<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Gödelian Letters - Medium]]></title>
        <description><![CDATA[Adventures in the exotic wonderland of incompleteness and paradox - Medium]]></description>
        <link>https://medium.com/godelian-letters?source=rss----e48cbbf77efb---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Gödelian Letters - Medium</title>
            <link>https://medium.com/godelian-letters?source=rss----e48cbbf77efb---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 09:17:31 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/godelian-letters" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Moral Economics vs. Economic Morality]]></title>
            <link>https://medium.com/godelian-letters/moral-economics-vs-economic-morality-aa1fe7038af6?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/aa1fe7038af6</guid>
            <category><![CDATA[social-norms]]></category>
            <category><![CDATA[alvin-roth]]></category>
            <category><![CDATA[mechanism-design]]></category>
            <category><![CDATA[morality]]></category>
            <category><![CDATA[economics]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Sat, 30 May 2026 12:28:11 GMT</pubDate>
            <atom:updated>2026-05-30T12:32:06.256Z</atom:updated>
            <content:encoded><![CDATA[<p>A critical review of Nobel laureate Alvin E. Roth’s book, <em>Moral Economics: From Prostitution to Organ Sales, What Controversial Transactions Reveal About How Markets Work</em> (Basic Venture, May 12, 2026), 368 pages.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AkORIBtwhiAoeVJW0vmx6A.png" /><figcaption>Alvin Roth, 2013. Source: <a href="https://english.pku.edu.cn/">Peking University</a>.</figcaption></figure><h4>An Exceptional Occasion</h4><p>Today marks an exceptional occasion, addressing an exceptional topic, and starring an exceptional economist.</p><ul><li>It is an exceptional occasion because it marks the first book review published on <strong><em>Gödelian Letters,</em></strong><em> </em>and it stands as our most extensive one to date.</li><li>It addresses an exceptional topic: the intersection of Morality and Economics. In an era defined by institutional and moral crises, this is a particularly timely topic.</li><li>Our star is an exceptional economist: Alvin Roth, an Engineer, Market Designer, and Nobel laureate.</li></ul><h4>A Brief <a href="https://www.nobelprize.org/prizes/economic-sciences/2012/roth/biographical/">Biography</a></h4><ul><li>Alvin E. Roth was born on December 18, 1951, in Queens, New York. He entered Columbia’s engineering school at age 16 before earning his Ph.D. at Stanford in Operations Research in 1974.</li><li>He became interested in matching markets, especially labor markets where prices alone do not clear supply and demand. Roth analyzed the U.S. medical residency match and showed how the deferred acceptance algorithm could produce stable matchings while giving strong incentives for truthful preference reporting.</li><li>At Harvard and later Stanford, he applied matching theory to school choice systems in New York and Boston, as well as to markets for law clerks and medical specialists.</li><li>He played a leading role in designing <a href="https://www.bbc.com/news/business-50632630">kidney exchange programs</a>, developing algorithms that enabled life-saving paired kidney donations and donor chains.</li><li>In 2012, Roth was awarded the Sveriges Riksbank <a href="https://www.nobelprize.org/prizes/economic-sciences/2012/summary/">Prize</a> in Economic Sciences in Memory of Alfred Nobel, shared with <a href="https://www.nobelprize.org/prizes/economic-sciences/2012/shapley/biographical/">Lloyd Shapley</a>, for the theory of stable allocations and the practice of market design.</li><li>Currently, he is a <a href="https://economics.stanford.edu/people/alvin-roth">Professor of Economics</a> at Stanford University and a <a href="https://siepr.stanford.edu/people/alvin-roth">Senior Fellow</a> at Stanford Institute for Economic Policy Research (SIEPR).</li></ul><p>More details can be found in:</p><ul><li><a href="https://profiles.stanford.edu/alvin-roth">https://profiles.stanford.edu/alvin-roth</a></li><li><a href="http://marketdesigner.blogspot.com/">http://marketdesigner.blogspot.com/</a>.</li></ul><h4>Moral Economics</h4><p>Alvin Roth’s <a href="https://www.hachettebookgroup.com/titles/alvin-e-roth/moral-economics/9781668658161/?lens=publicaffairs"><em>Moral Economics</em></a><em> </em>arrives with the authority of a Nobel laureate who genuinely changed the world.</p><p>Roth is among the growing number of economists who have ventured into the domain of morality and economics. See this list of <a href="https://medium.com/godelian-letters/selected-books-on-ethics-morality-and-norms-3fb54995790d">selected books</a> in this area.</p><p>In <em>Moral Economics</em>, Roth adds a distinctive voice to this literature by applying the tools of market design to some of society’s morally contested transactions — from prostitution to organ sales. He treats these as “repugnant transactions,” cases in which willing participants face opposition from third parties. He asks how different legal and regulatory regimes perform in terms of consequences, trade‑offs, and social support.</p><p>The book focuses on “morally contested markets” — transactions that some people wish to engage in, but others want to ban on moral or religious grounds. Roth labels such transactions “repugnant” when all direct participants are willing, yet third parties object and seek legal prohibition, often without experiencing direct, observable harm themselves; he distinguishes these from classic externality cases such as pollution, which he treats as involving “obvious harms” and therefore outside his main focus. His aim is to understand how moral objections shape markets and to explore how law and market design can respond.</p><p>Methodologically, Roth advocates an empiricist, benefits‑and‑costs trade‑off approach. He wants debates about repugnant markets to move beyond purely moral denunciations toward evidence‑based evaluation of alternative legal regimes, including experiments and cross‑country comparisons where possible.</p><p>Behind <em>Moral Economics</em>, one senses the confidence of a highly successful market designer: the hope that techniques which helped fix residency matches and kidney exchanges might also tame morally contested markets. This may well be the book’s greatest strength, but it is also, I will argue, its most vulnerable side. Since Roth’s stated aim is to “take moral and ethical concerns seriously,” this extended review tries to take the book equally seriously.</p><h4>Key Points of the Review</h4><ul><li><em>Roth’s tradeoff framework is itself morally contested. Yet it is not subjected to the same scrutiny that he applies to contested issues discussed in the book.</em></li><li><em>Roth excludes certain externalities from the discussion because they involve “obvious harms,” while admitting others as morally contested. However, the claim of “obviousness” is itself contestable.</em></li><li><em>Roth’s framework essentially calls for replacing norms with government policies, risking the destruction of trust — the very foundation of any successful policy.</em></li><li><em>The book’s analysis remains largely reductionist: it underplays collective-action problems and treats the commodification of human organs as a policy option. In such a reductionist world, dignity has no intrinsic value.</em></li><li><em>The book is highly selective in presenting case studies and empirical evidence consistently favoring its overall narrative and largely leaving contrary evidence out of view.</em></li><li><em>Emotions play a strategic role as commitment devices. Commodification implies detaching emotions from the underlying activities, thereby threatening the very commitments required for successful policies.</em></li><li><em>Given the utilitarian and reductionist leanings of the book, it would be more accurately titled “</em><strong><em>Economic Morality” </em></strong><em>rather than</em><strong><em> “Moral Economics”.</em></strong></li></ul><p>Let us now get into the details.</p><h4>Literature Deficit</h4><p>Among the early signs that catch a careful reader’s attention in <em>Moral Economics</em> is its disconnection from the rich existing literature on the topic. Specifically, most of the <a href="https://medium.com/godelian-letters/selected-books-on-ethics-morality-and-norms-3fb54995790d?postPublishedType=repub">books</a> on ethics, morality, and norms, several of which are by Nobel laureates, are not cited (the two exceptions are Sandel, 2012, and Satz, 2010). This is despite the fact that Roth explicitly presents his book as an attempt to think systematically about “moral economics.”</p><p>It is also striking that <em>Moral Economics</em> never mentions the <strong>Prisoner’s Dilemma</strong>, <strong>Social Dilemmas</strong>, or the <strong>Tragedy of the Commons</strong>, even though they are at the center of economic discussions on ethics and morality. This is especially true in game theory and experimental economics, areas in which the author has extensive expertise. These concepts are directly relevant to the book since many of the “morally contested problems” are variants of these very dilemma structures, as we shall see.</p><p>Although the idea of “repugnant markets” is Roth’s own contribution, it is by no means the first serious attempt to address the complex relationship between ethics, morality, and economics. We will see that this literature deficit recurs across many of the diverse topics the author engages with throughout the book.</p><h4>Reductionism</h4><p>The book implicitly adopts the worldview common to neoclassical economics: the world is essentially linear; social activities can generally be decomposed into separate components, each of which can be commoditized without significant impact on the whole. This is known as the <em>reductionist view</em>.</p><p>For example, in the organ‑market debate, the reductionist move is to treat a kidney as a market commodity. But human organs are not free‑floating commodities: they are integral to human life. If human life <em>as a whole</em> embeds unpriceable “dignity,” as we shall argue later, then human organs, in principle, are inseparable from the whole. Adopting a general rule that treats human organs as separable commodities disregards priceless human dignity.</p><p>In science, reductionism is, in many respects, an outdated view of the 20th century. An alternative worldview is that the world is essentially nonlinear (Waldrop, 2019). Physicists <a href="https://www.nobelprize.org/prizes/physics/1977/anderson/biographical/">Philip Anderson</a> and <a href="https://www.nobelprize.org/prizes/physics/1998/laughlin/biographical/">Robert Laughlin</a> won their Nobel Prizes, in part, for their pioneering work on <em>emergence </em>(Laughlin, 2005). This means that, in general, <em>the whole is different from the parts.</em></p><p>Nature is highly nonlinear, and there is no reason <em>a priori</em> to assume that the social world is linear; if anything, it would be even more complex than nature, as Richard Feynman (1955) and others pointed out.</p><p>Roth, of course, is free to choose the worldview he accepts. But he is not free to gloss over this matter and dismiss the alternative worldviews. This runs counter to the neutrality the author boasts of in the Introduction.</p><h4>Why Do We Need Social Norms?</h4><p>Roth approaches morally contested markets with the tools of match‑making market design. Given his stellar success in solving matching problems in kidneys and other markets, it would be natural to explore how the same tools can resolve the tensions in morally contested markets.</p><p>Beyond the familiar “<a href="https://en.wikipedia.org/wiki/Law_of_the_instrument">Maslow’s hammer</a>” problem, this strategy invites logical inconsistencies and risks defective policy design. To see this, we have to keep in mind that <em>there is no engineering without science</em>.</p><blockquote>A good engineer is like a good physician.</blockquote><p>A good physician will not prescribe medication or recommend surgery before understanding how the body normally functions and how the proposed treatment is meant to restore and enhance that function. Likewise, a good engineer will not reconstruct or redesign a system before understanding its normal operation, so that any intervention can genuinely restore or enhance its performance.</p><p>By analogy, a “moral engineer,” the kind of market designer that Roth envisions in his book, should not recommend redesigning a “repugnant market” before understanding the normal role of the involved social norms and moral commitments in the first place. Otherwise, the intervention risks treating symptoms in the market while damaging the very normative structures that make collective life possible.</p><p>But Roth offers little if any explanation of the normal functions of social norms. Instead, he considers social norms as <em>external constraints. </em>He attributes them to “religion” as if this is an unexplainable cause. But if people are rational, they would choose their values, religious or otherwise, rationally. If they were not rational, we would need to explain why they are not and how they could develop shared values despite being irrational. So, we are back at square number one: Why and how do these norms arise in the first place? Although Roth repeatedly mentions “social norms” in the text, no single reference in the Notes section is about “social norms.”</p><p><a href="https://medium.com/godelian-letters/selected-books-on-ethics-morality-and-norms-3fb54995790d">Selected Books on Economics, Ethics, and Norms</a></p><p>Here is a compact summary of the functions of social norms:</p><ol><li>They are essential for building and sustaining trust among community members (Fukuyama, 1995; Ho, 2021). Without trust, no economy can operate normally.</li><li>Based on trust, norms help sustain coordination and cooperation among members (Ben-Ner &amp; Putterman, 1998; Elster, 1989).</li><li>The immediate result of the two functions above is to reduce transaction costs, including the costs of monitoring, contracting, and enforcement. Social norms, including moral and ethical codes, work to compensate for market failure, as Ken Arrow (1969) points out.</li></ol><p>Collectively, social norms represent the <strong>social capital </strong>of the economy.</p><h4>Economics of Tragedies</h4><p>Social capital is arguably the most important common resource of the economy. It is a public good that cannot be produced through competitive markets.</p><p>There is a well-known problem in economics called the “<a href="https://en.wikipedia.org/wiki/Tragedy_of_the_commons">Tragedy of the Commons</a>.” It describes a situation where individuals, pursuing their own self-interest, overuse a shared resource (“commons”), ultimately depleting it. This leads to outcomes that are detrimental to the collective good, despite rational individual actions. <em>Local individual rationality leads to collective group irrationality</em>.</p><p>It has the same structure as the <a href="https://en.wikipedia.org/wiki/Prisoner%27s_dilemma">Prisoner’s Dilemma game</a>. These problems are so common in social sciences that they are called <a href="https://en.wikipedia.org/wiki/Collective_action_problem">Social Dilemmas</a>, also known as the Collective Action Problem. They are at the heart of economic discussions of ethics and morality. As pointed out earlier, Roth’s book is completely silent on all these concepts.</p><p>Once we recognize that the structure of a certain setting is that of Prisoner’s Dilemma, waiting for “sufficient reliable data” before acting would be irresponsible. By the time the data confirm the prediction beyond doubt, the pasture is barren, the fishery has collapsed, or the climate has irreversibly shifted. In such situations, the “more data first” methodology is exactly what a responsible economist should reject.</p><p>Roth, however, repeatedly insists that policy toward “repugnant markets” should be grounded in ex post comparisons of different legal regimes, and that we should avoid evidence‑free moral objections. He calls for “high‑quality evidence” and “sufficient data” to evaluate the consequences of bans versus legalization and regulation.</p><p>Climate skeptics rely on the same argument <em>structure</em> that Roth adopts, despite holding the opposite position: they demand more and better data before acting; they treat early‑warning models as speculative until validated ex post, etc. For them, the harm is not “obvious” nor easily measurable. Roth’s criterion can be turned against him using his own methodology.</p><p>Roth excludes such externalities from the book while admitting others on the grounds that the former entail “obvious harms.” But he doesn’t answer the question: obvious to whom and by what standard? By unilaterally deciding which harms are self-evident and which are “contested constraints,” he relies on his own unstated moral baseline. But this moral baseline was not determined by the evidence-first methodology he advocates in the book. This shows the inconsistency of the argument: Evidence is required for <em>some</em> norms that were selected <em>a priori</em> without evidence.</p><p>The same argument structure is invoked in the debates on the two worst economic crises in recent history: the Great Depression and the Global Financial Crisis. In a 2010 interview (Cassidy, 2010), Nobel laureate <a href="https://www.nobelprize.org/prizes/economic-sciences/2013/fama/facts/">Eugene Fama</a> — father of the Efficient Market Hypothesis — remarked about the Global Financial Crisis (GFC):</p><blockquote>“There’ll be a lot of work trying to figure out what happened [in the GFC] and why it happened, but we’ve been doing that with the Great Depression since it happened, and we haven’t really got to the bottom of that.”</blockquote><p>For eighty years, economists have been “collecting data” and trying to understand the Depression; another eighty may pass before they “get to the bottom” of the 2008 crisis. For those economists, measuring the externality costs of these crises is not “obvious” at all and might take forever before getting “to the bottom” of them.</p><p>A pro-market response to repeated financial crises and economic depressions often exhibits the same structural move as that of climate skeptics: when systemic harms threaten the framework’s policy recommendations, the response is to demand more data, treat causal claims as speculative, and defer action until the evidence is conclusive. A “moral skeptic” would follow the same strategy toward values inconsistent with their framework. This is not how science is supposed to work.</p><h4>Limits to Market Design</h4><p>Are there limits to market design? This might seem a strange question. But the limits are what distinguish science from non-science. As <a href="https://en.wikipedia.org/wiki/John_D._Barrow">John Barrow</a> (1998) states:</p><blockquote>“Science is only possible because some things are impossible.”</blockquote><p>Probably the first to warn of such limits was Garrett Hardin in his classic 1968 essay, “The Tragedy of the Commons.” He states that “the population problem has no technical solution; it requires a fundamental extension in morality.” Interestingly, the population problem that Hardin refers to is another example of a “morally contested problem,” overlooked by Roth. So, when Hardin says it has “no technical solution,” he is identifying a class of problems that cannot be resolved by engineering and incentives alone. Contrary to Roth’s approach, a “moral extension” is essential for solving this class of problems.</p><p>Why?</p><p>The Tragedy of the Commons problem has the same structure as a multi-person Prisoner’s Dilemma game. The Prisoner’s Dilemma exhibits a self-defeating structure at the individual level: each player’s dominance reasoning, applied step by step, recommends defection (<em>D</em>) regardless of the opponent’s choice. Yet, when both players choose to defect, the resulting outcome (<em>D</em>, <em>D</em>) yields each player a strictly lower payoff than mutual cooperation would.</p><p>The locally valid reasoning produces a globally inferior outcome for each individual considered separately. In one-shot games played by self-interested players, a <em>Pareto-optimal outcome is algorithmically unattainable</em>: no mechanism that respects each player’s local dominance reasoning can deliver the outcome that each player would individually prefer, because the local reasoning is itself what blocks the globally preferred outcome. As is well known in game theory, the Prisoner’s Dilemma illustrates the inevitable tension between <em>individual </em>and <em>group rationality</em>.</p><p>Hence, in such games, <em>no amount of clever market design can substitute for the moral extension</em> as Hardin points out<em>.</em> And since many morally contested problems are variants of the Prisoner’s Dilemma game, as we shall argue, it follows that these games cannot be solved by treating moral values as mere constraints to be relaxed or traded off at the margin, as <em>Moral Economics</em> often suggests.</p><p><a href="https://www.nobelprize.org/prizes/economic-sciences/2009/ostrom/facts/">Elinor Ostrom</a> won the 2009 Nobel Prize for her work on “governing the commons.” That is, for innovative designs attempting to solve the “collective action problem” inherent in the Tragedy of the Commons. As pointed out earlier, Roth mentions nothing about the Tragedy or designs for governing the commons.</p><p>Instead, Roth defines a “repugnant transaction” as one in which all direct participants want to engage in it, while third parties oppose it on moral or religious grounds, even though they are not directly harmed. Structurally, that definition places individual rationality at the center of market design and excludes or undermines considerations of <em>collective rationality</em>.</p><p>To see how a morally contested problem can be modeled as a Prisoner’s Dilemma game, consider Roth’s own signature domain — kidney transplantation — we can see a Prisoner’s Dilemma game hidden beneath the surface. Players face two options:</p><ul><li>Cooperation (<em>C</em>): Donate organs.</li><li>Defection (<em>D</em>): Sell organs.</li></ul><p>Why would mutual cooperation be superior to mutual defection? Because it preserves the dignity of the community members. (<em>C</em>, <em>C</em>) reflects reciprocal respect and conserved dignity. But when one party makes a sacrifice while the other makes money (<em>C</em>, <em>D</em>) or (<em>D</em>, <em>C</em>), it will be a bad outcome for the cooperator. The defector makes money while demanding to be treated with respect by the rest of the community — he or she wants to have the cake and eat it too. But if everyone reasons this way and switches to <em>D</em>, the norm of reciprocity collapses, and everyone loses their dignity and mutual trust.</p><p>Nobel laureate Ken Arrow argued that giving people two options, to donate or to sell, is better than forcing them to donate (see Sandel, 2012). Why would expanding the available options make things worse? This is the reductionist neoclassical framework in which options are linearly separable, and no feedback mechanism from one choice to the other affects the final outcomes.</p><p>As Sandel (2012) and Bowles (2016), among others, explain, there is a “crowding out” effect. The decision to donate presupposes a <em>reciprocal relationship</em> between the community members. I donate on the understanding that, if I or any of my close family members is in dire need, another member will donate. But if some members choose to sell, the reciprocal logic collapses. The choice set is inherently nonlinear and hence irreducible to separable choices.</p><p>The beauty of the kidney exchange system that Roth pioneered with others is that it respects this reciprocal logic, thereby greatly reducing the risk of crowding out.</p><blockquote>Roth’s exchange system combines the best of both worlds: reciprocity and matching.</blockquote><p>The commercial market system for organs, which Roth surprisingly seems to support in <em>Moral Economics</em>, destroys the logic of reciprocity.</p><h4>Design Impossibility Theorems</h4><p>Hardin is not the only one to announce an “impossibility theorem” related to market design. The <a href="https://www.nobelprize.org/prizes/economic-sciences/2007/summary/">2007 Nobel Prize</a> was awarded to three game theorists: <a href="https://www.nobelprize.org/prizes/economic-sciences/2007/hurwicz/facts/">Leonid Hurwicz</a>, <a href="https://www.nobelprize.org/prizes/economic-sciences/2007/maskin/facts/">Eric S. Maskin</a>, and <a href="https://www.nobelprize.org/prizes/economic-sciences/2007/myerson/facts/">Roger B. Myerson</a> for their foundational work, among other things, in mechanism design, of which market design is an applied branch.</p><p>While traditional game theory takes the rules of a game as given and predicts how strategic players will behave, mechanism design flips this approach: it starts with a desired outcome and works backward to engineer the rules and incentives that will induce or motivate self-interested players to produce that exact outcome.</p><p>The Nobel Committee’s <a href="https://www.nobelprize.org/uploads/2018/06/advanced-economicsciences2007.pdf"><em>Report on Mechanism Design</em></a> surveys a number of impossibility theorems — negative results showing that certain combinations of goals (such as Pareto optimality, incentive compatibility, and voluntary participation) cannot be achieved simultaneously under reasonable assumptions. From a scientific point of view, these impossibility results are not embarrassing anomalies; they are constitutive of the field because they delineate what can and cannot be done by any mechanism.</p><p>To these design‑theoretic limits, we can add the more foundational limits implied by Gödel’s and Turing’s theorems and related results in mathematical logic (see da Costa and Doria, 2014; Doria, 2017). These show that even in formal systems and using ideal computing (Turing) machines, there are well‑defined boundaries beyond which completeness, consistency, or decidability cannot be guaranteed.</p><p>Why are these limits important? Because they anchor both science and engineering. There is no serious engineering without a prior science of what is and is not achievable, and there is no serious science without accepted limits to what its methods can deliver.</p><blockquote>No engineering without science, and no science without limits.</blockquote><p>The list of identified limitations on what markets and mechanisms can do has grown steadily over the years. One might therefore expect <em>Moral Economics</em> to ask how social norms can be brought into market design as constructive resources for creatively navigating these limits. Instead, the book devotes substantial space to arguing why and how we should relax “ancient repugnancies.”</p><h4>Norms, Policies, and Design</h4><p>Roth’s framework effectively suggests that we ultimately translate contested moral values into government policies. But he fails to ask: On what bases would the same citizens appoint that government in the first place? What norms and values would guide their choices? What procedures would they agree to enact those policies?</p><p>A major function of social norms is to build social trust, as indicated earlier. This greatly reduces the costs of enforcement (den Butter &amp; Mosch, 2003). Trust is especially valuable for market design and contract structure (D’Acunto et al., 2022). Trust functions as an informal enforcement mechanism that reduces transaction costs, enables cooperation under incomplete contracts, and supports long-term exchange, as Benjamin Ho (2021) points out.</p><p>By replacing norms with policies, we risk destroying trust, the most valuable capital in the economy. This would create more problems than we aim to solve.</p><blockquote><em>Norms should be replaced by norms, not by policies</em>.</blockquote><p>If people are rational and they select the existing set of norms rationally, then market designers (to adopt Roth’s terminology) should cultivate an alternative set of rational norms that appeal to the same group of people. This is why understanding the nature and emergence of norms is essential for a successful market design.</p><p>Roth’s framework, however, tends to lean on the power of government policy to reshape behavior while treating underlying norms as exogenous constraints to be navigated rather than as objects of explicit normative design. This sits uneasily with an economic tradition that prides itself on free choice and voluntary exchange. Sustainable policies ultimately depend on the voluntary acceptance of new norms rather than on regulatory force alone.</p><h4>Policy Evaluation and Lucas Critique</h4><p>There is a deeper problem with the policy approach that Roth advocates.</p><p>Roth argues we should compare the data on legalizing, say, same-sex marriage, vs. banning them. Roth ignores the fact that the values on which people contest or support these repugnant transactions are deeply ingrained in their preferences. <em>When such deep preferences change, the data-generating process itself changes.</em> Comparing two communities with inherently different parameters is vulnerable to the well-known <a href="https://en.wikipedia.org/wiki/Lucas_critique">Lucas Critique</a>.</p><p>In 1976, Robert Lucas, a 1995 <a href="https://www.nobelprize.org/prizes/economic-sciences/1995/lucas/facts/">Nobel Prize</a> winner, shocked the economics world with a simple observation: policy interventions affect not only the costs and benefits of citizens’ actions, but also their expectations about the future behavior of others. A new economic policy, he argued, is not an intervention in a frozen model of how the economy works, but rather a change in the workings of the model itself (Bowles, 2016). The “data” that supported the old model were generated under conditions that the new policy will alter. Using that data to forecast the new policy’s effects is therefore epistemically circular. <em>The evidence presupposes the stability of exactly what the policy will disturb</em>.</p><p>The empirical regularities Roth relies on to design interventions — willingness to donate, to honor gift norms, to treat certain goods as outside the cash nexus — are not structural features of human behavior. They are equilibrium responses to a particular set of social norms and institutional regimes. Once the regime is reframed as open to cost-benefit revision, agents re-optimize, and the regularities that justified the intervention dissolve into the conditions the intervention created. <em>Evidence-based policy design, in domains where the evidence itself is a function of the policy, faces a problem that the discipline has known since 1976.</em></p><p>Intuitively, the Lucas Critique has a Gödelian flavor. Roughly speaking, Gödel’s second theorem warns that a sufficiently rich formal system cannot certify its own consistency from within; if it could, the system would thereby be inconsistent.</p><p>The Lucas Critique is an econometric cousin: If we treat the macro-econometric model, including the government, as a sufficiently rich self-modeling system, then asserting the model’s structural invariance from within encounters the same kind of self-reference problem that Gödel’s second theorem identifies in formal arithmetic. The model’s prediction of policy effects becomes part of the information environment in which agents act, and the resulting feedback undermines the parameter stability that the prediction presupposed.</p><h4>Sex, Intimacy, and Emergence</h4><p>In the standard neoclassical framework, the commercialization of human intimacy is treated as a straightforward exchange of services. Because the worldview is inherently reductionist, it assumes that the whole is merely equal to the sum of its parts; therefore, physical intimacy can be “unbundled” from emotional commitment, priced separately, and traded without systemic loss.</p><p>From a holistic perspective, however, the relationship between physical vulnerability, trust, and affection is highly nonlinear. Intimacy is a complex system characterized by emergence — the reality that the interaction of its parts creates a whole that cannot be deconstructed without destroying the thing itself. Detaching the physical act from its relational context requires overriding deeply ingrained cognitive and emotional structures.</p><p>The psychological and social costs of this detachment might be profound, and they are rarely distributed equally among participants. Studies show that women tend to be more attuned to emotional connection as a condition of a satisfying relationship, while men are more sensitive to sexual availability (Baumeister &amp; Vohs, 2004; Lippa, 2009; Wade &amp; Mogilsk, 2018). Within a reductionist framework, all emotional fractures can, in principle, be factored into the market price.</p><blockquote>The reductionist framework ultimately fails because it assumes a linear compensation can fix a nonlinear fracture.</blockquote><p>Surprisingly, none of these empirical studies on sex differences in emotional and sexual priorities is cited or discussed in <em>Moral Economics</em>, despite the author’s repeated calls for “high-quality evidence” to evaluate repugnant constraints. The empirical literature Roth does not engage is precisely the literature that would complicate the reductionist framework’s treatment of commercialized intimacy as a symmetric exchange of priceable services.</p><p>The same reasoning applies to motherhood. In a neoclassical world, motherhood can be linearly decomposed into its constituents without losses. In a complex world, motherhood is an emergent whole, and the systematic decomposition of motherhood to its constituents alters the very meaning of motherhood in ways that cannot be captured by simply reallocating and pricing its parts.</p><h4>Time Inconsistency and the Commitment Problem</h4><p>In standard economic models, self‑interested agents maximize utility at each point in time. Unless their intertemporal preferences are specially structured (exponential discounting), plans made earlier will not remain optimal later, so they cannot credibly commit to long‑term courses of action (see Hammond &amp; Zank 2014).</p><p>Although rationality is defined as consistency of choice, in time-extended transactions, the very nature of their rationality makes them inconsistent over time. For this reason, the problem is known as “time inconsistency” (it is closely related to the Lucas Critique discussed above). The time-inconsistency problem itself is a clear example of how individually consistent choices can lead to globally inconsistent behavior. It is an interesting instance of “the whole is different from the parts.”</p><p>In <em>Passions Within Reason</em>, Robert Frank (1988) argues that emotions (love, guilt, anger, etc.) can function as <strong>commitment devices</strong>, leading people to keep promises or forgo short‑term gains even when it is not in their immediate material interest. This helps solve time‑inconsistency and cooperation problems: if others know you are the kind of person who will stick to a promise out of loyalty or guilt, your promises become credible in ways that “cool rationality” cannot achieve (see also Peter and Schmid, 2007).</p><p>In a purely neoclassical, separable framework, emotions are often treated as noise or as preferences over outcomes; if commodification can “strip away” emotions and leave cleaner transactions, that might be efficiency‑enhancing. But if we take time inconsistency seriously, emotions are <em>functional</em>: they stabilize long‑run commitments (marriages, partnerships, norms of fair dealing). Removing or undermining those emotional ties can degrade the very commitment mechanisms required for successful policy implementation.</p><h4>Sex Trafficking &amp; Jeffrey Epstein</h4><p>On a reductionist view, detaching sex from its relational context is, in itself, neutral, and any cost can be priced into the wage. A complexity view insists that the detachment is itself costly. Performing sexual acts in a deliberately de‑intimate, commercialized mode requires overriding cognitive and emotional structures that evolved and were socialized to link sex with trust, affection, and commitment.</p><p>That override is possible, but it is not free. It limits how many people can or will perform commercial sex voluntarily, even at higher prices, and it inflicts psychological and social costs that the standard framework fails to register. The above analysis leads to an important conclusion:</p><blockquote>The supply of voluntary commercial sex is therefore <strong>inelastic</strong> at the levels of demand that legalization generates, and the inelasticity creates the gap that trafficking fills.