This article was originally printed in Issue 427 (January/February) of the Australian Mensa magazine, TableAus.

Dr Lynne Kelly is a science writer and Honorary Research Associate at LaTrobe University. Her work focuses on orality and the mnemonic systems used by ancient and modern oral cultures around the world. She is most famous for her theory that Stonehenge served the purpose of a memory palace for recording and transmitting knowledge by Neolithic Britons. This was the thesis of her book The Memory Code. It’s a bold idea and one I was keen to explore. Lynne was kind enough to indulge me.

Daniel Kilov: Your book begins with the observation that individuals in non-literate cultures frequently display an almost encyclopaedic knowledge of the land and skyscapes that they inhabit. Indeed, it seems that your journey into memory was motivated by the question of how this is/was achieved. What is the most impressive memory feat that you came across in your research?

Lynne Kelly: Two feats stand out as unbelievable achievements just because they really captured my attention, one in terms of the complexity of the information and the other in terms of the longevity.

The Navajo were able to describe a field guide to over 700 insects to ethnozoologists a few decades ago. The information not only included identification but also habitat, behaviour and stories integrating that insect into the cultural whole through mythology. Of these insects, ten were known because they cause bother such as gnats, lice, fleas or insects which attack crops. One was eaten, the cicada. All the rest were known because, as the ethnographers said, the Navajo love to classify. Knowledge for knowledge’s sake is not a trait of only literate humans. Once you add in all the other invertebrates, birds, mammals, amphibians, fish…the memory feat becomes even more extraordinary. Then add in a thousand or so plants, navigation, genealogies, laws, management of crops and domestic animals, hunting…the memory feat simply becomes overwhelming.

This feat particularly astounded me because I have attempted to identify insects and got nowhere. They are so much harder than the birds and mammals and other genres of critter. How many people can even identify the few hundred birds in their environment?

In terms of longevity, there are a number of examples of our Australian Aboriginal cultures retaining accurate descriptions of the changes in landscape dating back thousands of years. For example, the Dyirbal people from north-east Queensland have a story about the past when it was possible to walk across to the islands now called Palm and Hinchinbrook Islands. Geographers have since concluded that the sea level was low enough for this to be the case at the end of the last ice age, at least 8,000 years ago. It is simply mind-boggling to think that descriptions of events from so long ago could be retained so accurately without writing. I’m not surprised that these events tend to be linked to landscape formations because it is the landscape which acts as the primary medium for sequencing memory locations.

My research is about the mechanisms through which these feats are achieved.

Daniel: Your visit to Stonehenge was an important catalyst for the development of your theories, correct? Could you tell us a bit about your first visit to Stonehenge and why you think it is best interpreted as a memory space?

Lynne: I expected my visit to Stonehenge in 2008 to be fairly mundane. I was just visiting as a tourist with my husband Damian who had recently graduated in archaeology. I was researching my PhD on indigenous knowledge of animals through which I had started to glimpse the way non-literate cultures use the method of loci to memorise vast amounts of information. By that stage I had realised that the information they stored was so much more complex than I had ever realised, and I was including many genres along with zoology. My thesis was being derailed by a bigger question: how on earth did they memorise so much stuff?

I realised that out Australian Aboriginal cultures were using sequences of locations in the landscape as a memory device, sung pathways better known as songlines. At each sacred location along the hundreds of kilometres of songlines, they would perform a ritual, that is a repeated song, dance or story. It was these rituals which stored the pragmatic information I was looking at, encoded through mythology and integrated with spiritual beliefs. I also knew that the same methods were used by the ancient Greeks and Romans using locations within their built environment.

But what happened in the transition? Standing on Salisbury Plain that day, my thesis topic on animal knowledge in indigenous cultures got totally derailed as I added archaeology into the ever-expanding research topic. My precious animal knowledge was being engulfed by bigger questions. I realised that in order to settle, cultures would need to replicate the songlines in the local landscape to ensure that the information associated with this vast expanse of sacred locations was not lost. I realised that the perfect way to record the annual cycle of knowledge associated with the annual cycle of movement, seasons and ceremonies would be a circle of stones. That is why there are so many across the British Isles and Western Europe. The changes in the Stonehenge complex of monuments reflects exactly what you would expect of a memory palace as it moves from the needs of  a fairly open mobile society to a larger settled society, but that takes a lot of explaining.

