In 1427, the leaders of Florence ordered a complete recording of households within its domains: the number of household members, the wealth the household owned, and of course, the names of household members. This was not the first such recording, but it was one of the largest done up to the time, recording over 65,000 households. This recording was known as the Catasto, a term which refers to records of real property. In the twentieth century, David Herlihy, an historian of Renaissance Italy, began a project to catalogue the data from the Catasto. In the end, he made available to other researchers the data from the Catasto, including the names of all the household heads. This analyzes nearly 29,000 names from the "Condado" of Florence, the countryside surrounding the city proper (while I'm reanalyzing the city data, analysis of the raw data from the city itself are available online).
The names were relatively easy to enter into a database, as they were very standardized in the declarations of property in the Catasto. Most people are identified by a given name, followed by di (of) and their father's name (in the case of widowed women, often by their husband's name), and by a family name. Examples of names from declarations (not the database, but actual declarations) include:
| Berto di Guido Triniciavelli | Antonio Filippi |
| Benedetto di Piero degli Strozzi | Mariotto di Franciescho di ser Sengnia |
| Palla di Nofri | Marcho del Conte dala Lastra |
| mona Bartolomea | Antonio |
| Michele di Iachopo fabro | Magio |
| Biagio d'Antonio | Francesco |
A longer list of complete names from declarations.
Note that some have all three types of names (a given name, a patronym, and a family name), while others have only a given name and family name, a given name and father's name, or only a given name (these are most frequently family members of the declarer, making any further identification unnecessary).
Herlihy's project extracted those standard elements – given name, patronym, and family name – and entered them as part of an 80 column database, with 10 characters for each element. Name elements were truncated at ten characters. The connecting di was dropped from the patronyms, but elements like de, da, and della were retained in family names.
This leads directly to a discussion of some important limitations of these data. First, as the name elements were truncated at 10 characters, some must be reconstructed. This is not an enormous problem for given names, as most are shorter than that, but some must be reconstructed. For family names, it is a bigger issue, as some are quite long, requiring guesswork. Morever, to make constructing names more difficult, spaces and apostrophes in names (as in double given names or locative bynames) were removed to make them fit more neatly in ten characters. I have noted all extensions to names in brackets (note that in a few cases, eleven characters were accidentally entered for a name element). Spaces and apostrophes are not indicated with brackets, as all are reconstructed here based on the handful of fully transcribed Catasto declarations, other period documents, and modern practice.
The practice of truncating names and eliminating spaces creates special problems for reconstructed family names and the few examples of possible double given names. Locative family names are especially problematic, as the inclusion of da, della, and the like leave fewer letters for the substantive element, making them sometimes hard to identify. It has often been difficult to distinguish della from dell'A, and del from de L or from names beginning with Del-.
In addition, names were lightly normalized. While diminutives and other major spelling variations are represented, small spelling differences were eliminated. These spelling differences are systematic and follow certain rules:
A large number of given names are represented in this data; the 24,201 men in the Condado share 1362 names (counting each spelling variant separately) or 869 names (grouping spelling variants and diminutives together). Despite this, men's given names in this data are more concentrated than they were just a generation previous; the 26,367 men given as fathers and husbands in the document share 2105 names. Earlier Italian sources have even more variability; in 13th century Pisa (not part of the Condado, but part of the domains of Florence in 1427), the most common name accounts for only 4% of individuals (fewer than the top six names in this dataset). The five most common names for men (counting variants and hypocoristics with their base form) account for 37.8% of individuals, while the top 10 names account for 54.5% of individuals; to compare the top ten names in 13th century Pisa account for only 28.7% of individuals. To compare, in late 15th century Spain, the top 5 men's names accounted for 53% of individuals.
The top ten men's names (including in the count spelling variants and hypocoristic forms – see below for a discussion of hypocoristic forms) are:
| 1 | Giovanni | 2663 | 11.0% |
| 2 | Antonio | 2240 | 9.3% |
| 3 | Bartolomeo | 1467 | 6.1% |
| 4 | Piero | 1438 | 5.9% |
| 5 | Domenico | 1352 | 5.6% |
| 6 | Francesco | 1193 | 4.9% |
| 7 | Iacopo | 903 | 3.7% |
| 8 | Matteo | 692 | 2.9% |
| 9 | Michele | 656 | 2.7% |
| 10 | Andrea | 576 | 2.4% |
All men's given names by frequency.
All men's given names alphabetically.
Women's names are somewhat more diverse; the 2664 women share 389 names, which means that there are three times as many names per person as for the men. This is an unusual pattern, though one that is found more broadly in Italy. Not surprisingly, then, the most common names also account for fewer individuals. The five most common names for women (counting variants and hypocoristics with their base form) account for 31.8% of individuals, while the ten most common names account for 49.2% of individuals. To compare, in 13th century Perugia, the twelve most common women's names account for only 34% of the sample, while in 15th century Spain, the top 10 women's names account for 75% of individuals.
The top ten women's names (including in the count spelling variants and hypocoristic forms – see below for a discussion of hypocoristic forms) are:
| 1 | Giovanna | 220 | 83% |
| 2 | Caterina | 175 | 6.6% |
| 3 | Margherita | 161 | 6.0% |
| 4 | Antonia | 147 | 5.5% |
| 5 | Bartolomea | 143 | 5.4% |
| 6 | Francesca | 141 | 5.3% |
| 7 | Piera | 124 | 4.7% |
| 8 | (Elisabetta) | 68 | 2.6% |
| 9 | Iacopa | 67 | 2.5% |
| 10 | Niccolosa | 64 | 2.4% |
All women's given names by frequency.
All women's given names alphabetically.
