Repertory of Conjectures on Horace



Repertory of Conjectures on Horace is the result of a research project ‘A New Horace’ financed by The Bank of Sweden Tercentenary Foundation (Svenska Riksbankens Jubileumsfond). Grants have also been received from the Faculty of Humanities, University of Oslo. The work has been carried out at the University of Oslo. Head of project: Monika Asztalos. The Repertory has been compiled by Monika Asztalos, Victoria Gejrot, Gunn Haaland, Egil Kraggerud, Erik Sjaastad, Per Erik Solberg, Marianne Ophaug Wehus, Tor Ivar Østmoe, and Halfdan Baadsvik.

The idea to create a repertory of conjectures on Horace was Egil Kraggerud’s. Monika Asztalos received grants for a project involving several members. Victoria Gejrot, a graduate student in Latin, designed the database using the program FileMaker Pro. Four Research Assistants have registered conjectures: Erik Sjaastad, Per Erik Solberg, Marianne Ophaug Wehus, and Halfdan Baadsvik. Gunn Haaland, Librarian at the University Library of Oslo, has been most helpful in tracing publications; she has also registered conjectures proposed or discussed primarily in Italian publications. Tor Ivar Østmoe has written a PhD thesis, “Poeta noster. An analysis of discussions of conjectures in the Horatian commentaries of Lambin (1561 and -67), Bentley (1711), and Peerlkamp (1834 and -62)”, 2013, in addition to registering conjectures in the database. Each member of the project has checked a certain number of publications in the Bibliography as indicated by initials: MA (Monika Asztalos), VG (Victoria Gejrot), GH (Gunn Haaland), EK (Egil Kraggerud), MOW (Marianne Ophaug Wehus), ES (Erik Sjaastad), PES (Per Erik Solberg), TIØ (Tor Ivar Østmoe), HB (Halfdan Baadsvik). Michał Kosek, Tekstlab, University of Oslo, has provided assistance in preparing the database for publication online.

There are links from the Repertory to a Bibliography where users will find full references to publications that have been checked (more than 1900); the bibliography also lists around 1600 publications that have not been checked but are more or less likely to contain relevant material.

Users are enouraged to convey suggestions and criticism, or simply to draw attention to the existence of conjectures that have not found their way into the Repertory, by sending an e-mail to Monika Asztalos ( However, no guarantees can be given that all suggestions by the users will be followed up; it is a matter of time and resources. The Repertory is not, and perhaps never will be, exhaustive.

The database is to be referred to as:

Asztalos, Monika, Østmoe, Tor Ivar, et al. Repertory of Conjectures on Horace. University of Oslo, Oslo 2014. Downloaded 2023-11-28 from

The database

The poems by the Roman poet Horace (65-8 BC) have been part of the literary canon for two millennia; still, there is no consensus among scholars about their exact wording. The poet’s original is lost, and what remains from the 1500 years or so before printing was invented is ca. 250 manuscripts, viz. handwritten copies of other extant or lost handwritten copies, the oldest dating from the 9th century. Scribes have unwittingly introduced errors in the text and sometimes replaced what they found in the manuscripts they were copying with expressions that made more sense to them. As a result, the texts of any two manuscripts will differ in a large number of places.

It has been and is the work of classical philologists to try to determine which manuscript readings represent the original. If, in a given passage, they question the authenticity of all manuscript readings, they may propose a conjecture – that is to say, a reading not present in any manuscript – on the hypothesis that it represents what Horace actually wrote. Today classical philologists are faced with a great difficulty: It is virtually impossible, within a reasonable time span, for a scholar to find information about all conjectures that have been proposed on any given passage in printed works.

The Repertory of Conjectures on Horace is a searchable database that allows scholars to retrieve information about ca. 7500 conjectures proposed by, or attributed to, a named scholar in a printed work. The Repertory does not record manuscript readings except in certain cases (see Manuscript reading according to and Lemma in Guidelines). Some manuscript readings were undoubtedly introduced as conjectures by scribes reflecting on the text they were copying, but we can never be certain that a given manuscript reading is a conjecture, since scribes do not provide us with that information. The repertory is a tool enabling scholars to make searches like the ones presented in the five points below. Points 1-2 are of interest primarily to textual critics and readers of the text, 3-5 to historians of classical scholarship and scholars interested in the reception of Horace’s poems.

  1. The single most common question that scholars ask when faced with a difficult passage in Horace’s poems is whether any conjectures have been proposed on that passage. In the Repertory they will find a list of conjectures that scholars have either proposed as their own or attributed to someone else, with references not only to the publications in which they have been proposed or attributed but also to publications containing either discussions of the conjectures or defenses of the transmitted text against them.
  2. If a reading in the critical apparatus of any given edition is followed by a name, the assumption is that it is a conjecture proposed by the bearer of the name. Editors as a rule do not give references to all publications in which conjectures reported in their critical apparatuses have been proposed. By consulting the Repertory users will find references to those publications. They will also find that the reality is much more complex than the critical apparatuses would have us believe. One and the same conjecture has sometimes been proposed independently by different scholars, and quite often one and the same conjecture has been attributed to different scholars.
  3. For each conjecture in the Repertory, users can retrieve a list of publications in chronological order in which the conjecture is discussed. By searching not for conjectures but for units of poetry (words or lines of a poem, an entire poem), these lists of publications amount to bibliographies for students of the text of these units. The lists are not perfect, for it is somewhat arbitrary what constitutes a discussion, and not all relevant publications have so far been checked (more on this in Guidelines). But they reveal which places in Horace scholars keep coming back to. They also indicate that once someone has doubted the transmitted text and suggested a conjecture, this will spawn more discussion and more conjectures. The importance of individual scholars can be gauged by the number of publications in which their conjectures are discussed. The symbol ‘¤’ is used to indicate that a discussion is a defence of the transmitted text.
  4. Users can retrieve lists of all conjectures proposed by a given scholar.
  5. Users can search for conjectures of a certain type. For example, they can find out when remedies such as a change of line order, a new division of poems or a deletion of one or several lines were proposed. They will also be able to retrieve lines that scholars have added to the poems. Combining such searches with years of publication or names of scholars they will find material for studies in the history of classical scholarship.