Intentional Germline Modification & Evolution
Featured guest article by Dr Ben Griffin
Germline genetic modification—changing the DNA in eggs, sperm, or embryos so traits get passed down—comes with big promises and even bigger risks. On the one hand, it could eliminate genetic diseases and even enhance traits like intelligence or strength. On the other, we don’t fully understand how genes interact, and unintended mutations could have serious consequences that only show up generations later. The fear is that by trying to improve humanity, we might accidentally introduce new problems we can’t undo. Should we ban it outright because of these risks? Some argue yes, since the unknown dangers are too great. Others believe strict regulation, rather than prohibition, is the better path—allowing science to move forward while keeping safety in check. The real question is whether the potential benefits outweigh the risks of playing with evolution itself.
This debate is crucial for the future because it sits at the crossroads of science, ethics, and technology— and AI. As AI advances, it will likely play a key role in genetic modification by helping us analyse vast amounts of genetic data, predict the effects of edits, and even automate parts of the process. If we move forward with germline modification, AI could make it safer and more precise, reducing the risks of unintended consequences. On the other hand, AI might accelerate genetic engineering before we fully understand its long-term effects, pushing humanity into uncharted territory.
An Evolutionary Perspective is Essential to Answering “Should Enhancement Through the Intentional Genetic Modification of Germline Cells be Prohibited on the Grounds of Unintended Bad Consequences?”
Foreword
Over the last two years, I was fortunate enough to complete a MA in philosophy at Kings College London. Whilst particularly focusing on the philosophy of medicine and psychiatry – reflecting my clinical background – across the MA I had the opportunity to delve into a wide range of topics across philosophy. Concurrently, I developed a keen interest in evolutionary psychiatry. Introduced to the field around eighteen-months ago, I learnt that evolutionary psychiatrists hold the view that the application of evolutionary principles to understand what mental disorders are, how they are caused, and how they can be prevented and treated, should be a fundamental in psychiatry. Thanks to excellent books such as Good Reasons for Bad Feelings by Nesse, and Evolutionary Psychiatry: Current Perspectives on Evolution and Mental Health by Abed and St John Smith, as well as engaging with the evolutionary psychiatry special interest group (EPSIG) at the Royal College of Psychiatrists, I am now a firm believer in the fundamental importance of an evolutionary perspective to psychiatry.
However, as my MA progressed and my appreciation of evolutionary psychiatry grew, I increasingly realised that the importance of evolutionary principles extends far beyond just better understanding mental disorders. I came to view evolutionary theory as a profoundly important and versatile (although vastly underutilised!) epistemological resource, which when judiciously applied can bring critical new insights to a wide range of topics across philosophy. With this perspective in mind, I was able to produce what my assessor regarded as the best pieces of work of my MA.
What follows here is an essay I wrote for the ‘Topics in Applied Ethics and Politics’ module I completed during the second year of the MA, which focused on the theme of biomedical enhancement, and showcases how evolutionary principles can be used as a power epistemological resource for philosophers. Unlike biomedical treatments, which aims to restore normal functioning in individuals with medical conditions, biomedical enhancement is the use of biological and medical technologies to improve the performance, appearance, or capabilities of healthy people. The essay focusses on perhaps the most controversial form of biomedical enhancement; the intentional genetic modifications of germline cells (such as sperm, eggs, and embryos) to attempt to enhance healthy individuals. Proponents argue we should use novel biomedical technologies to make permanent, heritable changes to human DNA to attempt to enhance our natural capabilities. Examples discussed in the literature have included enhanced resistant to disease, improved intelligence, and improved physical strength and endurance.
I attempt to challenge the views of one of the leading authors in this area, Buchanan, who uses evolutionary principles to argue in favour of this form of biomedical enhancement. A lecture in which Buchanan argues in detail for his position can be found on YouTube, and may be of interest to some readers:
Introduction
The intentional genetic modification (IGM) of germline cells (GCs) – sperm, eggs, and embryos – to enhance the capacities of healthy humans is widely considered to be the most controversial form of biomedical enhancement. The recent discovery of the gene-editing technology CRISPR/Cas9 has opened up unprecedented possibilities in this area1, bringing this debate into acute focus. In Beyond Humanity?2, Buchanan provides one of the most comprehensive cases in favour of the IGM of GCs found across the biomedical enhancement literature. Here, I will critique his position, arguing that the IGM of GCs should be at least temporarily prohibited on the grounds of unintended bad consequences.
