Part & Whole

Jan Koenderink

D E C LOOTCRANS P RESS
Part & Whole
Jan Koenderink

D E C LOOTCRANS P RESS , MMXII

Front cover: Paul Klee’s “Brother and Sister” are intimately intertwined, like Siamese Foreword
twins. It would be impossible to pry them apart. Notice that they share a single heart!
This is a very tight pictorial composition.
This short E–Book is one of a series of short introductions. This one is
about a rather specific topic: Part–Whole relations in the visual arts. It is
still in a preliminary stage, given time—always in short supply—I plan to add
notes and Internet references.

Leuven, September 24, 2013 — Jan Koenderink

De Clootcrans Press
Utrecht The Netherlands
jan.koenderink@telfort.nl

Copyright © 2013 by Jan Koenderink

All rights reserved. Please do not redistribute this file in any form without my
express permission. Thank you!
Jan Koenderink
University of Leuven (KU Leuven)
pax / jan koenderink Laboratory of Experimental Psychology
Tiensestraat 102 – bus 3711
3000 Leuven
Belgium
First edition, 2013 jan.koenderink@ppw.kuleuven.be
Personal page
10 9 8 7 6 5 4 3 2 1

i
 What is a “work of art”? I am mainly concerned with pictures of some sort.
PART & WHOLE So what is a picture? This is a non-trivial topic by itself, one might devote a
voluminous book on it. I leave it at a somewhat superficial attempt here. The
topic is important enough that the reader might want to devote some serious
T HE WHOLE IS MORE THAT THE SUM OF ITS PARTS is understood thought to it.
as the key insight of the “Gestalt School” of psychology. One might Even if you are not interested in art at all might find the topic of interest.
say that the part–whole relation is at the very roots of the Gestalt de- If your interests lie elsewhere, you may yet consider to invest in Art. Many
scription of visual awareness. Thus one expects it to play an equally people do, it may be safer than the Golden Standard. Maybe it promises
important role in the visual arts, which thrive on visual Gestalts. higher gains than coffee, or even rare earths. As an investor you should be
Formally, the science of parthood relations is known as mereology. keenly aware of the ontological nature of the stuff you buy.
Hence, it is only natural to attempt a mereological account of the At first blush, one might venture that a painting is a physical object, say a
structure of works of the visual arts. But, perhaps unfortunately, canvas covered with pigments. Thus Maurice Denis (1890) famously wrote1 :
such an attempt is due to fail! There are various reasons for this,
perhaps the more interesting one being that one deals with parallel It is well to remember that a picture before being a battle horse, a
pictorial worlds. This is a natural result of the ontologically layered nude woman, or some anecdote, is essentially a flat surface covered
structure of the mind-world. Another way to intuit this outcome with colours assembled in a certain order.
is to understand that works of art are in important ways similar to
biological organisms. Strange to say, but many people actually believe to buy a nude woman when
they are actually buying a smudged canvas.
So paintings are physical objects. Of course they are! However, used as a
tea tray, such an object is evidently not a picture. In order to be considered a
Introduction picture, there should exist a double-sided intentionality.
Of course, this is a problematic term. “Intentionality” is due to Franz von
Brentano2 , and implies that thoughts are invariably about something. They
My discussion will be of an elementary nature, and be rather narrowly fo- are about your dog, Euro coins, golden mountains, or round squares. This is
cussed on parthood relations in the visual arts. In this introduction I first the “single-sided” intentionality, it is about your awareness.
explain the scope of the investigation, then discuss its very feasibility. “Double-sided” involves the awarenesses of different minds, the artist, and
the observer. I use “double-sided” so as to avoid confusion with Husserl’s
A DEFINITION OF TERMS notion of “double intentionality”.3 I do not use “shared intentionality” be-
Some terms may require introduction, since I use them in a perhaps unfa- cause artist and observer do not necessarily know each other, indeed, might
miliar manner. I summarily consider
1
what is a work of art? Définition du néo–traditionnisme. Art et critique, no. 65, 23 and 30 August 1890.
2
(Brentano, F. (1874). Psychologie vom empirischen Standpunkt. Verlag von Duncker &
what is science? Humblot, Leipzig.)
what is mereology? 3
Husserl, E. (1901). Third Logical Investigation, in: Logische Untersuchingen 2nd ed.,
before embarking on the actual discussion. Max Niemayer Verlag, Tübingen).

1
live centuries or continents apart. There is no collaborative aspect in the case The Face of Jesus has been spotted in the most unlikely places. Such por-
of art appreciation. traits are evidently “painted in the mind”, although they often give rise to
Thus this is what is implied by “double-sided intentionality”, very simply: rumors of miraculous events. I show an example5 (one of the numerous) “Je-
sus’s Faces”, this one spotted on the toilet door in the Ikea shop at Glasgow.
— the picture was made (this may be relaxed), and intended to be looked at It drew crows of spectators. This “Toilet Christ” indeed shows an example of
as a picture (this is essential), by an “artist”; a touching rendering of the Face of Jesus. Yet, striking as it may be, it is not
— the picture is looked at as a picture, by an “observer”. This is essential to be counted as a work of art!
too.

