Co-emergence of Our Knowledge of Mind and World (I)

Here are some excerpts from the report by Piet Hut and Roger Shepard to the Alfred P. Sloan Foundation, summarizing their activities in the academic year 1994-1995, concerning their investigations of limits to scientific knowledge. Their research project was titled: Co-emergence of Our Knowledge of Mind and World (I).

Our project was inspired through the workshop on `Limits to Scientific Knowledge', held at the Santa Fe Institute in May 1994. It was at this workshop that we met, and we started our series of dialogues soon afterwards. The question we focused on was the role of limits in scientific investigations, not only in specific disciplines, but in the whole approach to scientific research. Any type of research takes place against a backdrop of a whole world: the human society set within a natural environment. And, ultimately, any type of research outcome is reported and shared by individual minds.

The structure of the mind of the individual, the structure of society as a whole, the structure of nature, as well as the structure of the technological and mathematical tools we use to investigate nature -- all these structures are candidates for investigations into limits to scientific knowledge. In our particular project, we choose to focus on the structures of mind and nature, or more broadly mind and world. This seemed apt, given our respective backgrounds: Piet Hut, an astrophysicist interested in a physical characterization of the dynamics of star clusters, and Roger Shepard, a cognitive psychologist seeking a mathematical characterization of the human representation of generalization and imagined spatial transformations.

This broad view of an inquiry into limits, on all levels of scientific knowledge as well as cognition in general, lead us to organize a small informal workshop, in early March 1995. Besides the two of us, we invited two other alumni from the Santa Fe limits workshop, Robert Rosen and Otto Rossler, as well as Mel Cohen, a neurobiologist at Yale. Before and after this workshop, the two of us held a number of meetings during which we continued our dialogues.

Our results are not yet of the concrete type that one might expect from a research project within a well-established and tightly circumscribed discipline. Rather, we have learned to ask questions into the nature of limits in a way that is becoming more and more focused. We are satisfied with this type of result, given both the uncharted character of the terrain of study, as well as the fact that we could only engage in this study in a part-time fashion (since astrophysics and cognitive psychology remain our main fields of research). What is especially satisfying is the fact that we see a clear momentum building up towards further exploration of the questions we have begun to formulate, among the group of people we have communicated with so far.

Background

To start with the latter: the time seems ripe for widely interdisciplinary collaborations into the nature of scientific knowledge. Over the last few decades, cognitive science has been a trail blazer, combining the fascinating insights that have been obtained in such diverse discipline as computer science, brain studies, and cognitive psychology. But in the last five years, the scientific study of consciousness has accelerated the intensity and interest of this type of research. This much is clear from even a quick glance at the large number of books, articles, and conferences on the structure, evolution, and role of consciousness, since 1990.

Why this interest in consciousness, or in other words, in the workings of the mind? In the past, philosophers and other theoretical academic types have speculated endlessly about the mind, about the relation of mind and body, etc. But now, for the first time in human history, we are beginning to get experimentally involved. In physics, it was the respect for experimentation that started off the scientific revolution leading to Newtonian mechanics and beyond. It seems likely that cognitive science is on a similar threshold, in our own time. And experimentation and interpretation here take on many complementary forms: from the attempts to find geometric models for perception, to measurements of the dynamics of nerve cells, to experimentation with neural nets and other computational structures.

Against this backdrop, it seems timely to pause for a moment amidst the onrush of new results and ideas, and to reflect upon the nature of the overall project of trying to understand the mechanism of human consciousness. Yes, we live in a very exciting time, and breakthroughs on all levels abound. But at the same time there is the danger that all these new results can blind us and make us overlook the narrow restrictions that might still be built in into the approaches we customarily take in scientific investigation. As Phil Anderson expressed it to one of us, over lunch one day, in the form of a question: ``What are the Ptolemaic circles in current research?''

Looking back to previous historical periods, it is easy to see that whole disciplines were on the wrong track for long periods of time, either because of a lack of relevant data, or because of the overwhelming force of received opinion. An example of the first type was the state of biology in the seventeenth and early eighteenth century, before geological evidence began to show how species had developed over many millions of years. An example of the second type was the state of physics in the eighteenth and early nineteenth century, when all explanations were given in time-reversible terms, following the overwhelming success of Newtonian mechanics, while ignoring the abundant evidence that almost everything in nature is glaringly non-reversible. It was the need of understanding the technology of steam engines that gave a practical push towards an analysis of irreversible processes, which led to a rude awakening from the idealized time-symmetric slumber that physics had lapsed into.

Results

So what parts of our study of cognitive processes are prime candidates to be labeled by future generations as Ptolemaic circles? It is around this question that we can fruitfully present the results of our project so far. We will state them here in terse outlines.

• hidden assumptions make themselves felt indirectly, by leading to apparent limits in the terrain of study.

• further scrutiny of what seems to be `off limits' in a given area of study may point to the limitations in the methods used, rather than in the material that is being studied.

• realizing that limits are inherent in methods rather than in the structure of reality (mind and world), leads to an attitude of optimism.

