Original publication date!!!

I'm starting a new thread because I think I've posted about this in at least three other threads:

The "Original publication date" field has been added to Common Knowledge!

conceptDawg, I love you!

I know that sounds so over the top, but that's the first thought I had when I saw the field! I can't wait to have a day off from work so I can start filling in more of these fields you added! YOU ROCK!!!

You're welcome. Let me know if there are problems.

Awesome! Thanks!

Woo Hoo! Thanks!

When I think about it, I put the original publication date of a work in my copy's Date field. That way, in my profile stats, I get to see the distribution of original dates of publication (or I would if I were consistent about entering it). I leave my particular copy's publication date in the Publication field.

As for things that might apply to CK, though, I haven't yet found a satisfactory solution to the problem of "original publication date" for anthologies (which I don't consider to be original works in their own right). Average year of original publication? Earliest year of original publication?

Anthologies! A big problem for me in many respects. I hope someday there is a solution (she said hopefully...).

#7 - Since you can add multiple dates, so you might be able to add a date for each work and identify the work within parentheses. I say "might", because I don't know how many dates we can add.

Original Date would be even better as a proper DB field we could sort on. But I certainly welcome it as a CK field.

Anthologies: I look forward to the day when we get improved inter-work relationships, then I can enter the items I have, often several times, in various collections and anthologies and relate them to the collections/anthologies I have them in. And, give each one its own first publication date. (But then I'll start wanting to record in which source the first publication was, even if I don't own it!)

Hurray! For me, one of the most useful new fields.

I've seen (on at least one book) someone enter two entires for this date. One being the actual first published date and other being the copyright date. The explanation in brackets after the copyright date was to link with other books from the same year. This seems really silly to me. The functionality may not be best there yet to utilize this data but entering the same information twice is pointless.

11: Are they entering the same information twice? Or are the years different for original vs. copyright date? If they are different, then I'm not sure that I see a problem with it, but convince me....either way.

> 11 & 12 - I've seen a couple of these this morning too. I can see reasons to have more than one date in certain cases; I mentioned when first asking for this field that Leaves of Grass by Walt Whitman could be problematic, because he printed it, then added poems and reprinted, then had at least a third edition of it. I would think that those dates could be entered on separate lines.

Another case where I might accept more than one date is when a work isn't published until years after the author's death. I put in several dates for E. M. Forster novels the other day, but stopped at Maurice, because while it's in placed in MY library by the approximate date it was written, it wasn't *published* until after his death (if I'm recalling correctly), some 60 years later. Same for A Confederacy of Dunces - I have it shelved with the time it was written, not the time it was finally published.

I did see some entries in CK this morning where a title has both a copyright date, and a first paperback publication date - often just one year later. This helps some with recent books, but that seems more appropriate at an edition level than a work level, because as the dates move further in the past, works of lasting value will be republished multiple times, often with different introductions by different authors, and so on. To me, the phrase "original publication date" puts a limit on the kinds of dates that can be used there.

Sorry to go on so long, I get pretty particular about this one. ;-)

I put multiple dates on my Tom Sawyer/Huck Finn book original for each book.

I also put the 2 dates for some of my books that were printed in England 1-3 years before they were available in the US. I put both the British date and the US date of first published. Don't know if people like that or not.

I put both the British date and the US date of first published. Don't know if people like that or not.

I see the draw here since so many of us are in the US, but I'm not sure why the US date would be any more relevant from an unbiased standpoint than the Canadian, or Australian, or French, or, etc etc. I don't think it would necessarily bad to have available lots of publication dates, for lots of different countries, in fact I think this would probably be interesting. But I'm not sure that "original publication + US publication" would be such a good idea.

Actually it might be interesting to see when books were available in different countries, for a first published. The books I did this for are ones I buy as soon as they are available here in the US so those dates were the important ones for me. I also only found dates for British and US editions, no others. ( I certainly didn't mean to imply US dates were the only 'other' first dates that were important.)

