Tuesday, 24 May 2022

On AI and Disegno

My interest in drawing as 'disegno' is key to my thoughts regarding certain underpinning theoretical aspects of art practice. 

During the Renaissance central to the idea of drawing as 'disegno' was the use of drawings as the basic building blocks out of which a finished concept would emerge. On the other hand 'Colorito' was the direct application of colour (paint), to a support, (canvas or a plastered wall). This distinction has a philosophical dimension, one that provides us with a way to look at recent artificial intelligence (AI) applications as a contemporary form of 'disegno'

An artist who saw themselves as conceptually using drawing as disegno (e.g. Michelangelo) would expect fellow artists to have had a long training in drawing from life and from sources such as classical sculpture, together with an understanding of other forms of drawing such as geometry, compositional structuring, the development of complex figure poses and the use of mathematical proportion. Alongside these practical skills most importantly the artist would have the ability to draw from the imagination; these skills supported an artist's ability to build, construct and paint whatever ideas drawing had allowed the artist to visualise. Disegno both facilitated invention, and the capacity for visualising a concept, such as a building, a painting or a sculpture. It is the imaginative and intellectual core of the process, which elevated 'disegno' and those who practiced it, onto a par with other intellectually demanding areas of human endeavour. This was why artists were able to argue that there was a distinction between a craftsman and an artist and therefore demand more money for their work. This was a division of labour that could be understood by those in power who also no longer had to work with their hands to make a living. I personally think that because this argument was one focused on appealing to those in power who no longer had to get their hands dirty, that the distinction was taken too far and that the more art moved away from its craft base, the more it became estranged from those people who understood ideas via making or using their hands to think with. The more art became an intellectual pursuit, the more it forgot its earlier 'animist' relationship with materials, an understanding centred around the fact that materials had a life of their own. The spirit that used to be sensed in things as they were made was almost lost, but perhaps not forever and recent thinking in particular as laid out in the work of Tim Ingold, has pointed to a re-animating of the ‘western’ tradition of thought. I suppose what I'm circling around is the separation of making real things from the preparing for making them. I'm interested in how drawing allows the imagination to leap forward, and to envisage a possibility but drawing doesn't do the work for you, you will still have to carve that sculpture or get that building built. However entangled up with the material possibilities that confront us, is an ability to visualise possibilities and drawing is entwined within that envisioning process. 

I spent some time this afternoon talking to someone who had for years made a living by recycling old materials and re-making it into bags, cases and other useable objects. She had always done this, ever since she was a girl, growing up in post-war Britain during a time of rationing and a scarcity of materials, when it was normal to make do and mend. Her bags are beautifully sewn and lovingly constructed. I spoke to her about an idea I had for a bag of my own and she understood immediately what I needed, my old bag was falling apart and she could help not just revitalise it but rethink it. OK you might think but isn't this craft and not art? Perhaps there is an infinitesimal line between the two, but her remade bags are also ideas and she had very clear convictions about sustainability and the need for recycling, so it could also be argued that these objects did reflect an ideas base and that they contained within them a very clear point of view. The craftsperson is thankfully still with us and is still envisioning ideas as well as making them. 

Hanna Washburn: 'Imagining the Future' 2018.

Hanna Washburn: I'll Believe in Anything, 2018

Hanna Washburn's sculptural practice stems from a personal engagement with her own old clothing, these hand crafted pieces are about recycling. Many would argue that this is clearly art and not craft because there is invention and thought put into the work. But what if the forms of sculpture she uses are just that, forms, just like the templates a bag maker uses? 

Matteo Rattini: This sculpture doesn't exist: 2021

I was reminded of these distinctions when looking at the recent work of the artist Matteo Rattini, who has trained a neural network to create images of contemporary sculptures based on Instagram's algorithm suggestions. Basically he fed into a computer hundreds of images of modern sculpture and asked it to come up with a range of ideas based on the information it had received. These ideas were realised as 3D images using a CAD package and exhibited on screen as virtual sculptures. 