“Looked at as a picture” implies not looking “at” (as one might look at a tea
tray), but looking “into” , and entering a “pictorial world”.4 This is evidently
not generic looking. It is a special mode of using the eyes, that seems to be
The Face of Jesus discovered in a
limited to Homo Sapiens. I limit myself to this mode here.
toilet door at the Ikea shop in Glas-
Thus, an old stained wall, taken by itself, is not a picture, since the artist is
gow. Isn’t this great art? Doesn’t
lacking. Designating the wall an objet trouvé cures that. Indeed, mold on an
it address an important topic? Why
old wall often makes for striking pictures, and photographs of the topic often
don’t museum directors come up
sell easily in art galleries. Not that it changes the wall physically, not at all.
with great bids, as they often do in
But it contributes the artist’s intention.
cases that the public fails to under-
stand?

This is a photograph of molds grow-

ing on an old wall. Doesn’t it look
great? At least it makes a great pic- Not only the artist’s intention is important though. The observer’s inten-
ture, you may like the picture better tion is just as essential. Consider some examples. In Mark Tansey’s6 The
than the wall itself. Innocent Eye Test (1981), a cow is looking at Paulus Potter’s7 The Young Bull
(1647). The cow remains apparently unaware of the explicit erotic overtones
of this work. One concludes that in the bovine universe the painting is just
another irrelevant object, despite its life size and lifelike color. (Please keep
No work of art comes into existence as a cosmic accident, even the sunset
is not a work of art. Nor would a strikingly beautiful rendering of the sunset, 5
http://www.telegraph.co.uk/news/newstopics/religion/
due to peculiar fungus growth on some discarded board on a garbage heap, be 6373674/Jesuss-face-spotted-on-the-toilet-door-in-Ikea-Glasgow.
a work of art. html
6
Mark Tansey (born San Jose 1949), is a contemporary American painter.
4 7
Koenderink, J. J., van Doorn, A. J., Wagemans, J. (2011). Depth. i–Perception (special Paulus Potter (Enkhuizen 1625 —Amsterdam 1654) was a Dutch painter, specialized in
issue on Art & Perception) 2, 541–564. animals in landscapes.

2
in mind that Mark Tansey’s painting is a painting, rather than a “documentary involves additional layers of intentionality. Without a link, the work of art
photograph”!) doesn’t exist anymore. This applies to (once) famous pieces, currently known
only from ancient texts. Conversely, without the double-intentionality, the
physical object is just junk. This applies to various archeological items.
Thus, a work of pictorial art is a multifarious object. It is simultaneously
a physical object, and a double-sided intentional object. All these facets are
crucial. It is not a small matter “what a picture is”, it is fundamental to my
exercise.

What is science? I am singularly concerned with the science of perception

here. Even more restrictive, I deal mainly with the visual perception of pic-
tures. My interest is with the visual awareness of human observers looking at
art. Thus, I will limit the scope of science to “experimental phenomenology”9
of the visual arts.
Experimental phenomenology necessarily relies on “visual proofs”. Think
of the Necker cube10 , Kanizsa triangle11 , . . . . Here the “scientist” assumes the
role of the “artist”. Otherwise the “stimuli” of science have a similar status as
the “works” in the arts.
Mark Tansey, The innocent eye test (1981). The monochrome palette shows What science is, is also crucial to the exercise. Notice that main-stream
that this is actual art. Notice that the curtain just dropped from Potter’s Bull science is unlikely to agree with my conclusions here. It would attempt to
painting, whereas the cow couldn’t care less! The person at left wisely holds find an answer in brain scanning.
a cleaning utensil. Most of the guys in black are lab assistants, or interested
citizens. The guy in the white coat is the scientist, he doesn’t look at the cow,
What is mereology? Mereology (from the Greek “part”: µερoζ) is the for-
he is busy taking notes. This illustrates a serious experiment!
mal description of “parthood” relations. Assuming that you distinguish parts
and wholes in some domain, mereology proposes to describe the relation of
The art work is indeed also a physical object, although it need not be
unique. A postage stamp may exist in a million copies, whereas an easel University of Michigan Press, Ann Arbor.
9
painting (“oil on linen!”) is usually unique. However, to some extent, even Albertazzi, L. (2012). Experimental Phenomenology. An Introduction, in: The Wiley-
Blackwell Handbook of Experimental Phenomenology. Visual Perception of Shape, Space
a painting can be replaced with a copy, or a photograph. Many people are
and Appearance, L. Albertazzi (Ed.). Wiley, New York US.
happy enough with JPEGs on the Internet, as it saves them the hassle to visit a 10
Necker, L.A. (1832). Observations on some remarkable optical phaenomena seen in
museum. Switzerland; and on an optical phaenomenon which occurs on viewing a figure of a crystal
The physical object acts as a “pipeline” between artist and observer. Roman or geometrical solid. London and Edinburgh Philosophical Magazine and Journal of Sci-
marble copies of original Greek bronzes8 act as such a link. Notice that this ence 1(5): 329–337.
11
Kanizsa, G. (1955). Margini quasi-percettivi in campi con stimolazione omogenea. Riv-
8
Ridgway, B.S. (1984). Roman Copies of Greek Sculpture: The Problem of the Originals, ista di Psicologia 49(1), 7–30.

