Integrated Valuation When Choices Simultaneously Affect Both Oneself and Others

DIT (Dual Interest Theory) stands on solid empirical ground with testing of the Null Hypothesis of no role of the shared-with-the-other-interest in tempering self-interest economic driven choice rejected over and over for 4-decades. The search continues for more empirical evidence supporting or denying that DIT is a more empirically sound theory than SIT (Single Interest Theory used in mainstream Microeconomics) as demonstrated by rejecting that Null. And, if the Null is not rejected, the SIT also then stands on solid ground. So, said testing serves to test both theories.

So, in search of more empirical evidence supporting (or not) DIT, I have finished reading:  Rhoads, Shawn A., Lin Gan, Kathryn Berluti, Katherine O’Connell, Jo Cutler, Patricia L. Lockwood and Abigail A. Ma. "Neurocomputational Basis of Learning When Choices Simultaneously Affect Both Oneself and Others." Nature Communications 16 (October, 2025): 1–16.

The following is a Review, looking for empirical evidence supporting the two main DIT contentions of 1) nonallocability leading to nonseparability in outcomes, and 2) incommensurability. The nonallocable feature of economic choice is represented in overlapping isoquants and indifference curves as represented in a Market Forum.  The incommensurability is reflected in the fact that Other Forums operating on a higher plane are essential in giving context to the Market Forums, the latter operating in the domain of commensurability, e.g., price P of some measurable quantity of an input and/or product and good.  Other Forums are charged with the old problem of resolving the best balance in the jointly operant pushpins(material, quantifiable, price P, profane)&poetry(nonmaterial, nonquantifiable, value O, sacred), with the latter nonseparable.

Using the DIT Lens

Looking at some key quotes, and using the DIT Figures 1 and 2 to help look through the Dual Interest Theory (DIT) Lens, starting with the Abstract (p. 1):

“Computational modeling tested whether people simulate how other people value their choices or integrate self- and other-relevant information to guide choices. An integrated value framework, rather than simulation, characterizes multi-outcome social learning. People update the expected value of choices using different types of prediction errors related to the target (e.g., self, other) and valence (e.g., positive, negative).”  DIT sees the computational model in the research as examining choices made within the Human brain as represented in Figure 1, with path 0Z the attractor of best choice. Said choice is influenced by consideration of the situation of the other as represented on the integrated value framing represented on the higher plane of Figure 2. Think of the brain of the Econ as represented in Single Interest Theory (SIT) in mainstream (Micro)economics as operating only on path 0G, while the Human brain represented in Dual Interest Theory (DIT) in Metaeconomics is pondering what path to take, which will hopefully be some better path 0Z.  The Rhoads et al (2025) paper is about what path a Human would take given a changing social context, the latter represented on path 0M in Figure 1, as influenced by learning about and with the other as represented in Figure 2.

Then on p. 2: “Specifically, there were four categories of stimuli, although participants were not aware of this. These were: mutually beneficial (the option that most frequently yielded beneficial outcomes for both themselves and their partner), instrumentally harmful (the option that most frequently benefited themselves at a cost to their partner), altruistic (the option that most frequently benefited the partner at a cost to the self), and mutually costly (the option that most frequently yielded losses to both themselves and their partner).”  In DIT terms, mutually beneficial and altruistic can be on the same path 0Z. The mutually beneficial choice is some best point B in both Figures 1 and 2. And, every choice on path 0Z, including the best point B, is altruistic in the sense of a bit of sacrifice in the payoff from self-interest and bit of sacrifice in the payoff from the shared-with-the-other-interest as well. Losses to both self and the other are choices outside the realms of path 0G and 0M in the irrational zones.

On p. 4: “On average, people exhibit self-benefiting biases Across both tasks, participants most frequently selected the options that benefitted the self (i.e., the mutually beneficial option or the instrumentally harmful option… ).” DIT makes clear it is all about the own-self, and that is a matter of degree as to how much self-benefit is chosen.   All choice in the boundedly rational areas between path 0G and path 0M is self-benefiting, but some choices are better than others. It is about the attractor path 0Z in Figure 1 space, as influenced by the higher plane of value operating in Figure 2 space.