These are drawings, very convincing drawings yes, but they are still drawings. Once again I would argue 'disegno' or the use of drawings as the basic building blocks out of which a finished concept can emerge, is being used. Just as in the same way that Michelangelo would show his 'finished' drawings to prospective funders, in order to raise funds to get sculptural ideas commissioned, Matteo Rattini could use these images to persuade someone to fund their realisation by 3D printing. 

Thoughts about how a drawing allows the imagination to leap forward and how it can envisage a possibility are conflicted and yet stimulated by Rattini's work because his images look as if they have been already built and then photographed. But they have not, the visualisation process still doesn't do the work for you, someone will still have to carve that sculpture or get it made in some way. However that 'get it made in some way' has changed and these could be 3D printed. The virtual models are produced in such as way that every facet is recorded as a set of coordinates and these coordinates can be used to drive a 3D printer. If it is sent to be printed, the hand skills normally associated with crafting a work of art are redundant, replaced by a machine, at the moment still serviced by a technician but a machine nevertheless. But does this matter? Isn't the idea more important? 

I'm sort of mixing up two things here, the need to validate art as an intellectual activity, one that is about ideas and their realisation and the need to recognise the importance of material thinking. I'm hoping that materials might have embedded within them something that links to very old animist ideas, ideas centred on the vitalist nature of materials. Remember that Michelangelo saw his ideas in the marble and carved until he set them free. 

But we can also see the potential of 3D images within 2D drawings, some drawings are though harder to use to visualise 3D figures than others. 

Template for a teddy bear

The Leeds Arts University has a fashion course and I was always impressed with the pattern cutting technicians and their ability to 'see' three dimensional forms in the shapes of the flat patterns they were working on. 

The centuries-old art forms of origami (folding paper) and kirigami (cutting and folding paper) offer elegant solutions to this 2D to 3D problem, involving only folding and cutting to transform flat papers into complex geometries, but the more humble art of making flat templates for soft toys is also part of this tradition.

All of these issues become conflicted however as they enter the world of data manipulation. For instance the company Lalaland uses AI to create synthetic humans for fashion eCommerce brands, this is used to increase diversity in retail. Lalaland’s platform generates human like models of different ethnicities, ages, and sizes, customised according to customers’ body shapes, such as hourglass, apple, triangle, pear, or rectangle. A highly efficient automated workflow combines on-mannequin product shots with AI-generated model imagery to produce realistic model images, created in a fraction of the time required by traditional model photography. I.e. there is no need for an actual human model any more. In my earlier blog post on surfaces and body perception I looked at the process of measuring the body and creating a flat structure that could then be converted into a digital 'human', but now things have moved on; or have they? If you read the last post, you will be very aware of the issues I have with measurement and the collection of data. 

A flat map of a body created in Maya

Measurement in the case of the Lalaland images needs to be converted from flat data to 3D realisations. The skills a soft toy designer would need in order to think about how to realise a teddy bear, have been stretched to include the possibility of mapping any 3D form and 'people' data is fed into a computer, which can now make a very convincing 'human' used to model clothes for retail outlets.  
AI research has been defined as the field of study of intelligent agents or any system that perceives its environment and takes actions that maximise its chance of achieving its goals. In Lalaland's case AI-generated model imagery is used to produce realistic images of people. The perceived environment would have been restricted to images of people of all sorts of sizes and shapes, until enough data had been fed into the system for the production of a 'world view' determined by that data. Once the world view was in place, actions taken by the AI system that maximised its goals would be ones that produced answers that gave human body shape to its perceived needs. 

Lalaland site where you can customise body shape, ethnicity, hairstyle, identity, pose etc.

