In 1974 the BBC handed a 32-year-old educational consultant named Tony Buzan a television series and a companion book, both called Use Your Head. In it he described a way of taking notes that ran outward from a central image in curving, coloured branches instead of down the page in lines. He called it a Mind Map.
The technique spread further than almost any study method in history. It reached classrooms, boardrooms, exam-prep courses, and eventually a category of software that has been bought, forked, open-sourced and acquired for the last thirty years.
Then something odd happened to its history. Almost every account of where mind mapping came from — including the ones ranking at the top of Google today — repeats a lineage that runs Porphyry → Leonardo → Buzan. Two of those three are unsupported, and the most rigorous figure in the story is missing entirely.
This is what the sources actually say.
TL;DR: Porphyry never drew a tree — he wrote prose, and the diagram is a medieval addition. Nobody has ever produced Leonardo's mind map. Joseph Novak's concept map (Cornell, 1972) predates Buzan's mind map (BBC, 1974) by two years, is theoretically grounded, and is the one AI systems actually adopted. The best-controlled trial of mind mapping found no benefit. Try Mind Map view free →
🗺️ Mind Mapping in One Diagram
Mind mapping has two lineages, not one. The popular branch runs Porphyry (c. 270 AD) to Buzan (1974) to mind mapping software. The rigorous branch runs semantic networks (1966–69) to Joseph Novak's concept map (Cornell, 1972) to today's knowledge graphs. Almost every published history describes the first branch. Modern AI systems adopted the second.
The thesis of this article sits in that diagram: the branch everyone credits is not the branch that won.
📜 The Myth That Will Not Die: Porphyry Never Drew a Tree
Porphyry of Tyre did not draw the Tree of Porphyry. He wrote the Isagoge around 268–270 AD as prose, and Wikipedia's Porphyrian tree article confirms no illustrations occur in editions of his original work. The earliest surviving drawings are ninth-century. The name is thirteenth-century. Nearly every history of mind mapping still credits him.
Search "history of mind mapping" and you will read some version of this: Porphyry of Tyre created the Tree of Porphyry around 270 AD, making him the first mind mapper.
He wrote a book. The book has no pictures.
"No illustrations or diagrams occur in editions of Porphyry's original work; diagrams were eventually made, and became associated with the scheme that Porphyry describes."
— Wikipedia, Porphyrian tree
The scholarly manuscript project The Art of Reasoning in Medieval Manuscripts, run by the Huygens Institute, is just as direct:
"Although there is no evidence that Porphyry himself came up with this schematic device, it's most commonly associated with manuscripts of his Isagoge."
— Huygens Institute
What Porphyry actually wrote, around 268–270 AD, was the Isagoge — a short introduction to Aristotle's Categories that describes a method of dividing a genus into species by successive differences. It is a procedure in prose. Boethius translated it into Latin before 510 AD and wrote two commentaries on it, and through those the text became the standard logic textbook of Europe for roughly a thousand years.
Who drew it first, then?
Here the honest answer is we are not certain, and that uncertainty is itself more interesting than the myth.
The diagram historian Jean-Baptiste Piggin makes the strongest case for Boethius, around 520 AD, on the grounds that Boethius' commentary invites the reader to look at an accompanying visual figure and then refers back to it as the diagram "above." The earliest surviving drawn examples sit in ninth-century copies of that commentary.
"It is conceivable that Porphyry, who was born about 234 AD, devised the arbor porphyriana himself, but there is no recognized evidence for this traditional assumption, so it seems safest to give credit to Boethius as the draughtsman of the figure."
— Jean-Baptiste Piggin
The name is later still. "Arbor Porphyriana" was coined by Peter of Spain in the thirteenth-century Summulae Logicales — as Piggin notes, "750 years after Boethius' death." The familiar naturalistic version, with a trunk and foliage and sometimes a human figure, is a mid-thirteenth-century manuscript convention.