Also found people low on empathy-with, especially demonstrated psychopathy, were less willing to even learn about the gains or losses of others in a choice:  “Psychopathy is viewed as a multidimensional construct that reflects meanness, boldness, and disinhibition … explored how learning varied as a function of these sub-factors (p. 4).”  Found deeper degrees of psychopathy led to less learning about the other, less concerned, less engaged --- in effect, less willing to act out with consideration of the shared-with-the-other-interest of path 0M. Meanness especially  shifted most attention to path 0G outcomes.

People Integrate Self&Other into an Own-self Based Choice

As Rhoads et al (2025) say it:

“…computational and neural mechanisms of simultaneous learning for self and others … provide evidence that people learn to integrate self- and other-relevant information into a single value per choice. However, this value is updated asymmetrically according to distinct prediction errors, which encode information about the target (self or other) and valence (positive or negative). Participants who were more sensitive to unexpected positive and negative outcomes for others learned to make more prosocial choices and fewer antisocial choices. By contrast, as trait-level psychopathy increased, other-relevant learning rates decreased and self-relevant learning rates increased, suggesting a computational phenotype underlying the acquisition of antisocial behaviors in psychopathy (p. 7).” In DIT terms, the findings support an integration across the profaneF&sacredS at point B in Figure 2, for the normal Human but not so much for the psychopath, as associated with the learning on the higher plane of value O. It comes from consideration and influence at the more micro level, inside the Human brain processing in Figure 1, but learning in Figure 2  --- especially for the normal Human --- influences the eventual choice on some better path 0Z.  Harm to both is hardly ever chosen, except by psychopath style personalities, which is choice outside of the attractor region associated with path 0Z. Yet, the  own-self is always in play if some point on some path 0Z is chosen. And, psychopathy keeps operating close to, if not on some path 0G, and choosing something akin to point A.

“… people integrate self- and other relevant information during learning to guide future prosocial behaviors, suggesting that the brain combines self- and other-regarding

information into a common valuation signal, rather  than maintaining entirely separate valuation systems for the self and others when choosing options that have dual social outcomes. In other words, the way people distinctly encode self- and other-relevant outcomes that result from a particular behavior guides how desirable that same behavior will be in the future, regardless of whether the behavior is mutually beneficial or costly, instrumentally harmful, or altruistic (p. 8).” Well, yes, people have the attractor region associated with path 0Z always in play, unless completely a sociopath on path 0G especially when operating outside path 0G and close to the vertical axis. DIT would also posit it could also happen for the case of an extreme empath operating outside of path 0M and approaching the horizontal axis.

“… were guided by a single expected value, which was encoded by activation in medial prefrontal cortex, which is consistently found to track expected values of choices during learning

and decision-making in self-relevant …  and social contexts … findings provide a new account of its role in decision-making when learning to choose behaviors that simultaneously affect the self and another, sometimes in the same way but sometimes in ways that conflict (p. 8).”  Yes, the higher plane of influence addressed by the prefrontal cortex, from value O considerations in Figure 2 space (considering the situation of the other) coming to give the better choice on some path 0Z in Figure 1 space (internalized within the own-self, within the own-brain).

“Updating an expected value that integrates information about self and others might be most efficient in many contexts. It is possible that more complex tasks could require updating of differential values. For example, when learning in the context of social influence, one must additionally represent the value function that guides other people’s behavior (p. 8).”  DIT clarifies the internal balancing within the Human (brain) represented in Figure 1 considerations is resolved through consideration (especially cognitive empathy, but also affective empathy) operant in the Figure 2 space involving consideration of the situation of the other, considering the context of the social influence.

“… found a link between trait psychopathy and learning rates, a finding consistent with prior studies suggesting atypical learning for others with increased psychopathy and antisocial behavior (p. 9).”  To be expected, as DIT clarifies that the underdeveloped capacity for empathy-with the other will keep the person over in the path 0G region of the brain, and, perhaps, even irrationally outside of it and near the vertical axis of Figure 1, never even considering the other as reflected in Figure 2.

