Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 119
SESSION II
Dynamic Social Networks
OCR for page 120
OCR for page 121
Informal Social Roles and the Evolution and Stability of Social Networks
Jeffrey C Johnson East Carolina University
Lawrence A. Palinkas. University of California San Dieao
James S. Boster, University of Connecticut;
Introduction
From a series of cross-cultural studies on the evolution of group or network structure in Antarctic research
stations it is evident that despite similarity in natural environments (e.~., cold. isolation), organizational
goals (e.g., conducting science), formal organizational structure, physical settings, group size, and
duration of isolation across years group dynamics can vary dramatically from one group to another even
within the same physical and cultura] settings (Johnson et al. in press; Johnson et al. 20021. Thus. network
dynamics are largely a function of both formal and informal factors (e.g.- the emergence of informal
social roles) that have variable effects on the patterns of interaction and connection among network actors
and ultimately on such things as performance, productivity, morale, and individual psychological well-
being.
An important distinction needs to be made between formal, informal, and latent social roles in networks.
Formal social roles are those proscribed by groups, organizations, or cultures and are reflected in the
designation of formal positions (e.;,., manager, CEO). Although formal aspects of groups are certainly
important, much of network dynamics are the result of informal influences and the interaction between
formal and informal processes. Although early work in organization studies recognized a link between
latent and informal roles in organizations (Gouldner 1957; Becker and Geer 1960) there is a fundamental
difference between the two. Whereas all latent roles are informal, not all informal roles are latent. In this
sense there are in foal, as well as formal social roles, that are visible and active throughout a Troupes
existence. On the other hand, there are informal social roles that can be dormant or hidden emerging only
when circumstances or conditions warrant (e.g., due to external or internal events). The presence or
absence of informal social roles and the nature of hidden, do ant, or latent roles all have an impact on
network's emergent properties. Emergent properties here refer to the emergence of higher level group
phenomena, such as cohesion or other global structural properties, that stem from the characteristics of
lower level constituent entities, such as the mix of individual actors. The focus of this this paper is the
relationship between these emergent properties and the evolution of network structures as they relate to
such things as network stability, adaptability, and robustness.
In a study of the relationship between informal social roles and emergent properties of networks Johnson
et al. (in press) found that the evolution of globally coherent networks in Antarctic winter-over groups
was associated with group consensus on the presence of critically important informal social roles (e.g.,
expressive leadership) where global coherence is the extent to which a network forms a single group
composed of a unitary core and periphery as opposed to being fragmented into two or more subgroups.
Conversely, the evolution of multiple subgroups was associated with the absence of consensus on
informal socia] roles, above all the critically important role of instrumental leader and the lack of role
redundancy in expressive leadership. Thus the nature, multiplicity. and coincidence of both formal and
informal social roles in networks can have an impact on the evolution and stability of network structures.
More importantly, however, these network structures all have implications for such group outcomes as
productivity, group morale, efficiency. individual psychological well-being, group communication, and
group conflict (Johnson et al. 20021.
DYNAMIC SOCIAL N~TWO=MOD~WG ED CYSTS
121
OCR for page 122
Although there are a wide range of informal role properties and structural correlates we will concentrate
on 5 primary forms. Table 1 summarizes these properties in terms of the implications of their presence or
absence for group functioning in terms of their effect on network structure. Each of these properties will
be discussed in more or less detail in the examples the'; follow. Briefly, role complementarily is important
-for diminishing role competition and conflict. Whereas we may want homogeneity in the back;,rounds of
group members in order to diminish any bases for intra-group divisions (white collar vs. blue collar), role
complementarily or heterogeneity is desirable since it helps ensure that group member's roles fit in with
one another (Johnson et al. in press). Role consensus, or agreement on informal roles, reflects a lack of
role competition and conflict in that there is agreement on the set or sets of actors performing, such roles.
