Christopher K. Frantz, Saba Siddiki - Institutional Grammar

Institutional
Grammar
Foundations and Applications

for Institutional Analysis
Christopher K. Frantz · Saba Siddiki

Institutional Grammar
“Institutional Grammar is a must read for social scientists and legal scholars
who study the complex interplay between institutional arrangements and actors’
behavior, as well as scholars interested in the semantics of institutions. Beginning
with a careful exposition of Elinor Ostrom’s Institutional Analysis and Devel-
opment Framework and its affiliated grammar of institutions (aka IG 1.0), the
authors present IG 2.0, identifying how it corrects the shortcomings of IG 1.0,
as well as substantively expanding it to allow for the coding of regulative and
constitutive institutional statements at different levels of granularity. Importantly,
the authors include guidance on how to implement IG 2.0 and analyze the
coded data, making the grammar accessible and highlighting its practical value.
This book represents the definitive text on the grammar of institutions.”
—Edella Schlager, Professor, University of Arizona
“Frantz and Siddiki summarize the Institutional Grammar 2.0 for the systematic
analysis of institutions. In doing so, their book revolutionizes the study of
institutions and provides a sturdy foundation for building knowledge about
them. This book is without peers, and its impact on the analysis of institutions
will stretch across disciplines and extend far into the future.”
—Christopher M. Weible, Professor, University of Colorado Denver
“Institutions are ubiquitous but can be challenging to study. With their devel-
opment and description of the Institutional Grammar 2.0, Frantz and Siddiki
provide a de-facto open standard for a new field we might call ‘Computational
Institutional Analysis.’ This is a must-read for anyone interested in contributing
to this emerging area of social science.”
—Charlie Schweik, Professor, University of Massachusetts Amherst
Christopher K. Frantz · Saba Siddiki
Institutional Grammar
Foundations and Applications for Institutional
Analysis
Christopher K. Frantz Saba Siddiki
Norwegian University of Science Syracuse University
and Technology (NTNU) Syracuse, NY, USA
Gjøvik, Norway
ISBN 978-3-030-86371-5 ISBN 978-3-030-86372-2 (eBook)

https://doi.org/10.1007/978-3-030-86372-2
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer
Nature Switzerland AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the
Publisher, whether the whole or part of the material is concerned, specifically the rights
of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on
microfilms or in any other physical way, and transmission or information storage and
retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc.
in this publication does not imply, even in the absence of a specific statement, that such
names are exempt from the relevant protective laws and regulations and therefore free for
general use.
The publisher, the authors and the editors are safe to assume that the advice and informa-
tion in this book are believed to be true and accurate at the date of publication. Neither
the publisher nor the authors or the editors give a warranty, expressed or implied, with
respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps
and institutional affiliations.
Cover credit: Alex Linch/shutterstock.com
This Palgrave Macmillan imprint is published by the registered company Springer Nature
Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
The Institutional Grammar (IG) is a well-established approach for insti-

tutional analysis. It was conceptually defined by Crawford and Ostrom in
1995. Over the last quarter of a century, a rich body of work has applied
“the Grammar” in the context of policy analysis, computational modeling,
as well as in other domains. Throughout this process, scholars developed
associated methodological guidance, and also identified challenges to the
general applicability and conceptual validity of the IG. However, there
have not yet been attempts to view these developments and challenges
in concert, so as to consider how they link to the core conceptual and
methodological foundations of the IG, and more importantly, how they
can be reconciled in an integrated way to bolster the IG as a robust
approach for analyzing the structure and meaning of institutions.
This book responds to this gap by offering a comprehensive introduc-
tion to the foundational principles of the IG, how it has been used, and
the opportunities and challenges presented in existing research. But more
than just taking a retroactive view, the book reconciles the original IG
with a modernization, a New Institutional Grammar labeled “Institu-
tional Grammar 2.0,” that revises it for novel applications within and
beyond existing research fields. It aids in establishing linkages of the IG
to fields in which institutions are studied more generally, but in which the
IG has not yet found broad recognition or adoption.
The IG 2.0 provides the basis for a consistent and unambiguous
encoding and representation of institutional information independent
v
vi PREFACE
of application field and analytical method. Working toward this effort,

the refined Grammar offers both a platform for more comprehensive
characterization of institutional information as well as the ability to
systematically capture parameterizing information in the form of consti-
tutive statements, the latter of which was not at all considered as part of
the original IG. Overarching these revisions is the shift from a primarily
syntactic focus emphasized by the original IG toward a semantic perspec-
tive that explicitly focuses on institutional meaning, functions and effect of
statements, dissociating the “institutional” from a “linguistic” grammar,
enabling and preparing for advanced analyses not possible with the IG to
date. Responding to the diverse application fields of the IG, the different
features associated with this “New Institutional Grammar” are organized
by levels of expressiveness that respond to specific analytical needs, but
explicitly recognize its purpose in extending the original IG.
The structure and content of the book reflects this orientation by
initially focusing on Foundations of the IG, that includes both Craw-
ford and Ostrom’s Institutional Grammar and existing applications
(Chapters 1 and 2), before moving to an integrated discussion of the
conceptual refinements that form the core of this book (Chapters 3 to 6).
The latter part of the book focuses on the Application of the IG,
and offers methodological guidance for the encoding of institutional
information, both including general principles of research design as well
as operational guidance (Chapter 7), alongside supplementary resources
available via the book website (https://newinstitutionalgrammar.org).
Looking toward the adoption of the conceptual innovations, the substan-
tive part of the book ends with an exposition of existing applications
of the IG as well as discussion of prospective analytical opportunities of
the IG (Chapter 8). This includes applications that focus on fine-grained
elements (e.g., complexity metrics) as well higher-level applications that
provide the basis for systemic analyses of institutions.
In both its structure and orientation, the book reflects the empirically
observed convergence between independent approaches to institutional
analysis that rely on the IG. To this end, it is instructive that the authors
themselves approach this effort from different perspectives, including the
mainstay application of the IG in the context of public policy, with a
primary focus on a formal-legalistic perspective on institutions, and a
complementary socio-institutional perspective that emphasizes the anal-
ysis of simulated behavior in the context of computational modeling.
Central to the vision is the belief that the IG is equally suited to capture
PREFACE vii
diverse perspectives on institutions as the fabric of social systems, inde-

pendent of the nature of institutions, applied methods and techniques,
and analytical foci (e.g., levels of analysis). Recognizing and developing
the IG as an analytical paradigm, as opposed to merely a tool, the book
provides a foundational introduction to the IG that is intended to be
accessible to researchers that engage in traditional applications of the IG,
but also scholars who have no prior exposure and wish to explore the IG
and associated analytical opportunities.
Inasmuch as the book is written by the authors, the ideas repre-
sented herein have been developed with feedback and critiques from a
broader community of scholars equally committed to the IG. We specif-
ically acknowledge the Affiliates of the Institutional Grammar Research
Initiative (IGRI), with whom we have engaged in various valuable
and thought-provoking interactions, including through working groups,
research presentations and workshops, meetings, and individual conver-
sations. We are especially grateful to Ute Brady, Edella Schlager, Matia
Vannoni, Bartosz Pieliński, Seth Frey, Marcello Ceci, Doug Rice, Tanya
Heikkila, Chris Weible, Charlie Schweik, and Brenda Bushouse. We also
wish to thank IGRI interns that helped explore selected features of the
proposed refinements, including Anamaria Rizo, Beyza Gurler, Stephanie
Potts, Angelo Baldado, Umberto Tabalappi, as well as Johanne Bognøy,
who explored visual IG coding tools as part of her Master’s study. The
authors further wish to thank the anonymous reviewers of our IG 2.0
research, both the ideas presented in the proposal of this manuscript,
as well as complementary peer reviewed published work. We also wish
to thank the United States National Science Foundation (grant no.
1917908) for supporting this research and broader development of the
IG research program through a research coordination network grant
titled, “Coordinating and Advancing Analytical Approaches for Policy
Design.”
Finally, but certainly not least, we wish to express deep gratitude to
our families.
Gjøvik, Norway Christopher K. Frantz

Syracuse, USA Saba Siddiki
November 2021
Contents
1 Introduction 1
1.1 What Is Institutional Analysis? 6
1.1.1 Institutional Analysis by Discipline 7
1.1.2 Convergence Toward Interdisciplinary
Institutional Analysis 16
1.2 Institutional Analysis with the Institutional Analysis
and Development Framework 18
1.3 Primer on the Institutional Grammar 19
1.4 Overview of Chapter Contents 26
References 26
2 Review of Institutional Grammar Research: Overview,
Opportunities, Challenges 33
2.1 Guiding Research Questions 33
2.2 Analytical Approaches Used in Institutional Grammar
Research 36
2.2.1 Frameworks, Theories, and Concept
Measurement 36
2.2.2 Collecting and Analyzing Institutional
Grammar Data 38
2.3 Research Opportunities and Challenges 44
References 48
ix
x CONTENTS
3 Motivation for a New Institutional Grammar 53

3.1 Ontological Consistency 54
3.2 Toward a Comprehensive Representation
of Institutional Meaning 61
3.3 Grammars in the Linguistic and Institutional Sense 65
References 72
4 Institutional Grammar 2.0: Conceptual Foundations
and General Syntax 75
4.1 Conceptual Foundations 75
4.1.1 Levels of Expressiveness 76
4.1.2 Constitutive and Regulative Statements 79
4.2 IG Core 81
4.2.1 Regulative Syntax 82
4.2.2 Statement Combinations (Horizontal Nesting) 94
4.2.3 Regulative Institution Types 101
4.2.4 Delta Parameters in the Institutional
Grammar 114
4.2.5 Constitutive Syntax 116
4.2.6 Constitutive Institution Types 130
4.3 Summary of Chapter Content 133
References 135
5 Institutional Grammar 2.0: Deep Structural Parsing
and Hybrid Institutional Statements 141
5.1 IG Extended 141
5.1.1 Component-Level Structure 141
5.1.2 Institutional Statements vs. Institutional States 143
5.1.3 Reconstructing Embedded Institutional
Meaning 151
5.1.4 Object-Property Hierarchy 160
5.1.5 Property Types 164
5.1.6 Context Taxonomy 166
5.2 Hybrid Institutional Statements 172
5.2.1 Revisiting Constitutive Rules in Literature 172
5.2.2 Integrating Constitutive and Regulative
Statements 174
5.3 Summary of Chapter Content 187
References 190
CONTENTS xi
6 Institutional Grammar 2.0: Semantic Features

and Analytical Linkages 193
6.1 IG Logico 193
6.1.1 Semantic Specification 194
6.1.2 Semantic Annotations 201
6.1.3 Statement References 221
6.1.4 Statement Transformation Rules 224
6.1.5 Summary of Chapter Contents 232
6.2 Synthesizing the Institutional Grammar 233
References 240
7 Methodological Guidance for Encoding Institutional
Information 243
7.1 The IG Coding Process: Planning, Execution,
Assessment 243
7.2 Planning 244
7.2.1 Data Collection Considerations 244
7.2.2 Tool Support 245
7.2.3 Determining applicable Institutional
Grammar Feature Set 246
7.2.4 Determining Data Processing Practices 250
7.2.5 Coder Selection 251
7.3 Execution 251
7.3.1 Data Selection 251
7.3.2 Coding Platform Selection 252
7.3.3 Preprocessing Institutional Data 252
7.3.4 Operational Coding of Institutional Statements 260
7.4 Assessment 273
7.5 Concluding Remarks 275
References 276
8 Institutional Analysis and Applications 277
8.1 IG Core—Establishing Fundamental Institutional
Metrics 280
8.1.1 Component-Level Aggregate Metrics 280
8.1.2 Network Analysis 287
8.1.3 Additional Analytical Pathways 288
8.2 IG Extended—Structural and Behavioral Analysis
of Institutional Design 291
8.2.1 Structural Analysis 291
xii CONTENTS
8.2.2 Institutional Modeling 314

8.2.3 Discussion of Structural and Behavioral
Analytical Approaches 334
8.3 IG Logico—Semantic Analyses 336
8.3.1 Perspective Extrapolation in Institutional
Statements 337
8.3.2 Chaining Transformation Rules 342
8.3.3 Epistemological Linkage Through Institutional
Functions Analysis 344
8.3.4 Additional Analytical Opportunities 352
8.4 Summary of Analytical Approaches and Opportunities 353
References 358
9 Contextualization and Future Development
of the Institutional Grammar 363
9.1 The Institutional Grammar: An Analytical Paradigm
for Institutional Analysis 363
9.2 Future Directions in Institutional Grammar Research 369
9.2.1 Conceptual Directions 370
9.2.2 Disciplinary Directions 371
9.2.3 Catalyzing Theory and Framework
Development 374
References 375
Appendix A: Institutional Statement Structure 377

Appendix B: National Organic Program Regulation 379
U.S. National Organic Program Regulation 381
Glossary 389
Index 395
List of Figures
Fig. 3.1 Combined Institutional Grammar patterns example 57

Fig. 3.2 Combined Institutional Grammar patterns example 70
Fig. 3.3 Separated Institutional Grammar patterns example 70
Fig. 4.1 Levels of expressiveness in the Institutional Grammar 2.0 78
Fig. 4.2 Interaction of constitutive and regulative statements 80
Fig. 4.3 Continuous Deontic conception 84
Fig. 4.4 Direct and indirect objects in regulative institutional
statements 86
Fig. 4.5 Activation Conditions vs. Execution Constraints 88
Fig. 4.6 Regulative statement structure 94
Fig. 4.7 Nesting principles 100
Fig. 4.8 Regulative statement type structures 114
Fig. 4.9 Constitutive statement component relationships 130
Fig. 4.10 IG 2.0 Features by Level of Expressiveness 134
Fig. 5.1 Institutional statements vs. institutional states 146
Fig. 5.2 Schematic overview of nesting characteristics 152
Fig. 5.3 Institutional Statement Coding Process 154
Fig. 5.4 Institutional Statement Component Linkage to Action
Situation 159
Fig. 5.5 Exemplified linkage of objects and properties 162
Fig. 5.6 Property Typology 165
Fig. 5.7 Context taxonomy 170
Fig. 5.8 Institutional Statement Variants 186
Fig. 5.9 IG 2.0 Features by Level of Expressiveness 187
Fig. 6.1 Constitutive Functions Taxonomy 212
xiii
xiv LIST OF FIGURES
Fig. 6.2 First-order Linkages between Taxonomies and IG

Components 215
Fig. 6.3 IG 2.0 Levels of Expressiveness and Associated Features 234
Fig. 6.4 Structural Composition Patterns for Nested Institutional
Statements 235
Fig. 6.5 Institutional Grammar Overview 238
Fig. 8.1 Analytical Applications by Levels of Expressiveness 278
Fig. 8.2 Network structure of compliance excerpt 288
Fig. 8.3 Atomic statement distribution across subsections 289
Fig. 8.4 Atomic statement distribution across policy (Complexity
Landscape) 290
Fig. 8.5 Institutional State Complexity Metrics across Institutional
Tree 293
Fig. 8.6 Institutional State Complexity Aggregation 296
Fig. 8.7 IG Compositional Patterns Overview 303
Fig. 8.8 IG Compositional Patterns for Illustrative Statement 304
Fig. 8.9 IG Compositional Patterns for two statements (Vertical
Linkage) 306
Fig. 8.10 IG Compositional Patterns (Horizontal and Vertical
Linkages) 307
Fig. 8.11 IG compositional patterns example with selected
component information 310
Fig. 8.12 Excerpt of Conceptual Entity Organization in Scenario 312
Fig. 8.13 General conceptual mapping of IG components
to agent-based models 322
Fig. 8.14 Exemplary Execution Cycle of a farmer in the Organic
Farming Scenario 327
Fig. 8.15 Generated Agent Society 328
Fig. 8.16 Exemplary Institutional Statement Output 329
Fig. 8.17 Systemic Interlinkage of Institutional Function
Annotations 347
Fig. 8.18 Systemic Interlinkage of Action Situations 348
Fig. 8.19 Systemic decomposition with effect quantification 349
Fig. 8.20 Analytical Applications by Levels of Expressiveness 354
Fig. 8.21 Summary of analytical approaches and metrics based
on analytical focus 357
Fig. 9.1 Institutional Grammar 2.0 by Levels of Expressiveness
and associated Perspectives, Concepts and Analytical
Applications 367
Fig. A.1 Institutional Statement Structure in the Institutional
Grammar 2.0 378
List of Tables
Table 3.1 Ontological inconsistencies in the Institutional Grammar 60

Table 3.2 Comparison of key characteristics between linguistic
and Institutional Grammar 70
Table 4.1 Institution type characteristics 104
Table 4.2 Semantic distinction between norms and rules
in the Institutional Grammar 2.0 113
Table 4.3 Institution types of regulative and constitutive statements 133
Table 5.1 Types of nesting in institutional statements 150
Table 5.2 Indicators of Hybrid Statements 186
Table 5.3 IG Extended Features Summary 189
Table 6.1 IG Component Symbols 197
Table 6.2 IG Logico Taxonomies and affected Institutional
Statement Components 219
Table 8.1 Most frequently occurring component information
(regulative statements) 282
Table 8.2 Most frequently occurring component information
(constitutive statements) 282
Table 8.3 Most frequently paired institutional information
across syntactic components 283
Table 8.4 Statement stringency operationalization 284
Table 8.5 Constitutive-regulative Dynamics 286
Table 8.6 Degrees of State Variability for Logical Operators 295
Table 8.7 Institutional Tree Metrics 300
Table 8.8 Institutional State Regimentation based on associated
Logical Operators 301
xv
CHAPTER 1
Introduction
This book offers guidance on the analysis of institutions. Institutions

are rules, norms, and strategies that govern social systems (Crawford &
Ostrom, 1995; Frantz & Siddiki, 2021). They regulate behavior within
social systems, by structuring opportunities and constraints for actions,
and they establish features of social systems in which, and through which,
behavior occurs. They can be emergent – reflecting regularities in behav-
ioral manifestations that have developed among actors over time. Or, they
can be specifically crafted and applied in a deliberate attempt to alter
behavioral or systemic features or states of affairs. In this light, institutions
can be reflected in social norms that emerge and evolve within commu-
nities over time, or as laws or regulations that are purposely designed
and administered to influence social systems. Further, this quality of insti-
tutions means that they can be conceived of in terms of antecedents or
products of other features of social systems.
Given the salience of institutions in social systems and the recogni-
tion of their role as a structural foundation of every aspect of society,
their study has garnered the interest of scholars representing the gamut
of social science disciplines, including political science, public policy,
economics, and sociology, as well as fields such as computer science. As
well, the study of institutions has also been of interest to scholars whose
work lies at the intersections of these disciplines; for example, scholars
of political economy and computational social scientists. Core questions
© The Author(s), under exclusive license to Springer Nature 1

Switzerland AG 2022
C. K. Frantz and S. Siddiki, Institutional Grammar,
https://doi.org/10.1007/978-3-030-86372-2_1
2 C. K. FRANTZ AND S. SIDDIKI
variably associated with scholarship in these various disciplines that moti-

vate the study of institutions, include: What are the institutions that
communities use to govern behavior? How exactly do institutions oppor-
tune and constrain individual and collective behavior within particular
settings? What qualities of institutions make them more or less effec-
tive in meeting specific objectives? How do institutions interact with
environmental factors (e.g., resources, shocks) in order to produce effi-
cient outcomes? Are the institutions used in one community more or less
effective at achieving particular outcomes than others? What shapes insti-
tutional emergence and evolution? When and why do actors comply with
institutions? How does the specific nature of institutions interact with the
scale at which they operate (e.g., social groups, communities, society)?
Notably, the breadth of disciplines for which the study of institutions
is relevant has contributed to a theoretically and methodologically rich
body of scholarship. As a corpus, institutional scholarship relies on a
wide array of theories and methods, with varying epistemological roots,
reflecting the diverse disciplinary backgrounds of its contributors. This
theoretical and methodological diversity has led to the generation of novel
insights regarding institutional phenomena. At the same time, it has chal-
lenged the identification of generalizable insights, insofar as theories and
accompanying analytical approaches rooted in different disciplinary tradi-
tions often rely on distinctive conceptualizations and operationalizations
of institutions.
Efforts have been made to reconcile diversity in theoretical and
methodological approaches engaged in the study of institutions, while
at the same time respecting it, in order to foster the development of
generalizable knowledge regarding institutional phenomena. This is the
very effort that motivated the generation of the Institutional Grammar,
which is the focus of this text. The Institutional Grammar, originally
referred to as the “Grammar of Institutions,” was first proposed by Sue
Crawford and Elinor Ostrom in 1995 in an article published in the Amer-
ican Political Science Review. Essentially, the Institutional Grammar (IG)
offers an approach for the systematic characterization of institutions along
generalizable, constituent parts, fundamentally lending the ability to more
precisely define what institutions are, and more precisely predict how
institutions will influence behavior within different situations. Parts of
institutions are captured with syntactic components which combine to
form a “grammar” of institutions. Crawford and Ostrom posited that all
institutions, which are themselves assumed to be comprised of individual
1 INTRODUCTION 3
statements that regulate behavior or constitute features of the institutional

system, are made up of some configuration of syntactic components. The
particular array of components of which institutions are comprised deter-
mines how they are expected to compel behavior or parameterize social
systems.
The IG, while intriguing to many at the time of its publication, was
applied by few in the years following. Though the IG was introduced
in 1995, it was not until 2008 that the first academic journal article
explicitly incorporating a discussion of it was published (Smajgl et al.,
2008). This first article was shortly followed by another set of publications
(Basurto et al., 2010; Schlüter & Theesfeld, 2010). Noteworthy about
this first set of articles on the IG was that, although published in rela-
tively quickstep, articles were authored by distinctive research teams, in
different parts of the world, and by scholars with varying disciplinary
orientations. Smajgl et al. (2008) used the IG to characterize features
of rules endogenously generated in simulated environments. Schlüter and
Theesfeld (2010) addressed the premise of the IG’s embedded syntax.
Basurto et al. (2010) approached the IG from a methodological and use
case perspective, delineating a set of operational guidelines for applying it
toward the study of institutions taking the form of public policy.
Interest in the IG has surged over the last ten years since this initial
set of articles was published, and studies have largely followed one of
a few distinct tracks according with the different topics of this early
work. Indeed, extant IG research generally falls into one of three camps
tracking with the foci of Smajgl et al. (2008), Schlüter and Theesfeld
(2010), and Basurto et al. (2010). In one camp of IG research, scholars
engage it within the context of computational methods (e.g., agent-based
modeling), showcasing, as Smajgl et al. (2008) did, that the IG can
be used to analyze institutions that emerge in simulated environments,
as well as highlighting the utility of the IG in the parameterization of
agent-based models (Ghorbani & Bravo, 2016), and to analyze endoge-
nous emergence of “institutions in use” (Frantz et al., 2015b). In the
second camp of IG research, scholars have used it to study the design of
public policies. In the context of this pursuit, these scholars demonstrate
how policy design concepts can be operationalized using the Grammar’s
syntactic components, identify the amenability of using the Grammar to
study the kinds of highly structured statements found in policy docu-
ments, and offer practical guidance on applying the IG. While less
populated than the others, the third camp of IG research follows in
the steps of Schlüter and Theesfeld (2010) and offers critiques of the
Grammar’s syntax on an ontological basis, in general terms, but also in
connection with particular analytical applications (Frantz et al., 2013;
Frantz & Siddiki, 2021).
Given their differing objectives, as well as theoretical and method-
ological orientations, published research connected to the three different
camps of IG scholarship mentioned above has explored distinctive (albeit
not uncomplementary) kinds of research questions. One implication of
this is that the applicability of insights derived from studies in one camp to
studies in others is not always readily apparent. For example, for compu-
tational social scientists who are interested in modeling behavior in silico
within the context of institutionally derived constraints, or using the IG
syntax to characterize a limited set of emergent institutions, compre-
hensive guidance on how to manually code the entirety of statements
comprising a public policy, or how to leverage syntactic components
toward the measurement of policy design concepts, may be of limited use.
Further, for policy scholars who are primarily interested in capturing insti-
tutional information from highly structured statements found in policy
texts (i.e., only in the design, not application or interpretation, of formally
conveyed institutions), challenges to the ontological integrity of the IG
syntax that relate to how institutional statements are interpreted by policy
actors clearly may not be perceived to be that salient, or at least not
prohibitive.
Importantly, recent publications signal an impending paradigm shift
in the trajectory of IG scholarship. The hallmark of this recent scholar-
ship is that it integrates research and insights from the three camps of
IG researchers; in particular, through the engagement of computational
methods in the study of policy design, which, for some computational
applications, begs improvements in the ontological consistency of the
IG syntax (i.e., logically unambiguous definition of the meaning of
syntactic components) (see Chapter 3 for elaborated discussion) and
syntactic extensions to accommodate heterogeneous structures of state-
ments. Some use computational methods for automated coding of policy
design (Heikkila & Weible, 2018; Rice et al., 2021). Rice et al. (2021)
and have developed software specifically for the automated IG coding of
policy documents. In doing so, not only have they expanded analytical
opportunities relating to the application of the IG, they have developed
the first natural language processing software variant that is specifically
1 INTRODUCTION 5
suited to capturing the structure and semantics of the kinds of regula-

tive language unique to policy texts. Other natural language processing
software is well suited to parse and identify natural language observed in
prose, oral communication, and the like.
Other recent research extends previous scholarship that speaks to the
opportunity of using the IG for agent-based model parameterization by
offering empirical validation. Siddiki and Frantz (2019), for example, rely
on IG-coded policy data to identify actors and related choice sets and
outcomes in a simulated organic farming setting. Taken altogether, this
scholarship makes clear that accuracy in classification of policy text along
the IG’s syntactic components, as well as usability of IG-coded data in
model parameterization, will be enhanced with greater precision in the
definitions of its syntactic components, as well as the ability to opera-
tionally parse these syntactic components in more granular terms—both
of which are essentially linked to improved ontological consistency.
The kind of conceptual, methodological, and use case integration
observed in recent research, which effectively bridges the three aforemen-
tioned camps of IG scholarship, is notable for at least two reasons. First,
it can lead to further exchange among scholars currently using the IG to
guide their research, while at the same time fostering collaboration among
scholars from different disciplines. This integration may also encourage
uptake of the IG among social and computer scientists insofar as it high-
lights the versatility of the analytical use. The policy use case observed
in recent IG research may be particularly appealing to computer scien-
tists and computational social scientists seeking to develop empirically
grounded models. Engagement of computer scientists can bring to light
computational approaches and novel analytical features that can support
policy scholars in their study of policy design. In effect, IG-coded policy
data can provide the basis upon which to build simulations, let alone
establish a more robust foundation for the general computational and
logical treatment of IG-coded information.
But while recent scholarship has laid the groundwork for supporting
the possibilities highlighted here, we argue that significant progress from
this point on cannot be made without (i) resolution of ontological incon-
sistencies miring the existing IG syntax; (ii) adaptation of the IG to more
accurately account for heterogeneous structures of statements commonly
observed in written, and even orally communicated, institutions; and (iii)
adaptation of the grammatical form to make it computationally tractable.
Fundamentally, we argue that a revised specification of the IG is needed
at this time, and it is this very argument that motivates this text. In
this book, we present a revised specification of the IG, which we refer
to as the “Institutional Grammar 2.0,” or IG 2.0, signaling a structural
and paradigmatic shift to a New Institutional Grammar. Relatedly, we
provide a comprehensive description of the IG 2.0 and operational guide-
lines to support its usage. We also provide guidance on how to analyze
institutional data that has been coded according to the IG 2.0. Finally,
throughout the text, the IG 2.0 is contextualized with reference to the
existing – or original – IG, and related institutional analysis approaches
and scholarship.
In the remainder of this chapter, we (i) provide a brief introduction to
institutional analysis, describing what it is and how it is approached and
used by scholars with different disciplinary orientations; (ii) provide an
overview of the institutional analysis framework in which the IG embeds,
called the Institutional Analysis and Development (IAD) framework; and
(iii) offer a more detailed description of the IG as presented by Crawford
and Ostrom in 1995 and revised since then. We conclude the chapter
with an outline of this book.
1.1 What Is Institutional Analysis?

Broadly, institutional analysis refers to the study of institutions that govern
social systems. As stated earlier, institutions are defined as rules, norms,
and shared strategies that regulate behavior or constitute features of
governed systems; for example, by defining artifacts relevant within a
particular system, establishing positions that actors within a system can
hold, or establishing entities or venues in which system actors can interact.
In doing so, institutions foster regularity in behavior, which is regarded as
critical for enabling effective collective action in societies. Behavioral regu-
larity, in supporting understanding of what actors can expect of their own
and others’ actions, can minimize the transaction costs of decision-making
and interaction (Williamson, 1975), minimize the cognitive load associ-
ated with decision making (Simon, 1955), foster the generation of social
capital and reciprocity (Ostrom & Walker, 2003), and enable the efficacy
of monitoring and enforcement mechanisms (Ostrom, 1990).
Institutional scholarship acknowledges that institutions can be either
formal or informal in kind. Formal institutions (also referred to as insti-
tutions in form throughout this book) are institutions that result from
institutional decision-making processes engaged by recognized authorities
1 INTRODUCTION 7
(e.g., a public policy that results from a policy-making process engaged

by a legislature), and often codified in written form. Other examples of
formal institutions are organizational bylaws, treaties (Brady, 2020), or
written rules developed to govern online communities (Frey & Sumner,
2019).
Informal institutions (also referred to as institutions in use) are those
represented in social conventions or cultural habits—strategies, norms,
or rules that build on internal or decentralized social enforcement and
may not be formally codified (Ullmann-Margalit, 1977). Most of the
time, social systems are simultaneously governed by arrays of formal
and informal institutions that can vary in their extent of congruence
(Helmke & Levitsky, 2004), relative salience, and interactivity (i.e.,
dynamic development) over time (North, 1990). A single institution—
i.e., public policy or social convention pertaining to a particular topic—
can be comprised of one or more institutional statements that individually
detail what actors are required, permitted, or allowed to do within partic-
ular constraints, or constitute entities with varying degrees of specificity
and to varying extents.
Scholars engaged in the institutional analysis have been particu-
larly interested in identifying the presence and features (i.e., design)
of institutions, analyzing their emergence and/or antecedents, assessing
outcomes linked to institutionally governed behavior, and studying
instances and explanations of institutional non/compliance. Among the
various approaches used by institutional scholars for data collection
and/or analysis are formal modeling, historical and longitudinal case
studies, laboratory experiments, ethnography, interviews, and surveys.
This chapter will briefly review institutional analysis as rooted in different
disciplinary and methodological traditions.
1.1.1 Institutional Analysis by Discipline

In this section, we briefly review how institutional analysis is conducted by
scholars of public policy and administration, political science, economics,
sociology, social psychology, law, philosophy, and computer science. While
scholars of different disciplines emphasize distinctive factors in their
studies, rely on different epistemological assumptions, and/or engage
different analytical tools, there does seem to be overlap in the broader
institutional perspectives that motivate their work. Three institutional
perspectives predominate in institutional studies: historical institution-

alism, sociological institutionalism, and rational choice institutionalism.
Historical institutionalism emphasizes the role of temporal processes in
institutional evolution, and particularly how institutional formation and
evolution are shaped by features of the social and political contexts
in which they embed (Greif, 2006; Thelen, 1999). The contextual
orientation of historical institutionalism is also reflected in sociological
institutionalism insofar as the latter conceives of institutions as reflecting
habits of individuals that emerge within particular contexts to serve mate-
rial and non-material aims of individuals and communities (Zafirovski,
2004). Consistent with this view is the take that institutions embody
individual preferences and cognition. As noted by Campbell (2020):
“Institutions are thus repositories of taken-for-granted cognitive schemata
that shape people’s understandings of the world they live in and provide
scripts to guide their action.”
Whereas both historical and sociological institutionalism adopt an
endogenous orientation in their conception of institutions, rational choice
institutionalism embraces both endogenous and exogenous perspectives.
Taking each in turn, rational choice institutionalists conceiving of institu-
tions in endogenous terms see them as emerging in practice over time
and reflective of individuals’ strategic and utility-maximizing decisions
(Shepsle, 2009). Those conceiving of institutions in exogenous terms
view them as generated outputs specifically designed to shape behavior.
Recalling North (1990, p. 3)’s popular definition of institutions that
accords with this perspective, “[institutions are] … the rules of the game
in a society or, more formally … the humanly devised constraints that
shape human interaction.” An underlying presumption is that behavior
compelled by the imposition of devised institutions will in some way be
social welfare enhancing.
1.1.1.1 Public Policy and Administration

Public administration and policy scholars have dedicated substantial
effort to understanding the design, function, and impacts of institu-
tions engaged in various aspects of governance. The following discussion
provides a brief overview of distinctive tracts of institutional research
within public administration and policy, particularly policy process, schol-
arship. Extant public administration and policy scholarship in which
institutions are conceived of as behavioral constraints, as opposed to
organizations, draws inspiration from all of the dominant institutional
1 INTRODUCTION 9
perspectives. Within public administration research, the study of institu-

tions orients on theorizing and evaluating their role in structuring the
delivery of public goods and services as well as administrative settings, and
guiding the behavior of administrative actors that act within these settings.
With respect to administrative settings, institutions are assessed in terms
of their constitutive function; identifying the structural and procedural
boundaries of administrative settings in which a variety of deliberative
(i.e., collective decision-making) and operational (i.e., day-to-day prac-
tices) activities will occur (Toonen, 1998). With the latter, institutions
are investigated with respect to their specific role in shaping operational
activities, which can come to embody “administrative styles.”
Extending beyond behavior, public administration scholars highlight
the role of institutions in shaping how actors within administrative
environments interpret problems and solutions. Fundamentally, research
linked to this line of inquiry posits that understanding what occurs at an
operational level, as reflected in regularized patterns in behavior, requires
a complementary understanding of what is happening at higher levels of
decision making and activity. Public administration scholars also highlight
the relevance of institutions within the broader study of administrative
reform and change (Barzelay & Gallego, 2006; Howlett, 2003). An
important subset of this research, rooted in a sociological perspective,
examines how organizations adapt their form and practices over time in
accordance with community norms (DiMaggio & Powell, 1983).
Whereas the study of institutions within public administration schol-
arship focuses primarily on how institutions structure administrative
environments and govern the behavior of actors operating within them,
the study of institutions within public policy scholarship focuses on how
and why they emerge and change, their role in governing individual and
collective decision-making and behavior, and their impacts on broader,
systemic outcomes. As aptly summarized by Cairney (2012), those inter-
ested in studying institutions have exhibited an enduring interest in
responding to the following types of questions: What is an institution?
How does an institution influence individual behavior? How does an
institution become established in the first place? How does an institu-
tion change? How does institutionalism inform comparative public policy;
that is, how do institutions explain country-level differences? Policy
scholars adopting a historical institutional perspective, have posited and
responded to various questions linked to better understanding institu-
tional trajectories, such as how do previously enacted institutions shape
future institutions, politics, resource distributions, political learning, and

attitudes, and what are the reciprocal effects relating thereto (Mettler &
Sorelle, 2018). But while path dependency is a key underlying dimen-
sion of theories rooted in historical institutionalism, specific operational
guidance on how to model or otherwise analyze this dimension though
empirical application is limited. Policy scholars interested in the historical
trajectories of institutions have drawn heavily on the research of political
scientists focused on the same. The following section further describes
how political scientists have engaged in the study of institutions.
1.1.1.2 Political Science

Political scientists’ research on institutions has focused primarily on
contextualizing institutions; that is, identifying how institutions emerge
and develop in relation to context-specific social, political, and economic
factors. Similarly, political scientists have also been keenly interested in
identifying and understanding the political implications of the trajec-
tories along which institutions develop over time (Mettler & Sorelle,
2018; Skocpol, 1995). Another branch of political science research on
institutions focuses more specifically, on how institutions fit into decision-
making calculi of political actors. Yet a third branch of political science
research focuses specifically on the role of institutions in governing collec-
tive action in political processes (Olson, 1965; Ostrom, 2005). This
research details individuals’ motivations and incentives for engaging in
collective action in political processes and the role of institutions in
overcoming common collective dilemmas. Policy scholarship and polit-
ical science scholarship focused on assessing institutional dynamics have
considerable topical overlap.
1.1.1.3 Economics
Economists have made important contributions to the study of insti-
tutions, focusing primarily on how institutions influence the decision-
making and behavior of individuals, as well as how institutions shape
markets more broadly given their role in defining features of markets
and in governing collective action occurring with markets (North,
1990). More generally, the study of New Institutional Economics
(Hodgson, 2006, 2019; North, 1991) has aimed at developing explana-
tory approaches to analyze institutions at different scales; drawing on
the microscopic transaction cost perspective (Williamson, 1975), as well
as perspectives that support investigation of the effect of institutions
1 INTRODUCTION 11
on prosperity outcomes at macro scales, with increasing reference to

contextual factors such as resource governance that define or affect pros-
perity outcomes (e.g., Acemoglu & Robinson, 2012; Diamond, 1997).
Other aspects, interlinked with focus on historical institutionalism include
assessments of path dependence to support the explanation of modern
institutional arrangements (Greif, 2006; Milgrom et al., 1990), as well
as to motivate the evolution of institutional structure and enforcement
regimes (North et al., 2009).
Rooted in assumptions of rational decision-making, economic research
on institutions views actors as utility-maximizing agents that seek optimal
decisions within informational, resource, and importantly, behavioral
constraints. The latter assumes that individuals do not pursue decisions
with complete autonomy, but rather do so within systems of exogenously
applied or endogenously generated rules that outline behavioral opportu-
nities and constraints, govern their interactions with others, and constitute
aspects of markets with which they engage in transactions. Seen through
the lens of institutional economics, exogenous rules are those that are
created and applied by a governing authority, whereas endogenous rules
are those that represent strategic equilibria in decision-making (Aoki,
2007).
1.1.1.4 Sociology
Sociological research on institutions sheds light on the role of social
conventions and cultural habits—i.e., socially communicated or tacitly
understood strategies, norms, and rules—in shaping individual and collec-
tive behavior.
A pervasive theme in sociological studies is the interaction between
micro-level actors and macroscopic institutional arrangements, recog-
nizing the mutual dependence in the shaping of new behavior, and in
consequence coordination thereof, an interaction conceptually explored
by Coleman (1990) as well as Hedström and Swedberg (1998). Whereas
policy studies more generally emphasize the importance of the formal
institutional perspective as elaborated above, the sociological study inter-
prets institutions primarily as social norms (Ullmann-Margalit, 1977) that
emerge, and evolve based on continuous interaction between micro-level
entities, pre-existing or established formal institutions, and the associ-
ated frictions. Given the important role of theory generation in this
field, the study of the cognitive bases and social processes by which
norms, and, in extension, institutions emerge (Bourdieu, 1977; Giddens,
1984), reflect an essential linkage to the disciplines referenced above, but,

combined with the consideration of legal studies, lay the foundation for
the computational study of institutions more generally.1
Linking sociological applications to organizational studies specifically,
scholars also focus on the role of institutions in shaping forms and
practices within social organizations; for example, community or govern-
ment organizations. Sociological research on government organizations
applies sociological theory to the study of bureaucracy, contributing to
the advancement of public administration research on such. Institutional
perspectives gained prominence in the study of social organizations during
the 1990s, in part reflecting broader interest in institutions across the
social sciences, and in part as a reaction to the way that organizational
research was developing prior to this period. Leading up to the 1990s,
organizational scholars were drawing heavily on economics-based perspec-
tives, which emphasized the role of financial and resource constraints
in shaping organizational form and practices (Senge, 2013). Limited
attention was given to the role of cultural habits and social conven-
tions in shaping organizational dynamics, on which sociologists were well
positioned to contribute scientific insights.
1.1.1.5 Social Psychology

Social psychologists merge assumptions according to sociological institu-
tionalism and rational choice institutionalism with theories of cognition
and psychology to investigate, for example, how institutions designed
to govern social systems and interactions interface with cognitive and
psychological factors to inform individual decision-making and behavior,
as well as how cognitive and psychological factors inform how individ-
uals respond to (e.g., comply with) institutions that govern social systems
(Deci & Ryan, 2015). Recent social psychology research suggests ways
to extend the prevailing conception of rational choice decision-making
to offer a more complete accounting of the cognitive and psycholog-
ical factors that shape how individuals make decisions in institutionally
governed domains. For instance, this includes the effectiveness of enforce-
ment based on its facilitative vs. punitive nature (May, 2004), let alone
the traditional studies on the influence of authority to guide enforcement
or compliance behavior (Burger, 2009), as well as the general cognitive
1 Section 4.2.3 provides an extended discussion of the sociological perspective on

institutions.
1 INTRODUCTION 13
biases that influence individuals’ decision-making to an extent that it may

appear detached from any rationality (Kahneman & Tversky, 1972).
Consequently, research in this direction emphasizes the perceived
value of developing and/or implementing institutions that foster social
cooperation toward the resolution of social dilemmas in an individual’s
decision-making calculus (DeCaro, 2019). As well, it emphasizes ways
that institutions, by their design, can respond to fundamental human
needs (e.g., procedural justice) and thereby support the attainment of
institutional objectives (DeCaro et al., 2015).
1.1.1.6 Law
Legal scholarship has much to offer toward the study of institutions
generally, given that laws are ubiquitous kinds of formal institutions used
to govern social systems, but also specifically for institutional studies
employing the Institutional Grammar, given that legal scholars follow
diverse (and at times politicized) traditions of analysis (Huhn, 2014) that
are fundamentally focused on the content of the law. Given the traditional
appeal to interpret and apply laws in terms of sets of rules and the accurate
characterization of context (e.g., to reflect the activation or termination
of applicability of legal provisions), scholars in the area of legal studies
have an intrinsic interest in the formal characterization of law, including
the legal-theoretical perspective (e.g., Katz et al., 2021), a tradition that
has continued and is reflected in the subfield of Legal Informatics. Legal
Informatics (Katz et al., 2021) is particularly relevant in the context of
this book given that it addresses the use of computational methods for
assessing laws.2
Institutionally oriented legal scholars are principally interested in
understanding the design of laws, the circumstances under which laws
were devised (and corresponding contemporary interpretation), as well
as investigating conformance, or the lack of, among de jure and de
facto law (Cole, 2017), essentially focusing on the study of their effect.
Relating to institutional analysis are recent efforts in comparative legal
scholarship focused on the development of quantitative approaches for
measuring linguistic features of laws (Cooter & Ginsburg, 2005). These
approaches are considered valuable for comparative institutional anal-
ysis, insofar as they accommodate a comparison of values representing
2 Specific techniques and research directions are discussed below in the context of
Computer Science perspectives on institutions.
linguistic characteristics of laws found in different settings. This interest in

computational analysis of legal text is also observed outside of the compar-
ative legal analysis domain (Ceci et al., 2012). In step with syntactic
innovations presented in this book, some scholars interested in the use
of computational methods in legal research have given particular atten-
tion to representing linguistic features of constitutive (versus regulative)
institutions to accommodate machine interpretability (Ceci et al., 2018).
1.1.1.7 Philosophy
Philosophers have made important contributions to the study of insti-
tutions. Given the focus of this book, this brief overview of scholarship
on institutional analysis will highlight the contributions of philosophers
of language in particular (Austin, 2011; Searle, 1969). Philosophers of
language are interested, among other topics, in the bases and logic of
linguistics (Lycan, 2018). Relevant for the discussion in this book is
scholarship that examines how language is captured in different types
of “rules” (i.e., institutions) (Midgley, 1959). This scholarship highlights
fundamental differences in the meaning of rules that describe what may
or may not be done (i.e., regulate behavior), and rules that constitute
objects or behaviors within a system, essentially by defining what these
objects and behaviors count as (Grossi et al., 2006; Searle, 2018). In
addition to differentiating the meaning of rules that regulate behavior
from rules that constitute aspects of systems, Midgley (1959) also noted
the inherent non-violatability of the latter. Searle (2018, p. 52) highlights
the varying syntactic forms of regulative and constitutive rules; noting
that regulative rules typically exhibit a syntactic structure accordant with
an imperative form, whereas constitutive rules typically exhibit a declar-
ative syntactic structure of the form “X counts as Y,” or “X counts as
Y in context Z.” Searle also distinguished between brute facts and insti-
tutional facts; defining the former as facts that exist independently of
any institution, and the latter as facts within the context of a particular
institutional setting, and often expressed through constitutive rules. This
philosophy scholarship that addresses assumptions and logic of linguistic
meaning, linkages between linguistic meaning and syntactic structure, and
the interpretation of institutional meaning is foundational to the “New
Institutional Grammar.” 3
3 An extended discussion of the philosophical challenges is provided in Sect. 5.2.

1 INTRODUCTION 15
1.1.1.8 Computer Science

The final disciplinary orientation called out in this initial chapter is the
computational study of institutions. As many of the areas referenced
above, the role of computer science in the study of institutions is diverse,
but features cross-sectional linkages with these very areas.
Foregoing the general opportunities that computation offers more
generally (e.g., analytics, automation of data processing), these subfields
differentiate based on the focus of the operational application of institu-
tions in the context of socio-technical systems, the analysis of institutional
systems as they simulate the exploration of decentralized coordination
mechanisms, or variably explore formal properties of normative or legal
rule systems.
Discussed under the umbrella term mechanism design are approaches
that embed institutional features into software systems, and more specifi-
cally, socio-technical systems aimed at directly interacting with humans,
designed to display social behavior, or to operationalize legal institu-
tional rules. The approaches first dedicated to the explicit representation
of institutional concepts are captured under the label Electronic Institu-
tions (Noriega, 1997; Rodríguez-Aguilar, 2001), essentially translating
existing institutional arrangements into formal systems representing the
institutional features observed in real systems. The focus on the inter-
action with human actors, or resemblance of social systems, is reflected
by drawing on the principles of multi-agent systems (Ferber, 1999;
Shoham & Leyton-Brown, 2014). Contemporary developments loosely
linked with this tradition is the implementation of institutions in the form
of smart contracts (Szabo, 1997) as part of distributed ledger systems (see
e.g., Frantz & Nowostawski, 2016; Zheng et al., 2020), and the principles
of algorithmic governance (Katzenbach & Ulbricht, 2019) that aim at
representing conceptual and practical alternatives to automate, monitor,
or socialize governance functions assumed by human actors.
Evolving from the mechanistic view on institutions, the role of social
coordination mechanisms has found recognition in a set of approaches
presented under the label Normative Multi-Agent Systems (Boella et al.,
2006). Approaches under this label capture both the computational
exploration of normative processes, such as norm emergence and evolu-
tion, with variable focus on the cognitive or social bases, as well as a
focal emphasis on the norm concept itself, as captured in the concept
of norm lifecycles (Frantz & Pigozzi, 2018). The formal interpreta-
tion of norms is referenced as Norm Change (Boella et al., 2009) and
draws on fundamental principles of logic and argumentation to analyze

normative systems with focus on obligations and prohibitions, as well
as transitions between different normative statements, from a theoret-
ical perspective. The computational analysis of institutions in the form
of agent-based models (Epstein, 2007) (an approach explored in greater
detail in Chapter 2 and Sect. 8.2.2) intersects with this particular field, but
puts stronger emphasis on the empirical study of domain-specific scenarios
and the coordinative functions norms or rules play therein. Hence, unlike
the theoretical orientation of Normative Multi-Agent Systems, compu-
tational social scientists often identify as social scientists based on the
social-scientific nature (including theory and method) of their research.
The third and last approach referenced in this brief overview is the
study of formal institutions of a primarily legal kind. Laying the foun-
dations for both the implementation of institutional concepts in socio-
technical systems referenced above (e.g., information systems, robotics),
but also contributing to the legal-philosophical development of insti-
tutional concepts, this field is broadly captured under the label Legal
Informatics (Katz et al., 2021). Linked to reasoning techniques associated
with traditional Artificial Intelligence approaches as well as the modern
incarnation as data-centric Machine Learning techniques, a purview of
legal informatics is to extract legal information (e.g., policy documents),
and its logical treatment to accurately capture contextual interpretation
(e.g., based on temporal logics, event calculi). Objectives in this field
can be practical in kind (e.g., facilitating the parsing of legal texts) or
inherently theoretical [e.g., classification of legal statements (Ceci et al.,
2018), formal verification of legal constructs (Katz et al., 2021)]. Efforts
with relevance to the IG is the development of standards for the formal
representation (e.g., Palmirani et al., 2018) and computational interpre-
tation of law (e.g., Athan et al., 2015; Palmirani et al., 2011). Scholars
in this field often possess interdisciplinary backgrounds, involving logic,
philosophy as well as legal studies.
1.1.2 Convergence Toward Interdisciplinary Institutional Analysis

The purpose of the preceding discussion is in part to convey the salience
of institutional analysis across major social science disciplines. It is also in
part to convey the considerable overlap in core ideas explored by insti-
tutional analysts drawing on differing disciplinary perspectives, as well
as different institutional perspectives, some of which have found broad
1 INTRODUCTION 17
coverage in extant IG research, whereas others have found little. As will

be discussed in more detail in Chapter 2 of this book, these ideas are
also prominently showcased in Institutional Grammar research conducted
to date, and we demonstrate throughout this book that this explo-
ration is further supported by conceptual and methodological refinements
presented under the label Institutional Grammar 2.0, or IG 2.0.
Recounting these core ideas here. First, institutions are understood
to be exogenously and endogenously generated. Exogenous institutions,
are typically generated by a government authority, and are often repre-
sented in written form, such as in constitutions, laws, and regulations.
Endogenous institutions are generated through practice and experience,
reflecting individuals’ patterned behaviors, including behavior in response
to social queues and constraints. The formation of endogenous insti-
tutions is based on processes of experiential or social learning, paired
with informal enforcement that operates based on the ongoing social-
ization and internalization of the observed and experienced behavior.
Endogenous institutions typically present in the form of cultural habits
and social conventions, and social norms more specifically. Exogenous
institutions tend to be implemented, monitored, and enforced by the
governing authority that generated them. Endogenous institutions are
usually socially monitored and enforced.
A second core idea commonly represented in extant social science
research on institutions (and in part implied by the first one), is that
exogenously and endogenously generated institutions are subject to
change. Institutions of either sort are understood to evolve over time,
in more or less incremental ways in response to changes in aspects of
the social systems in which they embed. A third core idea represented
in social science research on institutions, augments the first and second
relating to the generation and evolution of institutions, respectively, by
highlighting the contextualized nature of both. Essentially, institutional
scholars recognize that institutions situate within contexts with partic-
ular social, political, economic, environmental, and other characteristics,
and context has implications for the specific design of institutions and
how they change over time. A fourth, somewhat related, idea explored
in extant institutional scholarship is that institutions shape decision-
making, behavior, and concomitant outcomes, alongside various personal
(i.e., cognitive), social, and contextual factors (Siddiki, 2014). Finally,
institutions have varying functional and structural properties.
In the following section, we provide an overview of one institutional

analysis framework, that explicitly incorporates the core ideas expressed
above, and related disciplinary insights, called the Institutional Analysis
and Development Framework. The relevance of this framework within the
context of this book is that it is the analytical framework in which the
Institutional Grammar is grounded.
1.2 Institutional Analysis with the Institutional

Analysis and Development Framework
The Institutional Analysis and Development Framework, or IAD frame-
work, is one of the leading analytical approaches for studying institutional
design, development, and outcomes (Ostrom, 2005). Linking to the
preceding discussion, one of the appeals of the framework is its interdis-
ciplinary foundations. Concept, theory, and methods linked to the IAD
framework are grounded in research from the fields of public adminis-
tration, political science, economics, psychology, and others. Borrowing
insights from these various fields, the framework accommodates explo-
ration of micro-, meso-, and macro-level institutional phenomena under
one analytical lens. In more specific terms, leveraging insights from
various disciplines, the framework articulates a model of individual
decision-making (micro level), factors hindering and enabling collective
action (meso level), and systemic factors shaping individual and collective
decision-making and behavior. Furthermore, the framework guides inves-
tigation of how micro-, meso-, and macro-level factors configure to shape
outcomes within systems with social and environmental features.
The IAD framework conceives of social systems as comprised of one
or more “action situations” that vary in their degree of interdependence.
An action situation is defined as a setting in which two or more actors
“are faced with a set of potential actions that jointly produce outcomes”
(Ostrom, 2005, p. 32). What occurs within action situations reflects the
institutions that are opportuning or constraining actors’ actions and inter-
actions. An institution itself is made up of one or more institutional
statements (described further in the following discussion) that indepen-
dently or configurally influence aspects of one or more action situations
(Siddiki et al., 2019).
To guide the institutional analyst’s understanding of institutions that
shape what happens within action situations, the IAD framework posits
a “rule typology” (Ostrom, 2005). Included within this rule typology
1 INTRODUCTION 19
are seven types of rules that accord with institutional functions pertaining
to actors and actor interactions within the context of action situations
(i) Position rules specify positions that actors can occupy within an
action situation; (ii) Boundary rules specify eligibility criteria for occu-
pying these positions; (iii) Choice rules specify operational actions linked
to actors occupying certain positions; (iv) Scope rules specify intended
goals or situational outcomes; (v) Information rules specify channels of
information flow; (vi) Aggregation rules specify guidance on collective
decision-making; and (vii) Payoff rules specify incentives tied to partic-
ular actions. These rules may operate configurally to guide individual
and collective behavior within action situations. Importantly, this rule
typology, that captures specific functions that different types of rules serve
within action situations, also eludes to meta-institutional functions that
different rules serve within action situations. As reflected in the defi-
nitions of the different rule types, some rules play a constituting, or
parameterizing, role in the context of action situations. For example, posi-
tion rules constitute positions that actors can hold. Some rules regulate
the behavior of actors within action situation. For example, choice rules
specify specific actions assigned to actors. This observation suggests the
value in considering the wider functions that rules play in relation to
action situations—whether they define the features of action situations
that actors act upon or in relation to, or whether they define actions in
the first place.
Because action situations are a focal unit of analysis under the IAD
framework, and what happens within action situations is presumed to be
largely shaped by institutions, the IAD framework offers multiple concep-
tual and methodological approaches for analyzing institutional design
that variably relate to the rule typology described above. One of these
approaches is the Institutional Grammar, which is described in more detail
in the following section.
1.3 Primer on the Institutional Grammar

As basis for the discussion of a revised specification of the Institutional
Grammar in this book, we provide an overview of the approach as orig-
inally conceived. Understanding of the original conception of the Insti-
tutional Grammar is important as the IG 2.0 as introduced in this book
invites for broader disciplinary adoption based on a shared perspective
on institutions, amounting to a new paradigm for conducting institu-

tional analysis using the IG. Specifically, it extends the original conception
toward improving its ontological consistency, ability to comprehen-
sively capture heterogeneous statement structures, and computational
tractability. Despite urging a paradigmatic shift with the referenced refine-
ments, the IG 2.0 should be seen as a complement rather than a substitute
to the original IG. Hence, an understanding of the original conception
is critical. It is also important for grounding the discussion of existing
research offered in Chapter 2 of this book, as extant research is almost
entirely based on the original conception. Yet further, this background is
crucial for contextualizing the refinements to the original IG presented
here.
As noted earlier in this introductory chapter, one of the primary
motivations of the Institutional Grammar was to lend clarity on the
definition of institutions through reference to institutional features, and
relatedly provide a systematic way of differentiating among different types
of institutions commonly analyzed by institutional analysts. Underlying
Crawford and Ostrom’s approach for defining institutions was an assump-
tion that institutions are comprised of abstractable and generalizable
constituent parts, and thus ancillary to the definition of institutions is
identification of these constituent parts. Crawford and Ostrom harkened
the analogy of genetic structure, referencing the value of understanding
the genetic code of institutions as basis for understanding their design
and influence on human decision-making and behavior. As institutions
are, or can be, captured in written language, a logical conceptualization
of the abstractable and generalizable parts of institutions was as syntactic
components that convey institutionally relevant meaning and configure
within an “institutional grammar.” This is in the same way that parts
of speech mapping to syntactic labels combine within, for example, the
English grammar. Within English grammar, parts of speech configure to
form statements, and according to grammar rules, there is a minimum set
of parts of speech needed to form a complete statement. In an analogous
way, in the Institutional Grammar, the focal unit of analysis is the “insti-
tutional statement,” that is comprised of some configuration of syntactic
components, some of which must necessarily be present. In this way,
these syntactic components are “necessary components of institutional
statements,” while others are deemed “sufficient components.” Crawford
and Ostrom defined an institutional statement made up of these various
syntactic components in the following way: “[An] institutional statement
1 INTRODUCTION 21
refers to a shared linguistic constraint or opportunity that prescribes,

permits, or advises actions or outcomes for actors (both individual and
corporate). Institutional statements are spoken, written, or tacitly under-
stood in a form intelligible to actors in an empirical setting” (Crawford &
Ostrom, 1995, p. 583).
The original presentation of the Institutional Grammar identified five
syntactic components of institutional statements: (i) Attributes: the actor
(individual or corporate) to whom an institutional statement applies. The
actor characterized in the Attributes component performs the action, or
set of actions, indicated in the statement; (ii) Aim: the action(s) that
the Attribute is linked to; (iii) Deontic: a prescriptive operator (e.g.,
must, may, must not) that indicates whether the action identified in the
statement is required, allowed, or forbidden; (iv) Condition: a temporal,
spatial, procedural, or other, condition that qualifies the action of the
statement; and (v) Or else: a payoff associated with performing, or failing
to perform, the action of the statement. These, and additional compo-
nents which have since been added to the Grammar, are often referenced
with their acronyms: Attributes (A), Aim (I), Deontic (D), Conditions
(C), and Or else (O). Referring again to the notion of necessary and suffi-
cient components of institutional statements, an institutional statement
must at least contain an Attributes, Aim, and Conditions component
to qualify as such. Thus, Deontic and Or else components are consid-
ered sufficient components. Drawing terminologically and theoretically
on game theory, institutional statements containing the necessary compo-
nents of Attributes, Aim, and Conditions are characterized as shared
strategies. Statements containing these necessary components and also a
Deontic are characterized as norms. Statements containing all five of the
referenced syntactic components are referred to as rules.
Two notable refinements to this list of syntactic components have been
published in the last ten years. Both of these refinements seem to have
been accepted by the Institutional Grammar user community, as indi-
cated by their incorporation into published research. The first of these
refinements was the introduction of the Object syntactic component by
Siddiki et al. (2011), referenced with the acronym B. The second refine-
ment was a reconceptualization of the Or else component by Frantz et al.
(2013). Each is discussed in turn. The Object component was intro-
duced by Siddiki et al. with the related aims of improving the ontological
consistency of the Institutional Grammar syntax, while, at the same time,
extending the extent to which institutional information can be captured,
as well as enabling more reliable institutional classification. In the absence

of a syntactic field to capture institutional statement clauses capturing
receivers of actions indicated in statements, these clauses were erroneously
captured within the Aim (i.e., statement action) or Conditions (i.e., qual-
ifier of statement action) fields.4 Siddiki et al. developed this syntactic
extension based on a study of public policies. In so doing, they built on
the work of Basurto et al. (2010) who offered specific operational guide-
lines for applying the Institutional Grammar to study the design of public
policies.
Rather than proposing a syntactic extension like Siddiki et al., Frantz
et al. (2013) introduced a syntactic modification in their presentation
of the Nested ADICO, or nADICO, concept. nADICO urges recon-
ceptualization of the Or else component of the Institutional Grammar
syntax to accommodate both conceptual and empirical observations. In
doing so, the nADICO conceptualization accommodates more detailed
description of institutions, while at the same time supporting analysis of
complex forms of institutional statements. Among the core features of the
nADICO concept, reflected in its labeling, is to treat institutional clauses
according with the Or else component under the original Institutional
Grammar as nested statements . This suggestion derives from empirical
observation. In many statements in which an Or else clause is present, that
clause is typically structurally and semantically akin to a complete institu-
tional statement. However, because the clause is closely tied (or directly
follows) from clauses according with ABDIC5 components in the state-
ment in which it is found, the notion is to capture this linkage through a
nesting conception. In this nesting conception, the statement in which the
Or else clause is initially encountered is treated as the monitored statement
and the Or else clause, which is subsequently represented as a complete
statement is referred to as a consequential statement . For illustration of
the nADICO concept, consider the following example statement, which
would be treated as one complete institutional statement under the orig-
inal Institutional Grammar: “Organic farmers must accommodate review of
records of farming practices during inspections, or else certifiers may suspend
4 To distinguish conditions from action qualifications, in computational studies quali-

fiers have been represented as properties of activities, rather than being conflated with
conditions (Frantz et al., 2015b).
5 The ABDIC acronym reflects the conceptual inclusion of the Object component by
Siddiki et al. (2011).
1 INTRODUCTION 23
organic farming certification.” This statement would be parsed as follows

along Institutional Grammar syntactic components:
Attributes = organic farmers

Deontic = must
Aim = accommodate
Object = review of records of farming practices
Condition = during inspections
Or else = or else certifiers may suspend organic farming certification
This example clearly elucidates the noted observation that clauses

linked to the Or else component often take the form of complete insti-
tutional statements. nADICO thus urges the delineation of this one
statement into two, with the monitoring statement, “Organic farmers
must accommodate review of records of farming practices during inspec-
tions” and nested consequential statement, “Certifiers may suspend organic
farming certification.”
The reconceptualization of the Or else component under the nADICO
also bears conceptual implications, particularly as related to the strategies,
norms, rules distinction posited under the original Institutional Grammar.
As previously described, strategies, norms, and rules as variant forms of
institutional statements are defined in terms of the presence or absence of
syntactic components, or rather clauses associating therewith. Norms and
Rules are differentiated by the presence of Or else clauses in the latter.
This differentiation is irrelevant under the nADICO conception, since
the Or else is in effect not treated as a syntactic component at all therein,
while explicitly admitting the notion of sanctioning provisions for norms.6
The nADICO conception is central to the IG 2.0, and is revisited and
described in further detail later in this book (Chapter 4).
Before concluding this brief overview of the IG, elements of which
will be expounded upon in much more detail in the remainder of this
book, it is worth mentioning what it means to have an institutional
syntax. Generally, syntax is a representation of structure. In a linguistic, as
well as institutional, sense, syntax is an arrangement of syntactic compo-
nents. Given the reference to syntactic components that liken to parts
of speech, and institutional statements that liken to sentences, it may
6 This is an extension of Schlüter and Theesfeld (2010)’s interpretation of the norms-

rules distinction, and is further discussed in Sect. 4.2.3.
seem that the IG simply offers another linguistic syntax according to

which one can understand the language of written and spoken institu-
tions. But this characterization would be incorrect. There are distinctive
differences between the institutional syntax offered through the IG and
traditional linguistic syntax, which enable the institutional analyst to
more comprehensively and usefully capture institutional structure and
meaning, all toward understanding what institutions are and their func-
tional role within social systems. For clarification, Institutional Structure
refers to arrangements of units of language that individually and config-
urally convey institutional meaning—Institutional Semantics —and which
can be analyzed at different levels. Within the IG, generally the focal unit
of analysis is the institutional statement.7 As this issue is addressed in great
detail in later chapters of this book, we highlight here only one of the key
differences between an institutional syntax that embeds in the IG and
a linguistic syntax. Whereas the parts of speech that accord with partic-
ular syntactic components in linguistic syntaxes are reflecting a somewhat
agnostic mapping of sentence information to linguistic components, the
syntactic components within an institutional syntax are specifically meant
to pick up on generalizable types of information conveyed within institu-
tional statements presumed to have important institutional function—i.e.,
bearing on how institutions will function or perform within a particular
domain.
The presentation here of the Institutional Grammar as an institu-
tional syntax, as differentiated from a linguistic syntax, is not offered as
a technical point. Rather, it is to highlight the richness of the informa-
tion that is captured from an application of the IG for the institutional
analyst. The syntax specifically captures information that is critical to
understanding how institutions shape behavior and interactions within
social systems. And, while presenting a relevant classification schema to
enable robust and reliable representations of institutions along select
features, the IG offers no prescription on specific kinds of questions
or phenomena the institutional analyst can explore leveraging IG coded
information. Further, while grounded in institutional theory and affili-
ated with the IAD framework, the IG offers no prescription about which
specific theories or analytical frameworks analysts should apply as part of
the assessment of IG coded data. It is these various features that make
7 This characterization is refined in Sect. 4.2.2 based on the introduction of atomic

institutional statements.
1 INTRODUCTION 25
the IG of broad appeal to institutional analysts with various research

questions, methodological approaches, and disciplinary backgrounds.
Independent of background and interest, the institutional analyst can
thus fundamentally treat the IG 2.0, in particular, as an analytical
paradigm—a way of approaching conceptualization, design, and appli-
cation of analysis, where the latter is also informed by epistemological
and disciplinary orientations. The Grammar offers a conceptual and repre-
sentational basis for viewing institutions, accommodates diverse analytical
applications across disciplines and domains, and supports diverse insti-
tutional theories and types of data as well as associated methodological
approaches and analytical techniques. Basically, even though the IG is
offering shared bases for representing institutions, the analyst can apply
different concepts and theories for guiding which aspects of institutions
are represented, as well as how to interpret institutional information; can
use different methodological approaches for collecting institutional infor-
mation, and; can use different analytical techniques for processing, manip-
ulating, and evaluating institutional information as linked to conceptual
and theoretical foci.
The remainder of this text provides a detailed overview of how the
IG generally, and the IG 2.0 specifically, enable the kind of integration
and diverse application referenced in this section. The following section
outlines the structure of the book. In navigating the contents of this book,
it may be useful to make note of the intended usage of this book. First,
this book is intended to provide a comprehensive exposition of the IG
2.0. Several chapters are dedicated to describing in detail the “concep-
tual foundations” of the IG 2.0. Conceptual foundations are organized
by the three levels of expressiveness on which the IG 2.0 orients that
vary in depth and focus of analysis. The book is also intended to serve
as a reference guide. Given that a central aim of the IG 2.0 is to support
diverse utilization of different features of it, and relatedly to accommo-
date varying analytical aims, institutional analysts may find it practically
helpful to reference sections of the book that are suited to his/her analyt-
ical objectives following a complete reading of the text. Supporting the
referential quality of the book, it contains two chapters dedicated to
“applications”, which too are organized around levels of expressiveness to
support selective references. The first of these application chapters focuses
on providing guidance on coding institutional statements according to
different levels of expressiveness. The second of the applications chap-
ters is focused on showcasing different ways that institutional information
collected through an application of the IG can be analyzed. Importantly,

it should also be noted that this book is intended to serve as the concep-
tual complement to a comprehensive operational codebook developed by
the authors of this text. The codebook (Frantz & Siddiki, 2020) provides
detailed guidance on the coding of institutional data according to the IG
2.0, and includes numerous examples of coded statements for reference,
beyond the illustrative examples provided here.
1.4 Overview of Chapter Contents

Following this introduction, Chapter 2 provides an overview of extant
IG scholarship. This overview describes questions that motivate existing
IG research, frameworks and theories used to guide it, concepts explored
therein, and how institutional analysts collect and analyze institutional
data. This chapter concludes with a presentation of key opportunities and
challenges of the original IG as noted by scholars who have engaged it.
Chapter 3 builds on the discussion in Chapter 2 by elaborating on central
challenges that the IG 2.0 is designed to address. Chapter 3 further intro-
duces the conceptual background on which features of the IG 2.0 build.
Chapters 4–6 provide a comprehensive conceptual overview of features
associated with each of the three levels of expressiveness: IG Core, IG
Extended, and IG Logico. Chapter 7, complementing the conceptual
discussion presented in Chapters 4–6, provides operational guidance on
the IG coding process, offering example coding according to different
levels of expressiveness. Chapter 8 offers guidance on how to analyze
IG-coded data and thereby seeding stimuli for further analytical opportu-
nities. Finally, Chapter 9 concludes the book, reflecting on the IG as an
analytical paradigm and noting opportunities for research leveraging the
Institutional Grammar.
References
Acemoglu, D., & Robinson, J. (2012). Why nations fail: The origins of power,
prosperity, and poverty. Crown Business.
Aoki, M. (2007). Endogenizing institutions and institutional changes. Journal of
Institutional Economics, 3(1), 1–31. https://doi.org/10.1017/s17441374
06000531
1 INTRODUCTION 27
Athan, T., Governatori, G., Palmirani, M., Paschke, A., & Wyner, A. (2015).
LegalRuleML: Design principles and foundations. Lecture Notes in Computer
Science, vol. 9203. https://doi.org/10.1007/978-3-319-21768-0_6
Austin, J. (2011). How to Do Things with Words. Oxford University
Press. https://doi.org/10.1093/acprof:oso/9780198245537.001.0001
Barzelay, M., & Gallego, R. (2006). From “new institutionalism” to “institu-
tional processualism”: Advancing knowledge about public management policy
change. Governance, 19(4), 531–557. https://doi.org/10.1111/j.1468-
0491.2006.00339.x
Basurto, X., Kingsley, G., McQueen, K., Smith, M., & Weible, C. M. (2010). A
systematic approach to institutional analysis: Applying Crawford and Ostrom’s
grammar. Political Research Quarterly, 63(3), 523–537. https://doi.org/10.
1177/1065912909334430
Boella, G., Pigozzi, G., & van der Torre, L. (2009). Normative framework for
normative system change. 8th International Joint Conference on Autonomous
Agents and Multiagent Systems (AAMAS 2009), Budapest, Hungary, May 10–
15, 2009, Volume 1, 169–176.
Boella, G., van der Torre, L., & Verhagen, H. (2006). Introduction to normative
multiagent systems. Computation and Mathematical Organizational Theory,
12(2–3), 71–79.
Bourdieu, P. (1977). An outline of a theory of practice. Cambridge University
Press.
Brady, U. (2020). Robust conservation anarchy: Comparing treaty institutional
design for evidence of Ostrom’s design principles, fit, and polycentricity (Doctoral
dissertation). Arizona State University. Tempe.
Burger, J. M. (2009). Replicating Milgram: Would people still obey today?
American Psychologist, 64(1), 1–11. https://doi.org/10.1037/a0010932
Cairney, P. (2012). Understanding Public Policy: Theories and Issues. Palgrave
Macmillan.
Campbell, J. L. (2020). Institutional change and globalization. https://doi.org/
10.2307/j.ctv131bw68
Ceci, M., Butler, T., O’Brien, L., & Al Khalil, F. (2018). Legal patterns for
different constitutive rules. Lecture Notes in Computer Science, vol. 10791.
https://doi.org/10.1007/978-3-030-00178-0_7
Ceci, M., Lesmo, L., Mazzei, A., Palmirani, M., & Radicioni, D. P. (2012).
Semantic annotation of legal texts through a framenet-based approach. Lecture
Notes in Computer Science, vol. 7639. https://doi.org/10.1007/978-3-642-
35731-2_17
Cole, D. H. (2017). Laws, norms, and the Institutional Analysis and Devel-
opment framework. Journal of Institutional Economics. https://doi.org/10.
1017/S1744137417000030
Coleman, J. S. (1990). Foundations of social theory. Harvard University Press.
Cooter, R. D., & Ginsburg, T. (2005). Leximetrics: Why the same laws are
longer in some countries than others. SSRN Electronic Journal. https://doi.
org/10.2139/ssrn.456520
Crawford, S. E. S., & Ostrom, E. (1995). A Grammar of institutions. Amer-
ican Political Science Review, 89(3), 582–600. https://doi.org/10.2307/208
2975
DeCaro, D. A. (2019). Humanistic rational choice understanding the funda-
mental motivations that drive self-organization and cooperation in commons
dilemmas. In Hudson, B., Rosenbloom, J., & Cole, D. (eds.): Routledge
handbook of the study of the commons. https://doi.org/10.4324/978131516
2782
DeCaro, D. A., Janssen, M. A., & Lee, A. (2015). Synergistic effects of
voting and enforcement on internalized motivation to cooperate in a resource
dilemma. Judgment and Decision Making, 10(6), 511–537.
Deci, E. L., & Ryan, R. M. (2015). Self-determination theory. In International
Encyclopedia of the Social & Behavioral Sciences: Second Edition https://doi.
org/10.1016/B978-0-08-097086-8.26036-4
Diamond, J. M. (1997). Guns, germs, and steel: The fates of human societies.
Norton.
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institu-
tional isomorphism and collective rationality in organizational fields. American
Sociological Review, 48(2). https://doi.org/10.2307/2095101
Epstein, J. M. (2007). Generative social science: Studies in agent-based computa-
tional modeling. Princeton University Press.
Ferber, J. (1999). Multi-agent systems - An introduction to distributed artificial
intelligence. Addison-Wesley.
Frantz, C., Purvis, M. K., Nowostawski, M., & Savarimuthu, B. T. R.
(2013). nADICO: A nested grammar of institutions. Lecture Notes in
Computer Science, vol. 8291, 429–436. https://doi.org/10.1007/978-3-
642-449277_31
Frantz, C. K., Keertipati, S., Purvis, M. K., & Purvis, M. (2015a). Subjec-
tive interpretation of directive terms: Methodology and preliminary results.
Proceedings - 2015 International Conference on Culture and Computing,
Culture and Computing 2015. https://doi.org/10.1109/Culture.and.Com
puting.2015.40
Frantz, C. K., Purvis, M. K., Savarimuthu, B. T. R., & Nowostawski, M.
(2015b). Modelling dynamic normative understanding in agent societies.
Scalable Computing, 16(4), 355–380. https://doi.org/10.12694/scpe.v16i4.
1128
Frantz, C. K., & Nowostawski, M. (2016). From institutions to code: Towards
automated generation of smart contracts. Proceedings - IEEE 1st International
1 INTRODUCTION 29
Workshops on Foundations and Applications of Self-Systems, FAS-W 2016, 210–

215. https://doi.org/10.1109/FAS-W.2016.53
Frantz, C. K., & Pigozzi, G. (2018). Modeling norm dynamics in multi-agent
systems. IfCoLoG Journal of Logics and Their Applications, 5(2), 491–563.
Frantz, C. K., & Siddiki, S. N. (2020). Institutional Grammar 2.0 Codebook.
https://arxiv.org/abs/2008.08937
Frantz, C. K., & Siddiki, S. (2021). Institutional Grammar 2.0: A specification
for encoding and analyzing institutional design. Public Administration, 99,
222-247. https://doi.org/10.1111/padm.12719
Frey, S., & Sumner, R. W. (2019). Emergence of integrated institutions in a
large population of self-governing communities. PLoS ONE, 14(7). https://
doi.org/10.1371/journal.pone.0216335
Ghorbani, A., & Bravo, G. (2016). Managing the commons: A simple model of
the emergence of institutions through collective action. International Journal
of the Commons, 10(1). https://doi.org/10.18352/ijc.606
Giddens, A. (1984). The constitution of society. Polity Press.
Greif, A. (2006). Institutions and the path to the modern economy: Lessons from
medieval trade. Cambridge University Press.
Grossi, D., Aldewereld, H., & Dignum, F. (2006). Ubi Lex, Ibi Poena:
Designing norm enforcement in E-institutions. In P. Noriega, J. Vázquez-
Salceda, G. Boella, O. Boissier, V. Dignum, N. Fornara, & E. Matson (Eds.),
Coordination, organizations, institutions, and norms in multi-agent systems ii,
107–120. Springer.
Hedström, P., & Swedberg, R. (1998). Social mechanisms – An analytical
approach to social theory. Cambridge University Press.
Heikkila, T., & Weible, C. M. (2018). A semiautomated approach to analyzing
polycentricity. Environmental Policy and Governance, 28(4). https://doi.org/
10.1002/eet.1817
Helmke, G., & Levitsky, S. (2004). Informal institutions and comparative poli-
tics: A research agenda. Perspectives on Politics, 2(4), 725–740. https://doi.
org/10.1017/S1537592704040472
Hodgson, G. M. (2006). What are institutions? Journal of Economic Issues, 40(1),
1–25. https://doi.org/10.1080/00213624.2006.11506879
Hodgson, G. M. (2019). Taxonomic definitions in social science, with firms,
markets and institutions as case studies. Journal of Institutional Economics,
15(2), 207–233. https://doi.org/10.1017/S1744137418000334
Howlett, M. (2003). Administrative styles and the limits of administrative reform:
A neo-institutional analysis of administrative culture. https://doi.org/10.
1111/j.1754-7121.2003.tb01588.x
Huhn, W. (2014). Five Types of Legal Arguments (3rd edition). Carolina
Academic Press.
Kahneman, D., & Tversky, A. (1972). Subjective probability: A judgment of

representativeness. Cognitive Psychology, 3(3), 430–454. https://doi.org/10.
1016/0010-0285(72)90016-3
Katz, D. M., Dolin, R., & Bommarito, M. J. (Eds.). (2021). Legal informatics.
Cambridge University Press. https://doi.org/10.1017/9781316529683
Katzenbach, C., & Ulbricht, L. (2019). Algorithmic governance. Internet Policy
Review, 8(4). https://doi.org/10.14763/2019.4.1424
Lycan, W. G. (2018). Philosophy of language - A contemporary introduction (3rd
edition). Routledge.
May, P. J. (2004). Compliance motivations: Affirmative and negative bases.
https://doi.org/10.1111/j.0023-9216.2004.03801002.x
Mettler, S., & Sorelle, M. (2018). Policy feedback theory. Theories of the Policy
Process. https://doi.org/10.4324/9780429494284-4
Midgley, G. C. J. (1959). XIV—Linguistic Rules. Proceedings of the Aristotelian
Society, 59(1), 271-290. https://doi.org/10.1093/aristotelian/59.1.271
Milgrom, P. R., North, D. C., & Weingast, B. R. (1990). The role of institutions
in the revival of the trade: The law merchant, private judges, and the cham-
pagne fairs. Economics and Politics, 2(1), 1–23. https://doi.org/10.1111/j.
1468-0343.1990.tb00020.x
Noriega, P. (1997). Agent-mediated auctions: The fishmarket metaphor (Doctoral
dissertation). Universitat Autonoma de Barcelona. Barcelona.
North, D. C. (1990). Institutions, institutional change, and economic perfor-
mance. Cambridge University Press.
North, D. C. (1991). Institutions. Journal of Economic Perspectives, 5(1), 97–
112. https://doi.org/10.1257/jep.5.1.97
North, D. C., Wallis, J. J., & Weingast, B. R. (2009). Violence and social orders:
A conceptual framework for interpreting recorded human history. Cambridge
University Press.
Olson, M. (1965). The logic of collective action. Cambridge.
Ostrom, E. (1990). Governing the commons: The evolution of institutions for
collective action. Cambridge University Press.
Ostrom, E. (2005). Understanding institutional diversity. Princeton University
Press.
Ostrom, E., & Walker, J. (2003). Trust and reciprocity: Interdisciplinary lessons
from experimental research, Russell Sage Foundation.
Palmirani, M., Governatori, G., Rotolo, A., Tabet, S., Boley, H., & Paschke,
A. (2011). LegalRuleML: XML-based rules and norms. Lecture Notes in
Computer Science (including subseries Lecture Notes in Artificial Intelligence
and Lecture Notes in Bioinformatics), 7018 LNCS. https://doi.org/10.1007/
978-3-642-24908-2_30
Palmirani, M., Sperberg, R., Vergottini, G., & Vitali, F. (2018). Akoma Ntoso
version 1.0 Part 1: XML vocabulary, OASIS Standard. http://docs.oasis-
1 INTRODUCTION 31
open.org/legaldocml/akn-core/v1.0/os/part1-vocabulary/akn-core-v1.0-
os-part1vocabulary.html
Rice, D., Siddiki, S., Frey, S., Kwon, J. H., & Sawyer, A. (2021). Machine coding
of policy texts with the Institutional Grammar. Public Administration. 99,
248-262. https://doi.org/10.1111/padm.12711
Rodríguez-Aguilar, J. A. (2001). On the design and construction of agent-
mediated electronic institutions (Doctoral dissertation). Universitat Autonoma
de Barcelona. Barcelona.
Schlüter, A., & Theesfeld, I. (2010). The Grammar of Institutions: The challenge
of distinguishing between strategies, norms, and rules. Rationality and Society,
22(4), 445–475. https://doi.org/10.1177/1043463110377299
Searle, J. (2018). Constitutive rules. Argumenta, 4(1), 51–54.
Searle, J. R. (1969). Speech acts: An essay in the philosophy of language. Cambridge
University Press.
Senge, K. (2013). The ‘new institutionalism’ in organization theory: Bringing
society and culture back in. The American Sociologist, 44(1), 76–95. https://
doi.org/10.1007/s12108-012-9170-5
Shepsle, K. A. (2009). Rational choice institutionalism. The Oxford Handbook of
Political Institutions. https://doi.org/10.1093/oxfordhb/9780199548460.
003.0002
Shoham, Y., & Leyton-Brown, K. (2014). Multiagent Systems: Algorithmic, game
theoretic, and logical foundations. Cambridge University Press.
Siddiki, S. (2014). Assessing policy design and interpretation: An institutions-
based analysis in the context of aquaculture in Florida and Virginia, United
States. Review of Policy Research, 31(4), 281–303. https://doi.org/10.
1111/ropr.12075
Siddiki, S., & Frantz, C. (2019). Understanding the effects of social value orien-
tations in shaping regulatory outcomes through agent based modeling: An
application to organic farming. Presented at the Workshop on the Workshop
(WOW) 6 Conference.
Siddiki, S., Heikkila, T., Weible, C. M., Pacheco-Vega, R., Carter, D., Curley, C.,
Deslatte, A., & Bennett, A. (2019). Institutional analysis with the Institutional
Grammar. Policy Studies Journal. https://doi.org/10.1111/psj.12361
Siddiki, S., Weible, C. M., Basurto, X., & Calanni, J. (2011). Dissecting policy
designs: An application of the Institutional Grammar Tool. Policy Studies
Journal, 39(1), 79–103. https://doi.org/10.1111/j.1541-0072.2010.003
97.x
Simon, H. A. (1955). A Behavioral Model of Rational Choice. The Quarterly
Journal of Economics, 69(1), 99–118. https://doi.org/10.2307/1884852
Skocpol, T. (1995). Protecting soldiers and mothers: The political origins of social
policy in the United States. Belknap Press.
Smajgl, A., Izquierdo, L. R., & Huigen, M. (2008). Modeling endogenous

rule changes in an institutional context: The ADICO Sequence. Advances
in Complex Systems, 2(11), 199–215. https://doi.org/10.1142/S02195259
0800157X
Szabo, N. (1997). Smart Contracts: Formalizing and securing relationships on
public networks. First Monday, 2(9). https://doi.org/10.5210/fm.v2i9.548
Thelen, K. (1999). Historical institutionalism in comparative politics. Annual
Review of Political Science, 2, 369–404. https://doi.org/10.1146/annurev.
polisci.2.1.369
Toonen, T. A. (1998). Networks, management and institutions: Public adminis-
tration as ‘normal science’. Public Administration, 76(2), 229–252. https://
doi.org/10.1111/1467-9299.00099
Ullmann-Margalit, E. (1977). The emergence of norms. Clarendon Press.
Williamson, O. E. (1975). Markets and hierarchies, analysis and antitrust
implications: A study in the economics of internal organization. Free Press.
Zafirovski, M. (2004). Paradigms for analysis of social institutions: A case for
sociological institutionalism, 14(3), 363–397. https://doi.org/10.1080/039
0670042000318250
Zheng, Z., Xie, S., Dai, H. N., Chen, W., Chen, X., Weng, J., & Imran, M.
(2020). An overview on smart contracts: Challenges, advances and platforms.
Future Generation Computer Systems, 105. https://doi.org/10.1016/j.future.
2019
CHAPTER 2
Review of Institutional Grammar Research:

