Environment and Planning B and me;or what is lost in data

Environment and Planning B, particularly in its previous guise as Planning & Design, was my ‘first love’ among academic journals.

I distinctly remember encountering the 1997 special issue on ‘urban systems as cellular automata’ on the sixth floor of the library at the University of Glasgow in pursuit of… well I don’t really remember what exactly: I may have been chasing down ideas on urban morphology in the context of an assignment on 3D GIS as part of my Masters study in Cartography and Geoinformation Technology. I had chanced upon space syntax, a perennial topic in ‘B’ and was doing my due diligence. This was a time when, upon finding a promising journal for a topic at hand, going into the stacks and skimming through all the tables of contents was still the most reliable way to make sure of not missing earlier work.

Primed as I had been by Fractal Cities (Batty and Longley, 1994) and before that James Gleick’s Chaos (1987) and Steven Levy’s Artificial Life (1992), that special issue’s table of contents instantly got my attention, and to this day Environment and Planning B 24(2) may well be the only issue of any journal that I’ve read cover to cover! The editorial commentary (Batty et al., 1997) and papers therein (Batty and Xie, 1997; Benati, 1997; Clarke et al., 1997; Couclelis, 1997; Phipps and Langlois, 1997; Portugali et al., 1997; Sanders et al., 1997; Wagner, 1997; White and Engelen, 1997) were essential reading for anyone interested in complexity and urbanism. They also set a high bar for what a special issue can be, which I’ve rarely seen matched. It’s too long ago now to remember the sequence of events with certainty, but it isn’t at all a stretch to say that that issue of this journal set me on the course that my career has since followed for over 25 years.

Since Planning & Design was, through Mike Batty the editor, akin to the in-house journal of the Centre for Advanced Spatial Analysis (CASA) at University College London where I pursued my PhD, ‘B’ continued to teach me what this mysterious academic game I was being apprenticed into involved. It was one of the first places I published my own research, and also the first journal I did peer-review for, participated on the editorial board of, and (albeit briefly) co-edited.

Sadly, I have to report this is also the first journal I’ve drifted away from. But I’ve thought hard about this, and it’s (mostly) not the journal, it’s me. As much by chance as by intention, over time I became more broadly interested in the discipline of geography, which this journal is somewhat a part of, but mostly is not. I’ve also remained more faithful to complexity science and its concerns than the journal has – at least that’s my impression. I attempted to verify this claim by doing some keyword analysis across paper abstracts over time (see Figure 1). My impression doesn’t really withstand scrutiny. The journal’s top keywords across all issues since 1982 (which rather mysteriously is when Scopus’s indexing starts) based on my inexpert analysis have been remarkably consistent over time. The term ‘model’ in particular is among the top three most frequent keywords in abstracts in every 5-year period since 1984. Other regulars include ‘planning’, ‘system’ and ‘land’, although the first two of these have declined substantially since the 1980s. Given the journal’s original title, ‘design’ and its attributes are surprising laggards (not even among the top 100 keywords). Keywords reflecting the topics that first drew my attention to the journal, such as ‘complexity’, ‘cell’ and ‘agent’ are notably absent before 1994 and peaked around 2010, but again lie well outside the top 100. A recent high-flyer has been ‘network’ which perhaps speaks to one direction in which much complexity-flavoured thinking has headed. The surge in the last decade in the use of ‘city’ and ‘data’ confirms the extent to which Planning & Design really has become Urban Analytics & City Science.OPEN IN VIEWER

At any rate, the change in emphasis, already underway but also underlined by the change in the journal’s title, seems to me to signal something deeper. Planning speaks to an essentially modernist confidence in our ability to design, manage and control the environment. Analytics and science seem – at least in theory – to respect the essential unpredictability of urban environments, the unpredictability which led to so many spectacular failures of modernist planning (Jacobs, 1961; Scott, 1998). Analytics and science suggest an attempt to understand how cities work, which is surely to be welcomed. However, I’m not so sure that contemporary confidence in our ability to wield ‘data analytics’ effectively in order to understand and manage urban environments is warranted. Complexity and its close cousin chaos recognised that unpredictability, unintended consequences, ‘butterfly effects’, emergence, and self-reinforcing positive feedbacks should be central to any attempt to get to grips with cities.

Of course, like everything else, these dynamic, processual features of urban environments – for that matter of all environments – can only be observed through data, and the analysis of data is central to detecting, characterising and ultimately, if it is considered desirable and possible, managing environments. However it is important to recognise the many limitations of ‘data’ in relation to the challenge of understanding essentially dynamic systems. This is a topic I explore at some length in Computing Geographically (O’Sullivan, 2024), so I won’t belabour the point here, but I consider it essential that our analytical approaches recognise the fluid, relational, and transient nature of the environments that can only be rather crudely captured in data. There is no end to data: it is always possible to ask for (and to need) more and to get lost in the data without necessarily coming any closer to knowing how to create better environments. Questions beget questions.

What seems missing from analytical approaches focused on data is the more speculative future-oriented perspective that the planning and design perspective of the original journal brought to bear. After my initial enchantment with urban complexity and chaos in the pages of this journal, it was the not necessarily very scientific design-oriented perspectives (shape grammars, space syntax, generative designs and so on) that appealed most. Design surely must be informed by data and by science, but like cities themselves, design is a profoundly human activity as much informed by intuition grounded in experience as by data. The current emphasis on analytics and science has left behind some of that sense of the journal speaking to audiences coming to cities as artists and scientists.

I’m guilty here of some sweeping generalisations. Hasty analysis of keywords is no substitute for closely engaging with the many fine examples of urban analytics and city science published in this journal in recent years. What is abundantly clear is that ‘B’ remains a huge asset to all of us with any interest in the complexities of cities whether from geography, planning, architecture, city science or elsewhere: long may it continue. I might have drifted away a little, but I still love Environment and Planning B!