|
FLHIG Home page |
List of Online Articles | Newsletter contents | Site map | Links |
| This article first appeared in SIGGNL 21 pages 4 to 13 | |||||
|
Further Comments
on the NCA Rules for Indexing Names |
|||||
|
Part I |
|
The review by Graeme Muirhead of the Rules ... appearing in the last issue of SIGGNL, and previously in the Library Association Record, has given rise to further correspondence which will appear in SIGGNL as space allows. I thought it might be useful to begin with a summary of the rules dealing with personal names, as that is the section of particular importance to indexers of family history. These rules are headed as follows: 1. Personal names |
|||
|
|
|
For present purposes we will ignore rules 7 and 8 and the appendices: while appendix 1 contains a useful bibliography, we do not have the room to describe it here. |
|||
|
Personal names |
|
Concerning personal names, the NCA rules prescribe that if a surname consists of more than one word, the last word will be selected as the entry element. While this agrees with AACR2 in many cases, there are exceptions as noted by Graeme Muirhead. The exceptions are however mainly for non-English names, and therefore perhaps not so important for most readers of SIGGNL. Hyphenated surnames are entered in NCA rules under the last word, but the hyphen is always retained, thus: Lewis, Cecil Day-. To my mind this is inelegant, as well as less easily located: most users would look first for Day-Lewis. |
|||
|
Forenames |
|
Forenames may be used as the entry element in much the same instances as in AACR2, such as royalty, saints or bishops, where there is no identifiable surname, and may instead be identified by such qualifications as toponyms, epithets or patronyms. These are described in the NCA rules as qualifiers, for which the AACR2 does not seem to have a collective term. The terminology of the two sets of rules is very different, and in many cases that of the NCA strikes me as easier to understand, although I am more familiar with AACR2. |
|||
|
Changes of name |
|
Another instance of varying treatment between the two rules concerns changes of name. AACR includes the rule under choice and form of name, giving in order of preference: 1) the name by which the author is generally identified in reference sources; 2) that by which most frequently identified in the author’s own works; or 3) the latest name used by the author. The NCA rules, however, have a separate rule on change of name (§2.3A.4), using as authority the last, latest or most recent name, with some exceptions for married women. |
|||
|
Pseudonyms |
|
In the case of pseudonyms, variations of treatment depend mainly upon the fact that AACR are mostly concerned with persons as authors, whereas persons covered by the NCA rules for pseudonyms may be of other categories, such as actors. This is in fact the basis for most of the variations between the two sets of rules, as one might expect, so there is no particular virtue in preferring the AACR approach, when considering rules designed for archivists. Compare for example the NCA requirement that where available all forenames must be recorded in full in the authorised name, to distinguish and identify the individual precisely, whereas the AACR favours the commonly found form. |
|||
|
Use of husband's initials |
|
The NCA treatment for a married woman identified only by her husband’s initials (§2.4.3) is I think quite different from any AACR approach. The woman’s unknown forename is represented by a dash, and her husband’s name is used as an epiphet, thus:
The AACR form, Johnson, Mrs C. (§46G3) is not accepted by NCA. |
|||
|
Dates |
|
The addition of dates, usually of birth and/or death, are defined in rather different ways in the two sets of rules, though the results are generally similar. NCA uses c instead of ca for circa, and defines its meaning as five years before or after the date given. |
|||
|
Family type entries |
|
The AACR deals with the “... family” type of heading as a footnote to collections of manuscripts, so the more detailed treatment in the NCA rules is more useful in indexing genealogical materials. The heading may comprise the following components:
Thus a full entry might have the following headings:
|
|||
|
Conclusion |
|
In compiling an index for books to be used mainly by family historians, one is left with the dilemma, which of the two sets of rules is preferable. Many indexers, including myself, have been trained as librarians, and therefore automatically think in terms of AACR. The only major difference is probably the choice of the entry name in compound surnames and those with prefixes as pointed out by Graeme Muirhead. To follow the order used in the text which is being indexed would be a logical decision, but what does one do with authors who are inconsistent? I am grateful to the staff of the Shropshire Records and Research Centre for the loan of the NCA Rules ... and for other assistance in this and other projects. |
|||
|
|
|
|
|||
|
|
