"Ancient Australian Aboriginal Memory Tool Superior to ‘Memory Palace’ Learning"

Thank you for the tag. I would love to have time to reply to everything here, point by point, but I can’t.

I only read the study today. I know the Aboriginal educator, Tyson Yunkaporta. I quote him in Memory Craft. The original paper pointed to by Josh quotes me and cites two of my books, The Memory Code and Memory Craft. My most recent, written with Aboriginal co-author, Margo Neale and the National Museum of Australia (Songlines: the power and promise) would not have been available when these researchers wrote the paper.

Your comments, ideas and Chris’ annotations are incredibly valuable for similar research that I am involved in. Does anyone have any objection if I point my colleagues to this discussion? It is a public forum, but I would still like you to be happy about it.

It is really interesting to compare with the research at the University of Melbourne (same city as Monash University where the research for this paper was done). The Honours student I am co-supervising is looking at the very issues you raise - but the project is not fully defined yet - along the lines of asking if songlines are different from the method of loci, and what are the implications in tertiary education. I tend to refer to songlines as memory palaces without thinking it through enough. I am very happy to see my work refined. I think the major difference is the use of song, dance and art, but that can be done in a memory palace. Although academics are also talking about the way knowledge is added, layer upon layer, ever more complex, over a lifetime. Songlines are used virtually and physically.

I am currently grappling with the academic papers I am being sent for the research - psychology research jargon is not my forte! The planning is underway for a very large scale cohort and dealing with long term complex knowledge, but it will take years. The work cited in this topic is just the start of a major push in Australia to get memory systems into education in an appropriate and pragmatic way, because it has the added value of our relationship with our Aboriginal cultures. As I know Tyson Yunkaporta and my books are cited, I am very reluctant to make any critical comment on a public forum, especially when I am so delighted to see these ideas being researched.

We have concrete evidence of these methods being used for tens of thousands of years, so not very ‘new age’. I can see that some might consider the use of art and song as airy-fairy if they haven’t tried them, and that might lead to a ‘new age’ label. I haven’t struck any resistance in any of the workshops I have been running including people aged from 9 to in their 80s.

One question raised was about why this isn’t happening at a much younger age. It is - but not as formally. I have written up the experience I had with one school on my blog. They used memory boards (lukasa) for science topics. They used art for multiplication tables and a songline for history.

The songline went through the bush, and I think might have indicated one of the differences which may come out of the research at the two universities. In that case, as with my History Walks, the songline acts as a continuous narrative. As described in my blog, the students created a timeline, but it was at specific locations they added new civilisations. They then compared them and looked at the implications for our contemporary culture, so took an overview rather than each location remaining independent. They also used song, dance and narrative, but not art in that particular exercise. They ended up with a woodhenge, testing out my ideas of ancient monuments such as Stonehenge acting as memory palaces / songlines …

I am giving teacher in-services on Indigenous memory systems - but adding in medieval and ancient Greek and contemporary to the mix.

I think that you are all raising really valid issues, but we are just starting the research into using these techniques in education and I would not be comfortable rushing off responses to your questions given my public role in this research. The applications need to be evaluated, so this is really early, radical stuff.

I really wish I could answer everything in detail here, given how much I value the opinions of those on this forum. Thank you so much for your insights - they are hugely appreciated and, unless someone objects, will serve to refine future research as I will relay them to the research team I am working with.



Yeah, it’s like prepping two groups for a “London Trivia” where one group gets to actually go site-seeing and for a pub crawl with a local guide and the other group instead gets Google Maps, a Lonely Planet guide, and a local for a Q&A session.


I moved your link into this thread, since it’s about the same study. :slight_smile:

1 Like

Oh well… I’ll just delete it if it already got posted, just wanted to share…

1 Like

I’ve been thinking a lot about an issue raised on this forum and by academics - are songlines really any different from memory palaces and the method of loci? Until now, I’ve used the terms interchangeably, without thinking it through. I discussed it with the researchers at Melbourne University and my Aboriginal co-author for ‘Songlines: the power and primise’. This is my thinking out loud - not a concrete theory!

