Category Archives: platforms

People Scopes, Platforms, and Research

I recently finished reading Pasteur’s Quadrant. The gist of the book is that the author, Donald Stokes, argues that the traditional (back to the Greeks) distinction between basic research and applied research is misguided.  Louis Pasteur didn’t make that distinction. He in fact was very much driven to solve real-world problems whilst also pursuing basic scientific understanding of the phenomena that he observed. The book does a great job of explaining the historical antecedents of the basic-applied distinction in the modern research-industrial complex, and I would highly recommend it to other researchers.

Stokes defines basic research as “experimental or theoretical work undertaken primarily to acquire new knowledge of the underlying foundation of phenomena and observable facts” whereas applied research is concerned with “the elaboration and application of the known … to convert the possible into the actual, to demonstrate the feasibility of scientific or engineering development, to explore alternative routes and methods for achieving practical ends.” But he argues that this one dimensional dichotomy is too simple and that it should be expanded to a two dimensional typology with consideration of use on one axis, and fundamental understanding on the other. The quadrant of this typology that is concerned with fundamental understanding AND considerations of use is termed Pasteur’s quadrant, or alternately “use-inspired basic research.”

I reproduced a diagram from the book that illustrates the typology:

Use-inspired basic research can advance both fundamental knowledge as well as technology. Which is a good thing because new (or better) technology enables new scientific questions to be asked. And the answers to those scientific questions can often lead to better technology designs. The scanning electron microscope (SEM) is a good example.

I think it’s likely that many interdisciplinary fields thrive at this intersection of applied and basic research and Human Computer Interaction (HCI) is no exception. A lot of HCI research seeks to harness fundamental knowledge for the design of interactive systems but at the same time use new technologies and interfaces to ask fundamental questions about people and interfaces (though these two phases do not always occur simultaneously, nor must they). Basic findings can trickle back to core disciplines (e.g. psychology, sociology), and other findings from other core disciplines can inform the designs and the engineering that goes into building the next generation of interactive systems.

Take a simple new technology that has had a huge impact on computational social science research: Twitter. Twitter is the computational social science “scope” that lets researchers ask all kinds of interesting questions about social psychology. Refining such knowledge could lead to a newer social scope (Twitter 2.0?) that is even better. Another example is Digg.com, which a few years ago was a technology that helped advance our understanding of information novelty and decay.

Real people-scopes working in naturalistic settings are essential for basic research as well as for driving technology forward. Academia (not just in HCI, but in other interdisciplinary social sciences) needs to get more strategic about building people-scopes, basically platforms that enable new human-centered questions to be asked, at scale. Unfortunately academia is not traditionally good at platforms. Right now I can only think of a few academic projects that have done this successfully: Movie Lens at University of Minnesota, Scratch at MIT, and maybe IBM has also had some semi-successful ones.

There are likely a number of reasons why academia is not that good at platforms: (1) grad students may not be around long enough to grow and maintain the system, (2) the risk of failure is immense and too high for a pre-tenure faculty to bear, (3) there are not enough sustained resources to maintain the systems, and (4) there are little to no marketing resources to support the acquisition of users. So there are incentive as well as resource issues here.

It may be that start-ups are simply a better source of new social platforms, since the market can quickly winnow out the unsuccessful ones, and the risk is externalized. But I think it may also warrant thinking about how funding agencies like the NSF might better support (e.g. through sustained resources and incentives) the construction of the next generation of people-scopes.