Library classification: DDC and Colon Classification

In short

Library classification translates the subject of a document into an ordinal notation so that documents on the same subject stand together on the shelf and in the catalogue. Two systems dominate the Indian syllabus: the Dewey Decimal Classification (DDC) — an almost-enumerative, decimal scheme of 1876, now in its 23rd edition and maintained online by OCLC — and the Colon Classification (CC) of S. R. Ranganathan, 1933, the world's first fully analytico-synthetic (faceted) scheme and India's main contribution to the theory of knowledge organisation. Classification is examined in every B.LISc/M.LISc "Knowledge organisation" paper (theory plus a practical number-building paper), in UGC-NET/SET Library and Information Science, and in librarian recruitment tests for KVS/NVS, universities, state public service commissions, DRDO/CSIR/ICAR libraries and public library services. In practice almost all Indian libraries shelve by DDC (the National Library, Kolkata, most university and college libraries), while CC is taught everywhere and used in a handful of libraries and in research on facet theory at DRTC, Bangalore.

Explained

Fundamentals: what classification does and the terms used

• Classification: arranging a universe of knowledge into a filiatory (progressive, related) sequence, then mapping each class to a notation so the sequence is mechanically reproducible.
• Class number (subject) + book number (author/title/year, often Cutter or Ranganathan's Book Number) + collection number = call number, the unique shelf address of a copy.
• Two functions: *shelf arrangement* (a linear order, one place per book) and *classified catalogue* (subject approach through a class index, usually by chain procedure).
• Three planes of work (Ranganathan, *Prolegomena to Library Classification*): idea plane (the subject and its analysis), verbal plane (the terminology/title), notational plane (the symbols). Errors must be traced to the right plane.
• Array and chain: an array is a set of co-ordinate classes derived by one characteristic (e.g. 541, 542, 543…); a chain is a sequence of classes of decreasing extension and increasing intension (600 → 630 → 633 → 633.1).
• Types of scheme:
  1. Enumerative — every class listed ready-made (Library of Congress Classification).
  2. Almost-enumerative — mainly listed, with limited synthesis (DDC).
  3. Almost-faceted / analytico-synthetic — schedules of isolates combined by rules (UDC).
  4. Rigidly faceted — fixed facet formula per main class (CC editions 1–6).
  5. Freely faceted — facets drawn as the subject demands (CC 7).
• Depth classification: classifying micro-subjects (periodical articles, reports) to many facets — CC's design goal; DDC's segmentation (prime marks) lets a library cut a long number short.

Dewey Decimal Classification (DDC)