Overview, Opportunities, Challenges
2.1 Guiding Research Questions

Application and development of the IG was initially delayed but has
recently burgeoned. The IG received little attention in the years immedi-
ately following its introduction by Crawford and Ostrom, as indicated by
a paucity of scholarship engaging it. In recent years, however, there has
been a notable surge in the number of journal publications featuring the
IG. Between 1995, when Crawford and Ostrom initially introduced the
IG, and 2015, only 14 journal articles addressing it in some way were
published. As of 2021, the number of publications (including journal
articles and book chapters) had increased to more than 70. These publi-
cations are authored by scholars affiliated with universities in different
parts of the world, including the United States, Canada, Germany, the
Netherlands, Mexico, Australia, New Zealand, and Pakistan. They explore
institutional dynamics in various contexts—e.g., privacy and knowledge
commons (Geary et al., 2019; Sanfilippo et al., 2021), natural resource
management (Clement et al., 2015; Pacheco-Vega, 2020; Watkins et al.,
2015), governance of nonprofit organizations (Siddiki & Lupton, 2016),
regulation of food systems (Carter et al., 2015), transportation regula-
tion (Basurto et al., 2010), tobacco use regulation (Roditis et al., 2015),
municipal governance (Feiock et al., 2016), the management of natural
hazards (e.g., flood mitigation) (Witting, 2017), among others. However,

Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2_2
the majority of studies to date focus on some aspect of environmental

governance.
A review of existing scholarship demonstrates the conceptual and
methodological versatility of the IG (Dunlop et al., 2019; Siddiki et al.,
2019). The IG has been used to operationalize various concepts of
enduring interest to institutional analysts, as described in more detail
later in this chapter. Further, the IG has been paired with a diverse array
of methodological approaches to investigate these concepts. However,
despite the noted diversity in conceptual and methodological approaches
employed in existing studies, there is considerable overlap in the specific
research questions that motivate them. Among the questions commonly
explored in IG research are: How are institutions structured? At whom
are institutions targeted? How do institutions afford decision-making
and behavioral discretion to institutional targets? When and how do
institutions emerge and evolve? How do institutions-in-form relate to
institutions-in-use? Is the syntactic specification, and related conceptual-
ization, of the IG ontologically and logically sound? These questions, that
have drawn the attention of scholars using the IG, are akin to those that
motivate the broader study of institutional theory and analysis, but the
specific features of the IG make it an especially appropriate tool for investi-
gating them. For example, the syntactic components that form the basis of
the original IG specifically capture actor and actor constraints. Relatedly,
through explicit identification of institutional statements as focal institu-
tional units of analysis, and identification of the abstractable components
(i.e., syntactic components) of the statements, the IG enables compar-
ison of common kinds of institutional information across statements.
This makes it especially well suited for investigating institutional change.
Essentially, these features of the IG offer the institutional analyst a system-
atic approach upon which to assess what about institutions specifically is
changing over time.
Under the broad research questions noted above that have motivated
existing IG research to date, IG scholars have explored the following
specific questions.
• Institutional structure: How are institutional statements configured

in institutional design (Carter et al., 2015)? What are theoretical
and practical implications of institutional statement configurations
observed in institutional design (Schlager et al., 2020)?
2 REVIEW OF INSTITUTIONAL GRAMMAR RESEARCH … 35
• Decision-making and behavioral discretion: How is decision-making

and/or behavioral discretion afforded or restricted through institu-
tional design (Dörrenbächer & Mastenbroek, 2019)?
• Institutional targets: How do institutions assign rights and responsi-
bilities to different actors (Siddiki et al., 2011)?
• Institutional emergence and evolution: How can the IG be used
to rigorously and reliably measure institutional change (Weible &
Carter, 2015)? What types of institutions emerge endogenously in
decision scenarios of different kind (Frantz et al., 2015; Ghorbani &
Bravo, 2016)? To what extent can the IG be used as a cognitive
representation of institutions? How are different institution types of
the IG conceptually linked through institutional formation processes
(Frantz et al., 2015)?
• Link between institutions-in-form and institutions-in-use: To what
extent do institutions-in-form and institutions-in-use overlap? How
do actors perceive and interpret institutions-in-form (Kamran &
Shivakoti, 2013)? What informs actors’ conformance with
institutions-in-form (Siddiki, 2014; Siddiki et al., 2012)? What
factors facilitate or hinder implementation of institutions-in-form
(Carter et al., 2015)?
• Ontological and logical consistency of the IG: Are norms and
rules syntactically and operationally distinguishable (Schlüter &
Theesfeld, 2010)? To what extent is the IG able to represent insti-
tutions comprehensively (reconciling the diverse granularity of
syntactic components), and capture the diverse forms of enforce-
ment regimes for formal and informal institutions (e.g., variation in
multiplicity, frequency, and diversity of enforcement signals) (Frantz,
2015)?
Reflected within these questions are the analytical interests of scholars

engaging the IG. As well, the kinds of questions entertained by scholars
using the IG also reflect responses to observed conceptual and opera-
tional challenges. For example, questions oriented on syntactic integrity
are specifically responding to perceived challenges of the IG syntax. The
remainder of this chapter offers a detailed discussion of the specific kinds
of analytical approaches scholars have used to address these questions.
2.2 Analytical Approaches Used

in Institutional Grammar Research
2.2.1 Frameworks, Theories, and Concept Measurement
The majority of IG studies published to date explore concepts,
phenomena, and theories associated with the IAD framework. IAD
concepts and phenomena of central interest to IG scholars are insti-
tutional design and institutional development, where the latter is typi-
cally operationalized in terms of institutional origination and evolution.
Regarding the former, scholars are using the IG to study the design
of both formal and informal institutions (overlapping with the concepts
of institutions-in-form and institutions-in-use, with an explicit differen-
tiation discussed in Section 4.2.3.3), though the majority of studies
published to date focus on formal institutions. Many IG studies have
used the IAD framework related Common Pool Resource Theory (CPR)
to guide their analysis (Ostrom, 1990). While the IAD framework offers
a generalized conception of the role of institutions in governing collec-
tive action, CPR theory offers a more specific set of assumptions about
how institutions are developed and engaged in collective action relating to
common-pool resource management. CPR theory highlights how certain
qualities, or design features, of institutions, can enable or hinder successful
common pool resource management. IG scholars have used CPR theory
as a basis for characterizing and/or modeling institutions engaged in
managing the commons.
IG scholars also engage conceptual typologies associated with the IAD
framework as a basis for characterizing individual institutional statements,
or groupings of institutional statements, to account for their functional
properties or role in action situations more broadly. Most frequently
engaged is the “rule typology,” (Ostrom, 2005) which identifies and
organizes the seven types of rules that configure in action situations
by their functional characteristics: position, boundary, aggregation, infor-
mation, choice, scope, and payoff . In most studies in which institutional
analysts engage the IG and rule typology in concert, institutional state-
ments are mapped to rule types. This kind of mapping urges some level
of initial parsing of an institution (e.g., public policy) into institutional
statements, which then can be interpreted to understand what role an
institutional statement plays with respect to an action situation (e.g.,
establish a position, regulate a specific behavior), and then classified by

rule type.
In some cases, institutional statement information corresponding to
syntactic components is used to support the rule type classification (Garcia
et al., 2019). Leveraging the Aim in particular in the classification of rule
type was explicitly directed by Crawford and Ostrom (2005), who suggest
that actions embedded within institutional statements (i.e., Aims) signal
the functional properties of statements more generally. IG studies that
couple IG and rule type coding enable a rich understanding of institu-
tional design as this approach offers one valuable way to contextualize
the role and meaning of institutional statements with respect to instances
of action situations. Finally, recent IG scholarship also relies on the IAD
related, Social-Ecological Systems (SES) framework (Novo & Garrido,
2014). Novo and Garrido link data collected through interviews with
IG-coded statements along SES variables. For example, they interpret
interview responses that relate to different SES factors with reference
to coded statements. Coded statements offer information that help the
authors contextualize their interview findings.
Institutional analysts have also drawn on non-IAD affiliated frame-
works and theories to guide their IG studies. For example, Carter et al.
(2015) draw on Sabatier and Mazmanian’s Policy Implementation Frame-
work to guide their examination of the design of laws and corresponding
regulations governing the U.S. organic farming industry. Schlager et al.
(2020) rely on local public sector contracting theory to evaluate diversity
in the designs of formal institutions governing watersheds in New York
state. Prior (2016) draws on Schwartz’s Value Framework (Schwartz,
2012) in assessing motivations for compliance with norms and rules.
Beyond the application of particular frameworks and theories not formally
affiliated with the IAD framework, institutional analysts have used the
IG to study various concepts; for example, policy compliance (Siddiki
et al., 2012), policy authority (Clement et al., 2015; Feiock et al., 2016),
policy coerciveness (Siddiki, 2014), policy divergence (Carter et al., 2015)
and polycentricity (Heikkila & Weible, 2018). In studies evaluating these
concepts, IG data are often employed in the measurement, or operational-
ization of concepts. For example, in Siddiki et al.’s study of the design
of policies governing the aquaculture industry in American states, the
stringency of institutional statements was measured based on the pres-
ence/absence of a Deontic within an institutional statement, as well as
the type of Deontic (e.g., must, may, must not) present.
The kind of concept operationalization described here is based on a

syntax level coding of institutions; that is, concepts are measured using
institutional information corresponding to specific syntactic components.
There are also cases where the concept operationalization is based on
a statement-level coding. Again referencing a study that couples the
IG with IAD’s rule typology, Schlager et al. (2020) measure behav-
ioral discretion through a count of institutional statements of particular
rule types. Schlager et al. also dissect institutional statements by syntactic
component.
2.2.2 Collecting and Analyzing Institutional Grammar Data

In the exploration and/or application of the research questions, concepts,
and theories described above, scholars rely on a common set of methods
for collecting IG data. Most IG studies rely on institutional data that are
coded at the statement and syntax level; i.e., institutions which have been
parsed into institutional statements, and then further parsed into syntactic
components. However, in these studies, scholars sometimes engage in
selective parsing at the syntax level, meaning that they may choose to
only decompose statements along select components. Heikkila and Weible
(2018), for example, only parse statements along Attributes, Aims, Deon-
tics, and Objects, in accordance with their particular analytical objectives.
The vast majority of IG studies published to date are based on the assess-
ment of formal institutions taking the form of public policy; meaning that
most studies are relying on IG coding of public policies of some kind.
Further, nearly every IG study of public policy published to date relies on
the manual extraction of institutional information from public policies;
that is, the manual coding of policy texts in accordance with the IG.
As noted previously, the line of IG research that focuses on the eval-
uation of policy text was initiated by Basurto et al. (2010) and shortly
after furthered by Siddiki et al. (2011). Many others have contributed to
extending this line of research, but notable about these earlier studies
is that they specify guidelines that institutional analysts should follow
when engaging in IG-based coding of the policy text. These guidelines
specify a general workflow for engaging in IG coding of policy text,
as well as offer guidance on pre-processing policy texts to make them
amenable to coding. These initial guidelines have largely been adopted
as convention in IG coding practice, as indicated by their codification in
the first publicly accessible IG Codebook (Brady et al., 2018), though
the IG 2.0 offers substantive modifications to these in accommodation

of syntactic refinements and associated analytical opportunities (Frantz &
Siddiki, 2021). The articulation of operational guidelines for using the
IG to code policy text was in itself an important methodological inno-
vation, insofar as the original presentation of the IG by Crawford and
Ostrom was largely conceptual; outlining the basic IG syntax and theo-
retical motivations thereof, but not offering specific practical guidance on
applying it, beyond a game-theoretical illustration.
Another, more recent, methodological innovation pertaining to the
collection of IG data from public policies has been the engagement of
computational approaches to support automated coding of policy text
(Rice et al., 2021). Rice et al. have developed an open-source tool for
automated classification of policy text according to the IG syntax lever-
aging computational text analysis and natural language processing. Their
approach is trained on IG-coded food system regulations (Siddiki et al.,
2015). Rice et al.’s automated IG coding approach offers significant
promise for future IG research. Further discussion on the opportuni-
ties and limitations associated with this approach is offered later in this
chapter.
Yet another notable methodological innovation that emerged from the
application of the IG to study policy text was the augmentation of the
IG by an additional syntactic component, called the Object . Formally
introduced by Siddiki et al. (2011), the Object was introduced primarily
to enhance reliability in institutional statement coding – in essence, by
clarifying the meaning and coding of the Aim, and accommodating the
classification of commonly observed but previously non-distinguished
institutional statement information with an additional syntactic compo-
nent. Previous to Siddiki et al. (2011)’s study, the focal action of an
institutional statement (i.e., the Aim) was not formally differentiated from
the receiver of the action. For example, for the statement “Farmer must
submit organic systems plan annually,” the entire clause “submit organic
systems plan” would be characterized as the Aim of the statement. The
inclusion of the Object within the IG urged the coder to effectively reflect
the syntactic difference between a word/phrase representing the action of
a statement from that which is functionally dependent on the action (i.e.,
the Object ). An analogous difference is found in the distinction between
sentence verbs and sentence objects in the English language. The identi-
fication of the Object has been common in IG coding practice following
the publication of Siddiki et al. (2011)’s study. Importantly, while Siddiki
et al. officially proposed the inclusion of the Object into the IG syntax,
others, as described in more detail below, had previously signaled the
analytical value associated with this differentiation (Smajgl et al., 2008).
To date, few scholars have used the IG to study informal institutions
in real-world settings; that is, to identify and code into the IG syntax
institutional statements reflected in social practices or conventions, and
which are tacitly understood or orally communicated rather than already
codified in written form (Watkins & Westphal, 2016). Watkins and West-
phal (2016)’s study showcases one application of the IG toward the
study of informal institutions. For their study, Watkins and Westphal use
interviews and participant observation of institutional actors of interest
to identify the shared strategies, norms, and rules they engage in the
practice of ecological restoration decision-making. These orally commu-
nicated strategies, norms, and rules are initially documented in written
form, annotated according to the IG syntax, and then interpreted for
patterns of interest to the authors. The challenge with this IG annota-
tion—in both the statement and syntactic classification—is that informal
institutions are not typically conveyed through speech in complete insti-
tutional statements. Rather, as Watkins and Westphal state, participants in
their study tended to convey how things get done non-linearly with anec-
dotes and personal assessments. The authors took descriptions of activities
recounted as such, translated them into institutional statements, which
could be subsequently parsed along IG syntactic components.
The approaches to collecting IG data described above have largely
been used in empirical studies of institutional design and phenomena.
The collection of institutional information for simulation-based IG studies
[e.g., those engaging agent-based modeling (ABM) (Epstein, 2007)]
may entail different approaches. Agent-based Modeling and Simulation
(Epstein, 2007; Gilbert & Troitzsch, 2005) reflects an analytical method
in which institutional arrangements (or any other social formations for
that matter) are reconstructed in the form of computational represen-
tations of humans, called agents, and placed in a setting that resembles
relevant features of the analyzed social system in an artificial society.
The latter is subsequently instantiated in a simulation in which agents
can exhibit the modeled behaviors and, by interaction, produce emer-
gent outcomes in the form of social phenomena or altered systemic states
evaluated by the modeler or experimenter. The experimenter can variably
adjust the parameterization to test different hypotheses, social structure,
environmental features, etc.1 In the context of ABM studies, the insti-

tutions of interest to a modeler, from which institutional information is
extracted, may be existing formal and informal institutions that s/he relies
on for the upfront parameterization of a model; that is, for the specifica-
tion of actors, actor activities, and actor constraints that will be reflected
in the modeling exercise. Alternatively, the modeler may be interested in
collecting institutional information from the institutions that are gener-
ated, or emerge, as a result of the simulated execution of the modeled
artificial society. In ABM, the patterned behaviors that emerge among
agents which have been assigned particular attributes and choice sets are
characterized as institutions, which can subsequently be classified along
IG syntactic components. Features of these designs are then typically eval-
uated in reference to systemic outcomes of interest to the modeler. For
example, Ghorbani and Bravo (2016) use the IG in both the up-front
parameterization of an ABM, as well as to characterize emergent insti-
tutions in a common pool resource setting. Ghorbani and Bravo (2016)
start their exercise with a set of institutional statements that follow the
IG syntactic structure—delineating what an actor is required, allowed, or
forbidden to do within a particular constraint. For example, “Actor must
appropriate n resource units at the time a condition is met.” Their study
evaluates how these statements are adapted over time by actors endowed
with particular characteristics and decision-making constraints. Thus, the
initial and resulting institutional statements can be coded in terms of the
IG.
Evident from this description, as well as specifically mentioned in extant
research that uses IG data as input, is that formal institutions are most
amenable to an IG coding, which may in part explain the relative abun-
dance of studies focusing on formal versus informal institutions. Formal
institutions usually contain directives following the typical “X must do
Y in context C” regulative structure, which is conducive to being coded
according to the A-D-I-C-O syntax. As noted above, coding informal
institutions into the IG syntax can be challenging as it may require the
analyst to engage, as Watkins and Westphal (2016) did, to reconstruct
institutional information into the typical regulative statement structure
referenced above. In a similar vein, those using the IG in the context
of simulations encode emergent patterns of behavior according to the
1 The principles and practical application is motivated in greater detail in Sect. 8.2.2.
IG syntax. In order to facilitate this, the modeler will need to consider

how modeling parameters map to syntactic components (e.g., agents to
Attributes, agent choices to Aims, etc.) as part of the model design
process. For the upfront parameterization of models, institutional analysts
may be relying on IG coding of formal or informal institutions.
Having reviewed conventions of data collection in existing IG research,
the following discussion addresses common practices in the analysis of IG
data.
The majority of IG studies published to date engage in descriptive
analyses of institutional information or institutional data that has been
coded according to the IG. Descriptive accounting of IG-coded data is
typically done at the statement and component level. At the statement
level, scholars typically identify the number of institutional statements
comprising their institution(s) of interest and summarize statements by
shared strategies, norms, and rules. Similarly, in descriptive accounting
at the component level, scholars typically report the frequency of infor-
mation by syntactic component; for example, indicating the Attributes,
Aims, and Deontics that occur most frequently within an institution of
interest (Siddiki et al., 2011). This type of analysis is useful insofar as it
signals qualities of institutions as indicated by the presence or absence
of particular types of information that correspond to different syntactic
components. At the same time, it does not support assessments of the
fundamentally configurable nature of institutional information; that is,
how institutional information relating to particular syntactic components
or statements combines to convey institutional meaning.
Recent studies that use network analysis in the assessment of IG coded
data help to address this gap, insofar as they are explicitly designed
to capture relations among institutional information. In these studies,
scholars are interested in connecting information that corresponds to
one or more syntactic components, within and across statements (Olivier,
2019). For example, using network analysis, one can assess which institu-
tional actors are connected by way of Aims or Conditions. Alternatively,
one can assess how a particular actor is connected to different Objects or
Aims found in an institution. Network analysis can be engaged to aid in
the visualization of institutional information linkages, and also to identify
qualities of “institutional networks.” Olivier (2019) uses network analysis
to determine “networks of prescribed actions” among actors as indicated
in formal rules governing the provision of high-quality drinking water in
two American cities, New York and Boston. As part of his analysis, Olivier
identified institutional statements in which the Attributes component of

the statement was linked to an actor captured in the Object field, by way of
the statement action (e.g., “Actor A must provide Actor B with a progress
report of activities undertaken.” ). In this statement, coded in accordance
with the original IG syntax, Actor A (the Attribute) is connected with
Actor B (the Object ) through the Aim provide. Olivier then constructed
Attribute-Object dyads represented in institutional statements, aggregated
these to form a Network of Prescribed Interactions, visualized them in the
form of network graphs, and calculated network statistics (e.g., density,
centrality) as related to concepts of interest.
Another kind of analysis found in extant IG research is the comparison
of formal and informal institutions. In these studies, scholars are typically
interested in assessing conformance among de jure and de facto institu-
tions (Siddiki, 2014; Siddiki et al., 2012). This fundamentally requires
that formal and informal institutions be captured in similar ways, so
as to facilitate a comparative assessment. To support this type of anal-
ysis, Siddiki et al. (2012) and Siddiki (2014) first coded regulations of
analytical interest using the IG, and then asked interviewees in the corre-
sponding regulatory domain to assess how their behaviors in practice
conform to prescribed or allowed behaviors specified within regulations
using a Q-Sort exercise (McKeown & Thomas, 2000).
Analysis of institutional information within the context of ABM focuses
on discerning characteristics of emergent institutions and then evaluating
systemic outcomes that result from the generated institutions exhibiting
particular features. For Smajgl et al. (2008), the IG offers a basis for
characterizing the types of institutions (actions/strategies or rules) that
agents are subject to in simulated environments, and a basis for modeling
institutional change. In their model, Smajgl et al. (2008) allow agents
the possibility of selecting among different actions and thereby estab-
lishing new sets of institutions within the simulations. A comparative
assessment of initial with emergent institutions, each of which is character-
ized according IG components, allows the authors to finely detect aspects
of institutions that evolve through the simulation. One syntactic revision
embedded in Smajgl et al. (2008) study, which was later formally intro-
duced by Siddiki et al. (2011), was the separation of the action of an insti-
tutional statement (i.e., Aim) from the receiver of the action, which they
like Siddiki et al. (2011) refer to as the Object. Smajgl et al., (2008)’s deci-
sion to separate the Aim and Object of modeled institutional statements
was analytically motivated—meant to aid in the detection of emergent
institutions—and was not accompanied by a suggestion to modify the

IG syntax. Nonetheless, by highlighting the pragmatic appeal of further
distinguishing syntactic components, they were calling out the need for
improved ontological consistency in the Grammar. Frantz et al. (2015)
use the IG to represent different forms of institutions governing trade.
They model historical trading scenarios that reflect differing formal and
informal institutional arrangements and associated interaction patterns
among agents associated with different cultural groups. They use this as
a foundation to test “historical hypotheticals” drawing on the postulated
interaction of those groups based on different enforcement principles.
The use of the IG in this work is twofold. It includes the conceptu-
alization of an endogenous institution formation process that links the
different institution types expressed in the IG (strategies, norms, rules) in
a refined form of the IG, capturing higher-level complexity of institutional
arrangements based on systemic interdependencies of institutional state-
ments. This provides the basis to establish an explanatory account of the
institutional formation process (why and how an arrangement came about)
(Frantz, 2020), as well as the opportunity to analyze the endogenously
generated institutional statements across different levels of social orga-
nization (individual, group, society). Importantly, as part of this work,
they present the Nested ADICO concept, which reconceptualizes the
Or else component in terms of a statement in its own right specifying
consequences for actors for acting or failing to act, as prescribed in other
statements – providing the basis for the representation of complex inter-
linkages of institutional statements. Through their modeling exercises,
both Smajgl et al. (2008) and Frantz et al. (2015) highlight the need
for syntactic specificity but also malleability when engaging the IG in
computational applications, a point that is described in more detail in
the following section.
The discussion to this point highlights the questions that have moti-
vated IG research, and describes the analytical approaches featured
therein. Looking forward, the following discussion highlights the oppor-
tunities and challenges associated with the use of the IG to conduct
institutional analysis.
2.3 Research Opportunities and Challenges

Development and application of the IG has occurred over a relatively
short period of time, but nonetheless, extant IG scholarship has clearly
and consistently conveyed the benefits and opportunities it affords for

institutional analysts. Existing IG research validates the utility of the
approach toward the systematic and rich characterization of institu-
tional design, particularly the design of formal institutions, toward an
understanding of institutional emergence and change, and toward the
understanding of institutional configurations (i.e., understanding of how
institutional statements link together to convey institutionally relevant
meaning). Applications of the IG in a variety of domains affirm that the
base IG syntax is generalizable; that is, that the IG can be applied to
study the design of institutions with different topical foci and analytical
approaches. This affirmation is crucial as it lends confidence that the IG
syntax is not overfitted to capture language within a particular institu-
tional domain, or setting. Additionally, the use of the syntax, and specific
syntactic components, toward the operationalization of a diverse array
of concepts that link to various extents with specific theories and frame-
works, highlights the analytical versatility of the IG. Moreover, reflecting
the analytical versatility of the IG is that institutional analysts have used
it to study both formal and informal institutions, and have demonstrated
the value of using IG coding, or coded data, within the context of a
variety of analytical techniques (e.g., social network analysis, agent-based
modeling).
The last decade of research also makes clear some of the central chal-
lenges of the IG. Some of these challenges have been explicitly called
out by institutional analysts, while others can be inferred by reviewing
the kinds of conceptual and practical adaptations scholars have made
to enable its usage in their research. A first challenge highlighted in
existing IG research are ontological inconsistencies in the IG syntax; that
is, ambiguities in the definitions of syntactic components. Siddiki et al.
(2011) and Smajgl et al. (2008) point out a syntactic hole in the orig-
inal Grammar prompting them to parse institutional statements along
an additional syntactic component, which they commonly refer to as the
Object. In their studies, both separate institutional data along an “Object”
component in addition to other IG components. For Siddiki et al. (2011),
the additional coding field encouraged reliability in institutional coding.
For Smajgl et al. (2008), the separation of actions from receivers of
action made it easier for the authors to detect the phenomena they were
interested in investigating in the context of agent-based simulation (i.e.,
institutional emergence). In this light, both research teams introduced
the additional component out of pragmatism, though this adjustment to
accommodate practical application of the IG also led Siddiki et al. (2011)

to make a case for formally incorporating the Object into the syntax. More
broadly, however, the syntactic adjustment suggests ontological inconsis-
tency in aspects of the original IG. It signals that fundamentally different
kinds of information were captured under one syntactic component (the
Aim), which became readily apparent when institutional analysts went to
code and analyze institutional data in accordance with the IG. At the same
time, coding practice also reveals lower coding reliability among Objects
and Conditions, relative to the other syntactic components, suggesting
that while the introduction of the Object offered some improvement to
the ontological consistency of the original IG, further refinement of the
definitions of Objects and Conditions is warranted. The IG 2.0 suggests
specific syntactic refinements to both components in light of observed
deficiencies.
Another ontological inconsistency presented in extant IG research
concerns the Or else component. As part of their computational modeling
exercise, Frantz et al. (2013) and Frantz (2015) challenge the idea that
the Or else is in fact a syntactic component at all. Instead, they suggest
that the institutional information that is coded into the Or else component
field, which conveys a payoff associated with a referenced action, typi-
cally takes the form of complete institutional statement. For example, in
the statement “Farmer must submit organic system plan by December 31st,
or else inspector must immediately suspend organic farming certification,”
“or else inspector must immediately suspend organic farming certification”
is the information that would typically be coded as the Or else as it
indicates the payoff (e.g., sanction) associated with failure to submit an
organic systems plan on time. However, this information also obviously
comprises a complete institutional statement in its own right, containing
an Attribute (inspector), Deontic (must), Aim (suspend), Object (organic
farming certification), and Conditions (immediately). For Frantz et al.
(2013), capturing this information as a separate institutional statement
is advantageous within an institutional modeling setting, insofar as the
non-Or else and Or else parts of the statement (under a conventional
coding) capture distinctive, albeit linked, activities performed by different
actors. Thus, they suggest that the Or else be reconceived as an institu-
tional statement of consequential kind that nests on another statement
of monitored kind, where the latter describes action required, permitted,
or forbidden by a particular actor and the former describes the conse-
quence for non-conformance with pre-/pro-scribed action. It is worth
highlighting that key ontological inconsistencies noted to date have been

presented by those engaging the IG within computational applications.
Broadly, this research suggests that the use of the IG within computa-
tional approaches necessitates syntactic and semantic clarity, as modeling
requires precision in parameter specification.
Another limitation of the IG noted by scholars is the time-consuming
and laborious nature of coding institutions by word or phrase to support
the classification of institutions at the statement and syntax level (Siddiki
et al., 2019). They have also noted the practical and scientific constraints
that stem from this. Practically, the resource-intensive nature of IG coding
limits the amount of institutional information that an analyst can reason-
ably collect for a single research study. Likely related, many IG studies
report on an analysis of a single institutional case. The associated scien-
tific limitation of this is that it precludes the institutional analyst’s ability
to produce generalizable insights. The recent work by Rice et al. (2021)
on the development of an automated IG coding tool can significantly
aid in addressing this challenge. Their approach effectively classifies insti-
tutional statement information according to the IG components. Here
too, however, classification is likely to improve with enhanced ontological
consistency relating to syntactic components.
In linking syntactic refinements with computational specific considera-
tions, the research and discussion presented above is essentially making a
case for improved computational tractability of the IG. This improved
computational tractability stems from the unambiguous definition of
syntactic components as this specificity enables modeling precision. Relat-
edly, implicit within extant research is that, in addition to stemming
from clarity in the definitions of syntactic components, computational
tractability may be enhanced with coding schemas that enable more parsi-
monious classification of institutional information; in other words, more
finely parsed institutional information within and across syntactic compo-
nents. This is essentially what Smajgl et al. (2008) were trying to achieve
in their separation of actions and receivers of actions, though this practical
adaptation also signaled a conceptual ambiguity in the Aim component.
Relating to other computational efforts, namely machine coding, parsi-
monious encoding of institutional data can also support greater accuracy
in automated classification. Essentially, if the machine coding approach is
trained on more finely parsed institutional information across a number
of clearly distinguishable syntactic components, this may in fact support
classification accuracy in application, while increasing the versatility of the

encoded data manifold.
An additional limitation of the existing IG referenced in extant schol-
arship pertains to its applicability to institutional statements that don’t
match a regulative form. Weible and Carter (2015) note that some
statements, rather than following a regulative form (“X must do Y in
context Z”), follow a constitutive syntax (“X is Y in context Z”). They
note, “Regulatory statements generally refer to an identifiable action, while
constitutive statements generally have no action, and instead define, label,
or describe a position or part of the physical world.” (Weible & Carter,
2015, p. 226). The distinction is an important one, and Weible and
Carter offer some guidance on how to deal with these types of statements.
Nevertheless, challenges remain in operationally characterizing consti-
tutive statements given that they fundamentally convey different types
of information. Further, existing coding guidelines under-emphasize the
inclusion of constitutive statements in institutional statement coding—
recommending, for example, discarding of definitional information at the
beginning of public policy documents—which may then yield incomplete
capturing of institutional information. Constitutive statements, including
their definition and operational treatment, are addressed extensively in the
characterization of the IG presented in this work.
Recounting these opportunities and limitations of the IG based on
extant research, the following chapter motivates their integration as part
of a refined New Institutional Grammar.
References
1177/1065912909334430
Brady, U., Basurto, X., Bennett, A., Carter, D. P., Hanlon, J., Heikkila, T.,
Lien, A., Chonaiew, S. M., Olivier, T., Schlager, E., Siddiki, S., & Weible, C.
(2018). Institutional analysis of rules-in-form coding guidelines (tech. rep.).
Center for Behavior, Institutions and the Environment. https://complexit
yasu.edu/sites/default/files/papers/cbie_wp_2018-006_0.pdf
Carter, D. P., Weible, C. M., Siddiki, S. N., Brett, J., & Chonaiew, S. M. (2015).
Assessing policy divergence: How to investigate the differences between a
law and a corresponding regulation. Public Administration, 93(1), 159–176.
https://doi.org/10.1111/padm.12120
Clement, S., Moore, S. A., & Lockwood, M. (2015). Authority, responsibility

and process in Australian biodiversity policy. Environmental and Planning
Law Journal, 32(2), 93–114.
ican Political Science Review, 89(3), 582–600. https://doi.org/10.2307/
2082975
Crawford, S. E. S., & Ostrom, E. (2005). A Grammar of institutions. In Ostrom,
E. (ed.) Understanding institutional diversity, 137–174. Princeton University
Press.
Dörrenbächer, N., & Mastenbroek, E. (2019). Passing the buck? Analyzing the
delegation of discretion after transposition of European Union law. Regulation
and Governance, 13(1), 70–85. https://doi.org/10.1111/rego.12153
Dunlop, C. A., Kamkhaji, J. C., & Radaelli, C. M. (2019). A sleeping giant
awakes? The rise of the Institutional Grammar Tool (IGT) in policy research.
Journal of Chinese Governance, 4(2), 163–180. https://doi.org/10.1080/
23812346.2019.1575502
Feiock, R. C., Weible, C. M., Carter, D. P., Curley, C., Deslatte, A., & Heikkila,
T. (2016). Capturing structural and functional diversity through institutional
analysis: The mayor position in city charters. Urban Affairs Review, 52(1):
129–150. https://doi.org/10.1177/1078087414555999
Frantz, C., Purvis, M. K., Nowostawski, M., & Savarimuthu, B. T. R. (2013).
nADICO: A nested grammar of institutions. Lecture Notes in Computer
Science, 8291 LNAI , 429–436. https://doi.org/10.1007/978-3-642-44927-
7_31
Frantz, C. K. (2020). Unleashing the agents: From a descriptive to an explana-
tory perspective in agent-based modelling. Springer Proceedings in Complexity.
https://doi.org/10.1007/978-3-030-34127-5_16
Frantz, C. K., Purvis, M. K., Savarimuthu, B. T. R., & Nowostawski, M. (2015).
Modelling dynamic normative understanding in agent societies. Scalable
Computing, 16(4), 355–380. https://doi.org/10.12694/scpe.v16i4.1128
Frantz, C. K., & Siddiki, S. N. (2021). IG 2.0 Resources. https://github.com/
InstitutionalGrammar/IG-2.0-Resources
Frantz, C. K. (2015). Agent-Based institutional modelling: Novel techniques for
deriving structure from behaviour (Doctoral dissertation). University of Otago.
Dunedin, New Zealand. http://hdl.handle.net/10523/5906
Garcia, M., Koebele, E., Deslatte, A., Ernst, K., Manago, K. F., & Treuer, G.
(2019). Towards urban water sustainability: Analyzing management transi-
tions in Miami, Las Vegas, and Los Angeles. Global Environmental Change,
58. https://doi.org/10.1016/j.gloenvcha.2019.101967
Geary, J., Reay, T., & Bubela, T. (2019). The impact of heterogeneity in a
global knowledge commons: Implications for governance of the DNA barcode
commons. International Journal of the Commons, 13(2), 909–930. https://
doi.org/10.5334/ijc.861
Gilbert, N., & Troitzsch, K. G. (2005). Simulation for the social scientist. Open
University Press.
Heikkila, T., & Weible, C. M. (2018). A semi-automated approach to
analyzing polycentricity. Environmental Policy and Governance, 28(4): 308–
318. https://doi.org/10.1002/eet.1817
Kamran, M. A., & Shivakoti, G. P. (2013). Comparative institutional analysis
of customary rights and colonial law in spate irrigation systems of Pakistani
Punjab. Water International, 38(5), 601–619. https://doi.org/10.1080/
02508060.2013.828584
McKeown, B. F., & Thomas, D. B. (2000). Q Methodology in assessment and
research. Summer Methodology Quantitative Applications in the Social Sciences
series.
Novo, P., & Garrido, A. (2014). From policy design to implementation: An
institutional analysis of the new Nicaraguan Water Law. Water Policy, 16(6),
1009–1030. https://doi.org/10.2166/wp.2014.188
Olivier, T. (2019). How do institutions address collective-action problems?
Bridging and bonding in institutional design. Political Research Quarterly,
72(1). https://doi.org/10.1177/1065912918784199
Ostrom, E. (1990). Governing the Commons: The evolution of institutions for
Press.
Pacheco-Vega, R. (2020). Governing urban water conflict through watershed
councils: A public policy analysis approach and critique. Water (Switzerland),
12(7), 1849. https://doi.org/10.3390/W12071849
Prior, J. (2016). The norms, rules and motivational values driving sustain-
able remediation of contaminated environments: A study of implementation.
Science of the Total Environment, 544, 824–836. https://doi.org/10.1016/
j.scitotenv.2015.11.045
of policy texts with the Institutional Grammar. Public Administration, 99,
248–262. https://doi.org/10.1111/padm.12711
Roditis, M. L., Wang, D., Glantz, S. A., & Fallin, A. (2015). Evaluating Cali-
fornia campus tobacco policies using the American college health association
guidelines and the Institutional Grammar Tool. Journal of American College

Health, 63(1), 57–67. https://doi.org/10.1080/07448481.2014.963108
Sabatier, P., & Mazmanian, D. (1980). The implementation of public policy: A
Framework of Analysis. Policy Studies Journal, 8(4), 538–560. https://doi.
org/10.1111/j.1541-0072.1980.tb01266.x
Sanfilippo, M., Strandburg, K. J., & Frischmann, B. M. (2021). Privacy as
knowledge commons governance. Governing privacy in knowledge commons.
https://doi.org/10.1017/9781108749978.012
Schlager, E. C., Bakkensen, L. A., Olivier, T., & Hanlon, J. (2020). Institu-
tional design for a complex commons: Variations in the design of credible
commitments and the provision of public goods. Public Administration, 99,
Schlüter, A., & Theesfeld, I. (2010). The Grammar of institutions: The challenge
22(4), 445–475. https://doi.org/10.1177/1043463110377299
Schwartz, S. H. (2012). An overview of the schwartz theory of basic values.
Online Readings in Psychology and Culture, 2(1). https://doi.org/10.9707/
2307-0919.1116
States. Review of Policy Research, 31(4), 281–303. https://doi.org/10.
1111/ropr.12075
Siddiki, S., Basurto, X., & Weible, C. M. (2012). Using the Institutional
Grammar Tool to understand regulatory compliance: The case of Colorado
aquaculture. Regulation and Governance, 6(2), 167–188. https://doi.org/
10.1111/j.1748-5991.2012.01132.x
Siddiki, S., Heikkila, T., Weible, C.M., Pacheco-Vega, R., Carter, D., Curley, C.,
Deslatte, A., & Bennett, A. (2019). Institutional analysis with the Institutional
Grammar. Policy Studies Journal. https://doi.org/10.1111/psj.12361
Siddiki, S., & Lupton, S. (2016). Assessing nonprofit rule interpretation
and compliance. Nonprofit and Voluntary Sector Quarterly, 45, 156–174.
https://doi.org/10.1177/0899764016643608
97.x
Siddiki, S. N., Carboni, J. L., Koski, C., & Sadiq, A. A. (2015). How policy rules
shape the structure and performance of collaborative governance arrange-
ments. Public Administration Review, 75(4), 536–547. https://doi.org/10.
1111/puar.12352

in Complex Systems, 2(11), 199–215.
Watkins, C., & Westphal, L. M. (2016). People don’t talk in institutional
statements: A methodological case study of the institutional analysis and devel-
opment framework. Policy Studies Journal, 44, 98–122. https://doi.org/10.
1111/psj.12139
Watkins, C., Westphal, L. M., Gobster, P. H., Vining, J., Wali, A., & Tudor,
M. (2015). Shared principles of restoration practice in the Chicago wilder-
ness region. Human Ecology Review, 21(1), 155–177. https://doi.org/10.
22459/HER.21.01.2015.07
Weible, C. M., & Carter, D. P. (2015). The composition of policy change:
Comparing Colorado’s 1977 and 2006 smoking bans. Policy Sciences, 48(2),
207–231. https://doi.org/10.1007/s11077-015-9217-x
Witting, A. (2017). Ruling out learning and change? Lessons from urban flood
mitigation. Policy and Society, 36(2), 251–269. https://doi.org/10.1080/
14494035.2017.1322772
CHAPTER 3
Motivation for a New Institutional Grammar
The version of the Institutional Grammar that will be presented

throughout this text is a revised version of the original developed and
presented by Crawford and Ostrom in 1995. The revised version—which
we refer to as Institutional Grammar 2.0 (IG 2.0), or New Institu-
tional Grammar—is modified by the book authors based on limitations
of the IG noted in extant scholarship, but is consequently contextual-
ized with the IG, leaving the reader with a comprehensive overview of
both the original IG and the introduced modification and extensions.
This chapter will offer a detailed account of the motivations informing
the revised IG specification and details theoretical and analytical implica-
tions thereof. The three core motivations of the revised specification that
will be described in this chapter, are (i) improved ontological consistency
of the syntax undergirding the IG; (ii) comprehensiveness of coverage
of institutional statements, paired with the intent to increase reliability in
coding; and (iii) increased computational tractability of the IG responding
to calls to engage computational approaches in IG-based coding and anal-
ysis, which includes the need for an alternative structural specification
of IG coded information to make its rich feature set analytically acces-
sible. As such, the revised specification inherently introduces elaborations
to the IG that support the disciplinary orientations and requirements
of the IG user community, which is increasingly multi-disciplinary. The

Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2_3
three aforestated motivations, and the issues to which they are linked, are
addressed through the revised IG specification—the IG 2.0.
3.1 Ontological Consistency

As relayed in earlier chapters of this book, the IG as developed by
Crawford and Ostrom (1995) offers a fundamental characterization of
components that in combination comprise institutional statements. These
components of an institutional statement include the Attributes (or
simply Attribute) (actor), Deontic (expectation about performed behav-
iors expressed as either obligation, prohibition, or permission), Aim
(action or outcome that is regulated), Conditions (circumstance under
which a regulated activity or outcome applies), and Or else (consequence
of noncompliance with the regulated statement).
While previous work has highlighted various challenges related to
the IG, including limitations in capturing institutional content suffi-
ciently, either based on absent components [Object component (Siddiki
et al., 2011)], insufficient differentiation between statement types (e.g.,
Schlüter & Theesfeld, 2010), and the inability to capture complexity
expressed in statements [e.g., Frantz et al. (2013) and Frantz et al.
(2015)], a notable concern not recognized in the extant literature is the
varying levels of granularity at which the Grammar components operate –
some being more general, and others more specific in kind –, and the
specificity with which they are defined. Shedding light on this incongru-
ence is of relevance, since it affects the institutional analyst’s ability to
apply the IG to unambiguously characterize components of an institu-
tional statement, and making those available for a systematic analysis of
institutional design. More essentially, this can be interpreted as a challenge
to the validity of the IG to reliably capture features of institutional design
in the first place, where such features, variably referred to as Institutional
Features, are institutional concepts, their relationships, and interaction
within the institutional setting that may be identified on component
(e.g., actors, actions referenced in components), statement (e.g., func-
tion/effects of statements) or system level (e.g., function of interlinked
statements).
Addressing this concern, this section discusses selected characteristics
of the IG that challenge its ontological consistency and prompts for
refinements proposed as part of this book. To establish a terminological
basis, ontological consistency is defined as the clear, logically consistent and
3 MOTIVATION FOR A NEW INSTITUTIONAL GRAMMAR 55
coherent characterization of syntactic components, where conceptual char-

acterizations, their relationships and (inter)dependencies are explicit and
unambiguous so as to avoid variable interpretation of encoded information,
or worse, potential contradictions independent of underlying analytical
objectives.1 Developing clarity on the conceptual level offers two essential
benefits for the long-term adoption and analytical value of the IG. First,
establishing an unambiguous conceptual understanding of the structure
institutional statements capture, alongside corresponding methodological
guidelines for the encoding of institutional content, the IG can be applied
more reliably across various domains with reduced attachment to project-
specific accommodations. This leverages broader usability of encoded
information for methodological developments, including the produc-
tion of a rich integrated data basis for the development of mechanisms
for partially or completely automated encoding of institutional state-
ments – which in turn reduces productivity hurdles associated with the
prevalent manual encoding of statements. The second benefit lies in the
development of the IG as an analytical tool. The flexible semantic char-
acterization of content (i.e., characterization of institutional meaning),
based on project-specific encoding conventions in the best case, and based
on coder-specific preference in the worst case, does not allow for gener-
alizations across a specific study. This thereby limits the development of
advanced analytical approaches that go beyond the statistical treatment of
component distributions, and rather operate on deeper structural aspects
of institutional statements; for example, those that focus on interlinkage
of institutional statements to derive a richer understanding of policy (as
later discussed in Sect. 8.2.1.4).
Drawing theoretical linkage to the understanding of ontological consis-
tency put forth, entertained here is the understanding that the ability to
describe and infer relationships between components hinges on the accu-
racy of characterization of components based on shared intrinsic prop-
erties [or “substantive kinds” (Ellis, 2007)] – properties that describe
what components “are” – or shared relational properties [“property
kinds” (Ellis, 2007)] – properties describing how components “relate”.
Ontological consistency in the IG context means that the core syntactic
components/operators of the IG – Attributes, Object, Deontic, Aim,
Conditions, and Or else – are each defined with sufficient clarity and
1 See also Agazzi (2011) and Ellis (2007).

uniformity, so that where those components occur in concert, there is

no ambiguity about the distinctiveness of semantic information that each
component conveys. This extends to the characterization of functional
relationships that components hold (e.g., how Aim and Object relate,
etc.). An implied – albeit challenging – objective is that syntactic compo-
nents are defined at a comparable level of granularity (i.e., some syntactic
components should not be more coarsely or finely defined than others),
where possible.2
The following discussion explicitly highlights concerns around onto-
logical consistency by drawing attention to three components of the IG,
namely Conditions, Object and finally, the Or else component.
Recalling its specific function as per Crawford and Ostrom, the Condi-
tions component defines “when, where, how, and to what extent an
[action] is permitted, obligatory, or forbidden” (Crawford & Ostrom,
1995, p. 584). At first glance this characterization appears non-
contentious, since this component captures any form of context charac-
terization as relevant for the analyzed institutional statement.
However, upon closer inspection, this component characterization
conflates an important distinction between conditions, or context in the
form of preconditions that lead to the activation of the institutional state-
ment, captured in an intuitive interpretation of the “when” and “where”
of the original definition, and a second form of context that characterizes,
or qualifies, the action specified in the institutional statement [e.g., “how
a statement is to be followed” (Crawford & Ostrom, 1995, p. 585)].
Drawing on Ellis (2007)’s characterization of ontology, conditions as
applied in the IG reflect different kinds, both in terms of their substan-
tive characterization (i.e., what they “are”) as well as their relationships to
other components or the statement as a whole (i.e., how they “relate”).
Figure 3.1 highlights the qualitatively different relationship between
the “Conditions of the first kind” that capture the preconditions leading
to the activation of a statement in the first place, and—from the
perspective of the statement—are exogenous variables that act upon the
statement as a whole.
“Conditions of the second kind” are qualifications of the Aim captured
in the first place. This is best motivated using a trivial example: Upon
2 Nevertheless, assuming complete parity remains a challenge.

Fig. 3.1 Combined Institutional Grammar patterns example
completion of the financial year, corporations must provide tax return

within three months, exemplary coded in the following:
Attributes: corporations
Deontic: must
Aim: provide
Object tax return
Conditions: Upon completion of the financial year; within three months
While this example explicitly identifies actor (corporations ), Deontic

(must ), action (provide), and Object (tax return) unambiguously, we can
observe multiple Conditions. While the existence of multiple conditions
is not problematic per se, the specific instances in this example reflect the
duality in characterization motivated above: Upon completion of the finan-
cial year highlights a condition under which a statement comprised of
the other components applies, whereas within three months qualifies, or
constrains, the obligation to provide a tax return, thus affects the state-
ment only when enacted. While seemingly concentrated on the compo-
nent itself, the ambiguous nature of the Conditions thereby indirectly and
variably affects or constrains, the interpretability of other non-condition
components, including Attributes, Deontic, Aim, and Object, whereas the
second form of conditions, the one qualifying the Aim bears stronger
relationships to the activity captured in the institutional statement.
However, despite the relational ambiguity that the Conditions compo-
nent exhibits, the substantive characterization of the kinds of context
characterizations (e.g., spatial, temporal, procedural, etc.) applies to both
conditions of the first and second kind—which we will subsequently label
as activation conditions and execution constraints, respectively.
Another component that introduces challenges to reliability on the
basis of ambiguous interpretation is the Object component. While it
essentially captures substantive institutional content, in the presence of
multiple objects in a statement (e.g., Inspector must send notification to

the applicant without delay, with notification and applicant as relevant
objects), prevailing coding practice (see Brady et al., 2018) shows a selec-
tive preference for specific objects either based on predefined criteria
(e.g., animacy) or study-specific heuristics. While useful to respond to
specific study objectives, it challenges reliable coding within, and more
importantly, comparative approaches across data sets. This concern is
interlinked with the observations related to the Conditions component,
since any object not identified as such in the corresponding IG compo-
nent (e.g., the applicant if notification is coded as Object and vice versa)
is necessarily coded as part of the Conditions component. As a conse-
quence, the selective coding of the Object enhances the ambiguity of the
Conditions component, both in terms of its variable semantic relationship
to other components (relational ambiguity) as well as substantive ambi-
guity of the latter component as it absorbs any parts of statements not
captured elsewhere, rendering it the function of a “fallback,” or “catch
all” component.
The following exemplary coding illustrates this observation:
Attributes: Inspector
Deontic: must
Aim: send
Object: notification
Conditions: to the applicant; without delay
The final component that drives an inconsistent treatment of insti-

tutional content is the Or else component. In contrast to ontological
inconsistencies identified for Conditions and Object components, the
Or else component highlights a consistency challenge related to granu-
larity. Borrowing the following example, Corporations listed in the Stock
Exchange must follow annual reporting provisions, or else may face sanc-
tioning by the Financial Oversight Commission, we arrive at the following
prototypical encoding:
Attributes: Corporations listed in the Stock Exchange

Deontic: must
Aim: follow
Object: annual reporting provisions
Or else: or else may face sanctioning by the Financial Oversight Commission.
The leading components (Attributes, Object, Deontic, Aim, Condi-

tions ) capture distinctive features necessary to reconstruct institutional
meaning in the first place, i.e., omitting any of those components
leads to a modification of institutional meaning (imagine the change
of Conditions ), or loss of meaning entirely (imagine the omission of
Attributes ).
Reviewing the Or else component, we first identify a semantic detach-
ment from the other components, in that it does not directly affect the
meaning of the leading components. A modification may well change
behavior of subjected entities when enacted (e.g., a Stock Exchange-listed
company would behave differently if it was not subject to consequences
for non-compliance), but it would not change the prescriptive content
of the leading statement (i.e., corporations still need to follow the provi-
sions). A more immediate point, however, is the structural incongruence
between the Or else component and the remaining components. Whereas
the leading components are necessary to construct a statement in the first
place, and while not always explicit, the Or else component itself reflects
the structure of an institutional statement in its own right. Drawing
on the example above, the Or else component can be expanded into a
self-contained statement as illustrated below:
Attributes: Corporations listed in the Stock Exchange

Deontic: must
Aim: follow
Object: annual reporting provisions
Or else:
Attributes: Financial Oversight Commission
Deontic: may
Aim: sanction
Object: corporation
It is important to note that this characterization requires the inter-

pretation of the “sanction” expressed in the original statement from an
institutional perspective, i.e., extracting the institutional meaning while
foregoing the linguistic expression. To this end, actor and subject to sanc-
tioning (Object ) need to be inferred, and thereby relying on the coder’s
ability to reflect on the institutional meaning of actions, involved roles,
etc., in order to reconstruct statements that express institutional infor-

mation following the uniform and unambiguously characterized syntactic
form of institutional statements.
While seemingly offering a source for reliability issues, only the recon-
struction of institutional acts (here, the sanctioning) in unified structural
form (here, in the form of the institutional statement components) allows
the explication of its essential features, and would otherwise pair the care-
fully parsed institutional information captured in the leading components
(i.e., corporations’ obligations) with a linguistic clause that at best signals
the presence of a sanction or incentive, but allows no further interpre-
tation in terms of involved actors, administered actions, as well as richer
contextual characterizations.
The essential insight conveyed at this stage is that the Or else compo-
nent displays structural incongruence in that it is a) more coarse-granular
than all other components, and b) limits the extraction of analytically valu-
able institutional meaning equivalent to the leading components. While
prominent in the Or else component, the structural embedding of insti-
tutional statements can similarly apply for Conditions components, albeit
for selected instances, an aspect we explore at greater depth in Chapter 5.
Table 3.1 highlights the different forms of ontological inconsistencies
as applicable for the IG components as discussed in this section.
While seemingly focusing on limitations of the original IG in terms of
structural aspects and implications of its application, establishing consis-
tency by resolving the challenges outlined above provides the basis for
a distinctive characterization of the IG in contrast to linguistic expres-
sion (an aspect we explore in the upcoming Sect. 3.3) and further marks
the consequent shift to a perspective that emphasizes the institutional
semantics when engaging in the encoding of institutional statements.
In doing so, the resolution of issues related to ontological consistency
enables a set of novel analytical opportunities toward the development of
Table 3.1 Ontological inconsistencies in the Institutional Grammar
IG component Type of ontological inconsistency
Conditions Substantive & relational inconsistency, structural incongruencea

Object Relational inconsistency
Or else Structural incongruence
a To variable extent
a systemic analysis of institutional design. Operationally, this includes the

resolution of semantic linkages of institutional statements established by
the Or else and the Conditions components as a basis to establish linkages
between institutional statements, either in the form of statement-level
or component-level interdependence and linkages, an aspect discussed
in Sect. 8.2.1.4. More immediately, however, is the resolution of onto-
logical inconsistencies, such as the ones outlined and motivated above.
Later discussion in the book specifically draws on these observations as a
motivation to introduce corresponding accommodations as part of the IG
2.0 outlined in Chapter 4.
Moving closer to the introduction of a refined characterization of the
IG, the following discussion picks up on the pivotal shift in perspec-
tive referenced throughout this section; the emphasis on meaning over
syntactic orientation of the IG.
3.2 Toward a Comprehensive

Representation of Institutional Meaning
The focus on the consistent conceptual characterization of components
discussed above invites for a reflection on what institutional statements
are, which features of an institutional system they capture, and the essen-
tial functions they exhibit. However, inasmuch as the discussion about
conceptual refinements motivates the consistent representation of institu-
tional content, this revision is further linked to methodological aspects as
relevant for the validity of coded institutional statements.
The following discussion addresses the analytical implications associ-
ated with the syntactic focus of the original IG (Crawford & Ostrom,
1995).
A practical consequence associated with the syntactic orientation is the
reliance on syntactic features alone when characterizing institutional state-
ments as different types; the IG suggests the identification of institution
types – strategies, norms, or rules – based on the presence or absence
of selected syntactic components of an institutional statement. While
such clear-cut characterization has theoretical appeal, in practice, the
specification of regulative statements does not follow the clear antecedent-
consequent (i.e., “do X, or else Y”) structure that constitutes rules in
the original IG. In other words, the activity and related consequences
may not always be captured in the same statement. Instead, we often
encounter examples where prescribed and punitive activities are sepa-

rated into separate statements. Consider the following example from the
context of organic farming:
Statement 1: “Certified organic farmers must comply with organic farming

provisions.”
Statement 2: “Program Manager may initiate organic license suspension or
revocation proceedings against a certified organic farmer that has violated
or is not in compliance with organic farming regulations.”
The first statement signals the compliance expectation attached to a

particular role, whereas the specification of potential sanctions in the
case of noncompliance are captured in an entirely separate statement
(Statement 2). Given the absence of an explicit syntactic Or else compo-
nent for each institutional statement, this separation leads to the mis-
characterization as two normative statements, as opposed to a rule, even
though the semantic interlinkage of both statements is unquestioned.
Reflecting on conventions of legal document structure more gener-
ally, writers may separate the specification of instructions and potential
sanctions in separate sections entirely, e.g., to aggregate compound sanc-
tions or to address different readership (enforcers vs. operational subjects
of regulation). In addition to the organizational features, the relation-
ships among statements oftentimes do not reflect a linear 1:1 relationship,
best illustrated with “blanket statements” that describe consequences of
violating or the invalidity of a collection of (or, in the extreme case, any)
other statements (e.g., salvatory clauses). An associated phenomenon can
be referred to as unidirectional referencing , in which only one state-
ment – generally the statement describing an obligation to monitor
behavior or a consequence of noncompliance – makes reference to the
activity it acts upon or governs (i.e., the statement prescribing behavior
to observed actors). Absent relevant contextual knowledge, instances of
statements that contain the behavioral expectation for the observed actor
itself, however, may simply not contain any explicit reference to the sanc-
tion. Consequently, such statement would be misconstrued as a “Norm”
following Crawford and Ostrom’s original syntactic classification.
An example of such is referenced by Siddiki (2014) who encountered
in her coding of regulation governing the practice of aquaculture a set of
statements that indicate that violation of any of the directives embodied
in the statements constituting the regulation will be subject to penal-

ties. As Siddiki notes, in the interpretation of Deontics, this means that
all statements contain an implied “must,” though the actual statements
may contain different Deontics. In addition, all those statements have an
implied sanction for non-compliance.
In practice, mere syntactic deviations or variations, such as the one
referenced above, lead to the dominant characterization of regulative
statements in policy documents as normative (in the Crawford and
Ostrom sense), simply because of the syntactic dissociation (and thus
presumed absence) of sanctions; for the purpose of encoding, the pres-
ence of an explicit Or else as expressed in the original IG specification is
then an exception, rather than the rule. These observations highlight that
the general reliance on syntactic features for both coding and classifica-
tion of statements creates challenges to validity, as the exclusive focus on
sentential structure is not able to resolve the semantic linkages obscured
by stylistic or structural divergence. Where the differentiation by insti-
tution type is deemed relevant, the characterization of statements based
on their contextualization in the wider statement context, capturing their
institutional meaning, is thus essential.
With specific focus on encoding and analysis, this underpins the
need for a paradigmatic shift in institutional statement coding from a
component-centric syntactic perspective to a semantic perspective. More
specifically, this implies a shift from a perspective that emphasizes the
retention of structural integrity of statements as expressed in policy by
adherence to syntactic features, construction of language, and embedded
stylistic nuances to one that emphasizes an accurate representation of
the institutional meaning of the statement – the function of the state-
ment in the wider institutional setting. An Institutional Setting is an
institutionally governed domain.
The objective to capture the meaning of a wider institutional setting
highlights a further concern that relates to the characterization of institu-
tional statements more generally. Inasfar as statements under the original
Grammar capture behavioral expectations, the comprehensive character-
ization of institutional settings requires the consideration of constitutive
statements that provide conceptual foundation on which regulative state-
ments draw to describe behavioral constraints. Exemplifying this, using
the statements below, we build on the example introduced before, where
the first statement describes a behavioral expectation in the form of an
obligation. The second statement complements the initial regulative one
by defining the underlying concept “certified organic farmer” in the

context of the policy.
Statement 1: “Certified organic farmers must comply with organic farming

provisions.”
Statement 2: “Certified organic farmers are farmers that have undergone a
certification procedure according to United States Department of Agriculture
National Organic Program Provisions.”
Statements of the latter kind, constitutive statements – without further

elaboration at this stage3 – essentially “set the stage” for the specifi-
cation of regulative statements in the first place. Seen from an institu-
tional perspective, constitutive statements describe, declare and (ideally)
relate all concepts (e.g., actors, actions, objects or mental constructs)
of relevance in an institutional setting, and potentially the institutional
setting itself (e.g., venue, infrastructure). Attempting to attain an under-
standing of the wider institutional setting thus relies on the existence, and
ideally antecedent specification (Searle, 1969), of conceptual specifications
in the form of constitutive statements that regulative characterizations can
draw from.
The concerns raised in this section, both the limitations of focusing on
a syntactic interpretation of institutional statements to infer their function
in the wider institutional context, as well as the absence of constitutive
statements that afford a comprehensive representation of the institutional
arrangements, including actors, actions and other concepts, motivate
the aforementioned paradigmatic shift toward a semantic perspective on
institutional statements. Such shift enables the departure from analyzing
formal institutions as collections of institutional statements and provides
the basis for a systemic analysis of institutional arrangements, in which
the functional linkages expressed in individual institutional statements
support the extraction, and, in consequence, reconstruction, of the insti-
tutional system in a “language” that enables the analyst to “interrogate”
the institutional arrangement from a wide range of perspectives and using
a diverse range of techniques.
Drawing on the observations made throughout this section, and the
motivations for an “institutional language” expressed alongside, as a final
3 The concept of constitutive rules is discussed from Chapter 4 onward.

discussion point, in the upcoming section we attempt to distill the objec-

tives and features associated with linguistic grammars and contrast those
with the objectives associated with the Institutional Grammar.
3.3 Grammars in the Linguistic

and Institutional Sense
As noted in Chapter 1, when referencing an institutional “grammar,” or
an institutional “syntax,” the reader may be inclined to draw linkages
to a grammar in the linguistic sense. In fact, to appreciate the analytical
opportunities leveraged by an “institutional grammar,” it is important to
realize the distinctive differences between a linguistic grammar, and the
Institutional Grammar, an aspect that Crawford and Ostrom’s seminal
work and the prevailing coding practice address only to a limited extent.
Providing a basis for a qualitative discussion, the central purpose of a
linguistic grammar is to capture word structure (morphology), sentence
structure (syntax), meaning (semantics), and finally practices around the
context-dependent use of language (pragmatics), with linguistic analysis
further considering its spoken production (phonology) as an elementary
aspect of language (Winograd, 1983).
However, inasmuch as linguistics agree on the distinctive features, the
way in which humans generate or extract meaning from those classes has
been approached from different perspectives. The currently dominating
linguistic perspective suggests the existence of an innate universal set of
syntactic primitives common to all humans that is combined based on
fixed grammatical rules, and, in this process, generates meaning accessible
to both producer and receiver, a grammar characterization referred to as
Generative Grammar, and commonly associated with the work of Noam
Chomsky (Chomsky, 1965; Jackendoff & Pinker, 2005). Essential to this
interpretation is that the syntax itself carries and produces the meaning of
the content based on the systematic combination, while, at the same time,
allowing for the isolated analysis of language based on different levels of
linguistic analysis (e.g., syntax, semantics).
A contrasting view on the production and interpretation of language
is the notion of the Cognitive Grammar (Langacker, 2008), often
referenced in conjunction with Construction Grammars (Hoffmann &
Trousdale, 2013), which puts the emphasis on a language-independent
interpretation by avoiding the isolated and sequential interpretation of
a text based on distinctive levels of language (i.e., syntax, semantics,
morphology), but posits that the understanding of language operates

across those levels based on general processes linked to neural activity
patterns. These processes are shared among individuals on a biological
basis, and relate to the establishment of psychological connections (associ-
ation), automated production of language based on entrenched structures
(automatization), the generalization of observed structures (schematiza-
tion), and the corresponding application of existing understanding to
categorize new expressions (categorization) (Broccias, 2013). All these
processes are integrated based on a contextual frame that guides the
production and corresponding interpretation. This focal emphasis on
processes and associated patterns motivates the dissociation from language
specifics (let alone specific languages or language families), but instead
suggests the mapping of expressions based on the abovementioned
schematic structures.
Returning to the Institutional Grammar perspective, we recognize
two aspects that are closely related to the interpretation of institutional
settings and construction and regulation of activity therein. Firstly, the
institutional meaning of a statement, i.e., its effect within an institutional
setting, requires a shared frame of reference. This is not grounded in
any specific domain, but in a shared conceptual framing, which, in the
case of the IG, is the action situation. The action situation, as intro-
duced in Chapter 1, provides the boundaries and typology of entities
involved, including actors and actions, and the bio-physical as well as
institutional context the action situation is embedded in. The construc-
tion of meaning within this shared frame of reference is based on shared
schematics, or patterns, which, in the case of the IG, are institutional
statements. This conception is inherently compatible with the institution
characterization offered by Campbell (2020) who interprets institutions
as cognitive schemata that organize individuals’ understanding of the real
world and provide guidance for their behavior (see Chapter 1), a perspec-
tive that is compatible with interpretations offered by Scott (2013) and
Castelfranchi (2014).
Institutional statements themselves, generally referenced based on their
syntax only, then carry not only the structural but also semantic integrity
needed for their consistent interpretation (i.e., any component has a
distinctive function and explicit semantic relationship to all other compo-
nents of the statement, such as the Attributes to the Aim). Observed
from this perspective, an Institutional Grammar is inherently agnostic of
language per se, but merely requires the extraction of institutionally rele-
vant information (however it is represented there) and its positioning in
the structural frame or schema that an institutional statement represents.
We see the latter, the essential detachment from specific forms of natural
language grammar (despite potential incidental overlaps), as a specific
feature that makes the Institutional Grammar a lingua franca for the
articulation of institutional arrangements. Its ability to do so comprehen-
sively and without ambiguity is then subject to structural and semantic
integrity of the institutional statement components themselves, an aspect
that has been motivated above and will addressed in the upcoming
chapters.
This fixed structural frame naturally limits the expression of non-
institutionally relevant information (e.g., speaker/writer characteristics,
style, open-ended sentence construction) and invariably focuses on struc-
tural units that capture essential institutional information. While linguistic
analysis following the generative tradition operates on word/token, clause
and sentence level, the Institutional Grammar recognizes component and
statement as corresponding elementary and compound units of analysis.
A practical consequence is that IG components do not necessarily only
correspond to specific words or tokens, but may furthermore map to
clauses. Using a running example to motivate this approach based on
the high-level characterization of the IG in the previous chapter, we can
suggest that.
Organic farmers must fulfill their reporting duties before the end of the year.
While the identification of the actor (Organic farmer), activity (fulfill )

and object (reporting duties ) provide a close match, the contextual char-
acterization (before the end of the year) is expressed in an extended
clause.
The observant reader may be inclined to infer patterns that map expres-
sions provided in natural language onto components of the Institutional
Grammar. While generally observable for simple statements (e.g., the
correspondence of the S-V-O structure on Attributes, Aim, and Objects ),
this mapping becomes unreliable with increasing statement complexity.
Relying, for example, on the assumption that a subject in the linguistic
sense always reflects the responsible actor of an institutional statement,
may oversee that an action (e.g., “received”) may signal an inverted direc-
tionality that in fact characterizes the entity as the receiver of action in the
institutional sense. Similarly, drawing on prepositional clauses alone for

characterization of context is challenging due to the polysemous nature,
various forms, and senses that those represent (Srikumar & Roth, 2013).4
Inasfar as morphology is concerned, and motivated more explicitly
in Chapter 5, a specific linguistic feature that is either applied selec-
tively (e.g., as a concession to style) or systematically (e.g., based on
usage conventions) is the notion of conceptual reification, a pattern
linked to the mental representation of language argued in the context
of the Cognitive Grammar referenced above. Conceptual reification, is
the expression of actions, mental constructs (e.g., beliefs) and other
nonphysical concepts in nominalized, or “thingified” form (Langacker,
2008).
Any reference to abstract concepts can of course invite for discourse,5
but in this context we reference (and subsequently discuss) commonplace
expressions found in policy text, such as Upon observation of a violation,
the inspector may revoke the license …
The Institutional Grammar seeks the explication of the essential institu-
tional content, which can involve the reconstruction of clauses to simplify
expression, formulate activities in active tense, but, most importantly,
the inference of implied components (e.g., actors, actions, context). For
example, to attain the idiomatic structure of an institutional statement as
far as introduced to this stage (i.e., identifying responsible actor, action,
etc.) without compromising semantic integrity, the leading conditional
clause of the abovementioned expression can be reconstructed as When
the inspector observes a violation, …, thus calling out the actor (inspector)
explicitly, while reconstructing the conceptually reified “observation” in
terms of an activity. The detection and resolution of such reification
necessarily requires an understanding of the institutional setting (actors,
activities, etc.) in order to make underlying institutional configurations
overt.
4 Foregoing further exploration at this stage, we will continue this discussion following
the introduction of the complete Institutional Grammar from Chapter 4 onward.
5 An analytical challenge at the core of sociological study (and in extension institu-
tionalism) is the separation of social facts from tangible ones (Durkheim, 1964), and
potentially leading to the “fallacy of misplaced concreteness” (Whitehead, 1925, p. 52)
by suggesting unjustified characterization of concepts (e.g., “the State”) as a cohesive and
consistent unit.
In natural language, the structure and choice of expression in a

sentence can be used to express nuance, develop coherent narratives, and
often makes concessions to pragmatics associated with specific domains
and the contextual use of language. This flexibility is desirable in the
context of human language, e.g., to carry narratives in novels, align-
ment of phonetic and semantic patterns in poems, and specificity in legal
expression. However, unlike natural language, the objective of the Insti-
tutional Grammar is clearly delineated: the reconstruction of institutional
configurations expressed in terms of potentially interlinked institutional
statements.
In doing so, the Institutional Grammar seeks structural consistency
that organizes the semantics of a statement alongside specific language-
independent syntactic patterns that carry distinctive institutional meaning,
framed by the institutional setting as shared conceptual reference. Instead
of imposing limitations on the analytical opportunities and extent to
which institutional information can be extracted based on the fixed
structural unit of institutional statements, in its revised form presented
throughout this book, the Institutional Grammar establishes interme-
diate patterns of statement linkage that allow the analyst to detect and, if
needed, reconstruct language systematically to accommodate the uniform
structure, but without compromising semantic integrity in the process
(from an institutional perspective).
The basic patterns of the Institutional Grammar referenced throughout
this section build on the characterization of institutional statements as
parameterizing (i.e., constituting) and/or regulating in kind. Broadly
conceived, parameterizing statements define an institutional setting,
including the actors, actions, status and artifacts embedded therein. Regu-
lating statements capture expectations associated with individual actors’
behaviors in the form of obligations, prohibitions, or permissions.
The mapping of language into schematic institutional statements is best
illustrated in examples. Reflecting on the example shown in Fig. 3.2, we
can observe the dual function the statement has with respect to behavior
regulation, on the one hand, namely the specification of duty, and the
parameterizing function on the other – here the specification of reporting
duties in the context of this institutional setting.
Assuming both regulating and parameterizing function, the sentence
shown here thus comprises of patterns reflecting two linked institutional
statements. Highlighting the dissociation of linguistic and Institutional
Grammar, these statements can be rewritten in the following form (see
Fig. 3.2 Combined Institutional Grammar patterns example
Fig. 3.3) without sacrificing semantic integrity from an institutional

perspective, while certainly affecting style and syntactic construction from
a linguistic perspective.
Table 3.2 highlights the essential differences between a grammar
understood in the linguistic and the institutional sense, and specifically
the varying levels of abstraction on which either grammar operates.
Fig. 3.3 Separated Institutional Grammar patterns example
Table 3.2 Comparison of key characteristics between linguistic and Institu-

tional Grammar
Characteristic Linguistic Grammar Institutional Grammar
Unit of analysis Sentence Institutional Statement

Objective Expression of nuance in Capturing institutional
content meaning structure and style
in detail
Facilitation Structural diversity based on Structural homogeneity based
flexible syntax on uniform structure &
linkage patterns
Contextual anchor Purpose and narrative of text Institutional Setting
Conceptual primitives Tokens/Words, Clauses, Components, Institutional
Sentences Statements
With the distinctive objective of the Institutional Grammar to capture

human coordination mechanisms generically, and the associated abstrac-
tion from the linguistic perspective, it holds the promise to operate
across languages, language families and various forms of context. While
introduced in the context of the English language (and hence appli-
cable to other Germanic languages), the original operationalization
of the IG in generic game-theoretical terms (e.g., actors, strategies)
(see Crawford & Ostrom, 1995), as well as the subsequent application
across other language families [Romance (Pacheco-Vega, 2020), Slavic
(Dunajevas & Skučiene, 2016)] support the claim that the IG is a lingua
franca for institutional analysis. This proposed conceptual generality,
however, should not overshadow the customization based on method-
ological considerations regarding (a) disciplinary traditions (irrespective
of linguistic context), and (b) potential language-specific conventions
that influence the construction – an aspect explicitly discussed as part of
the applied encoding of institutional content in Chapter 7.
Reflecting on the discussion to this stage, Chapter 1 provided the moti-
vational backdrop that highlights the contributions and applications of the
existing IG, alongside subsequent adaptations to the Grammar over time
that set the stage for this book. An overview of relevant research fields and
the distinctive contributions that the Grammar has offered is introduced
in Chapter 2. Building on these, this chapter has offered a synthesis of
various considerations, including challenges related to conceptual validity
(ontological inconsistencies), and an explicit characterization and conse-
quent distinction between grammars in the institutional and the linguistic
sense. This is augmented with an overarching call for richer and consis-
tent methodological considerations that respond to different forms of
data collection and analytical techniques, and motivate the shift from a
component-centric syntactic to a statement-centered semantic perspec-
tive. With this narrative in mind, in the upcoming Chapters 4–6, the
discussion will move toward the comprehensive conceptual introduction
of the IG that responds to aforementioned challenges, before attending
to methodological and analytical aspects and opportunities in Chapters 7
and 8, respectively.
References
Agazzi, E. (2011). Consistency, truth and ontology. Studia Logica, 97 , 7–29.
https://doi.org/10.1007/s11225-010-9295-x
Center for Behavior, Institutions and the Environment. https://complexity.
asu.edu/sites/default/files/papers/cbie_wp_2018–006_0.pdf
Broccias, C. (2013). Cognitive Grammar. In T. Hoffmann & G. Trousdale
(Eds.), The Oxford Handbook of Construction Grammar (pp. 1–14). Oxford
University Press. https://doi.org/10.1093/oxfordhb/9780195396683.013.
0011
Campbell, J. L. (2020). Institutional change and globalization. https://doi.org/
10.2307/j.ctv131bw68
Castelfranchi, C. (2014). Minds as social institutions. Phenomenology and the
Cognitive Sciences, 13(1): 121–143. https://doi.org/10.1007/s11097-013-
9324-0
Chomsky, N. (1965). Aspects of the theory of syntax. MIT Press (MA).
Crawford, S. E. S., & Ostrom, E. (1995). A Grammar of Institutions. Amer-
2975
Dunajevas, E., & Skučiene, D. (2016). Mandatory pension system and redistribu-
tion: The comparative analysis of institutions in baltic states. Central European
Journal of Public Policy, 10(2). https://doi.org/10.1515/cejpp-2016-0025
Durkheim, E. (1964). The rules of sociological method. Free Press.
Ellis, B. (2007). Scientific essentialism. Cambridge University Press.
nADICO: A Nested Grammar of Institutions. Lecture Notes in Computer
Science, 8291 LNAI , 429–436. https://doi.org/10.1007/978-3-642-44927-
7_31
Hoffmann, T., & Trousdale, G. (2013). Construction Grammar: Introduction.
In The Oxford Handbook of Construction Grammar. Oxford University Press,
Oxford (UK).
Jackendoff, R., & Pinker, S. (2005). The nature of the language faculty and
its implications for evolution of language (Reply to Fitch, Hauser, and
Chomsky). Cognition, 97 (2), 211–225. https://doi.org/10.1016/j.cognit
ion.2005.04.006
Langacker, R. (2008). Cognitive Grammar: A basic introduction. Oxford Univer-
sity Press. https://doi.org/10.1093/acprof:oso/9780195331967.001.0001
Pacheco-Vega, R. (2020). Governing urban water conflict through water shed

councils: A public policy analysis approach and critique. Water (Switzerland),
12(7). https://doi.org/10.3390/W12071849
22(4), 445–475. https://doi.org/10.1177/1043463110377299
Scott, W. R. (2013). Institutions and organizations: Ideas, interests, and identities.
4th edition, SAGE Publications Inc.
University Press (UK).
States. Review of Policy Research, 31(4). https://doi.org/10.1111/ropr.
12075
97.x
Srikumar, V., & Roth, D. (2013). Modeling Semantic Relations Expressed by
Prepositions. Transactions of the Association for Computational Linguistics, 1.
https://doi.org/10.1162/tacl_a_00223
Whitehead, A. N. (1925). Science and the modern world. Free Press (Simon;
Schuster).
Winograd, T. (1983). Language as a cognitive process: Volume 1 - Syntax.
Addison-Wesley.
CHAPTER 4
Institutional Grammar 2.0: Conceptual

Foundations and General Syntax
4.1 Conceptual Foundations

Building on the motivation set out in previous chapters, this chapter
focuses on the introduction of the Institutional Grammar concept, and
highlights its essential features. Most notably, it will draw on the latest
conceptualization of the Institutional Grammar, IG 2.0, that addresses
many of the observations raised in Chapter 3, leveraging a richer institu-
tional representation that moves beyond the mirroring of the linguistic
grammatical structure by capturing variable levels of structural and
semantic detail mediated via features stratified across a range of levels of
expressiveness. A term that finds use throughout the remainder of the
book is the institutional configuration. An Institutional Configuration
is a set of institutional statements that are directly, or indirectly linked
(e.g., based on implied reference) within or across action situations, and
can be analyzed at different scales (e.g., component, statement, set of
statements).
In this form, the Institutional Grammar addresses a set of essential
objectives, including:
• preempting ontological inconsistencies by offering refined syntactic

components that capture the content of statements consistently and
unambiguously,

Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2_4
• capturing institutional settings comprehensively based on both a

regulative syntax that expresses behavioral expectations, as well as
a constitutive syntax that reflects the parameterization of the system
by defining entities (such as actors, behavior, roles, venues, etc.) in
the first place, and finally
• affording accessibility for diverse use cases based on the computa-
tional tractability of institutional statements.
The Institutional Grammar aims at being an interface that allows

for diverse analytical use cases, spanning from the assessment of statis-
tical distributions of components, via the deeper structural assessment
of institutional design toward logical representations of institutions that
introduce analytical opportunities not accessible with the original Insti-
tutional Grammar, and thereby broadening the use cases for it along
and beyond the research strands outlined in Chapter 2. More specifi-
cally, making the Institutional Grammar amenable to diverse analytical
techniques (e.g., computational modeling, reasoning), the richer charac-
terization of the Grammar opens up to entirely novel applications of it,
such as the representation of “institutions in use” that complement the
traditionally primary concentration on “institutions in form,” as well as
novel structural analyses that draw on compositional details exposed by
institutional statements—aspects discussed at greater depth in Chapter 8.
4.1.1 Levels of Expressiveness

Realizing the diverse affordances of different disciplines and associated
techniques, the Institutional Grammar 2.0 recognizes three distinct Levels
of Expressiveness that vary in conceptual richness and focus that respond
to the levels of representational detail and complexity linked to various
analytical objectives.
These levels, namely IG Core, IG Extended (introduced in Chapter 5),
and IG Logico (introduced in Chapter 6), are progressively capturing
more detail about an institutional setting, and, consequentially, extract
richer structural information embedded in institutional statements. The
different levels further reflect a progression from a focus on struc-
tural features of individual statements and collections thereof toward
a perspective that focuses on systemic features of institutional configu-
rations, specifically capturing the meaning and interrelationships tacitly
4 INSTITUTIONAL GRAMMAR 2.0 … 77
expressed in institutional text. The levels are intended to be backward-

compatible, i.e., information encoded at higher levels of expressiveness is
also accessible to analysis at lower levels of expressiveness.
The IG 2.0 explicitly recognizes the potential dissociation between a
methodological perspective that primarily emphasizes the coding itself,
and an analytical perspective that orients the feature selection and other
methodological choices on the analytical objectives of a given study.
Whereas the latter reflects the scientific intuition of initiating any research
with research questions, and organizing the methodology as a reaction,
datasets may conceivably be encoded without a particular perspective in
mind and may be encoded in more general form that is intentionally
indifferent about the practical uses (e.g., large-scale data set creation).1
Emphasizing the analytical perspective, the analyst is tasked to identify
the level of expressiveness that aligns with her analytical objectives, where
the first two levels primarily focus on capturing institutional content,
with IG Core establishing a coding that focuses on capturing institu-
tional information in broad syntactic categories ensuring a consistent
characterization of individual components. In contrast to the high-level
parsing applied in the context of IG Core, IG Extended focuses on a
deep structural parsing that refines IG Core coding by revisiting indi-
vidual syntactic components to extract structures nested therein, building
up to the exploitation of the structural information to afford systemic
analyses (see Sect. 8.2.1). IG Logico features a semantic overlay that oper-
ates across all syntactic components parsed either on IG Core or Extended
level, building the basis for advanced analysis that afford the ontological
linkage with discipline- or domain-specific concepts (see Sect. 8.3). While
offering distinctive analytical opportunities, the levels do not exist in isola-
tion. Instead, movement between different levels reflects an incremental
extension of the feature set offered on preceding levels, an aspect captured
in the schematic representation in Fig. 4.1 that associates the coder and
analyst perspectives with the corresponding levels.
This, as a consequence, offers the flexibility to navigate various objec-
tives associated with the individual levels: institutional content captured at
greater structural detail (e.g., in IG Extended) can always be reduced to a
more coarse-grained form (e.g., IG Core), thereby essentially promising
“backward compatibility” across levels – and the opportunity to leverage
wider and broader use of coded datasets. Conversely, this implies “upward
1 Implications deriving from either perspective are discussed in Chapter 7.

