User:Mike Christie/Debates

Jump to navigation Jump to search


This section is to keep track of contentious issues; I might end up making a FAQ section out of this. At a minimum I plan to summarize the debate positions and any consensus.

Things to include in each summary:

  • What the issue is
  • What possible answers/solutions there are
  • Any future features or database changes that might address the issue
  • Current consensus, if there is one
  • What the help files say, including links to the relevant help files
  • Whether any of the outcomes are unacceptable -- e.g. if a moderator sees an edit of one type, should they reject it?

Issues that have been debated

Should title.title always match pub.title?

There are two separate fields in the database where a "title" can be recorded: one in the title table, and one in the publication table. These are referred to here as ttitle and ptitle, for brevity. Publications usually have "parent titles": that is, a book will have a publication record, which contains one or more title records, one of which is regarded as the title that represents the publication itself, as opposed to a short story within it. For example, this publication of Le Guin's collection "The Wind's Twelve Quarters" contains title records for each short story, but also a title record that has title-type of COLLECTION and has ttitle of "The Wind's Twelve Quarters".

Should the ttitle match the ptitle in every case?

Note that translated works are a separate issue which can lead to ptitles and ttitles not matching. See the separate translation debate, below.

Other than translations, there are two reasons why one might expect a ptitle to differ from the ttitle.

First, if the ttitle is used to track multiple different versions of a book, which are all the same story, under different titles or pseudonyms, then there would be different ptitles for that ttitle. For example, suppose you have two books: one is titled "A Midsummer Tempest" by Poul Anderson; the other is "A Mid-Summer Tempest" by Poul Anderson. Both these publications could be recorded under the single title "A Midsummer Tempest". The ptitle would still be accurate; the single ttitle would simplify the bibliography. The alternative is to use the variant title mechanism within the ISFDB to create a separate ttitle for each publication, and mark one as a variant of the other.

Second, if a book has a subtitle, the subtitle could be placed only in the ptitle. The ttitle would then record everything up to the subtitle. For example, Le Guin's "Tehanu: The Last Book of Earthsea" would have a ttitle of just "Tehanu".

Arguments for using a single ttitle for varying ptitles:

  • It can be a significant simplification of data entry and readability in the bibliography
  • No data is lost, because the ptitle still records the accurate data. (See below for a special case where data could be lost.)

Arguments against:

  • The variant title mechanism is the only way for the author bibliographies to display pseudonyms and alternative titles. For a title such as Lewis Padgett's "Robots Have No Tails", which was republished as "The Proud Robot, by Henry Kuttner and C.L. Moore, the title could only appear under the bibliography of one of those two author names.
  • Some information could be lost in certain circumstances. For example, suppose Le Guin's "Tehanu" appears in an omnibus with the full title and subtitle: "Tehanu: The Last Book of Earthsea". Either a separate ttitle is created for this version, just because the title appears in an omnibus, or else the "Tehanu" ttitle is used, which doesn't show the actual title printed in the omnibus. A similar comment applies to minor typographical variations, or pseudonyms, in novel titles in omnibuses.
  • Short stories and publication-length works would be treated inconsistently. Short stories, having no publication record, have no choice but to use the variant title mechanism.

There is not unanimous agreement on this issue. The help files currently assert that all differences in titles should become variant ttitles, under all circumstances, and that subtitles should be included in both ptitles and ttitles, so that they match. This is not a consensus view. Relevant help files are Template:TitleFields:Title and Template:PublicationFields:Title.

For now, there is no reason for a moderator to reject edits that adhere to either convention, at least for minor changes to the title's punctuation, and for subtitles. Significant changes to the title, and pseudonyms, should always require a variant title. However, so long as an edit does not obscure the truth, or delete information about the true state of affairs, then moderators do not need to reject edits or feel obliged to correct them afterwards to conform to either approach.

Should titles include subtitles?

See also should title.title always match pub.title?.

Often a subtitle is not widely used as part of the generally reported title for a work. For example, "Time Cat: The Remarkable Journeys of Jason and Gareth" is more widely known as simply "Time Cat". However, this is not universally true. "The Faded Sun: Kutath" is generally known by that full title, since there are three "Faded Sun" novels.

Should a subtitle be included in the title.title? What about the publication.title?

The argument for including it in the publication is that the publication record should be as close as possible to an exact representation of the publication. However, if it is included in the publication but not the title record, then they do not match; see should title.title always match pub.title? for the issues this raises. If it is included in both, then the bibliography will in many cases display a title that is not what most readers are familiar with. In addition, variant titles are then likely to appear which differ only in the presence of a subtitle; this is not interesting to the reader at the author bibliography level, but they will display there.

Possible future changes to the system that could address this include a separate field for recording subtitles. Distinguishing subtitles would allow improved bibliography displays while sacrificing no data.

Currently there is a consensus that the publication records should include the subtitle. There is not universal agreement on including them in the title as well, but most editors do not do so.

The help files say that the subtitles should always be included. This doesn't represent a consensus. The relevant files are Template:PublicationFields:Title and Template:TitleFields:Title.

The only edits that a moderator should be concerned about that relate to this issue are title or pub edits that remove a subtitle from a pub.title. These edits should be queried to determine if the subtitle exists on the publication. If it does, the subtitle should be left there. Adding or subtracting a subtitle to a title.title can be left to the editor's discretion at this point, as there is no consensus on this issue.

Is it always possible to identify the parent title for a publication?

Each publication in the ISFDB has an unambiguous list of content titles, which can be seen by editing the publication. One of these should be the "title reference"; that is, the title record which corresponds to the publication as a whole. There will generally also be title records that refer to partial contents such as art and shortfiction. The title record that corresponds to the publication is sometimes called the "parent" title record. It doesn't have to be unique to this publication -- for example "Nerves is a title record that displays five publications.

