Hemingway's library: done!

My personal library of 7,411 books is now completely entered into LibraryThing.

Note that this is the largest Legacy Library so far. Not that I have a competitive nature.

My thanks to the LT users who worked on this for the past three months: nperrin, who initiated the project; spookykitten (who added about 2,450 books); christiguc (2,350); Rullakartiina (1,350); jjlong (1,200).

I’m not familiar with the concept of “tags”, but jjlong and Rullakartiina have made a start on adding these to my library. Check them out. I wouldn’t have guessed I owned more poetry than fishing books, but there it is.

I’m enjoying browsing through my catalog- so much easier than rummaging through bookshelves in Cuba, or that storage shed behind Sloppy Joe’s Bar in Key West. I’m pleased to see my copy of Savage Landor’s An Explorer’s Adventures in Tibet, a Christmas gift from my mother in 1917, and now in a private collection somewhere.

Now, back to my poker game. Bill Faulkner is into me for ten bucks.

___

notes from the EH group:

Tags are a work in progress - subjective and somewhat hit-or-miss right now; volunteers for more tagging are welcome!

For more information on Papa’s library, and on his reading habits (reading obsession?), check out the introduction to our source material, the list compiled by Dr. James D. Brasch and Dr. Joseph Sigman of McMaster University, available through Boston’s John F. Kennedy Library, here.

Please let us know of any errors you find – in adding over 7400 books, there are bound to be one or two.

Finally, many thanks to nperrin for getting us started – unfortunately, she had to move across country soon after, and only had time to enter about 100 books. This doesn’t lessen her contribution in any way.

Absolutely fantastic.

Wonderful. Such a great effort, well done indeed!

Aww, thanks for all your help you guys, you were SO dedicated!!

And personally, I think this is the coolest legacy library so far. ;)

I'm going to get in on some of the tagging, let me know if I should know anything besides what's evident already from the catalogue. Thanks again!!!

Awesome. Yay for being done! ^^

Great! Thanks, everyone, and especially nperrin and jjlong, for organizing this.

I've just posted on the LT blog about the completion of EH's collection - if I screwed anything up or forgot anyone PLEASE let me know and I'll fix it!

When I compare my library with Hemingway's, I see that there are multiples of a lot of books in his library. Pride & Prejudice shows up twice, Vanity Fair three times, etc. Does this mean that Hemingway owned multiple copies, or is this a display error? I also see some of this duplication on other shared libraries. This is when I access the info through the Statistics-->Legacy Libraries link on my profile page.

Thanks to all who worked on this huge project! This is such a cool feature!

>8 fannyprice: Hemingway owned several multiples of books--yes, he owned two copies of Pride and Prejudice, etc. I can't say for sure about the other libraries, but I think that is true of them as well.

Indeed - where duplication shows up (at least in all the instances I've noticed) it's because the person owned multiple copies of the work (usually but not always different editions, sometimes different languages, &c.). Jefferson, for example, had at least six editions of The Iliad, three of Plato's Republic (French, Latin and English), and so on.

This reminds me that there's an inconsistency in the "shared" numbers - for example, Hemingway's profile page shows that he shares 68 books with christiguc, but her statistics/Legacy Library page shows that she shares 83 (showing EH's multiple copies each as "shared").

A bug, of sorts... it'd be nice to have the two numbers match. Maybe it should be the lower number, and read "shared titles."

Yeah, I'd noticed that too - I don't like that the numbers don't match, but I do sort of like that the statistics page shows the different editions/copies owned by the Legacy-subjects ...

I posted this comment on the Hemingway page, but maybe no one is checking there:

I notice that on April 1 you entered six copies of Cyril Connolly's "The Unquiet Grave", all apparently the same edition. If it's the case that Hemingway's library contains six identical copies of the book, would you please consider making only one entry, and indicating that there are multiple copies of it? I think there's an edit slot for number of copies.

It would decrease the bloat and make stats more meaningful.

>It would decrease the bloat and make stats more meaningful.

?

Do you mean this in terms of "books shared"? I can see that it's frustrating to think you share 6 books with Hemingway, only to find it's one book 6 times, but I don't see that it's a wise policy not to enter multiples, since it throws off the total volume count of the library and misrepresents the actual volumes held.

