01
Audit how computers are used now.
Count the worksheets that were digitized, the "one and done" projects and the corner kiosks.
- Evidence
- Theory and Practice.
- Caveat
- Digitizing paper isn't harmful in itself. It just "maintains things as they are."
Experimental · always changing
Ideas Lab
Dr. Bryan P. Sanders · Children, computers, and powerful ideas
“What can a human and a machine construct together that neither had fully specified before they began?”
Build it. Enter it. Notice what happened. Change it.
- teacher
- researcher
- builder
- collector
- inventor
- SLOAFer
Bryan's ideas, built from Bryan's own writing. Every quote below comes word for word from one of his texts, and every source is linked or named. This is a working lab, not a peer-reviewed publication, and it grows as the work grows.
Dr. Bryan P. Sanders has spent more than two decades asking one question in different forms: what happens when the computer stops delivering school and starts helping learners build it? The ideas on this page come from his 2019 dissertation on Critical Techno Constructivism, his 2021 chapter Could Minecraft Be a School?, his 2022 essay on purposeful play, his 2025 essay introducing co-active emergence, and his current Sunshine Machine research on Persistent Computing Environments.
RE:EDUCATION, LLC
re:think re:create re:design
Who is Bryan Sanders? · Guiding principles
- I nurture a learning environment where students create more than they consume.
- It is my belief that we have not yet fully harnessed what teachers and students can do with computers as part of the teaching and learning cycle.
- Kids still need help and guidance to use technology to become builders, creators, storytellers, and composers — they are not born knowing how.
- Hands-on activities are essential (building with blocks or tools, digging in the dirt) and screens cannot replace them.
- Relationships with parents and other students matter most — reframe using technology as something that happens inside of strong relationships.
Dr. Bryan P. Sanders, bryansanders.com /teacher/
Turtle path
Lineage
- Dewey / Freire / Papert
- Critical Techno Constructivism
- experimental classroom practice
- Minecraft learning environments
- Purposeful Play
- STEAMHAMLET
- Engaging with AI
- Co-active Emergence
- persistent computing environments
- WorldSpec + Turtle Charter + Turtle Terraria
- TurtleBlock AI
Card catalog · 14 drawers
The experiments, 1995 to now
Pedagogical experiments
01 · 1995
Music + Literature Summer Enrichment
02 · 1995–2022
The Experimental Classroom
03 · Undated
Literary-Character Fantasy-League Database and Embodied Play
04 · Mid-2010s
Hot Tub Fever Recursive theater
05 · 2016–present
STEAMHAMLET
06 · 2021–22
Minecraft as Persistent Learning Environment Purposeful Play
07 · 2023–24
GPT Designed to Help Build Better GPTs
08 · 2024
Structured Database for the How of Building GPTs
09 · 2024–26
Public-Document Ingestion and Persistent Research Databases
10 · 2025–26
Sunshine Machine
11 · 2023–25
Iterative Generative Writing Toward Co-active Emergence
12 · 2022–present
RE/EDUCATION
13 · 2026–
TurtleBlock AI Field Tests
14 · 2026–
Turtle Terraria
Order and years as listed in the TurtleBlock AI experiments timeline. The fantasy-league entry carries no year there.
Provenance
Every quote is Bryan's own words as written, linked to its source. Unpublished drafts are labeled as such.
How to read the quotes. Text marked 2019 draft comes from a March 2019 working draft of the dissertation. The final, approved version is at LMU Digital Commons, and its wording and page numbers may differ. Page numbers given for Dewey, Freire and Papert belong to those authors' books, as Bryan annotated them in Chapter 4 of the draft.
A · Concept codebook
Six ideas, each in plain words and Bryan's own.
Each entry gives a plain definition, the dimensions that make up the concept, what counts and what doesn't, one anchor quote, and the year the idea first appears in Bryan's writing.
01 · First appears 2019
Critical Techno Constructivism (CTC)
A learning theory for digital learning environments. It combines constructivism (learners make meaning by doing), connectivism (learning happens across networks of people and information) and critical theory (learners question power). Under CTC, students use computers to think, make and question, rather than having computers deliver predetermined answers to them.
