One of the central arguments of John et al.'s article is that proxy failure puts in jeopardy the survival not only of all kinds of human organizations but also of organisms. Their point is intellectually seductive for its rational approach: Poorly designed proxies that are not capable of guaranteeing the accomplishment of the goals for which they were created lead social and biological systems to underperform while regulators and agents play their cards. Regulators create tighter constraints, and agents “game” the proxies created by the regulators. Eventually, an entity of inherent and superior rationality, for example the market or evolution, would get rid of these systems through a definitive process of selection.

This view of human affairs ignores the fact that rationality is not humans’ strongest suit. What leads to a questionable relation between a goal and its proxy is not, in many relevant cases, the proxy, but the goal. “Proxy failure” dynamics is probably a good conceptual approach to physically highly testable goals (e.g., the three standard tests of death as a proxy for an individual's death), or goals and proxies embedded in systems ruled by constitutive norms (e.g., a checkmate counts as triumph in the world of chess, but a “maybe” does not count as getting married in a wedding ceremony). But when the goals become more abstract (and abstraction is a salient feature of institutional and organizational goals), the goals’ “operationalization” is always far from ideal.

For this reason, goals such as liberty, justice, wealth, hygiene, safety, or even sanctity have been translated into a clearly diverse repertory of proxies. The problem is not these proxies, but rather the inherent impossibility of translating an idea into a univocal, exclusive actual norm or standard, and also the human tendency to assume that the designated norm or standard is the only and definitive proxy for that idea. When one reads (Vann, Reference Vann2003) the details of the rat massacre in colonial Hanoi with which John et al. open their article, it becomes clear that tallying dead rats was not a faulty proxy – it was just a part of a repertory of exploitative practices aimed at constructing an ideal new France in Asia. French sewers in a tropical climate zone became a perfect vivarium for rats, and the failure of the attempt to control their population growth by sending natives to the filthy, infectious sewers to bring out dead rats was not because of a faulty proxy of the goal of killing rats but the consequence of an impossible goal. And the most plausible explanation of the natives’ “gaming” of their task is not a manifestation of rational, amoral self-interest, but a way of making everyone happy amidst an impossible-to-fulfill normative framework (to install France in Vietnam). Fernández-Dols et al. (Reference Fernández-Dols, Aguilar, Campo, Vallacher, Janowsky, Rabbia and Lerner2010) showed that the apparent hypocrisy of participants in an experiment in which they could “game” the experimenter about the outcome of flipping a coin (thus earning some underserved money) was a consequence of the participants’ perception of the experimenter's instructions as an arbitrary imposition; “gaming” did not happen when the instructions were clear and procedurally fair.

But even in examples such as that of colonial Hanoi, the social system does not necessarily collapse; and if it collapses is not necessarily because of the ways in which some of its core goals were translated into proxies, but typically because of the difficulties of providing an increasingly complex social system (a transplanted France) with the amount of energy needed for its survival (e.g., Tainter, Reference Tainter1988).

Why do systems not collapse even if the activities designed as proxies for some of its core goals are not working? The answer, for a social psychologist, is simple: Because we can create an illusion of accomplishment (Festinger, Riecken, & Schachter, Reference Festinger, Riecken and Schachter1956), and the social system can last for years, decades, or even centuries maintained by these illusions (Fernández-Dols, Reference Fernández-Dols, Ross and Miller2002). Such illusions can be protected by resorting to force or fraud, but they are much more effective when the level of abstraction of the goal itself forces the system to dictate arbitrary proxies that are cognitively legitimated as a moral or religious belief, or as a secular tradition. The more abstract these illusory core goals are, the better.

A prototypical example is religion. A recurrent theme in many contemporary movies about the Mafia is the criminals’ honest respect of Catholic rites (such as the equally honest respect of Muslim rites by jihadists). This blatant failure of the faithful practice of religious rituals to attain the sanctity of their practitioners could be considered by John et al. as an ultimate, extreme example of failed proxies. We do not need to watch mafia movies to know that respecting religious rites and precepts does not guarantee sanctity. But churches last for thousands of years and do not seem to be at risk of disappearing in the foreseeable future.

Space limitations prevent me exploring examples of other social structures that last for long historical periods despite the obvious flaws in their concrete translations of abstract goals. For example, the stock market as a translation of Adam Smith's invisible hand (Shiller, Reference Shiller2000), or contemporary democracies as a translation of an ideal of tolerance (Levitsky & Ziblatt, Reference Levitsky and Ziblatt2018).

The problem is not, in many cases, the quality of the proxy as a valid construct of a goal, but the goal itself. My point is that goals are frequently resistant, for different reasons, to any accurate representation through a valid proxy. And, most importantly, untranslatable goals, and their corresponding imperfect proxies, do not compromise the survival of the social systems that legitimate these goal–proxy tandems. In some cases, these loose tandems are a guarantee of the longevity of the system itself, because imperfect proxies allow people to survive goals that otherwise would be too demanding or even inhuman, as unfortunately many social large-scale political experiments have illustrated.