Personal names in Italian archives rarely function as the stable identifiers that modern researchers expect. A scholar may arrive with a name securely established in the secondary literature, search for it in an online catalogue or manuscript index, and retrieve nothing. The immediate conclusion may be that the individual is absent from the archive. Yet, the problem may lie not in the survival of the evidence, but in the form under which the person entered the documents, the finding aids or the modern information system. A person known today as Giovanni may appear in Latin as Iohannes, but also as Iohannis, Iohanni or Iohannem, depending on the grammatical structure of the sentence. He may be identified through his father, his place of origin, his office, his parish, his employer or the property with which he was associated. An archival description may subsequently place him under an authorised modern form that never appears in the original document. None of these layers is incidental. Each reflects a different process through which a historical person was made identifiable. The practical consequence is fundamental: searching for a person in Italian records is not simply a matter of generating spelling variants. It requires reconstructing the documentary forms through which a particular institution, writer or archivist could recognise that person.

Names often reach scholars through modern publications, museum catalogues, collection records or genealogical databases. These sources tend to provide a regularised form: a consistent spelling, a stable surname and, where possible, fixed dates. This is necessary for scholarly communication, but it can conceal the variability of the underlying documentation. The researcher who transfers that regularised name directly into an archival catalogue is therefore making an assumption: that the modern scholarly form corresponds to the form selected by the archive. Sometimes it does, but in other cases, the person may be described under a title, a family name, a patronymic or a form derived from the language of the documents. A hypothetical Giovanni di Marco might appear in a Latin act as Iohannes filius Marci. If the father had died, the notary might instead use Iohannes quondam Marci. A vernacular register might record Giovanni del fu Marco. These expressions are related, but they are not interchangeable strings. They contain information about filiation, the father’s status and the documentary language, and they may be indexed differently. The problem becomes more pronounced when a person is known to modern scholarship by a designation consolidated only after his or her lifetime. An artist may now be identified through a place-name, a dynastic surname or a conventional epithet that was used inconsistently, or not at all, in contemporary records. The relevant document may therefore be filed or indexed under the individual’s father, household, professional association or office rather than under the name familiar from the bibliography.

The relationship between a Latin and a vernacular name is often described too casually as translation. Iohannes is treated as the Latin version of Giovanni, Petrus as Pietro and Franciscus as Francesco. These equivalences are useful for discovery, but they explain only part of what happens inside a document. Latin names participate in the grammar of the sentence. Iohannes, Iohannis, Iohanni and Iohannem may be different case forms of the same name, not four independent variants. The form visible in a document may consequently depend on whether the person acts, possesses, receives, witnesses or is the object of an action. A full-text search that recognises only the nominative form can miss relevant attestations. Conversely, a researcher unfamiliar with the inflection may mistakenly interpret grammatical variation as evidence of several names or people. Early medieval Tuscan charters offer a particularly well-studied example of documentary Latin that differs substantially from classical and standardised medieval norms in spelling, morphology, syntax and lexicon. The Late Latin Charter Treebanks include more than five hundred original Tuscan documents from the eighth and ninth centuries, many written by professional notaries to record private transactions. Their language illustrates why documentary forms cannot be reduced to a simple opposition between “correct Latin” and “Italian”: they preserve linguistic developments, formulaic habits, local usage and large numbers of non-standard proper names. The point is not that every Italian archival project requires expertise in early medieval Latin. It is that documentary names are generated within linguistic systems. A researcher must first distinguish a genuine onomastic variant from a grammatical form, an abbreviation or an editorial normalisation. Otherwise, the search may expand indiscriminately while the evidential meaning of each form is lost.

A notarial act does not attempt to reproduce how a person introduced himself in ordinary conversation. Its purpose is to establish an identity sufficient for a legal, administrative or evidential transaction. The resulting name is therefore shaped by genre and institutional need. A person may be distinguished through filiation, origin, residence, profession, office, title, marital status, household or relationship to property. The descriptive element that appears most prominently may not be the one later chosen as the surname. Da Firenze may refer to provenance in one context, function as a relatively stable designation in another, or eventually become part of a family name. A patronymic may identify the person effectively within one generation without behaving like a hereditary surname. This is especially important when working with notarial registers, inventories, fiscal records, court papers and ecclesiastical documentation. The person is often present not as an autonomous biographical subject but as a party, heir, debtor, witness, parishioner, office-holder or relative. The source identifies the role required by the transaction. The distinction helps explain why a scholar may fail to find a documented individual even after testing several linguistic variants. The relevant records may not be organised around that individual at all. They may instead be structured around the notary, the contracting party, the estate, the institution, the deceased person whose succession is being administered, or the family archive in which the papers accumulated. Research on medieval legal charters likewise cautions against treating documentary passages as straightforward transcripts of spoken language. Formulaic structures, professional habits and the competence of the writer mediate what reaches the page, even where direct speech or witness testimony appears to be represented.

