Paper: KO &Processing: Classification

Module name: Work in the Idea Plane: Canons of characteristics

Module ID: LIS/KOP-C/5 (1)

Prerequisites: Understanding the basis of grouping, Division by logical

methods. Formation of classes based on purpose and utility.

Objectives: To have an acquaintance with the methods of formation of classes

and basis of making of groups

Keywords: Characteristics, Canons of characteristics, Dividing, Grouping, Idea

Plane, Planes of work.

Structure:

1. Grammar of Classification design

1.1 Planes of work

1.2 Characteristics

1.2.1 Definition of Characteristics

1.2.2 Parameter

1.2.3 Attributes

2. Canons given by Ranganathan for these twin tasks are:

2.1 Canon of Differentiation

2.2 Canon of Relevance

2.2.1 Relevance to the Purpose

2.3 Canons of Ascertain-ability and Permanence

2.4 Canon of Permanence:

3. Canons of Succession of Characteristics

3.1 The Canon of Concomitance

1
 3.1.1 Broader to narrower succession of characteristics:

3.2 Canon of Relevant Succession

3.2.1 Apply characteristics one by one

3.2.2 Order of application

3.2.3 Dividing Literature

3.2.4 No order is universal

3.2.4.1 Facet sequence

3.3 The Canon of Consistent Succession

3.3.1 Comments

4. Summary

5. Glossary

6. References and further Readings

7. Test Questions

1. Grammar of Classification design

S.R. Ranganathan veritably described his teacher W.C.B. Sayers (1881-

1960) as the first grammarian of libraryclassification, whereas Ranganathan

himself is universally acknowledged as the foremost grammarian of this field.

His logically trained mind turned the work of classification from an art into an

objective science.With his dynamic theory he objectified and mechanized the

process of designing and evaluating library classifications systems.

2
1.1 Planes of work

One of his important contributions is to divide the whole work of

classification into three planes, namely, the Idea Plane, Verbal Plane, and the

Notational Plane. He formulated clear canons, principles and postulates for

work in the each plane. All these laws are subject to the overarching normative

Five Laws of Library Since formulated by him and published in 1931. The

work in the idea plane, though invisible to a classifier, is the most important

one. It is the planning stage of the total work of classification which lays down

the foundation and sets the map of the classification system. In popular

parlance, it is the brain work of the whole process. Here a classification-maker

is concerned with dividing the entities into smaller and smaller groups until

every member gets uniquely individualized till no further division is possible or

required. Further,these are systematically regrouped into categories, traits and

isolates. The work of Idea Plane does not stop until each entity is properly

ranked by arranging them in arrays and chains.

1.2 Characteristics:

Classification at best is a paradoxical process. We divide to regroup, analyze

to synthesize, separate the constituents to reunite. The work of dividing entities

is done into categories and facets with the help of “characteristics"; and ranking

is done in arraysand chains with the help of principles andpostulates. The

whole process falls underthe purview of the Idea Plane and is carried out with

3
the help of four sets of Canons:

1. Canons of Characteristics(Four)
2. Canons of Succession of Characteristics(Three)
3. Canons ofArray (Four)
4. Canons of Chain(Three)
The canons of characteristics are concerned with the inherent qualities of the

characteristics to be used as the basis of division. A characteristic is a hammer

to break a mass of knowledge into smaller pieces. In other words it is the

criterion or basis of division.

1.2.1 Definition of Characteristics

Ranganathan defines a “characteristic”as “An attribute,quality, basis

with which a group may be divided into groups at least two”. Further he uses

the term "Division characteristics".

1.2.2 Parameter

The entire universe can be divided into two groups of living and non-

living. Here life is the characteristic. However, Ranganathan has interjected a

new term "parameter': "In the classification of a Universe of Isolate Ideas, each

of the successive characteristics used in arrivingat an isolate idea are, therefore,

sometimes referred to as the 'parameter' of its classification"'. The terms

"Characteristic" and "Parameter" mean the same thing.