“… finding demonstrates a potential basis in learning for the self-serving, uncaring behaviors that typify adults with psychopathic traits, most notably the persistent instrumental aggression that is uniquely linked to psychopathy. Unlike reactive aggression, which is emotional and spontaneous, instrumental aggression is deliberately performed to achieve a goal or reward despite the harm or cost to the victim …  present results suggest a computational phenotype of instrumental aggression among people who are distinguished by their relatively blunted learning to avoid behaviors that harm others or choose behaviors that benefit them, despite preserved ability to select behaviors that benefit themselves (p. 9).”  Again, DIT sees the self-serving choices on path 0G, and, in the extreme of mental illness represented in sociopathy and some forms of psychopathy, to operate in the irrational zone above path 0G.

People See Own-Self Embedded in Social Context

“It is important to note that in our task all decisions were embedded in a social context, because all choices carried joint social consequences even in trials when only self-relevant outcomes were displayed. Thus, participants were aware that their decisions affected their partner regardless of the outcome they saw (p. 9).” So, yes, in DIT terms, the social context of Figure 2 was put into view, and influenced the choice in Figure 1 (within the Human brain) choice. Figure 2 is all about the social learning, as the situation of the other is pondered, and, then, in normal people, works to strike some better balance in the choice in Figure 1 space.

“… observation corroborates work showing that people update the expected value of actions asymmetrically according to better- or worse-than-expected outcomes during self-relevant

learning (p. 9).” DIT clarifies that self-relevant learning starts by having empathy-with the other, which brings the perspective of the other into view as represented in Figure 2 space.  Said perspective taking then works to influence the choice within the Human brain in Figure 1 space.

“… (the research) provides a neurocomputational account of how people learn for others when required to consider how a single choice will affect themselves and someone else. These findings extend our understanding of the computational substrates of the types of prosocial and antisocial learning that occur frequently in daily life: learning behaviors that benefit others at a cost to the self, and learning behaviors that benefit the self at a cost to others. They suggest that learning about the consequences of our actions for oneself and others over time cannot as easily be teased apart as the correlates of value-based decision-making for self and others in a single moment … value signals cannot be disentangled at the level of decisions, but only at the level of outcomes, our findings show that learning to select actions that can help or harm both the self and others does not require the active representation and maintenance of another person’s value function separate from one’s own (p. 9).”  DIT sees it the same way, as in everything is internal to the own-self, as in the other person’s value function is internalized within the individual pondering what choice path to take. But, the ultimate choice in Figure 1 space --- within the Human brain… is influenced by both the cognitive and affective empathy-with the other as represented in Figure 2 space, the latter reflecting considerations evolving through time and having different influences on choices in Figure 1 (within the Human brain) space.

DIT Based Bottomline

Rhoads et al (2025) is providing clear and robust support for the DIT framing of Figures 1 and 2. It especially supports something akin to Figure 1, which points to a jointly operant self&other-interest. The Human brain has evolved both biologically and culturally to integrate value information associated with the potential for taking the perspective of the other.  In the real world, said perspective especially comes from interacting with others in the Other Forums to include family, community, country, civilization, and Humanity writ large, as represented in Figure 2.  Said dynamic and ever evolving value O in Figure 2 space, operating on said higher plane, works to integrate the value choice in Figure 1 space, the internal circuitry within the Human brain.  Said circuitry is represented generally by the ego-associated circuitry of path 0G and the empathy-associated circuitry of path 0M, the overlapping isocurves making clear said circuitry is absolutely integrated within normally healthy, mentally balanced and stable individuals. Such matters as psychopathy and sociopathy take the Human brain in regions beyond path 0G and/or path 0M.   The attractor zone between the two paths 0G and 0M represented in dynamic movement of path 0Z is the sweet spot of homeostatic balance, which itself is quite dynamic and ever changing.

It seems Rhoads et al (2025) are demonstrating several core DIT claims. So, I asked both ChatGPT and Claude “What am I missing?”

