Fiction and Poetry 3am Magazine Contact Links Submission Guidelines


3am Interview


"In a couple of seminars I've switched on the spot from arguing about the rightness of my algorithm, my approach, to saying please don't do anything to anyone on the basis of what I'm saying. All I've got to say is that the people who look at scientists' work and choose how to run their lives on that basis are probably making a mistake."

Richard Marshall interviews Daniel Glaser, Scientist In Residence At The ICA


3AM: A scientist resident at the Institute of Contemporary Arts in London. How did you end up here?

DG: I was born and bred in London, went to school in London, went to Cambridge -- in Cambridge you can do a part one and a part two -- you have to get a degree. I did a part one in pure maths, which was three years, and part two in English literature which was also three years. It's one of the advantages of the Cambridge system that you can do this radical switch. Also, you probably couldn't switch from English to maths because part two in maths depends on part one. Whereas part two in English doesn't depend on part one. So there are preferred directions for the switches. But it was ok for me because I did an English A level with maths and physics. That seemed a natural direction to go.

I guess I made the switch partly for social reasons. I didn't feel all that happy in maths lectures, looking around I didn't feel I connected with the people who went to them. It was also partly for the positive reason that I've always been interested in words and literature. The people I could study with there were Eric Griffiths and Jeremy Mall and their thing was what they would call historical Empsonianism. Basically, it was about how language works, practical criticism, reading things very closely, but trying to understand how the piece of text works in an historical context. It's driven by the words and the structure of the text, but you have to cite it properly. So that attracted me and I really enjoyed learning to read.

Then I was a waiter for about a year. At the end of that I decided I wanted to go back into academia, so I went back to do a masters in cognitive science at Sussex. In one sense it was the heyday of collectionism, which was neural network research: that's the idea that you look at computers and what they can do and you look at biological systems and what they can do, and you find that there are lots of things that biological systems can do a lot better -- even simple biological systems can do an awful lot better than a computer. For example vision, locomotion, moving around, understanding speech, making their way in the world. Computers are very good at number crunching -- and this is still to some extent true -- but they're bad at processing natural scenes and sound.

An example is that the tape recorder you're using works much better in a quiet room than in a noisy room and there's a documented phenomenon in brain research called the 'Cocktail Party Effect' which is that a biological system is good at picking out a conversation at a cocktail party and you can do that because you lock in to the content of the person you are talking to and very successfully block out all of the other stuff. Tape recorders and computers generally are very bad at that task. So what the connectionists thought was maybe the reason biological systems can do better than computers is because of the structure. So conventional computers have got a processing unit that gives a series of steps and what a biological system has got is a series of neurons and lots of connections, a web of neurons, and things don't go through in steps but rather there's a kind of parallel organisation and all sorts of stuff like that.

So they tried to build virtual machines, simulations of machines, which had that structure of neural networks and then used these networks to attack every problem under the sun. Vision, language, planning -- everything. Artificial Intelligence, which was the previous discipline, had become Neural Networks. That was one thread. The other thread is the philosophical thread. So Maggie Bowden and Andy Clark who alternate between Sussex and the University of Washington, they were cognitive philosophers, they were doing philosophy of connectionism. It's the case that certain technical developments, the rise of certain types of machine, give rise to certain philosophical issues and an example of that is in the history of representation.

I'm not a philosopher so I'll give it in a cod way. As I understand it, the issue of representation is how we associate things in our heads with things in the world. So I see a flower, I have a representation of that flower in my head, and you can use that as a representation in some ways. You can recognise flowers. Some would say you couldn't see flowers if you didn't have a representation of a flower in your head, maybe that's what seeing a flower means, and so on. Now what connectionism showed you was that you can have something like a representation of a thing, which is a bit inside the machine that is always activated when that thing in the world appears, but it's not an explicit representation. That is, you can't point to that bit of memory of the neural network that represents the flower.