</blockquote><p>The empirical pattern, established most prominently by Cho, Dreher, and Neumayer (2013) and by other <a href="https://orgs.law.harvard.edu/lids/2014/06/12/does-legalized-prostitution-increase-human-trafficking/">studies</a>, is that countries with legalized prostitution have higher inflows of human trafficking than countries with criminalized regimes, controlling for the relevant confounders. The effect is substantial and robust across specifications.</p><p>Germany and the Netherlands, the two large European countries that adopted liberal legalization regimes, became significant destination countries for trafficked women, with the trafficking flows scaling with the size of the legal markets. Sweden, which adopted the alternative approach of criminalizing buyers while decriminalizing sellers, saw trafficking flows decline. The pattern is not what the substitution logic predicts; it is closer to the opposite. Something is happening that the standard framework does not capture.</p><p>Although Roth discusses sex trafficking at length, the book omits any mention of the most famous sex trafficking scandal in the last 100 years, that of <a href="https://en.wikipedia.org/wiki/Jeffrey_Epstein">Jeffrey Epstein</a>. The scandal has probably more data than any other dataset, something the author keeps insisting on, and yet, when this particular data is available, the book is completely silent on it. At the same time, it documents other cases in surprising detail.</p><p>The federal proceedings, the 2021 Maxwell conviction, the unsealed Doe documents, and the extensive victim testimony together constitute one of the most thoroughly documented evidentiary records of trafficking in modern history. Moreover, it is one with direct bearing on Roth’s framework, since the case shows how the language of consensual “arrangement” and “transaction” can be deployed to launder coercion under conditions of extreme power asymmetry. A methodology that proposes evidence-based regulation of prostitution-adjacent markets owes its readers an engagement with the hardest available evidence about how such frameworks are exploited in practice.</p><p>Whatever the reason for the silence, its effect is to spare the book’s framework its sharpest test, and to reinforce the pattern, as we shall see, of empirical attention that is unevenly distributed across the cases that would confirm the book’s program and the cases that would complicate it.</p><h4>Dignity</h4><p>Dignity was mentioned 22 times in <em>Moral Economics</em>, based on search results in the Kindle edition. Yet, not one of these instances includes a clear definition or characterization of the term. It is interesting to note that the way dignity is invoked throughout the text gives the distinct impression that it functions as a “moral absolute” (which should be the case). However, treating it as such is inconsistent with the author’s explicit statement in the Introduction that the book excludes “moral absolutes.” We again notice methodological selectivity in the way moral concepts are addressed.</p><p>Dignity and commodification are at odds with each other. A commodity, by definition, is for the consumption of human beings. It is subordinated to their will and choice. Dignity implies superiority and the freedom to make a choice. A commodity has no such freedom.</p><p>In a reductionist world, we can decompose a human being from conception to death into separable, commoditized components. So, for example, sex, genetic material, a fertilized egg, childbearing, child-rearing, and human organs can all be traded separately without affecting the others. The total value of a person, on this view, is just the sum of the values of the parts. Within such a framework, turning more and more aspects of human existence into commodities is not conceptually problematic; it is simply a matter of extending markets into new domains. In such a world, dignity is an empty term.</p><p>In a complex world, the whole is greater than the sum of the parts. And the difference is generally unquantifiable or, more accurately, “uncomputable” in Alan Turing’s terms (more on Turing later). In such a world, a human being is immeasurably more valuable than its parts. <em>This difference between the whole and the parts defines the dignity of a human being</em><strong>.</strong></p><p>Roth’s <em>Moral Economics</em> is written squarely inside the reductionist framework. Because he never defines dignity or confronts the possibility that some social meanings are uncomputable and non‑priceable, dignity becomes an empty term in this analysis — something others care about, but that cannot be given real weight without breaking the additive, commodifying structure of mainstream neoclassical economics. For many readers, this is exactly why his framework feels alien: concepts that are central to their moral experience of the social world cease to have substantive content once reduced to utilities over parts.</p><h4>The Irreducibility of Dignity</h4><p>There are good reasons, grounded in computability theory, to argue that emergent properties are, in general, uncomputable. Rice’s theorem, building on Alan Turing’s work on the Halting Problem, shows that, roughly speaking, non-trivial semantic properties of computer programs are generally undecidable (Moore &amp; Mertens, 2011).</p><p>That is, for any non‑trivial semantic property of a program’s output, there is no general algorithm that can decide, for an arbitrary program 𝑃 and input 𝐼, whether the output of 𝑃(𝐼) has that property. In other words, no single mechanical procedure can, in advance, extract all meaningful higher‑level properties from arbitrary lower‑level code.</p><p>How does this relate to emergence?</p><p>Suppose we have a sufficiently rich program 𝑋 that specifies in complete detail every part 𝑇 of some complex system 𝑆. If we define an emergent property 𝐸 of 𝑆 as a non‑trivial property of the output of 𝑋(𝑇), then by Rice’s theorem, 𝐸 is, in general, uncomputable. No general program 𝐻 can look at the code specifying the parts and systematically correctly decide whether the whole will possess 𝐸.</p><p>What about dignity?</p><p>Dignity, as I have argued, is precisely such an emergent property. It is central to the difference between a human being taken as a whole and the mere aggregate of her parts — organs, capacities, behaviors, preferences — that might be traded or optimized in a reductionist model.</p><p>Dignity, therefore, is an irreducible property of human beings. Hence, it is, in principle, uncomputable. There can be no general mechanical procedure that tells us, from a complete description of lower‑level facts, whether a person is being treated with dignity. Particular judgments about dignity are still possible, but they will depend on <em>non‑mechanical</em> forms of understanding — judgment, perception, moral reasoning — that are not reducible to an algorithm. (Recall Hardin’s non-technical solution to the Tragedy of the Commons.)</p><p>This reinforces the earlier conclusion: many central social phenomena — intimacy, motherhood, personhood, dignity — are emergent in precisely this sense. A reductionist strategy that tries to represent them as sums of computable, linearly separable parts, to be priced and traded in a market, is not just morally thin; it is, in a strict sense, methodologically inadequate. It asks mechanical procedures to decide questions that, in principle, lie beyond their reach.</p><h4>Alan Turing</h4><p>It is interesting how Roth discusses the tragic death of the brilliant mathematician <a href="https://en.wikipedia.org/wiki/Alan_Turing">Alan Turing</a> (1912–1954). He presents the standard narrative: Turing’s 1954 cyanide death as suicide, driven by the shame and chemical castration following his 1952 conviction for homosexuality. Roth then delivers the punchline, “Repugnance can kill.”</p><p>Here again, we encounter a highly selective presentation of evidence that contradicts the author’s stated commitment to fairly exploring all sides of a controversy.</p><p>Since at least 2012, leading Turing experts, most prominently <a href="https://hpi.uq.edu.au/profile/338/jack-copeland">Jack Copeland</a> (2012), have argued that the inquest evidence is insufficient to support a firm suicide verdict. Key points in that counter-narrative include:</p><ul><li>Turing was actively experimenting with cyanide at home (electroplating and other chemical work); cyanide solutions were bubbling in his lab.</li><li>The half-eaten apple beside his bed was never chemically tested.</li><li>He was found in bed with undisturbed covers (cyanide poisoning typically causes violent convulsions).</li><li>No suicide note, and those close to him (including his mother) believed it was accidental.</li><li>His nephew, Dermot Turing, has explicitly stated he does not believe the conviction or hormone treatment caused suicide.</li></ul><p>All this information has been publicly available since 2012. Roth, writing in 2026, had access to this 14-year-old debate yet chose the rhetorically more powerful narrative that directly serves his <em>a priori</em> stance against “repugnance.”</p><p>Even if we agree that Turing committed suicide, Roth uses the colorful expression: “Repugnance can kill.” Compare this to the way he describes “medical assisted dying” as “death with dignity.” Why should suicide be cast in one context as a “repugnant act,” while in another as a civilized “death with dignity”? This semantic asymmetry shows that the author does not present the two sides fairly as he promises in the Introduction.</p><h4>The Hottest Places …</h4><p><a href="https://www.jfklibrary.org/learn/about-jfk/life-of-john-f-kennedy/fast-facts-john-f-kennedy/john-f-kennedys-favorite-quotations-dantes-inferno">John F. Kennedy</a> and <a href="https://www.crmvet.org/docs/mlkviet2.htm">Martin Luther King Jr.</a> frequently cited a line popularly attributed to the Italian poet <a href="https://en.wikipedia.org/wiki/Dante_Alighieri">Dante Alighieri</a> (1265–1321):</p><blockquote>“The hottest places in Hell are reserved for those who in time of moral crisis preserve their neutrality.”</blockquote><p>Roth never mentions this phrase in <em>Moral Economics</em>, nor does he offer his readers any criterion for when economists ought to abandon neutrality. He essentially offers only one framework: identify contested markets, collect data from different regimes, and evaluate the evidence before recommending policy.</p><p>It is instructive to ask: What would the book’s position be if it were written in the early 1960s, at the height of the <a href="https://en.wikipedia.org/wiki/Civil_rights_movement">Civil Rights Movement</a>?</p><p><a href="https://en.wikipedia.org/wiki/Ronald_Coase">Ronald Coase</a> (1910–2013) published his famous work “The Problem of Social Cost” in 1960, on the eve of the Civil Rights Movement in the US. In its common form, the <a href="https://www.nobelprize.org/prizes/economic-sciences/1991/coase/lecture/">Coase Theorem</a> states that, if agents are rational and the costs of transacting are zero, resources will be allocated efficiently, independent of how rights over those resources are initially distributed (see Medema, 2020). The theorem assumes that whoever holds the right doesn’t affect the final outcome — only who pays whom.</p><p>The theorem quickly became a cornerstone of neoclassical economics, celebrating market-based solutions to almost all problems. Coase won the <a href="https://www.nobelprize.org/prizes/economic-sciences/1991/coase/facts/">Nobel Prize</a> in 1991, and his ideas were critical in developing carbon markets and other arrangements to address externalities through the market mechanism.</p><p>Applied to the context of segregation, if a store had the legal right to display a “Whites Only” sign, patrons would theoretically have to pool their money to pay the owner to take it down. Conversely, if patrons had a legally protected right to service, the owner would have to compensate them to allow the discrimination. In either case, the market clears.</p><p>Under perfect market conditions, therefore, basic human rights and civil liberties are irrelevant. <em>Dignity has no intrinsic value in an ideal neoclassical world</em>. Only in an imperfect world, riddled with transaction costs and asymmetric information, do such concepts gain any useful meaning.</p><p>Under Roth’s framework of neutrality, imposing an absolute ban on segregation would not be good policy. Rather, the Movement should have studied communities with and without segregation, collected reliable data, and performed a thorough analysis to propose <em>regulating</em> segregation rather than banning it outright.</p><p>Luckily, the champions of the Civil Rights Movement did not embrace the Coase Theorem or neoclassical economics. They held firm beliefs in the <em>indivisible dignity of mankind </em>— regardless of race, culture, or religion — without requiring empirical evaluation to establish that civil rights are absolute.</p><h4>Moral Crises and the Crisis of Economics</h4><p>The world economy has been living through a sequence of moral crises for more than two decades: the Enron and similar scandals in the early 2000s, the 2007–2009 Global Financial Crisis, and emerging-market debt crises in 2019, among others.</p><p>These were not (only) crises of mismanagement and lax supervision; they were, at a deeper level, motivated and defended by values related to free markets, innovation, mathematical rigor, rationality, and many others (see Bookstaber, 2017; Partnoy, 2009; Skidelsky, 2020).</p><p>Sadly, the same values continue to dominate mainstream economics today, despite repeated calls to revise and revamp the field. A book on “moral economics” is expected to address the root moral problems of mainstream economics. Unfortunately, it doesn’t.</p><h4>Conclusion: Moral Economics or Economic Morality?</h4><p>There are two ways to analyze the relationship between morality and economics:</p><ul><li><strong>Subject economics to moral values:</strong> start from invariant moral commitments (about dignity, rights, justice) and ask what economic choices and organizations are compatible with them.</li><li><strong>Subject moral values to economic analysis:</strong> treat moral values as external constraints, evaluate the consequences with and without them, and design policy based on the tradeoff analysis.</li></ul><p>The first approach envisions “Moral Economics” as the practice of bringing moral values to economics, not the other way around. While there is a rich intellectual tradition dedicated to this approach (see White, 2019), the book pays little attention to it.</p><p>Roth follows the second approach. His objective is not to correct market failures or limitations; instead, it is to navigate around “repugnant constraints.” The book, therefore, would be more accurately titled <em>Economic Morality</em>, as it calls for scrutinizing moral values through economic tradeoff analysis. In doing so, Roth’s proposed approach itself becomes a “repugnance constraint” — a norm he fails to subject to the very evaluation methodology he applies to others. The utilitarian approach that Roth leans toward is itself morally contested, yet he fails to subject it to the very methodology he demands of other “repugnant constraints.”</p><p>It is interesting to note that, in his 2015 book, <em>Who Gets What and Why</em>, Roth mentioned “market failure” 11 times. In his <em>Moral Economics</em>, it was never mentioned. From the author’s own perspective, <em>Moral Economics</em> may appear as a natural extension of his earlier work in market design. Vigilant readers would immediately sense a change in perspective, from creatively correcting markets to mostly celebrating them.</p><p>In the introduction to his book, Roth tells an impressive story about his ability to fairly represent morally contested problems, then he writes:</p><blockquote>“… in this book I’ll continue to try to present all sides of a controversy fairly. I’ll surely fall short of that in some cases. (Disputes about morality seem to make it particularly difficult for disputants to fully appreciate opposing positions.)”</blockquote><p>In the same spirit, and especially in light of his parenthetical remark, I may close with a parallel confession:</p><blockquote>In this review, I have tried to be objective in discussing the author’s approach, although disputes over methodological issues may make it particularly difficult for authors to fully appreciate the opposing positions.</blockquote><p><em>I am grateful to Prof. Alvin Roth for taking the time to read an earlier draft of this review and offering constructive comments. Naturally, his engagement in no way implies endorsement of the views, methodologies, or conclusions presented here. While this review adopts a perspective different from the one Roth might have preferred, I nonetheless greatly appreciate his characteristic</em> <em>generosity and intellectual openness. All errors, omissions, and biases remain entirely my own</em>.</p><h4>References</h4><ul><li>Arrow, K. J. (1969). The organization of economic activity: Issues pertinent to the choice of market versus non-market allocation. In R. H. Haveman &amp; J. Margolis (Eds.), <em>Public expenditures and policy analysis</em> (pp. 59–73). Markham Publishing.</li><li>Barrow, J. D. (1998). <em>The impossible: The limits of science and the science of limits</em>. Oxford University Press.</li><li>Baumeister, R. F., &amp; Vohs, K. D. (2004). Sexual economics: Sex as female resource for social exchange in heterosexual interactions. <em>Personality and Social Psychology Review, 8</em>(4), 339–363.</li><li>Ben-Ner, A., &amp; Putterman, L. (Eds.). (1998). <em>Economics, values, and organization</em>. Cambridge University Press.</li><li>Bookstaber, R. (2017). <em>The end of theory: Financial crises, the failure of economics, and the sweep of human interaction</em>. Princeton University Press.</li><li>Bowles, S. (2016). <em>The moral economy: Why good incentives are no substitute for good citizens</em>. Yale University Press.</li><li>Cho, S.-Y., Dreher, A., &amp; Neumayer, E. (2013). Does legalized prostitution increase human trafficking? <em>World Development, 41</em>(1), 67–82.</li><li>Coase, R. H. (1960). The problem of social cost. <em>Journal of Law and Economics, 3</em>, 1–44.</li><li>Copeland, B. J. (2012). <em>Turing: Pioneer of the information age</em>. Oxford University Press.</li><li>D’Acunto, F., Xie, J., &amp; Yao, J. (2022). <em>Trust and contracts: Empirical evidence</em> (LawFin Working Paper №32; last revised March 17, 2025). Goethe University, Center for Advanced Studies on the Foundations of Law and Finance (LawFin).</li><li>da Costa, N. C. A., &amp; Doria, F. A. (2014). On an extension of Rice’s theorem and its applications in mathematical economics. <em>Advances in Austrian Economics, 18</em>, 237–257.</li><li>den Butter, F. A. G., &amp; Mosch, R. H. J. (2003). <em>Trade, trust and transaction costs</em> (Tinbergen Institute Discussion Paper No. TI 2003–082/2). Tinbergen Institute.</li><li>Doria, F. A. (Ed.). (2017). <em>The limits of mathematical modeling in the social sciences: The significance of Gödel’s incompleteness phenomenon</em>. World Scientific.</li><li>Elster, J. (1989). <em>The cement of society: A study of social order</em>. Cambridge University Press.</li><li>Elster, J. (2015). <em>Explaining social behavior</em> (Rev. ed.). Cambridge University Press.</li><li>Feynman, R. P. (1955). The value of science. <em>Engineering and Science, 19</em>(3), 13–15.</li><li>Frank, R. H. (1988). <em>Passions within reason: The strategic role of the emotions</em>. W. W. Norton &amp; Company.</li><li>Fukuyama, F. (1995). <em>Trust: The social virtues and the creation of prosperity</em>. Free Press.</li><li>Hammond, P. J., &amp; Zank, H. (2014). Rationality and dynamic consistency under risk and uncertainty. In M. J. Machina &amp; W. K. Viscusi (Eds.), <em>Handbook of the economics of risk and uncertainty</em> (Vol. 1, pp. 41–97). Elsevier.</li><li>Hardin, G. (1968). The tragedy of the commons. <em>Science, 162</em>(3859), 1243–1248.</li><li>Ho, B. (2021). <em>Why trust matters: An economist’s guide to the ties that bind us</em>. Columbia University Press.</li><li>Laughlin, R. B. (2005). <em>A different universe: Reinventing physics from the bottom down</em>. Basic Books.</li><li>Lippa, R. A. (2009). Sex differences in sex drive, sociosexuality, and height across 53 nations: Testing evolutionary and social structural theories. <em>Archives of Sexual Behavior, 38</em>(5), 631–651.</li><li>Lucas, R. E., Jr. (1976). Econometric policy evaluation: A critique. <em>Carnegie-Rochester Conference Series on Public Policy, 1</em>(1), 19–46.</li><li>Medema, S. G. (2020). The Coase theorem at sixty. <em>Journal of Economic Literature, 58</em>(4), 1045–1128.</li><li>Moore, C., &amp; Mertens, S. (2011). <em>The nature of computation</em>. Oxford University Press.</li><li>Partnoy, F. (2009). <em>Infectious greed: How deceit and risk corrupted the financial markets</em> (Rev. and updated ed.). Henry Holt and Company. (Original work published 2003).</li><li>Peter, F., &amp; Schmid, H. B. (Eds.). (2007). <em>Rationality and commitment</em>. Oxford University Press.</li><li>Roth, A. E. (2009). What have we learned from market design? <em>Innovation Policy and the Economy, 9</em>, 27–42. MIT Press.</li><li>Roth, A. E. (2015). <em>Who gets what — and why: The new economics of matchmaking and market design</em>. Houghton Mifflin Harcourt.</li><li>Sandel, M. J. (2012). <em>What money can’t buy: The moral limits of markets</em>. Farrar, Straus and Giroux.</li><li>Satz, D. (2010). <em>Why some things should not be for sale: The moral limits of markets</em>. Oxford University Press.</li><li>Skidelsky, R. (2020). <em>What’s wrong with economics?: A primer for the perplexed</em>. Yale University Press.</li><li>Wade, T. J., &amp; Mogilski, J. (2018). Emotional accessibility is more important than sexual accessibility in evaluating romantic relationships — especially for women: A conjoint analysis. <em>Frontiers in Psychology, 9</em>, 632.</li><li>Waldrop, M. M. (2019). <em>Complexity: The emerging science at the edge of order and chaos</em>. Open Road Media.</li><li>White, M. D. (Ed.). (2019). <em>The Oxford handbook of ethics and economics</em>. Oxford University Press.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aa1fe7038af6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/moral-economics-vs-economic-morality-aa1fe7038af6">Moral Economics vs. Economic Morality</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Évariste Galois Meets Kurt Gödel]]></title>
            <link>https://medium.com/godelian-letters/%C3%A9variste-galois-meets-kurt-g%C3%B6del-bd40a77b2483?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/bd40a77b2483</guid>
            <category><![CDATA[gödel]]></category>
            <category><![CDATA[incompleteness-theorem]]></category>
            <category><![CDATA[group-theory]]></category>
            <category><![CDATA[galois]]></category>
            <category><![CDATA[symmetry]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Sat, 02 May 2026 11:21:04 GMT</pubDate>
            <atom:updated>2026-05-02T18:21:48.514Z</atom:updated>
            <content:encoded><![CDATA[<p><em>A striking harmony between Symmetry, Group Theory, and the Incompleteness Theorem.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VqpZgG4_Tg32PXMN7v00Yg.png" /><figcaption>Source: <a href="https://www.shutterstock.com/image-photo/sea-ocean-gifts-on-wooden-background-1873594636?trackingId=c6bb2260-17db-4e42-9647-564d85fa1597&amp;listId=searchResults">Shutterstock</a>.</figcaption></figure><h4>Her Majesty Symmetry</h4><p>Beauty is truth, writes mathematician <a href="https://godelianletters.org/interview-with-ian-stewart">Ian Stewart</a> (2007). The crown jewel of beauty is symmetry. Throughout the ages, symmetry allured philosophers, artists, poets, mathematicians, and scholars of all fields.</p><p>During the 20th century, symmetry became the bedrock of modern science. The era witnessed major revolutions:</p><ul><li>Theory of Relativity</li><li>Quantum physics</li><li>Incompleteness of mathematics.</li></ul><p>While the role of symmetry in the progress of relativity and quantum mechanics is well established, far less attention has been given to the link between symmetry and incompleteness. With notable exceptions such as <a href="https://en.wikipedia.org/wiki/Douglas_Hofstadter">Douglas Hofstadter’s</a> <em>Gödel, Escher, Bach</em> (1999), Gödel’s theorems have often been treated mainly as formal results in mathematical logic, rather than as windows into beauty, structure, and insight.</p><p>This article explores the conceptual resonance between group theory and incompleteness, arguing that invariance and undecidability share a surprisingly common pattern. Recognizing this deepens our understanding of both.</p><h4>Change without Change</h4><p>So, what is symmetry?</p><p>Something is considered symmetrical when it looks the same from more than one point of view. For instance, a starfish exhibits symmetry because its five identical arms allow it to look the same even after being rotated by 72° or its multiples. Thus, symmetry relies not just on an object having a set of similar parts, but also on the specific arrangement of those parts.</p><p>Following <a href="https://plato.stanford.edu/entries/weyl/">Hermann Weyl</a>, Nobel laureate <a href="https://feynman.com/">Richard Feynman</a> (1963, p. 322), defines symmetry as follows:</p><blockquote>“A thing is symmetrical if one can subject it to a certain operation and it appears exactly the same after the operation.”</blockquote><p>Ian Stewart (2007) offers the following definition:</p><blockquote>“A symmetry of some mathematical object is a transformation that preserves the object’s structure.”</blockquote><p>A transformation is called a symmetry if the object remains invariant under the change induced by such a transformation. The set of such transformations is called a “group,” and it is the primary focus of <strong>group theory </strong>(Carter, 2009). The immunity of a symmetric system to a set of changes is also called “invariance,” since the object preserves its essential features just as they were before the transformation. Group theory, therefore, studies how different systems can be invariant under a specific set of symmetric transformations.</p><p>Nobel laureate <a href="https://www.frankawilczek.com/">Frank Wilczek</a> (2015) remarks that “change without change is the essence of symmetry.” This has the flavor of Gödel’s famous sentence, “<em>I am not provable</em>.” More on this later.</p><h4>A Tale of Two Geniuses</h4><p>The young <a href="https://www.britannica.com/biography/Evariste-Galois">Évariste Galois</a> (1811–1832) was a visionary pioneer who made key breakthroughs in what became known as group theory. He was the first to make decisive use of groups to solve a major mathematical problem: determining when polynomial equations are solvable by radicals. Unfortunately, he was unable to publish his major works during his short lifetime. He was just 20 years old when he was fatally wounded in a duel in Paris on the early morning of May 30, 1832 (Livio, 2005).</p><p>Sensing his death, he stayed up late that night writing a long letter to his friend Auguste Chevalier. In the letter, he summarized his work and pointed toward broader applications of group theory beyond algebraic equations. Throughout the margins of his manuscript, he scribbled the haunting phrase:</p><blockquote>“Je n’ai pas le temps” (“I have no time”)</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uc4aKdlSuO_Hpcmdo9moiw.png" /><figcaption>Left: Évariste Galois; image produced by Grok 4. Right: Kurt Gödel; image enhanced by Grok 4.</figcaption></figure><p>Fourteen years later, in 1846, Galois’s major work was finally published by Joseph Liouville, a highly respected French mathematician,<strong> </strong>in the <em>Journal de Mathématiques Pures et Appliquées</em>. This was the moment Galois went from a “failed student” to being “one of the greatest mathematicians who has ever lived,” in the words of Ian Stewart (2023).</p><p><a href="https://godelianletters.org/remembering-godel">Kurt Gödel</a> (1906–1978) was 25 years old when he shocked the world of mathematics by publishing his incompleteness theorem in 1931. Unlike Galois, Gödel’s paper was an instant success — something very unusual for a completely original and radical result. Although his theorem became widely accepted, many thought that incompleteness is a marginal problem that arises only at the far edges of pure logic. We will see that this is unlikely to be true.</p><p>The personality of the two geniuses could hardly have been more different. Galois was rebellious, politically charged, impatient, and often disorganized; Gödel was reserved, methodical, and almost obsessively precise. Yet both reached extraordinary depths at a remarkably young age.</p><p>Both saw mathematics not as the study of numbers and shapes, but as the study of structure and process. Symmetry is the summit of such a view. From that height, the details fall away, and the landscape resolves into a beautiful tapestry.</p><h4>Invariance and Undecidability</h4><p>Consider a perfect circle like the one on the left below. Suppose we rotate it by 60°. Can we detect that change? Unless we mark the circle, we cannot. The circle looks exactly the same before and after the rotation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/910/1*PJYRi_5CrteodSiPuQWvtw.png" /><figcaption>Can you verify that the right circle is the left after being rotated by 60°?</figcaption></figure><p>The statement, “The circle has been rotated by 60°,” can neither be proved nor refuted simply by observing the circle. Only if we have an <em>external</em> mark or reference can we verify it. Without these external landmarks, symmetric transformations effectively hide their own tracks. Thus, broadly speaking, the statement “The circle has been rotated by 60°” would be <strong>undecidable</strong>; there is no finite algorithm capable of determining its truth from within (see Boolos et. al., 2007). The collection of all such undetectable transformations (like rotations of the circle) forms a group; this group is literally the catalog of changes the system itself is blind to.</p><p>This points to a surprising link with Gödel’s Incompleteness Theorem. Roughly speaking, Gödel’s First Incompleteness Theorem dictates that, in any consistent, effectively axiomatized formal system rich enough to encompass arithmetic, there will be sentences, commonly referred to as “Gödel’s sentences,” that remain undecidable <em>within</em> the system. They can be decided only from <em>outside </em>the system, i.e., from within a higher system with additional axioms. By comparing group theory and the incompleteness theorem, we may claim that:</p><blockquote>Invariance and undecidability are two sides of the same coin.</blockquote><p>This is an interesting link between the two domains, but there is more to come.</p><h4>A Bug or a Feature?</h4><p>Although the two geniuses lived nearly a century apart, their approaches are surprisingly complementary. Gödel showed how undecidable statements can arise in effectively axiomatized systems rich in arithmetic. On the other hand, symmetry is an operation that leaves a mathematical system invariant; hence, a statement describing the change cannot be verified <em>within the system</em>. The two frameworks, therefore, approach undecidable statements from two opposing directions.</p><p>Gödel arrives at undecidable statements from inside a formal system using sophisticated techniques (to be discussed shortly). For Gödel, these sentences are <em>anomalies</em> to systems that were supposed to be “complete,” hence the term “incompleteness.” Group theory, by contrast, seeks out systems whose transformations leave them invariant. Statements describing such transformations are naturally undecidable <em>within</em> such systems. In group theory, these undecidable statements are <em>features</em> rather than <em>anomalies</em> of their systems.</p><p>While the systems studied by Gödel and those analyzed in group theory may differ substantially, the nature of the unverifiable statements in each is strikingly similar.</p><p>Both Gödel’s theorem and group theory investigate <em>blind spots</em> — what a system inherently cannot see. Yet, they leverage this gap in opposite ways. One treats it as the ultimate limit of formal proof; the other elevates it to the primary object of study, utilizing a symmetry group to measure indistinguishability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ifaN3Ais_Gv4hMuPhk4Tag.png" /></figure><h4>Magic Mirrors</h4><p>We are familiar with mirror symmetry: If you look in the mirror, your right and left sides are swapped. A “magic mirror” (also called a “<a href="https://medium.com/@kvshan23/the-true-mirror-image-what-it-tries-to-tell-you-6d91560c7269">true mirror</a>”) reflects you exactly as others see you — no swapping. In mathematics, the analog of such a mirror is a <strong>fixed point</strong>: a point that the transformation leaves exactly as it is.</p><p>Specifically, a fixed point of a function <em>f </em>is <em>x</em>* such that <em>f</em>(<em>x</em>*) = <em>x*</em>. For example,</p><ul><li><em>f</em>(<em>x</em>) = 3<em>x </em>has a fixed point at zero: <em>f</em>(0) = 0.</li><li><em>f</em>(<em>x</em>) = <em>x</em>²<em> </em>has fixed points at 0 and 1: <em>f</em>(0) = 0 and <em>f</em>(1) = 1.</li><li><em>f</em>(<em>x</em>) = 2<em>x</em>﹣5<em> </em>has a fixed point at 5: <em>f</em>(5) = 5.</li><li>The center of a circle is a fixed point of any function that rotates it.</li></ul><p>Before turning to Gödel and Galois, let us examine a concrete application of fixed points, namely, in economics.</p><p>Nobel laureates <a href="https://www.nobelprize.org/prizes/economic-sciences/1972/arrow/facts/">Kenneth Arrow</a> and <a href="https://www.nobelprize.org/prizes/economic-sciences/1983/debreu/facts/">Gerard Debreu</a> applied fixed-point theorems to economics to prove the existence, in theory, of market-clearing prices (Jehle &amp; Reny, 2011).</p><p>In simple terms, we imagine a market with an auctioneer. In the early morning, the auctioneer announces a price <em>p</em>. The function <em>f</em> adjusts the price based on the gap between supply and demand. Excess demand pushes the price up; excess supply pushes the price down. At equilibrium, supply equals demand, and therefore, no adjustment is made to the price at that point. Hence, the equilibrium price <em>p</em>* is a fixed point of the adjustment function: <em>f</em>(<em>p</em>*) = <em>p</em>*. At this point, the market becomes a “magic mirror” of the equilibrium price. In reality, of course, markets rarely reach such an equilibrium, but that is a <a href="https://godelianletters.org/market-reality-vs-economic-theory">different story</a>.</p><h4>Enter Gödel</h4><p>Gödel ingeniously engineered fixed-point machinery within formal logic to prove his Incompleteness Theorem. He created what is now known as the <strong>Diagonalization Lemma </strong>(Smullyan, 1994).</p><p>Roughly speaking, the Lemma states that in any effectively axiomatized formal theory rich in arithmetic, <em>T</em>, for any formula <em>φ</em>(<em>x</em>) with one free variable, there exists a specific sentence <em>γ</em> (the fixed point) such that the theory can prove</p><ul><li><em>γ </em>↔ <em>φ</em>(<em>⌜γ⌝</em>),</li></ul><p>where <em>⌜γ⌝ </em>is the numerical code (Gödel number) of that very sentence. The code serves as a “name” of <em>γ</em>. Here is a simple example. Let <em>φ </em>be the predicate: “is an odd number.” Then the sentence</p><ul><li><em>γ </em>↔ <em>φ</em>(<em>⌜γ⌝</em>)</li></ul><p>says that the code number of <em>γ</em> is odd. But since it is talking about itself, the sentence essentially says: “My code number is odd.” This elegant mechanism is how mathematicians achieve self-reference in formal systems. So far, nothing mysterious.</p><p>Gödel went far beyond such an innocent self-reference. He first defines a predicate, <em>Prov</em>, which says: “is provable within <em>T</em>.” This predicate holds exactly when its argument is the Gödel number of a sentence that has a formal proof in <em>T</em>. Accordingly, the sentence</p><ul><li><em>γ </em>↔ <em>Prov</em>(<em>⌜γ⌝</em>),</li></ul><p>says <em>γ </em>is provable, or essentially, “I am provable.” Gödel masterfully constructed the following sentence:</p><ul><li><em>g </em>↔ ¬<em>Prov</em>(<em>⌜g⌝</em>),</li></ul><p>where the symbol ¬ denotes negation. The sentence <em>g</em> is effectively saying: “I am not provable within <em>T</em>.” In Gödel’s (1931) words: “We are therefore confronted with a proposition which asserts its own unprovability.” What does that mean?