From: Sarah slimvirgin@gmail.com Subject: Re: [WikiEN-l] The boundaries of OR (contd) To: "English Wikipedia" wikien-l@wikipedia.org Message-ID: 4cc603b0612212250s7013522en8eb03e740941dcee@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1; format=flowed
On 12/21/06, Thomas Dalton thomas.dalton@gmail.com wrote:
No, you absolutely cannot do that, for reasons eloquently stated elsewhere. The claim that it is not in the Cornell University
Library
is a novel conclusion based on your own original research; this
seems
so trivially obvious to me that it astonishes me that others
would
claim otherwise. You might as well promote a novel claim in
physics,
and point people to the calculations you have made to prove your theory. If a reliable source says "the book is not found in the Cornell University Library", then quote them. Otherwise, move on.
What's novel about "It doesn't appear on the such results,
therefore
it isn't in the library"? Seems like an obvious conclusion to me... (might not be 100% reliable, depending on search terms, accuracy of the library's index, etc, but that doesn't make it novel).
You may have made a mistake. Or it may be in the library but not in the catalogue, or vice versa. That you conducted the search yourself makes publication of your results OR. We publish the mistakes of reliable sources, not of Wikipedians. :-)
The fact that mistakes might happen doesn't make it OR. You have to show how looking up a library catalogue is fundamentally different from looking up something in a book. I don't think you have done that. I think that an error in a library catalogue is not really a different problem from an error in a book, and an error in consulting the library catalogue is no different from an error in consulting the book.
To me, these are logically indistinguishable: 1. "Joe Blow died in 1923. [Franky: Joe Blow, A worthless life]" 2. "Library of Congress has the 1923 edition. [LofC catalogue]"
I think these negative assertions are indistinguishable too: 1. "Franky's biography of Joe Blow doesn't say when he died [Franky: Joe Blow, A worthless life]" 2. "Library of Congress doesn't have the 1923 edition. [LofC catalogue]"
Zero.
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
On Fri, 22 Dec 2006, zero 0000 wrote:
I think these negative assertions are indistinguishable too:
- "Franky's biography of Joe Blow doesn't say when he died [Franky: Joe Blow, A worthless life]"
- "Library of Congress doesn't have the 1923 edition. [LofC catalogue]"
Be careful with this argument, because someone could say the Library of Congress catalog is a self-published, non-peer-reviewed, source. (On the other hand, it's maintained by librarians who would get fired if they kept making mistakes, so in that sense it's not self-published, and it makes a non-contentious claim about the subject, so it should be allowed anyway.)
On 12/22/06, zero 0000 nought_0000@yahoo.com wrote:
From: Sarah slimvirgin@gmail.com Subject: Re: [WikiEN-l] The boundaries of OR (contd) To: "English Wikipedia" wikien-l@wikipedia.org Message-ID: 4cc603b0612212250s7013522en8eb03e740941dcee@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1; format=flowed
On 12/21/06, Thomas Dalton thomas.dalton@gmail.com wrote:
No, you absolutely cannot do that, for reasons eloquently stated elsewhere. The claim that it is not in the Cornell University
Library
is a novel conclusion based on your own original research; this
seems
so trivially obvious to me that it astonishes me that others
would
claim otherwise. You might as well promote a novel claim in
physics,
and point people to the calculations you have made to prove your theory. If a reliable source says "the book is not found in the Cornell University Library", then quote them. Otherwise, move on.
What's novel about "It doesn't appear on the such results,
therefore
it isn't in the library"? Seems like an obvious conclusion to me... (might not be 100% reliable, depending on search terms, accuracy of the library's index, etc, but that doesn't make it novel).
You may have made a mistake. Or it may be in the library but not in the catalogue, or vice versa. That you conducted the search yourself makes publication of your results OR. We publish the mistakes of reliable sources, not of Wikipedians. :-)
The fact that mistakes might happen doesn't make it OR. You have to show how looking up a library catalogue is fundamentally different from looking up something in a book. I don't think you have done that. I think that an error in a library catalogue is not really a different problem from an error in a book, and an error in consulting the library catalogue is no different from an error in consulting the book.
