The Academy of Management Annals

ISSN: 1941-6520 (Print) 1941-6067 (Online) Journal homepage: http://www.tandfonline.com/loi/rama20

Conceptualizing Emergent States:

Mary J. Waller, Gerardo A. Okhuysen & Marzieh Saghafian

To cite this article: Mary J. Waller, Gerardo A. Okhuysen & Marzieh Saghafian (2016):
Conceptualizing Emergent States:, The Academy of Management Annals, DOI:
10.1080/19416520.2016.1120958

To link to this article: http://dx.doi.org/10.1080/19416520.2016.1120958

Accepted author version posted online: 04

Jan 2016.
Published online: 27 Jan 2016.

Submit your article to this journal

Article views: 119

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=rama20

Download by: [Orta Dogu Teknik Universitesi] Date: 08 March 2016, At: 03:15
 The Academy of Management Annals, 2016
http://dx.doi.org/10.1080/19416520.2016.1120958

Conceptualizing Emergent States:

A Strategy to Advance the Study of Group Dynamics
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

MARY J. WALLER*
Schulich School of Business, York University

GERARDO A. OKHUYSEN
Paul Merage School of Business, University of California

MARZIEH SAGHAFIAN
Schulich School of Business, York University

Abstract
In this review, we aim to advance work on group and team dynamics by
examining how important elements of dynamism are embedded in the
current literature on emergent states in groups. We use the concept of emer-
gence as an organizing frame, building block, and critical lens, first summar-
izing key aspects of the extant literature on emergence, and then drawing
four core characteristics of emergent phenomena from this literature. We
use these characteristics to organize our review and examine how emergent
states are portrayed in the past decade of groups literature. We end by
exploring challenges to the development of a more dynamic perspective
∗
Corresponding author. Email: [email protected]

# 2016 Academy of Management

1
 2 † The Academy of Management Annals

and by offering specific suggestions to guide and advance future work on

groups and teams.

Behind complexity, there is always simplicity to be revealed.

Inside simplicity, there is always complexity to be discovered.
—Gang Yu
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

Introduction
Working in groups is a universal human experience. Ancient religious and his-
torical texts recount stories of social dynamics within groups (Watts, 2010),
and one of the first published studies of group behavior in social psychology
appeared in the late nineteenth century (Triplett, 1898). And yet, even
though the literature concerning groups and teams is beyond voluminous,
and even though individuals’ experiences in groups are ubiquitous, accurately
predicting outcomes for any given group, or even set of groups, has proven to
be a remarkably elusive feat. Countless managers and leaders of teams struggle
to apply what few actionable, practical tools are available to incrementally
reduce the uncertainty surrounding team performance. And this is the situ-
ation after over a century of published research on groups and teams.1 We
should know more.
This article joins an active conversation in the field that expresses deep dis-
satisfaction with current approaches to the study of groups and teams. This dis-
satisfaction stems from a wide recognition of the need for new approaches that
emphasize groups’ nonlinear, complex, and dynamic nature, and the seeming
inability of the field to adopt these approaches in its work. Most clearly stated
by McGrath, Arrow, and Berdahl (2000), the field of group dynamics is one
that primarily studies group statics—statics largely ineffective in linking the
complex dynamics within groups to group outcomes. The inability of the
field to advance past primarily static considerations is a problem because it
hobbles our capacity to meaningfully understand how groups perform their
work. This has deeper consequences for the literature on management and
organizations because much of the work in organizations happens in the
context of groups and teams. In fact, without a deep and accurate understand-
ing of how groups and teams function, many other areas of the literature on
organizations become necessarily suspect.
Responses to this situation have included calls for more emphasis on
dynamic elements (McGrath et al., 2000), expansive conceptual treatments
of groups as complex systems (Arrow, McGrath, & Berdahl, 2000), detailed
explanations of how principles of dynamism apply to groups (Cronin,
 Conceptualizing Emergent States † 3

Weingart, & Todorova, 2011), work on the use of different methodologies for
studying group dynamics (Kozlowski & Klein, 2000), and proposals for par-
ticular anchor points and approaches to begin this exploration (Humphrey
& Aime, 2014). Each of these contributions has sought to provide an
impetus for a broader adoption of a dynamic perspective on groups and
teams. Collectively, they define dynamism as interaction at different levels
over time that yields change and cross-level influence manifested in a variety
of ways (see Cronin et al., 2011; Kozlowski, Chao, Grand, Braun, & Kuljanin,
2013; McGrath et al., 2000).
As will be detailed below, our work here differs in three important ways
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

from the previous calls for new approaches to the study of groups and
teams. First, given that others more eloquent than ourselves have identified
the core issues in the current state of affairs in groups research, we do not
focus on elaborating the problems per se, although we do emphasize the
danger of allowing them to persist. Second, and similarly, given that others
more expert than ourselves in a particular method or methods have explicated
the use of these methods for the study of dynamics in groups, we do not advo-
cate particular methodologies here, although we do make note of some of the
previously discussed options. Finally, at the risk of being labeled by some as
pollyannaish and in contrast to many of the previous reviews, we begin with
the belief that fragments of dynamism do indeed exist in the current literature
on groups and teams; consequently, we collect and build on these fragments to
identify specific ways researchers think about dynamism in groups, rather than
calling for a wholesale quantum leap that would involve researchers initiating
simultaneous and radical changes in concepts, vocabulary, and methodology.
To collect these wisps of dynamism and arrange them in what seem to be
a range of stable sets or “building blocks” of dynamism present in current
thinking on groups, we focus in this article very intently on an area of the
groups literature that increases the probability of finding evidence of
dynamic thinking: the literature on emergent states in groups. Emergent
states have been defined as “constructs that characterize properties of the
team that are typically dynamic in nature and vary as a function of team
context, inputs, processes, and outcomes” (Marks, Mathieu, & Zaccaro,
2001, p. 357). We pay particular attention to the emergent nature of these
states and the ways in which their emergence exemplifies dynamism. Our
focus here is based on the belief that emergent constructs, by definition, rep-
resent the most dynamic conceptualizations in the groups and teams litera-
ture (Cronin et al., 2011).
Beginning with the highly cited framework of processes and emergent states
offered by Marks et al. (2001), we review the conceptualizations and treatments
of emergent states in groups from the publication of that article until present
day. We organize our review using four central characteristics of emergent
phenomena discussed in the literature on emergence drawn from the
 4 † The Academy of Management Annals

philosophy and complexity science fields. We also detail how the dynamic con-
ceptualizations of emergent states could be extended in future work. First,
however, we anchor our review in general concepts from the literature on
emergence ab initio.

Fundamental Emergence
Our approach is grounded in general expositions of emergence which, to our
knowledge, have not been completely leveraged in existing group dynamics lit-
erature; this approach, however, was advocated by Cronin et al. (2011) in their
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

recent critique of the teams literature, in which they suggested that a more
precise conceptualization of emergence could help advance explorations of
teams from a dynamic perspective. Compared to other elements they
described, emergent constructs are the most variable and dynamic, and form
over time through a synthesis of the interactions of team members (Cronin
et al., 2011). Tellingly, in spite of the potential of emergent constructs to
help advance our thinking in group dynamics, and in spite of their presence
in the literature, Cronin et al. (2011) concluded that authors fail to adequately
examine the dynamic nature of these constructs. Therefore, we facilitate the
beginnings of a more dynamic conceptualization of groups that is grounded
in an understanding of emergence in general and catalyzed by the group and
team literature’s multiple treatments of emergent states.
In general, the notion of emergence has a long and contentious history; spir-
ited discussions continue today in forums politely described as “discordant”.
The philosophy of science and complexity science literatures contain numer-
ous treatments and conceptualizations of emergence, many of which pertain
primarily to systems in physical science, biological science, or mathematics;
although a comprehensive review of these literatures is beyond the scope of
our work here, and although the literatures are replete with vigorous debates
regarding many open questions about emergence, we draw from their more
central and generalizable concepts. As succinctly put by Ablowitz (1939), the
theory of emergence
declares, in brief, that the whole is more than the sum of its parts . . . It
declares that there are levels of existence, and that there is a “tendency of
units of one kind in combination, to constitute units of a new kind, with
more complex constitution and new qualities due to the new together-
ness of the parts” (Sellars). (1939, p. 2)
More contemporary conceptualizations of emergence in social science can
be found in literature on complexity and complex systems (see Corning,
2002; Hodgson, 2000), which define emergence as “the arising of novel and
coherent structures, patterns, and properties during the process of self-organ-
ization in complex systems” (Goldstein, 1999, p. 49). These new phenomena
 Conceptualizing Emergent States † 5

that emerge in a system are not contained in any of the system’s parts, nor can
they be predicted simply from knowledge of the underlying level (Emmeche,
Køppe, & Stjernfelt, 1997; Lichtenstein, 2014).
Importantly, emergent phenomena are distinct from the process or mech-
anism of emergence itself (Kozlowski & Chao, 2012). In the context of
groups and teams, emergent states and other phenomena emanate or
“bubble up” from lower-level constituent properties that exist in the form
of individual or dyadic behaviors (Humphrey & Aime, 2014; Kozlowski &
Klein, 2000); furthermore, Marks and colleagues maintain that emergent
states “vary as a function of team context, inputs, processes, and outcomes”
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

(2001, p. 357), clearly suggesting that emergent states exist only when a team
has context, inputs, processes, and outcomes. This point may seem comple-
tely obvious; however, consider a situation in which a researcher measures a
construct such as team anxiety by aggregating individuals’ anxiety measured
before the team has been formed or has worked together. Strictly speaking,
one cannot conceptualize team anxiety as an emergent state under such con-
ditions. Thus, taking the work of Marks and colleagues together with our
reading of work from the philosophy of science and complexity science lit-
eratures, we suggest a broader point, namely that emergent team phenomena
emanate from behaviors of group members, and include emergent states, be-
havioral patterns, and structures. In other words, emergent team phenomena
may be (1) states such as collective cognition, affect, or other relatively
enduring properties, (2) behavioral patterns such as conversational routines,
conflict resolution, or feedback seeking, or (3) structures such as sub-groups
or hierarchies.
Important to our discussion here is the realization that emergent phenom-
ena in groups and teams exemplify “strong” emergence, partly due to the
notions of coherence and ostensiveness, which will be discussed later. Concep-
tualizations of strong emergence suggest that emergent phenomena arise from
micro levels, are both irreducible and autonomous from the underlying level,
endure for some period of time, and can exert downward causal forces or
“mutual causality” that affect the underlying parts themselves (Lichtenstein,
2014, p. 20, 112). For example, the emergent state of trust in a team may
exert downward causal forces on subsequent lower-level dyadic interactions
by increasing the probability of personal disclosures between team members.
This is congruent with McGrath and colleagues’ assertion that emergent
group-level variables “emerge from and subsequently shape and constrain”
lower-level dynamics (2000, p. 98). In contrast, theorists explain “weak” emer-
gence as not involving any downward influence as in situations of strong emer-
gence. For example, discrete behavioral processes in groups and teams could be
seen as products of weak emergence due to a lack of this subsequent downward
influence; although behavioral processes may lead to the emergence of
phenomena that could eventually influence the underlying level, they
 6 † The Academy of Management Annals

themselves are fleeting—not enduring over time—and do not exert subsequent

downward influence on the underlying micro level of team member
interactions.

