Since its creation, the Korean script has continuously evolved to meet the demands of each era. With the spread of digital environments, new readability measurement standards are required in the era of mobile devices, the internet, and variable fonts....
Since its creation, the Korean script has continuously evolved to meet the demands of each era. With the spread of digital environments, new readability measurement standards are required in the era of mobile devices, the internet, and variable fonts. However, existing research has primarily focused on analyzing static forms, failing to sufficiently address the dynamic visual perception process of readers. Therefore, this study integrates Rudolf Arnheim's theory of visual weight and direction with James J. Gibson's ecological perception theory to propose the “Readability Resistance Model (RRM)” to explain Korean readability.
The research proceeded in three stages.
First, I reclassified the grouping structures of 11,172 Korean characters (2,339 characters based on completed forms × 120 modern fonts) into six types using Arnheim's theory of visual weight and direction. Then, we quantitatively derived four indicators.
1) Visual pattern complexity (Arnheim's “form tension” as intensity),
2) Center of gravity deviation (Arnheim's "weight imbalance" as intentionality),
3) Gaze flow directionality (Gibson's optic flow guidance as contextuality);
4) recovery speed after resistance (Gibson's affordance recovery as recoverability).
Second, we used eye-tracking data from Yonsei University's VCC Lab (reanalyzed from Park et al., 2021; n = 52) to compare and validate metrics for Next Gothic (high resistance–fast recovery type) and Noto Sans CJK (low resistance–sustained type).
Third, I refined the RRM quantitative model for measuring readability by synthesizing the above results into the form VL = k·(VP/VR) · F. The research findings are as follows:
First, Daum Gothic exhibited high intensity and intentionality, resulting in a fixation duration of 248 ms (+42%) and a recognition rate of 87%. This quantitatively demonstrated the paradox that strong resistance enhances memory retention. Conversely, Noto Sans CJK exhibited low intensity and high recovery. It showed a total gaze duration of 1.18 seconds (28% less), confirming the superiority of the low resistance-sustained flow.
Second, principal component analysis (explaining 78.4% of the variance) applied to all 120 modern Korean typefaces consistently reproduced the same four indicators (intensity, intentionality, contextuality, and recovery) and six-type structure. This proves the generalizability of RRM.
Third, by synthesizing the measurement indicators and the results of the statistical analysis of the survey, the refined “Readability Resistance Model (RRM)”—quantified by the formula VL = k·(VP/VR)F—quantified the mechanism in Korean typography whereby intentional visual resistance simultaneously enhances attention and memory performance.
The conclusion of this study expands Korean character design research from static form analysis to dynamic perception and cognitive science. This was achieved by integrating Arnheim's and Gibson's approaches to visual and ecological perception with empirical indicators derived from Yonsei University's eye-tracking data. The four key findings are as follows:
1) Korean character forms are perceived as force structures.
2) Reading is a dynamic flow of visual energy.
3) The weight, direction, and balance of glyphs directly influence visual resistance.
4) RRM is a new analytical framework for interpreting legibility.
The proposed RRM can serve as a theoretical and practical standard for the design of next-generation typography, including digital displays and AI-based variable font generation. It is scalable for future comparative validation studies in multilingual and multimedia environments.