Daniel: I’ve heard people refer to your book as the ‘Stonehenge book’. I found your arguments for Stonehenge as a memory space compelling, but your theory is really much broader than that, and is intended to explain many different sites. Could you give us a quick sketch of your theory concerning the role of these architectural structures as landscapes-in-miniature?

Lynne: I felt pretentious enough thinking I had a new theory for the purposes of Stonehenge, so my self-doubt became almost crippling when I realised that if I’m saying this is the natural way for a culture to memorise information in the transition from hunter gathering to settle farming, then it wouldn’t only be the British Neolithic.

In the small-scale cultures I’ve been looking at, in particular Australian Aboriginal and Native American, power was in the hands of those who controlled information. Power was not maintained through individual wealth nor coercion through using violence as it has been in all larger communities since. I identified a set of indicators which would imply that a monument was primarily memory space. These indicators formed the basis of my academic monograph published by Cambridge University Press, Knowledge and Power in Prehistoric Societies.

Archaeologically, there should be no individual burials with wealthy grave goods. There should be no indication of a hierarchy where some members of the community were clearly far better off than others materially. If the major monuments built by these cultures were memory spaces, then that had to be the source of power. All of the Neolithic and American Archaic monuments met this requirement. They were all built in that transition from hunter gathering to farming, which is not a rapid transition.

Essentially what is required is a set of locations in a clearly defined sequence of memory locations. I started looking for all the other indicators for enhancing memory that I was learning about from Australian, Native American and other non-literate cultures.

Knowledge is performed in non-literate cultures. That is it is danced and sung and stories enacted. Monumental spaces are also performance spaces. Some form of acoustic enhancement is often present but what is essential is that there are both public and restricted performance spaces. As a member of a society is initiated higher and higher into the knowledge system, information is kept secret. This is essential to protect against the so-called Chinese whispers. This is particularly important when survival is dependent on the accuracy of information, such as how to handle extreme resource stress, long-term agreements with other tribes, resource boundaries and rights as well as changes in the distant landscape.

Timekeeping is an essential part of any society to run a calendar for ceremonies, agriculture and seasonal resource gathering. A wide range of methods are used in mobile cultures, but with settlement, maintaining a calendar according to the solstice, equinox and lunar alignments is the preferred method. Astronomical alignments are an essential part of the knowledge system. The timekeepers were always very powerful people.

All non-literate cultures use some form of handheld memory device, often abstract signs inscribed on stone or wood such as the Australian tjuringa, the African lukasa or the Native American songboard or birchbark scroll. Details of these devices were often hard to find as no one has made an inventory of these objects before and there is no common name for them. Wherever I looked, I found portable memory devices. Most used wood or stone decorated objects, while many used sculptures, sets of figurines and paintings on bark. Some cultures used bundles of objects while others attached information to the various arrangements of seeds or shells. The Inca used the incredibly adaptable knotted cord device known as the khipu. These mnemonic technologies are not used in isolation. Indigenous cultures integrate a range of mnemonic devices into a system where one method reinforces another.

All the indicators of a memory space were in the Stonehenge archaeology, as long as you consider the entire Stonehenge complex of monuments which the archaeologists say are linked, and as long as you look at the way they changed over the 1,500 years or so of use. These same indicators were also present in monuments all over the world which were used in the transition from mobile to small-scale settled cultures. Each site has to be analysed independently, considering the degree of settlement, size of the society, materials available and subsistence methods. I was able to show that the huge glyphs on the Nasca Desert of Peru also fit the same pattern. So do mounds and pyramids right across the Americas in the early stages of settlement, including the mound-building hunter-gatherer site of Poverty Point in Louisiana and many others. Understanding non-literate memory systems also explains the purpose of the extraordinary ‘great houses’ of Chaco Canyon in New Mexico and why the non-literate Inca managed to outshine the literate Aztecs and Maya and control a massive empire without writing. By adding in the way Pacific cultures use genealogies to structure information, the purpose of the moai of Easter Island becomes clear. Unravelling the purpose of these monuments is just a matter of acknowledging the critical role of memorising vast amounts of pragmatic information in the absence of writing.