The visualisation tools offered by Lalaland and other sites using AI would seem to me as being little different from those tools used in the Renaissance. Back then artists were using the new tool of perspective to visualise a building or a painting. However alongside the visualisation skills, most importantly, the artist would need the ability to draw from the imagination; as well as have an ability to build, construct or paint whatever ideas the process of drawing had allowed the artist to visualise. Disegno both facilitated invention, and the capacity for visualising a concept, such as a building, a painting or a sculpture. So does AI also allow for the development of a concept and invention? 
If by new concepts we simply mean looking at the existing world in a different way, then AI processes could be a very interesting way of developing ideas.  


Memo Akten is looking at how to visualise one of the fundamental problems we have as human beings. The problem of point of view. When we develop our world view we develop it from perceptions of the environment we inhabit. Let's say I was a lighthouse keeper, then my experience would be centred on the sea and breaking waves and rocks. Akten's video of his AI computer simulation as it processes data makes us very aware of what happens when data is coming in from only one narrow source. On the right side of the screen we have images made by a computer program that has been 'fed' with thousands of images of the sky, sea and breaking waves. This is the world that the program has 'learnt'. It expects therefore for everything to conform to the behaviour of rocks and sea and sky. It looks for waves splashing when they meet obstacles like rocks, it expects the swell and movement of water to be contrasted with the hard immovability of rocks. It expects the moving areas to reflect light, (the sea and sky) much more readily than the fixed points, (the dark rocks). Edges between moving surfaces and static ones are seen to be where waves break up and produce foam and splashes. The rhythm of sea movement is recorded over and over again, so that the software understands the nature of these rhythms and how they work. However when this 'point of view' is fed rather different information, such as in this case a hand moving a yellow cloth, the AI system can only read the new information in relation to the world view it has established. Therefore it translates the hand and cloth image into one of the sea's waves breaking over rocks. On the one hand you could say that this is a very imaginative interpretation of a hand and cloth on the other hand you might argue that this is simply a mistaken interpretation. It does however ask us big questions as to how our own interpretations of the world work. Are my own inventions or 'new' images simply results of my narrow experience of the world misreading new information? Are for instance the images I have been making recently in relation to 'interoception' simply a result of my own limited processing of experience and would I therefore see things very differently if I had a more 'scientific' understanding of the subject matter? The image below is from a series of images made recently in response to having Covid-19 and associated problems with breathlessness. 

Tuesday, 17 May 2022

The portrait as witness and control

From: Vaught's 'Practical Character Reader'

Drawing can lead us into many strange territories, one of which is 'witness description composite drawing'. In the United States artists are employed by police departments, usually in the forensic evidence department.  For instance witness description composite drawing is what the artist Harvey Pratt has spent most of the last 50 years working on. He has worked as a forensic artist for various police forces, whilst also being recognised as an accomplished master Native American Indian artist. In fact it was this dual role that first attracted me to his work.
The term 'composite art' was originally used when facial features (eyes, noses, lips, hair, facial shapes) were compiled to complete a total composite of a face. This procedure was to be used by a witness who was guided through a process of how to match these various facial features in order to form an image of a suspect. Witness description composite drawings were basically the same technique but were drawn free hand and based on an interview of a witness or victim to determine what someone might look like. These drawings are regarded as investigative aids and they are used to help narrow down suspect lists. Witness description drawing can be used not only to identify suspects but also identify tattoos, vehicles and other items that a witness might describe during an interview. 

Below are examples of witness description composite drawings made by Harvey Pratt beside photos of the subject later identified. 

Considered one of the leading forensic artists in the United States, Harvey Pratt spent over 50 years in law enforcement, completing thousands of witness description drawings and hundreds of soft tissue reconstructions. You could say that Harvey Pratt had an interesting portfolio, being a Cheyenne and Arapaho tribal member and also being recognised as an accomplished master Native American Indian artist. Both these aspects of his working life as an artist involved making portraits of people, some being one step removed both emotionally and in terms of experience and others made to celebrate his ethnic origins and feeling of belonging to a tribal culture. 