Fig. 4.1 Levels of expressiveness in the Institutional Grammar 2.0
compatibility” and affords synergies when attempting to draw on features

captured in higher levels of expressiveness: instead of recoding all state-
ments entirely, recoding merely involves revisiting the existing coding in
order to offer an incremental expansion of coding detail.
To realize this objective, the “Grammar” relies on an inclusive and
broadly captive base structure, rooted in the syntactic structure estab-
lished as part of IG Core that builds on the original IG by Crawford and
Ostrom.
However, inasmuch as IG 2.0 aims at providing a coherent and
consistent structure by resolving ontological inconsistencies discussed
in Chapter 3, it further aims at capturing institutional configurations
comprehensively. In this way, one can conceive of the institutional anal-
ysis occurring at the IG Core level as being of a more basic sort, while
the institutional analysis occurring at the IG Extended and IG Logico
levels as being of a more comprehensive sort with variable emphasis on
structure versus semantics.
4.1.2 Constitutive and Regulative Statements

The IG 2.0 recognizes institutional statements of both regulative and
constitutive kind, and thus defines institutional statements generally as
a statement that describes actions for actors within particular contexts
or parameterizes features of an institutional system within particular
contexts. As referenced throughout the preceding chapters of the book,
regulative statements describe actions for actors within particular contexts.
They may further indicate prescription and consequences related to the
referenced action. Constitutive statements, in contrast, build on the
constitutive rule conception (Cherry, 1973; Searle, 1969) and define insti-
tutional (f)acts, either based on brute (f)acts or based on existing insti-
tutional (f)acts (Cherry, 1973), defining behavior in the first place. More
broadly, in the context of IG 2.0 constitutive statements parameterize
features of the institutional system within particular contexts.
Whereas signing a lease contract, for instance, builds on the physical
(brute) act of signing a piece of paper, institutionally, it has distinctive
effects that transpire into the real world: it reflects the commitment and
establishes rights and liabilities associated with entering a lease, thereby
creating institutional facts that modify the status of the interacting partic-
ipants (e.g., making the signatory a “lease holder”). Without further
discussion at this stage,2 constitutive statements do not express behav-
ioral expectations in terms of duties and permissions as applicable to
individual actors or sets thereof, but emphasize the institutional setting
in a wider sense – effectively setting the stage of the “institutional play.”
This includes the specification of involved actors, behavior, and role
conceptions, but also the definition of objects or artifacts, among further
features of relevance in an action situation. In addition to the substan-
tive characterization of entities, constitutive statements can further define
relationships among any of such entities, including the embedding in
organizational (e.g., hierarchical) settings, and importantly, the specifi-
cation of status that the entity holds or attains in this institutional setting
(e.g., rights, powers, liabilities, etc.). Further, constitutive statements can
variably introduce (e.g., through definition) or modify (e.g., through
redefinition, or introduction of novel conceptual relationships) entities
antecedently, or do so dynamically, i.e., in an instantiated institutional
setting.
2 Discussions around the distinctive features and contrast to regulative statements are
provided in Chapter 7.
Fig. 4.2 Interaction of constitutive and regulative statements
In the idealized form, regulative statements coordinate behavior by

drawing on the entities defined in the institutional setting. More generally,
constitutive statements define and introduce entities in the institutional
action space, i.e., the actors and acts that carry institutional meaning
and effect, as well as the associated affordances, i.e., what any insti-
tutional act can possibly act upon, whereas regulative statements draw
the coordinative links between those entities based on an actor-centric
operationalization and the consideration of environmental constraints. In
essence, constitutive statements provide the fixtures on which regulative
statements anchor.
In this function, the performance of action based on behavioral
prescription can produce feedback effects that affect the wider institu-
tional setting, let alone potential external effects outside the institutional
system. Figure 4.2 abstractly highlights the complementary operation of
constitutive and regulative statements: artifacts are defined or modified by
constitutive statements, by drawing on brute (f)acts or, in self-referential
form, on already established institutional (f)acts. Regulative statements,
in contrast, exclusively draw on explicit or implicit institutional (f)acts3 to
capture behavioral guidance.
Recognizing the complementary role of both statement types, the
understanding developed at this stage provides a basis for introducing
3 This chapter returns to the discussion and accommodation of empirically-observed

interlinked constitutive and regulative statement structures in Sect. 5.2.
the comprehensive capturing of institutional arrangements based on

In response to the systemic view and complementary function of
constitutive and regulative statements introduced in IG 2.0, we deviate
from Crawford and Ostrom’s original characterization that focuses on
behavioral regulation only, and define an institutional statement as
follows: An institutional statement describes expected actions for actors
within particular contexts, or parameterizes features of an institutional
system.
Against this backdrop, the Institutional Grammar provides the neces-
sary “grammar” that affords a comprehensive representation that is
inclusive of both types of institutional statements, alongside potential
mixed forms (see Sect. 5.2), while capturing structural and semantic
detail.
Where all levels share the basic component characterization as defined
at the lowest level of expressiveness, IG Core, the subsequent levels intro-
duce additional features that refine individual components or require
characterization of the statement as a whole. In addition to offering an
interface for different analytical objectives, the encoding on various levels
invariably operates in a trade-off between human readability on the one
hand and cognitive load on the other hand. The transition from the
lowest level IG Core to IG Extended, for example, implies a consider-
able increase in cognitive effort during the encoding process due to the
endemic structural complexity IG Extended captures. The transition from
IG Extended to IG Logico, in contrast, likewise carries an increase in
cognitive load, primarily sponsored by the shift from a structural to a
semantic assessment of individual statements and arrangements thereof,
as well as selected concessions to machine readability that challenge the
accessibility of encoded text.
The intuitions underlying these trade-offs, in preparation for the
systematic selection of feature guidance as part of the analytical use
(discussed in Chapter 7) are best supported by introducing the individual
levels of expressiveness, alongside the associated feature set.
4.2 IG Core
Providing the basis for the Institutional Grammar, IG Core builds on the
original syntactic characterization of the Institutional Grammar offered
by Crawford and Ostrom (1995), in addition to the consideration of the
Object component (Siddiki et al., 2011), as well as the principles of Nested
Institutional Statements (Frantz et al., 2013). The objectives on this level

are the establishment of a comprehensive coarse-grained encoding by
taking into account statements of both constitutive and regulative kind,
resolution of ontological inconsistencies discussed in Chapter 3, most
notably the undifferentiated use of the Object and Conditions compo-
nents, as well as a conceptual revision of the Or else component. At the
same time, however, IG Core seeks to maintain general compatibility with
established coding practice (see Chapter 2).
Initiating the introduction of the basic syntactic components on
which the IG rests, we initially introduce all components relevant to
construct regulative statements, before continuing with the discussion of
components composing constitutive statements, and concluding with the
discussion of hybrid forms at the end of this chapter.
4.2.1 Regulative Syntax

In the IG, regulative statements that operate with the primary purpose of
specifying behavioral expectations consist of the following components:
Attributes, Deontic, Aim, Object, Context, and Or else.
To motivate the individual components and their composition into
institutional statements, we will selectively draw on the following stylized
examples of policy statements:
• Drivers must follow traffic regulations whenever driving on public

roads, or else they face a fine.
• All corporations, including the ones publicly listed in stock exchanges,
as well as private ones, must submit annual reports to the Tax Revenue
Service in a timely manner following the closing of the tax year, or else
the Tax Revenue Service may impose sanctions.
• Registered voters may cast their vote every four years.
4.2.1.1 Attributes
The Attributes describes the actor, whose behavior is regulated in an insti-
tutional statement. The actor can be an individual or corporate (e.g.,
juridical person), who either carries out, or who is expected (not) to
carry out (as indicated by the Deontic component) a given action (spec-
ified in the Aim component). The characterization of the actor can be
explicit (e.g., identifying a specific entity by name), be based on specific
attributes that the actor (or group of actors) holds, or be in anthropo-

morphized form.4 The Attributes component subsumes both distinctive
characterizations of the actor, e.g., based on role characterizations (e.g.,
Drivers, Citizens ), as well as descriptive or otherwise qualifying Proper-
ties associated with the actor (“Registered (voters)”, “(including) ones
publicly listed in stock exchanges ”) that can be of diverse kind (e.g., demo-
graphic characteristics, physical, or relational properties). Such properties
can further include qualifiers (e.g., older than 16 years), and be negated
(e.g., not older than 16 years), or be quantitative in nature (e.g., some,
all vs. no farmers). Independent of type and specificity, the number of
such properties is open-ended. However, while recognizing the different
characterizations of property information at this stage, on IG Core level,5
the identification and separation of properties from Attributes entity is
sufficient. It is further important to note that in practice the Attributes
component is oftentimes referred to in its singular form (Attribute),
essentially rendering the use synonymous to the plural form, especially
where the referenced actor is an unambiguously identified actor.
The presence of an Attributes component is compulsory for regula-
tive statements. However, the specification of the component may not be
explicit but be contextually implied.6
4.2.1.2 Deontic
The Deontic component explicitly defines whether an action of an insti-
tutional statement is compelled, restrained, or discretionary, and more
specifically, captures an actor’s “duty” (or lack thereof) to perform
a particular activity. The concept of the Deontic in the Institutional
Grammar is closely associated with the principles of deontic logic (von
Wright, 1951). Deontic logic (as formalized by von Wright) offers a
formal characterization of the logical relationships among the permissible,
obligatory, and conversely, the forbidden, based on their interdefinability,
i.e., the ability to define any two deontic primitives based on the respec-
tive other one. We can motivate this point by relying on standard deontic
4 A statement may, for instance, prima facie regulate the behavior of cars, while in fact
regulating the driver’s behavior, e.g., “Cars must stop at zebra crossings.”
5 Further features including a typology of properties, as well as relational characteristics
of properties are addressed in the context of IG Extended (Sect. 5.1) and IG Logico
(Sect. 6.1).
6 See Chapter 7 for an extended discussion of inferring contextually implied compo-
nents.
logic operators, with P signaling a permission, O an obligation, and F a

prohibition. Relating these primitives, the negation (¬) of an obligation
to perform or not to perform an action act (i.e., ¬O [act or ¬act ]) signals
the discretionary nature of this activity (P act ). Intuitively, if taxpayers are
neither obliged to file nor obliged not to file tax returns, it signals that
this activity is discretionary. However, if one were only to express an obli-
gation to not perform an act (O ¬act ), e.g., an obligation to not file tax
returns, this would be equivalent to a prohibition to deliver tax returns,
i.e., F ≡ O ¬act.
Both the intuitive accessibility and logical foundation support the
formal treatment of institutional statements. However, and despite the
intuitive accessibility, it is not without challenges. On the one hand,
deontic logic is met with a range of philosophical and logical challenges
(Hansen et al., 2007), referred to as paradoxes, in which logical conclu-
sions do not correspond to associated intuitions or produce contradic-
tions. However, a practical challenge to the discretized tripartite structure
of the Deontic is the observation that institutional statements invariably
express nuances in “Deontic strength” – the extent to which a Deontic
is discretionary or normative –, and thus inadvertently position expecta-
tions about regulated behavior along a Deontic continuum ranging from
prohibition to obligation, where the center of this continuum reflects
the optionality of such activity (as a further deontic modal in addition
to the obligatory, impermissible (read “prohibited”), permissible, ought
and omissible [McNamara, 2006]). While intuitively often conflated with
permissiveness, the logical relationships between obligation, prohibition,
and permission are incongruent: whereas an activity that is obliged can
be assumed permitted, the same cannot be said about an activity that
is prohibited; it is not permitted. Figure 4.3 schematically visualizes
Fig. 4.3 Continuous Deontic conception

this continuum, reflecting the discretized normative functions pro- and

prescriptions assume linked via variable levels of discretion.
Operationally, Deontic values associated with particular directives can
variably be parameterized based on explicit domain-specific legal conven-
tions (e.g., Bodansky, 2016), technical specifications (e.g., Bradner,
1997), or be based on empirical observations (see e.g., Frantz et al.,
2015a).
Beyond the immediate operationalization, the regulative effect of
the Deontic can be moderated by contextual factors, an aspect we can
exemplify using the following illustrative example: “Corporations must
follow regulations regarding environmental pollution, especially if situ-
ated in sensitive socio-ecological settings.” In this statement, the qualifying
clause “especially …” signals a differentiated interpretation of the Deontic
must (generally signaling an obligation with limited discretion), implicitly
elevating the level of prescriptiveness despite the in principle unchanged
Deontic. This example highlights that the analysis of the Deontic – while
conceptually intuitively accessible – cannot occur in isolation, but rests on
the contextualization with the surrounding statement components.
In the context of regulative statements, the presence of the Deontic
component is optional.
4.2.1.3 Aim
The Aim component reflects an activity, goal, or outcome regulated
by the specific institutional statement, and is associated with a given
actor specified in the Attribute component. Given the use of the Aim
component as the regulated activity, where constraints and guidance are
expressed in Context and Deontic components, a central prerequisite is
that the performance of the action specified in the Aim must be physi-
cally possible; this implies the negation of actions since the Institutional
Grammar describes behavioral regulation (and establishment thereof), not
physical laws.
Referencing the examples at the beginning of this section, the respec-
tive aims, or regulated activities, include follow (traffic regulation), submit
(annual reports), and cast (vote), all of which can in principle be
performed, or not performed (negated).
Naturally, the presence of this component is necessary for any regula-
tive statement.
4.2.1.4 Object
In many instances, institutional statements not only constrain actor
behavior, as expressed in the Attribute and Aim components but further
involve objects that are directly or indirectly affected by the perfor-
mance of the regulated activity. Objects, first introduced into the IG
syntax by Siddiki et al. (2011) and Smajgl et al. (2008) as a response to
observing the absent characterization of entities that are receiver by regu-
lated activity, draw the analytical linkage between behavioral prescription
in the form of Attribute and Aim, and entity affected by this behavior.
Given challenges to reliability in statements where multiple objects exist
(as highlighted in Chapter 3), in IG 2.0 Objects are recognized in two
distinctive forms: Objects that are directly affected by the action execu-
tion, reflecting the direct receiver of the action, are Direct Objects . Where
the application of a particular action to an object is targeted toward, indi-
rectly affecting, or otherwise experienced by another object, the latter
is identified as the Indirect Object . The principal relationships between
action (Aim), Direct and Indirect Object are highlighted in Fig. 4.4.
Illustrating the application using the above-mentioned examples,
traffic regulations are referenced in the activity follow. Similarly, the
activity submit is directed at annual reports. However, in this example,
the statement includes the tax revenue service as an entity affected by
the action-object application – the Indirect Object – as a receiver of the
annual reports. As with the Attribute component, objects can have their
own properties of diverse kind. Whereas IG Core focuses on the identi-
fication of such properties in the first place, richer characterizations are
offered in the context of IG Extended.
While seemingly a concession to linguistic structure, from an institu-
tional perspective, the central purpose of the Object is to make the linkage
between a responsible actor, regulated activity, and action receiver explicit,
where the semantic of the linkage (i.e., how direct and indirect object are
Fig. 4.4 Direct and indirect objects in regulative institutional statements

related in a particular statement) is captured in the activity (Aim compo-

nent). Naturally, institutional entities can thus find variable references
as Attribute or Object components in different institutional statements.
Where tax revenue service is the Indirect Object in the example statement,
it may be an actor (and thus Attribute component value) in a different
institutional statement. Objects can thus both be animate or inanimate
in kind, and further, be physical entities, or be abstract, reflecting beliefs,
observations, or other concepts such as procedures or institutional facts.
In consequence, abstract objects can be complex constructs in their own
right, and can further assume the structural form of institutional state-
ments themselves, an aspect discussed in the context of IG Extended (see
Sect. 5.1).
The presence of the Object component is optional for regulative state-
ments, since not every regulated activity in an institutional statement may
refer to objects.
4.2.1.5 Context
The Context component instantiates settings in which the focal action
of a statement applies, or qualifies the action indicated in an institutional
statement. Responding to the motivation set out in Chapter 3, the IG 2.0
resolves ontological inconsistencies embedded in the Conditions compo-
nent and introduces the distinction between context characterizations that
delineate the conditions under which the non-context part of the insti-
tutional statement applies – the conditions under which the statement
activates, referred to as Activation Conditions , aligned with the precon-
dition conception in Searle’s characterization of regulative statements, “If
Y, then X” (Searle, 1969).
Contrasting these Activation Conditions, Execution Constraints refer-
ence the qualification of activities during execution, thus imposing
constraints on the enacted Aim.7
Referencing the earlier example “All corporations, …, must submit
annual reports to the Tax Revenue Service in a timely manner following
the closing of the tax year,” we can identify two context clauses, where one
is the precondition for the submission of reports, namely “following the
closing of the tax year.” The second context clause “in a timely manner”
7 The treatment of Execution Constraints with respect to constitutive statements is

described in Sect. 4.2.5.
Fig. 4.5 Activation Conditions vs. Execution Constraints
qualifies or constrains the activity regulated in the institutional statement,

suggesting when the submission is to be performed.
Figure 4.5 highlights the variable linkage of activation conditions and
execution constraints to other institutional statement components, with
Execution Constraints constraining or qualifying behavior in execution,
and Activation Conditions highlighting the linkage to antecedents that
signal the applicability of the statement in its entirety.
The presence of this component is necessary for any regulative state-
ment, but, absent any explicit specification, they can be implied: where
absent, activation conditions imply their applicability “at all times ”; the
statement applies, or activates, under any circumstance. Where execution
constraints are absent, the regulated activity does not underlie further
constraints, and is executed “without constraints.”
The Context component, as an umbrella component capturing both
conditions for execution as well as moderating it, signal Statement
Context, as opposed to wider policy context as relevant for the domain
more generally.
This differentiated characterization reflects a conceptual deviation from
Crawford and Ostrom’s Conditions component, but affords the distinc-
tion between the context a statement is functionally embedded in, on
the one hand, and the contextual information a statement itself embeds.
Beyond the ontological clarity sponsored by this distinction, it enables the
reliable expression of conditional obligations/prohibitions (in an exten-
sion of pure obligations/prohibitions), thus offering the basis to establish
a formal linkage between the Activation Condition and the Deontic,
formally expressed in dyadic deontic logic.8 While dyadic deontic logic
is not without its own challenges (see e.g., Prakken & Sergot, 1997),
8 We will not explore the operationalization in detail at this stage, but rather signal that
the established ontological consistency enables such linkage in the first place.
it provides the basis for the consistent operationalization of the Context

component by uniquely identifying its respective functions within insti-
tutional statements and the logical linkages to other components, and
further draws it to a level of granularity that is more closely comparable
with the remaining IG components (recall the fine-granular Deontic and
Aim components).
However, despite the refined granularity offered by the distinction into
Activation Conditions and Execution Constraints, Context of either type
can be of variable kind, including temporal or spatial characterizations,
qualification of methodical aspects, as well as general, potentially overlap-
ping categorizations as events or states. Either Context type may further
be of variable complexity, e.g., in the form of actions as preconditions for
the activation of a statement, leading to a representation in the form of
institutional statements. However, not in all instances will those represen-
tations be institutional statements, but rather reflect institutional states
described in terms of the IG syntax. This specific aspect, alongside the
general categorization of context types, is subject to extended discussion
on the next level of expressiveness, IG Extended (Sect. 5.1.2).
4.2.1.6 Or Else
The final component of the regulative form of the IG is the Or else, which
captures any sanctioning or incentivizing provision associated with the
violation of the behavior (Aim) indicated in the institutional statement
regulated in pre- or proscriptive form (i.e., obligation or prohibition),
and further contextualized by the Context component. Sanctions, or
consequences, associated with the Or else vary in kind, and can include
physical sanctions, institutional consequences (e.g., revocation of rights),
be punitive or incentivizing, and can further emanate from diverse sets of
actors.
Deviating from the original interpretation of the Or else component
that concentrates on the indication of the substantive sanction content
(e.g., “…, or else receives fine.”), the refined characterization introduced
in IG 2.0 recognizes the structural equivalence of the regulated activity
and the corresponding sanctioning activity, both expressed in terms of the
same syntactic components introduced to this stage, namely Attribute,
Aim and Context components, alongside the selective use of Deontic and
Object components, referred to as institutional statements (with a more
refined characterization following the introduction of the components).
In the context of the Institutional Grammar characterization, this

provides the prompt to introduce the concept of nested institutional state-
ments based on Nested ADICO (nADICO) (Frantz, 2015; Frantz et al.,
2013), previously referred to in Chapter 1.
The notion of nested institutional statements recognizes the inter-
dependency of statements, both within and between institutional state-
ments. The explicit representation serves two interlinked objectives,
namely resolving ambiguities observed in natural language expression of
institutional statements, but also to reconstruct the different forms of
embedded institutional complexity that those afford. The essential feature
is the selective substitution of syntactic components with complete insti-
tutional statements. With a specific focus on the Or else component, we
can review the following statement introduced at the beginning of this
section: “Drivers must follow traffic regulations whenever driving on public
roads, or else they face a fine.”
Where the encoding of the initial part of the statement relies on the
component introduced to this stage, i.e., Attribute (drivers ), Deontic
(must ), Aim (follow), Direct Object (traffic regulations ), Activation
Condition (whenever driving on public roads ), the latter part, the Or else
(or else they face a fine) embeds complexity that – when made explicit –
can be expressed using the same components. Representing a sanctioning
behavior associated with the noncompliance with the prescription of the
leading statement, semantically, it exhibits a similar capacity for guiding
behavior, including the content of the sanction (fine), and, implicitly,
the actor responsible for its execution. While implicit in this statement,
imposing the same representation on the consequence forces the analyst
to makes the responsible actor overt, fostering an explicit and analytically
tractable representation. Where not clearly implied, the inference of the
enforcement actor occurs from the context the statement is embedded
in (e.g., surrounding statements, section the statement is embedded in,
or action situation). Assuming a traffic setting, the Or else expression
can thus be formulated as a complete institutional statement, resolving
to enforcement official must impose fine, where the responsible actor and
associated obligation are inferred. In an institutional statement for the
statement thus reads “enforcement official (Attributes ) must (Deontic)
impose (Aim) fine (Direct Object )”.
Combining both statements, and schematically visualizing the state-

ment linkage, the overall statement thus reads9
The interpretation of the Or else as an abstract component, or rather

logical operator,10 that links institutional statements in the form of a
material consequence thus serves the purpose of (a) making the structure
of the sanction provision explicit, including content, responsible actor, as
well as associated contextual conditions and constraints, and (b) drawing
an explicit linkage between both statements, characterizing the former
as a statement that is guarded or monitored by a separate complete
institutional statement that captures the consequences associated with the
violation of the preceding monitored statement.
This form of statement interdependency, the interlinking of moni-
tored statements – statements whose violation leads to the activation of
consequence – and consequential statements – statements that describe the
consequences of noncompliance with monitored statements – is referred
to as vertical nesting . Naturally, the level of nesting is not limited to a
single level but can capture the structural interdependencies of any depth.
Returning to the previous example, the enforcement official itself can be
subject to oversight by a superior, other official, or a combination thereof.
9 Note that in the following examples, prepositions and other linking language is asso-
ciated with the receiving component (e.g., “with traffic regulation”). This is discussed in
detail in Chapter 7.
10 This characterization is formalized in Sect. 6.1.1.
Returning to the discussion of the Or else, in IG 2.0, the Or

else component is abstract in that it requires the coder to encode
consequences, such as sanctions or incentives, in terms of institutional
statements, and thus acts as a logical connective between institutional
statements rather than embedding the content of the consequences (as
visualized in the previous figures). While this introduces the need to
reconstruct the institutional statement with respect to its institutional
meaning by inferring potentially implied actors, objects, and context, and
thereby imposing a cognitive load on the coder or analyst, it likewise
forces analysts to interpret arrangements in terms of consistent patterns
that are semantically anchored in the action situation and, in extension,
the institutional arrangements they are embedded in.
While the preceding discussion put primary focus on the structural
representation of consequences in institutional statements, Crawford and
Ostrom explicitly allude to the institutional consequences emanating from
some actor, such as another actor in an action situation. In the context of
the new IG, this interpretation is wider, given both the flexibility to char-
acterize consequences more richly in terms of institutional statements, but
more importantly, empirical applications of the IG to analyze “institu-
tions in use” suggest that sanctioning originating from non-human (e.g.,
natural) actors, such as sanctions from the natural environment, may be
central to guide conformance behavior (Watkins et al., 2015). Relatedly,
the observation of consequences for constitutive statements suggests that
consequences can further be existential in kind, such as signaling the
non-applicability of a statement itself, rather than imposing consequences
on a given actor – an aspect we revisit more explicitly in the context of
constitutive statements in Sect. 4.2.5.
Or elses have been of limited relevance in many studies that apply IG
coding (see Chapter 3 for an extended discussion). This can be attributed
to the lack of presence of Or elses and formulation in terms of Activa-
tion Conditions (see Chapter 7 as well as Sect. 6.1.4). It may also be
attributed to the commonly observed separation of regulated institutional
content – expressed in terms of the previously introduced “Attributes,
Deontic, Aim, Object, Context ” components – and the consequences for
noncompliance in policy documents, as well as their reconstructions in the
form of preconditions from the perspective of enforcing actors.11 These
11 This aspect is addressed in the context of statement transformations in Sect. 6.1.4

that showcase the logical interaction between preconditions and consequential statements.
aspects motivate the call for a semantic interpretation of institutional

content posed in Chapter 3.
Where provisions in policy text can be mapped into the structural
components introduced earlier (an aspect explored at greater depth in
Chapter 7), their linkage to consequences often requires a broader under-
standing of the coded document, since provisions related to sanctioning
may, in the best case, be captured in a different statement, or, more
commonly, provided in an entirely separate part, sub-part or section of
a given document dedicated to sanctions specifically. In addition, the
scope of monitored statements and corresponding consequential state-
ments may vary, with consequences expressed as “blanket statements” that
apply across a range of behavior regulated throughout the policy (e.g.,
fines associated with any violation of statements captured in the regu-
lation). In practice, a central challenge linked to the Or else is thus the
identification of the linkage between statements, rather than components
combined to reflect regulated behavior in the first place.
Notwithstanding these specific concerns, many of which we will revisit
throughout this book, the specification of consequences is not neces-
sary for all institutional statements, making the Or else optional in the
specification of institutional statements.
The introduction of the Or else as abstract institutional statement
component, or more accurately, as consequential logical connective linking
institutional statements, brings us closer to completing the characteriza-
tion of the structural makeup of (regulative) institutional statements more
generally, and the kinds of institutions they represent.
4.2.1.7 Regulative Institutional Statement Structure

The syntactic linkage of the components presented above provides us
with the basis to construct institutional statements that capture regu-
lated behavior, including the regulated actor as part of the Attributes
component, the action regulated in the Aim, the Object addressed and/or
affected by this action, the Deontic indicating whether the specified
behavior is compelled or restrained, as well as contextual characterizations
that signal the conditions under which a statement applies (Activation
Condition), as well as further constraints that qualify the enacted action
(Execution Constraint ). These components, in statement composition
provide the elementary structure that allows the description of how
behavioral conventions are described, restrained or compelled. Finally,
Fig. 4.6 Regulative statement structure
linking a consequence via the Or else – expressed using the same struc-
tural primitives – captures potential consequences applicable in the case
of noncompliance. The combination of those components (with optional
applicability of Deontic, Object, and Or else) features the essential regu-
lative structure of institutional statements, as schematically shown in
Fig. 4.6, and listed in order of presentation and typical appearance in
institutional data.
However, before offering a differentiated characterization of regula-
tive institutional that distinguishes between varying forms of institutional
statements, the following section highlights analytical challenges associ-
ated with complex institutional statements commonly found in policy
texts, and introduces conceptual approaches to address those challenges,
while, at the same time, providing the conceptual foundations referenced
throughout the remainder of this book.
4.2.2 Statement Combinations (Horizontal Nesting)

Inasmuch as the structure of institutional statements as introduced to this
stage renders conceptual clarity and consistency, in reality “people don’t
talk in institutional statements” (Watkins et al., 2015), and neither is a
policy written in such form.
However, instead of turning to stylistic and linguistic aspects related to
the coding addressed in Sect. 5.1 and Chapter 7, a more immediate aspect
is the way in which actors, objects, and circumstances relate to activi-
ties. In practice, and notably in form, actors can be involved in multiple
regulated activities (or multiple actors in the same activity). A common
sight in policy statements is the presence of component combinations,
such as … inspect and facilitate … in the case of activities (i.e., Aim
components), or certifiers, inspectors, and farmers for actors (Attributes )
or receivers of action (i.e., as Object components). While the stylistic
conflation of those component-level combinations oftentimes appear unam-
biguous to the human reader, they reflect endemic complexity that makes
it challenging to distinguish analytically between individual actors, activ-
ities, and associated compliance behavior. This is of particular relevance
when attempting the linkage or compliance assessment of actual “rules

in use” observed in terms of behavioral instances (i.e., specific activities)
as opposed to compound activity captured in component-level combi-
nations. To this end, the IG seeks to dissect the interpretation of such
combination by invoking a semantic perspective, challenging the coder to
firstly identify whether the combined components offer distinctive insti-
tutional information, as opposed to merely reflecting prose that does not
deserve further analytical attention. Exemplifying the latter, the expres-
sion perform regulation effectively signals the activity regulate, in which
case the semantics are captured in a single component (here Aim), as
opposed to being distributed across components corresponding to the
linguistic category (i.e., perform as Aim, and regulation as Direct Object ).
If both components are distinctive, but of the same component type (e.g.,
both are activities, objects, etc.), a second concern pertains to how those
expressions are logically linked.12
The analyst may ask: Do the referenced activities represent alterna-
tive actions? More specifically, are the actions exclusive alternatives (i.e.,
only one is to be performed)? Or, is the actor required to perform both?
Building on this, and showcasing empirically observable complexity of
provisions, does the multiplicity of actors, actions, and objects in a single
statement signal distinctive linkages between components? For example,
the expression … inspect facility and report violations … conflates both
activities and related objects, where the individual activities (inspect and
report ) are linked to distinctive objects (facility and violations, respec-
tively).
Since the IG affords the comprehensive representation of character-
istics of institutional relevance (as motivated in Chapter 3, and further
explored below), it provides the necessary mechanisms that capture such
semantic detail unambiguously.
The Institutional Grammar resolves this challenge by affording the
representation of logical combinations of institutional statements using
the following logical operators:
12 While the following examples focus on the Aim component, the principles equally
apply to any other component.
• AND (to signal conjunctions),

• OR (colloquially “AND/ORs”, to reflect inclusive disjunctions),
and
• XOR (colloquially “EITHER ORs”, reflecting exclusive disjunc-
tions).
Complementing these dyadic operators – operators that link two insti-

tutional statements – with negation (NOT ) as a central unary operator
that negates a specific statement, the IG inherits not only the ability to
assess compliance with specific provisions (e.g., a specific set of possible
choices or preconditions), but furthermore establishes a basis for the
logical transformation of statements based on the interdefinability features
that sentential, or propositional logic (Klement, 2004), offers – similar to
the interdefinability offered by deontic logic as discussed in the context
of the Deontic.13 We return to the discussion of richer use cases in the
context of IG Extended and IG Logico.
As far as the interpretation of statements is concerned, the analyst relies
on a set of guiding questions that afford the mapping of ambiguous
linguistic expressions to the precise logical characterizations provided
above:
• Do multiple distinctive components apply in conjunction (e.g.,

divide and conquer)?
• Does either of those apply optionally (e.g., monitor and sanction)?
• Do they reflect exclusive alternatives (e.g., accept or reject )?
The examples highlight how the interpretation can vary contextually

based on coder skills and background, but also knowledge of the coded
policy, including the ambiguous use (and interpretation) of and and or in
natural language (see e.g., Robbins, 201814 ), an aspect that is specifically
pronounced in enumerations of items due to the larger number of implied
linkages (e.g., “farmer may at any time apply for accreditation, seek reac-
creditation, or withdraw from the accreditation process” ), exacerbating the
reliance on contextual interpretation.
13 The transformation of logical statements is discussed in Sect. 6.1.4.

14 Robbins further explores the long-standing discourse on the use of AND/OR in the
context of contract drafting and legal writing.
Resolving the linkage of expressions that combine various components,

such as the example above (… inspect facility and report violations …), the
interpretation (i.e., whether “ands”, and “ors” convey semantic signals
that should be captured, or are merely stylistic – and reflect independent
institutional statements) can be unambiguously expressed as schemati-
cally visualized below. Here the combined expressions are separated, or
decomposed, into distinctive statements featuring all necessary and rele-
vant optional components, and linked by a conjunction (AND) indicating
an interpretation that assumes their conjoined applicability (i.e., making
overt that the “and” is not just stylistic in nature, but carries institutional
meaning).
While applicable for the decomposition of individual prescriptions, it

equally applies to more complex characterizations that link multiple alter-
native or complementary activities. Specifically notable, however, is the
combined use with the characterization of consequences, introduced as
vertical nesting in Sect. 4.2.1.6, e.g., to represent graduated sanctioning.
Amending the previous example for illustration (and approximating
the complexity of statements found in practice), we introduce a compli-
ance obligation for certified organic farmers, and augment the inspecting
officer with the discretion to apply variable sanctions depending on
the severity of potential noncompliance as indicated by the following
statement:
Upon certification, certified organic farmers must comply with organic

farming regulations. In the case of minor violations, inspectors may allow
for immediate correction, or, in the case of major violations, file a report
according to the relevant organic farming provisions.
Recognizing the exemplary separation of monitored and consequential

statements into separate sentences (as motivated in the context of the
Or else), we can visualize the institutional interpretation in the following
schematic form that links the decomposed statements (here using Or else
to signal the linkage of monitored and consequential statements, and the
OR signaling discretionary action15 ) in a form that makes the respective
institutionally relevant consequential linkage and associated action choices
explicit16 :
15 Here the choice of OR, i.e., AND/OR, is intentional to signal the potential co-
occurrence of both cases, i.e., minor immediately corrected violations, and major reported
ones, explicit.
16 The coding provided here is illustrative. Coding conventions, such as the handling
of linking terms, and principles to guide the separation of entities and properties are
discussed in Chapter 7.
Generalizing the introduced combinatorial forms for institutional state-

ments, we can retrace that such nesting is unconstrained with respect to
the depth of nesting – the linkage between monitored and consequential
statements (vertical nesting) – on the one hand, as well as with respect
to the breadth of statements linked by conjunctions and disjunctions
(horizontal nesting) on the other. These nesting faculties facilitate the
consideration of diverse sets of actors, activities, conditions, etc. – thus
providing the basis for capturing institutional arrangements of varying
complexity, while at the same time capturing distinctive information
comprehensively, conceptually reflected in institutional tree structures
(see Chapter 8). Figure 4.7 summarizes these principles schematically,
Fig. 4.7 Nesting principles
including the shifting foci of analysis expressed in statement relationship

pairs, where the first-order consequential statement may at the same time
be a monitored statement for potential second-order consequential state-
ment, which itself may be subject to higher-order oversight. In addition to
affording the mechanisms to capture institutional complexity at detail,17
the combined use of both forms of nesting thus affords a theoretically
unlimited representation of what can be referred to as structural insti-
tutional regress (Frantz, 2015), drawing linkages between different levels
and scopes of governance, relevant both for the analysis of “institutions
in form,” as well as “use.”
While this discussion put the focus on selective components for the
purpose of illustration (e.g., Or else), the described forms of nesting
apply to various other components (e.g., Context ), an aspect we draw
on later in this section, and specifically, in the context of IG Extended,
the level of expressiveness that aims at extracting fine-grained structure
from institutional statements using the nesting principles introduced at
this stage.18
Given the complex linkages of institutional statements that the nesting
concepts introduce, it is important to introduce a refined termino-
logical distinction between different forms of institutional statements.
Institutional statements, and more specifically, regulative institutional
statements, require the presence of a set of necessary components, namely
Attributes, Aim, and Context components, alongside further optional
17 We return to this discussion in the context of analytical applications of the

Institutional Grammar in Chapter 8.
18 The analytical benefits of nested statements will further be discussed in Chapter 8.
components. They can furthermore include nested structures of any kind,

where components are either combined, linked or entirely substituted by
institutional statements, such as the case for the Or else. An atomic insti-
tutional statement is an institutional statement that only contains one
of each necessary and (where applicable) optional components, where
none of these components is further decomposed into, or substituted
by, nested institutional statements. This implies that an atomic institu-
tional statement cannot contain an Or else component (since it consists
of a nested institutional statement itself), and neither any other compo-
nent that is decomposed into (and hence substituted by) full institutional
statements (such as a Context component containing the syntactic struc-
ture of an institutional statement). The latter nesting variation is yet to
be introduced in the context of IG Extended, and subject to further
discussion in Sect. 5.1. Under the broader conceptual umbrella of insti-
tutional statement, atomic institutional statements are complemented by
composite institutional statements, i.e., institutional statements that consist
of multiple atomic and/or combined statements, i.e., display any form of
the introduced nesting facilities.
4.2.3 Regulative Institution Types

4.2.3.1 Three Branches of Institutional Analysis
The introduction of institutional statements as a focal unit of analysis
reflects the grammatical correspondence to sentences in natural language.
However, as with the diverse forms and functions that natural language
expressions play in different speech acts (Searle, 1969) (and thereby
different communicative functions), the IG recognizes distinctive coordi-
native functions associated with different forms of institutions expressed
in terms of institutional statements. In their seminal article, Crawford
and Ostrom characterize three types of institutions grounded in different
analytical traditions. These three types of institutions – shared strate-
gies, norms, and rules – are reviewed in this section with reference to
institutional analysis perspectives (i.e., rational choice institutionalism and
sociological institutionalism) and contemporary approaches for engaging
in IG research (i.e., application of the IG to study public policy).
Crawford and Ostrom reference the “institutions as equilibria”
perspective, reflecting approaches to institutional analysis closely associ-
ated with “Rational Choice Institutionalism” (Shepsle, 2009). Primarily
drawing on theoretical foundations of Austrian Economics (e.g., Menger,
1963; von Hayek, 1945), a central theme is the interpretation of institu-

tions as equilibria that emerge through the interaction of players primarily
focused on utility maximization, with payoffs reflecting the performance
of various institutional outcomes. This analytical perspective is commonly
associated with rational choice theory, which, as an analytical technique,
relies on game theory (von Neumann & Morgenstern, 1947). At its
essence, essential subjects of study are the incentives that lead to moti-
vational alignment of players to produce stable interaction patterns (of
potentially varying efficiency) based on converging “shared strategies”.
Notable contributions in this field include Schotter (1981) and Calvert
(1995), as well as Greif (2006)’s work on medieval trading coopera-
tives. A common critique associated with a mere motivational focus on
institutional arrangements is the lacking recognition of path dependency
effects that challenge the efficiency that equilibrium-based approaches
presume (see e.g., Greif & Kingston, 2011), despite selected approaches
to alleviate this concern (see e.g., Greif & Laitin, 2004).
They also reference “institutions as norms.” In contrast to the equi-
libria view that focuses on individually motivated (and generally static)
preferences, the subject of study of normative systems are the social mech-
anisms that produce and sustain behavioral alignment, i.e., enforcement
mechanisms not administered by a recognized or appointed authority.
In this perspective on (social) institutions, the focus lies on the assess-
ment of normative conduct based on enforcement (in terms of incen-
tives or sanctions) that emanates from the social environment of actors
(Coleman, 1990), giving rise to the study of intersubjectivity of normative
belief (Tuomela, 1995), effective social sanctioning mechanisms (Axelrod,
1986),19 the effects across varying actor networks (Nee, 1998), along-
side the scalability of decentralized enforcement. Notable foundational
works in this area include Lewis (1969),20 Ullmann-Margalit (1977),
Coleman (1988), and more recently, Cialdini et al. (1991) and Bicchieri
(2006). While equilibria-based institutional analysis primarily has its roots
in economics, being associated with the study of sociology, the study of
norms applies across a wide range of social sciences, including economics
(Akerlof, 1976; Young, 1998), social psychology (Cialdini et al., 1991),
19 While seminal, a noteworthy challenge to Axelrod’s work has been made by Galan
and Izquierdo (2005).
20 A noteworthy critical account of Lewis’ game-theoretical operationalization is offered
by Gilbert (1989).
international relations (Finnemore & Sikkink, 1998), and for the study of
norm dynamics and life cycles (Frantz & Pigozzi, 2018). In this branch
of studies, an overarching assumption is the recognition of norms as
drivers for coordination, and their underlying dynamic nature, in contrast
to earlier characterizations of norms as relatively static.21 Social norms
have further been subject to legal studies, especially with respect to
their complementary role in motivating compliance behavior (Posner,
2000). Specifically, the facilitative role of norms for rule implementation
motivates the contrasting characterization of (social) norms as “informal
rules”, as opposed to the “formal rules” referenced in the third view on
institutions.
As described in the early parts of this book (see Chapter 2), most appli-
cations of the IG to date have been targeted toward the study of public
policy. This legalistic perspective on institutions has been descriptively
tagged as “institutions as rules” by Crawford and Ostrom (1995). In this
branch of institutional analysis, subjects of study include the regulations,
or formal rules, such as legal provisions devised in a legislative or collec-
tive action process, that define and constrain actor behavior. Building on
the traditions of New Institutional Economics (North, 1990; Ostrom,
1990; Williamson, 1975), the “rules” perspective roots in legal theory
(Hohfeld, 1913) and “Old Institutional Economics” commonly associ-
ated with Commons (1968), building on the fundamental premise that
formal rules assume the primary responsibility for structuring social coor-
dination. Following the establishing for formal rules, adaptive behavior in
terms of the “rules in use,” and corresponding accommodations based on
compliance mechanisms (e.g., regulatory compliance) are essential topics
of interest in corresponding legal and policy studies, a vast amount of
which is referenced in the earlier Chapter 2 of this book.
While characterized as distinctive branches of analysis as part of this
overview, it is important to note that the stylized “types” reflect the
primary foci of analyses, alongside the theoretical and methodological
toolbox associated with analysis of either type, recognizing the bene-
fits and associated trade-offs associated with either approach. Public
policy scholars, for instance, recognize the importance of “rules in
use,” but interpret legal rules as primary subjects of analysis that offer
explicit prescriptions for the moderation of social behavior. Consequently,
21 “[Norms] … are a part of the heritage that we call culture” (North, 1990, p. 3).
Table 4.1 Institution type characteristics
Institution type Subject of study Institution interpretation Primary driver

of coordination
Equilibria Individual Coordination outcomes Rational

motivations & as equilibrium strategies behavior
cooperation behavior of individual actors (at
any aggregate level)
Norms Norm emergence, Institutions as emergent Social
dynamics & outcome of environment
normative conduct decentralized
enforcement based on
internalization and
socialization
Rules Rule structure, Institutions established Regulatory/legal
compliance & rule in authoritative process environment
change by legitimate actors
scholars primarily interested in “rules in form” may hence apply content

analysis techniques to extract policy information, whereas researchers who
focus on the adoption “in the wild” will be drawn to appropriate empir-
ical methods to extract the “rules in use,” such as ethnographic studies,
structured interviews, etc.
Table 4.1 summarizes the distinguishing features for all referenced
branches, including underlying presumptions, and the institutional under-
standing put forth in the analysis.
4.2.3.2 Mapping Institutions to Statements

This stratification of institutions into the aforementioned types, and
specifically, the conceptual integration in a shared syntactic representa-
tion, provides the conceptual basis to understand the function the IG
assumes when responding to a diverse set of analytical perspectives. At
the same time, it provides the backdrop for the characterization of insti-
tution types in terms of institutional statements, affording the conceptual
mapping that constitutes the IG as a Grammar of Institutions.
The starting point for the characterization of distinctive institution
types is the recognition of institutional statements as basic units of
analysis, while, at the same time, recognizing that they can vary in
complexity, and specifically, with respect to the components used to
compose institutional statements.
As indicated earlier, the components necessary to form any institutional

statement include the Attributes, Aim, and Context components, consti-
tuting a Strategy.22 The Object component is optionally applicable for
all institution types. Offering a basic illustrative example, strategies can
capture a descriptive account of social behavior (Lewis, 1969), such as
citizens’ general compliance behavior:
Attributes: citizens
Aim: comply
Direct Object: with the law
Activation Condition: at all times
Execution Constraint: N/A
Reviewing this example, it is noteworthy to reiterate the special nature
of the Context component (see Sect. 4.2.1.5), which is comprised of
activation conditions and execution constraints. Where activation condi-
tions default to a characterization that suggests the applicability of the
statement at all times (as made explicit in this example), for execu-
tion constraints the absence suggests that no constraints whatsoever are
imposed on the activity during execution (here indicated as N/A).
Essentially, strategies express, or describe, conventional behavior, i.e.,
the fact that citizens adhere to the law, without making assumptions about
associated prescriptions, let alone the source of the underlying motivation.
As such, strategies can be of diverse nature and varying scope, providing
the basis for the most general operationalization of institutions afforded
by the IG.
The qualification of injunctive behavior is the role of the Deontic.
If augmented with a Deontic, e.g., to signal prescriptiveness or permis-
sion, institutional statements reflect the structure of what Crawford and
22 Crawford and Ostrom originally characterized those as “shared strategies” to signal

the general adoption. It is noteworthy to state, however, that strategies may not in all
instances be shared amongst all participants in a social setting, but potentially only adopted
by subsets of variable nature (see e.g., Ghorbani et al., 2013).
Ostrom devised as a Norm. Extending the previous example, the corre-

sponding normative form (and assuming a prescription in this instance)
suggests:
Signaling an obligation in this instance, the Deontic (as introduced in

Sect. 4.2.1.2) has the central role of establishing the regulative nature
of institutional statements based on the endowed capacity to guide and
direct behavior, either in the form of pre- or proscriptions (i.e., obliga-
tions and prohibitions) or as permissions, but also the nuanced expression
of the extent of prescription, let alone the regulatory weight based on
domain-specific conventions.23 However, while behavioral expectations
are made explicit, the associated enforcement in the form of sanctions or
incentives is not specified. While IG 2.0 deviates from this interpretation,
Crawford and Ostrom’s interpretation is based on the absence of a set
of distinctive qualifications that constitute sanctions, aspects we discuss in
the context of the rule characterization.
The third form of institutional statements in the institutional strata
is the Rule, and constituted by the introduction of the Or else as the
final syntactic component, which explicitly expresses the consequences
associated with noncompliance of the leading part of the institutional
statement. Where Crawford and Ostrom’s specification interprets the Or
else as a distinctive component carrying the sanctioning information,
in the context of the IG 2.0 the Or else component is special in its
function as a logical operator that links prescribed behavior with corre-
sponding consequences, where those consequences can be expressed in
the same structural form as the prescriptions they guard, i.e., as monitored
and consequential institutional statements, respectively.24 Extending the
previous example, the rule form is thus:
23 For a detailed discussion, see Sect. 4.2.1.2.

24 Recall Sect. 4.2.1.6 for an extended discussion.
As discussed in the context of Sect. 4.2.2, both the monitored and

consequential statements can be of varying complexity based on combi-
nations of atomic statements, e.g., to reflect the multitude of activities, or
to express graduated sanctioning provisions.
Returning to Crawford and Ostrom’s qualification for the existence of
an Or else not expressed syntactically, the specification of consequences
requires the backing by other institutional statements (an aspect the
nesting capability makes explicit), establishment of a range of punish-
ment and assigning authority and procedures for imposing the Or else,
and further assumes the existence of institutional statements that affect
the constraints and opportunities of the actors who monitor conformance
(Crawford & Ostrom, 1995, p. 586).
4.2.3.3 Revisiting the Norm/Rule Distinction

While the stratification based on the incremental introduction of syntactic
components offers conceptual simplicity and maintains structural distinc-
tiveness, it invites for a set of clarifications, both linked to the syntactic
focus and the requirements underlying a specification of consequences.
Reflections on the first aspect, the syntactic characterization, follow
from the discussion in Chapter 3 and empirical observations. When
engaging in the extraction of institutional statements from institutional
information, it is important to note that the characterization and linkage
of consequences (the consequential statement) may, in many instances,
not be understood based on syntactic grounds. When analyzing policy
text, for instance, the reader cannot necessarily expect that a consequence
for noncompliance immediately follows the specification of regulated
behavior. Instead, as discussed in Chapter 3, consequences (e.g., in
the form of sanctions) may be held in separate statements, sections,
appendices, or documents entirely – or may simply not exist. This is in
contrast to the remaining components, which are generally co-located
so as to signal the institutional content (actors, actions, conditions, and
constraints) in the first place; content relevant for the Or else is in prac-
tice rarely immediately linked to the monitored statement, but organized
in sections or parts dedicated to the specification of consequences (e.g.,
sanctions) – if present at all, i.e., formal rules do not necessarily carry
consequences (let alone explicit ones) (see e.g., de Moor, 2015).
Relatedly, and reviewing the structural characteristics of institutional
statements more immediately, consequences, e.g., in the form of sanc-
tioning provisions, are commonly constructed from the perspective of the
enforcer, with the noncompliance as an antecedent for any intervention.
In such instances, the consequential relationship is syntactically captured
wholly within a single institutional statement (with the conditional viola-
tion expressed in the (nested) Activation Condition) without any need for
the Or else component whatsoever.25 The following example illustrates
such case:
Attributes: enforcers
Attributes Properties: responsible
Deontic: may
Aim: impose
Direct Object: sanctions
Activation Condition: if organic farmers operate non-compliantly
Execution Constraint: as permitted by law
Beyond the structural considerations related to a purely syntactic inter-

pretation, we can observe conceptual challenges to the validity of the IG
for the representation of selected institution types.
The exclusive association of consequences with the Rule conception
highlights the focal orientation on formal institutions (where formal
pertains less to the form (e.g., written) in which institutions are repre-
sented (institutions-in-form), but primarily references the process by
25 The structural parsing of preconditions, or Activation Conditions, of institutional

statements will be discussed in Sect. 5.1.1.
which and forums in which those come about – as discussed in detail

later). However, the complementary dimension relevant for institu-
tional analysis are the informal institutions that, combined with formal
institutions embodied by actors "in the wild", reflect the institutions-
in-use. The interaction between institutions-in-form (e.g., institutional
designs expressed in policy) and -in-use is a recurring feature of analyses
commonly associated with the “institutions as rules” perspective, such as
North (1991) and Williamson (1975). Seen, for example, through the
lens of regulatory compliance, which primarily focuses on the assessment
of, and the mechanisms that facilitate, behavioral alignment with policy
objectives, the effectiveness of such mechanisms is in fact moderated by
perceptions about the enforcement (Siddiki et al., 2018). Further aspects
referenced by analysts primarily committed to policy studies include chal-
lenges to conformance based on the assumed appropriateness of rules to
govern particular domains (Young, 2016), as well as the capacity to recog-
nize and react to enforcement signals on the individual level in the first
place (Winter & May, 2001). Highlighting the analytical value further,
the interaction between laws and norms has further found attention by
scholars from the equilibrium camp (e.g., Acemoglu & Jackson, 2017).
Given the central role of social enforcement in the “norms” perspec-
tive on institutional analysis referenced previously, a general Grammar of
Institutions thus calls for the necessity to represent enforcement mech-
anisms not only for rules, thereby reflecting the legalistic view, but also
for the more commonly decentralized mechanisms applicable to norms –
as a proxy for the socio-institutional perspective. Foregoing the analysis of
consequences associated with norm violations (e.g., social consequences)
limits the ability to capture such institutions, and consequently challenges
the validity of the IG for the representation of institutional phenomena
at large. This call to reflect the presence of sanctions when analyzing
informal institutions using the IG has further been established in qualita-
tive studies in the field (Watkins et al., 2015), as well as studies of norm
emergence in artificial societies (Frantz et al., 2015b), both of which refer-
ence the necessity as essential to understand existing institutions or their
emergence in the first place.
Seeking a general and comprehensive representation of institutions,
the refined IG 2.0 responds to these observations and additional calls in
literature (Frantz et al., 2013; Schlüter & Theesfeld, 2010) in two ways.
This includes the explicit admission of the syntactic Or else as an

optional component both for norms and rules, and, in line with the moti-
vation in Chapter 3, calls for a distinction of both types on semantic (as
opposed to syntactic) grounds, as explored in the following.
Complementing Crawford and Ostrom’s qualification for formal rules,
and where consequences are present in the first place, the IG 2.0 relies
on a systematic differentiation of norms and rules based on an explicit
characterization of the monitoring party, as proposed by Schlüter and
Theesfeld (2010), where rules, for instance, presume a formally appointed
monitor (e.g., appointed in a legitimated legal forum), whereas informal
(e.g., social) appointment is sufficient for norms. While providing an
important distinguishing feature, the focus on the monitor as a differ-
entiating feature alone can nevertheless lead to the undifferentiated, or in
worst case wrongful, characterization of an institution, especially where
the monitor and enforcing entity are distinct – a common and often
intended characteristic of formal institutional arrangements. To this end,
the IG 2.0 further considers the conceptual separation of monitoring
and enforcing entity, originally proposed by Frantz et al. (2013) and
Frantz (2015), as a distinguishing feature between norms and rules. Here
conceptual separation implies the explicit recognition of both roles as distinc-
tive in the enforcement process; it does not imply that monitoring and
enforcing entity are necessarily different actors (or actor groups), but
rather aims at their characterization in an institution, an aspect that differs
for norms and rules. The benefits of this revision are valuable from both
a representational as well as an analytical perspective.
Representationally, the distinction rests on the observation that actors’
relationships in complex institutional arrangements may be non-dyadic,
and rather linked indirectly, with potential monitors detecting and
signaling potential transgressions that are subsequently addressed by a
responsible enforcer. In contrast to the perpetually acting monitor, the
enforcer may only become active whenever a noncompliance or violation
actually occurs. Drawing on financial audits as an illustrative example,
an auditor may act as a monitor of a corporation’s compliance and
report potential violations. Only upon violation of financial reporting,
the financial authority may then become active in its enforcing capacity.
In institutional statement parlance, this conception translates into the
following variably linked statements:
Monitor
As seen through this characterization, the first statement reflects the

linkage of the regulated entity and enforcer, whereas the second state-
ment describes the monitoring activities that lead to the enforcement in
the first place. In this representation, the linkage between violating actor
and the monitor is indirect, and expressed in separate institutional state-
ments, while the linkage between the regulated actor and the enforcing
entity is reflected by an Or else linkage as an institutionally mandated
consequence.26
In addition to the differentiated representation of the underlying insti-
tutional semantics, the separated treatment is motivated from an analytical
standpoint, since it provides the basis to represent the conceptual separa-
tion of powers, an aspect of central concern in legal studies [e.g., consti-
tutional (Michaels, 2015) and criminal law (Barkow, 2006)] as well as
public administration (Rosenbloom, 1983) – thereby leveraging extended
analytical opportunities not only for the distinction of norms and rules but
26 We will revisit this scenario for further conceptual nuance not introduced at this
stage.
for extended analysis from a legalistic or normative perspective, respec-

tively. In the context of rules, for example, it enables the assessment
of the extent to which formal institutional power is distributed across
appointed actors. From the perspective of norms, the call for the distinc-
tion of both functions is analytically useful, since both these roles are often
conflated in the context of social enforcement, and, by the very nature of
norms, fuzzily specified in terms of responsibility (if at all). In practice,
specific concerns in a normative setting include the uncertainty as to who
a monitor is in the first place, let alone the nature of the sanctions a
violator may face. Drawing on the trivial example of jay-walking, enforce-
ment of compliance may occur in varying ways, such as self-enforcement
(e.g., avoiding jay-walking based on moral grounds, or in the presence
of children), social enforcement (e.g., reacting bystanders), or, of course,
as legal enforcement (e.g., traffic officers). In addition to the different
types of sanctioners, it may furthermore be unclear how bystanders would
react, both in terms of multitude, as well as forms of sanctioning (see also
Posner & Rasmusen, 1999). Assuming the observation of a transgression
by a formally appointed authority (in contrast to the essentially self-
appointed bystanders), legal consequences operate within explicitly spec-
ified bounds that impose limitations on the extent of discretion enforcers
may employ, but also make those potentially calculable for potential viola-
tors. In addition to the formal appointment of monitors and enforcers for
rules, a noteworthy difference lies in the characterization of the nature
of enforcing actors, which may be anthropomorphic, be imaginary, or
ascribed to the social environment directly as reported in empirical work
by Watkins et al. (2015). Finally, the sanction, or consequences, them-
selves can vary in structure, with normative sanctions being of physical,
potentially economic, and emotional nature,27 as well potentially affecting
social status (and, of course, including any combination thereof). Formal
sanctions can additionally carry explicit institutional effects (in addition
to the effects of individual sanctioning), such as removal from appointed
positions, removal of privileges, potentially modifying the wider institu-
tional setting (e.g., collective consequences; modification of positions,
rights, and responsibilities more generally).28
27 See also Crawford and Ostrom (1995).

28 Related accounts have been reported empirically, e.g., by Anderson et al. (1977),
Siddiki et al. (2012), and Wodahl et al. (2015).
Reflecting the duality of institutional perspectives as well as the vari-

able complementary functions that monitors and enforcers play within
informal (normative) and formal (rule) settings, the IG offers a concep-
tually integrated socio- and legal-institutional perspective that is able to
conceptually capture the origin, content and function of both institutional
concepts, as well as their interaction based on congruence, competition
and conflict. Doing so establishes the basis for a systemic view on institu-
tions that not only opens up novel analytical opportunities with focus on
the interaction between the normative and legalistic perspective, but also
invites for richer analyses within either strand of institutionalism. The New
IG achieves this by affording norms and rules congruence in structure,
while retaining distinction in semantics.
Summarizing the preceding discussion, Table 4.2 collates the distin-
guishing characteristics of norms and rules, including the characterization
of actors as well as features of consequences.
Table 4.2 Semantic distinction between norms and rules in the Institutional
Grammar 2.0
Characteristic Norms Rules
Specification of Monitor Potentially unclear/fuzzy Clear specification

specification (Potentially
contextually implied)
Specification of Enforcer Potentially unclear/fuzzy Clear specification
specification (Potentially
contextually implied)
Appointment of Monitor Self- or informal appointment Formal appointment by
and Enforcera (ad hoc, informal private or public
forum/process) authority/legitimized
forum
Nature of Monitor and Physical, abstract or imaginary Generally human or
Enforcerb actor(s), natural environment organizational actor(s)
Relationship between Potentially unified or separate Unified or separate
Monitor and Enforcer entities, not explicitly specified entities, clear
specification
Consequences Potential uncertainty about Explicit specification,
frequency, nature (e.g., kind, including potential
intensity) and multitude discretion
a The differentiation of monitor types is an extension of the original characterization by Schlüter and
Theesfeld (2010)
b The refined actor characterization draws on Watkins et al. (2015)’s observations
This table has been adapted from Frantz (2015)
Fig. 4.8 Regulative statement type structures
Given the focal emphasis on the semantic distinction of institution

types as expressed in institutional statements, the syntactic representation
is visualized in Fig. 4.8. As motivated above, whereas norms and rules
share the Deontic, as well as the representation of consequences (i.e., the
abstract Or else linking consequences as separate statements) as common
syntactic features, strategies follow the syntactic form of Attributes, Aim,
as well as Context component decomposed into Activation Condition and
Execution Constraint, alongside the optional presence of Object variants
in either institution type. Semantic characteristics of strategies, discussed
to a limited extent to this stage, will be revisited in the discussion of
institution types for constitutive statements in Sect. 4.2.6.
4.2.4 Delta Parameters in the Institutional Grammar

One of the key features introduced by Crawford and Ostrom as part of
the original IG is the concept of delta parameters as a mechanism to oper-
ationalize the Institutional Grammar by dissociating the representation of
institutions, i.e., strategies, norms, or rules, from the incentives that lead
individuals to comply, or conversely, to violate institutions. A particular
reason for this dissociation was the observed limited accessibility of infor-
mation about the commitment to institutions in field settings; reflecting
the ability to describe empirically observed institutions at a reasonable
level of accuracy, whereas incentives that drive compliance behavior were
less overt, especially if no violation (and associated sanctioning) could be
observed (Crawford & Ostrom, 1995).
Delta parameters can occur in various forms, and at various levels

of granularity, as shown in the equations below and discussed in the
following:
δ o = δ oi + δ oe
δ b = δ bi + δ be
= δo + δb
At the highest level, delta parameters (deltas) reflect an aggregate

that integrates the (signed) payoffs associated with compliance (δ o , i.e.,
obeying) and noncompliance (δ b , i.e., breaking) with an institutional
statement, while further differentiating these payoffs into ones of internal
origin (δ oi and δ bi ), and others of an external source (δ oe and δ be ).
Internal sources, as understood in this context, are intrinsic motiva-
tions, emotions [“warm glow” (Andreoni, 1989)], as well as idealistic
motives. External sources, in contrast, reflect extrinsic motivators, such
as enforcement, economic or social utility (e.g., reputation enhancement,
honor).29
Aggregated, the delta parameter () is the operationalized choice as to
whether an individual conforms or abandons an institution it is subjected
to. Where all delta values are set to zero, naturally, no (dis)inclination of
conformance is signaled. Crawford and Ostrom specifically proposed this
operationalization to support their game-theoretical exposition. While it
has found limited adoption in extant IG research (in fact likely related
to the challenge to observe motivations in the first place as indicated
by Crawford and Ostrom), the concept is referenced as one potential
approach to the operationalization of the IG in dynamic institutional
analyses (see Sect. 8.2).
To this stage, we have introduced the general syntactic components,
and established basic forms of institutional statements, such as atomic
institutional statements that express the elementary unit of behavioral
prescription (or state description, as to be discussed in the context of
IG Extended), and combinations thereof to capture complex constructs
embedding multiple activities, e.g., actions performed at the same time,
discretionary activity, associated with regulated behavior (horizontal
29 A wider range of factors underpinning internal or external deltas are discussed by

Crawford and Ostrom (1995, p. 590).
nesting), or representing consequential relationships linking different

actors and/or acts in regulatory settings (vertical nesting).
Following the introduction of the basic structure of institutional state-
ments, this section provided an overview of the different forms that
institutional statements can take in terms of institution types, both with
the purpose of capturing institutions comprehensively in an “Institutional
Grammar”, alongside a review of the reflected institutional-analytical
branches and traditions, followed by a conceptual revision of the different
corresponding institution types to afford a more accurate semantic align-
ment and to provide the basis to extend the analytical depth of the
IG.
While this introduction focuses on the essential representation of
behavior regulation, a second aspect of institutions increasingly moving
into the spotlight of contemporary institutional analysis, the notion of
constitutive statements, has not been captured in the original IG. The
following introduction of a novel constitutive syntax thus moves beyond
the original IG conception, and complements the regulative side with the
purpose of establishing a comprehensive representation of institutions as
observed in theory and practice.
4.2.5 Constitutive Syntax

As introduced in Sect. 4.1.2, regulative statements regulate activities
of specified actors by specifying behavioral expectations or opportuni-
ties in terms of prescriptions, such as obligations and prohibitions, or
permissions, as well as the conditions under which those apply, alongside
the consideration of further constraints. Where explicit injunctions are
absent, regulative statements descriptively account for behavioral conven-
tions. Regulative statements thus offer a clear structural characterization
centering around the actor-action specification.
Constitutive statements, in contrast, focus on expressions pertaining to
the parameterization of the institutional setting, and in theory have been
discussed as “[defining] new forms of behavior” (Searle, 1969, p.33),
with a distinctive focus on the separate specification of regulated activity
and its definition in the first place, while others have noted the possible
limitations of such idealized separation (e.g., Giddens, 1984; Hindriks,
2009), an aspect we return to when discussing Hybrid Institutional
Statements (see Sect. 5.2).
What we can observe in practice, however, is that statements that have

such parameterizing role invariably exist in policy documents, and are
commonly associated with definitions that, in combination, capture essen-
tial properties of the wider institutional setting, which is subsequently
referenced in scenarios that regulate that very behavior.
Motivating the ideal structure of constitutive statements at this stage,
we can observe various commonplace patterns as found in definitions or
declarations:
• “[An] Ingredient [is any]... substance used in the preparation of

an agricultural product that is still present in the final commercial
product as consumed.”
• “The Committee shall consist of a President, Secretary, and Treasurer.”
• “There is hereby established a public Food Security Advisory Board.”
A central concept that accompanies us throughout the introduction of

a constitutive syntactic form of the Institutional Grammar is the Entity,
such as Ingredient, Committee, or Food Security Advisory Board. Comple-
menting the regulative view, constitutive statements signal a perspective
shift centered on an Entity in an abstract sense, that is either defined,
modified, or otherwise affected as part of the institutional statement. Enti-
ties can be actors, actions, objects, roles, venues, infrastructure, as well as
any other artifact of relevance, essentially reflecting the building blocks of
an institutional setting. In contrast to regulative statements that directly
link actor, activity, and context, such entities can be directly defined based
on brute (f)acts, i.e., ascribing institutional meaning to real-world enti-
ties (e.g., “substance used in the preparation …”), be abstract in kind
(e.g., status concepts such as institutional powers), or draw on established
institutional (f)acts for their specification (e.g., a position as a compound
entity linking roles, associated actions and venues, such as the Committee
in the example above). Entities and associated statements can further be
meta-constitutive in kind, such as parameterizing the policy itself (e.g.,
indicating life cycle characteristics, or linking novel to existing policy as
an amendment). However, it is important to note that the constitutive
nature of statements described here is independent of IAD concepts, and
not to be confused with the constitutional and meta-constitutional levels
of analysis in the IAD framework that govern the production of rules
applicable on the respective lower level. Constitutive and regulative state-

ments can exist on any IAD level of analysis, and define, modify, and
regulate entities of relevance on the corresponding level.
Paralleling Searle’s original “X counts as Y in context Z” qualifications,
the general structure that all constitutive statements share is the specifi-
cation of an entity that is defined, modified, or otherwise affected by the
institutional statement, the Constituted Entity. Where existent, Consti-
tuting Properties highlight the content of the parameterization of the
constituted entity – the definiens of the constitutive statement (with the
Constituted Entity as the corresponding definiendum). It is variable in
kind, and can be concrete (i.e., physical) or abstract (e.g., status) and
captures the feature or property linked to the Constituted Entity. The
Constitutive Function, loosely corresponding to the “counts as” in Sear-
le’s characterization, qualifies the linkage of a Constituted Entity with the
institutional setting, or where existing, with Constituting Properties, i.e.,
it defines the entity (e.g., “… is …”), describes structural aspects (e.g.,
“… consists of …”), modifies (e.g., receiving additional properties) or
affects in any other way (e.g., linking of status defined in Constituting
Properties, such as “… is assigned …”). These distinctive components
are complemented by the Context, which, similar to the same compo-
nent in regulative statements, is stratified into conditions that signal the
applicability of the statement (Activation Conditions ), as well as Execu-
tion Constraints that further qualify the Constitutive Functions (as the
equivalent to the Aim in regulative statements). In addition, the consti-
tutive syntax in the Institutional Grammar further recognizes the Modal
to signal the required or optional nature of the constitution expressed in
the constitutive statement. Finally, the Or else signals consequences for
non-fulfillment of the institutional statement content. Unlike a distinctive
sanction or incentive, as is the case for regulative statements, conse-
quences can be of existential kind (e.g., an entity does not come about)
and have a systemic impact (e.g., voiding the policy, or its applicability),
potentially leading to a re-parameterization of the institutional setting.
Combined, these components represent variable forms of statements
that characterize features of an institutional setting, but differ with
respect to the necessity or optionality of these features, as well as the
consequences for not recognizing or enacting constitutive statements.
Mirroring the structure of regulative statements, the individual compo-
nents will be introduced, alongside an explicit elaboration of the general
functional principles and purpose, as well as underlying assumptions and

scope of applicability.
4.2.5.1 Constituted Entity

The Constituted Entity is a central concept of the constitutive syntax and
identifies the entity, or subject, that is defined, modified (e.g., redefined
based on addition or removal of properties, characteristics, relationships,
as well as status), or otherwise affected by the institutional statement.
The Constituted Entity can be any entity of relevance in an institutional
setting, reflecting the most general concept of the IG. Entities can be
of different metaphysical types (e.g., abstract vs. concrete, animate vs.
inanimate), including:
• actors, both of animate or inanimate kind,

• actions/behavior,
• roles,
• objects,
• artifacts,
• venues,
• infrastructure (e.g., processes),
• status characterizations, as well as
• formal and informal institutions themselves.
Essentially, Constituted Entities provide principal entities relevant for a

comprehensive specification of the institutional setting, including contex-
tual characteristics, boundaries, positions, and actions, which are poten-
tially referenced in regulative statements (e.g., actors, roles, and actions
to be regulated), an aspect further discussed in Sect. 4.1.2.
Inasmuch as constitutive statements define entities of such basic kind
with reference to brute (f)acts and thereby introduce those in the institu-
tional setting (as institutional (f)acts or concepts), they can further define
or otherwise parameterize novel entities by referencing existing institu-
tional concepts and thereby capture the intrinsic structural complexity
of institutionally relevant concepts. A simple example is the concept of
the Role that may reference a set of actions the one occupying the
role can possibly exercise. Similarly, object definitions include affordances
(Gibson, 2013), expressed in the characterization of actions an entity can
be subjected to. A certification can, for example, be granted, reviewed, or
revoked. Another more complex example is the concept of a Position in an

action situation, which captures the linkage of roles, relevant actions, as
well as relevant venues a position relates to. Reflecting the parameterizing
function of constitutive statements specifically, the policies themselves (or
specific provisions, sections, etc. therein) can be Constituted Entities if
the statement assumes meta-constitutive function (e.g., by linking specific
provisions, or specifying the expiry of amendment of policies).
Constituted Entities can further hold additional properties (e.g., public
Food Advisory Board), similar to various components on the regulative
side (and specifically discussed in the context of the Attributes compo-
nent in Sect. 4.2.1.1). Conversely, Constituted Entities themselves can be
properties of other objects, actors, roles, etc. (e.g., actions associated with
an actor).
The presence of the Constituted Entity is required for any constitutive
institutional statement.
4.2.5.2 Modal
The Modal is an optional component in the constitutive syntax that
signals the extent to which the parameterization specified in the insti-
tutional statement is either required or optional. As such, the syntactic
function of the Modal mirrors the role of the Deontic in the regulative
syntax. However, unlike the regulative case, constitutive statements do
not restrain or incentivize a specific actor, but instead signal the necessity
of the parameterization (e.g., definition or modification of an entity) to
take effect. Conceptually, the Modal does thus not directly assign respon-
sibility to a particular actor or group of actors, but instead indicates that
an entity exists (e.g., by definition), has to exist or may exist.
Logically, the component makes reference to Modal Logic (Garson,
2021) more general, signaling the depersonalization of responsibility
associated with attaining the specified institutional state (e.g., intro-
ducing actors, actions, venues, etc.). Unlike its deontic counterpart, which
endows the responsible actor with the choice to perform a particular
activity, modals do not oblige or constrain a particular actor, but describe
objective necessities or optionalities in the context of the institutional
setting, so as to “bring the institutional game about.” In consequence,
the use of Modals can invite for a principled debate as to whether
violations of constitutive statements are possible in the first place (an
aspect the alethic logic, a specific branch of Modal Logic, rejects).
Recalling the IG’s broader objective, capturing institutional arrangements
comprehensively, and recognizing their systemic nesting, assessing the

impact of non-enacted institutional statements in terms of configura-
tional effect is of central analytical value. Moreover, since consequences
of non-fulfillment of constitutive statements generally permeate the wider
institutional setting in the form of a potential re-parameterization, an
expanded characterization of consequences is introduced in the context
of constitutive statements, and further discussed in Sect 4.2.5.6.
Noteworthy, however, is the potentially complementary nature of
constitutive and regulative statements in this context, where the constitu-
tive statement may specify that a particular entity is necessarily established,
but leaves it to a complementary regulative provision to operationalize
this constitution, e.g., by assigning specific responsibilities associated with
the entity defined as part of the constitutive statement to actors – a central
dialectic linked to the analysis of Hohfeld’s jural correlatives (Hohfeld,
1913). An example of such linkage is the abstract specification of a
right ascribed to actors or roles, and a corresponding regulative state-
ment specifying how this comes about. Building on the introduction of
Hybrid Institutional Statements in Sect. 5.2, analytical opportunities will
be discussed in Sect. 8.3.
A specific concern in the context of the Modal is its variable use
in linguistic expression. Its interpretation can vary from being merely
stylistic to serving a well-defined function in legal practice or convention.
Examples include the systematic use of directives in international treaties
(e.g., Bodansky, 2016), or specific meanings in the context of standards
(e.g., Björnsson & Shanklin, 2014; Bradner, 1997). This implies a devi-
ating default interpretation of the Deontic from the Modal. Whereas the
absence of Deontic values acts in a descriptive form associated with behav-
ioral conventions and strategies in absence of explicitly codified external
enforcement, in the context of constitutive statements, the absence of a
Modal generally implies the requirement or necessity for the state spec-
ified in the institutional statement to come about. For example, the
definition of an entity assumes that the entity has the associated ontolog-
ical status in the context of the institutional setting, i.e., it “has to” exist
as a concept. If a definition would be explicitly annotated with the indica-
tion that an entity may have a certain meaning, in constitutive statements
this would need to be made explicit.
4.2.5.3 Constitutive Function

An essential explicit component of any constitutive statement is the
Constitutive Function. Unlike Searle’s “X is Y in Context Z” syntax,
in which the functional descriptor expressed in the institutional state-
ment is fixed and carries implied “is” or “counts as” semantics (see also
Grossi et al., 2008), the Constitutive Function expresses the relationship
between Constituted Entity and the institutional setting by capturing the
effect of the linguistically associated expression on the institutional envi-
ronment. Given the purpose of constitutive statements to capture the
parameterization of an institutional setting, the explicit characterization
of the function that establishes the linkage between the entity created
(e.g., “is”), related (e.g., “is governed by”), or otherwise affected, and
the institutional setting based on the effect that the expression signals.
A central focus of the constitutive syntax in the IG is thus to capture
distinctive characterizations of different functions to analyze their func-
tional role with respect to the construction and/or (re-)configuration of
an institutional setting.
As indicated in the introductory paragraph and discussed here at
brevity,30 common constitutive functions include the indication of:
• Definitions (e.g., defining an actor, action, role, object, artifact,

status, venue, etc. – see Sect. 4.2.5.1) in intensional (e.g., “… is
…” ) or extensional forms (e.g., “… does …” ).
• Relationships of entities (e.g., “… resides within …” ), such as
embedding entities in an organizational context or defining entities
based on composition relationships (e.g., specifying that a committee
consists of a specific number of members).
• (Dis)Establishment of entities by specifying the circumstances under
which they come about, are suspended, or cease to exist (e.g.,
“… from 1st January onward, until 31st December” ) – effectively
reflecting an entity’s lifecycle.
• Modification of existing entities as part of a constitutive statement,
such as associating novel properties with an entity of concern (e.g.,
amending an action specification or modifying actor characteristics).
30 Constitutive Functions find explicit treatment in the Constitutive Functions Taxonomy

(Sect. 6.1.2.4) discussed in the context of IG Logico (Sect. 6.1).
• Conferral of status, such as assigning or conferring power (e.g., in

the form of authority), rights, or other forms of status to roles or
actors (e.g., Searle, 2005).
While the perspective on the effects of constitutive statements is

primarily linked to entities embedded in the institutional setting, consti-
tutive statements can have policy-level (or meta-constitutive) functions in
controlling life cycle aspects of the policy as a whole, i.e., defining when a
policy itself comes into force (and thereby activating the entities defined
therein), an aspect previously referenced in the context of the Constituted
Entity.
Common content and effects of C onstitutive F unctions are discussed
in greater detail in the context of the Constitutive Functions Taxonomy in
Sect. 6.1.2.4.
4.2.5.4 Constituting Properties