1.2.3 Attributes

An entity whether physical or conceptual may have a number of attributes,

4
both inherent (permanent) and given or assumed. The more complex or

sophisticated an entity, the more characteristics it will have. Therefore, the

selection of characteristics to form the basis of a classification is a very

judicious process. The success of classification and its intended purpose will

totally depend:

 Firstly,upon their skillful selection of characteristics, and,

 Secondly,on the order or sequence in which the various characteristics

are used.

2. Canons given by Ranganathan for these twin tasks are:

1 Canons of Characteristics

1.1 Canon of Differentiation

1.2 Canon of Relevance

1.3 Canon of Ascertainability

1.4 Canon of Permanence

2 Canons of Succession of Characteristics

2.1 Canon of Concomitance

2.2 Canon of Relevant Succession

2.3 Canon of Consistent Succession.

2.1 Canon of Differentiation:

It implies that acharacteristic common to the whole group should not be

usedasthe basis of its further division. A common characteristic, that is shared

5
by the entire group,will not be able to produce any division in the group of

entities. For example, being a student of library science cannot be a

characteristic to divide a library science class. Similarly, we cannot divide

crows on the basis of their colour as they are all black. But shades of variation

in the universal characteristic can be used on the basis of differentiation. For

example, though possession of twoeyes cannot be a characteristic of dividing a

group of normal human beings, but the shape and colour of eyes can be a good

basis for grouping. Hence we should select that characteristics which divides

the group into at least two groups. A class of library science students can be

divided by sender into males and females, or by their religion, etc.

2.2 Canon of Relevance

It is more important and more difficult. A chosencharacteristics may satisfy the

canon of differentiation, but may not be very useful. A differentiating

characteristic is not right or wrong; it is “more helpful” or “less helpful”

depending upon the purpose of classification. For example, a group of persons

canbe divided on the basis of their age, height, gender, skin colour, mother

tongue, religion, caste, race, nationality,educational qualification, and many

other numerous characteristics. But for the purpose of a desired grouping all of

them may not be relevant. For example, height, weight, gender, race and

physical strength may be relevant characteristics for selecting a class for

physical training, but for forming a class for teaching mathematics, these are

6
irrelevant. For this only relevant characteristics are educational qualifications,

or language known.

2.2.1 Relevance to the Purpose:

Simply speaking, the selection of characteristics should be correlated

only to the purposeof classification. But this cannot be done a priori. There is

no mechanical formula to determine the relevance of a characteristic.

Ranganathan admits that it is a real problem to select the relevant characteristics

amongst a large number of differentiating characteristics. It requires genius,

flair and experience to "reject the less helpful characteristics”. Relevance of a

characteristic entirely depends upon the purpose of classification. Therefore, the

concept of "Relevant Characteristic" itself is related with the purpose of the

classification to serve the needs of library users.Application of different

characteristics means a different classification. The real problem remains not

tofind the relevant characteristics but to find the more relevant ones to design

more helpful classification. More than finding the relevant characteristics, the

real difficulty lies in foreseeing the helpfulness of classification that we wish to

produce for the purpose. The problem is not of means but of ends.

2.3 Canons of Ascertain-ability and Permanence

Both these canons address themselves to the inherent qualities of the

characteristics to be applied. The Canon of Ascertain-ability asserts that the

characteristic itself should be tangible andobjective. It should be perceptible to

7
any of the five senses. The characteristic should not be illusory, mythical or

controvertible. It should be concrete, at best. For example, one's previous birth

may be a differentiating, or even a relevant, characteristic to divide human

beings, but certainly it is not ascertainable conclusively.Hence, it should not be

used as the characteristic of division. Characteristics should not be subjective.

Secondly, but less rigorously, a group of teachers should not be divided on the

basis of being brilliant or dull, nor literary authors could be classed as major or

minor. Not because these qualities do not exist objectively. It is rather mostly

due to the fact that these characteristics are very much subjective. These depend

upon the perceptions and opinions of the observer. A teacher who is dull for one

group of students may be a source of inspiration for some other, and vice versa.

Such a characteristic also violates the Canon of Reticence of Terminology in

the verbal plane. Hence such controversial and subjective characteristics lead

only to confusion and controversy, not to any acceptable classification. Such

anattribute is technicallynot fit to be a characteristics.