The other case when multiple dates might be useful is when the book is a compilation of several works originally published separately. I haven't done any of these yet, but I was thinking of indicating the title for each separate work in parentheses. This might become redundant, of course, if LT ever creates a function to distinguish parts of compilations.

IMO the original publication date should be just that. If we want copyright years and first date of publish in different languages (and possibly different countries but thats less important) then they should be additional CK fields rather than all tagging along in the same box. Maybe some of these things should be a 'normal' field not a CK field!

I think CK is most powerful when you can gather statistics from it. For example a feature similar to the average age of your books but based on CK data. If there are multiple pieces of data this kind of information cannot be computed accurately without extra work on an algorithm which cannot be perfect.

Basically I'm saying keep only the original publication date in the original publication date field.

So what is the original publication date of a book that was originally published serially (e.g., Great Expectations) -- the date of the first serial installment (December 1860), the last one (August 1861) or maybe the date that the full text was first published as a single volume (since, after all, the work touchstoned above represents the single-volume publication of this text)?

Isn't this inherent in our concept of a work? Seems to me that different editions of a work will have different original publication dates. If we were attaching original publication dates to a record of a single edition, there would be a single date. Is there a way for us to note the edition the date references?

notelinks @20:

Different editions of a 'work' will have different *publication dates*, not different *initial publication dates*. Except for edge cases like Great Expectations or the Sherlock Holmes stories where a serialized version was published first (and for those I'd put the date of the first publication in the collected form as the 'initial date') they'll only have one *intial publication*. It's not "When was Pride and Prejudice first published with the introduction by so-and-so, or first translated into Spanish"; it's "when was Pride and Prejudiced first published?"

There's not explanation of that opinion on the page where the original publication date is entered. I mentioned it because this may be one source of the confusion on the use of multiple dates.

I doubt that prepublication of some sections of a book are an edge case. That practice isn't at all unusual in nonfiction. Authors often compile collections of previously published articles with new material added in a new work. Finding the original publication date for those articles will definitely require more research-- and, especially when the possibilities of open access publication are considered.

I've been experimenting with this a bit.

A. #11 fluteflute & #12 conceptDawg: I put in two dates for Harry Potter and the deathly hallows. First I put the release date, since I know it. This is the only book for which I know the actual day of release. But, when I clicked on it, the date didn't link anywhere. So, I decided to add "2007" as a separate entry solely to provide a link to all 2007 dates. Questionable reasoning, but the intention was for it to be temporary. It's now been removed, I won't add it back.

B. So, conceptDawg, something to file away: "2007-07-21" does not link with "2007." Not sure how that should be handled.

C. I've put into multiple dates for the Chronicles of Narnia box set - one for each book.

D. I put in multiple dates for nonfiction books with several editions. All editions of these books are typically combined, and impossible to separate as most users don't add edition info. But, each edition is updated and typically quite different.

E. I've considered adding separate dates for each work in an anthology.

(I started this before # 23 was posted, but then couldn't submit. Re: # 23 item D, that relates to my note below: same title, same author, but changed or updated content ---> pain in the neck!)

22 notelinks wrote:
"There's not explanation of that opinion on the page where the original publication date is entered. I mentioned it because this may be one source of the confusion on the use of multiple dates."

Since this and other new fields were only added three days ago, it makes sense that there's not much in the way of guidelines (except the actual format of YYYY-MM-DD). For the most part, I agree with lorax at # 21, and fluteflute in # 18, that there is only one *original* publication date. But as I said in # 13, those cases like Leaves of Grass, where a later publication by the same author, with the same title, but some amount of DIFFERENT content than the earlier version, are at least one category where assigning a SINGLE date is problematic.

Part of me is thinking, "Darn it, let's just decide NOW what we're doing and have it put in the guidelines," but the fact that we're having the discussion and sharing different opinions (and some people who aren't in the conversation are putting in multiple dates) makes me think we should wait a bit longer, to see what shows up on the pages, and to allow users to share their rationales / examples where more than one date could be reasonable. That's how I feel for the moment, anyway. ;-)

By saying, "There's not explanation of that opinion on the page where the original publication date is entered", I didn't expect there to be one. This is an early discusssion of something new and we always have differences in opinion. When I wrote that, I was acknowledging that the use of original publication date still is opinion and I was emphasizing "original" instead of "initial". (Lorax had misquoted the terms used in his reply to my first message.)

