Many of the recent calls to reform humanities graduate education have focused on shortening the time to degree, offering more professional opportunities, and legitimizing alternatives to the traditional research model.

I would like to imagine one such alternative: the 20-year dissertation.

Before you stop reading, let me say that I agree unequivocally with the need to reduce the time to degree for graduate students in the humanities. Five years is a reasonable goal.

But, paradoxically, I think that a dissertation should take at least 20 years to complete. Here is how we can have it both ways.

In the medical sciences and social sciences, it is not uncommon to think in terms of grand challenges or big problems: cures for disease, energy independence, global health, clean water.

These problems are not only big but will require years of effort to solve; they require many people from many disciplines to work collaboratively with generations of scholars, often in labs and often with folks from public policy, the nonprofit and for-profit worlds, philanthropy, and government.

That kind of research proceeds incrementally and may, in fact, fail. Nonetheless, it is cumulative, with students coming in, working in labs for a short period of time, receiving their degrees focused on elemental aspects of the research process, and moving on.

In the humanities, we approach things in the opposite way: ever-smaller, ever-more-specialized problems, addressed by individuals working almost always in isolation, with the end product being a dissertation monograph. We peer into the microscope, focus, and train our students to do the same.

What if, instead, we asked ourselves the following question: What are the biggest challenges in a given discipline or field?

What would it mean to tackle these with generations of graduate students, working with faculty members and outside collaborators - across multiple disciplines and institutions - to come up with ways to understand, contextualize, and respond to these challenges?

Imagine, for example, a 20-year dissertation on the broad topic of the history of migration in Europe.

It would treat the topic historically, geographically, socially, culturally, and economically. It would focus on issues of race, the history of the nation-state, histories of exile and expulsion, technologies of mobility, assimilation, identity, gender, and representations of difference and otherness. It would be multilingual, transnational, and multimedia - and it would take decades to do well.

It would also have primary material, translations, annotations, commentary, interactive media, historical analyses, close and distant readings, data sets, interpretations of data sets, and maps.

It would be a forum for debate, a resource for course development, and a site for reviewing scholarly research of all types. In essence, it would be an integrated and collaborative research product.

Big projects like this already exist in certain humanistic fields, such as archaeology, where long-term fieldwork and broad-based collaboration are common.

In the digital humanities, foundational projects like the Perseus Digital Library are more than 20 years old. Many digitization projects of major textual collections are under way, allowing scholars an unprecedented scale for conducting research.

What would it mean to compare and analyze every textual variant, every cultural context, every author, every annotation, and every commentary on, say, the Talmud?

As new multidisciplinary fields, such as environmental humanities and urban humanities, gain traction, we might see 20-year dissertations on topics such as the emergence of the Anthropocene epoch (the impact of human activities on the earth's ecosystems, biodiversity, and climate); the cultural, social, and urban history of megacities, where nearly 10 percent of the earth's population now lives; or the grand challenge of designing and building a more democratic city.

In my own field, German-Jewish studies, the Holocaust is quite possibly the most documented event of the 20th century. As this documentation is steadily digitized, it gives rise to "big data" problems, which range from the historical and geographic to the linguistic, cultural, and ethical.

These are all 20-year projects that, in various ways, combine criticism, interpretation, and transhistorical, comparative perspectives with speculative, experimental, and projective problem-solving.

How would this work in practice? First, the projects require "humanities labs" in close collaboration with libraries and presses.

Maybe a few departments would establish a lab to support a handful of 20-year dissertations, based on faculty research areas, institutional resources, and intellectual interests.

Senior faculty members would need to play a crucial advisory role and work to define the scope of the lab and the dissertations emerging from it.

To secure legitimacy within the academy, the lab director would work with a university press right from the start to get the project under contract and plan for a series of timed releases, perhaps every two or three years.

These releases could take multiple forms (digital and print), but would put collaboration, multimedia authorship, quick dissemination, and broad public resonance in the foreground.

I could see a university press picking up a series of grand challenges in the humanities, published iteratively. The idea would be to produce a living project that sustains debate, discussion, research, and teaching.

Second, graduate students would be admitted to programs based on a series of 20-year dissertations under way at a given institution. UCLA might be known for its projects on one topic, and Stanford for another.

Students would be admitted into a humanities lab (perhaps instead of or in addition to a department) and become part of a team composed of graduate and undergraduate students, faculty members, librarians, technologists, and review boards.

Students would make designated contributions to the project, measured in flexible but clearly documentable ways, which would legitimize a range of intellectual work, including narrative and interpretative content as well as design, visualizations, maps, data sets and databases, textual markup and metadata, and platform development.

In essence, the vertically integrated team would be involved in something called "humanities knowledge design." This collective and collaborative knowledge would constitute a rigorous, long-term project to tackle a research question on the broadest possible scale.

Third, I imagine that students would begin as apprentices in their first year, taking classes related to the broad topic, mastering core competencies, and learning about research methods, digital tools, and the changing nature of public humanities scholarship through a robust connection to a digital-humanities program.

As they advanced in the program, they would become research associates and begin - with appropriate guidance, oversight, and peer review - to design the form of the scholarship and to publish on the project site.

Graduate students would be credentialed along the way, and every element that they created, authored, or contributed to would contain a unique identifier, making it possible to not only identify contributions but also to share, export, and reuse materials in a creative-commons environment.

These dissertations would have broad public resonance, involving potential collaborations with museums, libraries, historical societies, and professional organizations as well as the nonprofit and for-profit worlds.

When students finish their degrees, in five years, the "dissertation" would most likely not be done.

And that's the point: The research would continue beyond the student's five-year contribution; it would continue on with other graduate students and other faculty members and, perhaps, if the student was still compelled by the project, he or she could continue in a new capacity: as a postdoc or assistant professor - in essence, as a mentor, editor, and project collaborator.

The 20-year dissertation would ultimately have dozens of authors and be the product of many doctorates. In essence, it would embed specializations and individual scholarship into a significantly broader, interdisciplinary humanities-research context.

The results would transform the scale, duration, impact, and public resonance of humanities scholarship. And all without extending the time to degree.