Approaching the Power of Algorithmic Decision-Making

This essay is provided by Ajla Nesimovic, former student in the master program Organization Studies at University of Innsbruck, and based on her master thesis.

“This master thesis is a story,” are the first words of my thesis. If you are now frowning and thinking what the heck I am talking about, then you are definitely not alone. Once an inspiring person taught me that managing expectations could be helpful as it might give a sense of motivation and direction. Now that I have told you that my master thesis is a story you are probably expecting a lot of fairy tale and little scientific appropriateness. You are not that far off! I definitely write a lot about ambiguities and contradictions of theorists. In later sections, I critically reflect on my very own work and further identify it as an invention with a lot of ambiguities and contradictions too. Nevertheless, my supervisor wanted me to write a blogpost about my master thesis. I suppose, it’s because of the jokes.

The story is multilayered as it consists of various story lines which are differing from each other but are still overlapping and coexisting. My master thesis, therefore, can be read in many different ways: as a love letter to the study program Organization Studies; as an imaginary and intellectual debate between my AI professor and myself; as a story about myself; or as a story about algorithms. I am not offering these different opportunities to potential readers by accident, since this thesis was guided by an interpretation of Deleuze’s and Guattari’s process philosophy (1994).

Why story? Well, to answer this question, let us return to the very beginning, the reading into the master thesis’ topic. Reading various scientific articles about algorithmic decision-making within organizations sometimes felt like a lump in my throat. I had the feeling that researchers tend to embody a specific attitude when writing about algorithms. The attitude goes something like this: ‘Look, I am extremely careful with my words when trying to describe what algorithms are. The reason I do this is because I want to deliver clear and distinct statements that carry very specific and pointed meaning over algorithmic decision-making processes. This is not always easy for me to do, admittedly, sometimes I have to pause and think about for how exactly I want to describe the algorithmic practices, but because of how careful I am, basically any rational person that is reading my words will come to the meaning I was attending.‘

To put this in another way, the articles wanted to make me believe that there are stable and authentic meanings to words out there, the same way one might think that there is a stable and authentic reality out there. Do not get me wrong here. I am not judging researchers for trying to be precise with words, I am judging them because of occupying a “God’s Eye View of the World” (Putnam 1981), thereby assuming a prior existence of algorithmic decisions, representing the algorithmic world as it is in itself, and relying on a notion of truth as a desirable end-state. In is this sense, the articles wanted me to understand the scientific study of algorithmic decision-making as a cumulative project within a community of practice, where the main aim is the grounding of knowledge so that theory-building can proceed unimpeded. Science, the articles further tried to make me understand, enjoys cultural authority. The knowledge created within these scientific communities is perceived to be much more valuable than other forms of knowledge. Thus, the colonization of thought turned out to be a central feature within the modern science of algorithmic decision-making. Let this sink in for a moment. It is this subtle religious connotation that resonates with modern scientific approaches towards algorithmic decision-making that caused most of the lumps in my throat.

Whereas several centuries ago the legends and the myths, and then the authoritative voices of the church, provided the guiding ‘truths’ which sustained particular forms of life significantly different from that of our own, today it is science which speaks with the same kind of authority. (Chia 1996, p. 27)

Instead of representing algorithmic decisions by theorizing towards the grounding of knowledge, I wanted to intervene within the science of algorithmic decision-making by questioning the basic beliefs and conceptual categories upon which traditional approaches are built on. From this perspective, the truth about algorithmic decisions becomes secondary. The focus is rather placed on fundamental social organizing processes that inform algorithmic decisions. Thereby, I problematize the explanations offered by traditional theorists in diligently opening the space for all my subjective tendencies by adopting a passionate attitude within the inquiry process and using less familiar and unconventional research methods – the opposite of what traditional views would consider as good science. That’s why I called my own master thesis a story. “Irony, self-reflection and ‘playful seriousness’ now replaces the quest for ‘certain’ or ‘partially-true’ knowledge of an external reality” (Chia 1996, p. 17). My approach to this master thesis then can be considered as the art of “telling ourselves a story about ourselves” (Steier 1991, p. 3), while privileging an emergent and processual view of algorithmic decision-making.

Potential readers can therefore expect a very personal way of doing research, which will contribute to their very personal way of understanding algorithmic decision-making in contemporary organizations.  At least, I would wish that potential readers don’t end up with the same comprehension of algorithmic decisions since it was never a goal of this writing to dictate what algorithmic decision-making is or how it should be done. I rather encourage to embrace openness and difference if the phenomenon of algorithmic decision-making is to be beneficial for organizations and society.

Link to the master thesis:

  • Chia, R. (1996). Organizational Analysis as Deconstructive Practice. Berlin and New York: Walter de Gruyter.
  • Deleuze, G., & Guattari, F. (1994). What is Philosophy? New York: Columbia University Press.
  • Putnam, H. (1981). Reason, Truth and History. Cambridge: Cambridge University Press.
  • Steier, F. (1991). Research and Reflexivity. London: Sage.

One thought on “Approaching the Power of Algorithmic Decision-Making

  1. Wonderful post Ajla, and thank you for making you thesis publicly accessible.
    In the recent issue of Academy of Management Review (vol.45, no.1), there is a review essay of E.M. Forster’s “The Machine Stops”. The authors bring up the underlying assumption of rationality in algorithmic decision making. Perhaps this is also an interesting read for you.

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s