For some of these informal roles, redundancy is important in that removal of a single actor in a set of
multiple roles still ensures proper role function. As later examples will show, lack of replacement may
leave groups structurally vulnerable due to the loss of critically important roles that function to foster
group coherence or cohesion. Similarly. unforeseen internal or external events that threaten groups may
require latent or hidden informal roles that function to maintain structure] integrity during, these events, in
a sense informal roles in reserve being called on in times of need. Finally, the overlap between formal
and informal leadership is important for producing, group consensus on the Troupes instrumental goals and
objectives and for reducing role competition for leadership. A more in-depth discussion of these
properties can be found in Johnson and Finney (1986) and Johnson et al. (in press).
Table 1. Consequences of the presence or absence of network informal role properties.
Informal Role Properties Presence Absence
Role Complementanty People fit in with one another Role Homogeneity: Role
competitionlconflict ("Too many
chiefs and not enough Indians .)
Role Consensus Agreement on individual status. Role collisions leading to group
role, and function divisions
Role Redundancy For certain informal roles can Structural Vulnerability
enhance network adaptability
Role Latency Promotes adaptive responses to Structural Vulnerability
unforeseen events
Formal/Inforn~al Role Promotes agreement on group Role collisions leading to group
Ison~orphism goals and objectives divisions
Informal Roles and Emergent Properties: Examples of the Ties That Bind
In a dynamic world there is no guarantee of group or network stability over time despite the best
intentions of formal organizational efforts (Johnson and Parks 19981. Groups ares in a sense, constantly
under attack from both within and from the outside. Interpersonal tensions. for example, always have the
potential to tear at the very fabric of group structures from within, inhibiting adaptation to changing
circumstances and thereby have the potential to lead to divisive and less cohesive or coherent groups
(Johnson et al. in press). Different forms of deviance can work to either integrate or divide group structure
(Johnson and Miller 1983; Johnson et al. in press; Dentler and Ericksen 19591. Outside events or
organizations can also present threats in terms of such things as withholding resources or outside
meddlin;, in internal group affairs that can also impact a group's ability to adapt to chan;,ing conditions.
Although these outside influences can sometimes serve as group reference points that can foster group
cohesion (i.e., internal group alliance to fight outside threats in an llS versus sheen mentality) such
phenomena tend to be short-lived. and once such threats are diminished tend to exacerbate preexistin,
tensions and conflicts (Johnson and Finney 19861. In either case, the ability of a group to respond to such
challenges depends on number of important factors.
122
DYNAMIC SOCIAL NETWORK MODELING ED ^=YSIS
OCR for page 123
Example I: The Emergence of the Latent Informal Social Role of 'Court Jester'
We now turn to an example of a latent informal social role that contributed to positive group function
during a time of external threat. Johnson and Miller (1983) and Johnson and Finney (1986) described the
function of the informal social role of 'courtjester' for fostering group cohesion during periods of stress
in both fishers in an isolated fish camp in Alaska and in Antarctic expeditions. What makes these
examples important is the presence of informal latent social roles that fostered the emergence of group
coherence during a stressful event.
Every spring fishermen travel to Bristol Bay, Alaska in the pursuit of king, sockeye, and dog salmon,
among other species. The fishing is crowded, competitive, and has historically been very lucrative.
Bristol Bay, an arm of the Bering Sea, is extremely isolated and the purchasing, processing, and
marketing of salmon has traditionally been dominated by a small set of oligopolishc firms. Within recent
times, however. fishers have gained more power due to increasing competition from outside firms and the
organization of fishers under the Alaska Independent Fishermen's Marketing Association. Most fishers
are associated with large firms staying in bunkhouses within canneries or in fish camps during the salmon
season (2-3 months). One such fish camp was owned and operated by a large' powerful firm. The fish
camp was mulii-ethnic but dominated by a large group of Italians who had mostly emigrated from Sicily
to the U.S in the 1960's. The network of interest consists of 16 Italian boat captains. A more detailed
ethnographic background for this example can be found in Johnson and Miller (19831.