Constituting Properties complement the relationship between Constituted
Entity and Constitutive Function by capturing the characteristics func-
tionally attached or otherwise related to the Constituted Entity, reflecting
the linguistic function of a predicate that is grammatically linked to the
subject it describes. Reiterating the stereotypical definitional structure of
constitutive statements, the Constituted Entity reflects the definiendum,
whereas Constituting Properties hold the corresponding definiens. Given
the flexibility of constitutive statements to variably link brute (f)acts and
institutional (f)acts as well as institutional (f)acts to other institutional
(f)acts, the Constituting Properties of one statement may be the Consti-
tuted Entity in another. Whereas, for example, the signing of a paper
titled “Lease Contract” as a brute act (Constituting Properties ) may be
the basis to constitute the institutional concept of a lease (Constituted
Entity), defining its cancelation as an institutional concept (Constituted
Entity) relies on the existence of the institutional concept lease (then as
Constituting Properties ) in the first place.
In contrast to the Constituted Entity and Constitutive Function compo-
nents, the presence of Constituting Properties in a constitutive statement
is optional, since institutional concepts do not necessarily rely on a linkage
to existing brute or institutional concepts to come about (something may
just be declared as existing without any further qualification, e.g., “there
exists a Council”). As with Attributes, Object forms, and Constituted Enti-

ties, Constituting Properties can have nested or associated properties that
qualify or characterize the component information further.
Exemplifying the use of C onstituting P roperties in conjunction with
the previously introduced components, we can now systematically classify
the essential components of the statement “The Committee shall consist of
a President, Secretary, and Treasurer,” as shown in the following.
Constituted Entity: The Committee
Modal: shall
Constitutive Function: consist of
Constituting Properties: President, Secretary, and Treasurer
Activation Condition: at all times
Execution Constraint: N/A.
In addition to the variable nature of the Constituting Properties

component as referencing either brute or institutional (f)acts in the first
place, in this example, the observer will notice multiple elements that in
combination compose the Constituting Properties. As with other compo-
nents, the explicit capturing of such component-level combinations is
subject to encoding as discussed in Sect. 4.2.2 and, for the previous
example, is functionally equivalent to the expansion shown below:
4.2.5.5 Context
While the previous components for constitutive statements hold distinc-
tive functions that deviate from the interpretation on the regulative
side, both the Context component and the Or else are syntactically and
structurally identical, but their embedding in the constitutive statement
structure and functional relationship to other constitutive components
naturally varies, alongside an expanded characterization of both compo-
nents to accurately capture their use in the context of both regulative and
constitutive statements.
The Context component, as in the regulative case, is stratified into
Activation Conditions and Execution Constraints , with Activation Condi-
tions capturing the conditions under which the institutional statement
applies.31 In constitutive statements, Execution Constraints are linked
31 More precisely, Activation Conditions indicate the conditions under which the non-
activation condition part of a statement applies.
with the Constitutive Function (as opposed to the Aim as is the case for
regulative statements), both to qualify the constitutive process expressed
therein, but also to capture potential effects on the institutional environ-
ment associated with the specific constitutive statement. While this char-
acteristic equally applies to the use of Execution Constraints in the context
of regulative statements, the latter aspect is specifically pronounced in
constitutive statements due to the parameterizing function these state-
ments assume with respect to the institutional setting they describe. In
addition to the general qualification offered in IG Core, in the context of
IG Extended we will discuss richer categorizations to draw out additional
analytical value beyond the distinction into components that either reflect
the preconditions or qualification of statements.
The following example illustrates the stereotypical use of the Context
components in constitutive statements:
Constituted Entity: Council
Modal: shall
Constitutive Function: include
Constituting Properties: organic farming representatives
Activation Condition: From 1st January onward
Execution Constraint: to review chemical allowances within
organic food production standards
4.2.5.6 Or Else
The final component on the constitutive side is the Or else. Similar to
the regulative side, it is abstract in kind, and instead of capturing state-
ment content, it in fact reflects the logical linkage between statements that
prescribe monitored behavior, and statements that specify consequences
for noncompliance.
While consequences on the regulative side have been introduced as
capturing enforcement activity originating from actors or the environment
itself (see Sect. 4.2.1.6), consequences related to constitutive statements
are of different quality. Recalling the principal objectives, constitutive
statements parameterize the institutional setting by introducing or modi-
fying entities, such as actions, objects, artifacts, and other features as
discussed in Sect. 4.2.5.1. Any omitted or foregone parameterization
consequently leads to a modification of the wider institutional setting,
potentially rendering the parameterization invalid (or at least incongruent
with the institution in form), or, following Wittgenstein’s account (1983),

making institutional arrangements meaningless by design.
Reflecting this distinctive effect of non-fulfillment of constitutive
statements, the Or else operator thus allows for the specification of conse-
quences of existential quality, such as declaring the invalidity of the
policy – the institutional game entirely –, or, activation of consequences
in an institutional arrangement that embeds the invalidated institutional
arrangement. The notion of consequences in the IG is inclusive of meta-
configurational effects that transcend individual action situations. As with
regulative statements, and discussed in Chapter 3 as well as Sect. 4.2.3,
the specification of consequences, let alone their effect may not neces-
sarily be explicitly stated, nor be immediately evident. While higher-level
policy, outside the analyzed one, may govern the treatment of non-
fulfilled constitutive provisions more generally (as might be seen in legal
systems that recognize hierarchies of laws, and elevate the interpreta-
tion correspondingly) without explicit reference, the interpretation may
alternatively draw on dedicated sections in the policy document. This
may further include “blanket,” or “umbrella” statements, such as salva-
tory clauses, guarding the policy, regulation, or contract in its entirety
against invalidity of specific statements, alongside specific conventions
for document drafting that respond to a domain, type of document,
legal context, etc. Naturally (and indicated previously), the invalidity of
selected aspects of a policy can have a configurational effect by effectively
re-parameterizing the institutional setting alongside actors and activi-
ties governed therein. Integrating the kinds of consequences discussed
in the context of regulative statements (Sect. 4.2.1.6) and expanded
in Sect. 4.2.3, the IG essentially differentiates between existential and
non-existential consequences.
Beyond the meta-configurational function alluded to above, commonly
occurring existential consequences are ones that afford the addition,
removal or modification of a kind or type of entity (e.g., actors) to or from
an institutional setting, and, based on the modification of the institutional
design (as opposed to operational implementation and instantiation – the
“runtime structure”), have the parameterizing effect that can potentially
transcend action situations and encompass the wider institutional setting.
A basic example that highlights the implicit existential nature of
consequences is the following:
Committee must consist of at least three members [, or else committee does

not exist/come about].
Reiterating the more general form of non-existential consequences

implicitly referenced in Sect. 4.2.1.6, those are ones that affect individual
instances of entities, such as actors, e.g., based on the administration of
punitive measures originating from an enforcing actor, of human, corpo-
rate and anthropomorphized kind, that has a social effect (e.g., social
rebuke), carry economic consequences (e.g., fines), or affects other forms
of institutional status (e.g., demotion).
Reflecting these general categories, the following (regulative) examples
showcase typical use cases:
Cars must not exceed speed limit, or else enforcement officer may impose
fine.
Students must not plagiarize, or else instructor will assign student
failing grade.
Certified organic farmers must follow organic farming provisions, or
else certifier may revoke certification.
An important functional distinction of the effects of consequences

relies on the nature of the monitored institutional statement. In the case
of constitutive statements, the consequence can relate to the activation of
the Constitutive Function in the first place, i.e., modify or void the func-
tional relationship expressed in the leading institutional statement and
thereby implicitly affecting the Constituted Entity (e.g., by not bringing it
into existence). Alternatively, the consequence can target entities external
to the statement (i.e., have parameterizing function transcending the
specific statement). On the regulative side, in contrast, the consequence
can relate to entities internal (e.g., actor, action, object) or external to the
statement itself (and thereby implicitly link to the motivations that may
drive compliance [see Sect. 4.2.4]), but not affect the regulative function
the leading monitored statement reflects (i.e., the consequence does not
modify the Aim and the associated directive, the Deontic, of a regulative
statement).
Inasmuch as the Or else specifies consequences, its activation is contin-
gent on the noncompliance with (for regulative statements), or non-
fulfillment or -satisfaction of (for constitutive statements), the leading
monitored statement in the first place. Most notably, the role of the Modal
is central to assess non-fulfillment. As indicated before – and contrasting
the regulative side – the absence of an explicit Modal implies a necessity
to fulfill the provision contained in the monitored statement; it does not

signal optionality as is the case for regulative statements.
Widening the nature and forms of consequences for violation of consti-
tutive or regulative statements, variably originating from an actor, the
environment, or from the policy, or embedding the action situation itself,
offers the basis for an integrated treatment of regulative and constitutive
statements more generally, such as the specification of actor-centric (regu-
lative) consequences for the non-fulfillment of a constitutive provision, or
an existential consequence (e.g., non-applicability of specific provisions)
associated with a violated regulative statement. We will discuss the associ-
ated concept of Hybrid Institutional Statements and relevant theoretical
and conceptual implications in greater detail in Sect. 5.2.
4.2.5.7 Discussion
Reflecting on the essential role of constitutive statements as a comple-
mentary concept to regulative statements, it is important to note the
variable semantics expressed in the interactions of the statement compo-
nents. As signaled at the beginning of this section, constitutive statements
center around the Constituted Entity as a central concept, mediated by the
Constitutive Function. As with regulative statements, Activation Condi-
tions indicate the preconditions under which the statement applies. Where
relevant, Constituting Properties provide the necessary descriptors in the
form of predicates, existing entities or other forms of properties that
serve as input to the Constitutive Function, where the Constitutive Func-
tion signals the specific institutional semantics of the functional linkage
between Constituted Entity and Constituting Properties. Equivalent to
regulative statements, the function may further be qualified by Execution
Constraints, e.g., expressing how the function is performed, or signaling
the purpose, outcome, or secondary effects of the statement, beyond
the primary effect associated with the Constituted Entity. The Modal,
finally, takes the role in signaling the compulsory or optional nature of
the statement enactment. Figure 4.9 schematically highlights the concep-
tual relationships of the constitutive statement components, where dashed
lines represent potential influences. Where applicable, components are
further related to their Searlean correspondent.
Recognizing the basic constitutive syntax and the underlying existential
dimension begs the concern for a richer characterization of institution
types, akin to the strategies, norms, and rules stratification as offered by
Crawford and Ostrom, and refined in this work (see Sect. 4.2.3).
Fig. 4.9 Constitutive statement component relationships
Statements primarily carrying constitutive weight have an inherently

declarative role conveyed in the constitutive function, i.e., bringing about
a change of affairs permeating the institutional setting. As stated in
Sect. 4.2.3, constitutive statements provide the fixture on which regu-
lative statements anchor, e.g., by referencing actors, actions, venues, and
other contextual characterizations, the former establish and relate in the
institutional arrangement. Naturally, this generic purpose is reflected in
the divergence of functions constitutive statements can assume in oper-
ational settings, such as defining entities, assigning roles, or conferring
status.
Absent an explicit monitor responsible for enforcement (either by self-,
social, or public enforcement as on the regulative side), the consequences
associated with the violation of statements on the constitutive side inher-
ently carry parameterizing weight, albeit with varying effect and thus
impact on the institutional setting.
4.2.6 Constitutive Institution Types

Corresponding to the characterization on the regulative side, the distinc-
tion between different institution types is inherently grounded in the
arena from which the rules emerge and to which they apply; as opposed
to structural characteristics linked to individual components. The forums
that lead to the definition and adoption of constitutive statements can
vary by the degree of formality, allowing for the stratified characterization
of norms and rules on the grounds introduced for regulative statements.
The legitimation of normative provisions, for example, is inherently social;
the ontology for entities established in the social realm is socially estab-
lished, outside of a formally legitimized decision-making forum. This
forum is characterized by social influence that drives convergence toward
shared conceptions (Abrams et al., 1990), emerging from or defining the
shared social reality and thereby implicitly the identity of the underlying
group (Hogg & Turner, 1987; Turner, 1991). This involves, for example,
socially enforced consensus on what “violation” constitutes in the specific
socio-institutional context. Inadvertently – and not further discussed at
this point – norms naturally bear the potential to lead to divergent under-
standings, if not driving polarization among groups (Isenberg, 1986) – a
motivation for formal legitimation in the first place.
Congruent with the rule conception on the regulative side, the basis
of constitutive rules rests on their establishment in formally recognized
forums (e.g., legislative, courts, tribunal) that define and impose concep-
tualizations that form the basis for the construction and regulation of
institutional arrangements.
A special role in the context of constitutive institution types is asso-
ciated with strategies. On the regulative side, strategies reflect strategy
choices shared among a set of players, where the designation of these
strategies does not include any prescriptive component and the under-
lying motivations are inherently internal (or unknown), i.e., the nature
of the behavior is akin to descriptive norms (Cialdini et al., 1991) that
are conventionally adopted, but deviation from which does not carry any
form of sanction. Relately, the extent to which those strategies are shared
may vary.
However, contrasting the strategy characterization on the regulative
side, we recognize a different form of parameterizing statements on
the constitutive side. Where regulative strategies generally describe the
behavior of entities who are subject to the provision, constitutive state-
ments offer similar behavior characterizations of the source or forum
from which the statement originates or with which it associates, without
necessarily being subject to the provision itself. While parameterizing,
statements of this kind, include commitments, declarations of intent, or
recognition of states of affairs by the involved parties to a cause, action,
or aspiration linked to the policy of concern, or to establish a shared
understanding or laying out assumptions that provide the backdrop for
the interpretation of the broader institution. However, absent broader
qualification, constitutive strategies do not carry legal weight prima facie,
but may signal moral obligations.
An example of such constitutive strategy is “We, the city, recognize the

availability of food as a matter of public concern.”
In this statement, an actor – generally an entity intimately involved or
affected by the policy of concern – signals the recognition of the state
of affairs and concerns the associated policy aims to address, thereby
supporting the interpretation of the policy from the perspective of the
underlying intention, while, at the same time, avoiding legal exposure.
Central to constitutive statements is their parameterizing function: they
set the stage for the policy at large and contextualize the institutional
statements within an institution according to a broader state, meta-state,
or meta aspiration, providing the basis for their operational and poten-
tial judicial interpretation. Statements of this kind are commonly situated
in preambles of legal text or associated commentary. Unlike norm and
rule statements, constitutive strategies cannot be assessed with respect to
“fulfilment” in a strict legal sense, and thus, similar to their regulative
counterpart, do not carry explicit consequences.
Summarizing the preceding discussion, the following heuristics support
the operational identification of constitutive strategies:
• Consequences for their violation are non-existential, i.e., their viola-

tion does not void the policy or parts thereof.32
• These statements are de facto discretionary in kind since no obliga-
tions or consequences are explicitly imposed on any referenced or
implied actor.
• They may capture an aspiration that statements in the policy aim at
aligning with in spirit, providing the basis for their moral or legal
interpretation.
• The subject of the statement is generally involved with the policy
design process (e.g., as a member of decision-making forums,
authority, etc.).
It is important to note that such statements contrast with statements

of fact, i.e., statements stating the occurrence of actual factual circum-
stances – independent of associated intentions or objectives as is the case
32 The existential nature is here understood in the legal sense, and does not imply
unintended consequences such as limited confidence in policies, e.g., due to poor
implementation.
Table 4.3 Institution types of regulative and constitutive statements
Institution type Regulative variant Constitutive variant
Strategies behavioral conventions, commitments, declarations of

descriptive norms; discretionary intent, recognition of affairs;
discretionary
Norms injunctive norms and informally established
consequences; informal conceptualization and
monitoring and enforcement parameterization of institutional
features/setting; socially binding
Rules legal provisions and formally established
consequences; formal conceptualization and
monitoring and enforcement parameterization of institutional
features/setting; formally
binding
for strategies, or association with a particular actor. An example of such a

statement is “The meeting was held in Seville on the 5th November 1973.”
Completing the discussion on institution types, the preceding Table
4.3 summarizes the central features of both regulative and constitutive
statements.
The constitutive institution types as introduced here provide the func-
tional complement to the regulative structures that govern behavior, with
structures that define or otherwise parameterize entities (actors, actions,
environment). Both types of institutional statements represent idealized
forms, and while those can reflect the characterization of statements of
definitional (“… entity is defined as …” ) and prescriptive kind (“… actor
must perform …” ) in idiomatic form, in practice as well as theory do
we recognize divergent perspectives on the relationship and distinction
of those rule types (e.g., Giddens, 1984; Hindriks, 2009). To accommo-
date and potentially harmonize these debates, but more importantly, to
accommodate empirical observations, we introduce the concept of Hybrid
Institutional Statements , a feature discussed at greater depth in Sect. 5.2.
4.3 Summary of Chapter Content

This chapter introduced the conceptual umbrella of the Institutional
Grammar 2.0 by providing an overview of general principles, such as
the overarching objectives of ontological consistency, comprehensiveness of
parsing and representation, and advanced analytical uses based on compu-

tational tractability, as well as the levels of expressiveness that group specific
features of the IG 2.0 to manage the trade-off of conceptual compatibility
across those levels while accommodating specific analytical objectives.
Figure 4.10 provides an integrated overview of the features associated
with distinctive levels of expressiveness, which are highlighted in the
following.
Of these levels, this chapter introduced the lowest level IG Core as
the basic level of expressiveness (highlighted in Fig. 4.10) of the New
Institutional Grammar, and specifically highlights the syntactic compo-
nents for both regulative and the novel constitutive syntactic forms,
with the intent to resolve long-standing inconsistencies in representa-
tion (e.g., Conditions ), as well as to establish comprehensive encoding
of statements. To further abstract from stylistic features for compound
expressions, enumerations, etc., the IG 2.0 introduces the concept of
nesting to reflect components (e.g., multiple aims) and groups of compo-
nents (e.g., multiple actors and aims) in their elementary form (horizontal
nesting), as well as to represent consequential relationships between
different institutional statements (vertical nesting), substituting Crawford
and Ostrom’s original Or else component with a logical linkage of distinct
Fig. 4.10 IG 2.0 Features by Level of Expressiveness

institutional statements. Reflecting on the diversity of application of the

IG in extant work, alongside theoretical proposals, the IG 2.0 further
includes a refined characterization of institution types that moves beyond
a syntactic characterization of institution types, but establishes theoretical
validity by drawing on the underlying processual and contextual features
to distinguish between formal and informal institutions. At the same time,
the range and nature of consequences present in institutional statements
are extended in response to empirical observations as well as conceptual
necessity to accommodate the novel constitutive syntactic form.
However, despite the range of features introduced on this basic level of
expressiveness, statements encoded to IG Core may still embed untapped
structural complexity within or across components, either left untreated
by the basic parsing, or obscured by style. At this stage, the IG further
lacks mechanisms for the systematic categorization of Context as relevant
for the semantic anchoring and linking of institutional statements in an
action situation, or wider institutional setting. These aspects are subject to
the representational conventions introduced under the label IG Extended,
and introduced in the following chapter.
References
Abrams, D., Wetherell, M., Cochrane, S., Hogg, M. A., & Turner, J. C. (1990).
Knowing what to think by knowing who you are: Self-categorization and the
nature of norm formation, conformity and group polarization. British Journal
of Social Psychology, 29(2), 97–119. https://doi.org/10.1111/j.2044-8309.
1990.tb00892.x
Acemoglu, D., & Jackson, M. O. (2017). Social norms and the enforcement of
laws. Journal of the European Economic Association, 15(2), 245–295. https://
doi.org/10.1093/jeea/jvw006
Akerlof, G. (1976). The economics of caste and of the rat race and other woeful
tales. Quarterly Journal of Economics, 90(4), 599–617. https://doi.org/10.
2307/1885324
Anderson, L. S., Chiricos, T. G., & Waldo, G. P. (1977). Formal and informal
sanctions: A comparison of deterrent effects. Social Problems, 25(1), 103–114.
https://doi.org/10.2307/800471
Andreoni, J. (1989). Giving with impure altruism: Applications to charity
and ricardian equivalence. Journal of Political Economy, 97 (6), 1447–1458.
http://www.jstor.org/stable/1833247
Axelrod, R. (1986). An evolutionary approach to norms. The American Political
Science Review, 80(4), 1095–1111. https://doi.org/10.2307/1960858
Barkow, R. E. (2006). Separation of powers and the criminal law. The Stanford
Law Review, 58, 989–1054. http://www.jstor.org/stable/40040287
Bicchieri, C. (2006). The Grammar of society: The nature and dynamics of social
norms. Cambridge University Press.
Björnsson, G., & Shanklin, R. (2014). Must, ought and the structure of stan-
dards. In F. Cariani, D. Grossi, J. Meheus, & X. Parent (Eds.), Deontic
logic and normative systems (pp. 33–48). Springer International Publishing.
https://doi.org/10.1007/978-3-319-08615-6_4
Bodansky, D. (2016). The legal character of the Paris agreement. Review of Euro-
pean, Comparative and International Environmental Law, 25(2), 142–150.
https://doi.org/10.1111/reel.12154
Bradner, S. (1997). Key words for use in RFCs to Indicate Requirement Levels.
Internet Engineering Task Force (IETF). https://www.ietf.org/rfc/rfc2119.
txt
Calvert, R. L. (1995). Rational actors, equilibrium, and social institutions. In
J. Knight & I. Sened (Eds.), Explaining social institutions (pp. 57–93).
University of Michigan Press.
Cherry, C. (1973). Regulative rules and constitutive rules. Philosophical Quar-
terly, 23(93), 301–315. https://doi.org/10.2307/2218059
Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of norma-
tive conduct: A theoretical refinement and reevaluation of the role of norms
in human behavior. Advances in Experimental Social Psychology, 24(100),
201–234. https://doi.org/10.1016/S0065-2601(08)60330-5
Coleman, J. S. (1990). Foundations of social theory. Harvard University Press.
Coleman, J. (1988). Social capital in the creation of human capital. The American
Journal of Sociology, 94(1988), 95–120.
Commons, J. R. (1968). Legal foundations of capitalism. University of Wisconsin
Press.
2975
Finnemore, M., & Sikkink, K. (1998). International norm dynamics and polit-
ical change. International Organization, 52(4), 887–917. https://doi.org/
10.1162/002081898550789
nADICO: A nested grammar of institutions. Lecture Notes in Computer
Science, 8291 LNAI , 429–436. https://doi.org/10.1007/978-3-642-449
277_31
Frantz, C. K., Keertipati, S., Purvis, M. K., & Purvis, M. (2015a). Subjec-
tive interpretation of directive terms: Methodology and preliminary results.
Proceedings – 2015 International Conference on Culture and Computing,
Culture and Computing 2015. https://doi.org/10.1109/Culture.and.Com
puting.2015.40

Frantz, C. K., Purvis, M. K., Savarimuthu, B. T. R., & Nowostawski, M.
(2015b). Modelling dynamic normative understanding in agent societies.
Scalable Computing, 16(4), 355–380. https://doi.org/10.12694/scpe.v16i4.
1128
Frantz, C. K. (2015). Agent-based institutional modelling: Novel techniques for
deriving structure from behaviour (Doctoral dissertation). University of Otago.
Dunedin, New Zealand. https://hdl.handle.net/10523/5906
Galan, J. M., & Izquierdo, L. R. (2005). Appearances can be deceiving: Lessons
learned re-implementing Axelrod’s evolutionary approach to norms. Journal
of Artificial Societies and Social Simulation, 8(3), 2.
Garson, J. (2021). Modal logic. In E. N. Zalta (Ed.), The Stanford Encyclopedia
of Philosophy. Metaphysics Research Lab, Stanford University. https://plato.
stanford.edu/archives/sum2021/entries/logic-modal/
Ghorbani, A., Aldewereld, H., Dignum, V., & Noriega, P. (2013). Shared strate-
gies in artificial agent societies. In H. Aldewereld & J. Sichman (Eds.),
Coordination, organizations, institutions, and norms in agent systems viii
(pp. 71–86). Springer. https://doi.org/10.1007/978-3-642-37756-3_5
Gibson, J. J. (1986). The ecological approach to visual perception (1st ed.).
Psychology Press. https://doi.org/10.4324/9780203767764
Gilbert, M. P. (1989). On social facts. Routledge.
Greif, A. (2006). Institutions and the path to the modern economy: Lessons from
medieval trade. Cambridge University Press.
Greif, A., & Kingston, C. (2011) Institutions: Rules or equilibria?. In N.
Schofield & G. Caballero (Eds.), Political economy of institutions, democracy
and voting. Springer. https://doi.org/10.1007/978-3-642-19519-8_2
Greif, A., & Laitin, D. (2004). A theory of endogenous institutional change.
American Political Science Review, 98(4), 633–652. https://doi.org/10.
1017/S0003055404041395
Grossi, D., Meyer, J.-J. C., & Dignum, F. (2008). The many faces of counts-as: A
formal analysis of constitutive-rules. Journal of Applied Logic, 6(2), 192–217.
Hansen, J., Pigozzi, G., & van der Torre, L. (2007). Ten philosophical problems
in deontic logic. In G. Boella, L. van der Torre, & H. Verhagen (Eds.),
Normative multi-agent systems. Dagstuhl Seminar Proc. 07122.
Hindriks, F. (2009). Constitutive rules, language, and ontology. Erkenntnis,
71(2), 253–279. https://doi.org/10.1007/s10670-009-9178-6
Hogg, M. A., & Turner, J. C. (1987). Intergroup behaviour, self-stereotyping
and the salience of social categories. British Journal of Social Psychology, 26(4),
325–340. https://doi.org/10.1111/j.2044-8309.1987.tb00795.x
Hohfeld, W. N. (1913). Some fundamental legal conceptions as applied in judi-

cial reasoning. The Yale Law Journal, 23(1), 16–59. https://doi.org/10.
2307/785533
Isenberg, D. J. (1986). Group polarization. A critical review and meta-analysis.
Journal of Personality and Social Psychology, 50(6), 1141–1151. https://doi.
org/10.1037/0022-3514.50.6.1141
Klement, K. C. (2004). Propositional logic. Internet Encyclopedia of Philosophy.
https://iep.utm.edu/prop-log/
Lewis, D. K. (1969). Convention: A philosophical study. Harvard University Press.
McNamara, P. (2006). Deontic logic. Handbook of the History of Logic, 7 , 197–
288. https://doi.org/10.1016/S1874-5857(06)80029-4
Menger, C. (1963). Problems in economics and sociology. University of Illinois
Press.
Michaels, J. D. (2015). An enduring, evolving separation of powers. Columbia
Law Review, 115(3), 515–598. https://www.jstor.org/stable/43267874
de Moor, T. (2015). The dilemma of the commoners: Understanding the use of
common-pool resources in long-term perspective. Cambridge University Press.
Nee, V. (1998). Norms and networks in economic and organizational perfor-
mance. American Economic Review, 88(2), 85–89. https://doi.org/10.
2307/116898
North, D. C. (1990). Institutions, institutional change, and economic perfor-
mance. Cambridge University Press.
North, D. C. (1991). Institutions. Journal of Economic Perspectives, 5(1), 97–
112.
Posner, E. A. (2000). Law and social norms. Harvard University Press.
Posner, R. A., & Rasmusen, E. B. (1999). Creating and enforcing norms, with
special reference to sanctions. International Review of Law and Economics,
19(3), 369–382.
Prakken, H., & Sergot, M. (1997). Dyadic Deontic logic and contrary-to-duty
obligations. In D. Nute (Ed.), Defeasible Deontic logic, synthese library, vol.
263. Springer. https://doi.org/10.1007/978-94-015-8851-5_10
Robbins, I. (2018). ‘And/Or’ and the proper use of legal language. Maryland
Law Review (Baltimore, Md.), 77 (2), 311–336.
Rosenbloom, D. H. (1983). Public administrative theory and the separation of
powers. Public Administration Review, 43(3), 219–227. https://doi.org/10.
2307/976330
22(4), 445–475. https://doi.org/10.1177/1043463110377299
Schotter, A. (1981). The economic theory of social institutions. Cambridge

University Press.
University Press.
Searle, J. R. (2005). What is an institution? Journal of Institutional Economics,
1(1), 1–22.
Shepsle, K. A. (2009). Rational choice institutionalism. The Oxford Handbook of
Political Institutions. https://doi.org/10.1093/oxfordhb/9780199548460.
003.0002
Siddiki, S., Basurto, X., & Weible, C. M. (2012). Using the institutional grammar
tool to understand regulatory compliance: The case of Colorado aquaculture.
Regulation and Governance, 6(2), 167–188. https://doi.org/10.1111/j.
1748-5991.2012.01132.x
Siddiki, S., Espinosa, S., & Heikkila, T. (2018). Contextualizing compliance in
the public sector. Routledge.
97.x
rule changes in an institutional context: The ADICO sequence. Advances in
Complex Systems, 2(11), 199–215. https://doi.org/10.1142/S02195259080
0157X
Tuomela, R. (1995). The importance of us: A philosophical study of basic social
notions. Stanford University Press.
Turner, J. (1991). Social influence. Open University Press.
Ullmann-Margalit, E. (1977). The emergence of norms. Clarendon Press.
von Hayek, F. A. (1945). The use of knowledge in society. American Economic
Review, 35, 519–530.
von Neumann, J., & Morgenstern, O. (1947). Theory of games and economic
behavior. Princeton University Press.
von Wright, G. H. (1951). Deontic logic. Mind, 60(237), 1–15. https://doi.
org/10.1093/mind/LX.237.1
Watkins, C., Westphal, L. M., Gobster, P. H., Vining, J., Wali, A., & Tudor,
M. (2015). Shared principles of restoration practice in the Chicago wilderness
region. Human Ecology Review, 21(1), 155–178. https://doi.org/10.22459/
HER.21.01.2015.07
Williamson, O. E. (1975). Markets and hierarchies, analysis and antitrust
implications: A study in the economics of internal organization. Free Press.
Winter, S. C., & May, P. J. (2001). Motivation for compliance with envi-
ronmental regulations. Journal of Policy Analysis and Management, 20(4),
675–698. https://doi.org/10.1002/pam.1023
Wittgenstein, L. (1983). Remarks on the foundations of mathematics (G. H. von

Wright, R. Rhees, & G. E. M. Anscombe, Eds.). MIT Press.
Wodahl, E. J., Boman, J. H., & Garland, B. E. (2015). Responding to probation
and parole violations: Are jail sanctions more effective than community-based
graduated sanctions? Journal of Criminal Justice, 43(3), 242–250. https://
doi.org/10.1016/j.jcrimjus.2015.04.010
Young, H. P. (1998). Social norms and economic welfare. European Economic
Review, 42(3–5), 821–830. https://doi.org/10.1016/s0014-2921(97)001
38-4
Young, O. R. (2016). On environmental governance: Sustainability, efficiency,
and equity. Routledge. https://doi.org/10.4324/9781315633176
CHAPTER 5
Institutional Grammar 2.0: Deep Structural

Parsing and Hybrid Institutional Statements
5.1 IG Extended
Where IG Core focuses on an inclusive structural parsing capturing all
relevant statement features, IG Extended focuses on deep parsing of insti-
tutional statements, drawing structural linkages obscured by statement
structure and expressive style. As part of this, IG Extended initially focuses
on select components to facilitate a perspective shift that extracts addi-
tional institutional information, before reintegrating those coherently into
the institutional statement structure, fostering the basis for a systemic
perspective on institutional arrangements expressed in institutional state-
ments. The resulting statements are amenable to systematic statement
transformations relying on institutional content alone. A central second
aspect is the introduction of richer context characterizations based on
the systematic use of a Context Taxonomy that provides the basis for
the metaphysical linkage of components, where relevant. Analytically, this
offers novel opportunities; inasmuch as IG Core emphasizes and supports
a primarily descriptive analytical perspective, IG Extended puts a stronger
emphasis on a constructionist perspective on institutional statements.
5.1.1 Component-Level Structure

IG Core emphasizes statements at large with a primarily descriptive
focus. IG Extended puts individual components into focus in order to

Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2_5
extract deep structures that capture the interdependencies of institutional

statements, forming the basis for an analytical reconstruction of the insti-
tutional setting at large. Thus, what is meant by deep structural analysis, is
a comprehensive investigation of structural features of components, rather
than structural features of statements only, which fundamentally allows for
a more nuanced understanding of component and statement linkages, and
in consequence, the institutional setting.
Motivating the fine-grained parsing of institutional statements on
component level is the observation that the application of statements is
variably anchored in context situated outside (e.g., temporal, spatial, or
domanial characteristics) or within the analyzed action situation. Moti-
vating such contextual anchoring, let us draw on the following stylized
example:
When organic farmers violate provisions of the organic farming regulation,

certifiers may revoke violating farmers’ certifications following the procedures
laid out in the regulation.
Following the IG Core coding, the referenced statement takes the

following form:
AƩributes: cerƟfiers
DeonƟc: may
Aim: revoke
Direct Object: cerƟficaƟons
Direct Object ProperƟes: violaƟng farmers’
AcƟvaƟon CondiƟon: When organic farmers violate organic farming provisions
ExecuƟon Constraint: following the procedures laid out in the regulaƟon
The IG Core-based coding shown above explicitly identifies the
responsible actor(s) (certifiers ), alongside activity (revoke) and the receiver
of the activity, i.e., the direct object (certifications ), alongside the explicit
characterization of associated properties (violating farmers’ ). In addition,
the conditions under which this discretion applies are specified (When
organic farmers violate organic farming provisions ), as well are qualifica-
tions for administering the activity in the form of execution constraints
(following the procedures laid out in the regulation).
While this statement captures the core components of the institutional
activity explicitly, it does not make accessible the complex structure that
conditions the applicability of the institutional statement, and leaves it to
the interpretation by the analyst, as opposed to the coder, to parse the

specific conditions under which a statement applies.
Turning to the activation condition (When organic farmers violate
organic farming provisions ) specifically, it reflects an embedded structure
that closely resembles the syntactic structure of regulative statements as
reflected in the following coding1 :
When ...
AƩributes: organic farmers
Aim: violate
Direct Object: organic farming provisions
AcƟvaƟon CondiƟon: N/A
ExecuƟon Constraint: N/A
Recalling the institution types discussed in Sect. 4.2.3, the basic form
of (regulative) institutional statements comprises of Attributes, Aim, and
Context components, syntactically reflecting a Strategy expression. Recon-
structing the comprehensive statement, the analyst can thus observe
multilevel parsing of statements that applies to both the statement as a
whole, and to components individually.2
5.1.2 Institutional Statements vs. Institutional States

While the syntactic pattern of the previous example makes the linkage
of statements and structures nested therein explicit, it is important to
recognize that structural information embedded in the activation condi-
tion does not necessarily follow such clear patterns, and may in fact not
necessarily represent an institutional statement, such as a strategy, in the
1 Note that in the coding visualization following, the absence of activation condition
and/or execution constraint is indicated with “N/A.” In this case (or where absent
entirely), the default semantics apply (i.e., ‘at all times’ for Activation Conditions; ‘no
constraints’ for Execution Constraints ).
2 In addition to the recognition of the structural compatibility of the components with
the regulative syntactic form, on semantic grounds the specific example highlights the
configurational relationship of both decomposed statements leading to the characterization
of the statement as a sanctioning provision, exemplifying how a purely syntactic interpre-
tation of institutional quality can lead to mischaracterization (based on the presumed
absence of the syntactic Or else), an aspect discussed in Chapter 3, Sect. 4.2.3. The
book will revisit this aspect and associated transformation rules to address this concern in
Sect. 6.1.4.
first place. Most notably, a strategy in the institutional sense is thought

of as being applied by multiple actors. Albeit coded as strategies in the
individual sense, the characterization of individual choices may not be
understood as institutions, but rather as instances of behavior regulated
by institutions. Such behavior thus may or may not generalize into, or
be linked to institutional expressions regulating the behavior of different
kinds. Introducing some nuance, if the activation condition read
Where organic farmers fail to meet organic farming provisions due to

external circumstances, …
the actors’ behavior may not necessarily be interpreted as a Strategy,

since the actor’s noncompliance may no longer be a strategic choice,
but rather a consequence or a result of circumstance outside an actor’s
control, reflecting the lack of behavioral alternatives or options. Instances
of behavior (as opposed to behavioral regularities), in which refer-
enced behavior may or may not reflect choice, or reflect environmental
circumstances,3 are referred to as Institutional States – contrasting the
Institutional Statement that explicitly regulates behavior, or parameter-
izes the institutional setting more generally. While the previous example
referenced actor-centric behavioral state, a common observation in insti-
tutional statements is the characterization of the environmental state that
conditions the applicability of an institutional statement with or without
the involvement of actors relevant to the institutional setting. Let us
exemplify such environmental states:
Where the number of certified organic farmers per region exceeds a speci-
fied threshold, certifiers may appoint additional inspectors to ensure sufficient
inspection coverage.
Here the Activation Condition makes explicit reference to observa-

tions that cannot be attributed to individual actors, but rather reflect
an institutional state. However, similarly to the characterization of
behavior oriented at the regulative statement syntax, one can observe a
complex syntactic pattern reflecting the constitutive syntactic form (see
Sect. 4.2.5), as exemplified below.
3 Here environment is understood in the wider institutional sense, with its interpretation
not restricted to the bio-physical perspective only.
Analog to the behavioral variant of institutional states, the state-

ment exemplified here reflects an institutionally relevant state of affairs
(i.e., captures a potential situational state of an institutional arrange-
ment) without broader parameterization of the institutional setting (i.e.,
it does not reflect how an institutional arrangement ought to be), as a
precondition for the activation of the leading institutional statement.
Structurally, this institutional state, as a precondition for the discre-
tionary revocation of certifications, nests on the activation condition of
the leading statement as visualized below:
Generalizing the concept of institutional states as an instance-centric

counterpart to the more general institutional statements, one can broadly
differentiate both concepts alongside their variable forms based on two
dimensions as shown in Fig. 5.1. While the function of institutional
statements is general in nature (i.e., to regulate and parameterize the
institutional setting), institutional state reflects specific situational char-
acterizations, states of affairs, or instances, of institutional arrangements,
Fig. 5.1 Institutional statements vs. institutional states
and primarily serve a descriptive purpose. As indicated above, the contex-

tual application of institutional state references is flexible; it can, for
example, refer to observed factual arrangements (e.g., specific instances
of violations), mental conceptions (e.g., beliefs, suspicions), desired states
(e.g., to reflect the motivation or purpose of a regulated activity) or
outcomes (e.g., achieved institutional state).
Reiterating an earlier aspect, in addition to the differentiation based on
scope or purpose, Institutional State conceptually mirrors the distinction
between behavior-centric regulative forms, following the Attributes-Aim-
Context (AIC) pattern (alongside optional Object characterizations), and
environment-centric constitutive forms, following the Constituted Entity-
Constitutive Function-Context (EFC) pattern (alongside the optional
Constituting Properties characterization). Despite this conceptual distinc-
tion, and as indicated in the context of the motivating examples, environ-
ment and behavior state characterizations may not necessarily be separated
but can be interlinked or combined in as far as they reflect the state of
affairs of the institutional arrangement more broadly.
Notably, and complementing the discussion around Hybrid Institu-
tional Statements (Sect. 5.2) more generally, institutional statements can
reference institutional states of either kind, i.e., constitutive and regu-
lative statements can link to behavioral state characterizations, as well
as environmental characterizations. The previous example showcases the
explicit reference of a regulative statement to preconditions capturing

environmental state.
This third form of structural nesting, the substitution of individual
statement components by institutional statements or institutional states
parsed in terms of syntactic components of the Institutional Grammar,
is referred to as Component-level Nesting. While the stylized example
showed a simple linear linkage of precondition and institutional state-
ment, where combinations of activity or state exist, those are reduced to
individual (atomic) state characterizations and linked following the prin-
ciples of horizontal nesting (Sect. 4.2.2) introduced in the context of
The examples to this stage exclusively applied component-level nesting
in the context of activation conditions. However, while commonly
observed in the context of activation conditions, it equally applies to other
components, including Attributes, Object variants, Execution Constraints,
as well as the constitutive counterparts, namely Constituted Entities,
Constituting Properties, alongside the Context component that equally
applies on regulative and constitutive side. In addition, nested institu-
tional statements can be found on all Properties associated with any
component.
The following statement, for instance, highlights the use of
component-level nesting on an Object component:
Inspectors must ensure that certified organic farmers report annually on their
agricultural practices.
While not necessarily immediately evident, the receiver of the action

regulated in this statement is the desired behavioral strategy “… certi-
fied organic farmers report annually on their agricultural practices ” in its
entirety, which can be, similar to the leading statement “Inspectors must
ensure …” parsed in regulative syntactic form.
Representing the component-level nesting on the Object in visual form:
While showing the variable applicability of component-level nesting,

the previous examples may still suggest the linearity or focal presence of
nested structure within a single component only. A final example will
display the concurrent presence across various components, as well as
applicability across variable levels:
The Program Manager may initiate revocation proceedings against a certi-

fied operation when the Program Manager has reason to believe that a
certified operation has violated the Act or when a certifying agent fails to
enforce the Act.
This statement decomposes into a leading institutional statement (“The

Program Manager may initiate …” ), alongside two conditions (“… when
the Program Manager …”; “when a certifying agent …” ) under which
the activity described in the leading institutional statement applies. One
of the conditions further contains a nested object characterization (“… a
certified operation has violated the Act …” ). This latter embedding is an
example of institutional state representations reflecting mental concepts
(here beliefs or suspicions).
Applying the principles of component-level nesting, i.e., the identi-
fication of state or statement structure in the syntactic form of institu-
tional statements, the analyst can reconstruct the interlinked institutional
features of the compound statement.4
4 Note that activation condition and execution constraint are not explicitly specified
unless they carry distinctive value. Where entries are omitted, the default semantics apply
(i.e., ‘at all times’ for Activation Conditions; ‘no constraints’ for Execution Constraints ).
This example highlights the complexity embedded in individual

components, potentially to an extent that the central institutional content
is in fact captured within the structures that the nested components
embed.5
With the introduction of component-level nesting and the differenti-
ated characterization of institutional state, and statements more generally,
IG Extended represents a central mechanism to extract deep institutional
structure beyond the coarse-grained characterization of components as
5 While concluding the introduction of component-level nesting with this statement,

note that these aspects will be revisited using variants of statements referenced in this
section, alongside strategies for their systematic encoding in Chapter 7.
emphasized in the context of IG Core. Deep parsing of institutional state-

ments, as represented here, motivates the potential of IG Extended to
preempt the omission of institutionally relevant features from analytic
treatment, while, at the same time, extending the robustness of the IG
against effects of linguistic construction of institutional statements, by
systematically identifying and extracting institutional patterns irrespective
of expression, responding to a central motivation for the dissociation of
linguistic and institutional treatment of institutional statements discussed
in Chapter 3.
Before turning to the discussion of approaches to guide the systematic
extraction of institutional content, Table 5.1 summarizes the different
forms of institutional statement nesting as introduced as part of the
IG 2.0 concept, alongside central criteria for their application, such as
the underlying principles, purpose, and affected components and logical
operators.
Table 5.1 Types of nesting in institutional statements
Nesting type Principle Purpose Affected components/operators
Vertical Nesting Linkage between Semantic Or else

statements or disambiguation
parts of of governance
statements relationships
relaying regulated
monitored activity
or
parameterization,
and consequence
of violation
Horizontal Combination of Fine-granular all components; operators:
Nesting atomic decomposition AND, OR, XOR, including
institutional of institutional potentially combined negation
statements using information (NOT)
logical operators
Component-Level Substitution of Extraction of Attributes, Object, Constituted
Nesting individual institutionally Entity, Constituting
components with relevant Properties, Context
complete structural
institutional complexity
statement/state embedded in
expression individual
components
A central contrast of the component-level nesting introduced in the

context of IG Extended, and the previously discussed horizontal and
vertical nesting forms is that the latter operates on statement level, i.e.,
linking distinctive statements in varying forms. Component-level nesting,
in contrast, reflects a substitution of individual components within state-
ments, thus affecting their structural integrity, an aspect schematically
captured in Fig. 5.2.
Combined, the different nesting concepts provide the basis for the
logical treatment of institutional statements based on transformation rules
discussed in Sect. 6.1.4.
5.1.3 Reconstructing Embedded Institutional Meaning

The examples introduced to this stage leverage analytically valuable6 insti-
tutional content based on the recognition of institutional patterns in
the form of activity (AIC) or state (EFC) characterizations. Despite their
intrinsic complexity, in all examples showcased throughout the previous
section these uniform patterns were overt.
However, not in all cases, do those present themselves in explicit form,
but are obscured by stylistic features, such as varying tense, enumera-
tion of relevant terms, conflation of activities or states. Inasmuch as a
linguistic grammar offers the flexibility to construct conceptual constella-
tions of arbitrary complexity, it often relies on stylistic forms that deviate
from idiomatic patterns by offering efficiencies in expression that do not
challenge the human interpretation of the underlying content. The IG,
in contrast, aims at extracting the structural complexity in its raw form,
so as to reduce the influence of language, both with respect to style and
form, as a confounding factor that may otherwise obscure institutionally
relevant details (see Chapter 3).
Structuring the insights developed to this stage, while, at the same
time, preparing a methodological basis to retrace concepts presented in
the following, IG Extended emphasizes a constructionist perspective on
institutional analysis made explicit in the following high-level blueprint of
the institutional statement parsing process discussed in the following.
Any parsing commences at the statement level, assessing linguistic
expression with respect to the corresponding institutional meaning by
6 Analytical opportunities will be discussed in Chapter 8.

Fig. 5.2 Schematic overview of nesting characteristics

extracting classes of statements (e.g., statements, states), the classification

of such statements based on observed syntactic form (e.g., constitutive,
regulative), followed by fine-grained parsing of their elementary struc-
ture (i.e., identification of components and combinations thereof). While
presented as a linear process, any classification may potentially involve iter-
ative refinement of the statement characterization based on structural and
semantic knowledge developed during the parsing process. For example,
a statement overtly presented in regulative form may expose its param-
eterizing function when interpreting individual components, and may in
turn be interpreted as constitutive in kind.
Based on the high-level, or first-order, characterization of the general
statement, the decomposition is sequentially applied to the individual
components in order to detect and classify substructures embedded
therein, such as demonstrated for activation conditions and objects in
the previous section. This general process, i.e., the sequential parsing of
statements (including the iterative refinement of their characterization),
followed by the review of clauses captured within individual components
(according to the same principles) is highlighted in Fig. 5.3. Central in
this process is the retention of the relationships among individual compo-
nents (as facilitated by the idiomatic structure of institutional statements
of regulative or constitutive kind), but more importantly, the linkage
between the first-order statements and the structure embedded in specific
components. This provides the basis for the analytical reconstruction of
the internal structure of institutional statements,7 implicitly motivated
by the representation of institutional statements as institutional state-
ment trees. The statement visualizations provided throughout the previous
and the following sections offer an intuitive representation of these very
structures.
As observable from the process, the encoding of institutional state-
ments requires an understanding of institutional structures as expressed
in terms of (a) institutional statements and states, (b) their linkage, and
(c) their positioning within an action situation. With this background,
IG Extended opens a principled approach to capture institutional content
not overtly expressed. To motivate this approach, let us introduce the
following example:
7 The dissociated treatment of the leading provision without a potential complex

activation condition would render the statement non-sensical.
Fig. 5.3 Institutional Statement Coding Process
When an inspection of an accredited certifying agent by the Program

Manager reveals any non-compliance with the Act, a written notification
of non-compliance shall be sent to the certifying agent.
Overtly, this statement can be decomposed into two parts, the first
being the leading conditional clause (“When an inspection of an accredited
certifying agent by the Program Manager reveals any non-compliance with
the Act” ) representing the activation condition for the second clause, the
main statement (“a written notification of non-compliance shall be sent to
the certifying agent.” ).
While components of the Institutional Grammar are observable, this
statement presents the challenge that neither clause reflects the conven-
tional regulative (ADIBC) or constitutive (EMFPC) syntactic form.
Turning initially to the latter main statement, the challenge specifi-
cally relates to the implicitly characterized responsible actor, the Attributes
component of the statement. Contextually, however – where context both
refers to the statement in the narrow sense, and the policy in the wider
sense –, the parsing resolves the Program Manager as such actor, given
its evident role in detecting any noncompliance in the first place. On this
basis, the statement can be complemented by identifying its basic syntactic
form as regulative – given its primary focus on the regulation of behavior.
A second commonly observed aspect is the expression of activities in
passive form (… be sent …), given the accredited certifying agent, the
subject of any administered sanction is the subject in the linguistic expres-
sion. Here the essential role lies in the reconstruction of the statement in
its complete form so as to express the regulative content.
Applying these basic operational principles8 – the inference of implicit
components, and the reconstruction of activities from the perspective of
the responsible actor regulated in this statement –, the parsing process
renders the statement shown below (with the inferred/transformed
component values italicized).
AƩributes: Program Manager
DeonƟc: shall
Aim: send
Direct Object: noƟce
Direct Object ProperƟes: wriƩen; of non-compliance
Indirect Object: cerƟfying agent
Indirect Object ProperƟes: to the accredited
Following the blueprint outlined in Fig. 5.3, this leaves the explo-
ration of the second statement part (“When an inspection of an accredited
certifying agent by the Program Manager reveals any non-compliance
with the Act” ), necessitating a more substantive reconstruction of the
institutionally relevant semantics embedded in this statement.
Evaluating the conditional clause based on its general structure, this
statement indicates an institutional state – circumstance(s) that leads to
the activation of the consequence expressed in the earlier statement, iden-
tifying this statement as an Activation Condition. Similar to the first
statement part, we observe a provision centered on the Program Manager
as the responsible actor observing a potential transgression. However,
while overtly concerned with the detection of a noncompliance by the
accredited certifying agent, upon closer review, we can identify that the
initial part of this statement signals a precondition – the inspection – for
the identification of the noncompliance in the first place. Interpreting
the statement from a logical perspective, and in the conventional form
8 We will discuss these and further aspects in the upcoming Chapter 7.

of regulative statements, the Program Manager thus not only reveals

a noncompliance, but also performs the inspection, de facto rendering
the statement with two interdependent activities – the inspection, and
the potential revelation of noncompliance. Reflecting the institutional
content accurately, while accommodating the atomic institutional state-
ment structure that seeks the decomposition into elementary unique
components (i.e., the AIC syntax, alongside optional components, for
regulative statements), the statement can thus be reconstructed as “When
Program Manager reveals any non-compliance with the Act [by the accred-
ited certifying agent] [under the precondition that] Program Manager
inspects accredited certifying agent …”. Paying closer attention to the
initial activity (revelation), we can further observe its decomposition based
on the characterization of noncompliance, that, following the institutional
statement syntax, accurately decomposes into a complex object character-
ization, expressed as “When Program Manager reveals [that] accredited
certifying agent does not comply with the Act…, where “accredited certi-
fying agent does not comply …” reflects the decomposed object that the
revelation applies to. The specific signal that is indicative for the activity
characterization is the reference to the responsible actor (“by the accredited
certifying agent” ).
Deviating from the original expression, we can thus observe the
decomposition of the compound conditional expression into two expres-
sions that capture the fact that any revelation on the part of the Program
Manager is contingent on the inspection in the first place, where the reve-
lation of noncompliance represents a complex activity relationship in itself.
Represented in schematic form, the overall statement – decomposed in
terms of atomic statement form – decomposes as follows:
The compound linguistic expression of the statement part thus renders

two dependent institutional state characterizations that themselves are
the precondition for any sanction administered as a consequence, one of
which carries complexity by resolving noncompliance as a complex activity
in its own right.
At first glance, this form of reconstruction requires the furthest rein-
terpretation of institutional statements on semantic grounds, and certainly
demands the principled interpretation of the institutional setting based on
its contextualization with and reconstruction within the action situation
in the form of activity (“someone does” ) or state (“something is” ) primitives.
This example illustrates, however, how far linguistic expression can deviate
from the interpretation of institutional configurations in terms of struc-
ture reflected in institutional statements. The reconstruction of statements
as promoted in the context of IG Extended finds theoretical support in
the field of cognitive linguistics, as initially motivated in Chapter 3.
In cognitive linguistics – a branch of linguistics that considers the
embodiment of language essential in production and its interpretation9 –
the observed phenomenon is referred to as an instance of conceptual
9 Contrasting the firm separation of syntax, semantics and morphology, cognitive

linguistics demands for the consideration of all linguistic aspects to extract, or reconstruct
the meaning of an expression (Broccias, 2013).
reification (Langacker, 2008), observed here in the rephrasing of activi-

ties (“inspecting”) in substantive form (“inspection”). More specifically,
in this instance, conceptual reification captures the representation of
institutional statements in uniform patterns that leverage their interpreta-
tion from an institutional perspective. More specifically, in this instance,
conceptual reification means representing institutional statement infor-
mation in terms of institutional primitives understood by actors that are
subject to the institution (or monitors of that institution). This principled
composition of institutional information from atomic patterns further
showcases how institutional statement information assembles configurally
to inform regulation of behavior or parameterization of the institutional
setting, (re)creating the complex structural interlinkages that real-world
institutions represent.
Reification of linguistic expression introduces efficiency benefits by
conflating expressions, for instance by avoiding the reiteration of the
subject-verb-object form for all activities, especially if they are contextu-
ally unambiguous (e.g., where the activity is performed by the same actor,
and potentially on the same subject). A second motivation is of stylistic
nature, specifically, to draw attention to essential activity. In the given
example, the revelation of noncompliance is of central concern, in contrast
to the implied inspection. While expressions relying on those abstractions
are accessible to the human reader and desirable to guide attention and
interest, the Institutional Grammar as an institutional language draws out
this tacit, but nevertheless important, institutional information to make it
accessible for structural and semantic evaluation.
Whereas sentence structure as found in natural language grammars
allows for open-ended combination of lexical building blocks to produce
novel linguistic constructs and meaning, the Institutional Grammar, as an
analytical tool, differs in that it selectively focuses on the analysis of struc-
ture of an institutional setting. Doing so, the IG assumes a constructionist
perspective, in that the concepts captured in an individual institutional
statement do not only seek integrity of the referenced components within
an institutional statement, but further consider the configurational linkage
of statement components (such as actor and objects) and their meaning
(e.g., the institutional function of the activity “revealing noncompliance”)
beyond the scope of individual statements by conceptually positioning
(and thereby integrating) all relevant entities in a shared ontological struc-
ture that the institutional setting represents, and that any statement can
reference and draw from.
Fig. 5.4 Institutional Statement Component Linkage to Action Situation
To this end, the Institutional Grammar specifies a set of well-defined

institutional structures reflecting the skeletal institution types, all the
components of which are linked to ontological concepts and relation-
ships – the fixtures of the action situation – present as part of the
institutional arrangement that these statements describe or relate to. By
combining these basic structures based on a set of predefined patterns,
the IG can reconstruct the complexity of the institutional configuration,
and thereby capture the essential institutional meaning that the original
linguistic statement conveys. Figure 5.4 schematically reflects the variable
linkage of the elementary components of different institutional statement
structures to the action situation, while, at the same time, reflecting the
complementary nature of constitutive and regulative statements as central
representational and analytical units.
Following the extended motivation for the reconstruction as part of
the coding process,10 this section concludes with the visual reconstruction
of the illustrative institutional statement in its entirety, reflecting the two
10 Operational details are revisited in Chapter 7.

levels of nested activation conditions (and the nested complex object)

on which the applicability of the first-order statement – the sending of a
notification as a sanctioning activity – rests.
5.1.4 Object-Property Hierarchy

As signaled in the previous section, the extraction of institutional content
based on deep interpretation and reconstruction of institutional meaning
captured in individual statements extends beyond a mere mapping of
linguistic concepts onto their institutional counterparts, but the recog-
nition of a semantic linkage between entities, activities, and contexts
referenced across different statements.
Aspects implicit to the human reader, but not overtly captured by

institutional statements following the IG Core features, are the concep-
tual relationships that relate individual entities referenced in institutional
statements. To motivate this challenge, we will use an illustrative example:
Program Manager shall send a written notification of proposed suspension or

revocation of certification to certified organic farmer.
Encoded in the structural form proposed above, we arrive at
While most elements of this statement find direct correspondence in

institutional statement form, notably, the Direct Object Properties stand
out based on the complexity they embed. The initial property offers
a qualitative characterization of the Direct Object by characterizing its
written nature. On the other hand, the properties capture the organi-
zational relationship to other institutionally relevant concepts, including
the suspension and revocation in relationship to the certification, nuanced
structural relationships not captured in the syntactic elements of institu-
tional statements.11 While these concepts are only indirectly referenced,
they nevertheless expose conceptual features of the institutional setting,
alongside their relational linkages.
The statement, for instance, references the terms suspension and revo-
cation, both of which are linked to the certification, and more specifically,
both concepts are functionally dependent on the existence of the certifica-
tion; administering a suspension without an explicit or implied reference
to the status that is withdrawn or otherwise affected (here: the certifica-
tion) reflects an incomplete information basis. The entity referenced as
Direct Object in the exemplary institutional statement itself (notification)
likewise depends on the existence of the suspension or revocation, which
11 The implied logical linkage has been emphasized in the example above.
are inadvertently linked to the certification. Properties, in their simplest

form, can assume a qualifying function directly associated with the quali-
fied object. Alternatively, and capturing complex structural relationships,
the Institutional Grammar organizes the existence of principal forms of
relationships of entities in an Object-Property Hierarchy. Following prin-
ciples of conceptual ontologies (Agazzi, 2011; Campbell, 2005), entities
can consist of, or associate with any number of properties, where the latter
can either be objects themselves, and similarly signal relationships to or
dependence on other entities. Relationships can be of variable quality,
signaling both the independence of objects or, as indicated above, the
ontological dependence between entities (e.g., between suspension and
certification). Such hierarchy does not need to be rooted in an institu-
tional statement, but the entities referenced in an institutional statement
may anchor to this hierarchical structure on any entity (or multiple where
relevant).
Recontextualizing the discussion provided above with the initial
example, Fig. 5.5 offers an overview of the Object-Property Hierarchy
inferred from the Direct Object Properties component of the statement
referenced above that anchors on the notification as a direct object.
Fig. 5.5 Exemplified linkage of objects and properties

The value of making conceptual relationships overt is to leverage

features of institutional configurations that may or may not be explicitly
based on the specification in institutional statements alone. The Object-
Property Hierarchy plays across-cutting function in capturing institutional
structure comprehensively that operates across and, to some extent, inde-
pendent of institutional statements, while being anchored in statement
components. In addition to extracting tacit institutional information, the
lateral operation of the Object-Property Hierarchy compensates for short-
comings in policy specifications (e.g., missing policy information, tacit
context).
Given the general function of the Object-Property Hierarchy in the
context of institutional statements, structural information can be provided
as part of any property characterization across institutional statements.
While exemplified here for objects, structural complexity can occur in
various other components of the IG, namely Attributes Properties, Direct
Object Properties, Indirect Object Properties, Constituted Entity Proper-
ties, Constituting Properties Properties, but also as part of Activation
Conditions and Execution Constraints.
The principles of the Object-Property Hierarchy, and, in extension,
properties as prevalent across various statement components, can be
summarized as follows:
• Objects can have one or more properties.

• Properties can be simple or complex in kind.
– Simple properties offer qualifications of the associated objects.
– Properties themselves can be objects (and consequently possess
their own properties).
• Objects and their properties are linked by relationships of different
quality, most notably indicating functional, or ontological, depen-
dence (i.e., the existence of one concept depends on the existing of
the other), or independence.12
• Properties of objects (as objects themselves) can be logically linked
(e.g., signaling combined or alternative applicability13 ).
12 In the context of the Cognitive Grammar (Broccias, 2013; Langacker, 2008),

these variable relationships are characterized as conceptually dependent and conceptually
autonomous, respectively.
13 This aspect is illustrated in the referenced example.
5.1.5 Property Types

Building on the observed relationships between entities and their prop-
erties as captured in the Object-Property Hierarchy, and their flexible
association with institutional statements across various components, closer
observation of empirically encountered properties allows for a refined
classification that supports different analytical use cases. In addition to
recognizing the distinction between simple qualifications and complex
structural relationships, properties further carry quantitative information,
whether explicitly expressed in numeric or linguistic form (e.g., 1, one), or
as quantifiers (e.g., each, some, all, most ). Those features specifically, are
valuable from an evaluative perspective, e.g., to make the scope of applica-
bility of a given instruction or provision explicit, and to allow the explicit
assessment of regulatory compliance, and thus contrast the qualitative
features highlighted earlier.
Reviewing structural aspects of qualitative property characterizations,
we further observe a differentiation between simple and complex prop-
erties. Complex properties can afford a circumstantial qualification of
entities based on references to activity or state information. Exemplifying
the latter two aspects, an activity-centric property can pertain to activity
the entities engage in (e.g., “Participants engaging in non-compliant prac-
tice …” ), as well as the reference to internal or mental processes, such as
beliefs (e.g., “Inspectors who believe that operations are in violation …” ).
State-centric properties relate to externally attributed properties and can
carry status implications. An example for this is the statement “Partici-
pants without valid certification …”. While, in this instance, an external
entity (certification) is referenced, it primarily acts as a descriptor of the
associated actor (Participants ), as opposed to describing relational prop-
erties that functionally link these entities as introduced in the earlier
discussion of the Object-Property Hierarchy. Complex properties can
also describe the positioning of Objects within the relevant institutional
setting; for example, by situating those in relation to other institutional
statements. For example, “certification granted according to procedures
set forth in section A of policy X.”
Syntactically, the activity- and state-centric properties draw on the
Institutional State characterization (discussed in Sect. 5.1.2) using
either activity-centric regulative patterns, or the state-centric constitutive
patterns, thus conceptually integrating selected property types with the
fundamental syntactic patterns of institutional statements.
Figure 5.6 provides an overview of the proposed high-level character-

ization of properties as observed in select components (and properties
of components more generally). Naturally, observations that generalize
within or across particular domains, or signal features of specific relevance
for relevant analytical objectives, may naturally invite for the extension
or customization of this base typology in order to extract institutionally
relevant information embedded in individual components.
The mechanisms introduced to this stage provide the basis for the
deep structural parsing of institutionally relevant details embedded in the
institutional statements themselves, involving both the tacit expression of
activity leveraged by component-level nesting, as well as the explication
of conceptual relationships of referenced entities (Object-Property Hier-
archy), alongside a general characterization of properties with respect to
their function in relation to the entity (e.g., actor, object) they qualify
(Property Typology). Naturally, the structural decomposition introduced
here is general, and leaves opportunities for further refinement, including
accommodations that reflect source characteristics (e.g., policy docu-
ment, interview) – and implicitly language-specific aspects – or features
particular to analyzed domains.
Fig. 5.6 Property Typology

5.1.6 Context Taxonomy

Complementing features that focus on the institutional content internal
to institutional statements, an aspect that has not been reflected at this
stage are potential external linkages and contextual embedding. Insti-
tutional statements do not exist in isolation, but are integral to the
comprehensive characterization of an action situation and its embedding
in the wider institutional setting. To this end, IG Extended seeks the
contextual linkage of institutional statements with environmental features.
This specific aspect builds on the basic characterization offered in the
original IG (Crawford & Ostrom, 1995) that suggested the labeling of
content captured in the Conditions component as temporal, spatial, or
procedural in kind, essentially highlighting a basic contextual embedding.
Prevalent coding practice (e.g., Brady et al., 2018) relies on an extended
set of additional features commonly observed in institutional statements,
such as the “how” qualification of conditions.
As alluded to in Chapter 3, the Context component (previously Condi-
tions component) takes a diverse and important role in an institutional
statement by:
• specifying the conditions under which a particular activity takes

place (Activation Condition), as well as the qualification of enacted
activities (Execution Constraint ),
• potentially capturing those preconditions or qualifications in a form
that resembles institutional statement structure itself – reflecting
inter-statement linkages resolved via component-level nesting –, and
finally,
• providing the contextual anchoring of an institutional statement
within and beyond the action situation the statement associates with.
While the first two aspects have been introduced and motivated
throughout this chapter, a remaining aspect is the contextual embedding.
Drawing on empirical observation of diverse context characterizations in
policy statements, alongside the patterns recognized in established coding
practice, as well as the characterization of supersenses observed in human
language (e.g., Schneider et al., 2016, 2018), the IG includes a Context
Taxonomy that organizes context characterizations along general cate-
gories, allowing both for a general and fine-grained characterization of
contextual embedding.
We recognize four central categories along which context characteriza-

tions are organized. Those include:
• Substantive Context
• Procedural Context
• Aspirational Context
• Situational Context
5.1.6.1 Substantive Context

Context characterizations associated with Substantive Context relate to
contextual aspects pertaining to the wider institutional setting and its
metaphysical embedding, including the temporal, spatial, and, in addi-
tion, a domanial dimension. Temporal and spatial dimensions maintain
the same purpose as in the original Institutional Grammar, but are further
augmented with more specific characterizations of either kind. Whereas
temporal context characterizations reference points in time (e.g., at 8am),
time frames (e.g., from 9am to 5 pm), or alternatively, frequencies (e.g.,
annually), spatial characterizations, for instance, specialize into location
(e.g., at the town hall ), direction (e.g., on the way home), and path (e.g.,
through the town).
Reflecting the relevance of and frequent reference to domain-specific
aspects, the substantive context characterizations include different forms
of realms that variably capture contextual linkages captured in specific
components. Specifically, those may pertain to specific activities in the
form of activity realms (e.g., during accreditation), topical realms that
capture the theme of the domain (e.g., for drinking water), and poten-
tial forms of existential realms that reference features of the entities
embedded in the setting (e.g., during adolescence).
Context associations of these types capture general dimensions of
embedding of institutional statements or action situations in a wider insti-
tutional context (e.g., policy, domain), alongside the overarching general
temporal and spatial dimensions.
5.1.6.2 Procedural Context

Institutional design frequently references implied or explicit procedural
characterizations, where explicit may involve the explicit specification
of execution order (e.g., “first …, second …, …” ), a reference to
corresponding instructions (e.g., “… as specified in section 5(2)” ), or
methodical guidance that reflects essential information of the institutional

setting.
In addition to the characterization of procedural aspects as the first
category of the Procedural Context characterization, a second category
focuses on the method by which activities or processual aspects are satis-
fied. This instrumental orientation divides between means, or instruments
used to perform a particular activity (e.g., by car), and actor behavior
characterized as manners (e.g., by handshake).
5.1.6.3 Aspirational Context

Besides the common characterizations captured in the substantive and
procedural contextual embeddings, a third dimension is the aspirational
context that focuses on the motivation and objectives underlying a partic-
ular regulation or provision (e.g., in order to; for the purpose of ), and
hence labeled Purpose. Context classifications of this nature have diverse
uses, including the justification of regulated behavior specifically, or
signaling the objectives of the provision and thus providing the basis
for the interpretation of policies in their entirety (e.g., as constitutive
strategies discussed in Sect. 4.2.6).
Mirroring the aspirational perspective of the Purpose category, a
second, and to some extent complementary, category is the Effect. This
subcategory indicates, for instance, the satisfaction of an expectation set
out in previous statements (e.g., “If pollution thresholds are met, …” ).
5.1.6.4 Situational Context

The last context category is the Situational Context, which shifts the
perspective of statement context away from categories that reference
environmental universals or instrumental aspects. Complementing the
reference to motivational aspects commonplace in policy, the situational
context aims at capturing the state of affairs and distinct changes or transi-
tions specific institutional statements are linked to, with specific categories
including States and Events.14
States, as understood here, characterize specific constellations of
circumstances (states of affairs) of relevance for the institutional setting
14 The conception developed here builds on Mourelatos (1978)’s characterization,

where “Processes,” as a complementary classification to “Events,” are reflected in the
activities of the Aim and Constitutive Function components, respectively.
(e.g., as a precondition for an activity or change in institutional configura-

tion). These can be external or internal to the institutional statement, i.e.,
referring to the action situation or institutional setting more generally, or
variably referring to entities referenced in the institutional statement.15
Events, in contrast, reflect instantaneous situational occurrences
signaling a change in the state of affairs, including the instantiation of
novel entities, as well as indicating changes in their properties. Events can
thereby act as state transition markers both relevant for constitutive and
regulative statements.
As a central distinction, States persist over time spans, whereas Events
are inherently instantaneous in nature.
As suggested as part of the characterization, the context categories
offered as part of the situational context are general in nature, and aim at
capturing circumstances and changes therein on any scale (micro, macro),
and may optionally combine with any other context characterization of
the different categories (e.g., substantive, procedural, aspirational).
5.1.6.5 Integrating the Context Taxonomy

The context characterizations, or rather classifications, offered in the
context of IG Extended aim at reflecting the diverse and complex nature
of statement context, and the variable forms in which institutional state-
ments can be embedded in an action situation or wider institutional,
let alone bio-physical, setting. Such embedding is oftentimes complex;
a given statement can potentially be contextualized in multiple ways; i.e.,
statements can make reference to multiple forms of context.
The context characterizations proposed as part of the Context
Taxonomy apply to any C ontext component, especially since Execution
Constraints assume the primary role in qualifying (regulated) behavior
and (constitutive) functions. However, where contextual aspects are refer-
enced within Properties in particular (e.g., to offer a characterization
of the associated entity, e.g., “farmers operating West of the mountain
range” ), the taxonomy can equally be used to establish the corresponding
contextual characterization. Another component that may selectively draw
on the contextual characterization is the Object, Constituted Entity, and
Constituting Properties components (e.g., a state or event, or location
referenced in an Object component, or defined as a Constituted Entity).
15 It is noteworthy to state that states are different from facts in that states do not
necessarily signal truth in the epistemological sense.
Fig. 5.7 Context taxonomy
To synthesize and illustrate the use of the context types introduced

in this section, Fig. 5.7 provides an overview of all referenced types
and exemplifies their application for Activation Conditions and Execution
Constraints, reflecting the most common use cases for annotations based
on the Context Taxonomy. A detailed characterization of all categories
introduced here from a high-level perspective is provided as part of the
IG 2.0 Codebook (Frantz & Siddiki, 2020).
Summarizing the principles, the Context Taxonomy groups context
characterizations into categories that variably represent general reference
frames of the statement, including the substantive context, procedural,
and aspirational types. Contrasting these general anchor points, the assess-
ment of state transitions and situational activities on the micro level are
captured by situational context characterizations.
Specifically, this latter characterization offers a classification that may

of course apply conjointly with other context categories proposed in
this subsection. However, this particular category carries specific value
for the assessment of institutional dynamics based on state changes
(e.g., as potential signifiers for transitions between action situations) as
mediated by events (e.g., as a proxy for instances of enforcement interven-
tions or collective action), and can do so detached from domain-specific
information more closely associated with the other context categories.
The other two categories, notably, the characterization of methodical
aspects reflects the focus on specific behavior forms (i.e., the means by
which an activity is performed) as subject to regulation in the first place,
making the explicit identification of such qualifications important. Aspi-
rations aspects, such as the purpose, function, or anticipated effects of
regulation provide an evaluative basis relevant for litigation (i.e., whether
aspirations are sensibly addressed), as well as assessment of success of
a given provision (e.g., whether the objectives have been met), hence
rendering it, similar to the methodical characterization, as an important
feature in the assessment of institutional design.
This selection of categories discussed and exemplified is not exhaustive,
but rather includes a general set of dimensions as a basis for insti-
tutional analysis, striking a balance among characterizations of general
kind (substantive), empirically observed ones (procedural context, aspi-
rations) and ones capturing dynamics that context reflects (situational).
Where domain-specific aspects apply, the expansion and refinement of the
taxonomy for, or beyond, the specific application is relevant. Similarly,
and as relevant for specific analytical objectives, the analyst may draw on
these categories selectively (e.g., only capture situational context char-
acterizations) and at the level of general categories or specific context
types.
The Context Taxonomy offers a link between institutional statement
and its wider context in the first place, but also reflects the starting point
for shifting from a primarily structural perspective on institutional state-
ments and the entities characterized therein, to a semantic perspective
that focuses on the meaning embedded in institutional statements and
their systemic linkages, a perspective strengthened further in the context
of IG Logico in Chapter 6.
5.2 Hybrid Institutional Statements

The introduction of institutional statements as developed in this chapter
has primarily focused on the refinement of specificity across different
levels of expressiveness. A central novelty associated with the refined Insti-
tutional Grammar is the treatment of both regulative and constitutive
statements (with an initial introduction in Sect. 4.1.2), establishing the
Grammar’s ability to code policy comprehensively.
Throughout the discussion of the Institutional Grammar, this chapter
maintained the distinctive nature of both forms, in which constitutive
statements parameterize an institutional setting and expose the entities
and affordances on which the conceptual establishment of regulation, i.e.,
provisions in the form of obligations, permissions, etc., rests. In this ideal-
ized form, constitutive statements provide the fixture on which regulative
statements anchor, linking different actors, objects, artifacts operationally.
Empirically, the analyst will however recognize that the identification
of statements as either regulative or constitutive is not at all times
unambiguous. At the same time, the theoretical discourse invites for a
reconciliation that recognizes the opportunities for a distinctive treatment
of both statement forms, as well as its limitations. The IG harmonizes this
cleavage by introducing the concept of Hybrid Institutional Statements;
however, not before providing an overview of theoretical discourse and
empirical observations motivating this proposal.
5.2.1 Revisiting Constitutive Rules in Literature

While the conceptually separated presentation of constitutive and regu-
lative statements is useful to guide and develop a discrete understanding
of either form, any such introduction cannot ignore the ongoing debate
on the nature of constitutive and regulative rules (as referenced in the
literature). Building on earlier works of analytical philosophers such as
Wittgenstein (1983) and Rawls (1955), Searle (1969) proposed the
distinctive characterization of rules as rules that are either constitutive or
regulative in kind. Regulative rules, Searle posits, “regulate antecedently
or independently existing forms of behaviour […].” Constitutive rules
then “do not merely regulate, they create or define new forms of
behaviour” (Searle, 1969, p. 33). The essential aspect of his distinc-
tive conception is the separation and independent operation of rules of
constitutive and regulative kinds.
Objections to this distinctive characterization between the two types

of rules were raised on various grounds. Notably, Raz (2012) offered
a substantive critique based on the implied dependency between consti-
tutive and regulative rules, given that the parameterization necessarily
precedes the regulation, and thus implies that any regulative rule has a
constitutive correspondent that defines the behavior in the first place.
Given the functional interdependence of both representations, any sepa-
rated treatment then becomes artificial, challenging the validity of the
distinction and its analytical value. Variations of the concerns expressed
by Raz are mirrored in observations made by other researchers, however,
varying by level of rejection or reconceptualization. Giddens (1984), for
instance, as one of the leaner accounts, rejects the existence of consti-
tutive rules in their entirety given that any regulative statement de facto
assumes a descriptive function with respect to the underlying institutional
concepts. As Giddens (1984) puts forth, any rule prescribing employ-
ees’ working hours, for instance, makes implicit reference to fundamental
time governance principles associated with industrial societies (Giddens,
1984).16
A group of scholars that engage constructively with the concept of
constitutive rules are Ransdell (1971) and Cherry (1973). In contrast to
the conflating position held by Giddens, central concerns center around
the lack of distinctiveness associated with different forms of constitutive
rules. Cherry (1973) observes the differentiation into rules that intro-
duce brute affairs (aspects characterized earlier as brute (f)acts) into the
institutional setting in the form of institutional affairs in the first place,
and rules that draw on institutional (f)acts to constitute novel institu-
tional (f)acts. Drawing on the example of marriage as an institution, its
initiation relies on underlying brute rituals such as exchange of vows and
practices (such as passing of artifacts or performing ceremonies). The act
of divorce, in contrast, will inevitably rely on a pre-existent marriage (as
an institutional concept) in the first place, hence making reference to the
second kind of constitutive rules offered by Cherry.
Ransdell (1971), likewise offering a constructive account, specifically
suggests the avoidance of an explicit reference to an institutional (f)act,
the intermediate “Y” in Searle’s “X is Y in Context Z” characteriza-
tion. Eliminating the representational proxy, Ransdell suggests, avoids
16 Warnock (2020) assumes a similar position as Giddens.

conflated semantics, and instead produces distinctive rules that selectively

focus on connotation – describing institutional affairs as embedded in
context (e.g., intensional definitions) – and import – the pragmatics of
usage associated with an institutional construct (e.g., operational function
and interaction with other features in an institutional setting).
More recent critiques are offered by Ruben (1997) and Hindriks
(2009, 2013, 2015), centrally focusing on discounting the distinction
between regulative and constitutive rules as a mere linguistic artifact,
given the presumed lack of ontological validity. Targeting Searle’s concep-
tualization specifically based on the inconsistent characterization that
language plays in the interpretation of rules, Hindriks implies that any
rule can attain both constitutive and regulative function, irrespective of
linguistic form. In doing so, Hindriks nevertheless implicitly recognizes a
form of constitutive rules for the sole purpose of conveying status, strati-
fied into power-specification and power-conferring rules.17 With this status
account of constitutive rules, Hindriks (2009) provides a constructive
basis that recognizes a principal, albeit narrower, ontological distinction
between rules that assume regulative and constitutive function.
Summarizing essential aspects, critique rests on the strong linguistic
account offered in Searle’s conceptualization of constitutive rules (see
e.g., Hindriks, 2009), the variably narrow characterization of constitutive
rules (see e.g., Hage, 2018), or insufficient specificity (see e.g., Cherry,
1973; Ransdell, 1971). This short contextualization of extant concerns
around the distinction of regulative and constitutive rules underpins the
challenges underlying the characterization of statements as regulative or
constitutive (let alone the acceptance of this distinction in the first place),
but, at the same time, provides the backdrop for the integrative treatment
of both forms proposed in the following.
5.2.2 Integrating Constitutive and Regulative Statements

The presented challenges offer the basis for a clarification, and provide
retroactive justification for design choices associated with the constitutive
syntactic form introduced in Sect. 4.2.5, and implicitly, a basis for recon-
ciling concerns in extant debates. To this end, this section provides a basic
17 Naturally, these critiques have been met with rebuttals by Searle (2015).
discussion of essential features and offers a contextualization based on the

positions discussed in literature.
Reflecting on the constitutive form introduced in the IG, a central
aspect that, on the one hand, widens the applicability of constitutive
statements, while, at the same time, provides the basis for heuristic guid-
ance, is the abstract functional linkage between Constituted Entities and
Constituting Properties in the form of the Constitutive Function. Unlike
Searle (1969) and Grossi et al. (2008)’s characterization of the linkage
as “counts as,” the specific semantics of the function in a specific insti-
tutional statement depends on the effect the linkage plays in the context
of the institutional statement. The ability to distinguish between different
constitutive statements, and more importantly, to characterize their opera-
tional role in the institutional setting, the IG as presented here introduces
a categorization of Constitutive Functions based on prototypical linkages
captured in the Constitutive Functions Taxonomy (see Sect. 6.1.2.4).
The general distinction between constitutive and regulative statements
in the IG rests on empirically observed structural and semantic forms
prototypical of either type of statement. Absent the characterization of
a responsible actor (and the inability to infer such), the introduction of a
novel entity into an institutional setting, e.g., by defining it, reflects the
constitutive character of the statement based on form and effect on the
institution setting, as exemplified in the statement “A majority of Council
members constitutes a quorum”, introducing the concept quorum into the
setting.18
ConsƟtuted EnƟty: a quorum
ConsƟtuƟve FuncƟon: is (inferred)
ConsƟtuƟng ProperƟes: a majority
ConsƟtuƟng ProperƟes ProperƟes: of Council members
This exists in contrast to the prototypical regulative form “Council
members must participate in Annual General Meeting” that offers a clear
assignment of obligations, and is schematically represented as follows.
AƩributes: council members
DeonƟc: must
Aim: parƟcipate
Direct Object: in Annual General MeeƟng
18 This example shows properties hierarchically nested across multiple levels.