Now contrast that with the conventional computer. It's got a processing unit and it's got memory and you can point to the bit of memory or the bit of the disc which stores so and so's phone number, the picture of the flower or whatever. That's what you might call an explicit representation. There's a direct relationship. You can point at it. To the bit. Connectionism showed us that you can find and prove mathematically in many cases as well that you can find a distributed representation of something where you can't point to the explicit bit of the memory system that has it. Rather, it's in the activation of a thousand neurons. So that was the appeal of that insight. And it made philosophers think about the idea of representation in a different way.

So people became philosophers of connectionism, guys such as Xenon Pylyshyn talking about features and concepts, and representational examples of classes and stuff, there was some conventional philosophy of representation grew out of these empirical facts. So that was that. There were two things going in Sussex for me: neural networks and the philosophy of mind. At the end of Sussex there was a fork in the road. One option was to try and be a philosopher. And the other was to do something with actual intelligence systems. So the person who influenced that decision was someone whose grandmother was the best friend of my mother's grandmother. His mother was the best friend of my mother . . . anyway, four generations, stretching back to Germany in the early years of the twentieth century. He's a professor at Sussex and he said to me that he thought that philosophy was an old man's sport and while I was young I should go out and do some real work. By real work he meant neural biology. That means looking at animals' nerve cells and that sort of thing. So that's what I started to do.

What I was doing then was trying to find a way of looking at neural networks in biology -- not models of them but looking at some real ones -- so that was the turning point. I didn't try and become a philosopher. I became a neural biologist. Then I had to look for a place to do it. And there was a happy coincidence. My mother was born in Israel and I had always been curious about going there and hanging out. So I found a laboratory out there and found a guy out there who was doing things that I thought were going to be relevant to what I was doing. I went off and did a PhD. It was looking at networks of neurons in biological systems. A technique called optical imaging which is where you use an actual camera to look at the actual brain.

One technique relies on the fact that brain cells that are more active need more oxygen to supply more energy. And blood with more oxygen is redder than blood without oxygen. Blood without it is blue. So if you look at the colour of the brain, if you look at actual arterial blood going out from the heart rather than blood on its way back, so if you look at the colour on the surface of the skin, Venus blood, that's blood coming back, it's done its work, so by definition it's had the oxygen taken out of it by whatever organ it was supplying, but if you ever see arterial blood it's a vivid red. A very bright colour. It's not just pastely red, it's really really red! The colour of the brain if you look at it in detail tells you which bits are active. So you look at the brain, you show a stimulus and you observe which bits of the brain demand more oxygen. Which bits become redder. So you can look at the underlying neural activity directly.

Now the interesting thing about that story is that I made a good decision in my life but for completely the wrong reasons. So, I thought I was going to be looking at neural networks and that's why I was motivated to go to Israel, but it turned out that my neural network experience didn't play a role there, but I got out there. That ended up as seven years. Looking at the visual system, And that was trying to do functional architecture. Which is where you try and divide the brain up into small regions about a millimetre cubed which contain less than a million but more than a thousand neurons and trying to work out what each of them does. Some of them will be looking at colour, some movement, some will be looking at lines, horizontal lines, vertical lines, and so on. That's what I was looking at. In parallel with that, the place I was working raises a lot of its money from individual donors. We had a lot of them coming through the lab. So I was explaining what I was doing in the lab to them and had a spiel that I developed. That I enjoyed.

It also did a thing involving school kids from all over Israel. There were parties that came to visit the labs and also there were people who came to a regular science club. I was quite involved in them as well. Giving lectures every so often. I found that very rewarding. So explaining what I do to both scientists and non-scientists was, in simple terms, something that I really enjoyed. That was a strong motivating thing, as well as being a neuro-biologist which I was in Israel for seven years. I went into the army. Which is another story. I chose to become an Israeli and I became a paramedic. A four-month course. When you are an old person like me you get put into a reserve unit. There have been a lot of people in Britain, actually, who have been pulled up in the war against terrorism into territorial army units and they've been complaining terribly and their employers have been complaining terribly.