3
part to whole, and vice versa, as well as the relation between different parts Consider examples. Take the painting The Three Graces by Peter Paul
of a whole. Mereology has a venerable history, dating from at least the Pre- Rubens18 (1639). Here you have evidently three major objects, rolled into a
socratics on. single one. Obviously each single one of the graces is part of the group, and
Here I am especially concerned with mereology as it applies to experimen- so forth.
tal phenomenology, of which Gestalt psychology is the best known branch.
Franz Brentano12 might be considered a historical key figure. He was
strongly influenced by Aristotle, with influences from scholastic philosophy.
Brentano’s pupil Husserl wrote perhaps the first modern treatise on mereology
(Husserl, 1901). An exact formulation came with Leśniewski (writing in Pol-
ish) in 1916.13 Perhaps only with Leonard and Goodman1415 did mereology
claim a place of central interest in contemporary ontology and metaphysics. Peter Paul Rubens, “The
There is a voluminous literature on the topic. Three Graces” (1639).
These ladies possess the
Mereology is of central importance to Gestalt notions16 , which revolve
most desirable female
about part-whole relations. Thus one might a priori assume mereology to
bodies imaginable at the
be of interest in the descriptions of works of art, which may be understood as
time. Notice the strong
Gestalt complexes17 .
composition, especially
Mereology has no particular commitment to ontology, although here I will
the way the environment is
limit the discussion to pictures, albeit in a special, limited sense (see below).
used to “frame” the triad.
To give away my conclusion prematurely, I will argue that mereology —
Rubens could be trusted
perhaps surprisingly — is of very limited use in the study of pictorial art.
with a topic like this!
Worse still, it is not possible to come up with a unique, reasonable mereology
in this case.
C ONVENTIONAL WORRIES CONCERNING THE NATURE OF “ PART ”
With these definitions out of the way, I start the quest: What is a “part”?
Although this is indeed the key question of this investigation, it might perhaps
be considered a trivial question. Next look at Saul Steinberg’s19 Santa Claus. Does one have two major
12
objects merged into a single one? It seems impossible to use an eraser and
Franz Clemens Honoratus Hermann Brentano (1838 – 1917).
13
Leśniewski, S. (1992). Collected Works, S. J. Surma, J. T. Srzednicki, D. I. Barnett,
remove the Christmas tree from Santa Claus, or vice versa. (You may want
V. F. Rickey (Eds. and Transl.), Kluwer, Norwell MA. to try in Photoshop, it is an easy enough exercise.) Do Santa Clause and the
14
Leonard, H. S., Goodman, N. (1940). The Calculus of Individuals and its uses. Journal Christmas tree have independent existences at all? I would say no to that.
of Symbolic Logic 5(2), 45–55.
15 18
Goodman, N. (1958). On Relations that Generate. Philosophical Studies 9, 65–66. Sir Peter Paul Rubens (1577–1640), was a Flemish Baroque painter, and a proponent of
16
Metzger, W. (1975). Gesetze des Sehens, Verlag Waldemar Kramer, Frankfurt a.M. an extravagant Baroque style.
17 19
Koenderink, J. J., (2011). Gestalts and pictorial worlds. Gestalt Theory 33(3/4), 289– Saul Steinberg (1914 –1999) was a Romanian-born American cartoonist and illustrator,
324. best known for his work for The New Yorker.

4
Moreover, the drawing would be “lost” in case you could pry the tree and 2. the style is part of La Gioconda21 ;
Santa apart! In order to work as it does, both are needed. 3. the sky is part of El Greco’s View of Toledo22 ;
4. the craquelure is part of the painting;
5. the marble is part of the statue;
6. calcium is part of the statue (as marble is Ca CO3 );
7. cadmium yellow is part of the green patch;
8. the Eve–wing is part of the Ghent altarpiece23 ;
9. the sadness is part of the portrait;
Saul Steinberg’s Santa Claus is an
10. the title is part of the drawing;
amazing drawing! Steinberg could
11. the smell of varnish is part of the painting;
hardly be beaten at his game. His
12. the polishing is part of the (marble) bust.
drawings are amazing in their com-
. . . . and so forth! There’s no end to the complications.
bined simplicity and complexity.
Moreover, they’re invariably funny. This should make clear that the meaning of “parthood” is not immediately
It is the kind of humor that is be- evident, but more of a headache. It evidently calls for some careful thought.
yond verbal explanation, it is draw- Here I am mainly concerned with “part” in the meaning of “component”, or
ing pur sang. even “proper component”.
Part or component of what? At this point it becomes clear that ontology
cannot be ignored. According to the definition of “picture”, the mereology of
choice is going to vary.
T HE STRATIFIED STRUCTURE OF PICTURES
Pictures are neither real, nor ideal. They simply fail to fit the conventional
Here we have a problem. But, of course, this might just as well be said for metaphysical categories. I use the double-sided intentional structure to define
Rubens’ Three Graces. They exist only as a group. It would make no sense an alternative category. Many pictures came into being because of certain
to delete two of the Three Graces, you’d get a completely different painting. ideal meanings attached to them, not unlike poems. However, art appreciation
On further reflection you find that ordinary language has a variety of mean- is largely concerned with a more “anatomical” analysis. This goes to say that
ings in which the word “part” plays some role. Numerous examples are listed
21
in the philosophical literature. Standard (some famous) examples are a bikini, The Mona Lisa (La Gioconda or La Joconde, or Portrait of Lisa Gherardini (wife of
or a pair of shoes. Francesco del Giocondo)) is a half-length portrait of a woman by the Italian artist Leonardo
da Vinci (1452–1519).
Consider a dozen random examples: 22
El Greco, born Doménikos Theotokópoulos, (1541–1614) was a painter, sculptor and
1. the frame is part of La Grande Jatte20 ; architect of the Spanish Renaissance.
23
The Ghent Altarpiece or Adoration of the Mystic Lamb (Het Lam Gods, completed 1432)
20
A Sunday Afternoon on the Island of La Grande Jatte — 1884 (Un dimanche après-midi is a huge, complex Early Flemish polyptych panel painting. It was begun by Hubert van Eyck,
à l’Île de la Grande Jatte) is one of Georges Seurat’s (1859–1891) most famous works, and who died in 1426 while work was underway, and it was completed by his younger brother
is an example of pointillism. Jan van Eyck (born before c. 1395 – died before c. 9 july 1441).