Another specific aspect of Searle’s definition that is inconsistent

with empirical observation is the assumption that constitutive rules
antecedently define or otherwise introduce entities into an institutional
setting, which are then subsequently referenced by regulative statements.
Prima facie, legal documents appear organized in ways that separate the
parameterization of the institutional setting (e.g., as part of a ‘Defini-
tions’ section) from sections that group statements of regulative kind,
potentially aligned with action situation specifications (è.g., ‘Conflict
Resolution’). In practice, however, the linkage between statements of
regulative and constitutive kind in a document itself may at times be less
distinct, such as exemplified in the following stylized example:
“Certified organic farmers must submit a organic handling system plan,

which describes associated operational practices, including a list of substances
utilized during production, as well as monitoring practices.”
This statement specifies obligations associated with an actor, where

the artifact of relevance (here: the organic handling system plan) is
defined in situ, thus giving the statement both regulative and consti-
tutive character. Seeking conceptual clarity, such provision can naturally
be decomposed into distinctive statements as follows (applying both the
nesting of the Constituted Entity ‘organic handling system plan’ on the
Direct Object, as well as reflecting the composite nature of the constitutive
statement):
Naturally, and implied in the decomposition above, the statement can

be decomposed into two distinctive statements of different type, show-
casing an implicit inter-statement linkage of institutional statements (as
opposed to the explicit linkage afforded by Statement References (see
Sect. 6.1.3)).
Returning to the original critique, recognizing the linkage, but also
the distinctive structural institutional patterns, the entanglement can be
systematically resolved under consideration of the structural linkage. At
the same time, the ontological distinction between regulative and consti-
tutive statements can be retained. As a second aspect, this example
illustrates that the specification of entities and their reference (as part of
regulative statements) does not rely on the antecedent specification of
behavior in constitutive statements prior to its invocation in regulative
terms as originally suggested by Searle.
The IG 2.0 refers to statements that variably link regulative and consti-
tutive statements as Hybrid Institutional Statements , with the statement
shown above referenced as regulative-constitutive statement, reflecting the
order in which the specific statement is linked (i.e., a leading regulative
statements, on which the constitutive statement nests). The extent of
these linkages can be arbitrarily deep (e.g., higher-order statements can
nest on preceding nested statements).
Exemplifying the converse form of hybrid statements in constitutive-
regulative configuration, the following example provides an illustration:
“A non-compliance notification must include detailed specification of trans-

gressions, or else the non-compliance notification is invalid and affected
farmers may appeal the associated sanctions.”
Absent further contextual qualification of the responsible actor (e.g.,

in associated statements), this example is constitutive in kind since it
outlines the required elements of a noncompliance notification (in this
interpretation “must” signals the necessity that a notification contains
the specification of transgressions). While implicitly obliging an associ-
ated actor (an aspect potentially captured in a complementary regulative
statement), the statement per se provides an extensional specification of
the artifact ‘non-compliance notification.’ Drawing a linkage to the exis-
tential nature of consequences associated with constitutive statements, if
the requirement laid out in the monitored statement is not satisfied, the
non-compliance notification is invalid, i.e., does not have required status
to produce the intended institutional effect as an enforcement mecha-
nism. In addition to the existential consequence (which is characterized
as constitutive in its own right), and where interpreted as an explicitly
linked consequence of the invalidated notification, the non-fulfillment can
further provide the basis for associated behavioral consequences – here the
possible appeal or rejection of sanction on the part of the affected entity.
Decomposed into the corresponding linked institutional structures, the
statement composes as follows:
Working toward reconciling the varying perspectives on regulative and

constitutive statements as discussed above, the IG is able to address
selected concerns by explicitly recognizing the complementary function
of both statement types, and offering a structural integration based on
their systematic linkage both on component and statement level. Specif-
ically, the first example highlights how the definition of a novel entity
in constitutive form is interleaved with its referencing from a regulative
perspective.
The exemplified compositional approach offered by the IG enables
the dissociated treatment of statements for analytical purposes (i.e., the
selective focus on regulative or constitutive aspects only) while retaining
the structural integrity of the linguistic expression (i.e., the interpreta-
tion based on the original construction). With a focus on the operational
perspective (i.e., encoding institutional statement for the purpose of
analyzing institutional design), this representation is able to capture
complex statement constructions as observed empirically, affording the

linkage between the form in which institutional content is structured and
the representation of its meaning.
However, the linkage of statement structures of different kinds, as
introduced to this stage, does not resolve all challenges related to the
distinction between constitutive and regulative statements. Inasmuch as
the decomposition introduced above captures distinctive semantic aspects
associated with the regulative and constitutive forms, the interpretation
of institutional statements is further framed by considerations of interpre-
tational scope and contextual knowledge, as well as analytical preferences.
This specifically applies if statements offer limited structural distinctive-
ness, or are in fact amenable to trivial reconstruction that affects the
characterization as regulative and constitutive respectively.
The following example highlights such as variable interpretation basis:
“The functions of the Board shall be: (a) to implement the decisions of the
Health Assembly; (b) to perform any other functions entrusted to it by the
Health Assembly.”
Interpreted as parameterizing, the statement indicates the function of

the board, defining behavior associated with the “Board.” Borrowing the
constitutive structure, and with reference to the divergent Constituting
Properties component, the statement can thus be decomposed as:19
19 Note that this visualization, as the previous, favors the explicit decomposition into
atomic institutional statements; Chapter 7 introduces a syntactic form that concisely
captures the complexity embedded in the institutional statement.
Alternatively, and recognizing the specific activities assigned to the

entity, the analyst may intend to resolve the semantic linkage in regulative
form and encode the statement based on an operational interpretation,
resulting in the following decomposition:
Juxtaposing both forms, the reader may observe the systematic recon-
struction in both forms, and observe the variable focus the reinterpre-
tation introduces. An example is the representation of the Board that is
variably referenced in the context of the Constituted Entity “functions”
for the constitutive interpretation, or alternatively as the responsible
actor associated with the representation in regulative form. Similarly,
whereas the Constitutive Function emphasizes existential qualities for the
first statement variants, the second set ignores this characterization and
focuses on the operational activity that composes the “function” in the
constitutive statement.
This variable interpretation highlights operational challenges with the
distinctive characterization of constitutive and regulative statements.20
However, inasmuch as this dual characterization exposes threats for a reli-
able classification and encoding of institutional statements, and thereby
requires methodological affordances addressed in Chapter 7, conceptu-
ally it mirrors the variable interpretations of statements observed in extant
literature and the introduction of constitutive statements provided in
Sect. 4.1.2. Constitutive and regulative statements act complementarily,
where constitutive statements ontologically (or rather institutionally)
establish concepts, and regulative statements operationalize those. In
selected instances, such as the one exemplified, this duality in purpose
invites for the concurrent interpretation as both definitional and opera-
tional. However, instead of disqualifying any such ontological distinction
on this basis and challenging its operational value, it exposes the central
role the analytical perspective assume. Where the analysis primarily focuses
on features of the action situation, or institutional setting, such as the
nature and organization of actors, as well as environmental characteris-
tics, the interpretation may be biased toward a constitutive interpretation.
Where, in contrast, behavioral regulation and assessment of compliance is
of primary concern, the analyst may favor a regulative interpretation.
Given the IG’s intent to both support the encoding of institutional
statement both for circumstances where analytical objectives are well
defined, as well as the encoding of statement for general purposes, i.e.,
agnostic of specific application cases, such as for the creation of generic
datasets of institutional information, the IG supports the dual encoding of
20 A commonplace indicator for the variable interpretation, as in this case, is the

presence of a responsible actor endowed with activities referenced in the Constituting
Properties.
statements in both syntactic forms, rendering such statements polymorphic

institutional statements .
The introduction of the dual form reflects the potential dissociation of
parsing or encoding of institutional information and the ensuing analytical
uses, where the latter may selectively draw on either (or both) encoded
variants.
As an alternative to the analytical perspective as the basis for the vari-
able characterization of statements – here resolved by affording the dual
annotation – the IG highlights the variable classification of statements
based on the contextual knowledge and scoping of the interpretation by
the coder and analyst.
Another form of institutional statement commonly found in policy
documents are what could be referred to as “blanket statements,”
statements that reference and link other statements, e.g., to introduce
exceptions or impose limitations on more general provisions. Exempli-
fying such case, the following example showcases a listing of individual
statement, and a final statement indicating exceptions for the applicability
of all preceding statements:
(1) Organic farming operations must not utilize genetically modified

seeds.
(2) Organic farming operations may not process crops other than the
ones specified in Appendix A.
…
(10) Paragraphs in this section are not applicable to traces of genetically
modified material.
As observed based on this example, the final statement assumes a

parameterizing function based on its modification of existing statements
(i.e., the preceding statements). Naturally, the analyst is tasked to iden-
tify such statement as either type. Understood narrowly by itself, i.e.,
without consideration of contextual information (including the implicitly
referenced statements), the statement primarily assumes a parameterizing
function, since its effect is the modification of existing statements, and
thus implicitly the institutional setting that the existing statements define.
Taken in isolation, the statement would thus compose as follows21 :

ConsƟtuted EnƟty: Paragraphs
ConsƟtuted EnƟty ProperƟes: in this secƟon
ConsƟtuƟve FuncƟon: are not applicable
ExecuƟon Constraint: to traces of geneƟcally modified material
Understood broadly, i.e., under resolution of the semantic linkage to

the affected statements, an alternative encoding practice is the encoding
based on the target statement composition, rendering the following
structure:
AƩributes: Organic farming operaƟons
DeonƟc: may not
Aim: apply
Direct Object: paragraphs
Direct Object ProperƟes: in this secƟon
ExecuƟon Constraint: to traces of geneƟcally modified material
Applying a wider contextual consideration, the coder resolves the
embedded linkage to the referenced statement, and by doing so, super-
imposes the target structure on the encoded statement, which, in this
instance, renders the statement regulative.22
Similar to the previous case, such statement can thus be encoded as
polymorphic. In contrast to the previous case that emphasizes diversity in
analytical objectives as a source of classification preferences, the latter case
can be specified with greater consistency based on an agreed-upon scope of
interpretation applied in the coding process. Any scope of interpretation
can either be determined as wide or narrow, where wide interpretational
scope presumes the resolution of semantic linkages to other statements
embedded in the statement of concern, invariably extending the unit of
21 Note the characterization of traces of genetically modified material as Execution

Constraint, since it contextualizes the application of the referenced paragraph, as opposed
to modifying the paragraph directly (which would imply the encoding as Constituting
Properties ).
22 The encoding could conceivably extend further and alternatively afford the modi-
fication of the linked statement as “Organic farming operations must not utilize
genetically-modified seeds, with the exception of genetically modified material.”, where the
appended component is underlined. Note, however, that such reconstruction has signifi-
cant reliability challenges, including the iterative refinement and structural modification of
already encoded statements, as well as the de facto exclusion of the superfluous statement
whose institutional content has been absorbed into the refined statement.
analysis beyond the focal statement, to the policy/transcript (or subset

thereof) in the wider sense. A narrow interpretational scope consequently
focuses on the statement in isolation. Contextual information outside the
statement is not considered, and the effect of the statement is assessed
based on the semantics captured in the statement itself, ensuring high
levels of reliability, at the expense of not accommodating the resolution
of implied semantic linkages during the encoding phase, an aspect that
may be desirable or undesirable in the first place. Methodological consid-
erations and analytical implications associated with either preference are,
along with methodological guidelines, discussed in Chapter 7.
Concluding the conceptual discussion, the IG 2.0 recognizes a novel
hybrid form of institutional statement that augments the existing regu-
lative and constitutive forms. Doing so, it responds to the theoretical
discourse challenging the ontological distinctiveness of constitutive and
regulative statements in the first place, as well as reflects empirical obser-
vations related to the combined or variably parameterizing and regulating
effects of statements. Signals that indicate the hybrid nature of statements
can either be structural in kind, leading to the characterization of state-
ments as constitutive-regulative and regulative-constitutive respectively.
Beyond the structural perspective, ambiguity arises from the position
of the coder/analyst in the process. For statements of such kind, the
encoding may respond to the analytical objective, or variably permit an
encoding as a polymorphic statement to avoid concessions to a specific
analytical objective, while reflecting the dual function that the state-
ment of concern holds. The last form of hybrid statement reflects the
commonplace re-parameterization of existing statements by subsequent
provisions. Recognizing the interpretational scope as a central determi-
nant, the planning of any study shall consider the scope as a central
parameter of the study design, an aspect discussed in the supplementary
IG 2.0 Codebook. As for the varying analytical perspective, statements
could conceivably be encoded as polymorphic statements to seek dissoci-
ation of encoding from specific or predefined analytical objectives (e.g., to
establish a general-purpose dataset). Table 5.2 summarizes the discussed
points, along with candidate strategies to prevent ambiguous encoding
and associated reliability challenges.
Substantiating the presence of hybrid institutional statements as an
integrative feature, recognizing both the presence of regulative and
constitutive statements, as well as their combined use that addresses a
Table 5.2 Indicators of Hybrid Statements
Signal for hybrid statement Resolution strategies
Structure Encoding as constitutive-regulative or

constitutive-regulative statements respectively
Analytical Perspective Variably encoding from analytical perspective or
as polymorphic statement to establish
objective-agnostic set of coded statements
Interpretational Scope Interpretation according to interpretational scope
defined in planning phase, or as polymorphic if
interpretational scope is undefined
Fig. 5.8 Institutional Statement Variants
set of the highlighted theoretical and operational challenges, Fig. 5.8

positions the various forms on a continuum, with definitional and
behavior-regulating statements occupying the extreme strata, and the
varying integrative forms introduced throughout this section situated in
between. This includes hybrid statements that are based on the struc-
tural linkage of statements of varying kinds, generally with a dominant
regulative or constitutive effect, and a center occupied by polymorphic
institutional statements as forms of hybrid institutional statements that,
in the absence of methodological guidelines, can take either shape based
on varying contextual interpretation.
This integrative treatment of statements of all forms implies the
cross-cutting nature of the features previously described separately for
regulative and constitutive statements throughout this chapter (e.g.,
Object-Property Hierarchy, Property Types, Context Taxonomy); these
equally apply to hybrid statements in their varying forms. This harmo-
nization of all statement forms attends to one central objectives of the IG,
the comprehensive encoding of institutional settings. However, it further
exposes the susceptibility of interpretation to analytical objectives, and
the challenges associated with the establishment of a generic encoding

scheme that dissociates from, and thereby accommodates, a wide range
of analytical objectives and techniques.
The introduction of hybrid institutional statements concludes the
conceptual foundations of IG Extended, as well as the central structural
features of IG 2.0 specifically. However, before turning to the discussion
of semantic aspects of the IG, essential concepts introduced under the
label IG Extended are recalled and synthesized.
5.3 Summary of Chapter Content

Before turning to the highest level of expressiveness, IG Logico, in
Chapter 6, the breadth of features and conceptual background introduced
under the label IG Extended (highlighted in Fig. 5.9) invite for a review.
Where IG Core provided the basis for the general parsing of institu-
tional statements by specifying prototypical syntactic forms of institutional
statements, alongside consistent semantics associated with specific compo-
nents, and, more importantly, their interlinkage, IG Extended revisits
parsing on the component level in order to draw out detail and nuance
that the IG Core specification does not capture.
Fig. 5.9 IG 2.0 Features by Level of Expressiveness

To this end, IG Extended introduces a set of distinctive principles

that operate across components of both the regulative and constitutive
syntactic forms, and expose additional detail by making nested statement
structures on component level explicit (Sect. 5.1.1), and introducing a
distinction between institutional statements and the institutional states
these reference (see Sect. 5.1.2), a common occurrence in the context
of Activation Conditions. To realize these linkages as part of the coding
process, the deep parsing of IG Extended abstracts from the structural
alignment of linguistic and institutional form by motivating the recon-
struction of institutional content captured within components in terms of
the same syntactic patterns that apply to statements as a whole by drawing
on concepts of Cognitive Linguistics (see Sect. 5.1.3).
To motivate the principal application of deep parsing based on
component-level nesting – and departing from the primarily concep-
tual introduction of the IG – this chapter offered initial methodological
guidance (see Sect. 5.1.3) with respect to statement parsing under IG
Extended, an aspect that receives dedicated attention in Chapter 7.
Complementing the nested behavioral/existential structures captured
in institutional statements, IG Extended further identifies conceptual rela-
tionships of entities in the form of objects (or entities more generally)
and associated properties organized in various patterns referenced as the
Object-Property Hierarchy (Sect. 5.1.4) that is able to reflect variable
forms of complexity by signaling existential dependency relationships. The
extraction of conceptual relationships is further linked to a general Prop-
erty Typology (Sect. 5.1.5) that identifies component properties based on
their general nature as either quantitative or qualitative in kind.
A final cross-cutting feature is the Context Taxonomy (Sect. 5.1.6),
which draws the contextual linkage of components to the action situ-
ation or the wider institutional setting based on discussed substantive,
procedural, aspirational, or situational criteria.
Table 5.3 highlights the features introduced as part of IG Extended
in greater detail, and recalls components affected by the corresponding
features.
This review of features that enable the deep structural parsing prin-
ciples introduced with IG Extended includes the resolution of the
empirically observed interlinked use of regulative and constitutive state-
ments by conceptually integrating those in the form of Hybrid Institu-
tional Statements. The explicit recognition of this integration facilitates
a comprehensive treatment of institutional information, but furthermore,
Table 5.3 IG Extended Features Summary
Feature Purpose Affected

components/operators
Component-level Nesting Extracting Activation Condition,

behavioral/existential Execution Constraint,
information from components Direct Object, Indirect
based on uniform components Object, Constituting
Properties, Properties
associated with any
component
Object-Property Hierarchy Extracting conceptual Direct Object, Indirect
relationships between entities Object, Constituting
Properties, Activation
Condition and
Execution Constraint
(implicitly), Properties
associated with any
component
Property Typology Classifications of entity Properties associated
descriptors/properties as with any component
qualitative, quantitative, or
structural in kind
Context Taxonomy Context classification to Activation Condition,
capture linkage between Execution Constraint,
statements, action situations Direct Object, Indirect
and institutional settings Object, Constituting
Properties, Properties
associated with any
component
offers the opportunity to provide a conceptual response to the theoret-

ical rift surrounding the distinctive differentiation between regulative and
constitutive statements discussed in the chapter, and hence a proposal to
reconcile studies of institutions in theory and practice.
The upcoming Chapter 6 shifts the attention to semantic features of
the IG, enabling its conceptual and epistemological linkage and adap-
tation to specific analytical use cases, in addition to complementing the
concepts introduced intuitively to this stage with a formal characterization
that lays the foundation for the logical treatment of encoded institutional
statements.
References
Agazzi, E. (2011). Consistency, truth and ontology. Studia Logica, 97 , 7–29.
https://doi.org/10.1007/s11225-010-9295-x
Center for Behavior, Institutions and the Environment. https://complexity.
asu.edu/sites/default/files/papers/cbie_wp_2018–006_0.pdf
Broccias, C. (2013). Cognitive Grammar. In T. Hoffmann & G. Trousdale
(Eds.), The Oxford handbook of construction grammar (pp. 1–14). Oxford
University Press. https://doi.org/10.1093/oxfordhb/9780195396683.013.
0011
Campbell, Keith (2005). Ontology. In D. M. Borchert (Ed.), Encyclopedia of
philosophy. Macmillan Reference USA.
Cherry, C. (1973). Regulative rules and constitutive rules. Philosophical Quar-
terly, 23(93), 301–315. https://doi.org/10.2307/2218059
2975
Grossi, D., Meyer, J.-J. C., & Dignum, F. (2008). The many faces of counts-as: A
formal analysis of constitutive-rules. Journal of Applied Logic, 6(2), 192–217.
Hage, J. (2018). Two concepts of constitutive rules. Argumenta, 4, 1, 21–39.
https://doi.org/10.14275/2465-2334/20187.hag
Hindriks, F. (2009). Constitutive rules, language, and ontology. Erkenntnis,
71(2), 253–275. https://doi.org/10.1007/s10670-009-9178-6
Hindriks, F. (2013). Restructuring Searle’s making the social world. Philosophy
of the Social Sciences, 43(3), 373–389. https://doi.org/10.1177/004839311
1418299
Hindriks, F. (2015). Deconstructing Searle’s making the social world. Philosophy
of the Social Sciences, 45(3), 363–369. https://doi.org/10.1177/004839311
5575911
Langacker, R. (2008). Cognitive Grammar: A basic introduction. Oxford Univer-
sity Press. https://doi.org/10.1093/acprof:oso/9780195331967.001.0001
Mourelatos, A. P. (1978). Events, processes, and states. Linguistics and Philos-
ophy, 2(3), 415–434. https://doi.org/10.1007/BF00149015
Ransdell, J. (1971). Constitutive rules and speech-act analysis. The Journal of
Philosophy, 68(13), 385–400. https://doi.org/10.2307/2025037
Rawls, J. (1955). Two concepts of rules. The Philosophical Review, 64(1), 3–32.
https://doi.org/10.2307/2182230
Raz, J. (2012). Practical reason and norms. Oxford University Press. https://
doi.org/10.1093/acprof:oso/9780198268345.001.0001
Ruben, D.-H. (1997). John Searle’s the construction of social reality. Philosophy
and Phenomenological Research, 57 (2), 443–447. https://doi.org/10.2307/
2953734
Schneider, N., Hwang, J. D., Srikumar, V., Green, M., Suresh, A., Conger, K.,
O’Gorman, T., & Palmer, M. (2016). A corpus of preposition supersenses.
LAW 2016 - 10th Linguistic Annotation Workshop, held in conjuncion with
ACL 2016 - Workshop Proceedings (pp. 99–109). https://doi.org/10.18653/
v1/w16-1712
Schneider, N., Prange, J., Blodgett, A., Hwang, J. D., Moeller, S. R., Srikumar,
V., Stern, A., Bitan, A., & Abend, O. (2018). Comprehensive supersense
disambiguation of English prepositions and possessives. ACL 2018 - 56th
Annual Meeting of the Association for Computational Linguistics, Proceedings
of the Conference (Long Papers) (vol. 1, pp. 185–196). https://doi.org/10.
18653/v1/p18-1018
University Press.
Searle, J. R. (2015). Status functions and institutional facts: Reply to Hindriks
and Guala. Journal of Institutional Economics, 11(3), 507–514. https://doi.
org/10.1017/S1744137414000629
Warnock, G. J. (2020). The object of morality. Routledge. https://doi.org/10.
4324/9781003051053
Wittgenstein, L. (1983). Remarks on the foundations of mathematics (G. H. von
Wright, R. Rhees, & G. E. M. Anscombe, Eds.). MIT Press.
CHAPTER 6
Institutional Grammar 2.0: Semantic

Features and Analytical Linkages
6.1 IG Logico
With the introduction of features that are primarily geared at extracting
detailed structural information from institutional statements on (sub-)
component level, IG Extended expands on the principles of coarse-
grained encoding introduced as part of IG Core, and by doing so,
represents a backward-compatible extension of the latter; data parsed at
the granularity level of IG Extended can be collapsed to the coarse IG
Core. As a further characteristic, both IG Core and IG Extended share the
inter-component linkages and associated semantics, and thereby enable
the representation of component relationships within an institutional
statement.
An aspect that neither IG Core nor IG Extended focuses on, is
the expression of tacit properties as well as the linkages of entities
across statements, and conversely, the meaning of behaviors (in regulative
statements) or functions (in constitutive statements) evaluated through
distinctive analytical perspectives – a central feature of the highest level of
expressiveness, IG Logico.
Shifting from a primarily structural to a semantic perspective, IG
Logico introduces the underlying formal specification of the IG as an
initial aspect, as well as a set of additional features that can draw on this
specification and can be selectively applied. These include:

Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2_6
• Syntactic and semantic specification of the Institutional Grammar

• Semantic annotations of components based on general and domain-
/application-specific taxonomies
• Resolution of logical linkages between individual components
• Transformation rules
Whereas the formal specification of the IG seeks to establish conceptual

rigor of the IG by complementing the narrative introduction provided to
this stage, the remaining features focus on the epistemological integration
of the IG with specific domains and studies, as well as the opportunities
associated with the algorithmic transformation of statements enabled by
the formal specification. These features provide the basis for the discussion
of applications in Chapter 8. Readers primarily interested in pragmatic
aspects related to the application of IG Logico in IG studies – and less
so in underlying formal foundations – are encouraged to continue their
lecture from Sect. 6.1.2 onwards.
Motivating the initial aspect of IG Logico, the introduction of the
IG to this point in the book has implicitly highlighted semantic link-
ages between different components, starting with the stratification of
the Context component into both Activation Conditions and Execution
Constraints that variably relate to other components or the statement
as a whole as either preconditions for applicability, or as qualifiers of
the regulated activity (Aim) or C onstitutive F unction. Given the over-
arching objective to establish computational tractability of the IG (as part
of the IG 2.0), a formal specification of both the syntax and the seman-
tics is necessary to allow for independent, but conceptually compatible,
implementations and interpretations – enabling diverse applications of the
IG that can foster synergies not realized in current applications of the
IG, while nevertheless ensuring the consistent and unambiguous inter-
pretation of institutional statements. Maintaining compatibility across the
various levels of expressiveness, IG Logico relies on many of the concep-
tual features related to IG Core and IG Extended as introduced in
Sect. 4.2 and Sect. 5.1, respectively.
6.1.1 Semantic Specification

Since aspects of formalization are of primary concern for advanced analyt-
ical applications of the IG, conceptual principles are covered here, but
details are provided in supplementary resources at https://newinstituti

onalgrammar.org.1 As far as the syntactic treatment of institutional state-
ments is concerned, we will introduce those in Chapter 7 as part of the
methodological aspects revolving around the encoding of institutional
statements.
Reviewing the semantics that underlie the evaluation of institutional
statements is useful in order to understand the evaluation of institutional
statements from a logical perspective, in contrast to the intuitive access
based on the ABDICO,2 or ADIBCO, form that reflects the common
linguistic representation of institutional statements. Where the linguistic
perspective emphasizes an actor-centric or entity-centric interpretation for
regulative and constitutive statements respectively, the logical interpreta-
tion commences with the precondition for the execution of a statement,
the Activation Condition that signals the applicability of the remaining
statement. As an initial step for the specification of the semantics, we
thus introduce the symbols applied throughout the specification, starting
with the individual components without comprehensively reiterating the
definitions offered in Sects. 4.2.1 and 4.2.5.
In this context, A symbolizes the Attributes component, referencing
the actor whose behavior is constrained. In the specification, the Deontic
D is a function that moderates the extent to which the behavior specified
in the institutional statement is restrained or compelled. I is a func-
tion that represents the regulated activity that takes as parameters the
components it operates on, namely Direct Object, represented as Bdir, the
Indirect Object (Bind),3 and the Execution Constraint Cex that quali-
fies the behavior specified in I . Cac represents the Activation Condition
that specifies the applicability of the specified provision in the first place.
1 Resources to this book include a general overview of the IG, its formal specification,
methodological guidance for the encoding of institutional information, alongside coded
examples from the book chapters and associated software to support the parsing and
analysis of institutional statements (e.g., IG Parser), all of which can be found under
https://newinstitutionalgrammar.org.
2 This acronym reflects the common form in which the original IG, including the
additional Object component, is referenced.
3 The functional linkage between activity and objects is discussed in Sect. 4.2.1.4.
Where A, I , and the Context types Cac and Cex are required for any
regulative institutional statement,4 the remaining ones are optional.
Complementing the regulative perspective, E represents the Consti-
tuted Entity, F the Constitutive Function, and the Modal (signaling
optional or required/necessary nature of the activity described in the
Constitutive Function as described in Sect. 4.2.5.2) is represented as M .
As with the regulative side, F represents a function that takes Consti-
tuting Properties, represented as P, as parameter, alongside the Execution
Constraints Cex that qualify the function execution. As with the regulative
side, Activation Conditions Cac capture the conditions for applicability of
the statement.
Table 6.1 provides an overview of the symbols highlighted above and
used throughout this chapter.
On the basis of these components, the elementary form of an
institutional statement, an atomic institutional statement (stmt atm ), is
constructed as showcased in Eq. (6.1) in propositional logic. In its
base form, this can either (the alternative linkage is signaled by )
take the shape of a strategy of regulative or constitutive kind, vari-
ably represented in the Attributes-Aim-Context (AIC ) or Constituted
Entity-Constitutive Function-Context (EFC ) form. It can further assume
a normative form that includes the Deontic or Modal as additional
syntactic components, with Cac signaling the conditional activation of
the Attributes-Aim linkage further parameterized with Object variants and
Execution Constraints as shown in the first statement.
Atomic Institutional Statement
Regulative Strategy
Regulative Norm
(6.1)
Constitutive
Strategy
Constitutive Norm
Atomic statements (stmt atm ) of such kind can be combined into state-
ment combinations (stmt cmb ) as shown in Eq. (6.2), reflecting the notion
4 Note the inferred default values for Activation Conditions and Execution Constraints
in the absence of an explicit specification (see Sect. 4.2.1.5).
Table 6.1
Component Symbol Component Defined in
IG Component Symbols
Regulative Component Regulative
Symbols Components
A Attributes Section 4.2.1.1
D Deontic Section 4.2.1.2
I Aim Section 4.2.1.3
B dir Direct Object Section 4.2.1.4
B ind Indirect Object Section 4.2.1.4
C ac Activation Section 4.2.1.5
Condition
C ex Execution Section 4.2.1.5
Constraint
Constitutive Constitutive Components

Component Symbols
E Constituted Section 4.2.5.1
Entity
M Modal Section 4.2.5.2
F Constitutive Section 4.2.5.3
Function
P Constituting Section 4.2.5.4
Properties
C ac Activation Section 4.2.1.5
Condition
C ex Execution Section 4.2.1.5
Constraint
of horizontal nesting (see Sect. 4.2.2) in the form of statement combi-

nations. The linkage of these statements can occur in various forms,
reflecting the logical operators introduced in Sect. 4.2.2, namely conjunc-
tion (∧), inclusive disjunction (∨), exclusive disjunction (), and various
combinations of negated (¬) institutional statements.5 Statement combi-
nations, themselves, can comprise of atomic statements or already existing
statement combinations, or any mix thereof (e.g., (stmt atm ∨ stmt cmb )),
5 The negation of institutional statements will be of central concern in the discussion

of transformation rules in Sect. 6.1.4.
allowing for arbitrary forms and extents of (de)composition.
(6.2)
An institutional statement (stmt ) in its general form is then either of

atomic type (stmt atm ) or a combination (stmt cmb ) as shown in Eq. (6.3).
(6.3)
Based on this definition of institutional statements, we can then

finally introduce the consequential linkage of institutional statements in
Eq. (6.4), introduced as vertical nesting (see Sect. 4.2.2) that links a
leading monitored statement (stmt m ) with a separate consequential state-
ment (stmt c ) that activates in the case of violating (or not fulfilling) the
leading monitored statement (i.e., ¬stmt m ). Since both these statements
rest on the specification introduced before, monitored and consequential
statements can exhibit the same level of complexity (i.e., be themselves
atomic or combinations). Statements embedded in horizontally nested
statement combinations can themselves showcase vertical linkages to
statements outside the statement combination, reflecting a form of inter-
statement linkage (e.g., to reflect potential side effects associated with the
violation of specific parts of a statement).
Consequential Linkage of Institutional Statements

(6.4)
¬stmtm → stmtc Vertical Nesting
A further aspect introduced in the context of IG Extended is the

substitution of individual components by statements in their entirety
(i.e., either as individual statements or combinations thereof), an aspect

referred to as component-level nesting (see Sect. 5.1.1). This concept
revisits the formalization of individual components to articulate the vari-
able complexity those can exhibit. This approach is exemplified for the
Activation Condition component (showcased in Eq. [6.5]), given its
central role in the representation of inter-statement linkages of institu-
tional statements. Activation Conditions are furthermore special in that
their presence may be implicit, but nevertheless be required to constitute
a complete institutional statement.
In its basic form, and if not explicitly specified, an Activation Condi-
tion holds under all circumstances, semantically resolving to the Boolean
value true. Alternatively, an explicit atomic condition (e.g., “at 8 am”)
or its negation (¬Cac; i.e., “NOT at 8 am”) may be present, or any of
those atomic conditions may exist in various forms of logical combinations
(e.g., (expr ∧ expr), etc.). The final, complex variant of an Activation
Condition is an institutional statement (stmt ) that itself can be of atomic
and complex nature, including the negated form (¬stmt ). As signaled
here, component-level nesting captures the full spectrum of expressiveness
ranging from default values, simple atomic values to complex expressions,
and any combination thereof.
(6.5)
The component-level nesting semantics, reflected for Activation

Conditions at this stage, apply in similar form (i.e., adapted to the specific
component characteristics) to other syntactic components of the IG, as
well as to their respective Propert y subcomponents (where existing for
a given component). The complete semantic specification is provided in
equivalent form via the book website.6
More than offering a complete introduction to the semantic foun-
dations of the IG, the central purpose is to draw the link between the
intuitive access to the IG, presented throughout the preceding chap-
ters, and its sound formal specification of component characteristics,
their interrelationships, as well as the various forms of statement link-
ages that the Institutional Grammar affords. By doing so, it highlights
the general nature and applicability of the previously introduced concepts,
while at the same time sponsoring the precision needed to ensure unam-
biguous interpretation of encoded institutional information. This formal
rigor is particularly relevant to ensure compatibility and interoperability
of computational applications building on the Institutional Grammar,
leveraging the synergetic potential of future studies and applications of
the IG. Finally, it is essential to reiterate that the specification outlined
here is inclusive of both regulative and constitutive statement forms,
an aspect revisited in Sect. 5.2. Beyond the operationalization offered
here and elsewhere,7 this high-level semantic specification is not intended
to be limiting, but rather provides the basis for the introduction of
domain-specific refinements where analytically useful (e.g., specific actor
or context semantics) or operationally necessary (e.g., for computational
modeling), while at the same time assuring compatibility based on first
principles.
Following this initial discussion of formal characteristics of the IG
that invite for the computational and logical treatment of institutional
information (and returning to it in later sections), the following sections
focus on features of more immediate practical analytical value by high-
lighting the ability to encode the functional meaning and purpose of
institutions more generally, as well as allowing the analyst to draw theo-
retical linkages to her domain and application as part of the encoding.
6 The book website can be found under https://newinstitutionalgrammar.org.

7 The semantic specification linked via the book website (https://newinstitutionalgr
ammar.org) puts particular emphasis on individual component specifications not covered
here.
Central to achieving this is, as indicated in the beginning of this section,

a consequent shift to a semantic perspective when encoding and inter-
preting components and institutional statements, an aspect that builds on
semantic annotations as a central feature.
6.1.2 Semantic Annotations

Fundamentally, the IG is intended to support understanding of what insti-
tutional statements do within institutional settings by drawing on patterns
reflected in their structural features. Thus, offering a mechanism for over-
laying structural features with semantic characterizations suited to the
institutional setting as well as the analyst’s objectives, is critical. Specif-
ically, this involves the aggregation of institutional information across
specific components, the reconstruction of semantically rich relationship
networks, and furthermore allowance for the analytical mapping of state-
ments and concepts embedded therein to concepts and theory in the
target domain. Given the highest level of expressiveness that IG Logico
reflects, it is important to highlight that the features and concepts intro-
duced in this and other sections of this chapter allow for selective
application and refinement—albeit without concessions to methodolog-
ical rigor and consistency in application; conventions to define and specify
the use of features is subject to the methodological design of any study
applying the Institutional Grammar, an aspect discussed in Chapter 7, as
well as in supplementary resources.
To support the aforementioned mapping between structure and
meaning, or semantics, IG Logico revisits components of institutional
statements and proposes a set of classifications, here presented in the
form of taxonomies of annotations that are associated with specific (or
apply across) components to extract richer contextual meaning not acces-
sible based on the analysis of structure alone, or information of common
concern in the context of institutional analysis. A second purpose of this
incremental revisiting of institutional statements is to resolve inconsisten-
cies present in the prior conception and operationalization of selected
components in the Institutional Grammar.
A particular taxonomy that has found earlier introduction is the
Context Taxonomy. Explicitly introduced in the context of IG Extended,
its central purpose is to disambiguate the—by definition—conceptually
diverse Context component, building on the basic stratification present
in the original Grammar. Another typology introduced as part of IG
Extended is the Property Typology, affording the structural (and implicitly
high-level semantic) decomposition of the compound Properties concept

applicable across a wide range of components.
Taxonomies introduced in the context of this section include not only
the referenced variants, but extend to (and at times across) other compo-
nents based on the associated institutional features and their underlying
meaning or function. Complementing the annotation on component
level, selected taxonomies operate on statement level, so as to capture the
specific function of a statement in relation to other statements, or within
the broader institutional setting. The following sections iterate through a
set of such general taxonomies, alongside the scope of applicability and
contextualization of institutional and/or analytical value.
6.1.2.1 Animacy Taxonomy

An implicit characterization that the original Institutional Grammar relied
on, specifically with a focus on the Object component, Animacy is an
essential feature in the coding process. Recognizing the diversity of forms
that the revised Object component, now in the form of Direct Object and
Indirect Object, can take, semantic annotations do not limit either Object
type to reference animate or inanimate entities, but rather allow for the
flexible case-specific annotation thereof. The Animacy Taxonomy hence
recognizes two categories, animate and inanimate, that apply across
various components, including Attributes, Attributes Properties, Direct
Object, Direct Object Properties, Indirect Object, Indirect Object Prop-
erties, Activation Condition, Execution Constraint, Constituted Entity,
Constituted Entity Properties, Constituting Properties, Constituting Prop-
erties Properties.8
The value of such characterization specifically lies in the reconstruc-
tion of actor relationships based on the presence of animate actors across
the individual components (e.g., for social network analysis or agent-
based institutional models), distinguishing those from activities that link
actors and inanimate objects, such as artifacts, or abstract concepts. Where
analytically central, the analyst can thus filter institutional information
based on animacy characteristics for specific or across components in order
to extract relevant linkages, as opposed to being limited to particular
components that conventionally (but not exclusively) hold (in)animate
entities.
8 Given the implicit replication of these components in component-level nested state-

ments, or consequential linkage via the Or else, the same principles apply on those
levels.
6.1.2.2 Metatype Taxonomy

Objects, for instance, can not only be either animate or inanimate in kind,
but can be of concrete nature, or be abstract in kind (e.g., representing
mental concepts, such as beliefs or suspicions) and can furthermore appear
in varying forms of complexity, e.g., embedding complex institutional
content (e.g., “Inspectors must ensure that farmers report on their agri-
cultural practices ...”, with the nested Object structure represented in
italicized form), an aspect elaborated in Section 5.1.2.
To capture the variable use of selected components, the introduced
Metatype Taxonomy allows for an explicit characterization of entities as
either concrete or abstract in kind. Offering such distinction provides the
analyst with the opportunity to explicitly distinguish between objects,
actions, and states that produce detectable correspondents in the insti-
tutional setting, i.e., have institutional status, in contrast to ones that do
not (e.g., entirely mental concepts). Where introduced or referenced by
constitutive or regulative statements, respectively, institutional concepts,
such as a notification, have a function and effect in the institutional
setting, including a potential physical correspondence (e.g., in the form of
an artifact). Similarly, a violation may represent a concept that manifests
itself in the form of action in the institutional setting, i.e., is physically
observable both in terms of performance and effect. Contrasting this
position, purely mental concepts, such as beliefs or goals, may not carry
institutional status (or may not be sufficiently observable), and thus not
qualify as concrete in the institutional sense.
The value of this differentiation is linked to the objectives, and tech-
niques applied in the analysis, but aimed at analyses that intend to extract
richer systemic structure that extend beyond actor networks. The analyst
may seek to triangulate institutional concepts referenced in the institu-
tional setting, while distinguishing those from relevant actors (identified
via the Animacy Taxonomy), and mental concepts represented in Direct
Object or Indirect Object fields at the same time. Moreover, analyzing
the presence or absence of institutional concepts across components
offers a basis to draw insights on combinatorial or interaction effects of
components of particular types. For instance, drawing on this characteri-
zation can indicate as to whether decision-making parameters in an action
situation primarily consist of mental concepts (e.g., as signals for discre-
tion) – and thus pointing to high on subjective assessment of institutional
arrangements –, or are in fact grounded in tangible observable insti-
tutional concepts – pointing toward arrangements that can be robustly
and more objectively assessed by third parties or from the perspectives

of multiple actors.
Naturally, this annotation applies to any occurrence of entity references
across institutional statements, such as Attributes, Object variants, Consti-
tuted Entity, as well as Constituting Properties and Context components.
Table 6.2, at the end of this subsection, explicitly highlights the associated
components (alongside potential properties) explicitly.
6.1.2.3 Role Taxonomy

Whenever interacting in an institutional setting, or more narrowly, in
an action situation, actors necessarily assume positions, or roles therein.
Naturally, the specific nature and configuration of roles in terms of
responsibilities, liabilities, and other forms of status characterizations,
depends on the institutional setting and domain. Operationally, these
roles can shift situationally across different components, based on their
thematic positioning (e.g., an entity who may be the sender of an object
in one statement may be the receiver in another). These situational char-
acterizations can be identified in terms of general thematic roles (Carlson,
1984) that exist independently of the underlying domain. Specifically
the categorization of actors (and implicitly other entities) by thematic
role, recognized both theoretically (Carlson, 1984) as well as empir-
ically observed in natural language (with varying levels of specificity,
e.g., Schneider et al., [2018]), allows for a semantic layer that enables
a precise situational characterization and positioning of actors and enti-
ties as described by individual institutional statements. Within regulative
institutional statements, for instance, actors associated with the Attributes
component are the actors whose behavior is explicitly described and regu-
lated in the statement, whereas actors referenced in the context of Direct
Object are receivers of activity, or, specifically when referenced in the
Indirect Object component, experiencers of activity. Conversely, actors
may be referenced in Properties of varying kind, in which they assume
a relationship to the entity characterized, oftentimes in possessive form
(e.g., “operation owned by farmer” ). Referenced actors may, however,
also be represented as a beneficiary of an action contextually referenced
(e.g., in Context or Properties components), but not directly involved or
immediately affected by the regulated or specified activity. While largely
motivated for regulative statements, references to actors in varying forms
equally apply to components of constitutive kind. A specific example is
the reference to an actor that is endowed with a specific position, such
as being promoted to manager, and thus recipient of associated insti-

tutional power. Such individual is thereby directly advantaged in the
context of the action situation, whereas others, such as future subor-
dinates, are indirectly affected by experiencing the change in authority.
Where contextually inferable, the status endowment of the manager may
be to their advantage or disadvantage.
An essential benefit of a principal role structure is to allow for an
abstract classification of functional relationships associated with the actor
structure overtly presented in institutional statements. While conceptual
linkages between specific components exist (e.g., Attributes as originators
of activity), the relational dimensions in the role taxonomy apply across
components and properties. Such characterizations offer the basis for
the combination with linguistic concepts such as Semantic Role Labeling
(Fillmore, 1968).
As a set of general actor characterizations, the IG recognizes the
following annotations that showcase differentiated actor role and effect
characterizations, where role characterizations reflect the operational
linkage of different entities as expressed through the statement more
generally, and the Aim or Constitutive Function more specifically. The
complementary effect characterization captures the potential effect of the
instruction captured in the statement on directly or laterally involved enti-
ties, selectively reflecting that an individual is affected in the first place,
or qualifying the nature of the effect (either as being advantaged or
disadvantaged).
• Role Characterizations
– Originator/Causer/Agent – Entity from which action origi-
nates
– Recipient – Recipient of whatever is conferred (e.g., actions,
objects)
– Possessor – Owner of an object/entity (e.g., “house owner”)
• Effect Characterizations
– Experiencer – Indirectly affected actor (e.g., “observer of non-
compliance”)
– Advantaged – Beneficiary distinctively advantaged by referenced
activity/function; may not necessarily be recipient
– Disadvantaged – Maleficiary distinctively burdened by refer-

enced activity/function; may not necessarily be recipient
However, since role characterizations are functionally linked to

the underlying activities or functions, let alone their domain-specific
meaning, the general categories introduced here can only offer a coarse
categorization and characterization of linkages commonly encountered in
institutional arrangements. Both characterizations are thus neither exclu-
sive or exhaustive. Instead, they invite for adaptations or extensions that
consider domain-specific features (e.g., power relationships established
based on interactions), or capture specific semantic frames associated
with the activities and functions relevant in a given action situation or
broader institutional setting (e.g., by drawing on the lexical database
FrameNet [Baker et al., 2003]).
6.1.2.4 Institutional Functions

The IG taxonomies introduced to this stage essentially focus on the
contextual embedding and physical characterization of entities refer-
enced in institutional statements. The earlier taxonomies primarily attend
to features that are specific to entities and exist independently of the
contextual interpretation in the institutional setting (such as on the char-
acterization of an entity as animate based on its natural kind). The
Role characterizations draw the attention toward a contextual embed-
ding of individual entities, and thus draw a more explicit linkage between
the action situation and its elements. However, to move beyond actors,
objects and artifacts, and to better accommodate functional specifics rele-
vant to institutional analysis, a central focus lies on the activities that
are referenced in institutional statements. Given the functional diversity
those can exhibit (the Aim is inherently open-ended with respect to its
content), the semantic characterization from an institutional perspective
requires access to the meaning underlying the specific activities or func-
tions, as well as their effects on the institutional setting. To this end, the
IG includes a semantic classification of activities based on their institu-
tional function specific to both the regulative and constitutive syntactic
form, and referenced as Regulative Functions and Constitutive Functions,
respectively.
Regulative Functions
Offering generalizable categories for regulative statements, and more
specifically, the regulated activity captured in the Aim, the IG recognizes
the broad array of activities based both on the complex configurational

positioning of actors, as well as the brute nature of the actions themselves.
To this end, regulative functions act as a mechanism to apply concep-
tual and epistemological lenses on an institutional setting to introduce
functional primitives of specific analytical value. Seen through the lens
of regulatory compliance, for instance, stereotypical actors include opera-
tional entities that are subject to regulation (e.g., organic farmers), actors
assuming responsibility for monitoring (e.g., inspector) and enforcement
(e.g., program manager), where the latter variably exists in conflated or
separate form as discussed in Section 4.2.3.
Seen through the abstraction of primary concern for the evaluation
of compliance assessment, regulative functions can be used to indi-
cate whether the activities referenced in an institutional statement signal
compliance or violation behavior. Exemplifying this using the statement
“if organic farmers do not deliver documentation before the deadline,
Program Managers may suspend their certification,” the non-delivery
may in fact represent a violation, whereas a delivery on time (e.g., “if
organic farmers deliver documentation before the deadline, ...” ) can be
interpreted as compliance behavior. For all occurrences of institutionally
relevant activities, regulative functions are thus a consistent set of prim-
itives that reflect an abstract functional overlay that isolates the meaning
of the described activity in the wider context of the action situation, and
thereby offers specific insight into the coordinating purpose of a particular
statement.
Organized by distinctive classes of activities and associated positions, a
regulative functions taxonomy (based on the complementarity of selected
sub-elements also characterizable as an ontology) exemplarily focused on
compliance assessment includes functions related to compliance or viola-
tion events, monitoring, enforcement, as well as corresponding responses
(e.g., an actor’s rejection of enforcement), leading to a set of annotations
organized as follows:
• Compliance action – action reflecting compliance behavior

– Comply – action reflecting compliance
– Violate – action reflecting violation
• Monitor – action reflecting the institutional function of monitoring
– Detect compliance – action reflecting the detection of compli-
ance
– Detect violation – action reflecting the detection of violation

• Enforce – action reflecting enforcement acts
– Reward – action reflecting rewarding behavior (regulative-
incentivizing)
– Sanction – action reflecting sanctioning behavior (regulative-
punitive)
• Enforcement response – action reflecting responses to enforcement
outcomes
– Accept – action reflecting acceptance of enforcement outcome
– Reject – action reflecting the rejection of enforcement outcome
– Appeal (specialization of reject) – action reflecting appeal
against enforcement outcome
It is important to reemphasize that the nature of such abstractions shall

respond to analytical purposes. The general form introduced initially, for
instance, highlights compliance assessments, whereas other analyses may
primarily focus on transaction characteristics of a particular action situa-
tion, and thus emphasizing a sequential flow of interactions, which affords
a varying (or extended) set of labels to adequately reflect this in annotated
institutional information. Transgressing into formal territory, regula-
tive function taxonomies can further afford a mapping from linguistic
expression to symbolic representations that enable a logical treatment of
annotated information in algorithmic form or based on formal calculi,
thus closing the gap to advanced logical applications in disciplines that
have a formal approach to institutional analysis, such as Normative Multi-
Agent Systems (Boella et al., 2006) or Legal Informatics (Katz et al.,
2021) (see also the discussion in Chapter 1).
The abstraction levels applied to the different functional characteriza-
tions naturally vary. Applications can, for example, include superimposed
theoretical frameworks, such as Lowi’s policy typology (Lowi, 1964)
that focuses on variable redistributive functions. Adapted to individual
statements, regulative functions can draw an effective linkage to relevant
theoretical frameworks and thereby offer a robust linkage between under-
lying data and evaluation and interpretation in terms of the theoretical
frame of the study.
This extends, for instance, to analyses applying policy process theo-
ries (Weible & Sabatier, 2018), especially for circumstances in which the
researcher can identify direct linkages between framework and specific

relevant statements. In this context, it is noteworthy to restate that the
IG may not only be applied to policy itself, but rather institutions of
any nature as well as associated processes. A central prerequisite for the
applicability of the IG, however, is ability to operationally link theoretical
abstractions provided in theory of interest with institutional information
at the granularity level of individual institutional statements (or groups
or configurations thereof), an aspect of particular challenge for theo-
ries that operate on policy level at large. Barring this concern, analysis
of regulative functions can operate on a wide range of abstractions and
without linkage to particular disciplines or domains. Where analysts may,
for instance, be interested in the analysis of Process lifecycles, such as the
initiation or termination of activities of concern (e.g., for the purpose of
modeling those), these can be reflected with a corresponding exemplary
set of labels, such as:
• Process—Life cycle
– Initiate
– Interrupt
– Resume
– Conclude
Summarizing this discussion, institutional functions in general, and

regulative functions in particular, offer a basis to link theoretical
constructs relevant for a given field, domain, theory, or analytical frame-
work onto institutional statements, and more specifically, the activity or
function they convey. The central benefit is to draw an immediate analyt-
ical linkage between the micro-level activity that the statement captures,
and the associated macro-level observations and effects, or, complemen-
tarily, to make the linkage between analytical conclusions and underlying
institutional information transparent, making an appeal to methodolog-
ical rigor of the analysis. Drawing on this annotation capability affords
analytical benefits for traditional statistical analyses that primarily aim
at establishing regularities and metrics, to constructionist approaches to
analysis that aim at understanding policy by reconstructing it in silico,
and accommodates the logical treatment of statements by exploiting the
semantics the regulative functions reflect in order to draw explanatory
insights into the policy or to assess its conceptual integrity.
Constitutive Functions
Contrasting the regulative perspective, institutional functions of constitu-
tive kind vary in focus. As discussed in Sect. 4.1.2, constitutive statements
bring about entities, relate those conceptually, or otherwise parameterize
their role in the wider action situation. The contrasting characterization
necessitates a different set of labels that best capture the declarative form
and existential focus that Constitutive Functions reflect.
The effects of Constitutive Functions, reflecting the various ways
in which Constituted Entities and potential Constituting Properties are
linked, show a set of general patterns that transcend domain-specific
observations, and respond to the nature of C onstitutive F unctions as
parameterizing in kind. To this end, the Constitutive Functions Taxonomy
provided in this context aims at capturing the general functions that the
Constitutive Function component plays, while, at the same time, affording
opportunities for refinement with respect to specificity.
Capturing the essential effects, Constitutive Function labels are orga-
nized into two general groups, namely functions that reference entities
defined or otherwise parameterized, and functions that reference the
institution (e.g., policy) itself, de facto assuming a meta-constitutive role.
As far as the characterization of entities is concerned (and reiterating
central aspects discussed in Sect. 4.2.5), Constitutive Functions can either
define entities, such as actors, roles, actions, artifacts, venues, status, or
other objects or concepts relevant in the institutional setting. Constitutive
Functions can further establish Relationships between entities, reference
Lifecycle states or stages, as well as reference the Conferral of Status of
various kinds. These varying classes of functions can carry various special-
izations characterizing the functional semantics with greater specificity.
Entities, for instance, can be defined intensionally (i.e., based on what
they are), or extensionally (e.g., based on what they do). Similarly, refer-
enced relationships can reflect linkages of different strengths, such as
compositional (e.g., describing what an entity consists of ) forms signaling
conceptual dependence (e.g., existential structural dependence),9 func-
tional dependence (e.g., control relationships), or organizational forms
(e.g., describing hierarchical linkages or embeddedness) that assume
limited levels of existential dependency. Beyond the stratified lifecycle
characterization, various forms of status can be conferred. This includes
9 For a discussion of conceptual and functional dependence in the IG, please refer to
Sect. 5.1.4.
the assignment of explicit rights, authority (alongside associated respon-

sibility reflected in corresponding regulative statements), privileges, or
liabilities to other kinds of entities (e.g., parameterizing roles in terms of
status characteristics). It is important to note that the status endowment
as operationalized here applies to kinds or types of actors or roles more
generally (e.g., organic farmers as a role), as opposed to specific actor or
role instances only (e.g., a specific farmer). In addition to honorary status,
the institutionally relevant forms of status referenced here characterize
legal relationships (Hohfeld, 1913) conceivably endowed explicitly.10
Applying to institutions more generally (e.g., informal rules, policy),
constitutive statements assume reflexive functions, including the specifica-
tion of lifecycle characteristics, implied relationships to other institutions
based on substitutions or amendments, as well as signal the purpose
of the institution, or other relevant information related to the insti-
tution or associated states of affairs. Figure 6.1 visualizes the general
Constitutive Functions Taxonomy, alongside exemplary references to verbs
operationalizing the corresponding Constitutive Function.
The analytical value of analyzing institutional statements based on the
annotated Constitutive Functions is manifold. From a purely quantita-
tive perspective, a central analytical value drawn from the annotation
of Constitutive Functions lies on the isolated analysis of constitutive
statements and the distribution of prevalent ways in which entities
are established, and/or related in a given institutional setting. From a
constructionist perspective, a central value is the characterization of the
entities embedded in a given scenario, in the ideal case based on defini-
tional statements that introduce relevant entities, akin to the antecedent
specification suggested by Searle, but furthermore to distinguish those
from re-parameterizations that occur throughout the operationalized
action situation, including modified linkages of entities, and changes in
lifecycle status – thus providing the basis for the assessment of insti-
tutional dynamics and change. A noteworthy value of the distinctive
characterization of Constitutive Functions is the integrated treatment of
the complementary operation of constitutive and regulative statements,
where, for instance, constitutive statements create status modifications on
10 While Hohfeld references a wider range of relationships, it is noteworthy to highlight

the interdependence of the referenced legal concepts, i.e., they are effects of the jural
correlatives (e.g., an assigned power may affect another person’s liability if subjected to
the endowed power), and may not be explicitly stated in policy.
Fig. 6.1 Constitutive Functions Taxonomy
the one hand, and regulative statements afford the corresponding opera-
tional implementation. The logical perspective, in contrast, benefits from
the annotations in that they enable the analysis of institutions for consis-
tency, asking questions as to whether all referenced entities in a formal
institution (e.g., policy) are actually defined in the first place, or identi-
fying gaps in the specification of entities and their relationships, providing
the basis for a quality assessment of the institution. At the same time, the
logical treatment provides the basis to establish an explanatory account
of institutional analysis by extracting and relating semantic institutional
content on statement level.
As indicated above, the Constitutive Functions Taxonomy in the context

of the IG, and equivalent to the Regulative Functions, explicitly anno-
tates the Constitutive Function component, i.e., the semantic linkage of
Constituted Entity and, where existing, Constituting Properties. This sits
in contrast to higher-level taxonomies that classify statements more gener-
ally with respect to the functions and effects that these statements exercise
in the wider context of a policy, especially with respect to legal context,
nature of the document, and legislative forum. Notable examples for the
latter taxonomies include the work by Ceci et al. (2018) in the field
of Legal Informatics, and Brady’s (2020) taxonomy developed in the
context of conservation treaties.
6.1.2.5 Vertical Nesting Annotations

In contrast to the annotations offered in the previous section that
primarily focus on the annotation of individual components, other than
referencing existing taxonomies that aim at capturing constitutive content
semantically, explicit statement-level annotations have found limited atten-
tion, an aspect addressed by the remaining annotation schemes.
A central concept applicable across all levels of expressiveness is the
notion of vertical nesting (Sect. 4.2.1.6; Sect. 4.2.2), a feature that
provides the basis for the administration of select transformation rules (as
discussed in Sect. 6.1.4). Where applied explicitly, the parsing of institu-
tional statements in terms of monitored and corresponding consequential
statement invites for the unambiguous functional characterization of the
linked statements. While the principal linking mechanism of statements
based on the logical operation that the Or else reflects – the mate-
rial consequence of non-fulfillment of a monitored statement – implies
the characterization of statements as either monitored or consequen-
tial. However, where such inference is not automatically inferred, the
corresponding characterization can be made explicit. Complementing the
annotation as either monitored or consequential , the novel distinction into
monitoring and enforcing statements based on the nature and poten-
tial distinctiveness of monitors and enforcers (introduced in Sect. 4.2.3),
further invites for a more specific characterization of a statement as moni-
toring in addition, or independent of the annotation as consequential
statement. Statements can naturally occupy multiple such annotations,
including both being monitored and consequential (based on the linkage
to other statements), as motivated in Sect. 4.2.1.6,11 and exemplified in

Chapter 8.
The proposition of this tripartite characterization is twofold. Firstly,
from a theoretical perspective, it reflects the dissociated analysis of
enforcing and monitoring entity – representing empirically observed
governance regimes that either favor a conflated or separated positioning
of monitor and enforcer in action situations, an aspect of relevance
both for socio-institutional and legal-institutional settings. This aspect is
unspecified in the original IG, thus endowing the New IG with extended
validity. The second benefit is of analytical kind: the explicit recognition
of statement types leverages the ability to distinguish between state-
ments that are subjected to monitoring (monitored statements ), primarily
carry incentivizing content (consequential statements ), or merely signal
oversight without direct intervention opportunities (monitoring state-
ments ), offering an indication of a policy’s integrity on the one hand,
and assessment of facilitative or punitive measures embedded in institu-
tional statements, in addition to the dissociated treatment of monitor and
enforcer motivated above. Exploring the linkage of such oversight char-
acteristics across configurations, or the policy at large, further enables a
macroscopic assessment of the governance design.
6.1.2.6 Consequence Annotations

The widened characterization of consequence types as part of the IG
2.0 (Sect. 4.2.5.6) expanded the range of analytical features by making
explicit the general forms that consequences take, namely the interlinked
social and institutional consequences (e.g., economic or status moderation,
environmental), and configurational consequences (i.e., re-parameterizing
effects on the institutional setting).
Extracting this information provides the analyst with a second-order
analytical qualification of consequences based on their identification via
Vertical Nesting Annotations (Sect. 6.1.2.5). While hinging on the pres-
ence of consequences, corresponding annotations leverage insights into
policy effectiveness and, potentially, into underlying value systems [e.g.,
fear of natural consequences as an incentive (Watkins & Westphal, 2016)],
by providing the basis to assess the performance of an institution based
11 A statement that prescribes the administration of a sanction (i.e., a consequential

statement) can be a monitored statement by being subjected to oversight by another
monitoring or consequential statement.
on the nature of the consequences. Naturally, such annotation is valuable

only for scenarios that extensively rely on incentivizing features, but is
agnostic to the internal or external origin of such incentives and hence
invites for the applications that draw on Deltas (Section 4.2.4) to estab-
lish a fine-grained understanding of the incentive structure and its linkage
to institutional design.
As with most other taxonomies or ontologies discussed in this section,
the analyst may feel inclined to adapt the consequence characterizations
for the specific analytical use (e.g., by introducing refined subcategories).
6.1.2.7 Summary
The introduction of the statement-level annotations completes the
overview of the IG with respect to abstract semantic annotations applied
across syntactic components (and statements) that are organized in
distinctive taxonomies and typologies. The introduced annotation prin-
ciples signal the shift to an analytical interpretation by drawing the focus
away from a primarily structurally motivated analysis to one that empha-
sizes the semantics that specific statements embed, leveraging analytical
focus on the institutional meaning that actions, actors, and contextual
aspects reflect.
The taxonomies broadly associate with the syntactic components, and
in selected instances, specifically apply to individual components, but
more commonly, find their application across component groups. Fig. 6.2
showcases the high-level association of the introduced taxonomies (repre-
sented as a middle layer) with the syntactic components of regulative and
Fig. 6.2 First-order Linkages between Taxonomies and IG Components

constitutive components, with the Or else acting as a proxy for the under-
lying institutional statement to which the corresponding taxonomies
apply.
As broadly observed, all components, other than the Deontic/Modal
(an aspect discussed later in this section), may be selectively associated
with the taxonomies introduced in this section. Most prominent is the
characterization of actors based on two dimensions that relate to physical
and biological categorization (Animacy Taxonomy, Metatype Taxonomy),
and thereby accommodating the various forms of entities the IG can
reference (e.g., mental concepts). The Role Taxonomy captures variable
actor relationships and their positioning in specific statements as occurring
across Attributes, Object, Constituted Entity, and Constituting Properties
components.
The Context Taxonomy assumes the central role of structuring the
diverse nature of Activation Conditions and Execution Constraints refer-
enced under this label. As discussed in the context of IG Extended, classes
and categories in this specific taxonomy operate on general classes of
context either offering specific value for a nuanced analysis of transitions
between statements and action situations (Situational Context ), embed-
ding in physical and domanial context (Substantive Context ), alongside
further empirically observed classes, such as instrumental references or
specification of purpose (Procedural and Aspirational Context ). These
dimensions are not independent, potentially allowing for the annotation
of statement context along with multiple such categories.
The Institutional Functions taxonomies, specifically Regulative Func-
tions and Constitutive Functions Taxonomy, exclusively focus on the
semantic annotation of activities and functions represented in institu-
tional statements, drawing out the functions those statements play in a
given action situation, potentially detecting statements acting in config-
urational form in order to highlight complex functional arrangements.
Specifically, the Regulative Functions provide the basis for introducing
analytical perspectives that reflect aspects related to specific domains, and
theoretical applications.
The final group of statement-level annotations, represented by the
Vertical Nesting annotations (Sect. 6.1.2.5), as well as the Consequences
annotations (Sect. 6.1.2.6), emphasize the semantic qualification of
enforcement and oversight structures more generally, thus providing the
basis for a configurational assessment of institutional statements in an
institutional setting by drawing out distinctive institutional purposes the
individual statements reflect. The annotation of consequences offers a

second-order qualification of consequences that provides insight into how
such governance regimes are furnished.
It is central to note that the referenced taxonomies do not operate
in isolation, but can be combined systematically to support specific
analytical objectives by drawing links to existing theoretical frameworks
across disciplines. Illustrating such adaptation from a linguistic perspec-
tive, a categorization scheme that complements the interpretation of
institutional statements as compositional patterns (see Sect. 5.1), is the
notion of semantic roles (Carlson, 1984). The general thematic roles
embedded in this theory can, if superimposed on selected taxonomies, be
directly derived and made accessible for linguistic treatment of encoded
semantic properties.12 Inasmuch as the taxonomies thus augment indi-
vidual components with semantic contextualization, they do not exist in
isolation, but perform an integrative function, complementing the analyt-
ical picture, whether drawn from a perspective of logical and computa-
tional analysis, linguistic treatment, or statistical aggregation traditionally
associated with the IG.
As indicated repeatedly, the taxonomies as shown here are not exhaus-
tive, and lend themselves for domain-specific adaptation where useful,
including contextual considerations that span across multiple compo-
nents, or link semantics only accessible when reviewing multiple compo-
nents. Variably, the taxonomies can be augmented with or substituted
by ontologies that integrate domain-specific relevant concepts entirely.
Ontologies offer hierarchical relationships of concepts that enrich the
flat classifications offered by taxonomies with conceptual interlinkages,
underlying rules, and axioms that underpin the application (Staab &
Studer, 2009). By doing so, they provide the basis for logical applica-
tions such as querying institutional statements (individual, collections, or
systems thereof) based on simple or complex criteria, as well as affording
assessments of consistency of statements based on the associated ontology.
A specific example of a component whose effect can both be estab-
lished based on the annotation of an individual component, as well as the
compound interpretation, is the Deontic component. Where their seman-
tics are grounded in the underlying logic (e.g., McNamara, 2006; von
12 Whereas the Role Taxonomy captures actor-centric thematic roles, additional

thematic roles such as Instrument or Goal are directly or indirectly referenced by other
taxonomies (i.e., Context Taxonomy and Metatype Taxonomy).
Wright, 1951), if applied in the context of specific document forms (e.g.,

contracts, specifications), specific deontics may carry distinctive semantics
reflected in discretized form (see e.g., Bodansky, 2016; Bradner, 1997),
and be captured explicitly as part of the annotation process. Alterna-
tively, or in addition, such operationalization may leverage the conception
of deontics organized along a continuum as motivated in Sect. 4.2.1.2,
which may either be theoretically grounded, or be based on empirical
observations. The actual effect of Deontic values, however, can often only
be assessed in context, e.g., under consideration of Execution Constraints
and logical operators that may moderate or emphasize the stringency asso-
ciated with a particular Deontic. The Deontic “must,” for instance, may be
moderated in effect if combined with qualifications such as “especially for
cases ...” Similarly, where actors can choose to perform a set of actions
(e.g., ‘suspend’, ‘revoke’), e.g., based on an AND/OR combination of
individual activities, the effect of the Deontic “must” is moderated based
on the discretion to selectively apply one or both activities in the first place
(compared to a provision that those not offer such discretion). We discuss
selected analytical opportunities related to the combined interpretation of
components in Chapter 8.
It is important to note that the schematics shown in Fig. 6.2 focuses on
high-level, or first-order linkages between taxonomies and corresponding
components. It does not capture higher-order relationships, such as the
applicability of these taxonomies across P roperties associated with the
different components (e.g., Attributes Properties, Direct Object Proper-
ties, etc.). The basis for this is twofold, firstly based on the very nature
of P roperties to contextualize the entity they describe (e.g., based on
categories captured in the Context Taxonomy), as well as entities/actors
referenced in addition to the primary annotation (an aspect of partic-
ular relevance for the Role Taxonomy). The second reason is the potential
presence of nested institutional statements across a range of these compo-
nents (e.g., nested activation condition), the components of which can
naturally be categorized in the same form as the first-order compo-
nents. At the risk of sacrificing the conceptual simplicity presented above,
Table 6.2 makes the linkages between components, subcomponents, and
taxonomies explicit. It further includes the qualitative characterization of
Properties by the Property Typology introduced in 5.1.4 (and referenced
accordingly in Table 6.2).
Following the characterization of a wide range of general taxonomies
and annotations to provide a foundation for the qualitative analysis of
Role
tions
Context
Animacy
Metatype
Taxonomy
tions****
Properties
Table 6.2
Regulative Func-
Constitutive Func-
x
x
Attributes
x** x
x
x
x
(x)
Attributes Properties
Deontic*
Relevant components
Aim
x
x
x
(x)
Direct Object
x
x
x
x
(x)
Direct Object Properties

x
x
x
(x)
Indirect Object
x
x
x
x
(x)
Indirect Object Properties

x
x
x
(x)
Constituted Entity
x
x
x
x
(x) Constituted Entity Properties

Modal*
x
IG Logico Taxonomies and affected Institutional Statement Components
Constitutive Function
x
x
x
x
x
Constituting Properties
x
x
x
x
(x)
Constituting Properties Properties
x
x
x
Activation Condition
x***
x***
x
x
x
x***
x***
Or else*****
(continued)
Table 6.2 (continued)
220
Attributes
Attributes Properties
Deontic*
Aim
Direct Object
Direct Object Properties
Indirect Object
Indirect Object Properties
Constituted Entity
Constituted Entity Properties
Modal*
Constitutive Function
Constituting Properties
Constituting Properties Properties
Activation Condition
Or else*****
C. K. FRANTZ AND S. SIDDIKI
Vertical Nesting (x) x

Consequences (x) x
(x) Indicates implicit applicability (e.g., as nested reference to an actor or context; implied consequential relationships)
* The semantic operationalization of the Deontic component is discussed in this section, with reference to Sect. 4.2.1.2. Corresponding considerations
for the Modal can be found in Sect. 4.2.5.2
** While generally concrete in nature, an anthropomorphized abstract actor is a conceivable Attribute (e.g., Nature)
*** The referenced taxonomies implicitly apply for complex Context characterizations
**** The taxonomy variably applies to entities specified or referenced in institutions, or the institution itself (see Sect. 6.1.2.4). Similarly, where multiple
components are indicative for the identification of a function, the annotation applies on statement level
***** The Or else is a proxy for the underlying statement to which the annotation applies
institutional content directly, the following section focuses on the repre-

sentation of statement linkages, which play a central meta-analytical role
for analysis of institutional design.
6.1.3 Statement References

Institutional statements do not exist in isolation. While oftentimes unam-
biguously linked based on the situatedness within the same action situa-
tion, or co-location within an institution-in-form (such as a policy docu-
ment), a common feature of statements is the explicit reference to other
institutions, an aspect that is overt in policy documents, and includes
references to other policy documents, Acts, parts, or sections therein.
To afford the comprehensive capturing of institutions in general, IG
Logico recognizes two dimensions along which statement or institution
references are organized, namely
• Internal vs. external references (and, generally correlating with this

distinction, implicit and explicit references)
• Scope of references (1:1, 1:n)
The first aspect highlights whether a statement references other state-

ments within the same source (e.g., policy document, contract), or
reference a source external to the analyzed document/dataset, etc. A
second differentiation relates to the scope of a reference, i.e., whether
an individual statement references another specific statement, a collection
of statements (e.g., section), or a document entirely (e.g., Act, contract).
Beyond those two distinctive dimensions, operationally, the occur-
rence of such references further varies by the degree of explicitness. The
following expression
when an inspection of an accredited certifying agent by the Program

Manager reveals any noncompliance with the Act or regulations in this part,
…
features internal references of varying scope, referencing an Act in its

entirety, or a specific part. In this instance, such references are reflexive in
that they target the same policy or provision the reference is embedded
in, but do not make the Act explicit – inadvertently requiring manual
annotation/inference of the Act where analytically beneficial. Variably, an
implicit linkage can describe dependency relationships between different

statements that may not even be implied in the text but are apparent or
known to the reader (e.g., legal commentary).
This is in contrast to the following statement:
the Federal Environment Agency may grant temporary exemptions from §11
(1), first and fifth sentences …
While this statement is reflexive, the references to other provisions are

in explicit form.
The difference between both forms lies in the necessary contextual
knowledge the analyst requires about the statement origin. Exemplifying
this for policies, the initial reflexive implicit reference requires specific
knowledge about the association of a statement with a policy, as well as its
positioning within the policy to derive the Act referred to in general, but
more importantly, the positioning within the specific part. The second
example, in contrast, requires only limited knowledge to resolve the
reference explicitly.13
With the identification of general characteristics of policy references,
specifics relate to the nature of the annotation, which may diverge based
on the statement origin based on type (e.g., policy, contract, agree-
ment) and, reflecting on institutions more broadly, potentially form (e.g.,
written vs. oral).
Speaking to institutions-in-form specifically, the origin and nature of
the document, e.g., an act, deed, treaties, contract, etc., as well as the legal
setting (e.g., private vs. public actors) and governance level (State-level
legislation, international forums), affect the role and status of statements
within a document (e.g., as legally binding vs. merely signaling intent),
but, more centrally, define the structure of the document, such as recog-
nizing standards for the structural organization of acts into Act, Chapter,
Part, Sub-parts, Sections, and Paragraph (see e.g., the Document Drafting
Handbook14 for Federal Agencies in the United States).
13 Supporting the resolution of reflexive annotations, the operationalization discussed in

Chapter 7 encourages the annotation of encoded institutional statements with information
about the document structure they are embedded in (e.g., Section reference).
14 The latest version of the Document Drafting Handbook is available under https://
www.archives.gov/federal-register/write/handbook.
Looking at specifications and conventions associated with document

drafting, i.e., the production side, a set of standards have emerged for the
handling of archival information based on bibliographical reference stan-
dards, such as the Uniform System of Citations for legal texts, commonly
referred to as the “Blue Book” (Harvard Law Review Association, 2015),
or, representing the initiatives to foster machine-readable interpretation,
the Functional Requirements for Bibliographic Records (FRBR) (Madison,
2005). The latter is a referencing format used in other electronic stan-
dards for document annotation such as Akoma Ntoso (Palmirani et al.,
2018) that devises an entity-relationship model to represent document
linkages across hierarchical levels and document types.
Specific forms of referencing are highlighted in Chapter 7, without
narrowly imposing a specific general form of referencing, due to the
divergent forms of document structures, as well as established efforts
in this direction as outlined above. Corresponding conventions will
be collated in supplementary resources such as the IG 2.0 Codebook (and
continuously revised therein), but the broad and open use cases for the
IG invite for further exploration based on empirical applications. Further
exploration is particularly applicable in the context of institutions-in-
use, in which such linkages may more commonly be implied, be it
linguistically, or may at times only be inferred probabilistically, such
as signaling relationships between institutions based on structural simi-
larity or alignment on statement or component level, rather than any form
of explicit reference. This handling of references in institutions-in-use
specifically renders a potential opportunity for methodological exten-
sions (e.g., interpretation of implied references in interview or survey
settings), or the application of statistical and computational techniques
(e.g., machine learning).
Concluding the discussion, an essential aspect of the statement link-
ages is the resolution of linkages in parseable and consistent form. This
includes retaining the nature of relationships as either direct relationships
between statements, or relationships between individual statements and
collections (e.g., sections parts, Acts entirely). As indicated, references
may be implicit or explicit reflexive (i.e., reference statements within the
document), or external to the document the referencing statement is
embedded in. Essential for the parseable linkage, however, is the explicit
resolution of implicit linkages in order to afford the interpretation from
within an institutional statement (as unit of analysis), without relying on
extended contextual inference.
The value of resolving linkages between institutional statements relates

to the open-ended scoping of analysis of institutional designs across provi-
sions, e.g., to reference aspects that are externally defined, recognize
amendments of substitutions of documents, etc. Reflecting on policy
linkages endows the analysts with a mechanism to explicitly decide on
the scope of the study, i.e., deciding whether external policy linkages
reflect the boundary of analysis, or the extent to which the referenced
information is considered in the analysis.
Observed from a systemic perspective, the resolution of statement
relationships inevitably establishes opportunities to operationally link
institutions as part of the analysis, drawing inferences about the integrity
of a given set of statements, or conversely, its linkages, and of course
interdependencies with other institutions. Such analysis, in contrast to
most other concepts introduced in the context of IG Logico, can shift the
perspective from individual statements to the analysis of linkages between
institutions, and networks or systems of institutions more broadly –
potentially steering the attention away from semantic content toward
large-scale structural features of institutional design.
6.1.4 Statement Transformation Rules

The extended annotation and formalization of structurally deeply parsed
statements as part of IG Logico provides us with the immediate analytical
benefit not only to afford richer semantic interpretation of statements
with respect to their institutional semantics, but further allows us to
explicate selected logical transformations of institutional statements that
modify the form of the institutional statements without modifying the
underlying meaning.
IG Extended afforded the encoding of institutional statements in detail
by capturing nested structures across statements and inadvertently intro-
duced structural specificity that extracts institutional content. In this
process, it does, however, fundamentally retain the structure of the insti-
tutional statement in its parsing. Introducing a formal representation of
institutional statements, IG Logico is in the position to afford structural
transformations that can be selectively applied to allow for a uniform
expression of institutional statements (e.g., to facilitate analytical treat-
ment), make institutional meaning overt, but also to rephrase institutional
statements in ways that make them amenable for logical evaluation and
identification of logical consequences, let alone readability. This supports
the intent to gain accessible insight into the institutional content, but
also serves as a starting point for assessing the quality of institutional
statements.
To this end, this section introduces a set of statement transforma-
tions that, in conjunction with the formal semantics of institutional
statements (introduced in Sect. 6.1.1) and the introduction of semantic
component annotations (introduced in Sect. 6.1.2), provide the basis for
advanced analytical approaches discussed further in Chapter 8 that draw
on statement transformations specifically.
As part of this section, we initially introduce elementary syntactic trans-
formations in part to formalize concepts introduced intuitively previously,
but also to shed light on the variable forms in which statements can
be constructed while carrying identical institutional content, on which
analytical applications build.
6.1.4.1 Combination-Level Component Transformation

A transformation implicitly motivated as part of Sect. 4.2.2 is the
reconstruction of logically combined components within institutional
statements. The common occurrence of multiple logically linked actors,
activities, objects, and context characterizations prescribed in institutional
statements translates into a corresponding number of atomic institutional
statements.
The (intuitively annotated)15 statement
(Certified agents [OR] inspectors) may (review [OR] sanction) certified

operations.
effectively decomposes into four16 separate atomic institutional state-

ments, variably emphasizing divergence on actorship and activity,
resulting in the following statement linkage:
15 Stricter approaches for the comprehensive annotation of institutional statements are

presented in Chapter 7.
16 The number of statements can be determined using ac cc , where ac represents the
number of alternative choices per component (2), and cc the number of components that
hold such combinations (i.e., Attributes and Aim), i.e., 22 for this case.
“(Certified agents may review certified operations [OR] Certified agents

may sanction certified operations) [OR] (Inspectors may review certified
operations [OR] Inspectors may sanction certified operations)”
Drawing on the symbolic representation of individual components

introduced in Sect. 6.1.1, we can abstractly express the Component-Level
Combination Transformation as shown in Eq. (6.6), exemplified for the
Attributes component, and combinations of two components.
(6.6)
As indicated above, this transformation makes the already discussed

component-level linkages explicit; the transformation naturally applies for
all components of institutional statements, and equally regulative and
constitutive form.
6.1.4.2 Conditions-Consequence Transformation

Beyond the statement expansion offered by the Component-Level
Combination Transformation, the Conditions-Consequence Transforma-
tion is directly derived from the antecedent-consequent linkages opera-
tionalized via Activation Conditions, and in varying form linked in the
Or else component.
While the linkage between Activation Condition and leading institu-
tional statement is congruent in that the enactment of the antecedent
(i.e., the precondition) leads to the application of the remaining part of
the institutional statement (i.e., if-then relationship), the linkage between
the monitored institutional statement and consequential statement (the
statement linked via Or else) inverts the relationship (i.e., if not-then rela-
tionship) to represent the concept of institutional consequences associated
with the non-fulfillment of the specifications (i.e., behavioral regulation or

parameterization) offered as part of the leading monitored institutional
statement.
We can draw on the following statement as a motivational example:
Program Manager may suspend certification if certified organic farmer

violates organic farming provisions.
The directly parsed statement has the following structure:
The activation of the leading statement “Program Manager may

suspend certification ...” is contingent on the farmer’s violation. Impor-
tant to note here is that the focal entity of this institutional statement is
the Program Manager.
Alternatively, the statement can (with minor accommodations in
expression, and added elements held in square brackets) be reconstructed
as
Certified organic farmers [must not] violate organic farming provisions, or

else Program Manager may suspend certification.
Parsed visually, the reconstructed statement reads as follows:

This statement, originating from a positive antecedent-consequent

linkage, can thus be expressed with focal emphasis on the actor central
to the (previous) Activation Condition, the Certified organic farmer.
Reflecting the inverse consequential linkage afforded by vertical
nesting – the decomposition into a monitored and consequential state-
ment – can thus be made explicit by inverting the leading (then
monitored) statement’s Aim or Constitutive Function component.
This transformation leaves us with distinctive benefits. As a central
aspect, it provides the basis for greater abstraction from stylistic
features (e.g., writing style) encoded in institutional statement structure
(beyond the affordances introduced by IG Extended). More importantly,
however, and based on the structural organization (e.g., dedicated policy
section related to enforcement) or writing conventions and preferences,
writers may selectively construct punitive or incentivizing provisions
from the perspective of the monitor or enforcer, leading to a represen-
tation of monitored statements in terms of Activation Conditions (as
highlighted in the example above). The idiomatic form introduced in
the original IG by Crawford and Ostrom, in contrast, seeks the repre-
sentation of consequences from the perspective of the regulated actor,
associating any sanctioning provision with the Or else component. The
transformation thus allows reconstruction of institutional statements in
the idiomatic (i.e., intended) IG form, encoding institutional conse-
quences using vertical nesting, and recovering the semantic specificity the
Or else operator offers with respect to the representation of consequences.
The logical linkage via Activation Conditions is more general in kind,
not only including institutional consequential linkages, but any form of
precondition, including references to statement context more generally, as

well as context not expressed in terms of nested institutional statements
(i.e., atomic values, such as “at 8am”, but similarly across the categories
referenced in the Context Taxonomy in Sect. 5.1.6).
Supporting the call for ontological consistency (see Sect. 3.1), this
transformation thus affords the reconstruction of institutional statements
with greatest possible specificity – in a form that accurately reflects the
semantics underlying the specific components. Embedding this consider-
ation as part of the encoding process, resulting encoding may further bear
methodological benefits by rendering improved reliability. It is impor-
tant to note, however, that this transformation affords the inference of
an implicit Deontic (here ‘must not’) not present in the original struc-
tural form. This, on the one hand, presents itself as a source of reliability
concerns, but may furthermore raise epistemological considerations, since
it affords an explicit inference of information not present in the orig-
inal institutional data. This may lead researchers to raise potential validity
concerns around such reconstruction, e.g., based on conflicting disci-
plinary methodological conventions on content analysis. However, such
transformation may not necessarily need to find application as part of the
coding, but be performed in a separate step following the initial encoding.
Motivations for this decomposition may be to improve methodological
rigor based on reduction of task complexity and cognitive load on coders
(e.g., based on coder characteristics discussed in Chapter 7). Alterna-
tively, such transformations can be applied to existing, already encoded
datasets, with the intent to accommodate specific or novel analytical
objectives. As such, and where applicable in the first place, statement
transformations require explicit methodological consideration, but can
allow for flexible positioning in the research process.
Dissociating the generic encoding of institutional statements, transfor-
mations can be applied to respond to specific analytical objectives, and
as such can, for instance, afford uniform representations of institutional
statements that serve specific analytical techniques (e.g., uniform social
network relationship data), but conversely can be applied to reconstruct
statements from the perspective of different actors (perspective extrap-
olation) in order to reconstruct actor-specific characterizations of, or
perspectives on, institutions, an aspect discussed further in Chapter 8.
Equation (6.7) formalizes the transformation process, both for regu-
lative and constitutive statements. Referencing the regulative form, a
statement’s Activation Condition (i.e., the atomic statement Cac,2 {A1 D1
I1 Cac,1 })17 can be reconstructed in inverted form (i.e., A1 D1 ¬I1 Cac,1 )

and be the monitored statement, with the original leading atomic state-
ment assuming the role of the vertically nested consequential statement
(i.e., O{A2 D2 I2 Cac,2 }).
(6.7)
It is important to note that this transformation only applies for state-

ments that have one or more nested Activation Conditions.18 As a result,
consequential relationships can always be converted into condition-based
linkages, but not all conditional statements can be reformulated in an
institutional consequential sense,19 rendering the latter (as motivated
above) with greater specificity.
Drawing further attention to those concerns, it is important to empha-
size that the applicability of transformations needs to be assessed from an
institutional perspective, since it bears the risk of introducing logical falla-
cies – given the particular nature of the Or else linkage. The fact that a
transformation may in principle be possible is not sufficient to assume its
applicability. Rather, it requires consideration of whether the transforma-
tion makes sense institutionally. For instance, if the Activation Condition
exclusively references environmental features (e.g., time, location), the
activation of the non-context part of the institutional statement is a logical
consequence, but does not signal an institutional consequence.20 In
essence, while physical or brute laws may be representable in conditional
form, such linkage may not have an institutional correspondence and
hence not allow for the inference of a consequential relationship (see
discussion of brute versus institutional facts in Sect. 4.1.2).
17 Recall that D can be tacit when expressed in the Activation Condition.