So the honest sentence is: Porphyry described, Boethius probably drew, medieval scribes illustrated, and Peter of Spain named. Four people across a thousand years, compressed by modern blog posts into one wrong attribution.
❌ Myths and Facts: What the Top Results Get Wrong
Eight claims about mind mapping's history circulate on the first page of Google, and every one of them is wrong, unsupported, or misleading. Each row below pairs the published claim with what the primary source actually says: Porphyry's undiagrammed prose, Buzan's undated 1960s recollection, the refuted hemispheric premise, and a 10% recall gain whose confidence interval crosses zero.
Every row is a claim currently published on a first-page result, with the correction and the source. This table is the reason this article exists.
| The common claim | What the sources say | Status |
|---|---|---|
| Porphyry created the first mind map, c. 270 AD | He wrote prose. "No illustrations or diagrams occur in editions of Porphyry's original work." Drawings are ninth-century; the name is thirteenth-century | Wrong |
| Leonardo da Vinci used mind mapping | No folio, no historian, no example is ever cited. A standing public request from the mind-mapping community for one verifiable example was never answered | Unsupported |
| Allan Collins is "the father of modern mind mapping" | Collins and Quillian built semantic network models of human memory retrieval, tested by reaction times. It is a psychological theory, not a note-taking method | Wrong |
| Buzan invented mind mapping in the 1960s | The dated, documented public origin is 1974 — the BBC series and book. The 1960s date rests on Buzan's own later recollection with no contemporaneous artifact | Unsupported date |
| Mind maps work because they engage the right brain | Refuted. Nielsen et al. (2013, n=1,011): lateralisation is "a local rather than global property of brain networks" | Refuted premise |
| "Mind Map" is trademarked, so you cannot use it | The surviving general mark is Class 41 only — self-improvement courses. A 2006 UK attempt to extend it to software was opposed and did not succeed | Misleading |
| Concept maps are a variant of Buzan's mind maps | Separate lineages. Novak's programme began in 1972 from Ausubel's assimilation theory. The canonical Novak and Cañas paper never mentions Buzan once | Conflation |
| Studies prove mind mapping boosts recall | The most-cited study found a 10% gain whose confidence interval crosses zero — and lower motivation. A 131-student replication found no benefit | Overstated |
🔮 Ramon Llull and the First Machine for Thinking (1274–1308)
Ramon Llull (1232–1316), not Porphyry, is the genuine medieval ancestor of visual thinking, and almost nobody names him. The Majorcan philosopher published his combinatorial Art in 1274 and the Arbor Scientiae, sixteen trees of knowledge, in Rome in 1295 — two centuries before Leonardo and seven before Buzan, and with an ambition neither of them shared.
Llull built what he called the Art: a combinatorial system of rotating circles, triangles and squares for generating and testing propositions. The first version, Ars compendiosa inveniendi veritatem, appeared in 1274. In Rome between September 1295 and April 1296 he wrote the Arbor Scientiae — sixteen "trees" of knowledge, each with roots, trunk, branches, leaves and fruit, covering everything from the elements to theology.
Llull matters here for a reason that has nothing to do with note-taking. His diagrams were not records of thought. They were instruments for producing it — you turned the wheels and read off combinations. That is a different and more ambitious idea than either Porphyry's division or Buzan's recall aid, and it is the one that most resembles what AI agents do now.
🧬 Darwin's "I Think" (1837)
In July 1837, on page 36 of Notebook B, Charles Darwin drew a small branching sketch and wrote two words above it: "I think." It is the first record of common descent, and it is the most famous branching diagram in the history of science. It is not a mind map.
It is a claim about the world, drawn as a tree because the claim is tree-shaped. That distinction matters for everything that follows: a diagram is only as good as the relationship it encodes. Darwin's branches mean descended from. Buzan's branches mean, deliberately, nothing in particular.
