Yearly Archives: 2018

Pons-Moll (2018) – The limits of the free ride in morphophonemic learning. Evidence from Catalan

The limits of the free ride in morphophonemic learning. Evidence from Catalan
Clàudia Pons-Moll
direct link: http://ling.auf.net/lingbuzz/004357
December 2018
The free ride strategy (McCarthy 2005) is challenged when the input-output mapping(s) derived from morphophonemic dynamic alternations and which are potentially generalized to non-alternating items are not univocal, that is, when the alternating [B]s derive from more than one underlying representation. When restrictiveness (M >> F) leads to a non-consummated free ride and it is no longer exploitable, efficiency (F >> F, etc.) (Bermúdez-Otero 2003, 2006) comes into play to consummate, conclusively, the free ride.

Format: [ pdf ]
Reference: lingbuzz/004357
(please use that when you cite this article)
Published in: Handout of the poster presented at Going Romance 2018, Utrecht University
keywords: underlying representations, morphophonemic learning, free-ride, richness of the base, lexicon optimization, vowel reduction, non-alternating items, phonology
Facebooktwitterredditpinterestlinkedinmail

Pons-Moll & Torres-Tamarit (2018) – Catalan nativization patterns in the light of Weighted Scalar Constraints

Catalan nativization patterns in the light of Weighted Scalar Constraints
Clàudia Pons-Moll, Francesc Torres-Tamarit
direct link: http://ling.auf.net/lingbuzz/004354
December 2018
In this paper we explore phonological nativization patterns in Catalan loanwords, and we show, on the basis of a production and a judgment test, that the three processes under scrutiny (word-final –n deletion [ND], vowel reduction of unstressed mid-vowels [VR], and vowel laxing of stressed mid-vowels [VL]) interact in an asymmetrical way. We argue that these asymmetrical interactions can be straightforwardly formalized resorting to Harmonic Grammar with Scalar Weighted Constraints (Hsu & Jesney 2017, 2018), in which faithfulness constraints acquire an increasing relevance from the core to the peripheral strata and in which the scaling factor intervals for each stratum abstractly reproduce the most frequent and the least frequent patterns in loanword adaptation.

Format: [ pdf ]
Reference: lingbuzz/004354
(please use that when you cite this article)
Published in: Handout delivered at Going Romance 2018
keywords: loanword phonology, catalan, possible nativizations, impossible nativizations, harmonic grammar, scalar weighted constraints, phonology
Facebooktwitterredditpinterestlinkedinmail

Tilsen (2018) – Space and time in models of speech rhythm

Space and time in models of speech rhythm
Sam Tilsen
direct link: http://ling.auf.net/lingbuzz/004349
December 2018
How do rhythmic patterns in speech arise? Many representations and models incorporate a mechanism whose purpose is to generate a rhythmic pattern. Here an alternative is explored: rhythmic patterns arise indirectly, from spatial mechanisms which govern the organization of articulatory gestures. In pursuing this alternative, the roles of time and space in symbolic phonological representations are analyzed in detail, and conventional understandings of stress and accent are called into question. One aspect of rhythmic patterns in particular—the directionality of stress assignment—is examined closely. A novel dynamical model is developed, which proposes a reinterpretation of directionality and various other temporal phenomena.

Format: [ pdf ]
Reference: lingbuzz/004349
(please use that when you cite this article)
Published in: Cornell Working Papers in Phonetics and Phonology 2018
keywords: speech rhythm, metrical theory, rhythmic typology, prosody, dynamical models, phonology
Facebooktwitterredditpinterestlinkedinmail

Akita & Dingemanse (2018) – Ideophones (Mimetics, Expressives)

Ideophones (Mimetics, Expressives)
Kimi Akita, Mark Dingemanse
direct link: http://ling.auf.net/lingbuzz/004347
December 2018
Ideophones, also termed “mimetics” or “expressives,” are marked words that depict sensory imagery. They are found in many of the world’s languages, and sizable lexical classes of ideophones are particularly well-documented in languages of Asia, Africa, and the Americas. Ideophones are not limited to onomatopoeia like meow and smack, but cover a wide range of sensory domains, such as manner of motion (e.g., plisti plasta ‘splish-splash’ in Basque), texture (e.g., tsaklii ‘rough’ in Ewe), and psychological states (e.g., wakuwaku ‘excited’ in Japanese). Across languages, ideophones stand out as marked words due to special phonotactics, expressive morphology including certain types of reduplication, and relative syntactic independence, in addition to production features like prosodic foregrounding and common co-occurrence with iconic gestures. Three intertwined issues have been repeatedly debated in the century-long literature on ideophones. 1) Definition: Isolated descriptive traditions and cross-linguistic variation have sometimes obscured a typologically unified view of ideophones, but recent advances show the promise of a prototype definition of ideophones as conventionalised depictions in speech, with room for language-specific nuances. 2) Integration: The variable integration of ideophones across linguistic levels reveals an interaction between expressiveness and grammatical integration, and has important implications for how to conceive of dependencies between linguistic systems. 3) Iconicity: Ideophones form a natural laboratory for the study of iconic form-meaning associations in natural languages, and converging evidence from corpus and experimental studies suggests important developmental, evolutionary, and communicative advantages of ideophones.

