Newsletter no. 74, November 1999

Q: I recently purchased a "shaped" CD, the shape of which isn't the standard 4 3/4 in. circle, and it got me to thinking about how these things should be described. How do you reflect its dimensions in the 300 $c? Would you still give its diameter in inches (as for any other sound recording), or would you use centimeters? The container mentions that this type of disc is "Not playable in CD changer or car CD player." Would you put this information in a 538 note, or just a standard 500? The disc is (very) roughly triangular in shape. At its widest, the CD's diameter is about 4 1/2 in. The disc (for want of a better term) is an interview CD featuring the British heavy metal group Iron Maiden. It was published in 1997 by Sonotec for their Private Talks line, manufactured by Cuba GmbH, Berlin (in fact, they've trademarked the term "Shape CD"), and distributed by DA Music, Germany. The disc itself features a lineup picture of the band with their cadaverous mascot "Eddie" looming over them. The CD is shaped so that it's broad at the bottom and tapers off towards the top. For fans of "Star Trek: Deep Space 9," it looks very much like a top-down silhouette of the USS Reliant. (How do you like that for an analogy?). The digital information itself is still in circular rings, but takes up only 1/2 to 3/4s of the disc's surface. For an actual picture of this "disc" (as well as others), take a look at the following Web site: http://members.xoom.com/shapecdclub/shapecat9.htm (mine is in the upper left corner of Seite 9).

A: Presumably, even though the shape is a bit odd, the CD is still designed to play in most standard CD players (with the exceptions noted). That means that its widest diameter should still be considered the standard 4 3/4 in. and should be so described in the 300 subfield $c (not that 6.5D2 gives us much choice in the matter). I am guessing that the actual playing surface (with the pits and spaces to be read by the laser) remains circular within the odd shape. In accordance with 6.7B10, try to describe the physical shape, perhaps something like:

500 Disc is roughly triangular in shape.

Optionally, as you describe the shape, you could also describe the illustration and how it defines the shape of the CD, for example:

500 Disc is roughly triangular in shape, with portrait of the band across the base and its mascot Eddie at the apex.

If the term "Shape CD" appears on the item, it would be a logical quoted note (500). Additionally, include the statement about the disc not being playable in certain types of machines, either in quoted form or in paraphrase, whichever is appropriate. I think this note fits into the definition of, and could properly be coded as, 538. One also wonders if this is some sort of limited or otherwise special edition, with a standard round disc also being available. If there is some indication of that on the item, you can transcribe an edition statement or optionally formulate one as per 6.2B3.

Q: We have several fragments of scores that are not identifiable, except that John Philip Sousa is the composer and they seem not to be from the same work. We wondered if (a) all of these could be included in one record (yes, according to AACR2?), and (b) could we use the phrase "sketch score" to describe these fragments, since they appear to be incomplete motifs, not fleshed out. I have never heard or seen that phrase used before but that doesn't mean much. Many of these fragments are parts of vocal scores of Sousa's operettas, I think. Some have never been published.

A: You may compile such miscellaneous sketches into a single record, though you'll have to supply (I am supposing) some comprehensive (bracketed) title. If they can be identified as some particular genre of works, as you suggest, you would use the collective uniform title for that genre with the qualifier "(Sketches)" as per AACR2 25.35B. If they cannot be narrowed down, you'll need to be more general (such as "Selections (Sketches)"). The "(Sketches)" qualifier is not separately subfielded.

Optionally, as you describe the shape, you could also describe:

100 1 Sousa, John Philip, $d 1854-1932. 240 10 Operas. $k Selections (Sketches) 245 10 [Sketches for miscellaneous operettas].

If you do a search on "beet,sket" you'll find lots of examples to get ideas from. The specific phrase "sketch score" is not familiar to me, but there's no reason why you couldn't use it in a supplied 245 title.

