In Conversation event at Gimpel Fils, London
Saturday 22 May 2010

Edited transcription from a video document by Martin Pickles.

SPEAKERS (click to jump to individual presentations):
René Gimpel [RG]
Michaela Nettell [MN]
Tom Simmons [TS]
Dr Hugo Spiers [HS]
Helen Maurer [HM]
Dr Peg Rawes [PR]
Andrew McDonnell [AM]

[RG] Gimpel Fils is happy, proud and honoured to present the second Wellcome Trust scientists' and artists' [project] ... prompted by Hugo Spiers. And today we've got a beautiful display down in the showroom that is going to be the subject of the discussion by Michaela and Tom here. So I'm going to allow Hugo to make an introduction of the speakers, and to say that I hope that now we've done the second talk we may have an angle on these things, but that it's well worth getting the scientist and artist communities together on discussions such as these.

[HS] Thank you very much René. The first thing I'm going to say is that this is an In Conversation event, it's not a sequence of public lectures. Feel free to ask questions, raise your hand, if there's something you want to discuss or ask about please do that. I'm the scientist involved in the project, I'm one of the three people with Tom and Michaela here who had to put an application in to get funding out of the Wellcome Trust. We're very, very thankful they did because you're all here, because someone put a stamp on that said you can go away and do this. So we spent two years or so working on this project and I've been happy to put some of the science and ideas into this and really work very collaboratively with Tom and Michaela. And I'm going to hand over to Michaela now to introduce herself.

[MN] Hi everybody, thank you so much for coming. I am an artist. I work with moving image: film and video, often projections and more recently with glass as well, combining projections with glass to create sculptural installations and moving image works. Tom, Hugo and I have been working together on this project for a couple of years now. I know not all of you have seen the installation yet dowstairs, please do pop down at the end and have a look if you havenít seen it. For those of you who havenít seen it yet, I'm just going to describe what it is and how it feels, and also what the installation is doing. The project is called Pattern Completion and it is inspired by a theory of memory recall that Hugo introduced Tom and I to a couple of years ago. Basically itís an idea that we think the way our brain pieces together past experiences comes from taking a small fragment of an experience such as a name or a place and gradually building up a complete picture from that small fragment. Hugo presented this idea to us and thought we could create an installation that described it - in a useful way in terms of the science, but also in a beautiful and poetic way as well.

The installation downstairs consists of four glass spheres suspended in the gallery space. Behind the spheres is a video projector, throwing imagery onto the glass spheres. The imagery depicts four or five photographic scenes shot in woodlands and forests in East Yorkshire. [...] the spheres are sandblasted on the rear surface to allow the image projections to catch in the glass. And because the spheres are hanging in a cluster, and distributed like this, only portions of the woodland scenes are revealed in these distributed screens.

Not only are they just fragments of the scenes but the fragments are shuffled, so that to begin with when you walk into the space you'll see maybe more than one of these woodland scenes shuffled in the different surfaces. There are headphones in the installation space and there are sounds that accompany the image fragments and the sounds describe the forest environments as well.
As you watch the installation, and you listen to the sounds and watch the image sequences the different clips appear to move from one sphere to another, they start off quite jumbled up, quite shuffled and gradually they cohere into a complete scene. So maybe at the beginning a portion of a woodland scene that is a canopy might appear low down and a portion of the woodland scene that is a pathway might appear in one of the spheres at the top but gradually, gradually the images shuffle around until they find their places and the scenes become complete. And this process, which is controlled by a computer program, which weíll talk about later, is mimicking the Pattern Completion process that neuroscientists believe is going on in our minds.

Quite early on in the project we decided that forests would be an interesting environment to base the piece of work around. We chose forests for lots of different reasons but perhaps most importantly because we wanted the installation to be quite immersive and to feel when you walk into the gallery space that you are entering into a different kind of world. So we travelled round to various sites of woodland in England, we tried places in Essex and in [Surrey] and we tried places in Hampshire and we decided in the end to capture images and sounds from a nature reserve in East Yorkshire, which is really close to where Tom grew up, so for him there are lots of resonances of his childhood memories. But also for me, each time I go to this particular place it strikes me as a particularly evocative and imaginative area. It has these beautiful silver birch trees and rows of pines that are very evocative and fairy-tale-like. It's also extremely photogenic, this part of Yorkshire. We wanted the installation to feel quite otherworldly and quite poetic.

In terms of composition, our compositions have become quite classical in a sense, they often have strong horizon lines coming across the centre of the photograph and strong one-point perspectives leading off into the distance and this creates a beautiful photograph, but it also really helps when the images are projected into the spheres to have these strong lines and these strong directions. Because of the curved surfaces the images wrap around, there's a wonderful illusion of three-dimensions that's created from the flat photographs. The perspective lines going off into the distance and the strong horizon lines all help to add to this sense of depth and three-dimensionality. There's also an important aspect of leading your eyes off into the distance, and focussing maybe on clearings as well. We were really interested in ideas of walking and trains of thought and the wandering mind. We were trying to create an installation space that was quite open, so you enter into the space and your mind is free to wander and to daydream.

My background is in animation and I often work frame-by-frame with sequences of still images. We toyed with the idea of using film or using video to depict these environments and in the end we decided to use a time-lapse photography technique, which I think appeals to me as an animator because you have lots of sequences of stills that you manipulate into a moving image sequence. It allowed us to create quite an otherworldly sense of movement. So we went to these different forest environments and took sequences of photographs Ė some of the sequences were taken at intervals of three or four seconds, other sequences were taken in much quicker succession Ė then we took the sequences of stills and blended them. So that each of the photographs was expanded out to about two seconds each and then blended one to the another. So you get a hint of movement, it's very subtle, very small. I think when you first walk into the installation space the images appear to be static, or still photos. Then as you get drawn in, as you spend more time looking you gradually see little movements appearing. So the trees might sway a tiny bit, the grasses might move. There are a couple of scenes with water in and they create really mesmeric effects, the water appears to be almost like a solid mass just undulating slowly. We're quite interested in this sense of movement and the implications this has for dream-states and memory-states. We have some video clips as well. The images I'm going to show you are full-screen versions before they've been broken down and projected into the spheres so you can see a bit more clearly. The first one is quite a close up shot.

[Question from audience] Can I just ask a question about the installation? Because you've kind of put the hierarchy of the visual over the sound. I love the sound too, but I wondered what your reasoning was for having it silent in the room until you put the headphones on?

[MN] Tom is going to talk about the sound, and he's going to explain the importance of having the personal, intimate headspace that's created by the headphones. I think that it works well to have the headphones there, so that you can enjoy the installation in a quiet space too. When the gallery is empty it feels ever so peaceful, ever so quiet in there. It's the first time we've created this kind of installation in a white cube space, often we've had black spaces because of the projections, and in the white space it feels to me very light and very airy. And I think you can go in and enjoy the work silently too, and have your own thoughts, sounds and images going on in your head.

This shows a frame of our computer program that we used to divide these four scenes into the fragments that [you see] when you look at the installation. It's not that easy to see but there are different coloured squares overlaid across this scene. The different squares correspond to the different sizes of the glass spheres in the installation. We went through a process of choosing the elements of the image that we thought were the most interesting, and the visual details that we thought would communicate the sense of space best.

