TY - JOUR
T1 - The frame of reference for reading rotated text
AU - Mansfield, J. S.
AU - Legge, Gordon E
PY - 1996/2/15
Y1 - 1996/2/15
N2 - Purpose: Everyday experience suggests that reading is effortless even when text is rotated from its normal upright orientation. Such viewpoint invariance could be achieved if letter recognition in reading was independent of orientation. Alternatively, effortless reading may simply require that letter orientation is consistent with the word's frame of reference. We have investigated these alternatives. Method: We measured reading time (RSVP method) in 3 different orientation conditions: a) WORD, in which each word had a different orientation; b) RANDOM LETTERS, consisting of horizontally oriented words with randomly oriented letters; and c) UNIFORM LETTERS, consisting of horizontal words in which each letter had the same orientation but a different orientation was used in each word. The orientation parameter was selected randomly from a range ±θ, where θ was 0, 60, 120, or 180°. Words had 3, 6, 9, or 12 letters. Results: In the WORD condition, reading time was independent of orientation range θ. In the RANDOM LETTERS condition, reading time depended on θ and word length. Processing time per letter (determined from the slope of lines fitted to the reading-time vs word-length data) were 0.01 sec/letter when θ=0°, and 0.11 sec/letter when θ=180°. In the UNIORM LETTERS condition, processing time was less dependent on θ and word length (0.02 sec/letter when θ=180°). Conclusion: These data show that reading speed is invariant over word orientation. This invariance is not achieved via orientation-invariant letter recognition. Instead, orientation invariance requires that letters in a word have a consistent orientation even if this orientation is different from the word orientation. Effortless reading is possible provided adjacent letters share the same letter-centered coordinates.
AB - Purpose: Everyday experience suggests that reading is effortless even when text is rotated from its normal upright orientation. Such viewpoint invariance could be achieved if letter recognition in reading was independent of orientation. Alternatively, effortless reading may simply require that letter orientation is consistent with the word's frame of reference. We have investigated these alternatives. Method: We measured reading time (RSVP method) in 3 different orientation conditions: a) WORD, in which each word had a different orientation; b) RANDOM LETTERS, consisting of horizontally oriented words with randomly oriented letters; and c) UNIFORM LETTERS, consisting of horizontal words in which each letter had the same orientation but a different orientation was used in each word. The orientation parameter was selected randomly from a range ±θ, where θ was 0, 60, 120, or 180°. Words had 3, 6, 9, or 12 letters. Results: In the WORD condition, reading time was independent of orientation range θ. In the RANDOM LETTERS condition, reading time depended on θ and word length. Processing time per letter (determined from the slope of lines fitted to the reading-time vs word-length data) were 0.01 sec/letter when θ=0°, and 0.11 sec/letter when θ=180°. In the UNIORM LETTERS condition, processing time was less dependent on θ and word length (0.02 sec/letter when θ=180°). Conclusion: These data show that reading speed is invariant over word orientation. This invariance is not achieved via orientation-invariant letter recognition. Instead, orientation invariance requires that letters in a word have a consistent orientation even if this orientation is different from the word orientation. Effortless reading is possible provided adjacent letters share the same letter-centered coordinates.
UR - http://www.scopus.com/inward/record.url?scp=33750170193&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33750170193&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:33750170193
SN - 0146-0404
VL - 37
SP - S167
JO - Investigative Ophthalmology and Visual Science
JF - Investigative Ophthalmology and Visual Science
IS - 3
ER -