|
Review:
|
EDITORS: Calabrese, Andrea and Wetzels, W. Leo
TITLE: Loan Phonology
SERIES: Current Issues in Linguistic Theory 307
PUBLISHER: John Benjamins
YEAR: 2009
Lionel Mathieu, Department of Linguistics, University of Arizona
SUMMARY
“Loan Phonology” is an edited volume exploring the intricacies of lexical
borrowing. Written by specialists in loanword research, this collection provides
an accurate overview of this ever-growing area. Readers will appreciate its wide
theoretical and empirical perspectives, and gain better understanding of this
complex phenomenon. The book contains eleven articles offering insights into the
general processes of borrowing. Each is reviewed in turn.
“Loan phonology: Issues and controversies” Andrea Calabrese and W. Leo Wetzels
This short introduction provides a comprehensive overview of loanword phonology,
beginning with brief comments on the significance of loanwords to phonological
theory, and wasting no time in diving into core issues. The two dominant models
of loanword phonology, the phonological and perceptual stance models, are
discussed and compared. Proponents of the phonological stance model (Hyman 1970;
Lacharité and Paradis 2005; inter alia) contend that adaptations are
phonological in nature, ascribing a minor role to perception. Proponents of the
perceptual stance model (Silverman 1992; Peperkamp and Dupoux 2003; inter alia)
pursue the contrasting view that adaptations are primarily carried out by the
perceptual system of the borrowing language. Each article is briefly summarized
and situated within the loanword debate as it relates to the nature and locus of
adaptations. According to the editors, the conclusion that emerges is one of
“synergetic interaction” (9) between loanword models.
“Loanword adaptation as first-language phonological perception” Paul Boersma and
Silke Hamann
This article presents an account of loanword adaptations argued to be entirely
carried out by the phonological grammar of the recipient (L1) language. Relying
on an existing model of L1 processing (Boersma 2007), the authors show that
loanword adaptations in Korean can be handled by a single constraint-based
grammar that doesn’t necessitate any loanword-specific mechanisms or
constraints. The L1 model is bidirectional (perception and production) and
involves three levels of representation (phonetic, surface and underlying)
mediated by faithfulness, structural (~ markedness) and cue constraints. Cue
constraints are formally designed to relate phonetic representations with
surface representations and evaluate fine-grained phonetic details present in
output candidates. Focusing on the contrast between the absence of vowel
insertion in native Korean words and the presence of vowel insertion in Korean
loanwords, Boersma and Hamann argue that the former can be explained by an
interaction of faithfulness and structural constraints in production while the
latter can be accounted for by an interaction of cue and structural constraints
in perception. Crucially, structural constraints are active in both the
perception and production parts of the grammar; their respective interactions
with cue and faithfulness constraints give rise to the observable pattern of
adaptation. Perception and production are therefore not understood as mutually
exclusive components of a grammatical system but are linked to one another via
structural constraints. Their conclusion then is that “perception simply IS
phonological” (53), ultimately proposing a consensus between the two dominant
views of loanword adaptations.
“Perception, production and acoustic inputs in loanword phonology” Andrea Calabrese
This lengthy article covers the role of speech perception in loanword phonology.
Calabrese delves into the complex mechanisms at play when monolinguals perceive
and consequently adapt unfamiliar sounds and sound arrangements. The author
espouses a view of speech perception involving two parallel and interconnected
components: a bottom-up system, that “discriminates new, unfamiliar sounds and
sound strings from familiar, previously heard ones” (62) and a top-down system,
which becomes subsequently “active in analyzing those new or unfamiliar
configurations” (62). Throughout, he provides ample and technical details of the
inner-workings of perception, speech perception and echoic memory, interspersed
with concrete examples. The main import of Calabrese’s multistage model is that
both non-loanword-specific L1 bottom-up and top-down processes work hand in hand
in the computation and adaptation of foreign sounds. This model therefore
circumvents some of the issues articulated in the literature, especially
concerning the locus of adaptations.
“The adaptation of Romanian loanwords from Turkish and French” Michael L. Friesner
Friesner investigates internal and external factors responsible for the
adaptation of words of Turkish and French origin in Romanian. Drawing on a set
of eighty-five loanwords, the author considers the effects of social and
linguistic aspects that inform their adaptation. he shows that the difference in
the nature of the contact situation between the two source languages and
Romanian is revealing, especially as it concerns semantic domains. Turkish
loanwords tend to pertain to “common-place objects [and] aspects of the
government and the military” (119), some even assuming a negative connotation,
while loanwords from French “tend to refer to aspects of high society” (119),
generally exuding more positive qualities. Friesner then goes on to address the
assignment of stress and gender desinence in loanwords. With regard to stress,
Turkish loanwords may or may not retain the stress of the source language: if
preserved, these loanwords follow an adaptation pattern analogous to a closed
class of native Romanian words with final stress and no gender desinence vowel;
if not preserved, these loanwords adopt a gender desinence vowel and comply with
the metrical structure of Romanian. French loanwords, on the other hand, retain
the original stress more regularly and are always adapted with a gender
desinence vowel. When it comes to gender assignment, French loanwords tend to
retain their original grammatical gender, while Turkish loanwords are assigned a
gender on the basis of their phonological form. In addition, word-final segments
may at times inform the way Turkish and French words are adapted in Romanian.