This for me is the more exciting bit, more exciting than the maths of getting the sizes right! This is the more exciting moment in the process, when the video projections combine with the glass, the material of the glass in the space. The interaction of projected light and glass is something very magical. First of all, the way the different image fragments appear to curve around the glass forms so in effect you have these little capsules, these miniature worlds. Obviously the lines and the perspectives are slightly distorted but that creates really interesting effects. Because the images are morphing very slowly, the light is always changing a little bit and as it reflects and refracts around the glass shapes the colours are always changing as well. So there's this sense of transience and fluidity and everything being on the move all the time, which I think is something else that really relates to the feeling of memory, the idea that it's very transient, very kind of unreachable, that it's always changing and always a little bit different. You get that sort of ephemeral, fleeting feeling. The glass is hanging by fishing wire, so they're hanging in quite a fragile and delicate way. They move very easily if you walk close by them, the air currents allow them to shift and the constellation is always in motion really. So that adds another layer or dimension of ephemerality or transience to the installation, it's always changing a bit.

They also reflect each other, the spheres, you get little glimpses of one sphere in another and that also creates layers of associations and connections. I think I forgot to mention earlier on that the idea of having the cluster in the first place was to mimic, in a very basic way, the anatomy of the hippocampus. So these could be seen as very blown up and very simplified cells within the particular part of the brain that controls our memories. The idea that they might be seen as different brain cells and that they're each communicating with each other, each affecting each other and they are affected by each other. We went through lots of different variations of the cluster before we arrived at this particular configuration. It was really interesting finding out, you know, that the bottom one needed to go at the bottom and the little one needed to go at the top to create the right sense of balance and sense of weight. It feels very light, the whole thing feels as though it could just lift off, like bubbles.

I'm going to hand over to Tom now and he's going to talk a bit more about the sound aspect of the installation.

[TS] My name's Tom Simmons. I'm a sound designer and I also teach in an art college and I do research there too. I've been working with Hugo and Michaela on and off over the past four or five years doing various projects and this is not the first time we've used sound but it's certainly the first time we've used sound in this way, together. So I'm going to talk a bit about that process. I'll come back to the question about headphones in a little bit if that's okay.

The idea behind the sounds in the installation is partly to reflect the very detailed topography of the woodland and forest scenes that Michaela has depicted in the images, and a very specialist sound technique called binaural recording is used to gather the sounds. I have brought a pair of binaural microphones with me but I have locked them in the basement so, apologies for that! They are very small microphones that you put in your ears and when you go out collecting sounds and recording sounds the idea behind this technique is that it captures a very spacious sense of sound all around you and it is reproducible through headphones in the listening space. So when you hear in real life someone shuffling a chair over here, this is where the sound appears in your listening space, and when you hear a projector whirring over here, this is where this appears in your listening space, and so on. So this technique tries to recreate through headphones in a very intimate and quite immersive way, a very detailed sense of place and this was something we wanted to experiment with in this project.

Intriguingly it wasn't actually the technique we proposed using in the first place, it's something that we decided to get involved with about half way into the project. Initially we were proposing to develop a sound technique which would have been much more focussed on the glass spheres. And the initial tests we did we suspended very small speakers above the glass spheres so the sphere itself resonated with the sound recording. And this worked absolutely fine in some senses Ė bird sound of course is very high frequencies and glass resonates with quite a high frequency so this worked quite marvellously. But whenever rustling leaves came into the sound recordings or the sound of foxes barking or deer making noises or people - footsteps, or dogs, none of this was audible and this was a bit of a crisis so we weren't quite sure what to do about this and after a bit of umming and aahing we decided well, headphones! So we put this into practice. And we did a lot of binaural sound recordings, some of it in Norfolk, some of it in Yorkshire. And perhaps at this point we could have the first audio recording? This is just a straight binaural sound recording which, as we don't have any headphones here, isn't going to give a brilliant depiction of space, but hopefully it will sound very spacious.

So hopefully what's audible is, apart from a lot of birds, various bits of the forest environment. These speakers are also quite far [apart] so we're not picking up all the sounds. But in the installation downstairs with the headphones a very detailed soundscape is provided. And this partly reflects the very immersive nature of the woodlands and forests we were trying to capture, but also quite a lot of ambiguity about exactly where that might be and what might be going on in that space. One of the ideas we had was very much the installation - rather than simply mimicking, illustrating or reflecting the scientific process involved in Pattern Completion - should be conjuring up this process and making it available for people to experience or to have their own memories or to take their own memories from these types of placeholders, these types of images and sounds. So we didn't want to overly describe these environments, these places and the narratives that might be happening in them, rather we wanted to give suggestions for people to be able to develop their own.

In developing relationships with the images and with the glass spheres in particular, one of the techniques we wanted to try to do was to filter out very precise pockets of sound within this binaural sound recording. And this is almost like imagining a cone coming from the middle of your head out towards one of the glass spheres, and the sound almost being filtered on a conical shape out to this bubble. Hopefully when you look at the images, when you hear the sounds you should be aware that there are some very high frequency sounds which are accompanying the images that are going on in the bubble which is up here, and there are some quite bassy sounds going on in the images in the bubble here. And over on the left and the right the sounds are very spatialised too.

In sequencing the sounds and images in the installation we were quite aware that this process is incredibly complicated that happens inside our brains and this idea of Pattern Completion, which takes place in a fraction of a second, this incredible moment when you think, 'goodness me, what did I have for breakfast?' and then it comes back to you, 'yes it was some amazing toast and jam!' So this is a very slowed down version of this process. And in terms of dealing with all the possible combinations and permutations of images and image cues and sound cues and sequences that we could be working with in this project we thought if we do this all in Final Cut Pro, scene by scene, it's going to take us forever and actually this is not going to do anything other than illustrate our idea of the process. So what we decided to do instead, with Hugo's help, was to develop a program which employs a very simple version of a neural network. And this program helps us sequence the sounds and images in the installation. So this is deciding which sound will appear in which part of the listening space and which set of images will appear in which of the glass spheres.

There is a set of logical stuff up here. And then these objects in the middle, for anyone who's involved in the more computational side of things, are Kahonen map objects. These are self-organising maps, these are statistical, probability type objects that really just compute the likelihood of a particular thing occurring Ė in a particular bubble, a particular glass sphere or a particular space, at any point in time. What we were trying to do was to give this computer program various patterns which are over in this little box over here, and these are saying if everything is in the right place, which this is at the moment, this is what the pattern should be for the second scene Ė so the first clip plays back in the first glass sphere, the second in the second, the third in the third and the fourth in the fourth and all is harmonious. What we did to that was to add in quite a lot of noise to say well perhaps it's not quite so likely that this clip is going to play back here at this point in time, and we gave that information to this self-organising map, [which] is computing the likelihood of that pattern being brought back together at any point in time. There is a series of metrical steps at the top, which determine how quickly this happens. And the bottom part of this is just sorting out the sound and the image and which part gets played back where.

