Measuring the Degree of Translation Difficulty and Text Readability: A Statistical-mathematical Praxis

Understanding the difficulty level of translation tasks is of paramount significance in the field of translation evaluationproducts, translation pedagogy, and translation research. Truth be told, translation researchers and scholars do nothave access to any benchmarks to determine the difficulty level of source texts in process-oriented translation research.In this direction, this makes it intricate for a researcher to assess/evaluate the comparability of empirical results amongthese research studies. Text readability is mainly associated with translation difficulty level of a source text. Accordingto Sun and Shreve (2014: 99), ‘translation difficulty can be viewed as the extent to which cognitive resources are takenup by a translation task for a translator to achieve objective and subjective performance criteria’. Difficulty isconsidered an issue that it is firstly applied to tasks and thus task difficulty or task readability is more objective andmust be calculated through a number of statistical and mathematical formulae. Jensen (2009) has pointed out that textreadability/difficulty equations can be applied to measure the difficulty level of a translation. Text readability formulaesuch as SMOG, LIX, and so forth employ two strong factors for determining the level of a source text readability. Thesefactors include semantic factors with a focus on vocabulary and terminology difficulty, syntactic factors with a focus onsentence length. With this idea, the process of translation involves reading and re-verbalization (Wilss 1982) and inturn, the sources of translation readability/difficulty involve difficulty level of a text, factors associated with atranslator, and translation-specific difficulty (Baker 2011). In this light, the purpose of the present research paper istwo-fold: (i) this paper aims to identify and measure the level of translation difficulty through the application of pdocimology.P-docimology refers to the proportion of examinees who provide a correct answer for an item of interestdivided by the total parent population. To put theoretical p-docimology into practice, we have tested three types of text(easy, medium, and difficult) from English (L2) to Persian (L1) over a number of 20 translation participants using thePreselected Items Evaluation (PIE) method. The results showed that as the level of p-docimology went up, the level oftext readability decreased (a trade-off condition). This is mainly due to the fact that the bivariate correlation wasnegative and negatively skewed. And (ii) this research paper aims to formulate a statistical-mathematical equation tooutguess/predict the difficulty level of a translation without utilizing any complex statistical and mathematicalequations. To do so, we used NASA Task Load Index (TLX), a multidimensional scale for evaluating and measuring atranslator’s mental workload, to measure the difficulty level of a translation based on a holistic translation rating scale.

Keywords:

Text Readability , Translation Difficulty , P-docimology , NASA Task Load Index