About this version of the Tratte data.

The data file was initially conceived and largely coded by David Herlihy and his assistants at Harvard and Brown Universities in the twenty years before his death in 1991. The project was then resumed in 1997 by R. Burr Litchfield and Anthony Molho, colleagues of Herlihy at Brown. R. Burr Litchfield assumed practical direction of the later project and edited the master files; Anthony Molho provided advice and logistical support. After assessment of Herlihy’s computer files, and an initial small grant from Brown in 1997, financial assistance was obtained from the National Endowment for the Humanities, Division of Preservation and Access, in 1998-99 for the years 1999-2001 (PA23264-99). Claude Goldman of the Brown Center for Information Sciences made the computer files readable in FOCUS, a standard data management program, and managed the master files throughout. The project could not have proceeded without the valuable paleographic and archival assistance of Roberto Barducci who tirelessly checked printouts of the data against the original registers of the Tratte in the Florentine Archivio di Stato, made corrections that were entered into the master files at Brown, and coded missing data. Giovanni Ciappelli provided additional technical assistance in Florence. The Web page was designed by R. Burr Litchfield and programmed for the MySQL server by Carole Mah of the Brown Scholarly Technology Group under the direction of Elli Mylonas. The website is currently maintained by the Scholarly Technology Group.

David Herlihy’s conception of the office holding data file.

David Herlihy conceived the idea of computer coding the corpus of records surviving from elections of the Florentine Republic in the 1960s, when he was completing the monumental study of the Catasto of 1427 that he had undertaken in collaboration with Christiane Klapisch-Zuber of the Paris Ecole Pratique des Hautes Etudes: Les Toscans et leurs familles: Un étude du catasto Florentin de 1427 (Paris: FNSP Editions de l’école des hautes etudes en sciences sociales, 1978). He planned to use the material from the Tratte in a large work on the social bases of the Florentine political system. This was a grand subject, the political system of Renaissance Florence, which underlay its vibrant artistic and intellectual culture, and before his death in 1991 he enunciated an hypothesis, which he stated in an important article: "The Rulers of Florence, 1282-1532: Oligarchy, Democracy, Principate" (in A. Molho, K. Raaflaub, and J. Emlen, eds., Athens and Rome, Florence and Venice: City-States in Classical Antiquity and Medieval Italy. (Stuttgart: Franz Steiner Verlag, 1991). A common assumption of historians of Renaissance Florence through the ages, which has never been either fully substantiated or fully disproved, Herlihy began, is that the political system of the Republic was "oligarchic". Were the most important offices of the Republic, the "Tre Maggiori", held by a relatively small number of individuals and families, or was political influence diffused on a wider basis? He referred to the many historians who had contributed to this debate, and he stated that he planned to challenge the "oligarchic" assumption. "As I hope to illustrate," he wrote, "the numbers of citizens actively seeking office increased continuously, even spectacularly, from the early fourteenth to the late fifteenth century . . . . The increasing numbers of office holders have also proved puzzling, particularly since the city’s population was simultaneously in drastic decline. How could the number of citizens holding high office constantly grow . . . under regimes commonly regarded as oligarchic?" ("The Rulers of Florence", p. 197.) This is a large and important issue for interpretation of the Florentine Renaissance.

Study of the political culture of Renaissance Florence requires the ability to assess who held office, and the various ways in which the effective political elite changed, or was managed, over time. Since the eighteenth century there have been, of course, printed lists (priorists) of the names of individuals holding the offices of the Gonfaloniere di Giustizia or Priore in the Signoria. Herlihy intended to provide a broader assessment of the highest office-holding group by also searching out the names of men selected for the two subsidiary Collegi (the Buonuomini and Gonfalonieri di Compagnia) that completed the "Tre Maggiori" from the Giornali of the Tratte (the original records of names of men drawn by lot), and to assess for both the Signoria and Collegi not only the men "seated" in office, but also those rejected in the periodical drawings. He planned, of course, for computer analysis that would vastly expand what could be learned from these large encoded lists.

Florentine office holding operated at different levels: 1) the "Tre Maggiori" (Signoria and Collegi--the executive offices), 2) the administrative offices (the Otto di Guardia e Balia, Capitani di Parte Guelfa, etc., plus the Vicari and Podestà sent out to provincial towns in the Florentine Dominion), and 3) (at the lowest level) the Guild Consulates of the 21 major or minor guilds. In his study Herlihy planned to focus chiefly on the first and third levels: The "Tre Maggiori", and the Guild Consulates. The rationale of omitting the administrative offices was basically that men eligible for the "Tre Maggiori" were the same group eligible for selection to the administrative magistracies, and thus of themselves constituted the political elite. He was interested in the extent to which men involved in the elections for the guild Consulates penetrated to the level of the Tre Maggiori. Another reason for not coding the names of men selected for the administrative magistracies, although these were of course very important, was the practical consideration that the large number of names appearing in these offices would have made the data file too large to be manageable. After his death we found on Herlihy’s computers the following data files, which he had largely completed coding:

The series of Birth registrations was taken from declarations of age required of all office holders beginning in 1429. The Emancipations were taken from series of records reporting information about children declared to be ‘of age’ that had been coded and provided to Herlihy by Professor Thomas Kuehn, who had used them in a separate study of his own. The "Veduti" proved to be a late-seventeenth century compilation of names of individuals "seen" for the Tre Maggiori for the first time (taken from the volumes in the Archive of the Tratte currently numbered 130-33) to which we gave a lower priority because they were not an original source.

