doi:10.1016/j.tics.2012.03.007


TICS-1071; No. of Pages 3
Doing science making art

Bevil R. Conway1,2

1
Neuroscience Program, Wellesley College, Wellesley, MA, USA

2
Department of Neurobiology, Harvard Medical School, Boston, MA, USA

Scientific Life: My Word
It is a challenge to balance doing both science and art,
but I attempt both because each provides unique insight
into perception and cognition, often to mutual benefit.
The techniques I use in visual neuroscience (psycho-
physics, fMRI, microelectrode recording) are different
from those I use in visual art (etching, watercolor and
oil paint, glass and silk), but the process is always
empirical, testing hypotheses driven by observation,
trial and error, while striving to be creative.

Science has an apprentice system and its guilds establish
the goal of publishing in peer-reviewed journals. My sci-
ence interests dovetail two issues: the computations made
by brain regions responsible for vision; and color, a model
system for understanding how complex phenomena arise
from neural networks [1]. In addition to doing science, I
make art. The rules of making art are no longer well
codified, but this, I argue, may represent a boon to scien-
tists interested in perception and cognition.

When I started making art as a child, like many novices
I made figurative, representational work. I became reason-
ably good at observational rendering because I did it
obsessively. I was struck by the way artistic processes
are often described by non-artists who think art comes
about by a mystical, almost random act. In fact, the work of
most artists is the result of deep deliberation and few
accidents. Techniques are employed because they have
predictable results: drawing with lines is useful in depict-
ing object shape; painting a background green makes a red
object stand out; portraits often capture precisely the line
formed by closed lips because it provides a good cue to the
sitter’s identity. Deploying these strategies, the ‘lie’ of art
according to Picasso [2], is what making art is all about.
Some neuroscientists have started to mine art for insight
into neural computations – line drawings interpreted in
terms of the orientation selectivity of visual neurons [3];
color contrast effects in terms of double-opponent cells [4];
and portraiture in terms of face processing [5,6].

It is a curious feature of our perceptual cognitive appa-
ratus that insight into the artists’ strategies, whether
gained through art or science, does not seem to diminish
the power of the effects. In fact, the insight often enhances
appreciation, much like knowledge of winemaking
enriches the wine-tasting experience. On the one hand,
this is not surprising: we all return to our favorite visual
illusions even though we know what we will see in them
and may even have some knowledge of how they play on the
visual apparatus. On the other hand, it is often said that
visual delight comes from the element of surprise. That we
Corresponding author: Conway, B.R. (bconway@wellesley.edu).
can derive pleasure from both the familiar and the unex-
pected suggests that any single reductionist theory of
vision is probably incomplete. A similarly complex inter-
play is often at stake in celebrated works of art, which are
usually recognizable yet afford a seemingly endless source
of new discovery, inviting extended, repeated viewing.

Besides pleasure, art serves a useful function: looking at
pictures can make apparent the computations that the
brain must be solving, which can guide the experiments
we run. Making pictures, rather than just looking at them,
might be even more instructive, since the process of mak-
ing not only requires time, providing time to think, but is
also invested with a salience brought about by activating
muscles. Motor salience is important: pulling the muscles
in your face into a smiling position is far more effective in
bringing about happiness than just thinking about smiling.
The elusive yet powerful connection between cognition and
movement is what artists call muscle memory. By the same
token, I have found that in considering the brain’s mecha-
nisms of vision I have been inspired to take new
approaches to making art. A few examples are given at
the end of this essay.