Daniel: Inspired by many of the indigenous memory systems that you’ve observed, you’ve conducted a number of your own memory experiments. Could you tell us about some of these?

Lynne: I have taken examples of each of the mnemonic technologies I researched and copied them in order to understand how the technology worked, not just how the indigenous cultures used it. For example, I have encoded a field guide to the 408 birds of Victoria to a memory board made of wood, shells and beads, based on the African lukasa. The birds are encoded in taxonomic order and I am constantly adding more information to it. I don’t need the board in the field to use it—I know it so well. I would never have considered attempting this before I understood the effectiveness of abstract designs on hand held objects as a mnemonic device.

I have set up landscape sequences based on the Australian Aboriginal songlines. To one songline I have encoded the 250 countries and independent protectorates of the world in population order. That is a set of discrete locations. In another kilometre or so of the local neighbourhood, I walk through continuous time from 4,500 million years ago to the present, watching all sorts of events play out and nodding to a vast array of historic figures.

I have created a knotted cord device based on the Inca khipu/quipu, sets of objects to manipulate, a totem pole, a series of stones and many more. To each I am encoding practical information for contemporary life as that is what I felt I had to do to see what was happening in my brain as I implemented these mnemonic technologies. I am astounded how effective they are and how differently my brain works when I use them. I am also astounded that we don’t use these methods in education. I am seeing patterns and asking questions that would never have occurred to me without the information stored in memory first.

Daniel: Your beautiful description of the songlines of Australian Aboriginals, with their use of the landscape as a way of ordering and retrieving memory images, is strikingly similar to the techniques used by contemporary memory athletes. These techniques, of course, are thought to have originated in Ancient Greece. When did you draw the connection between the song-lines and the Art of Memory?

Lynne: I was originally doing a PhD as a creative science writer in the English program at LaTrobe University. I was looking at the way the stories encoded accurate details of animal behaviour, identification and ecology. Then I realised how many animals there were once you add in the hundreds of invertebrates. Then all the plants and so on. Given that I am blessed with an appallingly bad memory, I started asking how on earth they could remember so much stuff. That changed my research question and I stumbled on the role of songlines.

At the same time I was looking at Walter Ong and his book Orality and Literacy which led me to Frances Yates and The Art of Memory. It just seemed obvious that the two techniques were exactly the same: take a set of locations in a fixed order and encode information to each location. Make the stories vivid, the characters grotesque, vulgar, extreme, very active—anything which makes them memorable. That is a perfect description of mythology, which Walter Ong identified as a mnemonic technology. It went from there.

Daniel: One thing which sets the techniques of the Art of Memory apart from most contemporary mnemonics (ROYGBIV, for example, as an acronym for the sequence of hues in the rainbow) is its general purpose nature. The principles and techniques can be applied to practically any learning task. To what extent do you think the memory techniques of oral cultures share this feature?

Lynne: They share this feature completely. The stories tell of all the practical knowledge, but enmesh that information with the laws, ethical stories, spiritual domain and all use the same set of mnemonic technologies. One device can encode a huge range of knowledge domains. The African lukasa, for example, has the history of the Luba Kingdom at one level, battle strategies and animal behaviour, ceremonial and initiation cycles and then, higher and higher levels of training add more and more information. But the higher levels were so restricted to protect the accuracy of the information, that with all the lukasa experts gone, there is no way to know what they stored. They would not have told uninitiated ethnographers.

The difference between linking information to physical locations, be they in the landscape of on physical devices, is that you can add layer upon layer of information in an infinitely expandable way. You don’t need to learn them in order. If you take something like ROYGBIV, or Thirty-days-hath-September …, then you can’t add any more complexity to the database. Songlines, memory boards, Stonehenge…they are all essentially database structures with the data held in memorable oral forms: songs and stories, dance and mythology. That data is in variable length data cells, each one infinitely expandable. But, if human memory is lost, then only the database structure remains. The monument, cave art or enigmatic decorated object remains but the knowledge encoded to it died with the last elder.