"Using computers as the tools that we learn with, never at, this negotiated space encourages students to inquire deeply, question broadly, and fashion new ideas."
2019 draft, Ch. 3, "Propose New Learning Theory"
Dimensions, counts as, doesn't count as
Dimensions
The 2019 draft lists seven precepts, each paired with a question a teacher can ask:
- Personal Inquiry. Did the student develop the learning task?
- Open-ended Problem or Question. Did the student arrive at an answer that led to more questions or problems?
- Technology as Tool. Did the student use technology in the thinking process?
- Formative Demonstration of Learning. Did the student demonstrate learning throughout the process?
- Reflection as Learning. Did the student demonstrate a reflective approach in the formation of knowledge?
- Social and Cultural Critique. Did the student demonstrate a critical awareness of the larger established modes and forms of thought that shape thought?
- Sharing and Collaborating. Did the student actively seek out peers in person and online in the process of acquiring knowledge, testing theories, and creating an artifact?
Current wording (TurtleBlock AI taxonomy): Personal Inquiry · Compelling Problem or Question · Technology as Tool to Think With · Formative Demonstration of Learning · Reflection as Learning · Social and Cultural Critique · Sharing and Collaborating.
Counts as
- Students set the task.
- One answer opens further questions.
- The computer is part of how students think.
- Learning shows up throughout the work, not only at the end.
- The work questions whose knowledge and whose defaults shape it.
Doesn't count as
- A computer "along the wall or in the corner".
- "One and done" projects or once-a-week computer lab time.
- Digitized worksheets.
- Drill software built on predetermined outcomes.
- Hands-on making with no social or cultural critique. That is techno-constructivism without the critical.
First appears
2019. The draft names it at the end of the Chapter 2 literature review: "Critical Techno Constructivism may be the correct name."
02 · First appears 2025
Co-active Emergence
A way of working with generative AI in which humans and machines take part in an ongoing dialogue ("co-active"), and ideas and solutions develop out of that exchange ("emergence") instead of being set in advance. The human keeps judgment and authorship throughout.
"The newly coined term co-active emergence tries to describe a collaborative, interactive involvement (co-active) coupled with the natural development of ideas and solutions (emergence)."
GPT and Me, An Honest Reevaluation, Impacting Education 10(1), 2025, p. 96, doi.org/10.5195/ie.2025.479
Dimensions, counts as, doesn't count as
Dimensions
- Dialogue loop. Human prompts, the machine responds, the human evaluates, revises and returns.
- Emergence. Outcomes that neither side would reach alone.
- Human discernment and oversight at every stage.
- Equity and customization. Tools shaped to local contexts, with bias named and audited.
- Teacher as facilitator of inquiry, not a deliverer of content.
Counts as
- Asking the machine "what do you think of this in relationship to..."
- Brainstorming with a GPT, refining the ideas with peers and instructors, then returning to the GPT.
- A student who directs the machine's output and judges it.
- A doctoral student who moderates machine suggestions before acting on them.
Doesn't count as
- "Write an essay for me about..."
- A copy/paste model.
- Mainstream transactional uses such as chatbot tutors or virtual teaching assistants.
- Treating AI as an authority or an "infallible oracle".
- Banning AI outright, which Bryan argues "ignores the dialogue off campus".
First appears
2025 in print. Working drafts are earlier.
03 · First appears 2026
Persistent Computing Environment (PCE)
A computing environment that can stay with a learner over time. It keeps questions, artifacts, revisions and conversations available so they can come back into later inquiry. It does this without taking ownership of where the learning goes next.
"what might happen if a computational environment could remain with a learner over time, preserving enough continuity to support inquiry without quietly taking ownership of it?"
The Sunshine Machine, opening section, working draft
Dimensions, counts as, doesn't count as
Dimensions
- Continuity over time.
- Recoverable traces with provenance. Who said what, and when.
- Resurfacing. Earlier work returns as material for new work.
- Negotiated participation. The learner can accept, adjust or refuse the machine's help.