</p><p>If <em>g</em> were provable, then — assuming <em>T</em> is sound, meaning that it proves only true statements — <em>g</em> would be true. But <em>g</em> says of itself that it is not provable, so this would be a contradiction. Therefore, <em>g</em> is not provable in <em>T</em>. And since that is exactly what <em>g</em> asserts, <em>g</em> is true. We thus obtain a sentence that is true but unprovable within <em>T</em>. Truth, therefore, extends beyond proof.</p><p>The conclusion is that a formal system rich in arithmetic cannot be both complete (i.e., decides all its statements) and consistent (free of contradiction). If it is consistent, then there will be many sentences expressible within the system that are not provable. This means the system will be incomplete. If it were complete, i.e., all its statements are decidable (provable or refutable) within, it would involve contradictions and thus could not be consistent. This essentially is the First Incompleteness Theorem (Smith, 2020).</p><p>In a nutshell, Gödel invented a “Logic Magic Mirror” capable of reflecting the light of truth at the center of the most sophisticated formal systems.</p><h4>Symmetry and Impossibility</h4><p>There is a surprising pattern common in the works of Galois and Gödel.</p><p>The Galois group of a polynomial measures exactly which permutations of its roots the base field cannot detect — symmetries are, in essence, the indistinguishabilities of the system. If the roots are indistinguishable, no particular solution can be identified (Carter, 2009; Gallian, 2025). Just as you cannot tell whether a perfect circle has been rotated without an external mark, you cannot solve for a specific root if the system cannot distinguish it from its neighbors. Solving an equation means breaking the symmetry among its roots; when that symmetry cannot be broken from within the system, the equation is unsolvable.</p><p>A century later, Gödel followed a surprisingly similar terrain in formal logic: using a symmetry-like fixed-point construction, he revealed a boundary no sufficiently rich formal system can cross.</p><p>A comparison between the two achievements is telling: Just as sufficiently rich formal systems contain truths unreachable from <em>within the system</em>, sufficiently complex equations can contain solutions unreachable from <em>within the system</em>. These two geniuses, operating in entirely distinct domains, were illuminating the same pattern from different shores. We can, therefore, argue that, broadly speaking,</p><blockquote>Galois and Gödel used symmetry to reveal impossibility.</blockquote><p>The same pattern surfaces in physics, through the celebrated theorem of <a href="https://en.wikipedia.org/wiki/Noether%27s_theorem">Emmy Noether</a>. In 1918, Noether proved that every continuous symmetry of a physical system gives rise to a conservation law:</p><ul><li>Rotational symmetry yields conservation of angular momentum,</li><li>Spatial translation symmetry yields conservation of linear momentum, and</li><li>Time translation symmetry yields conservation of energy.</li></ul><p>And, as the theoretical physicist <a href="https://en.wikipedia.org/wiki/Brian_Wybourne">Brian Wybourne</a> (1992) observed, every such conservation law is, at heart, an impossibility:</p><blockquote>“The existence of a symmetry implies that it is impossible to devise an experiment to distinguish the before and after situation. If you succeed, then the symmetry does not exist. All the great conservation laws are associated with the assertion that a particular experiment is impossible.”</blockquote><p>For example:</p><ul><li>Conservation of angular momentum implies that no experiment can detect a preferred orientation in space.</li><li>Conservation of linear momentum implies that no experiment can detect a preferred location in space.</li><li>Conservation of energy implies that no experiment can detect a preferred moment in time.</li></ul><p>The relationship between symmetry and impossibility goes deeper than it appears. Every law of nature is, at heart, a prohibition, as <a href="https://plato.stanford.edu/entries/popper/">Karl Popper</a> (1959) points out. Without impossibility, there would be no laws and thus no science. As <a href="https://en.wikipedia.org/wiki/John_D._Barrow">John Barrow</a> (1998) states:</p><blockquote>“Science is only possible because some things are impossible.”</blockquote><h4>Fractal Truth</h4><p>Fractals are structures where the same pattern reappears at every scale: zoom in, and you see a smaller copy of the whole. No single snapshot captures all the details; there is always more structure further down.</p><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/1*8QxxZ4JEKjApeLsZWI4_sA.png" /><figcaption>A never-ending, computer-generated fractal image. Source: Shutterstock.</figcaption></figure><p>Gödel’s work shows something similar about truth in formal systems rich in arithmetic. Using his fixed‑point construction, he builds a sentence <em>g</em> that says, in effect, “I am not provable in this theory <em>T</em>.” If <em>T</em> is consistent, <em>g</em> is true but unprovable in <em>T</em>.</p><p>We can then strengthen the theory <em>T </em>by adding <em>g</em> (and other true but unprovable statements in <em>T</em>) as new axioms, forming a stronger theory <em>Tʹ</em>. In <em>Tʹ</em>, we can again talk about what is provable in <em>Tʹ</em>, build new fixed points, and uncover new true but unprovable sentences for <em>Tʹ</em> itself. Iterating this process, <em>T</em>, <em>Tʹ</em>, <em>Tʺ</em>, …, gives an open‑ended <em>hierarchy</em> of theories, each seeing more truths than the last, but none capturing them all at once.</p><blockquote>Gödel’s theorem implies that truth admits a fractal-like structure.</blockquote><p>In this sense, arithmetic has a <strong>fractal‑like structure</strong>: at each new “zoom level” (each stronger theory), familiar patterns of self‑reference and unprovability reappear, and there are always further layers of truth beyond the reach of any single formal system. This can be linked to <a href="https://npg.copernicus.org/articles/30/311/2023/"><strong>scale symmetry</strong></a> in fractals: a single pattern reappears at different levels, as a sentence talks about its own code, which in turn can be talked about again in richer languages.</p><p>Fractality may very well be woven into the very fabric of the universe (Frame &amp; Urry, 2016). Just as a fractal coastline hides new detail at every further zoom, arithmetic hides new true statements at every step to stronger theories; Gödel’s fixed-point construction is the mechanism that illuminates these “higher layers” lying beyond the grasp of any isolated formal system.</p><p>Hofstadter (1999) was among the first to recognize the fractal nature of truth. In <em>Gödel, Escher, Bach</em>, he writes (p. 71):</p><blockquote>“The boundary between the set of truths and the set of falsities is meant to suggest a randomly meandering coastline which, no matter how closely you examine it, always has finer levels of structure, and is consequently impossible to describe exactly in any finite way. (See <a href="https://en.wikipedia.org/wiki/Benoit_Mandelbrot">B. Mandelbrot</a>’s book Fractals.)”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/724/1*3d3QHHHcFKNDZgXIal09hA@2x.jpeg" /><figcaption>Image inspired by Hofstadter (1999, Figure 18, p. 71); generated by GPT-5.5.</figcaption></figure><p>John Barrow (1998) emphasizes the importance of the fractal nature of our knowledge:</p><blockquote>“Our knowledge about the Universe has an edge. Ultimately, we may even ﬁnd that the fractal edge of our knowledge of the Universe defines its character more precisely than its contents; that what cannot be known is more revealing than what can.”</blockquote><h4>But The Universe is Asymmetric?</h4><p>Can symmetry explain everything? The universe is full of asymmetries. The dominance of matter over antimatter, the arrow of time, the handedness of life — these are not symmetric phenomena (Goldberg, 2014). Can a thesis built on symmetry account for them?</p><p>The answer, drawn from modern science, is that symmetry and its breaking are not opposites but partners. The laws of the Standard Model are deeply symmetric; their ground states are not. A symmetric system, through a symmetric mechanism, can produce asymmetric outcomes. This is the lesson of spontaneous symmetry breaking, one of the most profound ideas in twentieth-century physics (Rosen, 2008).</p><p>Gödel’s theorem seems to fit this pattern. In a <em>complete</em> and consistent system, there is a symmetry between <em>expressibility</em> and <em>decidability. </em>Every sentence that can be meaningfully stated in the system would also be settled by proof within the system, and every proof would settle a sentence expressible in the system. Gödel showed that, for any consistent, effectively axiomatized system rich enough for arithmetic, this symmetry must break — and the mechanism of the breaking is built by a symmetry-inspired design: the fixed-point construction that produces the Gödel sentence.</p><blockquote>One kind of symmetry could break another.</blockquote><p>Symmetry, in other words, is what reveals the asymmetry between truth and proof. Symmetry does not stand opposed to its breaking; it is the engine that produces it.</p><h4>Conclusion</h4><p>Galois pioneered group theory, the modern language of symmetry. Symmetry has been linked to major revolutions in modern science. We argue that symmetry can also be linked to the Incompleteness Theorem, the most radical revolution in modern logic.</p><p>It is gratifying to see the connection between one of the most abstract theorems of logic and one of the most important drivers of modern science: Symmetry.</p><p>Richard Feynman concluded his book, <em>The Character of Physical Law </em>(1965), with the following insight:</p><blockquote>“Nature uses only the longest threads to weave her patterns, so each small piece of her fabric reveals the organization of the entire tapestry.”</blockquote><p>Symmetry is certainly one of nature’s longest threads. This article has been an attempt to follow that thread from Galois to Gödel, and to glimpse the larger pattern it reveals.</p><p><em>The basic idea for this article was born in April 2025. Ian Stewart gave me an encouraging nudge. A year later, Douglas Hofstadter and </em><a href="https://scholar.google.com/citations?user=07klrCcAAAAJ&amp;hl=en"><em>Fawzi Al-Thukair</em></a><em> offered generous and constructive comments on an early draft. I am grateful to all. I remain solely responsible for any errors or omissions</em>.</p><h4>References</h4><ul><li>Barrow, J. D. (1998). <em>Impossibility: The limits of science and the science of limits</em>. Oxford University Press.</li><li>Boolos, G. S., Burgess, J. P., &amp; Jeffrey, R. C. (2007). <em>Computability and logic</em> (5th ed.). Cambridge University Press.</li><li>Carter, N. (2009). <em>Visual group theory</em>. Mathematical Association of America.</li><li>Feynman, R. P., Leighton, R. B., &amp; Sands, M. (2010). <em>The Feynman lectures on physics</em> (Vol. 1). Basic Books.</li><li>Feynman, R. P. (2017). <em>The character of physical law</em> (F. Wilczek, Intro.). MIT Press. (Original work published 1965).</li><li>Frame, M., &amp; Urry, A. (2016). <em>Fractal worlds: Grown, built, and imagined</em>. Yale University Press.</li><li>Gallian, J. A. (2025). <em>Contemporary abstract algebra</em> (11th ed.). CRC Press.</li><li>Gödel, K. (1931/1962). <em>On formally undecidable propositions of Principia Mathematica and related systems</em> (B. Meltzer, Trans.). Edinburgh University Press. (Original work published 1931).</li><li>Goldberg, D. (2014). <em>The universe in the rearview mirror: How hidden symmetries shape reality</em>. Plume.</li><li>Hofstadter, D. R. (1999). <em>Gödel, Escher, Bach: An eternal golden braid</em> (20th anniversary ed.). Basic Books. (Original work published 1979).</li><li>Jehle, G., &amp; Reny, P. (2011). <em>Advanced microeconomic theory</em> (3rd ed.). Pearson.</li><li>Livio, M. (2005). <em>The equation that couldn’t be solved: How mathematical genius discovered the language of symmetry</em>. Simon &amp; Schuster.</li><li>Popper, K. (1992). <em>The logic of scientific discovery</em>. Routledge. (Original work published 1959)</li><li>Rosen, J. (2008). <em>Symmetry rules: How science and nature are founded on symmetry</em>. Springer.</li><li>Smith, P. (2020). <em>An introduction to Gödel’s theorems</em> (2nd ed.). Logic Matters.</li><li>Smullyan, R. M. (1994). <em>Diagonalization and self-reference</em>. Clarendon Press.</li><li>Stewart, I. (2007). <em>Why beauty is truth</em>. Basic Books.</li><li>Stewart, I. (2023). <em>Galois theory</em> (5th ed.). CRC Press.</li><li>Weyl, H. (1952). <em>Symmetry</em>. Princeton University Press.</li><li>Wilczek, F. (2015). <em>A beautiful question: Finding nature’s deep design</em>. Penguin Press.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bd40a77b2483" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/%C3%A9variste-galois-meets-kurt-g%C3%B6del-bd40a77b2483">Évariste Galois Meets Kurt Gödel</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Free Universe: Why Reality Begins with Freedom]]></title>
            <link>https://medium.com/godelian-letters/a-free-universe-why-reality-begins-with-freedom-c33019feb474?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/c33019feb474</guid>
            <category><![CDATA[incompleteness-theorem]]></category>
            <category><![CDATA[emergence]]></category>
            <category><![CDATA[free-will]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[existence-of-god]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Mon, 23 Mar 2026 11:28:33 GMT</pubDate>
            <atom:updated>2026-03-23T14:37:20.553Z</atom:updated>
            <content:encoded><![CDATA[<p><em>A firm belief in free will is imperative to overcome the potential dangers of AI.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-BapqiyENPVkVKnelol55w.png" /><figcaption>A clockwork universe. Source: <a href="https://www.shutterstock.com/image-illustration/fantasy-space-clock-machine-280016237?trackingId=81d1a722-5991-4885-8bbb-82f5bbc6aafa&amp;listId=searchResults">Shutterstock</a>.</figcaption></figure><p>Since Isaac Newton’s <em>Principia Mathematica</em>, science has relentlessly endeavored to uncover the fundamental laws governing the Universe. For the most part, the implicit vision is that the universe is a vast, deterministic machine.</p><p>A counterline of research emerged in the twentieth century with the discoveries of quantum mechanics and chaos theory. Uncertainty is irreducible, it concluded, and many natural phenomena are inherently unpredictable, contrary to the Newtonian clock world.</p><p>For both of these trends, however, free will remains an <a href="https://www.informationphilosopher.com/freedom/standard_argument.html">unresolved problem</a>. Free will is neither consistent with a clock worldview nor with quantum randomness or chaotic unpredictability. Consequently, for many modern thinkers, free will is simply dismissed as an illusion.</p><p>We propose an alternative view by building a framework that takes freedom as the foundation of reality rather than as a problem to be solved (away) from the scientific worldview.</p><h4><strong>A Ladder of Constraints</strong></h4><p>We propose a hypothesis about existing reality and examine how it can integrate key achievements of modern science and logic into a coherent framework.</p><p>Given freedom, we view the laws of nature as constraints. We, therefore, assume that existing entities within the universe are constrained by other entities. A falling apple is constrained by the Earth through gravity; electrons are constrained by the electromagnetic forces of the atom; biological cells are constrained by the rules of the host, and so on. These constraints are consistent with the laws of nature governing the interaction between physical objects.</p><p>For our framework, we make the following assumptions:</p><ul><li>For two interacting entities <em>A</em> and <em>B</em>, one is the dominant (net) constrainer of the other. Although mutual dependence is common, it is rarely the case that two entities have exactly equal constraining impact. The assumption that one entity has a net constraining impact on the other is, therefore, quite realistic.</li><li>If <em>A</em> constrains <em>B</em>, then <em>A </em>has fewer constraints than <em>B.</em> Intuitively, because the constrainer acts as the architect or the boundary-setter for the constrained, it must inherently possess a wider degree of freedom. As we shall see shortly, this is supported by sound developments of 21st-century science.</li><li>Since a constraint is either binding or not, constraints must be discrete steps rather than a continuous, infinite blur.</li></ul><p>From these assumptions, we can think of constrained entities as organized along a “constraint ladder”: As we ascend the ladder, each entity has fewer constraints than the one beneath.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/567/1*jlCxMNzD73ScMh2tsb8SFA.png" /><figcaption>A ladder to heaven. Image created by Grok.</figcaption></figure><p>As we shall see, the resulting architecture is surprisingly consistent with modern theories of science and logic.</p><h4>The Rank-Zero Existent</h4><p>We now ask a simple but important question: Does this ladder go upwards forever?</p><p>If we have a ladder built of discrete, distinct steps, and every step upward leads to a state with strictly fewer constraints, logic dictates that the ladder <em>must</em> eventually terminate. It cannot go on infinitely. It must hit zero. (A formal treatment is provided in Al-Suwailem, 2026.)</p><p>What lies at the top of the ladder? An entity with exactly zero constraints. This shows that absolute freedom is the highest level of reality. It is not difficult to see how this reasoning connects to <a href="https://cassyni.com/events/7cCqV1gKum6DF4i6ta86pV">ontological arguments</a> on the existence of God. However, these proofs often treat free will as a downstream anomaly to be resolved later.</p><p>By contrast, the present framework introduces a novel approach: it uses freedom itself, modeled as the absence of external constraints, as the ontological primitive. It demonstrates that the existence of a supreme, rank-zero entity can be deduced directly from the localized constraints we observe.</p><p>In classical philosophy, true freedom is not the liberty of random, arbitrary outcomes. True freedom is the ability to act entirely according to one’s own nature, without any external impediment or prior cause. An entity with zero external constraints is, by definition, absolutely free.</p><p>Regardless of the metaphysics of the existence of God, the conclusion that ultimate reality springs from absolute freedom is a liberating idea. Before going any further, we need to fine-tune this argument a bit more.</p><h4>A Gödelian Hierarchy</h4><p>Our proposed framework can be strengthened by the infamous Incompleteness Theorems of Kurt Gödel.</p><p>Roughly speaking, Gödel’s First Incompleteness Theorem states that within a formal axiomatic system rich in arithmetic, <em>S</em>, there will be a sentence (many, in fact), called the Gödel sentence, that can neither be proved nor disproved <em>within S</em>. A Gödel sentence <em>G </em>can be seen to be true only from a higher system, <em>S</em>ʹ, in which a new Gödel sentence, <em>G</em>ʹ, would emerge, <em>ad infinitum.</em></p><p>The conclusion is that mathematical truth is inexhaustible. No finite formal system can ever encapsulate the entirety of truth. Truth extends infinitely beyond formal logical reasoning.</p><p>Incompleteness isn’t a negative result, as it is often portrayed. Incompleteness can be viewed as providing a logical space for rational agency. To prove the sentence <em>G</em>, we have to move outside <em>S </em>to <em>S</em>ʹ, which is a rational move that must transcend formal apparatus. The progress of science can be seen as the progression from <em>S </em>to <em>S</em>ʹ to <em>S</em>ʺ … There is no closed end to the progress of science.</p><p>To be sure, Gödel’s Incompleteness Theorems do not <em>prove</em> the existence of free will. However, they can serve as a guiding light for rational inquiry. Incompleteness reveals cracks in any fixed formal framework—places where truth outruns what can be derived from a given finite rulebook.</p><blockquote>Incompleteness doesn’t hand us freedom — but it makes room for it.</blockquote><p>The search for truth is a driving motive of scientific progress: a North Star we can approach indefinitely, even if we never exhaust it. In this spirit, William Byers (2008) argues that mathematical truth can be explored yet resists final capture by any single formal, effectively specified system. If we treat this as a structural analogy, then the ontological counterpart of inexhaustible truth is unbounded freedom: the idea that reality cannot be fully “closed” by any final set of constraints, but remains open to new constraints, new organization, and new creation.</p><h4>Organizational Complexity</h4><p>Our proposed framework fits surprisingly well with modern theories of <em>emergence</em>.</p><p>Nobel laureates <a href="https://en.wikipedia.org/wiki/Philip_W._Anderson">Philip Anderson</a> and <a href="https://en.wikipedia.org/wiki/Robert_B._Laughlin">Robert Laughlin</a>, among others, were pioneers in emphasizing the role of <em>organization</em> in the emergence of structures governed by new rules and principles. In his article (1972), “More is Different,” Anderson argues that the reductionist approach tries to reduce every phenomenon of nature to a small set of fundamental laws and principles. Unfortunately, this approach is fruitless: there is no systematic way to derive the complex phenomena we measure and interact with from these fundamental laws.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NAgV0DwkE3Y4r3XG0cBGFQ.png" /><figcaption><a href="https://www.nytimes.com/2020/03/30/science/philip-w-anderson-dead.html">Philip Anderson</a> (left) and <a href="http://large.stanford.edu/history/kaist/photos/portraits/">Robert Laughlin</a>.</figcaption></figure><p>In his book, <em>A Different Universe</em> (2005), Laughlin offers a detailed account of the emerging “science of organization.” He writes:</p><blockquote>Since principles of organization — or, more precisely, their consequences — can be laws, these can themselves organize into new laws, and these into still newer laws, and so on.</blockquote><blockquote>The laws of electron motion beget the laws of thermodynamics and chemistry, which beget the laws of crystallization, which beget the laws of rigidity and plasticity, which beget the laws of engineering. The natural world is thus an interdependent hierarchy of descent …</blockquote><blockquote>Thus, the tendency of nature to form a hierarchical society of physical laws is much more than an academic debating point. It is why the world is knowable.</blockquote><p>To elaborate on how emergence relates to our Freedom Framework, consider electrons. An electron in isolation is identical to every other electron. But an electron in a Hydrogen atom plays a different role than that of an electron in an Oxygen atom, and the electrons in a water molecule, H₂O, play even more different roles. This functional diversity is critical for the emergence of complex systems and the associated hierarchy of constraints. Consequently, the electron has more constraints within an atom than the atom, which has more constraints than the molecule, and so on.</p><blockquote>As a component of the whole, the part must meet stricter conditions than the whole does.</blockquote><p>Similarly, different heart cells play different roles, as each occupies a different position in the heart. If we think of the heart as a complex network of cells, then this network would have critical nodes that are more important than other nodes. This shows that, within the heart, cells have different levels of constraints. These critical nodes have fewer constraints. The heart, in turn, has fewer constraints than each of its cells, while the body overall has yet fewer constraints.</p><p>The point is that in complex systems, members tend to have hierarchical levels of constraint, consistent with our framework.</p><h4>Saving Free Will</h4><p>Our framework makes freedom the fundamental ontological primitive of existence. In standard physicalist or mechanistic frameworks, determinism (constraint) is assumed to be the foundation, rendering “free will” an anomaly or illusion that must be painfully shoehorned in later. Our framework flips this approach.</p><p>This perspective offers a natural place for “free will”; it isn’t an exception to the rules of nature; it is simply a localized retention of the foundational reality. An agent with free will acts as a localized “source” of constraints (constraining its environment) without being perfectly symmetrically constrained by upstream variables.</p><p>The freedom framework reveals a deep, structural link between existence, free will, and creativity. If absolute freedom is the ontological bedrock, then the act of bringing reality into existence is a fundamentally free, creative act, the deliberate imposition of foundational constraints.</p><p>Furthermore, because the framework admits a multi-layered hierarchy of agents, finite free will operates as a localized retention of that foundational creativity. Agents do not merely passively experience reality; by exercising their local freedom to initiate novel sub-chains, they actively participate in the ongoing creative actualization of existence. In this model, the descent from the rank-zero source is not merely a restriction of possibilities, but a cascading, fractal expansion of creative agency.</p><h4>Free Will and AI</h4><p>A firm belief in free will is essential to overcome the potential dangers of Artificial Intelligence (AI).</p><p>AI is clearly useful for a wide range of civil tasks. As I argued in another <a href="https://godelianletters.org/God_devil_G%C3%B6del">article</a>, the real threat emerges when we put too much faith in it and <em>voluntarily submit</em> to the machine. In this case, AI is no longer an <em>assistant</em> but instead the governor of our business.</p><p>The view that free will is an illusion and that our actions are completely determined by mechanical processes provides a perfect environment for total submission to machines. If humans are nothing but biological machines, then surrender is easy. Why resist? Let the most efficient machine run the show. <em>That worldview doesn’t just permit submission to AI — it invites it</em>.</p><p>The only way to overcome the dangers of AI, therefore, is to liberate ourselves from the surrender to machines. Humans should remain in the driver’s seat.</p><blockquote>AI optimizes within constraints; free humans choose those constraints.</blockquote><h4>Conclusion</h4><p>A worldview that starts from freedom is able to hit two birds with one stone:</p><ul><li>It is able to explain natural laws, in line with modern science.</li><li>It can explain the undeniable experience of free will we all share.</li></ul><p>From a Philosophy of Science point of view, a theory that consistently explains a larger dataset is clearly preferable to one that explains only part of it and denies the remaining part.</p><p>By inverting the mechanistic assumption that makes the laws of nature the foundational basis, our framework makes freedom the ontological bedrock for existence. This allows us to explain <em>both </em>phenomena. Within this framework, finite free will is not an anomaly, but a localized microcosm of the unconstrained reality that grounds existence itself.</p><p>The biggest mistake of the mechanical worldview is not determinism per se. It’s the assumption that determinism is the starting point. A freedom-first ontology flips the direction: existence at its purest is unconstrained, and constraint is a derivative, organizing phenomenon.</p><p>In this view, the world is not a closed machine. It is an organized hierarchy whose constraints trace back to something that is not itself constrained. And if that is true, then free will is not an embarrassment; it is a clue.</p><p>We can build powerful machines. But we must not build a worldview that hands them our agency.</p><h4>References</h4><ul><li>Al-Suwailem, S. (2026). <em>God and freedom: A constraint ladder argument for an unconstrained existent</em> [Working paper]. <a href="https://www.academia.edu/164895317/God_and_Freedom_A_Constraint_Ladder_Argument_for_an_Unconstrained_Existent">Academia.edu</a>.</li><li>Anderson, P. W. (1972, August 4). More is different. <em>Science, 177</em>(4047), 393–396.</li><li>Byers, W. (2008). <em>The blind spot: Science and the crisis of uncertainty.</em> Princeton University Press.</li><li>Laughlin, R. B. (2005). <em>A different universe: Reinventing physics from the bottom down</em>. Basic Books.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c33019feb474" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/a-free-universe-why-reality-begins-with-freedom-c33019feb474">A Free Universe: Why Reality Begins with Freedom</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[God, the Devil, and Gödel]]></title>
            <link>https://medium.com/godelian-letters/god-the-devil-and-g%C3%B6del-705509a40895?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/705509a40895</guid>
            <category><![CDATA[creativity]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[tragedy-of-the-commons]]></category>
            <category><![CDATA[free-will]]></category>
            <category><![CDATA[prisoners-dilemma]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Sat, 03 Jan 2026 09:20:51 GMT</pubDate>
            <atom:updated>2026-01-03T09:20:50.333Z</atom:updated>
            <content:encoded><![CDATA[<p><em>If we were merely machines, the Devil would rule.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uTs_wYOSmJ5313k0Arl8Xg.png" /><figcaption>Escher: Angels &amp; Devils. Source: <a href="https://escherinhetpaleis.nl/en/about-escher/escher-today/circle-limit-iv-heaven-and-hell">Circle Limit</a>.</figcaption></figure><p>Gödel’s Incompleteness Theorems sparked rich debate about the nature of the mind, whether machines, via Artificial Intelligence, can achieve or even surpass human intelligence, and the resulting impact on civilization. Here, we argue:</p><ul><li>Machines will never reach the level of human intelligence.</li><li>The greatest threat to human intelligence is, surprisingly, human intelligence itself.</li><li>A society controlled by robots would be ruled by the “Devil.”</li></ul><h4>A Monumental Achievement</h4><p><a href="https://www.britannica.com/biography/John-von-Neumann">John von Neumann</a> (1903–1957), one of the most creative polymaths of the twentieth century and a key figure in the development of AI, was a close friend of the legendary <a href="https://medium.com/godelian-letters/remembering-g%C3%B6del-398fac905010">Kurt Gödel</a> (1906–1978). When presenting the <a href="https://en.wikipedia.org/wiki/Albert_Einstein_Award#:~:text=The%20Albert%20Einstein%20Award%20was,the%20trustees%20of%20the%20institute.">Albert Einstein Award</a> to Gödel on March 14, 1951, von Neumann remarked:</p><blockquote>“Kurt Gödel’s achievement in modern logic is singular and monumental — indeed, it is more than a monument; it is a landmark which will remain visible far in space and time.”</blockquote><p>He then explains the key achievement of Gödel:</p><blockquote>“Gödel was the first man to demonstrate that certain mathematical theorems can neither be proved nor disproved with the accepted, rigorous methods of mathematics. … He proved furthermore that a very important specific proposition belonged to this class of undecidable problems: The question as to whether mathematics is free of inner contradictions.”</blockquote><p>Von Neumann highlights the importance of Gödel’s result:</p><blockquote>“The result is remarkable in its quasi-paradoxical ‘self-denial’: It will never be possible to acquire with mathematical means the certainty that mathematics does not contain contradictions.”</blockquote><p>He then emphasized that Gödel’s results are not confined to mathematics but apply to <em>any</em> field that employs formal methods:</p><blockquote>“It must be emphasized that the important point is that this is not a philosophical principle or a plausible intellectual attitude, but the result of a rigorous mathematical proof of an extremely sophisticated kind.</blockquote><blockquote>Gödel actually proved this theorem, not with respect to mathematics only, but for all systems which permit a formalization, that is, a rigorous and exhaustive description in terms of modern logic: For no such system can its freedom from inner contradiction be demonstrated with the means of the system itself.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y7YAuKqtaz_5J9lTSyaJzw.png" /><figcaption><a href="https://www.sothebys.com/en/articles/john-von-neumann-the-last-great-polymath#:~:text=R%20egarded%20as%20the%20foremost%20mathematician%20of,a%20wealthy%20non%2Dobservant%20Jewish%20family%20in%20Budapest.">John von Neumann</a> (left) and <a href="https://attentionsw.org/about-andre-weil/">André Weil</a>.</figcaption></figure><h4>Weil’s Dictum</h4><p>One of the most influential figures in mathematics during the 20th century, <a href="https://www.britannica.com/biography/Andre-Weil">André Weil</a> (1906–98), quipped on Gödel’s results (Rosenbloom, 1950):</p><blockquote>“God exists, since mathematics is consistent; the Devil exists, since we cannot prove it.”</blockquote><p>Historically, the existence of God was often associated with gaps in our knowledge, usually referred to as the “<a href="https://en.wikipedia.org/wiki/God_of_the_gaps#:~:text=%22God%20of%20the%20gaps%22%20is,a%20comprehensive%20and%20coherent%20worldview.">God of the Gaps</a>.” Weil is pointing towards the opposite direction: The existence of God is linked to the remarkable consistency of mathematics.</p><p>The stunning harmony of mathematics with the physical universe has drawn the attention of many, including Nobel laureate <a href="https://en.wikipedia.org/wiki/Eugene_Wigner">Eugene Wigner</a>, who, in 1960, published an article titled “<em>The Unreasonable Effectiveness of Mathematics in the Natural Sciences</em>.” The existence of God, in this view, is derived from the fact that there is a deep, global order in the universe. The unifying order behind such a complex and diverse universe is what motivated Albert Einstein (1936) to lament:</p><blockquote>“The eternal mystery of the world is its comprehensibility. … The fact that it is comprehensible is a miracle.”</blockquote><h4>Gödel’s Ontological Proof</h4><p>Gödel developed an <a href="https://en.wikipedia.org/wiki/G%C3%B6del%27s_ontological_proof">ontological system</a> to prove the existence of God, which builds on arguments by Gottfried Leibniz, <a href="https://en.wikipedia.org/wiki/Ontological_argument">among others</a>. The proof was later sharpened by <a href="https://www.cs.cmu.edu/~scott/">Dana Scott</a>. The Gödel-Scott system, call it <em>G</em>, does not encode enough arithmetic and, thus, is not (directly) prone to Gödel’s Incompleteness Theorems.</p><p>However, <em>G </em>does not by itself describe a concrete world like ours. A formal theory of the physical world, <em>W</em>, will normally be rich in arithmetic to embody the laws of nature. To construct a global theory of God <em>and</em> the world, therefore, we must embed the ontological system, <em>G</em>, into a larger formal theory rich in arithmetic, <em>T = G </em>+<em> W</em>. Now, <em>T </em>is <em>directly</em> prone to Gödel’s Second Incompleteness Theorem: if <em>T</em> is consistent, it cannot prove its own consistency. Although <em>T</em> still proves the existence of God via <em>G</em>, the consistency of <em>T</em>, including <em>G</em>, cannot be proven from within.</p><p>So, either God is proved to exist uniquely, in which case neither our world nor the devil has any foothold, or our world coexists with God, in which case the devil sneaks in as the resulting global system cannot prove its consistency.</p><h4>Truth vs. Proof</h4><p>The key message of Gödel’s First Incompleteness Theorem is that truth extends beyond proof. There will be many sentences in the formal system that are unprovable <em>within</em> the system but can be seen to be true from <em>outside</em> the system (Smullyan, 1992).</p><p>For a formal system rich in arithmetic, <em>S</em>, there will be many sentences that are not provable within <em>S</em>, but can be proved in a higher system, <em>Sʹ. </em>But within <em>Sʹ </em>there will be yet new sentences that are undecidable within, but can be proved in a higher system, <em>Sʺ</em>, <em>ad infinitum</em>. The conclusion is that <em>mathematics is inexhaustible</em>. No finite logical system can ever prove <em>all</em> mathematical truths.