I was concerned to see in SIGGNL 20 that the National Council on Archives advocates departures from current standards for indexing names, as set out in the Anglo American Cataloguing Rules (AACR2) and supported in the 1996 British and international standard (BS ISO 999). This seems certain to cause confusion unless there is a clear definition of the types of index to which each of these conflicting standards should apply, and I can see no advantage in changing from current practices in genealogical indexes. My concern is the NCA proposal that all compound names, and names with separately written prefixes, should be entered under the last element. In England such names are usually listed under the first element or the prefix, which often becomes joined to the root name in the course of time. Thus Le Fevre from France often but not always progresses to Lefevre in England, and less often to Fevre. BS ISO 999, which I understand the Society of Indexers supported, refers to an earlier document to define standards for indexing names. This document and the subsequent AACR2 both recommend that in English-speaking countries compound and prefixed names should normally be listed under the first element. These were the rules quoted by Tony Rydings in SIGGNL 18, from which the NCA have chosen to depart. In his review Graeme Muirhead does question whether the NCA departures will really assist archive users; I will go further and suggest that they will handicap the users of genealogical indexes. In theory it is true that automated search systems could overcome the problem, but not all of them are all that sophisticated, and even the best ones can only retrieve a prefix if it has first been recorded; my own experience is that an indexer asked to separate the prefix from the surname can very easily lose it altogether. In any case most genealogists currently have to use fiche or printed indexes which can do no more than provide a cross-reference. Cross-references from known surname variants should always be quoted in an index, but they are not a complete solution to the problem. It is true that the conscientious index user will endeavour to search on all likely variants of a name, but they may be defeated by the sheer size of the task. In seeking my Le Maréchal family, I always check Le Marshall, but I usually abandon the task of checking every one in an endless list of Marshall. The NCA may have a need for rules which it can apply across national boundaries, but most genealogical indexes are based on national data. It seems to me that the advantages of the current practice of retaining prefixes and compounds in front of the root surname are too great to be lightly abandoned in English genealogical work, especially as this conflicts with the reference recommended by BS ISO 999. |
||||
|
|
|
|
|||
|
|
None of the standards mentioned are entirely satisfactory for genealogical purposes. Using the last word of a surname as the entry element, as the NCA Rules do, is not what we generally expect in Britain, though there are qualifications to this which I shall mention below. The other standards are designed for indexing books, periodicals or databases, or for cataloguing libraries. |
||||
|
Shortcomings of standards |
|
The American standard on basic criteria for indexes (ANSI Z39.4-1984) does not deal with the specific points discussed, but in the penultimate paragraph of §4.2, states that names of persons from countries or periods not following modern Western usage should be indexed according to the custom of the country or period. BS 3700: 1988, which is the latest British standard I have, states at §5.3.1.6, that compound and multiple surnames should be indexed under the first part, with any necessary "see" references from other parts, except where usage favours any other form. Custom and usage is not very helpful as a standard, because, unless the standard provides guidelines on how to decide upon the entry element, it requires the indexer to establish the relevant language, country or period, and research what is appropriate for that; as a professional book indexer, I don't mind doing this when indexing a book, because then I usually get paid for my time. However I would not recommend this procedure to an amateur indexer, unless they are prepared to meticulously research the appropriate custom and usage, and are quite confident they have got it right. The Concise AACR2 1988 revision, at §34C2, gives guidelines for various languages which are helpful to cataloguers. The Anglo-American Cataloguing Rules (AACR) gives similar guidelines at §46B. However, one cannot reasonably expect the user of a card catalogue, or the searcher of an index, to be familiar with indexing standards or cataloguing rules: they just have to bumble along as best they can. |
|||
|
Compound surnames |
|
As hinted above, what we usually expect in Britain and Ireland is for the first word of a surname to be used as the entry element. (However there are cases where compound names are indexed by the last part: for instance, the calendars at the Principal Probate Registry.) I don't feel it is particularly useful from a genealogical viewpoint to index compound names in hard copy indexes either by the first part or by the last part. Take Fuller-Acland-Hood, the surname of Baron St Audries; assuming that a searcher might be interested in any of these three surnames, cross-references would be included from each of its component surnames, with forenames as sub-headings:
Since there are unlikely to be many entries under Fuller-Acland-Hood, I feel it to be a waste of the [genealogical] searcher's time to insert these cross-references, and I would index as follows:
The cross-references would be replaced by a general direction in the introduction to the index to look at the end of the entries for a surname for compound surnames. |
|||
|
Titles of nobility |
|
BS 3700: 1998 states at §5.3.1.3 that if such [personal] names ... include a title of nobility, entry should be under the surname or under the title of nobility (according to the usage of the text), followed by other names, titles of address, etc. The same mutatis mutandis would apply to bishops and archbishops.