Q: In hand is a two-page fragment of a vocal score. Would this be described in the 300 field as "1 vocal score," despite the fact that it is obviously incomplete? Also, how best to describe it in a 500, as "fragment" or "score fragment" or some other phrase?

A: The way you've described this in the 300 ("1 vocal score ([2] p.); $c 35 cm.") seems fine, but I'm not sure if the copy you have is actually missing pages or if this item is simply something that was broken off in the middle and is as complete as the item ever was. If the former, look at AACR2 2.5B16 and at LCRI 1.7B20. If the latter, you might expand the "Score fragment" note to explain that it's a sketch that breaks off after so many measures, or whatever is appropriate.

Q: Many of the items in question are signed by Sousa at the end. If I make a 500 note to indicate that the work is signed, do I use quotation marks? For instance:

500 Signed at end: "J. P. Sousa, Washington D.C., March 14th 1905."

or are the quotation marks superfluous?

A: As I read the rule 1.7A3 on "Quotations," if you are formulating a note as you have with an introductory phrase, you would not use quotation marks.

Q: I have some questions about when to input a new bibliographic record. "Bibliographic Input Standards," 2nd ed. says about field 245 subfield $b, "absence or presence does not justify a new record." Does this mean that if the item you are cataloging lacks a parallel title but a parallel title does appear on the record in WorldCat, you are required to use the existing record anyway?

About field 300 subfield $a,"Bibliographic Input Standards" says, "variation in preliminary paging, post paging, or separate numbering do not justify a new record." Does this mean that if you are cataloging something with six preliminary pages and the record in WorldCat shows eight preliminary pages, you are required to use the existing record? The situation I have is a score that is published with title proper and accompanying material in German, vs. an OCLC record for this title with a parallel title in English and a different number of preliminary pages, presumably because the preface is in English as well as German. It seems as though "Bibliographic Input Standards" prevents a new record from being created in such cases. Is this true?

A: That section of the old "Bibliographic Input Standards" was superseded by the introductory chapter 4 (p. 45- 57; or http://www.oclc.org/oclc/bib/chap4.htm in the electronic version) of "Bibliographic Formats and Standards," "When to Input a New Record." One of those differences alone might not justify a new record, but evidence in the existing record of the presence of a preface in both English and German, versus your own item in hand with only the German language preface, would certainly justify a new record. Optionally, if you want to be a real stickler, you could indicate in your record that another version is available with the additional English language preface.

Q: In hand is a CD called "Music of the ancient Greeks," performed by an ensemble called De Organographia. The CD label has p1995. The back of the container has "c1995, 1997." My understanding is that I should infer 1997 as the date of publication, putting it in brackets in the 260. Then I'd put a comma, and record "p1995" after it. There are two records in OCLC which appear to be a match in every respect except the date. One is I-level and uses p1995 in the 260 AND it was input in 1995. The other is M-level and uses just 1995 in the 260, but it was input in late March of 1997. The only possible difference is in the notes. Mine include explanatory notes, including a description of each of the instruments, information on each piece, including the source, the translation for vocal items, and sometimes more, plus a couple of sentences about each performer. The first OCLC record mentioned above (#32913801) says only that there are explanatory notes and translations. Maybe that's the same as mine, maybe not. The other OCLC record (#38260033) says there are program notes, translations, and details of the instruments and sources. That clearly is the same as mine; the biographical information is so brief that it might not have been mentioned. I'd say that my item has to be different, at least bibliographically, from the one input in 1995. But is the second record in OCLC, input in 1997, the same or different from the first? Seems my item matches that second record, especially considering the notes match, and they could have failed to see the 1997 on the container.

So, the questions: 1) What's going on here? If the publisher did expand the notes when more copies were molded of the CD, they would definitely want to copyright them. That's the only thing I can think of. 2) How many records does OCLC want? Should I ask that the two existing ones be merged, then input a new record for my item? Or attach my holdings to the later record? I have run into this situation a fair number of times recently. It happens a lot on Hungaroton. The record in OCLC is a match except it has "p1996" and my copy has "p1997" or "p1996, 1997," or something similar. If I had some idea of why this is happening I could make a more informed decision.