Although I reject his conclusions regarding enhancement through the IGM of GCs, it is important to acknowledge that Buchanan’s broader discussion of enhancement has a number of merits. He identifies several key limitations endemic in the existing anti-enhancement literature:
1) Ignoring evolutionary biology and genetics: the tendency for authors to resort to ‘a pre-Darwinian view of Nature and our place in it’ (Buchanan, 2014, pp. 8)2.
2) Unclear argumentation and conclusions: ‘prominent figures in the debate persistently substitute high-sounding rhetoric for reasoning’ (Buchanan, 2014, pp. 2)2 ‘so… that it makes it very hard to tell exactly what they are arguing for or against’ (Buchanan, 2014, pp. 10)2.
3) Armchair theorising to sweeping conclusions: ‘the enhancement debate is… often breathtakingly naïve, from a methodological point of view… there often seems to be no awareness of the need for empirical evidence’ (Buchanan, 2014, pp. 8)2. There is a need to ‘resist the tendency towards sweeping condemnation or praise, and, above all, to start thinking hard about practical responses that… are true to the complexity of the phenomena, and realistic’ (Buchanan, 2014, pp. 12)2.
These astute points have significantly informed my methodology: I will proceed on the following basis. First, a discussion of evolutionary biology and genetic technologies will be integral to my thesis. As Buchanan argues, the view that one takes on enhancement through the IGM of GCs should be significantly informed by an account of what he terms unintentional genetic modification (UGM); the status quo, by which germline modifications occurs through evolutionary processes without deliberate human intervention. As shall be elaborated upon, how one understands UGM, and how it compares to the IGM of GCs, becomes a crucial determinant of the view one takes in this debate. Second, the discussion will be limited to arguing against pursuing enhancement through the IGM of GCs on the basis it is a uniquely risky type of biomedical enhancement. I do not seek to argue against the entire enhancement enterprise. Finally, I shall briefly outline a practical response in line with my assessment of the issue.
Buchanan’s Argument for the IGM of GCs is Comparative
It is key to appreciate Buchanan’s argument operates through comparison. His case consists of a weighing up of the risks and benefits associated with the IGM of GCs, against those of continuing with UGM. This accordingly entails a bleak assessment of UGM, juxtaposed against an optimistic account of the capabilities that IGM gives us to circumvent these limitations. This sets Buchanan up to argue that there is a significant likelihood that the goods resulting from the IGM of GCs will outweigh unintended bad consequences.
I shall attempt to show that both sides of his discussion are fallacious. First, I will argue that whilst he is correct that UGM is significantly constrained, Buchanan’s assessment of it is unduly negative. Despite admonishing anti-enhancement authors for relying upon misleading analogies, here Buchanan does the same himself. A discussion of adaptationism will be integral to making this point. Next, I will argue that by his own criteria, Buchanan’s claim that IGM will allow us to overcome many of the limitations of UGM is weak. He advances a largely theoretical argument which cherry picks concepts from evolutionary biology which support his case whilst ignoring those which do not, and fails to give consideration to the current practical limitations of CRIPSR/Cas9, the technology which would be used to conduct the IGM of GCs. Considering these gives us significant reason to think the IGM of GCs constitutes a uniquely risky form of enhancement, fraught with the risk of unintended bad consequences, which we should be opposed to.
Buchanan’s Assessment of the UGM of GCs: Undue Pessimism
To clear the way for his account, Buchanan seeks to dispel the common anti-enhancement ‘Master Engineer’ analogy for UGM as grossly inaccurate. Instead, evolution is cast as a ‘morally blind, fickle, tightly-shackled tinkerer’ (Buchanan, 2014, pp. 192)2. Evolution is said to be morally blind as human well-being is not the goal, and the process of maximising fitness may be achieved by any means necessary, even if suffering and death are incurred in the process. The only standard by which evolution selects for is maximising fitness; how likely a particular genotype is to leave offspring in the next generation, relative to other genotypes. The fickleness of evolution describes that it does not look ahead, but responds ad hoc to current ever-changing environmental demands. This is illustrated by the Red Queen Hypothesis: rather than moving towards an end point of optimality, species are constantly adapting just to cling on to survival as the selective environment perpetually changes beneath them3. Evolution is said to be a tightly shackled tinkerer as it is only capable of slow and incremental change, which is also subject to path dependency. Path dependency denotes that the development of traits across evolutionary time necessarily builds upon what has gone before, such that potential adaptions to new environments are constrained by what can be derived from organism's existing features. This leads to ‘the ubiquity of sub-optimal design’ (Buchanan, 2014, pp. 184)2.
This is illustrated by the Red Queen Hypothesis: rather than moving towards an end point of optimality, species are constantly adapting just to cling on to survival as the selective environment perpetually changes beneath them3
Buchanan states the purpose of presenting UGM in this way is ‘debunking the master engineer analogy’ (Buchanan, 2014, pp. 193)2, and acknowledges the limitations of analogies. Perhaps by his own admission then, the analogy he presents also functions as a rhetorical device, which seeks to paint a bleak picture of UGM. Like the master engineer analogy, this too is at the expense of accuracy. Whilst it is true that evolutionary processes do not intend to pursue moral ends – or any ends for that matter – it does not follow from this that evolution cannot produce things which we morally value4.
Buchanan may fairly respond that he is not committed to this claim. However, his presentation of this issue is equivocal. He grants that ‘if we plot the desirability of traits against their contribution to inclusive fitness, there may be significant overlap between the curves’ (Buchanan, 2004, pp. 192)2, before later claiming that ‘when UGM… produces what is valuable to human beings, it does so by sheer coincidence’ (Buchanan, 2004, pp. 192)2. It seems highly unlikely that the significant overlap Buchanan acknowledges is the result of sheer coincidence. Instead, it can be made sense of through adaptationism.
Adaptationism gives us reason to be significantly more optimistic about UGM that Buchanan allows for, denoting a family of views with the central tenet that the physical and psychological traits of organisms can be understood as evolved adaptations5. Natural selection is accordingly prioritised as the most important process at work in shaping the evolution of organisms. On an adaptationist view, we value the things we do because doing so has increased our fitness. Given this, contrary to Buchanan’s claim, it is no coincidence UGM can produce things we value as what we value has been selected for through UGM. A straightforward example of this is the ubiquitous human preference for life over death: we value life and disvalue death because doing so has increased our fitness.
Whilst there is abundant evidence that many features of the organisms found across the extraordinary breath of the natural world have been shaped by natural selection for adaptive functionality6,7, it should be noted that adaptationism has been subject to criticisms. In my view, these concerns should be understood as important caveats to an adaptationist view, but fall far short of arguments to completely dismiss adaptationism of any kind. A common accusation is that adaptationist accounts fail to recognise the importance of explanatory pluralism5. Attempting to reduce explanations for all characteristics of organisms to purported adaptive functions ignores the importance of constraints on selection, that traits are subject to trade-offs, and the role of other evolutionary processes such as genetic drift and frequency dependent selection. It is also essential to appreciate the constraining factor of adaptive lag. As captured in the Red Queen Hypothesis, environments are in perpetual flux so organisms can never become perfectly adapted to them. Particularly at times of more rapid environmental change, traits are at considerable risk of becoming significantly mismatched to their environment, with some which were adaptive in past environments being rendered neutral or maladaptive. Given the considerable changes in human environments since the Neolithic Revolution around 10,000 years ago – a tiny span of time in evolutionary terms – the concept of evolutionary mismatch is increasingly appreciated as important in understanding many of the diseases which afflict modern human populations8.
Given the considerable changes in human environments since the Neolithic Revolution around 10,000 years ago – a tiny span of time in evolutionary terms – the concept of evolutionary mismatch is increasingly appreciated as important in understanding many of the diseases which afflict modern human populations8.
