Orthographic codes are used in integrating information from the parafovea by the saccadic computation system.

Two experiments that examined the characteristics of information extracted from a parafoveal letter string by the saccadic system are reported. In Experiment 1, we observed the effect of the orthographic legality of the initial letter sequence on landing positions. The deviation of landing positions towards an orthographically illegal letter sequence was shown to occur in saccades when subjects were instructed to read a nonsense letter string. Moreover, it was shown that the initial placement of the eyes is influenced by the frequency of occurrence of the letter sequence and not by the frequency of the letters that formed this sequence. Experiment 2 followed up on the implication that readers use orthographic information only from the initial letters of the parafoveal string. We concluded that the saccadic computation system integrates orthographic information from the two initial letters.