It's been a long time since I wrote this newsletter. The previous missive was sent out on November 7th, almost three months ago. I have been writing the Socratus newsletter pretty religiously for most of these three months, and also writing a regular 'AI and cognition' focused newsletter on LinkedIn (a continuation of the podcast from last fall) but haven't found time to put words on personal paper until today.
Why now? What's new?
Glad you asked. What's new is that I have been thinking about Research Programs and feel I have something to say about them in general and about one in particular. But with a twist. What's a Research Program? According to Wikipedia:
A research program (British English: research programme) is a professional network of scientists conducting basic research. The term was used by philosopher of science Imre Lakatos to blend and revise the normative model of science offered by Karl Popper's The Logic of Scientific Discovery (with its idea of falsifiability) and the descriptive model of science offered by Thomas Kuhn's The Structure of Scientific Revolutions (with its ideas of normal science and paradigm shifts). Lakatos found falsificationism impractical and often not practiced, and found normal science—where a paradigm of science, mimicking an exemplar, extinguishes differing perspectives—more monopolistic than actual.
Lakatos found that many research programs coexisted. Each had a hard core of theories immune to revision, surrounded by a protective belt of malleable theories. A research programme vies against others to be most progressive. Extending the research program's theories into new domains is theoretical progress, and experimentally corroborating such is empirical progress, always refusing falsification of the research program's hard core. A research program might degenerate—lose progressiveness—but later return to progressiveness.
I find research programs to be the best combination of intellectual leadership, creativity and productivity on offer today. Some of them come with concrete goals - the Human Genome Project, for example - and others have more nebulous ones - Grothendieck's program to recast Algebraic Geometry, for example - but when they work, research programs are able to mobilize a distributed network of researchers along aligned (not identical) goals. In my intellectual life, I have been witness to three research programs at different stages of development:
The Computational Approach to the Mind, i.e., what became known as Cognitive Science, pioneered by Chomsky, Miller etc. Treating the mind as a computational device. This was well under way by the time I got to it and in retrospect, it was past its prime by then.
Embodied Cognition: same topic, different take. The mind isn't an abstract computational device but a concrete entity shaped by the kind of bodies we have. Originated with Varela and his collaborator's work on the Embodied Mind along with allied developments in Robotics by Brooks and others. Heterodox when I got to it but was on its way into the mainstream. Still not the mainstream (i.e., unlikely to get to you tenure at MIT/Harvard/Stanford) but a respected alternative.
The Scientific Study of Consciousness. The least programmatic of the three; initiated by philosophers more than scientists and has only recently become respectable enough that proper scientists are writing layperson friendly manuals on the topic. Has more room for wild variants than the previous two. You will never find quantum embodiment listed on anyone's academic profile but quantum consciousness has a Nobel Prize winner's imprimatur.
Thanks for reading Ranganaut! Subscribe for free to receive new posts and support my work.
Research programs are a thing and being involved in one can be good for your career. When done well, they are exciting ways to institutionalize the distributed, (relatively) non-hierarchical nature of knowledge production. Which brings us to the twist in my tale. We are taught to conduct research, but absolutely no one teaches you how to formulate a research program. The ones I know (like all the above) are the product of individual brilliance, social capital and privileged access to resources. Being at the right conference at the right time or studying with the right teacher might make all the difference between a key player and an NPC.
And even then, research programs are rarely formulated with the level of imagination or rigor that researchers bring to their personal research projects. There's mismatch between the needs of the collective (the research network) and the skills of the individual researchers. Which is why, you often find the definite chronicle of the program being written by science journalists or academics writing as synthesizers rather than creators. In contrast: here’s an original contributions from an earlier era.
What is to be done about this?
Research has become so complex that it needs presearch, a design and prototyping phase that precedes the production phase of research. That's even more important for research programs that have to choose goals, research pathways, theoretical frameworks and technology development appropriately and might end up investing enormous amounts of time, money and social commitment to the program. No one would ever build a commercial (or even an open-source) product without understanding its design constraints. Why should research be any different?
What is Presearch?
Design serves as a catalyst in the realms of engineering and commerce, opening up the adjacent possible for exploration and setting the stage for transformative shifts in behavior over time. Reflect on the evolution of web design: initially, web pages mimicked the static nature of printed pages. However, the adoption and normalization of scrolling revolutionized web design, leading to the concept of websites with infinite scrolling—a notion inconceivable in the tangible realm.
It's intriguing to observe that the corporate sector celebrates and integrates various professional roles—designers, architects, and more—who are dedicated to exploring the adjacent possible as part of their professional mandate. The sector recognizes complexity and has come up with ways of dealing with it. In stark contrast, the academic world exhibits an informal approach to this exploration. The absence of an academic equivalent to "knowledge designers" is notable, and researchers are seldom equipped with the methodology to navigate the terrain of research questions and ideas that are neither too radical to be dismissed nor too derivative to be dismissed as trivial. The concept of a knowledge studio, where seasoned researchers mentor and critique the inventive proposals of students, remains unrealized. While research seminars diligently assess the integrity of experimental design and the plausibility of alternative hypotheses, they seldom provide a forum for the critical evaluation of the novelty of ideas in the manner that design and art studios celebrate creative endeavors.
Presearch should be done in knowledge design labs on the same floors as experimental labs - drawing inspiration from forward-thinking initiatives like the Near Future Laboratory and the movement towards "design fiction," which involves crafting speculative documents and artifacts that, while nonexistent in the present, could very well materialize in the near future. Presearch explores research's adjacent possible, or in the case of a research program, a whole host of adjacent possibles.
I learned a lot from "Speculative Everything," a seminal text in the design fiction movement, which advocates for leveraging design as a mechanism not just for creating objects but ideas, and to conjecture about potential futures. More recently, I started reading ‘Discursive Design,’ about the design of artifacts intended for reflection instead of use. The authors of Speculative Everything and Discursive Design, through their design work, embody their concepts in tangible forms. Yet, there's no inherent limitation to extending this approach to speculative theories, experiments, and entire knowledge traditions—the essence of knowledge creation. Let’s conclude with a definition:
Presearch is the employment of design as a mechanism to generate ideas, theories, and broadly, to prototype knowledge creation.
Three final claims:
A presearcher is a philosospher and presearch is speculative philosophy.
Every research program should have an embedded presearcher.
Presearch demands distinct modes of communication, both within its research program (i.e., to other members of the program network) and to the world outside.
I wouldn't be going on an on about presearch if I didn't have a specific research program in mind, details of which will be shared in the next newsletter. Some clues as to the topic if you glance through this, this, and this.
PS: I am not sure of the frequency at which I will be writing this newsletter. No more than twice a week. No less than once a month.