by Chessie Alberti & WAW
Dancing Through Chaos
To deal with the increasing chaos, precarity, and uncertainty of rapid change in a higher education environment that is impacted by continually expanding generative AI tools, lately we have both been drawn to embodied practices which feel untouchable by AI: Somatic Experiencing and social dancing.
As Woodrós explained in prior posts (here and here), they have been studying Somatic Experiencing (SE), a particular type of supportive, one-on-one conversation.
Recently, Chessie started taking lindy hop classes and learning about the role of “lead” and “follow” in partnered dancing. Dance is also deeply somatic and embodied. Social dancing requires the ability to connect and respond to another human body in a one-on-one context, noticing the cues of a partner and responding.
On Being Embodied
So, we found ourselves reflecting: what can we learn from these partnered embodied practices, like dance and SE, that can help us understand what makes our 1-on-1 work go beyond what an AI large language model can provide to a specific person? After all, AI doesn’t (at this time) have a body. Humans are essentially nervous system mammals, and our need for other people in order to learn is why we have universities.
In the Office of Academic Support, our programs use a conversational framework called the WISE model. According to the WISE model, academic support conversations move through four stages: Welcome, Identifying goals and approach, Supporting the learner, and Ending with purpose. When we think about what human-centered learning support services have to offer in comparison with AI programs, it makes sense to consider our embodied experience within the WISE conversation cycle.
So, how do we consider WISE as an embodied practice?
Looking with Both “I”s
As we reflected and brainstormed ideas for this article, we kept coming back to the power of identifying an approach for each conversation. We realized that the I in WISE has two features: Identifying goals (the one we’re most familiar with) and Identifying APPROACH. Somehow, we realized, the “Identifying approach” piece of WISE kept falling out of our training materials or ending up in the S (Support the learner) part of the cycle.
When we “Identify goals,” we ask questions like: Why is the student at office hours? Why did your colleague want to meet with you? What is the student hoping to accomplish in their conversation with a peer coach or writing consultant? This way, we draw on the knowledge the person seeking support has about the challenge they are facing and the knowledge of the person offering support, working together to clarify the goals.
The second part, “Identify Approach,” is essential but we find it is easier to overlook. We typically have trained students on this component as a part of the “Support” phase, but believe it adds values to clearly root it in the “I” of “WISE,” and conceptualize the different shapes this “Identify Goal” and “Identify Approach” phase of the conversation might take.
“Im-Bodied” Idea Generation
It is here, as well, that we make distinctions between what AI can do and what is possible with a body, a “soma” – what somatic information and moves are available to us as social animals interacting with fellow mammals? The state of each nervous system entering the conversation and the extent to which the two people’s nervous systems can sync up in a productive way also impacts the approach that will be taken towards the goal. Maybe a conversation needs to start with a little venting, then move into problem solving or brainstorming, and then narrow the plan based on that. Maybe laughter is needed to dissipate tension so that deeper thinking can be possible.
These two areas intersect to help determine the most optimal conversation shape: Who knows more about the relevant part? How much deep thinking can be made available in the conversational space?
The combination of deciding the two “Identifies” is important, and there is also variety in the way the I and S might interact (Identify Goals, Identify Approach, offer Support). Some conversations are linear, following the progression we’ve outlined. In others, there is an ebb and flow between the different components – it might not be clear when identifying goals shifted into choosing an approach or offering support, as they do have interconnected components. Additionally, the model can be recursive, with several small cycles of identifying goals, identifying approach, offering support, and repeating, starting with identifying the next goal. That’s a little theoretical, so let’s put it in an embodied context.
Conversation as Dance
Actually–the shape of this conversation starts to feel a little bit like partnered dancing. At the beginning of the conversation, one partner takes the “lead” role, and one partner unfurls as the “follow.” In social dancing, being a good “lead” is less dictator-like than one might expect; the “lead’s” role is more about creating a “picture frame” for the “follow” to fill in. The “lead” offers cues that are responsive to the “follow” and within their skill level. A “follow” responds to the “lead,” filling in the space offered by each cue. In gender-inclusive “Everybody Leads, Everybody Follows” dancing culture, partners may switch roles depending on what they’re trying to practice–depending on the goals of their dance.
We believe the WISE model of 1-1 support conversations connects to this “Everybody Leads, Everybody Follows” culture, and Woodrós also sees connection to wisdom gleaned from the Somatic Experiencing model. Responding to the reality of what’s accessible and important in that moment and prioritizing based on that – still informed by deep topical or self-knowledge – allows a more compelling conversational “dance” to unfurl. Each person brings relevant, deep knowledge (whether self-knowledge or subject-specific knowledge) and uses that to create the necessary “picture frames” that will evoke a solution aligned with the goals. The nervous systems of these two social mammals inform, support, and enable this complex and multi-tiered exchange, mixing explicit and implicit knowledge sources to optimize the conversation for these two embodied individuals in this particular moment.
AI Not WISE Enough
This is so different from talking to ChatGPT! There, you have to be the “lead” and the “follow.” You have to define both the approach and the goals you want. You cannot receive coregulation from ChatGPT to support higher level thinking, but instead need to have prepared that ahead. It’s a different kind of conversation than the one that is possible with another person. Of course, it’s also quite complex to facilitate one-on-one support as a whole person with a nervous system of our own: ChatGPT isn’t going to have days where it’s less regulated and less able to effectively empower someone (but how good at empowerment is it to begin with, really?)
Perhaps a person under stress is more likely to fall into a “saviorist” mentality or tell someone “just go here instead, I can’t help you,” but AI is known to be sycophantic, offer hallucinations, or do the math wrong. If we human facilitators can remember that we are embodied and attend to the needs of our own nervous systems to support co-regulation, we can access our strength. Embodiment provides us the magic of the dance, and it also calls us to have the level of self-awareness that an experienced lead brings to that medium.
In conversation, we can also adjust to best support the other person to be as skilled and confident as they can be that day, supporting their agency. This might mean doing the conversational equivalent of switching roles so they can introduce a favorite dance step, setting them up for a solo spin, or supporting them in learning a new set of dance steps they’re stressed about and unsure of by providing firm direction and then gradually reducing the pressure of that guidance as they increase in confidence.
AI can’t dance like a person can, so it’s worth the effort it takes to have high quality conversations. One-on-one conversations can shift the world.