Abstract
This corpus study describes vowel phonotactics in Czech words. The results suggest that some probabilistic patterns are employed in Czech: some vowel combinations are overrepresented while others are underrepresented.
A syllable containing a short front vowel tends to be followed by a syllable with a long front vowel. A long front vowel is typically followed by a back vowel and a long back vowel tends to be followed by a short vowel; thus, an interesting circular dissimilative pattern can be observed.
An explanation of the phenomena can be facilitated by the Shannonian theory of communication. The analysis was performed both on words and word stems (i.e. words without endings), obtaining different results.