Abstract
The paper is focused on word-formation structure and its regularity and anomaly of new words (neologisms and occasionalisms) of a composite type (with a focus on quasi-compounds) occurring in current Czech Internet texts. Analysed were primarily media texts, included in the corpus syn2013pub in the Czech National Corpus.
The amount of formal variability and creativeness with the new quasi-compounds (namely the names of new products) is often significantly different from the (primarily) non-internet texts. The formation of some quasi-compounds appears in Czech word-formation system as anomalous and becomes similar or close to mechanic shortening or blending.
They indicate violations of regularity (often they are formed spontaneously), at the same time however, they become models for further formation. Many of them are created under the (still growing) influence of English with its word-formation models and with the aim to attract attention.
Therefore, they are used primarily in certain spheres of communication including communication on the Internet. The paper attempts to illustrate with selected examples, that using some components, foreign in particular, a whole series of names are currently created.