Counterintuitive Consequences Research Articles

Regularized Regression Can Reintroduce Backdoor Confounding: The Case of Mass Polarization

Regularization can improve statistical estimates made with highly correlated data. However, any regularization procedure embeds assumptions about the data generating process that can have counterintuitive consequences when those assumptions are untenable. We show that rather than simply shrinking estimates, regularization can reopen backdoor causal paths, inflating the estimates of some effects, and in the wrong circumstances, even reversing their direction. Recently, Cavari and Freedman (2023), argued that declining cooperation rates in surveys have inflated measures of mass polarization. We show that this finding is driven by large penalty terms in their regularized regressions, which leads to the estimates being confounded with time. Alternative methods do not show a clear positive or negative effect of declining cooperation on estimated levels of mass polarization.

Social Rationality and Human Reasoning: Logical Expressivism and the Flat Mind.

This paper attempts to reconcile the claims that the mind is both flat (Chater, 2018) and highly rational (Oaksford & Chater, 2020). According to the flat mind hypothesis, the mind is a mass of inconsistent and contradictory fragments of experience. However, standard accounts of rationality from formal epistemology argue that to be rational, our beliefs must be consistent, and we must believe all the logical consequences of our beliefs. A social account of rationality is developed based on Brandom's (1994) logical expressivism, in which respecting the norms of logic and probability theory is still central but where these standards apply to our public commitments in social dialogical contexts rather than to our individual belief systems (Skovgaard-Olsen, 2017). According to this account, even if someone's individual beliefs are inconsistent, they cannot be condemned as irrational if they acknowledge the inconsistency and seek to resolve it. It is shown how this approach interacts with people's fragmented and shallow world knowledge, and its social distribution yields some counterintuitive consequences, such as it sometimes being rational individually not seeking to resolve contradictions. Other consequences of this social expressivist approach are considered, including for dual process theories of reasoning, our view of beliefs, the status of logic, and Fodor's (1983) view of central systems. It is concluded that people can have flat minds and yet be highly rational.

Quantifying the risk of spillover reduction programs for human health.

Reducing spillover of zoonotic pathogens is an appealing approach to preventing human disease and minimizing the risk of future epidemics and pandemics. Although the immediate human health benefit of reducing spillover is clear, over time, spillover reduction could lead to counterintuitive negative consequences for human health. Here, we use mathematical models and computer simulations to explore the conditions under which unanticipated consequences of spillover reduction can occur in systems where the severity of disease increases with age at infection. Our results demonstrate that, because the average age at infection increases as spillover is reduced, programs that reduce spillover can actually increase population-level disease burden if the clinical severity of infection increases sufficiently rapidly with age. If, however, immunity wanes over time and reinfection is possible, our results reveal that negative health impacts of spillover reduction become substantially less likely. When our model is parameterized using published data on Lassa virus in West Africa, it predicts that negative health outcomes are possible, but likely to be restricted to a small subset of populations where spillover is unusually intense. Together, our results suggest that adverse consequences of spillover reduction programs are unlikely but that the public health gains observed immediately after spillover reduction may fade over time as the age structure of immunity gradually re-equilibrates to a reduced force of infection.

How to Do Without Encroachment

Abstract This paper defends a version of epistemic contextualism that accounts for the ordinary judgements and theoretical principles that motivate pragmatic encroachment. Adopting this contextualist view, we can avoid the counterintuitive consequences of pragmatic encroachment, while still preserving its attractive applications.

Freedom, foreknowledge, and betting

AbstractCertain kinds of prediction, foreknowledge, and future‐oriented action appear to require settled future truths. But open futurists think that the future is metaphysically unsettled: if it is open whether p is true, then it cannot currently be settled that p is true. So, open futurists—and libertarians who adopt the position—face the objection that their view makes rational action and deliberation impossible. I defuse the epistemic concern: open futurism does not entail obviously counterintuitive epistemic consequences or prevent rational action.