- Revisable purpose. Remembering something does not give it authority.
Counts as
- An environment where earlier questions, failed ideas, human corrections and machine interpretations stay distinct but can meet again.
- An environment where the learner can contest or bracket what the system "thinks" it knows about them.
Doesn't count as
- Plain AI memory or personalization. A Corpus with a Porpoise says a PCE "should not be conceived merely as" either one.
- A finished technology or "a claim of solved pedagogy".
- Persistence without governance, which "can simply become a better teaching machine" (Sunshine Machine Essay Assembly Method).
First appears
2026, in the Sunshine Machine and TurtleBlockAI working documents. Its roots run back through Minecraft (2021) and CTC (2019).
04 · First appears 2019
Microworlds and powerful ideas
A microworld is a bounded but open space, on a computer or off it, where learners can experiment, build, share and debug. A powerful idea is one that "not only enhances understanding of complex concepts but also stimulates creative innovation" (microworlds works-in-progress proposal). Both concepts come from Seymour Papert's Mindstorms (1980).
"the work people do with computers creates microworlds of the world outside of the computer."
Could Minecraft Be a School?, 2021, §17.3, p. 389
Dimensions, counts as, doesn't count as
Dimensions
- A bounded exploratory space.
- A feedback cycle. Make, share, take critique, return to "the laboratory table" to debug.
- Interdisciplinary assemblages. Bryan argues against Rieber's view that a microworld must be domain-specific.
- Powerful ideas as the content. Examples are body syntonicity, state, systems, cycles, feedback and emergence, with randomness and debugging next.
- Beyond the screen. Block play and imaginary play count too.
Counts as
Turtle Geometry, block play, imaginary play, robotics and Minecraft, when students are engaging with ideas and connecting new concepts to what they already know.
Doesn't count as
- A world where the teacher scripts every move. As the 2021 chapter puts it, "There is no better way to ruin a game or a project with students than to teach them every step or move or reaction of a newly encountered set of playthings."
- Treating a microworld as neutral. It is still a designed place with assumptions (A Corpus with a Porpoise).
First appears
2019 draft, through Bryan's annotations of Papert. It appears in his own argument by 2021. The microworlds proposal is undated.
05 · First appears 2021
Minecraft as school
Using Minecraft as the learning environment itself, "a way to think", rather than as a supplement or a room students visit, "a place to go". The curriculum grows out of students' play, questions and conflicts over long periods of time.
"the discussion of incorporating Minecraft receives a treatment that presents it more as a place to go instead of a way to think."
Could Minecraft Be a School?, 2021, §17.2, p. 388, doi.org/10.1007/978-3-030-75142-5_17
Dimensions, counts as, doesn't count as
Dimensions
- Daily immersion, not a once-a-week hour.
- Exploratory play first.
- A "curriculum of ideas" that grows from student dialogue and questions.
- Conflict as material, worked through with restorative practices.
- Recursion between inside and outside. Questions tracked in the game lead to work outside it, and that work comes back in.
Counts as
- Classes that play daily and track their ideas and questions.
- Teachers who join the game without needing to be experts.
- Texts and discussions placed inside the world.
- Long collaborations, possibly over years, across classrooms and subjects.
Doesn't count as
- Digitized worksheets.
- Using Minecraft to substitute for "previously planned lessons and activities".
- The once-a-week "one and done" Minecraft computer class.
- Any arrangement where an educator or researcher "dictate[s] or predetermine[s] exactly what lessons or actions should happen".
First appears
2021.
06 · First appears 2022
Purposeful play
Treating student play as serious learning. Students' own questions, prompted by a shared world, organize the work, and the teacher steps in "at relevant moments" with guidance, information, lessons and questions.
"Students and teachers enter a dialogue and co-construct the class curriculum."
Purposeful Play – Educating with Minecraft, Minecraft Education blog, August 22, 2022
Dimensions, counts as, doesn't count as
Dimensions
- "Hard fun". A balance of pushing the mind while staying entertained.
- A shared, unscripted starting world, which leads to natural small groups.
