Aim and material: In the present study, the data of the rating study presented earlier, where participants estimated the position of an object or phenomenon in vertical space by using the seven-point Likert scale, are used in order to systematically describe spatial information included in language units of different semantic categories. Background: The role of spatial semantics in language understanding is assumed by modern cognition theories and confirmed in experimental studies. Hypotheses: Based on conceptual metaphor theory, a number of hypotheses are proposed in the present study: different semantic categories of nouns (e.g., tools vs. animals vs. emotions etc.) should significantly differ in their spatial semantics as well; different semantic categories of abstract nouns (e.g., mental states vs. emotions vs. physical sensations) should also differ in their spatial semantics, as the latter is included in their conceptual structure; mental states and phenomena (e.g., imagination, thought or memory) should have higher values (i.e., be located higher in the virtual subjective space) than any other abstract concepts; emotional concepts (e.g., love, disgust or happiness) should be located higher than physical sensations (e.g., pain or softness); positive emotions (like joy or euphoria) should be located higher than negative ones (like feeling of guilt or disappointment). Methods: Statistical methods (parametric and non-parametric ones) are used in order to test the hypotheses. As additional cross-testing methods corpora data and expert assessment are included. Results: The results of the study confirmed all the hypotheses. A number of additional regularities were revealed: in general, abstract concepts get higher values on a scale, i.e., are located higher in the virtual subjective space than concrete ones; tool concepts are related more to the lower space, unlike sound concepts that are related to higher space. No difference was found between action concepts (like attack or running) and physical sensations. Discussion: The results obtained can also be explained in terms of other theories within the embodied cognition framework, as it is discussed in the conclusion (words as social tools by A. Borghi and F. Binkofski; ideas by G. Vigliocco and neurosemantic approach by F. Pulvermuller). The need in an integrative model and larger studies with other semantic categories is underlined.

Проведен анализ пространственной семантики различных категорий русских существительных, входящих в психолингви-стическую базуданных; особое внимание уделяется абстрактным концептам. Выявлены различия пространственной семантики наименований физических ощущений и действий, эмоций, ментальных процессов. Полученны ерезультаты обсуждаются с точки зрения отдельных подходов в рамках теории воплощенного познания – теории концептуальной метафоры, теории слов как социальных инструментов (WAT, Words As social Tools), нейросемантики.