I can't find the thread, but we settled this once, ages ago, by saying that if there are multiple copies of different editions (or where the editions are unknown), each should be added, but if there are multiple copies of a specific, known edition, that should just be noted in Comments (i.e. "6 copies").

This does not apply, of course, if the multiple copies of a particular edition (a pamphlet, for example) are bound in different volumes (we see this in Franklin's library or the Mather's, for example).

That said, I'm not too worried at all about cases where there are a few copies of something (i.e. less than ten or so) and they've been entered separately. I don't think it throws off statistics (in either direction) in any serious way.

I don't see that it's a wise policy not to enter multiples, since it throws off the total volume count of the library and misrepresents the actual volumes held.

It's HOW one enters multiples. There is the "number of copies" entry on the edit page. Aren't these counted toward total book count? Why not use this when it's included precisely for situations in which there are multiple identical copies?

I don't think it throws off statistics (in either direction) in any serious way.

But it does. The fewer total copies on LT, the more this kind of entry messes up meaningful connections.

And, while I'm complaining, I think the first-person voice in dead person profiles is creepy and silly. It's problematic enough that we don't know whether they'd care to have their libraries up for public perusal; let's not make it worse by imagining what they'd say about themselves.

> 17

There is the "number of copies" entry on the edit page. Aren't these counted toward total book count?

No. It's basically just a note to yourself, and doesn't show up in your statistics at all. When we discovered that, we decided to enter each book separately; otherwise, the total volume count for the library would be quite a bit low. (I remember another comment: "if he had six copies on his shelf at home, why not on his virtual shelf?")

In an ideal world, this slot on the edit page might contribute to stats, and the work page for The Unquiet Grave might read "Members: 149; total copies: 161".

We've not taken a heavy hand with the profile texts, as long as they contain the necessary elements to be granted Legacy status. Those have been left up to the individuals cataloging the libraries.

About profiles - and I can only speak for Hemingway's - I'm sure we followed Jefferson's lead; his was originally in first-person (though I see it isn't now, which makes sense, given the academic turn the older libraries have taken). Also, first person follows naturally from the headings "About me" and "About my library"; I always thought it might make legacy libraries a bit less intimidating to the casual browser (and, Ms W, I know you're anything but a casual browser... nice to know someone's looking through Hemingway's library).

Any other opinions on profiles like this? Creepy, or user-friendly? I can see it either way.

Yeah, I can see it either way too. We did change Jefferson's (and I think I've changed a few of the other ones I did at the beginning too) but I leave it up to the catalogers of each collection to decide how they want to phrase the profile texts.

the total volume count for the library would be quite a bit low.

I'm puzzled why anyone would care. Hemingway's catalogue now lists over 7K books--are we to understand that so many of these are identical multiples that without the redundancy this number would be "quite low"?

(I remember another comment: "if he had six copies on his shelf at home, why not on his virtual shelf?")

Not to nitpick, but maybe he kept the spares in the cellar, in a shoebox under the bed, or never knew they existed.

What if he'd had 100 copies, or heck, even only20? You'd have entered them all, separately, one by one?

Seems to me it would be more reasonable to indicate multiple copies in the edit slot provided for that, and then explain (if one so desires) on the profile that the actual volume count is higher, and what it is.

Let's try to think about what's most interesting AND most useful, when it comes to interactions between dead and live libraries on LT. Right now, in this instance (but understand this is only an example, I think the principle is important), we have dead Hemingway inflating the number of "The Unquiet Grave" copies, with a total under 200 (IIRC), and thereby making everyone else's connection to this work and other LTers who have it that much less significant. What's the point?

Interesting, isn't this?

are we to understand that so many of these are identical multiples...: I'd guess 300 or so; see here. It's not just The Unquiet Grave, for sure.

maybe he kept the spares in the cellar, in a shoebox...: or (just as possible) he was so taken with it that he ordered 50 copies, and urged them upon dinner guests or drinking companions he found intriguing? (I wish we knew.)

Are you just concerned with The Unquiet Grave? Or with Hemingway? Or do you plan to issue a site-wide appeal for all users to fix their data (say, this user, or this one? Is two duplicates okay, but 6 not? Would you recommend the site code be rewritten to prevent separate entry of duplicates?

In the end..

1. you're right. But I'm right, too. And Tim's wrong, for not making that "number of copies" entry factor in. Hi, Tim.

