# Layering and Directionality

## Metrical Stress in Optimality Theory

##### Brett Hyde [+–]

##### Washington University in St. Louis

###### View Website

Brett Hyde is an associate professor in the Philosophy Department at Washington University in St. Louis. He has authored numerous journal articles and book chaptersaddressing key topics in metrical stress theory, including prosodic layering, parsing directionality, the metrical grid, trisyllabic stress windows, nonfinality and extrametricality, and the Iambic-Trochaic Law.

The metrical grid, the prosodic hierarchy, and the devices that establish directional parsing effects are closely intertwined in metrical stress theory. The metrical grid is the structure that represents stress patterns. The locations of stressed positions on the grid are constrained by the positions of categories in the prosodic hierarchy. Both the metrical grid and the prosodic hierarchy are manipulated by constraints, such as alignment constraints, that establish directional orientations within these structures. Assumptions about the representations affect the behavior of the constraints, and the particular formulation of the constraints influences the ultimate configuration of the representations.

*Layering and Directionality* is unique in the OT literature in that it examines both halves of the equation. It addresses the formulation of constraints that produce directional parsing effects, but it also addresses assumptions concerning prosodic and metrical structure. The book presents and defends three central proposals: the Weak Bracketing approach to layering relationships between prosodic categories, the Optimal Mapping approach to the relationship between prosodic categories and the metrical grid, and the Relation-Specific Alignment approach to parsing directionality. The book is also be unique in its coverage of OT accounts, comparing the proposed approach to approaches that range from Generalized Alignment in standard OT to the more recent Iterative Foot Optimization couched within the framework of Harmonic Serialism. The book draws extensively on the typological literature to evaluate the predictions of the accounts examined.

**Series**: Advances in Optimality Theory

### Table of Contents

Prelims

Dedication [+–] vii

The metrical grid, the prosodic hierarchy, and the devices that establish directional parsing effects are closely intertwined in metrical stress theory. The metrical grid is the structure that represents stress patterns. The locations of stressed positions on the grid are constrained by the positions of categories in the prosodic hierarchy. Both the metrical grid and the prosodic hierarchy are manipulated by constraints, such as alignment constraints, that establish directional orientations within these structures. Assumptions about the representations affect the behavior of the constraints, and the particular formulation of the constraints influences the ultimate configuration of the representations. Layering and Directionality is unique in the OT literature in that it examines both halves of the equation. It addresses the formulation of constraints that produce directional parsing effects, but it also addresses assumptions concerning prosodic and metrical structure. The book presents and defends three central proposals: the Weak Bracketing approach to layering relationships between prosodic categories, the Optimal Mapping approach to the relationship between prosodic categories and the metrical grid, and the Relation-Specific Alignment approach to parsing directionality. The book is also be unique in its coverage of OT accounts, comparing the proposed approach to approaches that range from Generalized Alignment in standard OT to the more recent Iterative Foot Optimization couched within the framework of Harmonic Serialism. The book draws extensively on the typological literature to evaluate the predictions of the accounts examined.

Preface [+–] ix-x

The metrical grid, the prosodic hierarchy, and the devices that establish directional parsing effects are closely intertwined in metrical stress theory. The metrical grid is the structure that represents stress patterns. The locations of stressed positions on the grid are constrained by the positions of categories in the prosodic hierarchy. Both the metrical grid and the prosodic hierarchy are manipulated by constraints, such as alignment constraints, that establish directional orientations within these structures. Assumptions about the representations affect the behavior of the constraints, and the particular formulation of the constraints influences the ultimate configuration of the representations. Layering and Directionality is unique in the OT literature in that it examines both halves of the equation. It addresses the formulation of constraints that produce directional parsing effects, but it also addresses assumptions concerning prosodic and metrical structure. The book presents and defends three central proposals: the Weak Bracketing approach to layering relationships between prosodic categories, the Optimal Mapping approach to the relationship between prosodic categories and the metrical grid, and the Relation-Specific Alignment approach to parsing directionality. The book is also be unique in its coverage of OT accounts, comparing the proposed approach to approaches that range from Generalized Alignment in standard OT to the more recent Iterative Foot Optimization couched within the framework of Harmonic Serialism. The book draws extensively on the typological literature to evaluate the predictions of the accounts examined.

