As regards the observation of eye-movements through photographic recording in order to explain the reading process, much has already been done in the case of the English language; while in Japanese we have only one experiment by Shigeru Otomo, who used Japanese texts in Chicago, and whose results were only partially published.
In this paper the principal object in view is to observe eye-movements in normal reading of various Japanese texts, rapid reading and voluntary eyemovements being also taken into consideration. The method used is that of R. Dodge, photographing a beam of light refle-ion from the eye.
Subjects; Six subjects took part in the experiments; they are students or graduates of the psychological faculty of the University.
Readiny Materials: Selection A. Account of travel Selection B. Humourous story Selection C. Psychological monograph Selection D. Novel Selection E. Leading article Selection F. Natural science Selection G. Account of travel (only for rapid reading)
Results: The subjects are highly educated and have good reading ability, and yet there are no small differences of eye-movements between them. They are almost divided into two types which M. D. Vernon explained in the Medical Research Council (special report series, No. 148, 1930). But one point of Vernon's results, that one of the two types shows a shorter pause duration while the other a longer pause duration, can not be proved in my experiments. So long as my experiments have justified, the two types can be defined as follows: one tends to make a small and regular number of fixation pauses of irregular duration, while the other tends to make a large and irregular number of fixation pauses of regular duration.
Compared with normal reading, rapid reading decreases in duration of pause, and remarkably in number of pauses. But it must be noted that there lies individual difference in the degree of the decreases, and this individual difference seems to affirm the two types of normal reading.
Concerning the speed of reading, smaller number of pauses and shorter pause duration are required to rapid readers, and especially the former seems necessary. This corresponds to remarkable decrease of fixation frequency in rapid reading.
As to the voluntary eye-movements, the sujects were ranked both on the steadiness of voluntary fixation on a point and on the accuracy of voluntary movement from point, to point. Thus the subjects who make a larger number of pauses in reading, are generally less steady in voluntary fixation and less accurate in voluntary movements. My results appear to concur with Vernon's in the conclusion that “the individual types of motor process in reading are originally based upon underlying motor ability, and they may be cosiderably modified by training, and by further individual peculiarites”.
It is surmised that the reading of the Japanese language requires eyemovements of smaller angles and larger number of pauses per line than that of the English language. This fact must be taken into cosideration in researching for the reading of Japanese.
抄録全体を表示