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*PlK6XBz5I-aFmfEy" /><figcaption>Albert Einstein’s 70th birthday celebration at the Institute for Advanced Study, Princeton, March 14, 1949. Left to right: Howard Robertson, Eugene Wigner, Herman Weyl, Kurt Gödel, Isidor Rabi, Albert Einstein, Rudolf Ladenburg, J. R. Oppenheimer, and Gerald Clemence. Source: <a href="https://universityarchives.princeton.edu/2015/10/dear-mr-mudd-is-the-institute-for-advanced-study-part-of-princeton-university/">Historical Photograph Collection</a>.</figcaption></figure><p>The gulf between truth and proof sparked rich discussions about whether the human mind is a logical machine or embeds something “higher” than pure logic. Since the mind is finite, the logical apparatus of the mind, no matter how powerful, must fail to prove many (infinite, in fact) mathematical truths. Would the mind be able nonetheless to “see” these unprovable truths?</p><h4>Mind vs. Machine</h4><p>Gödel’s Incompleteness Theorems, as such, do not settle the question: Is the mind merely a logic machine? An additional axiom or principle must be invoked to formulate a full answer. Philosophers and mathematicians held different views on this matter.</p><p>Let us start with Gödel.</p><p>Gödel subscribed to Hilbert’s dictum that <em>all</em> mathematical truths are, in principle, knowable. Since no formal system can encapsulate all the truth, the mind must have faculties that surpass the powers of mechanical logic (Wang, 1996).</p><p>In light of Rosenbloom’s (1950, pp. 72–73) discussion, the non‑algorithmic margin of the mind beyond its mechanical logic apparatus may be technically narrow yet decisively critical in surpassing the limits of mechanical reasoning, much like the vital minority of causes that drive the majority of outcomes in <a href="https://www.juran.com/blog/a-guide-to-the-pareto-principle-80-20-rule-pareto-analysis/">Pareto’s Principle</a> or the 80/20 rule.</p><p>Gödel’s cautious stance regarding a theory of the mind stands in contrast to the more ambitious position taken by <a href="https://www.britannica.com/biography/Roger-Penrose">Roger Penrose</a>, a 2020 Nobel Prize winner. In his book <em>The Emperor’s New Mind</em>, Penrose offers a sophisticated theory to explain the brain’s superior ability to go beyond mechanical, logical capacity.</p><p>While Penrose clearly considers the brain to embed non‑algorithmic, non‑computable faculties, his theory entails that these faculties are essentially grounded in physical laws describable in the language of mathematics. With the discovery of these laws, Penrose argues, we can “come to grips with” the nature and behavior of the mind.</p><p><a href="https://iep.utm.edu/lp-argue/">Critics</a> see the sought‑after theory as a mechanical theory in disguise, effectively re‑introducing the kind of rule‑governed model his Gödelian argument was meant to exclude.</p><p>Beyond the standard criticism, there is a deeper issue that Penrose admits he had not paid enough attention to: free will.</p><h4>Free Will and Intelligence</h4><p>It can be argued that free will and intelligence are two sides of the same coin (see Simonton, 2013). How is that? The essence of intelligence is creativity: To conceive new ideas that do not mechanically follow from existing ones.</p><p>But this requires the will to explore the unknown, tolerate mistakes, and learn from errors. All this requires freedom of choice. Without it, there can be little creativity, and without creativity, the will can hardly be “free.” Freedom and creativity are deeply interlinked, and both break away from the mechanical model of the mind. A logic machine is supposed to be accurate; it does not make mistakes as a human does. Turing (1947) rightly points out,</p><blockquote>“If a machine is expected to be infallible, it cannot also be intelligent.”</blockquote><p>Note that we are not saying mechanical machines cannot produce unpredictable or unexpected behavior; they do (see Briegel, 2012). Turing’s theorem on the unsolvability of the <a href="https://brilliant.org/wiki/halting-problem/">Halting Problem</a> says exactly that: there is no general, mechanical procedure that can, for every Turing machine and input, predict from the code alone whether it will eventually halt or run forever.</p><p>Instead, we argue that no formal theory can encapsulate the creativity of the human mind and make it fully predictable <em>within that theory</em>. Formal logic systems, as per Gödel’s and Turing’s theorems, are forever prevented from achieving that level of sophistication. We are not saying the mind is impossible to predict. All we say is that the logic machinery, in principle, is incapable of doing so in a systematic manner. The mind frequently surpasses the machine.</p><h4>The Tragedy of <em>Homo Economicus</em></h4><p>With free will, we venture into the domain of free choice in the marketplace. Economic textbooks (especially graduate microeconomics) describe the ideal economic man or <a href="https://en.wikipedia.org/wiki/Homo_economicus"><em>homo economicus</em></a>. It is not difficult to conclude that <em>homo economicus </em>is a sophisticated machine or robot (technically, it is modeled as a Turing machine; see Velupillai, 2000). It follows a well-defined set of instructions for making a utility-maximizing choice of goods and services.</p><p>But what would happen if the market consisted of such sophisticated robots, as Nobel laureate <a href="https://www.nobelprize.org/prizes/economic-sciences/1995/lucas/facts/">Robert Lucas</a> (1988) envisioned? Robots excel in simple environments. In complex environments, the results could be disastrous. The most famous example is known as the “<a href="https://online.hbs.edu/blog/post/tragedy-of-the-commons-impact-on-sustainability-issues">Tragedy of the Commons</a>,” also known as the “<a href="https://www.britannica.com/science/game-theory/The-prisoners-dilemma">Prisoner’s Dilemma Game</a>.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/964/1*6IX8G0_SEVow6DX0kjY7ig.png" /><figcaption>The Prisoner’s Dilemma Game.</figcaption></figure><p>In this game, each party has two choices: To cooperate or to defect. Mutual cooperation results in a higher payoff for each player than mutual defection. The problem, however, is that unilateral defection produces a higher payoff to the defector and a worse payoff to the cooperator. Since each party wants to be in such a position, each player, if maximizing its own payoff, will choose to defect. The result is a worse outcome than mutual cooperation.</p><p>Mutual defection is a “dominant strategy,” meaning it is a mechanically “obvious” choice. So, if the game is played by self-interested robots, the outcome is decidable but Pareto-inferior. An allocation is Pareto-optimal if it is impossible to change the allocation so that some members become better off without making at least one member worse off. A Pareto-inferior outcome, therefore, is an allocation that makes at least some people worse off without making anyone better off. Mutual defection makes everyone worse off relative to mutual cooperation.</p><p>If the Prisoner’s Dilemma game is played between a group of farmers or fishers, say, who share common resources, then each member will be individually better off to consume more than their fair share in the commons. But if they all follow the same strategy, resources are depleted, and they all lose. The term “Tragedy of the Commons” describes the unfortunate fate of a community when its members behave selfishly. It is a standard “Fallacy of Composition” problem. <a href="https://www.britannica.com/money/Elinor-Ostrom">Elinor Ostrom</a> won the 2009 Nobel Prize for her pioneering work on the governance of the commons (1990).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*U6LYSX2-vORJh6CqfjcZJA.png" /><figcaption><a href="https://news.vt.edu/articles/2018/04/clahs-amartya-sen.html">Amartya Sen</a> (left) and <a href="https://www.britannica.com/money/Elinor-Ostrom">Elinor Ostrom</a>.</figcaption></figure><h4>The Wisdom of Pareto</h4><p>Sustaining Pareto-superior outcomes in environments such as those of the Tragedy-of-the-Commons often requires judgment and wisdom that go beyond the execution of any fixed set of decision rules.</p><p>To be clear, according to the <a href="https://plato.stanford.edu/entries/church-turing/">Church–Turing thesis</a>, any mechanical procedure — one governed by fixed, explicit rules — is algorithmic and can be simulated by a Turing machine. As we speak of fixed, mechanical rules, we are also speaking of algorithms in the Church-Turing sense.</p><p><a href="https://en.wikipedia.org/wiki/Rice%27s_theorem">Henry Rice</a> extended Turing’s results on the Halting Problem to show that any non-trivial semantic property of a program is undecidable in general. Da Costa and Doria (2014) apply these results to mathematical economics (see also Velupillai, 2000).</p><p>Because Pareto efficiency depends on the entire structure of outcomes, not just local rules, it’s a “non-trivial” property in the technical sense: no general algorithm can guarantee it.</p><p>In the Prisoner’s Dilemma game, mutual defection results from the mechanical rule of following the dominant strategy. Unfortunately, this mechanism yields an outcome that is Pareto-inferior to mutual cooperation. In the Prisoner’s Dilemma type of environment, therefore, no fixed-rules mechanism can guarantee mutual cooperation.</p><p>Even sophisticated fixed-rule strategies like <a href="https://en.wikipedia.org/wiki/Tit_for_tat">Tit-for-Tat</a> — which encourage cooperation in repeated interactions under ideal conditions — cannot <em>guarantee</em> mutual cooperation across all environments. They remain vulnerable to noise, exploitation, or short horizons, and can lock into Pareto-inferior cycles of defection. In complex, real-world settings akin to the Tragedy of the Commons, therefore, no purely mechanical procedure can reliably ensure Pareto-superior outcomes.</p><h4>A Cooperation Algorithm Never Halts</h4><p>An algorithm designed to produce mutual cooperation in a one-shot Prisoner’s Dilemma will never terminate; it will get stuck in an infinite loop. Cooperation, in this strict sense, is algorithmically unattainable. Here is why.</p><p>Suppose we program the following algorithm to implement mutual cooperation:</p><ul><li><strong>Step 1:</strong> Players start with a payoff-maximizing choice (Defect).</li><li><strong>Step 2:</strong> If the resulting outcome is Pareto-inferior (<em>DD</em> compared to <em>CC</em>), players choose the Pareto-superior option (Cooperate).</li><li><strong>Step 3:</strong> If the new choice is payoff-maximizing, <strong>Stop</strong>. Otherwise, go back to <strong>Step 1</strong>.</li></ul><p>This algorithm creates a paradoxical loop. The players start with Mutual Defection. Seeing that it is Pareto-inferior, they switch to Mutual Cooperation. But cooperation is <em>not</em> individually rational — it leaves each player vulnerable to exploitation. Therefore, the rule drives them back to Defection. The process never converges:</p><p><strong><em>DD</em> → <em>CC</em> → <em>DD</em> → <em>CC </em>…</strong></p><p>No matter where we start, <em>DD</em> or <em>CC</em>, the program will fail to halt. In Turing terms, this is a non-terminating algorithm, an instance of the unsolvable Halting Problem. In contrast, a dominant-strategy algorithm instantly halts: Compute the dominant choice (<em>D</em> for both) → output mutual <em>DD</em> → done.</p><p>The machine, seeking a solution it can compute, will inevitably choose the halting (but tragic) state over the infinite (but superior) one. To escape the loop, one must step <em>outside</em> the logic of payoff maximization entirely — a feat of <em>free will</em> that no Turing machine can perform.</p><h4>The Impossibility of a Paretian Liberal</h4><p>In 1970, <a href="https://www.nobelprize.org/prizes/economic-sciences/1998/sen/facts/">Amartya Sen</a>, who won the 1998 Nobel Prize, published “<em>The Impossibility of a Paretian Liberal</em>.” Sen showed that, under plausible conditions, individual freedom and Pareto optimality are logically incompatible within a social choice framework. Sen (2017) further demonstrated how this tension can be modeled as a Prisoner’s Dilemma.</p><p>Gödel showed that an axiomatic system rich in arithmetic cannot be simultaneously complete and consistent. Sen showed that a social choice system cannot simultaneously guarantee individual liberty and Pareto efficiency. Consistency and Pareto efficiency are both non-trivial global properties that cannot be fully decided from within the system. In both cases, the “view from the outside” is required to sustain the system as a whole.</p><p>Sen’s results illustrate the price of free will: There is no universal logical framework that guarantees both freedom of choice and collective wisdom. To maintain both, we need to transcend the rigid limits of logic machinery and employ the unique, non-algorithmic capacities of the mind.</p><h4>The Real Dangers of AI</h4><p>Contrary to a common perception (e.g., Yudkowsky and Soares, 2025), the above discussion indicates that machines can never surpass human intelligence, except in one case: if humans voluntarily relinquish their gifted wisdom and abdicate their responsibility to the machine. Only then can robots destroy our economy, in a scenario similar to the Tragedy of the Commons.</p><p>The erroneous belief that AI can achieve, or even exceed, human intelligence, combined with uncontrolled greed, could lead humans to voluntarily submit to robots. The most dangerous threat to human intelligence, therefore, is human intelligence. This paradoxical outcome, interestingly, echoes the pattern of Gödel’s theorems: contradictory self-reference can cause a system to fall apart due to logical inconsistency.</p><p>We can benefit from AI only if we maintain our human dignity and collective wisdom over the machine. Machines are only a means. Humanity is the end, as <a href="https://plato.stanford.edu/entries/persons-means/#TreaAnotMereMeanActiWron">Immanuel Kant</a> argued. Elevating machines to become the end goal transforms humans into mere means, an outcome the Devil will surely celebrate.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*8GlwtW4Um7SxEFvx" /><figcaption>Source: <em>The New Yorker, </em>2012</figcaption></figure><h4>Conclusion</h4><p>There are good reasons to support the view that AI cannot surpass human intelligence. However, AI can dominate human lives if people voluntarily submit to it. This could happen for two reasons:</p><ul><li>Overconfidence in technology, and</li><li>Excessive greed.</li></ul><p>Sadly, this is the same explosive mix that fueled the largest credit bubble in history and subsequently triggered its collapse during the 2008 <a href="https://www.britannica.com/money/financial-crisis-of-2007-2008">Global Financial Crisis</a>. To avoid repeating history, we need to proactively govern not only AI but also profit incentives, which, if unchecked, may tragically ruin human civilization.</p><p><em>The title of this article is borrowed from Benacerraf (1967).</em></p><h4>References</h4><ul><li>Benacerraf, P. (1967). God, the devil, and Gödel. <em>The Monist, 51(1)</em>, 9–32.</li><li>Briegel, H. J. (2012). On creative machines and the physical origins of freedom. <em>Scientific Reports, 2</em>, Article 522.</li><li>Da Costa, N. C. A., &amp; Doria, F. A. (2014). On an extension of Rice’s theorem and its applications in mathematical economics. <em>Advances in Austrian Economics, 18</em>, 237–257.</li><li>Einstein, A. (1936). Physics and reality (J. Piccard, Trans.). <em>Journal of the Franklin Institute, 221(3)</em>, 349–382.</li><li>Lucas, R. E. (1988). On the mechanics of economic development. <em>Journal of Monetary Economics, 22</em>(1), 3–42.</li><li>Ostrom, E. (1990). <em>Governing the commons: The evolution of institutions for collective action</em>. Cambridge University Press.</li><li>Penrose, R. (1999). <em>The emperor’s new mind: Concerning computers, minds, and the laws of physics</em>. Oxford University Press.</li><li>Rosenbloom, P. (1950). <em>The elements of mathematical logic</em>. Dover Publications.</li><li>Sen, A. (1970). The impossibility of a Paretian liberal. <em>Journal of Political Economy, 78</em>(1), 152–157.</li><li>Sen, A. (2017). <em>Collective choice and social welfare</em> (Expanded ed.). Penguin Books.</li><li>Simonton, D. K. (2013). Creative thoughts as acts of free will: A two-stage formal integration. <em>Review of General Psychology, 17</em>(4), 374–383.</li><li>Smullyan, R. (1992). <em>Gödel’s incompleteness theorems.</em> Oxford University Press.</li><li>Turing, A. M. (1947, February 20). <em>Lecture to the London Mathematical Society</em> [Unpublished lecture]. London Mathematical Society, London, England.</li><li>Velupillai, K. V. (2000). <em>Computable economics</em>. Oxford University Press.</li><li>Von Neumann, J. (1951). Tribute to Dr. Gödel. In J. Bulloff, T. Holyoke, &amp; S. Hahn (Eds.), <em>Foundations of mathematics: Symposium papers commemorating the sixtieth birthday of Kurt Gödel</em> (pp. ix–x). Springer-Verlag.</li><li>Wang, H. (1996). <em>A Logical journey: From Gödel to philosophy.</em> MIT Press.</li><li>Yudkowsky, E., &amp; Soares, N. (2025). <em>If anyone builds it, everyone dies</em>. Little, Brown and Company.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=705509a40895" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/god-the-devil-and-g%C3%B6del-705509a40895">God, the Devil, and Gödel</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Relativity of Uncertainty]]></title>
            <link>https://medium.com/godelian-letters/relativity-of-uncertainty-cfbdad477cd8?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/cfbdad477cd8</guid>
            <category><![CDATA[nested-universe]]></category>
            <category><![CDATA[complementary]]></category>
            <category><![CDATA[indeterminism]]></category>
            <category><![CDATA[free-will]]></category>
            <category><![CDATA[incompleteness-theorem]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Sat, 25 Oct 2025 20:19:44 GMT</pubDate>
            <atom:updated>2025-10-25T20:19:43.334Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Time relativity creates nested layers of uncertainty with ample room for free will.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*E7jvkVWJEGKNCqk48wYYKA.png" /><figcaption>Albert Einstein. Image edited by Grok 4.</figcaption></figure><p>I argued in an earlier <a href="https://medium.com/godelian-letters/time-risk-and-%CF%89-454882c48f83">article</a> that time and uncertainty are two sides of the same coin. Hence, the relativity of time implies the relativity of uncertainty. Here, we will discuss concrete applications of this relativity and its consequences for the nature of time and free will.</p><h4>Light Clocks</h4><p>The “Light Clock” setup was used by many physicists, e.g., John Wheeler, to illustrate the concept of time relativity.</p><p>Suppose we have two mirrors facing each other, one at the bottom and one at the top of a container. Suppose we manage to have a photon of light bouncing back and forth between the two mirrors. To make things concrete, suppose the vertical distance <em>D</em> between the two mirrors is D = 1 foot (30.48 cm). The speed of light is roughly one foot per nanosecond (Mermin, 2009). One nanosecond is one billionth of a second. This means that, in one second, the photon completes 1 billion trips between the two mirrors.</p><p>Now, suppose we have two light clocks with identical construction and setup. One is stationary, i.e., not moving. The other is moving at a speed of 50% of the speed of light (i.e., 1 foot per 2 nanoseconds). For a stationary observer, how would the path of the photon in each clock look like?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LpgXMnvKbd7FOCp8cNNs_A.gif" /><figcaption>The Light Clock. Generated by Wolfram (Mathematica) 14.3 and Claude Opus 4.1.</figcaption></figure><p>The figure above provides the answer.</p><p>For a stationary observer, each tick (i.e., each trip of the photon between the two mirrors) in the stationary clock corresponds to 1/1.155 ticks of the moving clock. This means that as the stationary clock completes two ticks, the moving clock completes 1.73 ticks. The photon has to travel a longer distance between the two mirrors from the perspective of the stationary observer.</p><p>How does this relate to uncertainty?</p><p>For the stationary clock, the probability distribution of the photon is a uniform distribution. The probability of spotting the photon at any point along the distance between the two mirrors is the same. So, if we divide the distance (1 foot) into 10 bins each of length 0.1 feet, the probability of finding the photon at any of these bins is 10%.</p><p>How about the moving clock?</p><p>We need to calculate the distance the photon travels between the two mirrors. For the moving clock, this distance is <em>γD</em> = 1.155 × 1 foot = 1.155 feet, where <em>γ</em> = 1.155 is the Lorentz factor.</p><p>Now, to divide this into bins, each equal to 0.1 feet, we need 11.55 bins. This means that the probability of spotting the photon at any single bin is now 8.66% instead of 10%. The probability dilutes by 1/<em>γ</em> as the path stretches by <em>γ</em>. Hence, we can see how the relativity of time translates to the relativity of probability.</p><h4>Horizontal Light Clocks</h4><p>Let us now do another kind of experiment.</p><p>Suppose the clock has two vertical right-left mirrors facing each other. The distance between them in the stationary clock is still 1 foot. The moving clock now has the light traveling from left to right (with the motion of the clock), then bounces back from right to left (against the motion of the clock).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZOTlYzbfNEWAK-SfVvJrCQ.gif" /></figure><p>Because the speed of light is constant, it takes the photon longer to travel to the right end than to travel to the left. While the width of the moving clock will shrink to 0.866 feet, the photon will spend 75% of the time moving from left to right vs. 25% moving from right to left. However, the photon spends equal time between the two mirrors: The asymmetry in direction (75% L→R, 25% R→L) is exactly balanced by the asymmetry in its speed relative to the mirrors, so the spatial distribution remains symmetric and, hence, the probability distribution is still uniform. But since the distance between the mirrors shrinks, the probability of spotting the photon at any 0.1-feet bin is 1/8.66 = 11.55% compared to 10% for the stationary clock.</p><h4>Light-cone Uncertainty</h4><p>Time relativity results in uncertainty at a more fundamental level. Karl Popper (1982) was the first to show how Special Relativity implies objective uncertainty about the past and the future.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QLRGmddPuJVgiazL3jFPfQ.png" /><figcaption>A spacetime diagram showing the light cones of an observer on a world line. Source: <a href="https://en.wikipedia.org/wiki/File:World_line2.svg">Wikipedia</a>. Modified by the author.</figcaption></figure><p>The above diagram is known as a ‘Minkowski Spacetime Diagram.’ Since the speed of light constrains information and all causal interactions, the diagram depicts the limits of what an observer at point A can know or influence.</p><p>Starting from point A, signals traveling at light speed expand outward like a growing circle (or sphere in 3D space). As time progresses into the future, this circle gets larger, creating the ‘Future Light Cone’ — representing all events that the observer at A could potentially influence.</p><p>Similarly, looking backward in time, only signals from a shrinking circle of past locations could have reached point A. This creates the ‘Past Light Cone’ — representing all events that could have influenced the observer at A.</p><p>Everything outside both light cones is causally disconnected from A — these events can neither influence nor be influenced by what happens at A.</p><p>Now consider a point B in the future of A. Can the observer at A predict with certainty what will happen at B? The Past Light Cone of B contains points, like C, that are completely out of reach for the observer at A. It is physically impossible for A to learn anything about C; yet, C does influence the future of A, which lies at B. This uncertainty about the future is built into the fabric of the universe as envisioned by the Special Theory of Relativity.</p><p>But this uncertainty, while objective and unbreakable for A, is relative: C might be in the Past Light Cone of another observer at point D, for whom the point A is completely out of reach. The relativity of uncertainty is fascinating — it is completely objective for each observer (A cannot possibly know C, D cannot possibly know A), yet what is knowable versus unknowable is relative between observers. What is uncertain for A may be certain for D, and vice versa.</p><h4>The Block Universe</h4><p>Although the uncertainty Popper identifies in Special Relativity is objective, it is relative to the observer. For a hypothetical observer from ‘outside the universe’ — one who could view the entire spacetime structure at once — the entire spacetime would appear certain and deterministic. This is the ‘Block Universe’ view, which <a href="https://shs.cairn.info/journal-revue-d-histoire-des-sciences-2018-1-page-79?lang=en">Einstein</a> seemed to embrace (Carroll, 2010).</p><p>The catch, however, is that we are not outside the universe! No observer within spacetime can physically occupy this ‘view from nowhere.’ This view, therefore, is not empirically testable.</p><p>Although the Block Universe view seems consistent with the Relativity Theory, it makes no testable predictions beyond those already provided by the theory. It is more metaphysics than physics — a philosophical interpretation rather than a scientific theory.</p><h4>Nested Universes</h4><p>Let us consider an alternative hypothesis.</p><p>Suppose that our entire universe is embedded within a meta-universe, perhaps as a substructure. Call our universe <em>U</em>₁ and the meta-universe <em>U</em>₂. <em>U</em>₂ is not a “parallel universe;” instead, it is a universe with possibly different features than ours, but similar enough to host our universe within. <em>U</em>₂ obviously is not physically falsifiable. So, we are talking meta-physics. But we are exploring a more reasonable hypothesis than the Block Universe.</p><p>From the viewpoint of an observer in <em>U</em>₂, <em>U</em>₁ seems completely determined. But from within <em>U</em>₁, uncertainty is impassable. How about <em>U</em>₂? If <em>U</em>₂ is similar enough to our universe, then similar uncertainty will emerge within. Only from the viewpoint of a higher universe, <em>U</em>₃, will <em>U</em>₂ appear completely determined, just like <em>U</em>₁ will appear from the viewpoint of <em>U</em>₂.</p><p>Of course, there is no reason to stop at <em>U</em>₃; there might be an infinite nested universes, just as there might be an infinite parallel universes. But the Nested Universes Hypothesis offers a more consistent and integrated view of existence. It is able to resolve the tension between uncertainty within and certainty without.</p><h4>Welcome to Gödel’s Wonderland</h4><p>The idea of Nested Universes is inspired by Gödel’s Incompleteness Theorems. Roughly, a formal axiomatic system rich in arithmetic, <em>S</em>, cannot be consistent and complete. If it is consistent, then there must be an infinite number of statements within the system that are true but unprovable within the system.</p><p>However, these statements are provable within a higher system, <em>Sʹ</em>, within which unprovable statements will emerge, which can be proved only in a higher system, <em>S</em>ʺ, <em>ad infinitum.</em></p><p>Viewing each formal system as a universe, we can think of <em>U</em>₁ as <em>S; U</em>₂ as <em>Sʹ; U</em>₃ as <em>S</em>ʺ; <em>ad infinitum</em>.</p><p>Just like the incompleteness of <em>S</em> is resolved in <em>Sʹ,</em> but does not resolve incompleteness within <em>Sʹ</em> itself, the uncertainty within <em>U</em>₁ is resolved in <em>U</em>₂, but does not resolve the uncertainty within <em>U</em>₂ itself.</p><p>The inherent uncertainty within each system or universe allows for the structure to accommodate change and growth. If higher universes influence events in U₁, those influences would look to us like <strong>randomness, novelty, or free will</strong> — because they are not predictable from within <em>U</em>₁​. Those same events might be fully determined by the laws of <em>U</em>₂​, but this determinism is invisible to <em>U</em>₁-bound observers. If this nesting continues indefinitely, then no single level ever has a final, closed, fully deterministic picture — there is always a “bigger context” that could generate novelty relative to the lower level.</p><h4>Whitney’s Embedding Theorem</h4><p><a href="https://en.wikipedia.org/wiki/Hassler_Whitney">Hassler Whitney</a> (1907 — 1989) was an American mathematician who did foundational work in manifolds, embeddings, immersions, characteristic classes, and geometric integration theory.</p><p>Whitney published his foundational theorem on embedding manifolds in 1936 in his paper, “Differentiable Manifolds”. In this paper, he proved that any <em>n</em>-dimensional differentiable manifold can be embedded in <em>R</em>⁽²<em>ⁿ </em>⁺ ¹⁾. He further refined this result in 1944, showing that for <em>n </em>≥ 2, an embedding in <em>R</em>²<em>ⁿ</em> is possible.</p><p>In other words:</p><ul><li>If we have an abstract smooth manifold <em>M</em>, there exists a smooth, one-to-one embedding such that <em>M</em> can be thought of as a “sub-manifold” of some higher-dimensional Euclidean space.</li><li>This is a structural guarantee that any smooth space can be faithfully represented inside a higher-dimensional space without self-intersections.</li></ul><p>By applying Whitney’s theorem to the Nested Universes hypothesis, each universe <em>Uⁿ</em>​ can be seen as a kind of “manifold” — a spacetime with its own internal causal and physical structure.</p><p>There is always a possible “embedding” of <em>Uⁿ</em>​​ in a larger space <em>Uⁿ </em>⁺ ¹​ with more degrees of freedom, which accounts for events in <em>Uⁿ</em>​​​ that are undecidable from within. This makes <strong>uncertainty both fundamental and relative</strong>: it is built into the projection from <em>Uⁿ </em>⁺ ¹​​ → <em>Uⁿ.</em></p><h4>Complementarity?</h4><p>There seems to be a kind of complementarity between each universe and its meta-universe.</p><p><em>U</em>₂ provides stability and predictability to <em>U</em>₁. Without <em>U</em>₂, <em>U</em>₁ would be unstable and unsustainable. On the other hand, <em>U</em>₁ will have internal freedom and creativity. Without <em>U</em>₂, <em>U</em>₁​ would lack stability and coherence. Without <em>U</em>₁​, <em>U</em>₂​ would be a dead, frozen block — no process, no drama, no growth.</p><p>It is not clear how this complementarity relates to <a href="https://www.thecollector.com/niels-bohr-principle-of-complementarity/">Niels Bohr&#39;s</a>. But it can be viewed as an interpretation of it. The Nested Universes Hypothesis is not just a hierarchy but a partnership, in line with <a href="https://futurism.com/john-wheelers-participatory-universe">John Wheeler’s</a> idea of a “participatory universe.” Each level stabilizes the one below and gives it coherence; each lower level animates the one above and gives it unfolding dynamism.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/895/1*Iu75zXTX-JnBmGnHgALaDg.png" /><figcaption><a href="https://www.biography.com/scientist/niels-bohr">Niels Bohr</a> (left) and <a href="https://www.bbc.com/mundo/noticias-61307108">John Wheeler</a>.</figcaption></figure><p>This turns complementarity into a universal principle:</p><blockquote>Each level of reality requires both its “inside” and its “outside,” and they cannot be reduced to one another.</blockquote><p>This mirrors Bohr’s complementarity in a higher-order way: determinacy and indeterminacy, stability and freedom, outside-view and inside-view — all mutually exclusive in description, but jointly necessary for a complete account of reality.</p><h4>The Mystery of Time — Resolved?</h4><p>Special Relativity states that there is no global time within the universe. But we feel the “beat of time” even though we cannot identify it. St. Augustine <a href="https://www.loebclassics.com/view/augustine-confessions/1912/pb_LCL027.239.xml?readMode=recto">remarked</a>: “What, then, is time? If no one asks me, I know. If I wish to explain it to one that asks, I know not.”</p><p>From the viewpoint of <em>U</em>₂, the global time of <em>U</em>₁ is visible and quantifiable. Within <em>U</em>₂, however, time is undecidable; only from the viewpoint of <em>U</em>₃ that the global time of <em>U</em>₂ is visible, and so on.</p><p>Within each universe, only local time can be decided. Global time is undecidable. But from the viewpoint of the meta-universe, it becomes decidable. Ultimately, however, the future is genuinely open.</p><h4>Gödel vs. Gödel</h4><p>The Nested Universes Hypothesis builds on Gödel’s Incompleteness Theorems. But Gödel has developed his own cosmological theory of the universe.</p><p>Gödel offered his exact solution to General Relativity in a paper he contributed to a volume in honor of Einstein’s seventieth birthday in 1949. Essentially, Gödel’s “<a href="https://www.space.com/rotating-universe-would-permit-time-travel">Rotating Universe</a>” is rigidly rotating about every point. The most critical consequence of Gödel’s universe is the possibility of time travel via Closed Time-like Curves (CTC). The existence of CTCs challenges the fundamental notion of a universal, linear, and objective flow of time, as it implies that an event can influence its own past (Gleick, 2016). Gödel concluded that, in his rotating universe, change —and therefore time —is an illusion.</p><p>Empirical evidence, however, does not support Gödel’s theory. As John Stachel (2007) points out, Einstein was aware of the theoretical possibility of closed time-like curves in his General Theory of Relativity long before Gödel wrote his paper. But Einstein dismissed it on a purely physical basis.</p><p>In commenting on Gödel’s Universe, he wrote: “It will be interesting to weigh whether these [solutions] are not to be excluded on physical grounds.” According to Stachel (2007), Gödel’s arguments against time can be matched by similar arguments against the reality of space. But no one can deny the existence of change in space.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/412/0*LxjJC2eDi8rYa-Bf" /><figcaption>Einstein (left) and Gödel. Source: <a href="https://www.art.com/products/p15420424-sa-i3748089/leonard-mccombe-distant-of-mathematicians-albert-einstein-and-kurt-godel-taking-a-walk.htm?upi=P6YRPH0&amp;PODConfigID=4990176">art.com</a>.</figcaption></figure><p>As philosopher <a href="https://scholarworks.brandeis.edu/esploro/profile/palle_yourgrau/overview">Palle Yourgrau</a> (2005) points out in his book, <em>A World Without Time</em>, Gödel was at once a mathematical realist and a temporal idealist. Gödel concluded from his Incompleteness Theorems that the “Platonic realm” of mathematics cannot be captured by formal systems.</p><p>When it came to cosmology, noted Yourgrau, Gödel took the opposite direction. His solution to Einstein’s equations captured the whole reality of the universe.</p><p>Gödel’s rotating universe is a beautiful mathematical object, but it implies:</p><ul><li>Every event is both past and future of something else (via closed time-like curves).</li><li>Therefore, change and becoming are illusory — only the 4D block exists.</li></ul><p>This view, however, clashes with what his own incompleteness results suggest about knowledge and structure:</p><ul><li>A block universe has no undecidable truths — it’s complete and timeless.</li><li>Yet his theorems show that any system rich enough to describe such a universe cannot be complete from within.</li></ul><p>Gödel’s cosmology, it seems, violates his own epistemological lesson: that reality can’t be fully captured by any single system. Gödel was a great fan of <a href="https://plato.stanford.edu/entries/leibniz/">Gottfried Wilhelm Leibniz</a>, who firmly believed in time as a relational concept. Ironically, Gödel’s cosmological solution parts ways with the spirit of Leibniz’s philosophy. Gödel’s rotating universe implies that a structure can be so warped as to allow CTCs, thereby destroying the very concept of temporal succession and flow that grounds the Leibniz relationalist’s view of the reality of time and change.</p><p>The Nested Universes Hypothesis is an attempt to formulate a time-friendly theory of the cosmos based on Gödel’s Incompleteness Theorems. Gödel’s theorem supplies the <em>logic</em> of openness; Leibniz’s relationalism supplies the <em>metaphysics</em> of becoming.</p><h4>Conclusion</h4><p>The relativity of time implies the relativity of uncertainty, and vice versa. Uncertainty is both relative and objective. We propose the Nested Universes Hypothesis to reconcile seemingly contradictory views on time, determinism, and free will.</p><p>This hypothesis offers a novel perspective that bridges the gaps between these complex concepts:</p><ul><li><strong>Free will within, determinism without:</strong> We experience genuine openness because we cannot transcend our epistemic horizon, even though our universe may appear determined from a higher level.</li><li><strong>Local time measurable, global time mysterious:</strong> Special Relativity correctly denies universal simultaneity within our universe, yet our intuition of time’s reality reflects something objectively true at the meta-level.</li><li><strong>Uncertainty that is both absolute and relative:</strong> The limits of knowledge are physically real for each observer, yet these limits themselves vary across observers and across nested levels.</li></ul><p>By leveraging Gödel’s Incompleteness Theorems as a framework, we gain a deeper understanding of the interplay between time and uncertainty. Just as formal systems exhibit incompleteness that can only be transcended by ascending to higher systems, universes may exhibit temporal and causal incompleteness that can only be resolved from higher ontological levels.</p><p>This approach not only enriches our theoretical frameworks but also invites further exploration into fundamental questions: What is the nature of the “embedding” relationship between nested universes? Can this hypothesis make contact with physics through quantum mechanics or cosmology? Does infinite regress pose a problem or reveal a deep feature of reality?</p><p>The Nested Universes Hypothesis transforms metaphysical puzzles into structured hierarchies, suggesting that what appears paradoxical at one level may find resolution at the next — while acknowledging that complete resolution may lie forever beyond any finite perspective.</p><p><em>Special thanks to Francisco A. Doria for helpful discussions.</em></p><h4>References</h4><ul><li>Carroll, S. (2010). <em>From eternity to here: The quest for the ultimate theory of time</em>. Dutton.</li><li>Gleick, J. (2016). <em>Time travel: A history</em>. Vintage Books.</li><li>Mermin, N. D. (2009). <em>It’s about time: Understanding Einstein’s relativity</em>. Princeton University Press.</li><li>Palle, Y. (2005). <em>A world without time: The forgotten legacy of Gödel and Einstein</em>. Basic Books.</li><li>Popper, K. (1982). <em>The open universe: An argument for indeterminism</em>. Routledge.</li><li>Stachel, J. (2007). <em>A world without time: The forgotten legacy of Gödel and Einstein</em> [Book Review]. <em>Notices of the AMS</em>, August.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cfbdad477cd8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/relativity-of-uncertainty-cfbdad477cd8">Relativity of Uncertainty</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Time, Risk, and Ω]]></title>