This field, attached to a work page, does seem more controversial to me than some of the others. It may simplify too much. How can we provide one date as the original publication of a work when, in some cases, textual critics cannot agree upon which is the original version of a text?

>25 jlane: notelinks
I'm sorry I misread you. Also, thanks for pointing out that lorax had used "initial" publication date. That didn't even catch my eye. I think many people, myself included, might use the terms interchangeably. I know that when I first asked for original publication date, I was essentially thinking "date when this work was first published."

Looking back at your message 20, I agree that it does go back to the concept of what a "work" is. Also, I think I've said "edition" at times when I meant "book" - actually, I did, above in message 13. A second edition, new edition, revised edition - those situations where the content included by the main author is different than the content of a prior published version - I can see where those could be considered separate "works." Where I said, "(including both copyright and paperback dates) seems more appropriate at an edition level than at a work level," I should have said "book level." In the great majority of cases, in the first paperback printing (not edition), the contents of the work first published in hardcover are the same. The same is true of the work Pride and Prejudice: no matter how many new hardcover or paperback or movie tie-in printings there are, with introductions by various literary scholars through the years, the primary content of that book is the same work that was first published in the early 1800s.

Where you say that there are cases where textual critics can't agree which is the original version, could you please share an example? I don't doubt what you say, but if you could provide an example, I think it would help me to better understand your perspective. I've mentioned Jane Austen (touchstone not working) and Walt Whitman (also not working) because most of my library and experience is fiction and literature. I'd like to look at it from another point of view, to learn something I hadn't thought of.

Especially for earlier publications, differences in typesetting and editing were substantial enough to affect meaning or intent. The ending of Dicken's Great Expectations might be an example. According to the story about it, Dickens changed the unhappy ending when he edited it for publication, but editors prefer it as he first wrote it. The Bible or ancient classics are others. If we use the date of the work, perhaps beyond dates printed in our own books, how do we determine original publication dates for those?

Edited to add:

>28 HeathMochaFrost:: HeathMochaFrost,

I've added to my message from today to, hopefully, make it clearer. The additions are in italics.

Especially for earlier publications, differences in typesetting and editing were substantial enough to affect meaning or intent. The ending of Dicken's Great Expectations might be an example of the question of which is the original version. According to the story about it, Dickens changed the unhappy ending when he edited it for publication, but editors prefer it as he first wrote it. The Bible or ancient classics are other examples of the difficulty of determining what was the original text. Those have been published in differing translations. If we use the date of the work, that may be earlier than dates printed in our own books, how do we determine original publication dates for those?

Dallas Willard's most recent book The Great Omission is an example of a non-fiction work that compiles previously published articles with added material for publication. The first chapter of that book, "Discipleship: For Super Christians Only?" was, according to credits, first published in the magazine, Christianity Today, on October 10, 1980. The copyright on the book is 2006. Other chapters were first published on various dates. Which would be considered the original date here?

Obviously, we could really overdo detail in this. I mention this to emphasize that making it too simple, establishing one firm rule won't work. Publication histories are too complex for that.

>27 jlane: notelinks
As EM Forster might have written, I feel I'm in a muddle. And, I need to go read my sons their library books & don't know when I can get back to this.

>12 conceptDawg: conceptDawg
You said we should convince you one way or the other, about whether it's acceptable to have more than one date entered in this field. I guess I feel a bit responsible for it, as I was one of the first to request it (as far as I know). Has the subsequent exchange caused you to lean one way or the other, or to say with any certainty what TYPES of cases could have more than one date?

Would having a few Combiners help us reach a consensus?

I just entered my pop-up copy of The Night Before Christmas. I used this book's original date (2002). Should I also be looking up the story's original date and adding it too?