So this in a way has been the meeting point in our collaboration. We've talked a lot about ideas, we've talked a lot about some philosophical and conceptual starting points for exploring memory and memory recall in particular. But in terms of actually dealing with a meeting point between sound and image and some of the neuroscientific concepts we've been exploring in the project, this is where we've tried to bring the two things together. So it's a rather strange meeting point between mathematics and computation of some very poetic ideas on our behalf and some very scientific ideas on Hugo's behalf. And I think on that note I'll hand over to Hugo to talk about the neuroscience.

[HS] The title of this section is Memories are made of this because this is very much what we're here to talk about. You've just heard about the mechanics, and the processes and thoughts that went into building this installation, but what are the ideas underlying this?

So the question is really what are memories? Tom mentioned this and Michaela mentioned this, that we carry around with us inside our head these thousands of memories that you collect over your lifetime. You can travel back in time, maybe ten, twenty, thirty years, back to an event that happened in your past and re-live that past, re-consider it. But we know from our research, and you can probably guess from your own experiences that our memory is not a rolling video tape, we're not taking in the world perfectly, repeating it out, perhaps unless you're an autistic savant. But if we're not, for most of us, we're not able to re-capitulate what happened to us in the past. We have to reconstruct it. So scientists are very interested in saying, well what are memories? How does this process work? How does this reconstructive process work?

Let me first of all start with the brain, this is a really beautiful brain, in fact, this one is my brain! In fact this was my brain, it was much nicer back in the past, it was taken when I was 25. If we zoom into this brain, into this incredibly curled up structure, if we really zoom into one bit of it we can see that there are these tiny little cells, our body is composed of cells and our brain is composed predominantly of neurons, here is an example of one neuron. And there are about ten billion of these neurons sitting inside your heads, all of you watching me now. And this is a type of neuron called a pyramidal cell because it looks a bit like a pyramid, scientists are not particularly imaginative when they name things. But they are very beautiful, scientists have been marvelling at them for many years, these things projecting out of this cell body here, all these amazing dendrites we call them, the protrusions that come out of the cell that collect all the information into this one little cell. And this one little cell has another thing coming out of it called its axon, projecting out of this one cell, and itís the only way it communicates to the rest of the outside of the cell, the rest of the whole world inside the brain, this one output. And thousands and thousands of inputs. So this one cell will project off to thousands of neighbours and friends that might be far across the brain, or that might be very close to it. There are ten billion of these all connected to fifty thousand other cells, thereís a lot of activity going on here, a lot of information processing and passing around inside your brain.

And if we look at one of these cells, here's an example, a more schematic diagram of a pyramidal cell. If you put an electrode into the brain, as a lot of scientists like to do, to try and find what are these things doing, what is the information these cells are carrying? We can record this and we can draw out what the electrical activity is and it sounds like that, if you record the information passing out of this axon and you pass this image through a sound filter it sounds like that. Quite spiky, electrical charged activity. So each of these cells is constantly bursting, there are ten billion of these constantly passing out this information around the brain. And if you take a sequence of these, you can see that they are effectively on or off, you can characterise this huge number of brain cells out there as either being on or off. So in effect there's a series of ones and zeros. It's a bit like The Matrix in a sense, that your brain is pulsing with these sequences of ones and zeros that are building up a mental model because what youíre seeing, my face coming into your eyes is not being recreated in your brain exactly as it is in the physical world. Your brain is making a huge number of assumptions about what youíre seeing. And also your memories do the same. They make a huge number of assumptions and fill in lots of gaps.

So what are the actual memories? Well we think the memories are stored in these points in the cells. These cells connect up, as I said, to many, many other cells. And when the axon reaches another cell it stops at a point with a synapse, and little tiny chemicals get released over the gap between one cell to the next. The electrical activity stops and the chemical activity takes over. And if this cell, for example this cell here if it has enough inputs to its dendrites, if it has enough cells shouting at it effectively, it says okay I'll send out an output and it will burst out some information. So these cells effectively filter in lots of shouting people at them and if they really get enough of it they'll create an output. And it's in these connections on these dendrites where we think these memories are stored as patterns. That's what your memories are in a sense, from a neurobiological viewpoint. It's the connections, the way in which the connections between your cells are configured.

So if we explore this a bit further we can say, I've told you that the cells communicate in ones and zeros, let's make an artificial cell, rather than having all this gubbins we can say here's the cell body, here's its dendrites and here's its axon, really simple. In fact you can put a network of these cells together. Here's a whole cluster of them, in fact they can appear like that inside the brain, in layers throughout your brain, they cluster around like this. As I said you might have inputs coming in, so here are the axons from four other cells coming into this set of cells, and we might also have other inputs coming in. I'm just denoting these points here as the synapses in this mathematical model I'm constructing now, what the brain might be like. Of course as I said there are ten billion of these but to make this simple Iíve drawn only four. We can also imagine there are more inputs coming in from somewhere else - anatomists spend their lives trying to look at these laminations of inputs of each of the cells. And here's another set of synapses coming in here. And the way this works is the cell listens to their inputs, they listen to the other information coming in to them, and they make decisions, if there are enough inputs then they get excited and fire off information to elsewhere in the brain.

So if you imagine this pattern of ones and zeros comes in from this set of connections here. So this cell gets excited and this cell gets excited and they terminate these connections. And at the same time up here there's another set of patterns coming in, two patterns converging on the same cells. And these cells down here are integrating patterns of inputs. What these cells in this network are set to do are only get excited when they get two ones arriving at the same time. So I donít know if any of you can work out which cell out of these four is going to get excited? Here's the answer Ė the one at the end at the front. Now this could be something like, this particular cell might be something that likes the smell of offal with the colour red. And this is the colour red coming in, and this is the smell of offal and it goes 'wow, thatís great, Iím going to tell the rest of the brain!' and the rest of these cells like something completely different. So that could be how sensation works.

The question is how does our memory work? Well one possibility is, as I said earlier, is the change, the strength by which these cells communicate with other cells. In this previous example I just had any two ones converging on the cell would be enough to make it get excited and pass on a message. But if we change the connections in this network, if we strengthen the amount of input that is sent to this cell then we can change the dynamics here, so that these two inputs are much more likely to lead, to be active in the network. You can change and add a whole sequence of strengthened connections so next time we might only have the 'one' here come in but now we get different outputs because the connections have changed their strength. And if we expand this out a bit, instead of just passing off to the rest of the brain as the bit of the brain I'm going to talk about, the hippocampus does, there's a very amazing piece inside the hippocampus which is very highly self-connected. This cell, and this is an example of one of those cells, sends out connections to lots of other cells, but it also goes back and tells itself and all its neighbours that it got excited. So if you imagine these connections going off here and looping back round on themselves, you get this diagram here. And if anyoneís got a project booklet with them they'll find this rather obscure looking diagram hidden within it. I'll hopefully make this diagram a little clearer to you as to what might be going on.