Format: [ pdf ]
Reference: lingbuzz/004347
(please use that when you cite this article)
Published in: Oxford Research Encyclopedia of Linguistics
keywords: ideophones, expressives, mimetics, typology, iconicity, syntax, phonology, semantics, morphology
Facebooktwitterredditpinterestlinkedinmail

Zymet (2018) – Lexical propensities in phonology: corpus and experimental evidence, grammar, and learning

Lexical propensities in phonology: corpus and experimental evidence, grammar, and learning
Jesse Zymet
direct link: http://ling.auf.net/lingbuzz/004346
December 2018
Traditional theories of phonological variation propose that morphemes be encoded with descriptors such as [+/- Rule X], to capture which of them participate in a variable process. More recent theories predict that morphemes can have LEXICAL PROPENSITIES: idiosyncratic, gradient rates at which they participate in a process—e.g., [0.7 Rule X]. This dissertation argues that such propensities exist, and that a binary distinction is not rich enough to characterize participation in variable processes. Corpus investigations into Slovenian palatalization and French liaison reveal that individual morphemes pattern across an entire propensity spectrum, and that encoding individual morphemes with gradient status improves model performance. Furthermore, an experimental investigation into French speakers’ intuitions suggests that they internalize word-specific propensities to undergo liaison. The dissertation turns to modeling language learners’ ability to acquire the idiosyncratic behavior of individual attested morphemes while frequency matching to statistical generalizations across the lexicon. A recent model based in Maximum Entropy Harmonic Grammar (MaxEnt) makes use of general constraints that putatively capture statistical generalizations across the lexicon, as well as lexical constraints governing the behavior of individual words. A series of learning simulations reveals that the approach fails to learn statistical generalizations across the lexicon: lexical constraints are so powerful that the learner comes to acquire the behavior of each attested form using only these constraints, at which point the general constraint is rendered ineffective. A GENERALITY BIAS is therefore attributed to learners, whereby they privilege general constraints over lexical ones. It is argued that MaxEnt fails to represent this property in its current formulation, and that it be replaced with the hierarchical MIXED-EFFECTS LOGISTIC REGRESSION MODEL (MIXED-EFFECTS MAXENT), which is shown to succeed in learning both a frequency-matching grammar and lexical propensities, by encoding general constraints as fixed effects and lexical constraints as a random effect. The learner treats the grammar and lexicon differently, in that vocabulary effects are subordinated to broad, grammatical effects in the learning process. For further developments, see my webpage: http://linguistics.berkeley.edu/~jzymet/

Format: [ pdf ]
Reference: lingbuzz/004346
(please use that when you cite this article)
Published in: UCLA Dissertation
keywords: variation, lexical propensities, maximum entropy harmonic grammar, hierarchical models, mixed-effects logistic regression, french, slovenian, morphology, phonology
Facebooktwitterredditpinterestlinkedinmail

Hussain (2018) — A typological study of Voice Onset Time (VOT) in Indo-Iranian languages

Hussain (2018) — A typological study of Voice Onset Time (VOT) in Indo-Iranian languages

Available online: https://www.sciencedirect.com/science/article/pii/S0095447017301936

Author: Qandeel Hussain

Abstract:

The stop consonants of Indo-Iranian languages are categorized into two to maximum five laryngeal categories. The present study investigates whether Voice Onset Time (VOT) reliably differentiates the word-initial stop laryngeal categories and how it covaries with different places of articulation in ten languages (two Iranian: Pashto and Wakhi; seven Indo-Aryan: Dawoodi, Punjabi, Shina, Jangli, Urdu, Sindhi, and Siraiki; and one Isolate: Burushaski). The results indicated that there was a clear VOT distinction between the voiceless unaspirated and voiceless aspirated stops. The voiceless unaspirated stops showed shorter voicing lag VOTs than voiceless aspirated stops. Voiced unaspirated, voiced aspirated, and voiced implosive stops were characterized by voicing lead VOTs. In the voiceless unaspirated and aspirated categories, palatal affricates showed the longest voicing lag VOT due to the frication interval of this stop type. In contrast, voiceless unaspirated retroflex stops were characterized by the shortest voicing lag VOT. There were no clear place differences in the voiceless aspirated, voiced unaspirated, voiced aspirated, and voiced implosive categories. The findings of the current study suggest that VOT reliably differentiates the stop consonants of all the languages that contrast two (voiceless unaspirated vs. voiced unaspirated: Pashto and Wakhi) or three (voiceless unaspirated vs. voiceless aspirated vs. voiced unaspirated: Burushaski, Dawoodi, Punjabi, and Shina) laryngeal categories. However, VOT does not consistently distinguish the stop consonants of languages (Jangli, Urdu, Sindhi, and Siraiki) with contrastive voiced unaspirated, voiced aspirated, and voiced implosive categories.

 

Facebooktwitterredditpinterestlinkedinmail

Lecturership (Assistant Professor) positions at Queen Mary U. of London

From: Adam Chong (a.chong@qmul.ac.uk)
Queen Mary University of London is currently inviting applications for lectureships in the Faculty of Humanities and Social Sciences. These strategic appointments are permanent positions and will include protected research time across the first three years. Within Linguistics, the university particularly welcomes strong applications in two areas: computational linguistics and first language acquisition (related to any sub-field, including potentially phonology). Full details can be found on LinguistList: https://linguistlist.org/issues/29/29-4770.html
Facebooktwitterredditpinterestlinkedinmail

1st Call for Papers for HISPhonCog 2019 (Seoul)

Hanyang International Symposium on Phonetics and Cognitive Sciences of Language 2019 (HISPhonCog 2019) from May 24-25, 2019 at Hanyang University, Seoul, Korea.

http://site.hanyang.ac.kr/web/hisphoncog/call-for-papers/2019

HIPCS (Hanyang Institute for Phonetics and Cognitive Sciences of Language), together with Department of English Language and Literature at Hanyang University, holds its second annual international symposium on current issues on phonetics and cognitive sciences of language—i.e., HisPhonCog 2019. The theme of HISPhonCog 2019 is “linguistic and cognitive functions of prosody and higher-order linguistic structures in speech production and perception in native and non-native languages.” Recent years have witnessed ample empirical evidence that one of the essential linguistic structural components that underlie the phonetic encoding-decoding process is prosodic structure. The fundamental assumption that underlies the theme is that the prosodic structure is an integral part of speech production serving as a frame for articulation which, for example, regulates groupings of phonological constituents and prominence distribution in conjunction with tonal/intonational specifications. This view of prosody entails further that a prosodic structure of an utterance is determined in reference to various other components of the linguistic structure of the language (e.g., phonology, morphology, syntax, information structure, and discourse structure) as well as extra- or para-linguistic factors (e.g., social indexical information and emotions). Thus, in order for the speaker to deliver a linguistic message to the listener successfully, the speaker must be able to encode the linguistic message in prosodic-structurally conditioned phonetic forms that reflect other linguistic and extra-linguistic structures, and the listener must in turn be able to exploit the resulting fine phonetic detail in decoding the intended linguistic message in reference to the prosodic structure that underlies the speech signal. This premise applies to speech production and perception in both L1 and L2.

We invite submissions for the symposium which explore any issues related to the theme of the symposium. There will be a special session on neuro-cognitive aspects of the role of prosody. We will also consider submissions that deal with other general issues in native and non-native speech production and perception.

Invited speakers for general sessions

Mary Beckman (OSU)

Edward Flemming (MIT)

Cecile Fougeron (Paris 3, Sorbonne, CNRS)

Martine Grice (University of Cologne)

Sun-Ah Jun (UCLA)

Jason Shaw (Yale University)

Michael Tyler (Western Sydney University)

Jie Zhang (University of Kansas)

Invited speakers for the special session on ‘neuro-cognitive aspects of prosody’

Karsten Steinhauer (McGill University)

Ferenc Honbolygó (Hungarian Academy of Sciences)

Planned satellite workshop (May 23)

  • Theoretical and practical issues on ToBI in Korean (organized by Sun-Ah Jun, Sahyang Kim and Taehong Cho).
  • Call for papers will be available on the conference webpage soon. Please check the conference webpage later.