It's quite funny really because the whole point of the territorial army is that if you're needed you get called up. In Israel if you're in the reserves, you do a month a year of reserve duty. So my unit now in Israel is seeing a lot of action. I think what's going on there is incredibly depressing. My politics are very left so I feel frustrated at not being there to help pull the country back from the brink of something awful. And that's actually a characteristic of someone who has become an Israeli -- that, paradoxically, the worse the situation is there, the more you feel you should be there. So I qualify on those grounds! But what I chose to be at that time was to be a paramedic. Which was very useful in that context. That was enjoyable.

But the ironical thing about that is that the only time I have ever really used my paramedic skill was in London one weekend. I was in Soho having a coffee and was a couple of hundred metres from the Soho bomb. I was the first paramedic on the scene. That was really like a war zone. So then I came to London and carried on with neuro-biology. Then more recently in the last year or son I started working in UCL on cognitive neuro science. I'm doing brain imaging. One day I got a call from a friend of mine who I've known for a few years saying 'There's someone I know who I'd never met or heard of who works as director of talks at the ICA, Rachel Cottal,and they've got this money together from the Wellcome Trust and also from the Gulbenkian Foundation and they've been thinking that whilst there have been a lot of artists in residence in a lot of scientific institutions there haven't been scientists in art ones!

If you look up scientists in residence on the internet you'll find a lot of American schools that have them, a few companies, but in terms of Arts Institutions as far as we know it hasn't been done before. Not extensively anyway. In some senses it's like the revenge of the scientist! For years companies have said, let's get an artist in, and in the board meeting the artist is seen as being a great investment because they're seen as thinking differently. And they cause havoc. The issue or the question is, there has been a lot of art and science collaboration and . . . Well, there was a nice moment where someone was making a presentation on this saying something about art-science collaborations in the mid- 90's and there was this, not irate but strongly worded comment or question at the end of that talk which said that, actually, in the 1960's we were doing stuff like this so young people thinking that they're being innovative are actually doing stuff that's been around a long time. It's clear though, partly because of the Wellcome Trust, that there have been an increasing number of art-science collaborations going on. One comment you could make about some of those is that they've been trying reasonably to make a thing that is both an artistic thing and a scientific thing.

And the problem can be that you spend a lot of time worrying about definitions of what is art and what is science. And also, you can get a little tied up with evaluation criteria. Should I be looking at it as a piece of art or should I be looking at it as a piece of science? Is it in fact rubbish in both senses where both people know what the other is talking about and so on? An interesting point that keeps coming up for me in all of this is that the disciplinary boundaries between art and science are actually mirrored within science where there are a lot of boundaries and barriers of the same type. A mathematician talking to a biologist has quite a similar experience to a mathematician talking to an artist. You don't speak the same language, you don't have the same points of reference, and so on.

What's interesting is in neuro biology where you're studying the brain or nervous systems, there has been quite a lot of emphasis on inter-disciplinary work, and that means mathematicians, biologists, neuro-scientists, whatever, getting together. And there's a danger with that that you end up with no one knowing what they're talking about. So again, I've had an interesting experience where I went back to Cambridge to talk to one of my old maths lecturers, and I told him about the problems we were having in these networks we were building in Sussex. And I described what we were doing and he couldn't believe it.

Because he said that was tried in the 1970's and I can give you the papers to show you that it's a really stupid idea. He was saying that he couldn't believe that there were serious artists who were using this technique. It's a perfectly reasonable first thing to do. But we've been there. So that's a problem with interdisciplinary work. The worry is that people don't know what they're talking about but that is a necessary consequence of the creativity that is implied. The idea of a scientist in residence is slightly different. I get to be a scientist by a conventional process. I am senior research Fellow at UCL in the institute of cognitive science. I get paid to be a scientist. So I'm a scientist. The scientists worry whether I'm a scientist or not. The ICA has the word Art in its title, so having a scientist in it is a science-art collaboration. So you get around the worry of defining what we are doing, it doesn't need to foreground definitions. It allows people who have expertise to define what's what.