ChatGPT as Integrator

Commented and nudge to ChatGPT:

Would like a Review of the attached Rhoads et al journal paper, the Review done through the DIT (Dual Interest Theory) Lens... looking for support for DIT, and findings counter to DIT, especially the DIT ideas of jointness in the jointly operant self&other-interest, on the way to jointly arising balance in the profane&sacred.

ChatGPT Found Support for DIT

ChatGPT claimed “… that Rhoads et al. offer substantial support for DIT's conception of jointly operant self- and other-interest, while also showing why jointness cannot simply be equated with balance, mutual benefit, or moral development. The most consequential finding is that people learning from choices affecting themselves and others appear to integrate information about both parties into a single expected value for each choice, rather than necessarily maintaining two separate value functions—one for themselves and another for the other person … highly relevant to Dual Interest Theory (DIT), especially if DIT understands self- and other-interest as distinguishable yet jointly operative dimensions of human motivation, rather than mutually exclusive or necessarily competing forces.”  The overlapping isocurves of Figure 1 make clear that DIT sees the jointly operative dimensions of human motivation.

And, the Rhoads et al research study and findings most useful contribution to DIT:  “Jointness may be a fundamental feature of social valuation, whereas balance is a variable—and perhaps developmental—achievement within that jointness … study makes the first proposition empirically plausible. It provides suggestive, but not direct, evidence about the second. And it does not investigate the culminating DIT proposition of jointly arising balance in the profane and sacred.”  True, about the latter:  DIT does frame the self-interest as mainly operant in the domain of the profane and the shared-with-the-other-interest, or simply other-interest as mainly operant in the domain of the sacred, as in profaneF&sacredS sorted out at point B in Figure 2, which influences the selection of point B in Figure 1.

ChatGPT goes on to explain the importance of distinguishing three things:

• Empirical findings: What Rhoads and colleagues actually measured and established.

• DIT interpretation: How those findings support, challenge, or potentially refine Dual Interest Theory.

• Theoretical extension: What DIT might predict about increasingly balanced self–other relations and their possible expression in the profane–sacred relationship.

The claim is that the separation matters, because the paper's computational definition of integration is narrower than DIT's richer conception of jointness.

On empirical findings, it is especially notable that the study “combined two features that might initially appear contradictory:

• A unified value guiding a particular choice.

• Differential sensitivity to information about who benefits or suffers.

… coexistence is especially revealing … (as in) is no requirement that self- and other-interest become indistinguishable in order to act jointly.”  In effect, the study supports the notion that two paths 0G and 0M do not have to be the same path, as SIT presumes (i.e., everything is about self-interest, so path 0M is subsumed under path 0G), a claim without any empirical standing.

And, the strongest support: Self- and other-interest are integrated in choice valuation “…

results favored an integrated-value model over models maintaining separate expected values for self and other (pp. 3–4) … ‘people learn to integrate self- and other-relevant information into a single value per choice’ (p. 7) … most significant convergence with DIT.”  Yes, in effect, Humans choose some point B in Figure 2, which influences the choice of some point B in Figure 1.

ChatGPT continues, that in  “… conventional interpretations (Note: Read SIT) of self-interest versus other-interest, an individual might be described as switching between two competing motives: sometimes acting selfishly, sometimes altruistically… DIT's jointness proposition suggests something different. Both interests may be active constituents of the same motivational process, even when they differ in their relative influence. Rhoads et al. provide computational evidence consistent with that possibility. The individual's learned valuation of an action can reflect information about outcomes for both self and other.”  The SIT conventional frame would say the other-interest represented on path 0M is subsumed as an argument within the self-interest only max U frame represented on path 0G. The empirical evidence in Rhoads et al. counters that claim.

As ChatGPT elaborates it, the DIT implication is: “The distinction between self-interest and other-interest need not imply that they exist as separate, independently operative motivational systems… can be analytically distinguishable but functionally integrated … support for joint operation … (but) … not yet proof of the stronger DIT claim that self- and other-interest are necessarily co-constitutive in every human motivation.”  Good point.  More research is needed as to said jointly operant self&other motivating every economic choice, although it is well-established it does motivate conservation choices as in Empathy Conservation.