🧠 The Science That Came First: Semantic Networks (1966–1969)
The rigorous line starts in a doctoral dissertation. In 1966 M. Ross Quillian submitted Semantic Memory at the Carnegie Institute of Technology, modelling word meaning as a network of nodes connected by typed links. Three years later he and Allan M. Collins published the reaction-time experiments that made the model testable — five years before Buzan reached the BBC.
That 1969 paper, "Retrieval time from semantic memory" in the Journal of Verbal Learning and Verbal Behavior, is the famous study where people confirm "a canary can fly" more slowly than "a canary can sing," because the network has to climb a level to find it. Collins and Loftus revised the strict hierarchy in 1975 with spreading-activation theory.
This is real cognitive science, and it is where the popular histories go wrong in a specific way: they credit Collins as an ancestor of mind mapping. He is not. He is an ancestor of the knowledge graph — the typed, labelled, machine-readable structure that turns up again at the end of this story. If you want that thread, we traced it separately in knowledge graphs for AI agents and in the history of agent memory.
🎓 Joseph Novak and the Concept Map (1972)
Here is the fact that reorders the whole history: the concept map is older than the mind map. Joseph D. Novak developed it at Cornell in 1972, two years before Buzan's BBC series, while running a longitudinal study of how children's understanding of science changes over time. He built it on David Ausubel's assimilation theory of meaningful learning.
He needed a way to see what a child knew at each stage. Out of that need came the concept map.
"Concept maps were developed in 1972 in the course of Novak's research program at Cornell where he sought to follow and understand changes in children's knowledge of science."
— Novak & Cañas, IHMC Technical Report
The mechanical difference from a mind map is one thing, and it changes everything:
"Words on the line, referred to as linking words or linking phrases, specify the relationship between the two concepts."
— Novak & Cañas, ibid.
Every edge is labelled. That turns each pair of nodes into a proposition — a statement that can be checked and can be wrong. A mind map branch says these two things are related somehow. A concept map edge says water is a solvent, and a teacher can mark that right or wrong.
Novak also allows what Buzan's rules forbid: multiple parents, and cross-links between distant regions of the map.
"Cross-links help us see how a concept in one domain of knowledge represented on the map is related to a concept in another domain shown on the map. In the creation of new knowledge, cross-links often represent creative leaps on the part of the knowledge producer."
— Novak & Cañas, ibid.
Learning How to Learn, written with D. Bob Gowin, followed from Cambridge University Press in 1984 and was translated into nine languages. In 1987 Novak began working with Alberto Cañas at IHMC, work that eventually produced CmapTools.
One detail is worth stating plainly, because it is the cleanest evidence that these are separate traditions rather than variants of one: the canonical Novak and Cañas technical report does not mention mind maps, or Buzan, a single time.
📺 Tony Buzan and Use Your Head (1974)
Buzan's contribution was not the diagram. It was the distribution. In 1974 the BBC gave the 32-year-old consultant a television series and companion book, Use Your Head — the first documented public use of the term "Mind Map." He codified the rules in The Mind Map Book (1993) with his brother Barry, and authorised one software product, iMindMap, in December 2006.
| Year | Contribution | Why it mattered |
|---|---|---|
| 2 June 1942 | Born, Palmers Green, Middlesex | — |
| 1974 | Use Your Head — BBC television series and companion book, prepared with the BBC Further Education Advisory Council | First public use of the term "Mind Map." A national broadcaster put a study method in front of a mass audience |
| 1970s–80s | Training, courses, licensed instructors | Turned a technique into an institution |
| 1993 | The Mind Map Book, co-written with his brother Barry Buzan | Codified the rules: one central image, one keyword per branch, curved lines, colour, imagery |
| Dec 2006 | iMindMap with Chris Griffiths | The only Buzan-authorised software, distributed through 1,000+ licensed trainers |
| 13 April 2019 | Died aged 76, Oxford, after a heart attack | — |
Two frequently repeated claims about Buzan deserve care.
The 1960s origin date. Several top-ranking articles date the invention to "the late 1960s" or "at some point in the 1960s." That rests on Buzan's own retrospective account of his student years. The dated, documented, verifiable public origin is 1974.