Our initial survey of the data files demonstrated amply that a checked, completed, and edited edition of Herlihy’s office-holding data, which could be made available on the World Wide Web (in a format similar to the Florentine Online Catasto of 1427 that we published from his Catasto data files in 1995), would be an extremely useful resource for researchers in Florentine history. We hoped initially to make all six series available on the Web. However, more detailed assessment of the data, and the practical considerations of time and money, led us to concentrate on the first three. Thus our Series 05 contains the records for the Tre Maggiori from the Archive of the Tratte (82,174 records); Series 06 the records for Guild Elections from the Archive of the Mercanzia (through 1497: 62,841 records), and Series 10 the records for the Birth Registrations from the Archive of the Tratte (21,386 records).

The Coding and Editing of the Data.

David Herlihy used a somewhat archaic method for coding the office holding data, although it was appropriate to the computer technology of his time. He took coded notes on the registers in the Archivio di Stato in notebooks which he transported to Harvard and later Brown, for data entry (using initially an Ohio Scientific Computer that had only 64K of RAM memory and data entry programs he wrote himself in BASIC). The files we retrieved from his computers were in a non-standard format that made them very difficult to read, and impossible to analyze, until the skill of Claude Goldman of Brown CIS succeeded in transforming them into standard rectangular files that could be read, analyzed, and edited in FOCUS, a data-management package. We soon became aware that there were many errors in the data (they had never been fully checked) and also some large gaps, due to the fact that Herlihy had not finished coding all the registers, and that the registers themselves were not entirely complete. Also the Archive of the Tratte had been reorganized since Herlihy had initially coded the material, so that the register numbers no longer corresponded to the coded ones. Some of the registers initially did not even have page numbers. We succeeded, however, in printing out the data in roughly the same order as it appeared in the original registers, and Roberto Barducci settled down to check the printouts against the original registers in Florence. He was able to make many corrections in the data, to complete the coding of registers Herlihy had not coded, and to find parallel sources in many cases to fill in names that were missing from the Giornali of the Tratte and the Mercanzia registers. The corrections were returned to Providence and edited into the master files. Herlihy had already begun to fill in missing names of Gonfalonieri di Giustizia and Priors from the priorista compiled and published by Rastrelli [Modesto Rastrelli, Priorista Fiorentino Istorico . . . (Florence, 1783)]. We checked all the names taken from Rastrelli against the Priorista di Palazzo (an original official record of names of members of the Signoria in the Archivio di Stato) and corrected some errors made by Rastrelli (although register numbers in the data file still refer to Rastrelli [RNUMs 0222-0224 being the three parts of Rastrelli, and RNUM 0001 the Priorista di Palazzo]).

Important disclaimer: It must be recognized at the onset that this is not an integral "edition" of the Giornali of the Tratte, priorists, and the other archival registers on which the data are based. The names in the registers appear in a Latinate form, whereas Herlihy, following the long tradition of translating the Latinate form of Florentine Renaissance names into Italian, translated all the names into Italian, and we followed his practice. He did the same for occupations and places of origin (although we returned many individual occupational titles to their original form). You can see on the name-slip visible on the first page of the Web site the name "Zanobius Benedicti Carocci de Strozzis", a name that appears in the data file in Italian as "Zanobi [di] Benedetto [di] Caroccio Strozzi". The different Italian spellings of names and surnames used by different coders have been standardized to a large extent. After use of the computer files, you will recognize that the computer analysis (the main advantage of the file) would have been impossible without some judicious standardization of names. There is further information about names in the Note on names.. Other information in the file is presented in the form of numerical codes; and the numerically coded information was also edited. Following Herlihy’s practice, dates in the file were all changed into the modern system of beginning the year on January 1st. Dates of entry into office taken from priorists after 1343 were edited to correspond to the reportage in the Giornali of the Tratte of the date of election, in the last days of the month before the beginning of the term, rather than the date when the term began. Designation of purses (major guilds, borsellino, minor guilds) have been edited to a certain extent when there were lapses in the Giornali or Mercanzia registers and the purse the office holder was drawn from was obvious from the way in which offices were customarily distributed. We have checked and edited the data carefully, and checked it for internal consistency. Some coding, checking, or editing errors still undoubtedly remain, but inconsistencies at this point may also reflect ambiguities in the original records.

As well, it must be recognized that this data file is not a final answer to the problems presented by Florentine names and office holding. There are many ambiguities in the original records that we have not attempted to resolve, although we hope that the computer analysis made possible with the file will permit researchers to make their own guesses and assessments (they may be better than ours would have been). You can download data from the master file and develop it further for your own research. It is, of course, always possible to refer to the original archival registers to resolve uncertainties through the register and page numbers provided for each case.

