10 min read
In this post, I discuss the notion of recursiveness in a roaming fashion. I address its implications for SaaS for community engagement in urban planning, as well as for planning as a discipline more generally.
The post ventures into some metaphysical chartered territory. But a bit of reflection can only do good, particularly in our data-obsessed world. Rather than throw the baby with the bathwater, this post argues the very best use of data is through human knowledge, if not wisdom.
Welcome to recursiveness
A most unsignificant step for humanity, but a breakthrough for me, was to find out about recursiveness. I first discovered the concept while learning basic Python at the beginning of the PhD on Coursera with the excellent teacher ‘Dr Chuck’.
More understated than loops, recursive codes essentially redefine the operations they enumerate. They transform themselves through sheer iteration. Loops go ‘round-and-round’ or ‘round-and-out’. But recursive codes are reflexive, self-(re)defining. They can therefore require intensive reflection to design. Recursive codes display an almost existential and self-questioning quality. Recursive codes can use more computing power than loops, which ironically take more physical space in a script. But that really depends on context, apparently. My hunch is that machines don’t really like to question themselves. A bit like humans, except more so. Which is the point of this blog post.
The term ‘recursion’ seems more common than ‘recursiveness’in mathematical, programming and linguistic circles (which can also be the one and same communities). The site WikiDiff sees recursion and recursiveness as related, with recursion defined on Wikipedia as “the act of recurring” and specially in mathematics as “the act of defining an object (usually a function) in terms of that object itself.” Recursiveness simply denotes “the quality of being recursive.”
Recursion, as a linguistic sibling of recursiveness, has been known to point toward infinity. The most quintessential example of infinity in maths is the Fibonacci sequence (or golden ratio), which can reach both into the infinitely small and the infinitely large, in both time and space. This numerical approach to infinity still has very practical applications in contemporary graphic design, among many other realms. Therefore, recursion is paradoxically infinitely finite. As poet William Blake wrote this in his mystical Auguries of Innocence:
Machinic, existential recursiveness
You may remember the classic science fiction film ‘2001: A Space Odyssey’ by Stanley Kubrick. Machines (in this particular instance: HAL 9000 – arguably the most famous ‘Heuristically programmed ALgorythmic computer’) are not particularly fond of recursive processes and discourses. Recursiveness might well be the ‘Achilles’ heel’ of all machines: the ‘I’ of the Robot, to paraphrase Isaac Asimov. Where ‘I’ is conceived as spiritual, this may even point to the ‘eye in the sky’, or the self-awareness that can learn to distance itself from itself (‘it’s self’), to see things more as they are, and for what they are, and so to take things less personally or literally. The superpowers of machines lie in hyper-rationality and algorithmic hyper speed, which can sometimes go out of hand, as both science fiction and the 2008 financial crisis illustrate in remarkably different ways.
Diving into the fundamental, existential level of experience is probably the bane of all machines. Machines think logically, or not at all. Perhaps it is truest of machines to claim: ‘I am, therefore I think’, to put Descartes’ saying on its head. To investigate and question the self in itself is to enter a ‘no machine’s land’. It is a personal battlefield where the only envisioned outcome might be, paradoxically, self-destruction. Descartes also left us with the great legacy of breaking down big problems into easily manageable chunks in a very logical and rational manner. To this day, the Cartesian approach to problem-solving approach is a great productivity hack, and it also provides the classic way to design and engage with machinic reasoning.
But logical reasoning is only part of the big picture. Descartes was both human and humane, being a religious/spiritual man. His work indicates that reason could only manifest in such a way that would accord with the inner, perhaps less ‘rational’ workings of faith. There are many publications on the topic (see for example the recent work by Aurélien Chukurian in French). In contrast, it seems machines must ‘sweat it out’ profusely when it comes to the ‘I’, and its non-material place in the wider universe.
What can a machine see about itself through its own eyes, as engineered by humans? It takes a human ‘I’ to spot and assemble the machine from a mountain of electronic debris. But truly, modern humans have deeply embraced of all manners of machine-based infrastructure, devices and applications to, one might guess, make their lives less burdensome and more enjoyable. Machinic infrastructure, hardware and software also provide the opportunity for humans to see themselves with a fresh look, through a pair of new appreciative robotic eyes, at the cost of relying solely on their own physical eyes. The artist falls in love with her/his creation – a sort of 21st century version of the Ancient Greek tale of Pygmalion. Another disconcerting example is the surreal romantic relationship between a man and his operating system, as explored in the spookily understated drama ‘Her’. The notion of machinic, algorithmic ‘identity’ is a perennial theme in electronic music, from the experimental music by Kraftwerk to the drum and bass by Seba.