5
pictures are composed of mutually heterogeneous strata. Here I identify four sometimes involves very specific historical facts. In the later case it
important ontic strata: merges into the domain of reflective thought, even the title of the work
might become relevant. There tends to be voluntary control.
1. Smallest relevant constituents. These are the strokes of a drawing, the
brush strokes of a painting, and so forth, as they are visually evident None of these strata is necessarily present in any given instance. There are
(in some paintings the touches are blended, in charcoal drawings the numerous instances in which each of these strata is relevant though.
“stump” may have been (over-)used). These are essential infima, the
structure of the paper, or canvas often being noticeable, but seen as part
of the physical object, rather than the double-intentional picture. For
photographs one might think of the film grain, or pixel structure, but
these are sometimes below visual acuity. Moreover, they are not essen-
tial. The maker often intentionally chooses a physical texture (rough
paper, film grain) such that it becomes part of the work. Such textures
are like an objet trouveé.

2. Simple meaningful units. Here one thinks of mutually dependent pairs

of strokes (common in “ovoid drawing”), a set of touches making up an
edge, and so forth. “Meaningful” involves a spontaneously felt relation
in immediate awareness. In some cases a single stroke (or gesture) may
also be a meaningful unit (like the indication of a breast in a Picasso24
drawing), in other cases (like hatching in a Dürer engraving) the simplest Three drawings by Salvador Dali. Notice how different these are! At some
units may contain many strokes. levels these are all drawings of humans. On a higher level they are all in-
volved in different events and represent different meanings. At a lower level
3. Salient Gestalts. Several (or many) simple meaningful units cohere in they are constructed from very different pen strokes.
non-trivial Gestalts, not necessarily standing for nameable parts, but
seen as significant geometrical configurations or even volumetric enti-
Consider three drawings by Salvador Dali25 , all representing human fig-
ties. These Gestalts may fluctuate on prolonged observation, as micro-
ures. These drawings have significant structure on all ontic strata. The reader
genesis organizes the presentations. In such viewing there appears to be
may find it rewarding to “dissect the essential anatomy” on each level. In
only a limited degree of voluntary control.
what sense(s) are these drawings “different”?
4. Represented entities. Objects, events, states of affairs, in some cases Indeed, the profile of “weights” that might be attached to the strata is a
plots or stories. Often of a generic, mainly affective nature, this level powerful indicator of “style”. The reader may check this by estimating such
profiles for the pictures used in this paper (e.g., compare Malevich26 , Pol-
24
Pablo Diego José Francisco de Paula Juan Nepomuceno Marı́a de los Remedios Cipriano
25
de la Santı́sima Trinidad Ruiz y Picasso (really!), commonly known as just Pablo Picasso Salvador Domingo Felipe Jacinto Dalı́ i Domènech, 1st Marqués de Dal de Pubol (1904
(1881 – 1973), was a Spanish painter, sculptor, printmaker, ceramicist, and stage designer – 1989), better known as Salvador Dalı́, was a prominent Spanish surrealist painter.
26
who spent most of his adult life in France. Kazimir Severinovich Malevich (1879 – 1935) was a Russian painter and art theoretician.

