/227

Manifestations and Near-Equivalents: Theory, with Special Attention to Moving-Image Materials

Martha M. Vee

Differences betlceen manifestations and near-equivalents that might be con­ sidered significant by catalog users are examined. Anglo-Amelican catalog­ ing practice concerning when to make a new record is examined. Definitions for manifestation, title manifestation, aud near-equivalent are proposed. It is suggested that current practice leads to making too many separate records for near-equivalents. It is recommended that practice be changed so that near-equivalents are more often cataloged on the same record. Next, differ­ ences between manifestations and near-equivalents of moving-image works are examined, and their Significance to users of moving-image works is assessed. It is suggested that tme manifestations result lchen the continuity, i.e., visual aspect of the work, or the soundtrack, i.e., audio aspect of the work, or the textual aspect ofthe work actually differ, whether due to editing, the appending of new material, or the work ofsubsidiary authors creating subtitles, neu; music tracks, etc. Title manifestations can occur lchen the title or billing order differs lcithout there being any underlying difference in continuity. Distribution infonnation can differ without there being any underlying difference in continuity, creating a near-equivalent. Finally, phYSical variants ornear-equivalents can occur when phYSicalfonnat differs without the involvement ofsubsidiary authors.

A manifestation of a work is a version differences that might be considered sig­ or edition of it that differs Significantly nificant by catalog users, and the ways from another version or edition. A near­ these differences have been handled by equivalent is used here to mean a copy of Anglo-American cataloging rules. the same manifestation of a work that Two types of users will he considered: differs from other copies in ways that do the general user, who is assumed to be not Significantly affect the intellectual or interested only in Significant differences artistic content. In this article (excerpted in the intellectual or artistic content of a from Yee 1993) I will discuss the kinds of work, or in Significant differences in the

MAHTIIA M. YEE is Cataloging Supervisor. University of California. Los Angeles Film and Television Archive. The author wishes to acknowlt'dge the assistance of Eric Aijala, Bob Epstein and Michael Friend. Manuscript received Novemlwr .'30,199.3; accepted for publication Febru­ ary 20, 1994: revised April 1, 1994. . 228/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /229

citation ofthe work; and other users inter­ HISTORICAL AND CURRENT PRACTICE definition had been devised that could be over of a long practice in book catalogs of ested in more minute differences, such as applied to nonbook materials not pro­ describing subsequent editions of a work the bibliographer-user, who might be in­ The practice ofcreating a new record for duced by the setting of type. with dashes to represent all elements of terested in physical evidence of the print­ each new edition of a work goes back to The current Anglo-American defini­ the description that are the same as in ing and publishing history ofa work, orthe the beginning of the use of unit records. tion reads (ALA 1988, 617): previous editions; the British Museum preservation officer, who might be inter­ Jewett's rule IV reads in part, "The whole (Books, pamphlets, fascicles, single sheets, catalog is perhaps the most readilyavail­ ested in binding or paper of differing title is to be repeated, for every distinct etc.) All copies produced from essentially able example ofsuch a book catalog. The qualities. Indeed, a theme running edition of the work" (Jewett 1852). Note, the same type image (whether by direct use of dashed-on entries is evidence of a through the article will be that of the however, an eqUivalent rule concerning contact or hy photographiC or other desire not to confuse users with multiple differences and similarities between bib­ the object of a cataloging record is never methods) and issued by the same entity. entries for nearly identical items; in a lilli' liography and cataloging. stated expliCitly in any published Anglo­ (Computer files) All copies embodying es­ dashed-on entry, only the significant dif­ Under current cataloging practice, the American cataloging rules. sentially the same content and issued by ferences in a new edition or issue are question of what is a manifestation of a The development ofa definition ofedi­ the same entity. noted, and the inclusion ofthe dashed-on tion in Anglo-American and international III work is essentially the same as the ques­ (Unpublished items) All copies made from entry on an existing record quickly and tion of what is the object of a cataloging cataloging codes demonstrates an attempt essentially the same original production concisely indicates to the user the degree record. Note in this connection, however, to come to terms \vith technological (e.g., the original and carbon copies of a to which the two items are identical. i that several writers (Wilson 1989, 9; change from the printing of books by the typescript). The use of dashed-on entries is also Layne 1989, 192-93) have proposed setting of type to the production of many (Other materials) All copies produced evidence of a desire to create fewer re­ Iii work-based records; Hinnebusch (1989) different kinds ofworks, including books, from essentially the same master copy cords and to Simplify cataloging. Further has proposed devising hierarchical by means of the many new methods of and issued by the same entity. A change evidence ofthe latter is found in the 1949 MARC records; Attig (1989) has dis­ duplication and reproduction that have 1II I in the identity of the distributor does not rules, which state, "To distinguish the 1 cussed the difficulty of linking MARC exploded into being in the course of the mean a change of edition. various issues of a given edition, any of a :1 unit records; and Yee (1991, 81) has dis­ twentieth century. Cutter's fourth edition I The latest ISBD(M) contains the fol­ \vide variety of details might need to be cussed the possible value ofmatching key­ contained the following definition of ecli­ lowing definition ofedition: "All copies of specified. However, atthe Library ofCon­ words in user-input, Imown-item searches tion: "A number of copies of a book, pub­ a publication produced from substantially gress it is not the policy, except in certain II!; on online catalogs against the set of rec­ lished at the same time and in the same the same original input and issued by the cases of rare books, to collect the various ords that make up a work. form. Alater publication ofthe same book same agency, whether by direct contact or issues of a given edition and consequently f A good deal of what follows will con­ unchanged is sometimes styled a different by photographiC orother methods" (IFLA no attempt is made to describe works in cern (1) the kinds of differences between edition, sometimes a new issue, some­ Universal BibliographiC Control and In­ detail sufficient to identify them as issues. I a document being cataloged (hence­ times a different thousand (4th thousand, temational MARC Programme 1987, 3). Various issues are added to the collection forward to be called an item) and docu­ 7th thousand)" (Cutter 1904, 19). The The fact that the definition no longer re­ as copies ifthe description ofthe first one ments that have already been cataloged 1908 rules were the first Anglo-American fers to the settingoftype seems to indicate cataloged fits those received later in all and are represented by surrogate records cataloging rules to adopt the bibliog­ an attempt to recognize the fact that cata­ details or in all details except tlle imprint in the database ofrecord, which can cause raphers' definition ofedition: "The whole logers have rarely been able to examine date or the form of the publisher's name the item being cataloged to be considered number of copies printed from the same and compare type settings or type image, or both. Ifthere are other differences, the a new manifestation, requiring a new rec­ set of types and issued at the same time" and that in fact they have relied on evi­ issues are generally treated as different ord, and (2) the kinds of differences that (American Library Association 1908, xiv). dence on title pages and preliminaries, editions" (LC, Descriptive Cataloging Di­ are felt to be so minor that the item can In 1941, the definition was changed so as and on paging or other extent measure­ vision 1949,9). be treated as a near-equivalent, which can to remove the requirement that the copies ment, to determine when two items were From 1949 forward, less and less em­ be described on a record that already be issued at the same time, in order to two different editions of the same work. phaSiS is placed on the distinction be­ exists. accommodate printing from stereotype or Dorcas Fellows, in 1915, described the tween issues and editions. The 1949 rules Historical and current practice will be electrotype plates ( ALA Catalog Code cataloging practice ofaddingeditions sub­ were the first to use the term item, as examined. The small amount of previous Revision Committee 1941, xix). sequently acqUired by a library to the card opposed to the more specific terms edi­ research on the question of the most reli­ This definition remained essentially for the first edition acqUired, using tion or issue, when referring to the object able \isible indicators of difference in unchanged until 1974, when the first of dashed-on entries (Fellows 1915, 132­ of a cataloging record: "The objectives of manifestation will also be desClibed. The the ISBDs ISBD(M) appeared, defining 37). This mention in 1915 might indicate descriptive cataloging are (1) to state the question of appropriate record-structur­ edition as "all the copies of a publication that the use of dashed-on entries was significant features of an item \vith the pur­ printed from one setting of type, or pro­ ing techniques to express differences be­ practiced in the construction of card cata­ pose of distinguishing it from other items tween items will be considered. Finally, cluced from one master copy, and issued logs prior to its fornlal introduction into and describing its scope, contents, and bib­ definitions of the following \vill be pro­ by one publisher or group of publishers. AACR in 1967, which allowed the more liographic relation to other items ...." (LC, I (An edition may comprise several impres­ posed: manifestation, title manifestation, limited practice of dashed-on entries for sions or issues, in which there may be Descriptive Cataloging Division 1949,7). and near-equivalent. First, historical and different issues of a given edition or for current practice will be examined in gen­ slight variations)" (lFLA Committee on The term item is a neutral and ambiguous reproductions (ALA 1967, 22.5-26). This eral terms. C~taloguing 1974,2). For the first time, a term that allows fleXibility in determining practice may be cor.sidered to be a hold­ what, in fact, to make the object of a II' :1

