I became interested in qualitative research long before I decided to study HCI. I did my undergrad in cognitive science, a magical place where science and humanities collide and give rise to big arguments about how to “properly” study the mind. It’s this space of total confusion, because it’s all about us. It’s 2019; we have all these fabulous tools for studying the things around us (plants and animals and even our physical bodies…) and created by us (literature and art and culture…), and we’ve been studying these things for thousands of years and we’re pretty good at it. But then, we turn to the human mind, that which we cannot see or touch but somehow is more real to us than anything else on the planet…and we are just trying to hack away at it using all these existing tools, but nothing is good enough.
Ta-da! that’s my undergrad in a nutshell. This hard problem about studying the mind, born in cognitive science, is also my theme of choice in art, design, and human-computer interaction (which I’m studying now).
I wrote my undergrad thesis (selfishly) about the structure of the self-concept, and how we can understand a multi-faceted sense of identity (i.e. feeling like your sense of self changes broadly in different situations). The takeaway? It varies for everybody, but we are all tremendously influenced by context. I’ll spare you the details, but after a large introduction of explaining why traditional scientific tests don’t work for self-concept research, the final chunk of the paper was dedicated to broad overview of some alternative, qualitative methods that might be more useful. Some of these examples included text analysis of internet searches, arts-based research via drawing and performance, and themes in Buddhist psychology.
And now, HCI at CMU. I know it’s called Human-Computer Interaction for a reason, but I came into this program with a healthy dose of skepticism about technology and will continue to push back on the idea that technology has to be everywhere, all the time. In an effort to seek some alternate perspectives outside the HCI bubble, I kind of stumbled into futures studies through two short courses, one with Stuart Candy and the other with Peter Scupelli, where I have been introduced to long-term thinking as a practice and given an array of tools to help us get there.
So, after this long-winded intro, this brings us to CLA (Causal Layered Analysis).
In thinking about how this tool applies to my design practice, I’m referencing a few helpful papers; their full citations are at the bottom of this page.
Causal Layered Analysis is “originally a futurist’s theory and method,”¹ but it’s overall design makes it applicable to a wide range of community-based (i.e. qualitative) research. It provides an “assessment of deeper individual and collective processes” which help create an understanding of social context that is necessary in human-centered research. Breen et al. call it contextualism, “the notion that people are not separable from context, [which] is contrary to lay understandings of what it means to be an individual.”
This contradiction explains why qualitative research methods a whole are slow to catch on. “Unfortunately, positivism remains the dominant scientific epistemology for the social and behavioral sciences, despite a longstanding critique concerning its applicability to understanding the complexities of social and community phenomena.” (a more eloquent way of saying what college-senior me took about 30 thesis pages to say).
So, CLA was developed to help combat the “fundamentally linear approach” between past and future that scientific positivism has assumed for generations. CLA instead gives us a “holistic consideration of complex social issues,” which is not just good for futures studies, but for all of social & behavioral sciences.
The 4 causal layers are as follows.
Litany — WHAT WE SAY. Surface-level events, contexts, and behaviors. Think newspaper headlines or direct observations.
Social/causal — WHAT WE DO. The relationship between people and settings, social systems, and structures.
Worldview discourse — HOW WE THINK. Perspectives, values, meanings, positions.
Myth/metaphor — WHO WE ARE. Underlying cultural aspects — archetypes, stories, symbols.
I won’t get too deep into this; if you want to understand this better, check out the references.
In general, the first two layers are the ones we tend to be the most comfortable with. For example, I like climbing, and I might say “I spend a lot of my free time at the climbing gym” (litany). Why? “Climbing makes me feel strong, and it provides a community of climbers — and women — that makes me feel empowered.” (social/causal). This seems like an insight, no? We usually say “wow, so deep,” pat ourselves on the back for our “aha” moment, and the conversation stops right there.
Case study: women’s participation in roller derby.
Breen et al. does a case study on women’s participation in sports. This particular case is about roller derby in Australia (quite specific, I know), but it was the first time I have connected futures thinking with the broader case which I care about deeply — women in the outdoors.
If I wanted to explain this fully, I would probably bore you to tears, but I want you to keep reading, so I’ll keep it succinct:
- Gender inequality is a thing
- I’ve spent some time working in the ski industry, where gender inequalities also exist, and manifest themselves in really interesting ways.
- To me, the outdoor industry is a tangible space where I can actually have an impact in the greater, more daunting realm of gender inequality.
- I also love skiing, so that helps.
In this roller derby case study, the researchers took interview transcripts with athletes and broke them down line-by-line. They categorized the lines into the four CLA layers, and then drew out themes to create a narrative.
This line-by-line analysis of interviews is something we already do in design research. We’re off to a good start.