Empirical results also support, as ChatGPT points out, a useful DIT principle:  “Jointness need not mean sameness, and balance need not mean equality of components. If both interests are jointly operant, their differential contributions need not dissolve their jointness. Indeed, an adaptive system may need to distinguish what is happening to oneself from what is happening to others. Differentiation could be a requirement for sophisticated joint regulation rather than evidence against it.”  Yes. DIT makes clear that what happens to the self gets more attention, as in a more primal ego-based self-interest.  So, one must sort out self-interest from other-interest, even though both arise jointly.

And, on what circuitry and brain organs, regions are in play:  “The neurological results also fit this interpretation … paper identifies both overlapping and distinct patterns of activation associated with value updating, including the ventral striatum, anterior cingulate regions, anterior insula, and amygdala (Rhoads et al 2025, pp. 5–8) … findings support the existence of differentiated neural contributions to an integrated computational process… do not, however, establish a single anatomical seat of the jointly operant self–other relationship.”  Many different organs and regions are integrating, working to find the best choice on path 0Z.

The experiment was organized as having four options: mutual benefit, instrumental harm, altruism and mutual cost.  The mutual benefit option was most frequently chosen. As ChatGPT characterizes it: “Mutual benefit is important because it demonstrates an outcome structure in which self-interest and other-interest are not in a zero-sum relationship. Both can be advanced through the same action. From a DIT perspective … is consistent with the possibility of mutually enhancing interests: The advancement of one does not require the suppression of the other.” Also, as DIT would predict,  “Rhoads et al. found that higher levels of subclinical psychopathic traits, particularly meanness, were associated with reduced sensitivity to information about outcomes for others.”  Yes.  Such reduced sensitivity means less attention to both cognitive and affective empathy essential to seeing the shared other-interest on some path 0M.

ChatGPT Also Raises a Cautionary Note

The following table was produced by ChatGPT. 

The really only concerning claim is the fact one cannot leap to claiming --- based on Rhoads et al findings --- that the participants in the study would make the same choices if the notion of profaneF&sacredS were put into the experimental questioning. That is, if the participants could in effect interact with some other, such that the FS interaction term could be tested for significance. The Claude assessment suggests FS could be tested in part from revisiting the data set, and, clearly an experiment designed specifically to test FS is needed.

ChatGPT also pointed out the finding that people in the experiment clearly did not choose something akin to a 50:50 balance.  Clearly the self-interest was more primal. So, while the results support a jointly operant self&other, the balance chosen can easily be, and most often was, tipped toward the self-interest. So, the path 0Z was often much closer to path 0G than to path 0M, with substantive sacrifice in the domain of shared-with-the-other-interest in favor of self-interest payoffs.  Also, DIT sees the possibility for dark empathy, which could explain how the meanness frame resulted in a perverse kind of balance in self&other-interest.   DIT does not inherently frame empathy-with as good, benevolent sharing --- it could be mean a kind of dark empathy based sharing used to in effect exclude if not outrightly harm “them” to favor the “us” within a narrower group.

Like ChatGPT framed it, after some deeper inquiry on said front, the “…  DIT framework becomes more powerful if jointness is morally open—capable of constructive and destructive expressions—while balance is the achievement that differentiates them.” Yes, it does, and, the DIT framework is open to both different moral and ethical considerations, all giving content to the shared-with-the-other-interest.  DIT stands on empirical ground, seeking what works best. 

ChatGPT Bottomline

Overall, ChatGPT finds substantive support in the experiment for DIT.  It points to the special need to design experiments to specifically look into the notion of jointness and nonseparability with respect to the profaneF&sacredS, in effect looking specifically at the role of the interaction FS. And, said interaction will likely stand the test, as it has stood that test in 4-decades of testing using personal interview and mail survey data, as well as controlled experiments in the economic laboratory.

Claude as Integrator

Comment and nudge to Claude:

Would like a Review of the attached Rhoads et al journal paper, the Review done through the DIT (Dual Interest Theory) Lens... looking for support for DIT, and findings counter to DIT, especially the DIT ideas of jointness in the jointly operant self&other-interest, on the way to jointly arising balance in the profane&sacred.

