Reader question: moving away from a memory space

[Click on all the images to get larger sizes.]

Miroslav Kalous from Prague in the Czech Republic, wrote and asked:

I’d like to thank you for the idea of “large memory spaces” which is really thrilling and I’m on the verge of building my own ones (one related to history till 1900, one for 1900+ years, one for specifically “all things Egypt” as that is a major country-project I’ve begun dealing with now).

However, I would also like to ask you one question before I begin, very practical one: unlike you (from what I understood between the lines), I don’t live at a permanent place; probably in 2 years I am going to move, then live somewhere else for other 3 years, then perhaps settling down for a longer time span at one place. As an experienced mnemonist, do you think it makes sense to start building the spaces where I live now? But what happens when I (or you) move? Re-writing all the loci spots into new palace/memory space is probably not realistic… and I am too much of a newbie to mnemonics to know if you can operate with, i.e. two complementary places. Also, I suppose, when moving somewhere else you lose the (critical?) advantage of going through the space and using them as “flashcards” prompting active recall of the stuff stored in there.

What a great question! I am so embedded in my landscape now that nothing would make me move. But as Miroslav points out, that is not practical assumption, especially for those much younger than me.

The first idea is to use public spaces which are unlikely to change. A quick check on Google images of Prague and – wow  – what a stunning city! The bridges across the Vltava River, as in the image above, looked wonderful to use as a set of memory loci.

There are a huge range of other possible solutions. These are often discussed on the Art of Memory Forum under “Method of Loci” – my favourite forum on the Internet

http://mt.artofmemory.com/forums/method-of-loci

One solution which was talked about in memory treatises written in the Middle Ages was to use an imaginary memory palace. One suggested way back then was to use Noah’s Arc as described in the Bible, but maybe something a little more contemporary is required.

Some people use sets of locations from their favourite films or books. It is a matter of creating the palace and a set of locations from that film or book using your imagination to add in extra locations or details. You would then, I expect, draw that memory palace and label it and keep it forever as your reference. You could even use Tolkein’s Middle Earth.

You could create your own imaginary world much as fantasy writers do. In fact, I have created imaginary worlds before when teaching science fiction and fantasy writing and I have just decided to try this as a memory experiment because I loved doing the maps and creating the worlds.

One quite common virtual memory palace is to use one from a video game. I’ve never tried this so I have no idea how it would work but I gather they can be very effective.

Another palace people use is this school or home from childhood and re-create these locations by drawing maps, just adapting any blurry remembering with imagination.

Commonly recommended in classical Greek and Roman, mediaeval and Renaissance times was using a famous building. Gothic churches were extremely popular and even designed with this use mind. Chartres Cathedral, as in the three images shown, is often discussed in these terms. 
You can use any streetscape. I would imagine the National Mall in Washington, for example, would work a treat. With the White House and all the Smithsonian museums and plenty of images online, you could easily create a memory palace that could be infinitely adaptable by adding the internals of each of the buildings if you wanted to expand it. There are visitor maps online for all the buildings. See below.

This is really fun thinking about all the possibilities, but I’ve got far too excited about creating my own fantasy world to write more. Sorry! Gotta go and start drawing!

 

Guest blog: experiments with memory

I am getting a lot of emails from readers which is so rewarding. Some are trying out the memory methods and are as astounded as I was about how effective they are.

A memory palace - From Emma Willard, The Temple of Time, 1846.
From Emma Willard, The Temple of Time, 1846.

Barry described his experiences. I will hand over the blog to him as he writes so well I don’t want to change a thing:
________________

I thought you might get a kick out of hearing how your work has impacted someone. It’s certainly had a powerful effect on me!

I’ve always been interested in the mystery of prehistoric civilisations, and of Australian indigenous culture before its catastrophic disruption by the Europeans. Your book has changed the way I see all of that. Myths and legends are not childish fantasies, but are multilayered storehouses of information! Astonishing, and yet, in retrospect, so obvious!

Anyway I could rave for ages about the insights into human history you’ve given me, but I will resist. I’ve been happily raving to practically everyone I know.

Of course, your book is a double-whammy — not only casting a new perspective on non-literate culture, but also painting an intriguing picture of the potential of using these long-neglected memory systems. I’d encountered memory palaces before, but they always seemed like too much hard work, and perhaps of dubious worth beyond remembering long shopping lists and playing cards.

