april09b.indd


Kristin Antelman and Mona Couts 

Embracing ambiguity . . . or not 
What the Triangle Research Libraries Network learned about collaboration 

One of the most long­lived cooperative groups in the United States, the Triangle 
Research Libraries Network (TRLN), is a col­
laborative organization of Duke University, 
North Carolina Central University, North 
Carolina State University (NCSU), and the 
University of North Carolina­Chapel Hill. 

Since the 1930s, TRLN members have col­
laborated to develop complementary collec­
tions. As a result, currently 70 percent of the 
titles held across the consortium are unique 
to a single institution. Expedited document 
delivery, direct borrowing, and collective 
electronic resource agreements encourage us­
ers at each institution to view all TRLN library 
collections as their own. Until recently, how­
ever, a key piece was missing: a TRLN­wide 
catalog. In the 1980s, TRLN had developed 
the first online consortial catalog, BIS,1 but, 
in the early 1990s, the libraries moved to a 
vended system; and by 2004, we were run­
ning three different integrated library systems. 

In 2006, TRLN undertook a strategic plan­
ning process and articulated an overarching 
vision, “Digital TRLN.” The vision was to 
design and create a digital knowledge and 
research environment characterized by ser­
vices and tools that facilitate access to large, 
multifaceted, print and electronic collections. 
The TRLN Council of Directors explicitly 
agreed to “embrace ambiguity” in our new 
endeavor, and to take bold steps toward our 
goals. A cornerstone of this initiative would 
fill in the missing union catalog piece using 
modern search technology. 

The Endeca software that NCSU had suc­
cessfully implemented in 2006 was licensed 
for TRLN and, in early 2008, “Search TRLN” 

was launched, featuring faceted searching 
and browsing, enhanced content, and request 
functionality. That went well, didn’t it? In­
deed. But we learned a few things about col­
laboration—and ambiguity—in the process. 

Relationships matter 
Ann Okerson notes that to succeed a consor­
tium must comprise members who trust each 
other and know each other well enough to 
take advantage of cooperation.2 The fact that 
TRLN has such a long history of successful 
collaboration is a tribute to its sustained, 
strong, and committed leadership. The 
provosts of each institution sit on the TRLN 
Governing Board, and, as a result, TRLN has a 
high profile within each institution as a model 
of longstanding, successful interinstitutional 
collaboration. TRLN’s Council of Directors, 
and its Executive Committee, meet regularly 
to oversee initiatives. Standing committees, 
task groups, and TRLN staff carry out the 
ongoing work. 

While the decision to purchase the Endeca 
software was made at the highest level of the 
organization, the library directors recognized 
that we were embarking on a deep collabo­
ration that had not existed for many years. 
There was full realization of the complexity 
of both the technology and the coordinated 
decision making that would be required. At 
lower levels, the project drew on relation­

Kristin Antelman is associate director for the Digital 
Library at North Carolina State University Libraries, 
e-mail: kristin_antelman@ncsu.edu, and Mona Couts 
is director of the Triangle Research Libraries Network, 
e-mail: mona@trln.org 
© 2009 Kristin Antelman and Mona Couts 

230C&RL News April 2009

mailto:mona@trln.org
mailto:kristin_antelman@ncsu.edu


ships built through years of TRLN­sponsored 
training and development events that bring 
staff together, familiarizing them both with 
each other and with the distinct cultures of 
each library. 

A good elevator speech 
In his excellent article on collaboration, 
“Consortia and Their Discontents,” Tom 
Peters cautions how the idea and reality of 
collaboration can get “out of whack.”3 This 
can stem from how effectively the initial goal 
is articulated, but also how it evolves through 
many meetings between players with their 
own objectives. Search TRLN was fortunate 
to have a clear and simple goal, embodied 
in the project’s name. And, given NCSU’s 
recent experience with the Endeca software, 
the roadmap to success could be envisioned. 

So, where’s the ambiguity? Recalling the 
broader Digital TRLN vision, from the outset 
of this project there were also implicit goals. 
It was clear that the Search TRLN project was 
building a shared platform. For what, exactly, 
was yet to be defi ned. 

