Abstract

English alveolar obstruents /t d s/ are wildly unstable in that their surface forms are rarely [t d s]. This paper focuses on the respective palatoalveolar set of surface forms [tʃ dʒ ʃ]. While many speakers produce these allophones in onsets before /ɹ/ (1) (Smith, et al., 2019), a smaller set of North American speakers, most notably near New York City and Montréal, also produce them before [w] (2) (Nelson & Flynn, 2021).

1) a. betray /bitɹeɪ/ → [bɪ.ˈtʃɹeɪ]

1. drone /dɹoʊn/ → [dʒɹoʊn]
2. sri* /ʃri/ → [ʃri] (*/sr/ is not permitted in English)

2) a. between /bitwin/ → [bɪ.ˈtʃwin]

1. Dwayne /dweɪn/ → [dʒweɪn]
2. swing /swiŋ/ → [ʃwiŋ]

This paper analyzes this phonological process as a case of reanalysis of enhancement cues and phonological features (Keyser & Stevens, 2006). In typical English pronunciation, lip rounding enhances [-anterior] consonants including /tʃ dʒ ʃ/. I therefore claim that these speakers have reinterpreted the phonetic rounding cue spread by rhotic and labio-velar approximants as indicating the phonological feature [-anterior] in preceding obstruents.

Thus, when these speakers produce a sequence of [+anterior] /t d s/ before rounded approximants /w ɹ/, they apply a phonological change (Bach & Harms, 1972) of [+anterior] to [-anterior], so that the rounding enhancement cue matches the surface representation. In English, this additionally forces the features [+strident] and [+distributed], as all English posterior obstruents carry these features, resulting in surface allophones [tʃ dʒ ʃ].

Abstract

Algonquian languages feature a typologically rare phenomenon known as OBVIATION, in which one DP is marked as PROXIMATE (with suffix -wa) and all other DPs are marked as OBVIATIVE (with suffix -yi). This contrast is frequently characterized as a means of tracking reference (e.g. Frantz, 2009; Goddard, 1984), empathy (e.g. Dahlstrom, 1991), point of view (e.g. Mühlbauer, 2008), and/or topicality (e.g. Genee, 2009). Since most languages only mark third person DPs for obviation, it has largely been described purely as a way of differentiating third persons. However, Blackfoot is unique in marking local (i.e. first and second person) pronouns for obviation, which raises the following question: if obviation truly tracks reference, should we not expect it to track an entity regardless of the person specification of the realizing DP?

Employing both fieldwork and a corpus study, the present study found that local pronouns in Blackfoot denoting SUBJECTS received PROXIMATE morphology while those denoting NON- SUBJECTS received OBVIATIVE morphology. Crucially, in cases of narrative direct speech where a speaker referred to themselves or their addressee, local pronouns often displayed different obviation markings than their co-referring third person DPs. This suggests that obviation marking on local pronouns tracks subjecthood rather than reference.

Abstract

Yeísmo is the merger of the sounds /ʎ/ (as in Italian “aglio”, less closely English “will you”, spelt <ll> in Spanish) and /ʝ/ (similar to “yell”, spelt <y>) in the Spanish language. Yeísmo and the distinction between the two sounds have been considered important regional markers by dialectologists when dividing Spanish into geographical variants. While yeísmo is now dominant in Spanish worldwide, there are regions of the Spanish-speaking world which conserve the distinction between /ʎ/ and /ʝ/. In Peru, the distinction between /ʎ/ and /ʝ/ is now fluctuating toward yeísmo.

This research consists of two components: A literature search that serves as background, and a corpus study focused on the department of Arequipa in southern Peru. The goal of the corpus study is to discover the current state of the ʎ/ʝ distinction in the region of Arequipa from a geographical perspective by sampling speech from speakers on videos found on the Internet that are associated to each province of the region, counting how many times speakers produce [ʎ], and producing average measurements of yeísmo by province.

Abstract

Stevens and Keyser’s Enhancement Theory (1989) holds that phonemic contrasts are communicated through primary features that may be enhanced through additional gestures. A classic example is that of [s] and [ʃ] in English, which are distinguished by the feature [anterior]. However, the [ʃ] in English is also typically produced with lip-rounding, which Stevens and Keyser (2006) argue makes the two sounds more acoustically distinct. Stevens and Keyser cite consonant confusion data from Miller and Nicely (1955), which shows that [s] and [ʃ] are unlikely to be confused; however, it is unknown how distinct they are without the secondary lip-rounding gesture on [ʃ]. It is also unknown how much lip-rounding could further enhance other fricative distinctions, such as between [f] and [θ], which are consistently one of the most difficult English consonant pairs to differentiate.

We will present initial plans for an audio-visual perception experiment to test the confusion of rounded and unrounded [f], [θ], [s], and [ʃ], and to quantify how much the addition of lip-rounding to these consonants may aid in disambiguation. The data should provide a clearer test of Enhancement Theory, and a novel examination of how it applies to other fricative contrasts in English.

Abstract

Persian has a distinctive subject-verb agreement system. The only slot which no agreement attaches to is PAST 3SG verbs. In the Tehrani dialect of Persian, though, a new marker ‘-esh’ homophonous to 3SG object clitic is optionally added to the end of PAST 3SG verbs as in (1). 

(1)    ye mard-i                         umad-(esh)

       one man-NONSPEC       came-(esh)

       ‘A man came.’

Some claim that it is a new agreement marker borrowed from the Persian object clitic paradigm to fill the empty slot in the agreement paradigm (Rasekh 2011, 2017). However, it still demonstrates some clitic-like properties like being optional on the verb and blocking other Persian clitics to follow unlike other agreement markers (Mahootian & Gebhardt 2019). I collected the judgement of 42 native speakers on items including ‘-esh’. The results reveal that although ‘-esh’ is still an object clitic and unable to select non-specific objects, it is in the stage of being reanalyzed as an agreement marker to fill the empty slot in the Persian agreement paradigm in that it can select subject arguments regardless of their specificity as in (1). 

Abstract

In Mandarin Chinese, the term ‘sentence-final particle’ has been used to describe a group of particles which always locate in the sentence-final positions and whose categorial status is unclear (e.g. their function depend on the specific extralinguistic context). Three examples containing sentence-final particles are given in (1a-c).

(1)

1. zhè  dōngxi    sānbaǐ                 yuán maǐ bù    laí      ne 

    this   stuff       three-hundred CL    buy NEG come particle

”This stuff cannot be bought with three hundred Yuan (believe me).”

                                                                       (Lu 1990, p.264)

1. Tā      zìjǐ      bù      yào     me.

      he      self      NEG  need    particle

”He does not need (one) (you should know this).”

                                                                        (Lu 1990, p.270)

1. Nǐ      juéde    zhème    gàn  duì     ha? 

    you    think   like.this   do    right  particle

”You think it is right to do this, eh?”

                                                                       (Yin 1999, p.99)

Mandarin sentence-final particles have been analyzed uniformly as sentence-final complementizers by a group of researchers (Paul and Pan 2017; Pan 2019). However, in the present paper, I draw evidence from co-occurring sentence-final particles to demonstrate that in Mandarin, sentence-final particles must co-occur in a fixed order. This observation casts doubts on the assumption that these particles are complementizers because treating particles as complementisers does not explain why particles must appear in a fixed order.