Charged with new enthusiasm, I decided to make some memory journeys of my own. I too normally have a rather vague and temporary kind of memory. Here’s what I’ve tried:

First memory path

I live in a small town in the Sunshine Coast hinterland, and often walk to my office in town — about a 10 minute journey. I took note of potential sites and took photos of them all. Then I added them to a spreadsheet and gave them all unique names. I then worked to be able to remember each in order.

Following your principle of marking 5s and 10s, I added special markers to every 10th item. Every 5 spots I make special by imagining them as extremely cold. This worked well, as any story I add is enhanced by the dramatic cold. I can easily remember where the “cold” sites are.

This path is now 118 stations long. I’ll make it longer but it will involve a lengthy hike into the outskirts of town where landmarks are further apart.

Periodic Table of Elements

As a test, I decided to memorise the elements. It’s not something that I particularly need, so I figured if I messed it up it wouldn’t matter. It took about three weeks, but I got there. The marker system makes it easy to jump to any point by atomic number. It piqued my interest and I bought a little pocket book about the elements, which I’m now using to add interesting facts to the stories.

Countries of the World

I liked this idea and decided to emulate it, using my existing memory track. I was worried that the Elements would interfere but to my surprise they made it even easier! Each station is now ready-made with extra meaning and personality that makes them distinct; so the countries and the elements just seem to reinforce each other without getting confused.

I’m still working on this one. I’m up to Bolivia (the Monkees singing “Daydream Believer” in a South American accent, compressed into a Ball of Ears and rolling around. It’s also the station for Lead, and fishing lines with lead sinkers are casting their hooks into the ears and pulling them around painfully).

I do like this journey, because the countries of the world are mentioned all the time, and now when I hear their names I think of their special place and I have a chance to add to it.

Ukulele Chords

I’m sick of not being able to remember the chords when I jam with people. I normally have to look them up on my phone. Now I just have to think for a moment and I have the chord I need.

I made a small circuit in my garden, with 12 stations, each representing a musical note. Each station has a totem animal to remind me of the note, eg “B flat” is Beetle. Each station has two stories, one for the minor chord and one for the major. The major story is high up, the minor story is low down or underground. I turned the finger positions for each chord into 4 numbers and converted them into words using a version of the “major system”. This gives me the basis for each story.

I guess I eventually I won’t need this system as I’ll have learnt it by rote.

(BTW did you know that the etymology of “rote” is unknown, and may have the same origin as “route”? Interesting…)

Future Plans

Next I would like to learn something about the natural world. such as all the known edible native plants of Australia. I don’t really want to make another great big memory trail, so I thought a portable memory device might be the way to go. If you can provide any guidance in the construction and use of lukasa-style devices I’d be very grateful.

Other ideas:
major stars by constellation
bones of the human body
muscles of the human body
planets and moons
geological time
history
trees of Australia
birds
fish
Spanish vocabulary
software design patterns (I’m a software developer)
That’ll do. I hope you found my account of adventures in memory land of value!

Thanks again for your magnificent work.
Regards
Barry

______________

Thank you for your magnificent email, Barry!

Stonehenge – they moved their memory palace from Wales!

Thank you to the many people who sent me links to the various reports of this discovery and commented on how wonderfully it suited my theory on the purpose of Stonehenge.

“Stonehenge was a Welsh monument from its very beginning. If we can find the original monument in Wales from which it was built, we will finally be able to solve the mystery of why Stonehenge was built and why some of its stones were brought so far.” Mike Parker Pearson, archaeologist who led the study.

sh-bluestones
Click on image to go to University College London website and the full story.

I could not be more delighted by this discovery. In my recent Cambridge University Press book, Knowledge and Power in Prehistoric Societies, and in my forthcoming book, The Memory Code, I offer a new theory for the purpose of Stonehenge and monuments around the world. The new findings in Wales fit the theory a treat.

My research is on the way non-literate cultures memorized vast amounts of practical information when they had no way of writing it down. All oral cultures used a combination of memory techniques and physical devices – their survival depended on accurate retention of practical information on plants, animals, navigation, genealogies, astronomy and timekeeping, seasonality, resource management, intertribal agreements and so on. The memory technology employed universally is the ‘method of loci’ or the ‘art of memory’, the use a sequence of physical locations to act as a set of mnemonic subheadings to the knowledge system. The information for each location is then stored in song and mythology, stories and dance – all kept in memory.

