The contrastive analysis of consonant clusters in Lithuanian and English has shown them to be distributed in accordance with Zipf’s curve: the relatively small number of clusters found in the texts account for as much as 75% of the whole inventory. The reason for this should be sought in lexical, morphological and phonetic peculiarities of the languages. It has been found that two-member consonant clusters have greater frequency. This can be accounted for by greater complexity of three-member clusters as compared to their two-member counterparts. Clusters containing affricates occur in fewer contexts than clusters with t-type consonants. The same holds for both clusters with doublefocus (š-type) consonants, and clusters with voiced consonants having additional articulation, e. g. /st/ ≫ /str/, /kl-/ ≫ /gl-/, etc. (where ≫ means that the preceding cluster occurs more frequently). Our analysis confirms the Zipf hypothesis according to which less complex clusters have greater frequency. The frequency of a word, a morpheme or their combinations is another factor which affects the frequency of the cluster.
The comparison has shown that Lithuanian makes use of less complex consonant clusters. English, where monosyllabic words predominate, uses more complex clusters.