1 I 230/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /231

record. Itwas first defined, somewhat cir­ rum 1990). The library community has ing codes have incorporated definitions whether their definition of edition corre­ cularlv, in the ISBD(G) as "a document, recently limited the Single-record ap­ for edition similar to that used by bibliog­ sponds to the needs of catalog u~ers has group of documents, or part of a docu­ proach to one type ofnear-equivalent, the raphers. Even though, as noted above, not yet been considered. Since catalogers ment, in any physical form. considered as reproduction (ALA, Task Force on Multi­ there is a trend away from the mention of have been unable to implement the bibli­ an entity and forming the basis ofa single ple Versions 1992). The tendency toward settings of type, the current definition of ographically accurate definition ofedition bibliographic description. The term docu­ reducing the number ofnew records cre­ edition for books in AACR2 still refers to carried in their glossaries all these years, ment is used here in its widest sense" ated seems to be in conflict. however, with type image: "In the case of books and they probably have been unable to satisfy (IFLA International Office for UBC the desire to simplify cataloging by teach­ booklike materials, all those copies of an the needs of those catalog users who are 1977, 2). The current Anglo-American ing cataloging staffs to make a new record item produced from substantially the bibliographers and textual scholars. For definition ofitem is "a document or set of any time there is a difference in the pub­ same type image, whether by direct con­ example, William B. Todd (1981, 48) documents in any physical form, publish­ lication statement, without discriminat­ tact or by photographiC methods" (ALA writes: ed, issued, or treated as an entity, and as ing, for example, between distributors 1988, 617). In practice, however, Anglo­ Without further analysis one may readily such forming the basis for a Single bibli­ and publishers, or among various types of American catalogers do not in fact carry accept a report. from a major research ographic description" ( ALA 1988, 619). date change. out textual comparisons to determine library, that through 1955 Melville's Moby The use of the concept of item might whether two items represent the same Dick ranged through 118 "editions." Upon edition. Itis very unusual for the cataloger proper investigation, however, one must represent a backing away from legislation TYPES OF DIFFERENCE BETWEEN of an item to look at more than its title conclude, with G. Thomas Tanselle, that I in the cataloging code itself on what the MANIFESTATION OR NEAR-EQUIVALENT object of a single description should be. page, preliminaries, overall paging and di­ all these NUC entries actually make up Both the Library of Congress and the Now that the historical context has been mensions, and any readily available cata­ only thirty-five editions. OCLC Online Computer Library Center, established and current practice has been loging records that might serve as surro­ Rather than dwelling on this gloomy Inc., have published fairly elaborate defined, the types of difference that can gates for other items tllat are candidates fact, however, one should consider the guidelines, which differ from each other, occur from one manifestation to another for representing the same edition. In more cheerful pOSSibility that perhaps to enable the cataloger to decide when to or from one near-equivalent to another other words, the Anglo-American cata­ current practice does meet the needs of make a new record (LC 1990; OCLC will be considered. loger is dependent on title page repre­ many non-bibliographer catalog users. 1993, 37-49); both sets of gUidelines at­ sentation or representation elsewhere in That is, the pOSSibility should be consid­ the preliminaries of a work in making de­ ered that differences in title page repre­ tempt to identify differences on title DIFFERENCE IN TITLE PAGE sentation, while they might not necessar­ pages andl)reliminaries or in phYSical de­ AND ITS CONNECTION WITH cisions about whether two items are cop­ ily reflect actual differences in the setting scription that should be considered so mi­ DIFFERENCE IN TEXT ies of the same edition or two different nor that it is not necessary to make a new editions. But title pages and preliminaries of the type beneath, might nevertheless record. These will be considered in more Catalogers give much more weight to a are not always reliable evidence. In some correspond to differences in citation and detail in the next section. difference in title page than do bibliog­ cases textual comparison of a number of searching practice on the part of those In summary, a historical review seems raphers. For bibliographers, the term edi­ different items, recorded in the form of catalog users who are not bibliographers. to reveal a trend away from use of the tion is used quite strictly to mean "all elaborate collations, is necessary to deter­ In practice, bibliographers and rare bibliographers' strict definition of edition copies resulting from a single job of typo­ mine whether two items are copies ofthe book catalogers create two records for two to a definition that takes into account graphical composition" (Tanselle 1975, same edition, in the bibliographers' strict items in the same edition with different newer technologies for duplication, re­ 17) regardless of differences on the title use ofthe term edition. Catalogers cannot title pages just as catalogers do. For exam­ 1'1 production, and distribution of works. It page. It has long been recognized that two afford to take the time to create elaborate ple, in the Eighteenth-Century Short copies of the same edition, using the bib­ I, I also seems to reveal a reluctance to legiS­ collations for current publications. Some Title Catalog Project, the object of a rec­ I late in the cataloging codes on the issue of liographers' definition, can have different rare book collections can afford to create ord is not just an edition, but an issue or the object of a single cataloging record. title pages. Bibliographers generally refer elaborate collations, but even in those col­ an impression. The real difference be­ And finally, it seems to indicate an attempt to such copies as issues or states (Gaskell lections, catalogers often are not able to tween catalogers and bibliographers lies to devise methods to cut down on the 1972,315-16). As far back as 1876, Cutter assemble all the copies usually necessary in the degree to which bibliographers number of cataloging records created to mentioned that such issues or states \vith to accurately clasSify editions, because, as clarify the relationships between issues describe the various issues, variants, im­ different title pages were referred to by Tanselle points out, "to establish such and editions. and use the terms according pressions, and reproductions ofan edition the Germans as title-editions (Cutter facts demands recourse to copies outside to their technical definitions, as compared of a work-in other words, to avoid creat­ 1876, 61, rule 135; Cutter 1904, 19; see the collection" (Tanselle 1975, 17). The to the tendency on the part of catalogers ing new bibliographic records to describe also LC, Processing Dept. 1946). It has question arises, then, whether catalogers to Simply report how an item describes minor variations between items. Perhaps long been recognized that the reverse can should retain the bibliographers' defini­ itself as to edition, whether the publisher the recent Multiple Versions Forum could be true: that two different editions, that is tion ofedition when thev do not have the has been bibliographically accurate or be seen as an attempt to respond to these two different settings of type, can be resources to identify and distinguish edi­ not, and to report differences in title trends; here the recommendation was masked by identical title pages (Le, Proc­ tions to that degree of accuracy. pages, but not differences in collation be­ made that near-equivalents be cataloged essing Dept. 1946; Blanck 1966; Jolley So far, only the question of what cata­ yond changes in primary paging. ("The on one record, using the new USMARC 1961, 10). logers can reasonably hope to accomplish First Phase" 1983, 11; Alston 1981,381). holdings format (Multiple Versions Fo- Since 1908, Anglo-American catalog- has been considered. The question of There are several ways title page I 1111

232/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /233

I series can be ignored, and an existing rec­ representation can vary, and distinctions when the only difference between two photoreproduction processes available ord that lacks the series, but is suitable in should be made. In cataloging most items, items was in the publication date, but this for all types ofmaterials make this pattern other respects for an item in hand, can be the following elements are transcribed provision was dropped later in the same of distribution widespread. Hagler de­ when present: title and statement of re­ used (OCLC 1993,47). However,OCLC's year (LC 1981, 3; LC 1990, 10). OCLC scribes the publishing practice of replac­ record matching algorithm does match on sponsibility, edition statement, imprint allows the cataloger to ignore differences ing ilie printed imprint on a title page by the series (O'Neill 1990, 11). (called publication, distribution area in in printing, manufacturing, distribution, a label or stamp for the U.S. or Canadian the Anglo-American CatalogUing Rules, or copyright dates, but not in publication publisher or distributor (Hagler 1963, second edition [AACR2]), and series. Edition and Imprint dates (OCLC 1993, 46). Wanninger 342). Several writers point out that it is not A number ofwriters over the past century points out tllat many duplicates are cre­ uncommon for issues of the same edition Title Proper and Series Title or more have noticed that certain differ­ \Vhen two items have different titles, one ences on title pages have more to do with ated in OCLC because separate records of a work to be issued with one imprint in \vith different reproduction dates are cre­ can make a good argument for creating indicating continuing availability ofa par­ England and another in tlIe United States separate records for each, even if they are ticular manifestation, rather than with any ated for photocopies and microfilms pro­ (McNellis 1985, 36; McPherson, Coyle, not two different editions in the bibliog­ difference in the copies of the manifesta­ duced on demand by University Micro­ and Montgomery 1982, 376). Changing rapher's use of the term; in fact, this has tion available. For example, publishers films International and the National distributors are particularly common with Technical Information Service (Wannin­ nonbook materials (Fothergill and always been standard cataloging practice. change dates and edition statements with­ ger 1982). Edward O'Neill found that The title is so important in citation prac­ out changing the setting of the type, to T. Butchart 1978, 180; Thaxter 1983, 19). In "the date of publication element, indi­ tice tllat it is felt to be wise to record on indicate that in the new year the work is fact, the definition of edition for nonbook vidually and in combination with other materials in AACR2R (based on ISBD) separate records all the different titles still available from the indicated publish­ fields, 'was responSible for the greatest indicates that "a change in the identity of under which a particular work has been er. For factual works, motives might be number of duplicate records," in a study the distributor does not mean a change of issued. Another way of stating this argu­ more unscrupulous, implying the work of duplicate records in tlIe OCLC data­ edition" (ALA 1988, 617). Unfortunately, ment is to say that issues or states with contained is more current than it is. base (O'Neill 1990, 11; O'Neill, Rogers, this particular provision ofAACR2 has not different titles on their title pages should Jewett noted the follOwing phenomenon and Oskins 1993). He also found "the edi­ been put into practice; OCLC, for exam­ be given separate records even though in 18.53 (140): tion statement, in combination with other ple, requires making a new record when separate records are not normally made It is frequently the case, that publishers, fields, was responSible for the highest per­ the distributor changes (OCLC 1993,45); for different impressions, issues, or states. after haVing stereotyped a hook, call every centage of duplicate records relative to as a result, OCLC contains numerous re­ Besides the importance for matching us­ thousand copies of it a separate edition, the number of records in the sample in cords for videocassettes of the same film. ers' citations to catalog records, another and, for twenty or more editions, there which it was present." Is it really necessary to create separate argument in favor of making separate rec­ may he no alteration in the hook, except in Sometimes two items are identical ex­ catalog records to record variation in date ords for title differences is that the rec­ the word expressing the number of the cept for variations in the name of the of issue, copyright date, or printing date; ords of these differences are of historical edition. and in the date. In such cases, it pu1)lisher. The 1949 mles allowed two or variation in distributor; or variation in interest in themselves; they could enable cannot he necessary to print a separate items to be described on the same record the name of the publisher, when there is historians and other scholars to trace the title for each pretended edition. if the only difference between them was history of a particular work, including the Differences in the various dates that no reason to suspect that the variation is in the form of the publisher's name (LC, associated with an actual difference in the various titles it has borne. appear on title pages, i.e., date ofpublica­ DeSCriptive Cataloging Division 1949,9). One (.'(mld extend the same argument tion, copyright date, and printing date, are intellectual content? Catalog records are The current LCRI distinguishes two cases: to cover differences in series titles. Series notorious for not reflecting an actual dif­ expensive to create. They are also expen­ (1) variant forms of name used concur­ titles might sometimes warrant less bibli­ ference in edition. OCLC has six records sive to maintain in large databases. Most rently by the publisher, in which case two importantly, multiple catalog records for ographic respect than title proper, how­ for Smollett's The Expedition ofHumphry items can be described on one record; and virtually identical items confuse users, in­ ever. A series title serves the dual func­ Clinker published by Century, all with (2) actual change in name ofthe publisher, cluding library staff, such as copy catalog­ tions of being (1) a unifYing principle for identical paging, the only differences be­ in which case two items must be given two ers and interlibrary loan assistants. It a number of intellectually related works, ing publication dates ofl902, 1903, 1904, separate records (LC 1990, 10). OCLC takes the user a long time to sort through and (2) a marketing tool for the publisher. 1905,1906, and 1907. One can be virtually allows the cataloger to ignore variation in a large retrieval set. Any given search is Sometimes a series title performs very certain that these are all the same edition, fullness ofpublisher's name (OCLC 1993, likely to bring up larger numbers of re­ little of the former function and a great but different issues with different dates, 45), although its machine matching algo­ cords than if these near-equivalents were deal of the latter. The Library ofCongress in order to indicate continuing aVailability. rithm probably would not do so (O'Neill The 1949 rules allowed two items that weeded out. The fact that multiple rec­ Rule Interpretation (LCRI) on when to 1990,11). make a new record indicates that a series differ only in imprint date to be treated as ords exist is likely to blind users to the fact title that is associated with just the soft­ copies and cataloged on the same record, Beginning in the 19th century, the use that a number of items listed separately bound or just the hardbound manifesta­ unless copyright date varied as well (LC, of stereotype plates and electroplates for are virtually interchangeable for most tion of a Cataloging-in-Publication title DeSCriptive Cataloging Division 1949, 9), printing made it possible for the same purposes. Differences in distributor and can be ignored (LC 1990). The OCLC When the current Library of Congress edition in the bibliographers' sense, Le., issue date that are unconnected to any I the same typesetting, to be issued by sev­ differences in the content are likely of I rules for when to make a new record indi­ rule interpretation was originally written cate that any difference in name of the in 1981, it forbade making a new record eral different publishers or distributors interest only to someone who would like (Steinberg 1974, 2i8-9). Now the various to acquire a copy of a particular • ! I 234/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /235

