The relationships between objective and subjective evaluations of the urban environment: Space Syntax, cognitve maps, and urban legibility

Abstract

ABSTRACT LONG, YIXIANG. The Relationships Between Objective and Subjective Evaluations of the Urban Environment: Space Syntax, Cognitive Maps, and Urban Legibility. (Under the direction of Perver K. Baran and Robin Moore.)

To some degree, the urban environment can facilitate/limit one?s orientation, depending on the structure and characteristics of the physical elements of the city. In this regard, Lynch?s concept of legibility has been fundamental in the urban design, planning, architecture, and environmental design fields for a few decades. Lynch argued that a strong legible city could facilitate humans? orientation in the city. However, urban design research has also criticized Lynch?s work for ignoring the relational characteristics between physical elements of the urban environment. Recent research has suggested that Space Syntax methodology could address the limitations of Lynch?s approach to urban spatial cognition.

The study focuses on exploring: a) the relationships between human cognitive representations and spatial configuration of the urban environment, and b) the effects that different spatial configurations have on legibility of the environment. Two research studies, correlational and experimental, were performed in this study. Two neighborhoods, one relatively intelligible and the other less so, in Changsha, China, were selected for the study areas. The Space Syntax approach was utilized to measure spatial configuration of the neighborhoods. Sketch maps, recognition tests, and interviews were used to measure individuals? cognitive representations and their perceived legibility of the environment.

Overall, the results of the correlational study indicated that there exists a positive association between cognitive representation and spatial configuration. In particular, the multiple regression analysis showed that global integration is the only significant variable explaining the variation in landmark scores, whereas local syntactical measures (local integration and connectivity) do not play a role in predicting landmark scores. For paths, the multiple regression analysis showed that local syntactical measures (local integration and connectivity) are the two significant variables explaining the variation in path scores, whereas global integration does not play a role in predicting path scores.

The results from the experimental study indicated that participants? mean path scores in the sketch maps, the mean scores of scene recognition, and the three mean scores of spatial cognition ability in a more intelligible neighborhood are much higher than those in a less intelligible one. However, no difference in landmark recognition was found between the high and low intelligible neighborhoods. Overall, the findings suggest that the more intelligible an area is, measured by Space Syntax as objective value of spatial configuration, the more legibly it is reflected in a human?s spatial cognition.