It is tempting to replace one overly simple assumption with another: once the researcher understands “the Italian system”, the names will become predictable. The historical evidence does not support that conclusion. Documentary practice varied by city, jurisdiction, period, institution and professional environment. A comparative study of notarial records from Paris, Toulouse, Mende, Turin, Florence and Livorno found substantial local differences in the distribution and use of notarial acts. In some instances, variation between cities within the same state was greater than variation across state boundaries. The implications extend beyond the typology of deeds. Local conventions affect how parties are introduced, how titles are handled, which relationships are foregrounded and how names are abbreviated or ordered. The same individual moving between courts, cities or institutions may therefore be described through different identifying features. The researcher should consequently resist treating one encountered form as a universal key. A naming formula learned in a Florentine series cannot automatically be transferred to Venetian, Roman, Neapolitan or Piedmontese documentation. Even within a single archive, different offices and documentary series may follow different conventions. This local and institutional variability is one reason why successful archival research depends on understanding the provenance and function of records, not merely the vocabulary they contain. A name becomes interpretable only when read within the type of document, administrative context and archival series that produced it. For this reason, Archivitaly assigns projects to researchers with direct familiarity with the relevant archival environment. Their expertise includes not only linguistic and documentary conventions, but also local finding aids, the internal structure of fonds and series, institutional histories and the practical habits through which records are described, requested and consulted.

Modern naming conventions encourage the expectation that each individual possessed a personal name followed by a stable hereditary surname. That expectation becomes progressively less reliable the further the research moves into medieval and early modern evidence. François Menant’s synthesis of naming in communal Italy emphasises that systems of designation were neither uniform nor unchanging. Different social groups used different combinations of personal names, family references and additional designations, reflecting wealth, political position, urban affiliation and collective identity. Many people were still designated by a single personal name in earlier periods, while more elaborate naming systems developed unevenly across regions and social environments. A second element may therefore be a patronymic, a place of origin, an occupation, a nickname or an indication of affiliation rather than a surname in the modern sense. Its recurrence does not automatically prove hereditary transmission. Nor does its disappearance necessarily indicate that the person has changed. The methodological risk operates in both directions. A researcher may divide one individual into several people because the additional designation changes. Alternatively, several homonymous individuals may be fused into one because a familiar name is treated as distinctive when it was in fact widespread. The problem is particularly acute where a limited repertoire of personal names recurs within the same family, profession or locality. A document referring to Petrus quondam Iohannis does not become securely identifiable merely because another record mentions a Pietro, son of Giovanni, within an approximately compatible period. The correspondence may be plausible, but further evidence is needed.

Some apparent variants originate not in language or social identity but in the material practice of writing. Medieval and early modern documents frequently abbreviate names, titles, kinship terms and formulaic expressions. The reader or editor must expand what is only partially represented on the page. Abbreviation systems were structured rather than arbitrary, but the correct expansion still depends on grammar, context, date and writing practice. The Oxford Handbook of Latin Palaeography describes the principal techniques, including contraction, suspension and specialised signs, and underscores the importance of interpreting abbreviations as part of the writing system rather than as missing letters that can always be restored mechanically.

This matters even when the scholar does not personally transcribe the manuscript. A name in a printed calendar, nineteenth-century inventory or modern database may already incorporate somebody else’s expansion. That expanded form should not automatically be treated as though every letter were visible in the original. The evidential status of the form therefore depends on where it was encountered. A diplomatic transcription, an editorially expanded text, a descriptive inventory and an authority record perform different operations. Researchers should preserve that distinction in their notes, particularly when an identification depends on one ambiguous reading.

The name encountered in a catalogue is not necessarily the name found in the source. It may be an authorised access point constructed to bring together several documentary forms.