2.5 Canon of Permanence:

Itstates that the characteristic though relevant and differentiating should

itself not be in a flux. It should be intrinsically permanent and of stable nature.

For example "fame" should not be used as a criterion for classifying literary

authors. Fame is not anything permanent. History of any literature clearly

shows that fame or obscurity is transitory and always changing shape like a

8
cloud. (Shakespeare and Milton were not reputed authors in their lifetimes). On

the other hand, many men famous when alive are forgotten the moment they

die. Persons may not be divided on the basis of the colour of clothes they are

wearing. It is changeable. They may be divided on the basis of their skin colour

or height, which are permanent. Natural characteristics are more reliable.

Ranganathan satirically cites the example that Indian politicians, who more

often than not change party affiliations to grab power, cannot be divided

permanently on the basis of their politicalideology they subscribe to. If such a

fluctuating characteristic is used as the basis, the members of the various groups

will become inter-migratory“aya-ram-gaya-rams”.And as a result, we will get

only a chaotic classification -- a classification in a flux: groups which are

intermixing, not the segregated ones.

However, "permanently fluctuating" nature of an entity can itself become a

characteristic. If it is their permanent feature, it itself is a good basis of

classification. "Migratory birds" can become a distinct group of birds on the

basis of their changing habitat. It is not the changing characteristic but the

"changing nature of entities" which can be used as the basis of division.

Individual chameleons cannot be classified on the basis of their colour; but the

chameleons as a group can be identified on the basis of their ever changing skin

colour.

9
3. Canons of Succession of Characteristics

To arrive at the desired grouping, or for individualization of entities, we have

to apply a number of characteristics one after the other. The next set of

canons is concerned with the sequence in which the various characteristics

(qualifying the canons of characteristics) are to be applied. These are:

(1) Canon of Concomitance

(2) Canon of Relevant Succession

(3) Cannon of Consistent Succession.

3.1 The Canon of Concomitance

It states that the two characteristics should not be concomitant i.e. the

two characteristics applied successively should not result in the same

grouping each time. Such a situation may occur if the two characteristics are

synonymous. For example, a teacher may at first divide a class in 2014 on the

basis of the students being less or above 20 years of age. It will result into

two groups. Ifthe nextcharacteristic is chosen as the year of birth being after

or before 1994, it will not produce any further grouping as both the

characteristics mean the same thing. A group divided on the basis of

male/female cannot be further divided on the basis of being men/women.

Tautology should be avoided.

3.1.1 Broader to narrower succession:

The successivecharacteristics should beapplied from boarder to

10
narrower. A group of females can be further divided into women, girls and

children.A stalemate in further division can occur if the two characteristics

are not applied in the order of their decreasing extension or increasing

specificity i.e. in the order of general to specific. For example, a group of

scientists may be divided on the basis of their being biologists and non-

biologists. But to apply the characteristic of being scientist to the biologists

will result in no classification, as all biologists are scientists first. Indians

cannot be further divided on the basis of being Asians. It will also violate the

canon of differentiation. Though Ranganathan has not statedthis explicitly,

but this principle is implied in his entire theory of classification. It is a

principle of logic firston which the classification at best is based. Many other

classificationists such as H.E. Bliss (1870-1955) and Sayers have stated this

canon explicitly. Ranganathan implies it as he thinks it to be too obvious to

be mentioned.

3.2 Canon of Relevant Succession

To divide a group minutely or rank every member uniquely we have to

apply a number of characteristics. These may be all differentiating,

permanent, ascertainable and respect the canon of concomitance, yet it may be

a formidable problem to decide the order in which the various characteristics

are to be applied one after the other.

11
3.2.1 Apply characteristics one by one:

One important caution is not to apply two or more characteristics

simultaneously, otherwise it will result in cross classification.For example, in

the following array four characteristics of clan, language spoken, caste, and

profession have been applied simultaneously:

Rajputs

Punjabis

Brahmins

Armymen

Therefore above group is motely, not coherent.In fact isno group in the

classification sense. It is not an array as it has no common genus. One could

bea Punajbi, Brahimin/Rajput and Army man at the same time.