            <link>https://medium.com/godelian-letters/time-risk-and-%CF%89-454882c48f83?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/454882c48f83</guid>
            <category><![CDATA[algorithmic-information]]></category>
            <category><![CDATA[global-financial-crisis]]></category>
            <category><![CDATA[special-relativity]]></category>
            <category><![CDATA[halting-problem]]></category>
            <category><![CDATA[omega]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Sat, 06 Sep 2025 11:12:16 GMT</pubDate>
            <atom:updated>2025-09-07T06:43:35.379Z</atom:updated>
            <content:encoded><![CDATA[<h4>Adventures in the Labyrinth of Incompleteness, Relativity, and Financial Instability</h4><p><em>The remarkable breakthroughs in logic, mathematics, and physics during the 20th century share critical insights that are particularly relevant to complex systems. These developments have been largely overlooked by mainstream economics. The price of this oversight is the flawed models that precipitated the Global Financial Crisis and may further contribute to future turmoil.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/931/1*mnxKkBAgreUMvPeDzJKjqw.png" /><figcaption>A medal designed by Stephen Wolfram and Hector Zenil in commemoration of the life and work of Gregory Chaitin in 2007. Source: <a href="https://www.automacoin.com/about.html">automacoin.com</a>.</figcaption></figure><h4>Time and Uncertainty</h4><p>Time and uncertainty can be seen as two sides of the same coin. Starting with simple probability statements, like “The probability of rain tomorrow is 70%,” we notice that any such statement must be about the future. So, uncertainty involves time.</p><p>How about the other way around? Does time involve uncertainty?</p><p>During the nineteenth century, it was commonly believed that, with full knowledge of the positions and velocities of all particles, the future was entirely deterministic. This concept was known as the “<a href="https://en.wikipedia.org/wiki/Laplace%27s_demon">Laplace Demon</a>.” During the 20th century, however, breakthroughs in science and mathematics left the “demon” largely unemployed.</p><p>Briefly, these developments include:</p><ul><li>Logical undecidability. The pioneering works of Kurt Gödel (1931) and Alan Turing (1936), among others, establish the fundamental limits of sufficiently rich mathematical systems (Byers, 2011). Karl Popper (1982) shows that such systems cannot consistently predict their own future.</li><li>Quantum uncertainty. In 1927, Werner Heisenberg formulated his Uncertainty Principle, which prevents complete foreknowledge of positions and velocities. Subsequent developments in quantum physics only strengthened this result. See Lindley (2008) and Carroll (2019).</li><li>Chaos. Many dynamical systems can be extremely sensitive to initial conditions. A tiny difference in the 8th or 9th decimal place could greatly affect the system’s behavior over long periods. See Gleick (2011) and Strogatz (2024).</li><li>Emergence. In many cases, the interaction of the parts results in properties absent from any of them. The whole becomes different from the sum of the parts. See Holland (2014) and Jensen (2022).</li><li>Fractals. Benoit Mandelbrot (1983) discovered that many phenomena, from tornadoes to stock markets, follow heavy-tailed distributions that increase the likelihood of extreme events compared to what the normal distribution predicts; see Nair et al. (2022). “Black Swans” are examples of such extreme events.</li><li>Collectively, these breakthroughs illustrate a world governed by what became known as “non-ergodic” systems, where no amount of past data can fully predict the system’s future behavior. More on this later.</li></ul><p>The overall result of these factors is that the future is fundamentally unpredictable. It can only be predictable <em>locally</em> or <em>qualitatively</em>. However, global predictability is logically and physically impossible. Therefore, when it comes to the future, time and uncertainty are two sides of the same coin.</p><h4>Relativity of Uncertainty</h4><p>In 1905, Einstein published his theory that became known as the Special Theory of Relativity (STR). In simple terms, STR states that there is no universal “now.” Simultaneity is relative, and therefore, there is no universal clock. If time and uncertainty are two sides of the same coin, then we should also expect a relativity of uncertainty. But what does that precisely mean?</p><p>Consider the most basic form of uncertainty: tossing a coin. Suppose we have an ideal machine that tosses a coin with equal probabilities of landing on a Head or a Tail. Moreover, each toss is completely independent of the previous and the next. The machine can perform any number of tosses without wear or bias.</p><p>Now, imagine another machine at a different location, and we want to ensure it is just as fair as the first. How would we verify this? We need to synchronize the two machines. For instance, we might instruct both machines to perform 10,000 tosses <em>simultaneously</em> and then compare the results. We calibrate the second machine as needed to ensure that the results are consistent.</p><p>The limiting distribution of heads and tails is generally independent of how long it takes to generate samples. However, fairness also requires that each toss be independent of the previous and the next. That independence is a property of the <em>process in time</em>. To compare two machines or verify fairness across space, we must align their generating processes in real-time. But aligning processes requires synchronized clocks. And clock synchronization at a distance inevitably depends on conventions about the one-way speed of light. From this perspective, process synchronization requires time synchronization — and that is precisely where relativity comes into play.</p><p><em>The result is striking</em>: if two tossing machines exist at different locations, they cannot be perfectly synchronized to generate perfectly consistent random distributions. This suggests that there is no universal tossing machine (no universal random-number generator), just as there is no universal clock. Randomness is relative, just as time is. Since coin tossing is the most basic form of uncertainty, the absence of a universal tossing machine means there is no universal probability distribution.</p><blockquote>Just as there is no universal clock, there is no universal coin-tossing machine.</blockquote><h4>Frame-dependent Uncertainty</h4><p>The relativity of uncertainty implies, <em>inter alia</em>, that two observers in different frames of reference may not agree on the nature of a random process. One observer may see independence of observations and symmetry in the distribution, while the other may infer correlations, skewness, or even heavy tails — and vice versa. This discrepancy is not due to a lack of data or miscalculations, nor is it an optical illusion. It is genuine relativity: each observer’s account is internally consistent and correct within their own frame of reference.</p><p>This is of the exact same nature as <a href="https://www.youtube.com/watch?v=K3MymDSzKZ0">time dilation</a> in Special Relativity: an observer in a relatively stationary frame views the clock in a moving frame as ticking more slowly than it appears to the moving observer. Both perspectives are equally valid, though they cannot be reconciled into a single universal ticking rate.</p><p>Likewise, in the relativity of uncertainty, one observer may see independence and symmetry in a random process, while another may see correlation or skewness. Each perspective is valid within its own frame, but there is no single universal description that all frames share.</p><p>The slowdown of time (or time dilation) is genuine and affects <em>all</em> natural processes. This is why, in the <a href="https://en.wikipedia.org/wiki/Twin_paradox">Twin Paradox</a>, the traveling twin returns to Earth biologically younger than his stationary twin.</p><p>If randomness is fundamentally tied to temporal sequencing, then a change in relative time flow will directly alter how processes of uncertainty unfold. Just as every clock slows in a moving frame, every random generator’s process changes with motion, leading observers to disagree about its nature (symmetry vs. skew, independence vs. correlation).</p><p>This disagreement arises because the temporal structure of randomness depends on the observer’s frame of reference. Consequently, what appears random in one frame may exhibit patterns or biases in another, challenging our understanding of randomness as an absolute property.</p><p>This relativity of uncertainty, where no universal random number generator can exist, finds an algorithmic parallel in Gregory Chaitin’s Ω, to which we now turn.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eODX3nph1hmTUV87LqM1EQ.png" /><figcaption>Source: <a href="https://www.scientificamerican.com/">Scientific American</a>.</figcaption></figure><h4>Enter Ω</h4><p>The absence of a universal probability distribution is surprising. Even more surprising is that this was (indirectly, at least) shown decades ago when Gregory Chaitin introduced his infamous number, Ω, in 1975.</p><p>What is Ω?</p><p>In short, Ω is the probability that a randomly chosen program in a universal programming language, <em>U</em>, will halt. Equivalently, it can be written as the weighted sum of the halting probabilities of all halting programs, where shorter programs contribute more weight than longer ones.</p><p>To illustrate, imagine an algorithmic economy <em>U</em>,<em> </em>where each program represents an investment project. If the program halts, the project is successfully completed and achieves its target. Otherwise, the project fails — it runs indefinitely, consuming resources but never producing final output.</p><p>This may suggest that we can use Ω to design an ideal investment portfolio. Unfortunately, this is not possible. While Ω is well-defined, it is uncomputable. More on this point later.</p><p>Ω has two levels of uncomputability:</p><ul><li>Turing-level uncomputability. This is the unsolvability of the Halting Problem established by Alan Turing and Alonzo Church in 1936. Since Ω is a weighted average of all halting programs in <em>U</em>, identifying these programs is undecidable.</li><li>Chaitin-level uncomputability. Ω encodes all the information about these programs in a maximally compressed manner. There is no program sufficiently shorter than Ω that can generate the digits of Ω. If there were, then the shorter program could be used to solve the Halting Problem.</li></ul><h4>Ω as a Universal Probability Distribution</h4><p>Ω encodes the probability distribution of all halting programs in <em>U</em>. Since halting is the precondition for extracting any information from computation, Ω represents the most basic kind of probability within a universal language. Any other informative probability distribution necessarily presupposes halting, and thus depends on Ω. If Ω is undecidable, then every other global distribution is necessarily undecidable too.</p><p>The undecidability of Ω reinforces our earlier conclusion about the absence of a universal random distribution. Probability is decidable only locally. Only a limited number of Ω bits can ever be computed.</p><p>Since any probability statement presupposes a temporal “before” and “after,” it follows that a universal timeline is also undecidable within <em>U</em>. In other words, Ω’s uncomputability might be considered as an algorithmic counterpart of Einstein’s relativity.</p><h4>Universality and the Speed of Information</h4><p>There appears to be a fundamental connection between universality and the speed of information transmission. If a universal property were fully decidable within a system, then in principle it would be available to all parts of the system simultaneously. That is, its propagation speed would be instantaneous.</p><p>But non-trivial properties (see below) necessarily involve non-trivial information, and information requires order. Algorithmic information, in particular, is built from sequences of digits whose content depends on their order. If all the digits were transmitted instantly, the order would collapse, and with it the information.</p><p>To preserve order, <a href="https://medium.com/godelian-letters/information-the-language-of-nature-and-markets-212e5a8b78b2">information must travel at a finite speed</a>. Yet once propagation is finite, the property cannot be simultaneously accessible everywhere in the system. This makes the universal property undecidable from within the system.</p><p>In this sense, undecidability is not an abstract quirk of formal logic but the direct consequence of a <em>trade-off between speed and order</em>: instantaneous transmission destroys information, while finite transmission destroys universality. The tension reveals a deep interplay between information, time, and system-wide coherence — one that underlies both Algorithmic Information Theory and Special Relativity.</p><blockquote>Instantaneous transmission destroys information, while finite transmission destroys universality.</blockquote><p>How does this relate to Ω?</p><p>Ω encodes the halting probabilities of all possible programs in a universal language. Each term in Ω corresponds to a halting program, weighted by the probability of generating it randomly.</p><p>The <strong>first term</strong> in this expansion corresponds to the shortest valid halting programs. From a temporal perspective, these are the programs that halt the fastest. Their probability weight is therefore tied to the <strong>shortest achievable halting time</strong> in the system.</p><p>Since probability presupposes time — every probability statement refers to the likelihood of an event <em>before</em> it occurs — this first halting term implicitly sets a lower bound on duration, and thus <em>an upper bound on the speed of information</em>. In other words, Ω indirectly encodes the maximum speed at which ordered information can propagate in the system.</p><ul><li>If information could propagate instantaneously, all events would collapse into simultaneity and probability would lose meaning.</li><li>If propagation is finite, then the first slice of Ω defines the earliest halting events, and by extension, the fastest possible information transmission.</li></ul><p>Thus, the speed of light in physics and the first halting probabilities in Ω play parallel roles: both enforce that information has a finite maximum speed, preserving order and preventing collapse into indeterminacy.</p><p><a href="https://medium.com/godelian-letters/gregory-chaitin-meets-albert-einstein-e0bb236cefb0">Gregory Chaitin Meets Albert Einstein</a></p><p>As we discussed in an earlier article, the logical upper limit on the speed of information in the (algorithmic) universe is undecidable from within our own programming language <em>U</em>. This is fully consistent with the undecidability of the one-way speed of light in Einstein’s framework.</p><h4>The Ivory Tower of Efficient Markets</h4><p>How does all this impact the world of economics and finance?</p><p>Chaitin’s Ω may seem like the most abstract concept imaginable. Economic theories, such as the <a href="https://www.investopedia.com/terms/e/efficientmarkethypothesis.asp">Efficient Market Hypothesis</a> (EMH), may seem, in contrast, closer to the real world.</p><p>Guess what? It is probably the other way around!</p><p>The works of Gödel, Turing, and Chaitin, among others, are technically abstract, true, but in substance, they highlight the fundamental limits of reason. The identification of these limits was a key driver behind the most advanced revolutions in human history: the development of electronic computers and the semiconductor industry. These developments culminated in the current AI revolution.</p><p>The Efficient Market Hypothesis, although it contributed to the progress of modern markets, was a key driver behind the persistent increase in financial instability. This culminated in the <a href="https://www.investopedia.com/articles/economics/09/financial-crisis-review.asp">Global Financial Crisis of 2008</a> (GFC). The EMH is still a foundational theory in modern finance despite mounting empirical evidence against it; see Shiller (2015).</p><p>Since it is considered an ideal theory, we can evaluate the key assumptions of EMH against modern mathematical logic. The key assumptions of EMH are:</p><ul><li>Uncertainty follows a normal (or log normal) distribution.</li><li>All agents assume this distribution in their models.</li><li>Information is transmitted instantaneously throughout the market.</li></ul><p>For decades, Benoit Mandelbrot has criticised the first assumption. Prices do not generally follow a normal distribution. Instead, they mostly follow heavy-tail distributions whereby extreme events tend to occur much more frequently than the normal distribution would predict. See Mandelbrot (2004) and Taleb (2010).</p><p>This explains why financial models consistently underestimated risk, which led to excessive leverage and speculation that fueled the bubble ahead of the GFC.</p><h4>Crash-free Models Triggered the Crash</h4><p>The first assumption would be harmless if not all agents adopted it. Apparently, there seems to be no reason to worry if all agents incorporate normal distribution or market efficiency into their models. If these global properties hold, why not embed them explicitly into the models?</p><p>Before Gödel, the answer would have been affirmative: There should be no harm in embedding these global properties into the model. But not after the discovery of the incompleteness phenomena.</p><p>Gödel’s Second Incompleteness Theorem, in particular, states, roughly, that a formal axiomatic system, rich in arithmetic, cannot assert its own consistency if it were to remain consistent.</p><p>Henry Rice (1953) extended the domain of Gödel’s Theorem from systems’ consistency to semantic, <a href="https://en.wikipedia.org/wiki/Rice%27s_theorem">non-trivial properties</a> of the system. Newton da Costa and Francisco Doria (2014) extended it further to many areas of mathematics, including mathematical economics.</p><p>These works demonstrate that for a mathematical model to remain consistent, it cannot assert a non-trivial, global property within itself. Doing so renders the model inconsistent. Despite the importance of this result, economic modeling is largely blind to this matter. Market efficiency, general equilibrium, normal distribution, and other global market properties are routinely assumed within economic models. But now we know that the resulting models <em>will be</em> inconsistent.</p><p>How does this logical inconsistency translate to economic behavior?</p><p>The short answer was provided by <a href="https://en.wikipedia.org/wiki/Jean-Philippe_Bouchaud">Jean-Philippe Bouchaud</a> (2008):</p><blockquote>“Ironically, it was the very use of a crash-free model that helped to trigger a crash.”</blockquote><p>As we have seen, the ubiquitous adoption of the normal distribution led to the systematic underestimation of risk, which itself led to the excessive buildup of risk and, eventually, to the crash.</p><p><a href="https://medium.com/godelian-letters/adam-smith-meets-kurt-g%C3%B6del-the-incompleteness-of-the-invisible-hand-f9e276342754">Adam Smith Meets Kurt Gödel: The Incompleteness of the Invisible Hand</a></p><p>It should be obvious now that the results of mathematical logic are not ivory-tower intellectual games. They have serious implications for economic modeling and market stability.</p><h4>Back to Ω</h4><p>A key insight of Chaitin’s Ω is that a universal probability distribution is undecidable within any universal programming language. This implies that economic models cannot assume a universal probability distribution (normal or otherwise) within these models. Doing so makes these models inconsistent, leading to mispricing of risks and thus market instability.</p><p>Moreover, Ω demonstrates why the third assumption— namely, the instantaneous transmission of information—is flawed. If universal probability distributions are undecidable, then the dissemination of information must be bounded by finite speed. This speed limit is not due to frictions, transaction costs, or irrationality; it is intrinsic to the logic of algorithmic information itself. In parallel, Einstein’s relativity tells us the same story physically: no signal propagates faster than light.</p><p>Taken together, AIT and STR show that financial markets cannot be universally efficient or instantaneously synchronized. Universality collapses under its own weight, leaving us with fragile and volatile markets. Instability isn’t just an accident of bad regulation or irrational actors — it is a result of the flawed architecture of sophisticated financial models.</p><h4>Ω Efficiency</h4><p>Suppose that each program in a universal language <em>U</em> represents an investment project in an algorithmic economy, as highlighted earlier. A program halting represents the project’s success, i.e., achieving its desired target.</p><p>In this case, Ω represents the ideal portfolio of the entire market. It incorporates all information about successful projects and their complexity-based probability weights in the most efficient manner possible. Ω achieves maximum information density, as it is algorithmically incompressible.</p><p>The EMH, in its strongest form, assumes that all information about all projects or assets is efficiently and instantly available across the market. Hence, an efficient portfolio of all market assets must be linked to Ω.</p><p>But Ω is uncomputable. No algorithm, mechanism, or process — including the collective action of a market — can calculate its value. Therefore, such an ideal index is fundamentally inaccessible <em>from both the inside and outside the market</em>. It follows that the market cannot be perfectly information-efficient.</p><blockquote>EMH implies that a market portfolio should track Ω. Since Ω is uncomputable, the market cannot be information-efficient.</blockquote><p>While EMH treats the market <em>as if</em> it tracks Ω, the reality is closer to fragments of Ω— locally useful approximations, but never the whole picture.</p><p>Although the EMH is no longer considered an accurate description of markets, it is still considered the ideal benchmark for analysis, particularly when it comes to behavioral “biases” or “anomalies.” We argue that, in the ideal world, the EMH is logically flawed.</p><p>How about the weak form of EMH, where only past information is reflected in current prices?</p><p>At best, this version of EMH corresponds only to fragments of Ω. These fragments are locally computable and useful, but they are not globally decidable. Since “market efficiency,” even in its weak form, is a non-trivial semantic property of the system, Rice’s theorem applies: such a property cannot be decided from within the system itself.</p><p>Embedding the EMH — even in weak form — into market players’ models renders those models logically inconsistent. This inconsistency leads to systematic mispricing of risks and, consequently, contributes to financial instability.</p><h4>A Perpetual Market Machine</h4><p>Suppose that strong EMH holds. Then Omega must be computable. If Omega were computable, then the halting problem would be solvable. What does that imply for the market?</p><p>Solving the halting problem means successful projects can be systematically identified from non-successful ones. But to do so means the market is able to generate net wealth at zero risk. Based on “<a href="https://www.lindahall.org/about/news/scientist-of-the-day/simon-stevin/">Stevin’s Principle</a>,” this means the market becomes a <a href="https://unacademy.com/content/upsc/study-material/physics/all-about-perpetual-motion-machine-of-the-second-kind-pmm2/">perpetual motion machine of the second type</a> (PMM2).</p><p><a href="https://en.wikipedia.org/wiki/Simon_Stevin">Simon Stevin</a> (1548–1620) was a contemporary of Galileo, one of the early scientists to recognize the impossibility of perpetual motion. Stevin’s principle states that a frictionless machine cannot produce net work unless it undergoes a net change. If the initial and final states of the machine are indistinguishable, the machine cannot produce work without external energy; otherwise, it is a PMM2 (Schumacher, 2010; Al-Suwailem, 2017).</p><p>To generate a riskless return, the capital must at the end of the period be in the same state at which it started. But this is impossible without external energy — capital cannot generate work without undergoing change. If it can, it becomes a PMM2.</p><p>In his book, <em>Meta Math!</em>, Chaitin (2005, p. 169) states: “I don’t think that you can really understand a mathematical result until you find your own proof.” I think the above provides a kind of “new proof,” so to speak, of the impossibility of solving the halting problem. If it were solvable, a PMM2 would become feasible. Since a PMM2 is impossible, the halting problem must be unsolvable.</p><h4>A Non-ergodic World</h4><p>Assuming a universal probability distribution implies that the future is already accounted for in the past. The past, the present, and the future are all governed by one universal probability distribution.</p><p>This is known as the “ergodicity hypothesis.” It states that the time average of a dynamical system tends to converge to the ensemble (space) average taken over all possible states (Gowers et al., 2008). Accordingly, future economic behavior can be reliably predicted (actuarially known) through a sufficient statistical analysis of already existing data.</p><p>A non-ergodic process is one where the future is not simply an extension of the statistical patterns of the past. A system is non-ergodic if it exhibits strong path-dependence, so time averages depend on initial conditions or early shocks. Complex systems, in general, are globally non-ergodic, whereby time and space averages may remain unequal even in the long run. These kinds of systems lead to what <a href="https://en.wikipedia.org/wiki/John_Kay_(economist)">John Kay</a> and <a href="https://en.wikipedia.org/wiki/Mervyn_King,_Baron_King_of_Lothbury">Mervyn King</a> (2020) describe as <em>Radical Uncertainty.</em></p><p>The undecidability of Ω suggests that algorithmic complex systems are non-ergodic from the perspective of agents within the system. While global averages might exist or be verifiable from outside the system, they cannot be computed or relied upon by internal observers. Thus, for practical purposes, such systems exhibit non-ergodic behavior. This explains why models assuming ergodicity (e.g., EMH) can fail: agents cannot access the “God’s-eye view” of global averages or other universal properties of the system.</p><p><a href="https://en.wikipedia.org/wiki/Paul_Davidson_(economist)">Paul Davidson</a> (2007 ) argues extensively for the rejection of the ergodicity hypothesis in economic systems. Nobel laureates <a href="https://www.nobelprize.org/prizes/economic-sciences/1972/hicks/facts/">John Hicks</a>, <a href="https://www.nobelprize.org/prizes/economic-sciences/1993/north/facts/">Douglas North</a>, and <a href="https://www.nobelprize.org/prizes/economic-sciences/1987/solow/facts/">Robert Solow</a> all agree that the economy, in general, is a non-ergodic system.</p><p>It is interesting to note how non-ergodicity relates to time relativity.</p><p>Ergodicity assumes that time and space averages are interchangeable in the long run. However, Special Relativity reveals that time and space are not symmetric; their relationship depends on the observer’s frame of reference. This asymmetry challenges the notion of a universal “long run” in which time averages can converge to space averages for all observers. This aligns well with the conclusions of previous sections.</p><h4>Summary</h4><ul><li>Time and uncertainty are two sides of the same coin: probability inherently involves future events, while the relativity of time implies that uncertainty is also frame-dependent.</li><li>The undecidability of a universal probability distribution, exemplified by Chaitin’s Ω in a universal programming language, parallels time relativity, showing that no universal random mechanism exists from within the system.</li><li>Mainstream EMH models violate these insights by assuming normal distributions and instantaneous information transmission within the models, ignoring logical and physical limits.</li><li>This leads to a Gödelian inconsistency, where models embedding universal, non-trivial properties become logically flawed. The result of this inconsistency is summarized by Bouchaud: Crash-free models helped to trigger the crash.</li><li>EMH implies that the index of total market value tracks Ω. Since Ω is uncomputable, the market cannot be information-efficient.</li><li>Economic systems are generally non-ergodic, meaning their future cannot be systematically predicted from past statistical patterns for embedded agents. This reinforces the relativity of time and, by extension, the relativity of uncertainty.</li></ul><h4>Conclusion</h4><p><a href="https://en.wikipedia.org/wiki/Arthur_Eddington">Arthur Eddington</a> (1939), in his book, <em>Philosophy of Physical Science</em>, notes:</p><blockquote>“It has always been the aim of scientific inquiry to trace a common cause underlying diverse phenomena.”</blockquote><p>This essay has embraced that aim, exploring common principles underlying diverse phenomena. Here, we have argued that the deep limits revealed by Gödel, Turing, and Chaitin are not mere abstractions but are inextricably linked to the physical limits of relativity and, consequently, to the inherent fragility of financial markets.</p><p>This journey has revealed a clear path forward: only by acknowledging the fundamental boundaries of complex systems — whether in logic, physics, or economics — can we hope to build resilient and stable markets. History has shown the cost of ignoring these limits: financial disasters born of overconfidence in flawed models.</p><p><em>I benefited from extensive discussions with GPT-5, Gemini 2.5 Pro, and DeepSeek 3.1 during the development of this essay.</em></p><h4>References</h4><ul><li>Al-Suwailem, S. (2017). Is economics a science? In F. A. Doria (Ed.), <em>The limits to mathematical modeling in the social sciences: The significance of Gödel’s incompleteness phenomenon</em> (pp. 207–265). World Scientific.</li><li>Bouchaud, J.-P. (2008). Economics needs a scientific revolution. <em>Nature, 455</em>(7217), 1181.</li><li>Byers, W. (2011). <em>The blind spot: Science and the crisis of uncertainty</em>. Princeton University Press.</li><li>Carroll, S. (2019). <em>Something deeply hidden: Quantum worlds and the emergence of spacetime</em>. Dutton.</li><li>Chaitin, G. (1997). <em>Information, randomness, &amp; incompleteness, papers on algorithmic information theory</em> (2nd ed.). World Scientific.</li><li>Chaitin, G. (2005). <em>Meta math! The quest for Omega</em>. Vintage eBooks.</li><li>da Costa, N. C. A., &amp; Doria, F. A. (2014). On an extension of Rice’s theorem and its applications in mathematical economics. <em>Advances in Austrian Economics, 18</em>, 237–257.</li><li>Davidson, P. (2007). <em>John Maynard Keynes</em>. Palgrave Macmillan.</li><li>Eddington, A. (1939). <em>The philosophy of physical science</em>. Cambridge University Press.</li><li>Gleick, J. (2011). <em>Chaos: The making of a new science</em>. Viking.</li><li>Gowers, T., Barrow-Green, J., &amp; Leader, I. (Eds.). (2008). <em>The Princeton companion to mathematics</em>. Princeton University Press.</li><li>Holland, J. (2014). <em>Complexity: A very short introduction</em>. Oxford University Press.</li><li>Jensen, H. J. (2022). <em>Complexity science: The study of emergence</em>. Cambridge University Press.</li><li>Kay, J., &amp; King, M. (2020). <em>Radical uncertainty: Decision-making beyond the numbers</em>. W. W. Norton &amp; Company.</li><li>Lindley, D. (2008). <em>Uncertainty: Einstein, Heisenberg, Bohr, and the struggle for the soul of science</em>. Doubleday.</li><li>Mandelbrot, B. (1983). <em>The fractal geometry of nature</em>. W.H. Freeman &amp; Co.</li><li>Mandelbrot, B., &amp; Hudson, R. (2004). <em>The misbehavior of markets: A fractal view of risk, ruin, and reward</em>. Basic Books.</li><li>Nair, J., Wierman, A., &amp; Zwart, B. (2022). <em>The fundamentals of heavy tails: Properties, emergence, and estimation</em>. Cambridge University Press.</li><li>Popper, K. (1982). <em>The open universe: An argument for indeterminism</em>. Routledge.</li><li>Schumacher, B. (2010). <em>Impossible: Physics beyond the edge</em>. The Great Courses. <a href="https://www.thegreatcourses.com/">www.thegreatcourses.com</a>.</li><li>Shiller, R. J. (2015). <em>Irrational exuberance: Revised and expanded third edition</em>. Princeton University Press.</li><li>Strogatz, S. (2024). <em>Nonlinear dynamics and chaos: With applications to physics, biology, chemistry, and engineering</em>. CRC Press.</li><li>Taleb, N. (2010). <em>The black swan: The impact of the highly improbable</em> (2nd ed.). Random House.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=454882c48f83" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/time-risk-and-%CF%89-454882c48f83">Time, Risk, and Ω</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Gregory Chaitin Meets Albert Einstein]]></title>
            <link>https://medium.com/godelian-letters/gregory-chaitin-meets-albert-einstein-e0bb236cefb0?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/e0bb236cefb0</guid>
            <category><![CDATA[einstein]]></category>
            <category><![CDATA[speed-of-light]]></category>
            <category><![CDATA[omega]]></category>
            <category><![CDATA[relativity]]></category>
            <category><![CDATA[algorithmic-information]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Thu, 14 Aug 2025 16:53:25 GMT</pubDate>