6
lock27 , Vallotton28 , van Gogh29 , Maillol30 , and Rosso Fiorentino31 ). I’ll show One usually writes the axioms in terms of a conventional notation system.
examples of their work below, it’s up to you to judge. Thus, with P standing for the parthood relation one would write
There are different aesthetic values, either positive or negative, attached to — Reflexivity Pxx
the strata. But what is more important, is that the strata are never seen in — Transitivity (Pxy ∧ Pyz) → Pxz
isolation, except for special cases, like art restoration work — but then the — Antisymmetry (Pxy ∧ Pyx) → x = y.
work is not a “picture” in the sense used by me.
Pictures are organic wholes, implying that the strata are interdependent. and so forth.
A superstratum contributes “context”, a substratum “material qualities”. A Such formulae decorate countless philosophical papers to best advantage.
Rubens painting is comparable to a polyphonic harmony. There is room for This avoids unfortunate misunderstandings, and allows algorithmic imple-
harmony and disharmony. This is crucial in aesthetic appreciation, which mentation as “symbol pushing”. I’ll skip such formalities here. Instead, I’ll
tends to be very noticeable in reading art critiques. consider possible objections, and other worries.
Although “higher” strata are (in some material) sense “made up” from el- Take REFLEXIVITY: what does it mean for an object to be part of itself?
ements of “lower” strata, it is not that the higher strata are in any way “ex- The intention of the axiom is to say that Michelangelo’s32 David33 (consid-
plained” by the lower strata. Each level has something “new” and unexpected, ered as part of the David) is just the David.
a fulguatio (in medieval terms), due to a spark of divine madness on the part But consider the famous Droste advertisement by Johannes Misset (1862 –
of the artist. That is why art is so interesting at all. 1922, a Duch painter). Droste is a cocoa and choclat factory based at Haarlem,
The Netherlands, started as a bakery in 1863, it became well known world-
Ground mereology wide during the 1920’s and ’30’s. It still exists. The advertisement became
famous for its design, and was used for many years.
The nurse was drawn by Misset after a pastel painting by Liotard34
There are many mereologies, just as there are many geometries, or logics. (1744/5) around 1900. The recursive design dates from 1904. Originally
Most mereologies have a common core though, that of a partial order. A the arm band carried a red cross (removed in 1914).
partial order is usually defined through three axioms, namely: Notice that the drawing repeats itself in the drawing, potentially infinitely
— Reflexivity Everything is part of itself ; many times! This is the reason for its fame. I vividly remember how I was
— Transitivity Any part of a part of an object is part of that object; fascinating by it when still a child.
— Antisymmetry Two distinct objects are not part of each other.
27
Paul Jackson Pollock (1912 –1956), known as Jackson Pollock (or informally as “Jack
the Dripper”), was an influential American painter and a major figure in the abstract expres-
32
sionist movement. Michelangelo di Lodovico Buonarroti Simoni (1475 — 1564), commonly known as
28 Michelangelo, was an Italian Renaissance sculptor, painter, architect, poet, and engineer.
Félix Edouard Vallotton (1865 – 1925) was a Swiss painter and printmaker associated
33
with Les Nabis. The David is a Renaissance sculpture created between 1501 and 1504, by Michelangelo.
29 The giant marble statue (nicknamed Il biancone by the natives) of a standing male nude is
Vincent Willem van Gogh (1853 – 1890) was a Dutch post-Impressionist painter.
30 5.17m high.
Aristide Maillol (1861–1944) was a French Catalan sculptor, painter, and printmaker.
31 34
Giovanni Battista di Jacopo (1494 – 1540), known as Rosso Fiorentino, or Il Rosso, was Jean-Étienne Liotard (1702 – 1789) was a Swiss-French painter, art connoisseur and
an Italian Mannerist painter in the Florentine school. dealer.

7
 congruent to infinitely other (larger or smaller) arcs of the same spiral. The
spiral (just a line — I use “line” in the sense of “curve”, or “trace”, not im-
plying rectilinearity) has no proper size. The spira mirabilis never ceases
This is the famous Droste advertise- to raise wonder. Jacob Bernoulli had it put on his grave stone in the Basle
ment design by Johannes Misset. It Münster. The motto Eadem mutata resurgo means “Changed and yet the
was printed on the front-face of the same, I rise again”. (Unfortunately, the spiral executed by the stonemasons
cocoa package, just like the one on is an Archimedean spiral. Bernoulli might turn in his grave.) An arc of a true
the tray, at left. But, of course, the logarithmic spiral is actually self-similar. It is congruent to arbibrary scalings
package in the picture again has a of itself. Perhaps that is why the spiral so frequently occurs in early (e.g.,
picture of the package on it. In prin- Celtic) ornamentation.
ciple one has an infinite regress, al- The spira mirabilis is part of itself !
though the technique of the printer
obviously can’t cope with that.
This is part of the grave
monument of Jacob
Bernoulli. It should have
been the spira mirabilis, but
unfortunately, the cratfsman
used an Archimedian spiral,
which lacks the remarkable
properties that so excited
Bernoulli.
Jean-Étienne Liotard’s pastel
painting of a maid serving co-
coa. Liotard was famous for
such renderings. The topic was
no doubt a reason for Misset
to “borrow” the design for the
Droste advertisement. Here is the true spira mirabilis. You may want to try
a magnifier on it.

This is an example on a high level. However, examples can just as well

be found at the lowest level. For instance, an arc of a logarithmic spiral is But there are worse problems. Consider the infamous duck-rabbit, usually

8
ascribed to Joseph Jastow (1863 – 1944) an American psychologist. Jastrow’s The duck-rabbit example is important in my arguments for two reasons. It
cartoon was based on one originally published in Harper’s Weekly (Nov. 19, shows that “pictorial worlds” are parallel worlds37 , in the sense that only one
1892, p.1114) which, in turn, was based on an earlier illustration in Fliegende instance is in immediate visual awareness, although this may vary from one
Blätter, a German humor magazine (Oct. 23, 1892, p.147).35 The “duck– presentation to another38 . On the ontic level where duck and rabbit live, they
rabbit” is in fact never seen as a “duck–rabbit”, but either as a duck, or a never meet. It also shows that awareness fluctuates between parallel worlds.
rabbit. The duck may be part of the duck, and the rabbit part of the rabbit, but Here the temporality is less important than the multifariousness39 .
is the duck part of the rabbit, or vice versa? Observers have either the duck or
the rabbit in mind at any one moment. The duck may give way to the rabbit,
or vice versa, in the blink of an eye.

The infamous duck-rabbit of Jastrow. Try to

“see” the duck and the rabbit. You can easily
see each separately. Now try to see both simul-
taneously (the “duck-rabbit”). So far I have met
nobody who managed to achieve that.

Do you “see” the red square? You may want to as-

sure yourself that is doesn’t exist as a physical ob-
ject. Is it in front or behind the black disks (who
don’t exist either). According to that the white at the
center will be white or reddish. The “two interpreta-
Felix Vallotton’s L’Irreparable.
tions” behave much like the duck-rabbit.