3.2.2 Order of applications:

As the name implies the Canon of Relevant Succession of

Characteristics depends upon the purpose of classification. There is nothing

right or wrong in the order of succession of characteristics (subject to the

observing the canons of concomitance and increasing specificity). At best,

two different sets in succession maybe termed as helpful or unhelpful; more

helpful or less helpful to the majority of the library users. "There are of

course", to quote Arthur Maltby, "several useful arrangements in some fields,

but the object is to offer the best of these".

12
3.2.3 Dividing Literature:

For example, division of the class literature (useful to the majority of the

readers) as prescribed in the DDC and the CC is successively by language, form

period and title of the work. This is relevant as the majority ofreaders, being

monolingual, are interested in the literature of only one language. Further,

within a given library the majority is interested in one form of the literature,

viz, either in drama, or novel, or poetry. Therefore, in such cases the relevant

succession of characteristics is Language, Form, Author and Work. The order of

these characteristics though popular is not an absolute one. A special

classification having a different set of users may change this order of succession

of these characteristics. For example, if a library is interested in the history of

world poetry irrespective of the language, then the relevant succession of the

above four characteristics will be

(Literature): (Form: Poetry): (Period): (Language)

As another example, if a library specializes in Nineteenth Century, World

literature, it may prefer the following order of characteristics:

(Literature): (Period): (Language): (Form)

Facet sequence in UDC is flexible, especially in case of auxiliaries, which can

be altered to suit local requirements.

3.2.4 No order is universal:

No sequence is perfect or serves all needs. A classificationist is like the

13
proverbial father having two daughters, one married to a potter and another to a

farmer. Farmer’s wife is urging her father to pray for rain to have bumper crops,

while potter’s wife asks his father to pray for sunshine for her pots to dry soon.

One obvious problem with this facet sequence is that an author, like

Rabindranath Tagore, who writes in different forms then the works of such an

author will scatter by form. In the DDC and the CC the plays, fiction and poetry

of Tagore are filed separately. The UDC on the other hand brings together

works of a single author at one place.

3.2.4.1 Facet Sequence:

But this succession of characteristics is subject to the logical principle of

facet sequence, viz, wall-picture principle, the cow-calf principle, and the rest

of them. For example, in the case of literature to separate the work facet from

author facet by an intervening facet will be sheer absurdity. To some extent, the

problem of the relevant succession is the problem of the relevant characteristic.

The various chosen characteristics are to be applied in the decreasing order of

their relevance or usefulness. The PMEST formula for sequence of categories is

in the decreasing order of their concreteness.

3.3 The Canon of Consistent Succession:

This canon is to ascertain the uniformity and consistency of

classification. It purports to be a bulwark against the changing tides of

fashion, to remain firmly unmoved by the whims and caprices of the person

14
who classifies. In simple words, this canon means that the order of succession

of characteristics once fixed should always remain the same.

3.3.1 Comments:

Firstly, a situation may never arise when we may have to apply the

same set of characteristics to a different set of entitiesfor dividing the latter.

Either the characteristics will be different or the universe of knowledge to be

classified will be different. If both are the same, it will mean the same

classification. If Ranganathan means that for a given set of objects the

sequence of characteristics should never be changed, he negates the value of

experience gained with time. It will mean that the sequence of characteristics

once chosen should remain the same disregarding whetherthat is found helpful

forthe majority of readers or not.

With experience and interaction with users, one may realize a better or

more helpful sequence of characteristics. This canon then is contrary to

Ranganathan'sown philosophy: he himself has been changing the sequence of

characteristics in the main class E Chemistry and Z Law of his Colon

Classification.