Harvey Pratt: Drawing of a Native American 

I'm sure some of you will be worried by the fact that someone like Harvey should be introduced under the umbrella of a fine art drawing blog, but one of the issues I am trying to open out is where are the boundaries between fine art, illustration and other occupations whereby similar skills are used? All of these people are called artists and the art world is prone to snobbery, so it is always worthwhile stepping back to explore how broad the church of the artist really is, and as we do so perhaps we also have to ask questions as to how does an artist contribute to society? 
A very similar profession to witness description drawing is of course courtroom drawing. In the UK photographs are not allowed to be taken whilst a trial is in progress. A courtroom sketch is defined as an 'artistic depiction' of the proceedings in a court of law and of course this means that news media have to rely on sketch artists for illustrations of the proceedings, which is probably the main way that the general public get to see hand drawn portraits of people. 

Priscilla Coleman: Naomi Campbell and Piers Morgan for ITN

In the UK, a courtroom illustrator is an incredibly niche trade, and a dying one. At the moment here are four professional courtroom sketch artists: Priscilla Coleman, Siân Frances, Julia Quenzler and Elizabeth Cook. However new laws were passed in January 2020 allowing cameras into the country’s Crown Courts, including the Old Bailey, to broadcast the sentencing remarks on high-profile criminal cases, which means that court artists are only needed for the less 'interesting' aspects of a court case. Drawing in court, or making an image of any kind, be that a photograph, doodle or otherwise, is illegal in British courts. So courtroom artists not only have to be skilled portraitists but they also have to have excellent memories for faces and places. To make their images, they take written notes during the case, before moving outside to the press room to draw their images as fast as possible, both to ensure accuracy and to enable their clients, (newspapers and broadcast media) to get images in time for various deadlines. 


Priscilla Coleman: Artist's notes from courtroom observations

I like to think about the notes taken by these artists as drawings in their own right. They are part of the process of remembrance and are used as triggers to ensure images are brought back as accurately as possible. 

The courtroom artist is a skilled compiler of facial images and they are making notes knowing they are going to construct an image of a face within a few minutes of making their notes. They understand how heads are constructed and have training and years of experience. Compare this to a crime witness's memory of a person that is recalled perhaps several days later or even more, that then has to be translated or communicated to an artist who then has to render the image. When I looked at communication theory I pointed out that Shannon and Weaver wanted to make sure that communication systems worked well. The model they came up with allows us to easily see where problems in a communication system could go wrong. 
In particular the information source needs to be working in such a way that it is decipherable by the information destination. So in this case we have someone who was probably very stressed at the time of the experience, having to recall the facial features of someone they may only have glanced at for a few seconds. Not only that, they will have had no training in the recognition and identification of facial features and no specialist language to help them either.  You should try this yourself, give a verbal description of someone to another student and ask them to make a portrait from that description. The receiver also needs to be good enough to understand or not distort the message coming through. In this case either the police artist uses their experience based on previous encounters or they use the composite model. The skills of the drawer and the ability of the drawer to put this information into an effective shape for transmission are again vital.  You very quickly begin to realise how amazing it is that any communication is made at all. But, and this is the real problematic issue, what both people in this communication system have in common is a set of preconceptions, stereotypes and ideas about what people look like and this is perhaps what they are really communicating. If we look at Vaught's Practical Character Reader's honest face, as seen at the top of this post, we see perhaps one end of a stereotype scale that the crime victim and police artist are measuring their ideas against. For instance in many experiments in relation to emotions instigated by facial type, people respond that they trust and find honesty in faces that are more symmetrical. Symmetry is at the core of a police artist's working system. Below is a verbatim account given by a working police artist of how they set out to do their work, the ruled pad is what they use as a starting point for every drawing. 

A police artist's ruled drawing pad

"I start the drawing by having the witness go through the FBI’s Facial Identification Catalog feature by feature. I always start by having the witness look through the FIC for the shapes of the suspect’s head, or face. This is very important because the shape of the head is one of the three main components to getting close to the suspect’s actual look. I stress it’s importance. I set my drawing up on my paper before every composite. This is also important because it keeps all the drawings to a certain scale and helps with symmetry".
 