The degree of completeness of the data. (The attached table shows the register numbers in the Archivio di Stato that were used to compile the three series, and the dates the number of cases taken from each.)

The Tre Maggiori. The data file is not entirely complete. Entries in the Giornali of the Tratte began in December 1345. Names of Gonfalonieri di Giustizia, Notai, and Priors before 1345 were taken from priorists, and thus the names of men "seen" but not "seated" in office between 1282 and 1345 are entirely missing. Although the Buonuomini and Gonfalonieri di Compagnia existed earlier, the earliest continuous series of names of men "seated" in these offices (in the Priorista di Palazzo) extend only from 1329 to 1342. After 1345 the only continuous source for names of men selected for the two Collegi is the Giornali of the Tratte. However, at there are some gaps in the Giornali (for the years 1348-49 for some years in the period 1355-1404) or missing pages from volumes (usually the first or last pages of extant registers) in the fourteenth and first years of the fifteenth century. In periods when material was missing from the Giornali we have supplied the names of Gonfalonieri di Giustizia, Notai, and Priors from priorists, but the names of men in the two Collegi are missing. Then, as well, between August 1497 and October 1512, the period of the Consiglio Maggiore, the Tratte were suspended, and we have had to seek names from other sources, so that names "selected" but not "seated" are again missing. The data file is virtually complete, however, for the fifteenth century and for the years 1512-1532. (The attached table shows for what terms names are missing for the Tre Maggiori.)

Finally, David Herlihy coded names selectively from both the Giornali of the Tratte and the Mercanzia registers in that he omitted a whole category of men "seen" but not "seated" in office. A large number of names were rejected from final seating through a "general Divieto" (RDRAW code ‘02’), that is for a reason not specified in the Giornali registers. This proved generally to have meant that the individual in question was rejected because: 1) he was already seated in another office, 2) he had held office too recently, or 3) a member of his immediate family was currently in office. David Herlihy generally included men with this particular Divieto up through Tratte 595 (1376-81), but only occasionally thereafter. This was probably because these drawings, although numerous, did not add much substantial meaning to the data file (the individuals in question were not minors, in tax arrears, away from Florence, dead, etc.). The management of the purses by accoppiatori happened before the drawings recorded in the Giornali, so political considerations were probably not involved. We estimate that in 1430 for the Tre Maggiori as many as 300 such initial drawings may have been omitted from the file, thus increasing the number of records for that year from the present 887 to about 1200 (perhaps as many as 30,000 additional records for the whole data file). Although we initially hoped to add these missing cases, we eventually concluded that David Herlihy had been right to omit them. However, we did not remove any cases with the RDRAW code 02 from the data file; 836 cases remain in the Tre Maggiori and 391 in the Guild elections, which users can utilize to explore this situation further.

Guild elections. Beginning in December 1393 the Mercanzia recorded the results of elections for magistrates of the Mercanzia itself and for the consuls of 20 of the 21 guilds (elections of the Arte dei Giudici e Notai were not supervised by the Mercanzia and, except for one election in 1393, they are entirely missing from the data file). The registers in the Archive of the Mercanzia (Tratte dei consolati delle arti) are not entirely complete. Out of what must initially have been nineteen volumes of elections between 1393 and 1538, five have not survived. There are consequently gaps in the data. The results of elections are complete from December 1393 to August 1421, from August 1429 to December 1443, from August 1465 to April 1474, and from April 1480 to December 1497. David Herlihy had coded the surviving volumes up through 1497. We did not code material after that date, although we have filled in names for 1433-34 that were missing from Mercanzia 83 from Mercanzia 109 "Bastardello di Tratte". The Mercanzia registers continue, after another gap between 1498 and 1507, through the end of the Republic. For the missing years in the fifteenth century, researchers could probably find sources to fill in names of men seated in the magistracy of the Mercanzia itself, but it is doubtful that the full spectrum of names could be reconstructed for the other guilds.

Birth registrations. The legislation of 1429 required all individuals seeking office to submit a declaration of their birth date to the Conservatori di Legge. Original declarations that may have initially been found in the archive of the Conservatori di Legge no longer exist; the registers in the Archive of the Tratte are largely alphabetical compilations made for internal use at the time of drawings by the officials of the Tratte. This makes these registers somewhat difficult to use, since some are more-or-less complete copies of others, and there is some variation in the amount of information provided. After 1457 the date when the birth was registered and the name of the notary who drew up the declaration were included in some volumes that appear to be the most original. Another problem arises from the fact that individuals sometimes registered their date of birth more than once, and sometimes with slight variations in the information provided. Many individuals (about 30 percent of the total who registered) never held office--they may have died or left the city before they reached the required age. To avoid confusion, we have removed exactly duplicate birth registrations (about 400 records) from the data file, leaving the record that had the earliest date; however duplicate registrations with varying information remain. At present we assess that the data file has birth dates for about 80 per cent of individuals who held office after the early fifteenth century.