Secondly, it is also contrary to the spirit of the canon or Relevant Succession of

Characteristics, as it states that the succession of characteristics should be

relevant to the purpose of classification.Ranganathan clarifies that "This canon

does not say that all the characteristics chosen as the basis of classification of

15
the subjects going with a Basic Class should be necessarily applied to all such

subjects. Many subjects do not take all the characteristics, one subject may

take some ofthe characteristics; another may take some others, and so on. The

demand of this canon is only that whatever characteristics are applicable to a

subject of the Basic Class should be applied in a sequence which has been

determined for all the characteristics applicable to one subject or another going

with the Basic Class concerned"". In spite of so much clarification, the Canon

of Relevant Succession may be violated, as this canon as enunciated in the

Prolegomena (EK1)reads: "The succession of characteristics should be

consistently adhered to, so long as there is no change in the purpose of

theclassification”

4. Summary:

Classification is a paradoxical process of dividing to make groups; analysis for

synthesis; breaking to reintegrate. For dividing an area/mass of knowledge or

some entities, we require a basis of division called characteristics. A

characteristic is a sort of hammer to break a group. To break a group into sub-

groups, we need a right kind of hammer. So what are the qualities of such a

hammer, and how to apply it? The guidance is provided by canons of

characteristics and the order in which they are to be applied given by

Ranganathan in the idea plane of classification work. Idea plane provides

theory based blueprints of the whole work of designing classification carried

16
out in three plans -- other being verbal and notational planes. These canons of

characteristics lay down that the characteristics chosen should be able to break

the group into atleast two sub-groups (Differentiation). The breaking should be

into larger to smaller groups. The characteristics chosen should not result in the

same grouping again (Concomitance). The characteristics applied should be

relevant to the purpose of classification (to serve the needs of majority of the

library users). The canon of ascertainability means the characteristic should be

tangible, objective and knowable. Lastly, the characteristic itself should be

permanent, not changing or be in a flux like rising or falling of mercury. The

chosen characteristic should not be fickle. Persons may be classified by their

skin colour but not by the colour of clothes they wear.

Further, for breaking a mass of knowledge, we have to strike the hammer

not once but many times successively to arrive at the right and desired

grouping. More than one characteristics are to be applied in succession. We

have to decide the order in which these are to be applied (Canons of succession

of characteristics). Two characteristics applied should not result in the same

grouping (Concomitance). The order in which these are applied should be

relevant to the purpose of library users. History of world can be arranged

chronologically irrespective of the country or it can be first divided by nation

and then by period. Both the approaches are correct depending upon the need of

the users. In fact no classification grouping serves all purpose or satisfies the

17
needs of all the users. To fulfil the needs of other users, librarians have devised

supplementary approaches in the form of indexes. Lastly the succession of

characteristics chosen once should remain the same as long as the purpose of

classification remains the same.

References and further Readings

Bavakutty, M. 1981. Canons of Library Classification. Trivandrum: Kerala

Library Association,. pp. 7-13.

Bronowski, J. 1960.The Commonsense of science.Harmondsworth: Penguin, pp.

54-55.

Buchanan, Brian.1979. Theory of Library Classification. London: Clive

Bingley, pp.17-37,52-53

Kaula, P.N.1980. "Canons in analytico-synthetic classification" International

Classification 7(3) 118-125

Langridge, D.W. 1992. Classification: Its Kinds, Elements Systems, and

Applications. London: Bowker- Saur, pp.84

Langridge, D.W. 1989.Subject analysis: Principles and procedure. London:

Bowker-Saur,. 146p.

Maltby, Arthur.1975. Sayers Manual of Classification for Librarians, 5th ed.

London: Andre Deutsch, p.32.

Ranganathan, S.R. 1967.Prolegomena to Library Classification, 3rd ed.

Bombay: Asia Publishing House, Sec EAI.

18
Ranganathan, S.R.1987. Colon Classification, 7th ed. Revised and edited by
M.A. Gopinath. Bangalore: Sarada Ranganathan Endowment for Library
Science.

Ranganathan, S.R.1967: A Descriptive Account of the Colon Classification.

Bombay: Asia publishing House, Sec. G17

Sachdeva, M.S. 1980.Colon Classification. 2nd ed. New Delhi: Sterling

Publish¬ers, pp. 88-93

Srivastava, A.P. 1992. TheoryofKnowledgeClassificationforLibrarians, 2nd ed.,

revised with the assistance of M.P. Satija, et al. New Delhi: The Learning
Laboratory,. 115p.

19