Acknowledgments [+–] xi

The metrical grid, the prosodic hierarchy, and the devices that establish directional parsing effects are closely intertwined in metrical stress theory. The metrical grid is the structure that represents stress patterns. The locations of stressed positions on the grid are constrained by the positions of categories in the prosodic hierarchy. Both the metrical grid and the prosodic hierarchy are manipulated by constraints, such as alignment constraints, that establish directional orientations within these structures. Assumptions about the representations affect the behavior of the constraints, and the particular formulation of the constraints influences the ultimate configuration of the representations. Layering and Directionality is unique in the OT literature in that it examines both halves of the equation. It addresses the formulation of constraints that produce directional parsing effects, but it also addresses assumptions concerning prosodic and metrical structure. The book presents and defends three central proposals: the Weak Bracketing approach to layering relationships between prosodic categories, the Optimal Mapping approach to the relationship between prosodic categories and the metrical grid, and the Relation-Specific Alignment approach to parsing directionality. The book is also be unique in its coverage of OT accounts, comparing the proposed approach to approaches that range from Generalized Alignment in standard OT to the more recent Iterative Foot Optimization couched within the framework of Harmonic Serialism. The book draws extensively on the typological literature to evaluate the predictions of the accounts examined.

Chapter 1

Introduction [+–] 1-37

The first half of Chapter 1 introduces three components of the proposed approach to metrical stress theory: Weak Bracketing (Hyde 2002), Optimal Mapping (Hyde 2002), and Relation-Specific Alignment (Hyde 2012a). Weak Bracketing is a departure from the standard Weak Layering (Itô and Mester 1992) approach to prosodic organization. Its key characteristic is that it requires exhaustive parsing of syllables into feet but it allows two feet to share a syllable. The Optimal Mapping approach to constructing the metrical grid represents a departure from the standardly assumed one-to-one correspondence between feet and stress (Selkirk 1980). Optimal Mapping allows feet to occur without grid entries and even allows them to share grid entire in some circumstances. Relation-Specific Alignment is an alternative to the original Generalized Alignment (McCarthy and Prince 1993) formulation for alignment constraints. By targeting misalignment only in specific structural contexts, Relation-Specific Alignment is able to produce the essential directional parsing effects while avoiding some of Generalized Alignment’s more significant difficulties. The second half of Chapter 1 presents the typology of quantity-insensitive stress patterns assumed throughout the book, the typology used to evaluate both the proposed approach and competing approaches. The typology, presented in terms of mirror-image iambic and trochaic stress patterns, is consistent with three key generalizations. First, in mirror image patterns with neither clash nor lapse, both members of the pair are attested. Second, in mirror image patterns with either clash or lapse, at most one member of the pair is attested. Finally, attested patterns with clash or lapse always have stress on the initial syllable, always leave the final syllable stressless, or both. Topics include: 1.1 Weak Layering and Weak Bracketing; 1.2 Mapping to the Grid; 1.3 Generalized Alignment and Relation-Specific Alignment; 1.4 Typological Generalizations; 1.5 Overview; 1.6 A Note on Tableaux