The Nobel Peace Prize nomination. This appears in his official biography and across profiles. It is structurally unverifiable: Nobel Foundation statutes seal nomination records for 50 years, and the Committee discloses nominees to nobody, including the nominee. Any qualified nominator can nominate anyone. A nomination is not an honour conferred. Report it, if at all, as an unverifiable claim.
⚖️ Mind Map vs Concept Map vs Spider Diagram vs Semantic Network vs Outline
Mind maps, concept maps, spider diagrams, semantic networks and outlines are routinely treated as synonyms. They are not, and the differences are mechanical rather than stylistic: how many parents a node may have, whether edges carry labels, whether the map is anchored by a topic or by a focus question, and — the consequence of all three — whether the thing can be wrong.
| Mind map | Concept map | Spider diagram | Semantic network | Outline | |
|---|---|---|---|---|---|
| Origin | 1974, BBC Use Your Head | 1972, Cornell | school convention, no single origin | 1966–69, AI + psychology | classical; via the Isagoge tradition |
| Key figure | Tony Buzan | Joseph Novak | none | Quillian, Collins | Porphyry → Boethius |
| Structure | one central image, radial curved branches, one parent per node | network, multiple parents, plus cross-links | central node, straight lines | directed graph of typed nodes | strictly nested list |
| Edges labelled? | No | Yes, always | No | Yes, typed | n/a |
| Anchored by | a topic | a focus question | a topic | a query | a thesis |
| Falsifiable? | Largely not | Yes | No | Yes | Partly |
Laid out side by side, the structural difference is easier to see than to describe:
MIND MAP CONCEPT MAP OUTLINE
──────── ─────────── ───────
┌─ branch ┌──[causes]──┐ 1. Topic
│ │ ▼ 1.1 Point
●─────┼─ branch [Node]─┤ [Node] 1.2 Point
│ │ ▲ 1.2.1 Sub
└─ branch └──[part of]─┘ 2. Topic one parent per node many parents, edges strict order,
edges mean "related" say WHAT the relation no cross-links
is, so they can be wrong
The practical consequence: a mind map is a fast, personal, divergent tool, and a concept map is a slow, shareable, testable one. Neither is better. They answer different questions. Michael Eppler's 2006 comparison in Information Visualization and Martin Davies' 2011 paper in Higher Education are the two peer-reviewed treatments worth reading if you want the full three-way distinction.
And here is the structural difference as a diagram — note that only one of these two can be wrong:
🧪 Does Mind Mapping Actually Work?
Probably, but by an unknown amount. The most-cited trial (Farrand 2002, n=50) found a 10% recall gain whose 95% confidence interval runs from −1% to 22%. The largest replication (D'Antoni 2010, n=131) found no benefit at all. The largest and newest meta-analysis (Dai 2026, 52 RCTs) reported very large effects and rated its own certainty of evidence "very low."
This is where most articles write "studies prove" and cite one paper without reading it. Here is the whole evidence base, with the caveats that make it honest.