            <atom:updated>2025-09-07T05:58:43.045Z</atom:updated>
            <content:encoded><![CDATA[<h4>Parallels Between Algorithmic Information Theory and the Special Theory of Relativity</h4><p><em>Einstein’s Special Theory of Relativity (STR), in substance, is an algorithmic theory of the speed of information. It is natural to ask how it relates to Algorithmic Information Theory (AIT). The analysis reveals striking parallels between the undecidability of the one-way speed of light and algorithmic undecidability, suggesting deep connections between relativity and computational limits.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/985/1*y0JRVZ74Rcc3CfyPbmM9lA.png" /><figcaption><a href="https://www.manhattanrarebooks.com/pages/books/2327/albert-einstein-yousuf-karsh/portrait-photograph-of-albert-einstein-signed-by-yousuf-karsh">Albert Einstein</a> (left) and <a href="https://www.youtube.com/watch?v=RKW_WgtB1oQ&amp;t=202s">Gregory Chaitin</a>.</figcaption></figure><h4>Is the Speed of Light Decidable?</h4><p>We all know that the speed of light is 299,792,458 meters per second. Isn’t it already determined?</p><p>In fact, the speed of light was <a href="https://en.wikipedia.org/wiki/Speed_of_light"><em>defined</em></a> in 1983 to be 299,792,458 m/s. The meter itself was redefined as the distance that light travels in 1/299,792,458 seconds. In this sense, it is decided. But it is not decided either mathematically or empirically — it is merely a convention.</p><p>To be precise, it is possible to empirically measure the round-trip speed of light, i.e., the speed from point <em>A</em> to point <em>B</em> and back again. But can we measure the speed of light from <em>A</em> to <em>B</em>? Surprisingly, the answer is no.</p><p>In this article, I argue that this impossibility is not only empirical but logical as well. Specifically, Gregory Chaitin’s Information Incompleteness Theorem may provide a solid foundation not only for the undecidability of the one-way speed of light, but also for key pillars of the Special Theory of Relativity.</p><h4>One Way? No Way!</h4><p><a href="https://www.veritasium.com/about">Derek Muller</a> elaborates in this <a href="https://www.youtube.com/watch?v=pTn6Ewhb27k">video</a> on why it is not physically possible to measure the one-way speed of light.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FpTn6Ewhb27k%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DpTn6Ewhb27k&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FpTn6Ewhb27k%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/a08dddf04626a70ed61bcc829d5e62e9/href">https://medium.com/media/a08dddf04626a70ed61bcc829d5e62e9/href</a></iframe><p>Here is a simple way to explain it:</p><ul><li>Suppose we set up two clocks: one at point <em>A</em>, the other at point <em>B.</em> The distance between the two points is known.</li><li>We fire a beam of light from <em>A</em> toward <em>B.</em></li><li>When the beam arrives, the clock at <em>B</em> records the arrival time and sends it back to <em>A.</em></li></ul><p>Since the distance is fixed, it seems we can calculate the speed of light, both one-way and round-trip. Right?</p><p>Not quite.</p><p>Here is the catch: To measure the one-way speed accurately, we must be sure that the two clocks, at their respective locations, are perfectly synchronized. But synchronizing distant clocks requires assuming something about the one-way speed of the very signal used to synchronize them. That’s the circularity: we can measure a one-way speed only if we’ve already assumed a one-way speed.</p><p>The only way around this problem is to measure the <em>round-trip</em> time. We install a mirror at point <em>B</em>, send a light beam from <em>A</em> to <em>B</em> and back, and measure the total time elapsed on a single clock at <em>A</em>. Then we simply <em>assume</em> that the speed of light is the same from <em>A</em> to <em>B </em>as from <em>B </em>to <em>A.</em></p><p>It should be noted that empirical measurements of the speed of light, from <a href="https://en.wikipedia.org/wiki/Ole_R%C3%B8mer">Ole Rømer</a>’s in 1676 to present-day experiments, have all been based on two-way (round-trip) travel times. No experiment has directly measured the <a href="https://en.wikipedia.org/wiki/One-way_speed_of_light">one-way speed</a> without relying on prior synchronization assumptions or conventions.</p><p>Einstein identified this problem in his 1905 paper (emphasis original):</p><blockquote><em>“We have not deﬁned a common “time” for A and B, for the latter cannot be deﬁned at all unless we establish </em>by definition<em> that the “time” required by light to travel from A to B equals the “time” it requires to travel from B to A.”</em></blockquote><p>In other words, the one-way speed of light — as we know it — is not measured or deduced. It is a convention.</p><h4>The Wonderland of Relativity</h4><p>The Special Theory of Relativity (STR) assumes key features of the speed of light: (i) the impossibility of measuring the one-way speed; (ii) the round-trip speed is empirically measurable; and (iii) this round-trip speed is constant for all observers regardless of their motion.</p><p>These properties have profound consequences; only a brilliant mind like Einstein could have uncovered them. For example:</p><ul><li>Simultaneity is not absolute but relative to the observer’s motion and chosen synchronization. There is no universal “now.”</li><li>Two events that happen “at the same time” for you might happen at different times for someone moving past you.</li><li>Moving clocks run slower relative to stationary ones.</li><li>A traveling twin ages more slowly than their stationary twin (the famous twin paradox).</li></ul><p>See Taylor and Wheeler (1992) and Muller (2016).</p><p>The relativity of simultaneity is <em>not</em> an extra assumption of Special Relativity. It is the unavoidable consequence of the fact that the one-way speed of light cannot be measured without already assuming it.</p><h4>Simultaneity is a Convention</h4><p>Simultaneity is at the center of our time-related activities. And it is at the core of Special Relativity. Einstein admits that “By means of a revision of the concept of simultaneity in a shapable form I arrived at the special relativity theory” (Jammer, 2006, p. 3).</p><p>In his book, <em>Einstein’s Clocks, Poincaré’s Maps</em>, Peter Galison (2003) documents how scientists and engineers in the nineteenth century struggled to define time objectively but failed. Both Einstein and Poincaré were grappling with the same issue: How do you define a common time across space when signals take time to travel? And both came to see that <em>simultaneity is not discovered, but defined.</em></p><p>For example, if you ask Google: “At 8:00 a.m. in London, what time is it in Jakarta?” it will answer: 2:00 p.m. We tend to think that the answer is based on a purely technical basis — something handled by satellites, atomic clocks, and engineering. But in reality, these systems are built on conventions.</p><p>Without an agreed-upon prime meridian (Greenwich), there would be no absolute way to define longitude — and thus, no meaningful way to say what “8:00 a.m. in London” even means globally. Moreover, the entire system of time zones, like the system of longitude and latitude, is a geopolitical and historical agreement, not a discovery of nature.</p><p>Without these conventions, there’s no coherent way to define “now” in two places — or even to know whether a message arrived “on time.” The reason behind these conventions is the undecidability of the one-way speed of light. See Jammer (2006) and Janis (2018).</p><p>The conventional nature of simultaneity stems from the undecidability of the one-way light speed, a concept that echoes undecidability in formal logic, as we’ll explore next.</p><h4>A Logic Journey</h4><p>It may seem odd that the one-way speed of light is undecidable, while the round-trip speed is both measurable and verifiable. But is this distinction logical?</p><p>Let’s turn to formal, first-order logic for insight. Consider the following question:</p><blockquote>For arbitrary first-order propositions A and B, is there a general algorithm to decide whether the implication <strong>A → B</strong> is logically valid?</blockquote><p>The answer is no. This is the well-known Decision Problem or the <a href="https://en.wikipedia.org/wiki/Entscheidungsproblem"><em>Entscheidungsproblem</em></a>. In 1936, Alan Turing and Alonzo Church, independently, showed that there is no general algorithm that can decide, for arbitrary statements, whether one formally implies the other. This was one of the key discoveries that defined the limits of computability.</p><p>But now, for the same propositions <em>A </em>and <em>B</em>, consider the following question:</p><blockquote>Is the statement <strong>A → B → A</strong> decidable?</blockquote><p>Surprisingly, not only is it decidable — it is a tautology; it is always true. (The connective <strong>→</strong> is right-associative, parsing the statement as <strong><em>A</em> → (<em>B</em> → <em>A</em>)</strong>, resulting in a tautology.) You can verify this using any formal logic engine, such as <a href="https://www.wolframalpha.com/input?i=If+A+then+B+then+A">Wolfram|Alpha</a>.</p><p>This highlights a key distinction: From a logical standpoint, a round-trip (from <em>A</em> to <em>B</em> and back to <em>A</em>) is fundamentally different from a one-way journey. This suggests that <em>open processes require external reference frames (which introduce undecidability), while closed processes are self-contained (enabling decidability)</em>.</p><p>Similarly, there is a key empirical distinction between a round-trip and a one-way signal. Just as a round-trip signal (like a light beam sent from <em>A</em> to <em>B</em> and reflected back) is empirically measurable, a round-trip implication in logic is structurally closed — and therefore decidable.</p><h4>The Algorithmic Einstein</h4><p>Although a rigorous definition of an <em>algorithm</em> was not formalized until the 1930s, Einstein’s approach in his 1905 paper was, at its core, algorithmic.</p><p>An algorithm is a finite sequence of ordered, deterministic steps to achieve a specific goal. That is precisely what Einstein focused on when analyzing how to measure the speed of light. He insisted that time must be measured through the synchronization of clocks via exchanged signals. His procedure wasn’t abstract — it was concrete, operational, and explicitly step-by-step.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XF0BE7MFxWpUgzfhFguhxQ.png" /><figcaption>Einstein at the Swiss Federal Patent Office (right), 1904. Source: <a href="https://www.ige.ch/en/about-us/the-history-of-the-ipi/einstein/einstein-at-the-patent-office">ige.ch</a>.</figcaption></figure><p>As Peter Galison (2003) notes, Einstein was likely influenced by his work at the Swiss Federal Office for Intellectual Property (1902–1909). There, patent applications required precise procedural descriptions of how inventions worked. Combined with the rapid pace of technological innovation — particularly in telecommunications and electromechanical timing — Einstein developed a procedural mindset that many physicists at the time had overlooked.</p><p>According to Galison:</p><blockquote>“There is a sense in which Einstein’s special theory of relativity has always been a machine, an imaginative one to be sure, but one suspended in a constantly evolving real skein of wires and pulses that synchronized time by the exchange of electromagnetic signals.”</blockquote><p>The algorithmic approach of Special Theory of Relativity (STR) opens the door to possible connections with the modern Algorithmic Information Theory (AIT), to which we now turn.</p><h4>Enter Chaitin</h4><p><a href="https://medium.com/godelian-letters/interview-with-gregory-chaitin-b4e63460c94b">Gregory Chaitin</a> is one of the pioneers who independently established <a href="https://en.wikipedia.org/wiki/Algorithmic_information_theory">Algorithmic Information Theory</a> together with <a href="https://en.wikipedia.org/wiki/Andrey_Kolmogorov">Andrey Kolmogorov</a> and <a href="https://en.wikipedia.org/wiki/Ray_Solomonoff">Ray Solomonoff</a>.</p><p>AIT aims to quantify the amount of information in mathematical expressions. As symbols are converted into a series of numbers, the theory seeks to detect general patterns in those numbers. Once those patterns are detected, the series can be compressed into a shorter program, whereby all the information of the original series is condensed into the shorter, patternless one. The original series is considered “random” only if it cannot be compressed into a shorter program.</p><p>AIT is concerned with the smallest program that can produce the original message. The length in bits of such a program that can produce message <em>x</em> is called the complexity of <em>x</em>, usually denoted as <em>K</em>(<em>x</em>). The complexity is therefore equal to the size in bits of the minimal programs producing the series.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pCM_wyaWnNwrMtXc28Jc9w.png" /><figcaption>Left to right: <a href="https://onionesquereality.wordpress.com/2009/12/20/ray-solomonoff-is-no-more/">Ray Solomonoff</a>, <a href="https://onionesquereality.wordpress.com/2009/12/20/ray-solomonoff-is-no-more/">Gregory Chaitin</a>, and <a href="https://valeman.medium.com/andrey-kolmogorov-master-of-uncertainty-and-order-1080618e1afe">Andrew Kolmogorov</a>.</figcaption></figure><p>Chaitin proved his own version of the Incompleteness Theorem based on the nature of algorithmic information. To demonstrate that a particular series of digits is random, we must prove that no small program exists to produce it. He shows that any formal axiomatic system, <em>S</em>, has an inherent limit. The system cannot prove that a string of digits has a complexity much greater than its own. Specifically, if <em>S</em> has a complexity of <em>N</em>, it cannot prove any string has a complexity greater than <em>N</em> plus a small constant.</p><p>Specifically, suppose <em>S</em> has complexity <em>K</em>(<em>S</em>) = <em>N</em>.<em> </em>In that case, it is impossible to prove that a particular series of binary digits is of complexity greater than <em>N</em> + <em>c</em>, where <em>c</em> is a constant that is independent of <em>S</em>. Hence, there will always be numbers whose randomness cannot be proved within <em>S</em>.</p><blockquote>A $150-theorem cannot be derived from a $100-axioms.</blockquote><p>The essence of Chaitin’s theorem is that a theory cannot <em>prove</em> that a statement contains more information than the information embedded in that theory. It is a kind of conservation law of information. In simple terms, if one has “$100-axioms” and a “$150-theorem,” then that theorem cannot be derived from those axioms (Chaitin, 1997, p. 113).</p><h4>Complexity and Speed of Information</h4><p>Measuring the complexity of information requires using an abstract programming language for expressing strings and making computations. This is done by fixing a Universal Turing Machine, call it UTM-<em>U</em>.</p><p>To simplify the discussion, consider an incompressible (random) binary string <em>x </em>with<em> </em>length <em>n </em>bits using the language of UTM-<em>U</em>. According to AIT, the complexity of <em>x</em> is (Chaitin, 1997, p. 173):</p><p><em>K</em>ᵤ(<em>x</em>) = <em>n </em>+ <em>K</em>ᵤ(<em>n</em>) + <em>c</em>ᵤ.</p><p>This means that the length of the shortest program to produce <em>x</em> equals <em>K</em>ᵤ(<em>x</em>), which is the number of bits needed to communicate the message <em>x</em>.</p><p>Suppose that the time needed to transmit each bit of <em>K</em>ᵤ(<em>x</em>) for a distance <em>d</em> in a vacuum is <em>δt</em>. Then the speed of transmitting <em>K</em>ᵤ(<em>x</em>) is <em>v </em>= <em>d</em> / (<em>K</em>ᵤ(<em>x</em>)<em>δt</em>).<em> </em>Since <em>K</em>ᵤ(<em>x</em>) measures the <em>shortest </em>program to produce <em>x</em>, then <em>v</em> is the <em>maximum </em>speed for transmitting <em>x</em>.</p><p>For a positive amount of information and transmission time, the maximum speed of information must be finite. That is, as long as <em>n </em>&gt; 0 and <em>δt </em>&gt; 0, then <em>v</em> &lt; ∞.</p><p>Practically, <em>n</em> will be much greater than 0. For example, the ASCII code for the word “Hi” is: 0100100001101001, which is 16 bits.</p><h4>Logical Limit of Information Speed</h4><p>But why do we have to have <em>δt </em>&gt; 0? Why should the time for transmitting each bit be finite?</p><p>Algorithmic information consists of <em>ordered </em>digits. For example, the ASCII binary code for the word “Hi” consists of 6 zeros and 10 ones. Rearranging the same number of zeros and ones, we can obtain a different word, “Be”:</p><ul><li>“Hi”: 01001000 01101001</li><li>“Be”: 01000010 01100101</li></ul><p>The two words have exactly the same number of 1s and 0s, but in different orders — obviously, order matters.</p><p>The transmission of ordered symbols requires that there be a maximum limit to the speed of information. Why?</p><p>Suppose it is possible to transmit information at an infinite speed. Then all symbols will be transmitted instantaneously. But this means it will not be possible to tell which symbol comes before the other. At excessively high speeds, bits will become effectively indistinguishable. <em>With infinite speed, order will be lost</em>. If order is lost, information is lost.</p><p>To preserve information, therefore, information must travel at a finite speed. A speed limit is necessary to maintain order and, thus, information. This means that we must have <em>δt </em>&gt; 0. Hence, we reach a crucial conclusion:</p><blockquote>The maximum speed of algorithmic information must be finite.</blockquote><p>This represents a major departure from Newtonian mechanics, which assumes no inherent limit on the speed of information (Taylor and Wheeler, 1992).</p><p>Next, we ask: Is this maximum speed decidable? Given a programming language UTM-<em>U</em>, and for a fixed <em>n</em>, <em>K</em>ᵤ(<em>x</em>) is then known and decidable. But how do we know the “programming language” of the universe?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*2QifzKN2A5X5vIt6bH6Fiw@2x.jpeg" /><figcaption>The spiral galaxy UGC 11397. Source: <a href="https://esahubble.org/images/potw2525a/">esahubble.org</a>.</figcaption></figure><h4>An Algorithmic Universe</h4><p>Many scientists and philosophers have explored the idea that the universe is fundamentally algorithmic — that its behavior can be described, at some level, as the execution of well-defined computational rules.</p><p>This view extends from Gottfried Wilhelm Leibniz all the way to Gregory Chaitin, David Deutsch, Edward Fredkin, Seth Lloyd, John Wheeler, and Stephen Wolfram, among others. See Chaitin (2005), Lloyd (2006), and Zenil (2013).</p><p>As we have already seen, Special Relativity is essentially an algorithmic theory. We may, therefore, conjecture that the universe itself operates according to algorithmic rules. In that light, modeling the cosmos as a giant computer system provides a powerful — and practically useful — approximation for exploring its fundamental limits.</p><p>Let us adopt this view and assume that the universe operates on its own Universal Turing Machine, which we’ll call UTM-<em>V. </em>Can we figure out the algorithmic complexity associated with UTM-<em>V</em>, namely <em>K</em>ᵥ(<em>x</em>)? Assuming <em>δt </em>is<em> </em>known, the value of <em>K</em>ᵥ(<em>x</em>) sets a lower bound on the length of any program that generates a given message. It thus connects directly to the maximum possible speed of information in the universe.</p><p>Here’s the problem: if we, as observers, are ourselves using another UTM-<em>U</em> to model reality, then from within UTM-<em>U</em>, we cannot fully deduce the internal code of UTM-<em>V</em>. This is a direct consequence of algorithmic undecidability — the same barrier that prevents us from solving the Halting Problem in full generality. From our perspective, <em>K</em>ᵥ(<em>x</em>)​ is undecidable.</p><p>Combined with the result of the previous section, we conclude that the maximum speed of information, according to AIT, satisfies three key properties:</p><ol><li><strong>It is finite</strong> — for a positive amount of information and transmission time.</li><li><strong>It is universal</strong> — assuming the universe runs on a fixed UTM-<em>V</em>.</li><li><strong>It is undecidable</strong> — because we cannot algorithmically deduce <em>K</em>ᵥ(<em>x</em>) from within our UTM-<em>U</em>.</li></ol><p>But these are <em>precisely</em> the same properties that the one-way speed of light exhibits in Einstein’s 1905 theory — reinforcing the deep connection between relativity and algorithmic limits.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S2WLDAX2sidjtMnW46VFfg.png" /><figcaption>From AIT to STR.</figcaption></figure><h4>The Mysterious Omega!</h4><p>Chaitin shows that <em>K</em>(<em>x</em>) is related to one of the most mysterious numbers of mathematics known as <a href="https://mathworld.wolfram.com/ChaitinsConstant.html">Chaitin’s number</a> — Ω.</p><p>What is Ω?</p><p>Chaitin argues that one must look not just at the size <em>K</em>(<em>x</em>) of the smallest (self-delimited) program that calculates <em>x</em> in a given language, but also at the algorithmic probability <em>P</em>(<em>x</em>) that a program picked at random calculates <em>x</em>. It turns out that the size of the smallest program for calculating <em>x</em> is essentially −log₂ of its algorithmic probability, the probability of calculating <em>x</em> by chance (in the language of UTM-<em>V</em>):</p><p><em>K</em>ᵥ(<em>x</em>) = −log₂<em>P</em>ᵥ(<em>x</em>) + <em>c</em>ᵥ.</p><p>Adding up the probabilities of all (self-delimited) programs that produce <em>x </em>over all <em>x</em>, we get the probability of <em>all</em> halting programs in UTM-<em>V</em>. That is, Ωᵥ = Σ<em>ₓ P</em>ᵥ(<em>x</em>). Ωᵥ, therefore, is the halting probability of all programs in UTM-<em>V. </em>And so, <em>P</em>ᵥ(<em>x</em>) is the slice of Ωᵥ for a specific message <em>x</em>.</p><p>According to Chaitin, Ωᵥ is logically irreducible. It cannot be compressed into any formal axiomatic system. For this reason, it is maximally uncomputable, although it may <em>look</em> computable.</p><p>Omega is a meta-mathematical construct — an oracle-like real number that encodes all halting truths about its universal Turing machine from an external vantage point. Like Gödel’s “I am not provable” sentence, it captures the limits of formal systems: it can be systematically defined but never fully computed or proved within any formal system in the language of the same UTM.</p><p>If the universe is algorithmic, then the <em>cosmic speed limit</em> for transmitting a given message <em>x</em> is determined by its algorithmic probability <em>P</em>ᵥ(<em>x</em>), which is part of the decomposition of Ωᵥ. We therefore may <em>speculate</em> that:</p><blockquote>The cosmic speed limit of information might be deeply connected to the most mysterious mathematical number, Omega.</blockquote><p><a href="https://medium.com/godelian-letters/time-risk-and-%CF%89-454882c48f83">Time, Risk, and Ω</a></p><h4>Can We Make a Round-trip?</h4><p>Can we figure out the “round-trip” speed of the universe’s UTM-<em>V</em>?</p><p>We can observe the effective round-trip time it takes to encode, transmit, decode, and confirm a message. This behavior is operationally verifiable and empirically stable, but <em>we cannot prove it reflects the maximum speed</em>. To do so would require deducing UTM-<em>V</em>’s internal rules — specifically, whether <em>K</em>ᵥ(<em>x</em>) for the observed process is minimal. This is algorithmically impossible from within the observers’ UTM-<em>U</em>.</p><p>Just as we measure light’s round-trip speed operationally, we can infer UTM-<em>V</em>’s effective computation time by observing physical processes (e.g., particle interactions, signal propagation). Yet both measurements reveal only the empirical outcome — not the underlying theoretical limit.</p><p>While we can measure the round-trip speed of light and treat it as a universal limit, we cannot <em>deduce</em> the true value of <em>K</em>ᵥ(<em>x</em>)​ because it is algorithmically undecidable.</p><h4>Conclusion</h4><p>Algorithmic Information Theory is not purely a mathematical theory. It provides an epistemological unification of mathematics and science, as Chaitin (2023) argues.</p><p>This article highlights how this unification could be applied. Our exploratory analysis reveals striking similarities between AIT and STR: both frameworks point to a finite, universal, yet undecidable limit on the one-way speed of information. This suggests that both AIT and STR converge on a foundational principle: the existence of a universal constraint on information transmission.</p><p>Perhaps the deepest lesson is that the universe’s most profound truths emerge not from what we can know, but from what we cannot. In this sense, Einstein and Chaitin are indeed kindred spirits — both revealing that <em>undecidability is not a barrier to understanding, but the very foundation of it</em>.</p><p><em>I am grateful to Gregory Chaitin for constructive comments. Any errors or omissions are my sole responsibility. The title of this article is metaphorical; Chaitin did not physically meet Einstein.</em></p><h4>References</h4><ul><li>Chaitin, G. (1997). <em>Information, randomness, &amp; incompleteness, papers on algorithmic information theory</em> (2nd ed.). World Scientific.</li><li>Chaitin, G. (2005). <em>Meta math! The quest for Omega</em>. Vintage eBooks.</li><li>Chaitin, G. (2023). <em>Philosophical Mathematics: Infinity, Incompleteness, Irreducibility.</em> Institute for Advanced Studies, <a href="https://cch.um6p.ma/wp-content/uploads/2024/01/Philosophical_Mathematics_Infinity_Incom_compressed-1.pdf">UM6P</a>, Benguerir, Morocco.</li><li>Einstein, A. (1905). On the electrodynamics of moving bodies. <em>Annalen Der Physik</em>, <em>17</em>, 891–921.</li><li>Galison, P. (2003). <em>Einstein’s clocks, Poincaré’s maps: Empires of time</em>. W. W. Norton &amp; Company.</li><li>Jammer, M. (2006). <em>Concepts of simultaneity: From antiquity to Einstein and beyond</em>. The Johns Hopkins University Press.</li><li>Janis, A. (2018). Conventionality of simultaneity. In E. N. Zalta (Ed.), <em>The Stanford Encyclopedia of Philosophy</em> (Fall 2018 Edition). Stanford University. <a href="https://plato.stanford.edu/">plato.stanford.edu</a></li><li>Lloyd, S. (2006). <em>The programmable universe: A quantum computer scientist takes on the cosmos</em>. Alfred A. Knopf.</li><li>Muller, R. (2016). <em>Now: The physics of time</em>. W.W. Norton &amp; Company.</li><li>Taylor, E., &amp; Wheeler, J. (1992). <em>Spacetime physics</em>. W. H. Freeman and Company.</li><li>Zenil, H. (Ed.). (2013). <em>A computable universe: Understanding and exploring nature as computation. </em>World Scientific.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e0bb236cefb0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/gregory-chaitin-meets-albert-einstein-e0bb236cefb0">Gregory Chaitin Meets Albert Einstein</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Can a Free Market Be Complete?]]></title>
            <link>https://medium.com/godelian-letters/can-a-free-market-be-complete-95134ff8cbad?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/95134ff8cbad</guid>
            <category><![CDATA[efficient-market]]></category>
            <category><![CDATA[undecidability]]></category>
            <category><![CDATA[incompleteness-theorem]]></category>
            <category><![CDATA[finite-automata]]></category>
            <category><![CDATA[free-market]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Sat, 05 Jul 2025 11:31:42 GMT</pubDate>
            <atom:updated>2025-07-06T09:28:58.611Z</atom:updated>
            <content:encoded><![CDATA[<p><em>A market system can be either free or complete, but cannot be both</em>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*g9y0iE1QWnX1cs6Donep0w.jpeg" /><figcaption>A 3D maze. Source: Shutterstock.</figcaption></figure><h4>Dreams of a Complete Market</h4><p>Is the market a panacea for all economic ills? Can we trade <em>anything </em>in the market?</p><p>This question lies at the heart of what is known as “<a href="https://medium.com/godelian-letters/the-mirage-of-complete-markets-e9301a540b36">complete markets</a>.” A complete market, in essence, is a world where every imaginable good can, in principle, be traded in every possible situation at competitive prices.</p><p>All trades in this imaginary world occur at the initial time (time 0). Agents trade state-contingent claims for consumption at all future dates and states, effectively agreeing on contracts that specify payoffs contingent on every possible future event. After this initial trade, no further trading occurs; the contracts are honored as states unfold over time. In such an imaginary world, every risk can be hedged, and every opportunity can be seized.</p><p>It is generally accepted by both economists and lawyers that almost all contracts are incomplete (Halonen-Akatwijuka and Hart, 2024). However, they do not agree on <em>why</em> this is the case or how to deal with it. In most studies, market incompleteness is assumed to be exogenously given due to imperfections or bounded rationality (see Magill and Quinzii, 2008).</p><p>In his survey of incomplete contracts, Nobel laureate <a href="https://www.nobelprize.org/prizes/economic-sciences/2014/tirole/facts/">Jean Tirole</a> (1999) remarks:</p><blockquote>“Incomplete contracting arguably underlies some of the most important questions in economics and some other social sciences … For all its importance, there is unfortunately no clear definition of “incomplete contracting” in the literature. While one recognizes one when one sees it, incomplete contracts are not members of a well-circumscribed family.”</blockquote><p><a href="https://en.wikipedia.org/wiki/John_Moore_(economist)">John Moore</a> (2016), moreover, notes: “We’ve had 25 years to come up with a watertight theory of contractual incompleteness and we haven’t succeeded yet. Eventually, someone will.”</p><p><a href="https://en.wikipedia.org/wiki/Francisco_D%C3%B3ria">Francisco A. Doria</a> and I (2020) argue that it is possible, in principle, to provide a rigorous framework for formulating the problem of incomplete contracting. The results, as we shall see, are remarkable.</p><h4>Contracts as Algorithms</h4><p>Let us start by examining the nature of contracts.</p><p>A contract, in essence, is a set of “if–then” statements. This means we can view contracts as algorithms or computer programs. This seems to be an interesting progress in formulating market interactions. If transactions are carefully designed and properly encoded as computer programs, then we can have a perfect and complete market. Right?</p><p>Wrong!</p><p>The idea that <em>if</em> we can overcome human weaknesses and rely instead on supercomputers to perform market transactions, <em>then</em> we can have a perfect market — rests on a deeply flawed concept. It was proved to be flawed not by economists but by first-class mathematicians almost a century ago, as we shall see.</p><p>An economy of robots, as Nobel laureate <a href="https://www.nobelprize.org/prizes/economic-sciences/1995/lucas/facts/">Robert Lucas</a> (1988) envisioned, will not solve all market problems nor answer all the difficult questions. A perfectly algorithmic market will face insurmountable limits imposed by pure logic.</p><h4><strong>Bug Hunt</strong>!</h4><p>Consider the following question:</p><p>A computer program <em>Q</em> is designed to screen through all publicly traded common shares worldwide, analyze each company, score them based on a predefined set of criteria, and print out a ranked list of all these shares. How do we know that this program is “perfect”? Put differently, how do we know that the program has no “bug” or will not produce a false ranking?</p><p>That’s easy, you might say. We need to develop another program, <em>V</em>, to verify the output of program <em>Q.</em> But how do we know that <em>V </em>is also bug-free? Another program, <em>V</em>₁, which, in turn, requires yet another program, V₂, …, <em>ad infinitum. </em>This means that the accuracy of program <em>Q</em> is algorithmically unverifiable.</p><p>This might seem obvious, but it was formally proven independently by <a href="https://en.wikipedia.org/wiki/Alonzo_Church">Alonzo Church</a> and <a href="https://en.wikipedia.org/wiki/Alan_Turing">Alan Turing</a> in 1936, long before the invention of electronic computers. Their work shows that the question: ‘Would program <em>P </em>execute and halt within a finite time?’ cannot, in general, be answered upfront by inspecting the code of the program. And it cannot be answered by simply running the program because we will never know when it will halt, if ever.</p><p>The results of Church and Turing follow the lead of <a href="https://en.wikipedia.org/wiki/Kurt_G%C3%B6del">Kurt Gödel</a> and <a href="https://en.wikipedia.org/wiki/Alfred_Tarski">Alfred Tarski</a>, among others, establishing the impenetrable limits of logic (see Davis, 2000).</p><p><a href="https://en.wikipedia.org/wiki/Henry_Gordon_Rice">Henry Rice</a> generalizes Church and Turing’s results, showing how the same problem arises with any <em>non-trivial property</em> of computer programs. A non-trivial property is a property that:</p><ol><li>Cannot be decided by simply inspecting the code (like if the code has the word ‘stop’ or the number 999, for example), and</li><li>The property holds for some programs but not others.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/615/1*-YXkcdttdEuAXrUpEBzlhQ@2x.png" /><figcaption><a href="https://www.cle.unicamp.br/cle/newton-carneiro-affonso-da-costa-nota-de-falecimento">Newton da Costa</a> (left) and <a href="https://www.math.buffalo.edu/mad/PEEPS2/doria_franciscoA.html">Francisco Doria</a>.</figcaption></figure><p><a href="https://en.wikipedia.org/wiki/Newton_da_Costa">Newton da Costa</a> and Francisco Doria (1991) further extend Rice’s Theorem to broad areas of mathematics and applied fields like mathematical economics, demonstrating that certain “interesting” or “useful” properties in these domains are generally undecidable, just as Rice’s Theorem shows for non-trivial semantic properties of programs.</p><p>Their work suggests that for various economic models and dynamic systems, algorithmic unpredictability exists for certain policies or properties, thereby challenging ideas of complete economic planning or prediction (da Costa and Doria, 2014).</p><p>These works show that, if we are interested in non-trivial properties of market systems, we are destined for disappointment: The vast majority of these properties will be undecidable; no algorithm can provide a definite answer to many of these questions.</p><h4>When Algorithms Go Wild: Flash Crashes</h4><p>Even if the computer program can execute normally, it will not be possible to tell in advance how it may respond to different stimuli. Programs may behave unexpectedly in some circumstances, even if the code was technically bug-free.</p><p>If market players were ideal computer agents formulating their decision plans as ideal computer programs, then such ideal agents cannot systematically predict the outcome of their plans (Al-Suwailem et al., 2018). A notable example is the phenomenon known as “Flash Crashes,” one of the most dramatic and mysterious phenomena in modern financial markets.</p><p>A flash crash is a rapid, deep, and volatile decline in security prices that occurs within a short time period, followed by a quick recovery. These events typically happen within minutes or even seconds, with prices plummeting dramatically before rebounding during the same trading day.</p><p>The most famous “flash crash” attributed to algorithmic trading occurred on May 6, 2010, in the United States, wiping out initially about $1 trillion. The DJIA recovered only about 70% of its decline by the end of the trading day (see the graph below).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/397/1*i39tSuV9C5FmRt37ya3b5g.png" /><figcaption>Dow Jones Industrial Index on May 6, 2010. Source: <a href="https://en.wikipedia.org/wiki/2010_flash_crash">Wikipedia</a>.</figcaption></figure><p>Based on data from <a href="https://www.nxcoredata.com/">Nanex</a>, mini flash crashes have <a href="https://www.researchgate.net/profile/Erhan-Bayraktar/publication/327119815_Mini-Flash_Crashes_Model_Risk_and_Optimal_Execution/links/5c4f713892851c22a39765f1/Mini-Flash-Crashes-Model-Risk-and-Optimal-Execution.pdf?origin=publication_detail&amp;_tp=eyJjb250ZXh0Ijp7ImZpcnN0UGFnZSI6InB1YmxpY2F0aW9uIiwicGFnZSI6InB1YmxpY2F0aW9uRG93bmxvYWQiLCJwcmV2aW91c1BhZ2UiOiJwdWJsaWNhdGlvbiJ9fQ">exceeded</a> 20,000 since 2006. A compiled list of the largest flash crashes is available <a href="https://claude.ai/public/artifacts/2f4895ca-c28a-4cd3-a948-6a85a7fa14cf">here</a>.