2. on the practical side, I spent most of the summer tagging and adding correct covers for Hemingway's library, and I can't imagine going back now, through this entire duplicate list, deleting extraneous copies, adding a comment to each, and adding a number-of-copies entry to each. I'm about Hemingwayed out.

ETA: Rullakartiina was a great help with the tagging. Sorry I took so much credit.

I see the point, but I really don't think it's a big deal. If it were 50 copies, that would be a little strange, but I'm not at all concerned about it as it is.

I'd guess 300 or so; see here. It's not just The Unquiet Grave, for sure.

And no one was concerned about this absurdity, how stupid and ugly it looks? "Hemingway" was racing "Sandburg" to some idiotic primacy?

I'm concerned about wherever this has been happening in Legacy libraries, because the dead have no say in how they'd like their libraries presented, what the multiple copies meant and so on. The responsibility with entering their catalogues in a reasonable fashion lies with us. Had I noticed this situation with a living LTer, I may or may not have felt impelled to ask them what's going on; the point is that I noticed it with Hemingway, who's dead, and ISN'T reponsible for the state of his catalogue on LT.

I spent most of the summer tagging and adding correct covers for Hemingway's library, and I can't imagine going back now,

If you let me have the password, I'll delete the extra copies of "The Unquiet Grave" (and add a note about it).

Anyway, I think there ought to be a general guideline for multiple copies that all Legacy Libraries can look to, so situations such as these do not occur.

If it were 50 copies, that would be a little strange, but I'm not at all concerned about it as it is.

Well, Jeremy, where do you set the limit, and why?

#22

Looking fleetingly at Hemingway's multiples list, there seems to be only a few examples such as "The Unquiet Grave", but there are quite a few bad combinations giving false multiples.

These libraries are entered using inventories or printed bibliographies. That for EH (here, from the JFK Library) specifically includes seven copies of the 1957 edition of The Unquiet Grave (plus a copy of the 1945 edition).

It's true, we don't know why the duplicates are included. But they are included in the list, and those who cataloged the library did so perfectly appropriately. I'm sorry you think it looks "stupid and ugly."

#25 - as I said above, up in #15, I think any more than ten would be a little strange, unless there were some particular reason. And even with less than ten, either entering multiple copies or making a note seems fine to me. It's really not skewing statistics in any appreciable way.

#27

And I'm sorry you take such a robotic attitude to LT. LT isn't a mirror image of physical libraries, nor is there any reason why it should be. The virtual libraries are interactive webs, and we ought to take care to make the interactions meaningful. Especially if we're taking these responsibilites for someone else's catalogues.

But they are included in the list, and those who cataloged the library did so perfectly appropriately.

No, they did not do so "perfectly appropriately", because the edit page in LT catalogues offers the perfectly appropriate copy number slot, and it was not used.

Why wasn't this slot used?

It's really not skewing statistics in any appreciable way.

Depends on the case, and I don't want to have this conversation every time there's "a case".

We need a guideline.

I suggest no true duplicates be allowed--not even two.

And I'll suggest that the guideline be this (for duplicates only of exactly the same known edition):

If you're using a printed bibliography/catalog/inventory, follow the inventory. If it lists multiple copies as separate entries, do the same (preferably including the entry number in Comments). If it lists one entry with multiple copies within, do that (in Comments or in Copy Number).

Do this within reason. If there are specific cases where a bibliography lists what seems an inordinate number of duplicate entries, even if numbered (I've suggested that the cutoff point be ten), consider making one entry and making a note in Comments about the number of dups.

> 26

there are quite a few bad combinations giving false multiples...

I do see this one:

Attack: a study of blitzkrieg tactics by Ferdinand Otto Miksche (New York: Random House, 1942.)
Paratroupes; l'histoire, l'organisation et l'emploi tactique des forces aéroportées by Ferdinand Otto Miksche (Paris: Payot, 1946.)

Thanks.

I can't find any more, much less "quite a few". If you mean, say, Marie Belloc Lowndes' The House by the Sea and Vanderlyn's Adventure, the latter is the American title of the former. As for the Yeats and the Byron, I'll defer to the combiners' group.

Speaking of statistics, these are all examples where Hemingway's multiples skew numbers "considerably", representing anywhere from 75% to almost 10% of copies entered:

http://www.librarything.com/work/4673468/25359406

http://www.librarything.com/work/2567618/25400389

http://www.librarything.com/work/2567618/25400389

http://www.librarything.com/work/1403443/25915281

As for "The Unquiet Grave", currently 157 total copies, Hemingway's catalogue lists 8 copies, of which 7 are identical. We can debate how much this skews the numbers, but there's no question that it does.