I have to say that I enjoyed adding all the reindeer in the character fields.

28:
I'm still not convinced either way. Honestly, I'm not sure we can reach a definite answer here. Message 29 is a good example of this. For now I'm going to stick with having multiple entries. We can revisit if we have some new facts or ideas.

I don't think #29 is difficult. It should be the original publication date of the work. Then we can shove all difficulties off into the "what is a work?" question :)

>30 conceptDawg:
Thanks for reviewing the discussion & posting your thoughts. I'll try not to bother you again so you can focus on Collections - yay!!!

>31 Amtep:
Amtep, I've been thinking along the same lines, that this is looking more like a question of "what is a work?" than whether multiple dates are appropriate. LT is called a "library-quality" catalog, but on the page about LT concepts, Tim says that the purpose is social - that people who own a certain "work," regardless of what edition it is, are likely to have things in common. (I'll look for the exact quote in a bit.) If some of the cataloging-specific details are obscured in favor of making connections with others, than I suppose multiple dates really don't matter much, as long as there's a note in parentheses defining what the date is - first edition, second edition, first UK edition, copyright, whatever. But I really DON'T want to see the dates of 50 different copies of Great Expectations! I really hope no one decides to go THAT route.

(Edited to add notes from the Concepts page. Under "What is a work?" it reads, "The purpose of works is social. Books that a library catalog considers distinct can nevertheless be a single LibraryThing "work." A work brings together all different copies of a book, regardless of edition, title variation, or language. This works system will provide improved shared cataloging, recommendations and more. For example, if you wanted to discuss M. I. Findley's The Ancient Economy, you wouldn't really care whether someone else had the US or the British edition, the first edition or the second." Then, under "What are some edge cases?" the first two sentences are, "Remember, the purpose of the system is social. Therefore, I feel that some edition or language differences are so major as to be socially significant.")

As always, more to say but no time. Hopefully later...

> 23 dchaikin

I would agree with your points C, D and E that in those cases at least, there could be multiple dates used.

>31 Amtep: Amtep

That's the way I would think too. You aren't adding information about your particular version of the work, but information about the work itself.

My view is that for the majority of fiction books there should only be one date - that when it was originally published, wherever that occurred and in whatever format. Otherwise the possibilities for "original publication date" begin to multiply - UK hardback, US hardback, UK trade paperback, superdooper edition with goldplated cover...

Okay, I may have made up that last one :-)

Yes, there are edge cases, but I would tend to go towards one date where possible and only add other dates where there is an appreciable difference in content between different editions, or where the "work" has been previously published in different parts already.

From comments made by Tim & Crew, the work/book dyad is likely to become a triad in the not-too-distant future: work/book/edition (with separate CK data for editions!). If and when this comes to fruition, #29 may be accomodated by assigning different dates to the work and the edition (this can be done even if the pop-up version is not combined with the storybooks). Of course, there will be new points of contention, but that will all be part of the fun...

I agree with 31.

Original publication date seems to clearly be a work-level parameter. If the pop-up edition is different enough to be considered a separate work -- as they usually are -- then it's different enough to get its own date.

Really these are the same issues combiners deal with all the time -- people more involved with the works tend to see differences between editions as being very important when the general public might not, so the general public wins out unless there's a particularly dedicated and stubborn combiner dedicated to a particular work.

I don't see anything in message 23 that justifies multiple dates. There should be one original publication date for a work. If a later edition has enough changes to give it a new date then it has enough changes to make it a new edition. The box set and the anthologies need to wait for a new work/subwork structure to accomodate them.

A comment I wanted to make that's NOT related to the question of multiple dates:

>5 drbubbles: drbubbles
I really like your idea of just using the standard Date box on the work page to put in the original publication date. If I'd thought of that, I might not have requested the new CK field. Since it was always populated already with the date of that edition or printing, I just never thought about changing it. When I first read your post, I had one of those "D'oh!" moments, like, Why the heck didn't I think of that? So, good idea!