In a multidimensional scalin,, of unconstrained judged similarity pile sort data among the 16 captains (not
shown) two distinct groups were clear based mostly on bunkhouse residence with the group from
bunkhouse 2 being highly among residents while those from bunkhouse 1 had moderate density (Johnson
and Miller 19831. There were a number of important exceptions including a fisher who lived in
bunkhouse 1 but who interacted most frequently with fishers in bunkhouse 2 and a fisher who lived in
bunkhouse 2 but socialized in bunkhouse 1. Another important actor is "fisher l" who was described by
captains during the pile sort task as "the bridge between the two groups" and "belonging to everyone".
Fisher 1 is the focal point in this example.
Table 2 shows the time]ine of events leading up to the emergence and disappearance of the role of court
jester over the course of the fishing season. Initially fishers are extremely busy during the period before
the opening of sockeye season preparing boats and gear. During this period there is little free time and
fishers work throughout the long Alaskan days (daylight until after midnight). Just prior to opening, the
fishers begin to negotiate season prices with firms. If prices have not been settled prior to season
opening, the fishers will stake until a satisfactory price can be negotiated. In the year of this study, the
strike was quite bitter and protracted (2 weeks of a 4 week season) creating an extremely stressful
atmosphere be.,,., captains could gross more than $10,000 a day). With the preparations for the season
mostly completed the fishers had little to occupy their time dunn';, the spike, contributing further to
potential conflicts among and between fishers and management.
Table 2, Timeline for the emergence and disappearance of the role of court; jester.
_ _
Preparation Strike Season Court Jester Strike Court Jester Role
for salmon Begins Opens, Role EndslFishing Role Rewarded
season (approx I Strike Still Emerges Begins (JUSt disappears with over
(May-June) week before in Effect prior to the limit fish
season (Tension, season peak)
opening) Boredom)
DYNAMIC SOCIAL NETWORK MODE ED THESIS
123
OCR for page 124
During this strike period fisher 1 emerged in the role of court jester and provided hours of comic relief for
the fishers of both bunkhouses.
Fishers from bunkhouse 1, for example, would come to bunkhouse 2 to
encage in a number of activities one of which included a mock funeral and burial in the tundra for fisherl
(Johnson and Miller 19831. Thus, fisher 1 provided comic relief and brought fishers together in positive
face to face interactions thereby diminishing chances for intra-group conflicts during this particularly
stressful period.
Once fish prices were settled the role of court jester receded as fishers immersed themselves in their
fishing, activities. However, despite the lack of need, for this role the role was still recognized as being
important to the function of the group. Dunng the remainder of the season fisher 1 received over limit
fish transfers from a wide range of fishers as a reward for his role as court jester. The receivers were
primarily from bunkhouse 2 with the exception of fishers (i.e., reflecting the relative status of members of
the respective bunkhouses). Among receivers, fisher 1 had a unique status and role in the group reflected
in his high indegree centrality in the fish transfer network. Although he received fish from a wide range
of fishers he tended to receive small amounts, generally bein;, at the bottom of the list behind those
related by kinship to the giver. This network of Italians was highly interconnected through kinship, and
fisher 1 had no kinship relation to any other members of the group, either captains or crew (Johnson arid
Miller 1983~.
There are a number of structural factors and actor characteristics that facilitated the emergence of role of
court jester with respect to fisher 1:
.
