Practical guidance 

Our database holds separate records for each of the different language versions of a self-translated "unit work", as long as these were printed as separate bibliographic objects (bilingual or multilingual editions are described in a single entry). The individual language versions of a single "unit work" are connected to each other through the "Linked to work(s) field". For instance, Jean Bodin's République has its own record, and so does its Latin counterpart, the De republica; but each records refers to the other in the "Linked to work(s) field" (see image below). 

A) Information included in each record 

Typically, each record in our catalogue displays the following information: [*]

• Author's name. For anonymous works, we give the name(s) of the presumed author(s) in square brackets. For pseudonymous works, we give the name(s) of the presumed author(s) in square brackets followed by the pseudonym. Any controversial cases are discussed in the Notes field.

• Author's dates of birth and death (when available - floruit [fl.] used otherwise).

• Title of the work. We use abbreviated titles for the sake of concision, but include additional details if and when relevant for our purposes. Whenever possible, we also specify if the language version in question is the original version (= source text, abbreviated ST) or the result of the translation (= target text, abbreviated TT).  

  • • The ST/TT markers directly follow the title of the work.

    • If directionality is uncertain, we write ST? and TT? 

• Linked to work(s). This field indicates the other language versions with which the work in question is associated. Users can refer to the relevant records for bibliographical information about them. 

• Directionality. This field indicates the languages involved in the self-translation and, whenever possible, the direction of transfer. We use the following conventions: 

  • • X>Y indicates that the work was first written in language X and then translated into language Y 

    • X&Y indicates multilanguage editions, where the different language versions are printed together in a single volume with no clear sense of directionality

    • X-Y indicates cases of simultaneous or near-simultaneous publication, as well as special cases of complex directionality 

    • X>Y? indicates cases of unclear, presumed, or debated directionality (usually discussed further in the notes) 

We also use the following abbreviations: 

LT = Latin; IT = (varieties of) Italian; FR = French; SP = Castilian Spanish; GR = ancient Greek; DU = Dutch; GE = German; SC = Scots; ENG = English; HEB = Hebrew; CH = Chinese; ARB = Arabic; ARM = Armenian 

• Type. Here we indicate if the work is (or is presented as) a self-translation; a self-translation revised by an allographic translator; or a semi-self-translation (collaborative; allographic supervised; allographic authorized). See "What's included" for more on these types. 

The following additional tags are used when relevant:

• • • Multilanguage edition (see "Special cases" below)

    • Simultaneous or near-simultaneous publication  (see "Special cases" below)

    • Partial (for cases where TT covered only part of ST)

    • Posthumous 

    • Plagiarized 

    • Pseudo-self-translation (see "Special cases" below)

• Explicit? Indicates if the work contains in itself (text or paratext) any explicit statement of authorial involvement (Yes / No). This has a bearing on whether a contemporary reading public would have been aware of its self-translated or semi-self-translated nature, and is still important today for determining attribution (with caveats - see "Special cases" below). 

• Grounds for attribution. We take a broadchurch approach to establishing what should count as a self-translation (see "What's included"). A work can be included on the basis of any one or more of the following: 

  • • Authorial statement in text or paratext;

    • Non-authorial statement in text or paratext; 

    • Epitext (e.g. epistolary correspondence) [**]

    • Textual analysis (this can include philological / palaeographical analysis of manuscript materials; close comparison of the printed versions revealing variants or additions that are likely authorial, such as new dedication letters; etc.) 

    • Tradition (this can include testimony from contemporaries or near-contemporaires; later bibliographical evidence; or modern historiography that has discussed the work as a self-translation

• Other authors. Normally reserved for allographic translators (named or anonymous), but other figures (e.g. editor, publisher, author of preface, artist engraver) can be indicated when they are relevant for the self-translated nature of the work. 

• Medium. Print or manuscript. 

• Date of composition (if known). 

• Date of publication. For manuscripts, we adopt the concept of scribal publication and give the relevant date if known. For printed works, we normally give the date of the editio princeps (for special cases, see below). For works published sine notis, we give the presumed date or date range (followed by ? or ca.) or write N/A.

• Place of publication. Normally as referred to the editio princeps (for special cases, see below). For works published sine notis, we write s.l.; if possible, we give the presumed place of publication in square brackets (if necessary followed by ?). False imprints are clearly flagged and we offer the presumed real place of publication in square brackets if possible. 

• Printer/publisher(s). Normally as referred to the editio princeps (for special cases, see below). For works published sine notis, we write s.n.; if possible, we give the presumed printer/publisher(s) in square brackets (if necessary followed by ?). False imprints are clearly flagged and we offer the presumed real printer/publisher(s) in square brackets if possible. 

• Format. 

• Reprints. Up to 1700, of the specific language version that is being described.  We pay special attention to revised or augmented versions that may contain authorial variants or additional material. As for any reprints that were not overseen by the author, but that can serve as indication of a work's success, we do not list these systematically but refer to existing bibliographical repertories for further detail.

• Subject, or subjects - many works in our corpus fall into more than one category. Subject classification is inherently arbitrary, but we try to be respectful of early modern taxonomies of knowledge while maintaining this field as intuitive as possible for modern users.  

• Paratexts. A brief description of paratextual materials included in the work, such as dedicatory letters, prefaces, addresses to the reader, indices and errata lists. Often a vital source of information about the (real or alleged) self-translated nature of the work. 

• Dedicatee. 

• E-book. Link to digitized copy, if available.