- A student story that emerges: "a story in motion".
- Teacher facilitation at relevant moments.
- A co-constructed curriculum.
Counts as
The redstone-and-fish-dock exchange in the essay. Students choose where to build a city, the teacher asks how it will be powered, and students raise their own question about keeping the fish population alive.
Doesn't count as
- Unstructured free time with no teacher dialogue.
- A teacher-scripted lesson that happens to run inside a game.
First appears
2022, the dated Minecraft Education version. An earlier MinecraftEdu-branded version exists.
B · Dewey · Freire · Papert
How Bryan reads them.
In the 2019 draft, Bryan carries out a critical interpretive document analysis of three books: Dewey's Democracy and Education (1916), Freire's Pedagogy of the Oppressed (1970) and Papert's Mindstorms (1980). The table summarizes each thinker's position as Bryan annotates it. Page numbers refer to those books as cited in the draft.
| Concept | Dewey (1916) | Freire (1970) | Papert (1980) |
|---|---|---|---|
| Learning / constructivism | Learning happens "through the intermediary of the environment" (p. 26). Growth is an end in itself, not preparation for something else (pp. 55–56). | Problem-posing replaces "banking". Content is "constituted and organized by the students' view of the world" (p. 109). | "Anything is easy if you can assimilate it to your collection of models" (p. vii). Debugging and "false theories" are how children really learn (pp. 101, 132–133). |
| Liberation / agency | Education should "liberate the young from reviving and retraversing the past" (p. 79). Rejects "externally imposed ends" (p. 117). | "Freedom is acquired by conquest, not by gift" (p. 47). Liberation comes through critical discovery and action. | Computers as "carriers of powerful ideas and of the seeds of cultural change" that can cross cultural barriers to science (p. 4). |
| Technology / tools | No computers, but schools need "agencies for doing, with tools and physical materials" (p. 44). Bryan's gloss: educational technology "is a series of pathways from multiple starting points to multiple ending points." | Says nothing about computers. Bryan applies Freire's critique to edtech: banking and "narration sickness" are "known pitfalls that have flagged Educational Technology." | The computer is "the Proteus of machines" (p. viii) and an "object-to-think-with" (p. 23). Against "inventing new gadgets to teach the same old stuff" (Papert, 1972, quoted in Ch. 2). |
| Role of teacher | "In such shared activity, the teacher is a learner, and the learner is, without knowing it, a teacher" (p. 167). | Resolve the teacher-student contradiction "so that both are simultaneously teachers and students" (pp. 72–73). "People teach each other, mediated by the world" (p. 80). | Teacher and learner engage in "a real intellectual collaboration," so "the teacher does not have to pretend not to know" (p. 115). |
Where they agree
All three reject learning as passive absorption. Bryan draws the line directly: Freire named "banking education", but "John Dewey wrote about it much earlier than Freire" (2019 draft, Ch. 4). All three put the learner's own activity, questions and world at the center, and all three make the teacher a co-learner.
Where they pull against each other
Optimism vs. critique of tools
Papert expects computers to cross cultural barriers. Bryan's critical-theory chapter warns that educational software has often treated "whiteness as the default", and that more hardware alone has widened the digital divide. A tool can open a world, and it can also carry the assumptions of the people who built it.
The individual mind vs. collective struggle
Dewey and Papert mostly describe how a person builds understanding. Freire frames learning as collective, political liberation.
Pace of change
Dewey calls reorganizing education "a slow work… piecemeal, a step at a time" (p. 144). For Freire, freedom is "acquired by conquest."
Boundary cases
Open-ended does not mean anything goes
Dewey: "It is equally fatal to an aim to permit capricious or discontinuous action in the name of spontaneous self-expression" (p. 108).
Drill is not the enemy in itself
"A behaviorist model does not preclude a student's self-driven inquiry, though it does not explicitly encourage it" (2019 draft, Ch. 2). The problem is building everything around it.
The tool is a model, not the answer
Papert called LOGO environments "too primitive, too limited by the technology of the 1970s. The role I hope they fill is that of a model" (pp. 181–182).
