Scrambled Bloom’s, Rhetorical Prompting, and the Creative Pulse of AI Writing
Still Bloom's, Still Bottom Up
When I first sketched out the “Scrambled Bloom’s” pyramid earlier today, I was intrigued by how shaking up the familiar order of cognitive skills could shine a spotlight on the recursive dance we do when writing with AI.
This evening, I was playing in the LinkedIn sandbox. My GSU colleagues1 posted about a presentation they gave that included an inverted Bloom’s Taxonomy. It was raining outside, so I needed something to occupy my neurodiverse brain. I wanted to see what a writing studies context might look like if we flipped Bloom’s taxonomy on its head and treated the higher-order skills, like evaluation and analysis, as the foundation of writing with AI. The more I iterated on that visual, I realized each stage—evaluating, analyzing, applying, comprehending, remembering—feeds back into itself. We don’t simply move from low-level tasks to high; we loop, pause, and loop again, constantly refining our prompts and deepening our understanding of how rhetorical choices shape meaning. This insight led me to scramble Bloom’s Taxonomy and combine it with my Rhetorical Prompting Method and Ethical Wheel of Prompting, forging practices that honor both critical thinking and creative flow.
In writing studies, we’ve long known that writing is not a linear process. We evaluate, then revise. We apply new strategies, then re-evaluate. We remember what’s worked in the past, remix it, and try again. What generative AI makes visible, especially when we treat it as a collaborative writing tool, is that creation happens through this cycle, not after it.
So in this post, I’m bringing together three threads:
A reimagined Bloom’s Taxonomy that starts with Evaluate and ends with Remember
My Rhetorical Prompting Method (RPM), which guides how we teach prompt engineering through rhetorical awareness
My Ethical Wheel of Prompting, which ensures AI-assisted writing remains fair, responsible, and human-led
And running through it all? Creation. Not as a capstone but as the ambient behavior that threads each moment of this recursive practice.
🧭 Evaluate
Just like other writers, I always start with the rhetorical situation: Who’s the audience? What’s the purpose? What’s the genre? From there, I craft an initial prompt for ChatGPT (or any other LLM) and evaluate what it gives me. Did it get the tone right? Are the facts sound? Does it align with my intention as a writer?
💡 This is the same kind of critical judgment Bloom placed at the top of his 1956 taxonomy—now flipped to the foundation, where it might be helpful in a writing process.
🛡️ Ethical Check:
Is the output free of bias or harmful assumptions?
Am I being transparent about how I’m using this AI tool?
🔎 Analyze
Now I take the AI’s output and analyze it like I would a student draft. What’s its structure? What rhetorical moves does it make? Is it using emotional appeals (pathos), credibility (ethos), or logic (logos)? Where are the gaps?
🛠️ This step trains writers to read like rhetoricians—critical, strategic, and attuned to nuance.
🛡️ Ethical Check:
Am I responsibly flagging misinformation or flawed reasoning?
Am I aware of the AI’s limitations and blind spots?
🛠️ Apply
Based on what I find, I apply a new strategy in my next prompt. I might adjust the tone, request a more inclusive example, or ask for a different genre. This is where prompting becomes a form of revision.
🎯 In Bloom’s terms, this is where we take what we know and use it in a new context. In rhetorical prompting, it’s where precision and experimentation meet.
🛡️ Ethical Check:
Am I being clear and purposeful in my prompt design?
Have I accounted for privacy or contextual sensitivity?
🔁 Re-Evaluate
After the second round of output, it’s time to re-evaluate. Did my changes make the draft better? Am I getting closer to what I need? I often ask students to compare drafts side-by-side—this is where growth becomes visible.
🧠 This is metacognition in action: writers becoming aware of how they think, revise, decide.
🛡️ Ethical Check:
Did I edit for usefulness, relevance, accuracy, and harmlessness?
Have I taken ownership of the final product?
🧠 5. Remember
Throughout this whole process, I’m constantly remembering what worked before: phrases that triggered better responses, rhetorical moves that helped the AI stay grounded. I take notes. I reflect. I build a personal knowledge base of what makes for an effective prompt.
📚 This isn’t just memory for the sake of testing—it’s practical, experiential learning, rooted in context and reflection.
🌟 Creation Is the Atmosphere
In this model, “Create” isn’t the final step. It’s the air we breathe at every stage. It’s why I had the word tiled in the background of the infographic. We’re not working toward creation—we’re creating as we go. Every prompt, every revision, every reflective pause is part of the composition.
Prompt engineering, done rhetorically and ethically, isn’t just a technical skill. It’s a writing practice. A recursive, reflective, deeply human one.
👩🏫 Want to bring this method into your own writing practice? Let’s talk.
💬 Reply here or drop me a message—I love collaborating on how we teach and use AI with integrity and imagination.
Works Cited
Bloom, Benjamin S., et al. Taxonomy of Educational Objectives: The Classification of Educational Goals. Handbook I: Cognitive Domain. David McKay Co., 1956.
With thanks to Michelle Kassorla and Eugenia Novokshanova.