This paper addresses the issue of underlying representations (URs) from a Guierrian perspective and the necessity of taking into account certain orthographic elements which are associated with phonological behaviours related to English stress. We propose that underlying vowels should be represented as abstract phonological objects and that the inclusion of orthographic consonant geminates and final mute <e> into URs should be considered. Additionally, we argue that the phonology should have access to morphosyntactic information, which implies that the input to the phonology should be polystratal. Eventually, after arguing that vowel reduction should occur after stress assignment, we will report the results of studies on vowel reduction and "stress preservation" showing that reduction is not systematic in unstressed syllables and that non-reduction can, in some cases, be attributed to the existence of a full vowel in a morphologically-related word or to a high frequency of the latter.

New instances of palato-alveolars have been reported in various regions and countries of the English-speaking world over the past few decades. Those new instances of /ʃ, tʃ, ʒ, dʒ/, which I propose to bring together under the common name of Instances of Contemporary Palatalisation (henceforth, ICPs), are the results of processes of palatalisation. I argue that ICPs can be seen as the products of a phonological process that finds its roots in the context of significant post-World War Two social and political changes, both within and beyond the British Isles. What is common to all four ICPs is not the input but the output of the process, that is to say systematic palato-alveolars. The phonological view adopted is therefore that Contemporary Palatalisation expresses product-oriented, rather than source-oriented generalisations (cf. Bybee, 2001: 126). I argue that ICPs are the continuation of a historic pattern endemic to English, which has systematically yielded palato-alveolars throughout the history of the language. This paper presents selected results of a corpus study based on a PhD project undertaken at the University of Lyon.

This paper investigates perceptions of speaker-indexical information from gender- specific phonetic variables in the absence of speakers' fundamental frequencies. The results revealed that listeners not familiar with a dialect under investigation were not sensitive to speaker-indexical information embedded in the phonetic variants. The results also showed that in their evaluations, male and female listeners often did not differentiate between localised and supra-local variants. Finally, the perceptual differences noticed between male and female listeners were not statistically significant.

This study uses eye-tracking to single out the role of 'wild' onomatopoeia in language development, as described by Rhodes (1994). Wildness–whereby extra-phonetic features are used in order to reproduce non-human sounds–is thought here to facilitate infants' understanding of onomatopoeic word forms, providing a salient cue for segmentation and understanding in the input. Infants heard onomatopoeic forms produced in familiar and unfamiliar languages, presented in a phonologically 'wild' (W) or 'tame' (T) manner. W forms in both familiar and unfamiliar languages were hypothesised to elicit longer looking times than T forms in both familiar and unfamiliar languages. Results reflect the role that onomatopoeia play in early language development: wildness was not found to be a factor in infants’ understanding of word forms, while reduplication and production knowledge of specific stimuli generated consistent responses across participants.

The paper proposes a structural analysis of English it-clefts which accounts for different agreement patterns with clefted personal pronouns corresponding to the subject gap in the cleft clause. The agreement patterns under investigation are patterns already reported in the literature (Akmajian 1970, Sornicola 1988) as well as those that were found in a questionnaire filled in by a group of native speakers in 2012. An attempt will be made to account for the agreement variations that appear in it-clefts within the theoretical framework of the Minimalist Program (e.g., Chomsky 2008) incorporating Kratzer's (2009) view on bound pronouns. Specifically, possessive pronouns will be argued to enter the derivation with valued φ-features. Reflexives, on the other hand, will require feature valuation. Case variations on the clefted pronoun will be accounted for with the three observations originally made by Quinn (2005).

Chilean Spanish speakers were audio and video recorded engaging in conversation to find out whether clicks in Chilean Spanish occur in sequences similar to the ones reported for English. Some other research questions are whether there are any tokens that function alongside clicks in which case it is relevant to see what their articulatory properties and interactional functions are. Finally, if there are gestures accompanying clicks, it is important to devise their functions. A methodological approach that combines the sequential techniques of Conversation Analysis (CA), and the phonetic techniques of impressionistic observation and instrumental analysis is employed with naturally-occurring conversation. The results show that clicks do have a regular distribution and are part of bigger meaning-bearing prosodic constructions in Chilean Spanish, which entails that they are indeed linguistic. Moreover, the similarities with English in the way a click helps to show "gearing up to speak", to index new sequences that are disjunctive with the prior, to signal trouble in finding words; and the differences such as the particular use of clicks used to display affect found for Chilean Spanish support this argument.

This study investigates phonetic aspects in the production and perception of Smiling Voice, i.e. speech accompanied by smiling. A new corpus of spontaneous conversations is recorded to compare the formant frequencies of Smiling Voice (SV) and Non-Smiling Voice (NSV); the hypothesis that smiling raises formant frequencies is proven to be valid also on spontaneous speech, after previous research found this hypothesis to be true for scripted speech. Then, two perception experiments are carried out to test the hypothesis that listeners can recognize instances of SV extracted from the corpus of spontaneous data and instances of NSV obtained from read speech. Once the hypothesis is confirmed, a second perception experiment is performed to attempt to locate the point, in an artificial continuum from SV to NSV, where such perception happens.

face and goat realisations from six female Bradford Panjabi-English speakers were observed to follow the characteristic pattern of other Asian- (and multicultural-) Englishes with significantly lower F1s than three Bradford Anglo-English females. A qualitative distinction with the closer kit and foot vowels is maintained, with younger PE speakers showing a greater degree of separation (more open face and goat). Both groups demonstrate variability in the degree of separation between the two vowel pairs (face and kit; goat and foot). The current paper considers whether closer face and goat realisations provide evidence for transfer or innovation within Panjabi-English. Closer realisations are considered to index a non-Anglo ethnic identity, as suggested in other studies into Multicultural-Englishes, with the relationship to kit and foot working in a complex way to index a further local identity.