(large image: 62 Kb)
Ligue des Bibliothèques Europeènnes de Recherche, Groupe des Cartothécaires de LIBER
TRA NSLATE ENGLISH to Français, Deutsch, Italiano, Português, Español! Explanation
© LIBER and author
For the Map Department of the Royal Library the decision to retro-convert was easily made, as we were asked by the head librarian himself to do so. The Map Department had already produced "The Danish National Bibliography. Cartographic Materials 1990", based on the map department's computer records in the Royal Library's bibliographic database REX. This being a success, it seemed natural to make the quinquennial cumulation of "The Danish National Bibliography. Cartographic Materials 1986-1990" in the same way. As the records for 1990 were already in REX, we had to retro-convert the entries for 1986, 1987, 1988, and 1989.
We were of course asked to make an estimate as to how much extra work this would involve. The
number of records in question was about 800 and after having made a few tests at reformatting
different types of catalogue cards we estimated that on average it would take 15 minutes per record
to
convert a catalogue card into danMARC (CM), our machine readable format for cartographic materials,
by keyboarding.
Realizing that this was a rough estimate, and that the project would involve quite an amount of
pre-editing and proofreading, we allowed for an overhead of 8 minutes per record for this.
Accordingly, we applied for what equals 1 fulltime staff member for two months or 272 what we call
"wage hours", i.e. including social time, holidays etc. As it is not realistic to assume that 1 person can
sit at a terminal and do this type of work for a whole day, we asked for two persons and were allotted
two clerical staff members familiar with cataloguing.
Preparations.
We have always maintained that the Danish National Bibliography for Cartographic Materials
was
a by-product of our catalogues. Before 1990, the annual publication was produced by way of copying
our catalogue cards and adapting these to the format of the bibliography by providing the entries with
special headings consisting of geographic name subject heading, subject heading, and scale, and the
price of the item, and removing local notes and classification codes etc. Since 1986 we have kept an
extra copy of the catalogue cards in their original form, because we already then had the idea that they
might come in handy for the cumulated edition.
We now collated the catalogue cards with the entries in their respective bibliographies, and since we are four librarians in the map room, each of us was responsible for one year as regards collating and amending the catalogue cards, and after the conversion, proofreading the prints of the formatted records. Since 1982 we have catalogued according to the AACR2(CM). But the interpretation of cataloguing rules can be very difficult and has given rise to many a discussion in the map department, succeeded often by a change of practice. Therefore, the earliest of the bibliographies did differ from the later ones, and as they were to appear in one cumulation, we felt it to be important that they were of a uniform standard. For example in 1986 and 1987 information about printer achieved from other than the prescribed sources of information would appear in the note area, whereas this information now appears in the edition field, i.e. field 260, in square brackets.
As the note area is divided into different note fields in the MARC format, we added the relevant field tags to the catalogue cards.
Illustration 1a: Catalogue card
1111-39-1986/6
Dinamarca de todo corazón : mapa e información de viajes 1986 / Oficina de Turismo
de
Dinamarca ; LegindKort/Folia.
Malestok 1:730.000.
Copenhague : la oficina de turismo, 1985 ([Herlev] : Recato).
1 kort : i farver ; 49 x 42 cm, falset till 21 x 10 cm.
Omslagtitel.
1- forts
1111-39-1986/6
Dinamarca de todo corazón : mapa e información de viajes 1986 /
Trykt recto-verso.
Kort over Bornholm indsat.
Med vejlaengdetabel og turistinformationer.
1986-427/2 2
Illustration 1b: Entry in the bibliography
Danmark. Turisme. 1:730.000.
Dinamarca de todo corazón : mapa e información de viajes 1986 / Officina de Turismo
de Dinamarca ; LegindKort/Folia.
Malestok 1:730.000
Copenhague : la oficina de turismo, 1985 ([Herlev] : Recato).
1 kort : i farver ; 49 x 42 cm, falset til 21 x 10 cm.
Omslagtitel
Trykt recto-verso
Kort over Bornholm indsat.
Med vejlaengdetabel of turistinformationer
As you will see from the entries in the bibliography, each entry has a heading consisting of geographic name subject heading, subject heading, and scale. But the catalogue cards have other headings and also additional information regarding classification number, shelf mark, acquisition number and local notes. This information had to be integrated into the computer record. We also added other relevant subject headings to the catalogue cards. So this meant a lot of preparatory work, and as we got the staff for the retro-conversion before we had finished amending all the records, we were under some stress getting the rest of the cards ready and at the same time beginning the proofreading of their work.
The clerical staff were familiar with cataloguing monographs, but the format for cartographic materials differs from that for monographs in many ways. Especially formatting map series in field 248, a field that is not in use for other materials than maps in the Royal Library, caused a certain amount of difficulties.
Illustration 2: Series entry
Danmark. 1:25.000.
Danmark 1:25 000 (4 cm) / Kort- og Matrikelstyrelsen.
Malestok 1:25 000 ; projektion UTM zone 32 og 33 (Europaeisk Datum 1950).