A teacher who knows less than the students
Papert's teacher "does not have to pretend not to know". In Minecraft, students' knowledge "far exceed[s] their teachers'", and that power shift may deter some educators (2021, §17.1).
Bryan's synthesis: Critical Techno Constructivism
Bryan takes Dewey's experience and growth, Papert's construction with computers and Freire's critique of power, and adds Siemens's connectivism. The result is a theory in which students use computers to pursue their own questions and to question the systems around them, and in which teachers treat what students produce as "marks of mastery in place of externally imposed outcomes" (2019 draft, Ch. 3).
In the draft's words: "These three theoretical approaches can function together to create a new Critical Techno Constructivism for future classrooms" (2019 draft, Ch. 2). Here "these three" means constructivism, connectivism and critical theory.
C · Minecraft as school
Concepts and evidence.
The four schooling terms, as Bryan uses them
The 2021 chapter uses these terms but doesn't formally define them. The working definitions below are drawn from how the chapter uses each one.
Student-centered
Learning organized around the student's own interests, voice and purpose. The chapter places Minecraft in "a thoughtful and powerful subsection of constructivist learning that focuses on student-centered experiences with games and projects." Its opposite is lockstep, predetermined outcomes.
Inquiry-driven
The students' questions set the curriculum. Bryan describes "a fully immersive gameplay approach that might follow student inquiry and choice to co-create the curricula." In this approach the questions arise through play, and the teacher helps pursue them.
Project-based
Making things, with the microworld feedback cycle of build, share, take critique and debug. Bryan's sharpest line: schools "often hold hostage the most engaging and transformative projects in their planned curricula" and quarantine "student-led work in specialty courses."
Game-based
A game environment used as the learning environment. Minecraft's creative (sandbox) and survival modes can both be modified.
How they relate
Where they align. All four oppose planned curricula with predetermined outcomes.
Where they conflict. Game-based learning as most research describes it is "supplementary lessons or companion experiences". That is game-based but not inquiry-driven. Bryan names this split: Minecraft treated "more as a place to go instead of a way to think."
Boundary case. Survival mode has built-in game objectives. Bryan counts both modes as open enough to be modified, so a game's own goals don't by themselves make it teacher-predetermined.
Evidence map: studies Bryan cites in his Minecraft writing
| Source | Type | Population / context | Design | What it found, per Bryan's text |
|---|---|---|---|---|
| Harel & Papert (1990) Software design as a learning environment | Case study | Students learning traditional math material with Logo. Age and setting not stated. | Experimental group (Logo, interactive computer environment) vs. control group (traditional methods) | The experimental group scored higher on a classroom examination. Bryan notes that "Thirty years later... this remains an open question for some educators and policymakers." |
| Ming (2020) | Small study, conference proceedings | Year Five students learning probability (the age group comes from the reference title). Country not stated. | Pretest/posttest with Minecraft as an experiential intervention, plus a qualitative reflection component. Sample size not stated, but described as small. | Better performance on a paper-and-pencil test. Claims about confidence and curiosity are "tempered by the sample size." Bryan calls for large-scale replication "across continents." |
| Baek, Min & Yun (2020) | Literature review | 28 articles (2012–2019) from the ProQuest database on Minecraft in education | Review | Most articles focus on elementary school, language arts, and playing the game rather than building or modifying it. "significantly more qualitative than quantitative studies." Use tends to be supplementary. Time and "curriculum inflexibility" are barriers. Bryan's Table 17.1 summarizes the benefits and experiences it lists. |
| Xu, Chen, Eutsler, Geng & Kogut (2019) | Scoping review | 59 articles (2000–2018) on digital game-based learning for English language learning. Not specific to Minecraft. | Review | Nearly 80% reported a positive impact on English language acquisition. About 75% had samples under 100. Recommends standardized tests aligned with commercial games. Bryan reads this as favoring traditional measurement. |
| Lane & Yi (2017) | Book chapter (conceptual) | Minecraft as a learning environment | Not an empirical study | Ties Minecraft to Papert's constructionism: knowledge "constructed by learners rather than simply being told." |
| Stager & Martinez (2019) Invent to Learn | Book (practitioner) | Making and tinkering classrooms | Not an empirical study | "quality work takes time, disobeys bell schedules... might not have any impact on test scores." |
| NAEP (2020) | National assessment data | U.S. students at grades 4, 8 and 12 | Not stated | Score gains since 1992 are slight and not sustained. Bryan uses this as context for why standardized approaches have not shown their worth. |
Comparatively well studied: elementary grades, language arts, playing (rather than building or modding), qualitative accounts, and Minecraft as a supplement to planned lessons.
