A cousin arranged for me to go to Kenya, to the bush camp of George Adamson. My project involved comparing the behaviour of wild lion cubs with three hand-reared cubs whose mother had been shot, and which George was raising for their eventual release to the wild. I spent about a month at the camp, studying lion behaviour.
There were various important experiences during this period that likely contributed to who I am, including the murder of George Adamson and two of his assistants by Somali bandits two days after I left camp. He was the first person I knew who had met a violent death. It was also during this trip that I contracted malaria, which eventually led me to write this book, but I had another formative experience that resulted from this trip that played out on my return to the UK.
Although I had a great time in Africa, I hadn’t collected much data for my undergraduate project on lion behaviour. The problem I faced was this. Once I had entered all my data on to the mainframe computer I had access to at university and had conducted analyses, I found that when I drew graphs there did appear to be some differences between the behaviour of the three hand-reared lion cubs and the observations I had collected from wild cubs from a nearby pride. However, when I conducted formal analyses, these differences were not statistically significant. What this means is that the differences I observed probably arose by chance. I started to ask myself how many litters of lion cubs I would have needed to collect data from for these results to be significant.
I explored this question by writing some computer code to simulate data. In particular, I used the computer to create a larger data set than the one I had collected, but which had the same statistical properties as the real data. Each column of the made-up data had the same average value as the real data, and the same degree of spread. Some of the columns in the real data were correlated, and the simulated data were correlated in the same way. I would then analyse the made-up data sets and see how big they needed to be to achieve statistical significance. I continued to add rows of data to my simulated data set until an analysis of it reached the threshold of statistical significance.
Much later in my career I learned that there was a much simpler way of solving this problem – an approach that statisticians called power analysis. Although I could have saved a lot of time if I’d known this before, I learned a huge amount about statistics from my simulation exercise. By doing the simulations I taught myself how to code in a computer language called Fortran. I also learned a lot about distributions of data, and about statistical tests. And just as at school I loved learning new things by trial and error, this made me determined to pursue a scientific career. I still get an adrenaline rush when I solve a problem or discover something new. I like the idea that no one else has ever known what I have just found out, however trivial it may be. These breakthroughs are very minor in the grand scheme of things and nothing I have done has had a major impact on the world, but that doesn’t lessen the sense of excitement.
Simulating the data taught me more about statistics than any of the lectures I attended, but it also made me realize that my undergraduate project was deeply flawed. Much in the same way that I cannot prove that a single event radically altered my life trajectory, it is not possible to study three hand-reared lion cubs and a similar number of wild ones and draw any meaningful conclusions. In the same way that no two people are the same, no two lion cubs are equivalent. For my project to have worked I would have had to record the behaviour of many more hand-reared and wild lions, but that was not possible. I described the limitations in the discussion section of my study, submitted my project, and this was enough for me to pass my degree.
Despite being well aware of the limitations of science in being able to assign particular events that happened early in life to personality traits or even the course of one’s later life, I still, irrationally, believe the narrative I describe above. I know I shouldn’t, but I do think that malaria made me decide to try to understand existence, and that a poorly designed undergraduate research project made me a more competent scientist. Can these narratives really be true? Would I have become a scientist and author if I did not have the mutation in GPR143 and had gone to the Perse school instead of Comberton Village College? Would I have become a mathematician rather than a biologist if I had not gone to Africa? Would I have decided to write this book if I had not caught malaria and nearly died? And did doing a dodgy undergraduate project really set me on a course to become an academic? I can never know because I can’t do the experiments required to test these narratives (I’d need many clones of myself), but I do know that all aspects of me are determined by genotype, environment or apparent chance, so perhaps I have constructed this narrative because it includes a little of each.
I have likely decided to believe in the narrative I tell to justify my approach to the science I do or because I wanted a reason for writing a book on what science can tell us about how we got here. As humans we have a strong tendency to look for causes that explain why something happened, and we do this because we want to live in a world we can understand. Most of us don’t enjoy change or the unexpected, and constructing narratives helps bring a sense of order to the unpredictable.
The narratives we tell serve two functions. They are used to justify our behaviours and events we have experienced, and we also use them to make sense of the world around us. Our narratives about our life trajectories can contain egregious acts committed by people that have prevented us from achieving our true potential. For example, if Fred Bloggs had behaved more professionally towards my friend Elmer, he would be a better person, would have won multiple Nobel Prizes, and would have wealth and power beyond his wildest dreams. Sure. Sometimes these slights may be real, other times they are imagined. There are very many well-documented historical cases where individuals have acted vindictively to prevent someone else from achieving success or being recognized for it, including in the sciences, and perhaps that is why we sometimes blame others for failures in our lives.
Isaac Newton is arguably the greatest scientist who ever lived. But he wasn’t a very nice person. He was paranoid and held grudges, particularly when it came to scientific precedence. He fell out with many other scientists who worked on gravity and calculus, including another great scientist of the era, Robert Hooke. Before Hooke died in 1703, he had been lauded by the Royal Society and a portrait had been painted of him that would help ensure his scientific legacy. Newton became head of the Royal Society in the same year as Hooke’s death and he oversaw the relocation of the Royal Society to new premises a few years later, and among the small number of things that went missing were Hooke’s portrait and many of his papers. Scientific historians have repeatedly suggested that these losses were not accidental but were overseen by Newton. He was angry, incensed that Hooke had succeeded while alive and accused him of stealing his ideas, and so Newton did what he could to ensure that Hooke was forgotten to history.
Fortunately, not everyone behaves so meanly. Some narratives can contain acts of kindness that are perceived to have altered the trajectory of a life. People who are generally happy with their lot and at home in their skin appear to be more likely to have narratives based on acts of kindness or on their ability to overcome adversity. In contrast, individuals who feel they should be performing better than they are hold narratives where the bad behaviour of others has prevented them from achieving their potential.