Connected Speech Phenomena PDF

Connected Speech Phenomena Assimilation It is a phonological process in which a consonant sound changes to become more like an adjacent sound in connected speech. It involves changing one one or more articulatory features (voicing, point of articulation, manner of articulation) of one or the two so...

Connected Speech Phenomena Assimilation It is a phonological process in which a consonant sound changes to become more like an adjacent sound in connected speech. It involves changing one one or more articulatory features (voicing, point of articulation, manner of articulation) of one or the two sounds. It’s defined by directionality—progressive or regressive—and by the phonological feature(s) being changed. Progressive assimilation: The sound that changes is affected by the one that precedes it. A mostly overlooked instance of progressive assimilation of voicing is the devoicing of laterals, rhotacised sounds and glides after voiceless consonants. Moreover, rhotacised sounds also change their MOA from approximant to fricative—which also happens when they follow a voiced consonant. [pʰl ̥iːz] [fɹ̝a ̥ ɪ] [kʰw̥ aɪṯ] [fju ̥ ː] [ḏɹ̝ɛḏ] Another instance is when [s, z] are followed by [ð], where the latter becomes [s, z], respectively. [ɪz zaṯ̚ ṯʰɹ̝u ̥ ː] [ðaṯs si aɪˈḏɪə] Similarly, when [ɫ] is followed by [ð], as alternative to making the dark ‘l’ dental, the [ð] can become a clear ‘l’. [ɔːɫ lə ṯaɪm] Lastly, we have progressive devoicing of the contracted forms of “is” and “has” after voiceless consonants (most commonly /t/). [wɒṯs]  [ðaṯs] Regressive assimilation: The sound that changes is affected by the one that follows it. Alveolar sounds are the most susceptible to this kind of assimilation. Labialisation: When [ṯ, ḏ, ṉ] are followed by [p, b, m] and sometimes [w], they are pronounced [p̚, b̚, m], respectively. [ðap̚ bɔɪ] [bab̚ pʰɜːsṉ̩] [ɒm ˈmʌṉḏeɪ] [ɪm bɜːˈlɪṉ] “Labio-dentalisation”: When [ṉ] is followed by [f, v], it is pronounced [ɱ]. [ˈɪɱfɪnəṯ] [ɪɱˈvaɪṯ] Dentalisation: When [ṯ, ḏ, ṉ, ɫ] are followed by [θ, ð], they are pronounced [t ̪̚, d̪̚, n̪, ɫ̪], respectively. [ət ̪̚ ðə ˈməʊməṉṯ] [bad̪̚ θɪŋz] [ˈan̪θəm] [ɔːɫ̪ˈðəʊ] Velarisation: When [ṯ, ḏ, ṉ] are followed by [k, ɡ], they are pronounced [k̚, ɡ̚, ŋ], respectively. [ðak̚ ɡɜːɫ] [ɹəʊɡ̚ kʰɹ̝ɒ ̥ sɪŋ] [ˈɪŋkʌm] [ʌŋˈɡɹ̝eɪṯ̚fəɫ]  Palatalisation: When [ṉ] is followed by [c, ɟ, j, tʃ, dʒ], it is pronounced [ŋ̘]. [bɹ̝aʊŋ̘ cʰaṯ] n [k ŋ̘ ɟɛṯ] [ɪŋ̘ ˈjeləʊ] [ɪŋ̘tʃ] [ˈɪŋ̘dʒəɹi] /s/ and /z/ followed by /ʃ/: When [s, z] are followed by [ʃ], they are pronounced [ʃ, ʒ], respectively. In the case of [z] assimilating into [ʒ], devoicing may occur. [bʌʃ ˈʃɛɫṯə] [ˈɪʒ ʃi]→[ˈɪʒ̥ ʃi]-[ˈɪʃ ʃi] Voicing: Only voiced consonants can become devoiced, it does not go the other way around. It mostly involves any voiced fricative [v, ð, z, ʒ] followed by a voiceless consonant, turning the former into their devoiced or voiceless counterparts [v̥, ð̥, z̥, ʒ̥ - f, θ, s, ʃ]. [hav̥ ṯə]-[haf ṯə] [wɪð̥ θɪŋz]-[wɪθ θɪŋz] [haz̥ ṯə]-[has ṯə] [ˈɪʒ̥ ʃi]-[ˈɪʃ ʃi] Reciprocal assimilation: Both sounds adjust mutually. The two most common cases are [s, z] followed by [j], resulting in [ʃʃ, ʒʒ]; and [ḏ, ṉ] followed by [ð], resulting in either [d̪̚d̪] or [n̪n̪], as an alternative to keeping [ð] and making a regressive assimilation of POA of the first sound. [ðɪʃ ʃɪə]  [ðəʊʒ ʒɪəz] [ɒn̪ n̪aṯ] [ṉiːd̪̚ d̪ɪs] Coalescent assimilation: Both sounds