A: It would be impossible to generalize about what might be going on here. In some cases, catalogers may have missed a later date, but it's just as likely that the later date was not there. In this case, since everything was the same in both records (I don't consider the 1997 input date on the ELvl "M" record to be a significant difference), I merged them. You are perfectly justified in adding a separate record with the bracketed "1997" date from the container.

Q: The Art Section a large public library wants over 450 titles (about twenty screens of 505s) keyed into an existing record so that every title in this set will be indexed in the OnLine catalog. Is this sort of thing done? Occasionally we enter song titles in 505s, but this request is for every title in every songbook and CD set to be put into the MARC records. I'm at a bit of a loss to explain why this is a distorted use of the MARC record, poor cataloging practice, and a possible cause of carpal tunnel syndrome among our Library Technicians. Just because it can be done to provide keyword access to titles is not the best reason for doing it. Suppose we could scan the titles from the packaging? There's also the consideration of impact on current cataloging production and limits to field length. I've suggested alternatives to such extensive indexing in the MARC record (conversion of the local card file to a separate database, or a separate song index on a locally-maintained database, or a partial listing of contents), to no avail. We've also pointed out that, since the set says it includes "all known recordings," there's no reason to particularize. Might this be the direction in which MARC records are going?

A: In trying to add such a contents note to the OCLC record, one would quickly run up against one or another of the system limits for the number of characters per field (1230 maximum), the number of variable fields per record (approximately 50), and the number of characters per record (4096 maximum). Additionally, the Library of Congress Rule Interpretation for AACR2 2.7B18 (covering contents notes, generally) gives some guidelines about formulating notes that are much more restrictive than is being suggested in your message. One legitimate possibility that might help would be to enter a separate bibliographic record for each of the twenty-four volumes (or some other sensible breakdown of the volumes), with a complete contents note for each (if it does not exceed system limitations). OCLC allows such "analytic" records.

Q: In the Sound Recording 007 field, subfield $c (REC 007/02, Original versus reproduction aspect), how do I figure this out? Are all mass produced items reproductions, or what? Is it important to code this element? Moving on to the Sound Recording 007 field, subfield $m (REC 007/12, Special playback characteristics), often a tape will say Dolby (or have the Dolby double D sign) but the choices for coding seem to be Dolby-A, Dolby-B, or Dolby-C. Do I assume one of these choices, or mark unknown, or other?

A: The subfields $c (007/02) of all the 007 fields have been made obsolete in the newly published MARC 21. OCLC has long recommended omitting this subfield when you formulate any 007 field. Eventually, we hope to do a scan to delete all instances of the subfield from WorldCat. In the Sound Recording 007 field, subfield $m, code "c" (Dolby-B encoded) is used for the standard Dolby noise reduction found on most commercial audio tapes (often indicated by the "double D" insignia). Dolby-A and Dolby-C are techniques used in special circumstances and would be explicitly indicated as such on the item.

Q: When is a song not a song? I have a couple of collections of instrumental versions of Hebrew and Yiddish songs. Do I still use the subjects "Songs, Hebrew" and "Songs, Yiddish"? If not, what would I use?

A: Check out the subject heading subdivision "Instrumental settings" (sh99001570, sh85066779, and in the LC Subject Cataloging Manual).

Q: I'm a little confused about geographical subdivisions. I've seen on some records that "$z United States" is added when the music is sung/produced in the U.S. regardless of the origin of the music. My instinct would be that if people in the U.S. were singing European songs to either indicate the European country/region or leave off the geographic subdivision altogether. Any words of wisdom on this?

A: The best guidance on this is LC's Subject Cataloging Manual, especially H 1916.3, H 1916.5, and H 1917. SCM H 1916.3 Section 3j specifically addresses geographic subdivisions, referring to other parts of the manual. Usually if the item itself calls attention to a national, ethnic, or religious aspect of the music (in the title, series, etc.), subject headings to bring that out are appropriate. Subdividing everything by "United States" is neither helpful nor appropriate.

