Lacy Paul de. Markedness: Reduction and Preservation in Phonology

Cambridge University Press, 2006. — 467 p. — (Cambridge Studies in Linguistics). — ISBN 0521839629, 9780521839624, 9780511242359.'Markedness' refers to the tendency of languages to show a preference for particular structures or sounds. This bias towards 'marked' elements is consistent within and across languages, and tells us a great deal about what languages can and cannot do. This pioneering study presents a groundbreaking theory of markedness in phonology. De Lacy argues that markedness is part of our linguistic competence, and is determined by three conflicting mechanisms in the brain: (a) pressure to preserve marked sounds ('preservation'), (b) pressure to turn marked sounds into unmarked sounds ('reduction'), and (c) a mechanism allowing the distinction between marked and unmarked sounds to be collapsed ('conflation'). He shows that due to these mechanisms, markedness occurs only when preservation is irrelevant. Drawing on examples of phenomena such as epenthesis, neutralisation, assimilation, vowel reduction and sonority-driven stress, Markedness offers an important insight into this essential concept in the understanding of human language.Aims
Challenges for markedness
Some markedness diagnostics do not work all the time
Marked elements are favoured
Markedness distinctions are conflated
Solutions
Competence
Typological frequency
Other frequencies
Diachronic change and loan phonology
Language acquisition and disorderC-markedness and synchronic alternations
Preservation of the marked
Markedness conflation and stringent form
Markedness reversals
Valid diagnostics
Invalid diagnostics
Markedness: an outline
Aims
Markedness hierarchies
The PoA hierarchy
Interpretation of [glottal]1
Phonological evidence for [N]
Comparison with Trigo (1988)
Evidence for dominance
Binary vs. multi-valued hierarchies
Markedness and environment
Theory and formalism
Constraint violations
Hierarchies and output constraints
Formal schema
Stringency and conflation
Hierarchies and faithfulness constraints
Previous theories
Structural descriptions
Multi-valued features
Multi-valued and binary features
Constraint form
Disjunction
Faithfulness
Environment and subcategories
DTEs and non-DTEs
Constraint form
Subsegmental and prosodic hierarchies
Manner of articulation: subcategories of constraints
Discovering hierarchies
Supra- and subsegmental hierarchies
Non-existing hierarchiesThe visibility of markedness reduction
An epenthetic typology
Glottal epenthesis: PoA beats sonority
Manner of articulation
The subordination of assimilation and dissimilation
Coronal epenthesis: sonority beats PoA
Why [t]?
Glottal sonority
Glottal Elimination is not related to PoACoda Sonority
Manner assimilation
Epenthetic labials and palatals
Impossible epenthetic segments
Epenthetic [d s zl]The output of neutralization
The irrelevance of faithfulness for neutralization outputs
Faithfulness
Ranking summary
Coronal promotion
More on conflict: voicing and sonority
Glottal neutralization
Miogliola /N/→[n]
Neutralization to [w]
Can labials and dorsals be promoted?Synchronic alternations
Mono-segmental morphemes in Koava and elsewhere
Suppletion
[N]
ConclusionsPreservation of the Marked in Yamphu
Neutralization
The need for PoM-faithfulness
Assimilation and Deletion
Beyond stops
Neutralization ranking
Typological implications
Inventories and neutralizations
Neutralization outside codas
The Glottal/Coronal Universal and disharmonic inventories
Nasal coalescence
Stop sonorization
Undergoers of assimilation
Unmarked undergoers and Preservation of the Marked9
Blocking labial and dorsal assimilation
Avoiding other outcomes
Direction of assimilation
Stop gemination
Marked undergoers and markedness reduction
Motivating assimilation
Analysis
Typology
Unmarked undergoers
Marked and mixed undergoer systemsOutput-only alternatives
Are PoM-faithfulness constraints enough?
Summary: overt and covert markednessPoA conflation
Blocking and conflation
Fixed ranking and conflation
Verb-noun asymmetries and supporting evidence
Preventing conflation: uvularsConflation prevention and vowel sonority
The Dutch data
Analysis
Preventing conflation
Permitting conflationDescriptionEvidence for stress
Sonority-driven stress I: attraction to [a]
Sonority-driven stress II: avoidance of…
Conflation
Alternatives
Representational approaches
Binary featuresBi-directional assimilation
Conflation
Faithfulness foixed rankingDescriptionCoalescence Generalizations
Details about the data
Analysis
Preservation of the Marked: dorsals
Faithfulness conflation: labials
Fixed ranking
Alternative faithfulness constraints
All Fixed Ranking theories
Minor PoA
Preservation of the Marked II: other features
Domain faithfulness and the majority rules problem
Domain faithfulness
Swedish and Pali again
Summary, extensions, and alternativesVowel epenthesis
The spectrum of epenthesis
High-sonority epenthesis in Coos
Alternative analyses
Low-sonority epenthesis
A sonority compromise
Epenthetic mid vowels
Universals of epenthetic quality
[o] epenthesis
[u] epenthesis
[œ]/[ø] epenthesis
Back unround epenthesis:…Vowel sonority in inventories
Harmonically contiguous vowel inventories
Unstressed vowel inventories
Gapped inventories
Disharmonic vowel inventories
Direction of vowel sonority neutralization
Sonorization and desonorization
Conflicting demands
Berguener Romansh unstressed ‘dispersion’
Mid vowels are not special
Impossible neutralization directions
ProsodificationNaming the theory
Predictions
Invalid diagnostics
Valid diagnostics
Variation
Labial unmarkedness and a theory of diagnostics
Markedness exists
Performance, not Competence
Many diachronic rights can create synchronic wrongs
No craziness
Representational complexity is not markedness
Targets
Placeless segments must undergo assimilation
Placeless segments cannot undergo dissimilation
Triggers
Blockers
Default variability and elaborated representation5
Default Variability
Velar unmarkedness
Restrictions on default variability
Other diagnostics for velar unmarkedness
Conflicting diagnostics
Conclusions
Non-contrastive markedness
Predictions
Inventory formation is neutralization
Which theory is right?
More on contrast-markedness
Markedness is absolute
A relative markedness approach to SLP creole
Dorsal undergoers
Unattested systems
Triggers
Conclusions
Markedness is expressed in both output and preservation constraints
The need for output constraints
Segmental vs. featural correspondence
Reference to the least marked
Conclusions
Markedness
Markedness in the futureSubject index
Language index