As far as anthologies, where every poem or story included is actually a literary work in its own right, being able to indicate the original publication dates of all the contents would be helpful (as jimroberts mentions in # 9). But as far as your comment that you don't consider an anthology itself a "work" - I agree it's not a literary work itself, but I see it as a specific collection of those smaller works. When editors decide what to include and exclude in a collection, and how to group and arrange the contents, they're making creative decisions. In those cases, I'd use the date when the collection in hand was first published.

Well, there's the case of Kant's Critique of Pure Reason, originally published in 1781 and, with a second, significantly revised edition in 1787. But it seemed to his readers that both editions were important, so at some early point someone had the bright idea of combining the two editions, and practically every edition since then, in every language, has been a conflation of the original two editions. Where they are identical, the text runs along as usual; where the differences are slight or not very lengthy, the text of the two editions is in parallel columns; and where the differences are lengthy or substantial, the text of the 1st is immediately followed by the text of the 2nd for that section.

This is a case where I think two dates should be used. I've long been recording this sort of information in my library in the comments field, in this case using the form "Orig. ¹1781, ²1787." (Of course, if it's absolutely necessary to choose only one date, it would have to be 1781, and prehaps this is the best we can do.)

There are plenty of other similar cases, like Leaves of Grass, or Unseen warfare, a work with an incredibly complex "original" publication history.

What these examples have in common is that the revisions between editions were made by the authors themselves, which makes these types of "editions" logically distinct from "editions" issued under the auspices of someone else. So it's not only a question of what counts as a "work" but also of what counts as an "origin." My feeling is that all changes made by the authors themselves are "original," so that it is possible to have a single "work" originally published over a range of dates.

what would people suggest using for plays, since hey are generally published a year or so after their premiere? Look at Shakespeare.

re #39

I would take a literal "publication" date for plays; it's not date written or date performed, it's date of original publication.

I'd do both with (published) and (performed).

Since I can't tell what (if any) consensus has been reached on the multiple date issue, can someone please clarify for me what I should do with this:

I wanted to enter the original publication date for Cynthia Ozick's (touchstone not working -- does this relate to the strange error message about LT's "roomful of trained monkeys" I get when I click the 'wrong' Cynthia Ozick?) book The Shawl. The book was first published in 1989. One of the two stories in this book was first published in 1980 and the other in 1983 (both in the New Yorker -- both also likely appeared in the O. Henry prize collections for those years, but I don't have them to confirm). I have only listed 1989 in the publication date field, and put the other two dates in the description field (which I'm not sure is what's it's meant for, but I didn't know where else to put them). Should the other two years also appear in the publication date field?

I'd say no. We're talking about the book, not about the components of the book. The latter would be a rather harder problem for anything with poetry or short stories in it.

Next we could have the original publication date for each work on the author page (with of course an option to sort on it).

Probably not worth doing until we have a lot more dates.

I notice that no one has mentioned the situation on "original publication date" that I have been working with. I have a fair number of books that are facsimiles of unique manuscripts, such as The Cloisters Apocalypse. I ended up putting in three dates:
1. The date of the original unique manuscript (e.g., 1320 ca. -- I put the ca. behind the date so it would sort properly)
2. The date that the text was first published (0095 AD seems to be the accepted date for the publication of the text of Revelations)
3. The date that the facsimile was published (1971).

Where possible, I have then been identifying the shelf mark of the manuscript in the book description section.
(edited to remove extraneous sentence fragment)

By the way, I really enjoy having the "Original Publication Date" Common Knowledge field available.

It allows me to see where original manuscripts fall with respect to others. I could see the Gabriel Bise published version of Gaston Phoebus's Book of Hunting, based on a French manuscript, and the Getty version of the same text with its own illustrations, being entered at different "Original Publication" dates, even though the underlying text is the same.

This allows LibraryThing to be used for manuscript comparisons, too (at least, for those mss. that have been published in facsimile).

Perhaps I am the only one that cares, but that delights my medieval-studying little heart!

Admittedly, I'm the only one entering these at the moment, but perhaps others will follow suite with ms. facsimiles that I don't have.

