An important finding is that more cohesive ties were used in the target language texts than in the source texts, in other words, the consecutively interpreted Hungarian target language texts hang together stronger as concerns reference. The results indicate that there were more cohesive ties per 100 words of the target language texts than for 100 words of the source language texts, in other words, interpreter trainees added grammatical cohesive devices to the TL texts.

As interpreter trainees progress in their training, they use fewer conjunctions. There seems to exist a shift in the use of grammatical cohesive devices from the use of connective devices to the use of referential ties as students progress in their training.

As concerns specific referential ties, students keep fewer of these ties from the SL text as they progress in their training, however, they also add referential ties to the TL texts that were not present in the SL texts. Additive connectives are used more frequently in the TL texts than in the SL text.

This signals that interpreter trainees are also active text producers, adding their own grammatical cohesive ties to the source language text.