I am beginning to think that the significant difference is that with songlines, learning is always done in the physical ‘memory palace’ which is constantly revisited. It can be recalled from memory, but is encoded in place. For me, that is way more effective, but I have aphantasia and very poor visualisation, so it may not be as big a factor for others.

So recalling your childhood home can be a memory palace, but not a songline.

It was today as I was doing Chinese vocabulary that it struck me. I tried to add words using the locations from memory because it was cold, and I didn’t want to go out. I know each of the houses in the songline, but adding vocabulary is way way easier when I walk and do the learning in the physical space. I couldn’t do it from home.

I can’t explain completely without explaining the whole system for Mandarin and Chinese characters, which I’ll do soon on my website, learning from others who have tried the same thing, and adapting. It has taken a year using it everyday to refine. Today, one of the words I wanted to add started with the Mandarin syllable ‘jin’ (1st tone). In my Person-Action ( PA ) system, that is a J-jester who turns to gold (ingot). It needed to be linked at a particular house which represents the radical - the basis of the associated character.

Had I simply imagined the house, I could have added the word. But by going there, I could look for something gold. There was a metal hose reel I had never noticed before. It is now, in my mind, the golden tool of a golden jester. The other syllable hooked on easily with its own PA. Plus experiences and emotional reactions happen when out in the physical locations. I had to add words associated with exercise one day. The fit looking owners of the house came out on bikes. Jackpot!

I am adding a large number of words to many of the locations, each with the syllables, tones, meaning, pronunciation and radical for the characters encoded - and then a story for the rest of the components of the character. Learning in place - or on Country as my Indigenous colleagues would say - is way more effective because I can always find something suitable, which adds to the complexity of the location and gives me a useful link.

I’m not sure if that makes sense, and there is no reason memory palaces can’t be used the same way. In competition, I used palaces for which each location was a simple concept and, for obvious reasons, I didn’t visit to encode. But I now realise that for everything else, even the 1,000 digits of pi, I always visited the locations, encoding during my morning walk. It was just so much easier.

My use of the different terms now involves only using the word ‘songlines’ for memory palaces where the encoding is done on Country.

Does that make sense?



I’m with Juliet on the whole (non-)issue. As Ms. Capulet puts it:

A rose by any other name would smell as sweet

If memory serves, there have been posts here before about the difference: Roman Rooms vs Memory Palaces vs Memory Journeys vs Method of loci. They’re not standardized terms in the first place, so it’s basically just people saying “to me it feels like” this is what this should mean. I mean when you say:

A bit like me saying that I used to call everything a multiplication but now I make a distinction by calling it “five squared” instead of “five times five.” The fact remains 5^2=5*5, so unless it’s an academic discussion there isn’t really much of a difference there.

I think that other discussion made a distinction that a Memory Journey is somehow in sequence whereas a Memory Palace is not. Believe me, when I traverse a “memory palace,” I do so in order. However, that does not mean that I couldn’t just tell you what’s in a certain location.

I know what you mean. I’ve stored all ~120 words of Toki Pona right then and there during an afternoon walk and I’ve done the same with pi: Paris - 2,000 Digits Of Pi walking through a single Arrondissement each day.

I still wouldn’t make that distinction though, because learning doesn’t happen only in the moment, but also when your brain synthesizes the information during sleep and I don’t physically go back to the locations each time just to review either.

Well, it could be if your parents haven’t moved and you could still go back to your old room. :wink: By the way, I find that example kind of overused… I know it’s what beginners get to hear when it is first explained to them, but memory palaces hardly rely on nostalgia and require childhood memories of school buildings, etc. to work.

In summary, I see where you’re going here but there is no requirement to ‘fill’ your memory palace in memory. Of course, make that distinction if you like… but aren’t we just just talking memory palace + context dependent memory? I mean, play a song while you are encoding onsite and then play it again the next day when recalling at home and use state dependent memory rather than context dependent memory to strengthen the connection. Doesn’t the method you’re talking about involve song and dance anyways… so I could take two out of three (location being the issue) with me?

Interesting comments, bjoern.gumboldt. I would like to think it is your way because then my writing in my books isn’t as sloppy as I feel it is having made this distinction today. But I do feel a significant difference if I encode in the physical world compared with when I do it just in memory. I don’t find it much different for one or two items, but when I start to stack more complex information, I find a massive difference. But whether that impacts on the language we use is another question.

