Distant Reading across Languages

Repeteme

Dataset Visuals Bibliography Team

Repeteme is an experimental construct that emerged from the discussions of the terminology used to describe the relationships between formal textual features found in the original and its translations as dynamic translatorial responses. The concept was used to model the database in which Faulkner’s novel is aligned with its translations by lexical repetitions arranged into repetition strings in the database. At the early stages of the project, it was felt that the concepts of lemma or headword were not accurate enough to capture the volatility and potential unpredictability of human response that should not be expected to match the prescriptiveness of bilingual dictionary entries and other codified representations of how language changes in translation.

In DRaL, the concept of repeteme may refer to 1) the cumulative effect that a repetition of words may have on a reader/ translator, and to 2) an entire repetition string which is a collection of variant word forms serving as a physical representation of that effect in the database. The construction of repetemes begins with the original text which provides the basis by which English repetition strings are compared with the choices made by translators. Repetition string is a physical representation of a repeteme in the database. The string constructs serve to track the patterns of human response to the effects that repetitions may have on a reader / translator. Occasionally, we use the terms repeteme and repetition string interchangeably depending on whether we want to explain how translatorial response relates to its overall effect or to the properties of repeteme in the original text.

Unlike isotopy in A.J. Greimas’ semiotic theory that captures repetition beyond the formal resemblance of words, repeteme is an abstract unit designed to capture psychosemiotic reverberations that formal repetition may have on a reader/ translator.

The following variants may constitute the formal expressions of such repetemes:

  1. inflectional word forms that occur on their own (e.g. fish/fishing), including homonyms (leaf/leaves/to leave);
  2. cognate words (e.g. kind/kindness; valuable/invaluable);
  3. compound words (e.g. bath(room));
  4. the forms that represent phonetic variation in spoken or dialectal discourse (e.g. mother/mammy/maw).

The need for such an unusual treatment of repetitions comes from the observation that the translators made dictionary-based and word-for-word choices as well as unusual options characteristic of creative or even misled interpretation. On the lexicographic level, compounds may be treated as dictionary entries separate from the words that comprise it. However, they are treated in a variety of ways across translations. The Lithuanian translator, for example, renders the instances of bathroom as vonios kambarys (bathroom) or vonia (bath). The compound bluegum as in 'bluegum children' in the Benjy chapter represents a far more interesting case for translation. The Digital Yoknapatawpha Project explains that the adjective refers to 'a black person whose lips and/or gums look blue'. Both Russian versions keep the meaning of 'blue' but deal differently with the second part. In the Russian version by Soroka, the word is translated in a literal manner as синедесный (adj. 'bluegum'), while Gurova renders it as синегубый (adj. 'bluelip'). Przedpełska-Trzeciakowska, the first Polish translator of Faulkner's novel, is as literal as Soroka, i.e. sinodziąsłe where dziąsła means 'gums'. Curiously, the Polish version by Polak contains an unexpected choice wilkołackie (adj. 'of werewolf') which loses any connotation of the blue colour. Instead, it evokes a very different imagery from horror folklore. Therefore, the decision in the DRaL project was to enter the compound twice in the database under the repeteme entries BLUE and GUM in order to track separately the retention of 'blue' and 'gum' in different contexts of the original.

It is by tracking translatorial choices along each repetition string that we can detect and observe what patterns they produce from shape of which illustrates individual preferences rather than systemic linguistic differences. The findings show varying degrees and patterns of repeteme awareness.

Repeteme pointers are words used to find a specific repetition string in the database and to label translatorial responses in the visualisations. Repeteme pointers are labels rather than headwords used in lexicography.

The choice of repeteme pointers for each repetition string is pragmatic rather than strictly methodological. Wherever possible, this is a basic unit of formal repetition that is closest to the root of the word forms that compose the repetition string (of repeteme variants) and can be looked up in the OED. Therefore, whenever only a plural form of a noun or a conjugated form of a verb is found in the original text, the repeteme pointer is that noun in capital letters and in singular form or the infinitive form of that verb in capital letters (e.g. “girls -> GIRL”, “fusses -> FUSS”). This means that sometimes repeteme pointer may be a form that does not actually occur in the original text.

Whenever a repetition string is formed of an unusual word form only (e.g. “projecking” found in the Benjy chapter of The Sound and the Fury) and no other similar word forms are to be found in the chapter or the OED, this form is used as repeteme pointer.