• Location. Physical location of the digitized copy (if available) or of the physical copy consulted and described.  

• USTC / ISTC number. Identifier of the specific edition described in the Universal Short Title Catalogue and/or the Incunabula Short Title Catalogue.

• Notes. A brief description of the work, particularly discussing any issues with attribution, directionality, transmission, contemporary reception, etc. 

• Status. In progress (including date of last update) or complete. 

B) The "French" corpus and the "Italian" corpus

Our database is divided into two parts, one for works produced in "France" (understood as Europe's wider Francophone area) and one for works produced in "Italy" (understood as the Italian peninsula + islands, and other predominantly Italophone areas). This subdivision is temporary, reflecting the way in which we split up the work between team members in the data collection phase. Our final printed repertory will bring the two corpora together in a single list, arranged alphabetically by author's name.

To assign works to one or the other corpus we follow this primary criterion: if the work was printed (or, for manuscripts, produced) in "France", we include it in the French corpus; if it was printed/produced in "Italy", we assign it to the Italian corpus. If one version was printed in France and the other in Italy, the rule of thumb is that the target text wins over the source text. Thus, a work for which the source text was printed in Rome and the target text in Paris will normally be included in the French corpus (barring other considerations). 

Under NO circumstances do we split a unit work into two, with the source text being assigned to one corpus and the target text to the other. But different unit works by the same author may be assigned to different corpora depending on where the target text was published: for instance, the Historie of Scotland // De origine, moribus, ac rebus gestis Scotiae by the Scottish bishop John Leslie will be included in our Italian corpus because the Latin TT was printed in Rome (1578), where Leslie was exiled at the time; hence, it satisfies the geographical (but not the linguistic) criterion. Similarly, Leslie’s Regiment of Women // De illustrium foeminarum authoritate belongs in our French corpus, for while it was written in English and Latin, the TT was printed in Reims.

For areas such as Piedmont and other parts of northwestern Italy, as well as certain areas of Romance Switzerland, where both Italian and French were widely used, we decide on a case-by-case basis whether to assign a work to one or the other corpus; in such cases, and whenever in doubt, we advise users to search both corpora. 

C) Special cases 

Multilanguage editions

We have a "multilanguage edition" whenever the various language versions of a self-translated unit work are presented together in a single volume, instead of forming discrete bibliographical entities. Not all multilanguage editions are or should be considered self-translations: we only include them in our corpus when attribution can be safely determined on the basis of explicit statements or other relevant evidence (see above, "Grounds for attribution"). Because with multilanguage editions we only have a single bibliographical entity to describe (corresponding to a single record in our catalogue), the "Linked to work(s)" is systematically left empty. Directionality of publication similarly does not apply in such cases. As for directionality of composition, it is usually difficult to determine - with some authors going to great lengths to obscure which version came first. For this reason, we limit ourselves to indicating which languages are involved using an ampersand (e.g. LT&FR).

Simultaneous or near-simultaneous publication

This case occurs when the 2+ language versions appear as distinct bibliographical entities, yet are published at the same time or in very rapid succession. Similar to multilanguage editions, directionality of publication is either non-applicable or difficult to ascertain in such cases, while directionality of composition may either be unknown or deliberately obscured by the authors themselves. Cases in which the two versions were supposed to be published simultaneously but contingent reasons (e.g. printers' delays) made this impossible also fall into this category. In all these cases, if directionality is unknown or unclear, we limit ourselves to indicating which languages are involved using a hyphen (e.g. LT-FR).

Multiple self-translations (into more than one language)

Some works in our corpus were translated not into one but into several languages. For instance, the Scottish bishop John Leslie (or Lesley, 1527-1596) translated his Treatise touching the right, title and interest of the mightie and noble princesse Marie into Latin, French, and Castilian Spanish (see French corpus). These various language versions are described individually in our database, but they are collectively understood as forming a single "unit work". Because multiple self-translation generates several target texts, we list these as TT1, TT2 etc. in chronological order (when known). 

Missing works (lost or projected) and pseudo-biblia

In the case of missing works, where one language version materially exists but the other does not, we have a single record in our catalogue, corresponding to the language version that has been preserved / produced. The missing version is acknowledged in the "Linked to work(s)" field: the title, if known, is preceded by the crux symbol  [†] to signal that the work is missing. We then specify, to the best of our knowledge, if the work was lost or projected; in what language it was (or was supposed to be) written; and whether it was the source text or target text. 

We adopt the same system for pseudo-self-translations when one of the two language versions appears to be purely fictitious. Uncertain cases are flagged in the "Type" field using a question mark (pseudo-self-translation?) and are further discussed in the Notes.

D) Technical tips for using our database 

Our database is powered by Airtable, a cloud-based platform for creating and sharing simple databases. The Airtable interface is quite intuitive and allows you to use our dataset in many different ways: you can browse the entire dataset at a glance, zoom in on individual items, run a keyword search, sort the data as you wish (e.g. alphabetically by author's name or chronologically by date of publication), or filter the data to focus on certain kinds of material (e.g. works on medicine, or "explicit" self-translations). You can even download the full dataset for offline use. 

If you need more guidance on how to operate our database, check out our FAQ section where we offer some simple tips. Please note that the best way to use our database is from a laptop or tablet: if you are using a smartphone, you will still be able to navigate our database but functionality will be limited. 

If you have a query that is not currently covered in our FAQ, please get in touch and we'll do our best to help you!