Now we come to this Pattern Completion, this word we've titled our project. What is this Pattern Completion? Well the idea of this network, this is a network that scientists devised, is that you could store patterns in the strength of these connections between the cells. So here's an example where these black connections here are very, very strengthened. This has stored a pattern - it might be someone's first kiss, for example, could be stored here. But if you had a completely different jumbled set of black synapses here that were strengthened, this one might be strengthened, this one here might not be, you could create billions of patterns to go into this network, and it would eventually collapse, this network if you fill it with too many patterns will not be able to remember the information. But it's able to store a remarkable amount of information, despite it not being set up on a standard computer system of memory, this is a memory system called a Recurrent Neural Network. And it's in fact the network that Tom showed in his diagram. Hidden behind the scenes in a box underneath this in the basement in our installation, Pattern Completion, is a neural network very much like this, processing away on the computer, passing through this information looping. So the next thing is weíve got patterns and we're trying to complete them. That's effectively what scientists have argued the brain might do. It has stored this pattern and it has to get back to it. And scientists have tried to describe this by saying youíve stored the pattern and you need a small fragment, so for example if you go to a party and the whole experience of having a party and the next day somebody says 'Lucy', and you go 'Lucy? ... oh yeah!' and the whole party floods back. That's the idea, the Lucy node in the network was enough to create a set of patterns, this cell got excited and then it went off and re-excited all its neighbours who happened to be connected to themselves, so they re-excited themselves and they went off again and re-excited their neighbours, and then those cells went off again and re-excited their neighbours and so on and so on. This system loops, from being given the name Lucy it will loop endlessly until it reaches a fixed state - it reaches a minima.

So the idea is that it reaches a minima. I'm going to show you the image, up until this project was released, on the web of Pattern Completion, how they thought this process works. Here's an image, if you search on Wikipedia for Pattern Completion, here's a set of connections, they've stored this pattern, maybe this is someone's, the birth of their child or something, represented in this random pattern of connections. And they've also stored this pattern Ė you should be able to see what that is, it's a recognisable pattern. And they've said okay, these are the two patterns stored in the network, and the idea is that this system of this neural network is you can give it a subset of the inputs and say okay given that input, which of these two patterns will it go back to, given that reverbarence in the system? And it does that on the first loop ... which is not bad, it's pretty similar, on the third loop, the second reiteration gets this, almost exactly back to where it was. This network has managed to realised that given that it should go to this state and not that. And you can get it to do that for a lot of other patterns. And theyíre quite impressive these networks, these were discovered in the late 70's, 80's. And here's for example a really noisy version of this image, and within one, two iterations it's jumped back to that state. Itís moved to a fixed state in the network.

And in fact you can characterise these states as energy maps, of pockets of where the states could be. These are four different patterns in the energy surface. This is very abstract and actually is not a very complicated concept, but I'm going to need four members of the audience to come up here and help explain how this works. Can I get four volunteers?

Here we have this piece of scientific apparatus that my son Max works with every day, this is entitled a marble run. And it comes with four glass spheres, a bit like the spheres down below but these ones are called marbles. So I'm going to give each of these people here a marble. So what I want to do now is have each person to come up and place their marble in this hole. If you'd like to go first ... and you ...
That is the idea, exactly what's being shown here, a dynamic attractor. Each of these marbles is a bit like a memory, a bit like the word 'Lucy' going into your memory and it reverberating and it going round and round each of these different states and possible states the network has, until it reaches the fixed state in the attractor network.

But our memories aren't like that marble run in fact, there's not one fixed state they can jump to. You could give someone the word ĎLucyí after that party and they go 'yeah I remember that, there was Dominic there,' and someone goes 'no, Dominic definitely wasn't at that party'. It's quite possible that their brain has gone into the wrong fixed state, and they're remembering a completely different party. All these weird facets of memory scientists uncover, these types of neural networks have been trying to explain how the network can make the kind of errors that we have in our daily life.

So the idea of Pattern Completion is to mimic this idea of these attractor states. When you look at the installation downstairs you'll be aware of these spheres hung a bit like a layer between the hippocampus. And theyíre moving, like those marbles down that run, each of the spheres is an image of it that's moving until it reaches that fixed state at the end. When the scene across the spheres and the sounds that you're hearing are coherent and they form a final, fixed state. So they start out jumbled and as Tom showed you there's a Kahonen neural network filtering through, jumping through these reverberant states.

The next bit I'm going to talk about is this bit called the hippocampus. It's the bit of the brain that scientists discovered in the 50's if you surgically go in and remove it, which you don't do anymore thankfully, you'll wake up with a dense amnesia. There's a patient called HM who had the surgery in the 50's and from the 50's to 2008 he never remembered a single thing that happened to him after that surgery. So we know this bit of the brain is crucial for memory. And we think that hidden within this are the cells that are responsible for this neural network Pattern Completion. If we cut through the hippocampus like this, take a cross section all the way through it, it looks a bit like this. The next thing I'm going to do is explain well what do they mainly do? You heard that crackling activity earlier and these cells do lots of fantastic things but for me one of the most fantastic things is the way in which they can represent space and remember space. So I'm going to spend the next bit of this talk covering how our brain creates a representation of space. Spaces like this room, your bedroom, your bathroom, the street out there, how does our brain piece these things together?

Scientists do that by recording animals like rats or it might be bats or it might be humans in virtual reality, and they can record their movements in a world through a camera and they can record the electrical activity in their brain. And so for example this is a view now, we're looking through the camera, through the lens down into a box, into a space where a little rat is running. We can track where he's run, so he started here and he's run to this point in space. As he runs through this point we get a burst, one of our cells we're listening to gets excited and fires off activity to the other cells it's connected to. And the rat keeps running and he wiggles around, he's looking for rice in this experiment, and he keeps running around all over the place and as he runs we can map out where these bursts of activity from an individual cell occur. And you'll realise that there's a certain clustering, this cell gets very excited when the little rat goes into this corner of the box. And very unexcited, that cell really didn't give a damn about this bit of the box, or elsewhere. And if we map out where it most fired, where it was most active, we get quite a beautiful cone-shaped field of activity in the environment, so one cell in this hippocampus likes this part of the box. And it's called a place cell.

In effect if you stick an electrode into the hippocampus you can record lots of cells simultaneously, you can listen to the bursting activity in lots of them. And here's that one cell we had before. This little cluster of red activity here. But actually if you put on other cells, where they liked, they didn't like the same places as this cell. Each cell, and there are hundreds of thousands of them in there, seems to like a slightly different part of this box. And it doesn't matter which way the rat is facing, it just likes that place. It doesn't matter what it's doing there, it doesn't matter if it's peeing or eating or running around. So these are known as place cells and they exist within the hippocampus.

Now the next part of the interactive nature of today's talk is trying to explain some of the properties to you. I'm not going to get anyone out of their seats this time but I am going to ask each of you to act. You're all going to become place cells in the hippocampus and I want you to integrate information coming in to you that I'm going to present to you via this screen. This screen is your input. And if you get something you like I want you to raise your hand, your hand going in the air is your output, your burst of activity. The information you're processing is do you recognise the scene, the place that I'm going to show you? If you do recognise it I want you to raise your hand, as soon as you think yes I know that place, not that generic view but that actual place.