A peer or a bishop could be indexed in several possible ways: by title, with a cross-reference from the surname; by surname, with a cross-reference from the title; or both by surname and by title. The last method - both by title and by surname - is the one I favour. I find the above entries somewhat cumbersome, and would use the following entries:
If there are numerous location references, then space and keying-in may be saved by using a cross-reference: this will also save errors, but meticulous checking is part and parcel of the indexing process. If genealogical or local-historical material contains a reference to a peer or a bishop, but not including their surname (Charles, Baron Chesham; Gerald, Bishop of Lincoln), then research (with all due care and diligence) the surname - reliable reference works are available - and index them both by title and surname. It is particularly important that the surname entry should include, where the text omits the surname, a note of the title. |
|||
|
Prefixes |
|
The NCA Rules stipulate that, except for mediaeval (pre-1500) names covered by §2.3E, names with prefixes should be indexes by the last part, which does at least have the merit of consistency. The British custom of treating prefixes as part of the name is also fairly consistent, but in cases where the prefix is lost (Le Fevre >> Fevre, Le Marechal >> Marechal), some means is required of grouping the prefixed and unprefixed name together: this can be done by inserting cross-references. I wonder whether it is appropriate to do this without checking that where a name seems to have lost a prefix, it has actually done so. However, it is quite usual in Britain to treat foreign names according to foreign usage, which rather spoils this apparent consistency. BS3700: 1988 states at §5.3.1.5 that where a surname includes an article or preposition or both, the practice of the language [or country] concerned should be followed. The ALA Filing Rules, list examples rather than giving rules. The Concise AACR2 1988 revision is rather more helpful at §34D, providing guidelines for various languages - the relevant AACR reference is §46E-F - which again are quite useful for cataloguers and indexers, but are obviously not known to users! R.N. Collison deals with prefixed names in an over-brief attempt to survey usage in various languages. Similarly, R.F. Hunnisett also sets out on p51 some rules for modern prefixes, though I have to say his explanation is much more obscure to me than the Concise AACR2 rules just referred to. His explanation of mediaeval, pre-1500 prefixes on the page before is much more satisfactory: for mediaeval material atte, de, de la, du, le, la and les are all to be ignored and the surname indexed by the main part, which appears to be the opposite of what the NCA Rules recommend! He also includes a most useful section on Scottish names at p70- 71, though it is not specified that the names such as McDonald, MacMillan, M'Turk and Maginnis which are to be filed as though they began with Mac need to be Scottish or Irish names (this is made clear, for instance, in the BLAISE filing rules). One trembles at the thought of some innocent indexer or cataloguer filing Tom M'boya under MacBoya, even though he was unquestionably African! Perhaps this is why the current indexing standards call for names of Scottish origin to be filed "as is",. Regarding prefixed names, I don't feel there need be anything sacrosanct about sticking to the English custom of not inverting any prefixes. English people are used to looking up foreign names, which are sometimes inverted and sometimes not. There is something to be said for deciding upon one rule - either never invert, or invert prefixes but only one at a time, or always invert, whatever the prefix - and sticking to it as ruthlessly as you dare, or as far as your copy-editor or your project coordinator will allow. Is not each defensible, provided that the user is forewarned by the introduction to the index? Each of them will have problems: in the first case, may not Le Fevre lose its prefix? - solvable by including a cross-reference from Le Fevre to Fevre - and in the second, may not Le Fevre become Lefevre? - solvable either by splitting Lefevre, or including a cross-reference, as before. Or must you treat, for instance, Mac, O' or ap as prefixes? Each will involve the indexer in some research. My own instinct is for all names with prefixes to be filed under the main part, to avoid the need for indexers and users to learn the rules for the circumstances in which prefixes may be inverted, and when they may not. This would group Le Fevre with Fevre (as Fevre, Le), with a cross-reference from the prefix, advising the user to look under the main part. (I don't advocate treating the Celtic Mac - son of - or Nic - daughter