Four Characteristics of Emergent Phenomena

The general literature on emergence suggests four characteristics of emergent
phenomena particularly useful for our discussion regarding emergent states in
groups and teams—namely, that these phenomena are global, coherent, osten-
sive, and radically novel (e.g. Bedau & Humphreys, 2008; Goldstein, 1999). The
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

global nature of emergent properties is based on the notion, fundamental to

emergence, that there are different conceptual levels of existence; emergent
properties are conceptualized to emerge from lower- or micro-level com-
ponents (Goldstein, 1999; McGrath et al., 2000). This concept is also referred
to as “supervenience” or “the sense in which emergent phenomena, while
having a distinct existence, nevertheless depend on more basic phenomena”
(Bedau & Humphreys, 2008, p. 10). The recent work of Humphrey and
Aime (2014) on team microdynamics, for example, emphasizes the need for
more attention in research on the underlying dynamics among team
members, and particularly on how they relate to and interact with each
other, in order to better understand subsequent (and emergent) organization
and teamwork.
The coherent nature of emergent phenomena pertains to their enduring
identity as wholes over some amount of time. Lissack and Letiche suggest
that a coherent phenomenon possesses “some rigor and sustainability
(that) may be temporary, but it is not entirely fleeting” (2002, p. 84),
making it possible for the phenomenon to exert subsequent downward influ-
ence on its underlying level. Logically, this endurance would also suggest
some inertia, making a sudden change of an emergent phenomenon difficult.
Additionally, Harper and Endres (2012) maintain that coherence involves
phenomena with components that interact in a unique and dynamic
fashion. As a result, each manifestation of “emergent phenomena will be
different to some degree from previous ones” (Goldstein, 1999, p. 50). For
example, a collective sense of trust may emerge over time as individual
team members interact with each other; although unique to the team and
dynamic, this emergent phenomena sustains and endures in a relatively
stable way over some period of time.
Due to their coherent nature, emergent phenomena are also ostensive,
meaning they are able to be experienced and recognized by team
members. A good amount of debate exists among theorists regarding the
requirement for emergent phenomena to be ostensive (Corning, 2002); if
no one recognizes it, has a phenomenon not emerged? However, for our pur-
poses here, considering that groups and teams are social systems, the ability
 Conceptualizing Emergent States † 7

of team members to sense, experience, perceive, or react to an emergent

phenomenon—that is, the ostensive nature of the phenomenon—seems fun-
damental for our examination of emergence. Ostensiveness can also include
the ability of team members to recognize, name, and/or discuss an emergent
property. For instance, Goldspink and Kay argue that emergence in social
organizations involves human agents with the distinctive ability to recognize
“self” from “other” and reflexively interact with their environments. Due to
this distinction, organizational participants can recognize or “notice an emer-
gent pattern of social behavior and explicitly denote it as a ‘norm’” (Gold-
spink & Kay, 2010, p. 56) and then go on to act on the basis of this
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

recognition of an ostensive emergent phenomenon, and potentially act to

change the pattern. This feedback path is, they suggest, unique to emergence
in social systems due to the self-awareness of human actors. Thus, two
important aspects of ostensiveness in relation to emergent states in groups
and teams are (1) the experiential nature of the state (i.e. group members
experience it and/or recognize it), and (2) the agency felt by group
members concerning their ability or desire to actively change the state
(Coen & Schnackenberg, 2012, p. 457).
Finally, the radical novelty of emergent phenomena suggests their features
have not been previously observed in lower-level phenomena from which
they emerge, and that they cannot be perfectly reduced back to their lower-
level constituent parts; such irreducibility, therefore, leads to the conclusion
that they are unpredictable (i.e. not perfectly predictable) from lower-level
inputs (Goldspink & Kay, 2010, p. 52; McLaughlin, 2008), and that they are
non-deducible from the lower level (Goldstein, 2013). Lichtenstein, for
example, describes this novelty of an emergent phenomenon as “meaning
that its properties transcend its components, producing outcomes that are
unpredictable and surprising even with a full understanding of the com-
ponents” (p. 19). Goldstein (1999) further suggests that outcomes of emergence
can be arranged on a continuum, loosely reflecting radical novelty, with simple
properties on one end (e.g. emergent states), patterns in the middle (e.g. recur-
rent sequences), and orderly structures or systems (e.g. persistent faultlines) on
the more complex end.
In the next section, we proceed with our review of emergence in the
groups and teams literature. To reiterate, in order to pursue our goal of
enhancing the study of team dynamics, in this review we concentrate on
one type of emergent team phenomena: emergent states. Following the pre-
vious organizing frameworks of scholars discussing emergence (e.g. Andriani
& McKelvey, 2009; Bedau, 2008) and most notably that of Lichtenstein
(2014), we characterize these efforts in terms of authors’ conceptualizations
and treatments of emergence, rather than in terms of the nature of the emer-
gent properties themselves.
 8 † The Academy of Management Annals

Emergent States in Groups

We began our work by identifying articles that cited Marks et al. (2001), pro-
ducing a set of more than 1000 articles. We reduced that by focusing on 110
articles that explicitly deal with emergent states, whether or not the authors
used that label. We proceeded to build our review by identifying the intersec-
tions between these articles and the four characteristics of emergence—global,
coherent, ostensive, and radically novel—we identified above. Through several
iterations, we developed the review with an emphasis on identifying work that
can be considered exemplary and examining both articulated and unstated
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

assumptions; additionally, we noted opportunities for further development

of knowledge about dynamism. Our aim, ultimately, was to develop a review
that usefully built on the emergence literature, providing both detail on the
conceptualizations and methodologies, but also offering clear pathways for
future work. Figure 1 depicts the four emergence characteristics and the
relationships among them in a group context.
We begin our review by exploring the global characteristic as a wide-
spread, fundamental and largely unarticulated assumption in most conceptu-
alizations of emergent states. We then explore the remaining emergence
characteristics in decreasing order of their articulation in the literature
reviewed—coherent, ostensive, and radically novel. Importantly, these charac-
teristics should not be considered mutually exclusive; instead, they are better
conceptualized as filters through which we see different aspects of dynamism
come into relief.

Figure 1 Characteristics of Emergent States in Groups.

 Conceptualizing Emergent States † 9

1. Global
The global characteristic of emergence captures the presence of different con-
ceptual levels of existence. In a general manner, emergent phenomena exist at a
higher or global level, and emerge from the dynamic interactions that take
place among components of an underlying level (Goldstein, 1999; McGrath
et al., 2000). In groups and teams, that underlying lower level exists in the
“microdynamic level” (Humphrey & Aime, 2014) where individuals interact
to coordinate and organize. These lower-level interactions give rise over time
to emergent states at the global level. In the literature on team and organiz-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

ational phenomena, the existence of levels is fundamental (Kozlowski &

Chao, 2012; Kozlowski & Klein, 2000), and the widespread adoption of this
assumption is indicative of multi-level thinking in organizational behavior
and related disciplines (Rousseau, 2011).
Research on emergent states incorporates this global characteristic through
the generally unarticulated assumption that different conceptual levels (indi-
vidual, interactions, group) of existence are possible. From a methodological
perspective and in the simplest terms, this assumption produces the
common practice of measuring emergent states by eliciting responses from
individual group members and aggregating these responses to the group
level. Most of the aggregation models present in the work we reviewed take
into account the distribution of individual responses, and we observed three
approaches used to characterize these distributions: convergence, dispersion,
and configuration. Authors tend to readily provide a clear justification for
their choice of aggregation approach; in case of convergence models, this is
done through assessment of within-group agreement indices, whereas for the
other two models the rationale can often be found when the authors define
and conceptualize the emergent state (see also Coultas, Driskell, Burke, &
Salas, 2014).

Global as convergence. When using convergence models, which are also

referred to as consensus models (Chan, 1998a) or composition forms
(Kozlowski & Chao, 2012; Kozlowski & Klein, 2000), researchers assume
that group members have homogenous experiences or hold similar perceptions
with regard to the group. Our review revealed that the majority of emergent
states in this work have been conceptualized and operationalized as convergent
(see also the review by Coultas et al., 2014). For some emergent states, the
shared character is explicitly incorporated into the conceptualization, particu-
larly when intended to capture dimensions of team climate. Psychological
safety, for example, is defined as “shared belief by team members that the
team is safe for interpersonal risk taking” (Edmondson, 1999, p. 354). Accord-
ingly, the common practice (e.g. Bresman & Zellmer-Bruhn, 2013; Faraj & Yan,
2009) is to collect members’ responses regarding their perception of the level of
psychology safety within their team. The next step is to take an average of those
 10 † The Academy of Management Annals

scores to create a group-level measure. Such aggregation is typically justified by

demonstrating an acceptable level of within-group agreement. The same meth-
odological approach has been used for other constructs such as cohesion (e.g.
Shin & Choi, 2010; Wei & Wu, 2013), trust (e.g. De Jong & Elfring, 2010;
Langfred, 2007), team empowerment (e.g. Kirkman, Rosen, Tesluk, &
Gibson, 2004; Mathieu, Gilson, & Ruddy, 2006), and collective efficacy (e.g.
Chen, Kanfer, DeShon, Mathieu, & Kozlowski, 2009; Porter, Gogus, & Yu,
2011).
The underlying assumption behind convergence models is that agreement
among members regarding a given aspect of group functioning (whether cog-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

nitive, affective, or motivational, whether conscious or not) provides strong

evidence that a group-level phenomenon exists. In essence, strong consensus
gives weight to the notion that a global property has emerged and that it
may have strong implications for team processes and outcomes. Nevertheless,
our review suggests that researchers often take the shared character of emer-
gent states as given, without always providing a rationale for the plausibility
of this assumption. For example, De Jong and Elfring (2010) hypothesized
the mediating role of different team processes between trust and team perform-
ance. They defined trust as “shared generalized perceptions of trust that team
members have in their fellow teammates” and used a convergence aggregation
approach. Although their conceptualization of trust does not explicitly explain
why consensus in members’ perception of trust is important, they describe
team-level trust as a shared construct emerging as a result of team membership,
social categorization processes, collective sensemaking, and other contextual
factors that shape the interactions among members.

Global as dispersion. Although much of the research in our review used

convergence as the underlying assumption for emergent states, we also
noted work that explicitly considered other possibilities. For instance, dis-
persion models (Chan, 1998a) reflect an effort to capture the heterogeneity
(or, statistically speaking, the variance) of responses that members provide.
Said another way, these models attempt to account for the absence of agree-
ment in members’ experiences or perceptions. Jehn, Rispens, and Thatcher
(2010), for example, introduce the concept of conflict asymmetry and
discuss why and how asymmetries in perceptions of conflict among
members have an effect on group outcomes (particularly performance and
creativity). Specifically, they show that accounting for asymmetry (in addition
to agreement) can help explain a greater amount of variance in performance. In
this case, the authors assess conflict asymmetry by calculating the standard
deviation for members’ scores assessing task and relationship conflict. An
important implication here is that consensus (i.e. a mean value) and dispersion
(i.e. a standard deviation value) in members’ perceptions characterize different
types of relationships between lower and higher levels.
 Conceptualizing Emergent States † 11

In research on mental models, dispersion approaches have been used exten-

sively, and one common approach is to elicit members’ knowledge of the
relationship between concepts that represent different aspects of taskwork or
teamwork, and use this information to construct representations of individuals’
mental models. Researchers can then compare these individual mental models
to assess the heterogeneity (i.e. difference) of knowledge structures
(Mohammed, Ferzandi, & Hamilton, 2010). For instance, Resick, Murase,
Randall, and DeChurch (2014) examined the role of mental models that
capture the collective understanding of the implications of key strategic
decision alternatives. The authors hypothesized that similarity among
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

members’ “strategy-focused team mental models” would mediate the effect

of team cognitive ability on information elaboration. To test this hypothesis,
they studied teams executing a strategic decision-making task and elicited
members’ knowledge of the relationships among a number of strategic
decisions, constructed a representation of each member’s mental model, and
calculated the degree of similarity across members’ models. Their results
show members with high cognitive ability are more likely to reach a
common understanding of their strategic options and develop similar team
mental models, leading to more effective information exchange.