Claude Found Support for DIT

While the researchers likely are quite unaware of DIT, the experiment asked a key question for a study framed by DIT thinking:  “Do people keep two separate value functions, one for self and one for the other (the “2Q” or “simulation” family), or one integrated value per choice updated by differently weighted prediction errors (the “1Q” or “integration” family)? Integration won decisively across every sample and model-comparison metric. The winning model had four learning rates, crossing self/other with positive/negative, and a single inverse temperature.”

As Claude then points out:  “For DIT this is (a) most relevant neurocomputational paper … the task is literally a Frisch-jointness design. Each fractal is a nonallocable input jointly producing a self-outcome and an other-outcome, the way sheep feed jointly produces mutton and wool. The … design embodies it. That makes the paper a fair test of what DIT says about jointness…”

Claude continues:  “The rejection of the 2Q models is, read correctly, a rejection of the two-ledger picture (that DIT) insists the “&” is not. In the 2Q models, self-value and other-value are kept as separate accounts (Note: as in Single Interest Theory --- SIT --- in mainstream economics) and weighed against each other at the moment of choice. That is precisely the “two separable accounts” reading (which distinguishes ) DIT from (SIT). The data say the brain does not work that way when one action carries consequences for both. Instead there is one value per choice, a single own-interest if you like, into which self-relevant and other-relevant information has already been folded. The … closing claim fits (DIT) own-interest framing almost exactly: learning to help or harm both self and other does not require maintaining another person’s value function separate from one’s own. On this reading, own-interest emerges from self&other-interest rather than from a sum of two independently maintained interests…” 

Going deeper into the Human brain, looking to circuitry and specific involvement of organs: “The neural results add a second line of support at the level of substrate. Several regions encode value updates from both self- and other-relevant prediction errors. Pregenual ACC encodes positive updates regardless of target. Subgenual ACC encodes other-relevant updates of both valences and also self-relevant positive updates. Ventral striatum encodes self-positive bilaterally and other-positive on the right. That is shared circuitry serving both interests, which is the neural analogue of a nonallocable input. You cannot cleanly assign pgACC or sgACC to ‘the profane’ or ‘the sacred.’ Meanwhile medial prefrontal cortex carries the integrated expected value at decision time. The large set of regions with negative value coding (dorsolateral PFC, SMA, anterior insula, precuneus, parietal lobule) is at least consistent with your third circuitry set, the deliberative and self-control system that monitors and adjusts the jointly valued interests. … hold that mapping loosely, because the authors interpret the negative coding as possibly an artifact of the emotional context rather than control. Still, the overall architecture fits DIT’s three-system picture better than it fits a single-interest maximizer: distinct sources, partly shared substrate, and a common valuation site.” 

The research results also suggest, as does DIT, that the sacred works to temper the profane:  “The cleanest DIT-relevant behavioral comparison in the paper is one the authors don’t emphasize. Mutual benefit and instrumental harm have identical self-payoff distributions; they differ only in what happens to the partner. In the four-option task, people chose mutual benefit 45.4% of the time and instrumental harm 29.1%. That 16-point gap, at zero profane cost, is a measurable sacred signal tempering profane pursuit. It is the soft-& in miniature: empathy-based consideration informing which of two equally self-serving paths gets taken. Mutual benefit is also the only option where the interest is genuinely shared, which is what DIT’s sacred domain specifies.”

Results also suggest, like does DIT, that a weak empathy-circuitry leads to dominance of the ego-circuitry:  “The psychopathy results support DIT’s claim that the sacred side is a real, independently varying dimension and not noise around self-interest. Higher Triarchic Psychopathy scores predicted lower other-relevant learning rates while self-relevant learning was preserved. This was the Target × Psychopathy interaction, significant in both tasks, and in the two-option task higher scores also predicted lower learning from negative prediction errors. The effect was driven mainly by the meanness factor. That is profane capture at the level of the individual learning system, a computational Scroogism, and it is what DIT predicts if the empathy-based circuitry can be weak while the ego-based circuitry remains intact.”