Chapter 2

Establishing Directional Orientations [+–] 38-91

Chapter 2 has two primary aims. The first is to address the formulation of alignment constraints, and the second is to establish the basic predictions of Symmetrical Alignment (McCarthy and Prince 1993) and Iterative Foot Optimization (Pruitt 2010). The chapter begins by examining the properties of alignment constraints under the standard Generalized Alignment (McCarthy and Prince 1993) formulation, considering both the formulation’s advantages and its shortcomings. It then examines the alternative Relation-Specific Alignment (Hyde 2012a) formulation. Relation-Specific Alignment allows alignment constraints to produce their essential directional parsing effects while avoiding the problems encountered under Generalized Alignment. Symmetrical Alignment is the primary example employed in demonstrating that Relation-Specific Alignment retains the ability to produce essential directionality effects. Iterative Foot Optimization is employed as an example to demonstrate the importance of parallel evaluation.Topics include: 2.1 Generalized Alignment; 2.2 Relation-Specific Alignment; 2.3 Symmetrical Alignment; 2.4 Advantages of Relation-Specific Alignment; 2.5 Iterative Foot Optimization; 2.6 Summary

Chapter 3

Prosodic Layering [+–] 92-138

Chapter 3 begins to establish the motivation for Weak Bracketing (Hyde 2002) by examining the Odd-Parity Input Problem (Hyde 2012b) in Weak Layering (Itô and Mester 1992) accounts. The Odd-Parity Input Problem is a set of problematic predictions that arise under Weak Layering in an effort to achieve exhaustive binary parsing in odd-parity forms. It can result in an unusual type of quantity-sensitivity or in the systematic conversion of odd-parity inputs into even-parity outputs. Examinations of quantity-insensitive patterns in the literature typically proceed under an idealized condition: all syllables are treated as if they are light. As a result, distinctions in syllable weight are not actually considered. Chapter 3 begins an examination of the effects of weight distinctions in Weak Layering accounts by reconsidering the predictions of Symmetrical Alignment (McCarthy and Prince 1993) and Iterative Foot Optimization (Pruitt 2010). When syllabic weight distinctions are taken into account, the unusual quantity-sensitivity of the Odd-Parity Input Problem emerges. In Symmetrical Alignment, heavy syllables are parsed as monosyllabic feet but only if they are in an odd-numbered position in an odd-parity form. In Iterative Foot Optimization, heavy syllables are parsed as monosyllabic feet but only if they are the last syllable in an odd-parity form to have their parsing status determined. Topic include: 3.1 Introduction; 3.2 The OPIP in Symmetrical Alignment; 3.3 The OPIP in Iterative Foot Optimization; 3.4 Summary

Chapter 4

Alternative Weak Layering Accounts [+–] 139-196

Chapter 4 completes the examination of the effects of the Odd-Parity Input Problem (Hyde 2012b) begun in Chapter 3. It examines the effects of the problem in two additional Weak Layering accounts: Asymmetrical Alignment (Alber 2005) and Rhythmic Licensing (Kager 2005). These two approaches differ from those considered in Chapter 3 in that they rely less on alignment constraints to produce directional parsing effects and more on principles, such as clash and lapse avoidance, that promote rhythmic well-formedness. The chapter demonstrates the effects of the Odd-Parity Input Problem become more exotic in Weak Layering accounts as the role of alignment is reduced. Topics include: 4.1 Asymmetrical Alignment; 4.2 Asymmetrical Alignment and the OHP; 4.3 Rhythmic Licensing; 4.4 The OHP in Rhythmic Licensing; 4.5 Summary

