This article is from
Journal of Creation 37(1):104–111, April 2023

Browse our latest digital issue Subscribe

What’s wrong with being wrong: part 4a—a more than cursory look into evolutionary ethics

by Marc Kay

Reciprocal altruism

Image: @TriversRobert, Twitter.com

Hamilton had argued altruistic acts could be explained since the actors involved shared a close genetic relationship, and natural selection would favourably work to perpetuate these genes. Robert Trivers (figure 1) criticized models like this because they were “designed to take the altruism out of altruism”.¹ Yet despite this misgiving, Trivers extended Hamilton’s idea by proposing that “under certain conditions natural selection favors these altruistic behaviours [directed toward distantly-related and non-related members] because in the long run they benefit the organism performing them.”² From this sophistry grew the now mainstream idea of reciprocal altruism.

Trivers famously posed a thought experiment involving a risk-cost-benefit calculus centred upon a drowning man and his rescuer. It was based upon the following propositions:

1. If no one attempts rescue, the man will almost certainly drown.
2. A rescuer, on the other hand, will almost certainly never drown.
3. In the unlikely event that a rescuer does drown, the victim will also.
4. In the very much more probable event of the rescuer surviving, the drowning man will too.

Trivers argued that if this were a one-off event the drowning man would never be rescued as the risk would exceed any future return. On the assumption, however, that rescuers may be exposed to drowning, the pay-off is potentially enormous:

“If we assume that the entire population is sooner or later exposed to the same risk of drowning, the two individuals who risk their lives to save each other will be selected over those who face drowning on their own. Note that the benefits of reciprocity depend on the unequal cost/benefit ratio of the altruistic act, that is, the benefit of the altruistic act to the recipient is greater than the cost of the act to the performer, cost and benefit being defined here as the increase or decrease in chances of the relevant alleles propagating themselves in the population.”³

Trivers’ thought experiment is designed to reduce complex behaviour into more manageable elements. It’s not that there is an immediate reciprocated act, but that in the long-term future, benefits may accrue to those who act ‘altruistically’ to those in need. It’s to this end that Trivers’ theory has almost universally been compared with the Good Samaritan parable, and approvingly regarded by Daniel Dennett as “the first step towards human promise-keeping”.⁴

On the surface the altruist gains nothing, not in the short term at least. Nevertheless, the altruist takes out something like an insurance policy: he may be throwing his resources away, even risking his life, by saving another, but if he ever gets into a predicament another altruist, who has similarly been helped, may be nearby and go to his aid.

Cheats do prosper … sometimes

Such explanations consistently raise the problem of ‘cheating’,⁵ where an individual ‘parasites’ off the altruists’ good-will. The cheater believes that her gains will always outweigh any possible future recriminations for not reciprocating, these latter usually understood to be zero, hence the decision to cheat—contrary to the adage, cheaters can prosper. Given the reproductive advantage of cheating, why is the world not full of cheaters? The evolutionist raconteur has several explanations.

One is that cheats can only continue to do so until a critical mass of altruistic individuals is reached. The altruists will then withdraw their favours, creating an environment in which cheats no longer flourish. Trivers contends that:

“Selection will discriminate against the cheater if cheating has later adverse affects [sic] on his life which outweigh the benefit of not reciprocating. This may happen if the altruist responds to the cheating by curtailing all future possible altruistic gestures to this individual. Assuming that the benefits of these lost altruistic acts outweigh the costs involved in reciprocating, the cheater will be selected against relative to individuals who, because neither cheat, exchange many altruistic acts.”^3,6

Another explanation is that, in response to the cheats’ ‘bad will’, the altruists evolve an ability to defend themselves:

“Once strong positive emotions have evolved to motivate altruistic behaviour, the altruist is in a vulnerable position because cheaters will be selected to take advantage of the altruist’s positive emotions. This, in turn, sets up a selection pressure for a protective mechanism.”⁷

These monitoring shields were the result of a brain physiology and are moral emotions, such as indignation, embodied, some believe, before morality was even possible:

“… the reason the moral emotions had to come about was to keep the adaptive behaviour of altruism working in the face of the equally adaptive behaviour of cheating. Once the behaviour of altruism had entered our ancestral species’ gene pool … it was inevitable that the behaviour of cheating would also come about.”⁸

Tenaciously, and quite conveniently for the evolutionary story, cheats, recognizing that they no longer rule the roost, and are being denied benefits they formerly could easily swindle from non-cheats, may even swap sides and become altruists, of sorts. According to Trivers, some cheats may become pseudo-altruists to deceive genuine altruists all over again: “Once friendship, moralistic aggression, guilt, sympathy, and gratitude have evolved to regulate the altruistic system, selection will favour mimicking these traits in order to influence the behaviour of others to one’s own advantage.”⁹