Hence, social, morphological and phonological factors join forces in shaping
Romanian loanwords of Turkish and French origin, leading the author to conclude
that loanword adaptations ought to be analyzed from a multitude of perspectives.
“Mandarin adaptations of coda nasals in English loanwords” Feng-fan Hsieh,
Micheal Kenstowicz and Xiaomin Mou
In this article, the authors rely on Mandarin Chinese (MC) adaptations of coda
nasals in English loanwords to tackle the phonology/phonetics debate of loanword
adaptations. In MC, a Rhyme Harmony constraint demands that, at an allophonic
level, the quality of non-high vowels agree with the place of articulation of
the following nasal coda. English loanwords possessing an incongruous VN
configuration present a conflict for the Rhyme Harmony constraint. From the
perspective of adaptation however, it represents a source of great insight given
that the two competing stances (phonological vs. perceptual) would predict
diverging outcomes. A phonological approach would favor faithfulness to the
nasal segment (because phonemically shared), while a perceptual approach would
favor faithfulness to the vowel (because phonetically more salient). This
verdict is this: in the majority of cases faithfulness to the vowel is preferred
-- the [front] vs. [back] quality of the English vowel is preserved whereas the
place of articulation of the MC nasal gets accordingly adjusted. English
non-salient schwas or wedges [ʌ] however do not determine the adaptation of the
MC coda. Assuming Richness of the Base, these facts are captured in
Correspondence Theory (McCarthy and Prince 1995), where output-output
faithfulness may outrank input-output faithfulness, allowing output candidates
to escape native phonological demands. Phonetic/perceptual considerations can
therefore play a role in determining native phonological contrasts.
“Korean adaptation of English affricates and fricatives in a feature-driven
model of loanword adaptation” Hyunsoon Kim
Drawing on previous research proposing a feature-driven model of loanword
adaptation (Kim 2007), this paper treats the Korean (L1) adaptation of English
(L2) affricate and fricative consonants. The adaptive treatment of English [s]
is first examined. Kim shows that [s] is adapted differently based on whether
the segment appears as a singleton or in a consonant cluster, arguing that the
reason for this lies in L1 perception where Korean speakers interpret an
acoustic durational difference of L2 [s] in terms of an L1 distinctive [±tense]
feature. In addition to the role of distinctive features in adaptation, L1
syllable structure is also shown to be involved. The adaptations of L2
prevocalic [ʃ] and postvocalic [ʃ, ʤ, ʧ] are then examined. Kim contends that
these adaptations reflect native syllabic requirements coupled with native
distinctive feature values. An analogous analytical perspective is taken to make
sense of the adaptation of English voicing contrast in affricates and
fricatives. The author then discusses the merits of his approach in light of the
current debate, showing that both approaches at times account as well as fail to
account for some of the patterns identified. Kim concludes that an intermediate
position is preferable, where loanword adaptations are carried out within both
the perceptual and phonological confines of the L1 grammar.
“The role of underlying representations in L2 Brazilian English” Andrew Nevins
and David Braun
This shorter article examines two intriguing phenomena in Brazilian Portuguese
English (BPE) and loanwords, namely “spurious affrication” and “rhotic
hypercorrection”. The relevance of these phenomena to models of loanword
phonology is noteworthy. In spurious affrication, an English sequence [tu] is
produced as [ʧu] in BPE and loanwords from English, even though affrication
before /u/ does not occur in Brazilian Portuguese (BP). Appealing to the
relative frontness of the English high back vowel (compared to that of BP), the
authors account for this BPE/loanword-specific occurrence by suggesting an L2
underlying representation containing a non-nuclear /i/. The resulting
representation of English [tu] in BPE and loanwords would be /tiu/, which would
triggering regular affrication. The other phenomenon, rhotic hypercorrection,
refers to the sporadic pronunciation of [h]-initial English words as [r]-initial
BPE words (e.g. ‘home’ [hom] -- [rom]). Nevins and Braun adopt the view that BPE
speakers posit an /r/-initial underlying representation for English [h]-initial
words. This underlying representation is generally subject to a native onset
debuccalization rule but can at times also be exempt, consequently revealing the
“true” underlying representation posited by BPE speakers. Together these two
puzzling phenomenon unveil an adaptation strategy rooted in perception yet
informed by phonological stipulations of the borrowing language.