Global as configuration. Configuration models, which have also been

referred to as compilation forms (Kozlowski & Chao, 2012; Kozlowski &
Klein, 2000), seek to capture the structure of an emergent state by assuming
that it is neither consensus (as used in convergence measurement models)
nor difference (as captured by dispersion measurement models) alone that
matters; instead, these models suggest the manner in which lower-level
elements are arranged has a profound influence on teams. Naturally, configur-
ation models capture divergence, but in a qualitatively different manner than
dispersion models that focus on variation alone—configuration models
examine the structural representation of lower-level responses rather than
relying only on arithmetic differences or similarities. For instance, Mell, van
Knippenberg, and van Ginkel (2014) distinguish between two structures of
metaknowledge, or transactive memory, in groups: centralized and decentra-
lized. Metaknowledge—the knowledge of who knows what in the team—is cen-
tralized when it is concentrated within one member, such as a leader who is
aware of each member’s expertise in a group. Decentralized metaknowledge
occurs when information about who knows what is distributed among team
members. The results of their study show that teams with centralized configur-
ation of metaknowledge outperform those with a decentralized configuration
when the distribution of task information among members imposes high
coordination demands.
In another example, Carson, Tesluk, and Marrone (2007) focused on shared
leadership and conceptualized it as an emergent state that characterizes the
 12 † The Academy of Management Annals

distribution of leadership influence across team members. In contrast to pre-

vious research using a convergent perspective of what the team as a whole pro-
vides in terms of leadership (D’Innocenzo, Mathieu, & Kukenberger, 2014),
Carson et al. (2007) argue that convergent approaches fail to capture the com-
plexities of shared leadership. Instead, they take a social network approach to
investigate the antecedents and consequences of different patterns of shared
leadership. To do so, they first elicited the perceived network of shared leader-
ship by asking each member to rate the extent to which each of his or her team-
mates exerted leadership in the team, and then combined these networks to
calculate the measure of density for the overall network. As evidenced by the
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

work of Mell et al. (2014) as well as that of Carson et al. (2007), using configur-
ation models appears to drive researchers to present an explicit articulation for
their choice of aggregation model when conceptualizing and operationalizing
emergent states of interest.
Convergence, divergence, or configuration perspectives each capture the
global nature of emergent states, involving lower and higher levels. In our
view, however, these perspectives suffer from two potential shortcomings.
The first is the common assumption of homogeneity of perception or experi-
ence among group members. This assumption imposes significant boundary
conditions for work incorporating such models, especially when it is left unar-
ticulated. The literature shows that, for instance, group members may have
widely varying perceptions of such relatively objective targets as time
(Waller, Conte, Gibson, & Carpenter, 2001) and safety (Huang et al., 2007);
to assume that group members’ perceptions of such subjective targets as emer-
gent group trust or cohesion must converge in order for an emergent phenom-
enon to exist seems tenuous.
Second, our review indicates that researchers are well aware of the existence
and importance of multiple levels when examining emergent states, but
detailed characterizations of the micro-level interactions that connect the
levels are generally absent. As Humphrey and Aime (2014) noted, most
research on groups and teams privilege treatments that rely on static aggrega-
tion of constructs to approximate emergent phenomena rather than investigate
and chart the dynamic interaction that gives rise to novel states, patterns and
structures. As a consequence we, as a field, have created a “black box” that is
often left unexplored; the methodologies used to explore emergent states fail
to capture the global essence of these phenomena. This issue is most clear in
research that measures individuals’ more stable characteristics such as intelli-
gence or ability and aggregates these measures to suggest the existence of a
team-level emergent state. While individual-level inputs provide a starting
point for member interactions and influence them greatly, they are likely insuf-
ficient to fully capture global emergent phenomena (Humphrey & Aime,
2014). Relegating the role of dynamism in emergence to an unspecified black
box hinders explorations and understanding of groups as dynamic entities.
 Conceptualizing Emergent States † 13

2. Coherent
As we described earlier, the coherent characteristic of emergent phenomena
highlights their persistent nature. For researchers of groups and teams, embra-
cing coherence as a property of emergent states indicates that we focus on how
they both endure and vary over time.

Coherence as endurance. In our review, we found that the groups and

teams literature does not include coherence as an overt organizing frame in
examinations of emergent states. For instance, we found no articles that expli-
citly articulate “coherence” to indicate endurance of emergent states; in con-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

trast, the idea of coherence as endurance and continuity has explicitly

appeared in work on emergent constructs at other levels of analysis (e.g. organ-
izational identity in Gioia, Patvardhan, Hamilton, & Corley, 2013). Work at the
team level typically includes an implicit assumption that emergence states
endure over time; researchers often measure the emergent state only at one
point in time, presumably based on the belief that one measurement is a
reasonable estimate of how the emergent state existed in the recent past of
the group, or will exist into the near future. Any other assumption would
render the research results applicable to and useful for only the one point in
time.
The assumption of endurance is more clearly operative when longitudinal
approaches are used in research designs. Our review identified a small
number of studies that include more than one round of measurement for an
emergent state—a step necessary to specifically establish or identify endurance.
Curşeu and Schruijer (2010), for instance, use a two-wave longitudinal data
collection design to explore alternative causal paths between trust and conflict
and to ascertain whether trust predicts the emergence of conflict or conflict
predicts the emergence of trust. Their findings support a causal order in
which trust (at Time 1) acts as an antecedent to decreased conflict (at Time
2). Moreover, their results suggest that trust at T2 was only predicted by
trust at T1, and not by conflict at T1, suggesting that the relationship
between trust and conflict may be unidirectional. From the perspective of
coherence as endurance, this exploration shows endurance for both of these
emergent states and seeks to map the relationships between them.
Mathieu, Heffner, Goodwin, Cannon-Bowers, and Salas (2005) use a differ-
ent longitudinal approach to examine the influence of the interaction between
the quality of mental models and mental model sharedness on team processes
and performance. In this work, student teams completed six missions on a
flight simulator and the researchers measured task and team mental models
at three different times. The data were collected in advance of three task epi-
sodes, each one encompassing two missions. In the initial phase of their data
analysis, the authors used repeated-measures analyses of variance for each vari-
able to assess the change in value over time. Their results show stability in the
 14 † The Academy of Management Annals

quality and sharedness of team and task mental models, which did not change,
clearly portraying how emergent states endure over time.
Lastly, we identified one article that provides a tentative but more direct link
to coherence as endurance. In their work, Okhuysen and Richardson (2007)
argue that emergent states should exhibit particular characteristics including
stability of meaning across time—that is, consistency in interpretation by
members of a group over time. Although not using exact language of “coher-
ence as endurance”, they argued that “though emergent states are, by defi-
nition, dynamic and mutable, the underlying domain of a given emergent
state should . . . remain constant over time” (p. 152). Focusing on group con-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

flict, Okhuysen and Richardson (2007) measured three types of conflict (i.e.
task, relationship, and process) in student groups at four different points in
time. To test for stability of meaning, they established measurement invariance
over the four measurement occasions (see Chan, 1998b, for further details
about this method). Such an approach illustrates one way in which endurance
of an emergent state can be operationalized.
By examining coherence as endurance, our review identified the exemplars
noted above, but also illuminated how as a field we seem to have no way to
measure or account for state absence—that is, for the possibility that states
may not have emerged or endured. Our current methods rely mostly on ascer-
taining the level of a state (at one or multiple points in time) from low to high,
always assuming emergence and endurance. For example, we ask members to
report the level of trust or conflict in their groups, rather than asking if trust or
conflict exist. Not surprisingly, most members comply by providing an answer
to what could be a leading question. Our science and approach assume that a
focal emergent state must be present—perhaps at a low level, but present none-
theless; we do not or cannot account for moments when emergence has not
occurred, although absence of a key property is central to understanding a
phenomenon. Perhaps, then, the emergence lens cannot be adopted wholesale
for use in groups and teams after all, as we do not seem to account for pre-
emergence conditions in groups, assuming instead that emergence occurs
within the first moments of micro-level interaction (Coultas et al., 2014), or
perhaps even before interaction, based on past experiences and expectations
members bring to a new group (see O’Neill & Allen, 2012).
If the period of time without emergent states in groups is so fleeting as to be
inconsequential, then the characterization of such states as emergent seems
inaccurate and ill-advised. However, the lack of attention to a pre-emergence
period in groups may be indicative of the so-called “hard problem” of emer-
gence (Mitchell, 2009)—the difficulty associated with specifying the actual
mechanism of emergence, whether in groups or in other contexts. Thus, and
as we suggest in more detail later, future research should peer more intently
into the mechanism, account for the past experiences and expectations of
members, and move away from assuming omnipresence of emergent states
 Conceptualizing Emergent States † 15

in groups. Our understanding and portrayal of the conditions leading to dyna-

mism in groups, we believe, will grow as a result.

Coherence as variability. The dynamism of emergent states is captured

through coherence as variability and reflects how states can change in intensity;
this characteristic of emergent phenomena has also been described as “dynami-
cal” (Goldstein, 2013), indicating that phenomena emerge and change as a
system evolves. In particular, the variability of emergent states distinguishes
them from more static characteristics such as structure. Marks et al. (2001)
characterize emergent states as “psychosocial traits” (although they prefer
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

the term “states” due to their mutable quality) that are affected by (and there-
fore vary according to) particular features of process, behavior, and context.
Similarly, Cronin and colleagues (2011) distinguish between emergent and
cumulative constructs, noting that emergent constructs derive from inter-
actions among members (making them highly dynamic) while cumulative
ones are based on more stable properties (such as the amount of time
members have spent in a group). Methodologically, examining variability
implies longitudinal approaches, such as work that examines the trajectories
of emergent states. Our review identified a small number of studies that
conduct more than one round of measurement and also focus on tracking
and understanding the variability of emergent states over time.
As one example of coherence as variability, Santos and Passos (2013) con-
ducted a longitudinal study examining the effect of team mental models
(measured at T1 and T3) and conflict (T2) on team outcomes (T5). In their
model one path linked mental model similarity (a group-level variable) at
the beginning (T1) to three different team outcomes, mediated by team
mental model similarity in the middle of the lifecycle (T3). Additionally,
they hypothesized that team mental models would become more similar (i.e.
converge) over time. They tested this hypothesis by tracing the development
of team mental models between T1 and T3, but found no evidence of major
change in team mental model similarity over time. The researchers speculated
that these group members might not have spent sufficient time together to
increase their common understanding. In this research, coherence as variability
is proposed as both an outcome from group processes but also as an explana-
tory factor for broader outcomes. As such, it shows how measuring emergent
states at two points in time incorporates coherence as variability in two ways.
First, it allows for theorizing change in the level of an emergent state; second, it
allows for theorizing how the later level of an emergent state may be related to
an earlier level.
Another example of how coherence as variability is present in research on
emergent states is provided by Zellmer-Bruhn, Maloney, Bhappu, and Salvador
(2008), who focused on perceived similarity—the degree to which members
view themselves as having few differences—and theorized under what
 16 † The Academy of Management Annals