Lower Status actor living in the higher status bunkhouse,
Worst fisher in the camp (deviant in terms of Productivity norms ala Homans ( 197411
No kinship relations to the :,roup,
Willingness to be the brunt of Jokes and pranks
~ 1 ~ _ . . ~ 1 · t · · ~ ~ ~ ~
~ ~ _ _ · ~ ~ ~
~ ~~;~c ~uuu-~ura~ bull ~~u~v~aua~ features allowed pilsner ~ tO play the role or court; jester without the fear of
violating norms, despite his formal status as fishing captain. What was important in this case was his
informal position and role as court jester bridged the gap between the two bunkhouses during a time of
extreme stress, thus significantly contributing to the group's structural integrity. Although there were
certainly other forces, such as extensive kinship relations, that also fostered group cohesion, the role of
court jester encouraged actual face-to-face interactions at a particularly critical time.
Example 2: The Bridging Informal Role of Clown in Antarctic Groups
The next example of a bridging informal role comes from research on the group dynamics of Antarctic
winter-over crews at the Amundsen-Scott South pole Station. Three separate years at; the station were
studied throughout the Austra] winter in which station crews were isolated for approximately 9 months.
Network data, informal role data and morale and psychological data were collected on a monthly basis
throughout the 9 months of the winter. Similar to the example above we concentrate on the role of
clown'. However, in this case the informal role of clown was recognized throughout winter in each of the
three years studied. The role term itself was derived from in-depth interviews with former winter-overs
prior to the beginning of the study (see Johnson and Weller 2002 for a discussion). This term and 10
other informal role terms (e.g., social director, work leader, peacemaker) were a part of a informal role
sentence completion task in which respondents were asked to associate each of his or her fellow winter-
over crewmembers with each of the 1 1 informal role terms.
Figure 1 shows a series of graphs revealing the connections among winter-over crew from year A in the
middle of winter (July) one of the most stressful pen oafs of the winter-over period (Palinkas et al. 1998,
2000a, 2000b, 2001, 2002; Johnson et al. 20021. The series of network graphs reveals the important
bridgin;, function of the clown role during this year (more than 67 % of the crew nominated this actor for
124
DYNAMIC SOCKS NETWO~MODELING AND ISIS
OCR for page 125
OCR for page 127
OCR for page 129
OCR for page 130
OCR for page 131
OCR for page 132
Representative terms from entire chapter:
role properties
the role of 'clown/comedian. This role was particularly important in maintaining connections between
trades crew (i.e., the contractors) and scientists (i.e., primarily NSF ;,rantees). The trades and scientists,
often jokingly referred to by the trades crew as "beakers" (as in laboratory beaker)- a distinction that has
historically been a potential source of intra-group conflicts in Antarctic winter-over groups.
What this series dramatically reveals is that the clown role functioned to bridge relations between the two
major social categories in the station. The crewmember who played this role was central to a wide range
of both trade and science crew. With respect to trade crew he linked different segments among members
of the trades themselves, particularly the two isolates to the lower left in graph II. Similarly, he provided
a link between the major cluster of scientists in the upper right of the graph both with other scientists,
particularly the isolated science personnel, and the primary trades contingent.
~ 0\0
°°o?& $
I. II.
Figure 1. The network structure in the middle
by) of the winter (July) derived from a stacked
T ~ Icorrespondenceanalysisof~eratingsof
~ / ~ ~ social interaction during each of the 9 winter
C ) ~' months. Edges are shown for rij > 4, where r
~ O Am/ ranges between O and 10. The graphs show (I)
C ~ ties among, scientists in the middle of winter
~ \ A\ ~ ~ for year A, (II) ties among trades in the
O /\ ~~ middle of winter with clown role highlighted,
/ (~S~ I and ~m' ~ ies between scientists and trades in
Clown he mic e of win er wi h c own role
~ | highlighted
III.
In addition to the clown noted in Figure 4, there were several other crewmembers who were nominated as
clowns by winter' s end in year A, 2 with high consensus (consensus c>0.66) and 4 with moderate
consensus (0.33
A. Thus, Years A and B displayed a high degree of role redundancy with respect to the informal role of
clown. This stands in stark contrast to year C where this infold ~.~.al role was largely absent.
Example 3: Expressive Leadership, Role Redundancy, and Adaptability
Id this example we examine the consequences of the de;,ree of expressive leadership available to groups.