And then, their analysis gets to the point of why CLA might be so useful. They recognized the same point I explored above in my climbing example — “if the intent was to explore mechanisms to support women’s participation in sport, a less complex analysis may have led us to the conclusion that supporting the development of a sense of community within leagues and exploring women’s empowerment might be options.” But this isn’t enough; CLA gets us deeper.
“CLA enabled the identification of more complex issues pertaining to broader cultural attitudes regarding the role and construction of women.”
And that’s huge. As I’ve experienced it, most work in the “women-in-the-outdoors” realm tends to focus a lot on that second layer — on the “community/women’s empowerment” level. Which is great, don’t get me wrong, but the problem is we have gotten stuck there. The term “women’s empowerment” has become increasingly saturated to the point where it may be losing meaning. When used by companies, it often transforms into a superficial marketing tactic (“we support women’s empowerment, buy our stuff because we love women!”) Used by women, it can sometimes calcify new stereotypes in place of the old ones we are battling against (“we are empowered and we are angry!” might feed into the negative stereotype that “feminists are scary” and discourage conversation).
So, this is a place where I could seriously see myself bringing in CLA to dig deeper and ultimately create a more productive dialogue. I’m not saying CLA will be the perfect answer, but it’s incredibly satisfying to find more tools like this in the space of qualitative research.
Sohail Inayatullah states two different applications of CLA — “the first is the analytic type wherein CLA is used as a research framework. The second is workshop-based wherein CLA is used in an interactive participatory environment.”² For example, this second type can be done with an organization’s stakeholders “to produce a more effective policy or a more inclusive vision statement.” The first type, as used in the roller derby study, is awesome on its own. But I am imagining a situation in which I could guide a team through an exercise to collectively explore the experience of women in the outdoors. Even as an informal framework for a meeting, this could be super powerful.
Through Terranova’s “aging workforce” case study³, I also learned that you do not have to step through CLA in a linear fashion, from litany to myth. The conversation can begin wherever the participants are most comfortable — in her case, it was easiest for the senior executives to think at the systems level. After this, they examined the litany layer by speculating on headlines of the future (how cool is that?!). Most importantly, this interactive form of CLA can help to “turn the problem from a predominantly negative focus into a more positive one.” This is exactly why I see the potential in CLA to help frame more productive discussions about uncomfortable topics around race, gender, etc. “in a non-threatening and non-judgmental way.” That last part is huge. It provides just enough structure to coax us out of our comfort zones and begin the conversations that may ultimately “lead to the inclusion of different ways of knowing.”
The challenges of CLA are part of the reason I think it’s so cool.
CLA is a relatively new tool; like any other new tool, it’s hard to convince people of its credibility. Likewise, it “requires the researcher to think critically and in greater depth than what might be expected from some other techniques.”¹ How can the researcher control for their own biases when doing this?
We had this discussion often in Dexign Futures. It is an especially relevant feeling for us as university students who are using this method for the first time and are acutely aware of our lack of experience with CLA. How can we think deeply and broadly about hard problems when we are limited to our own experiences? This is a super valid question, not just in futures studies but in HCI. In design research, we encounter the same issues in navigating hidden biases. We accept this, and we do our best to fill in the gaps by working in diverse teams and triangulating with other research methods. Likewise, Breen et al. help put us at ease a bit by reminding us that “analytical proficiency comes with practice” — and we have to start somewhere.
Beyond the concerns of a novice researcher, we run into the challenge of facilitating a CLA conversation within a social structure that is inherently resistant to it — much like the challenge of facilitating qualitative research in a world where scientific positivism is basically a religion. Described by Inayatullah, CLA “can be challenging to those who do not wish to include the subjective in the objective world,” such as “empiricists who see the world as either true or false (who insist on being right instead of being located in layers of reality)”².
Why is this challenging to these people? Because it “deconstructs their view that they hold the keys to the gates of enlightenment.”
But that is exactly what we’re trying to do, and that is all the motivation I need.
1. Breen, L., Dzidic, P. and Bishop, B. (2016). “Causal Layered Analysis.” Chapter 11 of Handbook of Methodological Approaches to Community-based Research: Qualitative, Quantitative, and Mixed Methods ed. Leonard A. Jason & David S. Glenwick.
2. Inayatullah, S. (2009). “Causal Layered Analysis: An Integrative and Transformative Theory and Method.” In Futures Research Methodology by Jerome Glenn and Theodore Gordon. Washington D.C, The Millennium Project.
3. Terranova, D. (2004). “Fathoming the Ageing Workforce Debate: Causal Layered Analysis in Action.” Journal of Futures Studies 2004, 9 (2), 37–42.