Positional maps: A quick step-by-step guide

Positional maps are part of sociologist Adele E. Clarke's situational analysis method, designed as a remedy to analytic paralysis (when we feel cognitively overwhelmed by a complex situation, like it happened to me here).

Their goal is to help the analyst see more clearly and discuss the messiness of a situation by detaching people and institutions from fixed positions (a “position” here meaning a viewpoint or a stance).

The method was developed as a tool to help researchers generate theory, but it works just as well as an experiment in itself, with no pressure to produce theoretical outputs or polished results.

How to do a positional map?

The basic idea is to build a graph with two axes, like any graph. But instead of plotting numbers, you will be plotting positions on an argument.

Step 1: Pick a debate that you want to understand better

OK, so this debate is your “situation.” For example, I took the question of how to protect kids and teens from social media harms.

 Or…

Step 2: Collect your data

Search, read, talk, collect information. For my experiment, I combined desk research with a video call with Estelle Pannatier from AlgorithmWatch CH, who generously walked me through some dimensions of the topic. I used DEVONthink to organize the documents (webpages, PDFs, notes), but any tools you usually use should do the job.

Step 3: Turn that tension into two axes

That’s a difficult thing to do. And you might find your real axes only later in the exercise. So don’t think too much about them at the beginning. Ideally, they should be axes of difference or controversy that you identified in your data; things that are attracting more vs. less of a certain concern. In my case, I tried these axes:

• x axis: need for social media regulation (more vs. less)
• y axis: target of intervention (users, people vs. platforms infrastructure/architecture)

I also tried:

• x axis: locus of intervention (institutional, legal vs. informal)
• y axis: construction of youth (fragile victims vs. capable partners)

Step 4: Plot the positions (not the people)

Once you have two axes you’re comfortable with, start plotting the positions you’ve found in the data onto your graph. Remember, you’re not plotting people (“the government is pushing for X” or “teachers argue for Y”). You’re writing arguments, stances, regardless of who holds them. For example, on my map it was “Ban it by law,” “Let the market regulate itself,” “No smartphones at school,” etc.

To help you with this step, you can try what Clarke calls a “situational map” first. Just list as many things as you can about the situation: people, objects, thoughts, documents, values, actions, whatever, regardless of their category. The more you can dissect the raw material, the faster the messy initial situation will start separating into distinct positions.

Step 5: Unlock some clues for your analysis

At this stage, you might start to see things more clearly. For example, to what extent are the positions carried by different actors? Does an actor hold onto several positions? The exercise can be repeated over time to observe how positions evolve, whether some disappear, appear, or are silenced. What are the relations between these positions? Do they clash, ignore or complement each other?

Step 6: Look for the empty spaces

Once you've plotted the positions that are being voiced, you can also ask: what positions are logically possible here, but nobody is actually occupying them? Silence on the map can be data. It might mean a perspective is being actively silenced, that it's unthinkable within the current discourse, or simply that no one has articulated it yet; this might become an opening for new argument or advocacy. On my own map, I started to identify some missing positions, but didn’t go far in analyzing what they could be. My axes didn’t feel solid yet, and I’d rather continue with step 7 before coming back to it.

Step 7: Repeat

Once you’ve done a first circle, come back to the beginning and repeat, collecting more data, finding new positions, refining your axes, etc. Or maybe you already feel unstuck, and move on.

Oh, and remember, this method doesn’t have to be strictly followed, just like I took some liberties here. After all, the goal is to find a way out of analytical overload; forget about methodological perfectionism.

Wanna know more?

If you’re interested in knowing more, this website built around Clarke's books is a good place to start, and maybe a chance to try mapping a tangled situation of your own.

Image reference

An iterative mapping exercise; the image above is the result of the first cycle of analysis. To be continued…