1
18 Application cases for multiple nested statements will be explored in Chapter 8.
19 The different forms of consequences explicitly supported in the consequential
statement are discussed in Sect. 4.2.5.6.
20 For example, the transformation of the statement “At 8am, parents must bring
children to school” would lead to the intuitively accessible, but institutionally non-
sensical conclusion that “It must not be 8am, or else parents will bring children to
school” by implying that this particular institution governs the concept Time.
6.1.4.3 Properties-Conditions-Transformation
A further transformation linkage embedded in the semantics of insti-
tutional statements requires the focus on Properties associated with
distinctive components. The Attributes Properties, if expressed in terms
of the structural form of institutional statements themselves, can equally
transform into Activation Conditions, and thereby simplify the Attributes
characterization, while at the same time concentrate the preconditions for
a statement’s applicability in the Activation Conditions component.
As with the previous transformation rules, this is best explored in
the context of an example, before formally presented in Eq. (6.8).
The example “Program Manager who believes that certified operations
violate organic farming provisions may initiate suspension proceedings”
essentially embeds the precondition for enacting the wider statement
in the Attributes Properties, (i.e., “who believes that certified operations
violate organic farming provisions” ) but could conversely be written
as “If the Program Manager believes that certified operations violate organic
farming provisions, Program Manager may initiate suspension proceed-
ings”. This principle equally holds for simple properties (e.g., “Citizens
older than 18 years may vote in federal elections.” translates into “If
citizens are older than 18 years, they [i.e., citizens] may vote in federal elec-
tions.” ). Equation (6.8) formalizes the transformation, in correspondence
with the narrative above, for Attributes Properties. While displayed here
for regulative statements, the transformation equally applies to constitu-
tive statements.
(6.8)
Similar to the previously introduced transformations, the reconstruc-

tion offered in this form leverages analytical opportunities, both related
to the uniform construction of institutional statements, but more impor-
tantly, as a basis for its combined application with the previously intro-
duced Conditions-Consequence-Transformation that operates on Activa-
tion Conditions and consequences expressed in the Or else component.
The transformations introduced at this stage showcase the semantic

integrity offered by the construction of institutional statements in terms
of structural patterns that can be flexibly recombined without sacrificing
the underlying institutional meaning. More importantly, it highlights the
essential objective of the Institutional Grammar, the abstraction from the
linguistic representation of institutional information, whether in written
or oral form. The semantic representation of institutional statements, as
motivated throughout this chapter, points to the analytical opportunities
based on applied techniques and study design that can draw on features
such as the combined use of transformation rules with semantic anno-
tations, both of which rely on the underlying consistent interpretation
and representation of institutional statements. Illustrative applications that
draw on, extend, and combine these transformation rules are presented
in Sect. 8.3.
6.1.5 Summary of Chapter Contents

The introduction of the transformation rules concludes the introduction
of IG Logico as the highest level of expressiveness of the Institutional
Grammar, and IG 2.0 specifically. It builds on the basic component struc-
ture derived from Crawford and Ostrom’s original syntax, which has been
ontologically refined and augmented with a complementary constitutive
syntax as part of IG Core. The deep structural parsing afforded by IG
Extended – with specific focus on embedded structural features such as
nested components and conceptual relationships between components
and properties – provided the access to all structural features of an institu-
tional statement. IG Logico completes this effort with a consequent shift
to a semantic perspective, extracting the essential institutional content by
introducing abstractions from the underlying structural features, affording
the selective extraction of generic and specific features based on semantic
annotations backed by extensible taxonomies, as well as introducing refer-
ential linkages to external statements. The statement transformations
introduced as the final conceptual feature facilitate the systematic recon-
struction of institutional statements as the strongest abstractions from the
underlying linguistic expression, while, at the same time, preparing the
IG for novel analytical applications (introduced in Chapter 8) that afford
varying representations to respond to advanced analytical necessities and
establish corresponding opportunities.
In addition to contributing semantic soundness to the IG and estab-

lishing the basis for the logical treatment of IG-coded institutional
information, IG Logico acts as an interface aimed at linking the IG
to diverse disciplinary uses, recognizing diverse analytical techniques, a
broad array of domain-specific theoretical frameworks, as well as scenario-
specific features not accommodated by any generic encoding mechanism.
To this end, IG Logico supports the adaptation of a range of features –
most prominently the taxonomies – in order to build the basis for
broad application, while retaining the necessary semantic integrity (e.g.,
ensuring the applicability of transformation rules) and retaining compat-
ibility throughout the levels of expressiveness (i.e., the option to reduce
complexity to lower, more coarse-grained, levels of expressiveness).
6.2 Synthesizing the Institutional Grammar

This chapter concludes the conceptual introduction of the Institu-
tional Grammar, with a specific focus on the Institutional Grammar 2.0
providing the basis for a comprehensive encoding of institutions in terms
of institutional statements under consideration of diverse analytical objec-
tives and techniques. The purpose of this chapter is to go beyond what is,
and to establish opportunities and develop aspirations regarding kinds
of questions and features that can be explored (i.e., what can be) in
institutional design across institutions that vary in “form” and/or “use”,
application realm (e.g., private, public order), level (e.g., local, federal,
international), and moreover, be addressed using different analytical tech-
niques that go beyond currently established practices and approaches.
To this end, the IG, in the presented form, reflects a New Institutional
Grammar, marking a transition in uptake, applications, and paradigms for
encoding and analysis of institutional design. The IG 2.0 is conceptualized
by levels of expressiveness that respond to specific analytical objectives,
and incrementally build on each other’s feature set, with institutional
information encoded at lower levels accessible to analysis at higher levels
of expressiveness. The levels of expressiveness and the associated feature
set is shown in Fig. 6.3, and discussed in the following.
Commencing with the initial level IG Core (introduced in Chapter 4),
the individual components of the IG are presented under consideration of
select refinements to resolve ambiguities identified both in existing work
as well as based on the authors’ and others’ observations – establishing
Fig. 6.3 IG 2.0 Levels of Expressiveness and Associated Features
ontological consistency within and among components. A notable revi-

sion in this context is the dissociated treatment of Activation Conditions
and Execution Constraints, the refined treatment of the Object compo-
nent, and the reconceptualization of the Or else as a semantic linkage of
individual statements (vertical nesting ), as opposed to being a distinctive
component. The identified components systematically combine to form
distinctive types of institutional statements, alongside a novel elementary
statement form, the atomic institutional statement, that captures essential
institutional meaning in its elementary form by decomposing nested insti-
tutional statements – aimed at capturing institutional content at detail,
and operating as the default unit of analysis.
With the same objective, comprehensiveness, in mind, the IG intro-
duces a complementary constitutive syntax that captures the specification
and parameterization of institutional features, in contrast to the regulative
syntax primarily focused on the regulation of existing behavior.
The features introduced at the IG Core level provide the basis for a
coarse structural analysis of institutional statements, primarily targeted as
a basis to establish refined statistical assessment in the tradition of the
original Institutional Grammar (see Chapter 2 for an overview).
IG Extended (introduced in Chapter 5), as the next higher level of

expressiveness, shifts the perspective on institutional statements by moti-
vating the analyst to move beyond a component-centric linguistic inter-
pretation, and to focus on the detection of richer institutional patterns
expressed in the syntactic components of institutional statements, intro-
duced to as component-level nesting, to draw out implied statement
linkages that signal the interdependency of activity in an institutional
setting.
Combining the varying forms of nesting introduced, IG 2.0 provides
the basis for the compositional analysis of structural features of institu-
tional settings, including combining the fine-grained and comprehensive
characterization of activity embedded in institutional statements (hori-
zontal nesting) with the bird’s eye perspective on statement interde-
pendencies that operate across statements, including the referencing of
preceding or embedded statements (component-level nesting), as well as
the linkage of institutional consequences (vertical nesting). With this shift
in perspective, IG Extended provides the foundations for a comprehensive
analytical reconstruction of structural features of the statements, and, in
consequence, the entire institutional setting in compositional form. This
principle is abstractly visualized in Fig. 6.4, with elementary composi-
Fig. 6.4 Structural Composition Patterns for Nested Institutional Statements

tional patterns shown on the left side, and an illustrative composition on

the right side, that further annotates the exemplary structural nature of
the statements.
The structural generalizations afforded by this approach essentially
reflect the Morphology of institutional statements, capturing the type of
the atomic structures of statements, alongside the ways in which they
are linked to capture the form of institutional statements. Capturing
individual statements, and their interlinkages with other institutional
statements, at large, exposes the structure of the entire institutional
setting, an aspect this book will return to in Chapter 8.
In addition to expressing structural information in terms of Institu-
tional State or activities, alongside inter-statement linkages, IG Extended
introduces a richer qualification of context based on the Context
Taxonomy, offering the basis for an extended embedding of statements
in a general context structure. By capturing a wide range of envi-
ronmental embeddings, the IG paves the way for a richer analysis of
context-dependent interpretation of institutional data within and across
datasets.
Combined, the features introduced under the label IG Extended facil-
itate deep structural analysis, providing the basis for the establishment of
advanced complexity metrics that capture the interlinkage of statements,
and their configurational arrangements, alongside advanced network anal-
yses that capture these features. Beyond these incremental refinements,
the analytical shift to compositional patterns enables a constructionist
perspective on institutional analysis as applied in the area of computational
studies, be it to afford richer structural analysis, or to simulate institu-
tional arrangements in artificial societies, either by parameterization or
the endogenous generation of institutional statements (e.g., Frantz et al.,
2015; Smajgl et al., 2008).
The introduction of the structural features captured under the label IG
Extended provides the basis for the introduction of Hybrid Institutional
Statements in Sect. 5.2; statements that combine features of regulative and
constitutive statements introduced throughout the chapter. The concep-
tualization of hybrid institutional statements responds to the surrounding
theoretical debate on the ontological distinctiveness of regulative and
constitutive statements, but more importantly, the empirical need to
capture statements not expressed exclusively in regulative or constitu-
tive form, or whose interpretation is contingent on the applied analytical
lens. The consideration of Hybrid Institutional Statements supports the
comprehensive representation of institutional features, a central objec-

tive of the IG, but further offers a response to the ongoing theoretical
discourse.
The highest level of expressiveness, IG Logico (introduced in this
chapter), moves beyond the structural assessment of statements, and
introduces a semantic layer that revisits encoded information with the
objective to extract not only structural detail (as afforded on IG
Extended), but detailed institutional meaning by systematically anno-
tating encoded components based on general extensible taxonomies
customized to respond to specific analytical objectives. Drawing partic-
ular attention to the linkages of statements within and beyond a given
policy, IG Logico further envisions the use of consistent annotations of
internal and external statement linkages to provide assessments under
consideration of their systemic embedding in the broader institutional
context.
Enabling the logical treatment of encoded information on this
level, formal syntax and semantics are provided, alongside a set of generic
transformations that retain institutional meaning. This affords further
abstraction from the underlying structural representation, an aspect rele-
vant for the logical assessment in the form of reasoning, and for validity
and integrity assessments across varying scopes, ranging from statements
to encoded institutional information in entirety (e.g., policy).
As for IG Extended, IG Logico works toward comprehensive catch-
ment of institutional settings, but extends to the establishment of compu-
tational tractability, affording the New Institutional Grammar transfor-
mational opportunities in the analysis of institutional design. IG Logico
accommodates this by preparing the IG for a wide range of analytical use
cases that rely on the tailored use of the Grammar, an aspect explored
further below.
Fig. 6.5 provides an integrated overview of the central concepts of
IG 2.0, organizing its features by components and syntactic forms of
statements across all levels of expressiveness. Appendix A complements
this with a semi-formal structural overview of institutional statements,
reflecting the variable forms (combinations, atomic) and kinds (regulative,
constitutive) of institutional statements and their structural linkages.
With the introduced concepts, the analyst may seek guidance on the
selection of features pertinent for a given study, but may also be left
with the impression that the individual features introduced as part of
Fig. 6.5 Institutional Grammar Overview
the IG are invariably organized by and linked to levels of expressive-

ness. The first concern, the need for methodological guidance, will be
addressed in the upcoming Chapter 7, which includes guidelines for the
general planning of studies using the IG, alongside considerations related
to analytical objectives, selection of relevant IG features, and preparation
and execution of the operational coding (see Chapter 7).
Responding to the need to customize the IG feature set, the IG 2.0
further includes principles, referenced as IG Profiles, that allow for the
accommodation or configuration of specific feature compositions. Inas-
much as the features are organized by levels of expressiveness, IG Profiles
act as aides to group features for analyses that do not follow stereo-
typical objectives, but instead require a flexible composition of features
that capture structural and semantic information of institutional design.
Reasons can be manifold, including varying study objectives and tech-
niques and available institutional data, and may inadvertently lead to the
need to selectively consider or forego features associated with a given
level of expressiveness. This can include the entire omission of compo-

nents from the analysis (e.g., if contextual information or consequences
are of limited concern), as well as the selective expansion of the feature
set by drawing on features introduced on higher levels of expressiveness.
Studies that apply the parsing of institutional statements on the IG Core
level could, for instance, selectively draw on the context characteriza-
tions based on the Context Taxonomy without adopting the remaining
features offered by IG Extended (which the Context Taxonomy is associ-
ated with).21 Alternatively, studies that operate on the IG Extended level
may selectively include features from IG Logico (e.g., statement refer-
ences), while, at the same time, ignore select features of IG Extended
(e.g., the Object-Property Hierarchy). The IG 2.0 Codebook (Frantz &
Siddiki, 2020) provides a comprehensive overview of IG features and
details the tailored specification of specific feature configurations.
While the tailoring of the IG to study design objectives is of central
concern, the accurate characterization of the flavor of IG is equally impor-
tant to ensure the replicability of the parsing and further treatment of
institutional information. Given the broad feature set, a precise feature
description is further important as a means to establish the principal
comparability (or lack thereof) of datasets originally encoded for different
analytical objectives in mind, e.g., for the purpose of comparative studies,
or to support machine coding efforts.22
This conceptual introduction of the IG, and specifically in its incarna-
tion as IG 2.0, provides a rich set of features that show the variable forms
in which the Institutional Grammar can extract institutional information
on a structural level in the form of institutional statements, or selectively
apply deep structural parsing to leverage institutional information that
serves advanced analytical objectives, such as the reconstruction of insti-
tutional settings and their structural analysis. Establishing computational
tractability, while drawing the semantic linkage to concepts embedded
in the analyst’s domain positions this refined, then, New Institutional
21 Such feature composition would be characterized as IG Core+C, with further details

discussed in the IG 2.0 Codebook (Frantz & Siddiki, 2020).
22 Considerations related to objective-agnostic encoding of institutional statements has
found brief discussion in Sect. 5.2 and will be further discussed in Chapter 7. The conclu-
sion in Chapter 9 will further reiterate the importance of rigid documentation of process
and data structures.
Grammar as a theoretically integrated concept that can serve as an inter-

disciplinary interface to facilitate the interaction of researchers of different
backgrounds based on a shared conceptual basis the IG provides. At
the same time, it can support novel analyses in areas that study institu-
tions in form and institutions in use, and their dynamics (based on the
harmonized representation of institution types). The layered, feature-rich
annotation afforded by the IG further enables the analysis of encoded
institutional information by means of different techniques, including ones
that may not have found application to date.
Following the comprehensive introduction of the IG in this and the
preceding chapters, Chapter 7 provides general methodological guid-
ance relevant for the design of studies applying the IG, and further-
more provides the analyst with an introduction into operational coding
principles. With the provided outlook, this chapter concludes the Foun-
dations part of this book, with the remaining chapters turning to the
Application of the reviewed concepts. As an initial step toward this
end, the following chapter introduces an accessible syntactic representation
that complements the primarily semantic focus emphasized throughout
this earlier part of the book. Chapter 8 then turns to the exposition of
analytical opportunities associated with the Institutional Grammar.
References
Baker, C. F., Fillmore, C. J., & Cronin, B. (2003). The structure of the framenet
database. International Journal of Lexicography, 16(3), 281–296. https://doi.
org/10.1093/ijl/16.3.281
Boella, G., van der Torre, L., & Verhagen, H. (2006). Introduction to normative
multiagent systems. Computation and Mathematical Organizational Theory,
12(2–3), 71–79. https://doi.org/10.1007/s10588-006-9537-7
Bodansky, D. (2016). The legal character of the Paris agreement. Review of Euro-
pean, Comparative and International Environmental Law, 25(2), 142–150.
https://doi.org/10.1111/reel.12154
Bradner, S. (1997). Key words for use in RFCs to Indicate Requirement Levels.
Internet Engineering Task Force, RFC 2119.
Brady, U. (2020). Robust Conservation Anarchy: Comparing Treaty Institu-
tional Design for Evidence of Ostrom’s Design Principles, Fit, and Polycentricity
(Doctoral dissertation). Arizona State University, Tempe.
Carlson, G. N. (1984). Thematic roles and their role in semantic interpretation.

Linguistics, 22(3), 259–280. https://doi.org/10.1515/ling.1984.22.3.259
Ceci, M., Butler, T., O’Brien, L., & Al Khalil, F. (2018). Legal patterns for
different constitutive rules. Lecture Notes in Computer Science (including
subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioin-
formatics), 10791. https://doi.org/10.1007/978-3-030-00178-0_7
2975
Fillmore, C. J. (1968). The case for case. In Bach & Harms (Eds.), Universals
in linguistic theory (pp. 1–88). Holt, Rinehart, Winston.
Harvard Law Review Association. (2015). The Bluebook: A uniform system of
citation (20th ed.). Harvard Law Review Association.
Hohfeld, W. N. (1913). Some fundamental legal conceptions as applied in judi-
cial reasoning. The Yale Law Journal, 23(1), 16–59. https://doi.org/10.
2307/785533
Katz, D. M., Dolin, R., & Bommarito, M. J. (Eds.) (2021). Legal informatics.
Lowi, T. J. (1964). American business, public policy, case studies, and political
theory. World Politics, 16(4), 677–715. https://doi.org/10.2307/2009452
Madison, O. M. (2005). The origins of the IFLA study on functional require-
ments for bibliographic records. Cataloging and Classification Quarterly,
39(3–4). https://doi.org/10.1300/J104v39n03_02
McNamara, P. (2006). Deontic logic. Handbook of the History of Logic, 7 , 197–
288. https://doi.org/10.1016/S1874-5857(06)80029-4
Palmirani, M., Sperberg, R., Vergottini, G., & Vitali, F. (2018). Akoma Ntoso
Version 1.0 Part 1: XML Vocabulary, OASIS Standard. http://docs.oasis-
open.org/legaldocml/akn-core/v1.0/os/part1-vocabulary/akn-core-v1.0-
os-part1vocabulary.html
Schneider, N., Prange, J., Blodgett, A., Hwang, J. D., Moeller, S. R., Srikumar,
V., Stern, A., Bitan, A., & Abend, O. (2018). Comprehensive supersense
disambiguation of English prepositions and possessives. ACL 2018 - 56th
Annual Meeting of the Association for Computational Linguistics, Proceedings
of the Conference (Long Papers), vol. 1, 185–196. https://doi.org/10.18653/
v1/p18-1018

in Complex Systems, 2(11), 199–215. https://doi.org/10.1142/S02195259
0800157X
Staab, S., & Studer, R. (Eds.). (2009). Handbook on ontologies. Springer. https://
doi.org/10.1007/978-3-540-92673-3
von Wright, G. H. (1951). Deontic logic. Mind, 60(237), 1–15. https://doi.
org/10.1093/mind/LX.237.1
1111/psj.12139
Weible, C. M., & Sabatier, P. A. (Eds.). (2018). Theories of the policy process
(4th ed.). Routledge. https://doi.org/10.4324/9780429494284
CHAPTER 7
Methodological Guidance for Encoding

Institutional Information
7.1 The IG Coding Process:

Planning, Execution, Assessment
Following the extensive review of the core features of the IG 2.0 in the
preceding chapters, this chapter turns to methodological aspects related
to the IG Coding Process. The coding process is described as consisting
of three stages: (i) Planning; (ii) Execution; and (iii) Assessment. Gener-
ally, the description provided in this chapter walks the institutional analyst
through the process of transforming institutional data into IG parsed
institutional information which can subsequently be analyzed using the
analyst’s technique of choice, and in accordance with his/her analyt-
ical objectives. Institutional data are those captured in institutions as
they normally exist in their respective domains. For example, public
policies which have undergone no processing or coding are considered
institutional data, as are written transcripts of descriptions by commu-
nity members of conventions that govern their behavior within a given
domain. Institutional information, in the context of an IG application,
is institutional data which have been processed and coded in accordance
with the IG (i.e., by institutional statements and syntactic components).
The final step involves an assessment of the encoding based on reliability
tests so as to establish consistency and quality of the coded institu-
tional information. Below, the individual phases Planning, Execution, and
Assessment of the IG coding process are discussed in turn.

Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2_7
7.2 Planning
The Planning phase of the IG coding process involves the institutional
analyst’s consideration of what and how institutional data will be coded, as
pertinent to his/her analytical objectives. As with any research study, each
institutional analysis is motivated by different analytical objectives that will
inform how data to be analyzed are to be collected, what tools will be
used for data collection, what aspects of the data will be analyzed, and
what data analysis techniques are appropriate for analyzing collected data
given the structure of the data as well as what insights the analyst seeks to
draw from them. For the IG application specifically, key aspects relating
to the identification of analytical objectives include the delineation of the
particular institutional dynamics the institutional analyst is interested in
evaluating as well as consideration of whether these institutional dynamics
will be discerned through an assessment of formal or informal institutions.
7.2.1 Data Collection Considerations

The design of the encoding process crucially depends on the nature and
source of the encoded data, as well as the scope of the data collected.
Associated considerations are discussed in the following.
7.2.1.1 Formal vs. Informal Institutional Data
The means for collecting formal versus informal institutional data can be
substantially different. For the collection of formal institutional data, the
institutional analyst is likely to rely on some form of archival research,
whereas the collection of informal institutional data is likely to entail
primary data collection (e.g., participant observation, interviews). Collec-
tion of the latter may also involve archival research; in cases, for example,
where an institutional analyst seeks to draw understanding of community
practices through a review of documents, field notes, historical accounts,
etc. For simulation studies that leverage institutional data for model
parameterization, the institutional analyst may rely on formal or informal
institutional data collected through the various means described above.
Simulations may also be the source of data collection, insofar as modelers
are often interested in evaluating institutional data that emerge through
the modeling exercise.
7.2.1.2 Scoping Institutional Data
One key consideration relating to the collection of institutional data
is that of institutional scope; or more specifically, consideration of
7 METHODOLOGICAL GUIDANCE … 245
sufficient capturing of institutional data to adequately assess institutional

dynamics of interest. This can be a challenge in the collection of both
informal and formal institutional data. When collecting informal insti-
tutional data, the analyst has to ensure through retrieval of multiple
observations that reported practices are shared among a relevant set of
actors within a particular domain, and that the analyst reaches a point
of “data saturation,” the point at which new institutional data are not
observed among collected observations. The first of these challenges is
ameliorated for the collection of formal institutions, as their “shared”
applicability is evidenced by their enactment and application by a entity
commonly recognized as a governing authority. A different challenge that
arises in the case of formal institutions is that relevant institutional data
may be spread out over a number of different policies. Understanding
of which formal institutions, or parts of formal institutions, the institu-
tional analyst needs to collect to support his/her aims, requires contextual
understanding.
7.2.2 Tool Support

The second aspect of planning is consideration of the tools/techniques
that the institutional analyst will use to generate and analyze institu-
tional information, which may be the same. Current practice tends to rely
on spreadsheet and text annotation software for generating institutional
information, and network analysis and statistical software for analyzing
institutional information.
On the generation side, the use of spreadsheet software enables
the institutional analyst to capture institutional information in tabular
form. Referencing current practice, in spreadsheets, institutional analysts
typically parse institutional statements listed by row by syntactic compo-
nents listed across columns (individual syntactic components are typically
captured in individual columns). This obviously necessitates manual
processing of institutional data (i.e., manually placing institutional data
into columns). The tabular representation of institutional information,
where information corresponding to particular statements and partic-
ular components is stacked in discrete cells, enables the analyst to
easily observe the specific information of which individual statements
are comprised, but also easily observe how specific information across
statements compares.
Text annotation software enables tagging of institutional data
according to IG relevant labels. Unlike spreadsheet-based coding, where
information is placed within different cells, the institutional analyst need

not engage in any manipulation of institutional data to engage in tagging.
Other requisites for tagging information include uploading of institu-
tional data into the annotation software in a software compatible file
format, and the creation of labels according to which institutional data
will be tagged. Generation of institutional information using text annota-
tion software may be more efficient than if using spreadsheet software,
given the lack of need to manually manipulate institutional data into
tabular form, but it does not readily support cross-statement comparisons.
For the purpose of coding institutional statements, specifically, this
book introduces a variant of the latter approach by drawing on a
purpose-built custom tool, IG Parser, alongside companion resources
(e.g., codebook, software), available under https://newinstitutionalgr
ammar.org, that facilitate the process. Details will be discussed in the
context of the operational encoding process.
An important criterion for the form (and potentially means) by which
data are encoded are determined by the analytical use, examples of which
are discussed in Chapter 8.
The brief overview of tools and techniques offered here merely serves
to show the different options used by convention, and variable utilities
thereof, that analysts might consider as they plan to engage in an IG
coding.
7.2.3 Determining applicable Institutional Grammar Feature Set

A third, important, aspect of planning for IG coding, is the identification
of the IG 2.0 levels and features that one will engage. In particular, the
institutional analyst must decide whether coding will be done according
to the IG Core, IG Extended, or IG Logico levels of expressiveness;
acknowledging that even when coding at one level, the analyst can engage
select coding features that associate with other levels (see discussion in
Sect. 6.2, and variable analytical use in Chapter 8). For example, the
institutional analyst may decide to code institutional data at the IG Core
level, but classify institutional information corresponding to the Context
component according to the Context Taxonomy (a feature associated
with coding at the IG Extended level). Similarly, during the Planning
phase, the institutional analyst may also decide to code institutional data
according to select features from within a particular level; for example,
only code Attributes, Aim, and Objects as specified at the IG Core level.
Again, the choice of levels and features to engage will be informed by
the analytical objectives, tools, and techniques the analyst plans to engage
in his/her study. Generally, however, the following summary of features
associated with different levels of expressiveness and related selection
heuristics may help guide the institutional analyst’s choice.1
7.2.3.1 Summary of Key Features by Level

IG Core (Chapter 4) facilitates coding following a fundamental syntactic
structure, capturing in basic form Attributes, Deontic, Aim, Object,
Context (differentiating between Activation Conditions and Execution
Constraints ), and Or else statements in the regulative realm and Consti-
tuting Properties, Modal, Constituted Entity, Constitutive Function,
Context and Or else statements in the constitutive realm. IG Core also
accommodates statement-level nesting (horizontal and vertical) as well
as Hybrid Institutional Statements.2 In practice, this involves parsing
institutional data to the atomic statement level (see Sect. 4.2.2), with
referencing of logical connections among atomic statements, which signal
how one or more atomic statements configure to form composite insti-
tutional statements. This level of encoding is suited for comparatively
simple institutional statements that largely follow the basic regulative or
constitutive structure. Of the three levels of expressiveness, IG Core is
considered to be the least computationally tractable; meaning that the
institutional information that results from coding of institutional data at
the IG Core level offers limited machine-interpretability, limiting most
analytical applications to component-oriented descriptive statistics (see
Sect. 8.1).
Coding at the IG Extended level (Chapter 5) enables the institu-
tional analyst to capture the syntactic structure of institutional state-
ments in greater detail (deep structure). For regulative statements, this
involves the fine-granular encoding of Actors and Objects, along with
complex property relationships. Furthermore, it enables for both regu-
lative and constitutive statements, a detailed encoding of Context, such
as the characterization of statement dependencies, and categorization
based on circumstantial aspects of conditions and constraints (e.g.,
temporal, spatial, procedural aspects). IG Extended also accommodates
1 Where acquainted, the reader may forego the following summary of key features.
2 While these are introduced in the context of deep structural parsing under IG
Extended, this is due to the lack of component-level nesting on IG Core level; the
principles of linking regulative and constitutive statements, however, are equally applicable
on IG Core level (albeit in more coarse-grained form).
component-level nesting, capturing of Object-Property hierarchies, and,

necessarily, hybrid and polymorphic institutional statements. Choosing to
encode on this level may be motivated by the complexity of institutional
data (e.g., complex statements involving embedded object linkages, or
extensive statement interdependencies), but also by analytical objectives,
such as the operationalization of the extracted institutional information in
advanced computational models that require the explicit actor and activity
representations, alongside richer context characterization. This latter
point reflects that IG Extended accommodates some computational appli-
cation, but therein focuses on a largely structural perspective, whereas IG
Logico augments this with a customizable semantic perspective.
Coding at the IG Logico level (Chapter 6) enables the analyst to derive
more sophisticated understanding of semantic relationships embedded
in and among institutional statements based on institutional statement
classification across syntactic components; for example, improved under-
standing of actor roles, explicit references between statements, as well as
inference of actor obligations tacitly expressed in the coded document.
As a point of contrast, whereas at the IG Core and IG Extended levels
syntactic classification of institutional statements is a final goal of the
encoding exercise, at the IG Logico level, the goal is to build on syntactic
classification by leveraging this coding toward the identification of insti-
tutional semantics that relay functional and/or relational information of
interest to the institutional analyst.
7.2.3.2 Heuristics for Feature Selection

Associated with the variable opportunity of coding at different levels of
expressiveness as described above, the following questions and associated
heuristics may be helpful in guiding at which level to code. These ques-
tions and heuristics should not be considered exclusively; coding choices
will be informed by various considerations reflected in different questions
and heuristics.
Heuristic 1: Based on the institutional analyst’s general understanding
of the institution(s) to be coded, are related institutional data observed to
be relatively simple or complex in kind? If institutional data are observed
to follow a relatively simple structural form, coding at the IG Core level
may be appropriate. If more complex, coding at the IG Extended or IG
Logico level may be more appropriate, given that the features associated
with these levels are more amenable to capturing statement complexity.
Heuristic 2: Is the institutional analyst interested in basic or compre-
hensive, and syntactic and/or semantic classification? If a basic syntactic
classification, the coder will likely find an IG Core level coding suitable.
If a comprehensive, syntactic, and semantic classification, the coder will
find an IG Logico level coding suitable. If a comprehensive syntactic clas-
sification and partial semantic classification, an IG Extended level coding
will likely be appropriate.
Heuristic 3: Is the institutional analyst interested in engaging manual
methods (e.g., manual coding or computer-assisted analysis) or computa-
tional methods (e.g., automated classification or agent-based modeling)
in either the generation or use of institutional information for analytical
purposes? If the analyst only intends to rely on manual methods, coding
at the IG Core level will likely be sufficient. If the analyst plans to rely
partially or fully on computational methods, coding at the IG Extended
and IG Logico levels is suggested.
To reiterate, the analyst need not be restricted to coding by one level.
While coding at one level of expressiveness, the analyst may also choose to
code institutional statements along select features corresponding to other
levels. Further, the analyst may choose to only code statements along
with a limited set of features corresponding with one particular level. This
limited coding may entail only coding institutional statements along select
syntactic components (e.g., Attributes, Objects, Aim). Critically, the IG
2.0, in comprehensively outlining coding features associated with three
levels of expressiveness, is specifically designed to accommodate flexibility
in IG coding to support diverse analytical aims.
One additional point worth highlighting relating to the choice of level
at which to code is the backward compatibility of coding at different
levels. Recall that parsing of institutional information corresponding to
syntactic components becomes more granular as one moves from the IG
Core to IG Extended to IG Logico level. That is, institutional informa-
tion undergoes a greater level of decomposition as one moves across these
levels. This means that decomposed information can be recomposed to
capture coding at levels prompting less expressiveness. The practical impli-
cation of this is that institutional analysts engaging a multi-pass coding,
coding at a different level with each pass, do not need to change the
basic syntactic coding of institutional data. Rather, the analyst will simply
further decompose and classify information linked to the basic A-D-I-
B-C/E-M-F-P-C syntactic components (referenced in idiomatic symbol
order) when moving from an IG Core to an IG Extended and/or IG
Logico coding. Alternatively, the analyst might merely concatenate (i.e.,
combine) information within specific syntactic categories to reflect coding
at a lower level of expressiveness.
7.2.4 Determining Data Processing Practices

During the planning phase of the coding process, the institutional analyst
may also choose to engage in more detailed considerations about how
institutional information will be captured as consistent with his/her aims,
tooling choices, and intended analysis. One overarching consideration,
which has several more specific considerations, is how closely institutional
information should match institutional data. For several reasons, coded
information may not correspond exactly to institutional data. One of these
reasons is that institutional data may contain implied information which
the analyst may wish to explicitly recognize when generating institutional
information.
7.2.4.1 Inferred/Implied Information

Take, for example, the following two statements, which may appear
consecutively in a policy document.
Farmer must submit an organic system plan.

In addition, a copy of the organic system plan should be retained.
The consecutive presentation of these two statements in a policy would

support the inference that the Attributes component of the second state-
ment is “Farmer,” just like in the first statement. Thus, when generating
institutional information (i.e., coding) for the second statement, the
analyst may wish to imply “Farmer” as the Attributes component with
some notation indicating that this is implied information. In practice,
the implication of information – across syntactic components is common.
However, the choice of coding tool may affect the possibility of doing so.
Some text annotation tools, for instance, only allow the analyst to label
information, not to manipulate text.
7.2.4.2 Handling Stylistic Features

Other times, the institutional analyst may wish to make accommoda-
tions to account for stylistic qualities of institutional data. For example,
use of plurals for types of actors (e.g., reference to “farmers” versus
“farmer” ) or selective stemming of information (e.g., “Organic Program
Official” versus “Official” ). In accordance with the analyst’s aims, it may
be useful to capture institutional information in singular/plural form,
stemmed/unstemmed, or lemmatized, form, even if this departs from
its presentation in institutional data. Finally, the analyst may choose to
eliminate prepositions or stop words (common terms that do not carry
distinctive meaning, such as “a”, “the” ) as institutional data are coded.

The latter two aspects may be of particular concern for analysts engaging
in computational methods (e.g., social network analysis, agent-based insti-
tutional modeling), but should in any case be carefully considered and
documented.
7.2.4.3 Wide or Narrow Semantic Interpretation
Yet another consideration that an institutional analyst may wish to enter-
tain during the planning phase is whether a wide or narrow perspective
will be adopted in statement coding. Essentially, with this consideration,
the institutional analyst is deciding to what extent broader institutional
context will be taken into account during the coding. This is of partic-
ular relevance when encoding constitutive statements, since the resolution
of implied statement linkages may lead to varying characterizations, and
correspondingly, coding, and ultimately hinges on shared knowledge
of context and coder experience. Details are discussed in the IG 2.0
Codebook (Frantz & Siddiki, 2020).
7.2.5 Coder Selection

A final aspect of the planning phase of the coding process we reference
is decision making regarding who will be engaged in coding, and relat-
edly, how many coders will be engaged. The institutional analyst will want
to ensure that coders have the requisite knowledge and experience to
engage in coding. As well, when multiple coders are engaged, the analyst
will want to make sure that mechanisms for determining, evaluating, and
responding to concerns about intercoder reliability are identified. Inter-
coder reliability is discussed in more detail in later parts of this chapter
(see Sect. 7.4). Especially where novice coders or coders with variable
experience/background are involved, initial pilot coding exercises should
be considered.
7.3 Execution
The previous section outlined various points of consideration for institu-
tional analysts as they prepare for coding. This section describes the actual
execution of coding.
7.3.1 Data Selection

A first, and critical, step in the coding process is the review of institu-
tional data that will be coded. Institutional data represented in policies or
written accounts of social conventions should be comprehensively read to

ensure that the institutional analyst has a baseline understanding of insti-
tutional content and context, before institutional data are parsed at the
statement and syntax level. This baseline understanding is crucial for the
analyst’s ability to discern how institutional information represented in
different institutional statements connects. Among the utilities of this is
the improved ability to infer information where the implication is required
to complete institutional statement coding.
7.3.2 Coding Platform Selection

A second step in the coding process is final decision-making regarding
the platform that will be used for coding. It is recommended that this
decision-making commence during the Planning phase. However, the
analyst may be prompted to make more specific decisions during this
step about which and how platforms will be engaged and/or designed
to support coding, following the data familiarization encouraged as part
of Step 1.
7.3.3 Preprocessing Institutional Data

A third step in the coding process involves organizing and basically
manipulating institutional data to make it more amenable to coding,
and in accordance with the institutional analyst’s study objectives. This
step is generally referred to as preprocessing. The organization part
of preprocessing entails an initial segmentation of institutional data by
section, theme, or other institutionally relevant demarcation, and by the
institutional statement.
7.3.3.1 Document Preprocessing

For public policies, this initial segmentation typically involves demar-
cating institutional data by the preamble, section, subsection, etc., and
then conducting an initial identification of institutional statements within
different sections of the document. For written transcripts of social
conventions, this initial segmentation might entail demarcating institu-
tional data by type of activity of thematic focus, and then conducting
an initial identification of institutional statements that share a common
activity or thematic focus. Here the coder is advised to acquaint
him/herself with conventions about document structure as pertaining to
the field of study (an aspect that has been discussed in Sect. 6.1.3).
7.3.3.2 Statement Preprocessing

For the initial identification of institutional statements, the analyst will
partition institutional data noting generally the presence of particular
content that corresponds to different syntactic components. A more
careful parsing with finer attention to syntactic information will be
conducted as part of the next coding steps. Generally, the analyst is
looking for clauses that contain at least Attributes-Aim-Context compo-
nent information or Constituted Entity-Constitutive Function-Context
component information that can be initially identified as institutional
statements, which may signal a composite institutional statement, but
where component-level combinations exist (e.g., two logically linked
activities), may also indicate multiple atomic institutional statements,
or variably two linked statements that express different content (actor,
activity, etc.) entirely.
Institutional analysts working with formal institutional data (e.g.,
public policies) may find that sentences map to institutional statements,
but this may not always be the case. Multiple statements can be present
within a sentence, for example. Analysts working with informal institu-
tional data (e.g., interview transcripts documenting social conventions),
may find relatively few expressions of complete institutional statements,
since oral speech often contains phrase fragments and elements of personal
conversational style that may need to be sorted. The practical implication
of this is that the institutional analyst may need to start implying informa-
tion for necessary components to generate complete statements, as well
as engage in additional preprocessing to account for stylistic features of
the data. Given the nature of data, and these associated tasks, it is critical
that the analyst has an understanding of the institutional setting.
Basic manipulation of institutional statement information during this
step entails removal of extraneous information from initially parsed insti-
tutional statements (e.g., removal of stop words as referenced above)
as well as a simple modification of institutional information to account
for punctuation features therein that may interfere with subsequent
institutional coding. Some punctuation may hinder accurate parsing on
institutional data by computers.
Some level of preprocessing is likely to be required for either manual
or automated coding; in any case, preprocessing can actually make coding
easier. Preprocessing can make coding easier insofar as the coder may not
have to accommodate in his/her coding unneeded or unwanted data or
stylistic qualities of institutional data that may challenge coding. Again,
identification of what will be considered useful data and how stylistic

features will be handled is urged during the Planning phase, once the
institutional analyst has collected institutional data and thus has a sense
of types and styles of content captured therein. Generally, easing the
coding process through preprocessing yields at least two notable bene-
fits. It can reduce the cognitive burden and time needed to engage in
manual coding, and it can improve accuracy and by extension reliability
of manual or automated coding.
A fourth step in the coding process is the verification of institu-
tional statements initially identified as part of step 3 to support syntactic
decomposition. Verification in this case means ascertaining that candidate
statements accord with defining syntactic and semantic features of regu-
lative and constitutive statements. Principally, this means verifying that
statements presumed to be regulative in kind at least contain Attributes,
Aim, and Context, and that statements presumed constitutive in kind
at least contain Constituted Entity, Constitutive Function, and Context
components. The coder may find it useful to begin labeling verified insti-
tutional statements as regulative or constitutive, or polymorphic, insofar
as the delineation can cognitively queue the coder to engage a particular
syntax for statement decomposition. The analyst may also wish to label
statements as monitored and consequential statements reflecting instances
of vertical nesting (see Sect. 4.2.1.6). To enable both of the tasks refer-
enced above – statement verification and labeling of statement type – the
institutional analyst can consider the following general heuristics (the IG
2.0 Codebook (Frantz & Siddiki, 2020) offers refined heuristics for the
distinction between statement types).
Constitutive Statement Heuristic: Does the statement introduce or
parameterize fundamental aspects of the action situation (bound-
aries, definition or modification of actors, actions, objects, artifacts,
and associated affordances; endowment of status, such as rights or
authority/power), and in doing so, define the positioning and constel-
lation of entities in an institutional setting in which potentially regulated
behavior is enacted? → If so, the statement is constitutive in kind.
Regulative Statement Heuristic: Does the statement signal behavioral
guidance for an unambiguously implied or explicitly identified actor by
specifying duty and discretion, alongside potential sanctions for trans-
gression? In doing so, does the statement draw on (i.e., makes implicit or
explicit reference to) actors, actions, objects, or artifacts in an institutional
setting? → If so, the statement is regulative in kind.
Practical Considerations
Practical considerations beyond the identification of individual statements
pertain to establishing standardized statement patterns irrespective of
the representation in the underlying institutional data. To this end, the
following practical considerations apply:
Statements may be expressed from the perspective of an enforcing
entity (e.g., Official may administer sanction if driver …). In such cases
practical guidelines should consider – either as part of the preprocessing,
or the operational coding – whether such statements are to be recon-
structed to be expressed in consistent form, e.g., from the perspective of
the monitored entity (e.g., Driver must not violate, or else official …) –
thus reflecting idiomatic Or else linkages of consequences. Alternatively,
the enforcement perspective can apply. Specifically where analytical appli-
cations benefit from uniform representation (e.g., computational use), or
even both forms (see Sect. 8.3.1 for operational details), a corresponding
guidance should be considered. Details on the associated transformation
is provided under Eq. (6.7).
Another important decision is the division between the entity and
property characterization in compound expressions. The “certified opera-
tion” can conceivably be treated as a compound entity, or be decomposed
into core entity (“operation” ) and associated property (“certified” ). While
the distinction is subject to the specific study, general considerations exist.
Decomposition is in principle desirable, e.g., to differentiate between
“certified” and “non-certified” operations for instance. However, the
decomposition must be meaningful and not dissociate multi-token proper
terms, or lead to a distinction that is mere syntactic reduction, but
ignores the polysemous nature of terms. Examples include chemical agent
and foreign agent, for which a separation into properties and entities
would not render semantically valuable criteria for distinction. The latter
specifically applies if the policy references variable properties for a given
entity (such as the previously mentioned variably ‘certified’ or ‘uncertified’
agent). Establishing this clarity as part of the study design is important to
improve the reliability of the encoded data. More details and additional
considerations are discussed in the IG 2.0 Codebook (Frantz & Siddiki,
2020).
Finally, the general handling of the coding inclusiveness is of rele-
vance. Does the coder retain, independent from potential reformula-
tion, the complete linguistic content of the statement (e.g., including
prepositions, articles, etc.), or are those removed during the coding
process? If retained, which component are they associated with (i.e., the
component they follow, or the one they precede)? Preferences should

be linked to the downstream analytical usage of institutional information
and be determined as part of the study design process to ensure reliable
coding. As indicated before, the IG 2.0 Codebook discusses associated
considerations.
The fifth step in the coding process is to actually engage in the syntactic
decomposition of institutional statements. In the following sections, we
provide examples of coded statements for reference. With each example
there is a brief commentary highlighting the central qualities of the
captured institutional information.
Before turning to the operational coding, a coding convention used
throughout the coding examples – as a formal correspondence to the
semantics presented under Sect. 6.1.1 – is introduced.
7.3.3.3 IG Script
The IG Script notation introduced in this context is an accessible formal
syntax for the encoding of institutional statements following IG 2.0 that
complements the semantic specification of the Institutional Grammar in
Sect. 6.1.1. and is a concise correspondence to the visual representation
used in the preceding sections. The principal syntax consists of explicitly
identified natural language elements that are parenthesized, alongside a
symbol signaling the associated component. A simple Attributes annota-
tion is:
To explore the coded statement in the IG Parser, please navigate to the URL:
https://newinstitutionalgrammar.org/book-examples#ed1ch7ex1
The combination of such annotations then reflects a statement, as
exemplified here
The ordering of components within a statement is arbitrary (as long as
the association with the statement is retained, and specifically relevant for
statement combinations introduced later). For instance, the coding
is treated equivalent to the example shown before. The purpose of

flexible reordering is to support the readability of encoded statements by
providing the option to align the order of encoded components with the
original text.
IG Script supports all IG components, namely
• Attributes (A)
• Deontic (D)
• Aim (I)
• Direct Object (Bdir)
• Indirect Object (Bind)
• Activation Conditions (Cac)
• Execution Constraints (Cex)
• Constituted Entity (E)
• Modal (M)
• Constitutive Function (F)
• Constituting Properties (P)
• Or else (O)
Component properties are indicated based on the suffix “,p”, for

instance:
Where components have multiple values (e.g., two actions), these are
explicitly specified as a logical combination signaled by the potential oper-
ators AND, XOR, or OR combination within the annotated component,
for instance:
Where components themselves exist multiple times, they are implicitly

treated as being AND-combined, i.e.:
This statement reflects the “accept and fulfill.”

The embedding of NOT as a special component reflects the negation

within its scope, i.e., applying to the component when embedded within
component annotation, or variably the statement if outside a component
annotation. For instance
negates the entire statement irrespective of operator position.
In contrast, the following
only applies within the first Aim element, i.e., applied within a
combination, the operator only refers to the associated element.
Generally, in IG Script, any non-annotated text other than logical oper-
ators is ignored for analytical purposes, highlighting the IG’s focus on
information that is essential to the characterization of the institutional
content. At the same time, it allows for the retention in coded data, e.g.,
to support readability.
Where multiple independent statements exist, their respective scope is
indicated with braces. As with the case of multiple components, unless
explicitly linked by logical operators, multiple statements are considered
AND-combined
Bdir
This effectively corresponds to the following encoding:
The final feature relevant for the interpretation is the notion of
component-level and vertical nesting. Where components embed entire
statements, the content is embedded in braces instead of parentheses and
augmented with the corresponding component symbol. For instance
reflects the nested consequence captured in the “O{ … }” element.
This principle applies to any other form of component-level nesting (e.g.,
“Cac{ … }”).
Finally, annotations associated with specific components are signaled
with square brackets, such as:
This applies analogously to statements (e.g., “O[stype=consequence]{
… }”).
Where multiple values apply for annotations, these are embedded in
nested square brackets
Where multiple annotations exist (e.g., by drawing on multiple
taxonomies), the corresponding annotations are comma-separated:
Where square brackets appear inside the annotated text (other than
with logical operators), they are conventionally used to reflect text recon-
structed or inferred as part of the encoding process. In the following
example, the value for the aim (‘[sends]’ ) is indicated as contextually
inferred or reconstructed
As indicated before, this specific use of square brackets inside annotated
text is merely conventional (as opposed to the mandated use to indicate
semantic annotations or logical operators where existing).
The complete syntax specification, as well as tool support for the

parsing of encoded institutional statements is provided via the book
resources websites.3 All IG Script-encoded examples in this and the
following chapters are directly processable by the purpose-built reference
parser implementation IG Parser ( see https://newinstitutionalgrammar.
org/ig-parser), and can be explored interactively by clicking on the
link below the corresponding IG Script-coded institutional statement
examples.
7.3.4 Operational Coding of Institutional Statements

The example statements in this section have been coded at the IG Core
level. By way of general coding guidance, as applicable to coding at
any level of expressiveness, it is recommended that the analyst begin
statement coding with identification of the Aim or Constitutive Func-
tion for regulative and constitutive statements, respectively. For regulative
statements, following the identification of Aim, it is recommended that
the analyst identify other necessary statement components – Attributes
(i.e., actor linked to statement Aim) and Context (i.e., qualifier and/or
instantiator of statement Aim). From there, the analyst should identify
the presence of sufficient component information (i.e., Deontic, Object,
and Or else information). In a similar vein, following the identification
of the Constitutive Function, it is recommended that the analyst iden-
tify other necessary components of constitutive statements – Constituted
Entity (i.e., Entity being constituted with the Constitutive Function) and
Context. From there, the analyst should identify the presence of sufficient
information (i.e., Modal, Constituting Properties, and Or else informa-
tion). The reasons for starting statement coding with identification of
the Aim or Constitutive Function are twofold. First, from a conceptual
standpoint, it is logical to commence coding along with necessary compo-
nents of institutional statements. Second, from a practical standpoint,
this information is typically explicit. The analyst is likely to encounter
few instances in which either the Aim or the Constitutive Function is
not explicitly present. Even other necessary information – Attributes
and Context /Constituted Entity and Context information – tend to
be more frequently missing, thus prompting inference on the part of
analyst. Attributes in formal institutional statements, for example, are
3 The IG resources website can be found under https://newinstitutionalgrammar.org.

often missing explicit reference given the tendency for statements to be

written passively in public policies.
In the following, selected examples are exemplified, with initial consid-
eration of regulative statements, before turning to the consideration of
constitutive statements.
7.3.4.1 Regulative Statement Examples
Each example is positioned as raw data, followed by a prototypical
encoding and elaboration. For the following examples, articles are kept
outside annotations (i.e., are ignored), and prepositions are associated
with the receiving component.
————————————————————
A student may request an Incomplete if the student has exceptional
circumstances that prevent the student from fulfilling all course require-
ments on time.
This example is of a complete institutional statement, in which all

necessary components for regulative statements (A-I-C) are explicitly
expressed.
————————————————————
Students are expected to attend all class sessions.
This example is missing explicit reference to Context. However, the

coding makes reference to an implicit Context—with an Activation
Condition “at all times,” and an Execution Constraint “no constraints.”
This is considered a default Context for instances of statements in which
the Context is not explicit. It is not uncommon to encounter state-
ments like this. Nor is it uncommon to encounter regulative statements in
which the Attribute is not explicitly stated. Attributes are typically missing
when statements are written in passive form. Attribute information can
usually be inferred from information provided in immediately preceding
or following statements, though sometimes the institutional analyst may
need to look beyond spatially proximate statements to derive needed
information. As indicated above, implied information is indicated by the

use of brackets within component coding.
This example also highlights an instance of Deontic information that
exists in clause form, rather than word form.
Additional practical heuristics that help distinguish Activation Condi-
tions from Execution Constraints are the semantic impact of their absence,
or the effects of reconstruction.
————————————————————
If physical access to university is restricted, instructors shall deliver courses
via online teaching platforms.
Initially, the coder will assess whether the omitted context (here the
physical access to university) changes the applicability of the regulated
part of the statement. If the applicability changed, the Context of concern
is an Activation Condition. A more specific assessment pertains to an
attempted reconstruction of an Execution Constraint in terms of an
Activation Condition: does the reconstruction in if-then form with the
candidate component as conditional lead to a modification of the substan-
tive meaning (or require extensive reconstruction) of the statement?4 If
such is the case, the Context of concern reflects an Execution Constraint.5
————————————————————
Instructors set course-specific policies on absences from scheduled class
meetings.
This example highlights a case in which the Direct Object of the state-
ment has a property assigned to it, and in which the property is essentially
4 In this instance “If via online teaching platform, instructors shall …” reflects a
substantive change in the meaning of the instruction.
5 Note that this statement foregoes the deep structural parsing of the Activation Condi-
tion associated with IG Extended. Coding corresponding to the latter are introduced later
in this section.
a purpose-oriented descriptor of the Object. As the Context Taxonomy

accommodates a purpose/function characterization – recognizing that
Context information sometimes describes the purpose or intent of an
Aim or Constitutive Function – the analyst may confront the question
of whether a clause like that included in the example statement (i.e., “on
absences from scheduled meetings”) is in fact an Object descriptor or char-
acterizes the activity’s purpose in the form of an Execution Constraint. If
confronted with such a question, it may be useful to recall that Execu-
tion Constraints are specifically linked to the Aim, and in extension the
Constitutive Function. Therefore, if the clause in question is activating
or qualifying the Aim or Constitutive Function in purposive or func-
tional terms, then it should be coded as Context, or more specifically,
as Execution Constraint. If however, as done in the example, the clause is
describing the purpose or function of an Object, then it should be coded
as an Object Property.
————————————————————
A state will rely on the advice of recognized clinical experts and scientists
to review and approve the safety and effectiveness of every vaccine that is
authorized by the federal government for distribution.
This example highlights instances where a statement contains logically
combined information associated with different syntactic components –
“recognized clinical experts [AND] scientists” with the Direct Object
Property field and “review [AND] approve” and “the safety [AND] effec-
tiveness” with the Execution Constraint field. Generally, this statement
also showcases that logically combined information can also be linked
to any syntactic component. Note that the Execution Constraint carries
further nested institutional state information. However, this is not decom-
posed on the IG Core level exemplified here (parsing of component-level
nested statements is discussed below).
————————————————————
A sailing vessel shall not impede the passage of a vessel that can safely
navigate only within a narrow channel or fairway.
This example shows an instance in which the Attribute of the statement
is actually not an animate actor (i.e., “sailing vessel”) but it is coded as
such because the statement phrasing animates (anthropomorphizes) the
inanimate actor, by ascribing it agency in lieu of its operator.
————————————————————
Businesses must submit a financial report annually, or else authorizing
body may suspend operating license.
This example presents a case that includes an Or else (i.e., consequential

statement) that nests on a monitored statement. The monitored statement
(“Businesses must submit a financial report annually …”) and consequen-
tial statement (“… authorizing body may suspend operating license”)
are both regulative in kind and are thus each coded in accordance with
regulative statement syntactic components. The monitored statement also
differentiates the Direct Object (“report”) from the Direct Object Prop-
erty (“a financial”). As a general heuristic, properties are distinguished
from Objects and Attributes when there are multiple kinds of particular
Attributes or Objects to which the Property is linked, and the Property
Qualifier plays an important distinguishing role. Using the example, if
a particular set of statements reference multiple types of reports, then
the “financial” qualifier is important for knowing which type of report
will be acted upon with the application of an institutional statement, and
this becomes easier to detect in the analysis process when the qualifier
is labeled as a property. If however, there is only one type of report – a
financial report – referenced in the set of statements, then the unique
classification becomes less important.
————————————————————
If farmer fails to submit an organic system’s plan by the end of the
calendar year, the certifier may suspend farmer’s operating license.
The referenced statement is consequential in kind; that is, it speci-
fies a sanctioning provision linked to a particular activity, or rather the
failure to perform a particular activity. The monitored activity itself is not
captured within a separate institutional statement, however, but rather
is captured within the Context (in this case, Activation Condition) of
the statement. In such cases – when the monitored activity to which a
consequential statement is linked is presented in the Context clause of
the consequential statement – it is advisable to engage by default to an
IG Extended coding of the statement, which accommodates treatment of
Context clauses as separable statements that can be syntactically parsed.
In this case, the Context clause is transformed into a complete regulative,
monitored statement to which the consequential statement represented
in the non-Context part of the statement is linked.6
7.3.4.2 Constitutive Statement Examples

————————————————————
Board Directors are “fiduciaries” of the organization they serve.
This statement presents an example of a prototypical statement in
which an entity is being defined. The example statement uses the Consti-
tutive Function “are.” Other common Constitutive Functions found in
such statements are “is” and “are defined as.”
————————————————————
All individuals 16 years of age and older that reside in the United States
are eligible by law to receive the COVID vaccine.
6 The formal qualification of this transformation is provided in Eq. (6.7) in Sect. 6.1.4.
This statement specifies status (captured in the Constituting Proper-

ties ), associated with a particular Constituted Entity through the Consti-
tutive Function “are eligible.” In so doing, the Constitutive Function in
essence ascribes rights to the Constituted Entity (i.e., “individuals 16 years
and older that reside in the United States”). The Execution Constraint
captures the contextual qualification of the Constitutive Function, i.e., how
this status is conferred.
Note that this statement offers a potential reconstruction in terms of
a conditional specification (following the Properties-Conditions Transfor-
mation principles highlighted in Eq. (6.8) in Sect. 6.1.4) by moving the
property characterization into the conditional clause (e.g., “if individuals
are 16 years and older and reside in the United States, they are eligible
…”).
————————————————————
A Renewable Energy Generation Unit may qualify as a Class II
Generation Unit subject to regulatory provisions.
This statement, by identifying how a Constituted Entity enters into

a particular position, essentially establishes eligibility criteria for occu-
pying a position. Coders will frequently find eligibility, phrased variably, as
the Constitutive Function in constitutive institutional statements, though
the statements may not only be referencing boundaries for occupying
positions.
————————————————————
An approval granted by the Department shall be a temporary approval
only.
This statement offers an example of a constitutive statement in which

“be” is the Constitutive Function. “Be” is a ubiquitous Function that
coders will find in statements defining Entities based on explicit definition
or ascription.
Where constitutive statements assume different purposes, such as

signaling a functional relationship, the encoding of the Constitutive
Function captures this quality.
————————————————————
Inspection procedures are regulated by provisions laid out in Section 2.
In this example, the quality of functional linkage is captured in

the Constitutive Function that links Constituted Entity and Consti-
tuting Properties explicitly. This similarly applies for the assignment of
organizational relationship (e.g., “is embedded in” ), status (e.g., rights,
authority), or any other form of functional relationship recognized for
constitutive statements as captured in the Constitutive Functions
Taxonomy (see Sect. 6.1.2.4).
————————————————————
Audits are financial inspections conducted by a certified authority
licensed according to this Act.
This example showcases intra-statement referencing. The statement
links information contained therein to “the Act” more broadly. Other
types of intra-statement linkages commonly observed in statements are
the linking of statement information to particular sets of other institu-
tional statements. In the case of formal institutions, linkages can be made
to the same Act in which the statement occurs, or a different Act. Or, in
the same case, linkages can be made to sets of other institutional state-
ments in the same or different Act. Note that this statement, as well
as all other preceding are coded according to IG Core principles. The
following statements will incrementally introduce features associated with
IG Extended and IG Logico.
7.3.4.3 Hybrid Institutional Statement Examples

Statements that embed both regulative and constitutive statements are
hybrid statements, with polymorphic statements reflecting a special form,
given that they can be variably expressed in regulative and constitutive

form. The following examples highlight various complex statements. Note
that the coding for the examples is indented to ease the interpretation for
the reader.7
————————————————————
The item in question must have the following characteristics: be duly
passed by the Board of Directors, establish rule or convention of general
application, and must be mission-consistent.”
This example in particular highlights instances in which an institutional
statement can be broken into multiple atomic statements, some of which
are constitutive in kind and some of which are regulative in kind. Specif-
ically, the component-level combination of Constituting Properties (by
AND operators) in definitions is a commonplace occurrence. Note that
the coded statement highlights the inference of implied component values
considered as part of the coding process, so as to make their interpretation
overt.
7.3.4.4 Polymorphic Institutional Statement Examples

As mentioned above, polymorphic institutional statements allow encoding
in terms of both regulative and constitutive statements, depending on
contextual interpretation, and more importantly, the focus on encoding,
with narrow focus immediately and only focused on the statement of
concern, or a wide focus that takes into account the nature of the state-
ments that a particular statement is linked to. This decision is a central
parameter to be determined as part of the study design (see Sect. 7.2).
7 IG Script allows for arbitrary formatting of encoded text.

A more detailed discussion is offered as part of the IG 2.0 Codebook

(Frantz & Siddiki, 2020).
Board members serve as organizational leaders.
Coded in regulative form (as a descriptive strategy), along with stylistic
adjustment, but under retention of component order (to visualize the
analogy to constitutive coding):
Coded in constitutive form (reflecting a constitutive rule):
————————————————————
Faculty are responsible for assigning grades.
Coded in regulative form:
Coded in constitutive form:
The two examples listed in this section, in being polymorphic state-
ments, represent the special case of being coded as regulative and/or
constitutive. Statements are generally polymorphic in kind if the actor
responsible for affording the functional linkage in the constitutive state-
ment can be unambiguously inferred. The notable difference is the
general characterization of how aspects are conceptually characterized.
The coding above exemplifies how each statement would be coded
according to the regulative and constitutive syntaxes. In this partic-
ular instance, the statements reflect the endowment of responsibility on
the constitutive side, which is operationalized as a duty (assuming the
establishment in a formally recognized forum) on the regulative side.
7.3.4.5 Component-Level Nesting (IG Extended)

Whereas the previous statements highlight the general coding of
components and illustrate distinctive considerations, the following state-
ment features the coding of structure embedded in components (see
Sect. 5.1.3).
————————————————————
When an inspection, review, or investigation of an accredited certifying
agent by the Program Manager reveals any noncompliance with the Act or
regulations in this part, a written notification of noncompliance shall be
sent to the certifying agent.
This statement is complex in that it requires the reconstruction of the

underlying statement from a structural perspective to reflect the insti-
tutional structure in terms of linked institutional state(ment)s following
the AIC schema.8 Specifically, this involves reconstructing the sequence
of events that condition the activity captured in the overall state-
ment, namely “inspection,” followed by “revelation of non-compliance,”
followed by “sending notification.” Reflecting this in the encoding, the
statement captures two Activation Conditions. The first is the perfor-
mance of an inspection (or variably review or investigation), which in
fact is the precondition for the second one (the revelation of non-
compliance).9 Moreover, the second Activation Condition exemplifies the
nesting on the Object component to capture atomic institutional informa-
tion captured therein (here the state specification of non-compliance, i.e.,
“accredited certifying agent is not in compliance with the Act …” ).
Additional observations include the use of component-level combi-
nations to express distinctive action or state alternatives captured in a
8 Modified or inferred component values are indicated using square brackets.

9 Recall that the ordering of components within atomic institutional statements is flex-
ible, i.e., the nested (first) Activation Condition could have been the first element in the
second higher-level Activation Condition.
single statement, alongside a reconstruction that resolves the perfor-

mance of distinctive monitoring activities in active terms (i.e., “inspects,
reviews, investigates” ). Note that reconstructions of such kind should
find consideration during the Planning Phase (see Sect. 7.2) to ensure
principal compatibility with analytical objectives and methodological
principles, alongside operational concerns for reliability (further details
will be provided in Sect. 7.4).
The encoding of deep structural information as presented here,
including the resolution of patterns within components (i.e., the two
Activation Conditions, and one Institutional State structure embedded in
the Object ), is a feature associated with IG Extended. As indicated above,
an extended discussion of the reconstruction principles, exemplified on a
variant of the same statement, is offered in Sect. 5.1.3.
While this statement retains the structure expressed in the original
form, the nature of the statement itself expresses a sanctioning provi-
sion, i.e., is consequential in kind. The coder (and the study designer
more generally) may consider affording the reconstruction in terms of
the idiomatic consequential form by transforming the statement to reflect
a linkage between monitored and consequential statements that accords
with the statement linkage afforded by the Or else operator. The principles
of this transformation are specified in Eq. (6.7) in Sect. 6.1.4, and further
generalized in Sect. 8.3.1 (to extrapolate the statement from multiple
perspectives). As indicated, the applicability of transformations as part of
the coding process should be explicitly specified as part of the Planning
step (see Sect. 7.2).
7.3.4.6 Object-Property Hierarchy (IG Extended)

A specific feature of IG Extended is the introduction of complex entity
relationships captured in the Object-Property Hierarchy (see Sect. 5.1.4).
Specifically, the coder will frequently encounter not only multiple prop-
erties associated with a given object, but also recognize their nestedness
across multiple levels, an aspect illustrated in the following example – an
example previously encoded using IG Core.
————————————————————
Audits are financial inspections conducted by a certified authority
licensed according to this Act.
The properties highlight noteworthy characteristics that warrant the
encoding according to IG Extended, and more specifically, the Object-
Property Hierarchy. Specifically, we can observe nested properties (i.e.,
“licensed according …” ), that is, properties that are associated with entities
that are themselves linked to entities referenced in properties (“certified
authority” ). Where multiple properties on a given level exist, and some of
those carry specific nested properties, the former (i.e., parent properties)
are indexed to retain the unique association between (parent) property
and nested (child) property. This encoding ensures the unique associa-
tion, but furthermore allows the coder to capture the dependencies of
individual entities within and beyond the statements they are referenced
in.
The principles apply analogously to other components that associate
with nested properties. Further examples exploring the full range of
features of IG Extended are provided in the supplementary IG 2.0
Codebook.
7.3.4.7 Semantic Annotations (IG Logico)

Highlighting the principles of semantic annotations and their operational
representation, the following encoding features the same institutional
statement shown before, but extends the encoding with distinctive
semantic annotations that accord with the various taxonomies introduced
in Sect. 6.1.2.
The example highlights the selective annotation, on the one hand,

as well as the ability to specify multiple annotations, as well as nuanced
second-order annotations linked to embedded value alternatives (e.g., Act
vs. regulations in this part). Naturally, the nature and extent of annota-
tion is subject to analytical objectives and necessities, and is discussed in
Chapter 8.
Notably, the overview of coded statements is intended to capture the
principal coding approach, but does not capture nuances and detailed
aspects of the coding process comprehensively. Rather, it links the
methodological considerations that pertain to the study design more
generally with the operationalization in terms of the coding exercise.
Detailed coding instructions and study design guidance are provided as
part of the IG 2.0 Codebook (Frantz & Siddiki, 2020). A comprehensive
overview of relevant resources, including the specification of IG Script as
well as tool support (such as the IG Parser reference implementation),
can be found under https://newinstitutionalgrammar.org.
Further considerations as part of the coding process – beyond planning
and operational coding – include the assessment of coding quality in the
form of reliability assessments, an aspect briefly discussed in the following
section of this chapter.
7.4 Assessment
Having provided an example of IG coding according with the three levels
of expressiveness, the discussion now turns to the third stage of the IG
coding process – Assessment. In this section, the discussion focuses on the
assessment of the validity of coded data and on the assessment of coding
reliability, before we turn to the analysis of coded data toward particular
research aims and using different forms of analytical techniques.
With assessments of coding validity, the focus lies on ensuring that
institutional information has been appropriately coded given aspects of
the institutional domain. This is especially important because, as noted
earlier in this chapter, analysts will often engage in the inference of institu-
tional information to complete statement coding, let alone apply variable
tacit conventions. Appropriate inference begs institutional and domanial
understanding. To assess coding validity, the institutional analyst may seek
to do additional archival research, particularly seeking information that
could aid in the verification of coding accuracy. Additionally, the institu-

tional analyst may also consider soliciting feedback on coded data from
domain experts.
Coding reliability is assessed through intercoder reliability testing.
Intercoder reliability testing (Gwet, 2014; O’Connor & Joffe, 2020)
basically assesses whether multiple independent coders generate the same
coding of institutional data. Reliable coding necessitates sufficient under-
standing of coding concepts, sufficient understanding of the institutional
domain (as domain specific knowledge is crucial for correct interpreta-
tion and inference of institutional information), common understanding
of how the institutional analyst will handle stylistic and other features of
institutional data referenced earlier in this section, as well as common
access to coding protocols, guides, or instructions. Intercoder reliability
assessments require at least two coders per institutional dataset. These
assessments may be easier to conduct when institutional analysts are
working in teams rather than independently, again given the need to have
domain-specific understanding. One convention in existing IG research is
to assess percent agreement on statement and syntax level classification.
It is recommended that a second coder code 10–20% of the institutional
data coded by a first coder. Eighty percent agreement among coders is
considered adequate agreement, by convention. Institutional analysts also
rely on particular statistical measures for ascertaining coding reliability; for
example, Cohen’s Kappa and Krippendorff’s Alpha (Krippendorff, 2018).
It is important that teams of coders all be made aware of coding decisions
made during the planning phase of an IG coding project (e.g., classi-
fication to accommodate different stylistic features), and that those are
sufficiently tested in pilot studies, as awareness and internalization of these
decisions can have significant implications for coding reliability.
A final point of consideration relating to intercoder reliability concerns
coding tools. Assessing intercoder reliability can be more or less diffi-
cult depending on the tool with which the coding has been conducted.
Institutional information captured in tabular form is easier to compare
and thus assess for reliability. In contrast, assessment of information that
is captured using text annotation tools inherently relies on the facili-
ties provided, and may, where absent, need to be further organized to
facilitate a direct comparison of information generated by different coders.
7.5 Concluding Remarks

This chapter provided a high-level overview of methodological aspects
pertaining to any study that intends to encode institutional information
in terms of the IG. The main focus of this overview was to provide
insights into essential points to be considered during study design in
order to ensure a robust design and an efficient coding process. In addi-
tion to principal planning steps, the chapter also highlighted operational
coding, as well as introducing a formal coding syntax (IG Script) to
support this process, before finally returning to discussions related to
the operational assessment of reliability in the coding process. Beyond
the high-level characterizations offered here, the analyst is encouraged
to review the supplementary Codebook (Frantz & Siddiki, 2020) for
additional and specific guidance on the operational aspects related to
the coding of institutional statements. In the codebook, one will find
guidance on the encoding of advanced concepts, such as complex data
structures, comparators and quantifiers, as well as an extended discus-
sion of encoding hierarchical relationships amongst properties, in addition
to a rich overview of coding heuristics for regulative and constitutive
statement classification only touched upon briefly in this chapter.
A concluding point related to the methodological process involved in
the encoding of institutional information is a note on documentation. The
analyst should consider a careful documentation of individual decisions
in the form of a customized codebook in order to support the coding
process itself and ideally to accompany the dataset, both to foster rigidity
and opportunity for replication of studies, supporting the call for method-
ological rigor of IG research. Moreover, it provides the basis to perform
comparative studies produced by different teams or based on existing
datasets, in which case detailed information about the encoded informa-
tion (data description) and ensuing preprocessing, let alone coding, is
essential to establish compatible and comparable datasets.10 Supporting
this process, the supplementary codebook provides a continuously refined
checklist that captures the essential aspects motivated throughout this
chapter.
10 These aspects are discussed in greater detail in Chapter 9.

References
Gwet, K. L. (2014). Handbook of inter-rater reliability: the definitive guide
to measuring the extent of agreement among raters, 4th edition, Advanced
Analytics.
Krippendorff, K. (2018). Content analysis: An introduction to its methodology
( 4th edition). SAGE Publications.
O’Connor, C., & Joffe, H. (2020). Intercoder reliability in qualitative research:
Debates and practical guidelines. International Journal of Qualitative
Methods. https://doi.org/10.1177/1609406919899220
CHAPTER 8
Institutional Analysis and Applications
The preceding chapters have introduced the Institutional Grammar as

a multilevel approach for capturing and expressing institutional design.
The Institutional Grammar supports the representation of institutional
information at multiple levels of expressiveness to respond to different
analytical objectives, either favoring the characterization of general struc-
tural features of institutions, the extraction of conceptual information
expressed in language, or the semantic layering that links theoretical
concepts of the application domain with the rigidity in analysis that the
IG affords. The general methodological guidance offered in Chapter 7
reflects the differentiated treatment of institutional data to arrive at
institutional information.
However, any analytical approach, especially if rich in concept and
methodological guidance, has to be assessed against the novel analytical
opportunities it offers. Responding to this demand, alongside the intro-
duction of associated analytical techniques that the refined IG is amenable
to, is the objective of this penultimate chapter.
This chapter approaches this essential discussion from two perspectives.
On the one hand, it will offer a characterization of analytical features
by level of expressiveness, incrementally enhancing the richness of the
insights that the institutional information encoded in terms of institu-
tional statements offers – and thereby establishing an intuitive sense of
the opportunities that analysts can leverage. On the other hand, it aims

Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2_8
Fig. 8.1 Analytical Applications by Levels of Expressiveness
at stimulating analysts to venture into unexplored research directions by

posing exemplary research questions that can motivate the selection and
adaptation of the IG feature set to respond to such questions. Inasmuch
as this chapter describes established analytical pathways that, in part, build
on the analytical traditions and techniques highlighted in Chapter 2, this
chapter aims at opening novel directions, providing the basis for a reori-
entation of institutional analysis more generally, exemplifying particularly
the analysis of policy design.
To address these objectives, this chapter is loosely structured along
the boundaries of the coarse-grained encoding sponsored by IG Core,
followed by the deep structural parsing enabled by IG Extended, and
the shift to a semantic perspective and disciplinary embedding high-
lighted as part of IG Logico. Figure 8.1 highlights the individual levels,
and exemplifies techniques that broadly align with the granularity of
parsed institutional information on corresponding levels. Preempting
the discussion of potential applications and associated research questions
in the context of the individual sections corresponding to the levels of
expressiveness, the analytical themes associated with the levels progres-
sively broaden, but initially emphasize the establishment of measures of
8 INSTITUTIONAL ANALYSIS AND APPLICATIONS 279
institutional design in the context of IG Core, the focus on under-

standing the structure of institutional design for IG Extended, and the
development of pathways toward explaining institutional design using the
facilities offered by IG Logico. The remainder of this chapter samples the
referenced methods, techniques, and their application to IG data.
In this process, the chapter makes an attempt to separate the analyt-
ical perspective from methodological considerations associated with the
encoding of institutional information (as featured in Chapter 7). Given
the diverse nature of the different analytical pathways presented here,
some may require additional processing steps. Where essential to the
analytical technique, relevant steps and associated considerations are
discussed.
The discussion in this section, and supporting analytical illustrations
presented herein, draw on a dataset of coded institutional statements
extracted from the United States National Organic Program Regula-
tion, hereafter NOP regulation (the complete excerpt is included in
Appendix B). This regulation, implemented by the United States Depart-
ment of Agriculture, governs organic food production in the United
States. The regulation outlines standards for organic food production, and
details protocols pertaining to the processing, handling, and labeling of
organic products.1 It also provides guidance on obtaining organic certi-
fication, accreditation of certifying agents, and monitoring and enforce-
ment relating to organic certification. The referenced excerpt pertains to
“Compliance” and describes the role of organic program managers and
certifying agents in monitoring and enforcement of regulatory standards
and also describes rights and responsibilities of organic operations subject
to monitoring and enforcement.
The excerpt contains 55 institutional statements, which altogether are
comprised of 259 atomic institutional statements. 203 of the atomic insti-
tutional statements were identified as regulative in kind and 56 were
identified as constitutive in kind. The discussion that follows reports on
analysis of these coded statements in the context of a more general discus-
sion of possibilities for analyzing institutional data parsed with techniques
and approaches that leverage specific features of the different levels of
expressiveness.
1 Source: https://www.ams.usda.gov/rules-regulations/organic.
8.1 IG Core---Establishing
Fundamental Institutional Metrics
The basis for the initial set of analyses discussed and proposed in this
chapter are the metrics introduced as part of extant work discussed in
Chapter 2, including the initial applications of Crawford and Ostrom
(1995)’s original IG for the analysis of policy, Basurto et al. (2010) and
Siddiki et al. (2011). In step with the structurally focused, and relatively
coarse-grained encoding of institutional data at the IG Core level, analyses
of institutional information generated at the IG Core level tend to focus
on general, structural depictions of institutional statements drawing on
basic parsing thereof along A-D-I-B-C/E-M-F-P-C components. Struc-
tural analysis of IG Core data relies on information corresponding to
individual or sets of syntactic components but typically engages aggre-
gate representations of this information across institutional statements.
Most often, this aggregation entails a summary of information linked
to syntactic components, but may also entail graphical representations of
how syntactic information from different statements links together. It may
also entail a translation of institutional information into specific kinds of
values (e.g., network metrics, numerical values) that can be summarized
and assessed in different ways and using different methods (e.g., network
analysis, statistical analysis).
Generally, structural analysis of IG Core data, which tends to orient
on generating and/or analyzing aggregate representations of institutional
information, focuses on discerning patterns in institutional design. Funda-
mentally, it is not about understanding the Attributes, Aim, Context,
etc. of a single institutional statement (unless perhaps an institution of
interest is only comprised of a single statement), but rather understanding
which components’ values (i.e., Attributes, Aim, Context, etc.) are repre-
sented within a configuration of statements that comprise an institution
of interest.
8.1.1 Component-Level Aggregate Metrics

The kind of structural analysis of institutional statement information
referenced here is well exemplified in existing IG studies that rely on
the original IG’s syntactic specification. The analysis engaged in existing
studies is a relevant reference for what is described in this section, insofar
as IG Core only presents a modest departure in component character-

ization and coding features from the original IG. At the same time,
however, analysis of IG Core parsed information, versus information
parsed according to the original IG, can lend more nuanced and complete
understanding of institutional design. This augmented capability stems
from the inclusion of additional syntactic features along which data can be
analyzed. Notably, the ability to capture different types of Objects (Direct
Objects and Indirect Objects ) and Contexts (Activation Conditions and
Execution Constraints ), as well as clearly delineate Attributes and Objects
from their defining properties. An additional valuable feature of the IG
2.0 is the specification of atomic institutional statements that can logi-
cally combine to form composite institutional statements (see Sect. 4.2.2).
Assessment of atomic statements comprising institutional statements, as
an alternative or in addition to the analysis on composite statement level,
can shed light on the structural complexity of institutional statements,
and concomitant implications, a point which is discussed in more detail
later in this section.
Nevertheless, a practice observed in existing studies which retains rele-
vance is the summary of institutional information by statements to derive
broader understanding of an institution in question. For example, analysts
can aggregate information corresponding to the Attributes component
to gain an understanding of the array of targets of a given institu-
tion. Similarly, analysts can aggregate information corresponding to the
Deontic component to generally infer prescriptiveness of a given institu-
tion. Summary by Aim component can signal focal activities referenced
within an institution. Summary by Objects can be used to discern focal
topics within an institution.
By way of illustration, Table 8.1 identifies which Attributes, Deontic,
Aim, and Direct Object component values are most frequently occurring
in the institutional statements that comprise the Compliance excerpt from
the NOP regulation of regulative kind. The reported information shows
that “Program managers” are the most frequently occurring Attributes,
“shall” is the most frequently occurring Deontic, “notification” is the
most frequently occurring Direct Object, and “operation” is the most
frequently occurring Indirect Object. Clearly, institutional analysts can
also rely on summaries of syntactic information represented in constitu-
tive statements to derive analogous qualities of information. Table 8.2
identifies the Constituted Entities (“notification”), Modal (“shall”), and
Constituting Properties (“subject to the provisions of section 1001 of title
Table 8.1 Most frequently occurring component information (regulative state-

ments)
Definition Attributes Deontic Direct object Indirect object
Most frequently Program manager shall notification operation

occurring information
corresponding to
syntactic components
Percentage of 0.34 0.38 0.40 0.37
statements in which
most frequently
is present
Count of statements in 70 78 82 74
which most frequently
is present
Note There are 203 regulative statements in the example dataset
Table 8.2 Most frequently occurring component information (constitutive

statements)
Definition Constituted entity Modal Constituted properties
Most frequently occurring Notification shall subject to the provisions of

information corresponding Section 1001 of title 18,
to syntactic components United States Code
Percentage of statements in 0.55 0.95 0.11
occurring information is
present
Count of statements in 31 53 6
occurring information is
present
Note There are 56 constitutive statements in the example dataset
18, United State Code”). This pattern among regulative and constitu-
tive statements is interesting, in that it conveys the focus on compliance
notifications across both kinds of statements – compliance notifications
are frequently being acted on by regulatory actors, as well as constituted
(e.g., defined in terms of properties).
Aggregated representations of institutional information coded at the

IG Core level need not only be done by individual components. Valu-
able insights may also be gleaned from assessing patterns in informa-
tion according with multiple syntactic components. For example, the
institutional analyst can summarize coupled Attributes and Aim data
to assess which actor-action linkages occur most frequently within a
given institution. The analyst can also analyze Attributes-Deontic-Aim
linkages to assess general levels of prescription associated with actions
associated with particular actors. Relating any of this institutional infor-
mation with Context information can elucidate the kinds and extent of
constraints assigned to particular actions. To illustrate the referenced anal-
ysis, Table 8.3 provides a summary of institutional information most often
linked within coded regulative statements. Patterns observed in frequent
institutional information pairings provide interesting general insights.
The observed coupling of “Program Manager” with “may” versus a
more prescriptive Deontic signals an ascription of discretionary authority
Table 8.3 Most frequently paired institutional information across syntactic

components
Most frequently paired component Attributes Deontic Aim Direct object

information
Attributes-Deontic Program may

manager
Attributes-Aim Organic send
program’s
governing
State official
Deontic-Aim shall send
Attributes-Deontic-Aim Program may initiate
manager
Attributes-Direct Object Organic notification
program’s
governing
State official
Aim-Direct Object send notification
Attributes-Deontic-Aim-Direct Organic shall send notification
Object program’s
governing
State official
Note There are 203 regulative statements in the example dataset

to Program Managers. Further, the observed coupling of “Organic

Program’s Governing State Official” and “send” signals the Official’s role
in transmission of institutionally relevant information/artifacts.
Importantly, referencing convention in existing studies, summaries
of information by syntactic component need not only be interpreted
in a strictly literal sense. Patterns reflected in summarized informa-
tion can also be used as operationalizations of concepts of interest.
For example, relative frequencies of Deontic information within a policy
document may be used to measure policy stringency, prescriptiveness,
and obligation, each of which can be used to understand how strongly
behavior is compelled (regulative statements)/necessity or (im-)possibility
of constitution (constitutive statements). These measures are signaled in
institutional statements based on different treatment/interpretation of
Deontic/Modal data. Table 8.4 shows the percentage of coded regula-
tive and constitutive statements containing Deontics/Modals of particular
kinds as a way to measure stringency. Note, stringency as operational-
ized here can be decomposed based on different values (e.g., “shall”,
“shall not”, “must”, “must not”). However, the interpretation of these
values as a quantification of stringency (e.g., suggesting that “must”
signals stronger levels of stringency than “shall”) relies on contextual
knowledge of domain- or discipline-specific conventions on the use of
specific Deontic/Modal values as discussed in Sect. 4.2.1.2.
Table 8.4 Statement stringency operationalization
Metric Definition Regulative Constitutive

measure measure
General policy Percentage of regulative and 0.623 1

stringency constitutive statements
containing non-permissive
deontics/modals (must, must not,
shall, shall not)
Stringency Percentage of regulative and 0.237 0.964
reflected with constitutive statements containing
“shall/shall not” “shall” or “shall not” deontics/modals
Stringency Percentage of regulative and 0.386 0.036
reflected with constitutive statements containing
“must/must “must” and “must not”
not” deontics/modals
Institutional analysts have used IG-coded data toward the operational-

ization of various concepts (see Chapter 2). One relatively understudied
concept in existing IG research that the IG 2.0 is particularly well suited
to support measurement of is policy complexity. Policy complexity has
been measured variably in existing scholarship by social scientists from
multiple fields of study. In this scholarship, policy complexity has been
defined in terms of the length of a policy (Huber & Shipan, 2002),
the size of a policy (Gerber & Teske, 2000), linguistic complexity of a
policy (Chun & Rainey, 2005), the degree to which a policy is perceived
as difficult to use and understand (Rogers, 2003, p. 242), and in terms
of number of actors and collaborative decision points indicated in insti-
tutional design (Shaffer, 2021). IG Core-coded data can be readily
leveraged to assess policy complexity defined in any of the above ways,
as well as in additional ways. Notably, while the discussion in this para-
graph is specific to assessing institutional complexity in policy studies,
further discussion of institutional complexity in this chapter offered in
Sect. 8.2.1.1 is more generalizable and speaks to other ways of quantifying
the complexity of IG-parsed institutional information.
For example, the analyst can seek to understand not only how many
actors are linked to particular actions of interest, but also the extent of
contextualization of actions as indicated by Activation Conditions and
Execution Constraints associated with these actions. To generate these
measures of complexity (i.e., complexity measured in terms of number of
actors and number of activation conditions and/or execution constraints
tied to different actions), the analyst will need to combine information
from logically linked atomic statements corresponding to individual insti-
tutional statements – thus first constructing a complexity measure at the
institutional statement level – and then the analyst can further aggregate
this information across institutional statements to derive a policy level
measure of complexity.
Aggregate descriptions of syntactic information can also be used to
assess linkages among regulative and constitutive statements, which, too,
can signal important qualities of institutions of interest. For example,
in a general assessment of policy comprehensiveness, the institutional
analyst can assess information commonly captured across regulative and
constitutive statements. Such assessments can be useful insofar as they
capture, for example, to what extent artifacts upon which actors are
pre-/de-scribed to act upon within regulative statements are defined
through constitutive statements. It may also be useful to identify the
relative number of constitutive and regulative statements. Table 8.5 iden-

tifies both coded regulative and constitutive statements. With respect to
the former, the table also indicates how many times co-occurring infor-
mation is referenced in institutional statements. Additionally, the table
indicates frequency of unique kinds of information conveyed in regulative
and constitutive statements; i.e., types of Attributes, Objects, and Entities
represented in institutional statements.
Table 8.5 Constitutive-regulative Dynamics
Metric Definition Regulative Constitutive