So the first image we're going to go to, raise your hand if you recognise it.
Okay so there are some clusters over here, there are a few people scattered here, there's a cluster over there, there's just silence here, some people here have absolutely no idea where that is. Never seen it before, didn't excite them. But other people got excited, and I'm the output cell gathering this. And this is Carlton Hill.
What about this? ... Very little activity in the hippocampus ... Okay, more is emerging, so there are one or two. This is a place that seems to evoke very little activity in this hippocampus here. And this is St James's Park.
Oh! Some places really do seem to create a lot of activity.
And this place? Complete silence ... one, two, five cells in the hippocampus, six? Taking a while for the brain to remember where this is ... This isn't Bavaria, this is actually Cardiff Castle!
This place? Two, three ... different clusters, getting little pockets again. This is The Meadows in Edinburgh.
This one? Oh a lot of activity in the hippocampus!
And the last image, where's this? ... I'm recognising a little cluster down here, it's a cluster I've seen before. This is Carlton Hill again in Edinburgh.

So there are several properties there, from that little scientific experiment, to get across about these place cells:

One is that each cell, so some of you liked multiple scenes there, each of you didn't like just one of those images, some of you didn't recognise any of them, some of you recognised more than one, some of you might have recognised all of them.

Seeing the same view, as we saw with Carlton Hill, some people I saw a little cluster that re-emerged in the hippocampus, the second time with a very different viewpoint.

And different groups, the main thing is the pattern was never the same there, I was getting a slightly different pattern each time across my hippocampus.

And this is really the groups of people, so if I was listening to you, listening to that pattern of activity, I will be able to tell based on who voted then, which one of those scenes you are actually looking at, without me actually seeing it.

And that's how we think the brain is processing this information, it's gathering these different bits of information in. And these cells effectively are like a constantly moving 'you are here' signal inside your brain, telling you 'you're by your bed, you're out in the street, you're in the Gimpel Fils gallery', they're constantly signalling that through the pattern of activity.

I've talked about this Pattern Completion process and these place cells. The reason I wanted to talk about place cells in the hippocampus is that the spheres downstairs, the reason we're interested that they're not flat photographs of scenes, because we project into the spheres they curve, they in essence make the place in the scene more of a place and in a sense that's what we think the cells in the hippocampus, they don't like views, they very rarely like views, they like places.

This is an example of the very few pieces of literature out there and this is a set of place cell maps. Each of these lines in this example here is one cell. This is some research by Tom Wills and colleagues in 2005. They recorded a whole ensemble of cells, here's shown eleven cells but they recorded up to thirty cells, in the same way that here we've got maybe sixty cells telling me information, they were doing it with thirty different cells voting on where they thought the animal was at any one time. And each of these cells we see the activity in each environment it visited. So there's a set of environments from A through to N. Environment A is a square box, they also introduced them to circular arenas, circular environments. So you can see this cell here for example it really likes the circle, it got excited the way you did perhaps on seeing Carlton Hill, but it totally didn't like the square, didn't fire a single action potential in this square. And the same is true of this cell here, it also didn't like the square but it really liked the circle. And each of these cells that likes the circle likes it in a slightly different place. And each of the cells down here likes the square, also in a slightly different place. But what struck the scientists in this study was as they slowly morphed these two worlds together, the square and the circle, as they slowly incrementally changed the wall shapes - you can see along the top here, they get slightly less and less square as you go this way and slightly less and less circular as you go this way until you reach some sort of middle point. But at a certain point, each of the cells in the network seems to say 'yeah, I'm pretty happy, that's still a circle, that's still a circle, that's still a circle Ė oh that is definitely not a circle'. And each cell in the network, generally the whole ensemble of cells all voting at the same time, jump on mass. They discriminate between the circle and the square despite the very, very tiny difference in the space. And it's the dynamics with which they switch, with which they jump between these states that made the scientists think these are place cells, this ability to know whether we're in Carlton Hill. The brain is jumping through this reverberant set of states. And on that note I'm going to stop.

[Question from audience] I'm really fascinated by this process. You said recognition, and memory of place and I'm thinking of aesthetics of place. Because I've never been to Edinburgh but I really recognise some of the architectural features, I have a kind of 'ooh' but I don't have a memory. I'm just wondering how many people have been to the place, or have an image of the place because they've seen it, I'm just wondering do you distinguish that?

[HS] That's one of those great questions when you come along as a scientist and it gets you thinking I could write a grant application on that! If I had been able to I would have removed that problem, but the fact that you spotted that is making me think well that's an untapped into area of research where do these cells have a certain property or is there some other bit of the brain that picks up on the generic Edinburgh-ness of a place or the very London-ness of an image. And we donít know, we literally don't know.

[Question from audience] Could it have been from memories of looking at photographs of the place? The last picture of Edinburgh I thought I knew, but I didn't, it was from a photograph. So you don't necessarily have to have been there.

[HS] Certainly, that's true. We remember whatever we're given in a sense and so you could remember something from a photograph and then later see it in actuality and go, wow, that's like I saw it in the photograph. That's not something that's been studied as far as I know, but itís quite interesting.

[Question from audience] There might be some issue with the fact that maybe new research is indicating that we feel before we think. And therefore somebody would feel, which would have huge effects, because most of the time during the presentation you've stated 'thinking' but there is evidence to say and show that you feel before you think.

[HS] That's another whole line of research in itself but you're right. When I talked about that process of going 'ah Lucy' and filling in the party that's what youíre alluding to, that's very conscious and I'm choosing it. We certainly think that this process of these cells jumping around in their network is occurring over milliseconds, it's very quick, that reverberant system is much more in a sense unconscious. And it's other bits of your brain that are interpreting that, and going 'well that just doesn't make sense' but that system doesn't, we think, do that. And there are patients who have amnesia, but their amnesia is confabulatory, so they are unable to inhibit these things and they think that for example when they're in hospital they think they're in a bank, and the doctors are their employees and they constantly tell them to go back and bring them money and they can't inhibit this failure to retrieve the wrong information.

[Question from audience] Are you able to artificially activate the networks? Can you use instrumentation to actually light up the network so that it remembers things that perhaps you wouldn't be able to remember without that stimulus?

[HS] Do you mean in a real situation, biologically, could you induce false memories into the network by stimulating them? Good question, again I don't know if anyone's tried that. Very tricky to do! I don't know if there are anyone here who knows anything about that? ... One last very quick question.

[Question from audience] I just wondered this idea about the programming and recognition of form and space, what implications that has for the notion of neuro-aesthetics?

[HS] Okay, I'll pass that on to my colleagues when they come to speak! We have other people who are interested in that.


[MN] One of the really exciting things about this project is when Tom, Hugo and I have come together to talk about some of these ideas from our individual disciplinary perspectives. Just there Hugo was talking about these ideas to do with place and to do with memory, and he applies his theories to the objects, processes and materials we've used downstairs to create our installation. Tom and I might be interested in how it feels to experience a memory and Hugo is interested in how it's actually working from a scientific point of view - and we all come together to explore this process. Loads of other practitioners in so many different fields are exploring these ideas as well and we've invited some of our friends and colleagues to talk to you this afternoon, to talk about their work and explain how they explore themes of memory and place through their practice. The first person I'd like to introduce is Helen Maurer, an artist who I've admired for a long time. Helen works with materials such as projected light and glass to create sculptural installations and video works that explore aspects of memory and place. I'm really fascinated by the way Helen manipulates these materials and combines them with found objects and architectural details in the places where she exhibits the work, to create new illusions, new environments and spaces that are really loaded with a sense of these placesí histories and pasts. So I think Helen is going to talk us through some of her projects and we're going to see some images and see some video works as well.

