Reflections on The Role of Models in Science
Note: This was originally written in response to an assignment for my course on the Collective Dynamics of Complex Systems at Binghamton University.
The Role of Models in Science (1945) by Arturo Rosenblueth and Norbert Wiener takes a wonderfully logical and systematic approach to reasoning about the role of models, and modelers, in science.1 For me, the assignment of this paper comes at the perfect time as I’ve been wrestling with fundamental questions about what types of models I’d like to create and why.
Up until this point, my research on blockchain systems (Bitcoin, Ethereum, etc.) has been heavily focused on the application of a methodology that demands a very deep and structured “engineering-style” analysis of systems as a hard prerequisite to the process of building models and then simulations.
It is an approach that might initially seem counter-intuitive in the face of the authors’ observation which reflects the common wisdom that “If a model has a more elaborate structure and is less readily amenable to experiment than the original system, it does not represent progress.”
On the one hand, I have found it rewarding to take such a rigorous approach to building an incredibly detailed and elaborate formal description of Bitcoin. Coming from a background in the social sciences, it has proved invaluable in deepening my technical understanding of a system that I previously understood primarily from an economic, political, and psychological perspective.
On the other hand, I’m finding that the process of analysis has not produced much clarity about what sort of models I want to build, what kinds of research questions I’m most interested in asking, or what types of experiments I want to run.
I’ve been working to zoom back out, step away from my “recreation” of Bitcoin, and reconnect with why I believe it is important for society to develop a deeper and more accessible scientific understanding of blockchain systems in the first place.
From this place of reflection, I found the work of Wiener and Rosenblueth to be inspiring, and thought-provoking. I’m asking myself various questions such as:
Am I more inclined towards the work of a theorist, focused on “moving up or down the scale of abstraction” or that of the experimenter, focused on “translating abstraction into experiment or vice-versa”?
Do the computational models I aim to build seem more like material models which represent simplified versions of complex systems with assumed similar properties, or formal models which represent symbolic/logical assertions of idealized simple situations. Are they a synthesis of the two?
In what specific ways do I aim to improve on existing approaches to modeling blockchain systems?
One of the most thought-provoking lines came at the very end of the paper, with the statement that “the main tool of science is the human mind and that the human mind is finite.” It is true that human intelligence still outpaces machine intelligence in many of the most important dimensions such as creativity. However, LLMs possess a tremendous and rapidly accelerating capacity to synthesize and summarize massive amounts of information in ways that greatly surpass the human mind.
I wonder if this development may require some re-assessment of the utility of building incredibly rich and detailed models in light of the powerful capacities of machine intelligence.
The only thing that I found myself missing from this paper was some insight into how the authors thought about the role of modeling in the social sciences. It’s an understandable omission given the time of its publication, but I know that Wiener went on to develop strong opinions on the utility of cybernetics in guiding society. I would have loved to have seen some of the authors’ thinking about how economists or sociologists at the time were using models.
This would be especially relevant for me given the nature of public blockchain systems as socio-technical ones whose emergent properties can only be properly understood by accounting for various interactions between unpredictable humans, rules enforced by rigid code, and the surrounding complex environment which includes the institutions and actors embedded in the global political economy.
Rosenblueth, A., & Wiener, N. (1945). The Role of Models in Science. Philosophy of Science, 12(4), 316–321.
I like how youʻre taking "a rigorous approach to building an incredibly detailed and elaborate formal description of Bitcoin" using George Mobusʻ model. As you point out, itʻs the first step. Wiener saw systems primarily as feedback loops and he worked with mathematical models. Modeling how a system changes through time (I guess that I should say "simulating" here) is an entirely different project. Algorithms in a computing environment offer all kinds of possibilities. Integration of various types–feedback loops, network/agent-based, etc.–is the big challenge.
Very nice excursion tour you've been on ... exploring the 1940's era when the processes of making behaviors, choices, and evaluation options ... alert self-awareness 'Reference Frames' themselves. Now it behooves you to appreciate the fuller context which gestated and birthed such intense creativity of "thinking and doing" during those focused years and decades (ca. 1890~1950's ; 1930-40's most intensely).