• History: devised by Melvil Dewey at Amherst College, published anonymously in 1876 (first edition, 44 pages); 23rd edition in print, 2011; Abridged Edition 15 (2012) for small libraries. The scheme is owned by OCLC (Online Computer Library Center), the editorial office sits at the Library of Congress, and policy is set by the Decimal Classification Editorial Policy Committee (EPC). Since DDC 23 the scheme is maintained as a continuously updated online product (WebDewey) rather than by fixed print editions; a Hindi translation of DDC 23 exists.
• Structure: 10 main classes → 100 divisions → 1,000 sections, pure decimal notation, minimum three digits, decimal point after the third digit.
  • 000 Computer science, information and general works · 100 Philosophy and psychology · 200 Religion · 300 Social sciences · 400 Language · 500 Science · 600 Technology · 700 Arts and recreation · 800 Literature · 900 History and geography.
  • Library and information science sits at 020; the DDC itself is classed at 025.431.
• Parts of the edition (DDC 23, four volumes): Vol. 1 — introduction, manual, Tables 1–6; Vols. 2–3 — schedules 000–999; Vol. 4 — Relative Index.
• Six auxiliary tables (numbers never used alone, always preceded by a hyphen in the table):
  • T1 Standard subdivisions (−01 philosophy and theory, −03 dictionaries/encyclopaedias, −05 serials, −06 organisations, −07 education and research, −09 history and geography/geographic treatment).
  • T2 Geographic areas, biography — India −54, Kerala −5483, Tamil Nadu −5482, Maharashtra −5479.
  • T3 Subdivisions for the arts, individual literatures, specific literary forms (T3A, T3B, T3C) — −1 poetry, −2 drama, −3 fiction.
  • T4 Subdivisions of individual languages and language families — −3 dictionaries, −5 grammar, −8 standard usage.
  • T5 Ethnic and national groups · T6 Languages.
  • (Table 7 "Persons" existed up to DDC 21 and was dropped in DDC 22, 2003; use T1 −08/−09 or the 920s instead.)
• Number building is additive, in a prescribed citation order: base number → facets listed under "add" instructions in the schedule → T1 standard subdivision → T2 area, etc. The schedule's own "add to base number…" note always overrides the general tables.
• Key rules and devices:
  • Hierarchical force: a note at a class applies to all its subdivisions.
  • Rule of three / rule of application / first-of-two rule / rule of zero — tie-breakers when a work covers several subjects (prefer the class covering all three; prefer the subject acted upon; prefer the one appearing first in the schedule; prefer a number without a zero in the subdivision).
  • Centred entries (►) cover a span of numbers (e.g. 592–599).
  • Relative Index: an alphabetical index that shows a term's *relations* — one term may lead to several numbers from different disciplines (e.g. "water" in 546, 553.7, 628.1).
  • Discipline-first: DDC classes by discipline, not by the subject in the abstract; "flowers" goes to botany, horticulture, art or literature depending on the treatment.
• Advantages: pure, internationally recognised, mnemonic (−54 is always India, 91.7 is always Russian in languages/literatures), continuously revised, relative location with infinite decimal hospitality in chain, enormous support (MARC field 082, records from LC/OCLC, CIP data), easy to teach.
• Limitations: Anglo-American and Christian bias (religion 220–280 for Christianity, 294 for all Indian religions); only 10 slots in each array, so no hospitality in array — new subjects are crammed in by long decimals (computer topics in 005–006); long numbers (ten digits and more); relocations and "Phoenix"/completely revised schedules between editions force reclassification; weak at micro-subjects and complex phase relations.

Colon Classification (CC)