Daniel: It seems to me that the Ars Memorativa presents some interesting challenges to your theories. If I’ve understood correctly, your theory predicts that once a group of people fully transition from a nomadic lifestyle to permanent settlements, they lose interest in memory systems. Even more so for those settled cultures that develop written language. Yet the Ancient Greeks are often considered to have developed the first ‘true’ alphabet (i.e. one involving the consistent use of letters for both vowels and consonants) as well as the most advanced expression of mnemotechnics.

To push this even further, it could be argued that the Art of Memory, which played an important role in European thought until the 17th Century, even experienced something of a renaissance after the invention of the printing press.  I’m talking here about the work of Giulio Camillo, Giordano Bruno and others in the hermetic tradition, but also Bacon, Leibniz and Descartes. What are your thoughts on this?

Lynne: I sincerely hope I have never given the impression that once the transition to permanent settlement is complete that any culture loses interest in memory systems. In fact, I draw heavily on Native American cultures, especially the Pueblo who have long ago left their hunter-gatherer lifestyle.

Before I continue, I also need to get picky about the term ‘nomadic’. A nomadic lifestyle is when a group of people wander through the landscape with the clear impression that they do not follow fixed paths. Australian Aboriginal cultures have not been nomadic for thousands of years. They are ‘mobile’. That is they move between a number of semi-permanent campsites over the annual cycle to optimise use of resources. Some of our 300 different Aboriginal cultures may live almost permanently in one location, while others are much more mobile. It depends very much on the habitat and availability of resources in one location. For me to talk about the method of loci, they must be constantly revisiting the same locations. A songline is essentially a set of locations that are known well enough to be able to sing in sequence and are revisited regularly.

I think the point you are picking up on is that I say that the monuments, such as Stonehenge and other stone circles, were abandoned once the society had grown larger and settled permanently to agriculture. The reason I give for abandoning the monuments is that much of the knowledge system is then maintained by specialists—the farmers, bakers, warriors and tradesmen. The society by this stage had become much more hierarchical and those in power maintain control through wealth and force. All the cultures I am talking about maintain power through control of information and therefore those in power could muster the resources to create the ancient memory palaces. Everyone in the society would have seen their value because of the dependence on those knowledgeable elites.  So what happens when the knowledgeable elite is no longer in power? They are still required to maintain a great deal of the information of the culture, especially the legal system, the history, much of the knowledge of plants and animals that still exist in the wild, astronomy, trade agreements and so on. In the British example, a thousand years after Stonehenge you have the Druids performing this role. But they are no longer the top of the heap. They serve the chiefs but are still powerful. They still use memory palaces, but these are embedded within the built environment, such as the Iron Age hillforts. You can see exactly this hierarchy in Pacific cultures and Native American and of course eventually in ancient Greece and Rome.

The development of writing and the impact on the memory spaces is one of the areas I’m looking into at the moment, but I will offer some opinions anyway in the understanding that I am more than willing to be corrected. Writing served to store tallies for trade and names long before it was used to record the narrative which told the stories of the culture in the performance mode that had always been used because performance is so much more memorable than a list of facts. So as you move from Homer to Cicero, Augustine and those you mention right up into the Renaissance, the nature of the memory techniques stays similar because that is how the human brain works.

However, you lose the complexity of the indigenous elders and their interwoven set of mnemonic technologies. And apparently that’s because of the genres of information which are still being stored orally. Slowly (very slowly) the performances move from being the encyclopaedic knowledge of the culture in purely oral cultures, to being a major repository for some aspects such as history and technology in Homer’s day to becoming primarily the medium for rhetoric, ethics and laws and eventually recalling screeds of religious beliefs. In order to retain power as the knowledge of critical pragmatic information is taken over by other members of society, the knowledge elites slowly adapt the mythological characters into gods and claim an exclusive link to them, thus retaining power in the role of priests.

The memory arts were never lost, they merely changed in their role within the power structure, the genres of information stored and their pervasiveness within the culture. We have gone too far. They need to come back and be embedded within our education system without losing any of the benefits that literacy has bought.