Physical bookshelves may be described as a fixed partitioned set S = {S1,S2, ... , Sn} where the Si are the fixed physical shelves defined by their length and height.
A collection of books may likewise be described as B = {B1,B2, ... , Bm} where now the Bj are components of a partition of the books into categories; fiction, poetry, psychogeography, chess etc. This partition is poorly defined as a given book may belong to more than one category: so, does Offa and the Mercian Wars belong to biography, or geography, or history, or politics? And as for Pond Life. The whole scheme is ad hoc and pays no heed to Melvil Dewey at all. That's probably a mistake.
Anyway, the problem becomes that of finding an allocation A of the Bj to the Si in a manner that is coherent and pleasing. And so that you can actually retrieve a book that you are certain you own, but cannot find. This can be tricky and is probably a highly specialised variant of the well-known knapsack problem.
So when you have a solution, you cling on tight to it, even if Offa's Pond Life. is confusingly located. The problem is that the book set is dynamic, so properly, B = B(t): usually, the bookset grows, with occasional small contractions as you realise that you're better off not caught with certain volumes. but on the whole, B and Bj have monotonic increasing cardinality.
Then from time to time, as shelves fill, A stops being viable. When this happens, we need a new allocation A', or - more seriously - a redefintion of the Bj, or - very seriously - the aquisition of new components to S; Sn+1, Sn+2,....
This happened today.
Library of Congress classmarks - which define a book's classification derive from the 600 subject field allotted to the book by their trained cataloguers. Dewey works similarly but can be taken go extreme lengths so the classmark can be ludicrously long (I recommend Kurtzweill's "Long Complication" to see a "classification off" in novel form). Booksellers are less concerned with accuracy and mainly use rough subject areas and the alphabet ( judiciously placing some titles where they think they'll sell rather than where they should live). A bookselling wag on the Wirral some 35 years ago placed the ible, Koran etc and atural history field guides in a section labelled "God and all His Works". I've been a cataloguer and a bookseller. As a boater my books are now arranged on the "where it will go" principal. It's chaotic but does not diminish the enjoyment of the contents. It took me 60 years to find that out. Perhaps the Two Ronnie's had it right all along.
ReplyDeleteYou've given me an idea
DeleteHmm, a tricky area, complicated by two factors, one being age of bookcase, the other being age of books. My approach has been alphabetical by author's surname (can cause some issues O'Rourke for example). But the main problem is relative size of book to shelf. My old bookcases have standard height shelves, maybe 8" at the top, with 10" and 12" shelves at the bottom. The standard 8" shelves work well for many older books, think traditional Penguin publications. But many more modern creations exceed that dimension. Whilst tempting to take STUPIDLY OVERSIZED books to the nearest bandsaw to trim them to fit. I have essentially had to resort to a dual start system. Ultimately I can see that I will spend much of my retirement (if it ever happens) getting further sawn fingers and splinters endeavouring to build taller bookcases with better spaced shelves to cope with the trend towards ever bigger books (and that's before I give in and go to large print).
ReplyDelete