• History: S. R. Ranganathan (1892–1972) — Madras University Librarian, founder of DRTC Bangalore (1962), National Research Professor in Library Science, author of the Five Laws of Library Science (1931). CC editions: 1st 1933, 2nd 1939, 3rd 1950, 4th 1952, 5th 1957, 6th 1960 (the edition still used for teaching and practice), 7th 1987 (Vol. 1 schedules, published posthumously, edited by M. A. Gopinath; incomplete and little adopted).
• Mixed notation: Indo-Arabic digits, Roman and Greek letters, and punctuation as facet indicators — hence the name "colon".
• Five Fundamental Categories (PMEST) — the postulate that every subject can be analysed into:
  1. Personality (P) — the "whole" or focal thing, identified by elimination; indicator comma ,
  2. Matter (M) — material, property or method; indicator semicolon ;
  3. Energy (E) — action, process, problem, operation; indicator colon :
  4. Space (S) — geographical area; indicator full stop .
  5. Time (T) — period; indicator apostrophe ' Facet sequence is P M E S T after the main class; a facet may recur in later rounds (a second [E] starts round 2) and at successive levels ([P1], [P2], [P3]).
• Main classes (about 43 in CC6): z Generalia, 1 Universe of knowledge, 2 Library science, 3 Book science, 4 Journalism, B Mathematics, C Physics, D Engineering, E Chemistry, F Technology, G Biology, I Botany, J Agriculture, K Zoology, L Medicine, M Useful arts, Δ Spiritual experience and mysticism, N Fine arts, O Literature, P Linguistics, Q Religion, R Philosophy, S Psychology, T Education, U Geography, V History, W Political science, X Economics, Y Sociology, Z Law, plus canonical and systems classes.
• Common isolates apply across classes — anteriorising common isolates (a bibliography, *a* for bibliography) file before the host class, posteriorising ones after.
• Devices for creating new isolates (sharp tools of synthesis):
  • Chronological Device (CD) — a time number as a focus (used for editions, theories named by period).
  • Geographical Device (GD) — a space number as a focus (e.g. a style of art by country).
  • Subject Device (SD) — a whole class number, in circular brackets, as a focus.
  • Alphabetical Device (AD) — the first letters of a proper name.
  • Enumeration / Mnemonic Device and superimposition of devices.
• Phase relations join two subjects in one document: general, bias, comparison, difference and (CC7) influencing/tool phases, shown by connecting digits (CC6 uses 0 with a, b, c, d); intra-facet and intra-array relations use the same idea inside a facet.
• Zone analysis of an array: digits of an array are split into zones (e.g. 1–8 for the first zone, then 9 as an octavising digit, letters as later zones) to give hospitality in array — Dewey's chief weakness.
• Canons, postulates and principles (the theory most often examined):
  • 43 canons across the three planes: for characteristics (differentiation, concomitance, relevance, ascertainability, permanence), for array (exhaustiveness, exclusiveness, helpful sequence, consistent sequence), for chain (decreasing extension, modulation), for filiatory sequence, and for notation (synonym, homonym, relativity, uniqueness, hierarchy, mnemonics, hospitality in array and in chain, mixed notation, faceted notation).
  • Principles for helpful sequence in an array: later-in-time, later-in-evolution, spatial contiguity, quantitative measure, increasing complexity, canonical sequence, literary warrant, alphabetical (last resort).
  • Principles for facet sequence: Wall–Picture (the wall must come before the picture on it — anatomy before disease), Whole–Organ, Cow–Calf, Actualised–Unactualised.
• Advantages: truly analytico-synthetic, so it classifies subjects never enumerated anywhere; brief numbers for complex subjects; strong hospitality in array and chain; expresses multi-faceted and phase-related subjects precisely; the theory behind it fuelled UDC revision, the Classification Research Group (CRG), BC2, PRECIS and modern faceted taxonomies and website navigation.
• Limitations: mixed notation with punctuation is hard to type, file and shelve (a comma or a dot is easily lost); not machine-friendly in MARC practice; unstable between editions (CC7 changed much and was never completed); no central revision agency or continuous updating after Ranganathan; schedules thin in science and technology; few ready-made numbers in shared cataloguing, so copy cataloguing is impossible.

Comparing the systems, and Indian practice

• DDC gives the classifier a number to find; CC gives a number to build. DDC's unit of work is the schedule entry; CC's is the isolate.
• Indian libraries that need interoperability (union catalogues, e-Granthalaya, Koha, SOUL 2.0, INFLIBNET's IndCat, NDLI) use DDC because its numbers travel with imported MARC 21 records (field 082 for DDC, 080 for UDC, 084 for other schemes such as CC).
• RRRLF-aided public libraries and most state/college libraries use DDC; CC survives in a few libraries and in teaching, and its theory remains compulsory in every Indian LIS syllabus.
• UDC (Brussels, 1905, from DDC 5) is the usual third scheme examined: decimal base with full synthesis, auxiliary signs + : / = (0…) and common auxiliaries; DDC 23 numbers must never be mixed with UDC notation in one sequence.

