The "necessary characteristics" of cyberinfrastructure for humanities and interpretive social sciences are
1. It will be accessible as a public good.
2. It will be sustainable.
3. It will provide interoperability.
and most important for us:
4. It will facilitate collaboration.
and finally and most interesting for us:
5. It will support experimentation.
Although cyberinfrastructure itself should be stable and reliable, it will need to support ongoing experimentation, and it will need to evolve. Researchers in the social sciences and humanities will need to experiment, and that experimentation will be crucial to bringing change to those disciplines. Institutions must encourage risk-taking by creating frameworks through which junior scholars and students are rewarded for ambitious research programs. Offering this encouragement means providing laboratories, postdoctoral grants, and other support that allows these research programs to be worked out and critically assessed. Institutions also need to allow their libraries and university presses to experiment and take chances in order to find more successful models of scholarly communication. It is important to foster a culture of experimentation by underwriting explicit mechanisms and traditions for capturing and sharing the lessons learned through innovation. True experimentation always carries with it the possibility of failure, as the necessary price for success, yet informative failures are essential to moving forward into the unknown, and they should be reported without prejudice and duly valued on that account. (p. 29)
In advocating for cyberinfrastructure, the ACLS called for digital technologies that go beyond presentation and visualization of extant materials, collections, or scholarship, which tends to reinforce the individual-centric model of DH. Instead part of its call was for what we would call a research environment or experimental system, that encourages interaction and "feverish archives": "A cyberinfrastructure for humanities and social sciences must encourage interactions between the expert and the amateur, the creative artist and the scholar, the teacher and the student. It is not just the collection of data—digital or otherwise—that matters: at least as important is the activity that goes on around it, contributes to it, and eventually integrates with it." (p. 11)