3AM: So is your role to introduce people to your ideas or to ideas knocking around in science generally?

DG: One of the liberating ways about how I decided to do this work was actually based on advice I got from my sister. She's like me in that we both did PhDs and are interested in public engagement. She chose to be a journalist and so works on 'Start The Week'. She chose to be a person who works in the media and the arts -- her name is Danielle Glaser -- with an academic pedigree, and I chose to stay with academia: it's a matter of who pays your wages. But we were both wanting to communicate with people. That's what turns me on. It's a fickle world. The more exposure you get in the media the more the media tries to knock you down. I think that I'm never going to be very much in the media world, but I worry about my engagement with that process. I don't want to be measured in those terms because they're very arbitrary and locally driven.

I'm very happy spending most of my time in academic research. It's great to have this terribly exciting thing on the side at the ICA. It only takes up a day a week. A few e mails, a few meetings, a few evenings. The ego thing was to let go of my own work. Of course its self-promotion on one level. I'm presenting the stuff and that's the ego trip, but I'm disconnected from selling myself and my work on the brain. That's not why I'm here. That was in the job description. After all, there are many perfectly respectable senior scientists who are experts on the brain who get to go on TV and talk about the brain and they're world experts. I know a lot of things about the brain but that's not how I got here. I got here because I'm relatively young. Curious and eager. Rather than being an established figure.

But what I do try and tell people about is what's going on in my part of the world. Especially since I work on vision and the brain and all that stuff there's a lot of curiosity in people about what's going on. Although I have to say that there's a lot of scientific curiosity about scientific stories with the public and although my area of study is quite fashionable I think that somebody who can explain why beer froths or why candle flames are half an inch tall and not three feet tall - people are interested in those kinds of stories in any case. So what's important is your ability to explain stuff not the fashionability of what it is you're explaining.

So partly what I'm doing here is telling people science stuff but if it's content stuff, my content stuff is the brain, that's what I know about. Another important element is telling people about the process of science. The reason why that's important is because one might say we're in something of a crisis in terms of people's views of science. The contemporary example is the MMR vaccine. There are people who think that because the British Medical Association has published a report that shows that there is no link between MMR and autism means that there's something to be worried about. Which I think is totally fucked. It's a really, really serious problem.

3AM: Like people and the lottery where people don't know about the numbers and the probability of winning that can be calculated. So the Lottery becomes a tax on the illiterate.

DG: Yes. So one of the events I'm setting up here is a thing on scientific risk. So I've got the former president of the Royal Statistical Society Robert Winston -- a medical scientist, a social psychologist, and the chief scientist at the Department of Transport -- talking about that issue. The question they'll be discussing is: is it really about quantifying the risk and then explaining it to people? How much do we know about the way we think about risk and assess risk. How much of that can be fed into a better interaction between scientists and people and government? I think it's dangerous to try and change the world especially through education. Education as a goal is good, but changing the world through education is problematic. But I still have leanings towards trying to do that!

3AM: So you're a dangerous man.

DG: Very dangerous! There's a whole load of science about and I think this is part of the story I'm telling which is there's a parallel between medicine and science -- I know about science -- scientists are seen by many people as having access to Truth. Now what's that? That's a misconception. We were talking earlier, before the interview, about Popper, so one of the things he showed was, picking up on Hume, what he showed was that you can't prove anything. It's just correlations. The fact that the sun rose every day for as long as anyone can remember doesn't mean anything about tomorrow.

3AM: Think about the turkey on the countdown to Christmas.

DG: Exactly. A good example.

3AM: Bertrand Russell.