This said, it is important not to commit the fallacy of applying mismatch to entire organisms. It is only specific traits which can said to be mismatched; many integral anatomical and physiological features of organisms essential for life necessarily continue to develop and function largely as they have done for millions of years. There are innumerable awe-inspiring examples of this when we consider our own bodies: from the steadfast output of our hearts, beating approximately 2.5 billion times in an average human lifetime, to the incredible complexity of our immune systems, locked in an ancient arms race with the ever-changing pathogens constantly attempting to gain footholds in our bodies. Whilst such systems are of course never optimal in the sense of being perfect, it seems difficult to deny they are not designed for clear adaptive functions. Moreover, it is increasingly recognised that given environmental flux has been a constant throughout evolutionary history, the potential for phenotypic plasticity – defined as the ability of individual genotypes to produce different adaptive phenotypes when exposed to different environmental conditions – has been significantly selected for9. So, whilst mismatch is undoubtedly an important concept, one should not draw from it the conclusion that modern environments have rendered all human phenotypic traits maladaptive, or that evolution has rendered phenotypic outcomes so rigid that environmental change necessarily always causes significant dysfunction.
In sum, although it is important to recognise some limitations, adaptationism contains an incredibly important kernel of truth in understanding living organisms. Distinguishing between strong and weak adaptationism can prevent the baby from being thrown out with the bathwater10. Strong adaptationism is the view that organisms are optimally designed, owing to natural selection alone. Weak adaptationism meanwhile asserts that organisms have generally been designed to maximise their fitness without implying they are optimally designed, accordingly making room for other evolutionary processes aside from natural selection. Whilst strong adaptationism cannot accommodate the caveats outlined here and hence is not a tenable view, weak adaptationism can acknowledge these as important cautions against the explanatory overreach strong adaptationism is guilty of, whilst maintaining the importance of natural selection and adaptive accounts of traits.
All of this this said, Buchanan might still insist that UGM is subject to major restraints. However, even if I have only succeeded in partially weakening the strength of his claim about UGM, this is of importance. The apparent favourability of IGM is created through its comparison with his unduly negative assessment of UGM. It is through strongly asserting the limitations of UGM that Buchanan sets himself up to argue there is an overwhelming mandate to use the IGM of GCs to enhance normal human capabilities. Here, I have attempted to provide a more nuanced account of the UGM of GCs by avoiding tenuous analogies, and addressed the ways he overstates its shortcomings. The weak adaptationist position I have presented here, and crucially the implications it has for Buchanan’s arguments, at the very least pushes us towards a more nuanced understanding of UGM than he allows for himself.
Buchanan’s Assessment of the IGM of GCs: Undue Optimism
The most misleading aspect of Buchanan’s account comes from his overly optimistic assessment of IGM: that IGM will be free us from many of the limitations under which UGM operates. He enthusiastically claims ‘IGM has the potential to go well beyond trait enhancement to the wholesale transformation of biological organisation’ (Buchanan, 2014, pp. 190)2 and expounds a long list of benefits that stand to be reaped. Particularly given his own demand for empirical evidence when sweeping empirical claims are being made, there are significant reasons to be highly sceptical of this as a possibility in the foreseeable future. Indeed, it seems the only way Buchanan can make such a naively optimistic claim is by failing to engage in any practical discussion of the specific technology which would be used to conduct the IGM of GCs, CRISPR/Cas9.
Since its discovery in 2012, CRISPR/Cas9 has created enormous amounts of excitement. Functioning like a pair of “genetic scissors” the technology, in principle, allows the genetic code to be cut at specific locations to allow specific sequences of DNA to be added or removed1. Faster, cheaper, more accurate, and more efficient than pre-existing gene editing technologies, CRISPR/Cas9 has opened up unprecedented possibilities in gene editing. However, major unresolved limitations associated with CRISPR/Cas 9 have largely prevented the hype surrounding it from being realised into novel genetic therapeutics, let alone viable human enhancements. Off-targeting describes unintended editing event(s) occurring at untargeted sites in the genome that are genetically similar to the targeted site11, to produce unpredictable and invariably negative phenotypic consequences. Immunotoxicity – where the body generates a potentially life-threatening immune response following administration – represents a further unresolved significant problem12. Whilst these technical issues are anticipated by many working in this area to be overcome in the coming years, for the time being they constitute significant practical pitfalls that must be considered in discussion about the IGM of GCs for enhancement.