I manifestation at any given point in time. liographies, commentators, illustrators, ence in the paging of a book is a sure sign have traditionally used holdings state­ Surely sources such as Books in Print are etc. Manifestations that themselves have the type has been reset. Resetting of type ments, dashed-on entries, or other similar more appropriate and more up-to-date for manifestations do not always have subsidi­ can easily lead to either intentional or techniques to communicate such infor­ this kind of information than catalog data­ ary authors, however. Sometimes a single unintentional alteration of the text. Re­ mation rather than asserting that such a bases can ever hope to be. There is one author or other creator can create several search done at the Library of Congress in difference created a new edition, requir­ caveat, however: users who need to find a different versions, or manifestations of 1946, to be discussed further below, ing a new record. The practice of adding particular edition because they have a ci­ one of his works, each of which can then clearly demonstrated that for books, pag­ reproductions to existing records by tation to a particular page number would go into several manifestations. Ravilious ing was the most reliable clue for detect­ means of dashed-on entries was actually benefit ifvariations in distributor and date mentions the "1919 version ofStravinsky's ing differences in edition, i.e., resettings codified in AACR in 1967 (ALA 1967, were recorded as near-equivalent-specific Firebird Suite," and Whiting mentions of type. A difference in the paging of a 22.5-26). The Library of Congress does III variations, so that they could be assured Wordsworth's versions ofthe Guide to the blXlk might well be ofinterest to even the not explicitly address the question of dif­ ii, I they have found the manifestation with Lakes (Ravilious 1975,47; Whiting 1980, general user. As has been said, it might ference in phYSical fornlat in its rule inter­ the paging they seek. 5). Thus, several writers have suggested indicate significant alteration of the text pretation for AACR2 TIlle 1.0 (LC 1990, IIIII that a conceptual level between work and itself. However, even if the resetting of 10); in practice, however, as long as Proc­ manifestation is needed, perhaps to be the type, which creates the new paging, essing Services at the Library ofCongress DIFFERENCE IN RESPONSIBILITY OR called version (Domanovszky 1974, 102; was cataloging audiovisual materials for OTHER DIFFERENCES SUBSTANTIAL has not altered the text, the user might I, Du Rietz 1974, 84; Richmond 1980, 33; need to find the correct manifestation in the libraries of the nation (a service no ENOUGH TO CREATE A VERSION Shineboume 1979,240; Whiting 1980,5). order to look up a citation to a particular longer provided by LC), the LC catalogers ill' Sometimes a manifestation can have its Barbara Tillett refers to the types of made a new record only for a difference I page number. own manifestations. Panizzi recognized versions described here as having either in general material designation (motion this in his rules for the arrangement of derivative relationships to their parent picture versus videorecording), and sum­ various manifestations under an author. works, or deSCriptive relationships. "De­ DIFFEHENCE IN PIIYSICAL FORMAT marized all videorecording formats avail­ For example, a particular translation of a rivative relationships are those [that] hold Sometimes the only difference between able on one record (Tucker 1982). The particular work could itselfgo into several between a bibliographiC item and a modi­ one item and another is a difference in archival mOVing-image catalogers at LC editions. Thus Panizzi's rule LXX read, fication based on that item.... One item physical format. This can be due to repro­ attach both videorecordings and motion "Editions by the same editor, or such as is derived from another when it enlarges, duction, in which a copy of an original is pictures to one record if the only differ­ are expressly stated to follow a specific abridges, or otherwise modifies the entire made for preservation or conservation ence is in phYSical format. OCLC, on text or edition, and editions \vith the same item or portions of it" (Tillett 1987, 43). purposes, or to make it available in an­ the other hand, encourages the creation notes or commentary, to succeed each However, she includes adaptations in this other useful format. It can also be due to of two records if there are differences in other immediately in their chronological category, while adaptations are generally simultaneous release in more than one the physical description between one order after the entry of that which is, or is treated as new works rather than as mani­ format, in order to reach different mar­ item and another (OCLC 1993,46-47). considered to be, the earliest." festations. She also includes editions, in kets. Examples would be a microform of The probable explanation for this ap­ Rule LXXII, dealing with the arrange­ the sense of resettings of type without a text, CD and audiocassette releases of a proach is that one of OCLC's primary ment of translations, reads, in part, differences in subsidiary authorship, sound recording, or a videocassette copy goals is to support interlibrary loan, in "Translations into the same language, and which have been discussed above. "A de­ of a motion picture. Sometimes reproduc­ which a potential borrower needs to their several editions, to be entered in scriptive relationship holds between a tions are made on a one-time basis either know precisely what the phYSical format conformity with the rules laid down for bibliographic item [or] work and a de­ by a particular institution for preservation of the item to be borrowed is. Since the the entries ofthe originals" (Panizzi 1985, scription, criticism, evaluation, or review purposes, or by an on-demand reproduc­ utilities have never had a holdings for­ 11). This kind of grouping together of all of that item or work, such as that between tion agency such as University Microfilms mat to allow the communication within the manifestations of a manifestation was an item and a book review describing it" International or the National Technical a Single record of information about never attempted in card catalogs, but one (Tillett 1987,57). Although most items in Information Service. Other times multi­ which formats are held by which institu­ wonders whether it could be done in on­ this category are new works about other ple copies or reproductions are issued and tions, their only recourse has been to fall line catalogs. works, Tillett includes here editions with made available by a reproduction/distri­ hack on encouraging the creation of a The "manifestations of manifesta­ commentary, which are sometimes bution agency. In any case, the purpose new bibliographiC record for every vari­ tions" under discussion tend to exhibit treated as manifestations of the same for reproduction or simultaneous release ation in phYSical deSCription. Now that authorship connected with the manifesta­ work, depending on the circumstances. in several formats is to produce a surro­ a US MARC holdings format exists, with tion rather than with the work itself. The gate for the item reproduced in order to a repeatable 007, the opportunity exists kind of authorship that can change with­ DIFFERENCE IN EXTENT make it more widely or readily available. to develop a hierarchically structured out causing change in the work itself has Certainly difference in phYSical format is single record to show differences in been called subsidiary authorship since It has long been recognized that differ­ of interest to users and should be commu­ physical format. A hierarchically struc­ the introduction ofthe ISBDs (ALA 1974, ence in extent, such as difference in the nicated to them, but whether it is neces­ tured Single-record approach would cut 24). Examples of subSidiary authors are paging of a book, can be with some fre­ sary to create a completely separate bibli­ down on repetition ofbibliographic data editors, translators, authors of introduc­ quency the only reliable clue that two agraphic record to communicate this when all that is different is the phYSical tions or notes, compilers of attacht'd bib­ items are significantly different. Differ- difference is open to question. Libraries format. II • II 236/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /237 I An unresolved difficulty with a hierar­ communicate the differences might be to at OCLC. O'Neill reports their current (Tillett 1987). Unfortunately only those chically structured record is that ofdecid­ make one record for both items, with re­ algorithm has a precision of .93, if the relationships that are revealed by way of ing what to describe primarily when the peated phYSical descriptions, rather than record similarity is setto.9 (O'Neill 1990, either explicit edition statements, expliCit item reproduced is not in hand, or when creating two full bibliographic records 13-14); in other words, 93% of the iden­ notes by the cataloger, or USMARC for­ no one item has primacy, as in the case of that differ onlyin the phYSical deSCription. tified pairs were duplicates, and presum­ mat coding were studied, and even those simultaneous release of several formats. This approach is more economical for da­ ably seven per cent of the pairs identified with explicit notes were only sampled. The two-tiered approach advocated by tabase managers to the extent that the cost as duplicates were not. However, the ac­ Differences between manifestations that the Multiple Versions Forum requires of managing large databases is increased companying recall was only .51; in other are revealed indirectly, for example, by a that one item be designated primary with by having to store, retrieve, and arrange words, the algorithm identified only 51 % statement of subSidiary authorship in the its phYSical description given in the first large numbers ofnear-equivalent records; of the duplicates in the sample. Unfortu­ 245 field or by other such implicit indica­ III tier; physical descriptions ofderivative re­ it is more economical to database users to nately, the similarity measure is not de­ tions-e.g., two items with the same title 1 1:1 produced items are given in the second the extent that they are charged for online scribed. Seven percent of pairs falsely and authorship but two different dates of tier. If a library has only a reproduction searching time, or for the number of rec­ merged seems of some significance and publication and extent statements, i.e., but not the original, it must construct ords they must access in order to make might not be tolerable in a high-quality paging-were thus excluded from study. some sort of phYSical deSCription for an decisions about usefulness of the records database. These are relatively important and fre­ item it does not have. If the technique for their purposes; if the user's time is quently occurring categories of manifes­ were ever to be extended to simultaneous considered valuable in its own right, it is tation. The major value of the study lies in PREVIOUS RESEARCH ON VISIBLE publications, e.g., an audiocassette and a more economical to summarize the differ­ the intellectual analysis of the types of INDICATORS OF MANIFESTATIONS CD issued at the same time, one of these ences on one record than to make the user relationships and the way they are com­ phYSical formats would have to be arbi­ look back and forth between two records In 1946, the Library of Congress pub­ municated in catalog records under cur­ trarily designated primary and described to see what the differences are. lished research on the frequency with rent practice.

I in the first tier. A more effective solution which title pages and collations of books might be to allow the physical description could be relied upon to indicate whether RECORD-STRUCTURING TECHNIQUES fields to be repeated on the second tier, RECORD-MATCHING ALGORITHMS two books containing the same work are and to allow the first tier to exist without In the section above, oeLe's record­ the same edition or not (LC, Processing One of the objectives of descriptive cata­ a "primary" physical deSCription. matching algorithm has been mentioned Dept. 1946). The Library studied 49 lOging is to communicate to users any dif­ BarbaraTillett includes physical variants occaSionally, and it has been compared to groups of books that had different title ferences between items that are J..'l1own to in the category ofequivalence relationships, OCLe's policies for catalogers concern­ pages, Le., that seemed to be difl'erent the cataloger and that might be ofsignifi­ which she defines as follows: "Equivalence ing when to make a new record. Record­ editions, based on examination ofthe title cance to most users. Conversely, insignifi­ relationships are those [that] hold (1) be­ matching algorithms are programs used in page, but had the same paging. The study cant difl'erences should not be made to tween exact copies of the same manifesta­ large bibliographic databases that collect revealed that of these 49 groups, 40 con­ look as if they were significant. A second tion of a work or (2) between an original records from many different sources sisted of either issues, reprints, type-fac­ goal is to communicate these differences work and reproductions of it, as long as (Coyle 1984; Coyle 1985; Coyle and Gal­ similes or copies, rather than true edi­ as economically and concisely as possible intellectual content and authorship are pre­ laher-Brown 1985; Hickey and Rypka tions, vvith reset type. In other words, (ALA 1967, 189). It seems self-evident served" (Tillett 1987, 27). 1979; Klemperer 1978; MacLaury 1979; there were a number of cases in which that the communication of a Significant Sometimes rather substantive differ­ McPherson, Coyle, and l\Iontgomery paging was a more reliable clue than title difl'erence to users with a Single line of ences can take place in phYSical format, 1982; O'Neill 1990; O'Neill 1991; Wil­ page variation as to whether two items text is preferable to the communication of especially when audio or visual works are liams and MacLaury 1979). The function actually represented different editions. a Significant difference with two full com­ transferred from one medium to another. of a record-matching algorithm is to iden­ Svenonius and O'Neill are engaged in puter screens of data that differ in only Asound recording can change from stereo tify duplicate records, records that repre­ a study of a sample of works from the one line of text. In the latter case, the to monophonic in the course of the trans­ sent the same manifestation of the same OCLC database, but their results have not users have to spend a good deal more time fer. A film that is transferred onto video work. There is some evidence that these yet been published (Svenonius and reading in order to figure out what the can experience considerable degradation are being designed to try to deal with O'Neill 1988). The purpose ofthe study is diHerence is. Some differences can affect of image. A still photograph in color can near-equivalency; for example, most to determine whether it is possible to pre­ a number of areas of the description; an experience a considerable shift in color match only on certain characters in the dict from clues easily accessible to the example is a language difference between values in the course of reproduction. It title field, not all characters (Coyle 1985, cataloger, such as paging or title page tran­ two film manifestations that can lead to can certainly be argued that these differ­ 59). However, there is always the possibil­ scription, when two items are the same the title and credits being in a different ences represent Significant difference in ity that such algOrithms might lead to work, text, or edition, i.e., typesetting. language, and to the need to code differ­ the intellectual or artistic content. How­ some merging of items that are truly dif­ Barbara Tillett's doctoral research on bib­ entlv a number of areas in the USMARC ever, if the differences resulting from this ferent manifestations, and also to lack of liographiC relationships includes a study for~1at. At times differences can be of kind of change are the only differences recognition of near-equivale,:cies that ap­ of the frequency with which equivalence, equal significance to the user, but that between two items, and the differences pear different to the algorIthms. Some derivative and descriptive relationships diHerence can be communicated vvith a can be clearly indicated in the phYSical research on the validity and reliability of are noted in cataloging done at the Li­ Single phrase in the phYSical deSCription; description, a more economical way to these algorithms is currently being done braryofCongress between 1968 and 1986 an example might he the difference II • 238/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /239