What separates expressive from instrumer~;al leadership is the arena in which each operates. Whereas
instrumental leadership generally deals with achieving organizational goals and objectives related to
work, expressive leadership, as its name entails, involves leadership and direction in a variety of mostly
non-work realms. Expressive leaders, referred to in the study as social directors, organized events that
were primarily expressive or leisure oriented (e.g., dinner parties, theme costume parties, sports
competitions, dance lessons, etc.~. ~ nese events brought people of venous backgrounds together in face-
to-face interactions that were of a positive nature.
Expressive Leader (October)
Expressive Leader (Apnl)
\ ~ ~
\
OF ~~,~
Figure 2a. Graph showing the position of the Figure 2b. Graph showing the position of the
single expressive leader two months after the start earlier nominated expressive leader at the end
of the Austral winter for year C. Edges are shown of the winter for year C. Edges are shown for
for rjj > 4, where ~ ranges between O and 10. r,; > 4, where r ranges between O and 10.
~:E=ressve Leader |
,~ ~ ~
Figure 2c. The network structure in the middle
\\~j~ \ | of e winter (July) for year A derived from a
f~ ~ stacked correspondence analysis of the ratings
, ~ ~ of social interaction during each of the 9 winter
Am) months. Edges are shown for rii > 4, where r
t/ Am I ranges betwee O and 10. The graph shows the
position of the 3-crew members who were
A\\ highly nominated as expressive leaders.
'\
an,
126
DYNAMIC SOCIAL NETWORK MODELING AND ISIS
Figures 2a & 2b show the position of the single expressive leader for year C in the beginning and end of
the Austral winter. In the beginning of the winter this actor is central to the group bnd=,in~ connections
between science and trades crewmembers. However, by the end of the winter this crewmember had
become less central and the group had fragmented into several subgroups with two primary divisions
being between science and trades crew.
This single expressive leader for this year dropped out of most: group interactions in about the middle of
winter because of problems with harassment, particularly from one crewmember (see discussion on
negative deviant; roles above). Whereas this expressive leader was an organizer of social events and
dinner parties in the earlier months of the winter, by mid-winter this crewmember had dropped out of
most social activities. particularly as the important organizer of events that [Functioned to brie,
leader and there was moderate consensus on three crewmembers including the station manager. Thus, in
the station manager as infold fatal work
l~a"~l, I~llc~lll~ ~ ~ull~lu<;~aule Degree or role consensus. 1ne moderate agreement for three crew for
year C, on the other hand, reflects role competition for instrumental leader that included the stahon
manager. It should be noted that initially year C had high consensus on the station manager as the
exclusive informal instrumental leader. This is important in that the lack of extensive informal role
properties in this year allowed for potential problems in the future stemming from interpersonal conflicts
and external threats of various kinds. Good leadership alone cannot produce stable, productive groups. It
requires a combination of informal roles, all improving the chances for the evolution of cohesive and
productive groups.