[HM] I've got a very bad memory Ė so I can't remember anything on my card here! I think it's interesting as an artist and also I teach part-time, what people make work about. I can't remember what I've had for breakfast, I couldn't remember who Jane was, or Laura -I don't seem to be able to retain anything! But I'm really interested in using memory as one of the things within my work. As Michaela said I often work with transparent materials and this is a view of some of the pieces that I use to play with and experiment with, to see what will happen. So I'm quite interested in process, in not having a clear idea before I start a project of what the outcome will be.

These are small models and for these particular works they're placed on overhead projectors. You probably remember those from school, a simple device that magnifies. So light passes through the glass scene and then [you see] the projected image. So you can see in the bottom right hand corner that the mug is placed on the projector. And you get a three-dimensional version and a two-dimensional version of the same space. You get different scenarios happening and I'm quite interested in that narrative. And those fragments, how they come together, it's almost like a slow animation the way they come together and change and shift and start creating something and it feels like trying to remember something, the process of making - trying things out and seeing what happens. There's also a different story, because of the nature of the way the overhead works that anything flat on the projector is upright in the projected image, so that in the model the boat is a wreck (this is called Cave Painting) and in the image is almost like a remembered view of that particular scene. And quite often things like the sea, I was casting wood to try and create a shed for something else and when I put it on the projector it became like the sea so youíre kind of given visual things.

Quite often I'm interested in there being a parallel experience. I think with projection it's a bit like remembering something, the actual process. I've made quite a lot of pieces with natural light, and like Michaela have used sandblasting as a way of holding the image. And with daylight sometimes, if I have an aperture and it's projecting onto another surface in a low light level, in hazy light level the image is quite out of focus and then the sun will come and suddenly it's focused. And I think there's a physical experience that I connect with in that process, and then it fades again, and that's very much like memory. Quite often when I'm making work I'm after a feeling. What I really want to do is go and kiss the wall, if it really works then I want to feel it Ė if a piece is really successful I'll want to kiss it! Also that memory isn't solid and it's trying to give form to something that's quite ephemeral. I always think of a Nick Cave quote, where he talks about trying to write God into existence, this idea of trying to find something through making, through creating.

How the work has developed is that I've become more interested in using real things, so rather than just having the model I wanted to sort of break down the boundaries between the projection and the real thing and so real objects have been introduced and it's this idea of what's real and what's projected. And how things are spilling in, I see the landscape spilling into the piece and trying to hold it back. And things like having two shadows of the boat projected, because you have a mirror passing through the object and the glass and the mirror both project so you get two boats from one source. Things like that all come through experimenting. I'm not sure of the science of that!

I think what's interesting about preparing for talks, it made me realise that space has always been quite an integral part of my practice. I've made quite a lot of site-specific projects. And recently I've just been working with some found objects. This was a chandelier that was in Manchester airport and I was given some of these drops to experiment with, to see whether or not I could make something for an art biennial at a stately home in Manchester. This was the original chandelier, a 60's Venetian chandelier, and then this is the condition I received it in! They were saying "have as many drops as you like!" you know, and when I went up there they were really broken. Two had been restored and then I just had this crate of glass to bring back and to try and make something with. And also to think about the new context and the history of the place and quite often I think it's about what the place brings and what it means to place these objects in a new environment.

This is the space in which the work was situated. It has an atrium, and the whole place is incredibly formal, the stately home. There are lots of views, I think walking around the grounds there were lots of interesting things that fed the work and one of them was how things were planted. There are two lakes for example and they're planted in a particular way so it looks like a river, so there's all this kind of play with scale, and apertures and views. They even bred sheep ten percent smaller in this place to create this sense of grandeur! So I was quite interested in the symmetry of everything and this framing. And this particular space gave you a long view and three doorways where you saw the work, and this view above.

When I went there I was talking to somebody whose father reinforced the ceilings at Manchester airport for the original chandelier. And she was telling me that 65,000 parachuters were trained at Tatton Park. And because of the idea that there's a landing, that my work was going to be placed and they're drops, there were lots of connections that I made. I also laid out the piece, it's almost like having to do an autopsy on this thing, when it came out of the crate. And the crate was kind of flat-packed. So in the end I used whatever drops I could recover or mend, and had them descending. I wanted to give the piece a gentle landing because it had been badly looked after. Everything in the house is really well taken care of - if you have a goblet there was a crate for it. So I thought that was what the place brought to this object. The title is Light Landing. I wanted the piece to land originally, and because for various reasons it didn't really transcend itself by lowering the piece, in the end I wanted to re-configure all the original elements. So I experimented with moving the light up and down. So what you're going to see in the video is the idea of it being re-configured really. The light bulb takes two minutes to go up and down the space and it lands the shadows. It was interesting meeting Michaela and talking about animation, because I'm really interested in animating still things, that seems to be persistent within what I'm thinking about within my practice.

There's a second piece called Set Dance, which was in an old Winter Garden in Eastbourne, again another site-specific project. And I was given the space for a week. It's usually an incredibly busy space, there are lots of different events and turn-arounds, there might be a darts conference on one day, and there are tea dances and I became in this idea of rotation. The space was originally a roller-skating rink and a Winter Garden. And when I went to visit there were lots of objects left lying around. I've always tried to make the most from the least really, and so if I can use found objects and they already have something, connections, then I'll try to do that. So I wanted to use things from second hand shops or things that were left in the space. And it's a theatre, it's got lots of different things, and there were lots of things lying around, I thought it would be interesting if the space is shut for the week what would happen if these objects got together and created a piece?

So there are the overhead projectors and the screens from the theatre. We found an old music box in a second hand shop and that was used to create the soundtrack. And it was quite interesting you could hear that before you experienced the piece and because you've got three light sources you've got three different projections happening simultaneously of the same objects. And there are two flowers, they became like two ladies dancing together. At one point they cross each other on the stage, because you get two projections from the one flower. It's often the things you can't anticipate that become the most interesting. These two flowers crossing each other on the stage, they were quite shaky as well and because of that context it was a bit like the tea dance.

[Introduction of Dr Peg Rawes by Hugo Spiers]

[PR] Thanks very much. I'm not going to talk for long as I know you've been here for quite some time. So I'll just make a few hopefully connective comments. As Hugo said, I work quite often with discussions around philosophy and this is really how the link between conversations Hugo has been looking at about perception and memory and space have linked in with my interests. And I really want to introduce just two discussions around aesthetic theories of space and time. Not actually from the contemporary period at all. In a way I'm taking you back in a historical sense to the eighteenth century and also to the late nineteenth century. Because, unlike the contemporary sense of neuroscience providing some really firm, concrete understandings of how the biological concept of sensation, and of a science of sensibility if you like, operates, these sciences of sensibility have actually been undergoing different transformations and developments in philosophy for a very, very long time. Those of you who know of philosophy and history will know it's not a new thing. And so in the eighteenth century there's an incredible sort of intensification of this and I really just wanted to say a few things about a couple of people who were working at that time.