Stonehenge was built in the transition from a mobile hunter gatherer society to a settled farming community. Mobile cultures used a range of landscape locations to store information, such as the Australian Aboriginal songlines. The ancient Greeks and Romans used their buildings and streetscapes in the same way, attaching information to each location and then recalling it by walking, or imagining themselves walking through their memory sites. Modern memory champions refer to their sequence of locations as memory palaces.

What happened when hunter gatherer cultures started to stay in one place, an essential development if they are ever to farm? They were no longer moving between their landscape locations over the annual cycle but didn’t yet have a built environment. The simplest thing to do was to replicate their landscape sequence locally, such as with a circle of stones or posts.

The original monument at Stonehenge is now considered to have been a circle of stones or posts, possibly the Welsh bluestones. The huge stones in the centre, the familiar sarsens, didn’t come to the monument for 500 years after the first circles.

I have argued in my PhD thesis and both books, that the bluestones were particularly suitable as memory locations because of the variety of textures and colours in their material made them visually so variable which is great for encoding information. I thought that the builders brought the stones and knowledge of the method of loci from Wales.

If Parker Pearson and his team are right, then they brought their entire memory palace!

I could not have hoped for a better development.

 

The Memory Code will be published by Allen & Unwin in July 2016 in Australia and later in the UK by Atlantic Books.

 

Knowledge, Power and Stonehenge

This blog is a response to questions from archaeologists from a talk I gave on Thursday. I addressed a crowd of over 200 at the Castlemaine Library on the topic of “Knowledge, Power and Stonehenge” based on my book. There were a number of archaeologists in the audience who were very positive in their response and have contacted me with questions that they didn’t get a chance to ask. Here are two of the questions:

Q: Last night you only briefly referred to the new stone arrangement reported from Durrington Walls. Can you expand on the way you see this setting fitting with the dichotomy you argued is seen in mnemonic monuments all over the world? (See the post below this one for more details of the new findings.)

DW-phase1-3_02
Part of the stone row at Durrington Walls, Ludwig Boltzmann Institute.

In monuments used primarily for memory purposes, I am always looking for ordered sequences of stones, posts or mounds to replicate the sequence of landscape sites used by mobile cultures. When they get to city size and clear hierarchies, my theory no longer holds.

The ethnographic record from small-scale oral cultures all over the world is unequivocal. There are always both public and restricted performance sites in which knowledge is taught and exchanged. The restricted sites are essential for two reasons (among others): to retain power for those initiated into the higher levels of the knowledge system and to avoid the so-called ‘Chinese whispers’ effect. Knowledge is corrupted if it is shared willy-nilly. Knowledge needed to survive severe resource stress, for example, is always held at the highest restricted levels. In the Australian mobile hunter-gatherer case, the public / restricted performance site dichotomy can be seen with the public corroboree grounds and highly restricted bora grounds. In Pueblo cultures, between plazas and kivas. And so on.

In terms of the Stonehenge / Durrington Walls complex of monuments:  Stonehenge became a highly restricted site when the huge sarsens arrived about 500 years into its use and everything was enclosed in the centre. At the same time, the superhenge Durrington Walls was built, giving a new public performance space. There was also a fairly restricted set of posts near Durrington Walls, known as Woodhenge.

The news a few days ago reported that at Durrington Walls a sequence of up to 90 standing stones had been found around the edge of the henge. This is exactly the sort of sequence of memory locations I am finding all over the world. The Durrington stones appear, from the reports available, to be separated so that each is encountered singly, as required for memory locations. This gives a much more defined public memory site at Durrington Walls than it was before, with restricted sites at Woodhenge, and even more restricted at Stonehenge. This complex works as a single site. Stonehenge alone won’t fit the theory I outlined at the talk and in the book.

Q: I understood from your talk that you believed that the memory techniques used were a product of evolutionary convergence and different societies developed these methods separately, not that they are 60,000 odd years old and left Africa at the same time as humans; what is your basis for that position?

trust-african-rock-art-600
The image links to the Trust for African Rock Art.

I confused you! Sorry! I believe that the human ability to memorise in this way probably dates to at least 60,000 years ago and is a critical part of human evolution – but I haven’t done that research thoroughly enough to claim that yet. There were evolutionary biologists in the audience who are very excited about this aspect and love what I am saying.

