The FRBR model application to Italian cataloguing practices:
general problems and its normative use
(January 2002)

0. RICA–Regole italiane di catalogazione per autori – Italian Authors’ Cataloguing Rules

The Regole Italiane di catalogazione per autori – RICA (Italian Authors’ Cataloguing Rules) has been published in 1979 and after then, it came gradually into use in the Italian libraries. RICA was promoted by the Ministry of Education, at the time responsible for the state libraries, and published by ICCU-Istituto centrale per il catalogo unico. The Commission, working for at least ten years at building up the code structure, was formed by librarians named by the Ministry and the code, before issuing, was approved by law: such an uncommon procedure was required in order to make its application mandatory for all state libraries in the country. In spite of this unusual way to proceed the code, as a whole, resulted conceived as a national set of rules, not specially oriented to one or another kind of institutions. Its priority was first of all that of harmonising different national cataloguing practices in the country and to put them in line with international principles and standards. Such features made its application easy all over the country also for other kinds of libraries: University, Regional, Special ones.

RICA, soundly based on Paris Principles, became effective in the production of the Italian National Bibliography-BNI’s records in 1981. Since its issue RICA’s use and application have been widespread in the majority of Italian libraries and we can therefore consider it a de facto national cataloguing code. To make a picture of the whole landscape it should be added that RICA is the cataloguing code used in SBN - Servizio Bibliotecario Nazionale, the on line Italian union catalogue, linking at the moment more than 1.380 libraries.

While RICA can be likely considered a consistent and stable set of requirements it has not necessarily been conceived as an ultimate point for what concerns cataloguing. In fact, in order to keep it in line with changing international principles and standards, a Standing Commission for RICA evolution was set up and started to work in 1997.

Though our cataloguing code, as far as the majority of codes born, bred and educated along the guidelines of the ParisPrinciples, has proved quite an effective structure up to now, RICA SC in revising it thought it advisable trying, if practicable, to redesign for it a more explicitly analytical frame according to the FRBR model. The code is in fact already providing, as optional, some of the FRBR requirements. Such provisions have, obviously, been ignored also in their more contained applications as long as cataloguing boundaries were limited by card catalogues possibilities. Moreover, due to long lasting traditions, they have been applied up to now only partially also in automated files.

The code is actually mainly dedicated to rules for choice and definition of author (personal and corporate)/title headings. The new structure of RICA will still be dealing mainly with access points, trying to cover all kind of responsibilities linked to new materials and media, and will provide for descriptive cataloguing, also if only at a general level, referring for this part to the recommendations of the various specialized ISBDs. The code will be reviewed in order to cover all kind of materials and of related linked responsibilities, varied in number and completeness according to different cataloguing level. We all think that the general principles on which the present code is based are still to be considered as valid.

The new code should provide a more analytical structure as a whole, with the possibility of developing a full scale FRBR model application. In our opinion, besides this, it should be also analyzed the parallel possibility of creating a less than full model, with bibliographic data which could be enhanced while re-used, in a process of cooperating activities, this would satisfy several needs included that of containing costs of production.

In order to verify the possible application of the model to the new structure of the code, the SC started since the beginning of 2000 an analysis of FRBR requirements. The work resulted in a set of considerations which have been discussed, and some of them are still under discussion, first of all with Tom Delsey, to whom the SC expresses its deep and sincere gratitude for the assistance, support, and consultancy supplied during the various phases of the work.

The paper has subsequently been analysed with the representatives of the Italian Libraries Association – AIB. The work is still in progress and both the Association and the SC will be cooperating, on this as well as on other points of the revision. The SC would also like to have Archives and Museums representatives involved in the analysis and in the discussion.

To conclude, we would like to make available the work done up to now, which should be considered for some points still in progress, available for a wider discussion at international level. It seems very likely that if other countries are planning, or are in the process of a revision of their respective cataloguing codes, incorporating FRBR provisions, the analysis developed could help in finding out possible common solutions and would permit exchanging a varied range of experiences.

1. Foreword about expression

The analysis put forward up to now by the RICA SC both theoretically, according to the entity-relationship model devised by the FRBRlogic, and practically, through the development of some examples, within the frame provided by the model, has emphasized, side by side with its undeniable innovative aspects, some general problems that we thought it useful to summarize. One of this is the question of balancing cost reductions (achieved through minimal level descriptions) with improved accesses (achieved through reflecting in greater depth the relationships between works and expression and the persons responsible for their creation or realization), It seems that the two goals do pull in opposite directions and for this reason, that each cataloguing agency has to determine how much it is prepared to invest in the cataloguing of an item. What FRBR tries to do is to provide a better understanding of the data that may be included in the catalogue record and to give some indication of the relative functional value of those data elements (or more correctly, of the attributes to which the data correspond) so that libraries can make more informed decisions about where to focus their investments.