conditions it would influence team outcomes. Their work was concerned with
two types of perceived similarity: (1) perceived social category similarity, based
on visible characteristics such as nationality, ethnicity, and sex; and (2) per-
ceived work style similarity, based on less-visible characteristics such as com-
munication style and work ethic. They hypothesized that members’ initial
perceptions of social category similarity would not change dramatically over
the life of a team, since these perceptions were based on observable and
static member characteristics. In contrast, they argued that perceived work
style similarity would vary as information about unobserved member charac-
teristics was revealed through interactions such as information sharing or con-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

flict. To test their hypotheses, the authors measured perceived social category
similarity and perceived work style similarity twice (at T1 and T3) and also col-
lected data for information sharing and conflict at T2 in a sample of first-year
MBA students. The results supported their hypotheses; in particular, perceived
work style similarity was found to decrease significantly over time, consistent
with the proposition that members would overestimate similarity in early
stages of groups’ activities but adjust their evaluations as differences were
revealed. The authors argued that such overestimation might be due to their
specific context by noting that MBA students might easily overestimate their
similarity to others at the start of their program. This study provides a good
example of explicit theorizing not only about the trajectory of change in emer-
gent states, but also how process variables, such as conflict and information
sharing, contribute to that variability.
Both the work of Santos and Passos (2013) and Zellmer-Bruhn et al. (2008)
are examples in which emergent states were measured at two points in time
and which clearly reflect coherence as variability. Other studies have used
measurement of emergent states at more than two points in time to explore
variability as a trajectory, explicitly articulating the functional form of emer-
gence. These studies go beyond describing evolution as a straightforward
matter of establishing increases and decreases in intensity, presenting a more
comprehensive view of the dynamics of emergent states.
In developing their arguments regarding team meeting attitudes (TMA)—
members’ thoughts, feelings, and evaluations of team meetings—O’Neill and
Allen (2012) hypothesized that trajectory dynamics depend on the experience
of members with meetings. In particular, they suggested that in teams com-
prised of members with limited meeting experience, within-group homogen-
eity and between-group heterogeneity on TMA would increase over time. To
test this hypothesis, they sought to capture variability in TMA convergence
by measuring it at four different times. In their work, they depict the results
of their analysis as a trajectory comprised of four points, plotting the level of
convergence of team members’ TMA—indicated by intraclass correlation coef-
ficient (ICC)—to highlight its change over time. Their results show that the
ICC values increased between T1 and T2 and subsequently became stable,
 Conceptualizing Emergent States † 17

partially supporting their hypothesis. O’Neill and Allen’s (2012) work,

however, is exemplary in another way. The development of a longitudinal
model of variability also required a more fine-grained articulation of the
dynamics underpinning the groups in the study. For instance, with multiple-
period values of TMA available, the authors explain how they faced a choice
regarding which one(s) to use as predictors for outcome variables (perceived
team meeting effectiveness, time spent in meetings, team potency, and team
performance) measured at T4. In conceptualizing the relationship between
different constructs, the authors articulated why using the shortest interval
(between TMA at T3 and outcome variables at T4) was appropriate, even as
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

they noted that the level of agreement on TMA was higher (and thus could
be considered more robust) at T1 and T2. Facing such choices caused the
authors to expand on their perspective regarding TMA dynamism; they
explained that the later measure, although perhaps quantitatively less robust,
was in fact a better representation of the phenomenon because it allowed
more time for team meeting interactions to influence the maturation of
TMA before its effect on other constructs was measured. In other words,
this research shows how measuring an emergent state at multiple time
points enables researchers to explore when and at what level it exerts its stron-
gest effects on team outcomes. Furthermore, this work elegantly demonstrates
how the conceptualization of variability is not merely a question of measuring
an emergent state at multiple times, but also requires additional theoretical
work to deepen understanding of other issues related to the dynamism of
teams.
A final example of work using trajectories to capture coherence as variabil-
ity is provided by Okhuysen and Richardson (2007). These authors argued that
emergent states incorporate coherence as endurance, as noted earlier, and also
articulated three characteristics that reflect different forms of coherence as
variability: variability over time, difference in patterns of variability among
groups, and variation in consensus across time. Using a four-wave longitudinal
design to measure three types of conflict (i.e. task, relationship, and process) in
student groups, the authors assessed the presence of each characteristic in each
form of conflict. To test for variability over time, they specified the trajectory of
change in latent means for each type of conflict and demonstrated how each
type of conflict unfolded along a distinct path. To test for difference in patterns
of variability, they compared the trajectories of each type of conflict between
high- and low-performing teams. Their findings confirmed differing patterns
of emergence between high and low performers, albeit more so for task conflict
than relationship and process conflict. Finally, variation in consensus across
time was tested by calculating ICC values for each type of conflict at each
measurement point, and the authors concluded that the level of within-
group agreement can vary across different types of conflict. This study offers
some initial hints regarding the aspects of an emergent state that may
 18 † The Academy of Management Annals

change over time (i.e. coherence as variability) and those aspects of it that may
remain constant (i.e. coherence as endurance).
Overall, there are very few articles that examine the trajectories of emergent
states, making it difficult to ascertain exactly how coherence as variability oper-
ates. One challenge associated with simultaneously considering the enduring
and variable nature of emergent states concerns possible inconsistencies in
the definition of coherence. For instance, some authors have proposed that
in the early life of a group, task conflict may be high. Over time, as the
group implements its approach towards its goals, task conflict may disappear
(see Jehn & Mannix, 2001). In other words, it is the very presence of the emer-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

gent state of task conflict that causes members of the group to behave in a way
that reduces (and perhaps even eliminates) that task conflict. In this example,
task conflict exhibits coherence as variability but, if it disappears altogether in
the process of execution, does it exhibit coherence as endurance? We would
suggest that it has indeed endured, existing long enough to influence sub-
sequent lower-level interactions. Coherence as endurance does not suggest per-
manence; as described previously, although coherence requires a certain
amount of temporal sustainability, it suggests a phenomenon that is neither
completely permanent nor completely ephemeral (Lissack & Letiche, 2002).
An emergent state could appear or disappear at any time; the point is that
the emergent state exists for some period of time in the group and can vary
over that time.
In sum, the literature on emergent states provides examples of the notion of
coherence as endurance and coherence as both endurance and as variability.
This helps us draw clearer boundaries around what emergent states in teams
are and do: they endure over some portion of time, they change over time,
and they have the potential to influence group member interactions that
occur subsequent to their emergence. This potential for influence places
them squarely in the “strong emergence” category, and also provides the
avenue for at least two resulting effects: team members’ experiences of an emer-
gent state, and the influence of an emergent state directly on team members’
behaviors—including their deliberate attempts to change the state. These two
possibilities are described in the following section on the ostensive nature of
emergent states.

3. Ostensive
In the general literature on emergence, system characteristics are said to be
ostensive when they can be sensed or perceived. Such a view is consistent
with the way in which the literature on groups and teams treats emergent
states as perceivable. However, there are important differences in how the
ostensive dimension operates in physical systems—realms that are the focus
of much of the literature on emergence—and in social systems such as
 Conceptualizing Emergent States † 19

groups and teams. For instance, in considering ostensiveness for a physical

system, the perceiver of an emergent property is an outsider to the phenom-
enon and the system in which it has emerged. This stands in contrast to
social systems such as groups and teams in which outside perceivers (such
as managers) are relevant, but so are the perceptions of components of the
system itself—that is, the people whose interactions create the lower level
from which states emerge. Moreover, while in physical systems the description
of emergence can be treated as “objective” and “factual”, in social systems the
perception of an emergent state is subject to interpretation by observers and
members. Lastly, and again in contrast to physical systems, people in social
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

systems also bring with them histories, motivations, desires, goals, and inten-
tions. As a consequence, they may choose to act in order to modify their social
circumstance, affecting the social system and emergent states through their be-
havior. Perhaps rather obviously, these differences are due to the presence of
people as participants and constituent elements of social systems (Goldspink
& Kay, 2010). As a consequence of these differences, the ostensive character-
istic must be extended somewhat to account for the perception of people
with particular intentions, goals, or desires. Thus, we suggest that a useful
way to think about emergent states as ostensive is to note that they are per-
ceived, and that such perception means they are experienced by people and
can influence human behavior.

Ostensive as experienced. Our literature review did not identify direct dis-
cussions of the ostensive nature of emergent states. However, we did note that
researchers typically operationalize their measurement through surveys,
relying on the assumption that members can experience emergent states and
respond accurately based on that experience. In fact, in our review we observed
that the vast majority of studies rely on such self-reports, with few examples in
which outside observers (i.e. supervisors, managers, or independent coders)
were used to determine the existence or level of a given emergent state. This
suggests that the “ostensive as experienced” nature of emergent states is
largely assumed to be true in this literature, and can be measured due to
members’ ability to reflect and report on their experiences.
Because the assumed ostensive nature of emergent states is most evident
through measurement, in our review we focused on different ways in which
this measurement is implemented. Thus, for instance, an important difference
in conceptualizing ostensive as experienced is evident when researchers choose
between direct-consensus or referent-shift consensus models of measurement
(Chan, 1998a). In both cases, researchers argue that convergence is the appro-
priate way to characterize an emergent state, and most often aggregate the
responses of members into a single mean value to reflect the collective percep-
tion of a group. When using a direct-consensus model, researchers ask
members for perceptions about themselves and their own experiences in the
 20 † The Academy of Management Annals

group. For example, studies that use individual affect measures (i.e. what one
feels) and average the scores across members to create a group measure use the
direct-consensus model. For instance, Peralta, Lopes, Gilson, Lourenço, and
Pais (2015) implemented a direct-consensus model in measuring team affective
tone. The authors hypothesized that team affective tone—team members’ con-
sistent or homogenous affective experiences—would moderate the relationship
between team innovation processes and team reputation among third parties
such as clients or supervisors. To test this hypothesis the authors asked each
member to rate the extent to which he or she experienced specific emotions
in the team through a Likert-type scale that included items such as “My
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

team made me feel happy” and “My team made me feel frustrated”. To
obtain a team-level measure, individual members’ responses were then aggre-
gated as an average for model testing.
In contrast, when using a reference-shift consensus model, researchers
capture the experience of the group as a whole by asking members to report
on or evaluate a group-level construct or experience in their team. Mitchell
et al. (2014), for instance, also examined affective tone as Peralta et al.
(2015) did, but focused on the negative elements. They proposed that team
members’ consistent or homogenous negative affective experiences moderate
the effect of interprofessional motivation—members’ drive to collaborate
across professional boundaries—on team effectiveness. The authors also pro-
posed that negative affective tone plays a moderating role in the relationship
between openness to diversity—members’ aggregate perceptions regarding
the value of recognizing alternative viewpoints and support for diversity—
and team effectiveness. The authors measured negative affective tone by
asking members to rate the extent to which specific negative emotions such
as irritation, distress, and hostility were evident in their team setting. In
other words, they asked members to report on the group’s overall experience,
rather than only their own, using a reference-shift model. The authors then
aggregated the scores to an average value to create a group-level measure of
negative affective tone.
This difference in measurement approaches is quite important, particularly
as we peer through the lens of emergence. Direct-consensus models of
measurement rely on the personal subjectivity of experience to characterize
the group. In other words, because a member has opinions, attitudes, and
agency, they are able to reflect on their own individual experience in the
group and provide an evaluation of that experience. However, when using
referent-shift consensus models, researchers ask members to respond by
reflecting on the experience of the group as a whole, requiring that members
be aware not only of their own experience but also that of other members,
and explicitly incorporate a degree of inter-subjectivity in the measurement.
In essence, the difference between direct and reference-shift models is based
on whether the experience in the group is fully shared, and whether all
 Conceptualizing Emergent States † 21

members have access to the experiences of others. A reference-shift model

requires that members be aware of others’ experiences as they form their
own evaluation of the group.