“Early bilingualism as a source of morphological rules for the adaptation of
loanwords: Spanish loanwords in Basque” Miren Lourdes Oñederra
Cast in the framework of Natural Phonology (Stampe 1979), this paper explores
the adaptation of Spanish loanwords in Basque. This study of the substitutions /
retentions of phonemes and phoneme sequences uses the formal concept of
‘phonological process’, “a mental substitution that responds to a phonetic, i.e.
physical, difficulty related to the articulation or perception of segments and
sequences” (197). Phonological processes are understood to be universal, but in
the course of acquisition can either be ‘allowed’ (i.e. the process applies in
order to circumvent a phonetic difficulty) or ‘overcome’ (i.e. the phonetic
difficulty survives, is integrated, and eliminates the related process from
competence). Bilingual acquisition of phonological processes may subsequently
result in their presence in phonology X but absence in phonology Y (or vice
versa), a scenario believed to be the source for loanword adaptations. The
author then delineates and exemplifies “three possible patterns of process
distribution between the two languages: a) when Basque keeps a process that
Spanish does not allow, b) when Spanish keeps a process that Basque has
overcome, and c) when both languages keep a process” (199). Oñederra then turns
to the application of substitutions that are no longer phonetically motivated
and have consequently become morphological rules. The analysis concentrates on
the fact that individual and collective (early) bilingualism is responsible for
the observed adaptations in Spanish loanwords.
“Nondistinctive features in loanword adaptation: The unimportance of English
aspiration in Mandarin Chinese phoneme categorization” Carole Paradis and
Antoine Tremblay
Akin to Kenstowicz and Mou (this volume), Paradis and Tremblay probe the thorny
debate of sound adaptations, providing corpus-based evidence from Mandarin
Chinese (MC). The fact that stop aspiration in English is phonetic
(non-contrastive), but phonological (contrastive) in MC, serves to examine
whether non-contrastive features of the donor language are relevant in phoneme
categorization in the recipient language. English loanwords in MC therefore
provide a favorable testing ground for both models of adaptations: the
perceptual stance model would make positive predictions (i.e. non-contrastive
aspiration in English yields contrastive adaptation in MC), while the
phonological stance model would infer otherwise (i.e. non-contrastive aspiration
in English does not inform phoneme adaptations in MC). Drawing on a corpus of
500 stops contained in 371 borrowings, Paradis and Tremblay arrive at the
conclusion that “both aspirated and unaspirated voiceless stops of English
systematically yield an aspirated stop in MC, whereas English voiced stops,
which are disallowed in MC, systematically yield a voiceless unaspirated stop”
(211). These findings lead the authors to support a phonological view of
loanword adaptations in which phonetic characteristics do not inform the outcome
of adaptation. Not entirely dismissing perceptual influences in loanword
adaptations, the authors conclude by arguing that both models present convincing
arguments and may actually only be facets of the same system. An appeal to
careful “conceptualization, terminology and methodology” (222) is made, in
addition to increased sociolinguistic considerations in the treatment and
analysis of loanword adaptations.
“Gemination in English loans in American varieties of Italian” Lori Repetti
Analogous to Friesner’s approach (this volume), this paper provides further
evidence for the need to consider multiple factors in loanword adaptations.
Using Optimality Theory, Repetti identifies three grammatical aspects
intervening in the determination of consonant length (gemination) in Italian
loans: the native lexicon, morphology and phonetics/phonology. Repetti first
remarks that some adaptations are conditioned by the native lexicon:
phonological similarity of foreign words with native lexical items, or foreign
words displaying word-final consonants that can be interpreted as Italian
suffixes, may result in consonant gemination. Also, native segmental and
syllabic demands can yield gemination in loans. The metrical structure of
English words is also at play in guiding the adaptation process: English v́cv#
and v́cvc# words generally do not cause gemination, but English v́c# (with final
obstruents) generally do. In the latter case, if the final consonant is a
sonorant, gemination is determined by the tenseness of the preceding stressed
vowel, where tense English vowels do not geminate the final consonant in the
Italian loans, and lax English vowels do. Even though it does not directly
address the on-going debate, Repetti sheds light on how various grammatical
components may intervene in the adaptation process of foreign words.