• this optimism stands in stark contrast to the prevailing sense of pessimism and crisis in many areas of pure and applied studies. Rather, it resembles the spirit of the eighteenth century, the Enlightenment, following the spread of Newtonian physics.

• an optimistic view of the role of reason, a form of rekindling of the spirit of the Enlightenment, is what we propose. Not only the type of reductionistic reason of the eighteenth century, although that continues to hold an important place, but also more holistic forms of rationality.

• one aspect of this more holistic approach is a reevaluation of the role of the subject, and the subject-object relation in cognitive processes. From quantum mechanics to robotics to virtual reality and to psychology: naive prejudices about the subject as something given outside the system have hampered progress by letting methods introduce limits that otherwise could be transparent.

• one example of questioning limits is this: in a computer simulation, one can assign part of the system to play the role of subject (information gathering and processing system) with the rest of the system being the `objective world', of which the subject forms part while observing it. As a `super-observer', designing and observing the simulation, we can map, control, and interpret the flow of information across the membrane separating subject and object within the simulation. Such investigations may shed fresh light on the various problems mentioned above, from questions in quantum mechanics to questions concerning the human mind and society.

• a wider question, in which the previous questions are embedded, concerns the nature of foundations. Can we view the scientific framework we work in as being grounded firmly on a set of foundations? Or is the basis for doing science something more contingent, and perhaps even partly arbitrary?

• our provisional answers to this last question go in the direction of a study of `horizons' rather than foundations. The metaphor of a horizon points to a peculiar kind of limit, something that surrounds us and at the same time recedes whenever we might try to reach it. It is ever there, as a limit that is set not by some theoretical determination, but rather by the very structure of our perceptual situation.

• are we ultimately limited by the way the mind structures our perceptual situation? Are the universal limits to the operation of the mind? And how would they show up in our study of nature, and in our study of life and consciousness? These are the type of questions that we are currently learning to ask. As a preliminary answer, we would say that we may not be limited, if the constraints of the mind reflect the universal or possibly necessary constraints of the world.

• our main goal for the coming year is to further sharpen this way of questioning, through further reflection among ourselves as well as through widening dialogues with interested experts in scientific disciplines outside our own areas of expertise.

Specific Activities

Sept. 6-11, 1994: Piet Hut visited Roger Shepard at Stanford. A preparatory visit, to explore the various possibilities for our collaboration. Lunch with Pat Suppes, one of the participants of the SFI workshop on limits to scientific knowledge.

Oct. 5-14, 1994: Piet Hut visited Roger Shepard at Harvard. While Roger was visiting Harvard, to give the William James Lectures of 1994, Piet joined him to continue the previous discussions. Several meetings with graduate students in psychology, philosophy, and computer science.

Oct. 21-24, 1994: Piet Hut visited Greg Chaitin at Yorktown Heights, together with a graduate student in mathematical logic, Mark Ettinger, from the University of Wisconsin at Madison. Greg gave a tour through his toy lisp programs, and demonstrated their use in algorithmic information theory.

Nov. 10-13, 1994: Piet Hut visited Greg Chaitin at Yorktown Heights. We further explored Greg's `epistemological mathlab'. We wrote an explicit version of non-trivial formal axiomatic systems that produces an infinite number of theorems, and we also wrote one that exemplifies the converse of the incompleteness results for lisp program size complexity.

Nov. 13-17, 1994: Piet Hut visited Roger Shepard at Harvard. Dinner with the philosopher Willard Quine. We also met Ned Block, philosopher at M.I.T., as well as the group of graduate students that discussed Roger's lectures.

Dec. 7-13, 1995: Piet Hut visited Roger Shepard at Stanford. We prepared the February workshop, putting together a rough agenda, and contacted the prospective participants. Lunch with Pat Suppes and Ariela Lazar, also a philosopher at Stanford.

Feb. 6-8, 1995: Piet Hut attended the Goedel conference in Boston. A fascinating meeting on the philosophical views of Goedel, that recently became available through the publication of Goedel's unpublished notes, left behind after his death.

Feb. 24 - March 6, 1995: Piet Hut, Bob Rosen, Otto Rossler, and Mel Cohen visited Roger Shepard at Stanford, for a one-week workshop, with in-depth discussions covering a wide range of limits-related topics. Our main theme was `The Paradox of Limits'.

Apr. 6-9, 1995: Piet Hut visited Roger Shepard at Stanford. Together with Mel Cohen, we continue our discussions, around the central topic `The Paradox of Limits'.

May 1-3, 1995: Ron Bruzina visited Piet Hut at Princeton. In preparation for the meeting on `Limits to Scientific Knowledge' in Abisko, Sweden, philosopher Ron Bruzina from Lexington, KY, visited Piet to start writing a joint paper on `An Inquiry into Limits: From Foundations to Horizons'.

July 11-16, 1995: Ron Bruzina visited Piet Hut at Princeton. After the Abisko meeting, Ron visited Piet once more, in order to continue writing their paper `An Inquiry into Limits'.

July 22-27, 1995: Piet Hut visited Roger Shepard at Stanford. They continued their discussions, and met Brian Smith, a computer scientist with interest in limits, and Guven Guzeldere, from Stanford's Center of the Study of Language and Information.