Ancient and patristic texts, with few exceptions e.g., epigraphical texts and some works on papyrus, have been passed down to us not in the original but through later manuscript copies, some even many centuries distant from the original. Since, through the entire time of transmission until a printed version, every work has undergone an inevitable process of progressive evolution or corruption, the purpose of a critical edition is to restore the original text to the degree possible, accompanied by an apparatus of documentation critical apparatus. In the first two centuries of the age of the printing press, editors had scant or no knowledge of the historical process of the manuscript transmission of the ancient work they were about to print.
More often than not the first editions editiones principes were based on a single manuscript, whose text was emended based on subjective criteria ope ingenii. The 18th c. began to see an awareness of the importance of the systematic study of the manuscript tradition. Later, the recognition that a text’s tradition proceeded genealogically, from one manuscript to another, and that the main factors that caused it to deteriorate during its transmission were recurrent and classifiable, allowed 19th-c. philologists to elaborate a scientific method of critical edition K. Lachmann’s method, which has two main stages: the recensio codicum and the constitutio textus. Recensio refers to the tracking down of all the manuscripts and also of indirect testimonies to the work, their historical, codicological and paleographical study, their collation, and the definition of their relations of dependence: the results of this work are represented in a stemma codicum genealogical tree of the tradition, which gives information useful or even decisive for the choice of lectiones.
When the manuscript tradition has been contaminated, the construction of the tree may not be possible. It is usually easier to reconstruct the path of the tradition of patristic-era works than that of ancient works, since we rarely have direct and certain evidence of the diffusion of the latter in the ancient period. For patristic works the archetype the reconstructable manuscript, from which the entire known or knowable tradition has descended is often a codex near to the original, or even the codex of the original edition, whereas for ancient works, when they exist, it is usually a codex from the period between the 5th and 8th c. The second phase of the editorial work is the constitutio textus, the point at which the editor, judging the authenticity of the various lectiones, performs the selectio, or perhaps the emendatio, if the entire tradition is corrupted. The results of the recensio and the criteria that preside over the constitution of the text are described in the praefatio of the edition.
The traditional rule, according to which the praefatio must be written in Latin, is today often ignored. The manuscripts used for the edition are listed in a conspectus siglorum. These characters letters of the Latin alphabet for preserved codices, Greek letters for reconstructable lost MSS serve to synthetically compile the apparatus criticus, where the variae lectiones which the editor did not approve are recorded, and the conjectures of the editor himself or of other scholars or past editors are indicated. The apparatus fontium contains internal references and parallel passages with works of the same or different authors, and the identification of citations, allusions and sources. A conventional system of diacritical signs highlights particular accidents that have happened to the text: omissions, interpolations, lacunae, irreparable corruptions. Texts passed down by a single witness inscriptions, papyri, palimpsests are usually published in a diplomatic edition. When there are various original redactions, it is possible and useful to publish them in parallel. Specific models have been worked out for the publication of fragments. Often the edition is provided with a specialized bibliography. It is considered important, and even more so before the recent advances in technology, that the critical edition be provided with good indexes.