This project is maintained by amartrics
In his book Scale: The Universal Laws of Growth, physicist Geoff West takes an interesting approach in describing the processes of human development. One could argue that West does not actually describe the complex, interconnectivity that supposedly characterizes development, as proposed by Owen Barder in his 2012 Kapuscinski lecture, but instead, summarizes every kind of developmental phenomena – organic, social, economic, and infrastructural alike – through “a common conceptual framework underlying [everything]” (4). To suggest that “all complex systems, from plants and animals to cities and companies,” actually obey the same few simple, linearly-governed rules may seem revolutionary when viewed beyond the scope of statistical scrutiny, but in reality, West’s argument makes a lot of sense in respect to development. Through several strong examples and the utilization of his scientific background, West reasonably and logically explains the scaling laws that govern virtually every observable life pattern in human society, from individual social metabolic rates (e.g., the amount of energy needed to sustain a modern human) to the growth of galactic cities. While not disproving the ideas of complexity and necessitated theory idealized by Owen Barder and Rob Kitchin, respectively, West’s presentation of development as a digestible and numerically-governed process proves outright the quantifiability and evolutionary nature of man-made systems and, in turn, the power and promise of big data projects as analytical tool for the furthering of human progress.
Big data is a nebulous term, to say the least, but in short, it is a field of data science focused on the extraction and evaluation of multifaceted data that is too complicated, too abstract, or simply too large to process through traditional means of processing. The recent explosion of data science in the realm of human development has allowed for the collection of never-before-seen data. More importantly, the predictive attributes of many algorithms currently implemented by advanced software have allowed for the breakdown of human processes, both past and present, to an almost minute level of quantifiable data points. It is this facet of data’s evolution and its rise as the fourth paradigm of science, as geographer Rob Kitchin puts it, that offers the most promise in terms of the advancement of human understanding, especially in relation to the constantly-shifting, interconnected networks, economic activity, and increasingly broadening social systems that govern the direction of contemporary development. The sheer power of predictive algorithms in the hands of big data software can be observed in the results of recent studies in Southeast Asia, which utilized a combination of Bayesian generalized linear modelling and neural networks to plot hotspots of female poverty, child stunting, neonatal mortality, and low literacy rates based on a relatively small pool of priori. Studies like this are able to ignore human-induced biases, such as those that have repressed surveys and examinations of gender inequality in Southeast Asia and in turn reduced the amount of available data on the topic, in order to create more accurate, unbiased, and geo-specific images of reality. The benefits offered by this project and similar endeavors cannot be discounted, and the relation of these numerical results to social networks, economic activity, and physical wellbeing in the region act as a perfect example of how much big data can touch in terms of explaining the interconnectivity of human development.
The Southeast Asia study brings up an important matter that is currently being debated across the field of data science. Though a purely empirical approach can provide rich rewards in the form of detailed data in some situations, it also holds potential hazards in disregarding the importance of human influence of data collection. As Rob Kitchin writes in his analytical response to Chris Anderson’s article, “The end of theory,” data cannot be utilized as an entirely self-supporting entity. Disregarding the natural pathways of experimentalist thought, theoretical modelling, computation, and exploration that have allowed for the incorporation and explosion of big data within modern society thus far in search of a world of data detached from bias undermines the power that this field of science offers in the specific context of human development. In contrast to the entirely data-based epistemology supported by Anderson, Kitchin argues that the inclusion of social sciences and a humanitarian mindset is essential to shaping the future of big data as it affects people, lest the niche attributes of localized culture, personal bias, and ultimately, human nature are to be overlooked in a new era of “data-driven” science (Kitchin). Relying on data alone for the breakdown of qualitative factors could spell disaster for under-researched topics, especially those that fall under the umbrella of human development issues. It is already far too common for “hot” research papers, analyses, and attractive techniques to overtake genuinely novel and insightfully conducted studies in the public eye. Placing stock in only what is generated by algorithms will only further the divide between issues that are popular or data-rich and issues that are necessary. Instead of moving towards the post-positivist approach presented by Chris Anderson, Kitchin asserts that there is still quite a lot of potential to be found in the theory-based exploration of existing problems, utilizing data-driven tools to support the critical review of underrepresented populations, problems, and data to create a better understanding of reality.
This conclusion is unequivocally supported by Geoff West, who, despite the impression readers may gather from his focus on the quantifiable characteristics of human behavior and development, clearly supports the cultivation of truly multidisciplinary research. The continual refinement of existing theories, West writes, must guide what big data is used for and how effective it can be; science, he argues, cannot advance without at least the semblance of mechanistic, unified theories. In contrast to the apparent attachment to quantification he demonstrates throughout his book, West acknowledges that a solely data-based future is ultimately not a realistic one due to the innately meritocratic nature of science. In West’s view, it is not enough to rely on the analytical constituents of big data alone; theoretical frameworks are necessary to understand any kind of collected data.
It is through this lens that West’s proposal of scale as a governing power over all phenomena – specifically, the idea that all fundamental, measurable characteristics of organisms, physiological quantities, and life history events and thus, their whole sums, scale with size along predictable, exponential trend lines – appears clearest in application to development. Scalability is the great equalizer, acting as a lens through which data scientists and development experts can analyze seemingly dissimilar phenomena, activity, and organisms across disciplines. Though the lifetimes of complex systems and organisms like cities, humans, animals, companies, and plants may demonstrate massive variances, they ultimately scale in the same way when examined as sublinear and superlinear expressions of data. Each of these unique manifestations of the modern world are comprised of individual components which, once aggregated, adopt collective characteristics identified as those of these components’ metaphorical sums; each are self-organizing and adaptable; and, most notably, each can be summarized concisely through the implementation of divergent forms of Kleiber’s law – a precise mathematical formula which relies on fractional power scaling. In the organic world, metabolic rates can be scaled “as a power law whose exponent is very close to the number ¾” (West 25), whereas city infrastructures can be scaled through exponential functions with degrees of 0.85 and companies through power laws with exponents of 0.9.
Big data is clearly a realm of great promise, but also great danger. Through a desire to detach from bias, societies may find themselves stuck in a developmental mire, unable to synthesize experiential observation with information recorded by increasingly intelligent machine learning systems. It is within the works of authors like Geoff West and Rob Kitchin that the rosiest vision of a future suffused with data appears. By integrating human understanding with machine processing, the overall reduction of unfreedoms and cultivation of the five key freedoms demanded by Amartya Sen for the achievement of true development may finally be observed. West notes that “… measurement has played a central role in the development of our understanding of the entire universe around us. Data provide the basis for constructing, testing, and refining our theories and models whether they seek to explain the origins of the universe, the nature of evolutionary processes, or the growth of the economy.” However, he also acknowledges the importance of combining quantifiable data – measurement in application – with the qualitative aspects of humanity as an irreplaceable tool of empathy to advance social understandings of complex adaptive, economic, and social systems.
Works Cited
Barder, Owen. “Development and complexity.” Center for Global Development, 15 Aug 2012. https://www.cgdev.org/sites/default/files/archive/doc/multimedia/Development_and_Complexity_Slides.pdf.
Kitchin, Rob. “Big Data, new epistemologies and paradigm shifts.” Big Data & Society, April/June 2014, pp. 1-12. Sage Publications, https://journals.sagepub.com/doi/10.1177/2053951714528481.
West, Geoffrey. Scale: The Universal Laws of Growth, Innovation, Sustainability, and the Pace of Life in Organisms, Cities, Economies, and Companies. Penguin Press, New York, 2016.