Q: How should we describe the upcoming "DVD-Audio" and "Super Audio CD" formats? Would the 007 field remain the same as for regular CDs?

A: In terms of physical description, both DVD-Audio and Super Audio CDs would have the same 300 field as those of standard audio compact discs, since they all look identical.

300 1 sound disc : $b digital, stereo. ; $c 4 3/4 in.

The designation of playback channels (stereo, mono, etc.) would depend on the individual item, of course. To distinguish either of these new formats, I would suggest using the 538 "System Details Note" to indicate which format is represented. Transcribe the format as it appears on the item.

538 DVD-Audio. 538 Super Audio CD.

Seeing that these formats are technically different and at least currently incompatible, they feel like the audio equivalent of the VHS/Beta difference in videos, so I think this treatment makes sense. The information I've seen on these formats has not said anything about the speed of these discs; if the speed is other than the standard for current audio CDs (1.4 meters per second), then you'd have to code 007 subfield $d with "z" for "other" or "u" for "unknown." If these discs are truly able to play back with multi-channel capabilities (that is, as many as six speakers, according to some of the advance hype), subfield $e would have to be coded "z" for "other." As far as I can tell from the available information, the remaining elements would not change.

Q: What do we do when a CD is re-issued under a different title than the original release, with the new title taken from a different song title on the CD, but the new title is misspelled everywhere on the new release?

In hand I have a CD with the publisher number "Classic 7724," published by Classic Sound, Inc. (Norcross, GA), performed by Art Blakey and the Jazz Messengers. The title on all parts of CD reads "Rucerdo;" the song list reads: Buttercorn lady, Rucerdo, The theme, Between races, My romance, Secret love. There are no dates whatsoever on the CD, but it is obviously a re-issue, although I've been unable to determine the date of the original issue. On the Web, I found an Art Blakey discography that lists and pictures a 1966 album titled: "Buttercorn lady" (Limelight 82034) with the exact same song list, except the one song is titled "Recuerdo" and not "Rucerdo." Recuerdo is a legitimate word in Spanish. As far as I know "rucerdo" isn't a word in any language. I feel sure the item I have in hand is a (possibly unauthorized) re-titled re-issue of the 1966 album. I really don't like having to exactly transcribe a title that is so obviously wrong and without meaning in any language. Here's how I'm thinking of describing the item:

100 1 Blakey, Art. 240 10 Buttercorn lady 245 10 Recuerdo $h [sound recording] / $c Art Blakey and the Jazz Messengers. 246 1 $i Title on disc label and container misspelled as: $a Rucerdo 500 Originally issued in 1966 under the title: Buttercorn lady.

Is this description acceptable?

A: Since you have determined through your research that the item is some sort of re-issue, incompetent though it may be, you should include that information in the record as you have in your proposed 500 field. Although I don't believe that the use of the original release's title as a uniform title would be called for, a related title 740 for "Buttercorn Lady" would be fine. The use of that uniform title under Blakey's name implies first of all that he is the composer of the entire collection and secondly that these pieces were composed, presented, and intended as a collection by that composer. Since Blakey is the main entry by reason of his being the chief performer rather than as composer (or so is my guess), a uniform title is not appropriate here. Regarding the misspelled title, AACR2 1.0F would have you transcribe the title as it appears, followed either with "[sic]" or "i.e." and a correction. The latter would seem to be the way to go in this case. LCRI 21.30J gives guidance about access to both the incorrect and corrected form of the title.

245 10 Rucerdo [i.e. Recuerdo] $h [sound recording] / $c Art Blakey and the Jazz Messengers. 246 3 Recuerdo

You could use the same "i.e." technique in the 505 field where the title is incorrect.