I’m co-supervising a postgraduate student as part of a research project where she is working on a literature review and looking for specific definitions. My Indigenous co-author told her that you can’t pin down Indigenous knowledge systems using academic terms. I have a feeling I am going to be caught in this debate for a while.

Thank you for your comments!


1 Like

I mean, that depends… here in Europe you’d ask for the vegetarian option if you wanted the non-standard dish. In a lot of places in India, you’d ask for the non-veg option if you wanted the non-standard one. The character “中” in the Chinese word for “China” (中国) means “middle” when used as a noun or “central” when used as an adjective. However, the prime meridian runs through Greenwich in England.

The German word for “meat” is “Fleisch.” The German word for “flesh” is also “Fleisch.” Not one of those cases where the one is feminine and the other is masculine either; simply the very same word. This does not mean in turn that Germans are okay with cannibalism. Add 1066 to the mix and all of a sudden English differentiates between meat on the table and livestock and the “cow” is suddenly “beef” when on a plate because of the French “boeuf.”

Then you got some Americans come up with a study where they want to see to what extent the grammatical gender influences a speaker’s perception of an object. I believe the word was “bridge” and they compared the German “die Brücke” (fem.) to the Spanish “el puente” (masc.) and noticed that German speakers used more “feminine” adjectives and Spanish speakers used more “masculine” adjectives when describing the characteristics of bridges.

Make of that what you will, but it’s interesting that people who speak a language that dropped their three genders a long time ago when Old English became Middle English come to these kinds of conclusions. I mean how about “the” water which happens to be “das Wasser” (neut.) in German, “el agua” (masc.) in Spanish, and “a água” (fem.) in Portuguese. Portuguese and Spanish share an 89% lexical similarity. Maybe they can run the study again and come up with an explanation as to why Spanish water is from Mars and Portuguese water is from Venus.

Long story short, both methods make use of our internal GPS system in the hippocampus… the rest is “just” details. It’s entirely up to you of course if you feel the need to eat “meat” or even “beef” or if “Fleisch” is good enough.

In which language? I mean the word “hippocampus” actually comes from Ancient Greek, but there it was still a horse-like sea monster and not part of the brain yet and just saying “because Frances Yates said so” isn’t a strong argument either. :wink: On the other hand, not all students speak Latin, Ancient Greek, or Biblical Hebrew to make sense of all that has been written.

I’ve seen some interesting arguments here before…

…and that wasn’t even the most complicated of Latin you could run into… either way, your postgrad is going to have her work cut out for her. Best of luck!

1 Like

I wanted to answer this one separately, because it didn’t have much to do with the other point.

I suppose you are already familiar with their findings but just in case you are not, you might want to look into the research that these good folks have done…

In 1971, John O’Keefe discovered the first component of this positioning system. He found that a type of nerve cell in an area of the brain called the hippocampus that was always activated when a rat was at a certain place in a room. Other nerve cells were activated when the rat was at other places. O’Keefe concluded that these “place cells” formed a map of the room.

We do have a couple of posts about it on the forum as well: Search results for 'place cells' - Art of Memory Forum

Which brings me back to memory palace + contextual memory, as already mentioned in my previous post. I for one however don’t think that I need a different word for my memory palace only because I walked through Paris when I did encode pi there; but if you want to have an easier academic discussion by having a separate word for it…

…but wouldn’t you change the definition of “memory palace” if you insisted on it being different because you are not allowed to physically go to the location? And if you are allowed to still visit the location… isn’t a songline a subset of the memory palace set, so a kind of memory palace?

Oh, now that is interesting! Maybe this is why I’ve been having so much trouble memorizing “Hamlet”. I put it into a place where I cannot go physically - a neighborhood around my former home. It is very hard to encode the information that way, and I have been struggling.

In contrast, the stuff I put in my current home or neighborhood is a lot easier to remember.

I guess I can add my one data point to that study. :slight_smile:


Are you actually in the physical space in your neighborhood when you encode the information? It could also be that you simply have more recent exposure to the locations, if you feel it’s easier when using either/or sitting at home.