One very striking feature of this data is the large number of hypocoristic (pet) forms. For many common names, hypocoristic forms are almost as frequent as the base name: there are nearly as many men with some hypocoristic form of Giovanni as men named Giovanni or Gianni, while there are actually more men named Meo than Bartolomeo, from which it is derived. The same thing can be found for women: the hypocoristic Cecca is more common than Francesca, from which it is derived.
Hypocoristics are formed in several ways. The most common form is a shortened form of the name, usually derived from the end of the name rather than the beginning of it: Vanni from Giovanni, Chele from Michele, Meo from Bartolomeo, and Lippo from Filippo. Some take this form and alter the initial consonant: Nanni from Vanni, Pippo from Lippo. The second major pattern is the addition of diminutive endings to a root name, generally –ino/-ina, -ello/-ella, -ozzo/-ozza, and –uccio/-uccia. Sometimes these dimutive endings are added to a shorted form: Vannuccio (from Giovanni), Menichello (from Domenico), Colino (from Niccolo), and Puccino (a double diminutive of a shortened form of Iacopo). In some cases, this creates diminutive forms whose root name cannot be clearly identified, such as Dino (a shortened form of the diminutive formed with the –ino ending from some name ending in –do), Tuccio (again, from some name ending in –to).
There are no clear examples of double given names in this data, unlike the situation in 1458 and 1480 (when additional Catastos were done). In other parts of the Catasto, a handful of either compound names or double given names are found for men, including the definitely compound Giovanniba[tista] (John the Baptist) and Micheleagn[olo] (Michaelangelo). Possible double given names include Gianbernar[do], Giovannifr[ancesco], Giovannipi[ero], Giovansimo[ne], Pieropagol[o], each of which occurs once in the data (making the incidence of double or compound given names about 0.01%). All the examples occur with the five most common men's names (except the special case of Michelagnolo). There are no examples of women with double given names, even in the data from 1458 and 1480.
There are two types of bynames found in this data: literal patronymics and family names. While literal descriptive bynames of various sorts can be found in other data sources from 15th century Italy, they are not found in this database. In the declarations, literal occupational bynames are often appended to the name, but the editors of the database treated them as not part of the names. While literal bynames were assuredly used for many purposes, they seem to have been at best extraordinarily rare in these declarations (perhaps because of the requirement of a literal patronymic byname).
Literal patronymics are formed by adding di 'of' in front of the father's name. Around 98% of individuals in this part of the Catasto database have a literal patronymic. Women are sometimes identified in the same way as the wife of their husbands; men are occasionally described with two generations worth of patronymics (identifying their father and grandfather). Occasionally in the declarations, the father (or other man) is identified with a title: the example Mariotto di Franciescho di ser Sengnia illustrates both ideas. The list of names used in patronyms here is similar but not identical to the list of given names; it can be found below. There are a handful of examples of literal matronyics: Lorenza, Giovanna, and Margherita are all found in the patronym database section. However, they are very rare (certainly less than 0.1% of all patronyms). When the father's (or husband's) name begins with vowel, di becomes d'.
Names used in Patronyms by Frequency.
Names used in Patronyms Alphabetically.
Matronyms.
Family names are quite varied and much rarer than patronymic bynames: fewer than 10% of individuals in this part of the Catasto database have family names (the number in Florence proper is somewhat higher). Family names seem to be limited to prominent families, and are rarely repeated. Even the most common family name, Martini, is held by only 44 individuals (out of nearly 27,000); only two family names account for more than 1% of the people with family names (in other words, excluding from the count the 90% of the heads of household with no family name). This suggests that family names are playing a special role here: identifying membership in specific prominent families, rather than describing who people are in more general terms.
Family names fall into several types: patronymics (which are formed differently than literal patronymics), locatives, occupationals, and descriptives. Patronymics are by far the most common; of the 1247 individuals having family names used by three or more individuals, over 90% (1138) were patronymic in nature. Patronymic family names are formed by putting the father's name into the old Latinized genitive case (which had fallen out of general use, and hence is not used to make literal patronymic bynames): Martino becomes Martini, Lippo becomes Lippi. Names that end in –o change to –i; names that end in –io or –e change to –i. Names that end in –i and –a are used in unmodified form, though in the Tratte names ending in –a also form family names by changing the –a to –i. It is not clear whether this is due to editorial standardization or a difference in the sources (both of which are from Florence, but across different time frames and reflecting different social classes)
Note that these names are formed from hypocoristic forms of name as well as the root forms. No preposition (di or otherwise) is used with these. These are by far the most common form of family name. Some patronymic family names are formed in this data by putting del in front of the unmodified form of the name: del Nero, del Drea, del Duccio. But these are much rarer than the first type.
Occupational family names are mostly also in the genitive case, as they are not so much a claim to an actual trade, but to descent from someone in an occupation. In this data, we see examples such as Cancellier[i], Calzerone (a rare non-genitive example), and Forestani. Again, these are sometimes formed with del in unmodified form: del Fabro and del Medico.
Locative family names are formed with a form of the preposition "of." The form broadly considered the most "typical" form, da, rarely appears here. Instead, family names with town names seem to be formed with del: del Cortona, del Milano, del Romagne. More common are names derived from generic toponymics, which are found with della (feminine) and dello (masculine), or more rarely dalla and dalle: della Caste[la], della Querc[ia], della Torre, della Valle. Note that when the toponym starts with a vowel, the final vowel of dalle/dalla is lost, and it is written as in dell'Abadia.
Descriptive family names are formed in two ways: with del or in the genitive form: del Grosso, del Rosso, Rossi.
Family Names by Frequency (with etymological notes)
Family Names Alphabetical.