they need placement decisions that cannot be made until the text is set. Naming them early is what keeps them from becoming a production emergency.

What happens when a figure is referred to but not supplied?

That produces a query only the author can answer, because nobody else knows what the figure was. A sentence that says the diagram below shows the sequence is unusable without the diagram, and the sentence cannot simply be cut without leaving a gap in the argument.

This happens most often with material the author intended to draw or to commission and never did. It also happens when a figure was deleted during revision and its mention survived elsewhere in the chapter.

The check that finds these is mechanical. Every figure mention in the text is matched against the supplied files, in both directions, and a figure supplied but never mentioned is caught alongside a figure mentioned but never supplied.

Running it early is worth the half hour it takes. A missing figure found during a copyedit stalls a chapter, while the same gap found by the author before delivery is a note on a list.

What happens when a quotation has no source?

The quotation stops where it is. An editor cannot supply a source for a line the author found somewhere, and guessing at one would be worse than leaving it blank.

Three outcomes are possible and the author chooses among them. The source is found and the note is completed, the quotation is cut, or the sentence is rewritten so that it no longer quotes. What is not available is publishing the quotation with an invented attribution.

Unsourced epigraphs and quotations remembered rather than looked up are the usual cases. Each is worth resolving before an edit rather than during one, because an unresolved quotation becomes a query, and queries are the slowest part of the process for everyone.

Quotations that were looked up years earlier deserve a second check. Wording drifts in transcription, and a quotation reproduced slightly wrong is a worse problem than one left unsourced.

Is there quoted material from other sources, and are permissions cleared?

The clearing rarely is, and the timing matters more than authors expect. Permissions have to be settled before typesetting, because material that turns out to be unavailable has to be cut or paraphrased, and replacement text written at that stage has never been through an edit.

What an editor needs at the start is the inventory rather than the clearances: what is quoted, from where, how long each extract runs, and whether permission has been sought. Structural editing includes determining permissions needs, so building that list is work an editor can share.⁴

Whether a given use requires permission is a legal question rather than an editorial one, and it is treated separately on the page about accuracy, real people, and permissions. Every document a client sends us is treated in confidence.