free html hit counter All Classification Schemes Have Bias | John Battelle's Search Blog

All Classification Schemes Have Bias

By - September 05, 2004

deweyAs David Weinberger notes. In particular, the Dewey Decimal System has inherent religious biases. I’ve done some research on Mr. Dewey as part of my book, and he was quite the bigot, it appears.

I wonder, 100 years from now, when folks are writing the history of indexes like Google and Yahoo, what biases will emerge?

Related Posts Plugin for WordPress, Blogger...

8 thoughts on “All Classification Schemes Have Bias

  1. quanta says:

    Where did Weinberger say Dewey a bigot? He wrote:

    “But it’s not because Melvil Dewey (1851-1931) was a racist…Dewey himself was a progressive on social issues. For example, he hired seven women at the Columbia University library

  2. Bill Brazell says:

    Speaking of Bigoted Giants of Communication, Samuel Morse, the brilliant inventor of the telegraph, donated a chunk of his earnings from that device to anti-Irish-immigrant groups like the “Order of United Americans” (aka the “Know-Nothings”). Describing the Irish as a teeming sea of ignorant, dirty, overly fertile drunks, he ran for office on a platform of keeping America ‘pure.’

    I think of Morse sometimes when Irish-American Pat Buchanan uses similar language to rail against today’s immigrants – and when a couple of my older Irish-American relatives agree with him. My relatives, Mr. Buchanan, and myself should all be grateful that Morse, the Pat Buchanan of the 1840s, did not get his wish.

    I have not sought evidence of anti-Irish bias in old telegrams, but I feel a ghostly twinge when I pass Morse’s statue in Central Park.

  3. Future people will most likely recognize the anti-male bias that’s not yet widely understood today. It’s there in all media, though the internet isn’t quite as bad as TV and newspapers.

  4. Mike says:

    Not a defender of Buchanon but to be fair I believe his issue with immigrants has to do with illegals. As for bigotry in the later 1800s it would not be surprising to find that the majority of adult males back then held various bias beliefs.
    In my travels and observations I find that little has changed in this way.
    If I were reading a description of the Dewey Decimal system in a book I would not want references to his value system, I would expect that he had flaws. If the book was about Dewey and how the system was developed I would be interested in knowing if his core beliefs were reflected in his work.

  5. gary price says:

    DDC can easily be expanded at the local level to create new classification numbers. DDC is what’s called a synthetic classification scheme. Take different parts of the scheme and merge them together as needed. Finally, DDC is always being updated.
    http://www.oclc.org/dewey/updates/default.htm

    Bottom line, yes, Dewey had many issues and like you John have read many accounts. However, this does not mean that other religions cannot be handled by the scheme. Weinberger is correct that a revamp adding would cause many problems and cost
    many $$$. Wicca, Jews for Jesus and other groups CAN be handled but with longer numbers.

    This is versus the other major library classification scheme called LCC. (Library of Congress Classification).

    This scheme is not as easy to expand. It’s considered an enumartive classification (it’s all spelled out).

    Classification does provide underlying subject access but in this country at most libraries, it’s primarily used to “mark and park” materials. In other words most libraries only use classification materials so they can place them on the shelves (call numbers) and bring like things together. Don’t get me wrong browsing and serendipity are KEY!!! Of course, most OPACS can be virtuall browsed by call/classification number (just like you would be physically walking along the stacks) but most people don’t search this way.

    What people call things AND how most people search for material (when searching by subject) is by using Library of Congress Subject Headings (LCSH). These headings are edited and voted upon by a group of librarians at LC.

    The biggest problem is that these subject headings often take time to come into official usage. Literary warrant is what drives most of this and they often don’t reflect how people speak. Of course, any library can use or modify these headings.

    If you would like to see some recent lists of LCSH terms being added, dropped, or modified this page has links to recent bulletins.

    Go to http://www.loc.gov/catdir/cpso/

    and scroll to:
    Library of Congress Subject Headings
    Select a bulletin by date.

    You’ll also see a link to read recent additions to the LCC scheme.

    Prior to online catalogs most catalogers limited records to only three subject headings. Why? Creating, filing, and maintaining catalog cards was a timely and inexpensive process. Now, with computerized catalogs, it’s much easier to do this. Computers have also made working with cross references easier.

    Other controlled vocabs and lists of subject headings exist. For example,

    ***ProQuest has its own thesaurus of terms as do many other databases that handle content from serials. LCSH and Dewey are not traditionally used for this type of material. That’s right, all of these schemes are used just for monographs, maps, and other types of material.

    ***Other classification, subject oriented, schemes exist. For example, ERIC not only offers subject access but also classified access to entries.

    Finally, Weinberger should have been a bit clearer (for the non-librarian audience) in his post. When he’s talking about school libraries, he means K-12. Most larger libraries (including most university libraries) use the LCC/LCSH schemes.

  6. gary price says:

    oops. I forgot to add that I agree with one of Weinberger’s conclusions, “Taxonomies are tools, so there’s no such thing as the One Right Taxonomy, just as can-openers aren’t more right than asphalt spreaders. By building in sufficient metadata

  7. alex wright says:

    For a deeper look at classification systems and cultural bias, try “Sorting Things Out: Classification and Its Consequences,” by Bowker and Star.

  8. Doug Cutting says:

    Google frequently defines its algorithm as “objective”, e.g. http://www.google.com/technology/, do there’s no bias there. I haven’t heard Yahoo! make this claim, so one must more closely watch their results for bias. Nutch punts altogether and says that objectivity is impossible, that transparent bias is preferable. How lame!

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>