"I then move on to the suspect’s eyes. I rush the witness just a little bit so they don’t make themselves crazy looking at all the different eyes. I explain the importance of choosing only the closest set of eyes. I let the witness know features can be changed, or fine-tuned, when the line drawing is completed. I make sure the witness understands they should concentrate but not get frustrated by all the choices".
 

"Once they choose a set of eyes, I sketch in the eyes and eye brows of the eyes they chose. I don’t want to nit-pick them on features. Eyebrows don’t normally make or break a likeness so, if they remember specifics I will note it, if not I skip the eyebrow section".


"I use a regular 2B pencil because it is harder than my finishing pencil, erases easier, and draws lighter. I have the witness start looking at noses while I sketch the eyes. I draw the nose while I have the witness looking for a mouth".


"I draw the mouth while the witness looks through the chins. I draw the chin and sides of the face, after referring back to the shape of the face. I always draw the top of the head with just the forehead. I always put generic ears on, unless they were a feature the witness remembered specifically".


"I get the witness to find the hairstyle somewhere in the book. Once the line drawing is completed I have the witness look it over. I make changes by erasing with the kneadable eraser and re-sketching until the drawing is as close to the suspect as the witness can remember. I have the witness rate the line drawing on a scale of one to ten, "ten" being a portrait (really close), and "one" being no where near close. Now, with the line drawing as close to the suspect as the witness can remember, I add skin tone and shading". 

"I use a 9b woodless pencil because it is soft, goes on dark, and smudges easily. I get good contrast between lights and darks. Shading takes practice. Achieving a good skin tone is important because it will make the composite believable. After, the composite has skin tone it should appear closer in the rating scale, or at least no change". 

In the example above it is interesting to see how the artist leads the crime victim through a series of decision making choices, ones that must influence the final outcome. 

What the artist is doing is very closely modelled on the 'Photo-fit' kit. 


You can of course download an app to do this on your phone

The Photo-fit' kit

The Photo-FIT kit for police forces was derived from the work of Jacques Penry. In the video clip below you can see him at work. The video is a timely reminder of how social convention was often thought of as being fact and that old beliefs in physiognomy, or the practice of assessing a person's character or personality from their outer appearance persisted for many years after they were shown to be false. 

Jacques Penry describes his system

Physiognomy as a 'science' was developed by Johann Caspar Lavater (1741-1801), his four volumes of 'Physiognomic Fragments' were popular throughout Western Europe, and they began to influence how people literally pictured themselves. In Central Europe, there was an obsession with silhouette pictures, usually created by cutting a person’s profile out of black cardboard. Lavater advocated the use of this system because it made face profiles easily measurable.
Lavater’s apparatus for making silhouettes 

The 'proof' that you were an upstanding citizen could then after your portrait had been made, be framed and put on your wall. 

Portrait by Robert Friend

Certain artists, such as Robert Friend, developed the silhouette portrait as a specialism and just as all the other elements in the images he made are indicators of class and status, the profile itself would have been slightly adjusted to ensure that the person depicted had the right sort of facial shape. 

Our continued need to control how we look in relation to how facial features are read within society means that for instance, because selfies distort the face many people are undergoing plastic surgery on their noses in order to make them smaller. Mobile phone photographs misrepresent looks, because from the distance of an arms length, most lenses will make the nose appear bigger in proportion to the rest of the face. People who are not photographers don't realise that camera lenses are not the same as normal eyesight. However the selfie becomes the 'reality' and is measured by social convention and that is itself determined by very old ideas that are still rooted in concepts that came from physiognomy. 

There is something weird about all these depictions of people. They are either abject in the sense that measured 'people' are completely without pride or dignity; or they are about lack of self worth. Julia Kristeva in 'the Powers of Horror' states that the abject refers to the human reaction to a threatened breakdown in meaning caused by the loss of the distinction between subject and object or between self and other. These images of faces seem to encompass all of these definitions and yet at the same time they reach out into a western European past that shows a continuing need to measure and determine ourselves through measurement. According to Foucault, knowledge is only possible within a vast system of power relationships that allow that knowledge to come to be. I.e. society allows these strange readings to continue and be used as if they are 'knowledge' because some powerful people find it useful. 

