Developers and everybody else: An attempt to bridge the chasm

Fig. 1. Feelings are facts. Wrote on Flickr, used under Creative Commons 2.0 license.

For those of us who are steeped in positivism, whether through academic training or by profession, the claim in figure 1, “feelings are facts,” is counter-intuitive. We are trained to imagine an objective reality,[1] a reality which is, in fact, really only the one that most of us happen to agree upon. The whole problem of ‘truth’ and thus of objectivity is in fact terribly problematic. Simply put, there is no theory of truth that withstands scrutiny.[2] And there is a terrible difficulty with any claim to objectivity in a view of reality that only permits a reliance on the very reality in question[3] and thus asserts an exclusive claim to adjudicate which methods of inquiry will be accepted as legitimate research methods, and what knowledge and what forms of knowledge will be accepted as legitimate.[4] Feelings, in this view, are subjective and therefore not true, and therefore false. They cannot be ‘facts.’

Reality intrudes, however, and even if its impact on research methods often remains constrained, the reality is that scientists increasingly recognize the limitations of the positivist view.[5] But of all the academic and professional fields out there, computer science might be the one most prone to positivist thinking. In significant part, computer programming involves decisions based on true-or-false propositions. All data reduces to numbers which are ultimately represented in binary. A computer itself seems immovably objective, mocking programmers’ and users’ pathetically subjective notions of what it should do and how it ought to follow the instructions in program code. Computer science, even more than logic itself, is a most quantitatively reductive field; this is, in fact, its purpose so that computers may assist with real-world problems.

Computer programmers professionally reside in the discrepancies between expectations of what a computer should do and what it, in fact does. They are caught in between that immovable mountain of what a computer can be told to do and (more often than not) fickle or even capricious users’ demands. When I was a programmer, systems analysts were people who helped mediate between users—ordinary human beings—and programmers, enabling programmers to focus more on negotiating—it’s never just writing—code with the ultimately impassive computer, a machine that is something more than a machine, but which cares not whether it does what humans want it to do, let alone whether it does so correctly.

It’s been three decades since I was last employed as a programmer/analyst, a role that combined systems analysis and computer programming. In that time, ‘systems analyst’ has come to mean something different and, to be honest, I have no understanding what the job title means now. And we speak less of programmers and more of ‘developers.’ Computers, of course, have become many orders of magnitude more powerful in every quantifiable dimension, but the fundamental problem of discrepancies between expectations and possibilities remains. I won’t call it intractable; research into artificial intelligence long ago made more progress than I thought possible, and there’s no way to forecast when ‘singularity,’ the moment when a machine or collection of machines becomes truly ‘intelligent,’ might occur, nor even really a satisfactory means of recognizing when computers enact rather than mimic autonomous intelligence. Even the much-vaunted Turing Test relies on convincing humans—which probably may be accomplished through mimicry much more easily than through actual accomplishment.

Part of the problem here is definitional. Intelligence itself is a social construction. We recognize as intelligent those whom we consider intelligent. This is a determination by consensus. Which is why, I am given to understand, the Turing Test is constructed the way it is. But in bypassing the problem, it sets a much lower threshold than can be considered satisfactory.

Autonomy, or self-direction, comes from need, even if that need lies at the top of Maslow’s hierarchy—self-actualization. In living creatures, need is expressed through feeling. It is part of what renders untenable the long-imagined distinction between mind and body.[6] So feelings are indeed facts and positivism is, even if on no other grounds,[7] an entirely inadequate view of reality.

The positivist view nonetheless prevails among many, especially those who rail against religion and assert the supremacy of scientific method over any other means of apprehending reality. I see it in those who deny that atheism—an assertion that no deity exists—is in fact based in faith. No evidence acceptable to positivism supports either the view that affirms or the view that denies a deity, yet these people attach the label ‘faith’ to one and not the other, arrogantly stigmatizing faith and thus contributing to a perception among social conservatives that they are being persecuted.

This arrogance, borne of a simplistically reductionist world view, carries over into developers’ relationships with users. The latter are stigmatized for lacking what developers consider a basic skill: coding. The developers’ assumption, ignoring a specialization of labor that has existed since the Neolithic, is that everyone can and should learn how to code. The capitalist libertarians among them should be ashamed: Specialization of labor is a necessary precondition to any systematic exchange of goods and services—a market. More generally, as I have learned from painful personal experience, not everyone is born to code; we have different inclinations and talents, this diversity is intrinsic to humanity, and we should welcome and encourage rather than devalue or in any way rank all of these inclinations and talents.

So imagine my surprise when I found the image above (figure 1) attached to a posting from Luis Villa, the Senior Director of Community Engagement at the Wikimedia Foundation. In it, he describes an OSFeels agenda:

  • Design and empathy (learning to build open software that empowers all users, not just the technically sophisticated)
  • Inclusive development (multiple talks about this, including non-English, family, and people of color) (so that the whole planet can access, and participate in developing, open software)
  • Documentation (so that users understand open software)
  • Communications skills (so that people feel welcome and engaged to help develop open software)[8]

Villa appears to have an unexpected clue about human development. He cites the ‘capabilities’ approach advocated by Amartya Sen and Martha Nussbaum,[9] an approach grounded in a notion that autonomous beings should expect to be able to develop to the limits of their capabilities, insofar as such development does not impose limits upon other beings’ capabilities. This is an approach to the problem of human rights that encompasses everything from basic physical needs to self-actualization needs, and Nussbaum, at least, describes how the approach might apply to non-human animals as well.[10] The notion that autonomous beings have different talents and inclinations is foundational to this approach. There is no place in it for developers’ prejudice against non-developers. I am astonished and impressed.

  1. [1]Bruce Mazlish, The Uncertain Sciences (New Brunswick, NJ: Transaction, 2007).
  2. [2]Bradley Dowden and Norman Swartz, “Truth,” Internet Encyclopedia of Philosophy, September 17, 2004,
  3. [3]Timothy Williamson, “On Ducking Challenges to Naturalism,” New York Times, September 28, 2011,
  4. [4]Yvonna S. Lincoln, “Institutional Review Boards and methodological Conservatism,” in Landscape of Qualitative Research, eds. Norman K. Denzin and Yvonna S. Lincoln, 3rd ed. (Thousand Oaks, CA: Sage, 2008).
  5. [5]Fritjof Capra, The Web of Life: A New Scientific Understanding of Living Systems (New York: Anchor, 1996); Colin Robson, Real World Research, 3rd ed. (Chichester, West Sussex, UK: John Wiley and Sons, 2011).
  6. [6]George Lakoff and Mark Johnson, Philosophy in the Flesh (New York: Basic, 1999).
  7. [7]David Benfell, “For those who still aspire to a natural science approach,” November 8, 2011,
  8. [8]Luis Villa, “Software that liberates people: feels about FSF@30 and OSFeels@1,” October 2, 2015,
  9. [9]Luis Villa, “Software that liberates people: feels about FSF@30 and OSFeels@1,” October 2, 2015,
  10. [10]Martha C. Nussbaum, Creating Capabilities: The Human Development Approach (Cambridge, MA: Belknap, 2011).

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.