Hi grendelkhan. Ok, I've put off writing this message for a couple of days, hoping to get my thoughts in order. They don't seem to be there yet, but I hope this is still helpful. Yes, what you say is true. And the most likely way to do this is some form of "copy-cataloging", which, for our present situation would most likely take the form of manually copying subject headings as found in the LoC. From experience, I would suggest that simply trying to keep up with ~50 titles a week, much less deal with the enormous back-log, would be enough to induce burn-out and frustration very quickly. I believe it is something of a misconception that aspects of cataloging (such as subject headings) are merely a matter of "paint-by-numbers" or "fill in the blanks". Every individual collection has its own oddities, and (assuming a well-developed catalog) its own local practices. Wallace's suggestion below, applied to subject headings, would not really be workable, as different libraries do not use the same set of subject headings. (For instance the National Library of Canada defines "Canadian Subject Headings", which are similar to LCSH but more focused on Canadian topics. So then, non-library-sciences people will ask "why not just use whatever looks good from whatever library you find it?" Then the issue of "controlled vocabulary" comes up. If you do not use a controlled vocabulary, you will soon reach a point where what you have is more like freely choosen "key words", rather than real subject headings. For more, you could look at: http://en.wikipedia.org/wiki/Library_of_Congress_Subject_Headings The biggest thing I would like to avoid would be having something that gives the impression of using LCSH, but is not. Library of Congress Classification is a different creature altogether. Keep in mind that it was designed to help with problems of organizing and shelving books in one particular library about 100 years ago. http://en.wikipedia.org/wiki/Library_of_Congress_Classification This can lead to (For example, many books of Canadian literature from the time period PG mostly works with are classified by the LoC as British, rather than in the small subset of American liturature that is set aside for Canadian.) Using the first two letters, as has been done for the first few thousand PG books, results in extremely broad catagories. This is better than nothing, but I can't help feeling that something that can respond better to our interests would be ideal. So, what do I mean by that? Some form of social cataloging, or folksonomy, or whatever buzzword you want to use. Yes, this has weaknesses as well, but given the nature of PG and the volunteers contributing to it, I think perhaps a better fit. Andrew On Wed, 23 Aug 2006, grendelkhan wrote:

...

From: "Wallace J.McLean" <ag737@freenet.carleton.ca> To: gutvol-d@lists.pglaf.org Date: Wed, 16 Aug 2006 15:04:55 -0400 Subject: Re: [gutvol-d] Re: Search by subject for ebooks?

Which any decent library already takes into account when cataloguing, which is why it would be easiest to catalogue under any and all pre- existing catalogue headings as given in major national library catalogues like LOC, AMICUS, British libraries, etc., etc., etc., unless, of course, there is a patently wrong catalogue entry. (That happens.)

If I recall (Andrew, please correct me if I'm wrong), there's was previously a cataloguer who was putting in that sort of thing, see (random example)

http://www.gutenberg.org/etext/2852

which is filed under "Holmes, Sherlock (Fictitious character) -- Fiction" as well as "Private investigators -- England -- Fiction", and its LCC class (PR, English literature). Andrew said that this hasn't been done in a while, but if there's interest, I'm going to be entering a Library Science program this fall, which I reckon would make me an incipient professional librarian--I'd be quite interested in applying cataloguing standards to the PG catalog.