Identical multiples are noise. We can't control noise everywhere, but why not control it where we can? That's all.

#31

Well, I disagree with your solution, but I hope the discussion at least made others aware of the problem.

#32

Examples of bad combination--some almanacs, apparently identically titled, but with different publication years

http://www.librarything.com/work/4665887/25308008

poetry collections

http://www.librarything.com/work/124524/25336046

travel guides

http://www.librarything.com/work/4706546/25575605

And there may be more, I was just skimming. Anyway, just remarking.

Combinations are tricky, since we don't really control them ... they can be made and unmade by anybody, or automatically. I try to go back after about every twenty titles or so and work on combinations, but between caching delays and other issues, sometimes those don't take or can't be made right then. If anyone's got free time and wants to go through Legacies and work on combinations, that would be a great thing!

I'm not so eager to split Byron's poems apart because they were published on different paper and with different typefaces. Someone who has all of Byron has all of Byron. Now, often, it's hard to tell just what's in those collections, but still...

Travel guides is an interesting one. The Let's Go from 2007 and 2008 are not far enough apart to warrant the social split LT talks about. I mean, you're at a party and discover than someone went to Turkey last year. You went the year before. You fall into a conversation about travel guides and she asks you "Man, doesn't the Let's Go suck!" Well, yes, it does, and it did the next year too...

But then, over years and then decades, the guide does change. Well, there's no solution to it, I think. We need a rule, but we shouldn't fool ourselves that it's not arbitrary.

Someone who has all of Byron has all of Byron.

One volume is titled "Complete poetical works", the other is unspecified. We were talking about Hemingway's duplicates, not the entire combined work. His two volumes could get combined with two different works.

As for the travel guide I linked, Hemingway's 8 copies are the only ones combined into this work, and years from 1948 to 1961 are combined together, under the now misleading title "Fielding's travel guide to Europe: 1957-1958".

http://www.librarything.com/work/4706546/members

Another case with his differently dated Michelin France guides, 1929 to 1954 (Hemingway is responsible for 5 out of 17 total copies, as this work is combined now--oh, and the work contains guides from as late as 2005):

http://www.librarything.com/work/199375/members

But bad combinations are a minor point, and we have Combiners to discuss them there.

Here's more on what's truly bothersome:

more absurdity with identical duplicates, 11 Ernest Hemingway entries for this book--and ONLY his entries:

http://www.librarything.com/work/4763687/members

One is in French, the 10 remaining ones are apparently identical Spanish editions.

For this one, 10 total copies on LT, of which 4 are Hemingway's identical listings:

http://www.librarything.com/work/1403443/members

Just a note, out of the 102 books Hemingway shares with T. E. Lawrence, there are:
3 copies of Charles Baudelaire’s “Les fleurs du mal”
2 copies of Miguel de Cervantes’ “Don Quixote of la Mancha”
2 copies of David Garnett’s “The grasshoppers come”
2 copies of George Moore’s “Avowals”
2 copies of Ezra Pound’s “Personae”
and 2 of Pound's “Selected Poems”

Which is only an inflation of about 7 books, not a particularly huge number all things considered, though I guess if these were the ONLY books they shared, it would be.

Just to use an example of how it might inflate the numbers since its been brought up a lot.

EDIT: Sorry, this is Kordo, I forgot to logout after working :(

Hemingway's three Baudelaires are all different editions of "Les fleurs du mal", not identical multiples.

ETA: If you look at "Multiple Works" under Statistics, you'll see the multiples with some identifying info appended, clarifying whether they are identical multiples or different editions.

I believe, though I'm not searching for old threads to find out, that I may be the culprit at the root of this. With such a massive listing of books, I remember being rather concerned that without entering the copies separately it would be too nearly impossible to make sure we had completed the catalog (since the multiple copies field doesn't affect the library total). Now that everything has been entered, however, I would not object to anyone going through and fixing identical edition multiples via the number of copies field. I do agree that it would help site-wide stats, an obsession of mine (though in a minor way--vaporware and dormant user bloat makes this obsession pretty fruitless in general).

So, Tim, a plea to make the number of copies field count toward the total. Not least because it just makes sense.