SYNC01-0JPRISH JeLKJ J J J Search Edit View Actions Options SID: 04036 OL SYNC01-0JPRISH JeLKJ J J J Beginning of record displayed. Search Edit View Actions Options SID: 040:56 OL Beqinning of record displayed. OLUC dt pat",/med Record 87 of 316 NO HOLDINGS IN CLU - 34 OTHER HOLDINGS OLUC dt pat",/med Record 260 of 316 OCLC: 4599336 Rec stat: e NO HOLDINGS IN CLU - 18 OTHER HOLDINGS Entered: 19790130 Reolaced: 19870805 Used: 19920217 OCLC: 9333787 Rec stat: n Type: 9 Bib Ivl: m Source: d Lang: eng Entered: 19830322 Reolaced: 19890505 Used: 19920609 Type mat: v Enc Ivl: I Govt pub: Ctry: miu Type: 9 Bib lvl: m Source: d Lang: eng lnt lvI: 9 Hod rec: Tech: 1 Leng: 171 Type lIlat: v Enc lvl: K Govt pub: Ctry: miu Oesc: i Accomp: Oat tp: r Dates: 1978,1969 Int lvl: Hod rec: Tech: 1 Leng: 171 1 040 GZR c GZR d m.c. d OCL Desc: a AccolllP: Oat tp: r Dates: 1982,1969 2 007 v b f dee b fag h h 0 1 040 YPL e YPL d m/e 3 019 4599427 2 092 791.43 b War, P 4 045 x4x4 3 090 b 5 090 b 4 049 CLUH 6 049 CLUH 5 245 00 Patton h Videorecording / c Twentieth Century-Fox FilII Corp.; 7 245 00 Patton. h [Videorecording] / e Twentieth Century-Fox FilII producer, Frank Mccarthy; director, Franklin J. Schaffner; screen story and Corp. screenlay, Francis Ford Coppola, Edaund H. North; music, Jerry Golds.ith. 8 260 Farllington Hills, Hich. : b Hagnetic Video Corp., c 1978, 6 260 Far.ington Hills, Hichigan : b 20th century-Fox :Video" c (Illade 1969] c1982. 9 300 2 cassettes, 171 .in. : b sd., col. ; c 1/2 in. 7 300 2 video cassettes (VHS) (171 lIin.) : b sd., col., ; c 12 in. 10 500 VHS. 8 511 1 George C. SCott, Karl Halden. 11 500 A videocassette release of the IIOtion picture. 9 500 Videodisc release of the 1969 IIOtion picture by 20th century- Fox. SYNC01-0JPRISH JeLKJ J J J search Edit View Actions Options SID: 04036 OL SYHC01-0JPRISH JeLKJ J J J End of record displayed. search Edit View Actions Options SID: 04036 OL End of record displayed. OLUC dt pat,../Illed Record 87 or 316­ NO HOLDINGS IN CLU - 34 OTHER HOLDINGS OLUC dt pat~../illed Record 26G..of 316. 12 500 Based on factual asterial froa the books Patton: ordeal- and NO HOLDINGS IN CLU - 18 OTHER HOLDINGS triulIIPh-, by Ladislas Farago, and A soldier's story, by Our N. Bradley. 10 sao Based on the books: Patton ordeal and triulIIPh, by Ladislas 13 511 George C. Scott, Karl Halden. Farago, and A soldier's story, by Our N. 8radley. 14 508 Producer, Frank t1cCarthy; director. Franklin J. Schaffner; 11 520 The WOrld War I I adventures of the controversial Allerican screenplay, Francis Ford Coppola. Edmund H. North; music, Jerry Go1ds.ith. general, George S. Patton. 15 520 Adventure drama of World War II American general George S. 12 500 Rated : PG Patton. 13 600 10 Patton, George S. q (George Smith), d 1885-1945. w cn 16 600 10 Patton, George S. q (George SlIith). d 1885-1945 x Oraaa. w 14 650 0 Feature films. en 15 650 0 War films. 17 700 11 Farago, Laaislas. t Patton: ordeal and triumph. w cn 16 650' 0 Video recoraings. w cn 18 700 11 Bradley, Omar Nelson. d 1893- t A soldier's story. w dn 17 710 21 Twentieth Century-Fox Film Corporation. w cn 19 710 21 Twentieth Century-Fox Film Corcoration. w cn Figure 1, page 2 Figure I, page 1

indicated to users by the fact that two reproduction, or that differ in distribution between the 70 millimeter and the Versions Forum 1990; ALA, Task Force records are identical except at the points information. See figure 1 for some exam­ panned-and-scanned non-wide-screen on Multiple Versions 1992). The third, the where the differences between two mani­ ples of catalo!,Ting done using this tech­ manifestation of a film. four- or five-tiered hierarchical technique, festations are being described. A record nique. The records in the examples were There are at least three techniques is not currently used in library cataloging, thus describes an item in such a way as to all found in OCLC. Under this technique, that could be used to indicate differences but has been in the past in blXlk catalogs. identify it as being in some ways the same users who are trying to sort out the various that are deemed significant. The first, the as another item or group of items and manifestations of a work must read separate record technique, is currently distinguishes it as being in some ways dif­ through each deSCription to see how it the most Widely used. The second, the ferent from another item or group of TIlE SEPARATE RECORD TECHNIQUE differs from the others; this involves read­ two- or three-tiered hierarchical tech­ items. This technique is currently used for ing through much repetitive information, nique, is currently being considered for Using the separate record technique, a true manifestations, for title manifesta­ deSCribing the aspects of a given manifes­ adoption for the deSCription of reproduc­ new record is made for every different tions, and for near-equivalents that differ tation that are actuallv the same for all tions, a type ofnear-equivalent (M ultiple manifestation of a work. Differences are in phYSical format for reasons other than manifestations. ­ • LRTS • 38(3) • Afallijestatiolls alld Near-E(/Ilivalellts /241 240/ LRTS • 38(3) • Yee

14. Video reproduction of a 35 mm film (fetc.) indicates portions of record here omitted for the only for reproductions, and even there, TIlE Two- OR TIIREE-TIEHED sake of brevity) HIEHAHCIIICAL TECIINIQUE only for reproductions of originals that have already been completely described. She wore a yellow ribbon [motion picture] / RKO ; producers, John Ford, Merian C. The very fact that two items are described Difficulties arise when one has to try to on two records is a signal that significant describe a reproduction without complete Cooper; associate producer, Lowell Farrell; director, John Ford; screenplay, Frank differences exist. Systems that allow du­ information about the original from which Nugent, Laurence Stallings. -- United States: RKO, c1949. plicate records, i.e., two records describ­ it derived. Ifthe use of this technique is 6 film reels (103 min.) : sd., col. ; 35 mm. ing the same manifestation, are confusing to be extended to other kinds of near­ Author, James Warner Bellah. to users for this reason. If two items are equivalents, such as simultaneous publi­ Cast: John Wayne, Joanne Dru, John Agar, Ben Johnson, Harry Carey, Jr., Victor described on the same record, the implicit cations in different physical formats, or McLaglen, Mildred Natwick, George O'Brien, Arthur Shields, Harry Woods, Chief Big near-equivalents with different distribu­ message is that their intellectual and artis­ Tree, Noble Johnson, Cliff Lyons, Tom Tyler, Michael Dugan, Mickey Simpson, Frank tors and distribution dates, the technique tic content is exactly the same. Under the McGrath, Don Summer, Fred Libbey, Jack Pennick, Billy Jones, Bill Goettinger, Fred two- or three-tiered hierarchical tech­ will probably have to be modified to ac­ Graham, Fred Kennedy, Rudy Bowman, Post Parks, Ray Hyke, Lee Bradley. nique, two items that do not differ in commodate near-equivalents of equal intellectual and artistic content, but only status, Le., with no identifiable original. Credits: Art director, James Basevi; musical director, C. Bakaleinikoff; photography, in phYSical format, are described on the See figure 3 for an example of a catalog Winton Hoch; editor, Jack Murray. same record; differences in physical for­ record created at the UCLA Film and All credits were supplied from: Film daily yearbook, 1950. mat are described in dependent near­ Television Archive, where no attempt is Safety film base; optical sound; filmed using the 3-color Technicolor process; Eastman­ equivalent records attached to the main made to identify an original. color print. catalog record. Ifthere are differences in Original running time was 103 min., according to: Film daily yearbook, 1950. the visible indicators associated with mere TIlE FOUR- OH FIVE-TIERED I. Ford, John, 1894-1973. II. Cooper, Merian C. III. Nl]gent, Frank. IV. Stallings, Jil difference in phYSical format, these, too, HIERARCHICAL TECIINIQUE Laurence, 1894-1968. V. Wayne, John, 1907-1979. VI. Dru, Joanne, 1923- VII. Agar, can be indicated in the dependent near­ John, 1921- [etc.] equivalent records. The two- or three­ The four/five-tiNed hierarchical tech­ tiered hierarchical technique is currently nique is a technique that was used in the being proposed as a two-tiered technique old book catalogs. Ifit is ~'onceived of as a • FlLM ARCHIVES - MP 619 - Reel 1-6 to deal with reproductions. The two tiers four-tiered technique, the four tiers consist of the catalog record and the de­ would be work-manifestation-near­ Reproduction (videocassette): [Los Angeles, Ca. : Taped by UCLA Film and Television I pendent records that describe both vari­ equivalent-holding; if it is conceived of Archives, 1988] ous near-equivalents and various copies as a five-tiered technique, the fifth tier I videocassette (103 min.) : sd., col. ; 1/2 in. held. Many think that for this new tech­ would be version in the old sense, that is, VHS. nique to work, it will have to be three­ a manifestation that itself has manifesta­ tions, such as the various editions of a tiered on implementation, with the sec­ • AUDIOVISUAL -- VC 201 ond tier, the near-equivalent tier, particular translation of a work. The five Figure 2 identifying various near-equivalents avail­ tiers would then be work-version­ able, and the third tier, the holdings tier, manifestation-near-equivalent-hold­ consisting of copies held and locations ing. In a sense, the unit of cataloging was attached to the appropriate near-equiva­ the work. Once a user located a work in 4 for an example from the British Mu­ each time it added a manifestation. The lent record. which he was interested, he could see seum book catalog. Multiple Versions Forum might represent The current implementation of the displayed the various versions, texts, edi­ The four- or five-tiered hierarchical a move in the direction of the four- or two- or three-tiered hierarchical tech­ tions, and phYSical variants of that work technique has not been used since the five-tiered hierarchical technique, al­ nique requires that one near-equivalent subarranged by language or subsidiary days of the book catalog, prior to the ad­ though the two-tiered approach that does he deSignated as the original, to be de­ author and then by date. In other words, vent of the environment in which we now not clearly differentiate between near­ scribed in the bibliographic record itself. records were arranged in such a way tllat live, dominated bv the unit record, bv equivalent-specific infilrmation and hold­ All other near-equivalents, those de­ manifestations that were most alike were shared cataloging: and by multiple n;­ ing-speCific information, but lumps them scribed at the second tier, are considered close together, and manifestations that tional databases. Such conglomerated together on the second tier, is not yet a to be derived from the original. See figure were most different were farthest apart. records, representing a work with the edi­ very elegant solution. So hlr, the library Concise entries for each version, text, etc., 2 for an example taken from the Guide­ tions held described in four or five tiers, communitv has taken a very conservative indicated only how it differed from those lines for Bibliographie Description ofRe­ would be difficult to use in shared catalog­ s~ope above it. Thus it was easy for a user to scan approach to defining the of poten­ productions adopted by the Committee ing the way it is currently practiced. be­ multiple entries and make an efficient tial application of the two-tiered ap­ on Cataloging: Description and Access cause each collection would hold differ­ choice of the best manifestation to suit his proach, limiting it to reproductions. (CCDA) at the ALA Annual Meeting in ent manifestations, and would have to or her purposes, in a listing of all the In the unit record. shared catalOging, July 1992. Because of the need to deSig­ re-edit and replace the whole work record multiple national datahll~e pn\ironment nate an original, this model works well manifestations of a given work. See figure

