Exclusions of categories of names
Mycenaean names were excluded mainly because of the difficulty of interpreting and presenting the evidence. Though the links between Mycenaean and post-Mycenaean names form an important part of Mycenaean studies, the project did not itself have the competence to undertake this important area of study, or to present the results within the framework of the published LGPN. We do, however, welcome the important work in this area by specialists, which we would like to reflect in these pages in due course.
Geographical names, though greatly in need of fresh lexical treatment, were considered, for many reasons, to lie beyond the scope of the present project.
Byzantine names: to have moved into the developed Byzantine system of 'baptismal name', 'family name' and 'additional name' would have been to move into a new world with new forms of nomenclature and different concepts of name-giving. It was always intended that LGPN would draw a chronological line at around 600 AD, that is, at the time of the Islamic conquests, when the old order disappeared. In practice, significant changes in the historical framework occurred at different times in different places. So, while much evidence from the 4th and 5th centuries AD is found in LGPN (for example, the names from the Sicilian catacombs, included in LGPN III.A), we have had to take each region on its own terms in considering the terminal date. This is a particular issue with LGPN IV, which includes Byzantion/Constantinople itself.
Totals: Each person in LGPN has a unique identifying id-number in the database. It is, of course, very easy to count up these id-numbers and give totals. Nonetheless, even with simple totals of persons, caution is required. The decision whether to identify homonymous individuals is a constant preoccupation with those who compile onomastica or prosopographies, or indeed indexes of epigraphical corpora. The totals depend ultimately on editorial judgments about homonymity.
When a person is classified 'ambiguous', it means that the name belongs to the group which may be borne by men or women, and that the context provides no clues.
The 'Total' is the number of names which receive a separate entry in LGPN, which are classified as 'primary' in the database.
Dialect forms stand as primary names, independant of the koine form. As a result, in the statistics given, what is essentially the same name may appear twice or more e.g. Asklepiades/Asklepiadas/Asklapiadas. A cross-referencing system between these different forms, an acknowedged need, and was introduced in LGPN III.B.
Orthographic variant forms, in contrast, do not get a separate entry, but are 'normalized', and recorded only at the end of the entry. In the database, they are classified as 'secondary'. They may be searched out and analyzed, but for the purpose of this web page they are too numerous to be handled easily. So, the statistics currently available offer only primary forms.
There is a further complication that a form which in one place or time is a dialect form may elsewhere and at another time be a variant. Sorry, but ancient Greek was like that.
In the database, every person has a gender. Since assigning gender was not part of the original research, the gender has been assigned by a computer program on the basis of name-termination. For most names, this method is reliable, but there are exceptions. In Western Greece, for example, there are masculine names ending in alpha. More significantly, names ending in -ις are usually genuinely ambiguous. Nikopolis, for example, was a name borne by both men and women. If neither the text nor the context supply the gender, the individual is classified as 'ambiguous'.
In the name statistics, names of this type may appear under 'male', 'female' and 'ambiguous'.