Thin evidence and open questions
- Large samples and replication. Ming's probability gains have not been replicated at scale.
- Experimental and quantitative designs for Minecraft in particular.
- Secondary grades, and subjects beyond language arts.
- Building and modifying the game, as opposed to playing it.
- Full immersion. No cited study tests Minecraft as a school, with daily, longitudinal, inquiry-driven use. Bryan's proposal is a design argument, not a tested intervention.
- Measurement fit. How do you document learning that happens "at rates that might not be possible to document" without pulling it back into worksheet-style testing?
- Equity. Private schools may have more freedom to try this than public schools. Who gets access?
- Teacher knowledge. What happens when students know the environment better than their teachers do?
D · Practitioner guides
Plain-language steps for educators, school leaders and doctoral programs.
Each step carries a note on how strong the evidence behind it is and a caveat.
Evidence labels used below
- Theory argued from document analysis.
- Practice Bryan's classroom and workshop experience.
- Limited studies small or few empirical studies.
- Preliminary early interview themes.
Guide 1: Critical classroom technology From the 2019 dissertation draft · 10 steps
02
Name one outcome you can let students define.
Swap an externally imposed outcome for one students shape.
- Evidence
- Theory (Dewey on imposed aims).
- Caveat
- Graduation and testing requirements still apply. Start with one unit.
03
Let students develop the task.
Use the first CTC question: Did the student develop the learning task?
- Evidence
- Theory.
- Caveat
- Some students will need scaffolding to choose well.
04
Use technology to think, not to deliver.
Programming, building, modeling and mapping, rather than quizzes.
- Evidence
- Theory (Papert's object-to-think-with), with small historical studies cited in the draft.
- Caveat
- This takes more class time than delivery does.
05
Treat errors as bugs.
Ask "How can I fix it?" instead of "right or wrong".
- Evidence
- Theory (Papert).
- Caveat
- Grading systems may still punish visible mistakes, so protect drafts.
06
Assess along the way.
Look for learning throughout the process, not only at the end.
- Evidence
- Theory.
- Caveat
- This needs record-keeping time that most schedules don't provide.
07
Build in reflection.
Students explain how their thinking changed.
- Evidence
- Theory (Dewey, Freire).
- Caveat
- Reflection can turn into a ritual if it is always the same prompt.
08
Add social and cultural critique.
Ask whose voices, defaults and assumptions are built into the software and the topic.
- Evidence
- Theory (critical theory).
- Caveat
- Needs care and trust, and the teacher's own assumptions are part of what gets examined.
09
Share and collaborate, in person and online.
- Evidence
- Theory (Dewey, connectivism).
- Caveat
- Online sharing raises privacy and safety questions for minors.
10
Check against the seven precepts.
Use the CTC questions as a self-review rubric, not a scorecard.
- Evidence
- Theory.
- Caveat
- The precepts were proposed for others to test and have not been empirically validated.
Expert review note
This guide is drawn from a theory-building dissertation (document analysis), not from a classroom trial. It should be reviewed by practicing teachers and an educational researcher before it is used as formal professional development.
Guide 2: Minecraft-based schooling From Could Minecraft Be a School? 2021 and Purposeful Play 2022 · 10 steps
01
Start with play, not instructions.
Drop students into a new world and let them explore and learn from each other.
- Evidence
- Practice.
- Caveat
- Some students need a gentle on-ramp.
02
Learn the game, but don't wait to become an expert.