EowynA: I just hit much the same problem with the Philobiblon - was the original publication date the date it was written (1345) or the date it was first published (in the modern sense, at least - 1473)?

It strikes me we'll have this dilemma for any work that predates printing or was originated in a manuscript-heavy culture (consider, say, Soviet-era samizdat) - or which just lay unpublished for a long time.

Strictly speaking we should use first printing date, I guess, but it doesn't seem right - we are, after all, looking at the work as a whole rather than at any individual manifestation of it, and we should use the date the *work* was created rather than when the first "conventional" manifestation became available...

Shimgray,

First, I tend to think that in the case of manuscript facsimiles, the form of the expression (the pictures, the hand used, etc) is usually more important than the words of the text, and often the owners of the facsimiles cannot actually read the Old English, Medieval Latin, etc. words on the page.

When considering a manuscript facsimile book published modernly, like the facsimile of The Book of Kells, the community of people that will connect through it are probably not ones who are reading it because they are devout Christians who wish to read a Latin version of the Gospels, but rather people who are reading it because they are calligraphers, illuminators, and students of Insular art.

Second, I am not familiar with the Philobiblon, but if the source of modern versions of it is from the 1473 publication, then I agree that really is the original publication date. But I'd probably include the date it was written, too, if that were known at all. Because to me, when a book is first available to the world is fundamentally its date of publication. Of course, some works were written and not made publicly available until much later (like Emily Dickinson's works), so clearly the date written is NOT date of Original Publication for some works.

Books were published (that is, multiple copies were made) thousands of years before the printing press was invented. I find it fascinating seeing the timeline of works written before Gutenburg that are still read today (we know they are read, or collected, because someone has cataloged the work in LibraryThing).

Is there a reason we need to speak strictly when filling that field, since pre-printing press we know the work wasn't "published" in the same sense as after. Technology is changing, so I'm content to include the first publication date on books first "printed" as an e-book in that field, too.

Bingo. We can't really talk of "publication" in the manuscript era; the best we have is a sort of evolutionary process, where surviving manuscripts are the "known fossils", until we get to first printing and it freezes into discrete editions. Effectively, they were "published" as soon as they were first written and made available for copying...

In this case, contemporary versions date to an ~1880 translation made by comparing multiple contemporary manuscripts, so goodness only knows what you'd take as the date there... it's really something that needs seven or eight footnotes to explain. The early printings are interesting but irrelevant in terms of modern textual versions, and I suspect the same is true for many such works where manuscripts still survive.

I'm tending towards "first published" being a vague field for whatever looks most appropriate! ;-)

I agree! :-)

For a book written after the invention of printing, that field can be narrowly defined. For books written prior, a lot less narrow.

I decided to post to this discussion with regards to dates in different languages. I think a work originally in Spanish should have that date as well as the date of publication in English translation. Thus -- La tabla de Flandes or the Flanders Panel by Arturo Perez-Reverte -- would be 1990 (esp), 1994 (eng). A problem is that there is no place for orignal language in the Common KNowlege section.

Is there a consensus yet about using serialization date versus first book publication date in the Original Publication Date field?
I've been entering Tarzan dates and titles and have been using book date rather than serialization date,but just noticed even Wikipedia uses different dates on the ERB page versus the Tarzan series page.

If I want to include the serialization info, would it be better as a second date, a comment under book description,or under disambiguation notice?

This is the first time I'd heard about the date controversy; I'd managed to miss the kerfluffle the first time around!

My personal feeling is that it should be the first publication as a book, ignoring the serial. Changes are often made between serial publication and print publication (Dune, Dune Messiah). And the serial installments, whether you list them individually or as their magazines, wouldn't be combined with the work in a book; since you wouldn't combine the works, you shouldn't use one's date for the other.

I think you can put info right with the first date of publication. So, you could put in

1953 (serial)
1955 (book)

I've done that on language.

Now how straightforward is that! You'd think I'd skipped reading the whole above discussion and forgotten about those useful little plus signs - must have been due to the 3am work attempt.

Seriously,thanks so much.