Also, I would echo post 19 re: the wording of the profile page. And I wouldn't have any kind of objection to amending that either.

ETA: I'm not going to look through all these right now, but are any of them instances where the location of the volume differs? E.g., one copy is in Key West, another in Cuba?

"LT isn't a mirror image of physical libraries, nor is there any reason why it should be." -post 29

Whoa. I thought the point of this was to replicate libraries as they existed at some point in time. I will only think of my LT library as complete when it mirrors the physical. If Hem had 7 identical copies, this should be recorded. We may not know why he had them, but it could prove interesting in the future, and could provide a point of departure for research...

blc

Absolutely identical copies can be marked at the bottom of the edit page—number of copies.

42: But as mentioned in previous messages, the number of copies field does not affect the total book count for the library, which is exactly why that method was not used for Hemingway. Can this be changed?

Ah. Sorry. Frankly, no—not now. The counting currently relies on counting from an index of book ids without having to scan the table itself. It's using a piece of data used everywhere on the site. Redoing it to count those would slow down something critical. It might only be a 10% slowdown, but it's not worth it for this.

Okay. Well as I said in post 40, I think this is not a big deal for Hemingway now that the library has been completed. But it would make me hesitate to use it for my own library if I did have multiples, not that that's a big deal either. Thanks.

While reading Hemingway on Writing, there are many references to specific books, many of these references could count as short reviews. Editor Larry Phillips combed Hem's correspondence, published articles, etc. to compile this book. For instance, in a letter to Edmund Wilson in 1923: "E.E. Cummings' Enormous Room was the best book published last year that I read." There is quite a bit along those lines. Could these "reviews" be included in Hemingway's library?

>Could these "reviews" be included in Hemingway's library?

I think that would be great!

I'd be happy to do them, if sent the particulars to Hem's library.

Sent. Thanks... this will be a great addition!

Done with books listed in library. However, gathering these, there are a few books that are not in the library, but do get a mention in letters, etc. For instance:

"In the meantime, since it is Christmas, if you want to read a book by a man who knows exactly what he is writing about and has written it marvelously well, read Appointment in Samarra by John O'Hara."
By-Line: Ernest Hemingway, pg. 184

PS, this is benjclark

So, what should be done about these "missing" books? There are 8 books explicitly stated (like Appointment in Samarra) that are not in his library. I left out (nearly) all general references to writers. He calls John Dos Passos all kinds of names, and others without referencing specific books. I left these out (mostly). Hemingway really did think a lof of Stendhal. There is also an exchange in one of his articles where he rattles of 30+ titles a writer "should have read... so he knows what he has to beat." Many of these titles are mentioned already in his "reviews", but some are not. I've not included these, as he disparages on some in detail later, so this list is not really an endorsement of quality, etc. Also, some of the recommendations on this list are vague: "all the good De Maupassant, all the good Kipling", etc.

I guess the question is that, though he mentioned reading those books, is there any proof that he owned them at any point? If not, I guess I would argue that you shouldn't add them as LibraryThing is for what books you own, not what book's you've read.

as LibraryThing is for what books you own, not what book's you've read

LibraryThing is for cataloging books, according to almost whatever criteria you want.

However, Hemingway's library reflects the books he is recorded as owning, so I would say that those books shouldn't be added. Since people can't review books without adding them to their library, I guess the reviews should just be left out, in my opinion.

>LibraryThing is for cataloging books, according to almost whatever criteria you want.

That's true for personal accounts, but we've been trying to enforce some consistency where Legacy Libraries are concerned. (At the very least, books read but not necessarily owned need to be clearly marked as such.)

I think the universal rule of Legacy Libraries should be that, just clarity. You can record books owned, read or fired out of cannons for all anyone cares. But they should be carefully marked for what they are, and if there is any chance of confusion, information on the profile that clarifies what the categories mean.

Yes, exactly. As long as it's made very clear, preferably both on the profile front page and also in the Comments field for relevant books, read or owned will work fine.

I think it's a tough call. While I would like to include them with an unowned tag or something, we know Hemingway must have read a lot of other stuff too that wouldn't be catalogued. We have a complete listing of his possessions; we couldn't possibly get a complete reading list.

What about tagging them NIL? Would it be possible to post all of the "reviews" together, like on a wiki on Hemingway's profile page?

Also, ran across five Hemingway books searching other special collections.

