On Friday 26th March 2021, the CPM welcomed geneticist, best-selling author, and broadcaster Dr Adam Rutherford to give our annual lecture, titled Race, Genomes & Data: The Bias Built into Science. If you missed this outstanding talk, you can still catch it on YouTube here.
In this blog, we revisit Adam’s lecture and discuss some of the key concepts.
Dr Rutherford introduced three fundamental topics for his talk: (1) data is not neutral, (2) science is always political, and (3) scientists lack the language necessary to approach some of these issues.
Human genetics is the study of similarities and differences between individuals, but differences are often not solely, or at all, driven by genetics. Consider, for example, differences in experiences of the COVID-19 pandemic. In many Western countries, the risk of death from COVID-19 was substantially higher for those from minority groups.
For example, one of the first large-scale studies of risk factors associated with COVID-19 death that was undertaken in the first phase of the pandemic in 2020 found that, after adjustment for other factors, Black and South Asian people in England were almost 1.5 times more likely to die from COVID-19 than people of white ancestry (hazard rate 1.48 (95% confidence interval: 1.29 – 1.69) and 1.45 (95% confidence interval:1.32 – 1.58) respectively) [1].
Most observers appreciate that these differential experiences of the pandemic are not likely to be driven by genetic differences alone and that socioeconomic factors play an important role. Indeed ethnicity is a social construct that is associated with countless factors that amplify, influence, mediate, or confound associations between health exposures and health outcomes. For example, there was a clear occupation-related gradient in death from COVID-19; London bus drivers, regardless of their ethnicity, experienced higher death rates from COVID-19 than did residents of London more generally [2].
What is much less widely appreciated is how challenging scientists find formulating and applying an appropriate vocabulary to understand differences associated with ethnicity. Dr Rutherford argued that race and ethnicity are both socially constructed concepts that exist because society enacts them. He noted that race, genetic population, ethnicity, geographic population and ancestry are often used interchangeably in everyday language. Despite this, there are no precise or widely accepted definitions or even conventions to express these concepts in science or medicine [3].
“There are no generally agreed norms of race language in science or medicine”
Differences in pigmentation associated with skin colour are often a tangible source of difference between people, and are a conventional basis on which “race” might be assigned to a particular person. However, there is little to no correlation between skin colour and the similarities and differences between people and wider populations. Moreover, genetic variants that influence skin pigmentation emerged hundreds of thousands of years before homo sapiens itself emerged as a distinct species. There is more genetic variation related to pigmentation within Africa than outside it. Racial groups, as traditionally conceived, do not correspond to inherited, biological sources of variation.
It remains the case, however, that experiences of COVID-19, social status, occupational achievement, and health outcomes differ in many societies by conventional classifications of ancestral background (even if misunderstood and misapplied). This clustering of resources and socioeconomic advantages reflects wider historical, political and economic processes.
Dr Rutherford also emphasised the role of science, in particular the scientific study of genetics, in developing notions of “race” . Carl Linneaus, the Swedish botanist and father of modern taxonomy, was the first naturalist to classify humans in the animal kingdom. In 1735, the first edition of his Systema naturae included “man” in the class of Quadrupeds and the order Anthropomorpha, later changed to Mammals and Primate respectively. Man was classified into four types according to geography:
- Europaeus albus: European white
- Americanus rubescens: American reddish
- Asiaticus fuscus: Asian tawny
- Africanus niger: African black
In the tenth edition of Systema naturae, Linneaus further characterised these four principal types of man in five categories: skin colour, physical traits, behaviour, manner of clothing, and form of government.
Not coincidentally (Linneaus was born in Sweden and spent most of his life there), the traits assigned to Europeans painted a more flattering picture of appearance and behaviour than those assigned to others. Europeans were wise and inventive, Asiatics haughty and greedy, and Africanus sluggish and neglectful. Indeed, the widely used and apparently neutral term ‘Caucasian’ was originally used (by Blumenbach) to characterise Europeans as beautiful (and by implication, more beautiful than other races).
Linneaus’ classification was arguably the origin of scientific racism, the expression of which did not escape some of the most eminent names in the history of biology and of genetics specifically. Thomas Huxley, Darwin’s “bulldog” and passionate defender of evolution and natural selection, attempted a classification of humanity in which he identified some ten or eleven “races” referencing skin colour and geographic distribution.
Huxley’s classification included distinctions of race based on skin colour within the continent of Africa, a very distant echo of which was felt in the extremely violent genocidal conflict in Rwanda between the Tutsi and Hutu groups in the 1990s. This distinction between these groups stemmed from the 1930s, when the Belgian colonial government classified the local population into three ethnic groups (majority Hutu, minority Tutsi and the small Twa group). The classification, as one might imagine, lacked any meaningful basis, notwithstanding the subsequent entrenching of inter-group differences by the mythologizing propaganda of colonial and proselytising Christian influences. Over one million people are estimated to have died in this conflict.
Darwin himself was reserved about contemporary efforts to identify what may be described as race. Noting in the Descent of Man (1871) “the greatest possible diversity” of whether there was a single race or many, he concluded “…that it is hardly possible to discover clear distinctive characters between them”. Darwin’s sober opinion was not to be last word in this field.
“If you teach or work in human genetics, the ocean from which human variation is drawn, you have little choice but to speak of race and the history of eugenics”
Many methods still used in genetics and indeed science more generally were motivated by an interest in eugenics. The legacy of some of these individuals has been both productive and pernicious. To this day, many buildings, statues and scientific achievement awards are named after these individuals. Dr Rutherford’s view is that “we should be able to have complex and nuanced views about people”.
Another key point that Dr Rutherford discussed is how “data is never neutral”. Every stage of experimental design, data collection, and data analysis are conceived by individuals who themselves are influenced by both implicit and explicit biases. Some of these biases may be structural biases in the way in which we do science that may have existed for centuries.
“Data is never neutral because it is curated, designed and harvested by humans”
Dr Rutherford’s talk began by noting the changing context surrounding genetics and debate on race. One is the wider political environment in which debates around “race” have, in some respects, grown more divisive. This has occurred despite (or perhaps in part because of) the tremendous growth of genetics as a scientific discipline. Efforts to reduce and redress the historical imbalances in society as well as in the practice of genetics were subsequently discussed in the CPM’s “Facing Disparities in Healthcare” online conference, which confirmed both the need for urgent attention to these issues, but also their complex and multi-faceted political nature.
References: