Error, a fundamental concept in the common errors method, refers to mistakes in the transmission of a text. As such an error is a type of secondary reading and in some cases is used as shorthand for secondary reading (West 1973, 32, note 3). An error can also be seen as an innovation in the textual tradition, although the latter term generally implies intention whereas error commonly implies an unplanned change. Generally speaking, errors represent copying errors although other processes of reproduction, such as dictation, can produce errors.
The common errors method uses significant, or kinship-revealing, errors to posit relationships among witnesses. Conjunctive errors indicate a common ancestor, that is they allow witnesses to be grouped together into families. Separative errors demonstrate the independence of a witness from another and allow the critic to distinguish relationships within a family. The determination of errors (or more clearly, secondary readings) and the relationships among witnesses allows one to root a textual tree, that is establish the position of the archetype in a stemma.
The term, error, finds disfavour among a number of textual scholars for a number of reasons, including the pejorative connotations associated with the word. Contemporary computer-assisted statistical methods generally employ not errors but variants, which are used to group witnesses. Consequently, these methods and their tools produce undirected trees or attempt to root the tree by a variety of methods.
Because of the centrality of the term in textual criticism, scholars have developed a rather extensive descriptive vocabulary to describe errors (and variants) and often the processes by which they arise (for examples, see types of errors). If one considers only the result (rather than the impetus), all errors (and indeed secondary readings and innovations) can be categorised as omission, addition, transposition and substitution.
The contrast between terms describing the impetus for an error and those examining the result, is evident in comparing homoeoteleuton and omission: homoeoteleuton describes an omission resulting from eye-skip between words with similar or identical endings; omission notes the absence of text but posits nothing as to how the phenomenon arose. Terms that describe the cause of a secondary reading lend little to categorisation, but can be helpful in assessing the significance of a reading.
– West, Martin L. 1973. Textual Criticism and Editorial Technique Applicable to Greek and Latin Texts. Stuttgart: Teubner.
In other languages
Unknown User (15conti)
At the end of the third paragraph: "or attempt to root the tree by a variety of methods."
This is something of a cop out. I remember both Teemu and Tara talking about rooting their trees using Stemsem and Stemmaweb, and a number of discussions related to this problem. But am not familiar enough with the tools and methods to remember the specifics on how this was done.
And while the pages for Stemsem and Stemmaweb both claim that the tools propose stemmata, I did not see an explanation of their rooting. (Indeed, under semstem there are two trees which are called stemmata in the figure).