This dissertation looks into how people acquire spatial knowledge from verbal descriptions of space and turn what they learn into accurate and precise mental representations. Certain personal characteristics, such as spatial abilities and native languages of people, were examined in detail to explain the variance in task performance. People in general were able to construct accurate and fairly precise mental representations of a regular space and an irregular space, though there were large individual differences among task performances. In a comparison of people from different cultures, Chinese participants built more precise mental representations than American participants on the same set of spatial estimation tasks, though the groups were similar in cognitive abilities.

This dissertation also investigates personal preference patterns for verbal descriptions of a familiar environment generated by individuals. For each participant there were four types of verbal descriptions of a familiar environment to rank: the route direction exactly as written by him/herself; an expert route direction written by investigators following criteria of good route directions; a sample good route direction selected from participant-generated route directions based on criteria of good route directions; and a sample poor route direction selected from participant-generated route directions based on criteria of good route directions. Results showed that a person preferred his/her own route direction as compared to route directions generated by other people, without realizing it was self-written. This self-preference was so strong that it existed regardless of the quality of the self-generated route direction. Next to this preference, people did prefer route directions that followed already established criteria of good route directions. A further study showed that people did not prefer route directions generated by authors whose personal characteristics were the same as their own. Therefore, there was something special about self-generated verbal descriptions of the environment.

In conclusion, people differ from each other significantly in the way they comprehend and generate verbal spatial information. Cultural groups differ as well, in that Chinese people tend to construct more precise mental representations of the environment they learn from verbal inputs than American people do.