| Study | Year | n | Finding | The caveat that matters |
|---|---|---|---|---|
| Farrand, Hussain & Hennessy, Medical Education | 2002 | 50 medical students | 10% greater factual recall at one week, 95% CI −1% to 22% | The headline number does not reach significance — the interval crosses zero. And "Motivation for the technique used was lower in the mind map group" |
| Nesbit & Adesope, Review of Educational Research | 2006 | 55 studies, 5,818 participants | Concept maps associated with increased knowledge retention | This is concept maps, not mind maps. "Significant heterogeneity was found in most subsets" |
| D'Antoni et al., BMC Medical Education | 2010 | 131 medical students | No benefit. Note-takers improved 0.21 SD while mind mappers declined 0.21 SD (p=.027) | Only 30 minutes of mind-map training — novices against habitual note-takers |
| Merchie & Van Keer, Contemporary Educational Psychology | 2016 | 5th–6th graders | Instruction in constructing maps improved strategy use | No significant free-recall gain. The benefit attaches to the teaching, not the diagram |
| Shi et al., Asia Pacific Education Review | 2023 | 21 studies | g = 0.67, a medium effect over traditional instruction | Small study pool; subject and level drive the result |
| Dai et al., Frontiers in Medicine | 2026 | 52 RCTs, 3,312 participants | Very large effects: theoretical knowledge SMD 1.45, procedural skill 1.68 | 98% of studies at high risk of bias. The authors' own conclusion: "overall certainty of evidence was very low" |
Those five results cannot be charted against one another, and that is worth stating rather than hiding. Each study reports a different quantity on a different scale:
| Study | Effect as the authors reported it | What that number measures |
|---|---|---|
| Farrand 2002 | +10% recall, 95% CI −1% to 22% | Percentage-point difference in factual recall at one week |
| D'Antoni 2010 | Note-takers +0.21 SD, mind mappers −0.21 SD, p=.027 | Change in score expressed in standard deviations |
| Merchie & Van Keer 2016 | Strategy use improved; free-recall gain not significant | A null result on recall, alongside a gain in strategy use |
| Shi et al. 2023 | g = 0.67 | Hedges' g, pooled across 21 studies, versus traditional instruction |
| Dai et al. 2026 | SMD 1.45 knowledge, 1.68 procedural skill | Standardised mean difference, pooled across 52 RCTs, certainty "very low" |
A percentage-point recall gap, a within-group shift measured in standard deviations, and two pooled standardised effect sizes are not the same unit and do not share an axis. Plotting them together would manufacture a comparison none of these authors made — which is the failure mode this section exists to correct.
The honest summary: mind mapping probably helps. The size of the help is unknown. The best-controlled trials are the least flattering, the largest and most recent meta-analysis rates its own evidence "very low certainty," and the mechanism Buzan gave for why it works is not the mechanism.
That is a less exciting claim than "studies prove mind mapping triples recall." It is the one the papers support.
The right-brain problem
Buzan's stated rationale — Radiant Thinking, engaging the whole brain — depends on hemispheric dominance. That premise did not survive.
"Lateralization of brain connections appears to be a local rather than global property of brain networks, and our data are not consistent with a whole-brain phenotype of greater 'left-brained' or greater 'right-brained' network strength across individuals."
— Nielsen et al., PLOS ONE, 2013 (n=1,011)
The myth is unusually sticky. A 2012 study of 242 teachers found 91% of UK and 86% of Dutch teachers endorsed hemispheric dominance as an explanation for learner differences — and that teachers with more general brain knowledge believed more neuromyths, not fewer.
A technique can work while its folk explanation is wrong. Both things are true here, and saying so is more useful than defending the folklore.
💾 The Software Era (1994–2019)
Mind mapping software began in 1994, when Mike Jetter wrote MindMan in Germany during leukaemia treatment. Free and open-source clones followed with FreeMind in 2000, the web and real-time collaboration arrived with XMind (2006) and MindMeister (2007), and the category ended by being absorbed: Corel bought MindManager in August 2016, and iMindMap gave way to Ayoa in 2019.