Note that the identifying the parent title, or title reference, for a publication is not really necessary in bibliographic terms. No information is lost if the database doesn't know which is the parent title. However, it makes navigation easier, because if you display one of the publications, such as this one, there is a link for "Title Reference" which points back to the parent title. Without this link it would be awkward to navigate from a publication up to the title record that shows all publications of that title.

However, there is no "marker" in the database that says a particular title record is the parent. So when scanning contents, if there's a title that matches the pub author, pub title, and pub type, it's assumed to be the parent title and treated as such. That includes displaying it for editing in the editpub screen, and displaying it as a title reference link in the publication display. If no such title is found, the title reference line is not displayed. There are many publications in the ISFDB that do not have correct title references, and since more can be created (temporarily at least) by good faith edits, it's not possible to prevent this situation completely. In addition, magazine records do not always have a title references as the EDITOR types are often merged into a single record for each year of magazine publication.

Existing problems are of two types: either the reference is missing completely, or there is a record present which should be the title reference but which is invalid (e.g. is an ANTHOLOGY title but the publication is a NOVEL). These are situations that editors should be aware of but which cannot be permanently fixed with the current design and usage conventions. Scripts to identify publications with missing titles can be run, and these should find and help fix the problem titles. If you find a publication with no title reference while editing, please consider adding the appropriate title record.

Should titles like "Introduction" have the publication title in parens afterwards?

Strictly interpreted, the rules for titles would create many thousands of record for "Introduction". If you're searching for an introduction written by Clarke, and you see a list of his essays with twenty "Introductions" listed, that's not very helpful in finding the one you're looking for.

A commonly used solution is to append the publication title in parentheses afterwards. For example, Le Guin's "Introduction" in her collection "The Wind's Twelve Quarters" would be titled "Introduction (The Wind's Twelve Quarters)". This approach could be taken for any title likely to be ambiguous, such as "Foreword", "Afterword", "Preface", "Editorial" and so on.


  • Helps disambiguate and avoids confusion
  • Straightforward to apply, especially if a list of titles is agreed for which this rule applies.


  • Breaks the rule that you enter exactly what you see, which is a good rule because it prevents differences of interpretation.
  • If the same publication is retitled, but the introduction is re-used, what is the title of the introduction?

An alternative that has been suggested is to have the software display the publication title in parentheses after the title itself, so that the data entered is just "Introduction", but in most bibliographic displays this will show as "Introduction (The Wind's Twelve Quarters"). This relies on the ability to identify the parent title unambiguously, which is not guaranteed. It has also been suggested that an additional field on the title record could be used to capture the parenthetical comment, in order to separate it from the "real" title.

The majority of editors on the ISFDB append the title in parentheses afterwards.

To do

  • Should titles have parenthetical comments that identify text revisions?
I'm leaning towards this one too. Only MAJOR edits though, if it's single content: additional short-fiction content (or removal) might be important too. BLongley 15:38, 14 Feb 2007 (CST)
  • How should chapbooks be displayed?
  • How should chapbooks be internally represented? In particular, should a chapbook holding a single title record it as a NOVEL regardless of length? If so, are there two title records for a SHORTFICTION printed in chapbook form?
  • When partial ISBNs are found on publications, can they be completed for entry into the ISBN field?
I do for SBNs. When there's more than a missing leading "0" (e.g. when the check digit gets replaced by a price), I don't as it would probably fail ISBN validation. BLongley 15:38, 14 Feb 2007 (CST)
  • Do translated publications (English -> foreign) get a separate title, or are they recorded under an English title?
  • How should Ace Doubles be represented? What if they contain anthologies or collections?
  • How should cover artists for Ace Doubles be represented?
  • How should boxed sets be represented?
  • How do we record sf works in non-genre magazines/anthologies/collections?
  • How do you tell when you're looking at a collection vs. an omnibus of works all of which are by the same author?
My rule of thumb: Omnibus is for previously published works with their own pubs. Any previously unpublished content makes it a collection. BLongley 15:38, 14 Feb 2007 (CST)
  • Do we record editors of collections, and if so, how?
  • Do we record cover designers, and if so, how?
Occasionally, and badly. :-/ See my comments on Robert Rankin books where the author did a sculpture that got photographed and then had the results "designed" into a cover. BLongley 15:38, 14 Feb 2007 (CST)

I always try and record cover designers, occasionally it's a company or design group, but I see no reason not to. It might be useful if there were some way to list them as designers though. Also, on few titles I've seen , the designer of a logo, or emblem or natty-looking calligraphy on the title is often listed ( especially on fantasy novels) I usually use add author as an option on this, but was wondering if I should list it on notes inteaad? Thomas conneely 16:59, 14 Feb 2007 (CST)

  • What's the deal with Tor ISBNs?
  • How do we handle fresh linking material between stories?
  • How do we handle section titles?
  • Can we use a printing history list in a late printing to enter earlier printings?
I believe we can, if we cite the source. I don't (yet), as it can add immense numbers of unverified publications at once. I am adding notes about previous printing histories as I find them to the Pub that I AM willing to verify. But I need a bit more experience before I can be sure what fields can be copied to other publications - obviously not price, maybe not page count and ISBN (until I find out what reissue and reprinting means to various publishers, and whether several imprints keep the same type-setting), and sometimes even format is up for debate. I THINK I can spot the first Hardcover editions of British publications when referenced but I'm not SURE yet. BLongley 15:38, 14 Feb 2007 (CST)
  • Does anyone think we should distinguish between publishers better, ie by adding say TOR UK to Uk published titles ( part of macmillan group) to distinguish them from the better known and longer established US imprint aka Tom Doherty? Or is this a waste of time? Thomas conneely 16:59, 14 Feb 2007 (CST)