</p><p>In his book, <em>Wait: The Art and Science of Delay </em>(2012), Frank Partnoy reports that a high-frequency trading firm relocated its offices to minimize the delay in order execution time, but the result was higher costs. It’s like a traffic system where being the first car at every light actually makes your commute slower because you hit all the lights at the wrong timing sequence. In complex systems simulations, having a higher speed limit may result in more congestion and, consequently, a lower average speed of cars (Gribbin, 2004). <a href="https://ccl.northwestern.edu/netlogo/models/TrafficBasic">Agent-based models</a> illustrate this phenomenon.</p><p>A related phenomenon is <a href="https://en.wikipedia.org/wiki/Braess%27s_paradox">Braess’s Paradox</a>. This paradox demonstrates that adding one or more roads to a road network can actually slow down the overall traffic flow through it. More formally, the idea behind Braess’s discovery is that the <a href="https://en.wikipedia.org/wiki/Nash_equilibrium">Nash equilibrium</a> may not equate with the best overall flow through a network.</p><blockquote>In complex environments, deterministic algorithms can result in unpredictable, emergent behavior.</blockquote><p>The moral of these examples is that in complex environments, deterministic algorithms can result in unpredictable, emergent behavior. Being algorithmic does not guarantee being predictable — it might be exactly the opposite.</p><h4>Market Boundaries</h4><p>The above discussion concludes that contracts, as algorithms, cannot be fully predictable and thus cannot be complete. Let us consider the market as a whole. Could it be complete?</p><p>If by completeness we mean that <em>any</em> object can be traded in a market, then it is obvious that the market cannot be complete in this sense; we cannot trade life, love, or beauty, for example.</p><p>However, it would be too restrictive to define market completeness in this manner. Instead, let us consider a less restrictive criterion: A market is complete if it is capable of identifying its own boundaries. That is, it can determine which objects can be traded in the market and which cannot. This appears to be a fairly realistic criterion.</p><p>Suppose we can design an algorithm that systematically identifies all contractable objects. By being contractable, an object becomes priceable and therefore tradable in the market. This implies that we can classify all objects in the economy into two mutually exclusive sets:</p><ul><li><em>A</em> = { <em>x</em> : <em>k</em>(<em>x</em>) = 1 }, the set of all contractable objects, and</li><li><em>B</em> = { <em>x </em>: <em>k</em>(<em>x</em>) = 0 }, the set of all non-contractable ones.</li></ul><p>The algorithm <em>k</em>(<em>x</em>), if exists, decides which objects are contractable and thus tradable in the market. A market cannot be considered complete if it is unable to solve such an important problem.</p><p>The algorithm <em>k</em>, however, must itself be contractable; otherwise, it will not be possible to identify the subject matter of the contract. That is, a contract must specify, for example, that “the counterparty must deliver the contractable object <em>x</em> such that <em>k</em>(<em>x</em>) = 1. Delivery of a non-contractable object is not acceptable.”</p><p>This implies that the set <em>B</em> is contractable because it is constructed using the algorithm <em>k,</em> which itself is contractable<em>.</em> This means that <em>k</em>(<em>B</em>) = 1, and thus <em>B</em> belongs to <em>A. </em>But this contradicts the assumption that the two sets are mutually exclusive. Accordingly, the function <em>k</em> cannot exist. No algorithm can universally identify which objects are contractable and which are not.</p><p>The undecidability of market boundaries reinforces the incompleteness of contracts since it is impossible to systematically identify the subject matter of the contract.</p><h4>Does It Matter?</h4><p>When markets are incomplete, competitive equilibrium may not exist in general, and even if it does, it may not be Pareto optimal, as Nobel laureate <a href="https://www.nobelprize.org/prizes/economic-sciences/2016/hart/facts/">Oliver Hart</a> (1975) argues. In addition, he demonstrates how, in an incomplete market, permitting some trades that were previously prohibited may make everybody worse off. Furthermore, equilibria of incomplete markets are highly sensitive to agents’ beliefs, leading to “<a href="https://en.wikipedia.org/wiki/Sunspots_(economics)">sunspot equilibria</a>,” where any inefficient outcome might be an equilibrium, as <a href="https://en.wikipedia.org/wiki/Jean-Jacques_Laffont">Jean-Jacques Laffont</a> (1989) points out.</p><p><a href="https://en.wikipedia.org/wiki/Samuel_Bowles_(economist)">Samuel Bowles</a> (2016) argues that the incompleteness of markets cannot, in general, be remedied by taxes and subsidies. Trying to complete the market when it is inherently <em>incompletable</em> might make things even worse. Attempting to complete an incompletable system from within is not only futile but potentially <a href="https://medium.com/godelian-letters/the-mirage-of-complete-markets-e9301a540b36">destabilizing</a>.</p><p>To ﬁll in the gaps of an incomplete market, we need trust and norms that transcend the market mechanism. Nobel laureate <a href="https://www.nobelprize.org/prizes/economic-sciences/1972/arrow/facts/">Ken Arrow</a> (1969) notes that social norms, including moral and ethical codes, serve to compensate for market incompleteness. Nobel laureate <a href="https://en.wikipedia.org/wiki/Amartya_Sen">Amartya Sen</a> (2017), moreover, remarks that the ultimate guarantee of individual liberty may rest not on rules, but on individual values.</p><h4>Make It Complete, Please</h4><p>So far, we have identified two sources of market incompleteness:</p><ol><li>Undecidability of non-trivial properties of algorithmic contracts.</li><li>Undecidability of market boundaries.</li></ol><p>Let us now reverse the argument and ask the following question: Suppose we can write algorithmically complete contracts. What would these contracts look like?</p><p>An algorithmically complete contract is a contract whose output is completely decidable. That is, we can decide all non-trivial properties of the outcome of the contract. This can be possible if the language of the contract is less powerful than standard computer languages we are familiar with.</p><p>Computer languages like C++, Python, and Java are examples of Universal Turing Machines, which are languages with maximum logical capabilities. According to the <a href="https://plato.stanford.edu/entries/church-turing/">Church-Turing Thesis</a>, no logical language can exceed the logical power of a Universal Turing Machine (UTM). Such languages are capable of performing all computable tasks. That is, any task that, in principle, can be implemented in a precise sequence of mechanical steps can be computed using a UTM.</p><p>But the price of maximum logical power is undecidability. Non-trivial properties of UTMs in general are undecidable, as we have seen.</p><p>Is there a computing language for which all properties are decidable? The answer is yes. <a href="https://www.geeksforgeeks.org/theory-of-computation/introduction-of-finite-automata/">Finite Automata</a> (FA) are computing structures that have limited logical powers, but for which almost every property is decidable (Reiter and Johnson, 2013).</p><p>Finite Automata are unable to encode and execute all computable tasks. Some tasks that can be organized mechanically in a step-by-step procedure, and thus computable by a UTM, will be beyond the capability of FA.</p><p>Hence, if contracts were in the form of Finite Automata, they would be complete. But in this case, the resulting market — now logically complete — can no longer implement many computable trades. Hence, it must be Pareto-inefficient relative to a UTM-based benchmark.</p><p>An economy of Finite Automata will be inefficient in the sense that there will be many Pareto-improving algorithmic arrangements for trade and production that the FA will systematically fail to perform. Thus, a complete market in this sense does not correspond to the efficient competitive market familiar in economic textbooks. An economy of Universal Turing Machines, on the other hand, will be able, according to the Church-Turing thesis, to capitalize on all <em>computable</em> tasks. See Table 1.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/538/1*oBXtUdGSV6VXqGWMQIpc1g@2x.png" /><figcaption>Table 1</figcaption></figure><p>This leads to a fundamental conclusion:</p><blockquote>A market system can be either efficient or complete, but not both.</blockquote><p>An incomplete market would also be inefficient, but not in a systematic manner. That is, there will be opportunities to achieve Pareto–superior allocations, but these allocations will not be generally computable within the system. However, with proper institutions, these inefficiencies can be reduced but never entirely eliminated. This illustrates why entrepreneurship plays a vital role in identifying Pareto-improving opportunities and enhancing market productivity.</p><h4>Ownership</h4><p>So far, we have defined what a complete market or an efficient market is. How about a “free market”?</p><p>A very practical definition is: A free market is a market in which private ownership is positively valued. Although it may have state ownership, private ownership is the norm.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/793/1*gLmC3TU6H1TggM7agMYCPA@2x.png" /><figcaption><a href="https://hart.scholars.harvard.edu/">Oliver Hart</a> (left) and <a href="https://www.slmath.org/people/17267">Sanford Grossman</a>.</figcaption></figure><p>Next, what is ownership?</p><p>Oliver Hart (2016) describes how, with <a href="https://en.wikipedia.org/wiki/Sanford_J._Grossman">Sanford Grossman</a>, they discovered the relationship between the incompleteness of contracts and the value of ownership. Why would a ﬁrm decide to sign a contract with another ﬁrm obtaining its services rather than buying it?</p><p>In a complete market with complete contracts and zero transaction costs, there is no need for firm ownership because every conceivable good or service produced by the firm can be contracted separately, thereby reducing total costs. In a complete market, why do we need firms in the first place?</p><p>Hart (2016, p. 372) reports that he and Grossman worked on the difference between ﬁrms and contracts for ten very intense days; “these were ten days that shook my world,” Hart writes. The result is that <em>ownership is valuable because it fills the gap in incomplete contracts</em>. If a contract is incomplete, then the right to decide about the missing parts of the contract is the residual control or decision right.</p><p>To be sure, when you buy a good or service, you own it. But why own the asset or capital producing that good? If you can fully specify all the details of the service you are after, why worry about the owner of the capital? It doesn’t matter if the owner is an individual or the government. Capital ownership will simply be irrelevant.</p><p>If we define a free market as one where private ownership prevails, then it does matter who owns the capital. Yet, if the market were complete, ownership would become irrelevant! So you can see the tradeoff between completeness and private ownership that defines a free market. This leads to another fundamental conclusion:</p><blockquote>A market system can be either free or complete, but not both.</blockquote><p>If contract incompleteness is viewed as simply arising from frictions or inefficiencies, then in an ideal world, a market system is indistinguishable from a centralized system. If they were fundamentally different, then there must be a significant reason why contracts are often incomplete.</p><p>Undecidabilities established by Gödel, Church, and Turing, among others, provide exactly that answer. In a perfectly logical world, markets are inherently incomplete, and thus, ownership is inherently valuable.</p><blockquote>In a perfectly logical world, markets are inherently incomplete, and thus, ownership is inherently valuable.</blockquote><p>If we think that private ownership is an indispensable property of a free market, then a complete market cannot be a free market, since in a complete world, ownership is dispensable.</p><p>In fact, in a complete world, ownership <em>must</em> be dispensable, because it is a constraint on the system. If the market is complete, such a constraint should result in an inefficient performance (in addition to the inefficiency resulting from the limited powers of FA, as discussed above). In contrast, the ownership institution contributes to optimal allocation of resources in incomplete markets (Table 2).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/549/1*gcESYBuJelw-ZIS00ZnNZg@2x.png" /><figcaption>Table 2</figcaption></figure><h4>Reason, Markets, and Free Will</h4><p>Can we have a free market without free will?</p><p>A great deal of ink has been spilled debating the problem of free will. However, the fact that each party to the debate aims to convince the other of their particular view is itself clear evidence that we have a choice. Without choice, the debate is pointless.</p><p>Nobel laureate <a href="https://en.wikipedia.org/wiki/Milton_Friedman">Milton Friedman</a> published his book, <em>Free to Choose</em> (1980), arguing that freedom is the key to a prosperous economy. Freedom is not a commodity per se; rather, it is an attribute of the choice process itself, which, according to Sen (1997), should be placed in a central position of choice theory.</p><p>Yet when we examine the textbook model of the “economic man” or <a href="https://en.wikipedia.org/wiki/Homo_economicus"><em>homo economicus</em></a>, we find an image of a mechanical calculator rather than an intelligent being capable of overcoming the limits of rigid logical rules.</p><p>Does freedom contravene the laws of reason?</p><p>According to Gödel’s <a href="https://plato.stanford.edu/entries/goedel-incompleteness/">First Incompleteness Theorem</a>, an axiomatic formal system rich enough to encompass arithmetic can be either complete or consistent, but not both. A consistent system, therefore, will have many true statements that cannot be proved within the system. No consistent formal system can encapsulate the entirety of truth.</p><p>The only way to verify the truth of such a statement is by “going outside” the system. That is, by integrating additional axioms that expand the system’s power. This is the essence of intelligence, and it is the key to free will. Free will gives us the choice to transcend the boundaries of a formal system — not to destroy it, but rather to extend it and expand its domain.</p><blockquote>Incompleteness is a feature, not a bug.</blockquote><p>The limits of reason, therefore, give us the freedom to explore creative means to overcome these limits. These limits provide ample room for free choice. The “Gödelian Space” created by reason’s limits becomes the domain where freedom operates. Reason and free will come into harmony rather than in conflict.</p><p>Market incompleteness explains the fundamental role of free choice and private ownership in market systems. Rather than seeing market incompleteness as a “market failure,” it is the engine for creativity and growth. Incompleteness is a feature, not a bug.</p><h4>Conclusion</h4><p>While economic theory celebrates completeness as the crown jewel of free market efficiency, a complete market, technically speaking, is neither free nor efficient.</p><p>The ultimate source of incompleteness is not frictions or bounded rationality. Rather, it is the inherent limits of pure reason. The most abstract market, where market players are robots equipped with the power of supercomputers, faces insurmountable barriers that often prevent rational decision-making. These limits are what secure the proper space for free choice and, consequently, vibrant, efficient markets.</p><p>The incompleteness of mathematics is one of the greatest discoveries of the 20th century. The implications of such a discovery for social sciences are yet to be fully internalized. We hope that the type of work discussed in this article provides a step towards bringing economics to the science of the 21st century.</p><p><em>This article is part of a series on </em><a href="https://medium.com/godelian-letters/economic-impossibility-theorems-1d3c3c431467"><em>Economic Impossibility Theorems</em></a><em>.</em></p><h4>References</h4><ul><li>Al-Suwailem, S., Doria, F. A., &amp; Kamel, M. (2018). Is risk quantifiable? In S.-H. Chen, Y.-F. Kao, R. Venkatachalam, &amp; Y.-R. Du (Eds.), <em>Complex systems modeling and simulation in economics and finance </em>(pp. 277–303). Springer.</li><li>Al-Suwailem, S., &amp; Doria, F. A. (2020). Can a free market be complete? In E. Bucciarelli, S.-H. Chen, &amp; J. M. Corchado (Eds.), <em>Decision Economics: Complexity of Decisions and Decisions for Complexity</em> (pp. 235–241). Springer.</li><li>Arrow, K. (1969). The organization of economic activity: Issues pertinent to the choice of market versus non-market allocation. In R. Haveman &amp; J. Margolis (Eds.), <em>Public Expenditures and Policy Analysis</em> (pp. 59–73). Markham Publishing.</li><li>da Costa, N. C. A., &amp; Doria, F. A. (1991). Undecidability and incompleteness in classical mechanics. <em>International Journal of Theoretical Physics, 30</em>, 1041–1073.</li><li>da Costa, N. C. A., &amp; Doria, F. A. (2014). On an extension of Rice’s theorem and its applications in mathematical economics. <em>Advances in Austrian Economics, 18</em>, 237–257.</li><li>Davis, M. (2018). <em>The universal computer: The road from Leibniz to Turing</em> (3rd ed.). CRC Press.</li><li>Friedman, M. (1980). <em>Free to choose</em>. Harcourt, Brace, Jovanovich.</li><li>Gribbin, J. (2004). <em>Deep simplicity: Bringing order to chaos and complexity</em>. Random House.</li><li>Hart, O. (1975). On the optimality of equilibrium when market structure is incomplete. <em>Journal of Economic Theory, 11</em>, 418–443.</li><li>Hart, O. (2016). Incomplete contracts and control. Prize lecture. <a href="https://www.nobelprize.org/">www.nobelprize.org</a>.</li><li>Hart, O., &amp; Halonen-Akatwijuka, M. (2024). More is less: Why parties may deliberately write incomplete contracts. <em>Cambridge Elements in Law, Economics, and Politics</em>.</li><li>Laffont, J. J. (1989). A brief overview of the economics of incomplete markets. <em>The Economic Record, 65</em>, 54–65.</li><li>Lucas, R. E. (1988). On the mechanics of economic development. <em>Journal of Monetary Economics, 22</em>(1), 3–42.</li><li>Magill, M., &amp; Quinzii, M. (Eds.). (2008). <em>Incomplete markets</em> (Vol. 1). Edward Elgar.</li><li>Moore, J. (2016). Introductory remarks on Grossman and Hart (1986). In P. Aghion, M. Dewatripont, P. Legros, &amp; L. Zingales (Eds.), <em>The impact of incomplete contracts on economics</em>. Oxford University Press.</li><li>Partnoy, F. (2012). <em>Wait: The art and science of delay</em>. PublicAffairs.</li><li>Reiter, E., &amp; Johnson, C. M. (2013). <em>Limits to computation: An introduction to the undecidable and the intractable</em>. CRC Press.</li><li>Sen, A. (1977). Rational fools: A critique of the behavioral foundations of economic theory. <em>Philosophy &amp; Public Affairs, 6</em>, 317–344.</li><li>Sen, A. (1997). Maximization and the act of choice. <em>Econometrica, 65</em>, 745–779.</li><li>Sen, A. (2017). <em>Collective choice and social welfare</em> (Expanded ed.). Penguin.</li><li>Tirole, J. (1999). Incomplete contracts: Where do we stand? <em>Econometrica, 67</em>, 741–781.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=95134ff8cbad" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/can-a-free-market-be-complete-95134ff8cbad">Can a Free Market Be Complete?</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Gödel, Fukuyama, and the “End of History”]]></title>
            <link>https://medium.com/godelian-letters/g%C3%B6del-fukuyama-and-the-end-of-history-238059a17e0d?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/238059a17e0d</guid>
            <category><![CDATA[incompleteness-theorem]]></category>
            <category><![CDATA[fukuyama]]></category>
            <category><![CDATA[gödel]]></category>
            <category><![CDATA[social-dilemma]]></category>
            <category><![CDATA[end-of-history]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Fri, 16 May 2025 12:38:36 GMT</pubDate>
            <atom:updated>2025-05-17T02:19:53.541Z</atom:updated>
            <content:encoded><![CDATA[<p><em>The greatest threat to human progress is the illusion of completeness and finality.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*szMdkER-1E19wMbsjDbcAg.png" /><figcaption><a href="https://www.ias.edu/scholars/godel">Kurt Gödel</a> (left) and <a href="https://www.britannica.com/biography/Francis-Fukuyama">Francis Fukuyama</a>.</figcaption></figure><h4>The End of History?</h4><p>In 1992, <a href="https://www.britannica.com/biography/Francis-Fukuyama">Francis Fukuyama</a> proclaimed <em>The End of History</em>, arguing that “liberal democracy” as a social system marked the final stage of humanity’s evolution. It is the end state of ultimate rationality and maturity. Although there will always be problems in implementing the system, conceptually, the system is complete.</p><p>Fukuyama writes extensively, and his opinions may evolve in directions that may appear, at least in part, inconsistent with the views presented in his <em>End of History</em> book (see Williams et al., 2016). I will, therefore, focus on the 2006 edition of the book.</p><p>Regardless of Fukuyama’s personal views, the notion that history might have a final, rational endpoint — whether in the form of liberal democracy, communism, or any other social system — continues to shape political thought and debate. Therefore, analyzing this thesis’s logical and philosophical foundations (and limitations) is a necessary task, independent of any specific author or philosopher.</p><p>In this article, I attempt to address Fukuyama’s argument from a modern science and mathematics point of view. This is in line with his approach of building the argument for the “End of History” on modern science.</p><p>In a nutshell, the End of History thesis is inconsistent with modern logic and mathematics and may negatively contribute to the general progress of human civilization.</p><h4>A Universal History</h4><p>Fukuyama (2006) builds his argument on the idea of the Universal History of mankind. Basically, he proposes that human history follows a coherent, progressive pattern toward what he calls liberal democracy. This isn’t random but driven by two fundamental mechanisms:</p><ol><li><strong>Modern natural science</strong> creates cumulative material progress through technological advancement, requiring rational social organization and fostering economic interdependence.</li><li><strong>The struggle for recognition</strong> — Humans’ inherent desire for dignity and acknowledgment (what Hegel called “thymos”) pushes societies toward systems that better satisfy this psychological need.</li></ol><p>Fukuyama argues that the same economic-technological forces Marx identified do indeed drive historical change, but he concludes they lead to capitalism rather than communism as the endpoint. Fukuyama writes:</p><blockquote>“What we have called the “logic of modern natural science” is in effect an economic interpretation of historical change, but one which (unlike its Marxist variant) leads to capitalism rather than socialism as its final result.”</blockquote><p>Since Fukuyama’s argument, in his own words, is a “variant” of Marx’s, but with the opposite conclusion, the two share critical vulnerabilities. They both:</p><ul><li>Propose a deterministic view of history with an inevitable endpoint, and</li><li>Claim to have identified the singular “correct” destination of human political evolution.</li></ul><p>We can trace the flaw in the arguments of both Marx and Fukuyama to systematically prioritizing one side over the other:</p><ul><li>Individual interest</li><li>Collective interest</li></ul><p>Marx believed that collective interest would dominate, while Fukuyama believed individual interest would dominate (see his book, <em>Liberalism and Its Discontents</em>). This shows why both thinkers’ deterministic frameworks fall short.</p><h4>The Impossibility of a Paretian Liberal</h4><p>In 1970, <a href="https://www.nobelprize.org/prizes/economic-sciences/1998/sen/facts/">Amartya Sen</a>, who won the Nobel Prize in 1998, published a paper with the above title. The 6-page paper is <a href="https://mappingignorance.org/2020/01/08/fifty-years-of-the-paretian-liberal-paradox/">considered</a> “one of the most important papers in the crossing of economics, philosophy, and political science.” According to Sen (2021), it has probably been read more than any of his other works.</p><p>Basically, Sen shows that individual freedom and <a href="https://www.britannica.com/money/Pareto-optimality">Pareto optimality </a>are logically incompatible. An allocation is Pareto optimal if it is impossible to change the allocation to make some members better off without making at least one member worse off. It is, therefore, a criterion for collective rationality. The theorem of Sen (1970) shows that the tension between individual rationality (free choice) and collective rationality in general is unsolvable.</p><p>Sen (2017) shows how the tension between the two dimensions can be modeled as a Prisoner’s Dilemma game. This game is well known as the classical example of the conflict between individual and group rationality. It is a benchmark for <a href="https://en.wikipedia.org/wiki/Collective_action_problem">Social Dilemmas</a> in general, which permeate all aspects of social interactions.</p><blockquote>Sen showed the logical incompatibility of individual liberty and collective optimality.</blockquote><p>The impossibility of a Paretian Liberal implies that no algorithm can systematically resolve the tension between the individual and the collective. Marx and Fukuyama each tend to prioritize one over the other, when in fact the problem is logically undecidable (see below). Both made the same category error by assuming their preferred value (collectivism for Marx, individualism for Fukuyama) could be universalized as history’s endpoint. They failed to recognize that prioritizing either necessarily sacrifices aspects of the other in ways that create inherent contradictions.</p><h4>Inescapable Dilemmas</h4><p>To illustrate the logical conundrum at the heart of social dilemmas, we outline below an intuitive argument inspired by the well-known <a href="https://plato.stanford.edu/entries/russell-paradox/">Russell’s Paradox</a>.</p><p>Suppose we have a universal rule <em>R</em> that systematically determines whether any (non-trivial) social problem <em>x</em> should be resolved collectively or individually. That is, for each non-trivial social problem <em>x</em>, the rule <em>R</em>(<em>x</em>) tells us whether to prioritize collective interest or individual interest. Hence, we can classify all social problems into two mutually exclusive sets:</p><ul><li><em>A</em> = {<em>x </em>: <em>R</em>(<em>x</em>) = 1} is the set of all social problems decided collectively.</li><li><em>B</em> = {<em>x</em> : <em>R</em>(<em>x</em>) = 0} is the set of all social problems decided individually.</li></ul><p>Now comes the question: How do we decide the rule <em>R</em> itself? Since <em>R</em> is a non-trivial social problem (a rule about how to decide social problems), it <em>must</em> fall into either set <em>A</em> or set <em>B</em>. (The two sets are exhaustive and mutually exclusive.) We will have one of two cases.</p><ul><li><strong>Case 1</strong>: <em>R </em>is decided collectively, so that <em>R</em>(<em>R</em>) = 1. This means that <em>R </em>∈ <em>A</em>. But this implies that <em>B</em> will also be in <em>A</em>. The reason is that <em>B</em> is constructed via <em>R</em>, as the definition of <em>B</em> shows. If <em>R</em> is decided collectively, then the very definition of <em>B</em> (problems resolved individually) becomes a collective choice, placing <em>B</em> within <em>A</em>. But this contradicts the assumption that the two sets are mutually exclusive.</li><li><strong>Case 2</strong>: <em>R </em>is decided individually, so that <em>R</em>(<em>R</em>) = 0, and hence <em>R </em>∈ <em>B</em>. This implies that <em>A </em>∈ <em>B</em>, which contradicts the definition of <em>A</em>.</li></ul><p>The conclusion is that a universal rule <em>R</em> is logically impossible. If <em>R</em> is universal, it must classify itself as a social problem. If it does, a contradiction inevitably emerges. Hence, <em>R </em>cannot exist.</p><p>Social dilemmas are not just practically difficult — they are inherently irresolvable. No single rule can consistently and completely resolve the tension between individual and collective rationality. The tension is not resolvable from within a single, universal schema. This aligns with the results of Amartya Sen indicated above.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QUgE9SwlnDTvlRdwI-rx4A.png" /><figcaption>The impossibility of a universal rule <em>R. Image generated by </em><a href="http://mermaidchart.com/"><em>Mermaidchart.com</em></a></figcaption></figure><p>Intuitively, the society must agree on the distribution of social problems across the sets <em>A </em>and <em>B</em>. This implies that <em>A </em>and <em>B</em> are decided collectively. This is not a problem with respect to <em>A. </em>But it creates a problem for set <em>B</em> since each of its members is supposed to be decided individually. It is like saying: You have the freedom to choose the meal you prefer, but the group is the one who decides the menu. The group can then limit the menu in a manner that defeats the individual’s freedom to choose.</p><p>It is important to note that the impossibility result concerns a universal rule <em>R </em>that can produce a complete <em>and</em> consistent classification of social problems. Local rules that can consistently decide <em>some</em> problems (and hence their domain is incomplete) are logically feasible. This aligns well with Gödel’s theorems, as we shall see.</p><h4>Modeling Universal History</h4><p>Fukuyama builds his argument on modern science, among other things. Modern science, in turn, is formulated in mathematics. The book of nature, <a href="https://www.jstor.org/stable/10.5749/j.ctt1d390rg.4?seq=1">Galileo famously noted</a>, “is written in the language of mathematics.” Mathematical models have been used extensively, not only for analyzing natural phenomena but for social and economic ones as well.</p><p>Fukuyama’s concept of Universal History aims to “find a meaningful pattern in the overall development of human societies generally.” It is therefore an attempt to construct a systematic “model” to understand and predict human history. Building a mathematical model of history seems to be the ultimate attempt to implement the concept of Universal History.</p><p>However, mathematics has its own rules, the most prominent of which are Gödel’s Incompleteness Theorems.</p><h4>Gödel’s Incompleteness Theorems</h4><p><a href="https://medium.com/godelian-letters/remembering-g%C3%B6del-398fac905010">Kurt Gödel</a> was one of the greatest mathematicians in modern history. He was exceptionally open-minded and had keen interests in various fields, including philosophy, physics, economics, and history, among others; see Dawson (1997), Wang (1998), and Budiansky (2021).</p><p>Gödel shocked the world of mathematics when he published his Incompleteness Theorems in 1931. Polymath <a href="https://www.britannica.com/biography/John-von-Neumann">John von Neumann</a> (1951) remarked:</p><blockquote>“Kurt Gödel’s achievement in modern logic is singular and monumental — indeed, it is more than a monument; it is a landmark which will remain visible far in space and time.”</blockquote><p>Von Neumann emphasized that Gödel’s results are not specific to the field of mathematics but are general to <em>any</em> field that applies formal methods:</p><blockquote>“Gödel actually proved this theorem, not with respect to mathematics only, but for all systems which permit a formalization, that is a rigorous and exhaustive description in terms of modern logic: For no such system can its freedom from inner contradiction be demonstrated with the means of the system itself.”</blockquote><p>With this brief background, let us take a look at Gödel’s theorems (Stillwell, 2016). Roughly speaking:</p><ul><li>Any consistent formal system capable of expressing basic arithmetic will contain true statements that are unprovable <em>within the system</em>. Truth extends beyond proof. This is the essence of Gödel’s First Incompleteness Theorem.</li><li>As long as such a system is consistent, it cannot prove its own consistency. This is the essence of the Second Incompleteness Theorem.</li></ul><p>Gödel knew well that his Theorems have far-reaching philosophical implications. In response to a proposal to generalize his theorems to human affairs, Gödel wrote in a 1961 letter draft, as recorded by his pupil Hao Wang (1998):</p><blockquote>“A completely unfree society (i.e., one proceeding in everything by strict rules of “conformity”) will, in its behavior, be either inconsistent or incomplete, i.e., unable to solve certain problems, perhaps of vital importance. Both of course, may jeopardize its survival in a difficult situation. A similar remark would also apply to individual human beings.”</blockquote><p>This shows that a deterministic theory of history — that is, one which attempts to prescribe the course of historical development according to strict, formal rules — must face the limits imposed by Gödel’s theorems. Let us see how this affects an ideal model of Universal History.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/567/1*qzqepBw6LVOvBOF9299BvA.png" /><figcaption>Source: GPT-Image-1.</figcaption></figure><h4>Cracks in the Crystal Ball</h4><p>Suppose we construct a sophisticated mathematical model of Universal History. The model successfully captures major historical trends. Statements or propositions in the model correspond to events in history.</p><p>This is a highly imaginative assumption, but it is possible nonetheless. Should such a sophisticated model ever exist, Gödel’s incompleteness theorems would apply to it just as rigorously as they do in mathematics. Rather than asking, “Can we model Universal History?” we ask: “Can the most ambitious model of Universal History ever be complete and final?”</p><p>According to Gödel’s theorems, there will be many propositions formulated in the language of the model that cannot be derived or refuted within the model. Since these propositions correspond to historical events, our model will fail to predict many of them. Truth, in principle, can be explored but can never be captured by a logical system, as William Byers (2008) elaborates.</p><blockquote>Truth can be explored, but cannot be captured.</blockquote><p>This shows that, even in the most optimistic scenario for the formalization and prediction of history, there will always be truths about the historical process that are undecidable within the system. Thus, the notion of a final, complete, and error-proof understanding of history is not just practically unreachable, but formally impossible.</p><p>This is not a defect in the model. No matter how powerful or sophisticated our model is, unpredictable events will always emerge, and they will not be rare. “Mathematics is inexhaustible,” asserted Gödel (Wang, 1998).</p><blockquote>History, like mathematics, is an open book — its final chapter can never be written from within its own pages.</blockquote><p>This means that unpredictability in history isn’t merely a result of practical limitations (insufficient data, computational power, or complexity) but reflects a deeper logical incompleteness inherent to any rich formal system capable of modeling historical processes.</p><p>Fukuyama’s error, then, is more profound than simply choosing individualism over collectivism. By assuming history has a determinate endpoint derivable from first principles, he’s implicitly treating historical processes as if they could be encapsulated within a complete formal system — something Gödel proved impossible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*Er3525aUUpx5Wv89-l293w.png" /><figcaption><a href="https://www.independent.co.uk/independentpremium/karl-popper-philosophy-karl-marx-b1999504.html">Karl Popper</a> (left) and <a href="https://fs.blog/book-club-fall-2021/">Amartya Sen</a></figcaption></figure><h4>An Open Universe</h4><p>The vision advocated above echoes the argument for indeterminism that <a href="https://plato.stanford.edu/entries/popper/">Karl Popper</a> defended in <em>The Open Universe </em>(Popper, 1982).