The Counterintuitive Consequences of Sex Offender Risk Assessments at Sentencing

Virginia adopted a risk assessment to help determine sentencing for sex offenders. It was incorporated as a one-way ratchet toward higher sentences: expanding the upper end of the sentence guidelines by up to 300 per cent. This led to a sharp increase in sentences for those convicted of sexual assault. More surprisingly, it also led to a decrease in sentences for those convicted of rape. This raises two questions: (a) why did sentencing patterns change differently across these groups, and (b) why would risk assessment lead to a reduction in sentence length? The first question is relatively easy to answer. While both groups saw an expansion in the upper end of the sentencing guidelines, only sexual assault had the floor lifted on the lower end, making leniency more costly. The second question is less straightforward. One potential explanation is that the risk assessment served as a political or moral shield that implicitly justified leniency for those in the lowest risk category. Even though the risk assessment did not change sentencing recommendations for low-risk individuals, it provided a ‘second opinion’ that could mitigate blame or guilt should the low-risk offender go on to reoffend. This decreased the risks of leniency and counterbalanced any increase in severity for high-risk individuals.

Increased gene expression variability hinders the formation of regional mechanical conflicts leading to reduced organ shape robustness

To relate gene networks and organ shape, one needs to address two wicked problems: i) Gene expression is often variable locally, and shape is reproducible globally; ii) gene expression can have cascading effects on tissue mechanics, with possibly counterintuitive consequences for the final organ shape. Here, we address such wicked problems, taking advantage of simpler plant organ development where shape only emerges from cell division and elongation. We confirm that mutation in VERNALIZATION INDEPENDENCE 3 (VIP3), a subunit of the conserved polymerase-associated factor 1 complex (Paf1C), increases gene expression variability in Arabidopsis. Then, we focused on the Arabidopsis sepal, which exhibits a reproducible shape and stereotypical regional growth patterns. In vip3 sepals, we measured higher growth heterogeneity between adjacent cells. This even culminated in the presence of negatively growing cells in specific growth conditions. Interestingly, such increased local noise interfered with the stereotypical regional pattern of growth. We previously showed that regional differential growth at the wild-type sepal tip triggers a mechanical conflict, to which cells resist by reinforcing their walls, leading to growth arrest. In vip3, the disturbed regional growth pattern delayed organ growth arrest and increased final organ shape variability. Altogether, we propose that gene expression variability is managed by Paf1C to ensure organ robustness by building up mechanical conflicts at the regional scale, instead of the local scale.

Wormhole Time Machines and Multiple Histories

In a previous paper, we showed that a class of time travel paradoxes which cannot be resolved using Novikov’s self-consistency conjecture can be resolved by assuming the existence of multiple histories or parallel timelines. However, our proof was obtained using a simplistic toy model, which was formulated using contrived laws of physics. In the present paper we define and analyze a new model of time travel paradoxes, which is more compatible with known physics. This model consists of a traversable Morris-Thorne wormhole time machine in 3+1 spacetime dimensions. We define the spacetime topology and geometry of the model, calculate the geodesics of objects passing through the time machine, and prove that this model inevitably leads to paradoxes which cannot be resolved using Novikov’s conjecture, but can be resolved using multiple histories. An open-source simulation of our new model using Mathematica is available for download on GitHub. We also provide additional arguments against the Novikov self-consistency conjecture by considering two new paradoxes, the switch paradox and the password paradox, for which assuming self-consistency inevitably leads to counter-intuitive consequences. Our new results provide more substantial support to our claim that if time travel is possible, then multiple histories or parallel timelines must also be possible.

Neural Teleportation

In this paper, we explore a process called neural teleportation, a mathematical consequence of applying quiver representation theory to neural networks. Neural teleportation teleports a network to a new position in the weight space and preserves its function. This phenomenon comes directly from the definitions of representation theory applied to neural networks and it turns out to be a very simple operation that has remarkable properties. We shed light on the surprising and counter-intuitive consequences neural teleportation has on the loss landscape. In particular, we show that teleportation can be used to explore loss level curves, that it changes the local loss landscape, sharpens global minima and boosts back-propagated gradients at any moment during the learning process.

Why Moral Paradoxes Support Error Theory

Moral error theory has many troubling and counterintuitive consequences. It entails, for example, that actions we ordinarily think of as obviously wrong are not wrong at all. This simple observation is at the heart of much opposition to error theory. I provide a new defense against it. The defense is based on the impossibility of finding satisfying solutions to a wide range of puzzles and paradoxes in moral philosophy. It is a consequence of this that if any moral claims are true, then a lot of highly troubling and counterintuitive moral claims must be in their number. This means that troubling and counterintuitive moral claims are everybody’s problem—not just error theorists’, but also their opponents’. Indeed, there is a sense in which this shared problem is worse for the opponents of error theory than for error theorists themselves.