Ostensive as influence. Capturing members’ experiences can be seen as a

starting point in exploring the ostensive nature of emergent states. A second
important dimension in the exploration of ostensiveness in emergent states
focuses on their potential impact. Although some of the work we have reviewed
has examined the relationship between emergent states and group outcomes,
this is not the only impact of emergent states in social systems such as
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

groups and teams. Instead, the impact of an emergent state can be seen most
directly as it affects the behavior of group members who react, consciously
or unconsciously, to their experience of that emergent state—as it influences
behavior, we note its ostensive character. But in reacting to an emergent
state, the behaviors members enact can also influence the path or trajectory
of that state. In other words, an emergent state is made dynamic by the influ-
ence it has on the behaviors of group members that, in turn, influence the
emergent state. In our review, we identified articles that provide examples of
this dynamic influence.
Earlier we described how Santos and Passos (2013) examine the interplay of
team mental models and conflict. The authors contend, for instance, that
members with similar team mental models are more likely to avoid disagree-
ments regarding personal and social issues—that is, the types of disagreements
that can engender or increase relationship conflict. This argument links an
emergent state (i.e. team mental model similarity) to a range of behaviors
that members engage in or avoid. In this manner, the ostensive nature of the
emergent state as experienced was captured, but so was the ostensive nature
as influencing human behavior. Santos and Passos (2013) build on the initial
link between the similarity of team mental models and conflict to develop a
recursive model, arguing for a link between relationship conflict and the sub-
sequent similarity of team mental models. They suggest that in groups with low
levels of conflict and tension, members may be more willing to spend time and
energy sharing information and discussing ideas which, in turn, help them
develop more similar team mental models. In developing an argument that
accounts for mutual influence, Santos and Passos (2013) work illustrates
how the ostensive nature of an emergent state includes an influence on
human behavior but also how human behavior influences the emergent state.
Tasa, Taggar, and Seijts (2007) present a second example showing how
emergent states can affect behavior and how that behavior, in turn, can
affect the emergent state. These authors explain that initial levels of collective
efficacy influence members’ subsequent individual-level teamwork behavior,
arguing that members are more likely to engage in activities that contribute
to the team’s task and goal accomplishment when they feel they are working
 22 † The Academy of Management Annals

in efficacious groups. The authors went further, suggesting that the perception
of belonging to an efficacious group came from members’ observing others as
they exhibited teamwork behavior. In other words, it was the behavior in the
group by members, observed by others, that caused collective efficacy to
emerge. The work of Tasa et al. (2007) as well as that of Santos and Passos
(2013) illustrates how members’ behavioral reactions to an emergent state,
whether in a conscious or unconscious manner, can potentially change that
emergent state for future interactions. Similar to many other articles we
reviewed, these authors are careful in their theorizing to avoid attributing
intentionality to members reacting to emergent states, instead opting for argu-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

ments that allow for either deliberate or unconscious reactions.

Alexander and Van Knippenberg (2014) take a stronger stance in incorpor-
ating the intentionality of members in their theorizing about emergent states.
They develop a conceptual model and propose that viewing the concept of
team goal orientation as a state, rather than a trait, enables us to better
explain radical innovation success in teams. As part of their argument, they
suggest that team goal orientation can be shifted by members in response to
changing situational demands, particularly those associated with radical inno-
vation. In this process, they argue, any member can act as a catalyst for making
such shifts, although leaders may be in a better position to engender substantial
shifts in team goal orientation. Such a view explicitly shows that an emergent
state such as team goal orientation can be experienced by members and can
also be influenced through their behavior, making it ostensive. Alexander
and Van Knippenberg (2014) provide an overt form of ostensiveness,
however, in which the emergent state is not only experienced but also recogniz-
able and subject to explicit evaluation and reflection by members. Moreover,
this form of ostensiveness is also seen in the articulation of how members
can explicitly work to change an emergent state through their behavior. This
work highlights the role of human agency—of members in general and a
leader in particular—in acting to deliberately change an emergent state in
the group, perhaps best reflecting why emergence is different in social
systems than in physical ones.

4. Radically Novel
Emergent phenomena are said to be radically novel because they do not exist in
their underlying levels, cannot be perfectly reduced back into a collection of
component parts, and cannot be perfectly predicted from these parts (Coen
& Schnackenberg, 2012). In other words, it is the inability to perfectly establish
the relationship between lower and higher level that makes emergent phenom-
ena radically novel. Take, as an example, team-level trust. In a given team, the
degree of trust among members cannot be dissected to determine the exact
interactions and sequences leading to it; these are perceptions and feelings
 Conceptualizing Emergent States † 23

catalyzed by behavior over time that result in one unique—and temporary—

manifestation of an emergent state. Two elements of radical novelty are par-
ticularly relevant in our literature review: the holistic nature and irreducibility
of emergent states.

Radically novel as holistic. Radical novelty in emergent states is holistic

because states are not simply static aggregations of lower-level elements such
as individual-level cognition, affect, or motivation. Rather, emergent states
originate from the interactions among team members as they work together
to achieve their collective goals (Marks et al., 2001). Crucially, however,
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

these interactions are also products of the broad context—that is, they are
influenced by elements of culture, task, or purpose (see Rousseau, 2011). In
our review, we did not discover work adopting a fully holistic perspective in
the treatment of emergent states, but we did identify work depicting the com-
plexity and difficulty of seeking a holistic perspective. In particular, we noted
the use of “blended” constructs (Mathieu, Maynard, Rapp, & Gilson, 2008)
that integrate emergent states and team processes.
Curşeu, Kenis, Raab, and Brandes (2010), for instance, compare the impact
of two team formation strategies—one based on members’ prior short-term
relations and another on member attributes—on team effectiveness, including
teamwork quality. In this case, teamwork quality is an umbrella term (Hoegl &
Gemuenden, 2001) characterizing the quality of collaborative work among
members. Teamwork quality is conceptualized as a multifaceted construct
comprised of several processes (i.e. communication, coordination, and plan-
ning) as well as emergent states (i.e. cohesion and team potency). Curşeu
et al. (2010) measure each facet of teamwork quality separately and then
combine them to create a higher-order variable after showing that they are
highly correlated. Their results show that teams that are formed based on
short-term relations demonstrate better teamwork quality.
Another example of a blended or hybrid construct is Lewis (2003) concep-
tualization of transactive memory systems (TMSs). In contrast to other con-
ceptualizations, Lewis (2003) defined TMS as the active use of transactive
memory and operationalized it as the integration of three interrelated
factors: members’ differentiated knowledge and expertise (i.e. specialization),
members’ beliefs about the reliability of other members’ knowledge (i.e. credi-
bility), and orchestrated processes used to combine this knowledge (i.e. coordi-
nation). In the parlance of Marks et al. (2001), specialization and credibility can
be considered emergent states while coordination reflects a team process. Lewis
showed, through the development of a measurement scale, a series of studies,
and confirmatory factor analysis, that TMS can be usefully treated as having a
three-dimensional structure underlying it.
In these two examples, hybrid constructs are explicitly conceptualized as a
blend of emergent states and team processes. However, in our review of
 24 † The Academy of Management Annals

literature we also identified a number of group-level constructs that are not

explicitly categorized as hybrid but have measurement scales that contain
elements involving both emergent states and processes. For example, Jehn’s
(1995) well-established scale for intragroup conflict was not intended to fully
separate cognitive and affective elements (i.e. emergent states) from processes
or behaviors, instead treating conflict as a whole phenomenon that incorpor-
ates both. As a consequence, conflict can be construed here as a hybrid con-
struct, akin to Curşeu et al. (2010) teamwork quality and Lewis (2003) TMSs.
As these hybrid constructs help capture the holistic nature of emergence,
their impact has broad reach. In other work we have discussed (i.e. Santos &
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

Passos, 2013), conflict appears as a central construct in explanations for the

development of emergent states (e.g. shared mental models). In those
studies, the hybrid nature of the measurement of conflict may remain unac-
knowledged, but it still exemplifies the manner in which a holistic perspective
on novelty is partially present in the field.
The use of hybrid constructs should not be considered unproblematic, though,
as these situations highlight how existing empirical approaches and the adoption
of an emergence perspective to study groups and teams can seem to stand at odds.
The core issue is that while emergent states and processes can be treated as con-
ceptually distinct, it may be difficult (and perhaps counterproductive) to separate
them as we try to capture the real world complexity of the phenomena we study,
as a holistic perspective on radical novelty emphasizes for us. But this quandary,
after all, is somewhat artificial—it merely highlights the tension that exists
between (1) the marked preference in our field for disassembling phenomena
into constituent parts for development of greater conceptual elegance and (2)
the goal of capturing the full complexity and richness of the phenomena.
Hybrid constructs appear to help bridge this tension by simultaneously describ-
ing the holistic reality of groups while providing conceptual clarity.

Radically novel as irreducible. A second principle of radical novelty in

emergence is irreducibility, or the notion that an emergent phenomenon
cannot be perfectly disassembled into its lower-level constituents. Perhaps
the most logical way to view irreducibility is that it is a natural consequence
of the holistic nature of radical novelty. As an emergent state is holistic, it oper-
ates in part as a temporary output of the interactions that occur at the lower
level. Importantly, though, knowing that an emergent state is in place is not
sufficient to describe the interactions that led to its emergence.
An interesting example of irreducibility is presented by Klep, Wisse, and Van
der Flier (2011), who distinguish between dynamic and static pathways in the
development of group mood. The dynamic path is characterized by interactive
affective sharing processes such as contagion, affective comparison processes,
and open communication about emotional experiences. In contrast, the static
path reflects how mood can stem from members’ similarity in personality or
 Conceptualizing Emergent States † 25

exposure to similar emotional events. The authors examined the effect of group
mood on three outcomes (i.e. belongingness, information sharing, and group
performance) and hypothesized that the strength of each effect would depend
on whether mood is the result of interactive affective processes or similarities
in team members’ affective experiences. Through an experimental study they
found that the dynamic path to group mood has a stronger influence on out-
comes than the static path. Nevertheless, their analysis did not detect any differ-
ence in the level of positive or negative mood across mood development
conditions (i.e. interactive versus non-interactive). Their results nicely highlight
how knowledge of group mood, in and of itself, is not sufficient to identify the
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

path through which it has developed (i.e. dynamic versus static) or the factors
that contributed to its development (e.g. interactive affective processes versus
exposure to common emotional experiences). As a result, in order to accurately
predict the influence of group mood on team outcomes, researchers need to not
only collect data on the current level of mood within the team, but also the poss-
ible paths for its development.
A second and more common way in which irreducibility is represented in the
literature can be traced to measurement and aggregation models typically used.
For instance, when emergent states are characterized using dispersion models of
aggregation, and particularly those models leading to a single-value measure,
they reflect irreducibility. Studies of mental model similarity make this most
evident. To establish similarity, mental model researchers elicit each team
member’s knowledge, for instance, of the relationship between task-related or
team-related concepts (Mohammed et al., 2010). This information can then
be used to construct the configuration of each member’s mental model using
representation techniques such as Pathfinder (Schvaneveldt, 1990), multidi-
mensional scaling (MDS) (Carroll & Arabie, 1980; Schiffman, Reynolds, &
Young, 1981), or card sorting (e.g. Hirschman & Wallendorf, 1982). Finally,
the degree of overlap among members’ mental models can be aggregated into
an index, such as the index of mental model similarity (Rentsch & Hall,
1994), which researchers report as a measure of the team-level mental model
(e.g. Lorinkova, Pearsall, & Sims, 2013; Mathieu et al., 2005; Resick et al.,
2014). In these cases, the configurational information that is initially derived
is reduced to a single measure, and this index represents the group-level emer-
gent state. As such, mental model research reflects a degree of irreducibility for
mental models; knowledge of a single-value index of mental model similarity in a
team cannot reveal complete information about the content and structure of
each member’s knowledge configuration.