In his discussion of the IGM of GCs, Buchanan’s failure to adhere to his own methodological recommendation of avoiding armchair theorising to sweeping conclusions goes hand in hand with a dubious theoretical discussion of the relevant evolutionary and genetic theories. He appears to cherry-pick concepts – namely modularity, canalisation, and redundancy – which support his argument whilst ignoring more fundamental theoretical aspects which all suggest that the IGM of GCs is going to be far more fraught with the risk of unintended bad biological consequences than he acknowledges. Indeed, further to the limitations detailed above, the highly complex genotype-phenotype relationship has already been recognised to significantly complicate IGM through CRISPR/Cas913. I briefly present four key complicating processes here to explicate this point.
Pleiotropy describes that almost all single genetic loci influence multiple, seemingly unrelated, phenotypic traits14. Marfan syndrome provides an example of pleiotropy: a mutation in a single gene causes a wide range of features including a characteristic body habitus, joint problems, eye problems, and an increased risk of heart disease15. A related concept is polygenetic inheritance, describing that the overwhelming majority of traits are coded for by multiple genes16. For example, height is contributed to by many genes across the genome rather than being determined by a single gene17. Further complicating matters is epistasis, which describes when the expression of one gene is modified by the expression of other(s)18; so, one gene can lead to different phenotypes depending upon the genes it is expressed alongside. Epistasis is at play in red-haired individuals: one gene overrides another to prevent the pigment associated with red hair, pheomelanin, from turning into eumelanin19. Finally, epigenetics means that environment significantly effects how genotype is realised as phenotype, by altering how the genetic code is expressed20. In the last several decades, a growing appreciation of the importance of epigenetic processes in determining how a given genotype can lead to a hugely different array of phenotypes has collapsed the traditional nature-nurture debate and decisively undermined crude genetic determinism. For example, it has been demonstrated that certain genes are associated with an increased risk of depression only if the individual has been exposed to trauma21.
Thinking more broadly about biomedical enhancement then, it is clear that attempting to modify the genome of GCs constitutes a uniquely risky mode of enhancement from a biological perspective. Compared with other types of biomedical enhancement, our ability to make specific beneficial phenotypic changes through editing the genes of GCs has been shown to be fraught with risks of unintended bad consequences. Even compared with the IGM of somatic cells, GC editing carries two significant additional risks. One relates to ontogenesis; changes made earlier in ontogenesis, with the editing of GCs representing the earliest point in ontogenesis possible, are much more likely to have cascading unpredictable consequences. The second is heritability. Whilst Buchanan points out that technologies exist which allow for inserted genes to be switched on and off, and furthermore it would hypothetically be possible for a given GC edit to be reversed, off-target effects of CRISPR-Cas 9 could not be.
Buchanan may argue that there are many benefits to the IGM of GCs which we stand to gain from which have been given insufficient attention in this account. However, as I have attempted to demonstrate here, notably writing 10 years after the publication of Beyond Humanity, the benefits he proclaims so confidently continue to remain largely theoretical possibilities only, unrealised due to unresolved technical limitations. Crucially, I am not seeking to make an absolutist claim that we should never pursue the IGM of GCs. The IGM of GCs may well become a vital tool in the betterment and survival of the humans species in the fullness of time. Rather, in reverence of the staggering complexity that must be overcome for the successful IGM of GCs to enhance normal human characteristics, I seek to argue that currently any suggestion that we are ready to embark on such a course is irresponsible and deeply hubristic.
I would also highlight these points against the IGM of GCs should not be interpreted as a wholesale rejection of the anti-enhancement enterprise. Other types of biomedical enhancement – for example drugs, implantable technologies, and brain-computer interfaces – are all much more likely to allow for specific phenotypic changes which could be more simply reversed, and would not be heritable. Whilst beyond the scope of this discussion it is my view, consistent with Buchanan’s overarching thesis, the cases for some other kinds of enhancements are indeed compelling. It is thus my contention that pursuing other forms of enhancement much less fraught with the risk of unintended bad consequences, whilst keeping the IGM of GCs firmly off the table until our understanding of genetics catches up with our ability to irreversibly alter genes, is the most sensible position to adopt.