</p><p>Popper develops an extensive argument for “The Impossibility of Self-Prediction.” Building on the work of Gödel and Alan Turing, Popper contends that, even with perfect information on present or past initial conditions, it is impossible to predict future states using deductive methods <em>from within</em>. Specifically, no calculator or predictor, machine or human, can deductively predict the results of its own calculations or predictions (see also Ismael, 2016).</p><p>The argument ultimately hinges on self-reference, which creates a “blind spot” whereby the prediction itself can influence the outcome, even for inanimate or physical systems.</p><p>Popper uses this argument to refute “historicism”: the view that human history is systematically predictable, which underpins the Universal History frameworks of Hegel, Marx, and Fukuyama. If complete self-prediction is impossible, whatever the complexity of the predictor, then this must also hold for any society of interacting predictors; consequently, no society of interacting predictors can predict its own future states of knowledge.</p><p>The course of human history, Popper argues, is strongly influenced by the growth of human knowledge. But we cannot predict, by rational or scientific methods, the future growth of our scientific knowledge. It follows, therefore, that we cannot predict the future course of human history (see also Popper, 1957). Where Fukuyama sees history culminating in liberal democracy, Popper sees an open process where new knowledge constantly rewrites the rules.</p><p>Amartya Sen’s undecidable tensions between individual and collective rationality mirror Popper’s indeterminism: both reveal systems that cannot resolve their own internal contradictions. This aligns with the logical incompleteness demonstrated by Gödel’s theorems.</p><h4>Fukuyama’s “End” as a Gödelian Trap</h4><p>Gödel’s Second Incompleteness Theorem states that no consistent formal system can prove its own consistency. Since we are part of the history we are trying to model, we lack the power to prove or even to assert the consistency of our model.</p><p>When we apply this mathematical insight to the modeling of Universal History, the implication becomes clear: absolute confidence in our framework is mathematically unjustifiable. The truly rational position must acknowledge the limits of rationality itself.</p><p>Those who proclaim perfect rationality unknowingly violate this mathematical boundary. By attempting to stand outside their own reasoning system to verify it, they commit a logical error that Gödel proved unavoidable. Paradoxically, absolute faith in rationality becomes an irrational position.</p><p>Fukuyama’s thesis represents precisely this kind of overconfidence. By declaring liberal democracy as history’s ultimate endpoint, it exemplifies the belief that humanity has achieved a consistent, rational political system that cannot be improved upon. This mirrors the mathematical error of a system attempting to prove its own consistency.</p><blockquote>True rationality invites epistemic humility.</blockquote><p>This approach embodies epistemic modesty, creating space for continuous learning, creativity, and improvement, rather than rigidly defending supposedly perfect systems.</p><p>Sustainable civilizations maintain this Gödelian humility. They build adaptable institutions that acknowledge the inherent incompleteness of their own rational frameworks. They understand that the claim “we have reached the end of history” is logically untenable.</p><h4>Limitless Growth and the Poverty of Diversity</h4><p>Fukuyama repeatedly points to the “unlimited accumulation of wealth” that was made possible because of the marriage of capitalism with modern science. Moreover, a civil society presents “a liberation of human acquisitiveness: man can accumulate not just what he needs, but whatever he wants without limit.”</p><p>This shows a major defect in this thesis: There are no inherent limits to knowledge, technology, or wealth. Fukuyama failed to study the limits of modern science established by science itself (e.g., Barrow, 1998). While claiming to ground his thesis in science, Fukuyama selectively embraces scientific enablement of growth while ignoring scientific discoveries about its limits. The very science he invokes as supporting liberal capitalism has actually established boundaries that his “End of History” thesis cannot coherently address.</p><p>By positioning “unlimited acquisitiveness” as liberation, Fukuyama fails to recognize that such unbounded accumulation contains within it the seeds of its undermining, as resource constraints, environmental degradation, and social inequality intensify.</p><p>By denying the inherent limits of knowledge and technology, Fukuyama reaches a logical corollary: Societies become increasingly homogenous. A “universal consumer culture” follows from a “universal evolution in the direction of capitalism.” Denying the diversity of cultures and economic systems is the precondition for imposing one model on all societies. This is the exact antithesis of “liberal democracy” that Fukuyama calls for.</p><p>This reveals the deeply problematic nature of Fukuyama’s argument: in the name of liberalism, he advances a deterministic framework that denies societies the very freedom of self-determination that liberalism supposedly champions. His vision of inevitable homogenization ultimately becomes indistinguishable from the totalizing ideologies he claims to transcend.</p><p>This shows why the theorems of Gödel and other great mathematicians (e.g., Chaitin, 1997, 2007) are essential for embracing diversity, creativity, and tolerance. Without epistemic humility, we risk falling into the totalitarian traps that we are seeking to avoid.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/350/1*x6G0r1r0KtB0Rsih36E62A.jpeg" /><figcaption>Kurt Gödel (right) with his brother Rudolf around 1908. Source: <a href="https://plus.maths.org/content/godel-and-limits-logic">John Dawson</a>.</figcaption></figure><h4><strong>Conclusion</strong></h4><p>The greatest threat to human progress is the illusion of completeness and finality. When we believe we’ve reached rationality’s endpoint — whether in governance, economics, or social systems in general— we violate Gödel’s insights and become vulnerable to catastrophic inconsistencies.</p><p>These insights offer a profound guide for sustainable civilization — one that remains open to progress, acknowledges its limitations, and never declares itself history’s final word. History, like mathematics, is an open book — its final chapter can never be written from within its own pages.</p><p>Within a Gödelian framework, the “struggle for recognition” that Fukuyama emphasized transforms into an unending quest for knowledge and discovery. Normal humans have an <a href="https://medium.com/godelian-letters/a-playful-mind-4b6641cd82f4">intrinsic preference</a> for adventure, learning, and exploring new possibilities. They enjoy creativity, discovery, and entrepreneurial activities.</p><p>In his family, the young Gödel was a highly inquisitive child, so much so that he was <a href="https://plus.maths.org/content/godel-and-limits-logic">nicknamed</a> <em>Herr Warum</em> (“Mr. Why”). His insatiable curiosity foreshadowed his groundbreaking Incompleteness Theorems, which demonstrate that truth is not a destination but a journey into the infinite. The end of history? Only if we stop questioning.</p><p><em>I am grateful to </em><a href="https://godelianletters.org/Interview_William_Byers"><em>William P. Byers</em></a><em> and </em><a href="https://en.wikipedia.org/wiki/John_Stillwell"><em>John C. Stilwell</em></a><em> for their helpful comments. All errors and omissions are my own.</em></p><h4><strong>References</strong></h4><ul><li>Barrow, J. D. (1998). <em>Impossibility: The limits of science and the science of limits</em>. Oxford University Press.</li><li>Budiansky, S. (2021). <em>Journey to the edge of reason. </em>W.W. Norton.</li><li>Byers, W. (2008). <em>The blind spot: Science and the crisis of uncertainty</em>. Princeton University Press.</li><li>Chaitin, G. J. (1997). <em>The limits of mathematics</em>. Springer-Verlag.</li><li>Chaitin, G. (2007). <em>Thinking about Gödel and Turing: Essays on complexity, 1970–2007</em>. World Scientific Publishing.</li><li>Dawson, J. W. (1997). <em>Logical dilemmas: The life and work of Kurt Gödel</em>. A. K. Peters.</li><li>Fukuyama, F. (2006). <em>The end of history and the last man</em>. Simon &amp; Schuster. (Original work published 1992.)</li><li>Fukuyama, F. (2022). <em>Liberalism and its discontents</em>. Farrar, Straus and Giroux.</li><li>Ismael, J. (2016). <em>How physics makes us free</em>. Oxford University Press.</li><li>Popper, K. (1957). <em>The poverty of historicism</em>. Routledge.</li><li>Popper, K. (1982). <em>The open universe: An argument for indeterminism</em>. Routledge.</li><li>Sen, A. (1970). The impossibility of a Paretian liberal. <em>Journal of Political Economy, 78</em>(1), 152–157.</li><li>Sen, A. (2017). <em>Collective choice and social welfare</em> (Expanded ed.). Penguin Books.</li><li>Sen, A. (2021). <em>Home in the world: A memoir</em>. Liveright Publishing.</li><li>Stillwell, J. (2016). <em>Elements of mathematics: From Euclid to Gödel</em>. Princeton University Press.</li><li>von Neumann, J. (1951). Tribute to Dr. Gödel. In J. Bulloff, T. Holyoke, &amp; S. Hahn (Eds.), <em>Foundations of mathematics: Symposium papers commemorating the sixtieth birthday of Kurt Gödel</em> (pp. ix–x). Springer-Verlag.</li><li>Wang, H. (1996). <em>A Logical journey: From Gödel to philosophy.</em> MIT Press.</li><li>Williams, H., Sullivan, D., &amp; Matthews, E. G. (2016). <em>Francis Fukuyama and the end of history</em>. University of Wales Press.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=238059a17e0d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/g%C3%B6del-fukuyama-and-the-end-of-history-238059a17e0d">Gödel, Fukuyama, and the “End of History”</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Interview with Ian Stewart]]></title>
            <link>https://medium.com/godelian-letters/interview-with-ian-stewart-c81386847856?source=rss----e48cbbf77efb---4</link>
            <guid isPermaLink="false">https://medium.com/p/c81386847856</guid>
            <category><![CDATA[symmetry-breaking]]></category>
            <category><![CDATA[symmetry]]></category>
            <category><![CDATA[science]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Sami Al-Suwailem]]></dc:creator>
            <pubDate>Sat, 19 Apr 2025 11:56:13 GMT</pubDate>
            <atom:updated>2025-04-19T11:56:13.526Z</atom:updated>
            <content:encoded><![CDATA[<p>“<em>The relationship between mathematics and the real world is mysterious and wonderful</em>.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*b79Kdb1n4Oh0QO1r32JPxA.png" /><figcaption>Source: <a href="https://www.sciencesetavenir.fr/fondamental/mathematiques/entretien-ian-stewart-l-incertitude-et-les-mathematiques_146149">Sciences et Avenir</a></figcaption></figure><p>We are honored to interview Professor Ian Stewart, an exceptional mathematician, scientist, and novelist.</p><p>Stewart is best known for engaging the public with mathematics and science through his many bestselling books, newspaper and magazine articles, and radio and television appearances. His accessible and entertaining style of writing has opened up a range of hard-to-fathom topics — including chaos theory and symmetry — to general audiences.</p><p>As an Emeritus Professor at the <a href="http://www2.warwick.ac.uk/">University of Warwick</a>, Ian splits his time between mathematics research and work to increase awareness of mathematics and science. His research themes cover bifurcation theory, pattern formation, and biomathematics, and he has made important contributions to catastrophe theory.</p><p>He has won several prizes for his efforts to further the public’s understanding of science, including the Royal Society’s <a href="https://royalsociety.org/awards/michael-faraday-prize/">Michael Faraday Prize</a> in 1995, the <a href="http://www.aaas.org/page/2001-award-recipients#Public">2001 AAAS Public Understanding of Science and Technology Award</a>, the London Mathematical Society’s Zeeman Medal, and Rockefeller University’s <a href="http://newswire.rockefeller.edu/2015/03/13/lewis-thomas-prize-to-honor-mathematicians-steven-strogatz-and-ian-stewart/">Lewis Thomas Prize</a> for writing about science.</p><p>He has published over 200 research papers and written or co-written more than 140 books, including many popular books on mathematics, several science fiction novels, and the <a href="https://en.wikipedia.org/wiki/The_Science_of_Discworld"><em>Science of Discworld</em></a><em> </em>series with Jack Cohen and Terry Pratchett.</p><p>You can learn more about our guest from his personal page: <a href="https://ianstewartjoat.weebly.com/">ianstewartjoat.weebly.com</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*Tvm7KXl2jOdD9-ca.jpg" /><figcaption><a href="https://manchesterhistory.net/architecture/2000/zeeman.html">Zeeman Building</a> is hosting the Mathematics Institute at the University of Warwick, UK.</figcaption></figure><p><strong>GL: </strong><em>You belong to a rare breed of mathematicians who blend rigorous logical capabilities with unusual skills in art and literature. How can our educational system better nurture this integration of “left-brain” and “right-brain” — so to speak — to support creative and healthy intellectual development?</em></p><p><strong>IS: </strong>Let me introspect a little first, considering what influenced my own intellectual development (if that’s not too posh a description) and career trajectory (ditto).</p><p>I’m a bit of a butterfly-brain: I flit from one thing to another. I’ve always had broad interests, but academically, it was science that most appealed to me, very strongly. At age six or so, I was fascinated by astronomy. I lapped up encyclopaedias whenever I could lay my hands on them. We lived near the sea, and I was interested in fish; along the coast, you could find fossils, and I collected those. Also butterflies, beetles, pressed wild flowers … if you could collect it, I did.</p><p>My father (who worked in a bank) had a practical bent, especially woodwork, and he was a competent artist, especially seascapes. He taught me to put up wallpaper. He found an old typewriter in the bank vault and was allowed to take it; he gave it to me and my friends, and we spent many hours writing things and typing them up. Bad humorous epic poetry, rules for invented card games, gazetteers of imaginary islands …</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c7TU7u0ZaSCBqSzjv56J9Q.png" /><figcaption>Left: Arthur (Nick) Stewart carrying baby Ian. Right: Ian at primary school (1951). Source: Ian Stewart.</figcaption></figure><p>I got interested in science fiction, and many years later, I started to write it. When I was 16 or so, I and three friends formed a guitar group, and at university I was lead guitar in another group for two years, which doubled my grant. So I’ve always been immersed in a rich variety of activities and interests, and open to new ones. (In later life, my wife and I took courses in Egyptology and geology, for example, and still engage in those activities.)</p><p>The key to this kind of mentality, I think — assuming that’s what we want to encourage, which is perhaps another matter — is being exposed to many different things from an early age. Also, crucially, learning to enjoy new challenges. You need to be self-critical, but not <em>too</em> self-critical. You don’t have to be the world number one; just reasonably competent. (Of course, you can aim to be world number one in some activity; there’s nothing wrong with ambition. But not at everything. If you’ve never failed at anything, you’ve not tried enough things.)</p><blockquote>“You need to be self-critical, but not <em>too</em> self-critical.”</blockquote><p>I’ll focus on mathematics but also on its relation to other areas. In many countries nowadays, the main emphasis when teaching mathematics is a rigid insistence on following specific rules for calculations. Not only must the answer be right: the method used to get it must follow the rule book exactly. This kills off creativity; in mathematics, there are usually many different ways to get the right answer. It’s not unusual for correct solutions to be graded down because the method isn’t the one in the book. Also, mathematics is often taught without any relation to any other areas of human activity.</p><p>There can be several reasons for this. A too prescriptive syllabus prevents teachers from introducing anything not in the rule book. Lack of confidence among some mathematics teachers causes them to fall back on the rule book, so — to be blunt — we end up with a student who doesn’t understand what they’re doing being taught by a teacher who doesn’t understand what they’re doing.</p><p>Basic mathematical skills are important, of course, but not to the exclusion of connections with other areas of human activity. Rigid segregation of subject areas adds to the problem. A more relaxed approach, allowing time to explore cultural connections and historical development of mathematics, in addition to routine problem solving, might improve matters. Something as simple as marking out a sports field involves quite a bit of mathematics, for instance. Doing mathematics need not be restricted to the mathematics classes.</p><blockquote>“There are mathematical patterns everywhere, but they’re not obvious until they’re pointed out.”</blockquote><p>There are mathematical patterns everywhere, but they’re not obvious until they’re pointed out. Next time you watch a dog or a cat or a horse moving, keep one eye on which legs move together and which don’t. Observe the pattern of footfalls. There are regular rhythms, patterns of motion called ‘gaits’. Look this up on the internet: there’s been a huge amount of gait science and mathematics, going back to Aristotle. When I watch TV documentaries about the natural world, I’m aware of an extra dimension that most viewers miss. Watch the legs!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5XNZGq7RwM3-Hg5uMGiqgQ.jpeg" /><figcaption>Watch the legs! Source: <a href="http://shutterstock.com/">Shutterstock</a>.</figcaption></figure><p><strong>GL: </strong><em>You have contributed extensively to many areas of mathematics and science, from pure mathematics to popular science writing. In an age of increasing specialization, how do you view the role of interdisciplinary studies?</em></p><p><strong>IS: </strong>Very important.</p><p>Interaction with other disciplines is the lifeblood of mathematics (and much else). Problems from other areas can inspire new mathematical ideas. Existing results and methods can be applied to new problems in the outside world.</p><p>My feeling is that over the past few decades, there’s been an increased appreciation of interdisciplinary activities. This has happened alongside specialization — which might or might not be increasing. In mathematics, there can be sudden unifications, in which previously separate specialisms turn out to be different aspects of the same thing.</p><p>Admittedly, in the past, a single (world-class) mathematician like Euler or Gauss could span almost every discipline in the subject, because a lot less mathematics was known then. But even today, some people manage to do impressive research in several disparate areas. And you get the impression that if they needed to know a new area, they’d have grasped it within a week or two.</p><p>It’s possible to combine specialization with interdisciplinary research. The trick — which is pretty much standard now — is teamwork. A team of specialists, talking to each other with a common goal, can achieve things that no single individual could.</p><blockquote>“A team of specialists, talking to each other with a common goal, can achieve things that no single individual could.”</blockquote><p>I’ve been involved in various interdisciplinary activities. The mathematics department at Warwick University spent decades trying to get support for interdisciplinary mathematics. Initially, these efforts failed — mainly because funding bodies tended to take decisions based on traditional subject areas. So an initiative in mathematical biology, say, would either go to a mathematics committee that didn’t understand the biology, or a biology committee that didn’t understand the mathematics.</p><p>With money always being in short supply compared to demand, the interdisciplinary proposal got rejected. Eventually, this changed, partly by setting up explicit funding for interdisciplinary research and partly by finding money to run small pilot projects. <a href="https://warwick.ac.uk/fac/sci/maths/research/miraw/">The MIR@W project</a> (Mathematical Interdisciplinary Research at Warwick) was one of the first such ventures.</p><p>At one point, I spent five years as part of a team developing new quality control methods for spring wire. This was totally out of my comfort zone, which was mainly a ‘pure mathematics’ attitude, although I was always interested in links to other areas. The project was based at SRAMA (Spring Research and Manufacturers’ Association, now the <a href="https://ist.org.uk/about/">Institute of Spring Technology</a>, in Sheffield).</p><p>Springs are made by coiling wire, but the industry had a serious problem. Namely, standard quality control measures failed to predict whether a given sample of wire could be made to coil into acceptable springs in a coiling machine. One of the engineers at SRAMA read one of my popular mathematics books and noticed a data-processing technique used to find chaotic attractors. He tried it out on data from spring wire, and it seemed to correlate with coilability.</p><p>He phoned me, and a year later, he secured a grant from the Department of Trade and Industry. He then set up a team of about 20 people: mathematicians, engineers, representatives of spring and wire-making companies, and an official from the DTI. We developed a rapport over many meetings, and the outcome was a quality control machine for spring wire, a method for controlling a coiling machine in real time to improve the regularity of the springs it was making, and a quality control test for strip metal.</p><p><strong>GL: </strong><em>In your book </em><a href="https://www.amazon.com/Why-Beauty-Truth-History-Symmetry/dp/0465082378"><em>Why Beauty Is Truth</em></a><em>, you explore symmetry as a unifying concept in mathematics and physics. Yet, as you’ve noted elsewhere, symmetry-breaking often leads to discovery and complexity. How do you think this dynamic tension between symmetry and asymmetry shapes the development of science and technology?</em></p><p><strong>IS: </strong>Symmetry is a very deep and beautiful concept that combines some extraordinary mathematics with aesthetics, science, and engineering.</p><p>Symmetries abound in the natural world, and underpin both relativity and quantum mechanics, the two most successful major physical theories. Yet the main mathematical formulation of symmetry, that of a ‘group’ of transformations, initially emerged from the pure mathematics of algebra with the ideas of <a href="https://fr.wikipedia.org/wiki/%C3%89variste_Galois">Évariste Galois</a>, killed in a futile duel just before his 21st birthday.</p><p>Essentially, a symmetry is a mathematical transformation of some object that leaves it looking exactly the same. ‘Change without change’, so to speak. Rotate a sphere through any angle about any axis, and it still looks like the same sphere, even though all of its points (except the two poles) have moved. Later, this formalism was linked to geometry, and the subject exploded in innumerable directions.</p><p>The phenomenon of symmetry-breaking has proved fundamental to the formation of patterns in nature, and also underlies frontier physics, such as the famous <a href="https://en.wikipedia.org/wiki/Higgs_boson">Higgs boson</a>, a fundamental particle that gives all other particles mass. Symmetry-breaking occurs when a system obeys rules that have some group of symmetries, but the states of the system — what it actually does — are less symmetric than the system itself.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BUBL2Qo3R0l9A4fymGwtPQ.png" /><figcaption><em>An illustration of a result detecting the Higgs Boson at the </em><a href="https://home.cern/science/accelerators/large-hadron-collider"><em>Large Hadron Collider</em></a><em> gathered on May 27, 2012. Source: </em><a href="https://www.britannica.com/science/Higgs-boson"><em>Britannica</em></a><em>.</em></figcaption></figure><p>Going back to the sphere: if a spherical shell — of metal or plastic, say — is compressed by equal forces at all points, all directed towards the centre, this system of forces has spherical symmetry. But if the forces become large enough, the shell buckles and loses its spherical shape. So the spherical symmetry is broken. In a mathematical model, the first stage of buckling creates circular ripples, symmetric about some axis. So the state of the system still has some symmetry: rotations about a single axis, rather than about any axis.</p><blockquote>“Understanding symmetries and how they break has had a profound effect on the development of science and technology by providing deep insights into pattern formation.”</blockquote><p>Understanding symmetries and how they break has had a profound effect on the development of science and technology by providing deep and general insights into pattern formation. Areas concerned include fluid dynamics, elastic buckling, biological development, liquid crystals, animal and human vision, and chemical engineering.</p><p><strong>GL: </strong><em>Some have argued that modern science increasingly places the “observer” at the center of theory — from quantum mechanics to relativity and even Gödel’s incompleteness theorem. Do you believe the role of the observer fundamentally changes how we should think about mathematical and scientific truth?</em></p><p><strong>IS: </strong>Historically, developments of relativity and quantum theory certainly changed scientists’ views of the universe in fundamental ways. Newton’s objective universal space and time, with simple mathematical models, turned out to be <em>too</em> simple. Interactions between observer and observed turned out to be far more important than had been assumed.</p><blockquote>“The role of the ‘conscious observer’ has been overstated.”</blockquote><p>That said, I think that the role of the ‘observer’ — especially a human one, a conscious one — has been overstated. I’ve seen sensible physicists claiming that if no one is looking at an apple on a table, it’s no longer there. It comes back next time someone looks at where it ought to be. I’ve seen people claiming that the reason why humans were created/evolved is to bring the entire universe into existence by observing it.</p><p>I think these ideas are nonsense. Not because I lack the imagination to throw away my naive view of the world, as presented to me by my senses. As a sometime science fiction writer, I can imagine far wilder things than those. No: because they make no sense. How does the apple ‘know’ somebody is looking at it? What if the observer is blind? And as for <em>us</em> being the reason there’s a universe — well, that’s exactly back to front.</p><p>The real issue as regards quantum mechanics underpinning what we perceive as a ‘classical’ world isn’t observers, but <em>measurements</em>. The same apparent paradoxes arise without human consciousness becoming involved. Conventional quantum theory represents a measurement as a ‘projection’ of the wave function onto an eigenfunction of an operator representing the system. This is like measuring the sound made by a clarinet by selecting just one of the sine functions in its Fourier expansion. It’s a simple, clean-cut mathematical operation — but it bears no relation to the messy reality, in which the measuring instrument is itself a gargantuan system of quantum particles, far more complex than the event it’s measuring. And impossible for us to understand on a quantum level, to boot.</p><blockquote>“The real issue as regards quantum mechanics isn’t observers, but <em>measurements.”</em></blockquote><p>Models of quantum systems have two ingredients. One is the Schrödinger equation for the wavefunction, which is always given huge emphasis. The other, generally neglected, is the ‘boundary condition’ imposed on the equation, which is a crisp and simple extra constraint. We are told repeatedly that the quantum world is fuzzy and probabilistic; yet the measurement process is represented as being crisp and tidy, with a precise yes/no outcome. This mismatch is, I suspect, responsible for most of the paradoxes and puzzles of quantum mechanics. The ‘just calculate’ attitude has taken root because (a) it seems to work, and (b) people don’t want to admit that no one really understands how a real-world measurement works on a quantum level.</p><p>For example, one common part of a measurement device is a half-silvered mirror (or prism). In model equations, this device either lets a photon pass through unaffected by the encounter or bends its path through a right angle and imposes a phase shift on its wavefunction. The predictions of such a model correspond very closely to what is observed. But the real-world interaction of a photon with a silvered surface is enormously complex, involving scattering off innumerable atoms. It’s hard to see how the outcome of all this can just be either the original wavefunction or a phase-shifted version, with equal probabilities.</p><p>At any rate, my feeling (as an amateur in this field) is that we need to pay much more attention to the modelling assumptions made when setting up boundary conditions and considering interactions with an instrument, and focus less on the elegance of the equations and crisp properties of their solutions. Yes, the existing approach ‘seems to work’. But perhaps we might ask: <em>why</em>?</p><p><strong>GL: </strong><a href="https://en.wikipedia.org/wiki/Eugene_Wigner"><em>Eugene Wigner</em></a><em> once wrote: “The miracle of the appropriateness of the language of mathematics for the formulation of the laws of physics is a wonderful gift which we neither understand nor deserve.” How do you see the relationship between mathematics and the real world?</em></p><p><strong>IS: </strong>Mysterious and wonderful.</p><p>Wigner understood this, talking about the ‘unreasonable effectiveness’ of mathematics in representing aspects of the universe. People often try to explain the effectiveness — but the real mystery is the unreasonableness.</p><p>Explanations often follow the line ‘mathematics is derived from nature, so it’s not surprising that it sheds light on nature.’ This is fair enough when the original motivation is directly related to the later conclusions, so that, for instance, the motion of the planets leads to calculus, which helps to explain the motion of the planets.</p><p>But this becomes unreasonable, in Wigner’s sense, when the motion of the planets leads to calculus which helps explain heart disease, fish populations in the Adriatic, the notes played by a violin, the spread of COVID-19, or the origins of the universe. All of which are things that calculus does, in fact, help to explain.</p><blockquote>“Mathematics becomes ‘unreasonable’ when the motion of the planets leads to calculus which helps explain heart disease, fish populations, or the origins of the universe.”</blockquote><p>Another answer to this conundrum is to argue that ‘the universe is made of mathematics’. I don’t buy it. First, that just replaces one question with an equally difficult one: <em>why </em>is it made of mathematics? Second, it endows a human invention (mathematics) with cosmic significance. I don’t believe we’re that important on a cosmic level. We’re only important to ourselves.</p><p>Of course, there are also many things about the world that mathematics is unable to explain, or even represent. But when it applies, it generally offers deep insights. Jack Cohen and I suggested (in a <a href="https://a.co/d/21XVbx6">book</a> of the same title) that part of the answer is that the human brain is a ‘<em>figment of reality</em>’ — rather than reality being a figment of our imagination, which some philosophers believe, along with physicists who think the apple disappears when we stop looking at it. That is: the brain evolved to represent the world around us; also, that a complex structure like the brain can function only in a universe that runs according to some system of rules.</p><blockquote>“The human brain is a ‘<em>figment of reality</em>’ — rather than reality being a figment of our imagination.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WXN1HXOQl25xYQ8tfBj-mw.jpeg" /><figcaption>Source: <a href="http://shutterstock.com/">Shutterstock</a></figcaption></figure><p>Mathematics seems to be one (rather recent) outcome of this evolutionary process, so recent that we’ve not had time to evolve mathematical ability. We get it by subverting abilities acquired through other activities, like swinging through trees or throwing rocks. So perhaps there’s some common ground between the two explanations I’ve just tried to rubbish: mathematics is something that the human brain has extracted from the real world, and it can, in turn, approximate some of the rules that the real world operates by.</p><p>The lesson from history is that whenever we get the feeling that our science and mathematics have finally got to grips with reality, reality gives us a big kick up the pants. I’m not trying to downgrade the importance of science: most other ways of coming to terms with the universe lack explanatory power and are little better than fairy-tales. But, to quote Shakespeare, there are more things in heaven and earth than are dreamt of in our philosophy. (By which he meant the whole panoply of human intellectual activity, of course.)</p><blockquote>“The lesson from history is that whenever we get the feeling that our science and mathematics have finally got to grips with reality, reality gives us a big kick up the pants.”</blockquote><p>So, ultimately, I retreat to my original statement: mysterious and wonderful.</p><p><strong>GL: </strong><em>Isn’t this the essence of Gödel’s First Incompleteness Theorem, namely, no matter how powerful our formal theory is, there will be always surprises?</em></p><p><strong>IS: </strong>As a statement about mathematics itself, Gödel’s Theorem makes it absolutely clear that we can never know the answer to every question. This reinforces my view.</p><p>However, I’m also thinking more generally about the relation between mathematics and reality. Or current leading theories of physics correspond amazingly well to observations, in those areas where precise and controlled observations are possible. But that doesn’t mean they’re an exact description of the real world. The biggest surprise, I think, would be if they actually are an exact description of the real world.</p><blockquote>“The biggest surprise would be if our leading theories actually are an exact description of the real world.”</blockquote><p><strong>GL: </strong><em>What three important pieces of advice do you offer to young non-mathematicians who want to learn and apply mathematics?</em></p><p><strong>IS: </strong>The same advice I’d offer to young mathematicians.</p><ol><li>Find a friendly mathematician — including a sympathetic mathematics teacher who actually knows something about the subject — and talk to them.</li><li>If you’re trying to understand a bit of mathematics and you get stuck, don’t just keep bashing away at the same obstacle. Don’t just ignore it, but put it on a metaphorical shelf in your mind, labelled ‘come back to this later’. Keep reading. Often something a few pages later will resolve the problem you were having. Or your subconscious will figure it out — the ‘lightbulb’ moment.</li><li>Be prepared to grapple with problems; don’t get discouraged when your attempts to solve them seem not to work. Confusion often leads to eventual enlightenment. You have to enjoy the battle, even when you don’t seem to be winning it. When light finally dawns, and you find a solution, all of that past effort will be worth it.</li></ol><p><strong>GL: </strong><em>I suppose you are one such friendly mathematician </em>😄. <em>We sincerely thank you for this enlightening and enjoyable discussion</em>.</p><p><strong>IS: </strong>You’re welcome!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c81386847856" width="1" height="1" alt=""><hr><p><a href="https://medium.com/godelian-letters/interview-with-ian-stewart-c81386847856">Interview with Ian Stewart</a> was originally published in <a href="https://medium.com/godelian-letters">Gödelian Letters</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>