on Dewey and complexity
The recent uproar (by librarians) following the decision of the Gilbert, AZ public library to abandon Dewey as the primary means of locating materials on the shelves demonstrates how we do (or do not) adapt to complexity.
The basic jist of the problem is summed up nicely in a Wall Street Journal article. In discussing the controversy, the author states that "it feeds into a broader, increasingly urgent discussion about libraries, where a growing number of patrons, used to Google and Yahoo, simply don't look for books and information the way they used to." As was the case with web design, in a complex system we must gather our information and adapt based on the environment.
As Bowker and Star pointed out in their book, Sorting things out: Classification and its consequences, classification works best when it is invisible. When classification starts to become visible you have a good indication that the classification system isn't performing its job. That's the problem here, Dewey (in this case, but it really could be any classification system common in libraries) doesn't work for the average person because it doesn't conform to our current mental models. It requires too much, "What does that mean?" or "Why is that?" processing to be viable for many uses. For example, why is computer science classified in the 004-006 range instead of the 500's (sciences) where it really belongs? Why when I perform a subject heading search for "Social network research" do I find books scattered throughout the library in spots as disparate as 658.472 G562c, 305.80072 M297, and 300.72 A244?
Adaptation is one of the ways we continue to make systems work in a changing environment. An example of this is the organization of books at Northwestern University. Here we use Dewey (which is odd for an academic library to begin with) but we do not have books arranged in strict call numeric order. The Dewey ranges are combined based on very high-level topic areas (humanities, hard science, natural science, social science, etc.) in an effort to bring disparate areas in the classification system together to make more sense of the order for our patrons. Is it perfect? No, but it is does make discovery easier in many areas that have been "broken up" due to anomalies of Dewey.
Rather than castigating the folks in Gilbert, we should be trying to learn from their approach. It's too bad the naysayers don't get that.
The basic jist of the problem is summed up nicely in a Wall Street Journal article. In discussing the controversy, the author states that "it feeds into a broader, increasingly urgent discussion about libraries, where a growing number of patrons, used to Google and Yahoo, simply don't look for books and information the way they used to." As was the case with web design, in a complex system we must gather our information and adapt based on the environment.
As Bowker and Star pointed out in their book, Sorting things out: Classification and its consequences, classification works best when it is invisible. When classification starts to become visible you have a good indication that the classification system isn't performing its job. That's the problem here, Dewey (in this case, but it really could be any classification system common in libraries) doesn't work for the average person because it doesn't conform to our current mental models. It requires too much, "What does that mean?" or "Why is that?" processing to be viable for many uses. For example, why is computer science classified in the 004-006 range instead of the 500's (sciences) where it really belongs? Why when I perform a subject heading search for "Social network research" do I find books scattered throughout the library in spots as disparate as 658.472 G562c, 305.80072 M297, and 300.72 A244?
Adaptation is one of the ways we continue to make systems work in a changing environment. An example of this is the organization of books at Northwestern University. Here we use Dewey (which is odd for an academic library to begin with) but we do not have books arranged in strict call numeric order. The Dewey ranges are combined based on very high-level topic areas (humanities, hard science, natural science, social science, etc.) in an effort to bring disparate areas in the classification system together to make more sense of the order for our patrons. Is it perfect? No, but it is does make discovery easier in many areas that have been "broken up" due to anomalies of Dewey.
Rather than castigating the folks in Gilbert, we should be trying to learn from their approach. It's too bad the naysayers don't get that.
Labels: change, complexity, service models
posted by Frank Cervone at
8:43 AM
0 Comments:
Post a Comment
Links to this post:
Create a Link
<< Home