Kobenhavn : KMS, 1957- (Kobenhavn : KMS).
kort : i farver ; 46 x 57 cm.
Serien omfatter 405 bl., alle kort er med UTM-referencenet, de nyfremstillede blade fas ogsa falsede
til 30 x 13 cm.
Signaturforklarinf pa hvert enkelt blad.
Med indlagte sognegraenser.
Hojdeforholdene er vist med hodjetal samt hojdekurver med 2,5 m interval.
Pr. bl. kr. 55,00.
1116 II SO :Roslev. - 2.1.1. - 1991.
1213 I NO :Hornsyld. - 3.0.1. - 1991.
1213 I SO :Stouby. - 3.0.1. - 1991.
1213 IV NV :Billund. - 2.0.1. - 1991.
1213 IV NO :Jelling. - 2.0.1. - 1991.
1213 IV SV :Vorbasse. - 2.0.1. - 1991.
1214 I NV :Silkeborg. - 2.0.1. - 1991.
1214 III NO :Norre Snede. - 2.0.1. - 1991.
1214 III SV :Give. - 2.0.1. - 1991.
1214 III SO :Kollemorten. - 1991.
1215 I NV :Klejtrup. - 2.1.1. - 1991.
1215 I NO :Onsild. - 2.1.1. - 1991.
1215 I SV :Tjele. - 2.1.1. - 1991.
Each of our map series has always appeared as one entry in the bibliography. This appearance of a multilevel record we preferred to maintain. To obtain this, we really had to strain REX, but we did overcome the obstacles. What we do in practice is to make an estimate for each map series as to how many sheets it will end up consisting of. As no one record in REX can be longer than a hundred lines, we allow a maximum of 50 map sheets into each record. Then we repeat the record and start on the next 50 sheets in numerical order, filling in only the 248 field and so forth and so on for the rest of the series. Each map series gets an identification number so that the individual records can be linked together in REX and in the bibliography by this number.
For this purpose we use field 016 sub-field a. For sorting the records we use field 016 sub-field b both to get the records sorted correctly in the bibliography and have them shown in correct order in REX. As we need only one first level series record for the bibliography we have to "throw away" so to speak quite a few first level records.
These are marked in field 016 with first and second indicators 00, whereas the record that has to be retained is marked with first indicator 9 and second indicator 0.
Ill. 3. Field 248 in the danMARC format for CM.
248 0A R Second level information related to a multipart item
*G Sorting key
*g Number or designation of volume or part
*a Title proper
*c Subtitle or other title-information
*p Parallel title proper
*x Title proper by another author
*e Other first statement of responsibility
*f Supplementary statement of responsibility
*t Parallel author
*w Edition statement
*m Scale
*i Edition statement
*j Year
*k Physiscal description
*l General notes
For future use we have made a sub-field capital G in field 248 which we fill in with a number for sorting purposes. In a new version of our computer system we hope to get a facility that will enable us to send a correction to a certain record in REX instead of having to download the entire record, amend it, and retransmit it. That would mean that we could send information to a record in REX concerning a single sheet in a map series. But for this information to find its correct place of order within the record we need the sorting key in field 248 sub-field capital G.
Before we started computer-cataloguing we only catalogued our map series for the bibliography and crossed them out on index cards. For the quinquennial cumulation of the bibliography, we used to take the 5 individual bibliographies, literally cut out the map series records and merge the 5 entries into one entry. As regards the retro-edition, we had already made records for many of the map series on computer for the 1990 edition, so these we could download and add to, but quite a few of the series we had to make from scratch, and realizing that it would be too time-consuming for us to teach others to do this, we ended up doing the map series ourselves.
Priming the clerical staff
They were both given a brief introduction to the Map Department and the maps. One preferred
to
work in the map room, the other at her own office. So for her I prepared two floppy disks and a
sample of what information were held in our special function keys. For the first records, we simply sat
together reformatting the records. They also got a small self-made manual and a copy of a catalogue
card with MARC-corrections on to use as an example.
Having formatted a record they would run their pen through the entry in the bibliography. At the end of the day, they would make prints of their work for proofreading. In the beginning, it was necessary to proof-read the corrected records once more, but later on it proved unnecessary. When a group of fifty records were correct, they would transmit them to REX and cross them out in the bibliography, after having checked that the records were updated in REX.
Ill. 4. The computer record.
s5 ; r1
001 00 *ax500015464
008 00 *as1985*bdk*lspa*se*tm*v0
089 00 *a1111*b39*c1986
245 10 *aDinmarca de todo corazón*cmapa e informací de viajes 1986*eOficina de Turismo de Dinmarca ; LegindKort/Folia
256 00 *aMàlestok 1:730 000
260 00 *aCopenhague*bla oficina de turismo*c1985*r<Herlev>*tRecato
300 00 *a1*nkort*bi farver*c49 x 42 cm, falset til 21 x 10 cm
530 00 *aKort over Bornholm indsat
530 00 *aMed vejlængdetabel og turistinformationer
559 00 *aTrykt recto-verso
559 00 *aTitlen gengivet fra omslaget
595 00 *aKF86
599 00 *bKBK910516
620 00 *aDanmark*bTurisme*c1:730 000
710 20 *aOficina de Turismo de Dinamarca
710 20 *aLegindKort/Folia
945 00 *aDinamarca*wDanmark
b01 00 *aKBK
k01 00 *aDanmark
k02 00 *aSpansk; Castiliansk
k13 00 *aBog;Mono
k14 00 *aKartografisk materiale, trykt;Kort
k16 00 *a
The project started mid-May, and by 23 August 1991 all records were reformatted and transmitted to REX. As our bibliography is normally published together with "The Danish National Bibliography. Books", our deadline with the publishing institution was 1 October.