Key points

• Classification - mapping subjects to an ordinal notation to give a filiatory shelf and catalogue order.
• Call number - class number + book number + collection number; identifies the copy, not the subject alone.
• DDC - Melvil Dewey, 1876; 23rd print edition 2011; owned by OCLC, policy by EPC, now updated online.
• DDC structure - 10 main classes, 100 divisions, 1,000 sections; decimal, minimum three digits.
• DDC tables - six: T1 standard subdivisions, T2 areas, T3 literature forms, T4 languages, T5 ethnic groups, T6 languages; Table 7 withdrawn in DDC 22.
• India in DDC - area notation −54; Hindi language 491.43, Hindi literature 891.43, history of India 954.
• Relative Index - alphabetical index showing one term's several disciplinary homes.
• Hierarchical force - a note at a class binds all its subdivisions.
• Hospitality - DDC has it in chain (decimals) but not in array (only 10 slots); CC has both, via zone analysis and octavising.
• CC - Ranganathan, 1933; CC6 (1960) is the working edition, CC7 (1987) incomplete; about 43 main classes.
• PMEST - Personality , Matter ; Energy : Space . Time ' — the facet sequence and their indicator digits.
• Rounds and levels - a repeated Energy facet starts a new round; repeated Personality gives [P1], [P2].
• Devices - Chronological, Geographical, Subject, Alphabetical, Enumeration; may be superimposed.
• Phase relations - general, bias, comparison, difference (and influencing in CC7) between two subjects.
• Three planes - idea, verbal, notational; diagnose every classification problem in the right plane.
• Wall–Picture principle - the more concrete, supporting facet precedes the dependent one.
• Five Laws of Library Science (1931) - the foundation from which Ranganathan derived the canons.
• Practice in India - DDC for shelving and shared records (MARC 082); CC for theory, depth classification and faceted-taxonomy design.

Formulas and examples

Example 1 — DDC, geographic treatment. *Economic aspects of agriculture in India.*

1. Discipline: economics of a sector → 338.1 Agriculture (economics of production).
2. No special add note for place here, so use Table 1 −09 (geographic treatment).
3. Table 2 for India = −54.
4. 338.1 + 09 + 54 → 338.10954. (A library may shelve the segmented form 338.1.)

Example 2 — DDC, language table. *A dictionary of the Hindi language.*

1. Language, not literature → 400s; Hindi language base number 491.43.
2. Table 4 subdivision for dictionaries = −3.
3. 491.43 + 3 → 491.433. (A Hindi *grammar* would take T4 −5 → 491.435.)

Example 3 — DDC, literature table. *Modern Hindi poetry.*

1. Literature of a language → 800s; Hindi literature 891.43.
2. Table 3B form subdivision for poetry = −1.
3. 891.43 + 1 → 891.431. Contrast with the history of India, 954, and the history of Kerala, 954 + 83 → 954.83 (T2 −5483, dropping the 54 already in 954).

Example 4 — CC, classic facet analysis. *Treatment of tuberculosis of the lungs.*

1. Main class: Medicine = L.
2. [P] organ: lung = 45 → L,45
3. [M] disease: tuberculosis = 421 → L,45;421
4. [E] problem/action: treatment = 6 → L,45;421:6

Example 5 — CC, rounds, space and time (the standard textbook illustration). *Research on the cure of pulmonary tuberculosis by X-ray, carried out in India in the 1950s.*

1. L,45;421:6 as above (round 1: P, M, E).
2. Round 2 matter — X-ray as the agent = ;253.
3. Round 2 energy — research = :f.
4. Space: India = .44; Time: 1950s = 'N5.
5. Final number: L,45;421:6;253:f.44′N5.

A simpler build: *Agriculture in India in the 1990s* → main class J + space 44 + time N9 = J.44′N9.

Where it is used

• Academic and college libraries (IIT/IIM/central university libraries, affiliated colleges under UGC CBCS institutions): DDC 23 class numbers entered in MARC 082 in Koha, SOUL 2.0 or LibSys; cataloguers and professional assistants build numbers daily.
• Public library systems aided by the Raja Rammohun Roy Library Foundation, and state library services under the respective state Public Libraries Acts (Tamil Nadu 1948, Andhra Pradesh 1960, Karnataka 1965, Kerala 1989 and others): DDC for shelving and open-access browsing.
• Special and research libraries — DRDO, ISRO, CSIR laboratories, ICAR institutes, RBI, NIC and corporate knowledge centres: depth classification of reports and articles, where CC-style facet analysis informs in-house thesauri and taxonomies.
• National-level digital services — National Digital Library of India, INFLIBNET's IndCat and Shodhganga, e-Granthalaya in government libraries: subject browsing trees built on DDC classes.
• Job roles and exams — Librarian/Assistant Librarian (UGC-NET/SET qualifying), Library and Information Officer, cataloguer, metadata and taxonomy specialist, information architect; practical classification papers in B.LISc/M.LISc exams and KVS/NVS, Railway, bank and PSC librarian tests.
• Beyond libraries — facet analysis from CC underlies faceted search filters in e-commerce and site navigation, enterprise taxonomies and controlled vocabularies.