It is peculiar kind of reflexive gaze, although a displaced or machine-mediated one. As I mention in the Prelude to this series of blog posts, an important component of Science and Technology Studies is to fully acknowledge the role of technology and otherwise inanimate ‘objects’ in (re)shaping the world, knowledge claims and even relations amongst people in society.
Overall, this philosophical detour might go straight to the heart of explaining the existential contours of machinic experience, including why recursive codes take more computer power to process, although being so short and beautifully crafted. Recursiveness can be translated in planning practice as reflexive practice, self-aware learning by doing, and continuous becoming as a practitioner or researcher (see the work by Donald Schön, John Friedman, Bengt Flyvbjerg and John Forester, among others). Recursiveness as a ‘quality’ seems therefore to be a core component of the human experience more than it is of the machinic existence.
But as the band Pink Floyd famously played, the alluring ‘welcome to the machine’ might really be a tight, constraining steel embrace that can crush us to the core. From George Orwell’s 1984 and the Blade Runner films to Terry Gilliam’s Brazil, the creative imagination has left us with ample warnings about undesirable futures-to-be where people would serve machinic states, rather than manage to keep friendly Robots and Co. as servicing real, fundamental human needs. As a leading researcher and consultant in the field of decision-support systems Power (2016) also warns of that the ubiquitous presence of technology could operate as a sophisticated Big Brother surveillance system. A tool remains a tool unless it controls the master.
A machine-driven world would also be a natural extension of the One-Dimensional Man (1964), a hyper specialised and rationalised conception and consumerisation of human individuals’ contribution to society, as critiqued by Herbert Marcuse of the Frankfurt School. As I discuss in a blog post about modern-day cyborgs, digital technology has become a bionic extension of who we are as citizens and professionals. This affects us all areas of life, not least of which placemaking and urban planning. Social media and digital addiction are a benign yet far-reaching example of humans and machines freely exchanging seats, which has taken a toll on people’s capacity to focus on work, mental health and the most deeply satisfying activities that truly matter to them.
So the question boils down to: Do the tools we use serve our common good? Or do we serve the tools that were initially supposed to serve us? Ang forward: Will machines continue to work for humans? Their narrow capacity for ‘recursive becoming’ seems to indicate ‘yes’. Machinic superpowers, particularly their capacity to process vast amounts of data, can indeed be an extension of human reason and sensibility and serve the common good. This is the self-consciously optimist position adopted by Sam Gilbert in his book Good Data.
From d8t8 and 1nform@at10n to kn0wledge and w1sd0m
I vividly remember the public VIVA/PhD defence of a friend and colleague at KTH in Stockholm whose external examiner asked, at a critical time of the examination: “What is the difference between data, information and knowledge?” I could not help but smile whilst feeling sympathy for my friend. My answer would be, without the stress of a VIVA situation, that data is raw information-to-be, and knowledge is to how to live that through real-world experience and reflexive learning. The next step on the machine-supported assembly line is wisdom, which is essential to make sense of lived experience (i.e. ‘knowledge’).
Today, there is an almost infinite amount of data available, hence the heightened/vertical infinite symbol ‘8’ in ‘d8t8’. This causes severe information overload. Clearly, machines can crunch vast amounts of data and extract useful information while humans are best at identifying patterns from lived experience. Big data is just too big for humans to learn without the help of machines, which ironically don’t ‘learn’. I remember a conversation with a data scientist who explained why ‘machine learning’ is really an oxymoron. One can only learn so much from ‘0s’ and ‘1s’.
What constitutes wisdom might vary for people, but cultural diversity and divergence about what wisdom ‘is’ should not stop us from trying to nurture our best collective insight to help heal our planet. The capacity for deep, mindful insight is distinctly human(-e), and a more solid basis for evidence-based planning than big data or information alone. The key point here is that machines seem to excel and outperform humans when it comes to capturing and number-crunching data and information. But humans will always have the upper hand (and presumably veto power) when it comes to knowledge, not to mention wisdom. Lest we should watch the Terminator film series.
In another blog post, I argue that today’s smart cities could learn from ‘smarter pasts’ now to craft smarter futures. The ‘past’ that lives into the ‘present,’ and which shapes our collective legacy through the Anthropocene, offers perennial, location-based wisdom for engineering, architecture and the design of the public realm. Even as we seek to innovate again and again, there is much that we can learn from the past and our ancestors, both in terms of successes and relative failures. To provide just one example, it is ironic that Plato’s ‘Republic’ and the Socratic ideal of dialogue, is one of many foundations to a communicative, democratic approach to planning, which has largely the underlying logic and pragmatic design of digital participation in urban planning. Better decisions could be said to arise from wisdom as the appreciation of high-quality information, as grounded in lived experience. The past lives on in today’s placemaking practices and cutting-edge technology in more ways than meets the eye or the shallow-thinking mind. As many others have argued, and my own research participants confirmed, information is not a low-hanging fruit of public participation, but rather a fundamental pre-requisite to high-quality community engagement Furthermore, the quality of information in planning processes, and the quality of the participation it enables, can accrue over time (see this great book by Nabatchi and Leighningher, and page 266 in my thesis).