The Italian NIERA (EPF) standards for archival authority records explicitly distinguish an authorised form of a person’s name from other forms. They accommodate personal names with surnames, patronymics, indications of provenance, assumed names, pseudonyms, lexical variants and parallel linguistic forms. The standards recognise that a person may have multiple names for legal, voluntary, attributed or historical reasons and that variants can consolidate over time. The international ISAAR (CPF) standard follows the same broad logic, providing separate fields for authorised, parallel, standardised and other forms of name.

For the researcher, this has two consequences. First, an authorised catalogue heading is a discovery tool, not a diplomatic transcription. It may deliberately normalise or reorder the name. Second, older inventories and local finding aids may have been compiled under different rules, or without a formal authority system at all. Two catalogues describing related materials may therefore select different access points for the same individual. Archival descriptions should consequently be treated as evidence about the organisation and interpretation of a collection, but not as substitutes for the documents they describe. As Randolph Head’s work on European record-keeping demonstrates more broadly, records and archives were created through historically situated practices of production, accumulation, arrangement and reuse. The modern finding aid belongs to that history of mediation rather than standing outside it.

This extends an argument made in Archivitaly’s earlier article, Archival Research in Italy: From Citation to Document: the catalogue is not the archive, and absence from a searchable interface is not equivalent to absence from the surviving documentation.

Recognising that Iohannes may correspond linguistically to Giovanni is necessary. It is not sufficient to prove that two attestations concern the same person. Identification requires a convergence of attributes. Chronology must be compatible, but chronological compatibility alone has little force where a name is common. Geography may strengthen the case, but only after distinguishing origin, residence, jurisdiction and location of the transaction. Filiation, spouse, profession, office, property, parish and networks of witnesses may provide stronger links. Explicit statements of succession, inheritance or continuity are stronger still. A useful report should preserve the degree of certainty rather than compress every correspondence into a positive identification. A possible match means that the names and circumstances are compatible. A probable identification means that several independent features converge and no equally plausible alternative has emerged. A documented identification requires an explicit connection or a sufficiently strong chain of evidence to exclude reasonable alternatives. These categories are not formal legal standards, nor can they replace source-specific judgement; they are a discipline of expression, they prevent the repeated use of a plausible name match from gradually transforming an hypothesis into an accepted biographical fact. The reverse problem is equally important. Slightly different forms should not be separated automatically. A change of grammatical case, language, abbreviation or descriptive context may leave the underlying identity untouched. The task is therefore not to prefer either unification or separation, but to test both possibilities against the documentary evidence.

A productive search strategy begins by separating retrieval forms from identity evidence. Retrieval forms include plausible Latin and vernacular versions, grammatical cases, graphic variants, abbreviations, patronymics, titles, places and professional descriptions. Their function is to enlarge the range of records that can be discovered. They should be generous enough to accommodate documentary variation. Identity evidence serves a different purpose. It tests whether the records discovered through those forms concern the same person. Here the method should be restrictive. Each proposed correspondence must be checked against chronology, relationships, location, office, property and the institutional context of the source. The distinction is essential because the same information can behave differently at the two stages. A shared patronymic may justify opening a file or examining a register, but it may remain too weak to support a conclusion. A title may be an effective search term while being chronologically transferable between several office-holders. A place-name may retrieve the right documentary cluster without proving whether it denotes origin, residence or family identity. For practical purposes, the researcher should also record where each form came from. A spelling read directly in an original document should not be silently merged with one supplied by a catalogue, a modern editor or a secondary publication. Preserving those layers makes later verification possible and often reveals how an apparent contradiction arose.

Personal names are among the most powerful routes into Italian archival collections, but they are not self-authenticating identifiers. They are linguistic forms, grammatical forms, social descriptions, institutional classifications and archival access points. A scholar may know exactly whom they are looking for and still search under the wrong documentary identity. Conversely, a familiar linguistic correspondence may create an identification that the surrounding evidence cannot sustain. The solution is not an exhaustive list of spelling variants. It is a layered method: distinguish the form used in the document from the form selected by the catalogue, distinguish grammatical inflection from onomastic change, and distinguish a useful retrieval term from proof that two records concern the same person. In Italian archival research, a name opens the enquiry, it does not complete the identification.