Summary of Literature Review

We began this literature review in response to calls for more dynamism in the
study of groups and teams (e.g. Cronin et al., 2011; McGrath et al., 2000). By
 26 † The Academy of Management Annals

using the four characteristics of emergent phenomena—global, coherent,

ostensive, and radically novel—as different filters through which to view the
existing literature on team emergent states, our review suggests that in some
ways, dynamism in this literature is alive and well and represented by much
more than the wisps of dynamism we had hoped to find. In other ways,
however, the representation of dynamism in the research we reviewed needs
development in terms of conceptualization, specification and measurement.
Figure 2 summarizes both the overall representation of these four character-
istics in the literature we reviewed as well as the key advantages and issues con-
cerning dynamism across them.
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

We sought to use the four characteristics to describe the topography of

dynamism across this research landscape, rather than to indict the literature
as a whole for a lack of dynamism. Our job now, as we describe below, is to
use this description as the basis for identifying specific areas for further
dynamic development.

Discussion
In an attempt to advance the conversation of dynamism in groups in a prag-
matic manner, sans ideology, we began our work with the explicit assumption
that some elements of dynamic thought already exist in the literature and have
the potential to provide a lever for advancing a dynamic perspective. We

Figure 2 Representation of Emergent Phenomena Characteristics in Existing Literature on

Emergent States in Teams.
 Conceptualizing Emergent States † 27

focused our review and analysis to work using emergent states due to its preva-
lence in numerous content areas within the groups and teams literature, by our
desire to be precise and specific in our characterizations of dynamism, and by
the potential role of emergent states as a solid and foundational building block
for dynamism in groups (Cronin et al., 2011).
In this discussion section we have two goals, both of which flow from taking
a critical stance to the current state of the literature and which are in keeping
with the original promise in the introduction of this article. One goal is to
quickly summarize the presence of the four elements of emergence—global,
coherent, ostensive, and novel—in the literature on groups and teams. The
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

second goal is to challenge the current work with an eye to advancing the
work of incorporating dynamism into the study of groups more broadly.

Current Strengths
We begin by considering how the four fundamental characteristics of emergent
phenomena are represented in the literature. First, the work we reviewed seems
to consistently reflect the overarching global characteristic of emergent
phenomena. Researchers interested in emergent states readily embrace the
notion that they exist on a different level than the interactions or components
that precede their appearance. This is not an insignificant situation; other fields
struggle with such multi-level conceptualizations—but our field has moved on
to embrace this fundamental assumption. Although we also observed some
challenges, such as the assumption of homogeneity for some aggregation
methods (see also Coultas et al., 2014) and a “glossing over” of micro-level
dynamics and interaction details (as detailed by Humphrey & Aime, 2014),
most work and assumptions in this area are consistent with the global charac-
teristic of emergent states.
Second, regarding coherence, the work we reviewed tends to assume that
emergent states are always present (e.g. O’Neill & Allen, 2012), that they
endure (e.g. Curşeu & Schruijer, 2010; Mathieu et al., 2005), and that their
change can be readily measured (e.g. Okhuysen & Richardson, 2007;
Zellmer-Bruhn et al., 2008). Such a view allows for a more free exploration
of the phenomenon, undoubtedly. However, it also reflects some potentially
problematic challenges. For instance, treating emergent states as always
present does not readily allow examination of their origin and the conditions
that engender some but might preclude others. Similarly, treating emergent
states as always present obscures the fact that some may disappear (or sub-
merge) when conditions in a team change. Lastly, and related to the first
two, a notable absence in the study of emergent states is explicit consideration
of temporal dynamics. For instance, there is little discussion of the duration in
time it may take for an emergent state to appear or to change. The absence of
timescales (Zaheer, Albert, & Zaheer, 1999) for emergent states reflects that, in
 28 † The Academy of Management Annals

this literature, we treat all emergent states as having similar temporal properties
and trajectories.
Third, regarding the ostensive nature of emergence, the work we reviewed
embraces the notion that emergent states are experienced by members (e.g.
Mitchell et al., 2014; Peralta et al., 2015) and, as such, have consequences for
subsequent behavior and outcomes (e.g. Santos & Passos, 2013; Tasa et al.,
2007). In fact, the idea that emergent states are experienced by members is
so pervasive (as is the global nature of emergence) that it often goes unre-
marked upon, and it is instead treated as a fundamental precept. Moreover,
we saw many instances in which researchers carefully specify (or, more accu-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

rately, avoid specifying) whether the behavior and outcomes that result from
emergent states originate through deliberation by members or in an uncon-
scious reactive manner. This leaves us with a curious omission: as a field, we
tend not to consider whether group and team members exercise reflection,
deliberation, or agency in their behavior within groups. However, this point
has considerable theoretical implications as well as pragmatic ones for man-
agers and trainers of teams (e.g. Alexander & Van Knippenberg, 2014).
Finally, regarding novelty, the work we reviewed reflects this characteristic
in two ways, adopting a holistic and irreducible perspective of emergent states.
Compared to the other three properties of emergence we considered, however,
novelty is most weakly treated in the literature on emergent states. In our
review, we noted that the holistic nature of emergent states was only rarely
adopted through the deliberate development of constructs that refused simpli-
fication and the desire to more accurately portray a global emergent phenom-
enon in its conceptualization (e.g. Klep et al., 2011). More often, work on
emergent states is focused on developing straightforward models and clear sep-
arations between the different conceptual elements used to examine the messy
reality of group behavior. As such, there is a tendency to divide and refine con-
structs, to separate them and, in the process, to conceptually fragment what is
an inevitable whole (e.g. Curşeu et al., 2010; Lewis, 2003). In terms of irredu-
cibility, the literature we reviewed seldom treated it explicitly. Although in
some cases the conceptualization of emergent states explicitly notes how
lower-level behaviors and interactions lead to the emergence of a new phenom-
enon (that is then treated as irreducible), most often we saw radical novelty
reflected in measurement and aggregation strategies that, for practical pur-
poses, rendered the emergent state irreducible (e.g. Lorinkova et al., 2013;
Resick et al., 2014).

One Major Challenge: Ourselves

Having noted the strengths of the literature that we examined, we now turn to
consider what we identified as a fundamental obstacle to the advancement of a
dynamic perspective in groups and teams research: our traditions and training.
 Conceptualizing Emergent States † 29

As we advanced in our review, we repeatedly noted that the obstacles to advan-

cing a dynamic perspective often lay in the difference between the traditional
way in which we conduct our science and an emergent perspective. Perhaps the
clearest example of this is the persistent pressure in our field to develop only
delineated and discriminant constructs through our analyses. For instance,
the use of a construct such as conflict (Jehn, 1995) that combines cognitive,
affective, and behavioral elements in its measurement might be seen as defec-
tive if we use a traditional analytical lens, one that requires conceptual pre-
cision and a thorough disassembly for the conceptual representation of a
phenomenon. However, as we noted earlier, the use of such hybrid constructs
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

also allows researchers to usefully capture a phenomenon in a more holistic

manner.
A second example of an obstacle that stems from our current research prac-
tices is the use of linear conceptualizations, models, and methods to study
groups and teams. The use of a linear framing imposes significant obstacles
when the goal is a more dynamic perspective; the two are simply not naturally
consistent. As Cronin et al. (2011) noted, “(f)eedback between and within con-
structs makes the study of emergence substantially more complex than the
typical linear model . . . ” (2011, p. 602). Linear approaches tend to work
best when simplifying assumptions are made and when relationships them-
selves are simple. In fact, McGrath et al. (2000, p. 98) described this as an
approach that captures reality as a series of “chain-like unidirectional cause –
effect relationships” that rarely approaches the complexity involved in
dynamism.
A third example of a conceptual obstacle in adopting a dynamic perspective
rooted in our current research practices comes from our inability to account for
absence. In most groups and teams research, our methods and approaches are
predicated on being able to see something happening. In contrast, however,
dynamic approaches require that we also hone our ability to see what is not
happening, such as emergent states that have not emerged or that have com-
pletely disappeared. If we assume an emergent state to be present by continu-
ally asking informants to report their perceptions of a state’s level rather than
(at least first verifying) the state’s existence, we starve ourselves of knowledge
regarding both the mechanisms of initial state emergence as well as of group
dynamics absent states. This risk, of course, assumes that group interaction
can occur for some period of time without a focal state emerging; if this is
not the case, then we need to more precisely determine what level a state
must attain in order to be minimally operative or what onset effects may
exist; additionally, if we assume that any duration of group interaction,
however brief, will produce some level of all emergent states, the entire
notion of emergent states in groups may be suspect as it slips into becoming
tautological.
 30 † The Academy of Management Annals

A fourth example, and perhaps one of the largest obstacles, is the persistent
desire in our field to develop predictive models of group and team behavior.
Such a desire is well established and often goes unmentioned, although we
see its presence in the causal language of our theories, the threshold values
we use to judge significance, and other elements. The desire for an ability to
predict is not without merit—after all, we are trying to develop knowledge
that people can use in practice. But one of the principles of the philosophy
behind dynamic systems is that prediction—as forward-looking expla-
nation—is difficult or impossible. Instead, these approaches are much more
amenable to explanation—that is, backward-looking explanation. Moreover,
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

in combination with some of our other limitations, the nature of emergence,

and the reality of the complexity of groups, prediction may well be nigh
impossible.
We present these examples merely to highlight the broad challenge that
faces our field as we seek to develop more dynamic models of groups and
teams; much of our current research practice can itself be an obstacle to the
adoption of a more dynamic perspective. The question naturally arises, then,
about how to achieve such advancement. A simplistic suggestion would be
to abandon—wholesale—previous perspectives, methods, and conceptual
approaches. Not only is such a suggestion unrealistic but, in addition, it
would neglect an opportunity to usefully integrate existing knowledge into a
more positive and well-developed dynamic perspective; as indicated by our lit-
erature review, much work in the field holds many lessons we can use and
leverage to advance a dynamic perspective. Moreover, as we noted in the intro-
duction, our intent has been to offer a pragmatic path forward for organization
scholars who seek to understand dynamism and improve team outcomes. In
this spirit, we now move to a small set of suggestions that—while recognizing
the obstacle that our own training and traditions impose—can allow for some
meaningful steps forward.