The mystery of first appearance

In such just-so stories, it is unsurprising that reciprocal altruism becomes fixed and dominant, and cheaters are rare. However, as many have pointed out, the real problem is how the altruistic gene was not eliminated on first appearance. Trivers, in a moment of plain speaking regarding his idea’s shortcomings, nevertheless falls back onto the omnific powers of circular reasoning:

“There is no direct evidence regarding the degree of reciprocal altruism practiced during human evolution nor its genetic basis today, but given the universal and nearly daily practice of reciprocal altruism among humans today, it is reasonable to assume that it has been an important factor in recent human evolution and that the underlying emotional dispositions affecting altruistic behaviour have important genetic components.”¹⁰

E.O. Wilson admits frankly to the problem: “How critical frequencies [of an altruist gene] are attained from scratch remains unknown.”¹¹ Faced with the logically insurmountable, Wilson believes that the answer must lie out there, somewhere, and then throws into the mix all candidates:

“The machinery for bringing the gene frequency up to the critical value [over which it isn’t immediately eliminated] must lie outside the game itself. It could be genetic drift, which is entirely feasible in semiclosed societies, or a concomitant of interdemic or kin selection favoring other aspects of altruism displayed by the cooperator genotypes.”¹²

In developing their models, theoreticians take for granted the already established presence of altruism genes. At once both question-begging and tautological, the maths posits “a nonzero frequency of helpers above which selection tends to increase their frequency, and below which selection tends to decrease their frequency.”¹³ In further tautological clarification, in order for the gene to spread, the number of helpers must be greater than a ‘sociality threshold’. In other words, once an exceptionally rare (and theoretical) altruistic allele proliferates to the extent it can keep on increasing in numbers, it will continue to multiply.

Theoreticians habitually use abstruse language to mask their failure to adequately explain their case. For example, Brown et al., addressing Boorman and Levitt’s work (see below), state:

“[the authors] demonstrate the existence of a threshold value of the social allele which must be overcome for sociality to evolve. Since this threshold frequency would usually prevent takeover by a rare mutant allele, the authors suggest the ‘takeover cascade’ as a plausible mechanism for successful invasion, whereby population viscosity would allow the social allele to first establish itself locally and then slowly invade neighboring populations.”¹⁴

Suggesting a rare-as-hen’s-teeth altruistic allele could eventually conquer its own territory, and then its enemies’, by virtue of their model allowing it, is not a heuristic exercise; it’s a smoke and mirrors word feast.

As for Boorman and Levitt’s paper, the authors give life to a much needed altruistic insurrection by arguing that if an altruistic allele is a homozygous recessive (hr) and that not meeting another hr is lethal, then one hr linking up with another will be a fitness-enhancing event because cooperation will necessarily take place.¹⁵ Given the fundamental rule of deductive logic that there is nothing in the conclusion that isn’t already contained in the premises, Boorman and Levitt’s model doesn’t appear to challenge that modest truth.

Suggested models unfailingly rely on a series of unreasonable assumptions designed to break through this imagined threshold. For example, populations are of infinite size,¹⁶ have the population subdivided,¹⁵ have interactions between individuals occurring on a non-random basis, meaning an unrealistically high frequency of helpers aiding other helpers,¹⁷ have near-perfect memory of encounters with cheaters,¹⁸ or ensure altruists rarely interact with strangers.¹⁹ It would seem that such models hardly reflect real-life situations in which people are more likely to be members of large populations where interactions are entirely random and little is known about the other, particularly in one-off meetings.

Occasionally arguments are put forward containing an already conveniently propitious disposition. Peck and Feldman acknowledge that previous models involving large, randomly mixed populations cannot raise the level of helpers from the initial low frequencies. They concede that their “intention in presenting the model is not to suggest that the performance of helping behavior should be a ubiquitous phenomenon in natural populations, but to identify a set of biologically reasonable circumstances which, when they occur, favour the evolution of helping behaviour.”²⁰

Reciprocal altruism isn’t really

Let me restate the obvious. Trivers’ altruism bears no resemblance to ordinarily understood altruism in which expectations of return are absent from any interaction with another. Reciprocal altruism is merely a quid pro quo contract, and this species demands a return for its investment. Giving without receiving makes no sense in an evolutionary world where the highest ideal is the perpetuation of one’s gene pool.²¹

Dennett, however, argues that objections to this entirely miss the point. Reciprocal altruism, rather than being the full-blown ‘left-hand-not-knowing-what-the-right-is-doing’ altruism, is a transitional phenomenon that serves as “a useful stepping-stone on the progression” to the real deal (or as he calls it, the ‘real McCoy’).⁴ Not realizing his self created chicken-or-egg problem, Dennett dodges the problem of self-serving behaviour becoming its opposite. Critics have underscored the unbridgeable difference between the two. In his response, Dennett argues that once a creature acquires “advanced cognitive abilities—a rather specific memory capable of reidentifying one’s debtors and creditors, and the capacity to spot a cheat”,²² the difficulty just vanishes.