--- The World Stage was dealing with several co-present "competitions" at the same time: Polity organizations. Economic practices. Private vs Social priorities {everybody deserves noble inclusion -versus- "need to know" only. (!)} Struggles over "who" has decision authority over different parts of the Territories where humans live. Wealth production ~ wealth extraction ~ wealth utilizations. ... Information and Influences involved in Applications .. involved in improving "information and influences". Casual 'speed of social life activities" versus Ramped up activity ('early bird gets the worm" mentality competition).
--- Both major overt competitor groups juggling with these subsets of focus and performance .. the Axis and the Allies ... made different choices on priorities and goals even though handling the same topics to coordinate along the way. --- The Allies .. reacting to "first strike" actions by the Axis .. had to catch up and catch up FAST. Not only respond, but do it swiftly and competently when they were decades behind the curve the Axis groups initiated ca 1920+.
--- It was accomplished ca 1940-43 by bringing all the required components of social organization intensely together and synchronized under a Procedures, Policies, Actions MANDATE GROUP of Academics~Education, Finances, Industrial Production, narrow Rules of Citizens participation, Science tools and skills .. as initially available .. pressed to speedy creativity, improvements and novel implementations (overall: Improved Competence and Power to face and defeat the Axis (and then the Communist Dictatorship movements -- post WWII).
In the United States (fortunate in semi-isolation -- no destructive warring happen in American territories) .. one man was chosen to coordinate the Massive Effort: Vannevar Bush.
I will leave it to you Shingai, and your readership here, to learn about him, his life, and what he did ... as a man -not- in the limelight known by the public at large, but pivotal in designing, orchestrating and behind the scenes conducting of the totality of efforts .. especially the Intense Creativity .. imagined, and then brought to fruition in many fields and areas of human endeavors. They would prove required not only DURING the military & information struggles for survival in the active physical war years, but absolutely foundation setting for everything attended to and to be accomplished, improved, reached for, re-modeled, implemented and expanded .. AFTER that war.
It was subsequent to Bush's hidden design directorship of Allies (America primarily) of "how to win the struggles in all spheres of action" .. that we got .. cybernetics, computers, hyper information networks, NACA -->> NASA, AEC (atomic energy commissions), Fed Govt funding of intense academic research, Cybernetics, national roads network, Coordinated food production, rocketry and then space efforts, Lend-Lease relations with the worlds nations, NATO, United Nations, electrification of the nation, massive telephony connection of the nation, massive implementation of standardized (or close to it) Management Practices across the nation, national resources management and investment choices .. media intensification for commercial advertising and promotion of dedicated economics. .. Media intensification of information and forms of communicating together to not just Creative Arts --- but subliminally ... Social Propaganda aimed at population guiding and control. And not to be underrated or forgotten .. the "unfinished festering" of the 1860's Civil War (which the Union won militarily, but lost socially!) ;; human rights for all citizens :: states rights retained which undermined federal national authority over universal social policies ;; separation of church vs non-clerical social organization .. ethics, morals, rules, judiciary. bigotry over decency. personal liberties and rights and opportunities for all; ...
... anyway ... back on focus:
THE ... SEMINAL WRITING overseeing Rosenblueth and Wiener was openly presented to the American public and the world in 1945 by Vannevar Bush. He wrote "As We May Think" -- as the Allies warring efforts were winding down and a post-war society and situations had to make best use practices of the advances made during the war. The imminent 'peace years' -- needed a single clear enunciation of .... "things to come", of how to re-direct war efforts to peace efforts ... ((altruistically!!)).
You can read it here: Atlantic Monthly
https://www.theatlantic.com/magazine/archive/1945/07/as-we-may-think/303881/
It made waves when first published .. and co-distributed more broadly in LIFE Magazine.
But the limelight faded as "daily life" took people's attention, over conceptual considerations of the same things.
I'd be very anxious to read your thoughts and impressions of Bush's prognostications and recommendations, Shingai .. in a "PART II of "Reflections on the Role of Models in Science". And anyone else's here too! ... Jamie 01/31/2025