It is the implementation using sequences of posts, stones or mounds for sets of sequenced memory locations which I believe was developed independently. These monument types don’t appear in the archaeological record until the last 10,000 years or so. I think the evidence is there for the landscape being used as a sequenced set of memory locations for much longer than 10,000 years, but it is the specific implementation of the method locally on settlement which I believe has been developed by different societies independently.

pp-post-circle-web
The image of the barrels marking one of the 27 post circles at Poverty Point, Louisiana, by Jenny Ellerbe. Used with permission.

The posts circles in the plaza at the mound site of Poverty Point in Louisiana, for example, weren’t copied from the British Neolithic despite their similarity in dimensions and the separation of the posts to stone and post circles in the British Neolithic. They developed this implementation because it is an incredibly effective method (the method of loci) that has never been bettered, and we all share the same brain structures.

Singing the knowledge – Yanyuwa kujika

A wonderful collaboration between the Yanyuwa people and Monash University is online and enables us to glimpse the singing tracks of their culture. The Yanyuwa live 0n the Gulf Carpentaria in Northern Australia.

Animations of the songs can be seen at the Monash Country Lines Archive.

yanyuwa-kujika-monash1

As described in his 2010 book, Singing Saltwater Country John Bradley has mapped over 800 km of singing tracks in a three decade long association with the Yanyuwa people of Carpentaria.  The sung pathways through Country are referred to by the Yanyuwa as kujika and described as a ‘Yanyuwa way of knowing’ and as the ‘key to rich, complex and intricately related knowledge systems’. For one kujika, Bradley recorded over 230 verses, with knowledge stored in layer upon layer, the more complex knowledge gained with initiation into higher levels. Every detail of the landscape is described and stored in the sung narratives. Fixed in place by the very landscape they describe, the kujika act as the link connecting all songs in a sequence.

This is the method of loci in its most expansive form. Every aspect of the knowledge is encoded in these songs. This is one of the few glimpses of primary orality available in the world today, when this method of knowing almost certainly served every human community for tens of thousands of years.

The complexity of Australian Aboriginal knowledge has been hugely underrepresented until recently. We need to ensure it is understood before it is lost forever.

Bradley wrote about the experience of learning an Australian Yanyuwa kujika:

So much knowledge was being presented to me, at many levels and intricately interrelated. I was struggling to find words for much of the material as it was deeply encoded and dependent on other knowledge.

…I was amazed by the detail of this kujika, especially of the different species of sea turtles, their life cycle and habitats; it was a biology lesson in sung form.

The first stage of the Tigershark Dreaming  and then the second represent just a tiny part of the Manankurra kujika. Part One is linked to the image above, with part 2 below. Further songs are available at the Monash Country Lines Archive.

yanyuwa-kujika-monash2

Although there is clearly a spiritual dimension to the kujika, it is so beautifully clear that these songs give an intensie knowledge of the geography for navigation and identification and behavioural details of the animals in each microenvironment. The techniques of treating the cycad seeds to take them from deadly to edible are also mentioned. These are a rational people with great depth of knowledge.

It is the singing tracks and the depth of pragmatic knowledge which is the basis of all I write about in Knowledge and Power in Prehistoric Societies (Cambridge University Press, 2015).

 

Adding the Dowitcher – a comparison of memory aids

A small wading bird landed in a lake and started a twitch unlike any before in Australia. Social media went into overdrive, as bird watchers scrambled to travel for hours in the hope of a glimpse of the one long-billed dowitcher among the sharp-tailed sandpipers at Lake Tutchewop, between Swan Hill and Kerang in northern Victoria.

"Long-billed Dowitcher"
The long-billed dowitcher and sharp tailed sandpipers. Photograph with permission of Paul Dodd. Lots more stunning photographs on his website: http://paul.angrybluecat.com/Trips-and-Locations/2014/Lake-Tutchewop-Nov-2014/

Why all the excitement? This is the first record ever of a long-billed dowitcher in Australia. A new tick, a lifer, for every twitcher who managed to see it. They turned up in droves.

Readers of this blog will know that one of the memory challenges I have undertaken is to memorise the 407 birds in Victoria using a memory board. The list in my head is in taxonomic order with family names in Latin. So what happens when a new bird inserts itself in the middle of one of my existing families?