A Practical Response to the Risk of Unintended Bad Consequences
At present, the legal frameworks for the IGM of GCs varies substantially across different countries22. Whilst the general trend has been towards blocking the IGM of GCs, grey areas exist. For example, in the US whilst Federal Law prohibits the use of federal funds for research on human germline gene therapy, currently there is no law or regulation that bans germline gene editing conducted through private funding. This has led to self-styled ‘biohackers’ to be the first to experiment with these technologies23. Inconsistent legislation across different countries also creates a risk of ‘ethics dumping’24, whereby the IGM of GCs could be exported from more regulated higher-income to less regulated lower-income countries.
To mitigate the risk of unintended bad consequences, an urgent coordinated international response is required. Through international summits, treaties could commit countries to robust standardised policy banning both private and state sponsored biomedical enhancement through the IGM of GCs, for a period of time whilst further research is conducted, before the moratorium is reviewed. Whilst the suggestion of a moratorium could be criticised on libertarian grounds, I have sought to demonstrate here the extent of the unique risks associated with the IGM of UCs constitute a strong case for the prudence of such a policy intervention, until such time the risks are better understood.
To mitigate the risk of unintended bad consequences, an urgent coordinated international response is required. Through international summits, treaties could commit countries to robust standardised policy banning both private and state sponsored biomedical enhancement through the IGM of GCs, for a period of time whilst further research is conducted, before the moratorium is reviewed.
Conclusion
Whist uncertainty remains about what forms of biomedical enhancement will be possible through the IGM of GCs, this uncertainty forms the crux of the case for an international moratorium prohibiting it. Through considering core concepts from evolutionary biology and genetics which are not given due attention by Buchanan, I have attempted to highlight here just how uncertain and hence fraught with the risk of unintended bad consequences the IGM of GCs currently is. Consistent with spirit of both Buchanan’s arguments and the rebuttal I have presented here, philosophers working within this area must take seriously the importance of having an adequate understanding of evolutionary theory and genetics. Although philosophy and science are all too often unhelpfully pitted against one another, this discussion has sought to demonstrate the important of cross-disciplinary work in attempting to come to sound judgements upon the ethics and politics of biomedical enhancement.
If you enjoyed this article and would like to discover more about Evolutionary Psychiatry please consider:
subscribing to our Substack to receive regular content updates
visiting the webpage of the Evolution and Psychiatry Special Interest Group within the College of Psychiatrists of Ireland
visiting the webpage of the Evolutionary Psychiatry Special Interest Group within the Royal College of Psychiatrists
exploring a Youtube playlist on curated presentations by the Evolution and Psychiatry Special Interest Group within the College of Psychiatrists of Ireland
exploring the Youtube page of the Evolutionary Psychiatry Special Interest Group within the Royal College of Psychiatrists
exploring the Evolving Psychiatry podcast
First, nice work.
Second, my working knowledge of Red Queen was adaptation relative to a competitive species, not simply environment. You seem to position it in absolute terms which wouldn’t make complete sense.
Third, you mention CRISPR-Cas which of course came from a genetic system which modifies its own genome in bacteria. There are other species which bypass stochastic evolutionary processes and modify their own genome, rare as it may be. Oxytricha create new genomes during reproduction. Under stress bdelloid rotifers self-modify germ lines to accelerate stochastic evolution (why did they get stressed). And, humans have transposition activity that’s not fully understood in germline cells.
One could say that germ line modification simply stems from a genome that’s sufficient to create that system, it’s both natural and a consequence of the level of organzation which has occurred.
Science is littered with the history of humans believing they are not animals or are special agents - and then finding out they aren’t. Heliocentrism, evolution, deep time, consciousness, behavioral genetics, language, tools use, rationality, embodied mind, etc.
However acting on our natural ability may have negative selection consequences. I suspect it’s not probable to know if a given germline change improves fitness universally.