Q: I really like the ability to make a new record from an existing record, and would like to use it in cases where the bibliographic information is useful, but the item I am cataloging is in a different physical format, so that the record Type needs to be changed. When I simply type over the existing code for "Type" and send the change, the system will not allow me to keep that change, even before I type "new" (to make a new record). Instead it gives the error message, "Not authorized to change record format." I'm pretty sure it can be done, but I don't know the right way to do it.

A: In 1998 OCLC installed a new capability that allows users to change the Type Code on all unlocked or locally edited bibliographic records to any other valid Type Code value, but this technique cannot be used on a workform (that is, a record that is not yet added to the database). The announcement of this is in OCLC System News as item "Type Code Editing Capability" (just type "news" and <send> in Passport and browse the list of news items). When you are doing a "new" command, you can change the Type simultaneously with that "new" command by entering "new wfXX" where "wfXX" is the workform command for the Type you want. So if you have a Books format record and want to do a "new" command to change it to a Sound Recording, for instance, type "new wfmj" and the information will transfer to a Sound Recording workform. This is documented in Technical Bulletin No. 209, section 2.1 (on the OCLC Web site at http://www.oclc.org/oclc/tb/9417pris/9417.htm#2.1), and is being incorporated into the new third edition of the "OCLC Cataloging Service User Guide," which is currently in the works.

Q: There is an OCLC record that seems to have a problem with the 028 fields. It indexes only the first 028 field (which has a set number and a double-dashed range for the individual numbers in the three-CD set) for a music number search. I Enhanced the record and retyped the next 028 field, but it still does not index this line.

A: Only the first 028 is being indexed in this instance because it is a case where the range of music publisher numbers increases in increments not by the final number, but instead by the next-to-last number. Although there are only three discs and three individual numbers associated with them (5 56221 2, 5 56222 2, and 5 56223 2), when these are entered as a range (5 56221 2--5 56223 2), every number in between (up to the system limit of twenty numbers) is also indexed (that is: 5562212, 5562213, 5562214, 5562215, etc.) As "Bibliographic Formats and Standards" says ("Ranges of consecutive numbers in increments of more than one," p. 0:74, http://www.oclc.org/oclc/bib/028.htm in the electronic version), such numbers must be entered separately in order to index correctly. I've revised the record, splitting the numbers into individual 028 fields, which all should now index correctly, and added an explicit 500 field for the whole range.

Q: "Bibliographic Formats and Standards" says that the first language code in field 041 must agree with the Language code in the fixed field, and that, in the absence of a field 041, the system uses the code in the Language fixed field to determine filing in certain title and heading fields. What happens when the 245 title is in one language, but the 041 subfields $a or $d and/or the fixed field code is for the content of the item in hand and it differs from the 245 title? Does this confuse the system as far as indexing? This scenario happens to me a few times a month. For example, a sound recording of Rossini opera excerpts called "Rossini in Vienna." The 245 is in English, but the 041 subfield $d and the Language fixed field coded "ita." Basically, it is an American label production with program notes in English, but the sung content of the album is Italian.

A: As long as the relevant filing indicators are correctly coded (in fields 130, 222, 240, 243, 245, 440, 730, 740, and 830), the system will generally be able to keep its indexing ducks in a row. There may be cases where an initial article incorrectly included in a corporate heading would cause the heading to be indexed wrong. For instance, if the Language code for an item was "N/A" and there was a 710 that read "I Musici", the system would not know that the "I" is an Italian article meant to be ignored. (The correct heading would have been "Musici" [n83129444], anyway.) Something similar can happen when an initial article is incorrectly included in one of the indexed title fields that lack a filing indicator (fields 212, 246, 247, 780, and 785). For instance, if the Language is coded "fre" and there is a 246 that includes an initial article "The," it will be included as part of the indexed title since it is not recognized as an English article meant to be ignored. In most cases, if you make sure the filing indicators are correct and that initial articles are neither included in corporate headings nor transcribed in fields without filing indicators, there should be no indexing problems.