Yet, without so much as a hint of shame, Marc Hauser sums this up:

“When we lack kinship, we act nicely if we have some guarantee of a reciprocated return. This isn’t an act of kindness. Reciprocity is an act of self-interest, because it is driven by the expectation of a fair return: food for food, backrub for backrub, babysitting for babysitting. From the gene’s-eye view, the way to think about the evolution of moral behaviour is to think selfishly. Instead of asking ‘How can I help you?’ ask ‘How can my helping you help me?’ In the simplest case, you would compare two strategies, moral versus immoral, and tote up the number of babies for each. If the moral strategy wins, both in terms of reproductive output and in terms of its immunity to immoral invaders, then selection will favour the moralist and eliminate the immoralist. Life isn’t that simple, but the logic of the argument is.”²³

A world according to Hauser, if true, is a world which few would prefer to live in, and Michael Ghiselin’s now infamous comment “Scratch an ‘altruist’ and watch a ‘hypocrite’ bleed”²⁴ would arguably sum up reality more frighteningly and accurately.

And this dystopian vision is that of two famous sociobiologists. Repudiating a realist moral objectivity, Ruse and Wilson draw upon reciprocity, but a nuanced ‘genteel’ one:

“… the expectation [of reciprocity] does not necessarily employ a crude demand for returns, which would be antithetical to true morality. Rather, I expect you (or God) to help me because it is right for you (or God) to help me, just as it was right for me to help you (or obey God). The reciprocation occurs in the name of morality.”²⁵

But theirs is a crude caricature of genuine morality. Both ‘demand’ and ‘expect’ have the same ontological function. Without an external moral objective base, something both deny, the quid pro quo transactions based on these being ‘right’ is nothing but a non-cognitivist²⁶ prescriptive meta-ethic wanting something, eventually, in return for being ‘good’.

Winning is not the most important thing; it’s how you play the game

So how could kindness for kindness’s sake, real altruism, take a first step? In the late 1970s Axelrod²⁷ set up a computer tournament and asked people to design programs which represented the way creatures related to each other. Tit for Tat, the most successful, operated on an antithetical Golden Rule principle of being initially nice and then subsequently mirroring the response of the other. Wright, explaining Tit for Tat, describes it as “do[ing] unto others as they’ve done unto you”.²⁸ Consequently, if the other returned nastiness, you reacted identically; if you received a kind approach, you returned the favour. The game was lauded as a demonstration of how a population of ‘altruists’ couldn’t be invaded²⁹ and thus the population could continue to operate as an evolutionarily stable strategy (ESS).³⁰ Selfish cheats were excluded, and gracious bodies were brought in by cooperating until the cheats had to change to survive. The significance of an ESS was that it could, so it was suggested, explain cooperative behaviour without calling upon higher levels of selection.

Of course, the resemblance to altruism, the initial kindness, is hardly convincing—it was already present from the beginning as a result of being introduced as part of the program. From a computer program, Janet Richards makes the leap to full-blown altruism by claiming, “the details are not essential to our purpose here, which is simply to answer the claim that evolution cannot produce unselfishness.”³¹

Faced with a complete disjunct between the indulgences of the theoretical and the clarity of the real, some opt for the optimism of fideism. Dennett well understands the complexity of simulating our ancestral struggles toward morality. The problems are exceedingly compound, yet he nevertheless encourages us to believe there are evolutionary solutions because “We may be quite sure that [the existence of such an evolutionary explanation] is true, for here we are.”³²

Kitcher likewise ignores important issues such as the origin of the underlying genetics and the unlikelihood of two similar altruistic alleles arising at the same time: “provided that we have a pair of DA [discriminating altruist] mutants, discriminating altruism can invade a population of defectors.”³³ DAs are not pure altruists but will interact with someone as long as they haven’t previously defected on a DA. In other words, they are unforgiving. Kitcher’s elaborate equations, contrived to calculate payoff values, strain to obtain traction in explaining genuine altruism or what he calls ‘hyperaltruism’. The best is an admission that “we seem to have shown that selection will favour ‘golden-rule’ altruism of a discriminating kind (treating the other as oneself so long one has no basis for thinking that the other will not do the same).”³⁴ Isn’t this nothing more than a well-weighed you-scratch-my-back-if-I-scratch-yours interaction? In a disingenuous admission of defeat for his reductionist maths, Kitcher limits the real-deal altruism, what he calls ‘the more subtle forms’, to that “portrayed by historians and writers of fiction”.³⁵