Anyone can find the book you have quoted, and look on the page, and see if it says what you claim it says. The best footnotes actually quote the section in question, so the confirmation is easy, and the book is published, we know the edition and publication date, and the contents of that specific publication are not going to change. However, we can't easily reproduce the search you Zero0000 has done on the library catalogue, and even if we could, there's no guarantee that the catalogue will have the same contents when *we* search it. On the contrary, one can guarantee that the contents *will* have changed.
Databases, library catalogues, and similar ephemera are complex beasts that have little in common with printed works.
Jay.
On 12/22/06, zero 0000 nought_0000@yahoo.com wrote:
<snip>
The fact that mistakes might happen doesn't make it OR. You have to show how looking up a library catalogue is fundamentally different from looking up something in a book. I don't think you have done that. I think that an error in a library catalogue is not really a different problem from an error in a book, and an error in consulting the library catalogue is no different from an error in consulting the book.
To me, these are logically indistinguishable:
- "Joe Blow died in 1923. [Franky: Joe Blow, A worthless life]"
- "Library of Congress has the 1923 edition. [LofC catalogue]
And for both these cases, cite as finely as you can: down to the page number of the exact edition for the book, and with the LoC card number (it'll be found in the full record, next to the ISBN if one exists) for the other.
I think these negative assertions are indistinguishable too:
- "Franky's biography of Joe Blow doesn't say when he died [Franky: Joe Blow, A worthless life]"
- "Library of Congress doesn't have the 1923 edition. [LofC catalogue]"
There is one fundamental difference: unless you're writing *about* the biography or the LoC itself, why on earth would you include such a statement in an article? IMHO, the absence of information in the few sources you may have consulted doesn't constitute a useful statement for a global encyclopedia. (And the fact that your sources were few is not a negative judgement on your research skills; it's simply a fact of life -- it's likely *none* of us will be able to consult all the potential useful sources of information on any given topic unless the topic is so small as to be basically NN in its own right. Hence the usefulness of collaboration).
As far as "the ultimate truthfulness" of these two statements: you're totally right, they both depend on your skills as a researcher. For the book statement, it depends on your reading comprehension abilities and ability to search an index/read the complete book. For the LoC statement, it depends on your ability to search. In general, we trust people's abilities to read, and we may be able to trust their abilities to search, as well, but there seems to be some confusion about what a library catalog is.
A catalog is not like a book (format aside) because it is a continually updated series collection of collated and indexed metadata about other works, whereas a book (except for other collections of information or printed databases, like the phonebook) is usually a narrative about some particular topic. A library catalog serves three major purposes: discovery (what are some books about serial killers?); identification (what is the name of the book about serial killers written by Brown?); and location (does the LoC have the book by Brown about serial killers, and if so, where is it?) Mistakes can enter at any of these stages, information can come from a variety of places at any of these stages, and you have to be entirely clear that you've got the right edition at stages 2 & 3. That's not to say that for your statement (LoC doesn't have the 1923 edition) I wouldn't take your word for it -- modern catalogs have been designed so that it's easy to find things, just like it's easy to find web pages in the catalog of them we affectionately call Google (though I think we all know how hard it is to track down one rare snippet of information on the web, and try finding a rare non-English title in LoC sometime).
However! There is one major caveat: not everything in a library is cataloged and present in their online catalog. Not even (especially?) in the LoC. Every library I have *ever* run across in my short career has a cataloging backlog of some sort. Cataloging is hard and tedious and manually done by trained professionals who take a lot of care in making sure they have the right book, which is why catalogs are generally accurate, and irregardless of this particular question or references in Wikipedia that's worth remembered if one is trying to be comprehensive in one's researches :) And, as Jayig and I have pointed out, holdings change, and sadly there is no "permanent link" or "compare diffs" in most library catalogs; the best you can do is cite a date. *No* library catalog will claim that "this is a full and accurate record of everything we have," and if they do, they're either very small or lying. -- phoebe
p.s. the LoC catalog *is* peer-reviewed, because their records are downloaded into worldcat and may be checked by many other people. Mistakes are reported and fixed :)