In a statement published yesterday, Michael J. Klag, the Dean of Johns Hopkins Bloomberg School of Public Health writes that has “directed that the POPLINE administrators restore “abortion” as a search term immediately.” He is also launching an inquiry to determine why the change occurred. And also:
The Johns Hopkins Bloomberg School of Public Health is dedicated to the advancement and dissemination of knowledge and not its restriction.
As reported here, it seems that someone in the USAID’s Office of Population and Reproductive Health made a complaint about two abortion-related articles in the database that they felt were too close to “advocacy.” Those articles were removed from the database, but someone decided to go the extra step and make abortion a stop-word. The logic behind that decision has not, to my knowledge, been explained.
Classifications are political instruments… all classifications make epistemological, ethical, and political statements; there is nothing new to this. The library blogshere seems to argue that POPLINE’s move is unprecedented and unacceptable… get a grip; what is the ethical assumption behind Dewey’s religion section? I don’t see any ethical justification in the introduction to LCSH…
The first statement is absolutely accurate. For more on that, I refer you to Sorting Things Out: Classification and its Consequences. Hope Olson’s work also springs to mind. The bias present in tools like the Dewey Decimal Classification and the Library of Congress Subject Headings is glaring. We were just discussing this in my cataloging and classification course on Wednesday when we dove into DDC for the first time. If you are not familiar with the Religion schedule in DDC, it goes pretty much as follows:
- 200 Religion (general concepts, applicable to any religion, though the assumption of the classification is made fairly clear in captions such as: 202.11 God, gods, goddesses, divinities and deities)
- 210 Philosophy and theory of religion (the basic numbers in this span are for these topics in Christianity. To indicate other religions, one adds more digits to the end of a number.)
- 220 Bible
- 230 Christianity Christian theology
- 240 Christian moral and devotional theology
- 250 Local Christian church and Christian religious orders
- 260 Christian social and ecclesiastical theology
- 270 Historical, geographic, persons treatment of Christianity Church History
- 280 Denominations and sects of Christian church
- 290 Other religions
Just a little biased. Just a tad. Not ok. They are working on it. The foreword to DDC22 does address what the Editorial Policy Committee is doing about the bias of the classification since they recognize they have responsibilities to diverse users. Due to the basic structure of the classification, however, they can’t alter it much without causing the classification numbers of huge numbers of already classified items to suddenly become out of date and wrong in the current system.
But anyway… comparing this to the POPLINE issue is apples and oranges. In DDC, “other religions” are still there. They still have numbers, even if those numbers are all crammed into the 290s. It is not as though someone decided to remove 297 Islam, Babism, Bahai Faith from the schedule altogether, as though it does not exist. From reading the subject heading change list every week, it seems any concept with literary warrant, no matter how bizarre, will be added to LCSH. We may not like the preferred term chosen, or the references made, but the concept will be represented there somehow.
But the outrage wasn’t about the formal terminological structure behind POPLINE. They did not alter their indexing controlled-vocabulary to reflect a change in politics or society, etc. If they had changed the preferred term in their thesaurus from abortion to murder of the unborn (or whatever they call it) with a see reference from abortion, many people would be offended but it would not be censorship.
They made the term abortion a stopword in the indexing, effectively removing the term from the index altogether, including the indexing language, as though the it carried the same semantic weight as the or and. A plain keyword search for abortion returned nothing, despite the presence of thousands of titles and abstracts in the database containing the term.
This was not a matter of bias or politically incorrect/offensive terminology in a knowledge representation. Instead they were just waving their hands saying that topic doesn’t exist in this database at all, nothing to see here. It mainly looks like someone freaked out that the database’s USAID funding might get yanked because people could find any information about abortion in the database, so they decided not to let anyone find any information on “that word.”
Sadly this is very precedented, but absolutely not acceptable. If we catch any reputable information providers doing this sort of thing, it is well worth an outrage.