Epigenetic factor competition reshapes the EMT landscape

The emergence of and transitions between distinct phenotypes in isogenic cells can be attributed to the intricate interplay of epigenetic marks, external signals, and gene-regulatory elements. These elements include chromatin remodelers, histone modifiers, transcription factors, and regulatory RNAs. Mathematical models known as gene-regulatory networks (GRNs) are an increasingly important tool to unravel the workings of such complex networks. In such models, epigenetic factors are usually proposed to act on the chromatin regions directly involved in the expression of relevant genes. However, it has been well-established that these factors operate globally and compete with each other for targets genome-wide. Therefore, a perturbation of the activity of a regulator can redistribute epigenetic marks across the genome and modulate the levels of competing regulators. In this paper, we propose a conceptual and mathematical modeling framework that incorporates both local and global competition effects between antagonistic epigenetic regulators, in addition to local transcription factors, and show the counterintuitive consequences of such interactions. We apply our approach to recent experimental findings on the epithelial-mesenchymal transition (EMT). We show that it can explain the puzzling experimental data, as well as provide verifiable predictions.

Toward a practical legal rule revision in legal debugging

Computational law researchers have long been interested in developing machine legal reasoning systems and machine-readable rule-bases to simulate legal reasoning given by judges. Basically, legal rule-bases rely on literal interpretations of statutes. In practice, however, literal interpretation of statute can have counterintuitive consequences. To deal with such consequences, English courts (the courts of England and Wales) have developed rules of judicial interpretation for reinterpreting statutes i.e. choosing or modifying the meaning of the statute’s wording (golden rule) or determining the statute’s purpose (mischief rule), and these rules have not been investigated in computational law. Previous research proposed the concept of legal debugging to formalize detection and resolution of counterintuitive consequences in legal rule-bases. Legal debugging consists of two steps. The first step is culprit detection, which involves interacting with a judge in order to identify a culprit, which is intuitively a part of statute that can be determined as a root cause of counterintuitive consequences. The second step is culprit resolution, which involves revising the rule-base to resolve a detected culprit. In this paper, we match two inductive logic programming approaches to two rules of judicial interpretation. The first approach is a bottom-up approach, which involves identifying an exceptional situation of the case and generalizing the legal rule revision. We match the bottom-up approach to the golden rule by using knowledge bases in the form of legal ontologies. The second approach is a top-down approach, which involves determining the purpose of the statute and refining the legal rule revision so that the revision fits within the context of the statute. We match the top-down approach to the mischief rule by using knowledge bases in the form of metarules. By using these matches, legal debugging can perform more practical legal revisions during the culprit resolution step.

Evolutionary theory and feminism

It is often argued that evolutionary theory and feminism remain in tension, since the evolutionary view of human nature is hard to reconcile with the feminist view. The goal of this paper is to demonstrate that this thesis is false. This goal is realized by reconstructing a certain anti-feminist evolutionary argument (whose descriptive conclusion is the “patriarchal” picture of male and female nature, and the normative conclusion is the claim that given the deep differences between men and women the feminist postulates cannot be achieved) and providing its critique. The argument is based on three premises: a theory of parental investment, the assumption of a relatively large (compared with other species) men’s parental investment in ancestral environments, and the uncertainty of paternity. Its (descriptive) conclusion is the claim that men are “by nature” much more polygamously disposed, much more desirous of power (over the opposite sex), and much more aggressive than women. The paper presents several objections to this argument. The first objection questions its internal coherence. The second one points at its counterintuitive (not supported by empirical facts) consequences. The third one criticizes one of the assumptions of the argument, i.e. the assumption about a relatively large (compared with other species) men’s parental investment in ancestral environments.

The inefficiency of plea bargaining

AbstractThe fundamental assumption underlying plea bargaining is that it is ‘efficient’. Even ardent critics of plea bargaining often accept the assumption that it is efficient in being cost effective. For proponents, beliefs about plea bargaining's efficiency are used to argue that it is beneficial, even necessary, on pragmatic grounds. From this assumption about efficiency, volumes of laws and policies have been created. This article uses empirical data to demonstrate the flaws in this foundational tenet of plea‐bargaining logic. The article shows that, in reality, plea‐bargaining practices have imported their own inherent inefficiencies into the criminal process. These inefficiencies are detrimental not only to cost effectiveness but also to justice and fairness. Foremost among these inefficiencies are the (some argue iniquitous) game‐playing cultures that, ironically, policymakers have unintentionally fostered in striving for efficiency. The article demonstrates the pressing need to better recognize and understand these counterintuitive inefficient consequences of plea bargaining.