The first person is Immanuel Kant. What's interesting about Immanuel Kant is that he investigates or constructs this science of sensibility that he calls aesthetics and what he is doing is attempting to construct an incredibly precise understanding of what constitutes not a formal cognitive process of thinking but something that he calls aesthetic judgements. And these notions of judgements are actually something really quite different from our notion of logical, of procedural or conceptual thought. They may be much more in keeping with concepts of percept. And he calls it aesthetic judgement. Now what's interesting about these discussions on aesthetic judgement is that when he discusses these he uses concepts of space particularly in order to illustrate this, and in particular in relation to the notion of the imagination. The imagination is not a term that is necessarily used in terms of a formal analysis perhaps of place cells and their activities, but nevertheless there is an understanding that the imagination is an operation that is also resident, or it comes out of, the hippocampus working. And what Kant was doing was looking at how the imagination engages with, or allows us to engage with our external environment. And allow us to produce both productive and also reproductive images. This particularly has been taken into a discussion of the aesthetics of taste, but he also talks about this analogue capacity, which is a kind of agitation. So the notion of constructing a sense of place or a sense of self in this aesthetic experience comes through a series of sort of agitated approximations that the imagination attempts. And when he talks about the sublime, the agitation actually fails. The imagination becomes this kind of inadequate state. But nevertheless at various times in his analysis of the imagination, where he really engages in it as a very productive and powerful mode of engaging with the world. It is not thinking, it is a different kind of process. And it's also particularly obviously related to the image, the construction of images, which are distinct from our conceptual, verbal understanding. And this is something that he talks about particularly in a later text called The Anthropology from a Pragmatic Point of View, which is a wonderful book where he talks about rainbows and moonlight and really quite ambient and quite altered states of reality.

So in terms of the eighteenth century there's also another group of people who I'm struck by thinking of today who were called the Jena romantics and they were led by someone called Friedrich Schlegel and they came out of discussions about aesthetics from German Romantic thinking. I think what's interesting again is this is almost this precursor, sort of prototype of some of these discussions. They were really trying to engage in concepts of nature and how we operate in often very immersive states of embodied modes of activity or senses of the world. But what they did was to construct these poems, these fragments, which were incredibly meticulous in their minutiae they were very small, almost aphorisms, Iím particularly thinking of work by Friedrich Nietzsche. One of the things that's quite interesting is that this idea of this fragment almost constitutes a sort of infinity, a sense of totality that is more excessive than its size, than the fragment that is there. So there's a sense of them being a very sort of minute sense of the world, but its capacity to immerse you, to give you a sense of the feeling that is way exceeding your own bodily limits was something they were really fascinated with. They were also quite interesting because they were really argumentative and they had massive fights! And they were quite an intense group in their own right they didnít sort of operate within certain codified behaviours, they were disruptive and they sought disruption.

Now I want to come forward another hundred years to the philosopher who's particularly being discussed in terms of discussions on time and memory and that's Henri Bergson. Bergson for me is interesting not because he is just purely a person who discusses time and therefore supersedes the notion of space, and this is a way a lot of people respond to him now, particularly if people have read him through Deleuze's work, he tends to be seen purely a philosopher thinking of time. But Bergson actually trained as a geometer so his concepts of the importance of time come out of his study of mathematics of space and of topology. Particularly he's working after the time when Riemann has been dealing with very wonderful developments in mathematics, topological thinking. And he's also working at a time when Hertz is really making incredibly interesting strides in terms of audio understanding and electromagnetic construction of materialisations of the world. So I think he is an interesting individual because he's also very connective, he's someone who understands that though there may be orders of thinking, of prioritising space above time, actually the reality is that there could be other ways in which these intuitions link. The other reason to mention Bergson is that like Kant, though in complete disagreement with Kant, he is fascinated by concepts of time which he calls durée he is interested in this but also in terms of the concept of matter as notions of intuition again of concepts of engaging in the world or sensory perception, which is not just a notion of cognitive thought.

And one of the things which I also think is interesting about Bergson is that he was thinking about the construction of images in sound. There's an essay called Time and Free Will where he talks about how the concept of memory is actually something that is very specific in its construction as a sonic image. And he talks about the swing of a pendulum being something that can actually be accounted for as a series of instances of moments which he then uses to illustrate the tolling of a bell. And what he is interested in is talking about time not as something that is sustained or a series of representations that can be captured and fixed, but that time and memory is something which is a constant, vital concept of how we engage in the world. And he's in many ways seen to be attempting to describe very precisely notions of life. So in some ways a kind of precursor again to a kind of biological interest in how we engage in the world.

The other thing that some of you if you know Bergson's work may have picked up on already is that there is this iteration if you like of how memory is constructed, which he describes in a conical form of emergence and so it's very interesting to hear that you both use that term and it would be fascinating to know whether Bergson is considered to be one of the precursors of that, or if he's there some where? I've not come across that. But certainly it's a notion of how images are made, not in a sequential or linear form but through this very conical form of intensification that moves both upward and downward through a state of different levels of intensity different kinds of calibration. I think the other thing to mention about Bergson's discussion and the installation downstairs is that one of the things Bergson is interested in is the construction of the body as a result of these relationships to memory and image. And he talks about this in terms of the notion of the body-image, so he constructs the concept of the body out of these experiences. And one thing I think is interesting about the sound images and obviously the visual images in the installation is that they are both embodied and disembodied. There is both a reality there and also a simulation of reality. And I think that given this interesting discussion of how the digital/analogue relationship works, particularly of course if you go back to the zeros and ones, there's a really quite fascinating link into the notion of the virtual, which is something which comes into that discussion. Okay, I'll stop there.

[MN] Now, last but by no means least I'd like to introduce Andrew McDonnell who is a writer based in Norwich. When we were thinking about our compositions with our photography you'll remember I said we were really interested in clearings and pathways and ideas of leading your eye on a journey, and ideas of trains of thought and creating a sense of openness where your imagination or your memory might be activated. There are overlapping themes with those ideas in Andy's practice. I know that walking, pathways, clearings play an important part in his creative writing process. And there are also other themes to do with memory and dream-states that he'll pick up on as well. Andy is going to perform one of his own poems and possibly some fragments of poems by other writers.

