Assessing Projects in Relation to the Leadership Framework
Rough project comparisons, along the principles of the framework, are set forth in the following paragraphs.
Willingness to confront reality
Though all four of the projects cited here faced major challenges of scope, NSDL and DLESE seemed less able than the others to gain agreement on how to limit expectations during startup. Furthermore, NSDL and DLESE—both tied to the “library” legacy—were immediately embroiled the problems of purpose and identity brought upon every library by the realities of Internet, Google and other disruptive technologies [4]. Absent such legacies, and with somewhat operationally oriented leadership (at NSF, on steering committees, etc.), NSFNET and Unidata had more latitude for confronting the realities that startups must face, especially the reality that pleasing every collaborator and constituent is literally impossible.
A simple, coherent strategic concept (Collins’ “hedgehog concept” [5]) – At least in hindsight, the four cited projects might be ranked for simplicity of concept (from simple to complex) as follows: NSFNET, Unidata, DLESE and NSDL. In the first two cases, I think simplicity was encouraged by NSF’s grant structure, whereas the grant structures for NSDL and DLESE tended to favor diversity over coherence. This led to imprecise alignments among the collaborators’ visions and priorities, which in turn compounded the difficulties DLESE and NSDL had in gaining collaborator agreement on value propositions and target constituents.
Foreshadowing of this difficulty in NSDL may be discerned from the previously mentioned NRC report by comparing, for example, the following two quotes from Appendix B. Note Arm’s implicit focus on “library” as (expertly selected) content in contrast with Lynch’s focus on “library” as place for transformative action.
- “All materials in the library will be selected by members of the library staff. Sometimes, selection will be at an item level, at other times by groups of material. The method of selection and the selection criteria will be stored with all material, so that users will know why each item is in the library.” [NRC/Arms 1998]
- “A library [serving] undergraduates in science, engineering, mathematics and related subjects (and their teachers) is primarily an engine for enhanced teaching and learning rather than an opportunity to transform the broad system of scholarly communication and publishing. … In my view, much of the challenge here is how to supplement, extend, and enhance traditional textbooks and related educational materials (such as problem sets or experimental exercises) in the digital environment.” [NRC/Lynch 1998]
Willingness to say “no”
Admitting that this is highly subjective, I think the NSF structuring of the NSFNET and Unidata grant programs provided better cover and support for leadership decisions about what not to do, in comparison with how DLESE and NSDL were shaped. In the NSFNET case, one might cite the decision not to support (at least initially) certain protocols (such as BITNET, SNA and X.25), even though they were currently in use or represented international standards. In Unidata, a number of (controversial) decisions were made about which data-visualization and analysis packages could or could not be fully supported. In contrast, I think both digital libraries found it hard to gain agreement and support for decisions about focus, and this situation was exacerbated by the committee structures mentioned below as well as the absence of a clear “hedgehog” concept for each library.
Clear identification of who holds responsibilities
In a similarly subjective vein, I think that the NSF Program Officers for DLESE and NSDL felt more constrained (than those for NSFNET and Unidata) in clearly designating leadership responsibility. DLESE and NSDL also adopted committee structures that, rather than being advisory, were chartered to play roles more akin to governance, and this effectively diffused the leadership responsibilities.
Advancement as a cumulative process
Fortunately for NSFNET and Unidata, NSF support remained steady and strong through difficult, controversial times. Similar attitudes have been evident in NSDL and DLESE, but for the latter, the duration of this patience may have been insufficient. True community-wide enthusiasm for Unidata took something like a decade to achieve, and one of Unidata’s key low-level software packages (NetCDF) required some 10-15 years after its first release to achieve widespread use (in the geosciences). Personally, I was saddened by the cessation of NSF/GEO funding for DLESE at a point when evidence of educator enthusiasm was ramping up and important ideas were beginning to emerge on such important topics as:
- Fostering and supporting educationally and scientifically effective uses of numeric data (i.e., key concepts for extending the traditional notion of library content).
- Creatively linking library resources to educational standards and sophisticated characterizations of learners’ conceptual development. [Butcher et al., 2006]
- Modeling the library less as a place for finding and accessing content and more as a place for development along the lines envisaged by Lynch (as quoted above).
- Building community among (sometimes isolated) science educators.
An enduring purpose
As mentioned earlier, the term “library” has been a mixed blessing for DLESE and NSDL. On the one hand, it evokes a history and legacy that is nothing if not enduring. On the other hand, clear statements of purpose for digital libraries such as NSDL and DLESE have been devilishly difficult to articulate, much less to adopt in collaborative settings, because Google and similar technologies have so disrupted the traditional approaches to finding and accessing information [Lagoze et al 2005]. Fortunately for Unidata and NSFNET, their original purposes—simple as they were—withstood the test of time.
Immutable core values (around which to advance)
Perhaps this item should not be in the analysis framework, because I perceive little variation among the examples. Each of the four programs was built around core values shared by all stakeholders, including the NSF Program Officers and the key collaborators. The only exception may have been conflicting views in DLESE about the place of traditional peer review in collection building. For some, this value was so central that its absence would fundamentally diminish the value of DLESE. For others, emerging possibilities for materials evaluation (employing technology-enabled networks of trust, feedback from teachers and learners, well constructed usage statistics, etc.) held such promise in terms of scalability and end-user engagement that traditional peer review (especially as a library function rather than a publisher function) seemed contrary to educational advancement [6].
Primary alignment with a single value discipline (among operational excellence, customer relations, and product leadership)
NSF’s rather prescriptive grant program for NSFNET was aligned primarily with operational excellence, capitalizing on the relative maturity of TCP/IP. Unidata’s earliest focus was on customer relations, and the alignment successfully shifted toward product leadership (i.e., leading-edge software tools) during the 1990s. In contrast, I think community leaders and NSF all contributed to NSDL and DLESE ambiguity regarding value proposition and value discipline. In making this judgment, I note that none of the value disciplines is easy to achieve without clarity about concept, purpose and target constituents, reflected in the other elements of this framework. For example, returning to the quotations cited above, Arms’ perspective seems to lead toward operational excellence, while Lynch’s seems to imply product leadership.
[4] James Keller anticipated aspects of this problem early in the NSDL planning stage: “The ability of the digital library to recast the learning process by extending across traditionally distinct institutions and processes raises the question of whether the word ‘library’ remains an appropriate moniker. The so-called ‘digital library’ will, if properly designed, quickly become a part of most, if not all, scholarly aspects of the university and the larger research community.” [NRC/Keller 1998]
[5] Collins, in Good to Great (referencing an Isaiah Berlin essay, “the Hedgehog and the Fox”) asserts that the most successful companies are more like hedgehogs than foxes. The latter are characterized as seeing a complex world and acting on many levels, where hedgehogs integrate their decisions and actions under a single concept or unifying vision, yielding remarkable clarity and simplicity. [Collins 2001, pp 90-120]
[6] Though written in the context of NSDL planning, Keller’s previously cited contribution to the NRC report anticipates this tension: “Technical development of the key functional elements of a national digital library to support undergraduate SMET education can largely be fulfilled with existing web-capabilities. It can be built with existing knowledge and tools, but successful use of this resource will require a fundamental rethinking of the larger learning process in which it is intended to operate. This would affect some of the defining elements of the university, including peer review, standards for professional achievement, and undergraduate teaching.” [NRC/Keller 1998]