We deem it opportune, before starting a thorough examination of such problems, to remind briefly some points, which appeared to be particularly difficult to be approached, and for which the SC has provided for necessary clearance through a direct exchange of opinions with one of the consultants responsible for the development of the FRBRmodel, Tom Delsey. The points, listed below, are mainly concerned with the entity expression .

  • The “realization” of a work represents an expression, indipendently by its resistration (i.e., the extemporary delivery of a speech, and the improvisation of a composer at the piano would qualify as an expressions of works, regardless of whether they were recorded or not).
  • We normally deal with recorded expressions, and as a result, the evidence we have of an expression is normally in the form of a manifestation, or more precisely, in the form of an item exemplifying a manifestation. In practice, the distinction between the manifestation and the expression is often blurred, because of the common practice in cataloguing to treat most manifestations as though they contained only one work.
  • To clarify the concept of expression, and the distinction between it and work, on one hand, and manifestation on the other, it is helpful to look at the attributes and relationships associated with expression: form, date, language of the expression are all associated with the actual “realization” of the work, as distinct from the “conceptualization” of the work. Similarly the relationships defined between expression and person are those associated with the “realization” (editing, performing, etc.), as distinct from those associated with the “conceptualization” (composing, etc.).
  • The model is structured so as to permit the representation of aggregate and componend entities in the same way as entities viewed as integral units are represented. That is, an aggregate work or a component of a work may be substituted for the entity labelled as work in the diagrams, and the same kind of substitutions may be made for aggregates and components of expressions and manifestations.
  • In thinking now to what all this might mean in terms of cataloguing rules, it helps to look first at how the attributes of the expression have been reflected in conventional cataloguing records. In general, catalogue records (as distinguished from bibliographic records emanating from the disciplines of textual criticism and analytical bibliography) tend to reflect only the more obvious distinctions between expressions (e.g., the distinction between the textual form of a work and that same work expressed as spoken word in an audio recording). In more recent practice, distinctions of that kind are sometimes reflected (indirectly, at least) in a fairly prominent position in the description, through the general material designation. In other cases, those kinds of distinctions (related to form of expression, language of expression, etc.) are recorded simply as notes. Certain attributes of the expression (e.g., language of expression) are reflected in additions to uniform titles, where they become part of what FRBR refers to as organizing elements, as distinct form simple descriptive elements.
  • As FRBR points out (in section 3.2.2), the degree to which bibliographic distinctions are made between variant expressions of a work will depend to some extent on the nature of the work itself, and on the anticipated needs of users.
  • Regardless of how the distinction between one expression and another is reflected in the catalogue record, it can only be reflected through the attributes that are associated with the entity expression as defined in the model (i.e. form, language, etc.). Attributes associated with either the work or the manifestation per se are of no value in distinguishing the expression. It should be noted, however, that because catalogue records tend not to differentiate between expressions on the basis of less obvious differences, such as textual variants, there is no guarantee that two expressions are in fact the same simply because the one or two attributes noted for each correspond.
  • In answer to the question about determining which manifestations embody the same expression, it follows that the expressions embodied in separate manifestations can only be considered to be the same if the values for the attributes and relationships that are associated with expression as an entity are the same (e.g., if the language of expression is the same, and the relationship to the person responsible for translation is the same).
  • The title and the author relationship, however, operate at a different level, inasmuch as they indicate (though not always reliably) that the expressions, whether the same or different, are at least expressions of the same work.
  • The primary value of recognizing expression as an entity within the model is that it focuses attention not only on the attributes of the expression, but perhaps more importantly that it highlights the relevance of the relationships between the expression and the person(s) responsible for the expression.

2. From “authors”’s to "works and expressions"’s cataloguing?

RICA SC has noted that while, at first sight, the main new structural element of the FRBR model is represented by the introduction of the entity expression, the whole model appears to give great relevance to the identification and distinction of all the first group’s entities, from the work to the item level, i.e. to the entities generally named with titles.