If I’d used one of the standard mnemonic methods, such as creating a rhyme for all the birds, I would have real trouble adding in a newcomer. What would happen if they found a new king between Richard I and John and needed to add him into the famous mnemonic for the monarchs of England:

Willie, Willie, Harry, Stee,
Harry, Dick, John, Harry three;
One, two, three Neds, Richard two;
Harrys four, five, six… then who? …

More complex memory methods, such as using an acronym or a linked sequence of items (the Link Method)  would also let me down here. It is very difficult to add in a new comer into the chain.  By using the complex of indigenous methods as I do, adding the dowitcher was a breeze.

side-view-lukasa
My memory board – a perfect size to hold in one hand.

I describe the way I memorise the birds in the post Memorising Birds. Each bead on the memory board (based on the African lukasa) represents a family. I traverse the beads in a set order. I can sing the 82 family names: Dromaiidae, Anatidae, Megapodiidae, Phasianidae, Podicipedidae … 82 of them. As I sing them, I can see the memory board in my mind, I know it so well. Mythological stories associated with each bead tell me how many species there are in the family. If there are more than 4, I use a journey, the method of loci, to memorise all the members of the family, again in taxonomic order. I had difficulty believing this would work when I first read about the lukasa. I now know that it is amazingly effective.

I had to add a new bird, the dowitcher. It is Scolopacidae, along with the 25 sandpipers, stints, godwits and the like in the family. The house along my sandpiper songline with the godwits and sharp-tailed sandpiper (the birds most like the dowitcher) has two small palms in front which I have linked to the two godwits. I needed something new, and soon noticed a third small palm, mostly hidden by other bushes. That palm was, within a minute, linked to the dowitcher. Hear the word ‘dowitcher’ now and my brain instantly sees that palm which allows me to locate the bird in the taxonomy. Into the ‘mythology’ I have already created around that location, I encode everything I learn about the dowitcher.

Using a range of mnemonic technologies system reflects the way non-literate cultures encode such a vast store of information about animals, plants, laws, navigation, astronomy, timekeeping and all the other practical knowledge I talk about in Indigenous mnemonics. It seems as if it would be confusing, but it isn’t. It seems as if it is more work than just memorising by constant repetition, but it is far far easier and far more robust.

I love this stuff!

Post and stone circles – everywhere

These barrels mark the places where a massive timber circle once stood. Just like timber and stone circles all over the UK, Ireland and Western Europe. But where is it?

pp-post-circle-web

So where is the plaza with this familiar form of monument? Louisiana. USA. Constructed by hunter gatherer fishers amid their mounds and massive earth works (note the scale on the plan below!), the Poverty Point culture never farmed.

10-1-Poverty_Point_graphic-web

Plan of the earthworks, mounds and post circles at the World Heritage Site of Poverty Point, Louisiana. Image (c) Lynne Kelly.

The Poverty Point Site consists of massive earthworks built  3,400 years ago: five mounds shown around six C-shaped ridges enclosing a huge plaza. Within the plaza, 25 – 30 timber circles were built, but were not all standing at the same time. Like all knowledge sites, it was constantly changed.

The geometric design of Poverty Point is unique – there is nothing like it anywhere else. It is a masterpiece. When constructed, the Poverty Point earthworks were the largest in North America, the major political, trading and ceremonial centre of its day.

Why did people in America’s southeast build monuments so reminiscent of those built by Neolithic cultures in the UK, Ireland and Western Europe?

Because this is the best way to create the necessary memory spaces if an oral culture is to settle and replicate a knowledge system once based in the broad landscape.

The image of the barrels marking the post circle from Jenny Ellerbe can be found on the Poverty Point World Heritage Initiative document which can be downloaded from the site – just click on the image below. The plan of the site has been adapted from that document as well.

poverty-point-websiteIt wasn’t only in Louisiana that post circles were built by Native Americans. One circle which has been reconstructed is in Illinois, at the mound building site of Cahokia. The ancient Native American city was active long after Poverty Point, from about 600 to 1400 AD. Archaeologists even name the timber circle Woodhenge after a wooden post circle in Wiltshire, England.

cahokia_woodhenge2
Woodhenge: The reconstructed post circle at Cahokia, southern Illinois.

The oral cultures in North America and the European Neolithic were so far apart in both space and time that they almost certainly had no contact with each other. It is no coincidence that they used very similar structures at the ceremonial sites. These are practical monuments which served a very practical purpose. They are memory spaces used to aid memory of all the practical, scientific, historic and spiritual knowledge of the culture.