The next phase took place in the computer department and the map department respectively.
Ill. 5. Procedure in the computer department.
(large image: 62 Kb)
The data had to go through various processes in the computer department. The year previous we had already made a conversion table, so that the formatted records would appear in the proper form for the bibliography, and also in the correct sequence. But as more data have to be handled, more errors will show. So the first proof had to be checked both for system errors and formatting errors. The proofreading was done in the map department by the three librarians proof-reading a third each, and then by me proof-reading the entire bibliography. Then we collated our corrections and downloaded those records that had to be adjusted, corrected them and retransmitted them to REX. The system errors I would discuss with the computer department for possible solutions. The computer department now repeated their procedures and a new proof was made for the final proof-reading. Errors found in this phase can only be corrected in either the word-processing program or in the desk top publishing program Ventura, where the lay-out for the bibliography is made.
As we had a deadline with the publishing institution, our biggest problem in this final phase was the
fact that our online system frequently was "down" during this period, causing us a considerable amount
of stress. A short period before deadline however, we were told that the publishing institution had
decided not to publish a quinquennial cumulation of The National bibliography for books anymore, and
as the bibliography for maps is bound together with this publication, our bibliography would not be
published either!
You can probably imagine the feelings that took hold of us when we learned about this. We had all
put
our hearts into the project and worked really hard to achieve a good result. Of course the work had
not
been done in vain, the records had been retro-converted and that in itself was something, but...
As some of you already know, a solution was found. We decided to publish the bibliography ourselves!
We had it photo-typeset at the university computer centre, and subsequently printed in the Royal
Library's own printing office. This procedure also allowed us more freedom in regard to choosing
lay-out, binding etc.
As regards distribution, we chose to give all public library systems in Denmark a free sample, as well
as the relevant research libraries and the institutions and persons traditionally on our mailing list, with
an accompanying letter explaining the circumstances leading to this procedure. If more samples of the
bibliography would be required, it would then be at a cost. This solution was comparatively inexpensive for the library, and we hope, of good public relations
value.
Conclusion
Exactly how long time it took to reformat a record on average, including the time we spent on
enriching the records and the proofreading, we have estimated to half a "wage hour" or 22 net
minutes. As it was done for two purposes, the bibliography and our online computer base, it was more
time-consuming than ordinary retrospective conversion. The bibliography must appear as a
homogeneous unit, because the entries are in print next to each other, as opposed to an online
catalogue. Also it took some time before the clerical staff became familiar with the technical terms
used for cartographic materials, the format for maps, and the working process in general. They soon
became very professional, so should we have continued retro-converting, the time spent per record
would have been considerably lower.
Retro-conversion by keyboarding as opposed to scanning is a slow process, but has the advantage, when done in-house, that relevant information can be added, such as subject headings, subsequent classification codes, etc. I do, however, foresee many difficulties when it comes to the retro-conversion of our handwritten catalogues. Some of them are even in gothic lettering, the cataloguing differs from AACR2, the classification codes are out of date, and the information level in general is lower than that of today. In my opinion though, it would be all too time-consuming to compare the catalogue card with the item itself. So albeit on a lower level, I think it preferable to have all catalogues converted to our online system than to have both card catalogues and an online catalogue. As our card catalogues are arranged according to area and subject, we hope, that in a future conversion process done by scanning, it will not prove too difficult to enrich the computer records with this information.
The retro-conversion policy of the Royal Library in general.
The policy of The Royal Library as a whole is to have all catalogues retro-converted by the year 2000,
primarily by means of scanning. In 1991, the library started a scanning project, and by now, September
1992, already 300,000 catalogue cards have been scanned, reformatted by computer, and are retrievable
in a sister computer to REX. The computer programme for conversion of ASCII-strings to our marc
format was developed by the Royal Library itself, and the standard of a reformatted record is on a very
high level, needing only cursory proofreading and few subsequent corrections. Since Autumn 1991, the
Royal Library has offered an automatic lending system to our readers, so the books can be ordered
or
reserved online. Therefore it is of great importance, that as many items as possible can be found in
REX. This of course puts great demands on the computer capacity, which is in the process of being
enlarged at the moment, allowing for a total amount of 2,230,000 records. Of the scanning project as
such I am not at liberty to unveil much. But in December 1992, a seminar on this specific subject will
take place at the Royal Library, throwing more light on the matter.
Back to GdC Homepage