Phonology Circle will meet on Mondays in 32-D831 from 5-6pm, unless otherwise noted. The Phonology Circle is a weekly forum for the presentation of current research in phonology and phonetics. If you want to receive the email announcements, or have any other comments about Phonology Circle, please email Michael Kenstowicz.
Edward Flemming
The grammar of coarticulation
It is common to appeal to phonetic phenomena such as coarticulation to explain generalizations about phonological typology. For example Hayes (1999) invokes coarticulatory velum lowering in nasal-obstruent sequences as part of an explanation for the prevalence of post-nasal voicing of obstruents. Ohala (1993, 1994) hypothesizes that coarticulation is the ultimate diachronic source of many types of assimilation including vowel harmony, and properties of these processes consequently derive from properties of coarticulation.
These two examples represent distinct modes of explanation in relating phonetics to phonology: Ohala argues for a process of phonologization in which phonetic phenomena can be reinterpreted by listeners as phonological (Hyman 1976, Ohala 1992), while Hayes argues that phonetic considerations of effort and perceptual distinctiveness provide the grounding for phonological constraints. In spite of the differences between these lines of explanation, both take properties of coarticulation (and other aspects of phonetics) as given and then use them to explain phonological generalizations. This is directly apparent in the case of analyses based on phonologization but also applies to Hayes’s Grounding theory because evaluating effort and distinctiveness for standard phonological representations requires assumptions about patterns of coarticulation in the realization of those representations.
Taking coarticulation as given is problematic because coarticulatory patterns are language-specific and just as much in need of explanation as the phonological patterns they are supposed to explain. Moreover, when we try to fill this explanatory gap by providing an analysis of coarticulation, we find that coarticulation is part of grammar, and that the analysis of the cross-linguistic typology of coarticulation is similar to the analysis of phonological typology – specifically we find evidence that coarticulatory patterns are shaped by variation in the weighting of conflicting, universal constraints (cf. Optimality Theory). So explanations of phonological patterns in terms of coarticulation are not explanations in terms of extra-grammatical phenomena, they are attempts to explain properties of one part of grammar (phonology) in terms of properties of another (phonetics). Given the similarities between the phonetic and phonological components, it is more productive to look for the fundamental constraints that explain both rather than trying to explain phonology in terms of phonetics (or vice versa).
These arguments will be developed in the context of two case studies of cross-linguistic variation in coarticulation: (i) tonal coarticulation and assimilation and (ii) coarticulatory fronting of vowels by coronals.
Hayes, Bruce (1999). Phonetically driven phonology: the role of Optimality Theory and inductive grounding. In Darnell, Moravcsik, Newmeyer, Noonan, and Wheatley (eds.) (1999). 243-286.
Hyman, Larry M. (1976). Phonologization. Alphonse Juilland (ed.), Linguistic Studies Offered to Joseph Greenberg: Second Volume: Phonology. Studia linguistica et philological 4. Saratoga, California: Anma Libri.
Ohala, J.J. (1992). ‘What’s cognitive, what’s not, in sound change’ In: Kellermann, G. & Morrissey, M.D. (eds.) Diachrony within synchrony: language history and cognition. Frankfurt/M: Peter Lang Verlag. 309-355.
Ohala, J. J. 1993. Coarticulation and Phonology. Language & Speech 36, 155-170.
Ohala, J. J. 1994. Towards a universal, phonetically-based, theory of vowel harmony, ICSLP 3, Yokohama, 491-494.
David Hill
Matching minimalities: quantitative correspondence in Ancient Greek textsetting
This preliminary talk has three main ingredients: a new empirical finding, an observation, and a simple analytical concept. The finding is that in Ancient Greek vocal music, the mapping between syllable rime type and musical quantity (the number of grid positions occupied), already known to be tight, is too fine-grained to be captured by a binary L/H weight contrast, or even by a skeletal classification of rime structure (V, VC, VV, VVC), since not all VC rimes behave alike.