| Tool | Year | Founder | The detail worth knowing |
|---|---|---|---|
| MindMan → MindManager | 1994 | Mike Jetter | Written in Germany during leukaemia treatment, in weeks of isolation. Mindjet founded 1998 with Bettina Jetter; sold to Corel August 2016 with ~16 million users |
| FreeMind | 2000 | Jörg Müller + contributors | Started 7 May, first ran 25 May; the earliest release on SourceForge is dated 27 June 2000. Made mind mapping free and cross-platform |
| XMind | 2006 | Shenzhen / Hong Kong | Open-sourced XMind 3 in November 2008; won Best Commercial RCP Application at EclipseCon 2008 |
| MindMeister | 2007 | Hollauf & Vollmer | Built on a gap they identified in 2006: mind mapping had been popular since the 1970s, and no collaborative online tool existed |
| iMindMap | Dec 2006 | Chris Griffiths with Buzan | The only Buzan-authorised software. Sources disagree on the year — Wikipedia says December 2006, its own successor's marketing says 2007 |
| Freeplane | 2007 | fork of FreeMind | The still-maintained continuation of the codebase |
| Miro (as RealtimeBoard) | 2011 | Khusid & Shardin | Renamed after Joan Miró in 2019; $400M Series C in January 2022 |
| Coggle | 2013 | Cambridge, UK | Three graduates frustrated by the hierarchy of text documents |
| Ayoa | 2019 | Chris Griffiths / OpenGenius | Cloud successor to iMindMap, which was sunset with support through 2022 |
The arc is the same one that ran through spreadsheets and HyperCard: a single-purpose desktop application, then a free clone, then a collaborative web version, then absorption into something larger.
About that trademark
The widely repeated claim that "Mind Map" is trademarked and off-limits is misleading. The surviving general registration is narrow — Class 41, covering courses in personal and intellectual awareness and self-improvement. In 2006 an attempt was made in the UK to extend the mark to computer software and other classes; it was opposed on genericness grounds and did not succeed.
One caveat, stated because the rest of this article is sourced and this part is thinner: the detailed account of that opposition comes from a single first-person report by a named participant, and I could not retrieve the IPO decision document to confirm it. If trademark specifics matter to you, check the UK IPO and USPTO registers directly.
🤖 The AI Era: The Machines Chose Novak
Between 2023 and 2026 three things genuinely changed, and all three landed on Novak's side of the family tree. Generation flipped direction, so the map is now a model's output rather than a person's input. Branching became the reasoning substrate through Tree of Thoughts (NeurIPS 2023) and Graph of Thoughts (AAAI 2024). Retrieval became graph-shaped through Microsoft's GraphRAG, open-sourced in 2024.
A fourth thing gets asserted constantly and did not happen. Each in turn.
Generation flipped direction. The map used to be the input a person made. It is now an output a model produces from a PDF, a URL or a video. Xmind acquired Chatmind in May 2023 and rebranded it Mapify in June 2024; AI generation now ships in most major tools, including AI flowchart makers and PDF-to-mind-map converters.
Branching structure became the reasoning substrate, not the display. Tree of Thoughts (NeurIPS 2023) reframed problem-solving as a search over a tree of intermediate thoughts. Graph of Thoughts (AAAI 2024) generalised that to arbitrary graphs. The diagram stopped being how a person looks at the thinking and became how the thinking is done.
Retrieval became graph-shaped. Microsoft open-sourced GraphRAG in 2024: a model extracts entities and relations into a knowledge graph, which is then traversed to answer questions flat retrieval cannot reach.
The clearest artifact of the convergence is a 2023 paper literally titled MindMap, which prompts a language model with a knowledge graph and reports that the method "elicits the mind map of LLMs, which reveals their reasoning pathways based on the ontology of knowledge."
Here is the irony worth sitting with. Every one of those AI systems uses labelled, typed edges. That is Novak's concept map, not Buzan's mind map. The academically rigorous cousin that the popular history skips is precisely the one the machines adopted — because a machine cannot do anything with an edge that means "related somehow." It needs the edge to say what the relation is.
And the thing that is not supported: claims that AI-generated maps improve learning. No study tests an AI-generated map against a hand-drawn one. Given that Merchie and Van Keer found the benefit attaches to the act of constructing the map, automating construction may remove the active ingredient. That is an open question, not a finding — and anyone selling you the opposite is ahead of the evidence.
🧬 What This History Means for How You Work
Three practical conclusions fall out of the record, and none of them is the one the folklore suggests. Format should follow the job rather than a self-diagnosed thinking style, because the 2013 fMRI work closed that door. Labelled relationships earn their extra cost the moment a map leaves your head. And switching views should never mean rebuilding the content.
