Abstract
This article focuses on the description of the frequent types of the stereotypical reactions in everyday communication. Such reactions represent certain repeating patterns typical of spontaneous speech. They are not dependent on the specific speaker and represent a strategy by which the conversation partner takes the floor. These formulas are searched for in spoken corpora of Czech using the most frequent bigram to je ‘it is’ further combined with other word classes. Thus, trigrams to tetragrams are searched for which are repeated in a conversation and whose function can be inferred from the context. Some of these structures belong to idioms and have their idiomatic meaning but they are usually open patterns using a wider alternation of lexical elements (e.g. combinations with evaluative adjectives). One of these elements tends to be central, which means that they can be detected using only this method. Another important feature is that the most common form of an n-gram is stable, and other words, such as discourse markers are rarely inserted into it. From this it can be concluded that these reactions can be used as compact units, enabling speakers to react quickly in a conversation and express their support, agreement or to prepare for opposition to their partners.