Smith draws attention to problems in the simple representation of game theory³⁶. He concedes that strategies played out with only two genetic polymorphisms are far more manageable and analyzable than ones with more than two. His modelling assumes asexual inheritance and, if reformatted for sexual reproduction, depends on an ESS being produced through a homozygous genotype. More complications arise in asymmetric contests where participants possess unequal characteristics, such as differences in sex, size, and age. Also, where there are more than two engaged at any one time this increases the complexity of strategy determination.³⁷

Other game theorists have grappled with the difficulty of altruism’s spread among unrelated individuals. If you knew beforehand how another would relate to you, then you would know if an alliance would enhance your fitness or reduce it. Trying to predetermine the strategy of the other player is termed ‘the problem of common knowledge’ in which the action the other chooses depends on what she thinks you’ll do, which depends on what you think she’ll do … and so on. Calculating the possibility of someone’s cheating or honesty, on this infinite regress scenario, means there is no best strategy, and if so, morality would seem that much more unlikely to have arisen. But, of course, we do demonstrate other-concerning action and trust toward non-kin, and so theorists came up with the Prisoner’s Dilemma (figure 3).

Image: Composite of images by Ron Lach, Pexels.com

Rats always win … well, almost always

Imagine the following scenario. You are arrested with a friend and accused of a serious crime. The arresting authorities have less than certain evidence but are relying on one of you ratting the other out. Separated, neither you nor your friend have any idea how the other will respond when interrogated. If both of you keep silent, each will receive a moderate sentence; if both finger each other, slightly more; and if one admits nothing while the other says the first committed the crime, the silent one gets hit with the severest of sentences while the rat gets off scot-free. What’s the best course of action?

From any one individual’s perspective the best outcome is to cheat and blame the other. As a pair, however, it’s to cooperate with each other and not confess in order to receive the lightest of sentences. Of course, one cannot be guaranteed that the other will not cheat because one cannot know, hence the dilemma. Thus, no matter what the other does, in terms of pay-off, the best strategy is to cheat and to pursue your own self-interest, at least for one-off interactions.³⁸

Single encounters raised a significant problem. The maths meant it would be impossible for cooperation and altruism to arise if meetings were single, period, because the strategy of always defecting would produce an uninvadable population utterly resistant to a good guy’s charms. In other words, the cheat will never transform to an altruist because he quickly moves on to swindle yet another sucker. To remove this obstacle, evolutionary strategists claimed, perhaps realistically, that a considerable percentage of our interactions are not one-off but multiple, and often with the same person. Once the assumption that the number of games is not fixed is removed, thus generating the idea that there is a reasonable probability there will always be another game, an ESS will arise from the application of a Tit-for-Tat response.³⁹

Prisoner Dilemma-styled scenarios have not escaped criticism, particularly when they’re brought in as a basis to understand the real-life problem of morality’s rise. Shortcomings include the mandatory concomitant development of players possessing the necessary machinery and genetics to store and process memory of participants and outcomes,⁴⁰ learning capacities, the problem of the co-evolution of the communicative abilities between sender and receiver, the unpredictability of winners, and the crudeness of measures of behaviour “with no regard for subtle gradations, intervals between events, repetitions of events, proximity of interactants and their orientations, or context.”⁴¹ Barlow and Rowell concluded their paper with the following candid comment: “The bottom line is whether the game theoretical approach has led to useful new predictions, to unanticipated hypotheses to be tested. Our assessment is that so far it has not.”⁴²

Others have discovered that limited environmental cues, the inability to process information in a consistent way, information overload and fatigue can bring about mistrust and thus affect decisions even in initial interactions. Einhorn further doubted whether the usual explanation for the first cooperative response in Tit-for-Tat, one motivated from payoff calculations, is always adequate. After all, according to the devotees of Tit-for-Tat, testing the waters by being nice on the first encounter is a dominant strategy. Dominance occurs when a participant’s response is the best tactic for that player no matter how the other plays. However, Einhorn noted the possibility of the unintentional reaping undeserved rewards: “It would seem that one also needs to include a probability of choosing the cooperative response of error.”⁴³ He was also critical of the superficiality of evolutionary game theory: “the major limitation of game theory in describing the evolution of behaviour lies in its static formulation of the process.”⁴⁴

