Subject: RE: [dita] Are indexterm ranges backwards incompatible?
Paul, If you look at the major style guides and Nancy Mulvany's critical text, Indexing Books, you'll find that professional indexers always point to ranges as reflecting an important and extended discussion of a subject and that a series of point page references does not convey the same information, but points to individual mentions of a topic. Ranges are not designed to make the index easier to read but to convey a level of importance about the subject matter. As a reader of indexes, I also assume the same level of importance of a range of pages. That's where I tend to look first, unless the index uses bold (another convention in print) to indicate a key page on the subject. Many indexes, of course, do both. JoAnn JoAnn T. Hackos, PhD President Comtech Services, Inc. 710 Kipling Street, Suite 400 Denver, CO 80215 303-232-7586 firstname.lastname@example.org joannhackos Skype www.comtech-serv.com -----Original Message----- From: Grosso, Paul [mailto:email@example.com] Sent: Wednesday, August 16, 2006 8:01 AM To: firstname.lastname@example.org Subject: RE: [dita] Are indexterm ranges backwards incompatible? > -----Original Message----- > From: Tony Self [mailto:email@example.com] > Sent: Wednesday, 2006 August 16 01:42 > To: firstname.lastname@example.org > Subject: RE: [dita] Are indexterm ranges backwards incompatible? > Thinking laterally... just an idea... The primary purpose of > presenting to the reader a range (pages 34-37) as opposed to > a series of points (34,35,36,37) is to indicate the greater > importance of the range. I disagree. The primary purpose of merging page numbers in an index is to make the index easier to read. Ranges do not semantically imply importance. > Expressing that semantically, the index markup > should simply indicate the content's importance. This would > mean the prolog index entry would be expressed as: > <prolog><metadata><keywords> > <indexterm important="true">cheese</indexterm> > </keywords></metadata></prolog> > if the author wanted to indicate that this is a major > discussion of the index term. I agree one could have markup that allows the indexer to indicate importance. And I suppose an implementation could allow the user to decide how to reflect that importance in the resulting index. I would still argue that reflecting importance by using or not using ranges would be completely lost on almost all index readers and would therefore be a bad choice for how to reflect importance. > > A DITA publishing tool may choose to provide the facility to > output contiguous "important" indexes as a concatentated > range (34-37), or perhaps in bold (as some "traditional" > indexes do), or perhaps only indicate the first page number, > or treat them like any other "point" index. > > Would the above mean that there would be no need for the > index-range element? In some ways, there is no "need" for an index range element. It is mostly a convenience to allow indexers to mark a sequence of pages. As an alternative, the indexer could just repeat the pointwise indexterm at the beginning of each paragraph within the range. But the fact that many markup applications allow some form of index ranges indicates it is a "nice to have" for users. Importance and ranges are orthogonal concepts. paul