A Path Forward
In this final section of our review, we move to consider four different areas for
additional work by researchers that can meaningfully advance a dynamic per-
spective for research on groups and teams. We avoid suggestions (such as more
extensive use of longitudinal designs) that have been well treated by others.
These four areas are best understood as examples rather than as a definitive
list. In our judgment, they appear to be fruitful areas but, importantly, other
researchers may find other areas that are more relevant or interesting to
pursue. In each case, we offer a connection between our suggestion, current
approaches, and potentially important future developments. Consistent with
the conclusions we have drawn from our literature review and the obstacles
we have identified, these suggestions generally advocate mixed approaches
 Conceptualizing Emergent States † 31

that seek to balance the habits and traditions of current practice with a
dynamic view.
Our first suggestion is a difficult one: the reconsideration of some founda-
tional assumptions in the manner in which we conduct our research. One
example that forces such a discussion is the conceptual and empirical develop-
ment of holistic, hybrid models. Advancing a dynamic perspective can benefit
from identifying aspects of group functioning that are best characterized
using hybrid constructs that involve cognitive, affective, and behavioral
elements such as conflict (Jehn, 1995) and TMSs (Lewis, 2003). However,
such advancement requires that the field contend with identifying the appro-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

priate boundaries for dismantling and fragmenting real-life phenomena into

conceptually tractable (and appealing) constructs while preserving the
complex and holistic understanding of the phenomenon. Creating and main-
taining this balance between figure and ground, between gestalt and reduction,
is complicated, but our field is not alone in facing a move toward a middle way.
For example, after years of struggling to explain the dynamic strategic value of
organizations’ information technology (IT) capabilities, many IT researchers
are finding greater utility in combining the relative strengths of reductionist
and holistic explanations (Fink, 2010). Similarly, as the goal of many areas
of environmental science has shifted from increasing farm yield with reductio-
nistic approaches (traditionally favored by scientists and journal editors) to
producing yield sustainably in order to solve larger problems, many argue
for combining holistic models with reductionistic science (Jordan, 2013).
Such a combined approach might evolve for the study of groups and teams
by integrating (1) some selection of phenomenon-driven computational (or
agent-based) models that provide a view of emergence mechanisms and out-
comes (Coen, 2006), (2) system dynamics models that detail flows and recur-
sion (Cronin et al., 2011) and (3) limited, targeted, reductionistic models that
use lower-level inputs to help predict aspects of global outcomes.
Second, an area in which significant advancement has already occurred is in
the consideration of different forms of aggregation in groups, signaling its
recognized importance in the field. Such work, however, has typically been
applied using a hypothetic-deductive lens that focuses on prediction in our
science. This has meant that we are often left in a place to predict an aggrega-
tion pattern and then test to make sure that it is reliably present in all of the
groups in a sample before we proceed with some type of overall model analysis.
A useful area of inquiry could develop in establishing methods for aggregation
that are agnostic from priors—that is, methods that do not require uniform pre-
dictions across groups to account for the different configurational patterns for
groups in a sample but can still be used as inputs into broader modeling
approaches. The ability to propose models that do not pre-specify (but
rather detect) particular aggregation configurations would allow researchers
to theorize a broader range of emergence.
 32 † The Academy of Management Annals

Third, continuing with another suggestion to improve on our field’s usage

of linear models, we also see a need for work that expands consideration of out-
comes in groups. Current work, through its reliance on linear explanation, often
causes researchers to impose a “one-size-fits-all” prediction for outcomes in
groups. This is not surprising, as it is consistent with nomothetic approaches
that seek to describe populations (and their distributions) rather than specific
entities within those populations (that is, idiographic approaches). However,
dynamic approaches require that we adopt practices that allow for similar
starting situations but different outcomes—hybrid approaches that can identify
multiplex relationships between antecedents and consequences. For instance,
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

given the different configurations of mental models that are possible in a

group it would be useful to consider the different possible outcomes they
can lead to and the conditions under which each one of them is likely to
appear. A second path that could be useful, as a starting point, would allow
researchers to specify ranges of outcomes. A more complicated form of this
suggestion would seek to develop models in which alternative outcome
arrangements can be linked to similar antecedents.
The final suggestion we make is perhaps the one that most clearly separates
emergence in physical systems from emergence in social systems. Because the
lower-level components of a dynamic group are people and their interactions,
our field needs to find ways to account for those people and interactions in a
manner that represents them better. Our review shows how groups researchers
studiously eschew ascribing intention in descriptions involving emergent states
and behavior, although we often recognize that group members may be acting
in ways that advance their values, goals, and wishes. In other words, we often
model persons as group members who react and “go along” in their inter-
actions. However, we need to be able to account for members as persons who
have pre-thought ideas and preferences, as persons whose previous experiences
are sedimented in their minds and influence them as they make sense of new
situations in the group. Such approaches are indispensable as we seek to under-
stand why and how members contribute to the lower-level interactions in
groups. In a related vein, we need conceptual and empirical development of
work that explicitly considers the agency and volition of individuals as
members of groups. Our field needs to find ways of conceptualizing and mod-
eling people who are reflective, who learn, who intend, and thus who may act
deliberately and dynamically to change their groups and emergent states.

Limitations
This review is accompanied by limitations in at least three areas. The first area
concerns the scope of our review. Our goal was to piece together a coherent
picture of dynamism in the study of groups and teams as currently reflected
in the literature on emergent states; however, although we considered the
 Conceptualizing Emergent States † 33

focus on emergent states as a way to maximize the probability of locating evi-

dence pertinent to our goal, we limited our initial set of articles to those citing
Marks et al. (2001), and further narrowed our scope to include only those
articles that dealt implicitly or explicitly with emergent states. Clearly, pub-
lished work exists in our field that contains conceptualizations of dynamism
without entertaining notions of emergent states. Thus, we present our work
as current thinking about dynamism in the groups and teams literature,
based on a logical-but-constrained sample.
Additionally, we oriented our work here by referring to the general work on
emergence, drawing from published articles, chapters, and books in the philos-
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

ophy of science and complexity literatures. The existing literature on emer-

gence is voluminous and, in many areas, quite contentious and rife with
ongoing debate regarding even the most fundamental of precepts. We con-
ducted a targeted reading of this work, consciously seeking those areas with
concepts that seemed most pertinent to our goal here, but other authors
might have made different choices.
Finally, although we took care to limit our preconceptions, even the most
carefully constructed literature review is subject to author biases (Rousseau,
2011). Our work is not a meta-analysis, although meta-analyses are also
limited to those variables previously considered important for study, and are
limited in the number of variables considered.

Conclusion
“The whole is more than the sum of the parts” is a phrase often used to
describe the existence of emergent phenomena; some have used the same
phrase to encapsulate the area of group dynamics in general. One could,
in fact, argue that any construct which theoretically exists at the group
level represents an emergent phenomenon in some way. Emergence is inex-
tricably tied to group and team dynamics, and it is with emergence that we
began our search for dynamism in the literature on groups and teams. We
found and organized many examples of emergence as portrayed in this litera-
ture and suggested ways in which conceptualizations of dynamics in groups
could be extended.
Kuhn (1970) maintained that less mature fields of study begin their devel-
opmental trajectory by coalescing around shared definitions of key constructs.
Any casual perusal of the group dynamics textbooks available today would
suggest that such clarity is present in this literature (allowing for some variation
across authors and studies), and is represented in staple topics such as cohe-
sion, conflict, composition, norms, mood, identity, biases, creativity, bound-
aries, socialization, and many more. A natural next step in the field’s
developmental trajectory would be to specify the interactions across constructs
and over time—in other words, to explicate the dynamics of groups, which, we
 34 † The Academy of Management Annals

hope to have argued well here, must rest on a more refined and careful explica-
tion of emergent phenomena, including emergent states.
However, as identified by many others (e.g. Cronin et al., 2011; McGrath
et al., 2000), this progression of the field has been thwarted by forces
ranging from lack of methodological tools to university performance metrics
that focus chiefly on numbers of publications for professional advancement
and lead to more expedient and static approaches to research on groups. We
agree, but also suggest that the enduring curiosity of those drawn to study
the complexity and beauty of group intricacies peeks through the shadows
cast by these forces, and that this curiosity has left a trail of dynamic evidence,
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

particularly in the work concerning emergence in groups. What we have

attempted to do here is simply follow that trail, piecing together the evidence
and building the outlines of an emerging mosaic, and hoping to suggest a path
forward.

Note
1. Throughout this paper, and reflecting much of the existing literature, we use the
terms “group” and “team” interchangeably.

References
Ablowitz, R. (1939). The theory of emergence. Philosophy of Science, 6(1), 1 –16.
Alexander, L., & Van Knippenberg, D. (2014). Teams in pursuit of radical innovation: A
goal orientation perspective. Academy of Management Review, 39(4), 423– 438.
Andriani, P. & McKelvey, B. (2009). Perspective—From Gaussian to Paretian thinking:
Causes and implications of power laws in organizations. Organization Science, 20,
1053– 1071.
Arrow, H., McGrath, J. E., & Berdahl, J. L. (2000). Small groups as complex systems.
Thousand Oaks, CA: Sage Publications.
Bedau, M. A. (2008). Downward causation and autonomy in weak emergence. In M.
Bedau & P. Humphreys (Eds.), Emergence: Contemporary readings in philosophy
and science (pp. 155– 188). Cambridge, MA: MIT Press.
Bedau, M. A., & Humphreys, P. (2008). Emergence: Contemporary readings in philos-
ophy and science. Cambridge, MA: MIT Press.
Bresman, H., & Zellmer-Bruhn, M. (2013). The structural context of team learning:
Effects of organizational and team structure on internal and external learning.
Organization Science, 24(4), 1120– 1139.
Carroll, J. D., & Arabie, P. (1980). Multidimensional scaling. Annual Review of
Psychology, 31, 607 – 649.
Carson, J. B., Tesluk, P. E., & Marrone, J. A. (2007). Shared leadership in teams: An
investigation of antecedent conditions and performance. Academy of
Management Journal, 50(5), 1217– 1234.
 Conceptualizing Emergent States † 35

Chan, D. (1998a). Functional relations among constructs in the same content domain at
different levels of analysis: A typology of composition models. Journal of Applied
Psychology, 83(2), 234 – 246.
Chan, D. (1998b). The conceptualization and analysis of change over time: An integra-
tive approach incorporating longitudinal mean and covariance structures analysis
(LMACS) and multiple indicator latent growth modeling (MLGM).
Organizational Research Methods, 1, 421 – 483.
Chen, G., Kanfer, R., DeShon, R. P., Mathieu, J. E., & Kozlowski, S. W. (2009). The
motivating potential of teams: Test and extension of cross-level model of motiv-
ation in teams. Organizational Behavior and Human Decision Processes, 110(1),
45 – 55.
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

Coen, C. A. (2006). Seeking the comparative advantage: The dynamics of individual

cooperation in single vs. multiple team environments. Organizational Behavior
and Human Decision Processes, 100, 145 – 159.
Coen, C. A., & Schnackenberg, A. (2012). Complex system methods for studying multi-
team systems. In S. Zaccaro, M. Marks, & L. DeChurch (Eds.), Multiteam systems:
An organizational form for dynamic and complex environments (pp. 453– 479).
New York: Routledge.
Corning, P. (2002). The emergence of “emergence”: Now what? Emergence, 4(3), 54 – 71.
Coultas, C. W., Driskell, T., Burke, C. S., & Salas, E. (2014). A conceptual review of
emergent state measurement current problems, future solutions. Small Group
Research, 45(6), 671 – 703.
Cronin, M. A., Weingart, L. R., & Todorova, G. (2011). Dynamics in groups: Are we
there yet? The Academy of Management Annals, 5(1), 571– 612.
Curşeu, P. L., Kenis, P., Raab, J., & Brandes, U. (2010). Composing effective teams
through team dating. Organization Studies, 31(7), 873– 894.
Curşeu, P. L., & Schruijer, S. G. (2010). Does conflict shatter trust or does trust obliterate
conflict? Revisiting the relationships between team diversity, conflict, and trust.
Group Dynamics: Theory, Research, and Practice, 14(1), 66 – 79.
De Jong, B. A., & Elfring, T. (2010). How does trust affect the performance of ongoing
teams? The mediating role of reflexivity, monitoring, and effort. Academy of
Management Journal, 53(3), 535 – 549.
D’Innocenzo, L., Mathieu, J. E., & Kukenberger, M. R. (2014). A meta-analysis of differ-
ent forms of shared leadership– team performance relations. Journal of
Management, 20(10), 1 – 28.
Edmondson, A. (1999). Psychological safety and learning behavior in work teams.
Administrative Science Quarterly, 44(2), 350 – 383.
Emmeche, C., Køppe, S., & Stjernfelt, F. (1997). Explaining emergence: Towards ontol-
ogy of levels. Journal for General Philosophy of Science, 28(1), 83– 117.
Faraj, S., & Yan, A. (2009). Boundary work in knowledge teams. Journal of Applied
Psychology, 94(3), 604 – 617.
Fink, L. (2010). How do IT capabilities create strategic value? Toward greater inte-
gration of insights from reductionistic and holistic approaches. European
Journal of Information Systems, 20, 16 – 33.
Gioia, D. A., Patvardhan, S. D., Hamilton, A. L., & Corley, K. G. (2013). Organizational
identity formation and change. The Academy of Management Annals, 7(1), 123–
193.
 36 † The Academy of Management Annals