[AM] In a nutshell, I'm halfway through my first year of my PhD in Creative and Critical Writing and I'm particularly interested in walking's relation to the speech acts like poetry. So for example I'm interested in Situationist / Interventionist walking, the idea of déreve, the idea of psychogeography, as a way of understanding the city through events and through emotions and feelings rather than through its cartography, through its actual physical map. The other thing I'm interested in is something called the fugue. The fugue is a psychogenic disorder first diagnosed in the mid-late 19th century. A chap called Albert Dadas, who was a gas fitter, kept disappearing from home. There's a fantastic book on him called Mad Travellers: A History of Mental Transient Illness by Ian Hacking and it's worth tracking down because it talks about how he walks for miles away from home, doing 70km a day, couldn't remember where he'd been, ended up in places like Moscow. Brought back, under Dr Tissie, hypnotised he told the story of how he'd been away on his adventures but he had no recollection unless he was hypnotised, it was very strange. So I'm interested in that idea as well. And what I want to do is talk about the idea of poems and clearings. So if you think about the idea of poetry being a periphery, our edges, and that when a poet wants to say something, the audience on the other side of the edge and the message arrives in the clearing between the two. I was also thinking about asyndeton, which is a linguistic term, which talks about ellipses and spaces in between. For example on the underground we travel between stations, so we have an idea of the city as based on an underground map, particularly in London if you donít live here you have an idea of space so the actual points of reference between each station is how we base our maps around the city.

What I want to do is read a little bit from a Penelope Shuttle poem, called Building a City for Jamie. Here she tried to recollect poems from memory of dreams that she'd had and in his particular poem there's this character called Jamie, we're not sure who he is, he's like a child-like monster a very strange character who she's showing maternal instincts towards. I want to read a couple of bits, I want to get the idea of the clearing across, the idea that she's bringing something to this clearing and we as readers are trying to understand what's being put across. So this is from the second part:


I lean over the plans of the city,
explaining everything to Jamie.
'Your city will have wish-bone weather,
the happiest backstreets anywhere in the world,
with houses painted in such anticipatory colours,
freshly-caught fish, lily, young green, high flying blue!
Here the moon will be no stranger,
a heart-shaped wall will surround Jamie's city ...
Look at all this scaffolding,
the cold cages of steel
jambs and stanchions,
the webs and ichors of architecture,
dog-shat-on-sand-heaps, stacks of bricks,
sketched out foyers, half homes, spires sighing
and moaning to be in their rightful places ...
See it all roll towards the radiance of the built, Jamie!'

'Oh!' says Jamie,
opening his eyes very wide.


So here in a sense she's getting across the actual construction of the city to this imaginary character and also to the reader. And then what's interesting in the last stanza is that she seems to have awoken perhaps from the poem, from the idea but Jamieís still there, and she has this warning about memory in there:


'This is your city, Jamie.
Here it is.'

Oh Jamie, you are getting sleepy and not listening!
Snuggle down in my arms,
safe in this city that I built from memory.
Tomorrow when you wake in our rooms of interesting squalor,
asking: 'Was there really a city, a city for Jamie?'
Iíll say lightly, 'No, you were just dreaming, baby. There is no city.'

'No city!' youíll frown, shivering,
as I shake my head and lie to you for your own good,
'No city?'

No city. Of course not.


This is something I've discovered, this sense that when a poet works from memory or bring back ideas of dreams, they also bring back warnings about memory and how we use it. Another poet I want to read from is W S Graham, and there's a sense of an oubliette, I was thinking about the hippocampus earlier and the sense that there's almost this memory prism sometimes, there's a danger of being trapped within that and what that means. I think what happens in poetry, they come across these ideas, they obviously meditate on them before writing and some strange things occur. Iíll read through it and I'll explain a bit about the typography at the end. This is called Loch Thom:


Just for the sake of recovering
I walked backward from fifty-six
Quick years of age wanting to see
And managed not to trip or stumble
To find Loch Thom and turned round
To see the stretch of my childhood
Before me. Here is the Loch. The same
Long- beaked cry curls across
The heather-edges of the water held
Between the hills a boyhood's walk
Up from Greenock. It is the morning.

And I am here with my marnmy's
Bramble jam scones in my pocket.
The Firth is miles and I have come
Back to find Loch Tom maybe
In this light does not recognise me.

This is a lonely freshwater loch.
No farms on the edge. Only
Heather grouse-moor stretching
Down to Greenock and One Hope
Street or stretching away across
Into the blue moors of Ayrshire.

And almost I am back again
Wading the heather down the edge
To sit. The minnows go by in shoals
Like iron-filings in the shallows.

My mother is dead. My father is dead.
And all the trout I used to know
Leaping from their sad rings are dead.

I drop my crumbs into the shallow
Weed for the minnows and pinheads.
You see what I will have to rise
And turn round and get back where
My running age will slow for a moment
To let me on. It is a colder
Stretch of water than I remember.

The curlews cry travelling still
Kills me fairly. In front of me
The grouse flurry and settle. GO BACK


At the end of this poem Graham has written the 'GO BACK GO BACK FAREWELL LOCH TOM' in capitals. He was interested in how language could work anyway, and he did a lot of language games, almost like cryptic crossword puzzles with some of his poems. But also what's interesting about this poem, written in his later life when he was living in Cornwall, so it's remembering Greenock, which are almost two opposite ends of the country to each other. It's an interesting sense of memory and how we recall things and the danger of recalling, the danger of nostalgia, the dangers of the melancholy of remembering and heís saying maybe we should keep going in the ongoing moment.

The last thing I'd like to read is a quick poem of mine, which is in the book that Tom and Michaela have published for the event. It's called The Fetch. And this is a character that has perhaps not heeded the warnings and has gone back to try and find the city that they loved and remembered and then they're confronted with this monstrous version of themselves and nothing else. The Fetch was originally performed with a poem called The Doppelganger, The Doppelganger is the same poem line by line backwards and it was written for two voices, the idea being that it's done by mobile phones to an event and two people walk in opposite directions and then come back again, the poem dictating how far they walk.


I tried to remember
how I loved in this city;
I tried to fetch
back memories
from a state between sleep
and the moth-bothered lamps;
I slipped out the city
and it slipped from inside me
from the suitcase I carried at my side
from my mouth and from my eyes;
last night I said goodbye to the city
to the town planners, the marching bands
and the Sunday Traders
to my wife and her canary
to the vertiginous music
to the smoke free bars
to the giant strawberries
to the floodlit car parks,
the graffiti, the pigeons and cycle paths
and soon I was crossing dual carriageways
out into the fields
and then on further into marshes
taking a line
drawn from overlaid maps
like scarves over mirrors
they would show the way,
to fall between worlds
- I was told the droving lanes
would suddenly appear
the greenways, by-ways,
the ghost lines
the bittern booming in the reeds
of Poly-Albion -
but the maps were no longer
in my suitcase
the city had held them back.

And then it appeared to me
a fetch, a doppelganger
of the city I had loved and left
and like a face in the water
it looked at me strangely
as if it were my hands
which had been trying to drown it
in itís own reflection,
and stood before me shyly,
not wanting to catch my eye,
so I entered in and my
suitcase became heavy
with new maps and possibilities
but when I checked them
there were blueprints of buildings
that had never been
tramways and railway lines
dropped in my lap like spaghetti
and I realised the city
could never be recovered from memory
it was a ghost train
without the ghosts, or I, a giant
rampaging in toytown
that had long before seen
the last eviction
and in the empty streets echoes
I caught a glimpse of what was me
as I turned like Goya's monster
with myself
between my


[Excerpts of Building a City for Jamie and Loch Thom taken from Building a City for Jamie, Penelope Shuttle, 2003, Oxford Poets and W S Graham Collected Poems, Matthew Francis (Ed.), 2004, Faber & Faber]