In Felix Vallotton’s L’Irreparable the sad atmosphere pervades the whole

Pictorial objects are not self-identical in an obvious way. Their identity 37
Koenderink, J. J., (2001). Multiple visual worlds. Perception 30, 1–7.
38
does not necessarily satisfy Leibniz’s Law36 . A “presentation” is the content of momentary visual awareness. Presentations are time-
less, so I mention instances of presentations, but will not imply any temporal order.
35 39
Jastow’s publication is: Jastrow, J., The Mind’s Eye, Popular Sci. Monthly 54, 299–312, I say this because it is a moot point to what extent presentations may be held to be
1899. temporally ordered. Each presence is complete, and timeless in that it contains its own past
36
The so called Identity of Indiscernables. and future. One presence is fully independent of any other one.

9
picture. The title appears not just part of, but one inseparable part with the apart, the functional meaning of the part for the whole dwindles. Thus, a
picture (“irreparable”). How to handle such cases? fingernail is part of a finger, the finger of a hand, the hand of an arm, the arm
Consider TRANSITIVITY: is a part of a part of an object a part of that ob- of a torso, the torso of a figure. But is a fingernail part of the figure? Typically,
ject? Judging from the present state of the Venus de Milo, or the Belvedere it will not be painted.
Torso. The Aphrodite of Milos, or the Venus de Milo, is an ancient Greek Of course, there are exceptions. The fingers of a mannerist figure tend to
statue. It was created sometime between 130 and 100BCE. It is believed to be important for the finger posture they modulate the posture as a whole. This
depict Aphrodite, the Greek goddess of love and beauty (“Venus” to the Ro- is evident in Giambologna’s sculpture (1529 – 1608, born as Jean Boulogne,
mans). The marble sculpture is slightly larger than life size. Its arms and but incorrectly known as Giovanni da Bologna. Giambologna was a sculptor,
original plinth were lost following the discovery. The Belvedere Torso is a known for his marble and bronze statuary in a late Renaissance or Mannerist
fragment of a nude male statue. It is signed by an Athenian sculptor “Apol- style).
lonios son of Nestor”. The statue is documented in the collection of Cardinal In contradistinction, in Maillol’s (Aristide Joseph Bonaventure Maillol
Prospero Colonna in Monte Cavallo, Rome from the 1430’s. As these an- (1861 – 1944) was a French Catalan sculptor, painter, and printmaker) draw-
cient sculptures show, small parts are not necessarily “functional”, they can ing of Juno the hands have no further parts. Neither are they being missed.
be missed.

Aristide Maillol Juno. This

is a typical “sculptor’s
drawing”. Notice that Mail-
Giovanni da Bologna Ratto lol doesn’t lose any time
delle Sabine (1579–1583), over the fingers. Different
Loggia dei Lanzi, Florence. from Giambologna, it is of
Notice the posture of no relevance to him.
extremities, hand, find-
ers, feet, toes. They add
significantly to the total
impression. Scale-related issues often come up in experimental phenomenology. They
often have to do with bottlenecks of structural complexity. A well known
example is found in Minsky and Papert’s Perceptrons (Minsky, M, Papert,
R. (1969). Perceptrons, MIT Press, Cambridge MA.). How many blobs are
there? This is an easy task for the top row (left one, right two). But how
about the bottom row? Most observers don’t succeed in parsing Minsky and
It is somewhat of a scale matter. When the whole and the part are wide Papert’s spirals. Has the double spiral “two parts”? Not in visual awareness.

10
 been held that the marble and the statue stand in exactly such a relation. Of
course, any sculptor would disagree. However, mereological loops seem not
a priori ruled out. Saul Steinberg’s Santa Claus is certainly as much part of
the Christmas tree as the Christmas tree is part of Santa Claus. Santa Claus
and the Christmas tree are distinct objects, yet they are evidently part of each
other. This is different from the duck-rabbit in that Santa Claus and the Christ-
mas tree exist simultaneously. It is a structure like the centaur, or the mermaid.

Here is a wonderful draw-

ing by Vincent van Gogh.
Notice how he handles the
textures! The point I want
to make here is the exis-
tence of a (not explicitly
indicated) vanishing point.
It serves (among more) to
“point out” the house.
This is about blobs, especially their numerosity. You will find it trivial (I
hope) to see that there is one blob at top left, and two blobs at top right. But
how about the bottom row? So far I have not met anybody who manages to
see the number of blobs at a glance. Here we run into a basic bottleneck of
the human mind. Finally, there is an obvious problem with “parts” that are important in vi-
sual awareness, but which are only implicit in the art work. An example is a
vanishing point that plays an important role in a composition, but is not explic-
In the pathologies of visual awareness this shows up as a topoagnosia, a
itly indicated. This is evident in de drawing Cottage garden (1888) by Vincent
soul-blindness to topological structure40 .
van Gogh. Other examples include the centers of circular arrangements, or the
Consider ANTISYMMETRY: can two objects be part of each other? It has
patterns of gaze directions between the protagonists of a depicted scene, for
40
This is occasionally referred to as “tarachopia”, meaning “scrambled vision”. instance Jacopo Pontormo’s Entombment (1525–1528) or Rosso Fiorentino’s