There is only so much knowledge in planning practice, if any, that can arise from data capture and rule-based reasoning. Bengt Flyvbjerg (2006) famously wrote the landmark methodological paper ‘Five Misunderstandings about Case-Study Research’ about what constitutes planning expertise. He compellingly argues that planning practitioners derive their expertise, sensibility and know-how from the cumulative experience of countless, unique cases. Thanks to such aggregate, context-specific experience, they can identify overarching patterns and tacit rules. Textbooks, on the other other hand, provide ready-made theories and rules of thumb that do little justice to the information-rich (and truly exceptional) nature of case-based knowledge. Reflexive, human(-e) experience is what differentiates data and information from knowledge, not to mention wisdom. Such distinctly human(-e) experience lies at the heart of communities of practice (e.g. see the landmark book by Etienne Wenger), craft-based knowledge (e.g. see Making and other work by Tim Ingold) and reflexive practice (see the Reflective Practitioner by Donald Schön).
Another landmark paper is the paper by Carl Benedikt Frey and Michael A. Osborne (2013) ‘The Future of Employment: How Susceptible are Jobs to Computerisation?’. The authors predicted that in the contemporary era of continuous automation, the jobs that will be most resilient are those that require extensive interpretation. The emerging trend toward ‘data-driven planning’, underpinned by national policy and location-based data, could make more explicit the inherently human(-e) component of evidence-based planning. This paper by Nochta and colleagues (2020) about digital twins makes the case for a strong socio-technical approach to urban analytics, favouring a locally relevant ‘challenge-led´ perspective rather than one led by data alone.
Evidence relies perhaps less on academic theories or binary algorithms – not matter how advanced or elaborate these may be. Healthy evidence arguably relies more on human judgement and a moral sense of care for fellow sentient beings. The evidence for sound planning is not self-evident either, whether rational, linear, algorithmic, data-steered, as in oversimplified textbooks or in blunt, out-of-touch policies. Rather, evidence may be emerging, contingent, sensitive, agile as well as sensible and well-reasoned. A seasoned practitioner makes use of both sides of the brain and can think both fast and slow. Not that I can speak from experience… But this is the observation I have made from my many research participants across three continents.
The intellectual and advocacy of progressive planning as a discipline has repeatedly sought to unite heart and mind, which has come to shape the world of SaaS for community engagement. As we journey from data to wisdom, it matters to combine the caring, activist, advocacy-based, climate-aware qualities of the heart, with the reasoned, informed, evidence-based, open, politically and economically astute capabilities of the mind. One can cite two civic tech that exemplify this approach, among many others: 1) the Helsinki-based Maptionnaire; and 2) the US-based Neighborland platforms. Maptionnaire openly advocates the principle and ambition of enabling citizens to shape the decisions that influence them by providing a “digital heart for building cities together”, while Neighborland’s mission is “to empower residents to shape the development of their neighborhoods.” Many other Civic Tech have similar pursuits. Decades of planning history, theory and case-based experience inform the very design of these platforms, and their ongoing use across all manners of planning projects. To cite but one example, a recent post by Nouvella Kusi on the Commonplace blog presents the landmark Ladder of Citizen Participation by Sherry Arnstein (1969), which is proof of the enduring salience of (almost) timeless principles to inclusive planning. All civic tech evangelists also explicitly recognise the risk of exclusion through ‘digital only’ approaches. My own cumulative research findings also showed that real value of SaaS lies in leveraging hybrid and blended approaches to participatory planning, where the evidence base is (at least partly) shaped by urban residents.
In zer0 sum, there is an almost infinite amount of data out there to help inform the evidence base for planning. Machines have a great role to play in transforming that ‘d8t8’ into readily usable information. Likewise, humans naturally possess an equally infinite potential to transform information into wisdom, by way of living knowledge, both as individuals and communities. The key is to make sure data and wisdom can align to make better decisions for/with both people and planet.
This points to the potential of SaaS and other PlanTech to leverage ‘hum@ne planning efficacy’, which will be the topic of a new post.
Flyvbjerg, B. (2006). Five Misunderstandings About Case-Study Research. Qualitative Inquiry, 12(2), 219-245. doi:10.1177/1077800405284363
Nabatchi, T., & Leighninger, M. (2015). Public Participation for 21st Century Democracy. New Jersey: Jossey-Bass.
Power, D. J. (2016). “Big Brother” can watch us. Journal of Decision Systems, 25(sup1), 578-588. doi:10.1080/12460125.2016.1187420