(From the introduction) During years coherence in texts has been understood as being woven through the different cohesive devices detected and classified within different types of cohesion. The origin of this classification can be found in the model proposed by Halliday & Hasan (1976). It included the following types of cohesion: conjunction, lexical, reference, ellipsis and substitution. This basic classification of the different types of cohesive devices has not varied through the years. In fact, we still find this basis in recent studies of discourse cohesion such as in Carter’s et al. (1997), although with some important remarks, as we will see. Lexical cohesion started to gain importance when it was discovered that this type of cohesion reflected the way texts worked beyond the propositional level. This implied an awareness of the semantic bearing that cohesive devices had in relation to their abstract connections out of the text. An example of this can be found in Haynes (1989). He made a further classification grouping the five types of cohesion within just two: lexical cohesion (including lexical cohesion itself) and grammatical cohesion devices (namely: substitution, conjunction, ellipsis and reference). The relation that existed among the elements belonging to the lexical cohesive type can be reflected according to Carter et al. (1997) by means of: direct repetition, synonyms or near synonyms, superordination (hyperonyms), antonymy, specific-general reference, ordered series and Whole-part relations. As opposed to the grammatical cohesion type, these relations within lexical cohesion provided a description of the boundaries that existed among different items belonging to the same conceptual field. Lexical cohesion may be the most appropriate device to maintain and reinforce the “aboutness” of a text, since it reflects the different lexical relations among the events or elements that the text is dealing with. (Publisher Book Description)