Reclassification after a DDC edition migration in a grant-aided college library
A grant-aided arts, science and commerce college in Pune district runs a single library of 52,000 volumes, all classified in DDC 19 and shelved accordingly. The library migrated to Koha eighteen months ago, but the OPAC still displays the DDC 19 numbers, and a new batch of postgraduate students in computer science and management cannot locate material: the computer science books sit under 001.64, while the syllabus and the publisher's catalogue use 004–006. Faculty in education and law report similar misfits. The principal has secured a one-time grant of ₹6 lakh for 'modernisation', and the NAAC peer team is expected within eighteen months. Staffing is three professionals and two semi-professionals, with no new posts sanctioned. Roughly 2,100 volumes of Marathi literature and 900 of history are heavily used, but their class numbers have barely changed between editions; the 300s, 500s and 600s are both the fastest-growing and the most affected. Full reclassification of all 52,000 volumes has been estimated by the librarian at 3.5 person-years.
1. Compare full reclassification, no change, and selective reclassification, and recommend one with reasons.
Read the worked answer
Full reclassification. 3.5 person-years against three professionals, of whom perhaps 1.5 FTE is actually free from circulation, reference and acquisition work. In an 18-month window it is not achievable, and while it runs the shelves are split between two editions - the worst possible state for a user.
No change. Cheapest and fastest, but the misfit is concentrated exactly where the collection grows fastest (004–006 computer science, 340 law, 370 education, 65x management). Koha already exposes the inconsistency, so 'no change' is not a neutral option; it is a visible defect during a peer-team visit.
Selective reclassification. Reclassify whole class ranges where (i) the schedules diverged materially between DDC 19 and DDC 23, (ii) stock growth is high, and (iii) circulation is high. Leave 800s literature (Marathi at 891.46 has been stable) and 900s history largely untouched apart from spot checks.
Recommendation: selective reclassification plus a written rule that every new accession from a stated date carries DDC 23 numbers. Approximately 12,000–15,000 volumes, achievable by one professional plus one semi-professional in 14 months with student assistants for shelf work. The ₹6 lakh is better spent on student assistance and reprinting shelf labels than on a project that cannot finish. The committee should minute that the criterion is documented consistency, not uniformity of edition - which is what matters in a NAAC file.
2. Which parts of the collection should be reclassified first, and what documentary evidence makes the choice defensible?
Read the worked answer
Prioritise by a composite rule: magnitude of schedule change × rate of growth × intensity of use. In practice that means computer science (old 001.64 arrangements against 004–006), education (370), law (340), management (658 and its subdivisions), engineering and medicine.
Evidence to place before the library committee:
- OCLC's published summary of changes between the two editions, showing which classes were relocated or given phoenix schedules.
- Side-by-side relative-index entries for a sample of 50 titles, showing what each edition gives as the class number.
- ILMS circulation by class range over three years, which shows where the misfit actually hurts users.
- Shelf-list counts per class, which converts the decision into a work estimate and a cost per class range.
- Syllabus-to-class mapping for the postgraduate programmes that complained.
The committee is not being asked to approve a preference; it is approving a rule. If the same rule is applied to every class and the counts are attached, the choice of what to reclassify becomes auditable and reversible. Anything left at DDC 19 should be listed as a deliberate decision with a review date, not left as an unexplained gap.
3. How should the catalogue and OPAC handle the transition so users are not misled?
Read the worked answer
- Never let one number silently stand for another. Record the classification in field 082 with the correct second indicator and the edition in subfield $2 (for example $2ddc23 or $2ddc19), so the edition travels with the record.
- Retain the superseded number in a local field or a public note ('Former class number: 001.64'), and build a searchable superseded-number index in Koha so a reader holding an old reading list can still find the book.
- Reclassify whole ranges in one operation, not scattered titles. Move a range with an 'in transit' location, relabel the shelves, then release it - otherwise the shelf order breaks and the reader loses the browsing function that classified arrangement exists to provide.
- Signage and a one-page map at the entrance, plus a short orientation for the postgraduate batch that raised the complaint.
- Announce the going-forward rule ('new books from 1 July carry DDC 23') so staff and users understand why two editions coexist temporarily.
- Keep the shelf list and a reclassification log (accession number, old number, new number, date, initials). This is the audit trail if the process is questioned, and it makes the work resumable if staffing changes.