Interview questions

• Q: What is the difference between an enumerative and an analytico-synthetic scheme? A: An enumerative scheme lists every ready-made class number; an analytico-synthetic scheme lists isolates and rules, and the classifier synthesises the number for the subject in hand.
• Q: Expand PMEST and give the indicator digit for each. A: Personality (comma), Matter (semicolon), Energy (colon), Space (full stop), Time (apostrophe) — cited in that order after the main class.
• Q: How many tables does DDC 23 have, and which one was removed earlier? A: Six tables (T1–T6); Table 7 for persons was dropped in DDC 22 (2003), its work now done by T1 −08/−09 and the 920s.
• Q: What does "hospitality in array" mean and which scheme has it? A: The ability to insert a new co-ordinate class into an existing array; CC provides it through zone analysis and octavising, while DDC's ten-digit array does not.
• Q: Why is DDC called a *relative* location scheme? A: The number denotes the subject, not a fixed shelf; books move as the collection grows, keeping subject neighbours together — unlike fixed location by shelf and slot.
• Q: What are Ranganathan's three planes of work? A: Idea, verbal and notational planes; a problem in classification must be solved in the plane in which it arises.
• Q: State the Wall–Picture principle with an example. A: The facet whose existence the other depends on comes first — anatomy (lung) before disease (tuberculosis), hence L,45;421 and not the reverse.
• Q: Build a DDC number for "public libraries in India". A: 027.4 (public libraries) + T1 −09 geographic treatment + T2 −54 India → 027.40954, subject to any add note in the schedule at 026–027.
• Q: Why do most Indian libraries use DDC rather than CC, despite CC's theoretical superiority? A: DDC is continuously revised, has pure typeable notation, and arrives free in imported MARC records; CC has had no central revision since 1987 and no pool of shared numbers.
• Q: What is a phase relation? Give a type. A: A relation between two whole subjects in one document — for example a bias phase ("statistics for librarians"), where one subject is presented with a bias towards the other.
• Q: What is segmentation in DDC? A: Prime marks in a printed number showing where a library may truncate it, so a small library can shelve 338.1 instead of 338.10954.
• Q: Which MARC 21 field carries a DDC number, and which carries other schemes? A: 082 for DDC (084 for other classification numbers such as CC, 080 for UDC, 050 for LC).

Common mistakes

• Forgetting the −0 before a standard subdivision, or adding T1 when the schedule has its own "add to base number" instruction that supersedes it.
• Repeating digits already present in the base number when adding an area notation — the history of Kerala is 954.83, not 954.5483.
• Writing a DDC number with fewer than three digits (54 instead of 954) or putting the decimal point anywhere but after the third digit.
• Mixing up CC indicators — using a colon for Matter or a comma for Energy; losing the apostrophe or full stop when writing on a spine label.
• Treating the five fundamental categories as subject names: Personality is determined by the method of residues within a main class, not by "a person".
• Confusing array with chain, and so misapplying the canons (exhaustiveness and exclusiveness belong to an array; modulation to a chain).
• Classifying by the concrete thing rather than by discipline in DDC, and quoting an old edition's number without checking relocations and revised schedules in DDC 23 or WebDewey.

Related topics

Cataloguing: AACR2, RDA and MARC 21 · Information retrieval and indexing · Five Laws of Library Science and Ranganathan's contribution · Universal Decimal Classification and Library of Congress Classification · Subject headings, thesauri and controlled vocabularies · Library automation: Koha, SOUL and e-Granthalaya.