Getting too difficult, and I am terrible at semantics. I would have seen memory palaces as a simplified version of songlines, but my Indigenous colleague has now pointed out that sometimes songlines are taught before the person goes on Country - that’s how they navigated: teach the songlines and you can then travel knowing where food sources and waterholes are. So my theory is falling down already. But the songline always involves movement (not always dance), song, narrative and a mesh of genres of information.

I am back to thinking that ‘memory palace’ is a good general term, understanding that they can be implemented in a variety of ways. The academic study of the MoL, songlines and the implications for education, is just starting, so I will keep having my ideas changed over the next few years, I expect. They are looking at fairly large cohorts for testing.

I also have students in various schools testing memory palaces, one doing a formal research project. I’ll get that report later in the year. But she is creating a palace around the school buildings and physically visiting locations. Different subjects take off through different doors around the central courtyard. I don’t have all the details yet.

Thank you so much for your input. I am adding it all to the mix being explored by the Melbourne Uni academics.


1 Like

How interesting, meepster. I am really valuing the inputs here. My thinking gets changed and refined with each comment.

1 Like

Yes, I mention the O’Keefe and Moser place cell research in my books. I don’t think I have taken it far enough yet, though. This is a hugely important development. I am struggling to keep up with all the related research, while also doing my own. It is so good that the topic is so popular.



I agree with @LynneKelly that studies into what structures these techniques as such is really just beginning.

Remember: ancient Egypt had many dynasties. Perhaps one day our times will be but a grain of sand in some other dynasty by some other generalized term, like “Ancient World Wide Web, First Dynasty.”

My hope is that one day - sooner than later - schools will have courses on memory.

My theory is that they should involve three themes:

History of memory, theory and practice.

Starting at around age 10 until high school age should be more than enough to create a revolution of shared knowledge and understanding that exceeds the Internet revolution by many magnitudes of order.

If you think that statement sounds dramatic, it is. But just do the exponential math and note how much more the human brain can do in terms of abstract thinking compared to computers. If only the human brain has remembered enough to run the calculations.


It is exciting to see ancient people knew things that we didn’t and we still don’t know many things. I wonder if they had ancient visualization techniques and other stuff. Also I didn’t see it is posted but there is interwiew of the researcher on YouTube.



Schedule to follow-up with Tyson Yunkaporta next week. He was the one who taught the Aboriginal techniques.

Hopefully we’ll have that second interview out by June 15th or 16th.

I’ll ask him what you’re wondering about, along with many other questions I have myself. It would be great if people compiled any others they have.


Hope I get you before you speak to him.
I remember seeing a program taking about memory it showed a memory study of school children in Australia playing Kims Game and Aboriginal children were brilliant.
I just wondered if he might have an idea of what they might be doing differently? Faced with 32 items how would he remember them.
I think there is an issue here for me with terms. A memory palace is a place real or imagined that you use as a peg board. Then you plan a journey in that space which you connect to what you are trying to remember. For me the PAO system, Memory Palaces and journeys are required to work together for this all to work.
I find it worrying that the place needs to be real however if we know that mentally practicing a sporting activity is actually better than more practice, is it maybe that you need to run through it more in your head? I am quoting the Basketball study were they actually found that the group that did shooting in their head rather than extra practice actual preformed better than the other 2 groups.


Sorry I missed this before recording with Tyson, @Minotaur. He did give a list of different approaches. I think we had a great conversation that will be a boon for the serious mnemonist with more than wordly concerns at heart. Wordly concerns are covered too.

There are terminological issues. The word “real” is one of them, since no Memory Palace is truly real in any solid way. But some people seem to handle using actual references better than others - I happen to be one of them.

But the ultimate “real” is space itself.

For example, anything we encode is experienced spatially - or at least so it seems to me.

Imagine, for example, a 00-99 PAO with no Memory Palace.

If you encode something with the sun - say 02 - then when you encode your 03 and 04 target image, these will be experienced in some kind of spatial relationship with one another. I don’t see how this could not be the case, or how mental rotation would work without some level of spatial reference.

1 Like

Linking information to a narrative and a place may help memory more than linking information to a place alone.

Ofcourse you going to remember more if you add more associations, mental effort and focus to it. Otherwise the memory palace wouldn’t have any effect in the first place.