Attributes Direct Indirect Constituted
Object Object Entity
Atomic Count of unique 8 26 10 8

Diversity instances of syntactic
information (e.g.
“agent”, “official”)
Property Count of unique 14 86 20 21
Diversity component-property
combinations (e.g.
“certifying agent”,
“certifying agent
whose accreditation
is suspended …”)
Occurrences Count of references 70 82 74 31
of Most to most frequently
Frequent occurring
Component component value
across institutional
statements
Most Most frequently Program Notification Operation Notification
Frequent occurring manager
Component component value
Total Count of statements 203 203 124 56
Number of containing explicit
Components reference to
syntactic component
Total Count of regulative 203 56
Number of and constitutive
Atomic atomic institutional
Institutional statements
Statements
8.1.2 Network Analysis

The discussion up to this point provides summaries of institutional infor-
mation corresponding to different syntactic components in nominal or
numerical form. As referenced above, scholars also rely on graphical
summaries of information, which help to visualize connections among
institutional information. In recent years, scholars engaging the IG
in institutional analysis have increasingly relied on network analysis to
support the generation of visualizations and assessment of institutional
patterns found therein (Chapter 2). Institutional information can be
engaged flexibly in network analysis, with syntactic information corre-
sponding to nodes or edges depending on the analyst’s objectives.
Further, the analyst may rely on information that corresponds to one or
more syntactic components to support network visualization and anal-
ysis. For example, the analyst can connect Attributes to Objects treating
both as nodes, and relying on edges to convey frequency of association
across institutional statements. As an extension, the analyst can connect
Attributes and Objects, while capturing Aim information on edges. Essen-
tially, this analysis would convey the specific action through which actors
act on objects.
Figure 8.2 provides a network visualization of statements from the
Compliance excerpt of the NOP regulation. All statements from the
excerpt are represented in the network graph. Attribute, Direct Object,
and Indirect Object information are used for nodes, and Aim informa-
tion is used for edges. Arrows in the graphs show directionality in activity,
which allows one to detect, where Attributes and Objects are actors, which
actor is the “Originator” of an activity, and which actor is the “Recipient”
of an activity (see reference to Role Taxonomy in Sect. 6.1.2 and further
discussion in Sect. 8.3).
Note that this type of analysis and representation of institutional infor-
mation is uniquely afforded with IG Core coding, over the basic syntactic
parsing recommended with the original Grammar, because it enables the
reliable specification of actors in an action-receiving role. The analyst may
wish to superimpose this role recognition onto network graphs, as well as
superimpose other kinds of labels that accord with various taxonomies
presented as part of the IG 2.0, which would essentially help portray
notable semantic information along with structural information. Finally,
the thickness of lines in the graph show the directional frequency of
Fig. 8.2 Network structure of compliance excerpt (An enlarged version of this
figure can be found under https://newinstitutionalgrammar.org/book-figures)
syntactic information pairings across institutional statements. One addi-

tional benefit of network analysis is the ability to generate network
statistics (e.g., density, centrality)—quantitative metrics that convey qual-
ities of networks. In essence, these statistics support assessments of
institutional foci.
8.1.3 Additional Analytical Pathways

Finally, to support understanding of institutional patterns that draw on
institutional information captured at the IG Core level, the analyst can
correspond qualitative institutional information to numerical values to
support usage thereof in statistical analyses, where numerical values accord
to independent or dependent variables. Basically, the analyst can translate
institutional information into discrete values which can be treated individ-
ually or indexed – through combined treatment of discrete values – and
then incorporated into statistical analysis (e.g., one could represent degra-
dation of Deontic stringency based on a numerical scale). Discretization
of institutional information in numerical form will necessarily require the
analyst to a priori identify what characteristics/qualities of institutional
information will map to different numerical values. Such decisions may
be tied to theoretical, conceptual, and/or empirical foci.
Zooming out to capture broader institutional dynamics, the institu-
tional analyst may also be interested in capturing general patterns in the
organization of institutions. For example, the analyst may perceive some
value in identifying the number of statements present within different
sections of a policy, as another – albeit general – way of capturing core
institutional foci. For illustration, Fig. 8.3 shows how institutional state-
ments are distributed across subsections of the Compliance section of
NOP rule. The figure indicates that over half of all institutional state-
ments pertain to noncompliance procedures for certified operations and
noncompliance procedures for certifying agents.
Additionally, extending the preceding discussion of policy complexity,
the analyst may wish to capture patterns in institutional statement
complexity by assessing the number of logically connected atomic state-
ments a single statement is composed of. Figure 8.4, for example, shows
Fig. 8.3 Atomic statement distribution across subsections (An enlarged version
of this figure can be found under https://newinstitutionalgrammar.org/book-fig
ures)
Fig. 8.4 Atomic statement distribution across policy (Complexity Landscape)

(An enlarged version of this figure can be found under https://newinstitutiona
lgrammar.org/book-figures)
the breakdown of atomic statements by institutional statements included

in the NOP regulation Compliance excerpt. Descriptively, the graph
shows that statements in this excerpt are comprised of between 1 and 20
atomic statements. Substantively, an analyst can use such information for
assessing complexity of instruction conveyed in policy language, recalling
that each atomic statement conveys a unique instruction – an aspect
explored at greater detail in the following section. This can have impli-
cations for compliance with statements, implementation of statements, as
well as cognitive effort associated with the interpretation of statements.
Even more broadly, and practically, statement complexity as measured by
the number of atomic statements of which an institutional statement is
comprised, can have implications for mere readability.
Another analytical possibility in the assessment of broader institutional
patterns is evaluation of the presence and qualities of linked monitored
and consequential statements. Firstly, the analyst may be interested in
ascertaining whether a set of institutional statements has consequen-
tial statements reflected within it. This fundamentally conveys whether
institutions reference enforcement for instances of institutional non-
conformance. Secondly, where consequential statements exist, the analyst
may wish to analyze features of enforcement; for example, who is respon-

sible for enforcement (through assessment of Attributes information),
how much discretion is afforded for enforcement (through assessment of
Deontic information), and the severity of enforcement (through assess-
ment of information corresponding to various components). For the
latter, the analyst may find it useful to assess severity of referenced
actions, and whether action contextualization tempers the Deontic. For
example, the NOP rule references both the potential for enforcement
personnel to “suspend” or “revoke” certifications, actions that clearly
convey varying levels of enforcement severity. Activation Conditions and
Execution Constraints can temper Deontics, by qualifying when, where,
how, etc. varying levels of prescription should be applied.
These latter aspects, the extended treatment of atomic statements, as
well as Activation Conditions and Execution Constraints , are subject to
the advanced metrics established in the upcoming section.
8.2 IG Extended---Structural and Behavioral

Analysis of Institutional Design
Following the essential characterization of the basic structure of insti-
tutional statements under IG Core, which often entails an aggregated
representation of institutional information corresponding to different
syntactic components, the advanced coding introduced under the label
IG Extended leverages additional features associated with this level to
support a richer characterization of institutional statements drawing out,
in particular, their compositional qualities. Among those features are
component-level nesting and complex property characterizations. In this
light, a central promise attached to the use of IG Extended is the extrac-
tion of deep structural information. This section differentiates between
two types of techniques for evaluating the fine-grained parsing of institu-
tional statements according with IG Extended. The first is structural, and
the second is behavioral. Both are discussed in turn below.
8.2.1 Structural Analysis

Structural analysis works towards the extraction of embedded structural
features relevant for extended analytical treatment. A central premise of
the nesting facilities of the IG is that the complexity of institutions is
embedded within statements, but also includes their linkages, allowing the
analysis of institutional statements both on atomic, as well as on a compo-

sitional level, transcending individual institutional statements and moving
towards a systemic perspective based on the linkage of IG statements.
8.2.1.1 Complexity Analysis

Commencing the overview of approaches to structural analysis using the
IG, an initial innovation sponsored by IG Extended is the assessment
of the complexity of the underlying information beyond the aggregate
metrics introduced in Sect. 8.1. To motivate both complexity assess-
ments based on its decomposition, this discussion draws on a variant of a
complex institutional statement previously encoded in Chapter 7:
Given the self-evident complexity of the statement, the analyst may ask
for a qualification of these linkages, as well as the conceptual mapping
relevant in the analytical domain. Where the latter will be subject to
later discussion, the former is an aspect addressed at this stage. While
IG Core introduces metrics that capture, for example, the presence and
quantity of the institutional information expressed and linked in atomic
institutional statements, IG Extended expands this characterization based
on additional decompositional facilities and their qualification. A central
premise of the decomposition introduced in the IG is that it embeds
nuanced, and institutionally relevant information. While this can neces-
sarily be interpreted in terms of the institutional content, i.e., the specific
actors involved, actions performed, contextual aspects, a varying perspec-
tive is the focus on the qualitative characterization of the linkages between
statements or institutions themselves. Beyond the specific content, the
nature and cardinality of linkages between statements or parts thereof
reflects analytical approaches best captured under the label Structural
Institutional Analysis.
Motivating this principle of structural analysis, Fig. 8.5 decomposes

the institutional statement in an Institutional Statement Decomposition
Fig. 8.5 Institutional State Complexity Metrics across Institutional Tree (An
enlarged version of this figure can be found under https://newinstitutionalgr
ammar.org/book-figures)
Diagram, a vertically oriented tree diagram that navigates through indi-

vidual components in fixed presentation order (here, ADIBC), and high-
lights the various navigational paths associated with individual compo-
nents (e.g., multiple action alternatives), and, where components are
nested, expands these into nested scopes embedded within the parent
statement. The nested scopes can either represent institutional state-
ments, or, as showcased here, institutional states (see Sect. 5.1.2). More
specifically, in this instance, these representations capture nested Acti-
vation Conditions. The purpose of this representation is to indicate the
composition of a statement in terms of its nested structure, while, at the
same time, highlighting state variations within all scopes the statement is
composed of (represented as dashed boxes in this example, and labeled
according to the level of nesting).
8.2.1.2 Institutional State Complexity

Reviewing this structure reveals various navigation paths that represent
the potential institutional states (or more accurately state compositions)
that this specific statement captures (see Sect. 5.1.2 for details on Insti-
tutional States ). Analytically speaking, this includes the most elementary
states of affairs (e.g., instances of behavior or existential state constel-
lations) that the statement describes, and in this case, regulates (since
the statement is regulative in kind). Notably, this includes all interaction
patterns of entities referenced in the institutional statement. Drawing on
the illustrative example, this pertains to the implicit interaction between
the Program Manager (making the top-level choice) and the activities of
the Certified Operation or Certifying Agent that the Program Manager
reacts to.
Analytically, the decomposition afforded under IG Extended thus
reveals the complete, atomic complexity of the institution captured as
far as captured in the focal statement, and can be descriptively captured
in a set of general metrics, captured under the umbrella Institutional
State Complexity. While the statement essentially relies on component-level
nesting on the Activation Conditions, with the two distinctive Activation
Conditions operating on the same level, reflecting an operational example
of horizontal nesting . Contrasting both Activation Conditions, the left
one further displays component-level nesting on the Direct Object that
reflects the content of the belief that the Program Manager can base
the initiation of proceedings on, capturing the variable complexity the
statement embeds.
In step with this, the discussion of analysis associated with IG Extended

introduces a set of complexity metrics that can be applied to individual
statements to explore the action/state variations captured in a state-
ment. The foundations of the assessment center on the logical operators,
and their aggregation across levels. Essential parameters thus include the
number of levels of a statement, which, for the illustrative example, are
three in total, and two levels of nesting (Level 1 and 2). The second
aspect relates to the nature of the logical operators. The semantics of
the different operators signal different extents of variability, or insti-
tutionally speaking, discretion or choice. AND-combined activities, for
instance, leave the actor without any discretion other than to perform
both actions, reducing the number of potential choices to one. If two
options are combined using the XOR operator, the actor has the choice
between those two alternatives, whereas the OR operator extends the
choice set to three, including the application of either action, as well
as their combination. Table 8.6 generalizes the State Variability for all
logical operators and any number of options, where k is the number of
options that are logically linked.
Referring to the flexibility associated with the different logical opera-
tors as Degrees of State Variability, the sum of these freedoms within a
given scope, then aggregated by statement level (across scopes on a given
level) reflects the possible states/activities the level captures. Generalizing
further, the product of these states determined per level represents the
Table 8.6 Degrees of State Variability for Logical Operators
Logical operator Degrees of State Variability Description
AND 1 The AND operator requires the

co-occurrence of the logically linked
component options; without discretion
on the part of the actor or situational
state variation
XOR k The XOR allows for the exclusive
choice of either of the k logically
linked component options
OR 2k −1 The OR allows for any combination
of the k logically linked component
options
k represents the number of options for a given component

number of states the statement captures as a whole. Given the embed-

dedness of levels within other levels, this aggregation is performed from
the deepest nesting level upwards, leading to the incremental aggregation
of scopes across levels to arrive at the statement-level metric.
Illustrating such complexity assessment for the statement shown in
Fig. 8.5, Fig. 8.6 schematically captures the individual aggregation steps
discussed in the following (with the left schema in Fig. 8.6 showcasing the
state characterizations associated with the individual states, and the right
schema illustrating the cumulative aggregation per level). Commencing at
the lowest level, Level 2 comprises of two logical combinations. Given the
presence of OR linkages for both options, Level 2 has a State Complexity
of six (i.e., two OR alternatives allowing for three state alternatives each).
On Level 1, the left branch of the tree allows only for a single state charac-
terization (the fact that the Program Manager believes ). Multiplied with
the State Complexity of the nested scope on Level 2, the left branch
captures six possible states. The right Level 1 branch, reflects the linkages
between different components and likewise displays a State Complexity
of six (two OR alternatives). Combined across both branches, Level 1
reflects a State Complexity of 12. The top-level statement offers a single
discretionary choice, but ensures exclusivity: the Program Manager can
decide to initiate either a suspension or revocation; not both. The degree
of variability on the top level is thus two. Combined with the State
Complexity on the lower levels, the overall statement captures possible
24 state constellations that the actors referenced in this statement can
Fig. 8.6 Institutional State Complexity Aggregation

variably be positioned in based on their individual choices and reactions,

or environmental circumstances. Imagining hypothetical constellations, if
both Activation Conditions were AND-combined (i.e., both need to be
met for the Program Manager to exercise any discretion at all), the State
Complexity would reduce to 12.
Reiterated algorithmically, the calculation of the Institutional State
Complexity for a given statement is initiated on the deepest nesting level,
initially determining State Complexity for individual Atomic Statements,
before integrating these per Nesting Level and finally aggregating those
to the entire Statement:
1. Identify atomic statement on deepest nesting level

2. Calculate Degree of State Variability for each logical linkage within
institutional state(ment) scopes
3. Calculate sum of the Degrees of State Variability per state(ment)
scope
→ Statement Scope Complexity
4. If multiple scopes exist on a given level (i.e., are horizontally nested),
calculate sum across scopes
→ Level Complexity
5. Multiply with State Complexity of next higher level (determined
using Step 2–4)
6. Repeat from Step 2 for next higher level, unless the top-level
statement is reached
→ Statement Complexity/Institutional State Complexity
Captured formally in Eq. (8.1), the Institutional State Complexity

is the sum of Degrees of State Variability for each operator
(DoV(operator(opIdx, optionCt))) within any level (where opIdx is a
unique identifier of the operator on a given level, and optionCt the
number of component value options linked via the logical operator),
k
multiplied across all levels of an institutional statement ( level=0 ).

k
m
I nstitutional State Complexit y = DoV (operator (opld x, optionCt))
level=0 op I d x=0
(8.1)
Reviewing this metric, it is important to note that the calculated states

are all possible state constellations captured by the institutional state-
ment as a whole, taking into account all parameters considered in an
institutional statement (e.g., environmental context, discretion, scopes
of applicability, etc.); institutional states do not only correspond to
intentional choices at the disposal of any one actor, or any actor at all.
Inasmuch as this metric provides a quantitative assessment only,
analysts may seek a qualitative characterization of the features that
produce the complexity. More specifically, the analyst may be interested to
learn where this complexity resides, and secondly, its characterizing effect
in the regulation or parameterization that the statement represents.2
The original IG characterizes abstract purposes of statements with
respect to the action situation they are embedded in based on distinc-
tive Rule Types (Ostrom, 2005). However, the refined decompositional
schema provides the basis to extract differentiated insights on the sub-
statement level (e.g., quantifying different features and affordances that
an institutional captures – such as discretion on the part of the actor,
as well as scopes of applicability). For the analyst, this offers the oppor-
tunity to either retrace the emergent complexity, or to selectively focus
on specific sub-statement metrics that respond to the analyst’s objec-
tives. Reviewing the components individually provides the basis for such
qualitative assessment.
As discussed before, the initial variation on Level 0 essentially reflects
different forms of sanctions that a Program Manager can impose in
response to noncompliance, with two action alternatives, signaling a
moderate form of discretion (XOR). Where the analyst is drawn to assess
the conditions (and their variable necessity) that lead to the application of
a statement, the statement may reflect the extent to which variable condi-
tions apply – not only their presence, but their variable presence (with the
strongest possible degree of variability in this case). Analytically, this can
provide an indication of the specificity of the circumstances to which the
statement applies, both indicated by the quantity of conditions, but more
importantly, by their dependencies (e.g., whether all conditions need to
be fulfilled, or any).3
2 For a quantitative differentiation of institutional states themselves (as opposed to

component categories), the reader is referred to advanced analyses presented under
Sect. 8.3.3.2.
3 Section 8.3.3.2 enhances this metric with a richer quantification of condition content.
The analyst can similarly be focused on elements embedded in nested

structures, such as the potential Activity State Variability, where the latter
reflects the potential states the provoking actor can be situated in; this
choice is not at the discretion of the responsible actor identified at the
top level.4 Similarly to the Conditions Variability, the metric signals the
exclusivity of either state either in terms of options (where those are OR-
combined), or in terms of necessities (where those are AND-combined).
Finally, the analyst may draw interest from analyzing contextual rela-
tionships, with particular focus on the specificity of contextual state
qualifications based on the Execution C onstraints. Reflecting the elemen-
tary form of reporting, the decomposition of such assessment can occur
on any level. Table 8.7 summarizes the Institutional Tree Metrics, both
including the statement-level complexity metrics explored in the begin-
ning of this section, as well as the component-level metrics introduced
before.
8.2.1.3 Institutional State Regimentation

By shifting from the aggregate perspective of State Complexity toward a
differentiated treatment on component level, as shown in Table 8.7, we
can not only analyze the institution from the perspective of possible states
it can capture, but rather get a sense of the specificity of the component
specification. For AND-combined activities, for instance, the variability in
state is limited; any compliant actor is obliged to perform both activities –
without any discretion; such linkage is captured as a single institutional
statement. The number of options, however, offers relevant insight about
the specificity of any component characterization, e.g., specificity about
the activities involved, specificity about the constraints that apply, speci-
ficity about the conditions that lead to the activation of the statement.
When reviewing the specificity of statements carefully, it is important
to note that the number of options and their linkages interact: Where
co-occurring (i.e., linked by AND operators), a higher number of compo-
nent value options (e.g., three, four, five) increases the specificity of the
institutional specification, thus introducing level of detail, and implicitly
parameters that require explicit consideration whenever the statement is
enacted (e.g., by actors subjected to the provision), or evaluated. If linked
by operators that afford a great extent of discretion (e.g., OR), an increase
4 The conceptual counterpart on the constitutive side is the Constitutive Function State
Variability.
Table 8.7 Institutional Tree Metrics
Metric Value Description
Institutional Tree Depth 3 Indicates the maximum nesting depth,

i.e., the maximum to which components
are substituted by institutional
state(ment)s
Institutional States 24 Indicates the maximum number of
institutional states described in the
statement
Discretion Option Count Counts the number of instances where
Level 0 2 actors have discretionary choices (per
level)
Discretion Extent Activity Variability (per level)
Level 0 2
Conditions Option Count Indicates the number of activation
Level 0 2 conditions (per level)
Conditions Variability Indicates the State Variability of
Level 0 3 activation conditions (per level)
Activity State Option Count Indicates the number of activity states
Level 1 2 (per level)
Activity State Variability Indicates the State Variability of activity
Level 1 3 states (per level)
Application Option Count Indicates the number of application
Level 1 2 alternatives (per level)
Level 2 2
Application Variability Indicates the State Variability of the
Level 1 3 application (here in the Object
Level 2 3 component) (per level)
in options signals the opposite effect, i.e., an increase in options increases

State Variability, and inadvertently reduces operational constraints based
on the flexible linkage of options (e.g., an increasing number of alternative
activities implies greater discretion on the part of the actor).
The Institutional State Regimentation metric captures this interaction
of the Variability of Institutional State (i.e., the number of possible states
a statement captures) with the extent to which these states are complex
and thus implicitly constrain, or regiment, an entity subjected to the
provision by dividing the number of options for any component-level varia-
tion (Option counts in Table 8.7) by the State Variability signaled by the
logical operator (see Table 8.6), as shown in Table 8.8.
Table 8.8 Institutional State Regimentation based on associated Logical Oper-

ators
Logical operator Regimentation Description
AND k Regimentation for AND-combined options are the

1
number of options divided by the single state that
AND-combined options reflect, effectively resolving
the number of options
XOR k Regimentation for XOR divides the number of
k
options by the number of state variations,
rendering a constant regimentation of 1
OR k Regimentation in the OR case reduces with the
2k −1
number of options (due to the non-linear increase
in state variability)
k represents the number of options for a given component
Whereas two AND-combined options lead to a regimentation of 2

(since both need to be taken into account), the XOR regimentation is
constant, i.e., 1 (with the number of options, the flexibility to choose
either of those increases to the same extent), and the OR linkage signals
a reduction in regimentation (while a larger number of options exist,
those lead to a disproportionate increase in state variation), here 0.6667.
Introducing four options, for instance, the corresponding levels of regi-
mentation are 4, 1, and 0.2667, indicating a high level of regimentation
for complex AND-combined options, a constant level for XOR-linked
options, and a considerable reduction in regimentation for OR-combined
options.
The aggregation of the Regimentation metric highlighted in Eq. (8.2)
operates analogue to the aggregation applied for State Variability high-
lighted in the previous section, and offers a basis for the comparative
assessment of individual statements or aggregates thereof (e.g., collections
of statements, entire policies).

k
m
optionCt
I nstitutional State Regimentation =
DoV (operator (op I d x, optionCt))
level=0 level=0
(8.2)
Reflecting on the introduced metrics, whereas the State Variability

metric introduced under the umbrella of Complexity Metrics focuses on
the identification of the total number of permissible states that an institu-

tional statement captures, the second State Regimentation contextualizes
this with a more accurate characterization of the degree to which these
states are constrained, or regimented, allowing the analyst to extract
differentiated insights such as identifying how broadly, or narrowly insti-
tutional state specifications are devised in the context of the analyzed
institutional setting, and to what extent they impact the regulated actors
or other entities subjected to those institutions.
These objective metrics only operate on logical linkages as well as the
general IG institutional statement structure, and can thus find general
application without disciplinary considerations (e.g., knowledge about the
domain or case). This dissociation from specific use cases implies an essen-
tial assumption: the metrics do not reflect the quality of the options they
operate on; they are considering the number of options and their link-
ages, but do not consider whether activities or states are qualitatively
different (e.g., harder to perform) from each other, or even whether one
activity/state subsumes the other. A qualitatively richer characterization
of components that draws on the application domain as an epistemolog-
ical basis (e.g., to introduce weights for particular options) is discussed in
Sect. 8.3.3.2.
It is, however, valuable to motivate specific application cases at this
stage. Focusing the assessment on activities, for instance, the logical link-
ages between action alternatives signal discretion on the part of the actor.
Interpreted in context, the distinctive analysis may reveal the representa-
tion of graduated sanctioning based on the variable effect (e.g., payoffs),
as well as the applied logical operator. This specifically applies to the XOR
that signals exclusive applicability of distinctive action choices, in contrast
to the arbitrary combination that the OR suggests (see Sects. 4.2.2 and
4.2.3).
Central aspects of the operationalization highlighted above are gener-
alizable. However, the choice of metrics, and the decision on which level
reporting occurs, are subject to analytical objectives, contingent on the
available data (see Chapter 7), and the employed level of expressiveness.
As implied throughout this section, the extraction of structural informa-
tion akin to that reported above requires encoding on the IG Extended
level. Any complexity assessment performed based on IG Core-coded
data is limited to the assessment of top-level component combinations,
as well as the metrics related to Activation Conditions. However, while

the IG Extended coding lends itself well for detailed structural analysis,
the statement-level data can be analyzed using the same metrics intro-
duced in the context of IG Core. Without illustration at this stage, the
complexity metrics can be aggregated across statements, and further-
more consider their linkages, selectively focusing on selected samples
of the encoded sources, or the policy as a whole. Similarly, complexity
can be captured in Policy Landscape Diagrams that draw out the State
Variability or Regimentation for individual statements or components,
alongside conventional distribution charts, the principles of which have
been discussed in Sect. 8.1.
8.2.1.4 Systemic Analysis

Beyond the micro-level complexity metrics introduced in the previous
section, based on the comprehensive nesting characteristics on the one
hand, and the focal emphasis on the interpretation of institutions as
compositions of uniform patterns that atomic institutional statements and
states represent, IG Extended affords a shift to a systemic perspective in
analysis. To this end, the analyst can abstract from the composite insti-
tutional statement as a de facto boundary of the systemic structure, and
instead, interpret the institutional setting entirely in terms of composi-
tional structures that aggregate from the atomic structures identified in
the encoding process, and explored exemplary in the previous section.
To schematically visualize such compositional analysis, we can draw
on the abstract compositional notation introduced in Chapter 5, the
elementary patterns of which are shown in Fig. 8.7.
Based on this visual notation, and drawing on the illustrative statement
discussed earlier (and repeated below), Fig. 8.8 exemplifies the compo-
sitional structure. The linked elements are annotated with the type of
Fig. 8.7 IG Compositional Patterns Overview (An enlarged version of this

Fig. 8.8 IG Compositional Patterns for Illustrative Statement
state/ment structure they represent (e.g., statement, state, constitutive,

regulative).
Instead of leveraging structural detail on statement level, composi-
tional analysis focuses on the interaction of the different compositional
elements. Following the structural analysis referenced above, the institu-
tional statement (bounded by dashed lines) captures the selective features
of state/ment representations introduced in Sect. 5.1.2, here specifi-

cally Behavioral State and Regulative Statements, and makes the essential
dependency patterns explicit. Central to this representation is not the
exclusive focus on an individual statement, but rather the expression of
compositional linkages of the provisions that statements capture. To moti-
vate this approach, a second statement (annotated in IG Script Notation)
is drawn into the analysis:
Explored semantically, the statement reflects an obligation on the part
of the certifying agent. Taken on its own, this statement is regulative in
kind, without showing any explicit linkages. Taken in context, however,
the obligation to report is central to the role of the certifying agent, the
violation of which may lead to the initiation of suspension or revocation
proceedings against a certified operation. The interpretation of context
here literally references the content of the Activation Conditions attached
to the original illustrative statement (“when ... a certifying agent ... fails to
enforce the Act or regulations in this part” ). While structurally distinct, in
the context of the action situation both statements thus integrate based
on an implied vertical linkage, with the original statement referencing the
initiation of suspension or revocation proceedings acting as a de facto
consequence of the noncompliance with the newly introduced statement.
Compositionally, the statements thus exist in the configurational structure
showcased in Fig. 8.9.
Given the potentially vast scope of institutional information and the
diverse tacit interlinkages, reliable inference of the semantics is facili-
tated by (a) decomposing of institutional statements in atomic forms,
and, more specifically, by (b) systematic detection of statement or state
patterns within components of institutional statements as afforded by IG
Extended.
Extending this compositional linkage of statements, the following
statement introduces additional obligations on the part of the Program
Manager:
Fig. 8.9 IG Compositional Patterns for two statements (Vertical Linkage)
“When an inspection ... of an accredited certifying agent by the Program

Manager reveals any noncompliance with the Act or regulations in this part,
a written notification of noncompliance shall be sent to the certifying agent.”
Encoded in IG Extended form that affords the reconstruction of the

activation condition part of the statement, the statement is encoded as:
Extending the compositional perspective further with the inferred
consequence that foregoing the notification of the noncompliant actor
implies a valid initiation of noncompliance proceedings, this can be
represented as a constitutive institutional state encoded as follows:
Figure 8.10 visualizes the compositional integration of the referenced

statements. While indicated in the schema for the sake of conceptual
linkage of elements to underlying statements, the systemic positioning
and organization of statements based on variable linkages essentially
Fig. 8.10 IG Compositional Patterns (Horizontal and Vertical Linkages) (An

enlarged version of this figure can be found under https://newinstitutionalgr
ammar.org/book-figures)
reflects the institutional structure of the action situation, composed of,

but transcending the individual statements.
Reflecting on the analytical opportunities associated with the composi-
tional patterns presented as part of the systemic analysis, the analyst may
ask a set of questions. Firstly, the systemic analysis provides a general
overview that leverages the linkages that exist within institutional infor-
mation encoded from documents, transcripts, etc. – seeking a general
impression of the institutional features. To this end, the analyst may be
interested in the nature of the linkages between individual state char-
acterizations, establishing a sense of the general foci of the institution.
Building on the discussion of analytical opportunities for IG Core-coded
data on how different aggregate institutional information signal different
institutional foci, drawing on institutional linkages captured under IG
Extended provides the facilities to expose additional nuance associated
with aforereferenced metrics.
• What is the nature of linkage between individual state character-

izations within and across statements, i.e., which statement types
dominate the analyzed setting, and how are they configurally related
(i.e., are regulative and constitutive statements linked, for example
in complementary function)?
• How is complexity (as determined as shown under Sect. 8.2.1.1)
distributed across different parts of the analyzed setting? Which
aspects of the action situation are central to the overall institutional
function, and which ones are lateral? Which features of the action
situation are more frequently visited (in analysis of institutions in
use)? What implications does this have for the overall assessment of
stringency?
• To what extent does the action situation contain explicit or implied
consequential features (potentially signaling primarily punitive or
facilitative emphasis of the analyzed setting)? Can the analyst identify
exceptional circumstances that are not governed by the institutions
in place?
Naturally, many of the observations intersect with the high-level anal-

ysis of component interactions on IG Core level (see Sect. 8.1); a key
objective of the systemic analysis, however, is to draw an explicit linkage
between the institution as expressed in institutional information (indepen-

dent of its source) and the institutional structure across arbitrary levels
of organization, including individual statements, selective combinations,
action situation, as well as institutional setting in its entirety.
This provides the analyst with higher-level opportunities alluded to
above, and further expanded here. A specific example of such analyses
is the detection of structural holes in the institutional information. This
may include the assessment of the density with which statements are inter-
linked. Based on the detection of isolated statements, or clusters thereof,
the analyst may be motivated to explore the underlying reasons, which
can, for instance, be characteristic for the specific institutional information
of interest at large, or be an artifact of the sampling performed as part of
the study (see Chapter 7), e.g., in terms of preconceived action situa-
tions. A further alternative consideration is the quality of the institution
itself, including its completeness, cohesiveness, or extensive reliance on
external institutional references that carry the complementary operational
regulative information.
While the compositional analysis, as introduced to this stage, eases
the assessment by abstracting from details, the analyst may nevertheless
be inclined to get a richer qualitative picture of the identified config-
uration, and selectively focus, or zoom in, on specific features of the
individual atomic statements to provide an intuitive insight of the under-
lying operational semantics. Figure 8.11 exemplifies this by calling out the
responsible actor (Attributes ) and activity (Aim) information, allowing
the reader to reconstruct the essential principles of the original state-
ment information alongside their configuration. Without illustration at
this stage, and drawing toward an enriched quantitative assessment,
the presentation can selectively be linked with complexity information
introduced in Sect. 8.2.1.1.
8.2.1.5 Extracting Conceptual Organization

Beyond the focus on the structural linkage between statements as a basis
to afford a systemic reconstruction of the institutional setting, institu-
tional information encoded on the refined IG Extended level provides
additional structural information about the institutional system. As intro-
duced throughout this book, the IG recognizes the presence of entities
of general kind that are (ideally) captured in the dedicated constitutive
statements that provide the basis to parameterize the institutional setting.
Fig. 8.11 IG compositional patterns example with selected component infor-

mation (An enlarged version of this figure can be found under https://newinstit
utionalgrammar.org/book-figures)
Examples of entities include actors, actions, objects, and artifacts as rele-

vant in an action situation, or institutional setting more generally, but
further extend to any form of entity originally discussed in the context of
the constitutive syntax (see Sect. 4.2.5), including actors, actions, objects,
artifacts, roles, infrastructure, status characterizations, etc.
In practice, however, in addition to the well-defined role of constitutive
statements, structural features pertaining to the conceptual organiza-
tion of institutional statements are captured throughout all forms of
institutional statements, most notably as part of Properties attached to
individual components. Realizing the importance of the tacit characteri-
zations, IG Extended draws on the Object-Property Hierarchy introduced
in Chapter 5. Combined with the Property Typology introduced in
Sect. 5.1.5, the analyst is able to categorize and extract the organizational
skeleton that the explicit and tacit linkages expressed throughout state-
ments, components, and Properties reflect. The Property Typology extracts
diverse kinds of Properties based on various types, including the general
distinction into Properties that are of quantitative and qualitative kind,
and where qualitative information is identified, stratifies such into simple
and complex information. Simple information is generally expressed in

adverbial form, whereas complex information essentially presumes an
embedded structural linkage to other entities in the institutional setting.
To illustrate this collation and organization of conceptual information,
the following statement is revisited (under consideration of Property Type
annotations):
This statement recognizes a set of different entities, including the

actors and objects established in an institutional setting, but also their
extended qualification based on properties. Actors identified as part of
the statement include the Program Manager and certifying agent. As
signaled based on the Properties annotation attached to individual enti-
ties, certifying agents can be accredited. Reviewing the qualification of the
noncompliance (“any” ), this reflects an operational linkage to the state-
ment the object is embedded in, as opposed to being a property that
is conceptually attached to the object itself (i.e., conceptually or func-
tionally dependent as discussed in Sect. 5.1.3). Contrasting this form of
qualification, the notification of noncompliance indicates a specific form of
notification that is conceptually linked to noncompliances observed in an
institutional setting, but it furthermore does not reflect the only form of
notification in statements, as evident from other statements such as the
following example:
When correction of a noncompliance is not possible, the notification of

noncompliance and the proposed suspension or revocation may be combined
in one notification.
Sampling across statements based on property and component type

qualifications, let alone the explicit referencing of entities in all compo-
nents, the analyst can reflect the conceptual organization of the entities,
as exemplary showcased (for the statements referenced in this section) in
Fig. 8.12.
Fig. 8.12 Excerpt of Conceptual Entity Organization in Scenario (An enlarged

version of this figure can be found under https://newinstitutionalgrammar.org/
book-figures)
An essential feature of the conceptual organization, or ontology of

concepts referenced in the encoded data, is the isolation of institutionally
relevant concepts on the one hand, and the implied semantic interlinkages
on the other. Concepts prevalent in the referenced institutional setting are
the referenced actors, as well as specific institutional acts available to an
official, such as the certification and proceedings. As part of this specifica-
tion, the sample makes reference to specific artifacts associated with any of
such acts, as well as a distinctive set of policy references occurring across
statements. The excerpt provided here, as indicated above, is selective in
that it focuses on the categorization of individual entities. Beyond this,
however, it offers insight into the interlinkages between the referenced
entities based on the configuration of these entities expressed in institu-
tional statements. The qualification by properties for actors, for instance,
suggests their respective characterization as accredited (for certifying
agents) and as either certified or uncertified (for operations). Specifically
the latter state is linked to the activities associated with the institutional
act performed by other actors, with the certification taking a focal role,
since it is implicitly referenced by an operation’s property. As far as the
structure suggests, certification can either be suspended or revoked; in
the context of the institutional setting, suspensions or revocations exclu-
sively associate with certifications. Other acts or artifacts may, however,
be qualified in terms of the function they play. Signaling a suspension
notification, for instance, highlights an implicit linkage between the noti-
fication and certification mediated via the suspension expressed as part of
the notifications. The links in Fig. 8.12 reflect these implied conceptual
linkages.
Recognizing the general conceptual structure of entities, the analyst

has an overview of the fixtures an institutional setting is comprised of.
By making the implied linkages captured in the complex relationship char-
acterizations expressed in the component annotations explicit, it provides
the basis to establish the conceptual integration of constitutive and
regulative statements, where the former define the structure, and latter
regulate the actors embedded in this structure. From this perspective,
the extraction of the conceptual organization of an institutional setting
operates orthogonal to the compositional statement linkages highlighted
in Sect. 8.2.1.4; instead of emphasizing the statement, and implicitly,
institutional actor interdependencies, this structure offers an entity-centric
organizational blueprint as a total view of the explicit and implicit organi-
zational linkages. Complementing the analysis on the compositional side,
the analyst may use this to draw insights about specification gaps in a
given policy, sample, e.g., identifying in how far entities, actions, etc.
are explicitly defined or merely referenced, let alone their relationships
as implied by institutional statements. Specifically, the latter aspect can be
realized by applying reasoning techniques that operate on the extracted
conceptual structure and associated semantics (see e.g., Staab & Studer,
2009).
This view on the organizational structure completes the sampling of
static approaches to the systemic analysis as afforded by IG Extended
features. The deep structural representation provides the basis to analyze
individual statements on atomic level based on general complexity metrics
that are further broken down to individual components in order to
attach the semantics that meet the analyst’s objective. The compositional
analysis builds on the structural decomposition based on the various
nesting capabilities and provides a structural overview of the institutional
patterns across various scopes, starting from individual statements, and
generalizing to arbitrary levels of statement aggregation. The conceptual
organization contextualizes the institutional linkages of statements.
All of the analytical approaches highlighted in this section can be
variably combined, but nevertheless focus on the static perspective of
analysis. To draw a linkage to the analytical opportunities that specifically
IG Extended affords for the study of institutions in simulated behavioral
settings, the next section sketches methodological pathways toward such
study, alongside exemplary research questions.
8.2.2 Institutional Modeling

Building on the extended insight sponsored by the deeper structural
analysis that IG Extended affords, both with respect to the quantifica-
tion of complexity and representation of institutional structure, a specific
promise associated with IG coding related to the computational use
of the encoded information is to move beyond a mere processing of
the coded data. Instead, IG Extended data can further be leveraged
to generate data based on computational representations that essentially
reflect synthetic institutions in use, offering the opportunity to comple-
mentarily or supplementarily engage in experimental studies to leverage
and derive institutional insight. The concept underlying the behavioral
use of IG data (whether coded or generated) is Institutional Modeling, a
brief background on which is provided in the following.
8.2.2.1 Background
Institutional Modeling, as referenced here, captures approaches that
emphasize the study of specific or systemic structure of institutions, with
primary focus on the description and/or re/construction of institutional
arrangements. Associated analyses may be static (i.e., focus on structure)
or dynamic (i.e., focus on behavior and/or change in structure) in kind.
With the primary emphasis on computational approaches, it builds on
the traditions of electronic institutions (e.g., Noriega, 1997; Rodríguez-
Aguilar, 2001) that implement institutions in mechanistic form to afford
automated enforcement, essentially reflecting the computational equiva-
lent to mechanism design and implementation, which de facto guarantees
regulatory compliance by design. Other approaches, such as the ones
emphasizes in this context, include the self-governance based on social
mechanisms built into computational models (e.g., Grossi et al., 2006;
Savarimuthu & Cranefield, 2011) in order to respond to questions related
to the socio-cognitive processes that lead to the emergence of institutions
more generally (e.g., Frantz, 2015; Morales et al., 2015), apply to char-
acteristic institutional arrangements (e.g., Ghorbani & Bravo, 2016), and
cover the socio-institutional and legal-institutional perspective (Frantz &
Pigozzi, 2018; Morris-Martin et al., 2019), respectively.5
5 A broader background on the computational modeling of institutions is provided in

Chapter 1.
8.2.2.2 Agent-Based Institutional Modeling

With the focus on computational approaches to institutions, this section
specifically references the computational modeling of artificial societies
that conceptually represent the real-world institutional setting, refer-
enced as Agent-based Institutional Modeling. Agent-based modeling, or
more accurately Agent-based Modeling and Simulation (Epstein, 2007;
Gilbert & Troitzsch, 2005), more generally, reflects an interactionist
approach to modeling and aims at reconstruction of macro-level patterns
(e.g., normative and institutional arrangements) based on micro-level
interaction between the associated entities (e.g., enforcement, decision-
making) (see e.g., Gilbert & Troitzsch, 2005) in a simulation setting.
The principal approach of drawing analytical insight from micro-level
information signals the paradigmatic compatibility of IG research with
agent-based modeling (see preceding discussion).
Naturally, such reconstruction requires the representation of agents
as proxies for humans, organizations, or any other form of entity that
exercises agency. Derived from richer representations in the area of multi-
agent systems (Shoham & Leyton-Brown, 2014), agents are generally
characterized by the ability to show social behavior based on their action
and reaction, potentially including the ability to communicate indirectly
(mediated via the environment) or directly (via communication). They
may further display a certain extent of autonomy in their decision-
making (see e.g., Castelfranchi (1995), Ferber (1999), and Jennings and
Wooldridge (2000)). In practice, the cognitive abilities of agents can
vastly differ and generally exist in a trade-off with the representational
necessities and computational bounds in terms of number of agents.
An essential determinant is the contextually required level of cognitively
plausible behavior displayed by an agent (see Epstein, 2014).
In addition to the agent, models of artificial societies rely on the pres-
ence of an environment that can, similar to the agent concept, be more
or less abstract, ranging from an accurate representation based on spatial
models, realistic resource capacity, to the effective absence. In addition
to agents and environment, scenarios are further characterized by the
relationships that exist between agents (Ferber, 1999).
An extension that is specific to the application of agent-based models
in the context of institutional modeling is the representation of institu-
tional concepts. While general applications to institutional analysis using
ABMs have been broad (e.g., Janssen (2009) for an example; Smajgl
and Barreteau (2017) for an overview), applications related to the IG
have been more limited. Models related to the IG more specifically have
covered the full range of institution types of the IG (see Sect. 4.2.3) and
may variably be represented external to the agents as part of the environ-
ment (e.g., Ghorbani & Bravo, 2016), represented in a dedicated agent
representing a proxy for the institution (see e.g., Smajgl et al., 2010),
or be distributed across agents (e.g., Frantz et al., 2015). Generative
processes associated with institutions, decision-making, and enforcement
mechanisms may exist in variable forms. Institutional models may further
be general in kind, or respond to modeling needs with respect to specific
forms of institutions. Notably, Ghorbani et al. (2013) propose a method-
ological process that specifically supports the formulation of common
pool resource problems in terms of agent-based models.
To this end, associated models build on a computational variant of
the Institutional Analysis and Design Framework (IAD), termed IAD+,
to generate templates of partially populated agent-based models based on
rich preconfigured structures that accord to the IAD framework.
8.2.2.3 Analytical Opportunities

This brief overview of the conceptual principles may leave the reader
with an initial impression of the principal opportunity to “replay” human
behavior in arbitrary scenarios that feature various forms of institutions,
but should also leave the sense that agent-based models are inherently
flexible, making them variably specific, or reflect institutional settings that
are able to respond to open-ended sets of questions. This is in contrast
to the structural analyses introduced for IG Core and IG Extended that
primarily offer descriptive insights into institutional arrangements, inde-
pendent of the source of the information (e.g., institutions in form,
institutions in use, various types). Moreover, the analysis is static in
kind in that it operates on fixed statements, or snapshots of institutional
statements. Agent-based models, in contrast, assume a flexible dynamic
approach that extends to the nature in which institutions are represented
or introduced in the system, as well as the analytical opportunities based
on the purposes they inherit from agent-based modeling more gener-
ally (see Edmonds and Meyer, (2017), J. Epstein (2008), and Gilbert
(2004)). While use cases and applications are diverse (e.g., illustration,
prediction), their value and orientation is contingent on objectives, as well
as available theoretical foundation, and data used to parameterize such
models. Targeted specifically at the institutional perspective, agent-based
models offer the following relevant purposes:
• Replication/Validation – Agent-based models can be used to recon-

struct institutional arrangements reported in the literature to lend
support and suggest validity based on a different techniques comple-
mentary to the ones reported in the literature basis (e.g., qualitative
study).
• Hypothesis/Theory Testing – Relatedly, agent-based institutional
models can be used to “replay” behavior, and by doing so, retrace
or postulate processes of path dependence in institutional decisions,
test hypotheses about the origin of the observed behavior (e.g.,
where factual information is missing, analyses are based on anecdotal
accounts, or validation of analyses is sought). The ability to replay
simulations arbitrarily often, and in varying constellations, provides
the basis for iterative refinement of models in response to theoretical
insights.
• Theory Development – Alternatively, or complementary, modelers
may look ahead, and postulate future realities in the form of “what
if” studies that “fast forward” models of human societies to serve as
a basis for prediction on the one hand, but, based on targeted exper-
imentation, can support theory development based on observed
phenomena.
In addition to the establishment of reliability, falsifying hypotheses,

ABMs can be used to develop explanatory accounts – an aspect show-
cased later – based on their dynamic execution principles, that can operate
complementary to traditional static analyses offered by the previously
introduced approaches, both to enhance the analysis, but also to populate
agent-based models in the first place.
A feature specific to agent-based models is the ability to flexibly
combine theory and data (Tolk, 2015), thus providing the opportunity
to compensate and complement theoretical features with data where exis-
tent, and conversely, fill gaps in data with model features derived from
theory. This combined interactive use of theory and data provides an intu-
ition of the principal flexibility that agent-based modeling affords, which
has earned simulation as the associated method the label of the “third
way of doing science” (Axelrod, 1997), complementing the dominating
traditional deductive and inductive approaches.
8.2.2.4 Institutional Modeling Principles

Turning to the practical use of agent-based institutional models in the
research process, this section establishes a basic understanding of the
process, provokes questions relevant in this process, and provides the
reader with a sense of how IG data can be used to support this process.
Building on the principal building blocks of agent-based models,
agents, their organization, and the environment, alongside the represen-
tation of institutional structures, the modeler is initially tasked to derive
the relevant information to variably support the design, parameterization,
and analysis of output generated by institutional models.
Methodological Considerations
As with any research, the identification of analytical objectives is
necessarily the primary concern. This is associated with the ques-
tion of whether the model is parameterized with de facto exogenous
institutional information, or variably produces, or generates, institutional
arrangements as part of the model execution.
Where input in the form of institution statements is sought, the struc-
ture of the resulting model critically relies on the presence of relevant
data, as well as the level of detail at which data is available. Another aspect
relates to the mapping of data to relevant entities that defines how the
data is used. Implied in this consideration is that institutional information
alone, as potentially collected based on encoded statements is insufficient
to design and parameterize a model entirely. In fact, the modeler is chal-
lenged to identify which parts or elements in the model can be populated
with institutional information (if any).
While information captured in collected institutional statements is
useful to characterize aspects of actor behavior, the extraction of this infor-
mation depends on a set of principal questions related to the nature and
quality of the data to guide the ensuing processing.
• Which institutional features and domain(s) do the data represent?

– This question is central to the model design, since it constrains
the key features in the institutional setting to be analyzed. What
domain is the setting associated with? Does the institutional
setting include consideration of environmental characteristics
(e.g., resources, topology), or is the environment inherently

abstract?
• Is the representation complete?
– Essential concern associated with this question is whether the
provided information is partial, and thus requires complemen-
tary information derived from context, literature, or based
on other information sources. It effectively signals the need
for additional data collection. An alternative approach is the
targeted exploration (e.g., based on assumed theorized insti-
tutions) to fill conceptual gaps.
• Does the data describe “institutions in form” or “institutions in
use”?
– This question relates to the nature of the institutions, specif-
ically, whether it reflects institutions in form or institutions
in use. Institutions in form invite for the implementation of
institutional statements within an institutionally parameterized
environment, e.g., implementation through a dedicated actor
that effectively reflects the governing authority. Institutions
in use, in contrast, inform the structuring of agent behavior
directly, generally represented in terms of execution cycles that
agents enact, and in which institutional information is reflected.
• Does the data include sanctioning information, and if so, of which
quality?
– While linked to the nature of the institution types (Sect. 4.2.3),
a central concern is whether sanctioning mechanisms are
present in the first place, and more importantly, what the
intended effect is. Do they reflect economic sanctions, status
moderation or include other forms of social, emotional, or even
existential consequences? Aspects of such nature have funda-
mental impact on how an agent is structured, the cognitive
makeup it needs to provide (e.g., an agent being able to feel
emotions is of different nature than an agent experiencing
economic sanctions), and what the agent is in the first place,
including its type (e.g., human, organization, State), multitude,
and diversity (homogeneous, heterogenous). Another consider-
ation is the origin of sanctions or consequences: if originating
from the natural environment, such facility must be considered

in the institutional model (how do agents sense the environ-
ment, or should the environment be represented as an agent in
its own right?).
Stratifying the discussion, and moving from the model design more
generally to individual elements, modeling concerns center on the associ-
ation of model elements and institutional information, notably asking the
following questions:
• Which actors are involved in the modeled setting, and what are their
characteristics?
• How are actors in an institutional setting organization-
ally/structurally related?
• Which actions can those actors perform? Which entities or environ-
mental features do or can they act upon (i.e., interact with, or react
to)?
• Which environmental characteristics are relevant in the institutional
setting?
With answers to those general questions, the modeler is in the position

to link the institutional data with elements in the model.
The mapping of information naturally depends on the nature of the
model and the kinds of questions the modeler intends to seek responses
to as part of the analytical process. As part of the design process, as indi-
cated above, the modeler has to consider the central aspects that are part
of the model, and establish which information can be sourced from insti-
tutional information alone, and which relies on contextual information
derived or sourced otherwise (e.g., based on literature, qualitative studies,
experimentation).
Looking at institutional information captured in institutional state-
ments more specifically, the modeler can draw on some natural mappings
that derive from the component-level syntactic characterization, the
statement-level classification, as well preliminary structural analyses, akin
to the kinds mentioned above.
Given the objective of the IG 2.0 on the IG Extended level to capture
institutions comprehensively, the operationalization of collected institu-
tional information is twofold, firstly contributing to the model structure
more generally, i.e., the identification of relevant environmental and insti-

tutional characteristics that institutional statements reference; the fixture
of the institutional setting. The second aspect relates to the runtime
parameterization with contextual information.
Focusing on the model design initially, it is important to state that
agent-based models implicitly reflect a closed environment (i.e., in simu-
lation settings, unlike open multi-agent systems, agents generally interact
within the system only in order to be able to develop deterministic models
in which behavior can be replicated). The scope of this environment,
however, is flexible. Naturally, it will provide the boundaries of the insti-
tutional setting in the wider sense, but it is at the designer’s discretion
to determine the specific scope, e.g., whether a scenario entails multiple
or only a single action situation. Where multiple action situations are
reflected, their differentiated consideration is important to unambigu-
ously identify and associate characteristics specific to the individual action
situations (e.g., positions).
Inasfar as the entities (e.g., actors, objects, actions, etc.) within the
model are concerned, such structural information is primarily derived
from individual component references captured in statements of both
constitutive and regulative kind. Specifically agency-centric regulative
statements bear a conceptual and structural commonality with agent
behavior that has motivated the uptake of the IG for computational
studies in the first place. This specifically includes the representation of
actors in Attributes (in the contemporary interpretation of the IG), activ-
ities in Aim, and receiver of activities in the Object component variants.
For other instances, the mapping may be less symmetric, including the
specific objects that institutional statements can hold, which can, where
annotated as animate, often assume the role of an actor, and conversely
represent environmental features, alongside the Activation Conditions
and Execution Constraints that reference environmental aspects of mate-
rial (e.g., bio-physical environment) or non-material kind, as well as
the potential institutional environment outside the analyzed setting
(i.e., the one that embeds the analyzed institutional setting), let alone
features of the institutional setting represented in the environment (as
discussed above). Importantly, where components capture nested state-
ments (e.g., based on component-level nesting (see Sect. 5.1.1)), those
are linked equivalent to first-order components. Figure 8.13 offers a
general overview of principal linkages between components in the IG and
ABM constructs.
Fig. 8.13 General conceptual mapping of IG components to agent-based

models
Reviewing the primary associations between components and ABM

constructs, regulative statements offer a more immediate linkage to
agent-based concepts due to the explicitly actor-centrism, and the
comparatively diverse characterization of systemic features by constitu-
tive statements, including actor-centric entities, and institutional as well as
environmental features. It is further important to note that this mapping
is general, but should be operationalized with specific cases or domains in
mind so as to generate an unambiguous structural linkage of institutional
information and model.
Turning to the use of institutional statements for the operational
parameterization of behavior regulation, especially if directly injected into
agent behavior (and thereby more closely associated with the dynamic
runtime structure of the model), a set of considerations needs to be taken
into account. To afford accessible modeling, institutional statements as
compound units, require decomposition to a level that corresponds to
the atomic unit of behavior agents can engage in (i.e., ideally in the form
of linked atomic institutional statements), institutional information is still
expressed in variable form, including stylistic preferences, writing tradi-
tions, forms of text, and other aspects discussed in Chapter 3. However,
barring other forms of syntactic preprocessing (e.g., stemming of concepts
where an automated mapping is considered), the representation in insti-

tutional models requires very specific methodological considerations
pertaining to the embedded or implied semantic information:
• Statements may be expressed in passive form, requiring a reconstruc-

tion in active tense, an aspect of principal relevance in the context of
the encoding of statements more generally (see Chapter 7).
• Statements may variably express sanctions in terms of conditional
statements, as opposed to the Or else structure assumed for insti-
tutional statements, demanding for a reconstruction in inconsistent
form that separates environmental circumstances as activators of
behavioral regulation from consequences of noncompliances.6
• Individual statements are necessarily expressed from the perspective
of one actor (i.e., the entity captured in the Attributes component),
but may carry implications for multiple actors (e.g., actors referenced
in the Object or Context components). Such statements require
corresponding reconstruction based on the principles of Perspective
Extrapolation, to be discussed in the upcoming Sect. 8.3.
• Properties associated with individual components have particular
operational relevance, since they, a) hold information that offers a
substantive characterization of the actor (e.g., nature of the actor),
b) express content that can be reconstructed in condition terms
(e.g., beliefs), and c) may capture quantitative information that offer
a qualification of regulative content of an institutional statement
itself (e.g., “no actor,” “some actors,” “all actors” ) that is relevant
for the compliance assessment in the model (independent of how
monitoring and enforcement are implemented).
• Finally, semantic annotations associated with individual components
(e.g., role characterizations), inform about the semantic linkages
that a model needs to afford (e.g., the ability to transmit objects
between entities, identification of parties affected by any transac-
tion), or context characterizations that are of institutional relevance
(e.g., temporal, spatial representations, distinctive states/transitions,
or other forms of circumstances), or offer enhanced characterization
of activities (e.g., contextual references to methods or tools that the
6 Relevant encoding practices are discussed in Chapter 7, and the logical treatment is
offered in Sect. 6.1.4.
modeler may consider representing). As indicated in Sect. 6.1.2, in

preparation for (or accommodation of) the modeling task, annota-
tions may be enhanced to afford a customized mapping in response
to analytical objectives or characteristics of the modeled setting,
or the modeling activity (e.g., reflecting distinctive functional roles
of agents, or accommodating concepts specific to the employed
modeling platform).
Naturally, the resolution of such challenges depends on the level of

expressiveness at which information is encoded. Specifically for the resolu-
tion of interactive linkages, the representation in the form of IG Extended
is important. However, the modeler can, as implied above, selectively
draw on features derived from higher levels of expressiveness.
Beyond the focus on the individual statements, the structure of entities
and their relationships modeled at design time, i.e., before the execu-
tion of the model as part of a simulation, can be derived based on the
structural analyses introduced earlier in this section. For instance, the
conceptual organization of entities (see Sect. 8.2.1.5) provides a concep-
tual blueprint of the relationships and the nature of the entities relevant in
an institutional setting, and can inform the selection of relevant elements
for a modeling exercise, and afford a specific mapping beyond the general
mapping presented above.
Moving closer to the design of agent behavior, the recognition of
action interdependencies is of relevance to model behavioral cycles that
agents engage in. Whereas institutional statements may express behavioral
guidance, in this context the primary focus is the presence and interde-
pendencies of activities more generally, aspects that are captured in the
compositional patterns introduced as part of the Systemic Analysis (see
Sect. 8.2.1.4), and support the design of general execution models that
describe agent behavior, specifically with respect to co-occurring activity,
alternatives, and consequential linkages.
A further aspect relates to the actual implementation of institutional
features in the system, including the consideration of the representation
of institutions in the first place (e.g., centralized, decentralized), as well
as the cognitive abilities that agents have with respect to the institutions,
signaling the diversity of capabilities captured in agent-based models more
generally:
• Do agents rationalize institutions explicitly, can they potentially

modify those representations, or reason about those (e.g., develop
attitudes toward those, or merely execute the rules)?
• Are agents able to violate institutionally prescribed behavior? How
are the consequences represented (an aspect discussed above)? Does
the model reflect graduated sanctioning?
• How (and by whom) is the violation of behavior assessed? How
is the observation of violations and the enforcement of sanc-
tions implemented (performed by independent entities (see
Sect. 4.2.3), centralized or decentralized)?
Responses to these questions will help the modeler decide on the

complexity of the agent model, alongside potential reasoning capabili-
ties (Balke & Gilbert, 2014), and the representation of relevant concepts
(e.g., affective behavior, economic principles, etc.), as well as central
system design decisions (e.g., presence of monitors, enforcers), which
may, similar to the considerations of agents earlier, be built into the envi-
ronment, as opposed to being represented as agents themselves (e.g.,
dedicated or distributed enforcement). Central here is the flexibility
that the modeler has with respect to the detail and comprehensiveness
of representation, leading variably to models that abstractly focus on
particular types of institutions more generally (e.g., Ghorbani & Bravo,
2016), the analysis of specific institutional arrangements that draws on
concrete environmental features (e.g., Smajgl et al., 2008), or empha-
sizing an inherently internal representation of institutions and foregoing
an explicit representation of the environment entirely (e.g., Frantz et al.,
2015). While previously only theoretically explored by Crawford and
Ostrom in the context of their original game-theoretical operational-
ization, the notion of Delta parameters (see Sect. 4.2.4) is a potential
candidate to operationalize institutional conformance based on environ-
mental or internal factors. Central in the context of modeling, however,
is the appropriate parameterization of associated payoffs in the insti-
tutional setting. In contrast to game-theoretical applications, however,
ABMs in principle allow for the dynamic adjustment of payoff structures,
enhancing the operational value of delta parameter in ABMs specifically.
Agent concepts are furthermore often enriched with learning capabilities
(e.g., based on reinforcement learning [Watkins, 1989]) to resemble more
plausible human behavior.
Having briefly highlighted the opportunities associated with ABMs,

this flexibility comes at the risk of developing models that offer insufficient
grounding based on compatible theory and empirical evidence (poten-
tially leading to validity problems), and more immediate risks associated
with the complexity of the programmed simulation model itself (making
verification, i.e., assessing the correctness of the implemented model itself,
more challenging). An extended range of problems, alongside mitigation
strategies, is discussed by Edmonds and Meyer (2017).
Exemplifying aspects of the modeling process, agents are designed in
terms of distinctive execution cycles specific to each actor in the system.
The execution cycle is parameterized based on contextual information
that describes empirically, theoretically, or hypothetically, how typical enti-
ties of a particular kind, including the strategies and choices they make
based on individual (ir-)rational operation, learning, social influence, or
whichever concept a modeler wishes to see implemented. It is this flex-
ibility that makes agent-based models, in principle, amenable to address
questions in any of the domains referenced in Chapter 1.
To illustrate a principal structure of a decision-making process an agent
engages in, Fig. 8.14 describes the internal process of a farmer in an
Organic Farming setting, who can choose to apply for organic farming
certifications, can decide to behave (non-)compliantly based on indi-
vidual preferences, and, on similar bases, engage in social monitoring.
Illustrated here conceptually, such behavior would then be implemented,
typically using a general-purpose agent-based modeling platform, such
as NetLogo (Wilensky, 1999), MASON (Luke et al., 2005), alongside
numerous others,7 or implemented directly using a general-purpose (e.g.,
Python, Java) or specialized programming languages (e.g., Julia).
Analytical Considerations
Based on parameterization, the instantiated (i.e., running) model then
recreates a network structure of the agent society (artificial society) in
which the agents behave according to the implemented behavior, poten-
tially reflecting complex social interrelationships based on organizational
features (agent relationships) embedded in the model. This is visualized in
Fig. 8.15 for different types of agents in the illustrative Organic Farming
7 See Abar et al. (2017) for an overview of various agent-based modeling and simulation
platforms.
Fig. 8.14 Exemplary Execution Cycle of a farmer in the Organic Farming

Scenario
Fig. 8.15 Generated Agent Society
scenario, which is comprised of few certifiers (in red color), the inspectors
(blue) endowed with monitoring tasks, and the large number of organic
farming operations they monitor (green).
Analysis of such models generally occurs based on the produced
output, and generally in aggregated form and as time series, e.g., to
reflect dominant behavior, formed structural groupings (akin to the figure
above) that can be analyzed using social network metrics, or environ-
mental characteristics, such as resource levels, economic outcomes, etc.
Fig. 8.16 Exemplary Institutional Statement Output
on various levels of aggregation. The temporal dimension highlights the

dynamics of such aspects over time, allowing the experimenter to observe
convergence towards stable (or unstable) states that generally form the
basis for further analysis (e.g., equilibria as indicators of effective institu-
tions in use). Robustness of simulation results is established by sensitivity
analyses across relevant parameter ranges and repeated runs.8
Specific to the discussion in the context of the IG, however, the
output can occur in the form of traces represented in the structure
of institutional statements. Interpreted in a dynamic system such as a
social simulation setting, such statements thus reflect the “institutions in
use,” capturing the institutional reality within the artificial society, while
reflecting a considerable level of complexity. Drawing on the key features
introduced as part of the IG 2.0, Fig. 8.16 highlights a set of institutional
traces that exploit the structural features of IG Extended and illustrate
8 For further introduction to social simulation principles see Gilbert and Troitzsch
(2005) and Railsback and Grimm (2011), for advanced methodological considerations
refer to Edmonds and Meyer (2017).
the opportunities associated with the direct use of institutional statements

in ABMs to parameterize as well as to generate institutional information
using the same structural representation.
Providing intuitive guidance on the interpretation, transcribed directly,
the first statement suggests that “Farmers must apply for certification, or
else farmers will earn profit of less then 500,” signaling the rationaliza-
tion that the principal application for certification is attractive based on
observations and experience.
The second statement is more complex and suggests that “certified
farmers must not violate organic farming rules once their certification
has been approved, or else certifiers will revoke the certification.” While
pragmatically expressed here, the Activation Condition (Cac) in this state-
ment essentially retraces the principal activities involved in the certification
process, signaling the interdependence of the individual activities that
lead to approval of certification as a precondition for compliance expec-
tations in the first place. Made explicit, the statement has the following
structure:
The third statement contextualizes and suggests that certified farmers

themselves should engage in peer monitoring once they received certifica-
tion, while laying out a more complex justification for such behavior based
on the unfolding dynamics (i.e., because of observed enhanced violation
behavior, or increasingly observed suspensions of certifications following
their approval).
The fourth statement highlights, for instance, an effect of learning.
Unless explicitly injected as part of the model design, agents do not have
any priors about compliance expectations, and may learn that compliance
with organic rules only applies if they are actually certified (i.e., they may
violate if they are not certified). The fifth statement exemplifies potential
notions of cognitive dissonance that agents can express, since it signals
motivations not to apply for certification, which is balanced by the leading
first statement that captures the motivation to apply. Aggregated across
both statements (note the numeric representation of the Deontic), the
agent can thus develop a differentiated and complex picture of the moti-
vations for compliance. Conceptually, the principles discussed here mirror
the principle of Delta parameters (see Sect. 4.2.4), an aspect that is further
discussed in Frantz et al. (2015) and Frantz (2020).
The final statement highlights potential observational capabilities
across entities captured in agent-based models, here, for instance, relating
to the certifier.
Summarizing this exposition of the stylized application, the reader
may appreciate the principles of agent-based institutional modeling more
generally, but also get an impression of how the IG can be integrated in
this process, both to inform the representation of characteristics of the
environment, agents and their relationships, and, of course, institutional
characteristics, whether used as parameters for a model (i.e., exogenous
to the model), or as institutions in use (i.e., endogenously generated).
Naturally, institutional information alone offers only partial infor-
mation for the model generation. Centrally, the underlying research
question should guide the modeling process, which informs how and
where institutional information is injected in the modeling process. The
design and parameterization of agent-based models necessarily relies on
extended contextual scenario information, including, for instance, infor-
mation about the number of agents in the simulated physical system,
behavioral characteristics not captured in institutional information (e.g.,
underlying motivational bases, assuming cognitive makeup), including
lifecycle patterns of agents (e.g., daytime structure, lifetime), frequen-
cies of interaction, duration of the simulation, aspects sourced from
complementary information as indicated above, or plausibly established.
Concluding this overview of the potential use of the IG in Agent-
based institutional models, both the IG and the principles of ABMs
exhibit compatibilities that makes their complementary use attractive.
Both ABMs and IG information, specifically with the refined structure
described in this book, are inherently flexible in the degree to which they
capture and express complexity. Institutional statements, as compositional
patterns based on uniform statement units, focus on the representation

of essential institutional information captured in the well-defined units of
institutional statements (with additional specificity sponsored by annota-
tion features). For the analyst, this offers conveniences to use agent-based
models as synthetic alternatives to human societies, affording broader
experimental opportunities, de facto unlimited opportunities to “replay”
the experiments, as well as drawing direct insights from the running
model.
Central to analytical approaches that embed agent-based modeling is
the ability to integrate institutional concepts in variable form as discussed
above, i.e., either to exogenously parameterize the simulated institu-
tional environments (as motivated in the discussion of methodological
considerations highlighted before), or to interpret “institution as the
output” of such models, signaling the institutions that emerge either in
response, or independent of institutional information that served as input.
In addition to the principal flexibility to include institutional informa-
tion either for the input or output of models, a central difference to the
previous methods is the inherently dynamic nature that, on the one hand,
allows the observation of dynamics over time, identifying the conditions
under which (and for how long) stability emerges. While the indication
of convergence of a dependent variable is common for many agent-
based models, institutional models specifically draw on the (in)stability
of institutional arrangements over time as an important reported variable.
From the perspective of the institutional analyst specifically, complex
social models are able to represent not only a wide range of institutional
facets, but are also able to integrate disciplinary perspectives that have
previously been represented distinctly in Chapter 1, for example, affording
the integrated consideration of economic, social-psychological, and legal
aspects in a model, and produce emergent results based on the employed
interactionist metaphor, an aspect that is mirrored in the IG. In addi-
tion to being agnostic to specific disciplinary perspectives, the structural
homogeneity offered by the IG makes it a candidate to serve as input,
i.e., to parameterize agents and/or environment, or as output (as shown
above).
Summarizing, agent-based institutional modeling drawing on the IG
as representational means:
• can capture dynamics over time, potentially including runtime

dynamics,
• can apply the complexity metrics introduced throughout this
chapter, depending on the complexity of the output (detail of
component representation, statement interlinkage of atomic state-
ments),
• can be aggregated on arbitrary levels of social organization, variably
representing individual agents (micro level), groups thereof (meso
level), or the society at large (macro level),
• can represent formal institutions, i.e., rules devised in a collective
action setting (e.g., Ghorbani & Bravo, 2016), or informal institu-
tions, based on norms that groups of individuals or the society adopt
over time (e.g., Frantz et al., 2015).
Contextualizing the analytical value of ABMs in contrast to real-world

societies, computational models allow for the controlled replay of experi-
ments, which can expose institutional information of arbitrary complexity
and accuracy at levels of detail that may be hard to establish in real-world
settings in the first place (see Watkins & Westphal, 2016). In addition
to making institutional information apparent to the experimenter, agent-
based models in principle allow for the linkage to the structures (e.g.,
agents, environmental preconditions) that caused them in the first place,
giving them explanatory power (Frantz, 2020), beyond the primarily
descriptive foci of other approaches. The reader should, however, be
left with the impression that, unlike people, agents can in fact “talk in
institutional statements,” affording artificial societies a special role in the
empirical study of behavior using the Institutional Grammar.
8.2.3 Discussion of Structural and Behavioral Analytical

Approaches
Inasmuch as the previous discussion focuses on dynamic approaches to
use or generate institutional information in terms of the IG, similar to
the static analysis principles highlighted before, approaches in this section
focus on the deep structural analysis that builds on a fine-grained parsing

of statement features.
On this basis, the structural analysis approaches introduced in
Sect. 8.2.1 provide fine-grained metrics of institutional statements that
can leverage the features of the IG to quantify institutional complexity in a
differentiated way that makes distinctive complexity features accessible to
the analyst. At the same time, the deep parsing that operates on the funda-
mental institutional structures motivated in Sects. 5.1.2 and 5.1.3 exposes
the linkages between individual statements in order to facilitate a systemic
analysis that enables an abstract assessment of the structural linkages based
on compositional representations, as well as a general organizational view
on the entities embedded in the institutional setting.
Building off the perspective on institutions as systems of institutional
statements, Institutional Modeling, and more specifically Agent-based
Institutional Modeling, was introduced as representative of a wider
range of dynamic approaches to institutional analysis, such as System
Dynamics (Forrester, 1971), alongside methodological considerations
linked to the processing of institutional information. Where static anal-
yses focus on institutional statements as input for assessment (e.g., to
establish complexity metrics), dynamic approaches provide the potential
to produce, or generate institutional statements as output (e.g., reflecting
emerging or evolving institutional arrangements). More specifically, the
dynamic analysis allows for novel analytical opportunities by emulating
behavioral assessments (e.g., based on simulated societies) that are repeat-
able, and can be used to test hypotheticals impossible or unknown in
real systems. Dynamic approaches can thereby draw on and be applied
in concert with static analytical approaches, e.g., to evaluate snapshots of
institutional system constellations.
Central to the approaches introduced in this section is the principal
applicability on arbitrary levels of aggregation, ranging from the analysis
of individual statements’ complexity to the composite structural repre-
sentation of institutions. Central difference, however, lies in the primarily
descriptive perspective of static analysis of existing arrangements and the
generative perspective of institutional models that responds to a broader
range of additional objectives (e.g., replication, hypothesis testing, etc.).
The exposition of approaches to structural and behavioral analysis in

this section has been organized with primary focus on the exploitation
of features introduced as part of IG Extended, i.e., deep structural infor-
mation. However, despite this extended structural focus aiming at more
comprehensive institutional analyses, most of the metrics, approaches, and
techniques are general in kind, and put limited emphasis on epistemo-
logical integration with theories or frameworks linked to the analyst’s
domain. The upcoming section addresses this by introducing a semantic
perspective on institutional information that draws on general institution-
ally relevant concepts, but further offers the analyst the opportunity to
tailor their analytical approaches. Important to note, however, is that
the introduced concepts do not inherently depart from the techniques
proposed in this section, but can selectively extend analytical opportuni-
ties or supplement methodological features, or be applied independently,
especially where the application is tailored toward objectives that focus on
explanatory accounts related to specific institutional statement content
(or collections thereof), and less on the reconstruction of structural
complexity at large.
8.3 IG Logico---Semantic Analyses

The final section of this chapter concentrates on the semantic perspec-
tive imposed on institutional statements. Most notably, motivations for
analysis on this level are opportunities to respond to research questions
that speak to the conceptual integrity of the encoded institutional infor-
mation itself, and attempts to evaluate and explain the meaning of the
institutional content captured in individual statements.
While the previous section initiated the extraction of features based
on a systemic perspective, primary emphasis lay on structural features.
This section complements this perspective by overlaying the structural
with a semantic perspective, initially to further abstract from linguistic
representations by exposing biases embedded in structure and alleviating
those by transforming statements accordingly. An initial approach in this
direction is the extrapolation of actor perspectives (referenced throughout
earlier sections of the book) that are not immediately overt in institutional
statements, followed by opportunities to draw out semantic features of
institutional statements more explicitly to allow for explanatory accounts

by developing conceptual and theoretical links to the analyst’s domain.
8.3.1 Perspective Extrapolation in Institutional Statements

Whether in policy, language, or generated by agents (see Fig. 8.16), state-
ments are necessarily formulated with a particular entity in focus (e.g., the
actor), and the effects that this actor has on its environment as commu-
nicated via the activity it can perform, and, of course, the target of such
activity (i.e., the object).
If attempting to work toward the reconstruction of the institutional
setting under consideration of all involved and directly or indirectly
affected actors, it becomes prudent to provide mechanisms that cannot
only capture language abstractly in compositions of uniform structural
patterns as discussed throughout this book, but to flexibly reconstruct
these statements themselves to explicitly characterize the effect each
statement has on all involved actors.
To this end, a central feature introduced under the label IG Logico is
the logical characterization of the IG, and an associated set of transforma-
tion rules (see Sect. 6.1.4) that enable the extraction and reconstruction
of statements in variable form. The approach introduced in this section
builds on these basic transformations and highlights how their combined
use can yield more complex operations of distinctive analytical value,
beyond the reconstruction to establish consistent representations of
statements (as discussed in Sect. 6.1.4 and Chapter 7).
We explore this borrowing a variant of the running example explored in
the context of the structural analysis under IG Extended (see Sect. 8.2.1),
expressed in IG Script format below.
Structurally, this corresponds to the following institutional tree struc-
ture:
This compound statement consists, as discussed earlier, of a range

of atomic states and a leading statement. Notably, this statement is
constructed from the perspective of the Program Manager as the focal
entity. Institutionally, however, this statement invites for select obser-
vations. Firstly, this leading statement is conceptually a consequential
statement, since it specifies a sanctioning provision targeted against non-
compliant entities (certified operation & certifying agent ). Secondly, it
implies obligations on the part of all three actors referenced in this
statement. In order to reconstruct the statement from the perspective

of all involved actors (Perspective Extrapolation), the logical transforma-
tions introduced as part of IG Logico provide a basis for the extraction
and reconstruction of three distinct statements that reflect the specific
provision from individual perspectives. Essential hereby is the implied
representation of institutional consequences (e.g., behavioral or existen-
tial consequences as a response to the nonfulfillment or violation of
provisions).
The basic algorithm to facilitate Perspective Extrapolation includes the
following steps:
Exemplifying this for the referenced statement, the leaf statements are
“certified operation has violated the Act ”

“certifying agent fails to enforce the Act ”
Following the reconstruction principles outlined above, the first state-

ment is constructed as follows (with inferences highlighted in bold
font):
As illustrated in the reconstruction, linguistic adjustments (e.g., to

accommodate the reconstructed (linguistic) grammatical sentence struc-
ture), are explicitly tolerated; essential for the application of the IG is that
the institutional meaning associated with the individual components and
their composite form is not affected (as elaborated in Sect. 5.1). While
this flexibility of abstracting from linguistic structure highlights a central
feature of the New IG, it is important to ensure that methodological
aspects or assumptions do not prohibit any deviation from the original
structural form of language.
Reconstructed in an analogous way, the second statement reads:
Combined with the original statement that had the Program Manager
in focus, the semantics associated with the original institutional statement
have been transformed to capture all inferred associated obligations on
the part of all referenced actors. While applied specifically to a statement
that displays diverse actors, the extrapolation equally applies to nested
statements that reference the same actor as the leading statement.
The value of the principle highlighted above is both of methodological
and analytical nature.
The methodological value lies in the ability to make institutional
text written from a specific perspective useful to operationalize institu-
tional models that assume a multi-actor perspective, an aspect relevant
for the construction of agent-based models as discussed as part of the
methodological considerations highlighted in Sect. 8.2.2.4.
Applied analytically, the extrapolation of provisions from multiple
perspectives offers a range of potentials, including the facility to identify
or remove intentional or unintentional bias embedded in the text itself.
Independent from intentional biases based on the application domain
(e.g., regulation addressed at enforcement), the emphasis on enforcement
and monitoring personnel, for instance, could be hypothesized to signal
a primarily punitive perspective, whereas the perspective of policy subjects
may signal enabling function and assume a more facilitative perspec-
tive. Similarly, the reconstruction may expose broader systemic biases
not evident at first sight, such as an over- or underregulation of specific
parties referenced in a regulatory document. The reconstruction could,
for instance, indicate the complete lack of oversight for any given party
referenced in the document, and thereby capture power disparity not
overtly captured in the text. Finally, in the design process, the recon-
struction could aid the assessment of policy quality by seeking validation
for a consistent representation from the perspective of all policy targets.
Fundamentally, however, the logical treatment provides the basis to
afford an extreme abstraction from the underlying linguistic form still
captured in the structure of statements encoded to IG Extended. IG
Logico-based analysis facilitates the analytical concentration on semantic
interpretation only, whereas the logical operations afford the abstrac-
tion from the underlying structure extracted as part of the deep parsing
afforded by IG Extended.
8.3.2 Chaining Transformation Rules

Beyond the specific reconstruction highlighted above, the extraction of
tacit actor perspective for the purpose of analysis, the chained application
of the various transformation rules introduced under Sect. 6.1.4 enable
the realization of complex reconstructions of institutions for diverse
analytical needs.
An example of such chained application of transformation rules is
the conversion of hybrid, and in the extreme case, polymorphic insti-
tutional statements into their corresponding regulative or constitutive
forms, albeit with the potential necessity to infer missing component
information.
To illustrate this approach, the reader may consider the following
statement:
Reflecting the structural form of a constitutive statement, it can
undergo a set of transformations to be represented in the consequen-
tial form. As a first step, the Properties associated with the Constituted
Entity can conceivably be reconstructed as Activation Conditions based
on the Properties-Conditions Transformation (see Eq. (6.8)), rendering

the coding as:
Parsing the Activation Conditions component deeply (i.e., according
to IG Extended), the statement can be expanded to:
At this stage, conditional features previously embedded in the state-
ment are overt. In addition, given the structural patterns embedded in
the Activation Condition, a possible reconstruction in existential terms is
possible by invoking the Conditions-Consequence Transformation:
Moreover, if the contextual reinterpretation of this statement as

regulative is possible (based on contextual knowledge of the institu-
tion) and admissible (e.g., to accommodate specific analytical use), the
following interpretation as a regulative-constitutive hybrid statement is
conceivable:
Note, however, that this reconstruction leads to an inference of
a responsible actor, i.e., the assumption that Councils themselves are
responsible for soliciting minority representation, a characterization that
may not necessarily be justifiable without further contextualization of the
original statements. However, beyond the discussion of the contextual
permissiveness of this specific reconstruction, the example highlights the

principal possibility to combine different transformations to accommo-
date specific methodological requirements (e.g., reconstruction for the
use in institutional models), or to widen analytical objectives based on
the extraction of the variable representations of coded statements.
Naturally, any application of transformation rules, especially where
involving the inference of elements relies on the permissiveness of such
transformations on methodological grounds, an aspect that should be
established and specified as part of the study design process.
Moving beyond the systematic reconstruction of statements that offers
distinct methodological and analytical opportunities, IG Logico further
introduces features that can be combined to establish novel macro-level
analyses grounded in the analyst’s domain or discipline, an example of
which is featured in the following.
8.3.3 Epistemological Linkage Through Institutional Functions

Analysis
The fundamental reconstruction facilities and the introduction of
semantic annotations as part of IG Logico provide the basis for the
customized application of the IG to respond to domain- or discipline-
specific analytical objectives. While IG Logico introduces features that are
of particular value when encoding has occurred on detailed levels, the IG
is devised to support flexible composition of feature sets (see discussion
in Sect. 6.2, and with more detail, in Frantz and Siddiki (2020)). If, for
example, only annotations of functions of statements, i.e., their Aim and
Constitutive Function, are of analytical value (while, for instance, fore-
going the deep structural parsing introduced under IG Extended), the
semantics captured in the selective annotation of Aim and Constitutive
Function can nevertheless be made accessible for downstream analysis.
8.3.3.1 Epistemological Linkage

