The analysis is mainly concerned with the issues of incorrect linking, devoicing of the final lenis consonants and aspiration. The number of cases of linking within the text is forty-two (42), the number of cases where the final lenis consonant is partially devoiced is twelve (12), and the number of cases of aspiration is also twelve (12). The comparison of the degree of devoicing, incorrect linking and not aspirating will be made for both nationalities, for individual speakers and also for the reading part and spontaneous speaking (answering the questions). The Czech and Slovak participants’ pronunciation will be assessed and compared and the hypotheses of the most common mistakes of Czech and Slovak speakers based on the differences between English, Czech and Slovak phonetic systems will be confirmed or disproved. The percentages of the error-making will be counted for both nationalities as well.
4.1.1 Czech speakers
As it is examined in the theoretical part, the Czech language does not aspirate fortis plosives /p, t, k/. This is one of the main areas where Czech speakers make mistakes in English pronunciation. Czech participants of the research did not correctly aspirate especially in spontaneous speech (answering the research questions). However, the error rate is slightly lower in the reading part. The aspiration of the phoneme /p/ is usually the weakest compared to the other two phonemes, nevertheless, it should retain its aspirated character in word-initial positions. Czech speakers neglected this and thus the phoneme /p/ could be confused with the phoneme /b/ in their pronunciation. The aspiration of the phoneme /k/ was often omitted as well, but interestingly, most of the speakers aspirated the initial consonant of the word college, but neglected the aspiration in the words Cambridge and captain. The aspirated phoneme /t/ has a rather hissing sound, which poses a difficulty for the speakers as well. Two of the three speakers aspirated /t/ in the word too, but all of them neglected it in the word time, where a diphthong follows the phoneme /t/.
Each of the Czech speakers devoiced the final lenis consonants, shortened the length of the preceding vowel and incorrectly pronounced lenis consonants at the end of words as fortis consonants. This mainly occurs in spontaneous speaking; the error rate is again lower in reading. The most common mistake was primarily made in devoicing of the final phoneme /s/ which should be pronounced as [z], such as in the words because, studies, and voyages. The assimilation connected to the final devoicing poses a significant problem as well. The speakers devoiced the final consonants in the words such as afterwards, pounds, relations and years. The speakers did not have significant difficulties with the preposition of, which none of them pronounced as the word off.
The Czech participants of the research did not have any significant difficulties with the correct pronunciation of the nasal sound [ŋ], especially when linking with a following word; only one of the participants mispronounced the sound as [-nk] in most of the words ending with /ng/.
Linking of neighbouring words poses a challenge for Czech speakers in continuous speech. All participants had significantly bigger difficulties with linking two words when the first of them ends in a lenis consonant and the following word begins with a vowel. The most common mistakes occur in words such as college in, years and, pounds a (year), voyage or (two). Linking of two words the second of which begins with a consonant is not an area in which the speakers made many mistakes; the participants had problems only with the correct linking of the words encouraged me, where the final /d/ was often incorrectly omitted.
4.1.2 Slovak speakers
The Slovak speakers did not have such difficulties with aspiration as the Czech participants in the research. Only one of the three Slovak participants neglected aspiration more than his Czech counterparts. The rest of the Slovak speakers had either same or slightly lower percentage of mispronunciation in aspiration. The most common mistake of the Slovak speakers is omitting the aspiration of the phoneme /p/, just as it is the most common mistake of the Czech speakers. However, they managed to correctly aspirate the phoneme /t/ in both syllable-initial cases contained in the text (too, time), only one of the Slovak participants neglected the aspiration in these words. The aspiration in spontaneous speaking is neglected by the Slovak speakers as much as it is by the Czech speakers.
The Slovak speakers struggled with voicing of the final consonants just as much as the Czech speakers, the error rate being only slightly lower in the speech of the Slovak participants. The errors made in this area were rather the same: devoicing after a vowel (words such as studies, voyages) and assimilation (words such as pounds, years, relations). There were, however, not as many errors made in the spontaneous speaking as there were in the speech of the Czech participants.
There are no significant differences in the cases of incorrect linking between the Slovak and Czech speakers in the reading part. Slovak speakers did not link about a third of the cases of linking in the text, linking of two words the second of which begins with a vowel was less problematic for the Slovak speakers than it was for the Czech speakers, as for example in the words charge of, pounds a (year), voyage or (two). Interestingly, the Slovak speakers performed much better in linking when speaking spontaneously, as opposed to their Czech counterparts.
4.2 Results of the research
The observations of the research have been collected, counted and expressed in the tables below. The speakers were divided into two groups according to their nationalities. The individual tables for each nationality then assess the three phenomena analyzed in the thesis: linking, final lenis consonant voicing and aspiration. Two of the tables contain numbers of incorrectly pronounced cases of the three main areas when reading the text and speaking spontaneously (answering the questions). However, the results obtained from the spontaneous speech of the speakers are only informative and cannot be compared to the results of mispronunciation in reading, since the answers of the participants were short and did not contain all of the analyzed areas of pronunciation. The remaining two tables show the error rates for each speaker in each area and average error rates for both nationalities. Despite the great effort in assessing the recordings, it is possible that some errors may have been overheard.
The following table shows numbers of cases of linking, final lenis consonant voicing and aspiration contained in the text which was read by each participant. Linking is divided into two groups depending on whether the following word of the two neighbouring words begins with a consonant or a vowel.
Table 1: Numbers of cases of linking, final lenis consonant voicing and aspiration contained in the text
Number of Cases
Final Lenis Consonant Voicing
The two following tables show the numbers of mistakes made by the Czech and Slovak speakers in each area of the analysis. The numbers of mistakes made in spontaneous speaking are included; however, they are only informative since the speakers’ answers differ and do not contain the same numbers of cases of each phenomenon.
Numbers in brackets signify numbers of cases of given phenomena contained in the text
The numbers of errors were counted and the percentage of mispronunciation was then deduced for each speaker and both nationalities. The error rates do not include mispronunciation in spontaneous speaking for reasons mentioned above. The percentages of mispronunciation are expressed in the two following tables.
Table 4: Individual and average error rates of the Czech speakers
Final lenis cons. devoicing
Speaker 1 (F)
Speaker 2 (F)
Speaker 3 (M)
Table 5: Individual and average error rates of the Slovak speakers
Final lenis cons. devoicing
Speaker 4 (F)
Speaker 5 (F)
Speaker 6 (M)
The tables indicate that Slovak speakers are more proficient in aspiration and to a lesser extent in final lenis consonant voicing and linking of two neighbouring words where the following word begins with a vowel. However, the tables show higher error rate for the Slovak speakers in linking of two words the second of which begins with a consonant.