Claude Also Raises a Cautionary Note

The perhaps biggest disconnect is that the research did not specifically test the FS interaction.  Results seem to imply the profaneF&sacredS arise jointly, but no specific test was made of the proposition. It is left to conjecture, in that the two domains did come out as joint outcomes.

And, a result that needs specific clarification in DIT explanations:  “… the brain does appear to reduce self&other to one common currency … (DIT already does) treat incommensurability as secondary to nonallocability, so this need not be damaging. The defensible DIT position is incommensurability at the source and commensuration through cognitive control at choice. But (DIT analysts need to) state that explicitly, because a mainstream reader will take this paper as evidence for a single neural utility.”   DIT clarification here is that the commensuration through cognitive control in Figure 1 (within the Human brain) space at some point B is influenced by considering the perspective of the other as depicted at point B in Figure 2.  And, while this experiment does not outrightly support that claim, it is suggestive. So, it does  not have to be a single neural utility like SIT claims, but rather an integration in Figure 1 space made possible from Figure 2 considerations.

Also, nudging empathy-with the other did not give as strong a result as DIT would posit:  “The null results on empathy are the hardest finding for DIT. The pre-registered prediction that QCAE cognitive and affective empathy would predict other-relevant learning rates failed. The exploratory test of social value orientation also found nothing. DIT grounds the sacred domain specifically in cognitive empathy, so this is a direct miss rather than a peripheral one. Individual differences on the sacred side appeared only at the dark end, as callous meanness reducing learning, not at the bright end as empathy increasing it. Small samples and self-report measures are a fair caveat, but the test was pre-registered and should be taken seriously. It suggests that what varies across people in this setting is the absence of callousness more than the presence of empathy, and DIT’s empathy-to-institutions chain may need to accommodate that asymmetry.” Well, DIT can easily accommodate it, as it allows for not only positive and good but also dark empathy. The meanness is a form of dark empathy, and, clearly it was at play.

Claude suggests going to the data set and use it to test the significance of the FS interaction, test for nonseparability, like in a null hypothesis that the payoff in the realm of self-outcome is separable from the payoff in the realm of other-outcome: “The data and code are on OSF (osf.io/2gnv4) and the pyEM fitting package is public. That makes the decisive DIT test feasible without new data collection. You could add a model in which the self-relevant learning rate, or the subjective value of a self-outcome, is modulated by the option’s learned other-value. A simple version would include a multiplicative self × other term in the outcome utility… would then compare it against the winning additive 1Q model using the same Bayesian model comparison. If the interaction model wins (one has found the) FS signature in a neurocomputational learning task … If it doesn’t, the null of separability stands … which DIT’s own framing accepts. Either result is informative. A natural experimental follow-up would show both outcomes on some trials, which is the design the authors themselves call for, since nonallocability should be most visible exactly when people see the joint outcome.” It is highly likely the null of separable payoff would be rejected, based on 4-decades of testing such interactions using personal interview and survey data, and, especially the experimental laboratory testing that was used under controlled experiments, all pointing to nonseparability, i.e., pointing to F&S arising jointly on some path 0Z in Figure 1.

Claude Bottomline

Claude concludes:  “The paper is genuinely helpful to DIT on one front. It shows that when one action jointly affects self and other, the brain builds a single integrated value rather than running two ledgers, using partly shared circuitry and a common valuation site, with an independently varying sacred-side sensitivity whose absence produces psychopathic learning. It should not be cited as evidence of nonseparability, because its model of integration is the additive weighted sum DIT rejects, and the design hides the joint outcomes where interaction would show. The weak altruism and the empathy nulls are real counterweights. A DIT reanalysis of their open data is, in my view, the highest-value next step this paper invites.”

So, the research must continue. The main goal is to ascertain the role of nonallocability leading to joint outcomes. And, the outcome of that research then affects how one addresses the incommensurability --- the pushpins&poetry, profane&sacred problem at the core of the economic choice problem..