of - in this way.) There are objections to this: one is that indexers may lose the prefix during the indexing process. The solution to this is to ensure that the indexing guidelines clearly state that the prefix should not be lost, and that the checking process picks up any errors and omissions. Alternatively, inversion of prefixes and insertion of cross-references should be dealt with by the project coordinator. (Cross-references from, for example, Le Fevre to Fevre, if there are any, will need to be inserted anyhow.) Another objection is that Marshal, le would be lost among a sea of Marshals, which could be avoided not just by ensuring that the prefix is not omitted, but by ensuring that Marshal, le is filed separately from Marshal, and particularly from Marshalle:
|
|||
|
Consolidated prefixes |
|
The main difficulty from my point of view is what to do about surnames such as Vandermoewe, Lefevre and Verwoerd - the latter is always spelt as one word even in the original language. The surname may still lose its prefix, when a cross-reference will be needed. Should an indexer be prepared to be so ruthless as to separate a prefix in a one-word surname? Should they, for instance, file Lefevre as Fevre, Le-? The major pitfall is that it requires the indexer to interpret what they find in the material, though I would put it a different way: to check whether what appears to be a prefix is in fact so. I feel it is unlikely that any indexer would be so foolish as to invert all one-word surnames beginning with the letters of a prefix, whether they are consolidated prefixes or not, but maybe there are some ticklish cases which I cannot call to mind at the moment. Readers may feel that these consolidated surnames are best left alone, whatever is decided about prefixed surnames in general, because a cross- reference to the main part will be included in either case. Both amateur and professional indexers may find themselves in a situation where they need to carry out some research in order to decide how to file a surname in the index. This is particularly true if they feel, as I do, that rules (and standards) are made to be broken: this should only be done if you know exactly what you are doing, and the needs of the user are constantly born in mind. |
|||
|
|
|
|
|||
|
ENDNOTES |
|
||||
|
(1) |
|
Rules for the Construction of Personal, Place and Corporate Names; National Council on Archives, 1997. |
|||
|
(2) |
|
Anglo-American Cataloguing Rules, 2nd edition, 1988 revision, edited by Michael Gorman and Paul W. Winkler; or the Concise AACR2 1988 revision, prepared by Michael Gorman, both published jointly by the American Library Association, the Canadian Library Association, and Library Association Publishing Ltd, 1989. |
|||
|
(3) |
|
Guidelines for the content, organization and presentation of indexes; BS ISO 999, 1996. |
|||
|
(4) |
|
Names of persons: national usages for entries in catalogues, 3rd edition; London: International Federation of Library Associations, 1977. Supplement, 1980. |
|||
|
(5) |
|
Anglo-American Cataloguing Rules, British Text. London: the Library Association, 1967. |
|||
|
(6) |
|
R.N. Collison, in Indexes and Indexing, advised that compound names should be arranged under the first part, with cross-references from subsequent parts. However he would not have excluded arranging them under the last part, if there were good reason for doing so, or under all parts, if there were only a few locations. |
|||
|
(7) |
|
A point made by R.N. Collison in Indexing books: a manual of basic principles; New York, John de Graff, 1961. |
|||
|
(8) |
|
ALA Rules for Filing Catalog Cards ..., Pauline A. Seely (ed); second edition abridged. Chicago: American Library Association, 1968. Perhaps the unabridged version gives more guidance; I do not know, never having seen it. |
|||
|
(9) |
|
ALA Filing Rules, [prepared by] the American Library Association Resources and Technical Services Division Filing Committee. Chicago: American Library Association, 1980. |
|||
|
(10) |
|
R.N. Collison, Indexes and Indexing. London: Ernest Benn, 1972. |
|||
|
(11) |
|
In Indexing for Editors; London: British Records Association, 1997 reprint. (Archives and the User no 2.) |
|||
|
(12) |
|
BLAISE filing rules, [prepared by the] British Library Filing Rules Committee. London: the British Library Board, 1980. |
|||
|
(13) |
|
See BS3700: 1988, §6.2.1.5, and the Chicago Manual of Style, 14th edition, cited in Nancy Mulvany, Indexing Books; Chicago: University of Chicago Press, 1994. |
|||
|
(14) |
|
See Hans H Wellisch, Indexing from A to Z, 2nd edition revised and enlarged; New York; Dublin: H.W. Wilson, 1995 at p294-7 and p450. Indexing personal names is a complex subject which Wellisch deals with at p345-77. |
|||
|
| |||||
|
Page updated 10 April 2005 |
|||||
|
|
|||||