http://site.hanyang.ac.kr/web/tcho/hisphoncog-2019

Support for international participants:

  • Free local hotel accommodation (3 nights) for international presenters affiliated with a foreign institute/university, travelling from abroad. One additional night may be provided, pending the budget availability, to those presenters who also attend a satellite workshop on May 23. (The detail will be sent to qualified individuals along with an acceptance letter.)
  • No registration fees(including 2 lunches, refreshments and one banquet)

Timeline

  • Deadline of submission of a two-page long abstract: March 10,2019
  • Notification of Acceptance: No later than March 31, 2019
  • Free Registration with free accommodation: No later than April 10, 2019
  • Symposium dates: May 24-25, 2019

Abstract Submission Instruction:

Free Registration by April 10

  • Pre-registration should be made by no later than April 10, 2019 to be guaranteed for free accommodation (for international presenters) and free registration (for all participants and audience).
  • Pre-registration form that arrives after April 10 may still be considered for free registration and accommodation, depending on the budget and availability. Please contact us at hanyang.hipcs@gmail.comif you miss the deadline but still would like to register in advance.
  • On-site registration will be possible for small fees, but with no guarantee for lunches and banquet admission.
  • For further information about how to register, please check the website.

Local Organizing Committee

Taehong Cho (Chair, Hanyang University, Seoul)

Sahyang Kim (Hongik University, Seoul)

Say Young Kim (Hanyang University, Seoul)

Jonny Jungyun Kim (Hanyang University, Seoul)

Contact: Jonny Jungyun Kim at hanyang.hipcs@gmail.com

Organized by HIPCS (the Hanyang Institute for Phonetics and Cognitive Sciences of Language); Department of English Language and Literature, Hanyang University, Seoul, Korea

URL: http://site.hanyang.ac.kr/web/hisphoncog/call-for-papers/2019

Facebooktwitterredditpinterestlinkedinmail

Kawahara (2018) – Teaching phonetics through sound symbolism

Teaching phonetics through sound symbolism
Shigeto Kawahara
direct link: http://ling.auf.net/lingbuzz/004342
December 2018
Teaching introductory phonetics classes can be challenging for several reasons. One reason is that students are introduced with many new concepts such as place of articulation, manner of articulation, and the obstruent/sonorant distinction. Remembering these classification terms can be overwhelming and/or boring. Another challenge is that while many students taking phonetics classes are humanity major students who often do not like mathematics, understanding phonetics does require basic background in mathematics and physics. In this paper, I summarize my pedagogical attempt to lower the psychological boundary of students against learning these concepts by making use of sound symbolism in introductory phonetics class. Analyses of sound symbolic patterns can be presented using materials that students are familiar with (e.g. Disney characters and Pokémon monsters), so that the students feel that classification terms that phoneticians use are “real”. Furthermore, since some sound symbolic patterns are demonstrably grounded in the articulatory and acoustic natures of particular sounds, we are able to teach some important articulatory and acoustic principles. Finally, statistical analyses of the Pokémon dataset, which emerged from this teaching strategy, help us illustrate some important statistical skills.

Format: [ pdf ]
Reference: lingbuzz/004342
(please use that when you cite this article)
Published in: Proceedings of ISAPh
keywords: pedagogy, sound symbolism, sonorants, voiced obstruents, labiality, pokémon(astics), statistics, phonology
Facebooktwitterredditpinterestlinkedinmail

Akkus (2018) – Variable embedded agent in Sason Arabic

Variable embedded agent in Sason Arabic
Faruk Akkus
direct link: http://ling.auf.net/lingbuzz/004341
December 2018
The paper investigates the syntax of an indirect causative construction, “make”-causatives in Sason Arabic (SA), with a focus on the syntactic status of the implicit embedded agent and the embedded structure. Differing from the analyses of other languages for similar constructions, the study demonstrates that this construction embeds both an active and passive VoiceP. It contends that the implicit embedded agent may be introduced (i) as a full DP in Spec,VoiceP, being subject to Romance ECM-type restrictions, and providing striking evidence of A’-movement feeding licensing relationships, or (ii) as a free variable à la Heim (1982) generated in the Voice head itself. The latter possibility also raises implications regarding licensing, suggesting that licensing of a grammatical object is dissociated from the projection of a specifier.

Format: [ pdf ]
Reference: lingbuzz/004341
(please use that when you cite this article)
Published in: submitted
keywords: causative, implicit arguments, licensing, a’-movement, sason arabic, semantics, syntax, phonology
Facebooktwitterredditpinterestlinkedinmail