I • 'I i

" 242/ LRTS • 38(3) • Yee LRTS • .38(3) • Manifestations and Near-Equivalents /24:3

Arter tomorrow COMMAND-> Type HELP or press PFl for options. OTHER ENTR(IES): 1. 80rzage, Frank. 2. LeVien, Sonya, 1898-1960. 3. Howe, James Wong. 4. Leverett, George. 5. Darling, William, b. 1882. 6. Duty, GUy S. 7. Hanley, James F. (James Frederick), 1892-1942. 8. -- Long Display Screen ------­ Clancey, Margaret. 9. Farrell, Charles, 1901- 10. Nixon, Marian, 1904­ Record D13 of 2316 Screen 1 of 11. Gombell, Minna, 1892-1973. 12. Collier, William, 1866-1944. 13. HUll, Josephine, 1886-1957. 14. Golden, John, 1874-1955. 15. Stange, Hugh After tomorrow / Fox Film corporation ; Frank Borzage production ; directed by Stan1slaus. 16. Fox Film Corporation. Frank Borzage ; screen play by Sonya Levien. United states : Fox, C1932. COP(IES) HELD IN FILM COLLECTION: Drama; feature. 1. 35 mm. nitrate print. 9 reels of 9 (79 min.) (ca. 9000 ft.) "Based on the stage play by John Golden and Hugh S. Stange." b&w opt sd., The players: Charles Farrell, Marian Nixon, Minna Gambell, William Collier, Sr.; Josephine Hull; William PaWley; Greta Granstedt; Ferdinand Munier; Nora AVAILABILITY: Individual use only; no projection. Lane. :IOTES: Studio print. Old Archives location no.: 46-AA-2. CREDITS: Photography by James Howe; sound recorder, George Leverett; art Press ENTER for the Next Screen director, William Darling; costumes by GUy S. Duty; music by James Hanley. {Editor, Margaret Clancy, i.e. Clancey}. Current Search: fsu ucla preservation Production credit in brackets supplied from copyright catalog, 1912-1939. COMMAND-> "Western Electric System." Playing time on release was 79 min., according to: Film daily yearbook, 1933. End of last reel contains exit music. Type HELP or press PFl for options. Press ENTER for the Next Screen -- Long Display Screen ------_ Record D13 of 2316 Screen 4 of CUrrent Search: fsu ucla preservation COMMAND-> After tomorrow Type HELP or press PF1 for options. CONDITION: Poor condition: single-system work print used as projection print; splice at every cut (800th condition report, 4/17/86). LOCATION: R-F63-R13-1 INVENTORY NO: M15105 -- Long Display Screen ------­ Record 013 of 2316 screen 2 of 2. 35 mm. safety print. 5 reels of 5 (ca. 9000 ft.) : opt sd., b&w After tomorrow AVAILABILITY: ProJection only; no individual viewing. II :IOTES: Copy added from inventory record without viewing or inspection. CONDITION: Missing footage compared to 79 min. original release length? III!I Copyright: Fox Film Corp.; 12Feb32; LP2881. "Passed by National Board of Review." Running time supplied by Booth. Excellent condition (Melnitz condo , II PROGRAM NOTES: Charles Farrell and Marian Nixon star as two young people report, 7/27/1989). prevented from marrying by the selfishness of their parents in an adaptation REPRODUCTION: Reproduced by UCLA Film and Television Archive from 35 of a play by John Golden and Hugh S. Stange. For many years, this film was rnm. safety prsv camp dupe neg (XFE504 -512 M). Reproduction for by i 1:11 thought to be lost, until the last surv1ving nitrate print was discovered in preservation purposes permitted Twent1eth Century-Fox, 1982. the Fox vaults and turned over to the UCLA Film and Television Archive.to be Press ENTER for the Next Screen copied for preservation. With Minna Gombell, William Collier, Sr., and Josepnine Hull (later famous as one of the two murdering aunts in Arsenic Current Search: fsu ucla p~eservation and old lace) recreating a role she had played on the stage. COMMAND-> PRESERVATION HISTORY: Preserved at UCLA. GENRE(S): Features. SUBJECT(S): UCLA preservation. Type HELP or press PF1 :or options. Press ENTER for the Next screen -- Long Display Screen ------_ Record 013 of 2316 Screen 5 of Current Search: :SU ucla preservation COMMAND-> .;fter tomorroW"

~~CATION: R-A3-12~-3 !:~e HELP or press PFl for options. INVENTORY NO: MJ5412 35 nm. safety print. : ~eel of 8 lea. 1000 ft.) : opt sd., -- Long Display Screen ------­ b&w :;OTES: Copy addeci f::-01:l inventory record without view1ng or inspec't.ion. Record 013 of 2316 Screen J of CCNOITION: Incomplete. Figure 3. pa~~ 2 Figure 3~ page 1

in which we have been living for the past determine where a new manifestation of unit records, but decisions about the rela­ well for machine linking. because so many century, it is difficult to devise elegant a work should fit among other manifesta­ tionships ofmultiple unit records. Or per­ works are !"tiven title main entries under solutions. Perhaps it would do no harm to tions of that work, and to devise a c(ll1dse haps it would be pOSSible to de"ise some AACR2 rules, and there are so manvcases dream of a distant future in which all description that indicated only how it dif­ other wav to record and share information ofdifferent works that have the sanie title. libraries share a Single "irtual catalog, fered from other manifestations. Once about he;w an item being cataloged is re­ Because the filUr- or five-tiered hierar­ with searching and display mechanisms this record had been fitted into place, its lated to other items already cataloged in chical technifjue is currently impractical, for library patrons that could, on demand, place in the arrangement ofall manifesta­ the national databases. Current tech­ the f(lllowing discussion of recommended suppress items not in the local collection. tIons would he fixed felr all users of the niques for relating, invohing alphabetic cataloging techniques assumes a choice ,1'1 Part of the cataloger's task would be to catalog. In effect, we could share not just matching on main enhies, do not work hetween the separate n'conl technique ~JL • I, i 244/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /245

SMOLLET (TOBIAS GEORGE) ---The second edition. London: W Johnston; Salis­ --See WRIGHT (Thomas) M.A., F.B.A. History of the bury: B. Collins, 1772. 12614. see. 9. reigns of George IV, and William IV, being a continuation 3 vol. 12°. of Hume, Smollett, and Miller's History of England, etc. Anonymous. [ca. 1838]. 1500/88. pp. xi, 637. 8°. --[Another edition.] Dublin: A. Leathley, etc., 1774. AN ESSAY ON THE EXTERNAL USE OF WATER. 12612. dd. 13. I 2 vol. 12°. ! --An essay on the external use of water. In a letter to , i Anonymous. Dr. 0000 with particular remarks upon the present method of using the mineral waters at Bath in Somersetshire, etc. , , --The expedition of Humphry Clinker, etc. 1775. See i London: printed for M. Cooper; sold by D. Wilson: Bath: sold by Leake & Frederick, 1752. C. 123. k. 3. supra: [Collections.] The select works ofT. Smollet, etc. pp. 48. 4°. vol. 7,8.1776.12°. 1578/1925. ~ ---[Another edition.] Edited, with introduction and notes, by Claude E. Jones. Reprinted from Bulletin ofthe Institute Anonymous. of the History of Medicine, etc. Baltimore: Johns Hopkins ---The expedition of Humphry Clinker. By the author of Press, 1935. 7483. r.10. Roderick Random. London: W Johnston; Salisbury: pp. 31-82: plate; port. 37 cm. B. Collins, 1779. 1807/4538. 2 vol. 8°. ~ THE EXPEDITION OF HUMPHRY CLINKER ---The expedition of Humphry Clinker. By the author of --The expedition of Humphry Clinker, etc. London: T. Becket; Roderick Random. London: w: Johnston; Salisbury: J. Pridden, 1681 [1781]. 1807/3762. ~ I ' B. Collins, 1671 [1771]. C. 95 sa. 8. 2 vol. 12°. 3 vol. 13°. Anonymous. I The date is correctly printed in vol. 2,3. I Anonymous. --[Another edition.] Dublin: J. Ezshaw, etc., 1781. 012642. pp. 86. --The expedition of Humphry Clinker. By the author of 2 vol. 12°. Roderick Random. The second edition. London: W Johnston; Anonymous. Salisbury: B. Collins, 1771. C. 175. m. 15. I, 3 vol. 12°. ~ I --The third edition. London: T. Longman, and G. Robinson, Anonymous. 1683 [1783]. 12650. a. 78. 3 vol. 12°. --[Another edition.] Dublin: A. Leathley, etc. 1771. Anonymous. 1484. bbb. 11. 3 vol. 12°. Anonymous. --[Another edition.] Dublin: w: Sleater, etc., 1784, 85. 1471. de. 44. ---[Another edition.] Dublin: A. Leathley, etc. 1771. 2 vol. 12°. 1478. c. 41. Anonymous. 3 voL 12°. Anonymous. Vol. 2 is a duplicate ofthe preceding. I Figure 4, p"Ae 1 II Figure 4. page 2 II, • 246/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /247