DG: So what Popper said was, ok, that's fine. All you can do is try and falsify things. Now actually, most scientists don't engage with that way of thinking. That's for philosophers not scientists. But that's what seems to be going on. So here's an example. As a brain scientist, I spend a lot of the time thinking about the statistics of brain imaging which is about how can I reliably come up with an accurate prediction about which bits of the brain are active on the basis of this data. So we took these pictures -- we put the head in the scanner, got them to count to a hundred, we saw this bit of brain activity, the question is: can we say on the basis of that experiment that this is the counting bit of the brain? And that's just a statistical question.

Now, since I work closely with the institute of neurology and they're connected to the hospital, and every so often someone from Great Ormond Street hospital will say 'We've got this group of kids and what you're saying is really interesting, and we're wondering whether we can give this new drug to these kids? We want you to do a brain scan and decide which kids to give the drugs to on the basis of that.' And I'm saying 'Whoa!' And I'll switch. And this has happened in a couple of seminars where I've switched on the spot from arguing about the rightness of my algorithm, my approach, to saying please don't do anything to anyone on the basis of what I'm saying. All I've got to say is that the people who look at scientists' work and choose how to run their lives on that basis are probably making a mistake.

That's why there's often a difficulty with being both a doctor and a scientist. People who try to be a both are often in trouble. A doctor's job is not to access Truth or anything, but to treat the patient. The risks are set up very differently. So interestingly, on the radio, in the middle of the Foot and Mouth epidemic, the head of the National Farmers' Union came on the Today programme and said he was really disappointed in the scientists because a year ago what they were saying was x and y and now what they're saying is z and w. And he was saying he wanted some right answers from scientists as to what the right thing to do is. If he's talking about scientists then there's a public example of what I believe to be true in general of a mistaken impression of what scientists can do. So this is a rather long-worded way of explaining that part of my role here is to explain the brain which is what I work on and part of my role is also to explain what scientists do all day. And the kinds of things that scientists are comfortable saying and what they're not. That's cultural partly. And political too.

3AM: So what is happening in science now and how are you being received?

DG: There is a curiosity among non-scientists about science. There is an asymmetry in that artists are more interested in science than scientists are interested in art. That's quite interesting and is probably about the power relationship in culture between science and the arts. Science has this funny privileged status right now which isn't necessarily that healthy. But that may account for the imbalance in curiosity. I think it's important for scientists to get out more and come to the ICA or whatever. One of the things I like about being here is that I've invited people with whom I work down here to events and they've participated in very interesting ways. Very often in debates involving people who aren't scientists, people can just go off on one. The discussion can go to places where there are actually facts of the matter about it.

The thing I did last month 'Games and the Body' there was some discussion as to whether your mind could change your body. Now, one of the guys I brought with me had a background in Sports Psychology and is now a neuro-scientist and he was able to say that there is data to show that if you think of lifting your finger a hundred times and then you lift your finger then you will lift it harder than if you hadn't done the mental exercise. That was an example of where the discussion was going round and round and there was no agreement and the scientist was able to introduce facts to stop that. So I think there have been lots of positive interactions that have been engendered by collaboration. Bringing scientists into the ICA. The ICA has a long history of looking at scientific issues. Historically, the ICA thought of itself as a laboratory and a place of culture in its broadest sense. That's happening in a very practical sense now.

For example, towards the end of the year we're going to do some experiments on people's reactions to catastrophic situations. Putting it on the back of September 11. We're going to show some films and some stills and make some objective measurements of people in an art gallery in the ICA looking at these images and trying to look at how people's perception of risk is changed by these images. So that's one example of how it's part of the cultural stream, but it's where a scientific question impinges on the cultural scene. It's what people should be interested in.