The observation, surprising at first, is that the meter of Greek song is demonstrably quantity-insensitive. Its currency is an abstract prominence alternation, which does not map directly to syllable weight. Quantity sensitivity emerges from the way that text, meter and time grid inter-correspond. The existence of a class of songs defined by a tempo specification--half time--from which L syllables are categorically barred, but which are nevertheless completely normal metrically and in text-to-time grid alignment, confirms that Greek song meters do not care about the weight of the rimes they align with. The fine-grained mapping mentioned above is therefore primarily a text-to-time grid phenomenon.
I use these phenomena to explore the viability of a notion of "(non-)minimality correspondence." The rough idea is that there is a cross-modal notion of minimality defined in some domains, and that objects in those domains that correspond with each other are required to both be minimal or both be non-minimal according to the appropriate definition. I apply this notion to quantity in textsetting, using a cardinal definition of minimality for meter and the time grid and structural definitions of minimality in rime.
Nabila Louriz
He Loves Me… He Loves Me Not: Irregularities nasal vowel adaptation in Moroccan Arabic
The aim of this talk is to analyse the repair strategies for nasal vowels in French loanwords into Moroccan Arabic.
The claim in the literature has been that in languages that lack phonemic nasal vowels, the latter is repaired as a sequence of oral vowel + nasal consonant (VN, henceforth). That is the nasal vowel undergoes the process of “unpacking” and is adapted as VN (Paradis & Lacharite 1996, Paradis and Prunet 2005, Rose 1999, to name but a few).
Paradis and Lacharite (1996) introduce evidence from French loans in Fula, Kinyarwanda, and Moroccan Arabic to show that the nasal vowel is “universally” adapted as a sequence of VN as long as the relevant requirements of the Theory of Constraints and Repair Strategies (1986) (TCRS, hereafter) are met, namely the Threshold Principle. Subsequently, Rose (1999) introduced a structural account of Root node deletion/preservation to explain nasal adaptation, presenting evidence from Fula and Kinyarwanda. He basically claims that the nasal part of the nasal vowel (which is the result of unpacking) is preserved when there is an available licenser, and deleted only when there is none.
I will bring these two approaches together to account for the adaptation of nasal vowels in French loanwords in Moroccan Arabic. The latter does not seem to adopt one single strategy to “fix” the ill formed segments. Consider the following examples:
(1)
MA |
French |
Gloss |
gufel |
gonfler |
swell |
klakson |
klaxon |
horn |
kwansa |
coincer |
to block |
zbiktur |
inspecteur |
inspector |
kofra |
Coup-franc |
out-of-bounds |
fran |
frein |
brake |
It seems that there is no uniform repair for adapting nasal vowels in Moroccan Arabic. It is repaired is as (i) VN, (ii) V, (iii) or deleted altogether. Maybe this is what drives some researchers to blame it on “complicating analogical factors” (Paradis & Prunet, 2005), or morphological factors (Heath, 1989). I shall present an analysis that can account for the different strategies manifested in the examples above. Namely, one that incorporates both phonology and phonetics. I will discuss how phonology and phonetics interact in accounting for the asymmetry manifested in the loanwords data.
Gillian Gallagher
Identity and laryngeal phonotactics
In this talk, I look at phonotactic restrictions on the cooccurrence of laryngeal features (aspiration, ejection and implosion). Many languages disallow roots or words with two distinct consonants with the same laryngeal feature, *k'-t'. Some languages with this restriction also disallow identical consonants with the same laryngeal feature *k'-k', while other languages allow identical consonants, k'-k'. I show that the (un)grammaticality of identical consonants sharing a laryngeal feature (k'-k') correlates with the (un)grammaticality of consonants differing only in that laryngeal feature (k'-k). In all the languages in MacEachern's (1999) survey, one of these forms is ungrammatical and one grammatical. The trading relationship is shown in (1).