In 1902 Vaught's 'Practical Character Reader' was published. It's hard to believe that this is a Twentieth Century publication, as it seems almost Medieval in its reasoning. 




It is interesting to compare and contrast the total humbug of Vaught's 'Practical Character Reader' above with the images below as set out by a contemporary company advertising its skills in plastic surgery. 


These images are meant to demonstrate how rhinoplasty can improve your facial harmony

A repose frontal mask

Some experts on facial harmony use a measuring system based on the golden ratio, the one above does that. You are asked to overlay the Repose Frontal Mask (also called the RF Mask or Repose Expression Frontal View Mask) over a photograph of your own face to help you apply makeup, to aid in evaluating your face for facial surgery, or simply to see how much your face conforms to the measurements of the Golden Ratio.

It is not a far remove from stating that broad headed humans and animals are vicious or people with convex curves to their noses are deceitful, to stating that a broad nose or one with a convex curve is un-harmonious. As we will see when we come to the 
main anthropometric points for facial harmonisation, the faces that 'scientists' use to take the measurements from, fit a very narrow stereotype, and by extension when we begin to look at equity in facial recognition technologies, we find that these technologies were once again invented by people who see white faces as the norm and the further away from that norm we get the less accurate is the technology. No matter how far measurement is taken in order to develop objectivity, it would seem to be obvious that we cant escape bias and subjectivity in our reading of faces. If so it is probably much better just to accept that and spend more time exposing ourselves to other people's ideas of beauty and character, so that at least we come to realise that what we think is only a point of view and that that old cliche 'beauty is in the eye of the beholder' has more going for it than we tend to think. 

Measurement and faces is a long running concern. Durer's 'Four Books of Human Proportions' ('Vier Bücher von Menschlicher Proportionen), is another attempt to set out to give mathematical precision to the visualisation of faces. 



Durer

Durer, like the forensic police artist, was attempting to develop a system that could capture the almost indefinable essence of what it is to be an individual. 

Our current obsession with faces and the selfie, as already suggested, leads us into plastic surgery, an area where facial measurement is an obsession, such an obsession that it has led to a specific discipline, 'faciometrics'. Faciometrics consists of making direct measurements from predefined anthropometric points. In order to do this a facial ruler is used and measurements linked to standardised references for quantitative analysis. The relationship between these measurements then guides facial interpretation. A Practical Guide for Orofacial Harmonisation, states that 'with these normative parameters, we can guide harmonisation procedures and recover facial proportions, making facial features more similar to the reference.' 'In this way, we will achieve more individualised planning that will be a more assertive approach in the proposed treatment.'  See: https://www.mathewsopenaccess.com/full-text/faciometrics-a-practical-guide-for-orofacial-harmonization

Beauty Setup ruler (A) Caliper (B)



The table below sets out the main anthropometric points for facial harmonisation.

Points

Initials

Description

Trichion

Tr

A point at on the hairline in the midline of the forehead

Glabella

Most prominent midpoint between eyebrows

Endocanthion

En

Inner commissure of palpebral fissure (left and right)

Exocanthion

Ex

 

Outer commissure of palpebral fissure (left and right)

Pupil

P

The black circle in the center of the iris.

Nasion

The midline point of the nasal root and nasofrontal region.

Pronasale

Prn

Most prominent midpoint of nasal tip

Nasal Alare

Al

 

Most lateral point of alar contour (left and right)

Zyghion

Zy

 

The point of the most lateral soft tissue overlying the zygomatic arch (left and right).

Subnasale

Sn

Midpoint of columellar base at junction of upper lip

The point on the midline of greater concavity in the facial contour of the upper lip, between the subnasale point and the upper lip.