Chapter 5

Restricting Clash and Lapse [+–] 197-244

Chapter 5 discusses the three components of the proposed approach?Weak Bracketing (Hyde 2002), Optimal Mapping (Hyde 2002), and Relation-Specific Alignment (Hyde 2012a)?in fuller detail and carefully examines the approach’s predictions. First, the chapter demonstrates that Weak Bracketing allows the proposal to avoid the effects Odd-Parity Input Problem (Hyde 2012b) discussed in Chapters 3 and 4. The ability of disyllabic feet to overlap allows for exhaustive binary parsing of odd-parity forms in a way that does not result in the problematic patterns that arise under Weak Layering. The chapter then demonstrates how Weak Bracketing combines with Relation-Specific Alignment and Optimal Mapping to provide the foundation for an effective analysis of quantity-insensitive binary patterns. It also explores the role of two asymmetrically formulated constraints: INITIAL-GRIDMARK (Prince 1983), which insists that initial syllables be stressed, and NON-FINALITY (Prince and Smolensky 1993/2004), which insists that final syllables be stressless. INITIAL-GRIDMARK and NON-FINALITY play a key role in predicting iambic-trochaic asymmetries. Topics include: 5.1 Conditions and Constraints; 5.2 Avoiding the Odd-Parity Input Problem; 5.3 Outline of the Crucial Rankings for Binary Patterns; 5.4 Minimal Alternation and its Variants; 5.5 Maximal Alternation and its Variants; 5.6 Summary of Predicted Patterns

Chapter 6

Accent Windows [+–] 245-288

Chapter 6 presents the Relation-Specific Alignment (Hyde 2012a) approach to trisyllabic and other accent windows and demonstrates that it provides a general account of the phenomenon where alternative proposals do not. Having examined the predictions that arise under the different approaches to prosodic layering in Chapters 3-5, Chapter 6 returns to the Relation-Specific Alignment formulation, introduced in Chapter 2, and to an unexpected result that emerges in the context of opposite-edge alignment. Like its same-edge counterparts, Relation-Specific Alignment’s opposite-edge constraints avoid the difficulties that emerge under Generalized Alignment (McCarthy and Prince 1993). In the opposite-edge case, however, the remedy has some surprising consequences. Opposite-edge constraints have the effect of confining instances of one of the aligned categories to a ‘window’ at an edge of a form established by an instance of the other aligned category. The opposite-edge effect provides a simple, general account of trisyllabic accent windows and other types of windows. While alternatives such as extended lapse avoidance (Gordon 2002, Kager 2005), weak local parsing (Kager 1994, Green 1995, Green and Kenstowicz 1995) and non-finality (Prince and Smolensky 1993/2004) provide effective analyses in some cases, they do not provide analyses for the range of windows accommodated under Relation-Specific Alignment. Topic include: 6.1 A General Approach to Trisyllabic Windows; 6.2 Full Windows; 6.3 Truncated Windows; 6.4 A Morphology-Based Window; 6.5 Summary

Chapter 7

Conclusion [+–] 289-306

The metrical grid, the prosodic hierarchy, and the devices that establish directional parsing effects are closely intertwined in metrical stress theory. The metrical grid is the structure that represents stress patterns. The locations of stressed positions on the grid are constrained by the positions of categories in the prosodic hierarchy. Both the metrical grid and the prosodic hierarchy are manipulated by constraints, such as alignment constraints, that establish directional orientations within these structures. Assumptions about the representations affect the behavior of the constraints, and the particular formulation of the constraints influences the ultimate configuration of the representations. Layering and Directionality is unique in the OT literature in that it examines both halves of the equation. It addresses the formulation of constraints that produce directional parsing effects, but it also addresses assumptions concerning prosodic and metrical structure. The book presents and defends three central proposals: the Weak Bracketing approach to layering relationships between prosodic categories, the Optimal Mapping approach to the relationship between prosodic categories and the metrical grid, and the Relation-Specific Alignment approach to parsing directionality. The book is also be unique in its coverage of OT accounts, comparing the proposed approach to approaches that range from Generalized Alignment in standard OT to the more recent Iterative Foot Optimization couched within the framework of Harmonic Serialism. The book draws extensively on the typological literature to evaluate the predictions of the accounts examined. This chapter summarizes the main results of the previous chapters and discusses their consequences for an OT approach to metrical stress. Topics include: 7.1 Relation-Specific Alignment; 7.2 Weak Bracketing; 7.3 Optimal Mapping; 7.4 Additional Support for Alignment Constraints; 7.5 Directions for Future Research

End Matter

References [+–] 307-316

Index [+–] 317-321