A central focus of the analysis under IG Logico is the embedding or
linkage of institutional information to theoretical foundations, frame-
works, or other applications of interest to the analyst. IG Logico intro-
duces various taxonomies, most of which have general value. Notably,
however, Institutional Functions reflect the special role of drawing in
concepts from the analyst’s domain, theories, or frameworks and linking

those to individual statements, or specific components such as Aim
or Constitutive Function, and thereby introducing the epistemological
linkage between the Grammar and the theory/domain of interest. This
mechanism of IG Logico thus implicitly affords the IG the role of a theory
integrator, firstly by drawing the linkage in the first place, but in exten-
sion, affording the potential conceptual linkage between theories inasfar as
the different theoretical linkages overlap; extending beyond the mere
application of the IG for a distinctive analysis, but rather involving it for
the purpose of Theory Building.
Offering a practical display of the principles of drawing conceptual
linkages based on the taxonomies introduced under Sect. 6.1.2, this
discussion draws on a variant of the running complex example intro-
duced in Sect. 8.2.1.4, alongside the transformation exemplified in the
previous section, and further augments those with coding annotations
associated with IG Logico. Central in this encoding is the application of
regulative functions. As indicated in Sect. 6.1.2.4, the central purpose of
regulative functions is to annotate activities and functions with analytically
relevant functional characterizations sponsored by disciplinary epistemo-
logical frameworks, constructs (e.g., in qualitative studies), or other
mapping that best responds to analytical necessities, whether abstract
theoretical, or concrete operational.
The annotation exemplified here operationalizes the regulative func-
tions to reflect features relevant in the context of regulatory compliance,
notably identifying activities that signal violation, enforcement activities,
as well as potential monitoring and responses to enforcement. In the
example, two annotation types are used: regulative functions as well as
statement references.9
9 Details about coding conventions are discussed at greater detail in Frantz and Siddiki
(2020).
Both the regulative functions and the statement references are
further annotated to signal the substantive differences in component-
level combinations (e.g., “has violated [OR] is not in compli-
ance” as severe and moderate forms of violation, respectively, i.e.,
“I[regfunc=violate[severe,moderate]] ...” ). Analytically, such annotation
is able to introduce abstractions that map the linguistic expression
onto concepts of analytical value, while retaining the nuanced second-
order characterization associated with the respective variations. A second
example is the annotation of “initiate” as a sanctioning activity based on
contextual interpretation. Given the implicit specification of the alterna-
tive activities in the Direct Object Properties (i.e., “suspension [XOR] revo-
cation” ), the second-order qualification as “mild” or “severe” sanctions is
applied on this component.
This example highlights the operationalization of the encoding based
on semantic principles informed by the epistemological lens of the analyst.
Drawing on those values the analyst can now engage in advanced analyses
that augment the structural perspective (if of relevance) with a semantic
one, for instance exclusively concentrating the analysis on the function
and interaction of activities that carry some form of regulative function,
as conceptualized in Fig. 8.17. The schema highlights the focal treatment
of atomic statements with emphasis on the associated regulative function
(i.e., violate, sanction), making the interaction between those activities
explicit.
Fig. 8.17 Systemic Interlinkage of Institutional Function Annotations
This level of semantic analysis allows for the tracing of interaction

within an action situation, but also supports movement toward the
systemic perspective that no longer emphasizes interaction of statements
within an action situation, but further captures transitions between distinc-
tive action situations as effectuated by activities of interest. Exemplifying
such transition, assuming the noncompliance and ensuing sanctioning
activity, the sanctioned entity (certified operation) may conceivably appeal,
provoking a transition to a novel action situation centered around dispute
resolution (see Fig. 8.18), which is distinct from the original action
situation focused on compliance. Any subsequent statement may hence
involve a different set of actors, positions, actions, and further features
not relevant in the original action situation.
This focal shift on a primarily semantic perspective on activities, actors
(e.g., role characterizations), or other entities of interest highlights the
analytical value of IG Logico even for cases in which institutional infor-
mation is not comprehensively annotated – showcasing the selectively
dissociated feature sets of the different levels of expressiveness (e.g., if
only macro-level systemic analysis of regulative functions is of concern).
Fig. 8.18 Systemic Interlinkage of Action Situations
8.3.3.2 Semantic Systemic Analysis

Contrasting this abstract application of regulative functions, the use of
features of IG Logico on highly structured institutional information
encoded according to IG Extended offers extended analytical opportu-
nities that build on the analyses highlighted in Sect. 8.2.1.1. An example
for such augmentation is the quantification of the categories captured in
the second-order labels attached to combinations of activities and states.
Exemplifying this on the previously reconstructed statement, the focus
lies on the quantification of variable activities or scope references, essen-
tially the effect, in addition to capturing the state variability itself (see
Sect. 8.2.1.1). As implied in the qualifications of the regulative function
annotations, the varying options can carry different semantic strengths,
such as the differentiated characterization of different forms of violation
as either severe or moderate. Operationally, this provides the basis for a
systematic quantification not only of the extent of choice, but also the effect
that either choice can have. Figure 8.19 explicitly highlights this for the
annotated alternatives, with a simple mapping based on numeric values
illustrating quantification of severity and scope for different component
alternatives, respectively.
Fig. 8.19 Systemic decomposition with effect quantification (An enlarged

version of this figure can be found under https://newinstitutionalgrammar.org/
book-figures)
While the quantification exemplified here is illustrative, any such quan-

tification should be empirically or theoretically grounded, and factor in
features of relevance for the purpose of analysis (e.g., payoffs, external
effects). Moreover, any value association should be calibrated across the
institutional setting so as to provide the operational basis for the compar-
ative treatment and quantification across statements. This supports the
key premise of IG Logico to link the operationalization to concepts and
evidence drawn from the analyst’s domain and use case, signaling the
tailored use of the IG.
Use cases for this quantification are manifold. A more immediate
value is the use in parameterizing computational simulations, espe-
cially if operationalized in conjunction with the Delta parameter (see
Sect. 4.2.4) as a proxy for the integrated calculation of payoffs. In such
models, the quantified effect of activities can reflect the feedback an
agent experiences in response to a performed activity (e.g., violation).
However, for static models the quantification bears extended opportuni-
ties beyond the State Variability and Regimentation metrics introduced
in Sect. 8.2.1.1 by qualifying not only the linkage between alterna-
tives, but the options themselves. This moves the analysis of institutional
regimes beyond commonplace stringency assessments based on Deontic
values alone (as discussed in Sect. 8.1). Rather, it enables the comparative
evaluation of sanctioning regimes under consideration of graduated sanc-
tioning specifically, providing the basis to operationalize the associated
discipline-sponsored epistemology.10
However, such assessment is not limited to activities, or sanctioning.
Instead, it applies to any set of semantic annotations that follows ordinal
or higher-order scaling (e.g., interval scaling), or the corresponding
second-order qualification of the annotation labels as illustrated above.
A second example that highlights opportunities to develop metrics of
statement dependency in terms of linkages to other statements is the
annotation “Bdir[ref =[act,part]]((with the Act [OR] regulations in this
part))” referenced as Scope of Reference, which variably references parts
of the Act the statement is embedded in, or the Act in its entirety. Here,
the varying scope of both references is apparent, leading to a differ-
entiated assessment of this alternative. This characterization can be of
10 The fact that actions are OR-combined does, by itself, not provide any information
about the relative strengths or effects of sanctions.
particular use in contexts where external references are numerous or a

granular treatment of the varying scopes of references is of value due to
implications for the interpretation of the analytical outcome.
Beyond extending the principal quantification of effect to other
components, it enables their integrated assessment in generalized form,
e.g., combining the assessment of Deontics (e.g., may vs. must) with
potential sanctioning alternatives (e.g., fine vs. imprisonment) and their
scopes of applicability (e.g., per violation vs. for any violation). Given
the operationalization based on domain-specific information, the quan-
tification of individual component values allows for a differentiated
comparative treatment of statements (e.g., identifying high levels of
stringency signaled by the Deontic with narrow scope of applicability
signaled by the Context, or conversely moderating stringency levels across
wider scopes). This integrated quantitative treatment establishes the basis
for a comprehensive analysis of institutions in terms of their systemic
interlinkages and interactions by reflecting the effect of the individual
components contained in an institutional statement, while at the same
time, seeking epistemological embedding in the analyst’s domain, or
addressing methodological requirements or techniques.
Such integrated treatment is particularly valuable for components that
themselves embed complex institutional structures. An example of those
are Activation Conditions, since those reflect the variable applicability
of provisions under specific provisions, both in quantity (i.e., number
of conditions under which particular provisions apply) and their speci-
ficity (i.e., the broad vs. narrow nature of the condition based on the
circumstances it describes, e.g., general applicability vs. specific times
and places). Execution Constraints can similarly embed complex context
characterizations (e.g., signaling scopes of applicability, instrumental char-
acterizations, etc.) that qualify the applicability and execution of activities
or functions captured in the statement.
Irrespective of the specific instances highlighted in this example, the
quantification based on second-order qualification of semantic annota-
tion (as afforded by IG Logico) is a general principle that allows for
customized operationalization of parsed institutional statement content,
especially if parsed at greater levels of expressiveness (i.e., IG Extended).
8.3.4 Additional Analytical Opportunities

This overview of different analytical applications that specifically draw on
features offered by IG Logico is, as any of the preceding sections, illustra-
tive and selective. The breadth of the exposed features, including semantic
annotations of various kinds, as well as the exemplified statement trans-
formations allows for richer analyses that further combine these features.
However, other overt opportunities include the reconstruction of state-
ment reference networks based on the external linkages of statements,
alongside the scoping information as motivated in the previous section,
drawing on techniques highlighted in Sect. 8.1.
Central to the use of IG Logico for analytical purposes is the breadth
of customization opportunities that can respond to field studies that
rely on the mapping of carefully designed constructs onto the structure
of institutional statements. Similarly, institutional modelers can draw on
taxonomies that respond to specific model structures or model domains
(e.g., CPR models) in order to link entities operationally to encoded
institutional information. Conversely, institutional models can themselves
produce additional annotations as part of their own output to facili-
tate a richer contextualization of their analysis by drawing immediate
epistemological relationships generated from within the model.
The annotations of IG Logico further provide the basis for analyzing
semantic gaps hinting at institutional quality concerns. A more central
application is the detection of disparities in accountability and oversight
as reflected in the reconstructed institutional statements reflecting an arbi-
trary actor position that makes not only structural, but also semantic
gaps in any set of specifications overt based on absent consequential
statements.
Further examples of richer semantic treatment of encoded information
include the identification of incongruent specifications of jural correlatives
(Hohfeld, 1913), such as detecting the presence of rights as annotated
based on the Constitutive Function taxonomy, and the linking of the asso-
ciated duties potentially represented in the form of regulative statements.
Naturally, such mapping requires semantic references that operate across
the distinctive institutional statement type annotations. Documenting
this linkage, however, can be afforded using the provided customiz-
able semantic annotation mechanics, especially under consideration of
the second-order qualifications illustrated above.
In addition to richer sets of taxonomies and ontologies for the

annotation of particular components, the development of higher-level
ontologies (Staab & Studer, 2009) can offer contextual integrations of the
semantics currently captured in isolation, and thus provide the basis for
semantic reasoning across institutional statements and the entities linked
therein, developing the IG further to support explanatory accounts of
institutional analysis. Applying this perspective, conceptual aspects such
as the extracted entity structure based on constitutive statements and
Object-Property relationships, as well as behavioral aspects that link those
concepts, are integrated via general or domain-specific ontologies that
draw on the semantic annotation principles introduced under IG Logico
(see Sect. 6.1.2). These can provide the basis to query institutional infor-
mation based on complex logical criteria that operate on the encoded
and semantically annotated institutional structure and the implied link-
ages. By introducing stricter hierarchical organization of concepts, their
interrelationships, associated rules and underlying axioms, ontological
classification schemes close the gap to perform legal reasoning (or logical
reasoning, such as consistency assessments) on the encoded institutions
and institutional setting more generally, an aspect centrally linked to well-
established techniques in the area of Legal Informatics (Barabucci et al.,
2010; Katz et al., 2021).
8.4 Summary of Analytical

Approaches and Opportunities
This chapter provided an overview of different approaches to institutional
analysis using IG-coded data. The chapter intentionally abstracted from
the explicit reference to analytical frameworks, but rather attempted to
develop an exposition of analytical directions as well as techniques that
can draw on the features of the Institutional Grammar.
More specifically, the overview provided in this chapter responds to
the variable objectives and trade-offs associated with the different levels
of expressiveness introduced as part of the Institutional Grammar and
reiterated in Fig. 8.20.
The initial section Sect. 8.1 built off contemporary analytical
approaches that draw on the original IG, with specific focus on the
descriptive analysis of component value distributions (e.g., dominant
actors, actions), as well as combinations thereof (e.g., dominant actor-
action associations). Drawing on the nesting features introduced under
Fig. 8.20 Analytical Applications by Levels of Expressiveness
the IG 2.0 specifically, basic analyses further provide the novel oppor-
tunity of drawing out a basic Policy Landscape that provides a sense of
the complexity embedded in different parts of documents or datasets. As
a final aspect, this section works toward the reconstruction of the insti-
tutional network based on involved actors, objects, artifacts, and their
interactions, or in fact statements more broadly (moving toward struc-
tural analysis). Beyond providing a broader overview of the institutional
setting, this representation draws out implicit hierarchical linkages based
on the directionality of activity. The generated networks can further be
described based on conventional network metrics.
Moving beyond basic analytical approaches to institutions as offered
based on IG Core coding, analysis of institutional information encoded
based on IG Extended works toward developing a comprehensive under-
standing of institutions as systems of interlinked statements. The analysis
specifically draws on the structural features of institutional statements
exposed via deep structural parsing. This includes leveraging extended
complexity metrics that operate on interlinked atomic institutional state-
ments, as well as on component level, in order to extract institutional
detail not accessible by a coarse-grained statement-level encoding. This
includes the detailed assessment of potential institutional states captured

by a given institutional statement as a proxy for the complexity and
regimentation of a statement. Beyond the micro-level assessment of
statement-level complexity, the compositional analyses introduced in this
section motivate the resolution of inter-statement linkages in order to
reconstruct institutions in the form of systems – exposing the overall
structure of an institution. Complementing this systemic view on insti-
tutions, conceptual organization diagrams are introduced to make the
fixtures, or elements, of an institutional setting accessible.
These static forms of analysis provide the basis for dynamic assessment
in agent-based institutional models. Specifically, the nature of research
questions and objectives associated with institutional models, such as
replication of reported scenarios, and the testing of hypotheticals, is of
concern. An essential aspect of this illustration is to draw attention to
methodological considerations as far as relevant and applicable to IG
information. Concluding this explanation, exemplary results in the form
of institutional statement output is discussed to seed intuitions about
the opportunities underlying institutional modeling approaches. A specific
feature that links the coding on IG Extended level and agent-based insti-
tutional modeling specifically is the systemic perspective and the selective
focus of analysis operating across arbitrary levels of aggregation.
A central objective of analyses under the label of IG Extended is
the detailed representation of institutional structure, providing the basis
for comprehensive institutional analyses that can draw on arbitrary struc-
tural features, on various levels of aggregation (micro, meso, macro) and
further augment the static analysis with a dynamic perspective, e.g., in the
form of behavioral models to help understand existing systems, or explore
hypothetical futures.
The final Sect. 8.3 focuses on the features introduced by IG Logico,
with particular focus on the semantic transformations of statements to
afford a differentiated perspective based on various involved actors using
the principle of Perspective Extrapolation. Underlying motivations include
the methodological support for the development of institutional models,
as well as distinctive analytical opportunities associated with the exposure
of bias as expressed in institutional information. At the same time, the
transformations introduced in this section provide the basis for advanced
systemic analyses that build on the structural analyses established in
Sect. 8.2. However, they extend those by introducing semantic anno-
tations that not only capture the semantic categories associated with the
distinctive associated taxonomies, but moreover, operationalize the anal-

ysis quantitatively to extend the assessment of variability or discretion as
established under IG Extended. This notion of transformations specifically
broadens the pathway toward to date unknown comparative analyses that
can apply within and across sets of statements, or across policies more
generally, by facilitating their transformation and reconstruction to estab-
lish semantic compatibility as a starting point for comparative analyses in
the first place.
Central to the analysis under Sect. 8.3 is the explicit focus on the
semantic perspective, and thereby provides a twofold purpose. On the one
hand, the abstract treatment of institutional information permits complex
reconstruction of institutional statements to accommodate analytical and
methodological needs. On the other hand, it draws in the epistemological
perspective of the analyst based on disciplinary framing, mapping of theo-
retical constructs, as well as accommodation of methodological concerns.
A practical example of the second aspect is the linkage of the IG to
concepts from the area of Legal Informatics (Katz et al., 2021), affording
legal reasoning based on legal-theoretical constructions superimposed on
the analyzed institutional setting and domain-specific features.
Hence, IG Logico not only completes the levels of expressiveness by
providing the semantic perspective to augment and contextualize the
structural detail offered by the lower levels, but also offers an open
disciplinary interface that is targeting researchers of any disciplinary back-
ground, whether primarily studying institutions in form, in use, whether
applying formal or empirical techniques, whether focusing on a micro-
scopic or macroscopic analysis, whether approaching the study from an
academic or practitioner perspective.
Figure 8.21 organizes the introduced metrics and analytical tech-
niques based on the degree to which they facilitate a general analysis or
more comprehensive analysis of institutions, and furthermore differenti-
ates between the static, dynamic, or formal approaches and techniques to
institutional analysis as discussed in this chapter or envisioned for future
adoption in the context of IG research (with the left side featuring funda-
mental analyses based on IG Core, moving toward progressively more
complex ones toward the right side). The characterization is not abso-
lute, but rather offers orientation with respect to the recommended level
of institutional detail (in terms of levels of expressiveness) when applying
Fig. 8.21 Summary of analytical approaches and metrics based on analytical

focus
these techniques. As set out in the beginning of this chapter, IG Core

provides the basis to measure institutional design that advanced tech-
niques can leverage to establish structural understanding of this very
design (using features from IG Extended). IG Logico takes the final step
toward explanation of institutional design at any scope (e.g., individual
statements, institutional system at large) and through various theoret-
ical lenses and varying degrees of formality. Whereas metrics directly
operate on encoded information, specific techniques generally require
methodological accommodations in the form of preprocessing or trans-
formation to make the data analytically accessible. Inasfar as structurally
and conceptually compatible, the introduced metrics and techniques are
then combined as part of the analysis.
Looking ahead, rather than reading the discussed approaches as finite
or comprehensive on their own, the illustrations and expositions in this
chapter highlight opportunities for the richer analysis of institutional
information, and aim at stimulating novel analytical applications that may
draw and move beyond the introduced metrics and techniques to develop
novel perspectives on the rich institutional information encoded using
the refined Institutional Grammar – treating the IG as a interdisciplinary
interface amenable to diverse perspectives and objectives.
References
Abar, S., Theodoropoulos, G. K., Lemarinier, P., & O’Hare, G. M. P. (2017).
Agent based modelling and simulation tools: A review of the state-of-art soft-
ware. Computer Science Review, 24, 13–33. https://doi.org/10.1016/j.cos
rev.2017.03.001
Axelrod, R. (1997). Advancing the art of simulation in the social sciences.
Complexity, 3(2), 16–22.
Balke, T., & Gilbert, N. (2014). How do agents make decisions? A survey.
Journal of Artificial Societies and Social Simulation, 17 (4), 13. https://doi.
org/10.18564/jasss.2687
Barabucci, G., Cervone, L., Palmirani, M., Peroni, S., & Vitali, F. (2010). Multi-
layer markup and ontological structures in Akoma Ntoso. In P. Casanovas,
U. Pagallo, G. Sartor, & G. Ajani, (Eds.), AI approaches to the complexity
of legal systems. complex systems, the semantic web, ontologies, argumenta-
tion, and dialogue. AICOL 2009. Lecture Notes in Computer Science, Vol.
6237. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-
16524-5_9
1177/1065912909334430
Castelfranchi, C. (1995). Guarantees for autonomy in cognitive agent architec-
ture. In M. J. Wooldridge & N. R. Jennings (Eds.), Proceedings of the workshop
on agent theories, architectures and languages, ATAL’94 (pp. 56–70). Springer.
https://doi.org/10.1007/3-540-58855-8_3
Chun, Y. H., & Rainey, H. G. (2005). Goal ambiguity in U.S. federal agencies.
Vol. 15(1), 1–30. https://doi.org/10.1093/jopart/mui001
2975
Edmonds, B., & Meyer, R. (2017). Simulating social complexity - A Handbook.
Springer, Cham. https://doi.org/10.1007/978-3-540-93813-2
Epstein, J. (2008). Why model? Journal of Artificial Societies and Social
Simulation, 11(4). http://jasss.soc.surrey.ac.uk/11/4/12.html
Epstein, J. M. (2014). Agent_Zero: Toward neurocognitive foundations for gener-
ative social science. Princeton University Press. https://doi.org/10.23943/
princeton/9780691158884.001.0001
Ferber, J. (1999). Multi-agent systems - An introduction to distributed artificial
intelligence. Addison-Wesley.
Forrester, J. W. (1971). Counterintuitive behavior of social systems. Theory and

Decision, 2, 109–140.
Frantz, C. K. (2020). Unleashing the agents: From a descriptive to an explana-
tory perspective in agent-based modelling. In H. Verhagen, M. Borit, G.
Bravo, & N. Wijermans (Eds.), Advances in social simulation. Springer
proceedings in complexity. Springer, Cham. https://doi.org/10.1007/978-3-
030-34127-5_16
Frantz, C. K. (2015). Agent-based institutional modelling: Novel techniques
for deriving structure from behaviour (Doctoral dissertation). University
of Otago. Dunedin, New Zealand. https://ourarchive.otago.ac.nz/handle/
10523/5906
Gerber, B. J., & Teske, P. (2000). Regulatory policymaking in the American
States: A review of theories and evidence. Political Research Quarterly, 53(4).
https://doi.org/10.1177/106591290005300408
Ghorbani, A., Bots, P., Dignum, V., & Dijkema, G. (2013). MAIA: A framework
for developing agent-based social simulations. Journal of Artificial Societies
and Social Simulation, 16(2), 9. https://doi.org/10.18564/jasss.2166
Gilbert, N., & Troitzsch, K. G. (2005). Simulation for the social scientist. Open
University Press.
Gilbert, N. (2004). Agent-based social simulation: Dealing with complexity. The
Complex Systems Network of Excellence, 9(25), 1–14.
Grossi, D., Aldewereld, H., & Dignum, F. (2006). Ubi Lex, Ibi Poena:
Designing norm enforcement in e-institutions. In P. Noriega, J. Vázquez-
Salceda, G. Boella, O. Boissier, V. Dignum, N. Fornara, & E. Matson (Eds.),
Coordination, organizations, institutions, and norms in multi-agent systems ii
(pp. 107–120). Springer. https://doi.org/10.1007/978-3-540-74459-7_7
Hohfeld, W. N. (1913). Some fundamental legal conceptions as applied in
judicial reasoning. The Yale Law Journal, 23(1). https://doi.org/10.2307/
785533
Huber, J. D., & Shipan, C. R. (2002). Deliberate discretion? Cambridge
University Press. https://doi.org/10.1017/cbo9780511804915
Janssen, M. A. (2009). Understanding artificial Anasazi. Journal of Artificial

Societies and Social Simulation, 12(4).
Jennings, N. R. (2000). On agent-oriented software engineering. Artificial Intel-
ligence, 117 (2), 277–296. https://doi.org/10.1016/S0004-3702(99)001
07-1
Katz, D. M., Dolin, R., & Bommarito, M. J. (Eds.). (2021). Legal informatics.
Luke, S., Cioffi-Revilla, C., Panait, L., Sullivan, K., & Balan, G. (2005).
MASON: A multiagent simulation environment. Simulation, 81(7), 517–527.
https://doi.org/10.1177/0037549705058073
Morales, J., López-Sánchez, M., Rodríguez-Aguilar, J. A., Vasconcelos, W., &
Wooldridge, M. (2015). Online automated synthesis of compact normative
systems. ACM Transactions on Autonomous and Adaptive Systems, 10(1), 2:1–
2:33. https://doi.org/10.1145/2720024
Morris-Martin, A., De Vos, M., & Padget, J. (2019). Norm emergence in
multiagent systems: A viewpoint paper. Autonomous Agents and Multi-Agent
Systems, 33(6), 706–749. https://doi.org/10.1007/s10458-019-09422-0
Noriega, P. (1997). Agent-mediated auctions: The fishmarket metaphor (Doctoral
dissertation). Universitat Autonoma de Barcelona. Barcelona
Press.
Railsback, S. F., & Grimm, V. (2011). Agent-based and individual-based
modeling: A practical introduction. Princeton University Press.
Rodríguez-Aguilar, J. A. (2001). On the design and construction of agent-
mediated electronic institutions (Doctoral dissertation). Universitat Autonoma
de Barcelona. Barcelona.
Rogers, E. M. (2003). Diffusion of innovations (5th ed.). Simon; Schuster.
Savarimuthu, B. T. R., & Cranefield, S. (2011). Norm creation, spreading and
emergence: A survey of simulation models of norms in multi-agent systems.
Multiagent and Grid Systems, 7 (1), 21–54.
Shaffer, R. (2021). Power in text: Implementing networks and institutional
complexity in American Law. The Journal of Politics. https://doi.org/10.
1086/714933
Shoham, Y., & Leyton-Brown, K. (2014). Multiagent systems: Algorithmic, game-
theoretic, and logical foundations. Cambridge University Press.
97.x
Smajgl, A., & Barreteau, O. (2017). Framing options for characterising and
parameterising human agents in empirical ABM. Environmental Modelling
and Software, 93. https://doi.org/10.1016/j.envsoft.2017.02.011
Smajgl, A., Izquierdo, L., & Huigen, M. G. A. (2010). Rules, knowledge and
complexity: How agents shape their institutional environment. Journal of
Modelling and Simulation of Systems, 1(2), 98–107. http://www.hyperscie
nces.org/JMSS/Iss.2-2010/JMSS-3-2-2010.pdf
rule changes in an institutional context: The ADICO sequence. Advances in
Complex Systems, 2(11), 199–215. https://doi.org/10.1142/S02195259080
0157X
Staab, S., & Studer, R. (Eds.). (2009). Handbook on ontologies. Springer. https://
doi.org/10.1007/978-3-540-92673-3
Tolk, A. (2015). Learning something right from models that are wrong: Epis-
temology of simulation. In L. Yilmaz (Ed.), Concepts and methodologies for
modeling and simulation: A tribute to Tuncer Ören (pp. 87–106). Springer.
https://doi.org/10.1007/978-3-319-150963_5
Watkins, C. (1989). Learning from delayed rewards (Doctoral dissertation).
Cambridge University. Cambridge (UK).
Watkins, C., & Westphal, L. M. (2016). People Don’t Talk in Institutional
Statements: A Methodological Case Study of the Institutional Analysis and
Development Framework. Policy Studies Journal, 44, 98–122. https://doi.
org/10.1111/psj.12139
Wilensky, U. (1999). NetLogo. http://ccl.northwestern.edu/netlogo/
CHAPTER 9
Contextualization and Future Development

of the Institutional Grammar
9.1 The Institutional Grammar: An Analytical

Paradigm for Institutional Analysis
Institutional analysts from public policy, political science, law, philos-
ophy, computer science, among other disciplines, have engaged institu-
tional analysis extensively to generate understanding of the institutions
that govern social, natural, and/or artificial domains. Three perspec-
tives predominantly guide institutional analysis as conducted by scholars
from different disciplines – rational choice institutionalism, historical insti-
tutionalism, and sociological institutionalism. Associated with each of
these three perspectives are assumptions regarding what institutions are,
how institutions function within communities in which they embed, and
how institutions change. These different perspectives have been embraced
by scholars from various disciplines, such that while the discipline-
specific theories and frameworks with which they are integrated may
reflect field-specific foci, they tend to be motivated by common research
questions.
Among the questions that scholars commonly explore, are: What
are the qualities of institutions used to govern behavior within social
systems? How do institutions emerge? When and how do institutions
change? Among the aspects that differentiate research engaging these
questions that is conducted by scholars with different disciplinary back-
grounds is how they go about responding to these questions – i.e.,

Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2_9
the methods undertaken to investigate them. This trend is reflected in

contemporary IG research. Whereas policy scholars have engaged content
analysis of policy text (leveraging manual extraction of institutional infor-
mation according with the IG syntax) to study institutional change,
computer scientists and computational social scientists have studied the
same through agent-based modeling (using the IG to characterize model
parameterizing institutions or emergent institutions).
Recent IG research highlights an integration of disciplines in institu-
tional analysis, as evidenced by kinds of methodological crossovers that
are novel in this line of research. More specifically, scholars appear to be
expanding the array of analytical techniques they are using, leveraging
methods linked with various disciplines to augment their assessment of
institutional data. Instances of this include the engagement of compu-
tational approaches for the coding of public policy text (Rice et al.,
2021), the engagement of manually coded policy content in agent-based
modeling exercises (Siddiki & Frantz, 2019), and the usage of network
analysis to reflect interlinkages among various kinds of institutional infor-
mation. These examples essentially highlight one of the draws to the IG,
which is that its structure makes it amenable to address diverse ques-
tions, concepts, and methods. They also then highlight how analysts have
exploited this opportunity, by leveraging techniques that may have been
to date uncommon in IG research, and even variably common in their
own disciplines.
Against this backdrop, the IG attains the function of an integrator to
produce novel insights that can draw on expertise no longer within but
across domains, marrying the theoretical foundations and methodolog-
ical rigor established in empirical domains with the formal approaches
to reasoning and representation of artificial societies found in legal and
computational domains.
To this end, the “Grammar” may no longer just be a “tool,” but
rather an interdisciplinary interface that has to reflect linkages to diverse
analytical objectives, and be open to diverse methodological approaches
at the same time. At the core of such accommodation, however, lies
an unambiguous and consistent characterization of the concept that
all such approaches build upon, a representation of institutions with
broad conceptual validity, whether looking at institutions as legalistic or
socio-normative concepts, approaching the analysis formally or empiri-
cally, basing the analysis on individual statements (or collections thereof),
or selectively favoring a systemic institutional perspective. Offering a
9 CONTEXTUALIZATION AND FUTURE DEVELOPMENT … 365
basis for such integrated IG Concept, this book introduces the Institu-
tional Grammar based on a refined perspective that departs from the
original primarily linguistically motivated view on institutions, to one
that progressively abstracts from linguistic structure and captures institu-
tions by reconstructing institutional structure based on semantic features
(see Chapter 3). At the same time, conceptual integration requires the
recognition of abstractions that expose analytically essential features,
but without being overly prescriptive, and hence constraining. Navi-
gating diverse analytical needs, the IG, in its refined form as IG 2.0,
is built on three premises, namely establishing ontological consistency of
the syntactic form, capturing institutional information comprehensively,
and finally, accommodating novel analytical applications by making the
IG computationally tractable.
To tailor it to use cases, the IG introduces distinctive levels of expres-
siveness that capture features that incrementally expose structural and
semantic detail responding to these premises. Whereas the basic level, IG
Core, introduced in Chapter 4, primarily focuses on retaining compati-
bility to the original IG, it resolves selected ontological inconsistencies
observed in literature (e.g., Schlüter & Theesfeld, 2010) or practice (e.g.,
Frantz & Siddiki, 2021), while establishing a baseline of informational
detail accessible for analyses based on the atomic institutional statement.
At the same time, the revised regulative structure derived from Crawford
and Ostrom’s original IG is complemented with a constitutive structure,
which accommodates the representation of the observed variation in insti-
tutional statement function that is also reflected in structure. More plainly,
the IG 2.0 is explicitly acknowledging that institutional statement form
follows from statement function. Notably, however, recall that in the IG
2.0, the regulative and constitutive statement syntaxes are not treated
as entirely separable, but rather part of an integrated syntax to reflect
interdependence among the functions of regulative and constitutive state-
ments within an institutional setting. Whereas the latter parameterizes, the
former describes opportunities and constraints within those parameters.
IG Extended (introduced in Chapter 5) builds on this feature set by
advising the deep structural parsing of institutional information, providing
a fine-granular representation of institutions focused not only on indi-
vidual statements, but moreover interpreting institutions as systemically
linked patterns of institutional states and statements, alongside concep-
tual information that defines the structure of the institutional setting
more generally. In addition to exposing the setting, IG Extended further
establishes the semantic embedding of statements in this institutional

setting based on taxonomic context characterizations. The comprehensive
capturing of structural detail in IG Extended necessitates methodological
adjustments in the way in which institutional information is interpreted
and represented, which includes the potential reconstruction of language
to complete the shift of representing institutional information in terms of
uniform institutional patterns (as opposed to linguistic patterns) that are
accessible to algorithmic treatment.
With the established structural detail, IG Logico (introduced in
Chapter 6) completes the abstraction of institutional information
expressed in terms of the IG from its underlying linguistic origin by
not only extracting structural information, but complementing it with
an unambiguous semantic specification of institutional statements that
enables a treatment independent from the original structural represen-
tation. Establishing the basis for an algorithmic treatment of institutional
statements, IG Logico introduces a set of general structural transforma-
tions applicable to encoded institutional information.
On the basis of these concepts, IG 2.0 establishes a consistent ontolog-
ical basis in the form of a composable feature set grounded in institutional
theory and philosophy, from which analysts can draw based on their
varying methodological and analytical foci.
Beyond the conceptual perspective, as an analytical paradigm, the IG
2.0 also provides methodological guidance by systematizing the encoding
of institutional information by identifying the central design parameters
relevant for any study, including choice of feature sets of relevance, corre-
sponding preprocessing steps, as well as conventions to enhance reliability
in the encoding process (see Chapter 7). This is complemented by opera-
tional facilities that include a dedicated encoding notation, IG Script, that
is format-agnostic (i.e., is not linked to a particular output format such
as tabular output), both human- and machine-readable, and supports the
variable use of features associated with different levels of expressiveness
(i.e., being permissive about selective feature use). Associated with the
methodological development are calls for a consistent documentation of
encoded information based on the recorded study design parameters and
dataset description (an aspect that is further elaborated in Sect. 9.2.3).
Beyond the encoding, methodological guidance extends to the analysis,
which includes the choice of relevant IG features as well as their linkage
to constructs central for distinctive analyses, a selection of which was
discussed in Chapter 8, without being prescriptive in analytical objec-

tives or choice of specific techniques. With the shared conceptual basis for
any analysis that the IG sponsors, the methodological adaptation (specific
forms and extent of data collection, analytical approaches and associated
techniques) opens the IG for broad disciplinary adoption. To support the
latter, the semantic annotation scheme in IG 2.0, and IG Logico specif-
ically, broadens applicability of the IG by providing the infrastructure
to establish an epistemological grounding of encoded institutional infor-
mation by linking it to theories, frameworks, and concepts drawn from
specific disciplines, or superimposing constructs specific to the analysis at
hand.
Figure 9.1 presents the levels of expressiveness, alongside envisioned
primary foci of coder and analyst, as well as concepts and analytical
applications associated with the specific levels.
Combined, (i) the ontological basis of the IG based on the original
structure and refinements introduced in this book, (ii) the methodological
principles for both the encoding of institutional information and anal-
ysis that provide the parameters for capturing diverse and situationally
tailored applications of the Institutional Grammar, and (iii) the linkage to
knowledge and concepts anchored in the analyst’s domain, position the
Fig. 9.1 Institutional Grammar 2.0 by Levels of Expressiveness and associated

Perspectives, Concepts and Analytical Applications (An enlarged version of this
Institutional Grammar as an analytical paradigm. Generally, an analyt-

ical paradigm is a shared ontology that guides conceptualization, design,
and application of the analysis, while being informed by and compatible
to varying epistemological and disciplinary orientations, and associated
methodological preferences. The IG as introduced here is thus charac-
terized as an analytical paradigm. Fundamentally, while offering a shared
conceptual and representational basis for viewing institutions, the IG also
accommodates diverse analytical techniques and applications rooted in
different disciplines and domains (see Chapter 1), ability to link to diverse
institutional theories and frameworks (including and beyond the IAD),
and types of institutional data (e.g., regulative, constitutive; institutions
in form, institutions in use).
The collation of features and principles is thereby equally looking
back at the history of IG research as well as looking at opportuni-
ties lying ahead. The IG concept introduced here builds on extant IG
research introduced in Chapters 1–2, including theoretical developments
and empirical insights, diverse analytical branches and techniques. The
background discussion further reflects the broader interpretation of insti-
tutions and institutional analysis in diverse disciplines, including ones the
IG has found limited attention in to date. Chapters 3–6 describe the
conceptual foundations of the IG, including its original form as well as the
refinements introduced to meet the aforementioned analytical flexibility,
alongside associated theoretical justifications (e.g., statement reconstruc-
tion based on conceptual reification, institutional states vs. statements,
hybrid institutional statements).
Completing the conceptual Foundations part of the book, the second
Applications part that spans across Chapters 7 and 8 operationalizes the
analytical versatility of the Institutional Grammar by providing method-
ological guidance for study design that is specifically focused on the
collection and encoding of data, before shifting to the illustration of appli-
cation cases and novel analytical opportunities linked to the conceptual
innovations introduced as part of the IG. However, instead of guiding,
or prescribing specific forms of analyses, the latter chapter initially exposes
extensions to statistical applications of the IG, before traversing diverse
techniques such as network analysis, as well as institutional modeling.
Drawing on the extended feature set, this chapter further introduces
substantively novel complexity metrics (State Complexity, Regimentation)
pertaining to the structural assessment of institutions.
It finally turns to algorithmic opportunities associated with institu-

tional information encoded at high level of detail (e.g., transformations,
reasoning on ontologically embedded information), and principles of the
epistemological anchoring of institutional information.
With the integrative perspective of concept, method, and analysis, a
paradigmatic view on the IG signals a coherent perspective that aims at
being inclusive of all extant IG research (and researchers), while providing
a conceptual basis of the Grammar on which novel (and potentially not
yet identified) analytical directions can orient. By doing so, it more imme-
diately positions the Grammar as an interdisciplinary interface that has
the potential to foster novel synergistic directions in institutional analysis
(including and beyond ongoing efforts), with the possibility of estab-
lishing a transdisciplinary perspective on institutional analysis that more
systematically draws on and unites disciplinary views on institutions.
9.2 Future Directions

in Institutional Grammar Research
Concluding the conceptual introduction of the Institutional Grammar as
an analytical paradigm and a conceptual bridge between disciplines, what
remains is to look ahead.
What is presented in this book conveys an evolution of the original
IG as presented by Crawford and Ostrom (1995), not a substitution.
The original IG presented a syntactic structure – a container for institu-
tional language – initially operationalized in the context of game theory.
It, however, took institutional analysts thirteen years to find analytical
use of the original IG as a vehicle to capture institutional meaning,
initially in agent-based models, and more prominently in policy studies.
At first glance, these applications are distinct; distinct enough to raise the
question of how they relate to a common approach. But, this variable
application signaled an early illustration of the interdisciplinary potential
of the Grammar.
The IG 2.0 is poised to exhibit similar opportunity. It refines the
original Grammar to drive its conceptual integrity based on empirical
observations, but more importantly, aims at offering greater validity by
extracting essential institutional features without concessions to the way
in which they are expressed. With this aspiration (moderated by different
levels of expressiveness, or rigidity), the rich set of novel features position
it as an integrator that harmonizes areas of institutional analysis observed
so far, as well as an enabler to invite novel applications from fields not

considered as of yet.
Living up to such promise necessarily requires cognizance about the
current state of affairs and challenges associated with use of the IG, as
well as the need to glance ahead. The consequential next steps pertain
to: (i) addressing concerns related to the IG concept and its applica-
bility; (ii) attending to the disciplinary adoption of the IG more generally
and the contextualization of the IG with institutional theory such as the
IAD more specifically; and (iii) entertaining opportunities leveraged for
institutional analysis as a whole.
9.2.1 Conceptual Directions

Turning to the first, the IG concept as presented in this work integrates
methodological and representational needs of applications we observe
today in a set of conceptual extensions that are integrated across levels of
expressiveness. While empirical application will have to show the concep-
tual value of these refinements (or a subset thereof), we can see early
instances of the transpiring value recognition of select features within
the community. The distinction of Activation Conditions and Execution
Constraints (see Sect. 4.2.1.5), for example, is a basis to afford a mean-
ingful behavioral operationalization of the IG in computational models.
While policy studies have attended to the ontological distinction to a
limited extent, the first studies leverage those for the representation of
statement dependencies (e.g., network studies). Aspects observed on the
data collection side include the conceptual clarity and reliability spon-
sored by select feature innovations, such as the stratified Object concept
(see Sect. 4.2.1.4).
A byproduct of the conceptual refinements is the necessity for method-
ological and operational adjustments. The extraction of complexity of
institutional statements (e.g., statement combinations) leaves the coder
with the same – challenging reliable manual coding. Alleviations are
proposed as part of this book (IG Script), and are supported by ongoing
developments in the form of supplementary software and operational
guidance. Similarly, machine coding efforts currently concentrated on the
original IG require revision to accommodate the structural refinements of
IG 2.0.
Features of the IG 2.0 that require incremental accommodation of
empirical necessity and observation are the provided taxonomies. Devised
with a balance of generality and usability in mind, i.e., introducing

concepts that are useful across a wide range of domains without “over-
fitting” a particular domain, the practical use will drive their further
refinement, let alone novel contributions. An example of a general variant
is the Context Taxonomy. In many instances, however, the introduced
taxonomies invite for refinement to establish sufficient analytical value
for specific studies and domains. In other instances, the specification of
taxonomies explicitly rests with the analyst (e.g., Regulative Functions
Taxonomy).
9.2.2 Disciplinary Directions

Interdisciplinary approaches elicit a variety of considerations, some of
which are described below. An initial consideration is the value of the
IG for field research. Scholars (Watkins & Westphal, 2016) have outlined
the challenges of encoding interviews in terms of institutional statements,
pointing to methodological concerns, as well as highlighting features
specifically related to institutions-in-use (e.g., natural consequences as
sanctions). The refined IG integrates these aspects in the refined charac-
terization of institution types that – structurally speaking – puts norms and
rules on “eye level,” inviting for richer studies in real-world settings. This
includes both traditional physical field studies, as well as online studies
drawing on the internet as data source for social studies.
Given the overt motivation for the IG 2.0 to be accessible to the
study of institutions of any kind, the exemplary focus on policy in this
book warrants some qualification. The examples and themes discussed
throughout this book largely draw on policy text, i.e., institutions-in-
form. While this exemplification and thematic focus are intentional, their
inclusion is not meant to suggest that the IG is not applicable to different
forms of institutions. In fact, this book draws on policy characteriza-
tions specifically, since (a) this approach recognizes language as a central
means to communicate institutions, and (b) it does so in complex and
highly structured form. Developing a robust structural representation
based on complex (legal) language, it is presumed that other institutions
can usefully be captured in the same general structure, albeit with supple-
mentary accommodations. Challenges related to the encoding of spoken
language in particular, and outside the scope of this discussion, naturally
urge specific methodological accommodations.
A second important aspect related to the value proposition of the IG

as applicable across disciplines is the linkage to institutional theory. The
original IG is embedded in the Institutional Analysis and Design (IAD)
Framework. The principled extension of the original IG by IG 2.0 as
a structural representation of institutional information, and the institu-
tional statement as a basic unit of analysis, maintains this compatibility,
and particularly on the IG Core level. Without further contextualization
at this stage, promising empirical questions revolve around the linkage
between different forms of institutional statements (i.e., regulative and
constitutive) and rule types that position individual statements in the
action situation (Ostrom, 2005). Important to note here is the differ-
entiated treatment of the concept of rules between the IAD framework
generally and the IG specifically. Typically within discussions of the IAD
framework, the term is used interchangeably with the term institutions,
and thus for example, the rule typology (Ostrom & Crawford, 2005)
is leveraged for classification of institutions/institutional configurations
conveying particular kinds of information. As explained in Sect. 4.2.3,
rules in the IG have a more specific meaning. Nevertheless, institu-
tional analysts can map rule types (e.g., information, choice, aggregation,
etc.) to institutional statements of different types (i.e., strategies, norms,
rules). There may also be value in ascertaining if and how semantic
annotations of parts of institutional statements as introduced in the IG
2.0 aid in mapping them to different types of rules. This could be
semantic annotations drawing on the taxonomic or ontological classi-
fications (e.g., Context Taxonomy, Function Taxonomy) or semantic
annotations drawing on theory that accords with the analyst’s research
aims.
Yet still, the institutional analyst can assess to what extent reliable
representation of vertical nesting allows for the rigorous evaluation of
monitoring and enforcement mechanisms in IAD studies; when moni-
toring and enforcement activates, who carries out each, complementary
activities, among other aspects of monitoring and enforcement. This
may be of particular interest to institutional analysts which draw on the
IAD framework affiliated Common Pool Resource theory and associated
design principles (Ostrom, 1990). On the assessment of design principles
specifically, application of the IG to study these can lend measurement
clarity insofar as IAD scholars typically operationalize design principles
through evaluation of rule configurations while not always clarifying the
unit of observation upon which rule configurations are measured. The IG

can aid in the measurement of configurations as explained in the book,
and thereby even aid in construct development. Additional aspects that
invite for exploration are the conceptual linkages between elements of the
IAD and the systemic perspective of capturing institutional information
introduced as part of IG Extended.
While referencing the linkage between the IG and the IAD framework,
it is also useful to differentiate levels of action in the IAD framework
(Kiser & Ostrom, 1982; Ostrom, 2005) and levels of expressiveness under
the IG 2.0. Under the IAD framework, the levels of action framing is
used as another way of delineating types of rules, essentially by their
functional properties. Whereas the IAD’s rule typology offers a basis
for classifying the different types of rules of action situations, the levels
of action typology differentiates between rules of operational, collective
choice, and (meta-)constitutive sort. Those forms are inherently inde-
pendent. Level of actions distinguish hierarchically embedded forums of
distinctive purpose (e.g., decision-making vs. defining rules for decision-
making), within which the IAD rule type characterizations (e.g., choice,
position, boundary) apply. Institutional statements of different types –
strategies, norms, rules – can be characterized as different types of rules
that situate at different levels of action.
Contextualizing this discussion of the IG-IAD linkage, this section
merely highlights existing and possible linkages among concepts of the
IG and concepts and taxa of the IAD. However, the discussed features
are inherently opportunities for the development of the IG in relation-
ship to the IAD. Broader reconciliation of IG with IAD concepts is a
separate matter, and issues related to this are best addressed by scholars
seeking to apply and develop the IAD framework.
Instead of seeking reintegration into any specific theoretical framework,
the IG, as presented here, shows compatibility, while at the same time
displaying independence. In fact, it seeks interaction with existing and
novel theoretical frameworks, making the IG a potential catalyst not only
for the linkage to diverse theory, but also a sponsor of theoretical integra-
tion across disciplines, as far as they are able to respond to the granularity
of institutional information that the IG exposes (i.e., are able to intro-
duce or link theoretical constructs to taxonomies, let alone statement or
component characterizations).
9.2.3 Catalyzing Theory and Framework Development

The flexibility of the IG opens the opportunity to look ahead and foster
novel research opportunities that move beyond realizing the usage across
disciplines by happenstance. Rather, interdisciplinary linkages in IG studies
can now occur by design, with an aim toward explicitly identifying concep-
tual and methodological compatibility (or lack thereof) in concepts,
theories, and methods linked to different disciplines. This is critical for
contextualizing and retaining understanding of scientific findings within
and across fields of study. Essentially, what is being suggested here is a call
to more explicitly reference how diverse concepts, theories, and methods
are being flexibly integrated within the context of IG analyses to provide
institutional analysts with a full appreciation of the interdisciplinary and
discipline-specific contributions being offered. This point relates more
broadly to considerations endemic to integrative science generally; that
is, the need to clarify how integrative science advances knowledge on the
integrated as well on what is being integrated.
The perspective shift to understand the IG as an analytical paradigm
is not only an observation of empirical trends, but it urges for method-
ological rigor in application. The reasons for this are multiple: as with any
research field, the trust endowed into the findings of any study rely on
the trust in method, and essentially rigor, associated with data collection
and analysis. One challenge, or variably opportunity, in this context is to
drive community practices around the documentation of the data collec-
tion and coding process. As evident from the methodological discussion
around data collection presented in Chapter 7, any study design requires
a rich set of considerations, some of which apply across different projects,
and others that may diverge across discipline, teams, etc. While present in
previous IG research, the extended feature set of IG 2.0 makes it all the
more pertinent to make such choices explicit to establish a reproducible
research process,1 and to make studies in the field as a whole more robust.
Beyond the focus on methodological rigor as disciplinary credential,
operationally, robust data collection and processing facilitate the use of
data across domains and techniques. This can apply on data structure
level (e.g., different encoding), relate to semantic variation (e.g., varying
1 The IG 2.0 Codebook (Frantz & Siddiki, 2020) includes a guide that highlights
essential considerations in the study design process, alongside further conceptual resources
and instructive resources under https://newinstitutionalgrammar.org.
interpretation of components), let alone data processing (e.g., exclusion

criteria for statements, selective use of IG features). The IG speaks to
those concerns by affording the corresponding descriptive facilities (see
Sect. 6.2 and Frantz & Siddiki, 2020). However, the value does not
lie in the concept, but in practice. Establishing reliable specifications of
the form in which data are represented, alongside data descriptions that
capture the features of the data, will enable synergy effects in broadening
the opportunity to perform large-scale comparative studies on the one
hand, but also provide the basis to make institutional information inter-
operable across techniques, or at least reliably signal whether they are
not.
Taken together, the different directions highlighted above need to
work in concert to ensure the conceptual integrity based on the estab-
lished ontological consistency, broad epistemological applicability, and
scientific credibility based on methodological rigor, constituting the value
of the IG as a reliable yet flexible means to analyze institutions. And, with
the extent to which the Institutional Grammar has evolved over the past
decades, and in its revised form reflecting a New Institutional Grammar,
there is the opportunity to transform research in the area of institutional
analysis for decades to come.
This is supported by the increasingly diverse set of disciplines and
researchers attracted to the IG, all of which are united in one aspect, their
interest for the analysis of institutions. Accommodating these disciplines’
and communities’ varied needs and objectives moves the IG closer to what
it aspires to be: a lingua franca of institutions. The Institutional Grammar
exposes institutions as what they are: sets of institutional statements that
systemically interlink and embed to govern social systems at any level of
organization, or put more generally, the lowest common denominator of
society.
References
2975
Frantz, C. K., & Siddiki, S. (2021). Institutional Grammar 2.0: A specification
for encoding and analyzing institutional design. Public Administration, 99,

Kiser, L. L., & Ostrom, E. (1982). The three worlds of action: A meta-theoretical
synthesis of institutional approaches. In E. Ostrom (Ed.), Strategies of political
inquiry (pp. 179–222).
Ostrom, E., & Crawford, S. E. (2005). Classifying rules. In E. Ostrom (Ed.),
Understanding institutional diversity (pp. 186–216). Princeton University
Press.
Press.
of policy texts with the Institutional Grammar. Public Administration, 99,
Schlüter, A., & Theesfeld, I. (2010). The Grammar of institutions: The challenge
22(4), 445–475. https://doi.org/10.1177/1043463110377299
Siddiki, S., & Frantz, C. (2019). Understanding the effects of social value orien-
tations in shaping regulatory outcomes through agent based modeling: An
application to organic farming. Presented at the Workshop on the Workshop
(WOW) 6 Conference.
1111/psj.12139
Appendix A: Institutional Statement
Structure
The following Fig. A.1 offers a semi-formal characterization of insti-

tutional statements in the Institutional Grammar 2.0 in the form of
an UML class diagram which highlights the possible representational
forms and kinds of institutional statements. Institutional statement forms
include atomic statements and combinations thereof (i.e., different types
of nesting), whereas atomic statements can be of regulative or constitutive
kind (and where combined, hybrid). The structure supports combinations
of statements of either kind (horizontal nesting), features consequen-
tial linkages of statements of either form and kind (vertical nesting),
as well as supporting component-level nesting on selected components
and properties. Note that statements (i.e., atomic and combinations) can
be negated, and be logically linked based on logical operators. A high-
resolution version of the statement structure is available via the book
website (https://newinstitutionalgrammar.org).
© The Editor(s) (if applicable) and The Author(s), under exclusive 377
license to Springer Nature Switzerland AG 2022
https://doi.org/10.1007/978-3-030-86372-2
378
APPENDIX A: INSTITUTIONAL STATEMENT STRUCTURE
Fig. A.1 Institutional Statement Structure in the Institutional Grammar 2.0 (An enlarged version of this figure can be
found under https://newinstitutionalgrammar.org/book-figures)
Appendix B: National Organic Program
Regulation
The following information is an excerpt of the United States Depart-

ment of Agriculture National Organic Program Regulation,1 specifically
pertaining to compliance. This excerpt forms the basis of the coded data
used in the analysis in Chapter 8.
1 Source: https://www.ams.usda.gov/rules-regulations/organic.
https://doi.org/10.1007/978-3-030-86372-2
U.S. National Organic Program
Regulation
Compliance
§ 205.660 General
(a) The National Organic Program’s Program Manager, on behalf of
the Secretary, may inspect and review certified production and
handling operations and accredited certifying agents for compliance
with the Act or regulations in this part.
(b) The Program Manager may initiate suspension or revocation
proceedings against a certified operation:
(1) When the Program Manager has reason to believe that a certi-
fied operation has violated or is not in compliance with the Act
or regulations in this part; or
(2) When a certifying agent or a State organic program’s governing
State official fails to take appropriate action to enforce the Act
or regulations in this part.
(c) The Program Manager may initiate suspension or revocation of a
certifying agent’s accreditation if the certifying agent fails to meet,
conduct, or maintain accreditation requirements pursuant to the
Act or this part.
(d) Each notification of noncompliance, rejection of mediation,
noncompliance resolution, proposed suspension or revocation, and
suspension or revocation issued pursuant to §205.662, §205.663,
https://doi.org/10.1007/978-3-030-86372-2
382 U.S. NATIONAL ORGANIC PROGRAM REGULATION
and §205.665 and each response to such notification must be sent

to the recipient’s place of business via a delivery service which
provides dated return receipts.
§ 205.661 Investigation of Certified Operations

(a) A certifying agent may investigate complaints of noncompliance
with the Act or regulations of this part concerning production
and handling operations certified as organic by the certifying
agent. A certifying agent must notify the Program Manager of all
compliance proceedings and actions taken pursuant to this part.
(b) A State organic program’s governing State official may investi-
gate complaints of noncompliance with the Act or regulations in
this part concerning organic production or handling operations
operating in the State.
§ 205.662 Noncompliance Procedure for Certified Operations

(a) Notification. When an inspection, review, or investigation of
a certified operation by a certifying agent or a State organic
program’s governing State official reveals any noncompliance with
the Act or regulations in this part, a written notification of
noncompliance shall be sent to the certified operation. Such noti-
fication shall provide:
(1) A description of each noncompliance;
(2) The facts upon which the notification of noncompliance is
based; and
(3) The date by which the certified operation must rebut or correct
each noncompliance and submit supporting documentation of
each such correction when correction is possible.
(b) Resolution. When a certified operation demonstrates that each
noncompliance has been resolved, the certifying agent or the State
organic program’s governing State official, as applicable, shall send
the certified operation a written notification of noncompliance
resolution.
(c) Proposed suspension or revocation. When rebuttal is unsuccessful
or correction of the noncompliance is not completed within
the prescribed time period, the certifying agent or State organic
U.S. NATIONAL ORGANIC PROGRAM REGULATION 383
program’s governing State official shall send the certified opera-

tion a written notification of proposed suspension or revocation
of certification of the entire operation or a portion of the oper-
ation, as applicable to the noncompliance. When correction of a
noncompliance is not possible, the notification of noncompliance
and the proposed suspension or revocation of certification may be
combined in one notification.
The notification of proposed suspension or revocation of certi-
fication shall state:
(1) The reasons for the proposed suspension or revocation;
(2) The proposed effective date of such suspension or revocation;
(3) The impact of a suspension or revocation on future eligibility
for certification; and
(4) The right to request mediation pursuant to §205.663 or to file
an appeal pursuant to §205.681
(d) Willful violations. Notwithstanding paragraph (a) of this section,
if a certifying agent or State organic program’s governing State
official has reason to believe that a certified operation has willfully
violated the Act or regulations in this part, the certifying agent
or State organic program’s governing State official shall send the
certified operation a notification of proposed suspension or revo-
cation of certification of the entire operation or a portion of the
operation, as applicable to the noncompliance.
(e) Suspension or revocation. (1) If the certified operation fails to
correct the noncompliance, to resolve the issue through rebuttal
or mediation, or to file an appeal of the proposed suspension or
revocation of certification, the certifying agent or State organic
program’s governing State official shall send the certified operation
a written notification of suspension or revocation.
(2) A certifying agent or State organic program’s governing
State official must not send a notification of suspension or revoca-
tion to a certified operation that has requested mediation pursuant
to §205.663 or filed an appeal pursuant to §205.681, while final
resolution of either is pending.
(f) Eligibility. (1) A certified operation whose certification has been
suspended under this section may at any time, unless otherwise
stated in the notification of suspension, submit a request to the
Secretary for reinstatement of its certification. The request must
be accompanied by evidence demonstrating correction of each

noncompliance and corrective actions taken to comply with and
remain in compliance with the Act and the regulations in this part.
(2) A certified operation or a person responsibly connected with
an operation whose certification has been revoked will be ineligible
to receive certification for a period of 5 years following the date of
such revocation, except, that, the Secretary may, when in the best
interest of the certification program, reduce or eliminate the period
of ineligibility.
(g) Violations of Act. In addition to suspension or revocation, any
certified operation that:
(1) Knowingly sells or labels a product as organic, except in accor-
dance with the Act, shall be subject to a civil penalty of not
more than the amount specified in §3.91(b)(1)(xxxvii) of this
title per violation.
(2) Makes a false statement under the Act to the Secretary, a State
organic program’s governing State official, or a certifying agent
shall be subject to the provisions of Section. 1001 of title 18,
United States Code.
[65 FR 80637, Dec. 21, 2000, as amended by 75 FR 17560,
Apr. 7, 2010]
§ 205.663 Mediation
Any dispute with respect to denial of certification or proposed suspen-
sion or revocation of certification under this part may be mediated at
the request of the applicant for certification or certified operation and
with acceptance by the certifying agent. Mediation shall be requested in
writing to the applicable certifying agent. If the certifying agent rejects the
request for mediation, the certifying agent shall provide written notifica-
tion to the applicant for certification or certified operation. The written
notification shall advise the applicant for certification or certified opera-
tion of the right to request an appeal, pursuant to §205.681, within 30
days of the date of the written notification of rejection of the request for
mediation. If mediation is accepted by the certifying agent, such medi-
ation shall be conducted by a qualified mediator mutually agreed upon
by the parties to the mediation. If a State organic program is in effect,
the mediation procedures established in the State organic program, as
approved by the Secretary, will be followed. The parties to the mediation

shall have no more than 30 days to reach an agreement following a medi-
ation session. If mediation is unsuccessful, the applicant for certification
or certified operation shall have 30 days from termination of mediation to
appeal the certifying agent’s decision pursuant to §205.681. Any agree-
ment reached during or as a result of the mediation process shall be in
compliance with the Act and the regulations in this part. The Secretary
may review any mediated agreement for conformity to the Act and the
regulations in this part and may reject any agreement or provision not in
conformance with the Act or the regulations in this part.
§ 205.664 [Reserved]
§ 205.665 Noncompliance Procedure for Certifying Agents
(a) Notification. When an inspection, review, or investigation of an
accredited certifying agent by the Program Manager reveals any
noncompliance with the Act or regulations in this part, a written
notification of noncompliance shall be sent to the certifying agent.
Such notification shall provide:
(1) A description of each noncompliance;
(2) The facts upon which the notification of noncompliance is
based; and
(3) The date by which the certifying agent must rebut or correct
each noncompliance and submit supporting documentation of
each correction when correction is possible.
(b) Resolution. When the certifying agent demonstrates that each
noncompliance has been resolved, the Program Manager shall
send the certifying agent a written notification of noncompliance
resolution.
(c) Proposed suspension or revocation. When rebuttal is unsuccessful
or correction of the noncompliance is not completed within the
prescribed time period, the Program Manager shall send a written
notification of proposed suspension or revocation of accreditation
to the certifying agent. The notification of proposed suspension or
revocation shall state whether the certifying agent’s accreditation
or specified areas of accreditation are to be suspended or revoked.
When correction of a noncompliance is not possible, the notifica-
tion of noncompliance and the proposed suspension or revocation
may be combined in one notification. The notification of proposed

suspension or revocation of accreditation shall state:
(1) The reasons for the proposed suspension or revocation;
(2) The proposed effective date of the suspension or revocation;
(3) The impact of a suspension or revocation on future eligibility
for accreditation; and
(4) The right to file an appeal pursuant to §205.681.
(d) Willful violations. Notwithstanding paragraph (a) of this section,
if the Program Manager has reason to believe that a certifying
agent has willfully violated the Act or regulations in this part, the
Program Manager shall send a written notification of proposed
suspension or revocation of accreditation to the certifying agent.
(e) Suspension or revocation. When the accredited certifying agent
fails to file an appeal of the proposed suspension or revocation of
accreditation, the Program Manager shall send a written notice of
suspension or revocation of accreditation to the certifying agent.
(f) Cessation of certification activities. A certifying agent whose
accreditation is suspended or revoked must:
(1) Cease all certification activities in each area of accreditation
and in each State for which its accreditation is suspended or
revoked.
(2) Transfer to the Secretary and make available to any applicable
State organic program’s governing State official all records
concerning its certification activities that were suspended or
revoked.
(g) Eligibility. (1) A certifying agent whose accreditation is suspended
by the Secretary under this section may at any time, unless other-
wise stated in the notification of suspension, submit a request to
the Secretary for reinstatement of its accreditation. The request
must be accompanied by evidence demonstrating correction of
each noncompliance and corrective actions taken to comply with
and remain in compliance with the Act and the regulations in this
part. (2) A certifying agent whose accreditation is revoked by the
Secretary shall be ineligible to be accredited as a certifying agent
under the Act and the regulations in this part for a period of not
less than 3 years following the date of such revocation.
§§ 205.666–205.667 [Reserved]
§ 205.668 Noncompliance Procedures Under State Organic
Programs
(a) A State organic program’s governing State official must promptly
notify the Secretary of commencement of any noncompliance
proceeding against a certified operation and forward to the Secre-
tary a copy of each notice issued.
(b) A noncompliance proceeding, brought by a State organic
program’s governing State official against a certified operation,
shall be appealable pursuant to the appeal procedures of the State
organic program. There shall be no subsequent rights of appeal to
the Secretary. Final decisions of a State may be appealed to the
United States District Court for the district in which such certified
operation is located.
(c) A State organic program’s governing State official may review and
investigate complaints of noncompliance with the Act or regula-
tions concerning accreditation of certifying agents operating in the
State. When such review or investigation reveals any noncompli-
ance, the State organic program’s governing State official shall send
a written report of noncompliance to the Program Manager. The
report shall provide a description of each noncompliance and the
facts upon which the noncompliance is based.
Glossary
Action Situation An Action Situation is defined as a setting in which

two or more actors “are faced with a set of potential actions that jointly
produce outcomes” (Ostrom, 2005, p. 32).
Activation Condition An Activation Condition is the condition under
which the non-Context part of an institutional statement applies. Insti-
tutional statements can contain one or more logically linked activation
conditions.
Analytical Paradigm An Analytical Paradigm is a shared ontological
basis that guides conceptualization, design, and application of analysis,
where the latter is also informed by epistemological and disciplinary
orientations, as well as methodological preferences or necessities.
Atomic Institutional Statement An Atomic Institutional Statement is
an institutional statement that only contains one of each necessary and
(where applicable) optional components, where none of these compo-
nents is further decomposed into, or substituted by, nested institutional
statements.
Coding Coding, or Encoding, as referenced in this book, is the parsing
of institutional data corresponding to different syntactic components,
relying on interpretation and meaning of institutional data (including
explication or inference).
https://doi.org/10.1007/978-3-030-86372-2
390 GLOSSARY
Components Components are syntactic elements of an institutional state-

ment that capture abstract and generalizable units of language. Indi-
vidual types of components convey distinctive kinds of institutionally
relevant meaning.
Composite Institutional Statements A Composite Institutional State-
ment is an institutional statement that features any nesting facility,
including horizontal, vertical and component-level nesting.
Conceptual Reification Conceptual Reification is the expression of
actions, mental constructs (e.g., beliefs) and other nonphysical concepts
in nominalized, or “thingified” form.
Constitutive Statement Constitutive Statements parameterize features
of an institutional system within particular contexts. Constitutive insti-
tutional statements may also describe possibility and consequences
associated with the referenced constitution.
Deep Structure Deep Structure reflects the lowest-level composition of
units of language (e.g., components) that capture institutional content,
i.e., structural features of components, rather than structural features of
statements only.
Execution Constraints Execution Constraints reference the qualification
of activities during execution, thus imposing constraints on the enacted
Aim or Constitutive Function.
Formal Institutions Formal Institutions are institutions that result from
institutional decision-making processes engaged by recognized authori-
ties appointed in a legitimized forum (e.g., a public policy that results
from a policy making process engaged by an appointed legislature).
Hybrid Institutional Statement Hybrid Institutional Statements are
statements that are variably comprised of configurations of regulative
and constitutive institutional statements.
IAD Framework The Institutional Analysis and Design Framework is an
analytical approach for studying institutional design, development, and
outcomes.
Informal Institutions Informal Institutions are those that emerge from
social processes and are represented, for example, in social norms and
conventions, and rely on internal and decentralized social enforcement.
Institution Institutions are rules, norms, and strategies that govern
social systems.
Institutional Analysis Institutional Analysis refers to the study of insti-
tutions that govern social systems.
GLOSSARY 391
Institutional Configuration An Institutional Configuration is a set of

institutional statements that are directly, or indirectly linked (e.g., based
on implied reference) to the institutional setting, and can be analyzed at
different scales (e.g., component, statement, set of statements).
Institutional Data Institutional Data are those captured in institutions
as they normally exist in their respective domains.
Institutional Features Institutional Features are institutional concepts,
their relationships and interaction within the institutional setting that
may be identified on component (e.g., actors, actions referenced in
components), statement (e.g., function/effects of statements), or system
level (e.g., function of interlinked statements).
Institutional Grammar The Institutional Grammar is a theoretically
grounded analytical approach for the characterization of structure
and meaning of institutions of diverse forms and kinds, including
institutions-in-form, in-use, and covering strategies (or conventions),
norms, and rules as different institution types. The Institutional
Grammar is distinct from a linguistic grammar in that the units of
language captured in components associate with institutional meaning
and functions, which are structurally combined to potentially inter-
linked institutional statements. Analysis can occur on variable levels of
aggregation.
Institutional Information Institutional Information is institutional data
which have been processed and coded in accordance with the Institu-
tional Grammar.
Institutional Modeling Institutional Modeling captures approaches that
emphasize the study of specific or systemic structure of institutions, with
primary focus on the description and/or re/construction of institutional
settings and arrangements. Associated analyses can be static (i.e., focus
on structure) or dynamic (i.e., focus on behavior and/or change in
structure) in kind.
Institutional Semantics Semantics, as pertaining to the IG, refers to
meaning conveyed in institutional statements that reflects understanding
of the institutional setting in which statements are observed.
Institutional Setting An Institutional Setting is an institutionally
governed domain.
Institutional State An Institutional State refers to a state of affairs
in the form of behavior (e.g., observed behavior) or environmental
392 GLOSSARY
circumstances. These may condition the applicability of an institutional

statement or otherwise be referenced as part of such.
Institutional Statement Institutional Statements describe actions for
actors within particular contexts, or parameterize features of an insti-
tutional system within particular contexts. Institutional statements are
comprised of components and take the form of atomic institutional
statements (i.e., statements that do not display any form of nesting) or
composite institutional statements (i.e., statements displaying any form of
nesting).
Institutional Structure Institutional Structure refers to arrangements
of units of language organized in components that individually and
configurally convey institutional meaning (Semantics) and which can be
analyzed at different levels.
Institutional Syntax Syntax, as pertaining to the IG, refers to the rules
that govern the configuration of syntactic components that are the
structural units of which institutional statements are comprised.
Institutions-in-Form Institutions-in-form are institutions as embodied
in form (e.g., as written text), independent of the specific process by
which they came about.
Institutions-in-Use Institutions-in-use are institutions as embodied in
practice, irrespective of their formal or informal origin.
Intercoder Reliability Intercoder Reliability (ICR) is a numerical
measure of the agreement between different coders regarding how the
same data should be coded (Connor & Joffe, 2020).
Levels of Expressiveness Levels of Expressiveness in the Institutional
Grammar 2.0 reflect distinctive levels of conceptual richness and focus
that respond to the levels of representational detail and complexity
linked to specific analytical foci. Levels of expressiveness are backward-
compatible, i.e., information encoded at higher levels of expressiveness
is accessible to analysis at lower levels.
Nested Institutional Statement Nested Institutional Statements are
institutional statements that are logically combined via conjunction,
in/exclusive disjunction (horizontal nesting), in a consequential rela-
tionship, with one statement (or combination thereof) being the
monitored, and the other one (or combination thereof) the consequen-
tial statement (vertical nesting), or are structurally embedded within
components of higher-level institutional statements (component-level
nesting).
GLOSSARY 393
Ontological Consistency Ontological Consistency is defined as the

clear, logically consistent and coherent characterization of syntactic
components, where conceptual characterizations, their relationships and
(inter-)dependencies are explicit and unambiguous so as to avoid vari-
able interpretation of encoded information.
Ontologies Ontologies are explicit specifications of relationships of
different qualities (e.g., hierarchical) between classes of concepts
exhibiting different qualities, associated rules, and underlying axioms on
which formal reasoning can operate.
Polymorphic Institutional Statement Polymorphic Institutional State-
ments are statements that can be represented as both constitutive and
regulative statements (subject to analytical objectives), i.e., characterize
along the components associated with the corresponding syntactic form.
Regulative Statement Regulative Statements describe actions for actors
within particular contexts. Regulative institutional statements may also
indicate prescription and consequences related to the referenced action.
Index
A C
Action situation, 18 Codebook, 38
Activation Conditions (Component), Coding ‘Institutions in Use’, 41
57, 87, 125 Coding process, 153
Activation Conditions Cognitive Grammar, 65
(Component)–constitutive Cognitive linguistics, 157
discussion, 125 Combination-level component
Activation Conditions transformation, 225
(Component)–regulative Common Pool Resource Theory, 36
discussion, 87 Complexity analysis, 292
Agent-based Institutional Modeling, Complexity metrics, 285, 294
315, 335 Component-level combinations, 94,
Agent-based Modeling, 45 253
Agent-based Modeling and Component-level combination
Simulation, 40 transformation, 226
AIC pattern, 146 Component-level nesting, 294
Aim (Component), 21, 85 Composite Institutional Statement,
Akoma Ntoso, 223 101
Animacy taxonomy, 202 Composition, 76
Artificial society, 40 Computer science, 15
Aspirational context, 168 Concept measurement, 285
Assessment (Coding phase), 273 Conceptual organization analysis, 309
Atomic institutional statement, 101 Conceptual reification, 68, 158
Attributes (Component), 21, 82 Conditions (Component), 21
https://doi.org/10.1007/978-3-030-86372-2
396 INDEX
Conditions-Consequence Epistemological linkage, 344, 375

Transformation, 226 Equilibrium, 104
Consequence annotations, 214 Execution (Coding phase), 251
Consequential statements, 22, 213 Execution Constraints (Component),
Constituted Entity (Component), 119 57, 87, 125, 291
Constituting Properties (Component), Execution Constraints
123 (Component)–constitutive
Constitutive Function (Component), discussion, 125
122 Execution Constraints
Constitutive functions (Institutional (Component)–regulative
functions), 210 discussion, 87
Constitutive Functions Taxonomy, Exogenous institutions, 17
211, 212
Constitutive rules, 131
Constitutive statements, 48, 79 F
Constitutive strategies, 133 Formal institutions, 6
Constitutive syntax, 116 Formal reasoning, 353
Construction Grammar, 65, 158 Formal rules, 103
Context (Component), 87, 125 Functional Requirements for
Context (Component)–constitutive Bibliographic Records (FRBR),
discussion, 125 223
Context (Component)–regulative
discussion, 87
Context taxonomy, 166, 169, 201 G
Crawford, S.E.S. and Ostrom, E., 21 Game theory, 102
Generative Grammar, 65
Generative Institutional Analysis, 335
D
Data collection, 244
Data processing (Coding), 250 H
Deep structural analysis, 151 Heuristics for feature selection, 248
Degrees of State Variability (Logical Historical institutionalism, 8
operators), 295 Horizontal nesting, 94, 150, 197, 294
Delta parameter, 114 Hybrid institutional statements, 133,
Deontic (Component), 21, 83 172, 178
Direct Object (Component), 86
Dynamic institutional analysis, 115
I
IG Core features, 134
E IG Extended features, 189
Economics, 10 IG feature customization, 247
EFC pattern, 146 IG Logico features, 193
Endogenous institutions, 17 IG Script, 256
INDEX 397
Indirect Object (Component), 86 Institutional Tree Metrics (Complexity

Informal institutions, 7 metrics), 300
Institutional analysis, 6 Institutions, 1, 375
Institutional Analysis and Institutions as equilibria, 101
Development Framework (IAD), Institutions as norms, 102
18 Institutions as rules, 103
Institutional complexity metrics Institutions-in-form, 76
(Complexity metrics), 294 Institutions-in-use, 76
Institutional configuration, 75 Institution types, 101, 133
Institutional design, 36 Institution types (constitutive), 130
Institutional development, 36 Institution types (regulative), 101,
Institutional features, 54 114
Institutional functions, 206 Intercoder reliability, 274
Institutional functions analysis, 344 Interpretational scope (wide and
Institutional Grammar (IG), 2, 19, narrow), 184
21, 53, 69, 75
Institutional Grammar 2.0, 53, 75
Institutional meaning, 61, 151 L
Law, 13
Institutional modeling, 314
Legal informatics, 13, 16
Institutional nesting principles, 100
Levels of expressiveness, 76
Institutional patterns, 151
Linguistic grammar, 65, 70
Institutional semantics, 24
Logical operators, 95, 295
Institutional setting, 79
Institutional state, 143
Institutional State Complexity M
(Complexity metric), 297, 298 Machine learning, 16
Institutional statement, 20, 79, 100 Metatype taxonomy, 203
Institutional statement coding, 260 Methodology (Agent-based
Institutional statement coding modeling), 318
(Hybrid institutional statements), Methodology (Coding), 243
267 Modal (Component), 120
Institutional statement coding (IG Monitored statements, 22, 213
Core), 261 Monitoring statements, 213
Institutional statement coding (IG
Extended), 270
Institutional statement coding (IG N
Logico), 272 Natural language, 69
Institutional statement trees, 153 Nested ADICO (nADICO), 22
Institutional State Regimentation Nested institutional statements, 22,
(Complexity metric), 299, 301 82, 90, 101, 147, 218, 229, 234
Institutional state vs. statement, 143 Nesting levels, 294
Institutional structure, 24 Network analysis, 287
398 INDEX
New Institutional Economics, 10 Reference dataset, 279

New Institutional Grammar, 48, 53, Regulative functions (Institutional
233, 375 functions), 206
Normative Multi-Agent Systems, 15 Regulative statements, 79
Norms, 63, 104, 107 Regulative statement structure, 94
Norm vs. rule distinction, 107 Regulative syntax, 82
Relational ambiguity, 58
O Research questions, 33
Object (Component), 21, 39, 45, 86 Role taxonomy, 204
Objectives (Institutional Grammar Rules, 104, 108, 133
2.0), 76 Rule typology (IAD), 18
Object-Property Hierarchy, 160, 162
Ontological consistency, 54
Ontological inconsistency, 46
Ontology, 217, 353 S
Or else (Component), 21, 89, 126 Scoping institutional data, 244
Or else (Component)–constitutive Semantic annotations, 201
discussion, 126 Semantics, 24
Or else (Component)–regulative Semantic specification (Institutional
discussion, 89 Grammar 2.0), 194
Semantic systemic analysis, 348
Shared strategies, 102
P
Perspective extrapolation, 337 Situational context, 168
Philosophy, 14 Social-Ecological Systems Framework,
Planning (Coding phase), 244 37
Policy complexity metrics (Complexity Social network analysis, 45
metrics), 285 Social psychology, 12
Policy Implementation Framework, 37 Sociological institutionalism, 101
Political science, 10 Sociology, 11
Polymorphic institutional statements, Spreadsheet software, 245
183 Statement references, 221
Pre-processing institutional data, 252 Statement transformation rules, 224,
Procedural context, 167 342
Properties-Conditions Transformation, Strategies, 105, 133
231 Structural analysis, 291
Property types, 164 Structural consistency, 69
Public policy and administration, 8 Structural incongruence, 59
Substantive ambiguity, 58
R Substantive context, 167
Rational choice institutionalism, 8, Syntax, 23
101 Systemic analysis, 303
INDEX 399
T U
Unidirectional referencing
Text annotation software, 245 (Institutional statements), 62
Unit of analysis, 101
Theory integration, 345, 374
Tool support, 245 V

Vertical nesting, 91, 151, 198, 213
Transformation rules, 224, 342 Vertical nesting annotations, 213

Christopher K. Frantz, Saba Siddiki - Institutional Grammar

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Christopher K. Frantz, Saba Siddiki - Institutional Grammar

Uploaded by

Copyright:

Available Formats

Institutional

Foundations and Applications

Christopher K. Frantz · Saba Siddiki

ISBN 978-3-030-86371-5 ISBN 978-3-030-86372-2 (eBook)

Cover credit: Alex Linch/shutterstock.com

The Institutional Grammar (IG) is a well-established approach for insti-

of application field and analytical method. Working toward this effort,

diverse perspectives on institutions as the fabric of social systems, inde-

Gjøvik, Norway Christopher K. Frantz

3 Motivation for a New Institutional Grammar 53

6 Institutional Grammar 2.0: Semantic Features

8.2.2 Institutional Modeling 314

Appendix A: Institutional Statement Structure 377

Fig. 3.1 Combined Institutional Grammar patterns example 57

Fig. 6.2 First-order Linkages between Taxonomies and IG

Table 3.1 Ontological inconsistencies in the Institutional Grammar 60

This book offers guidance on the analysis of institutions. Institutions

© The Author(s), under exclusive license to Springer Nature 1

variably associated with scholarship in these various disciplines that moti-

statements that regulate behavior or constitute features of the institutional

suited to capturing the structure and semantics of the kinds of regula-

1.1 What Is Institutional Analysis?

(e.g., a public policy that results from a policy-making process engaged

1.1.1 Institutional Analysis by Discipline

perspectives predominate in institutional studies: historical institution-

1.1.1.1 Public Policy and Administration

perspectives. Within public administration research, the study of institu-

future institutions, politics, resource distributions, political learning, and

1.1.1.2 Political Science

on prosperity outcomes at macro scales, with increasing reference to

1984), reflect an essential linkage to the disciplines referenced above, but,

1.1.1.5 Social Psychology

1 Section 4.2.3 provides an extended discussion of the sociological perspective on

biases that influence individuals’ decision-making to an extent that it may

linguistic characteristics of laws found in different settings. This interest in

3 An extended discussion of the philosophical challenges is provided in Sect. 5.2.

1.1.1.8 Computer Science

draws on fundamental principles of logic and argumentation to analyze

1.1.2 Convergence Toward Interdisciplinary Institutional Analysis

coverage in extant IG research, whereas others have found little. As will

In the following section, we provide an overview of one institutional

1.2 Institutional Analysis with the Institutional

1.3 Primer on the Institutional Grammar

on institutions, amounting to a new paradigm for conducting institu-

refers to a shared linguistic constraint or opportunity that prescribes,

as well as enabling more reliable institutional classification. In the absence

4 To distinguish conditions from action qualifications, in computational studies quali-

organic farming certification.” This statement would be parsed as follows

Attributes = organic farmers

This example clearly elucidates the noted observation that clauses

6 This is an extension of Schlüter and Theesfeld (2010)’s interpretation of the norms-

seem that the IG simply offers another linguistic syntax according to

7 This characterization is refined in Sect. 4.2.2 based on the introduction of atomic

the IG of broad appeal to institutional analysts with various research

collected through an application of the IG can be analyzed. Importantly,

1.4 Overview of Chapter Contents

Workshops on Foundations and Applications of Self-Systems, FAS-W 2016, 210–

Kahneman, D., & Tversky, A. (1972). Subjective probability: A judgment of

Smajgl, A., Izquierdo, L. R., & Huigen, M. (2008). Modeling endogenous

Review of Institutional Grammar Research:

2.1 Guiding Research Questions

© The Author(s), under exclusive license to Springer Nature 33