(1) k'-k' <--> *k'-k
*k'-k' <--> k'-k
I argue that the pattern in (1) results from the interaction of phonotactic constraints with *two* kinds of laryngeal faithfulness constraints: faithfulness to individual features (standard Ident[F] constraints) and faithfulness to word level laryngeal contrasts (this is a new idea)
Peter Graff
Evidence from Artificial Grammar
In this talk I present the results of a 4-week-long artificial language study which was set up to investigate the deeper motivations of sound change. In the experiment 18 native speakers of English were asked to learn an Artificial language with an obstruent system exhibiting a 3-way VOT contrast {ph, p, b, th, t, d, kh, k, g}. Subjects were recorded weekly over a period of 3 weeks. After the third week, subjects were assigned to three groups, which were each taught a different “Dialect” of the artificial language; in the “Northern Dialect” voiced stops spirantized, in the “Southern Dialect” voiced stops nasalized, the third dialect acted as a control. After a week of training, subjects were recorded speaking their respective Dialects. Based measurements of closure duration and VOT of voiceless aspirated and voiceless unaspirated stops before and after treatment, I conclude that:
I will provide an analysis of the different types of systemic adaptation utilizing conjoined MINDIST constraints in the spirit of Flemming (1995) and conclude that the complex computation of contrast warrants the postulation of a system-wide Dispersion requirement, which I will formalize as a Sysdist constraint. This approach is in line with more recent proposals of weighted cumulative markedness metrics (e.g. Coon and Gallagher, 2007).
Joan Mascarò
A prosodic analysis of stress-dependent harmony
In some harmonic systems the trigger or the target must be a stressed vowel. These systems have been analyzed as limited by a prosodic domain, the stress foot, but more recently as long-distance assimilation of the stressed vowel to an unstressed vowel (Walker 2005, 2006), grounded on the need of "weak trigger" positions to realize their feature content on prominent positions. I will examine the evidence presented in favor of a weak trigger analysis and discuss additional evidence that suggests that a prosodic account should be preferred.
Joan Mascarò
A prosodic analysis of stress-dependent harmony
In some harmonic systems the trigger or the target must be a stressed vowel. These systems have been analyzed as limited by a prosodic domain, the stress foot, but more recently as long-distance assimilation of the stressed vowel to an unstressed vowel (Walker 2005, 2006), grounded on the need of "weak trigger" positions to realize their feature content on prominent positions. I will examine the evidence presented in favor of a weak trigger analysis and discuss additional evidence that suggests that a prosodic account should be preferred.
Franz Cozier
Encoding perceived contrast between CC-clusters and simplified counterparts in coda CC simplification
This paper examines grammatical constraints on word-final consonant cluster inventories (VC1C2#). Crosslinguistically, languages such as Trinidad dialectal English (TE), African American English, Cameroon English, Quebec French, and Catalan show striking consistency in the set of clusters that are illicit word-finally as shown in (1) (cf. Côté 2004, Green 1992, Bobda 1994, Mascaro 1976). Languages that ban these clusters do not release their final stops (cf. Archambault & Dumochel 1993). This makes it seem likely that simplification is related to the perceptibility of C2 in the absence of release. The central claims of this paper are (1) that C2 deletion is triggered when the distinctiveness of VC1C2 and VC1, as a function of phonetic cues, falls below a particular threshold and (2) that speakers encode this perceptually based difference between simplified and preserved clusters in their grammars. Experimental results will show how the synchronic grammar of TE reflects the historical simplification process. A second experiment will confirm that perceptibility is not just something that causes loss of C2 over time but that the grammar attributes simplification to perceptual difficulty raised by unreleased C2’s
Adam Albright
Chaotic evolution in an unbiased learner
A premise of channel-base) explanations of typology is that isolated misproductions or miscategorizations may cause the signal to deviate from the speaker’s original intent in a way that may be misinterpreted as a phonetically natural change. For example, /np/ may be perceived as [mp] due to articulatory overlap and the difficulty of distinguishing coarticulated [np] from [mp]. Over time, deviations are assumed to create patterns corresponding to cross-linguistically common processes, which may then be learned and reinforced even by unbiased learners. Numerous studies have investigated whether human infants or adults behave like unbiased learners, while less attention has been paid to a prior question: are series of misperceptions actually sufficient to create the patterns observed typologically?