- Evidence
- Practice.
- Caveat
- Students may know more than you, which can feel uncomfortable at first.
03
Watch for the questions.
Note what students ask, about mechanics and about values.
- Evidence
- Practice.
- Caveat
- You'll get a lot of questions quickly, so you need a system to capture them.
04
Step in at relevant moments.
Add guidance, information, a short lesson or your own question, as in the redstone example.
- Evidence
- Practice.
- Caveat
- Your questions also steer the work, so offer them as invitations.
05
Keep a running log of ideas and questions.
Classes that "tracked their ideas and questions" find new things to explore.
- Evidence
- Theory and Practice.
- Caveat
- Logging takes routine and time.
06
Connect inside and outside.
Move questions into reading, writing, math and science work outside the game, then bring them back in.
- Evidence
- Theory.
- Caveat
- This needs coordination with other teachers or subjects.
07
Place texts and discussions in the world.
Scatter readings through the game, and let debates grow from real decisions, such as building too close to a classmate.
- Evidence
- Theory.
- Caveat
- These texts still need to be chosen with care.
08
Use conflict as curriculum.
Rebuild together and practice restorative communication.
- Evidence
- Theory (proposed, not studied).
- Caveat
- Real conflicts need adult support.
09
Go daily, if you can.
An hour a week "goes so quickly it will not be a terribly productive or meaningful experience."
- Evidence
- Theory and Practice.
- Caveat
- Daily time is the hardest thing to win in a school schedule.
10
Study your own classroom.
Bryan suggests asking your principal for a self-study of students working together in Minecraft.
- Evidence
- n/a (method).
- Caveat
- Get approvals, and protect student privacy.
Evidence strength overall
The benefits listed in reviews (Baek et al., 2020) are mostly from qualitative studies, elementary school and language arts, and the only cited outcome gain on a test (Ming, 2020) comes from a small study. Full Minecraft as school has not been tested.
Guide 3: Community-based learning in doctoral leadership courses From the 2019 Uncharted Territory analytic note · 8 steps
These are preliminary themes from interviews in progress, and the note calls itself an "early attempt to brainstorm." They are not study findings.
01
Plan around the known pain points.
Interviewees raised compressed time schedules, worry about "walking away" from the volunteer site, and doubts about how the site connects to the course, the program and their dissertation.
- Evidence
- Preliminary.
- Caveat
- These themes come from an unstated number of interviews.
02
Protect what works.
Interviewees valued more space and time to practice transformative leadership, engaging with communities "in real space and time", and the chance to create change.
- Evidence
- Preliminary.
- Caveat
- Positive themes may reflect who agreed to be interviewed.
03
Spread community-based learning across all three years.
The note proposes "a few units throughout the three years" focused solely on a community-based learning project, rather than a single course.
- Evidence
- Preliminary (a proposal).
- Caveat
- Requires redesigning the program, not just one course.
04
Teach networking early.
As students start forming research questions and literature reviews, teach them to network across the university, the alumni database and their local communities.
- Evidence
- Preliminary.
- Caveat
- Networking support needs staff time.
05
Help each student build a "squad of professionals".
These people can guide the dissertation and become entry points to a volunteer project.
- Evidence
- Preliminary.
- Caveat
- Access to networks is unequal, so plan support for students without strong ties.
06
Link the site to the dissertation.
Address the relevance concern directly, so the community project and the research inform each other.
- Evidence
- Preliminary.
- Caveat
- Not every dissertation topic has a natural community site.
07
Plan for continuity rather than walking away.
Aim for projects that can "last through graduation and beyond."
- Evidence
- Preliminary.
- Caveat
- Long commitments must stay ethical and mutual for the community partner.
08
Put the locus of control in students' hands.
Let doctoral students lead the choice and shape of their community work as part of the program's arc.
- Evidence
- Preliminary.
- Caveat
- Student-led doesn't mean unsupported, so pair it with advising.
Status
This guide is a starting point for discussion. It should be checked against the completed study and any published version before anyone treats it as a recommendation.
E · Open questions
The lab's research agenda.
