The paper approaches an important issue of the phonological similarity of words, relevant for current research in phonotactics, word recognition, production and acquisition, by analyzing the data collected in an experiment in which 30 native speakers of Polish were asked to provide phonologically similar words to 80 nonwords. The study demonstrates that the uncovered patterns of phonological similarity (segment substitutions, deletions and additions, the use of bigrams, trigrams and quadrigrams, noncontiguous sounds and segment metathesis) go beyond the commonly employed concept of neighbourhood density and point to the need to revise the current approaches to phonological similarity of words. It is argued that the experimental results can be attributed to the considerably more complex phonotactic and morphological structure of Polish than English.

The present study is the first attempt at examining the perception and evaluation of 10 internationally known political and religious leaders’ English pronunciation. 40 Polish students’ assessed their speech samples in terms of the degree of foreign accentedness, comprehensibility and acceptability. We examine whether the following factors affect the assessors’ judgements: their personal attitude to the speakers, the students’ level of English proficiency and the genetic proximity between between the speakers’ and the listeners’ L1s combined with the raters’ familiarity with foreign accents of English. It is demonstrated that the listeners’ attitude to the speakers has no impact on the ratings of the samples’ comprehensibility and accentedness, but plays an important role in their evaluations of acceptability. The participants’ level of English proficiency is crucial for their assessment of comprehensibility, but not accentedness and acceptability. Finally, the genetic proximity between the involved languages and the listeners’ familiarity with varieties of foreign-accented English are shown to be relevant for all the presented accent jugdements.