Johnston pointed out that evolutionary game theory addressed the interaction phase to the exclusion of how the variations arose in the first place. He showed how theorists create technical jargon to circumvent problems of origins. Installing the appropriate genetic variability from the beginning is easy when terms like ‘mutant strategy’ or ‘incipient variations’ are buried within an explanation.⁴⁵

Some don’t bother concealing their weighted conditions. One paper, without warrant, includes a genetic predisposition to indiscriminately assist. This theoretical structure then immediately ‘learns’ to discriminate against non-reciprocators. This skill comes not from a single allele but “generalized traits such as enhanced memory”, and as a result reciprocal altruism becomes widespread.⁴⁶ Similarly, Hamilton and Axelrod, following a highly phrenetic theoretical brief about variations of games, postulate: “Once the genes for cooperation exist, selection will promote strategies that base cooperative behaviour on cues in the environment.”⁴⁷ In combination with a dazzling series of matrices, equations and application of game theory, they assign ‘arbitrarily low initial frequencies’ of helpers that increase in numbers, making no mention of just how the ‘helpers’ came into existence. Apparently, they see nothing prejudicial about including several good Samaritans ab initio.

Furthermore, a population of egoists is an evolutionarily stable strategy and can never be invaded by a co-operative one, even a Tit-for-Tat one. Some games actually do produce better yields for selfish participants. One such experiment “highlight[ed] the possibility that the selfish behaviour of psychopaths can lead to greater short-term rewards than prosocial behaviour [and if] psychopaths notice this advantage, this very experience may act as a reinforcement of antisocial behaviour.”⁴⁸ Extending the question-begging further, to get reciprocal altruism over the theoretical hurdle, interactions between individuals must occur repeatedly, a faculty for discriminating against cheaters and recognizing conspecifics must be up and functioning, as well as a mechanism to suppress or avoid the choice of accepting aid but not returning it. Still, this form of altruism promoted by its adherents bears no resemblance to the biblical understanding of good will to your neighbour or, indeed, to your enemy. Co-operative strategies such as Prisoner Dilemmas and Tit-for-Tat are, as O’Hear remarks, “very far from genuine altruism, or from morality at all … [and] are clearly for the mutual self-interest of the parties involved.”⁴⁹

 Image: Muhammad Mahdi Karim, Wikimedia / GFDL 1.2

Let me recap. From one perspective, game theory is intricate and well-honed. From another, however, it only successfully satisfies the evolutionary mind due to a smoke-and-mirrors strategy of question begging into existence the parodies of altruism. Dennett’s circularity of argument is worth repeating: interaction strategies of these computer simulations of morality’s development more or less accurately reflect evolutionary history and present reality because “here we are [today]”.³²

Conclusion

Evolutionists faced a counterfactual: how can neo-Darwinism, the aim of which is to place more of one’s genetic material into one’s flesh-and-blood posterity, account for acts of kindness which extend far beyond consanguinity? Such good deeds were counterintuitive to evolutionary expectations and yet Good Samaritan acts do occur.

Trivers, however, raised the possibility of unconscious quid pro quo acts through his reciprocal altruism explanation. He imagined that a good turn by an actor ‘deserved’ to be returned by its recipient to some needy other at a later date and place. This altruistic mentality would spread throughout the population and become, ultimately, a potential pay-off for the initial cost of the actor’s assisting the recipient as well as a future corporate benefit to the larger community.

The problem of the parasitical cheater was exposed, and the best evolution proponents could offer was that the rogue, if he didn’t reform his swindling ways, would be frozen out of ‘polite’ society and left to defend for himself.

Notwithstanding these ‘cavils’, the most vexing hurdle remained: where is the original allele(s) that directs an individual to rise above selfishness and by what process did it come into existence?

Desperate to overcome the mounting counterarguments, game theory was exploited to vindicate, at least in theory, how individual acts of altruism could develop and spread. The relevance of this to real life scenarios was thin at best.

Once again creationists have nothing to be troubled about. Evolution’s materialist quiddity continues its (failed) battle to unseat ethics’ essence; namely, it’s very much non-natural ontology.

In the next part, I will continue the analysis of reciprocal altruism and its almost complete reliance on animal studies. Evolutionists appreciate it is far too problematic for morality to have arisen in humans de novo. Consequently, it is expected animals possess a kind of proto-morality.

Related Media

15 Questions for Evolutionists -- #11 How did blind chemistry create mind, meaning, morality?