On Cotnoir’s two notions of proper parthood

A.J. Cotnoir has argued that we should distinguish between two notions of proper parthood: outstripped part and non-identical part. Outstripped parthood is an asymmetric relation, but non-identical parthood is not. We argue, first, that the intuitions Cotnoir uses to motivate these notions do not always give the right verdict; and, second, that systematic reasons for distinguishing these two notions of parthood have further counter-intuitive consequences. This means the distinction between two notions of proper parthood currently lacks adequate motivation.

Blue collar with tie: a human-centered reformulation of the ironies of automation

When Lisanne Bainbridge wrote about counterintuitive consequences of the increasing human–machine interaction, she concentrated on the resulting issues for system performance, stability, and safety. Now, decades later, however, the automized work environment is substantially more pervasive, sophisticated, and interactive. Current advances in machine learning technologies reshape the value, meaning, and future of the human workforce. While the ‘human factor’ still challenges automation system architects, inconspicuously new ironic settings have evolved that only become distinctly evident from a human-centered perspective. This brief essay discusses the role of the human workforce in human–machine interaction as machine learning continues to improve, and it points to the counterintuitive insight that although the demand for blue-collar workers may decrease, exactly this labor class increasingly enters more privileged working domains and establishes itself henceforth as ‘blue collar with tie.’

Promotion of behavioural change for health in a heterogeneous population.

Public health policy often involves implementing cost-efficient, large-scale interventions. When mandating or forbidding a specific behaviour is not permissible, public health professionals may draw on behaviour change interventions to achieve socially beneficial policy objectives. Interventions can have two main effects: (i) a direct effect on people initially targeted by the intervention; and (ii) an indirect effect mediated by social influence and by the observation of other people’s behaviour. However, people’s attitudes and beliefs can differ markedly throughout the population, with the result that these two effects can interact to produce unexpected, unhelpful and counterintuitive consequences. Public health professionals need to understand this interaction better. This paper illustrates the key principles of this interaction by examining two important areas of public health policy: tobacco smoking and vaccination. The example of antismoking campaigns shows when and how public health professionals can amplify the effects of a behaviour change intervention by taking advantage of the indirect pathway. The example of vaccination campaigns illustrates how underlying incentive structures, particularly anticoordination incentives, can interfere with the indirect effect of an intervention and stall efforts to scale up its implementation. Recommendations are presented on how public health professionals can maximize the total effect of behaviour change interventions in heterogeneous populations based on these concepts and examples.

Luck Theory: A Philosophical Introduction to the Mathematics of Luck by Nicholas Rescher