11
Moses defending the daughters of Jethro, 1523. Notice pattern of lines of
regard.

Il rosso Fiorentino got the idea

for this composition from a sar-
cophagus at the Pisa Campo
Santo. I use it as an exam-
Jacopo Pontormo painted ple of “virtual connections” be-
this fresco in a small chapel tween parts, here the various fig-
in Florence. It is remarkable ures. The connections are indi-
in many respects, e.g., no- cated in a variety of ways, the
tice the color scheme which directions of regard of the fig-
is evidently “off”. I use it ures being one.
here as an example of a vir-
tual (not explicitly painted)
circle. Perhaps perversely,
at its center is nothing, ex-
cept for a cloth.
“Amodal completions” are everybit as common in the arts as they are in the
stimuli of experimental phenomenology. A famous example are the “vanish-
ing girls” of Phillips. Clarence Coles Phillips (1880 – 1927) was an American
artist and illustrator. He is known for his stylish images of women, created
for advertisements and covers of popular magazines. The outlines of the van-
ishing girls are often largely absent. Yet you “see” the outline in sharp detail!
The woodcuts of Felix Valloton are another famous example. His L’Argent
“Amodal completions” are everybit as common in the arts as they are in the is a good example. Notice the righthand edge of the man (the lefthand edge is
stimuli of experimental phenomenology. A famous example are the “vanish- just as interesting).
ing girls” of Phillips. Clarence Coles Phillips (1880 – 1927) was an American Contemporary artists do much the same thing, although they are much less
artist and illustrator. He is known for his stylish images of women, created tied to conventions. Chico Hayasaki is a Tokyo based, active illustrator. I
for advertisements and covers of popular magazines. The outlines of the van- show a fashion design: notice the contour. Where is the body? The ambiguity
ishing girls are often largely absent. Yet you “see” the outline in sharp detail! greatly increases the dynamism of the picture.

12
One of Clarence Coles
Phillips amazing “vanishing
girls”. Notice the blue
dress against the blue
background! The dess
looks sharply delineated,
although its outline is only Chico Hayasaki drew this amazing figure.
explicit at the girl’s bosom, The “contour” fails to fit the interior! Yet
and its hem. Otherwise the nothing looks wrong here. The result is a
outline is apparent through very lively, dynamic figure. It is as if a time-
the threads of the spider’s slice of life has been encapsulated here.
web, which run at right
angles to it! No “edge
detector” of the familiar
type would ever “get” it.

Felix Vallotton’s L’Argent.

Notice that the back of the
man’s suit has not been
Peter Jeroense evidently enjoys the “lost
drawn, yet looks “present”.
contour”, much like Phillips. However,
Of course, the back of the
Jeroense combines this with appreciable
woman’s dress was never
two-dimensional simplification of shape.
drawn either, but this is
The result is a strong design that looks a bit
“explained” through the oc-
forced, which may well be one of Jeroense’s
clusion by the man’s left
objectives. Forced and sophisticated are
side. Both nonexistent out-
closely related.
lines are “amodally com-
pleted”, though in different
ways.

13
 Peter Jeroense is an active Dutch illustrator. In his fashion design one easily the object might make visual sense, for instance, you occasionally see it at
parses the big black triangle. The hidden details attribute a pleasant challenge the bottom of “slide shows” running on your computer screen. In this case
to your vision. many of the parts make some visual sense, although the Gestalt measure of
All these are the result of “object creation” in microgenesis. “Prägnanz”41 various largely over the set. The intended objects in this “vi-
Thus, there are some grave problems right from the beginning. Many, sual proof” of the Pythagorean theorem involve one square, two diamonds,
though not all, of these problems are due to the mutual dependence of the and a triangle (see figure legend). These are the “natural parts” in this case,
various ontic strata of the pictures. These are the interesting problems from defined by the artist’s (in this case scientist’s) intention. Notice that an ob-
the perspective of artistic appreciation. server need not be aware of this.
What exactly the “problems” are, depends on your current perspective.
When considering the planar pattern of pigments, the difficulties are not in-
A visual (partial) proof of the Pythagerean
surmountable. This is the image processing, or restorer’s perspective. But
theorem: the square erected on the hy-
this is hardly of much interest from a psychological point of view. When you
pothenuse is made up of four rectangular
consider visual awareness the problems soon exceed your capacity to handle
tiles, each of the squares erected on the
them.
sides of two. Thus one needs only to notice
Awareness changes at a dozen beats a second. The mereological structure
that 2 + 2 = 4!
of a picture changes from presentation to presentation, even for a single ob-
server. Different observers have different awareness, the artist had yet another.
And so forth.
Here is a subset of the tiles of the previous
figure. It makes for a nice symmetric figure
Split and Merge that may appear as a single Gestalt. Notice
that it is a “bikini-like” object in that is is
composed of disjunct pieces.
Mereology is is all about “parts”.
I show a hexagonal lozenge “made up” of six congruent triangles. The set
of all combinations of triangles from the figure has 6! (over seven hundred) In the case of works of visual art, some parts are often much more “natural”
members. Consider the lower yellow triangle. The center square is erected on than others. For instance, sawing up the Mona Lisa for a jigsaw-puzzle yields
its hypotenuse. The red and orange diamonds have been erected on its sides. (intentionally) “artificial” parts. One desires to have the mereology respect
Splitting the diamonds on their vertical diagonals, one notices that the figure the visual “naturalness”. Consider some examples.
contains six congruent triangles. Thus, the area of the center square equals The Mona Lisa jigsaw puzzle has eight pieces. How “natural” are these
the total area of the diamonds. A visual proof of the Pythagorean theorem parts? What if the jigsaw puzzle (not the Mona Lisa) is the artwork? Then the
requires “mental parthood juggling”, mostly in visual awareness, rather than pieces are “natural”, at least to the artist.
reflective thought.
Another example happens to be a disconnected (“bikini-like”) object. It 41
Wertheimer, M. (1923/1938). Untersuchungen zur Lehre von der Gestalt II. Psycholo-
is an object that is the merge of two disjunct triangles. In some contexts gische Forschung 4, 301–350.