SaveSave

Uluru as a set of memory locations

I’ve talked about the method of loci – a set of locations in the landscape used as memory aids – the most effective memory aid known. I believe that the singing tracks of the Australian cultures, the sacred trails of the Native Americans and sacred paths found in cultures around the world served the needs of memory in exactly the same way.

uluru-ian-161-1000
Uluru, Central Australia. Photo: Ian Rowland.

The massive natural Central Australian monolith, Uluru, is easily visualised as mythological landscape with a path encircling the entire rock. The pathway around Uluru is nearly nine kilometres long. The many crevices and indentations around the base are each linked with stories. As I detail in my research and forthcoming books, the songs and stories of non-literate cultures are the means by which a vast store of information is retained, much of which is knowledge of plant and animal classifications and characteristics, navigation, weather, tides, a calendar, a pharmacopoeia, rules and ethics, genealogies all integrated with history and religion.

Australian anthropologist, Charles Mountford, described how almost every feature on the surface of Uluru (at that time referred to as Ayers Rock), is named, acting as a mnemonic for mythological story.

uluru-mountford-800

It is so easy to imagine walking the rock, each crevice and point reminding you of a story and the encoded knowledge. Over years of learning, the amount of knowledge encoded in this sequence of locations would become vast.

The Anangu traditional owners describe Uluru as part of their knowledge system, Tjukurpa, which they explain has many deep, complex meanings including the law for caring for each other and their Country, the relationships between people, plants, animals and the physical features of the land, the past, the present and the future.

It is not necessary for the Anangu knowledge specialists to be walking the rock to recall the stories. The sequence of sites is so well-known after years of learning that they can travel and part of the perimeter in their memories whenever they want. This is the art of memory exactly as described by the ancient Greeks.

 

 

Medieval memories – illuminated manuscripts

smithfield-decretals

I have always found the illuminated manuscripts of the Middle Ages to be some of the most beautiful artistic creations of all. It was only through the research on memory spaces that I discovered that some scholars consider their design to be intricately interwoven with the need to memorise the content. The manuscript pages were designed as miniature memory spaces.

I spent an inordinate amount of time staring at the glorious reproductions in books on illuminated manuscripts.

In the Middle Ages, the memory arts changed purpose from the oratory of classical times to become the domain of the monks wishing to memorise great slabs of religious tracts. Monks were expected to memorise, at a minimum, all 150 psalms, a task which took somewhere between six months and three years.

The heavily illustrated handwritten manuscripts were seen as a prompt for medieval memory when books were extremely rare and horrendously expensive. The words were enmeshed in images which match the classical recommendations for making information far more memorable: grotesque and violent acts along with fanciful beasts, strange figures, gross ugliness and extraordinary beauty. It was common to have each chapter start with a coloured initial, alternating between red and blue, with repeated letters each having their own design, such as in the Smithfield Decretal shown above.

kells-canon

Around 331 A.D., the Roman historian, Eusebius of Caesare, created lists of chapter numbers indicating areas of agreement and of difference between the Gospel accounts in  Matthew, Mark, Luke and John. These chapter numbers were typically presented as elaborately decorated tables, designed to be memorised. The lists of numbers were written between illustrations of columns with arches above, replicating the classical Greek advice to use intercolumnar spaces as locations for memory images. The vertical spaces were then divided into small rectangular spaces each holding no more than five items, the maximum number suggested for retaining in memory for a single location.

A huge range of memory devices were recommended in the many treatises written during the Middle Ages for monks, and later for a whole range of students. Alphabets, the zodiac, bestiaries … so many different sequences were used for miniature memory locations, often illustrated in the elaborate medieval style. Much of the art decorating churches were also aimed to aid memory. More of that in future posts!

References:

Carruthers, Mary, The book of memory: a study of memory in mediaeval culture (2008), second edition, Cambridge University Press, Cambridge.

Carruthers, M & Ziolkowski, JM (eds), The medieval craft of memory: an anthology of texts and pictures (2004), University of Pennsylvania Press, Philadelphia.

de Hamel, Christopher, A history of illuminated manuscripts (1994), Phaidon, Oxford.

Yates, Frances, The art of memory (1966), Routledge and Kegan Paul, London.