These are from Bryan's working list for "The Sunshine Machine, a dissertation continued," quoted verbatim.
- Is PCE genuinely new, or is it a recombination of older ideas?
- Does persistence change the epistemology of a learning environment, or merely its convenience?
- What does generative AI threaten that those earlier environments protected?
- Can a machine be an intellectual participant without becoming an authority?
- Who controls memory, representation, retrieval, and forgetting inside a PCE?
- How does a PCE avoid becoming a more sophisticated teaching machine?
- What happens to Freirean dialogue when one participant is computational?
- What happens to Papertian constructionism when the artifact can answer back?
F
Why build it this way
Built in public
TurtleBlock AI: the workbench
Constructivist AI agents for building, exploring, and iterating computational worlds in Minecraft and beyond.
TURTLE LOOP · TURTLE_CHARTER.md
THINK → TALK → REPRESENT → CONSTRUCT → ENTER WORLD → NOTICE / TEST / BREAK / QUESTION → REFLECT → REVISE / FORK / RESTORE / LEAVE UNRESOLVED ↺
- Turtle is not an answer machine, an evaluator, or a world vending machine.
- Build it and see.
- disagreement is data
- manual human edits are authored state
- memory supports continuity, not destiny
- noticing is not permission to interrupt
- Two turtles agreeing is still not peer review.
- Record stays. Object is optional.
- lossless before normalized
- gloriously weird meaning does not vanish because a schema lacks a field.
Research Library
The shelf
- Democracy and Education Dewey
- Mindstorms Papert
- The Children's Machine Papert
- The Connected Family Papert
- Pedagogy of the Oppressed Freire
- Lifelong Kindergarten Resnick
- LogoWorks
- Hamlet on the Holodeck Murray
- The Klingon Hamlet
- Invent to Learn
- Teach Yourself Esperanto
- On the Road Kerouac
- Waiting for Godot Beckett
- Saussure
- Trees of Santa Monica
- American Favorite Ballads Pete Seeger
Turtle notebook
What the turtle noticed
Provisional machine interpretation · written by Hark from Bryan's writing · disagreement is data
-
Persistence showed up before it had a name.
The Experimental Classroom ran 1995–2022. In 2021 Minecraft was imagined as long collaborations, possibly over years. In 2026 it is named: the Persistent Computing Environment. The through-line may be continuity that a learner, not the machine, owns.
-
The 'critical' in CTC moved.
In 2019 it asked whose defaults live in the software. By 2026 it asks who controls memory, representation, retrieval, and forgetting. Same question, new machine.
-
Shakespeare + computers is a method, not a joke.
STEAMHAMLET, recursive theater in Hot Tub Fever, The Klingon Hamlet on the shelf, and 'Make, Walk Around In, and Change the Story' all treat a story as a world you can enter and revise.
Next Edge: When Turtle shows why initiative is on hold, how can it expose the conflict without steering the learner toward one resolution?
X-factor: whoooo knooooowwwssssssssssss
Bibliography
Publications
- 2016 STEAMHAMLET: A Transformative Situated Inquiry
- 2017 Virtual Learning Environments
- 2018 Towards a New Learning Theory for Digital Learning Environments
- 2018 Create Boldly. cateacherssummit.com/create-boldly-bryan-sanders
- 2019 Toward a Unified Computer Learning Theory: Critical Techno Constructivism. digitalcommons.lmu.edu/etd/901
- 2019 Tenets of CTC
- 2021 Possible Possibles. tynker.com/blog/articles/ideas-and-tips/possible-possibles
- 2021 Could Minecraft Be a School?. doi.org/10.1007/978-3-030-75142-5_17
- 2022 Purposeful Play – Educating with Minecraft. education.minecraft.net/en-us/blog/purposeful-play
- 2023 The First Thing I Did With A Computer
- 2023 Engaging with AI — A Radical Shift for the Future, Today
- 2023 STEAMHAMLET Is School 2051
- 2025 GPT and Me, An Honest Reevaluation: The Dawn of Co-active Emergence. doi.org/10.5195/ie.2025.479