Although work at the interface of neuroscience and art
can be fruitful in uncovering the nuts and bolts of vision
and in inspiring new art, my hunch is that it will be
impossible to gain insight into aesthetics simply by asses-
sing what tricks artists use to represent the world or by
looking at patterns of brain activity generated when people
look at ‘beautiful’ or ‘ugly’ pictures, for the simple reason
that aesthetic judgments are mutable and shaped by cul-
tural context. A red spot might pop out when surrounded
by green, but whether this color combination is considered
beautiful would seem to be dependent on the cultural
baggage with which it is delivered. That aesthetic judg-
ments are contingent on context should not come as a
surprise to neuroscientists: almost every sensory phenom-
enon is characterized by a contextual relationship (as the
red-on-green example illustrates). Making unqualified
associations of brain activity and aesthetics may be not
only naı̈ve, but also detrimental to the integrity of science,
and I think it misses the point of art. Against this back-
drop, the nascent field of neuroesthetics is struggling to
establish the terms by which it can be productive. Certain-
ly, to be so it will be important for scientists, art historians,
philosophers and artists to negotiate meaningful points of
contact.

Picasso quipped that he could do a decent portrait of a
good friend without looking at him. Making art, as Picasso
showed, is a cognitive act that has an indirect relationship
with seeing. The success of Picasso’s portraits depended on
hundreds of hours of careful observation, not while making
1

mailto:bconway@wellesley.edu


Scientific Life: My Word Trends in Cognitive Sciences xxx xxxx, Vol. xxx, No. x

TICS-1071; No. of Pages 3
art but in the course of life with his friends. From these
hours, Picasso was able to extract a friend’s ‘essence’, which
he pinned down with marks that bore little resemblance to
those one might expect from a photograph, but were re-
markably more recognizable. Through his work, Picasso
advanced the Modernist spirit that placed a premium on
subjective experience. Picasso’s portraits are celebrated
even though few of us actually knew, or even saw, his
subjects. How did Picasso’s private cognitive work trans-
late into pictures that had, and continue to have, a pro-
found effect on us? The legacy of Picasso is evident in
contemporary art, almost all of which is concerned with
ideas, some entirely divorced from materials (conceptual
art). Art today is freed from the constraints of the codified
apprentice system of an earlier time. This would seem to
provide an even richer data set for neuroscientist interest-
ed in brain mechanisms of cognition, as artists pursue the
imperative of continuous innovation, constantly seeking
new modes of expression, exploring the full range of cogni-
tive space. Yet few neuroscientists have equipped them-
selves with the skills to understand contemporary art.
Scientists are not alone: many people find contemporary
art esoteric and impenetrable. Most of us gravitate instead
to work that has an objective basis, which is why galleries
in vacation hotspots show pictures that look like some-
thing, whereas major London galleries show sharks float-
ing in formaldehyde. The penchant of the uninitiated for
figurative work and the culturally literate for conceptual
work underscores the challenges facing the field of neu-
roaesthetics.

Liberated from the professional requirements of many
artists to make work that meets the public’s expectations, I
have enjoyed making a wide range of work using a wide
range of media. Figure 1a shows a watercolor inspired by
the brilliant foliage seen on a hike through the rolling hills
of New Hampshire in the fall. An art critic looking at the
work remarked that the picture was ‘full of air’, which I
attributed to the vacant regions of paper left white. These
white regions buffer against chromatic induction, so that
(a) (b)

Figure 1. (a) New Hampshire hills, watercolor on paper, 2007. (b) A Rake’s Progress IV (At

2

the color of each mark as it appears in the final picture is
similar to the color of the mark as it was made during the
painting’s development. The importance of the white un-
finished spaces to color relationships was obvious to the
widely recognized colorist Paul Cézanne. Yve-Alain Bois
recounts an exchange between the artist and his dealer,
Vollard, concerning the two small spots on the hands in a
portrait that had not yet been covered with pigment.
Cézanne said ‘maybe tomorrow I will be able to find the
exact tone to cover up those spots. Don’t you see, Monsieur
Vollard, that if I put something there by guesswork, I
might have to paint the whole canvas over starting from
that point?’ Bois goes on to emphasize that another famous
colorist, Henri Matisse, ‘was well aware that the appar-
ently ‘‘unfinished’’ quality of Cézanne’s canvasses had an
essential function in their construction’ [7]. Tackling color
is a case in which art history, art practice and neuroscience
can mutually benefit [4].

