The key purpose of this work is to examine the interrelationship between the field of text linguistics and translation with specific reference to computer translation. The question arises whether machine translation can ever be a practicable and reliable substitute for human translators. Based on this premise, this study assesses the effectiveness of machine translation software in its ability to translate the nuances of text linguistics from a source language to a target language. Following a literature review of text linguistics covering a number of textual analysis models and translation studies including machine translation, the primary research utilises a qualitative research methodology by means of text-based assessment. Three text samples are drawn from each of the following six text linguistic categories: register, pragmatics, semiotics, text type, genre and discourse. Using three leading Arabic translation software programmes the text samples are subjected to a comparative evaluation.