Crista

Philtri

Cph

The point at each elevated margin of the philtrum just above the vermilion line

Labiale superius

Ls

The midpoint of the vermilion line of the upper lip

Stomion

St

Midpoint of the labial fissure between gently closed lips

Cheilion

Ch

Lateral extent of labial commissure (left and right)

Labialeinferius

Li

The midpoint of the lower vermilion line

Gonion

Go

 

The most lateral point on the mandibular angle (gonial angle). Its location is close to that of the bony gonion. (left and right)

B’

The point on the midline with greater concavity on the facial contour of the lower lip, located between the lower lip and the soft chin. It is the deepest point of the mentolabial fold.

Pogonion

Pg′

The most anterior midpoint of the chin.

Mento

Me′

The lowest point on the midline of the soft tissue of the chin. This is the lowest point in the measurement of facial height.


Within a few moments we have moved from the abject to a set of concerns that are all about beauty. Durer was looking for harmonic proportion, and so is the plastic surgeon. The drawings done by the police forensic artists are an attempt to gain control of what is unknown, an attempt to find the criminal face, a project I would have thought doomed by subjectivity. But so are the standards of beauty we see set out in the supporting information for those dealing with faciometrics. These images, usually of white women, are yet another problematic concept. The unconscious bias in our actions and beliefs, is not in fact unconscious, its pretty easy to see that the models these facial features are based on are of a certain type and the type reflects the dominant characteristics of the people who control the technology. 

Drawing as both witness and control shows us that what we would normally see as activities outside of fine art drawing as a discipline, are in fact areas that highlight issues for any artist doing something as apparently straightforward as making an image of someone's face. As soon as I begin to draw I'm reminded of those hours of practice as a student, always being told to measure the situation again, always been pushed towards an idea of accuracy. Strangely the more you were able to grasp things like perspective and get what you were drawing 'right' the more you were asked to now push past superficial 'likeness' and to now look for the action of perception, the traces of head movement that you initially had tried to iron out but which now became vital to the 'life' of the looking and its capture in the drawing. 

Drawing of a face using eye tracking technology

What we are actually doing when we look at a face is several things at once. Initially checking fight or flight impulses, which will be done almost before you can begin to think about what you are doing, and then checking out a whole host of other bits of incoming information, status and class, usefulness, sexual interest, kinship, interest in me, similarity and difference, etc. etc. and as we do this we scan, trying to pick up changes in eye shape, frowning, smile lines etc etc so that we can establish a proper relationship with this encountered other, one we can build upon even further when we both open our mouths. But look at how different the static measured drawings are to the drawings measured by eye tracking technology. It is now much harder to tell what sort of person we are looking at because we are concerned with action and a series of events and not an isolated moment. 

I have recently been exploring portraiture myself, an area of art practice that is full of clichés and a perceived need for 'likenesses' and one that on the whole I have always suggested students steer well clear of, but people persist in making them, so rather than avoid the problem, I've decided to  tackle it head on. The first issue of course being 'likeness'. Likeness nearly always comes up when in conversation with someone from outside of the art business. "It looks just like a photograph" being many people's judgement call when it comes to likeness and of course the best way for an image to be "just like a photograph" is to work from one. This personally seems a redundant exercise so I never use photographs. Likeness is also measured against a normally seen static image, so I never ask my sitters to keep still, I just respond to the interaction of an exchanged conversation and as materials come into play I let the materials interact with myself and what I am experiencing. The images therefore become records of time spent with someone, rather than depictions of them. This feels much more honest and about people or encounters rather than renderings and copies. I want to avoid ways of working that tend to get people looking at themselves as if seen in a mirror, and checking for likeness. I want people to begin looking at the images as if they were new things, new figures formed from paint and ink and pens and brushes, things that talk their own language and which you have to look at very carefully if you want to listen to what they are saying. 




From an ongoing series of encounters with people

Drawing and photography or an attempt to think about why someone might work from a photograph