Just to pick up on a couple of points from this discussion. The introduction of title as a name component was done, somewhat gratuitously, to provide a parallel with the World Directory format. It has not yet been consigned to hard print anywhere, and we (COMCIFS) could back out this suggestion if it's likely to prove contentious. Comments? Arguably the best database-inspired solution would be to have a large collection of datanames to cover every eventuality (..._name_family, ..._name_dynastic_part, ..._name_professional, ..._name_maiden, etc); but this would require a large amount of annotation, add a very large extra number of datanames (since "names" crop up in several contexts), and run somewhat counter to the existing practice in small-molecule CIF. It was also clear from the World Directory that not everyone is able to separate their personal name components into the family/given names scheme that is familiar to us; and, as Lynn has pointed out, other cultures do have quite different conventions for assigning personal names. In reality, as I think we have all discovered by experience, it is a very knotty problem because individuals can be inconsistent even in their self-identification, and journals may amend the presentation of names slightly to modify their house style. Further, the primary author on a paper may be the one who consigns to print the names of coauthors, even when never having met them. So it may well be that entries for Smith, E., Smith, E.B., Jones, E. B., Smith-Jones, E., Smith Jones, E.-B., Jones(Smith), E. Jones (Smith), B., Jones nee Smith, E. B., Jones (formerly Smith), E. may all refer to the same Lizzie Jones-Smith that we all know and love (and who is rather frustrated over her citations record, we presume!). There are further problems caused by transliteration of non-Latin alphabets (especially if the rules change as with place names, where L'vov becomes Lviv and Peking Beijing), and indeed with the representation of diacritic and ligatured characters. Incidentally, the reason we put "dynastic component" in with "family name" is that 'von', 'ben', 'ibn', 'de' and so forth might be considered either, and we really prefer to see 'de Gaulle, Charles', rather than 'Gaulle, Charles de'. But then, the pedant would insist on 'Gaulle de, Charles'. Clearly, one could go on ad infinitum exploring these many different cases. The CIF convention (minus the "title" part, if you will) has served us tolerably well at Acta, since it is at least parseable. Is there support for going ahead with the convention 'family_name_with_dynastic_component, given_names_or_initials', and drawing up a list of illuminating examples that cover most cases, this last to be done by those who have a profound interest in the matter? Brian _______________________________________________________________________________ Brian McMahon tel: +44 1244 342878 Research and Development Officer fax: +44 1244 314888 International Union of Crystallography e-mail: bm@iucr.ac.uk 5 Abbey Square, Chester CH1 2HU, England