[Publib] Re: Good-bye to Dewey

[Andy Barnett]

late in the discussion, but I wanted to think this through


The Perry Branch of the Maricopa County Library 
District has generated a  lot of discussion 
lately for totally dropping the Dewey Decimal 
system. You have to give Maricopa County credit 
for trying out new ideas. After attending the 
Public Library Association conference in Phoenix 
in 2004, I came back prepared to adapt some of 
their concepts for use in our library. But note: 
adapt and not adopt. Maricopa is growing at such 
a pace (basically adding a new branch site a 
year) that they have no choice but to drink from 
the fire hose. The rest of us can implement the future at a more sedate pace.

Let's start with what they seem to be doing. The 
new branch will have 24,000 items. Since it is 
sure to be heavily media oriented, I would expect 
no more than 6,000 of those to be adult 
non-fiction. No one expects DVDs, CDs or fiction 
to be in Dewey order, so the adult non-fiction is 
the heart of the discussion. Moving Children's 
non-fiction to display collections would be 
innovative, but much less of a concern. 
Children's non-fiction is a much smaller 
collection and less specialized. While Perry says 
there will be fifty display collections, some of 
those will be used for fiction, so let's estimate 
that there are twenty-five display collections 
for those 6,000 adult non-fiction books.  That 
works out to less than 250 books per display collection.

There are several things that strike me about the situation.

Perry is a small library, certain to be heavily 
focused on current and popular materials. For a 
collection of 6,000 adult non-fiction items, 
eliminating Dewey in display style collections is 
an option. Perry's collection is necessarily 
thin, indeed, thin by design. Maricopa could have 
built fewer larger branches, but decided in favor 
of a larger number of smaller branches.

Perry is a branch. They can rely on larger 
collections elsewhere in the library district to 
meet the demand for in-depth or non-current 
material. In general, the smaller the library, 
the more it relies on items permanently stored 
elsewhere in the system. For a branch Perry's 
size, it would not be surprising if 35% of what 
it checked out came from elsewhere. This 
definitely affects the size and type of the 
permanent collection needed on site.

For a highly visible library in a high tech city, 
it would not unusual if 25% of all circulation 
resulted from items requested over the Internet, 
with many of these coming from other libraries in 
the system. A great deal of Perry's need for 
in-depth non-fiction will be filled this way. 
Patrons will order a title online and pick it up 
at the branch without knowing or caring where the 
book "lives" permanently. Even among in-library 
users, many are willing to wait a few days for 
delivery of a title they really need. 
Inter-library delivery is as fast as any Internet 
seller and free to the patron.

If Perry's display collections run 250-300 
titles, they are the right size. My experience is 
that that size display is large enough to attract 
attention and fill casual needs, but small enough 
to browse effectively. Richmond (B.C.) Public 
Library speaks of "critical mass" in a 
collection. The difference is that in the 
Richmond model, items age out of the display 
collections and move to regular stacks. It is 
quite possible that items will age out of Perry's 
collections too, but be discarded instead.

Perry is planning to shelve items alphabetically 
by author in its display collections, while other 
libraries using display collections continue to 
use Dewey. The trade off is obvious. It is easy 
to shelve items in author order. If a patron 
knows the author, then author order is easy for 
them too. Author order does violate one of the 
expected rules for shelving – similar books 
should be shelved together. If a health section 
has six books on diabetes, four on schizophrenia 
and three on Parkinson's disease, shelving by 
Dewey puts the books in subject order, but author order mixes them at random.

No matter how many display collections one 
creates, there are titles that don't fit anywhere 
easily. This really is not an objection to 
developing display collections. Even in a badly 
chosen collection, a book benefits from being 
displayed. Patrons who want it can use the 
catalog to identify the chosen location. It is a 
problem if a book is permanently in a poorly 
chosen collection. In that case, it might be 
better off in a Dewey arranged stack, where it 
would at least be near similar titles.

In a properly sized display collection, the 
difference between Dewey and author order might 
not make too much difference. The collection is 
designed to be small and browsable. As long as 
the collections stay small, it will remain 
workable. Larger collections demand Dewey. If a 
library had a mere thousand cookbooks, author 
order would create a jumble of ethnic, regional 
and diet specific titles. It is likely that Perry 
plans to remain in that "sweet spot" where author order will work.

Display collections are a great idea. Items on 
displays will circulate more than items not 
displayed. Few works flourish in the stacks. 
Stacks are designed for effective storage, not 
effective browsing. If a smaller library can 
shelve everything in display, keep those 
collections current and rely on collections at 
other libraries for depth, it will be a circulation dynamo.

If a library is larger than Perry, it can 
implement many features of their display 
collections, following the pattern established by 
Richmond (B.C.) Public Library and others. New 
and popular items are placed in display 
collections, retaining Dewey classification. As 
items age, they are moved to regular stack areas, 
where they will remain available but will not 
generate as much use. The displays are gateways 
to the stacks. Readers can find a new thriller 
and go to the stacks for more by the author. They 
can find one book on Alzheimer's and use the 
Dewey number on it to find the rest of the library's holdings.

Will we see a spate of libraries dropping Dewey 
and moving to display style collections? I don't 
think so, though Maricopa might well be on target 
with this branch. At a very low level, size will 
preclude libraries from breaking their entire 
non-fiction into small collections. A relatively 
small non-fiction collection of 30,000 titles 
would require about one hundred collections. That 
many collections would be confusing, not helpful, 
to browsers. Dewey provides real advantages in 
providing fine-grained browsing. The War of 1812, 
home schooling and Japanese cooking all deserve 
their own space on the shelf, but are too small 
for their own collection. Dewey gathers them 
together and keeps them in context. Display 
collections would mix them into broader topics.


Andy Barnett, Asst. Director            McMillan Memorial Library
490 East Grand Ave. Wisconsin Rapids, WI  54494
                 www.mcmillanlibrary.org