One book in Delaware wound up in his bibliographer's papers. One bookseller records 3 books. Well, one certainly, one was certainly presented to EH, but doubtful he used it, one was his Father's but certainly of interes. There is one book recorded in the library, but was sold at PBA auctions in '03, (w/ great photo!), information that could be useful.

Can I send these to someone who helped with Hemingway's cataloging, so they are handled uniformly with the other books?

"NIL." If you're going to use a tag, I'd suggest "not in library" not an abbreviation of it.

Would something else be better? I use NIL *sparingly* in my own library. I don't like it, but since collections.... anyway, I'd love to hear some suggestions about these books. I think they should be included, since he actually admitted to reading them. Who knows, this may be a useful thing to track down copies that may have gone into the market some time.

> 58 Nice work on finding the five other books... Drs Brasch & Sigman did try to ferret out stray copies, so I'm wondering if you've checked these to see if they did find these five, or not...

As for the review of say, Appointment in Samarra, I'm like nperrin; do we want to change the catalog from "books Hemingway can be proved to have owned" to "anything he ever read"? I'd be inclined to create a wiki page, a la Marie Antoinette's, to post these.

Nothing's set in stone, though, is it.

61: Actually, the wiki page is a good idea. We could easily link to it from Hemingway's profile, something like "Check out this list of books Hemingway read and commented on, but did not own at the time of his death."

I think the wiki idea is a very good one. If it's incredibly important that the books be in the library, a tag plus a note in the Comments field probably wouldn't hurt (and I agree with Tim, use the whole tag, not an abbreviation).

>61 jjlong:

I think that's a bad idea to put the books on the wiki instead of in the catalog, because you lose all of the interactive features of the catalog. (I.e., his reviews won't show up on the work page, he won't be linked as a legacy author from the work page, it won't be factored into his library stats, it won't show up as a book "in common" on each individual users' stats pages.)

Truthfully, I've never understood the supposed purity of cataloging only books owned and not read-but-not-owned. I understand wanting to put only good data into LT, generally, and into the Legacy Libraries project above all, but, to apply the ever-cited "cocktail test" do you genuinely think you have more in common with someone who owns a physical copy of a book (but may or may not have read it) than you do with someone who you know has definitely read it and even written a review of it?

Yeah, I agree with Katya*. If a madman burned down Hemmingway's library would we put it all on the wiki because it's no longer around? Granting that the distinction between statuses is useful, interesting and important, I see no reason to exile Hemmingway's read-but-not-owned-at-death books. If that's the decision, let's start by cleaning out Jefferson's LoC gift—he "read-but-not-owned"-ed those too. Also, he doesn't own anything, since he's dead.

*On nearly everything. You and Jeremy need to meet sometime. The combined bibliographic excellence might cause an earthquake.

The key, as we've said above, is just making sure that it's obvious even to the most casual searcher/browser where the information is coming from. I do agree that having everything in one place is ideal (and more useful), but being a purist myself, I understand the dilemma.

For most of the Legacy collections, of course, we haven't any idea whether the owners read the books or not (except in the surprisingly-frequent cases where Jefferson writes to an author and thanks them for the book but says he'll probably never read it, haha).

I hate to throw out another "it depends," here, but I think it will depend on the catalogers working on each library. If people are just working from a single list, they can input those. If additions are found later, by them or by others, I'm inclined to add them, since it increases the usefulness of the catalog (just as I'm inclined to and really enjoy adding information about copies which have managed to survive, like the ten or so I was able to find from Priestley's collection).

For Hemingway's specific case, it may be enough to add a note in the "About my library" section noting that the majority of the books are from the JFK Library list, but that some additional books Hemingway read and commented on have been added based on whatever the source is you're using (presuming it's a reputable source). Then the books could be added to the library (with suitable qualifications, like not knowing the edition, him not owning it, &c.).

That suggestion made, though, I'm not going to set that route in stone - it's up to the Hemingwayers.

>The key, as we've said above, is just making sure that it's obvious even to the most casual searcher/browser where the information is coming from.

I had this issue with the Joseph Smith library, since the books came from a number of sources. In that case, I tagged every book with the source, then explained the sources in the profile. (Of course, all of those were owned, so I didn't need to clarify that in the catalog.) What about tagging the books with the source and with something like "read but not owned"?