Rarely do I make art to illustrate perceptual phenomena,
but occasionally on reflection I become aware of the deploy-
ment of strategies (lies in Picasso’s sense) that prompt me to
think about the brain mechanisms underlying perception.
Figure 1b shows an etching I made by painting a sugar-ink
solution directly onto a copper plate at Tanglewood, Mas-
sachusetts. The piece explores the relationship between
marks made to represent objects and marks made to com-
municate written words – hand writing. The artistic issues
at stake here are not unlike those of the transformation of
retinal signals into position-invariant object codes and the
cognitive function of forming symbols.

Another example of my artwork tiptoes towards sculp-
ture (Figure 2). In these 3D drawings, I define a set of
conceptual rules on which I base the construction of a cube,
which I make out of glass and silk. The approach is a nod to
Sol Lewitt and minimalism, but came about from the
echoes of computational neuroscience, which has charac-
terized object identity as ‘read out’ from neuronal spikes
[8]. Does a set of instructions describing how the spikes are
read out adequately capture the representation of an
TRENDS in Cognitive Sciences 

 Tanglewood, While They Had Dinner with Ellsworth Kelly), sugar-lift aquatint, 2007.



TRENDS in Cognitive Sciences 

Figure 2. GB #5, four rows of four tufts of red silk, glass and silk, 2005–2012. Photo credits: (left) Stewart Clement; (right) Joanne Rathe (Boston Globe).

Scientific Life: My Word Trends in Cognitive Sciences xxx xxxx, Vol. xxx, No. x

TICS-1071; No. of Pages 3
object? By analogy, what is the difference in impact in how
we feel about or expect we will feel about something
pictured in the mind’s eye versus something physical, solid
and real? Four rows of four tufts of red silk arranged in a
grid and suspended by glass micropipettes. What does this
look like? Thinking about smiling versus actually smiling.
In this work, I am interested in the subtle and unavoidable
variability that arises not just in the execution of an
instruction (‘noise’ according to computational neurosci-
ence) but also in the surprises that accompany the realiza-
tion of something previously only visualized. This
variability is the fingerprint of the handmade, and the
surprise is a reminder that vision is only complete on
experience. Unlike my science, which sets out to ask and
answer questions, my art just seeks to raise questions. But
I strive to be creative in both pursuits and to enrich the
experience of seeing.

It is a joy attempting to do something original, in both
art and science. And life can be rich partly because of the
variety, a lesson I learned from my mentor David Hubel.
Finding something new in art or science, writing about it,
teaching, developing the manual creativity to make a
woodblock, etching, or electrode: it’s all great fun.

References
1 Conway, B.R. et al. (2010) Advances in color science: from retina to

behavior. J. Neurosci. 30, 14955–14963
2 ‘Picasso speaks’ [statement to Marius De Zayas] (1923) The Arts, May,

pp. 315–326 [reprinted in Barr, A. (1946) Picasso – Fifty Years of His Art,
Museum of Modern Art, pp. 270–271]

3 Cavanagh, P. (2005) The artist as neuroscientist. Nature 434, 301–307
4 Conway, B.R. (2012) Color consilience: color through the lens of art

practice, history, philosophy and neuroscience. Ann. N. Y. Acad. Sci.
DOI: 10.1111/j.1749-6632.2012.06470.x

5 Balas, B.J. and Sinha, P. (2007) Portraits and perception: configural
information in creating and recognizing face images. Spat. Vis. 21,
119–135

6 Conway, B.R. and Livingstone, M.S. (2007) Perspectives on science and
art. Curr. Opin. Neurobiol. 17, 476–482

7 Bois, Y-A. (1993) Matisse and ‘arche drawing’, In Painting as Model,
MIT Press, p. 48

8 Hung, C.P. et al. (2005) Fast readout of object identity from macaque
inferior temporal cortex. Science 310, 863–866

1364-6613/$ – see front matter

doi:10.1016/j.tics.2012.03.007 Trends in Cognitive Sciences xx (2012) 1–3
3

http://dx.doi.org/10.1111/j.1749-6632.2012.06470.x
http://dx.doi.org/10.1016/j.tics.2012.03.007