I I and the two- or three-tiered hierarchical format, do not make a new record. This work. Pre-code 1930s films were cut prior had its title changed, to Call It Murder. technique. approach could both save money and to reissue or rerelease (McElwee 1990, pt. Another example is the 1936 filr.1 of As help users. 3, 139). Foreign films of the fifties and You Like It. When originally issued, Elisa­ sixties were felt to be too racy for U.S. beth Bergner received top billing. By the PROPOSED DEFINITIONS MOVING-IMAGE MATERIALS audiences and were cut before being time the film was reissued several years Based on the above discussions of user shown. Sometimes unacceptable words later, her costar, Laurence Olivier, was a needs, the following definitions are pro­ Above, we have attempted to define were dubbed out on the soundtrack. bigger star and was given top billing on posed: manifestation, title manifestation, and In other cases, manifestations were reissue prints. Making two separate cata­ Manifestation: The set of all items that near-equivalent in general terms that created for different regions. Newsreels log records for the original release and the represent the same work and do not differ would apply to all materials. Now, the were issued in several manifestations \vith reissue documents this difference in bill­ in intellectual and artistic content from kinds of differences that can occur be­ regional stories for shOwing only in a cer­ ing order for film historians, Billing each other in a way that would be consid­ tween manifestations or near-equiva­ tain part of the country. Leni Riefenstahl changes are interesting in their own right, ered significant by most users of the col­ lents of moving-image works, and that made several different manifestations of and therefore probably worth recording lection. An example of an insignificant might be significant to users, will be Olympia (1936); the German manifesta­ in our catalogs; as McElwee puts it, "Bill­ difference in intellectual and artistic con­ discussed and categorized based on the tion included more minor events so as to ing changes charted a player's rise and fall tent, i.e. one that would not create a new previously developed definitions. show more German victories than did the from the original release ofa feature to its manifestation, might be correction oftype­ There is much anecdotal evidence in Spanish and English manifestations. reissue years later" (McElwee 1990, pt. 3, setting errors or misspellings. An example the HIm literature concerning the exist­ 140). However, billing changes do not al­ of significant difference in intellectual and ence ofvarious manifestations. The rights ways plea~e the actors and actresses in­ II TYPES OF DIFFERENCE BETWEEN artistic content, i.e., one that would create to most moving-image materials belong to volved. A recent news item in the Las I MOVING IMAGE MANIFESTATIONS OR I a new manifestation, is creation of a dis­ for-profit corporations that are perfectly Angeles Times indicates that Kevin Cost­ NEAR-EQUIVALENTS 'I tinct manifestation by the original author, willing to edit these works to be shown to ner is suing a video firm for using his name III e.g., a revised edition. various markets in various formats, as long Manifestations of moving-image works prominently in marketing and di~tribut­ Title manifestation: The set of all items a~ they think a profit can be made. Prior can be created in many of the same ways ing the 1985 film Chasing Dreams, in that represent the same manifestation of to the era oftelevision, companies such as that manifestations ofother kinds ofwork which Costner had a minor role ("Costner ,1'1'II the same work and that have identical chief Film Classics and Realart Pictures ac­ can be created. Let us consider some spe­ sues video firm" 1990). ",! 'I sources of information, other than distri­ quired the rights to distribute olderstudio cific categories of difference. bution information; two items that have titles to neighborhood theaters and drive·· DIFFERENCE IN EDITION 011 the same intellectual and artistic content, ins to fill out double features. According ",1'1I DIFFERENCE IN TITLE AND ORDER DISTRIBUTION STATEMENT: 1'1'1', ! but differ in title or statement of responsi­ to McElwee, cuts would be made when OF CIlEDITS: TITLE MANIFESTATIONS TIlUE MANIFESTATIONS AND bility, are two different title manifesta­ necessary to accommodate time limita­ NEAIl·EQUIVALENTS tions. tions in the double-feature format It is very common for films to be reissued Near-equivalent: The set of all items (McElwee 1990, pt. 3,140). When televi­ or rereleased under new titles, and for Educational and informational films will that represent the same manifestation of sion became a medium of distribution for television programs to be rebroadcast un­ sometimes carry edition statements on re­ the same work and that have identical dis­ films, they were edited to remove profan­ der new series titles. At least one reason vised and updated editions. Until re­ tribution information and phYSical charac­ ity, sex, violence and product identifica­ for the reissue of films under new titles cently, such statements have been very teristics; two items that have the same tions (ifLucky Strike was a sponsor, Hum­ was a desire to prevent a member of the rare on theatricallv released films or net­ intellectual and artistic content and iden­ phrey Bogart couldn't be seen smoking audience from realizing ahead ofUme that work broadcast tel~vision programs. Close tical chief sources of information other Camels), and then footage was either he or she had already seen the film. McEl­ Encounters ofthe Third Kind (1977) had than distrihution information, but differ in added or removed to enable them to fit wee mentions, for example, that Chaplin a much-advertised rerelease in 1980 as distribution information, such as edition into standard time slots between commer­ and Pickford shorts were reissued under "The Special Edition," which had indeed statement, publisher. distributor, or date, cials; wide-screen films would be "panned a variety of misleading titles (McElwee been re-edited by Spielberg. The recently or in phYSical characteristics, such as pa­ and scanned," a process in which only a 1989,59,3). Maltin indicates that films are released reconstructed manifestations of per, type, binding, film base. or medium of portion of the wide-screen image is se­ being retitled on video "to lure unsuspect­ films such as A Star is Bom (1954) and reproduction, are two different near­ lected for showing on the small TV screen ing renters" (Maltin 1989, viii). Lawrence of Arabia (1962) have had equivalents. (Haserot 1989, 49). Airline manifestations Films would also be reissued or rere­ prominently displayed manifestation If we could adopt the definitions are edited for sex, violence (especially air­ lea~ed with the credits altered. McElwee statements, as have the director's cut above, we could cut down considerablyon plane crashes), language, and length. indicates that after Alan Ladd, Marilyn manifestations being released on video. the number ofnear-equivalents cluttering Sometimes different manifestations Monroe, and Humphrey Bogart became Television manifestations and airline our databases. A principled approach were created for censorship reasons. In big stars, earlier films in which they manifestations, on the other hand, do tlle early days ofstates' rights distribution could be taught as follows: Make a new played minor roles were reissued or rere­ not carry expliCit manifestation state­ I: of motion pictures, each state had its own record only if title, authorship or extent leased with their names given top billing ments, and the industry has resisted re­ censorship board; in effect, there were I (paging for books) changes; if the only above the title. He mentions one such cent attempts to get 'it to label such state-specific manifestations of each change is in publisher, date or phYSical Bogart Him, Midnight (1934), which also manifestations.

Ii, I II 248/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-EqUivalents /249

Distribution statements often differ on by a film scholar can be added in such is Born (1954), Intolerance (1916), and into another language. It also includes si­ moving-image materials without there be­ a way that it can be switched on or off. Way Down East (1920), involved the sub­ lent films with intertitles that have been ing any other difference. In today's world This latter would be equivalent to an stitution of stills for footage missing even translated from another language. The in­ of\ideocassette distribution, one suspects edition of a textual work 'vitll com­ after exhaustive searching (Everson 1990, tertitles on silent films can differ from one that rights to videocassette distribution mentarv. Ii; Gunning 1984, 19; Haver 1983, .33; manifestation to another in ways other frequently change hands without any ac­ Some 'physical format differences, Stanbrook 1989-90,29). than translation. For example, Gunning companying change in the work being dis­ such as colorization or panning and scan­ Sound films can have their soundtracks indicates tllat Way Down East (1920), tributed. This was undoubtedly true, as ning of wide-screen films, might be con­ altered in ways that do not involve differ­ Griffith's silent film, existed in manifesta­ well, in the 16 millimeter market that pre­ sidered by some to be a difference in the ences in the footage. The MCA Home tions with several different sets of inter­ ceded today's videocassette market. intellectual and artistic content, because Video videocassette and videodisc mani­ titles (Gunning 1984). Gillett notes that they affect the visuals so radically. These festations of Dracula (1931) include an "intertitles could be altered to smooth issues ,viII be discussed further below. original soundtrack suppressed before re­ over censorship problems from one coun­ DIFFEHENCE IN ACTUAL CONTENT: 'I!! The follO\ving, then, are examples of lease because it contained more groans, try to another" (Gillett 19ii-i8, 38). I, TnUE MANIFESTATIONS I alteration that can be said to create new bone cracks and other horrible noises than There have been several English transla­ " There are basically three ways in which an manifestations: were considered acceptable for audiences tions of soundtrack into subtitles for edited film work can be altered in such a of the 1930s. Grand Illusion (l93i) and Breathless way as to create differences in the intel­ Manifestations with Editing (1960), some more accurate than otllers. lec'tual and artistic content significant Causing Differences in the Addition of New Material In the late 1920s and early 1930s, during enough to create a new manifestation of Continuity or Track Appended to Work the transition from silent to sound films, I the \~()rk. This category includes short manifesta­ Some of the new director's cuts being re­ films were often released in silent and 'I • The film can be edited to change the tions, and manifestations censored or ed­ leased on video include such additions as sound manifestations. The sound mani­ continuity. For example, footage can ited for television or airline showing or for an interview with the director, outtakes, festations would have music tracks and be deleted, or "cut," but with the inclusion in double features. Informa­ rehearsals, shot setups, and auditions some dialogue added. The new FIAF original continuity, or order of shots, tional films with new footage added to (Fleming 1990). Blackhawk reissues with rules consider difference in language, mu­ preserved. This would be equivalent update them would fall into this category. historical introductions were mentioned sic, or dialogue to create "an item with to abridgement of textual works. Also, Sometimes footage is added to films. Dobi above. McElwee mentions several silent minor changes," one that is to be de­ footage can be added, but with the indicates that the 1948 reissue of Nanook films thatwere rereleased in tlle sound era scribed on the same record as the item Ii original continuity preserved. This of the North (1922) included outtakes with prologues (McElwee 1989,594). He without such differences (FIAF 1991,41). would be equivalent to enlargement from the original footage that were not in also mentions that Puhlic Encmy (1931) Since such "minor" differences can be as­ of textual works. Finallv, alternate the original release (Dobi 19ii, ll). Air­ was rereleased in 1954, heavily censored sociated with differences in subSidiary footage can be substituted. The most line and television manifestations might and \\ith a cautionary foreword'(McElwee authorship (translator, composer, writer common example of this is a manifes­ sometimes require the addition offootage 1989,596). of intertitles, etc.), this practice seems tation released with two different to bring them up to contractual length or Rebroadcasts of television programs dangerous. It could lead to no access un­ endings. to fit specified time slots. MCA added a might be considered to be special cases in der the names of subsidiary authors, or to • New material can be appended to the two-minute dream sequence to Rear Win­ this category. When tele"ision programs misleading access ifa user 'retrieves a rec­ work. For example, Blackhawk reissues are rebroadcast, the commercials, public , dow (1954) after Hitchcock had died in ord on which only one holding of several of early motion pictures often include order to make the film fit into television service and station announcements, etc., is of interest; it might be hard ft)r the user I historical introductions. time slots. The rerelease of Phantom of that are broadcast at regular intervals to tell which holding is ofinterest when a • Finally, changes to the soundtrack, or the Opera (1925) in a sound manifestation throughout the program are different. great deal ofholding-specific information subtitles can be carried out by identi­ required the shooting of new footage to Since commercials and the like can be is buried far down the record in the notes. Hable subSidiary authors, or the cast replace silent footage where there were very revealing social and historical docu­ Anotller consideration is that the coding can change slightly. The soundtrack or sound synchronization problems. In the ments in tlleir own right, they are often for language in the USMARC format is titles can be either translated or re­ course of restoring Toll ofthe Sea (1922), analyzed in contents notes and made ac­ record-specific. Ifholdings have different written entirely. (The term titles is UCLA found it necessary to reshoot the cessible by means of analytical title added language characteristics, there is no way used in the fiI~ world to mean either last scene (Slide 1992, 109-10). Films re­ entries in television cataloging. Thus, it is at present to code for them all. subtitles on sound films, or inter­ leased in several different manifestations useful to treat each rebroadcast of a tele­ In addition to the voices of tlle per­ titles-frames of textual matter ap­ fall into this category, as well. One exam­ vision program as a new manifestation and formers, the soundtrack ofa film also con­ pearing between frames of picture­ ple is Fred Niblo's Blood and Sand (1941), create a new record for it. tains music and sound effects, and these, on silent films.) Differences in the released with two different endings too, can be different from one manifesta­ soundtrack other than differences in (American Film Institute 19i1, 69). An­ Manifestations with tion to another. Films originally issued text can occur. A silent film can have other example is Legal Eagles (1986), Subsidiary Authors silent can be reissued or rereleased with a a music track added or changed, or the which had its ending completely changed This categ~ry includes dubbed and subti­ music track. The same film can be reis­ I sound effects portion of the sound­ for television shOWings (Maltin 1989, viii). tled manifestations in which the sound­ sued or rereleased with several different track can be changed. A commentary Several recent restoration projects, A Star track in one language ha~ been translated music tracks. Few of thp silpnt films came • I,! ~