In this talk, I report a series of simulations designed to address this question. An unbiased inductive learner was used to investigate what patterns might arise in languages partway through a phonetically motivated change. I consider languages with a typologically dispreferred contrast such as [np] vs. [mp], with [np] words occasionally reanalyzed as [mp]. I explored the properties of hypothetical languages at a stage with a 3:1 preference for [mp] by generating 1,000 artificial lexicons, each containing 50 words with nasal+[p] clusters. Lexical items were randomly constructed to obey basic syllable constraints, with a skew towards shorter (di- or trisyllabic) words. In all of these languages, there is a 75% tendency for labials before [p] (nasal place assimilation). The question of interest is whether there are even stronger statistical patterns, due to coincidences elsewhere in the word. To test this, I submitted all 1,000 languages to an inductive model of phonological constraint discovery, which compares words that share a particular property (such as [n] or [m]) to determine the best predictors in the surrounding phonological context (such as a following [p]). It emerged that in 678/1000 languages, the algorithm found specific contexts that were more reliable predictors of nasal place. If taken seriously and extended productively to derived contexts, this constraint could lead to highly unnatural alternations. Thus, it appears that rather than leading to neutralization, phonetically natural changes may be derailed, creating unnatural statistical correlations that may be picked up and extended by an unbiased learner. I consider several biases which would prevent the learner from being lead astray by such patterns, and provide a more accurate account of the attested typology.
WCCFL Practice:
Claire Halpert
Overlap-Driven Consequences of Zulu Nasal Place Assimilation
I examine the behavior of two noun class prefixes in Zulu, um- and iN. Nasals in iN- prefixes must become homorganic to a following C; they delete when there is no adjacent C. The nasal in um- prefixes surface in all contexts, with invariant features. I assume that the former nasals are underlyingly placeless and can only surface when place-assimilated. All Zulu homorganic NC clusters where the nasal has undergone place assimilation display secondary effects on C. I focus on three effects: de-aspiration of aspirated consonants, affrication of fricatives, and loss of implosion. The presence of these effects in assimilated NC clusters contrasts with their absence in non-place-assimilated mC clusters, suggesting that the secondary effects are directly linked to the process of place assimilation. I propose that nasal place assimilation in Zulu results in close transition between N and the following C, allowing place gestures of C to overlap the nasal, providing it with a place of articulation. A side effect of close transition, however, is that other of C’s gestures also overlap the nasal. This overlap becomes problematic for recovering nasality (cf. Silverman 1997, Browman & Goldstein 2000), particularly when it is overlapped by aspiration and glottalization. I argue that as a result of this close transition, some of these problematic features are lost from the entire cluster. In contrast, mC clusters remain in open transition, avoiding instances of problematic overlap.
Guillaume Thomas
An Analysis of the Xiamen Tone circle
The tone system of Xiamen presents systematic tone sandhi, organized in a circular fashion. This process is notorious for appearing to rely on a noncomputable function that is not analyzable in classical OT (Moreton: 1999). The analysis of circular shifts of this type faces two challenges: one is the identification of the constraints that motivate the circular move. The other challenge is completeness: the analysis of the Xiamen circle must evaluate 3,125 candidates, the number of permutations with repetitions of the five citation tones (assuming that the set of possible sandhi tones is identical to the set of citation tones).
I will a present a simple analysis of the phenomenon, that uses Contrast Preservation (Lubowicz: 2003) and a modified notion of Faithfulness, in a grammar where each candidate is a scenario that represents a mapping between each of the five citation tones of Xiamen, and its corresponding sandhi tone. Using a simple Perl script (written with the indispensable help of Adam) to automate the OT analysis, I will demonstrate that my analysis picks out the attested winner among the 3,125 candidates.