I think that technology is a real worry -- by technology I mean things that are possible, think about cloning. We shouldn't worry that we're going to discover things about ourselves that we didn't know before. For example, some of the sets of worries people had about evolution when it started off were 'Oh my God, we're from apes!' The questions are well defined - who are we? Where did we come from? And so on. You can ask those questions biologically or in terms of astro-physics, whatever. The worry about technology says that these are roughly political decisions, these are issues that are democratically decided or whatever. You make a political decision to fund nuclear research, for example. The bomb was invented not just by scientists. It was invented in the Manhattan Project which was incredibly well funded. You know, they put the scientists in a desert, they were given loads of money and told to get on with it. It's a political decision as to whether we should be funding nuclear research, chemical research, agricultural research -- that's the question. So even before we decide which technologies we are going to outlaw we need to see this point. It's about knowledge and power and democracy is about giving power to the people. And the people need knowledge so you need to have separate projects in which people can be informed so they can participate in the decision-making process. It's interesting how art can create a buzz and having a scientist here can also create a buzz.

3AM: Do you think that there is a crisis in the arts at the moment?

DG: It's interesting to think that some higher mathematics and nuclear physics is in the same kind of crisis as some conceptual art. But with worse PR! The question is, at its core, how can we open it to a greater number of people? Now if you ask how many people really understand conceptual art, who really know what's going on, and understand nuclear physics it's probably round about the same number of people. And half of them are lying! So, there is an interesting communication issue that goes back to what I was saying before. There are these boundaries within science that are very powerful. They're about language, they're about belonging, they're about who gets appointed to get things done.

3AM: Is there a need for the return of the old man's game, philosophy?

DG: Well, Popper, for instance, is a good philosopher of science, but no scientist I know bases their work on him. There are some philosophers, for example, who claim that you cannot find a scientific account of consciousness. It's to do with subjectivity and so on. They have an argument which says there is no story which you can tell me, which is a scientific story, that will explain consciousness for me. A scientist is interested in finding what's called the neural correlate of consciousness which is the brain circuit that does consciousness. Now to the philosopher who says there will never be such an account, should he just move on to another question and give up? No serious scientist is ever going to let a philosopher tell him what he's going to work on. The philosophers are going to be important in focusing on what we're doing and also to bring clear thinking and interesting questions, pointing out contradictions, they should be doing all that stuff. But whether you're going to bet on what they say…?

So, I'm not someone who doesn't think there will be a neural correlate of consciousness. I think it's a relatively interesting project. Perhaps not as interesting as a lot of people seem to think it is, but it's quite interesting. I do think, however, in working out what it is we've worked out, what a discovery may mean, I think the role of philosophy in explaining what our relationship to the outside world is, what thinking is like, issues of representation, is always going to be a need for them. I don't think there's going to be a scientific story at the end of which the philosophers are going to go home. Neither group is going to put the other one out of work!

If you read Don Juan by Byron, he slags off Bishop Berkley and his view of physical reality and actually raises some philosophically respectable objections to Berkley's position, but in the sense of ridicule. I think philosophers have always been willing to put things out there and been prepared to be wrong but in interesting ways. So I don't think they're going to be out of work these days soon!


Dr Daniel Glaser is Scientist in Residence this year at the ICA. He is a Senior Research Fellow at the Institute of Cognitive Neuroscience, University College London, where he makes images of the human brain in action. He is interested in how expectations change the way our brains process visual information about the world. Please take a minute to participate in his ICA research into the collective unconscious, by going to his Brainjuicer link. (It only takes 2-3 minutes). He is funded by the Wellcome Trust and the Calouste Gulbenkian Foundation.

Related links:
Daniel Glaser at University College London
Cafe Scientifique

Your Name:
Your Email:
Enter your email address above for 3 AM MAGAZINE'S Monthly Newsletter. Each time a new issue is posted, we'll let you know. (Your email address will be kept confidential!)

home | buzzwords
fiction and poetry | literature | arts | politica | music | nonfiction
| offers | contact | guidelines | advertise | webmasters
Copyright © 2005, 3 AM Magazine. All Rights Reserved.