1. The format should follow the job, not your identity. You are not a "visual thinker" or a "list person" — the 2013 fMRI work closed that door. Mind maps suit divergent capture. Outlines suit sequential argument. Concept maps suit anything you need to be checkable. Picking one format for life is the actual mistake.
2. Labelled relationships are worth the extra effort when the map has to leave your head. For personal recall, unlabelled branches are fine. The moment a teammate, a client or a model has to read it, an unlabelled edge is a guess. This is exactly why the AI stack converged on typed edges.
3. Switching views should not mean rebuilding. The reason mind mapping software fragmented into a dozen apps is that each one owned its own file format. Every switch was an export.
That last one is the problem Taskade was built around. A project is one set of nodes that renders multiple ways — List, Board, Table, Mind Map, Docs (Beta), Gantt Chart, Calendar, Actionsheet and Org Chart. You outline in List while you draft, flip to Mind Map to see the shape, and switch to Board to assign it. Nothing is exported, converted, or recreated — it is the same content seen differently. Gantt Chart, Actionsheet and Org Chart are on paid plans; the rest, including Mind Map, are on Free.
The second half is what Llull was reaching for in 1295 and could not build: a diagram that does something. In Taskade Genesis you describe what you need in plain language and get a running system — a project that holds the structure, AI agents that reason over it with persistent memory, and automations that act on it. Agents run on 15+ frontier models from OpenAI, Anthropic, Google, and open-weight providers, with Auto choosing per task, and connect to 100+ integrations. The map stops being a picture of the work and becomes the work.
Taskade is free to start, with no credit card. Pro is $10 per month billed annually.
🛠️ How to Actually Draw One
Five rules survive the evidence: Buzan's originals, minus the neuroscience that did not hold up, plus Novak's focus question. The gain in Merchie and Van Keer's 2016 work attached to the act of constructing the map, not to the finished artifact — which means a rough map you built beats a beautiful one you were handed.
- Start from a question, not a topic. Novak's focus question is the single most transferable idea in this history. "Marketing" produces a shapeless map; "How do we reach developers who have never heard of us?" produces a useful one.
- One keyword per branch. This is Buzan's rule and it survives, because a phrase closes off associations that a single word leaves open.
- Label the edges when anyone else will read it. Even sparingly.
→ causes →,→ blocks →,→ depends on →. - Let it be ugly first. Merchie and Van Keer found the gain comes from constructing the map. Tidying is not the part that works.
- Convert when you are done diverging. A map is a poor final artifact. Flip it to a list or a board once the shape is clear.
If you want to go deeper on the visual-thinking thread, we have the full history of Mermaid.js and diagrams-as-code, a guide to AI second-brain tools, a comparison of outliner apps, and the broader story of systems thinking — the discipline that asks what the arrows between your boxes actually mean.
🔗 Further Reading
Every factual claim in this article traces to one of the sources below. The primary documents matter most: Novak's own technical report states the 1972 origin, the Huygens Institute manuscript project dates the first drawn trees, and the FreeMind release archive gives tamper-evident software dates that marketing pages routinely get wrong.
Primary and scholarly sources
- Novak & Cañas, The Theory Underlying Concept Maps — the inventor's own account, free and authoritative
- IHMC, The Origin and Development of Concept Maps — the canonical 1972 origin statement
- Stanford Encyclopedia of Philosophy: Porphyry and Llull
- The Art of Reasoning in Medieval Manuscripts — Huygens Institute, on when the tree was actually drawn
- Jean-Baptiste Piggin on the arbor Porphyriana — the Boethius case, argued in detail
- Nesbit & Adesope (2006) — the field's foundational meta-analysis
- Nielsen et al. (2013), PLOS ONE — the study that retires the hemispheres claim
- FreeMind release archive — primary, tamper-evident software dates
Related reading on Taskade
- History of Mermaid.js: Diagrams as Code
- Knowledge Graphs for AI Agents
- The History of Agent Memory
- What Are AI Agents?