A sense of urgency 
Also helpful in syncing idea and reality is 
a shared, ongoing sense of urgency. Each 
library planned to use this platform for its 
local catalogs. The fact that this project 
paved the way by doing much of that work 
certainly inspired a sense of urgency. And it 
is never hard to find passionate stakeholders 
in the catalog. The not­insignifi cant upfront 
financial investment in software, technical 
consultation, and time­limited project staff 
also played a part in motivating us to keep 
the project meeting its targets. 

“Us” and “me” 
In “The Lone Ranger is Dead: Success Today 
Demands Collaboration,” Betsy Wilson points 
to the need for collaborators to change their 
perspective from “me” to “us.”4 To a certain 
extent, Search TRLN was able to dodge this. 
A fortuitous aspect of the project was that 
each institution was able focus on both “us” 
and “me.” While the “us” was the consortial 

catalog, the “me” was the local “skin,” a lo­
cally scoped view to the catalog. The local 
skins (each of which looks quite different 
than Search TRLN!) served more than once 
as a safety valve in disagreements about the 
TRLN interface: when anyone felt they were 
making a concession, the local skin could 
stand as the shining model of perfection yet 
to come. 

More risk . . . 
We knew that successful collaborators should 
assume shared (and roughly equal) risk, and 
risk was certainly present in this project. 
First, everyone had to buy into the idea of 
outsourcing the hosting of their catalog to 
a network service provider in the Triangle. 
This was a requirement of the technical ar­
chitecture, but also served to level both risk 
and responsibility. The project did introduce 
quite a bit of change into day­to­day ILS 
operations as well, one aspect of which is 
that the primary administrator of the TRLN 
application works for the consortium, not any 
of the member institutions. The exact nature 
and dimension of this risk will probably not 
become clear until unanticipated things begin 
to happen! 

. . . with less control 
Successful collaborators must also sacrifi ce 
some autonomy. The libraries had to coor­
dinate on numerous data decisions in order 
for search and navigation to work and to 
create a good user experience. A number 
of these decisions have implications, some 
well understood and some not so much, for 
what each institution can do in their local 
implementations. TRLN history teaches us 
that collaboration can be most challenging in 
the area of data coordination, where for good 
reasons (or not!), each institution employs 
significantly different bibliographic control 
practices. Again in this case, our path was 
eased by the nature of the Endeca technol­
ogy. The shared “database” is simply an index 
created from merged catalog data extracts, 
with each ILS still serving as the authoritative 
data store. Nevertheless, the degree of future 

April 2009  231 C&RL News 



autonomy each library will lose was intention­
ally not deeply explored during the project. 

Roles and skills 
Leveraging complementary deep skills within 
the libraries was critical to meeting the go 
live date. In part by design, but also hap­
penstance, each library ended up taking the 
lead role in one area. One focused on the 
Endeca technical architecture, another on 
graphic design, another MeSH and NLM, and 
another on usability testing. 

Wilson asks if we can be taught to col­
laborate. Certainly, collaboration requires 
both specific skills and a certain mindset. For 
example, negotiating, being able to see trad­
eoffs, and being comfortable sharing control. 
Collaborators learn how to cross boundaries. 
Search TRLN involved many players: a steer­
ing committee, three task groups (whose 
chairs sat on the steering committee), and a 
newly hired project librarian (who fortunately 
possessed project management and people 
skills as well as technical skills). With each 
task group needing to be representative of all 
institutions, as well as all types of libraries in 
the consortium, these groups were not small. 
In retrospect, discussion of or training in the 
skills needed for successful collaboration 
would have been beneficial for everybody. 
It could be said that training in collabora­
tion is training in how to manage—if not 
embrace—ambiguity. 

