Subways & datasets
My friend Amber got back to our mid-size, mid-east city from a trip to New York some weeks ago, and she was wondering what it would take for our mid-size, mid-east city to get a subway system. Curious about the correlation between city size and commuter trains, she looked it up and came across a report examining the relationship between subways and urban growth.
This report summarized the findings of two researchers, Marco Gonzalez-Navarro and Matthew A. Turner, whose 2016 study examined “the relationship between the extent of a city’s subway network, its population and its spatial configuration” (Gonzalez-Navarro and Turner, 2016). It’s an interesting look at these data, but what particularly caught my attention was the second paragraph:
The authors developed a unique data set describing all subway systems in the world. Their data includes information on underground, surface, and above-ground rail transit lines, and identifies the latitude, longitude, and date of opening for every station. To examine the effect of subways on urban development, they utilize data from satellite images of the world at night (drawn from NASA satellites), and then calculate “light intensity gradients” for each city to measure its centralization. Finally, they draw upon data for population changes from the United Nations.
Turner and Gonzalez-Navarro’s paper drew from lots of different areas of study by taking cross-cuts of pertinent data and combining them for a new perspective. I imagined each of these areas as being its own point on some kind of knowledge map.
But really, if I zoomed in on any point, each area is essentially a huge set of points that make up the topic.
New areas of study are created by selectively comparing points (within the dataset or between datasets), and the relationship between them to creates a network.
To reach this level of connectivity and innovation, we need to have built the datasets out considerably. Satellites didn’t appear: they were built after years of exploration into the physical realities of fire and the wheel, the theoretical realities of gravity and friction, and the metaphysical existence of the questioning human mind — which drives us to examine the inner limits of ourselves and the outer limits of the universe.
When Aristotle set himself the task of classifying all living organisms, he was really laying the foundation for the science of today. As incredible as his work was, it was thin and incomplete compared to the information (and information structures) we have now. Obviously: we only have that power because he effectually gave it to us. Without Aristotle’s work, I wouldn’t have the power to Google the bird that just landed on my deck — and be presented in 30 seconds with its mating habits, migratory patterns, average life span, natural habitat, and number left in the wild.
Data builds on itself. The thin and incomplete foundations we start with give rise to thicker, richer, better layers of understanding.
What I’m getting at is that simple data points, collected, connect to create slightly more complex relationships, which we can call data sets. Every relationship between data points and data sets can become its own point, giving rise to ever more points and sets and opportunities for discovery; and when these data sets become connected, they form a network.
And all of a sudden, we have an exponential number of possible connections and relationships to explore.
This study of subways has never been done before, probably for several reasons (like, it’s boring?). But really, only recently have all those sets of data become available. This study couldn’t have been done 100 years ago: we didn’t have the next levels of data points to look at because some of this stuff didn’t exist yet. It had to be developed.
At this point in the world, we have so many data points, sets, networks, and networks of networks of networks, that there are really an infinite number of relationships to explore and write niche studies on.
So we have the satellites; the calculations for the light gradients; the records of stations; the records of population; the geographical locations of the stations and of the rail lines; the servers to store all this information; the computers to parse it. We have all the millions of points contained in each of those things — the millions of thoughts and skills and ideas to develop and improve each piece, each function, each system. And now, we even have Gonzalez-Navarro and Turner cross-sectioning these points to combine them into a new network, describing new relationships that could become just one set, one part of a future exploration of— I don’t know, hmm, networks and connections between them?
Who knows! Anything is possible!
The data points, to sets, to networks start looking remarkably like brain neurons, which is interesting because that’s how intelligence grows: by making new pathways between neurons and then strengthening those connections. (Interesting that innovation and intelligence can be visually represented in a very similar way.) Every new connection offers the opportunity for more new connections, at an ever-multiplying rate.
This became a very personally applicable thing to me as I thought about the impact of networks in the context of design. AIGA recently released their “Design Futures” feature, analyzing trends and directions in the light of new technology and theory, and I found the observations fascinating. Trend 1: Complex Problems by Meredith Davis makes a strong case for our developing focus on designing networks; Davis’s report describes the increase in creating broad systems that reach across whole fields of study — relating psychology and traditional design and web development and science. This has happened, I think, as we move away from concrete design (print) into more abstract interfaces (for example, in-home AI). “Today’s design work extends to the design of services and communities of interest that interact through new models of communication,” says Davis. “The elements that make up a system are usually easy to see, but the relationships among them are often informal or invisible and require research.”
The future of design is this deep research applied, as designers “build connections across disciplines when design knowledge is insufficient for the problem at hand,” and promote collaboration between teams “comprised of experts from many fields” (Davis 2018).
From those system interactions, we’re moving to designing a lot of frameworks. “[Visual systems] work addresses problems not only at the level of components, but also at the level of social and technological systems through which diverse audiences and stakeholders engage in a variety of interpretive tasks,” says Davis. “More than a coordinated collection of visual elements (logos, typefaces, etc.), these systems communicate how an organization is structured, its position with respect to competitors and the larger culture, and its perceptions of the people who provide and use its services.”
As design becomes ever more multi-disciplinary, the production of any conceptual system of ideas, released to interact with the other systems, becomes — well hey — a network of systems, made up of smaller inputs.
And you know what you just did by sticking with me through all this? You just created a new network in your own head, of subway systems and urban development, neural pathways, and moving-beyond-graphic design.
The combinations are endless.