- History of Spreadsheets
- History of HyperCard
- Systems Thinking Explained
- Claude Shannon and the Birth of Information Theory
- Free AI Flowchart Maker: 13 Tools Tested
- Best AI Second Brain Tools
- PDF to Mind Map, Free
❓ Frequently Asked Questions
Twelve questions cover what people most often ask about this history. The short answers: Buzan named the mind map in 1974, Novak's concept map predates it by two years, Porphyry drew nothing, Leonardo's mind map has never been produced, and the memory evidence is considerably weaker than the marketing.
Who invented mind mapping?
Tony Buzan named and popularised it in 1974 through the BBC series and book Use Your Head. He did not invent visual note-taking, which is far older, and he did not invent the concept map, which Joseph Novak developed at Cornell in 1972 — two years earlier.
Did Porphyry create the first mind map?
No. Porphyry wrote the Isagoge around 268–270 AD as prose containing no diagrams. The earliest surviving drawn trees are in ninth-century copies of Boethius' commentary, and the name arbor Porphyriana was coined by Peter of Spain in the thirteenth century.
Did Leonardo da Vinci use mind maps?
There is no evidence. The claim circulates widely but no article citing it names a folio or a historian, and a long-standing public request for one verifiable example has never been answered.
What is the difference between a mind map and a concept map?
A mind map has one central topic, one parent per node, and unlabelled connections. A concept map allows multiple parents, adds cross-links, and labels every edge so each pair forms a testable proposition. That labelling is why a concept map can be marked right or wrong.
Does mind mapping actually improve memory?
Less clearly than usually claimed. The most-cited study found a 10% gain whose confidence interval crosses zero, and lower motivation. A 131-student replication found no benefit. A 2026 meta-analysis of 52 trials reported large effects but rated its own certainty of evidence as very low.
Is the left-brain right-brain explanation true?
No. Nielsen et al. (2013), using resting-state fMRI on 1,011 people, found lateralisation is a local rather than global property of brain networks. This refutes Buzan's stated mechanism, not necessarily the practice.
Is "Mind Map" a registered trademark?
The surviving general mark is narrow — Class 41 only, for self-improvement courses. A 2006 UK attempt to extend it to software was opposed and did not succeed. Check the UK IPO or USPTO registers directly before relying on specifics.
When was the first mind mapping software released?
MindMan, later MindManager, was written in 1994 by Mike Jetter in Germany during leukaemia treatment. FreeMind followed as the first widely used free option, with its earliest SourceForge release dated 27 June 2000.
Who was Tony Buzan?
A British author and educational consultant, born 2 June 1942 and died 13 April 2019. He introduced the mind map through Use Your Head in 1974, co-wrote The Mind Map Book with his brother Barry in 1993, and authorised iMindMap.
How did AI change mind mapping?
Generation flipped direction — models now produce maps from documents. More significantly, branching structure became the reasoning substrate through Tree of Thoughts and Graph of Thoughts, and retrieval became graph-shaped through GraphRAG. All of these use labelled typed edges, which is Novak's concept map rather than Buzan's mind map.
Is a mind map better than an outline?
It depends on the job. Maps suit divergent capture; outlines suit sequential argument. The most useful setup holds one structure and lets you switch the view, so the same content is an outline while you draft and a map while you explore.
What replaced standalone mind mapping software?
Increasingly, a view inside a broader workspace. In Taskade a project renders multiple ways, including Mind Map, over one set of nodes, so switching view never means exporting between apps.
Porphyry described a method and someone else drew it. Llull built a machine for thinking and nobody could run it. Novak labelled his edges and the machines eventually did. Buzan gave the technique a name, a television series, and the wrong explanation for why it worked — and reached more people than the other three combined.
The lesson that survives all of it is small and durable: a diagram is only worth what its arrows mean. Get the relationships right and the picture takes care of itself.
▲ ■ ● — Memory, Intelligence, Execution. Start free with Taskade Genesis →