2.50/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /2.51

with a score, but some did. Various mani­ ences might be said to create a new mani­ to sound, prints would be released \vith the Wind, and Bambi---eliminating un­ festations of these silent films can exist festation on the grounds of difference in both optical soundtracks and sound-on­ wanted noise and converting mono­ with various performances of the same intellectual and artistic content. Haserot sound recording disks for the smaller phonic, or Single-channel, sound into ste­ original score. Stanbrook discusses a new would seem to support this approach theaters that had not yet converted to the reo" (Spotnitz 1990, 56). McElwee also recording ofthe original score for Alexan­ when she writes, "The addition of color newer sound equipment. Other examples mentions the Perspecta Process, which der Nevsky (1938), for instance, and men­ . .. radically alters the fUm's nature by of physical variants eligible for this treat­ simulated stereo and was applied to reis­ tions a video manifestation of Battleship changing the language through which a ment might be monophonic and stereo sues of Gone With the Wind (1939) and Potemkin (1925) that allows one to choose film communicates" (Haserot 1989, 50). soundtrack variants, or panned-and­ The Jolson Story (1946), and a reissue of to hear either the score originally com­ However, this argument might also be scanned 16 millimeter prints or videocas­ Disney's Fantasia (1940) with a Dolby posed for the film by Edmund Meisel, or used to argue that the creation of a black­ settes of wide-screen motion pictures in track that apparently annoyed Disney­ the score by N. Kryukov that was associ­ and-white print of a film originally in which a wide-screen image has been philes (McElwee 1990. pt. 3, 139). Cer­ ated \vith the film for years (Stanbrook color, frequently done in the days when cropped at the sides, losirtg part of the tainly these differences should be com­ 1989-90,31). The restorations ofNanook television was black and white, creates a picture, or even cut slightly differently. (A municated to users so that they can make of the North (1922), and Lucky Star new manifestation, or that two black-and­ wide-screen sequence of two people hav­ sure they have the proper equipment or I (1929) include newly composed contem­ white prints of a silent film, one with tint­ ing a conversation, with both people on can Simply make choices, hut whether porary scores on the soundtracks (Dobi ing, are two different manifestations. screen, and with no cuts, can become a they need to be treated as distinct mani­ 1977, 14-16; Benson 1991). On the 1942 Videotransfer of a color film alters its sequence in which one person is shown on festations is questionable. Spotnitz quotes reissue of Gold Rush (1925), Chaplin's color values and preservation ofa Techni­ screen at a time, with cuts at each point several people who differ in their views Ii' voice was substituted for the original in­ color film must be done on Eastmancolor where one person stops talking and the about the advisability of tampering with I' tertitles (McElwee 1989, 594). The thrust stock using a completely different color other begins. Apparently panning and original soundtracks in this way, but ends I of North Star (1943) was completely process, because the earlier color process scanning is also occasionally used to make with a quote from director John Milius, I changed on rerelease as Armoured At­ is no longer available in the United States. a wide-screen film out of one originally who says, "As long as it sounds baSically I tack, largely by editing the soundtrack. Certainly these bits of information about released at the standard \vidth. The .50th the same, it doesn't matter" (Spotnitz I' The reconstruction of the director's cut the prints are important and should be anniversary reissue of Snow ·White (1937) 1990,56). I,' of Lawrence of Arabia (1962) involved communicated to users, but whether a in 1987 was "vertically panned and Restoration of earlier film formats, getting some of the original actors to new record is necessary to do so is another scanned to fit the 1.8.5 frame," according such as 22 millimeter, 17.,5 millimeter and ,,11 rerecord eight minutes of their dia­ question. I, to Joseph McBride [McBride 1992].) Let­ 28 millimeter, can involve blOWing the logue, which was then mechanically al­ Other substantive differences in physi­ terboxing, adding black borders to the top image up to a standard 35 millimeter im­ 1 tered to compensate for the way the ac­ cal format could result in near-equiva­ and bottom of a wide-screen image to age. Silent films were shot through hand­ I' [ill tors' voices would have changed over the lents, as well, if they could be concisely allow it all to fit into the bounds of a CRT cranked cameras and then projected years (Stanbrook 1989-90, 31). Fantasia indicated in the phYSical description. Ever screen, actually preserves the original through projectors that were hand­ (1940) was apparently released in 1982 since the 1950s, the motion picture indus­ wide-screen image, so although the fact cranked by projectionists to match the "\ • IIIII with a new performance of the musical try has been trying to devise means to that letterboxing has been done should be original cameraman's speed. Thus film " ill track conducted by If\vin Kostal, replac­ make fIlms spectacular enough to draw recorded as holding-specific information, speed can vary a great deal. In order to II ing Leopold Stokowski's original perform­ the audience away from their television I," it cannot be argued that it creates a new restore silent films that are not at the II ance (Phinn 1990,86). sets and back to the theaters. Various manifestation per se. It can certainly be modern 24-frames-per-second speed, the wide-screen processes, 3D, and various argued that these differences represent a films are often step-printed (a process in kinds of stereo sound have been the re­ I" PHYSICAL VARIANTS: significant difference in the intellectual or which a Single frame is repeated) up to 24 sult. Usually these processes reqUired that NEAR-EQUIVALENTS artistic content. However, if this kind of frames per second. Again, these kinds of the tllms be projected using special equip­ difference is the only difference between differences seem to produce near-equiva­ Sometimes rather substantive changes ment that not all theaters would have. two items, and the difference can be lents, rather than new manifestations. take place in physical format. For exam­ Thus, the films were often issued in sev­ clearly indicated in the phYSical deSCrip­ More controversial changes to film ple, color films can be reissued as black eral different formats. Three-dimensional tion, a more economical way to communi­ speed apparently occur in tlle process of and white, black-and-white films can be films were often released in both a 3D cate the difference might be to make one "time compression/expansion" of picture colorized, and silent films can be reissued manifestation and a non-3D manifesta­ record for both items, \vith repeated and "lexiconing" of sound in order to with music and effects tracks. Such differ­ tion for running in theaters that did not physical descriptions, rather than creating shOtten or lengthen a film to fit broadcast ences can substantially affect the quality have the correct projection equipment to two full bibliographic records that differ formats. As Stanley Richards puts it, "The show 3D films. The same is true of\vide­ ofimage in what are essentially visual ma­ only in the physical description. art of an actor's performance is bound up terials. The addition of soundtrack to a screen films and, more recently, films on The equivalent for soundtracks of col­ in timing. Every momentofprincipal pho­ 70 millimeter film that would also be is­ silent film actually introduces an element orization is "stereo-ization." Spotnitz sued in 35 millimeter for theaters that tography is concerned with timing---{)f of subsidiary authorship (e.g., the ar­ writes about sound engineer Rick Chace, could not project 70 millimeter. All of staging, ofperformance, ofcamera move­ ranger of a music track), and a colorizer who since 1984 "has electronically trans­ these could easily be treated as near­ ment. Filmmakers labor in postproduc­ might also be considered a subSidiary figured the soundtracks of some 300 author. For these reasons, such differ­ equivalents. In the days of the transition tion for months and even years to finesse filmS-including Casahlanca, Gone ·With the exact timing of cuts and pacing of I 1,1; • 252/ LRTS • 38(3) • Yee LRTS • 38(3) • Manifestations and Near-Equivalents /253