years A and B In terms there was high agreement with respect to
1 ~ ~ ~ ~ A ~+ ~ ~ 1 _ ~ ~ ¢ . _ 1 _ _ Ant
N ~ nightie
0 ''my ,i
Hi ~ ~.,-
1 / \ ~
~ W~·.7~
A
B
Bar Group Figure 3. Final (October) group structures for
3 l - Group ~ winter-over crews for the 3 showing the position
of station manager and the delineation of groups
i\ \~ as determined by the winter-overs in in-depth
1~ . ~ $~/ interviews (from ~ correspondence analysis of the
ratings of illteraction). Year A formed a single
/ ~ group (global coherence), while years B and C
had varying degrees of subgroups. Year C had
subgroups deterniined primarily on the basis of a
science versus trades split. The library group
. / consisted mostly of scientists while the bar group
\J consisted of trades crew members (e.g.,
Biomed Grcup carpenters, plumbers, mechanics, etc.) [from
Johnson et al., in press]
C
Informal Role Properties and Group Outcomes
In this section we examine the consequences of the informal role properties discussed above. In an earlier
paper, Johnson et al. (in press) demonstrated that informal role properties of these winter-over groups
were important for understanding the evolution of coherent group structures. Globally coherent structures
were associated with role consensus, role heterogeneity, expressive leadership, functional deviance, and
an overlap between formal arid informal instrumental leadership, while local coherence (presence of
128
DYNAMIC SOCKS N~TWO=MOD~G ED TRYSTS
network fragmentation) was associated with the lack of extensive informal social roles, negative
deviance, and role competition for instrumental leadership. In this case. year A s structure evolved to be
highly coherent over time, year B's structure was moderately coherent, while year C's structure became
locally coherent over time, fragmenting into several subgroups (Figure 31. Aside from these structural
consequences of informal role properties, what effect do these :factors have on such things as group
morale, psychological well-being, and group productivity?
Figure 4 is a graph of changes in 'overall morale' of group members over the 9 months of the winter.
Whereas there is no significant difference between years initially tF=2.297, p<0.2] there is a dramatic
decline in morale for year C over the 8.5 months of the winter. Although each year experiences a
downward trend in morale as mid-winter approaches year C shows a much more dramatic decline
relative to the other two years. In a two way repeated measures analysis of variance there is a significant
difference between years EF=19.375 p<~.0001] in overall morale. More importantly, however, there is an
interaction effect between year and month rF=2.495 D<0.0021 indicating si~nificnntlv rliff~r~.nt line.nr
~ , ~ , ~ ~ . , ~ _ _ ~ . ~ . ~ _ - .
, . . .. . . . . . .
trends In morale over the winter. Although year A declines slightly in morale as the winter progresses, by
August (when the sun first appears) the trend is upward. This is in contrast to the other two years that
experience further downward movements in morale oniv re.ho''nclina :~t the 'earl of thin `~'int~r_~`lPr in
anticipation of station opening. However this mend is much larger for year C than year B.
~ ~ _ ~ A^~— _— ~^—_ _~ ~^ Ace_ ~~ ^~ V ~ —~ 1~]
A further example of the consequences of differences in informal role structures across the 3 years can be
seen in a comparison of anger and tensionlanxiety over the winter. Figure 5 is a scatter plot of the
relationship between group means for tension/anxiety and anger over the winter gained from the profile of
mood states (POMS) instrument that was administered on a monthly basis (Palinkas et al. 1998, 2000a,
2000b, 2001, 2002; Johnson et al. 2002~. A 68 percent confidence ellipse is shown for each of the years.
Year A has the least variation in these psychological measures over the course of the winter and has the
lowest overall values across the three groups. Year C has somewhat more variability in these
psychological indices over the winter and is the most extreme overall in values over the winter months.
Year B has a high degree of variation over the winter and is In between the other years in terms of the
magnitude of psychological indices.
a' 4
o
_
a)
O 3
2
--' I
-
· },
\1 !
~~ _
~-r
.S ~
.
T
T
~ .-.,1
o
5
YEAR
10
MONTH
15
C
A
B
DYNAMIC SOCIAL NE:TWORK MODELING AND ANALYSIS
Fissure 4. Comparison of
"overall morale" over the
Austral winter across the
three years of South Pole
crews (intervals shown as
standard errors)
129
~ 6
LU
X
Z
_
U)
Z
llJ
4
3
2
I octoDe Ore ''-1~ i--—,,
octot~
august ~g |
septinb~, - -
ma~h fix Any
I
1 1
3 4
/ march
august
June-- \
~ . _
1 I aP9olctober
~ ~jUIY./ 1
I ogre /
! Q~6eptemb,~
~ J /'
~X~ o~aDn'
/
5 6 7 8 9
ANGER
Figure 5. Relationship
between anger and
tension/anxiety over the 8
months for the three years
studied (years are
encompassed by 68
percent confidence leve]
ellipses).