“Nasal harmony and the representation of nasality in Maxacalí: Evidence from
Portuguese loans” W. Leo Wetzels
Wetzels explores the complex system of Maxacalí nasality in light of loanword
adaptations from Brazilian Portuguese (BP). Taking up the discussion spurred by
previous researchers on whether nasality is represented underlyingly or not,
Wetzels turns to loanwords from Portuguese in the hope of reaching a conclusion
on the matter. In short, the concern revolves around the lexical (underlying) or
derivable (surface) status of nasality for consonants, vowels and/or both in
Maxacalí. In light of the adaptations of BP words containing nasal segments the
author is compelled to posit nasality as a contrastive feature of Maxacalí
vowels, but not of consonants. The ensuing analysis employs autosegmental
association lines that spread the nasal feature to surrounding, tautosyllabic
segments, obligatorily for codas and optionally for onsets. The exploration of
loanword adaptations enables the linguist to settle a debate over the underlying
presence or absence of featural properties that would have been unresolved based
on data from the native lexicon.
EVALUATION
This volume presents state-of-the-art research in loanword phonology, addressing
the issues underlying most publications in this domain. The theoretical and
empirical breadth are both commendable. Some articles are technical and
theoretical while others are more descriptive (sometimes leaving a formal
analysis for future work), offering a good balance of perspectives across
contributors. The diversity of theoretical approaches (e.g. Optimality Theory,
Perceptually-driven models, Natural Phonology, Feature-driven model) is
particularly welcome, and shows that loanword adaptations can be, and should be,
viewed through multiple analytical lenses. Loanwords in widely studied
languages, such as Korean and Mandarin Chinese, are examined, but loanwords in
languages like Basque and Maxacalí are also represented. Furthermore, even
though the source language oftentimes tends to be English, a few articles do
investigate loanwords from other languages -- yet still belonging to the
Indo-European language family. A wider array of lending languages would have
been valuable. But despite the diverse empirical coverage and analytical
treatments, contributors refer to each other’s articles making this book cohesive.
Corpus-based studies are well represented. but this volume includes only limited
psycholinguistic / experimental research, like that spearheaded by Dupoux and
colleagues (Dupoux et al.1999, Peperkamp et al. 2008). Since, in the words of
Kenstowicz and Suchato (2006: 921) “loanword phonology takes on the status of
something akin to an ‘experiment of nature’”, it would be reasonable to expect
discussion of more experimentally-based studies investigating loanword adaptations.
In short, loanword adaptation is a complex phenomenon that necessitates
inclusive analytical treatments, making explicit note of the interaction between
various grammatical and extra-grammatical aspects of both the donor and
recipient language. This collection therefore provides testimony to the health
of, and prospects for, future loanword research.
REFERENCES
Boersma, Paul (2007). Cue constraints and their interactions in phonological
perception and production. Rutgers Optimality Archive 944.
Dupoux, Emmanuel, Christopher Pallier, Kazuhiko Kakehi, Yuki Hirose and Jacques
Mehler (1999). Epenthetic vowels in Japanese: A perceptual illusion? Journal of
Experimental Psychology: Human Perception and Performance 25 (6): 1568-1578.
Hyman, Larry (1970). The Role of Borrowings in the Justification of Phonological
Grammars. Studies in African Linguistics 1: 1-48.
Kenstowicz, Michael and Atiwong Suchato (2006). Issues in loanword phonology: A
case study from Thai. Lingua 116 (7): 921-949.
Kim, Hyunsoon (2007). A feature-driven model of loanword adaptation: evidence
from Korean. Ms., Hongik University.
Lacharité, Darlene and Carole Paradis (2005). Category preservation and
proximity versus phonetic approximation in loanword adaptation. Linguistic
Inquiry 36: 223-258.
McCarthy, John and Alan Prince (1995). Faithfulness and reduplicative identity.
In Jill Beckman, Laura Walsh Dickey and Suzanne Urbanczyck (eds.). Papers in
Optimality Theory (University of Massachusetts Occasional Papers 18). Amherst,
Mass.: Graduate Linguistics Student Association, 249-384.
Peperkamp, Sharon and Emmanuel Dupoux (2003). Reinterpreting Loanword
Adaptations: The role of perception. Proceedings of the 15th International
Conference of Phonetic Sciences. Barcelona: Causal Productions. 367-370.
Peperkamp, Sharon, Inga Vendelin and Kimihiro Nakamura (2008). On the perceptual
origin of loanword adaptations: evidence from Japanese. Phonology 25: 129-164.
Silverman, Daniel (1992). Multiple scansions in loanword phonology: evidence
from Cantonese. Phonology 9: 289-328.
Stampe, David (1979). A Dissertation on Natural Phonology. New York: Garland.
ABOUT THE REVIEWER
|
| |
ABOUT THE REVIEWER:
Lionel Mathieu holds a BA in linguistics from St. Cloud State University
(MN) and an MA in linguistics from the University of Arizona, where he is
currently a third year PhD student. His research interests focus on
loanword adaptations from a theoretical and experimental perspective,
phonology, psycholinguistics and second language acquisition.
|
|
|
|
|
|