Reviewed by: Luck Theory: A Philosophical Introduction to the Mathematics of Luck by Nicholas Rescher Steven D. Hales RESCHER, Nicholas. Luck Theory: A Philosophical Introduction to the Mathematics of Luck. Cham, Switzerland: Springer, 2021. xxxiii + 90 pp. Cloth, $70.00 Usually luck is thought of as something chancy, such as an unlikely occurrence, or something that might have easily not happened. Other kinds of luck seem like inexorable fate, like the bad luck of coming of age just in time for a wartime draft. In his latest book on luck, Nicholas Rescher focuses on the first kind, stochastic luck. In his first book on the topic, Luck: The Brilliant Randomness of Everyday Life (1995), Rescher defended a probability account of luck, one of the few contemporary philosophers to do so. Here he is back for more of the same, but centered on neither the philosophy nor the psychology of luck. Instead, his approach in his new book is mathematical and formal. He is convinced both that “virtually all consequences of the theory are intuitively acceptable” and also that “virtually all intuitively acceptable features of luck can be verified in the theory.” The claim that all the consequences of his theory are intuitively acceptable will come as a surprise to defenders of moral luck. In half a paragraph, Rescher dismisses the topic entirely: “[S]trictly speaking, there is no such thing as ‘moral luck’” (repeated twice). There is a long history of attempting to formalize and mathematize our understanding of luck. The origins of probability theory are due to Cardano, Pascal, Fermat, de Montmort, and others trying to solve gambling problems, quintessential examples of good and bad luck. By 1718 de Moivre had declared that good luck meant nothing more than chancy success (being “a Gainer at play”). Now, many of the early probability theorists took themselves to be eliminativists about luck: To understand the probability calculus was to recognize that luck is not a real thing. By contrast, Rescher thinks luck is real and is giving a theory of what it is. But what’s wrong with the eliminativist approach? If probability theory tells us all about chanciness, and luck is just chanciness, why hang on to an antiquated notion like luck? Rescher’s answer is familiar in the luck literature: Not everything improbable is a matter of luck. It is improbable that a coin toss would come up tails three times in a row, but unless you had a wager riding on it, it’s not a matter of luck. This feature is commonly known as a [End Page 152] significance condition, and Rescher endorses it. His mathematics of luck is built out of the probability calculus plus a calculus of value. The project of providing a mathematics of luck smacks of a solution in search of a problem. Providing a formal machinery lends an unjustified air of precision, as if at last we are reading the book of nature in its original language. For example, Rescher’s Basic Luck Equation is λ{O} =| O| – E. This equation means that the luck of an outcome (λ{O}) is equivalent to the difference between that outcome’s yield for an agent and the expectation of that yield. This equivalency sets up a linear relationship, but why think that’s how luck works? Maybe luck should be measured on a logarithmic scale, as the Richter scale measures the power of earthquakes, or luck could be a heteroscedastic variable that varies unequally as the result of yield and expectation. Or it could be that luck is something else altogether. Rescher writes that “lack of control is neither a sufficient nor a necessary condition for luck,” for instance, but he barely engages with the considerable literature on this idea. For Rescher, the inevitable cannot be lucky (Theorem 5). While initially appealing, that has some counterintuitive consequences. For example, if the world is deterministic and everything is inevitable, then nothing is lucky. Intuitively we are lucky that the constants of the universe hit the sweet spot of making life possible, but not for Rescher, given that those constants were nomologically necessary. We are neither lucky that God exists (since his existence would be necessary) nor unlucky...

Knowability Without Rigidity

The most straightforward interpretation of the principle of knowability is that every true proposition may be known. This, taken together with some intuitively appealing ideas, raises a problem known as the Church–Fitch paradox. There is a wide variety of alternative interpretations of the principle of knowability that have been offered to avoid the paradox. Some of them are based on rigidification of certain aspects of what is knowable. I examine three proposals representing this strategy, those by Edgington, Rückert and Jenkins. Edgington defines what is knowable as a proposition prefixed by the actuality operator. Rückert and Jenkins maintain that what makes a proposition knowable is the possibility of knowing de re (Rückert) or recognizing (Jenkins) the state of affairs that renders the proposition actually true. In both cases, the link to the actual world (or situation) rigidifies what is knowable in some aspect or other. I argue that all three theories have strongly counterintuitive consequences, and I offer an interpretation of the principle of knowability that is both free from rigidity and immune to the Church–Fitch argument.

When team sponsorship of cause-related messages leads to unfavorable fan reactions: the interplay among competency, cause orientation, and message source

ABSTRACT Research question One general conclusion from existing studies is that if a team spends more resources and efforts on socially responsible initiatives, fan responses are more likely to be favorable. In the current study, however, the researchers tested a boundary condition in which consumer responses toward a team's sponsorship of cause-related messages may be paradoxically unfavorable. That is, we sought to explore when and how ‘doing good’ (cause sponsorship) might not complement ‘doing well’ (performance). Research methods Two online experiments were conducted employing participants through MTurk. The main experiment was a 2 (sponsor competency: low vs. high) × 2 (cause orientation: philanthropic vs. promotional) × 2 (message source: team-controlled channel, sponsor vs. third-party channel, news media) between-subjects design; three specific cause categories (community service, donation, and pro-environmental activism) were utilized as a within-subjects factor. Four NFL teams were identified based on their cumulative athletic performance records. Results and findings The results revealed that for high (vs. low) competency teams, philanthropic (vs. promotional) cause sponsorship induced more favorable consumer responses. Also, communication with fans through sponsor-controlled channels exacerbated unfavorable fan reactions, especially in the case of competitive teams’ philanthropic cause sponsorship initiatives. Implications The current study helps broaden the existing knowledge of CSR effectiveness by exploring the seemingly counterintuitive consequences of cause sponsorship. The outcomes of the study also offer important suggestions for managers regarding potential approaches for sports organizations to strategically allocate their resources toward developing effective CSR initiatives.