YEAR
o C
x A
— B
Table 3 summarizes the comparison between the three years win respect to the informal role properties
discussed above. Year A had by far the richest combination of role properties, followed by year B. Year
C lacked many of the informal role properties that are hypothesized as important for the proper
functioning of groups. It is clear from this analysis that the presence of informal role properties lead to
more desirable group outcomes. Year A had higher levels of morale and lower levels of tension, anxiety
and anger. This was followed by year B that had more moderate levels of informal role properties and
correspondingly lower levels of morale and higher levels of anger, tension, and anxiety. Year C actually
staked with the most coherent structure of the three years at the beginning of winter (Johnson et al. in
press) but the structure fragmented into several subgroups over the course of the winter that mostly
reflected the Hades versus science categones. Although year C initially displayed a cohesive structure, it
lacked the informal role properties needed to effectively deal with the inevitable internal and external
threats faced by groups (e.g., negative deviance, interpersonal conflicts, meddling from outside agents or
agencies).
Table 3. Summary of informal role properties present in the 3 years at the South Pole Station by
winter's end.
Informal Role
_% -
rropert~es
Role Complementmity
Role Consensus
Role Redundancv
Role Latencv
Role Isomorphism
Year A
High
High
Hip h .
Moderate
High _ _
Year B
Moderate
High
High-Moderate
Moderate
Moderate
.
Year C
Low
Low
Low*
Low
. Low
*Redundancy is low in terms of expressive leadership but high in terms of instrumental
leadership leading to role competition for the work leadership position
130
DYNAMIC SOCIAL NETWORK MODELING kD ISIS
Concluding Remarks: Informal Roles and Implications for Network Evolution, Stability, and
Adaptability
This paper provided a brief discussion and examples of the importance of the recognition of the iniDorTnal
and latent role properties of networks for understanding network evolution and change particularly with
respect to the direction of chance (i.e., the evolution of network fragmentation or coherence). The
examples above have suggested that groups that have a rich mix of informal role properties fair better in
terms of both the evolution and stability of group cohesion or coherence. ~ addition, these properties
appear to enhance the ability of groups to adapt to a number of both internal and external threats to group
structure and function. Whereas this research has largely been interested in the factors that keep networks
together (e.g., cohesion, coherence) and productive, such an understanding can similarly be used to
achieve just the opposite, the fragmentation or disruption of network structures. Steve Borgatti (this
conference) has been concerned with the removal of nodes and its effect on network fragmentation and
inter-node distances. Kathleen Carley has recently examined the importance of node insertion in
disrupting network structures and the importance of extra-network information in predicting network
change and adaptability (e.g., a talk at the Cambridge Colloquium Complexity and Social Networks in
December 20011. In either case, an understanding of the relationship between structural properties and
role properties of the kind discussed here can aid in producing more informed theories on network
fragmentation and disruption. It is important, for example, to know not only the network or nodal
properties of actors, but also their informal role properties. Such combined information can lead to a
deeper understanding of the potential impacts of both node removal (e.;,., the removal of expressive
leaders that bridge various categories of actors) and insertion (e.g., increased role competition for
instrumental leader) on network structures. This entails knowledge of related but distinctly different forms
of social structural data.
The examples described here have been relatively simplistic in that they involved well-bounded and
endurin;, groups with little or no movement in or out of the network. Despite their simplicity, however.
such examples can help inform us about the importance of informal role properties across a wider range
of network contexts. Further research is needed on the types of informal role properhes at; work in
networks in other settings (e.~,., terrorist networks). In addition we need a better theoretical
understanding of the structural correlates of these informal rode properties and their ultimate effect on
network evolution and stability.
Footnotes
1. This research was supported by the National Science Foundation Cants BNS-901 1351 and OPP-
9610231. In addition? a portion of this work was sponsored by a grant from the National Aeronautics and
Space Administration grant NAG-5457 1. We would like to warmly thank the winter-over crews of
South Pole Station for their kindness, friendship, and generous cooperation.
DYNAMIC SOCIAL NETWORK MODELING ED CYSTS
131
References
H.S. Becker and B. Geer. Latent culture: A note on the theory of latent social roles. Administrative Science
Quarterly, Volume 5. Issue 2 Sep. 1960. 304-313.
R.A. Dent]er and K.T. Erickson. The function of deviance in smal] groups. Social Problems, 1939 99-107
A.W. Gouldner. Cosmopolitans and locals: Towards an analysis of latent social roles. Administrative Science
Quarterly, Volume 2, Issue 3 Dec. 1957, 281-306.
G. Hogans ( 19744. Social Behavior, 2n~ edn' New York: Harcourt Brace Jovanovich.
J.C. Johnson and S. Weller. Elicitation Techniques in Interviewing. (2002) In Handbook of Interview Research (J.
Gubnum and J. Holstein, eds.), pp 491-514, Sage:Newbury Park.
J.C Johnson and M.L. Miller. "Deviant Social Positions in Small Groups: the Relation Between Role and
Individual, ~ Social Networks 5:51-69, 1983.
J.C. Johnson and B.R. Finney. "Structural Approaches to the Study of Groups in Space: A Lookat Two Analogs,
Journal of Social Behavior and Personality 1~3~:305-347, 1986.
J.C. Johnson, J. Boster, and L. Palinkas. "Social Roles and the Evolution of Networks in Isolatedand Extreme
Environments. The Journal of Mathematical Sociology (In Press).
J.C. Johnson and D. Parks. "Communication Roles, Perceived Effectiveness, and Satisfaction in an Environmental
Management Program." Journal of Computational and Mathematical Organization Theory VOL 4~3) 223-239, 1998.
J. C. Johnson, L.A. Palinkas,., J. S. Boster, S. Rakusa-Suszczewslci, V. P. Klopov, X. Quan Fu, and Usha
Sachdeva. Network Evolution. Social Roles, and Well-being in Isolated Groups: A Cross-Cultural Comparison. .
Paper presented at the Scientific Committee for Antarctic Research (SCAR) meeting. Shanghai, China. July 2002.
L.A. Palinkas, E.K.E. Gunderson, 3.C. Johnson, and A. Holland,. "Behavior and Perfonnance on long Duration
Space Flights: Evidence from Analog Environments." Aviation. Space. and Environmental Medicine Vol. 71, No.
49: September 2000.
L.A. Palinkas, E.K.E. Gunderson, A. Holland, C. Miller, and J.C. Johnson. "Predictors of Behavior and Performance
in Extreme Environments: The Antarctic Space Analogue Program." Aviation Space. and Environmental Medicine
Vol. 71, No. 41: 1-7, April 2000.
L.A. Palinkas, J.C. Johnson, J.S. Boster, and M. Housea]. Longitudinal Studies of Behavior and Performance
Dunug a Winter at the South Pole. Aviation Space. and Environmental Medicine 69~1 ) 73-77, 1998.
L.A. Palinkas, J.C. Johnson, and J.S. Boster. Social Support and Depressed Moods in Isolated and Confined
Environments. In The proceedings of the 52n~ International Astronautical Congress, IAFflAA-OlOG3.b.01, Oct.
2001.
L. A. Palinkas, J. C. Johnson., J.S. Boster, S. Rakusa-Suszczewski V. P. Klopov, X. Quan Fu, U. Sachdeva.
Cultural Differences in Mood and Social Support in Antarctica. Paper presented at He Scientific Committee for
Antarctic Research (SCAR) meeting, Shanghai, China. July ?002.
132
DYNAMIC SOCIAL NETWORK MODELING ED ISIS