On the contrary, authors’ cataloguing – by its same definition - has traditionally concentrated its attention more on the responsibilities of literary and not-literary products, and on their functions. This is a typical and partly compulsory choice made by cataloguing codes still in use, due to the not flexible context offered by card catalogues. The choice is also derived by the fact that, lacking the possibility of using other elements of selection now offered by automated systems, works, often with not distinctive names/titles, would have created more problems for their identification than authors. Only through automation and the possibility of combined searches (author + title) it is possible to recognise to the entity title the same value assigned from ever to authors. Up to now a particular attention has been dedicated to develop tools for defining and managing standard forms representing names of persons and bodies, while titles (as far as designating literary/textual entities) have been for a long time considered mainly as a sort of "surrogate" for names of authors; as a consequence, tools for defining and managing their standard forms have been developed more slowly and in a partial and limited way.

Since long ago modern cataloguing rules consider it a usual standard the fact that each author should represent a well identified entity, to which corresponds a proper authority registration, linked to the bibliographic record not as a simple attribute of this last one. This happens notwithstanding empirical inquiries demonstrated that, generally, the majority of authors registered in a catalogue of considerable dimensions are linked to only one publication and do not present variant forms of the name. This means that managing all authors through an entity-relation model it is unanimously considered to be the more opportune and efficient solution, notwithstanding the fact that only in a minority of cases the author is acting as a grouping entity and the fact that not always there is a real need for creating relations between more than one name or forms of a name for the same author. On the contrary, for titles we commonly accept the rule according to which only some titles are registered as an authority entity, i.e. are controlled as autonomous entities.

This evident difference of attitude is framed in a general structural asymmetry: from a logical point of view, the relation of responsibility should link the registration of the entity author (e.g. Dante Alighieri) to the registration of the entity work (e.g. Divina commedia) and therefore the registrations of the various editions will be consequently linked to this last one. On the contrary, in the major part of the systems the registrations of the entity author are linked directly to those of the single editions, or in some cases, as for example in SBN, they are linked to both levels (to the work, represented by its uniform title, and to the edition).

Moreover, it important to note the following points, at present:

  1. cataloguing rules are considering optional the analytical description of the content of a single publication: so it happens for the description of works contained in a collection or issued as subsidiary of a principal work; in this way not all the works of an author, contained in a publication, are registered but only those presented or recognised as the main work;
  2. for what concerns works, with or without an author, registered in a catalogue, both filing systems, the one based on the title proper and the other one on categories, do not completely fulfil Paris Principles’ requirements for which the same work has to appear in the same place of the catalogue.

Consequently, in searching by authors, catalogues are usually presenting an orderly sequence of specific editions (more than the works of an author) or an orderly mixture of single editions and of their classes (as works, represented by uniform titles). On the other hand, an analytical and rigorous development of relations/accesses, as required by the FRBR model, does not appear realistically applicable if not in computerised files.

Undoubtedly, in the FRBR model titles are acquiring weight and importance ever since recognised to them. The key of the whole model is that it has been developed mainly with the scope of creating a correct relation between each author and his/her work, or expression of the work, to the realisation of which he/she has contributed (as translator, editor, etc.).

3. FRBR model’s innovations with regard to its application to a catalogue

The most innovative points of the model are, in our opinion, the following.

  1. It provides for responsibility relations to be, exactly and correctly, created/linked to the proper entity, be it a work, an expression or a manifestation, i.e. with the title which specifically represents them. We could briefly say: “for each author its title”.
  2. The content of items is firstly analysed from a textual point of view, distinguishing its integral or separate and separable components, without making, at least as a principle, any difference between the “primary” and the complementary or subsidiary ones (i.e., introductions, appendixes, comments, notes, illustrations, etc.), all bound to the be analysed in the same way in terms of work, expression and manifestation.
  3. Consequently, a neat distinction is made between responsibilities referred to a specific (separable) component and those transversal, realised in giving to the work a specific form (expression), through the action of an editor, translator, performer, etc.

4.For what concerns transversal responsibilities the link is made with the expression, for the separable ones with the work.The more accurate attribution of responsibilities for each entity of the first group (work, expression, etc.) means a diminished importance of the distinction between principal and secondary responsibilities and a shifting of the point of view: from the attribution of the principal responsibility of a publication to the individualization of the principal component of the same, in the case we wouldn’t want to catalogue each one of its components (as a matter of fact we are not bound to individuate the principal component as such: each author is the principal one in respect to his/her work or to the specific realization of this in an expression).