In 2018, I was very (very) lucky to be selected as one of 60 or so researchers for a British Academy Postdoctoral Fellowship out of, accordingly to the Academy, over 700 applicants. This was unexpected and deeply humbling. And I will be forever grateful to the good people at the British Academy for taking a gamble on me — especially as they were kind enough to let me radically change the topic of my fellowship from "Online Climate Change Denial/Misinformation by Social Movement Organisations" to "Network Formation in Traditional Human Populations" near the second year of the three-year fellowship (2018-2020). In this short post, I partly reproduce the brief research summary from my final report for my fellowship (Grant Num: pf170158). For anyone reading this post who is thinking of applying - I very strongly encourage you to do so. The British Academy is a fabulous organisation/funder and their postdoctoral fellowships are a great way to carry out whatever "Blue Skies" research project you are passionate about.
The objective of my postdoctoral fellowship (PDF) was to determine the extent to which the formation of human social networks might vary across environments using: (i) simulation-based methods for modelling network dynamics; and (ii) data from low- and middle-income countries (LMICs) on relationships between adults in rural villages.
To meet my objective, I completed three sole-authored papers.
Published in the journal Social Networks, my PDF’s flagship paper used data on friendship amongst 4,713 rice-growing smallholders in 162 rural villages in China to show that the most important determinants of who farmers choose as their friends (i.e., key sources of occupational support) systematically varies with the amount of agricultural land under their control. Crucially, my finding challenges commonly-held assumptions about friend choice — i.e., that humans near-universally prioritise reciprocity (i.e., responding in kind) and transitivity (i.e., befriending friends-of-friends) — to instead show that this process can be highly context-dependent.
In a second paper, also published in Social Networks, I used data on the provision of tangible aid amongst all 108 adults of a village of horticulturalists in Nicaragua to show that the patrons/beneficiaries of those who live together are distinct such that it is ill-advised to assume that the sources/targets of help for any one coresident is representative of their household. This finding is non-trivial as scientists often use “household heads” to map networks spanning villages in societies across the globe in order to carry out vital research on the link between social support and wellbeing. However, my results indicate that uncritical application of this simple data-collection strategy can easily combine with local context (e.g., norms around living with adult kin) to increase the risk of measurement error.
Finally, in the third paper (currently undergoing peer review), I use the Nicaraguan data to evaluate the power of genetic kinship and direct reciprocity over who humans help. Breaking with research using study designs focused on dyads (i.e., pairs of individuals), I find that helping family and responding in kind are less important to who one helps compared to the dynamics of the tangible aid network itself (e.g., helping residents helped by many others). My findings provide, arguably for the first time, clear, quantitative evidence of the supremacy of supra-dyadic network dynamics over genetic kinship and direct reciprocity. But, following my results from China, my Nicaraguan findings also indicate that the salient determinants of the formation of social support networks likely reflects contextual factors linked to variation in cooperativeness (e.g., in the case of my study, religious homogeneity and a high level of genetic relatedness between villagers).