Collaboration and ambiguity 
In his book The Ambiguity Advantage: What 
Great Leaders Are Great At, David Wilkinson 
explores ambiguity in the context of leader­
ship styles. In the collaborative leadership 
style, ambiguity is often associated with 
“talk, not much action,” and the tendency 
to listen to all voices equally.5 Where the 
goal is to innovate, and on a tight timeline, 
the TRLN collaboration drew on a number 
of characteristics of “generative leadership,” 
the highest level of leadership in Wilkinson’s 
framework. Generative leaders are able to set 
goals and leave the route to achieving them 
open, yet they understand the difference 

between a learning situation and a decision­
making one.6 

Search TRLN is an example of a success­
ful collaboration in a new area for many 
libraries, namely, core services. Collaborating 
on a core service, as opposed to collection 
development, requires a different kind of 
staff support, unfamiliar risks and roles, and 
perhaps a more uncomfortable—or at least 
unfamiliar—level of ambiguity. In taking a 
new and ambitious direction such as Digital 
TRLN, the consortium was implicitly relying 
on its strong center of gravity: a long­standing 
culture of collaboration, not just cooperation, 
with shared agendas and values. The leaders 
had confi dence that we could accommodate 
significant tensions as the routes to achieving 
our goals were explored. Indeed, the Search 
TRLN project relied on TRLN’s strong center 
of gravity to help us survive the centrifugal 
forces of complex collaborations. 

Part of this project was to fi gure out when 
to embrace ambiguity and when not to. Peters 
says that “often the best collaborative efforts 
occur when there is lively disagreement and 
debate, not consensus”7—and this project did 
include some level of conflict. There was one 
steering committee contingent that wanted 
the group to delve more deeply into options 
and one that simply wanted to accept the task 
group recommendations as is and keep to the 
implementation timeline. It rapidly became 
clear that the former path was strewn with 
too much ambiguity, and so the project took 
the latter path. Too much ambiguity in the 
first phase could have jeopardized not just the 
timeline but also the whole project. 

Our whole world is ambiguous 
Libraries are awash in ambiguity these days, 
and no more so than in regard to our catalogs. 
It would be hard to find an area of academic 
librarianship now where the status quo was 
more despised, and the way forward more 
uncertain. There are at least two dimensions 
to this, the more obvious one being that, as 
a profession, we realize that we have poor 
discovery tools that have not kept pace with 
either Web search or user expectations. TRLN 

232C&RL News April 2009



librarians were in agreement that our catalogs 
were bad, and that what NCSU had in its En­
deca catalog was, if not the answer, at least 
an improvement. The harder challenge is that 
the very concept of the catalog is in transition. 
Implementing a “next­generation” catalog 
doesn’t answer the question, what should a 
library catalog be anymore? 

Still, all one can do is forge ahead, using 
vision, talent, and enthusiasm for ambiguity to 
chart a way forward. Ever aware that competi­
tion lurks beneath even happy collaborations, 
we hope and expect competition to thrive 
through the local skins, and we embrace the 
challenge of integrating innovation conceived 
in the individual libraries with innovation con­
ceived centrally. We also realize the imperative 
to remain flexible and not allow ourselves 
to become wedded to a model of the union 
catalog that may not make sense much longer. 

As we head into phase two and beyond, 
there are plenty of ideas within TRLN for what 
content and services belong inside, around, or 
outside of Search TRLN. And when the time 
comes to define boundaries—between TRLN 
and other catalogs, digital collections, institu­
tional repositories, and the broader world—we 
are as well positioned as we could hope to 
be as a consortium to move from ambiguity 
to action, and maybe back again. 

Notes 
1. Joe A. Hewitt, “The Triangle Research 

Libraries Network,” North Carolina Libraries 
42 (Summer 1984). 

2. Ann Okerson, “Strength in Numbers: 
Library Consortia in the Electronic Age,” IDT 
Net conference, Paris (2000). 

3. Thomas A. Peters, “Consortia and Their 
Discontents,” The Journal of Academic Librari­
anship 29:2 (March 2003). 

4. Betsy Wilson, “The Lone Ranger is Dead: 
Success Today Demands Collaboration,” C&RL 
News 61:8 (September 2000). 

5. David Wilkinson, The Ambiguity Advan­
tage: What Great Leaders are Great At (New 
York: Palgrave Macmillan), 2006: 36. 

6. Ibid., 121. 
7. Peters, 113. 

April 2009  233 C&RL News