sequences to within a fraction ofa second. practice based on local needs. To the de­ WORKS CITED rithm. In: ASIS '85. ed. Carol A. Parkhurst. Time compression/expansion (literally gree that there are emerging standards, White Plains, N.Y.: Published for the running the film in slow or speeded-up they conflict. For example, OCLC's input Note: The follOWing abbreviations have been American Societv for Inf{)rmation Science I~dustry motion) is so totally damabring to every standards are quite different from the used in the citations for this paper: by Knowledge Pubs., 77-80. single moment of the filmmaker's vision LCRI on when to make a new record. ALA American Librarv Association Cutter, Charles A. 1876. Rules for a printed that to allow for its use ... is an insult to Current library practice calls for making LC Library of Congr'ess dictionary catalogue. In Public libraries in the United States of America: Their his­ intelligent thinking men and women eve­ two full catalog records for two items that American Film Institute. 1971. The American tory, condition and management: Special rywhere" (Richards 1990). If time com­ differ only in distributor or phYSical for­ report, Department (:fthe Interior, Bu reau pression/expansion and lexiconing result mat. On the other hand, current archival Film Institute catalog of motion pictures produced in the Unitcd States. Feature of Education. Part II. Washington, D.C.: in any loss of image or soundtrack, they moving-image practice calls for recording films, 1921-19.'30 New York: Bowker. Govt. Print. Off., :3-89. certainly should be considered to create a on the same record two items that have American Library Association. 1908. Catalog ---. 1904. Rules for a dictionary catalog. new manifestation; if not, however, per­ signiHcantlv different intellectual and ar­ rules: Author and title entries. Amt'rican 4th ed. Washington, D.C.: Govt. Print. Off. haps they could be considered to create ti;tic conte~t, such as two films that are in ed. Boston, Ma~s.: American Library Dobi, Steve. 1977. Restoring Robert Fla­ near-equivalents. different languages or have different mu­ Assn., Publishing Board. herty's Nanook of the north. Film library sic tracks. ---. 1967. Anglo-American cataloging quarterly 10, nos. 1 and 2: 11. Since the time ofTewett, catalog codes rules. North American text. Chicago: Domanovszky. A. 1974. Functions and objects CONCLUSIONS REGARDING :'1'1 have avoided formufating rules concern­ American Library Assn. of author and title catalogUing Budapest: MOVING-hIAGE MATERIALS ing the object of a record. Some might ---.1941. Catalog Code Revision Commit­ Akademiai Kiado. It appears, from anecdotal evidence in the argue that it is a healthy thing to allow tee. AL.A. catalog rules: Author and title Du Rietz, Rolf. 1974. The concept of biblio­ literature, that the follOwing kinds of dif­ local practice to vary. Public librarians entries, prelim. American 2d ed. Chicago: type. Text 1: 78-92. III American Library Assn. ference can occur between film items rep­ Everson, William K. 1990. Intolerance. Film~ "I could probably make the case that the ---. 1974. Anglo-American cataloguing II] resenting the same work: majority of their users are not particular in review 41, nos. 1/2 (Jan.lFeb.): 1-20. I rules. North American text. Chapter 6, Fellows, Dorcas. 191.5. Cataloging rules, pre­ • Title manifestations can occur when as to which edition (setting of type) they separately published monographs. Chi­ the title or billing order differs with­ lim. ed. Unicersity ofthe State ofNew York read, and that the long practice in public cago: American Lihrary Assn. bulletin 586 (Mar. 1): 7-175. out there being any underlying differ­ libraries of making one record for each ---. 1988. Anglo-American cataloguing FIAF. 1991. The FIAF Cataloguing rules for rules. 2d ed. 1988 revision. Chicago: ence in continuity. text is adequate for their users. On the film archives, compo and ed. Harriet Vv. • Distribution information can differ other hand, it is pOSSible that research American Library Assn. Harrison for the FIAF Cataloguing Com­ without there being any underlying libraries, with their current financial ---. Association for Library Collections & mission. Munich: K.G. Sauro Technical Services. Cataloging Commit­ difference in continuity, creating a problems and cataloging backlogs, should Fleming, Charles. 1990. Directors' [sic] cut. tee: Description and Access. Task Force near-equivalent. consider adopting practices similar to American film 15, no. 6 (Mar. 11). these for purposes of economy. The Mul­ on Multiple Versions. 1992. Guidelinesfor • True manifestations can occur when hihliographic description of reproduc­Fothergill, Richard. and Ian Butchart. Non­ materiaL~ the continuity, Le., visual aspect ofthe tiple Versions Forum might represent a tions. May 17. 1992 draft prepared by the book in libraries: A pradical work, or the soundtrack, i.e., audio slight trend in this direction, although the Cataloging Committee: Description and guide. Hamden. Conn.: Linnett. 1978. aspect of the work, or the textual as­ field has taken a conservative approach in Access Task Force on Multiple Versions, Gaskell, Philip. 1972. A new introduction to pect of the work actually differ, limiting application to reproductions. If and adopted in June 1992. bibliography. New York: Oxford Univ. Pro whether due to editing, due to the all institutions were to decide to create Attig, John. 1989. Descriptive cataloging rules Gillett, John. 1977. Munich's cleaned pictures. appending of new material or due to new records only when Significant differ­ and machine-readable record structures: Sight and sound 47, no. 1 (Winter): 37-39. the work of subsidiary authors creat­ ence in eitherintellectual and artistic con­ Some directions for parallel development. Gunning. Tom. 1984. Rebirth of a movie. ing subtitles, new music tracks, etc. tent or identification occurs, codification I n The conceptual foundatiom ofdescrip­American film 10, no. 1 (Oct. 18-19): 93. • Finally, of this practice would help to standardize tice cataloging, ed. Elaine Svenonius. San Hagler, Ronald. 1963. Local autonomy. Li­ phYSical variants or near­ Diego: Academic Press: 135-48. eqUivalents can occur when physical it. Now that many of us are using large brary resources & technical services 7: Benson, Sheila. 1991. A glOWing lucky star 340-49. fClrmat differs without the involve­ bibliographiC databases such as OCLC for rises again. Los Angeles times, Nov..5. ment of subsidiary authors. shared cataloging and interlibrary loan, Haserot. Karen E. 1989. The colorization of Blanck, Jacob. 1966. The title-page as hihliog­fllm: Technical, legal, historical, and so­ such standardization could have a practi­ raphical ecidence Berkeley, Calif.: School ciocultural considerations. Techne 3 cal benefit for copy catalogers and interli­ of Librarianship, University of California. SIIOULD TIlE OBJECT OF TIlE (Spring) 45-52. brary loan clerks and those patrons who Costner sues video firm. 1990. Los Angeles RECOHD BE CODIFIED? Haver. Ronald. 1983. A Star is born again. bendlt from their activities if the stand­ times, Feb. 7. American film 8, no. 9: 29-33, 59. Coyle. Karen. 1984. Consolidation of mono­ There is a long history ofvariant practice ardization could be done in a Simple, ele­ Hickt'y, Thomas B., and David J. Rypka. 1979. graph records in the UC online catalog. with reganl to the object of a record, i.e., gant, and principled way that could be Automatic detection of duplicate mono­ DLA bulletin 4.2: 10-13. different institutions have different poli­ explained to copy catalogers and interli­ graphic records. journal of library auto­ ---. 1985. Record matching: A discussion. cies on when one item is sufficiently dif­ brary loan clerks. If codification of the mation 12: 125-42. Information technology and lihraries 4: ferent from another to require a new re­ obje~t of a record is attempted, however, Hinnebusch, Mark. 1989. METAMARC: An 57-59. I' cord. It is one area in which cataloging it should be based on a rationale of em­ t'xtension of the MARC ({)rmat. Informa­ institutions are still free to follow local pirical research. Coyle, Karen, and Linda Gallaher-Brown. tion technology and libraries 8: 20-33. '/ 1985. Record matching: An expert algo­ I ,Ii • LRTS • 38(3) III Manifestations and Near-EqUivalents /2.5.5 254/ LRTS Ii> 38(3) Ii> Yee

Richards, Stanley, 1990. New definitions of Todd, William B. 1981. Scholarly uses of the IFLA Committee on Cataloguing. 1974. leni,rth keys derived from title strings.Jour­ preservation. Daily variety, Aug. 31, 12. national union catalog: A bibliographic ISBDM: International standard bibliog­ nal of library automation 12; 143-55. saga. In In Celebration: the National Un­ raphic description for monographic publi­ Richmond, Phyllis A. 1980. AACR2-a review Maltin, Leonard. 1989. Leonard Alaltin's TV Th~journal ion Catalog, Pre-1956 Imprints, ed. John cations, 1st standard ed. London: IFLA article. (if academic librarian­ movies and video gUide. 1990 ed. Bergen­ shill 6: 30-37. Y. Cole. Washington, D.C.: Library ofCon­ Committee on Cataloguing. field, N.J.: New American Library, 1989. gress, 48. IFLA International Office for UBC. 1977. Shinebourne, John A. 1979. A Critique of McBride, Joseph. 1992. New technique brings Tueker, Ben. Letter to Martha Yee, Oct. 13, ISBDIG): General international standard AACR. Libri 29: 231-.59. Pinocchio back to life. Daily variety, June 1982. bibliographic description: Annotated text. Slide, Anthony. 1992. Nitrate won't wait. Jef­ 11,35. Wanninger, Patricia Dwyer. 1982. Is the London: IFLA International Office for ferson, N.C.: MeFariand. McElwee, John P. 1989. Theatrical re-issues. OCLC database too large? A study of the UBC. Spotnitz, Frank. 1990. The Sound of Two films in review 40: 593-96. efIect of duplicate records in the OCLC IFLA Universal Bibliographic Control and In­ Hands Clapping. American Film 15, no. 9: ---. 1990. Theatrical re-issues, Part 3. system. Library resources & technical ternational MARC Programme. 1987. 56-57. Films in review 41: 139-45. services 26: 353-61. ISBD(M): International standard bibliog­ Stanbrook. Alan. 1989/90. As it was in the be­ Whiting, A. D. 1980. AACR2-edition state­ raphic description for rrwnographic publi­McNellis, Claudia Houk. 1985. Describing re­ ginning. Sight and sound 59, no. 1: 28-32. cations. Rev. ed. London: IFLA Universal productions: Multiple phYSical manifesta­ ment. Catalogue & index 59: 5-6. tions in the bibliographical universe. Cata­Steinberg, S. H. 1974. Five hundred years of Williams, Marth~ Eo, and Keith D. MacLaury. Bibliographic Control and International printing. 3d ed. New York: Penguin. MARC Programme, British Library Bibli­ loging & class(fication quarterly 5; 35-48. 1979. Automatic merging of monographiC Svenonius, Elaine, and Edward T. O'Neill. data bases-Identification of duplicate re­ , 1 ographic Services. McPherson, Dorothy S., Karen E. Coyle, and ,I! ! Teresa L. Montgomery. 1982. BUilding a 1988. Clustering eqUivalent bibliographiC cords in multiple files: The IUCS scheme. Jewett, Charles Coffin. 1852. Rules for pre­ record~, merged bibliographie database: The Uni­ Annual revielL' ofocLC research, Journal oflibrary automation 12: 1.56-68. paring catalog cards. In Smithsonian re­ 1987-1988,6-8. port on the constrnction of catalogues of versity of California experience. Informa­ Wilson, Patrick. 1989. The objectives of the tion technology and libraries 1: 371-80. Tanselle, G. Thomas. 1975. The bibliog­ catalog and the means to reach them. In libraries and of a general catalogue; and raphical concepts of issue and state. Bibli­ their publication by means of separate Multiple Versions Forum. 1990. Multiple Ver­ The conceptual foundations of deSCriptive sions Forum report: Reportfrom a meeting ographical society of America. Papers 69: cataloging, ed. Elaine Svenonius. San Di­ stereotyped titles with ru les and examples. 17-6.5. Washington, D.C. Mimeographed type­ held December, 1989, Air/ie, Virginia. ego: Academic, 1-16. Thaxter, Dick, quoted in Verna Urbanski. script transcription, Univ. of Michigan, Washington, D.C.: Network Development Yee, Martha M. 1991. System design and cata­ 1983. Clinic on AV editions held during Dept. of Library Science, Course 305. n.d. and MARC Standards Office, Library of lOging meet the user: User interfaces to I'ii ALA. Newsletter (On-line audiovisual I, ' ---. 1853. On the construction of cata­ Congress. online publie access eatalogs. Journal of catalogers) 3.3 19-20. logues. 2d ed. In Char/es Coffin Jewett and OCLC. 199,3, Bibliographic formats and the American society for information sci­ American lib rarianship, 1841-1968, ed. standards. Dublin, Ohio: OCLC. Tillett, Barbara B. 1987. BibliographiC rela­ence 42: 78-98. Michael H. Harris. Littleton, Colo.: Li­ O'Neill, Edward T. 1990. Duplicate Detec­ tionships: Toward a conceptual structure ---. 1993. Moving image works and mani· braries Unlimited, 1975: 131-55. tion. Annual review of OCLC research ofbibliographiC information used in cata­festations. Ph.D. diss., University of Cali­ loging. Ph.D. diss., UCLA. fornia, Los Angeles. Jolley, L. 1961. The principles ofcatalogUing. 1989-1990: 13-14. I New York: Philosophical Library. ---. 1991. Duplicate Records in Union Klemperer, Katharine. 1978, Bibliographic Catalogs. Annual Review (if OCLC Re­ J spec({ications for consolidation ofrecords. search 1990-1991: 11-14. Berkeley, Calif.: Univ. of Calif. Division of O'Neill, Edward T., Sally A. Rogers, and W. Ii Librarv Automation. Michael Oskins. 1993. Characteristics of Layne, Sa'ra Shatford. 1989. Integration and duplicate records in OCLe's online union Objectives of the Catalog. In The Concep­catalog. Library resources & technical tual Foundations of Descriptive Catalog­services 37: 59-72. ing, ed. Elaine Svenonius. San Diego: Aca­ Panizzi, Antonio. 1985. Rules for the compila­ demic Press, 185-96. tion of the catalogue. In Foundations of Library of Congress. 1981. 1.0. Cataloging cataloging: A sourcebook, ed. Michael Car­ service bulletin 11: 3. penter and Elaine Svenonius. Littleton, ---. 1990. Edition or copy. Cataloging Colo.: Libraries Unlimited, 1-14. Service Bulletin 49: 10. Phinn, Kevin. 1990. Shot by shot [column]. Library of Congress. Descriptive Cataloging Disney restores Fantasia, its most ambi­ Division. 1949. Rules for deSCriptive cata­tious feature. Premiere, Nov., 84-89. loging in the Library (if Congress Washing­ Ravilious, C. P. 1975, Analysis and evaluation ton, D.C.: LC, DCD. of eXisting catalogUing codes. In A survey Library of Congress. Processing Dept. 1946. of eXisting systems and current proposals Studies of Descriptive Cataloging: a Re­for the catalogUing and description ofnon­ port to the Librarian of Congress by the book materiaL~ collected by libraries, with Director ofthe Processing Dept. Washing­ preliminary suggestions for their interna­ ton, D.C.: Govt. Print. Off, tional coordination. Paris: United Nations II MacLaury, Keith D. 1979. Automatic merging Educational. Seientific and Cultural Or­ of monographiC data bases-use of fixed- ganization. II til •