merge into one. This only happens between the alveolar stops /t,d/ and the yod /j/, turning them into affricates with the corresponding voicing of the stop (/tʃ/ and /dʒ/, respectively). [bəˈtʃuː] [ˈḏɪdʒə] Elision It is the phonological process in which speakers omit sounds within the flow of speech to facilitate rapid or casual pronunciation. These elisions do not occur randomly, but in a set of particular contexts. Elisions may be historical or contextual (rule-guided), and both consonants and vowels may be elided, but for the test, we are only concerned with contextual consonant elisions. Elision of /h/: /h/ is commonly elided in the weak forms of after a consonant, but not when they are in absolute initial position or when they are in an accented syllable. [ɟɪv ɪm ɪz bʊk] Elision of /t/ and /d/: In rapid, casual speech, /t/ and /d/ are commonly elided at word boundary or word-internally when they are the central element of a 3-consonant cluster or sequence. [lʊk̚ ṉaɪs] [kʰəʊɫ ˈkʰɒfi] [faʊŋ wʌṉ] or [faʊm wʌṉ] In cases such as the last one, elision may create a context for assimilation.  There are, however, three key contexts in which /t/ and /d/ are not normally elided: 1. When in a sequence of 3 consonants, the third element is /h/. [baḏ ˈhabɪṯ] 2. When /t/ and /d/ are preceded by a vowel and followed by /h/. [ðaṯ haʊs] 3. When /nt/ and /lt/ are followed by a consonant without assimilatory influence. In these cases, /t/ may be glottally reinforced or glottalised. [ḏəʊṉʔṯ̚ ɹʌṉ] or [ḏəʊṉʔ ɹʌṉ] Elision of /t/ in “not” contractions: /t/ is commonly elided at word boundary when it is the final sound of the contraction “not” and the following word begins with a consonant. [wəʊŋ gəʊ] [kʰɑːm baɪ] [ḏəʊŋ̘ ʃaʊṯ] Elision of /ð/: /ð/ is, in casual styles, elided in unaccented [ˈɟɪvəm] [ˈʃəʊəm] [ˈṯʰeɪkəm] Elision of /v/: The /v/ in is commonly elided before unaccented [ðə] and other consonants as well. [ˌaʊṯəðəˈweɪ] [ˌpʰiːsəˈpʰeɪpə] [əz ə ˈmaṯəɹ ə fak̚ṯ]  Elision in and : Both of these words have alternative forms with elision of their dental fricatives /θ, ð/. [mʌṉs] [kʰl ̥əʊz] Elision in numerals: Words like don’t elide /θ/ but may instead elide the preceding consonant. [fɪθ] [ṯʰw̥ ɛɫθ] Linking Also known as liaison, linking refers to the phenomenon where sounds are linked across word boundaries to facilitate fluidity in speech. It happens when a word ends with a consonant and the next one begins with a vowel. [əṉ ͜ ˈəʊɫḏ ͜ əʊk ṯʰɹ̝i̥ ː] For this test we should also consider instances of linking ‘r’. In non-rhotic accents (like GB), word-final is pronounced when the next word begins with a vowel. [aɪ ˈḏəʊṉṯ̚ cʰɛː] [aɪ ˈḏəʊṉṯ̚ cʰɛːɹ ͜ əˈbaʊṯ ɪṯ] What we don’t need to consider for the test are instances of linking /j/ (palatalisation) and linking /w/ (labialisation). Intrusion Also known as epenthesis, intrusion refers to the insertion of a sound at word boundaries or word-internally where these sounds are not suggested in word spelling. Epenthetic sounds can facilitate smoother pronunciation by creating extra articulation points. /p/, /t/, /k/ and /r/ are used as epenthetic sounds in English. We are only concerned with intrusive ‘r’ for this test, though. Intrusive ‘r’ occurs in non-rhotic accents like GB. An /r/ sound (which is not represented in the written form of the word) is inserted between /ə, ɪə, ɔː/ and any following vowel. [ði aɪˈḏɪəɹ əv] [ˈḏɹ̝ɔːɹɪŋ] [ˈɹɔːɹ ˈʌŋ̘jəṉz]