Well I'm not going to set it in stone either, but I think there's a pretty clear quality difference between what's going to come from the JFK Library list and what's going to come from elsewhere. In one case, we know the exact edition of the book owned, which house it was kept in, whether it belonged to Hemingway or his wife, and, for many of them, where he ordered them. In another case we know that he read and commented on them, but we don't know any of the bibliographic data. My take on it is: without bibliographic data, why bother cataloging it? Without any hope of ever getting an exhausted list of things he read and didn't own, why bother cataloging them? We have the ability to say, "This is a complete catalog of books Hemingway owned on his death." But instead we will be saying, "This is a complete catalog of books Hemingway owned on his death, plus some other stuff that he just happened to mention somewhere."

Obviously I am a purist, even though I have given in and put read but not owned books in my own library. And it wouldn't bother me to put these in for Hemingway, especially since it seems like, maybe, maybe, maybe soon we would get to segregate them with a collection. But I also don't see the point, when we have such a clean and complete list (of a particular type) now.

>67 Katya0133:

Are we including the golden plates in Smith's library? Read but unowned...

>Are we including the golden plates in Smith's library? Read but unowned...

Ha! No, I think not, although I did once try to write up a fake MARC record for them.

Also off subject, but we did a record for a Horcrux in our museum collections database. I suggested it be classified as Immortality, Tools and Equipment.

Ok, to bring it back...
I think of myself as an LT purist too. No "read-not-owned" in my library. However, these books for Hemingway's library are very small in number, are written mostly by folks in his circle, or on subjects that influenced his writing. In short, anyone interested in Hemingway will probably like having this information. We're talking less than 10 books in a library of over 7000!

"But we don't know any of the bibliographic data. My take on it is: without bibliographic data, why bother cataloging it?" - nperrin

But that is a big part of the fun! Why not include what we know so perhaps we'll be able to hunt up the answers. It takes the stumper out of the cataloger's hands and into the hands of anyone with an interest.

For instance, the Gertrude Stein book he wrote about in '26 was only available in one of two editions. One from NY, the other in Paris, where he was living in 1925 when the book came out. Taking it to this point allows the next step in the research to happen.

>In short, anyone interested in Hemingway will probably like having this information.

I think that's the most succinct explanation for why I'd like to include this information. I get that it's more tidy and neatly defined if we only include the books physically present in his catalog, and I certainly don't want to see Legacy Libraries which consist of "books X owned, plus books they read, plus books they wrote, plus books I think they probably read . . . etc."

But I've always seen these libraries as a means, not as an end of themselves, and if the point is to give us a window into their literary consumption, then books that they read are certainly as much (or more) a part of that as the books they physically owned. In fact, I'd say that the books they read are probably a better indicator of who they were as readers than the books they owned, it's just that it's usually easier to reconstruct the physical ownership, so we default to that.

As far as not being able to reconstruct bibliographic data goes, the beauty of the works system is that we can link their reviews, etc. to the larger LT catalog without needing to know the specific bibliographic details. (And I think a library of an interesting person is still worth reconstructing as much as possible, even if some bibliographic information is lacking.)

Ok. The "missing" books are now added. Sources are noted with the "review". Editions are "bests". Some are pretty good, others doubtful. One book, published in 1923 only had two editions by the time of the 1925 quote, one in English the other edition in Czech. I've already mentioned the Stein book. The Flaubert books, he gave the French titles in his quote, so I went with French editions close to the time he was living in Paris. As close as I could find. I've tagged these "missing" books as "not in library" and all of these reviewed books as "reviewed". None of these are sacred, so blaze away.

Oh, also, I still think it would be nice to have all of these quotes on one page perhaps linked from EH's profile page. Especially if it was a wiki so others could add to if new quotes are found. I have no idea how to do this, but I'm happy to send along my compilation.

On the editions - please don't enter bad data - if you don't know the specific edition, don't put any data in Publication or Date, and make a note in Comments that the specific edition is unknown (you can note there that there are a couple of possibilities, or whatever).

Have now removed edition info and replaced with "Edition Unknown" and then put the possibles in the comments section. Have also updated the profile with a short note about the source of the reviews, etc.

Thanks :-)

My pleasure. I've got some more goodies in the pipeline.

A fun post from the Ransom Center on a book owned by Hemingway:

http://ransomcenter.tumblr.com/post/59686482977/though-the-story-may-sound-like-...