Language scale is a sequence of alternative words or clauses that are ordered according to lessening of informativeness, intensity or volume of some feature, for instance, the sequence splendid, beautiful, and good form the lexical scale. Scales are rather numerous and intrinsic to every language, they are commonly used for measuring during reasoning and speaking. Similar to other internal components of a language, cognitive aspects of these scales are unobservable directly and may be investigated only indirectly by linguistic, psychological or neurophysiological means. The paper considers the former. After summarizing the current state-of-Art in studying language scales, several observations about cognitive aspects of these linguistic phenomena are discussed. The paper considers semantic and pragmatic aspects of scales, although specially focuses on the latter, particularly on interlingual comparisons and studying accessibility of the components of a scale by the means of associative vocabularies. The main ussie of the paper consists in studying hidden features of language scales by the means of two types of instruments: linguistic corpora and associative dictionaries. Linguistic resources include data from Russian, English, and French languages.