Goldspink, C., & Kay, R. (2010). Emergence in organizations: The reflexive turn.
Emergence: Complexity and Organization, 12(3), 47 – 63.
Goldstein, J. (1999). Emergence as a construct: History and issues. Emergence, 1, 49 – 72.
Goldstein, J. (2013). Re-imagining emergence: Part 1. Emergence: Complexity &
Organization, 15(2), 78 – 104.
Harper, D. A., & Endres, A. M. (2012). The anatomy of emergence, with a focus upon
capital formation. Journal of Economic Behavior and Organization, 82, 352– 367.
Hirschman, E. C., & Wallendorf, M. R. (1982). Free response and card-sort techniques
for assessing cognitive content: Two studies concerning their stability, validity,
and utility. Perceptual and Motor Skills, 54, 1095– 1110.
Hodgson, G. M. (2000). The concept of emergence in social science: Its history and
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

importance. Emergence, 2(4), 65– 77.

Hoegl, M., & Gemuenden, H. G. (2001). Teamwork quality and the success of innova-
tive projects: A theoretical concept and empirical evidence. Organization Science,
12(4), 435 – 449.
Huang, D. T., Clermont, G., Sexton, J. B., Karlo, C. A., Miller, R. G., Weissfeld, L. A.,
Rowan, K.M., & Angus, D. C. (2007). Perceptions of safety culture vary across
the intensive care units of a single institution. Critical Care Medicine, 35, 165–
176.
Humphrey, S. E., & Aime, F. (2014). Team microdynamics: Toward an organizing
approach to teamwork. The Academy of Management Annals, 8(1), 443–503.
Jehn, K. A. (1995). A multimethod examination of the benefits and detriments of
intragroup conflict. Administrative Science Quarterly, 40(2), 256– 282.
Jehn, K. A., & Mannix, E. A. (2001). The dynamic nature of conflict: A longitudinal
study of intragroup conflict and group performance. Academy of Management
Journal, 44(2), 238 –251.
Jehn, K. A., Rispens, S., & Thatcher, S. M. (2010). The effects of conflict asymmetry on
work group and individual outcomes. Academy of Management Journal, 53(3),
596 – 616.
Jordan, C. F. (2013). An ecosystem approach to sustainable agriculture. Dordrecht:
Springer Netherlands.
Kirkman, B. L., Rosen, B., Tesluk, P. E., & Gibson, C. B. (2004). The impact of team
empowerment on virtual team performance: The moderating role of face-to-
face interaction. Academy of Management Journal, 47(2), 175 – 192.
Klep, A., Wisse, B., & Van der Flier, H. (2011). Interactive affective sharing versus non-
interactive affective sharing in work groups: Comparative effects of group affect
on work group performance and dynamics. European Journal of Social
Psychology, 41(3), 312 – 323.
Kozlowski, S. W., & Chao, G. T. (2012). The dynamics of emergence: Cognition and
cohesion in work teams. Managerial and Decision Economics, 33(5– 6), 335– 354.
Kozlowski, S. W., Chao, G. T., Grand, J. A., Braun, M. T., & Kuljanin, G. (2013).
Advancing multilevel research design capturing the dynamics of emergence.
Organizational Research Methods, 16(4), 581 – 615.
Kozlowski, S. W. J., & Klein, K. J. (2000). From micro to meso: Critical steps in concep-
tualizing and conducting multilevel research. Organizational Research Methods, 3,
211 – 236.
Kuhn, T. S. (1970). The structure of scientific revolutions (2nd ed.). Chicago, IL:
University of Chicago.
 Conceptualizing Emergent States † 37

Langfred, C. W. (2007). The downside of self-management: A longitudinal study of the

effects of conflict on trust, autonomy, and task interdependence in self-managing
teams. Academy of Management Journal, 50(4), 885– 900.
Lewis, K. (2003). Measuring transactive memory systems in the field: Scale development
and validation. Journal of Applied Psychology, 88(4), 587–604.
Lichtenstein, B. B. (2014). Generative emergence: A new discipline of organizational,
entrepreneurial and social innovation. Oxford: Oxford University Press.
Lissack, M. R., & Letiche, H. (2002). Complexity, emergence, resilience, and coherence:
Gaining perspective on organizations and their study. Emergence, 4, 72 – 94.
Lorinkova, N. M., Pearsall, M. J., & Sims Jr., H. P. (2013). Examining the differential
longitudinal performance of directive versus empowering leadership in teams.
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

Academy of Management Journal, 56(2), 573– 596.

Marks, M. A., Mathieu, J. E., & Zaccaro, S. J. (2001). A temporally based framework and
taxonomy of team processes. Academy of Management Review, 26(3), 356–376.
Mathieu, J. E., Gilson, L. L., & Ruddy, T. M. (2006). Empowerment and team effective-
ness: An empirical test of an integrated model. Journal of Applied Psychology,
91(1), 97 – 108.
Mathieu, J. E., Heffner, T. S., Goodwin, G. F., Cannon-Bowers, J. A., & Salas, E. (2005).
Scaling the quality of teammates’ mental models: Equifinality and normative
comparisons. Journal of Organizational Behavior, 26(1), 37– 56.
Mathieu, J. E., Maynard, M. T., Rapp, T., & Gilson, L. (2008). Team effectiveness 1997–
2007: A review of recent advancements and a glimpse into the future. Journal of
Management, 34(3), 410 – 476.
McGrath, J. E., Arrow, H., & Berdahl, J. L. (2000). The study of groups: Past, present,
and future. Personality and Social Psychology Review, 4(1), 95 – 105.
McLaughlin, B. P. (2008). The rise and fall of British emergentism. In M. Bedau & P.
Humphreys (Eds.), Emergence: Contemporary readings in philosophy and
science (pp. 19 –59). Cambridge, MA: MIT Press.
Mell, J., van Knippenberg, D., & van Ginkel, W. (2014). The catalyst effect: The impact
of transactive memory system structure on team performance. Academy of
Management Journal, 57(4), 1154– 1173.
Mitchell, M. (2009). Complexity: A guided tour. Oxford: Oxford University Press.
Mitchell, R., Boyle, B., Parker, V., Giles, M., Joyce, P., & Chiang, V. (2014).
Transformation through tension: The moderating impact of negative affect on
transformational leadership in teams. Human Relations, 67(9), 1095– 1121.
Mohammed, S., Ferzandi, L., & Hamilton, K. (2010). Metaphor no more: A 15-year
review of the team mental model construct. Journal of Management, 36, 876– 910.
Okhuysen, G. A., & Richardson, H. A. (2007). Group conflict as an emergent state:
Temporal issues in the conceptualization and measurement of disagreement. In
K. J. Behfar & L. L. Thompson (Eds.), Conflict in organizational groups: New
directions in theory and practice (pp. 145 – 179). Evanston, IL: Northwestern
University Press.
O’Neill, T. A., & Allen, N. J. (2012). Team meeting attitudes: Conceptualization and
investigation of a new construct. Small Group Research, 43(2), 186– 210.
Peralta, C. F., Lopes, P. N., Gilson, L. L., Lourenço, P. R., & Pais, L. (2015). Innovation
processes and team effectiveness: The role of goal clarity and commitment, and
team affective tone. Journal of Occupational and Organizational Psychology,
88(1), 80 – 107.
 38 † The Academy of Management Annals

Porter, C. O., Gogus, C. I., & Yu, R. C. F. (2011). Does backing up behavior explain the effi-
cacy–performance relationship in teams? Small Group Research, 42(4), 458–474.
Rentsch, J. R., & Hall, R. J. (1994). Members of great teams think alike: A model of team
effectiveness and schema similarity among team members. In M. M. Beyerlein &
D. A. Johnson (Eds.), Advances in interdisciplinary studies of work teams: Theories
of self-managing work teams (Vol. 1, pp. 223 – 261). Greenwich, CT: Elsevier
Science/JAI Press.
Resick, C. J., Murase, T., Randall, K. R., & DeChurch, L. A. (2014). Information elabor-
ation and team performance: Examining the psychological origins and environ-
mental contingencies. Organizational Behavior and Human Decision Processes,
124(2), 165 – 176.
Downloaded by [Orta Dogu Teknik Universitesi] at 03:15 08 March 2016

Rousseau, D. M. (2011). Reinforcing the micro/macro bridge: Organizational thinking

and pluralistic vehicles. Journal of Management, 37, 429– 442.
Santos, C. M., & Passos, A. M. (2013). Team mental models, relationship conflict and
effectiveness over time. Team Performance Management: An International
Journal, 19(7/8), 363 – 385.
Schiffman, S. S., Reynolds, M. L., & Young, F. W. (1981). Introduction to multidimen-
sional scaling: Theory, methods and applications. New York: Academic Press.
Schvaneveldt, R. W. (1990). Proximities, networks, and schemata. In R. W.
Schvaneveldt (Ed.), Pathfinder associative networks: Studies in knowledge organ-
ization (pp. 135 – 148). Norwood, NJ: Ablex.
Shin, Y., & Choi, J. N. (2010). What makes a group of good citizens? The role of per-
ceived group-level fit and critical psychological states in organizational teams.
Journal of Occupational and Organizational Psychology, 83, 531– 552.
Tasa, K., Taggar, S., & Seijts, G. H. (2007). The development of collective efficacy in
teams: A multilevel and longitudinal perspective. Journal of Applied Psychology,
92, 17 – 27.
Triplett, N. (1898). The dynamogenic factors in pacemaking and competition. American
Journal of Psychology, 9, 507 – 533.
Waller, M. J., Conte, J. M., Gibson, C. B., & Carpenter, M. A. (2001). The effect of indi-
vidual perceptions of deadlines on team performance. Academy of Management
Review, 26, 586– 600.
Watts, E. J. (2010). Riot in Alexandria: Tradition and group dynamics in late antique
pagan and Christian communities. Oakland, CA: University of California Press.
Wei, L. Q., & Wu, L. (2013). What a diverse top management team means: Testing an
integrated model. Journal of Management Studies, 50, 389– 412.
Zaheer, S., Albert, S., & Zaheer, A. (1999). Time scales and organizational theory.
Academy of Management Review, 24, 725 – 741.
Zellmer-Bruhn, M. E., Maloney, M. M., Bhappu, A. D., & Salvador, R. B. (2008). When
and how do differences matter? An exploration of perceived similarity in teams.
Organizational Behavior and Human Decision Processes, 107, 41 –59.