Translation as a production-driven activity Michael Carl and Barbara Dragsted ISV, Copenhagen Business School, Dalgas Have 15, DK-2000 Frederiksberg Abstract It is controversial in the translation process literature whether human translation is an iterative process of comprehension-transfer-production or whether and to what extent comprehension and production activities may occur in parallel. This paper suggests a novel interpretation of Tirkkonen- Condit’s literal default rendering procedure. We compare text copying with translation activities under the assumption that text copying is a prototypical literal default rendering procedure. Both activities require decoding, retrieval and encoding of textual segments but translation requires in addition a transfer step into another language. Comparing user behaviour obtained in copying and translation experiments, we observe surprisingly many similarities between the two different activities. Copyists deviate from the default literal text reproduction into more effortful text understanding, and much of the translators’ behaviour resembles that of copyists. Instead of the comprehension-transfer-production cycle we observe that production problems trigger comprehension revision, during translation as well as during text copying. Introduction According to the eye-mind hypothesis (Just & Carpenter, 1984) 1 there is a strong correlation between where one is looking and what one is thinking about. The eye-mind hypothesis is controversial, on the one hand it is well known that the observation of a correlation between two events does not imply a causal relation, and hence no conclusion can be made regarding the existence or the direction of cause and effect only from the observation that two events, e.g. gaze location and mental processes, correlate. On the other hand, the strong correlation assumption between gaze and mind has be questioned, e.g. by (John R. Anderson, Dan Bothell, and Scott Douglass, 2004) who find that longer gaze durations do not correlate with greater problems of memory retrieval. In Translation Process Research it has often been stated, that gaze location reflects the focus of attention of the translator (cite). That is, when the gaze focusses on the source text (ST) the mind is involved in ST comprehension process, and when the gaze is directed on the TT, the mind is involved in text production processes, where longer gaze durations reflect bigger translation problems (cite). These assumptions fit well a stratificational process model of translation, which states that at any one time the translator either reads (understands) the ST or is busy with typing its translation. Craciunescu et al. (2004), for instance, claim that “the first stage in human translation is complete comprehension of the source language text”. Only after this complete (i.e. deep) comprehension is achieved can the translation be produced. Similarly Gile (2005), suggests a stratificational translation process model, in which a translator iteratively reads a piece of the ST and then produces its translation. First the translator would create a “Meaning Hypothesis” for a ST chunk (i.e. a Translation Unit) which is consistent with the “context and the linguistic and extra linguistic knowledge of the translator” (p. 107) for which then a translation can be produced. Also Angelone (2010) supports that translators process in cycles of comprehension-transfer- production and that “uncertainties” of translators can be attributed to any of the comprehension, transfer, or production phases. He claims that “non-articulated indicators, such a pauses and eye- fixations, give us no real clue as to how and where to allocate the uncertainty” [p.23] 1 “there is no appreciable lag between what is fixated and what is processed"