14
 Sigmund Freud splits a vulture-shape from Leonardo’s group The Virgin
and Child with St. Anne, a painting by Leonardo da Vinci (The Virgin and
Child with St. Anne, 1508)42 . Notice the “vulture”, is it a “natural part”? Few
people would consider that visually obvious. Freud wrote a monograph to de-
fend this split (Freud, 1910). As might be expected, it depends on Leonardo’s
childhood. We conclude that “naturalness” may well vary over observers.
Edgar Degas43 painted a landscape (1892) that can be readily seen as a
reclining female nude. Notice the supine female figure, buried in the hills.
A jigsaw puzzle based on the famous Mona Perhaps it represents “Mother Earth”. The way you might parse the painting
Lisa. evidently depends upon your vision.

Notice the reclining female

nude in Edgar Degas’ land-
scape. Such anthropomor-
phic landscapes in no way
are in the visual arts.

Sigmund Freud’s phantasy on

Leonardo’s The Virgin and Child
with St. Anne is really fancy. Few
reader will buy it. I don’t. But is is—in The “hills and dales”44 are different from the “natural body members”45 .
itself—a valid exercise in mereology, The landscape and the female figure inhabit parallel universes. We conclude
not less reasonable than the jigsaw
42
puzzle. Freud, G. (1910).Leonardo–Eine Kindheitserinnerung des Leonardo da Vinci,In the se-
ries: Schriften zur angewandten Seelenkunde, Heft VII, Leipzig, Vienna.
43
Edgar Degas (1834 – 1917) was a French artist famous for his work in painting, sculp-
ture, printmaking and drawing.
44
Maxwell, J.C. (1870). On Hills and Dales. The Philosophical Magazine 40(269), 421–
427.
45
Hogarth, B. (1958). Dynamic Anatomy, Watson-Guptill, New York.

15
that even for the single observer “naturalness” may vary greatly according to object. The “eye measure” merge is much cleaner (a warm colored almond
the pictorial world the observer happens to be in. It will fluctuate. shape) than the straightforward Photoshop merge (almost an anatomical car-
There are two generic methods by which microgenesis generates additional diac shape) shown at right.
parts. They are (roughly) characterized as split and merge. Splits and merges
occur all the time in experimental phenomenology, the Pythagorean theorem
figure illustrates this well.

Splitting an object intuitively yields two additional objects, namely the ob-
ject that was “split off”, and a “remainder”. Of course, there are numerous
problems with such an idea. For instance, can you be certain that a “remain-
der” exists? The “ground” in a figure-ground split is not an object. Is there an
end to splitting? Are there rockbottom “atomic” objects?
The intuition is that splitting an object only makes sense if there is a re-
mainder. But then, the figure-ground split doesn’t split an object, because the
remainder (the background) is not a proper object of visual awareness. This
is captured by defining proper parthood of an object with respect to a whole
such that the whole should at least contain another part, not overlapping with
the object to split off. Thus any “proper part” is “supplemented” by another,
disjoint part. That other part then is the “remainder”. Most people would
agree to add this to the axioms of ground mereology, promoting it to minimal
mereology. For the aforementioned reasons, this “common sense” mereology At left Perino del Vaga (ca 1535) Madonna col Bambino. At right I have
appears ill fit to handle interesting cases from the visual arts. used Photoshop’s “Posterize” tool to define a “natural” segmentation. Here
“natural” means “based upon the raw pixel values”, you could not arrive at
Merging two objects yields an additional, novel object. Of course, there segmentations like the Freud vulture this way!
are numerous problems with such an idea. For instance, consider the object
obtained when you “merge” the two outermost triangles of the Pythagorean
theorem. Does it make (visual) sense? How many objects will you arrive at In the painting by Salvador Dali’s Apparition of face and fruit dish on
when you continue the process indiscriminately? Is there an end to merging, a beach (1938) numerous odd, and sometimes surprising merges appear to
a “universe” that contains everything? “make (visual) sense”. After cognition kicks in they further merge into name-
In the visual arts that would be the “gist”, but is it unique? In the sim- able things. Prolonged viewing will suggest various different part–whole rela-
plest case one might do a “blur” and “posterize” to force a “sum”, but the tions. They do not necessarily start from nameable parts, often “recognition”
“eye measure sum” is different. I show an example based on a painting by occurs after the Gestalt formation. The splits and merges largely occur on a
Piero di Giovanni Bonaccorsi (“Perino del Vaga”; 1501–1547), Madonna col level just above the rockbottom marks. Yet they lead to transitions from one
Bambino, ca 1535. The virgin and child visually merge into a single, unified parallel world to another at the highest level. “Wholes” and “parts” interact.

16
 megapixel, and a hundred-megapixel blue sky photograph look exactly the
same.

Malevich’s famous Black Square. It repre-

sented a breakthrough in the visual arts.

Salvador Dali (1938) Apparition of face and fruit dish on a beach. Keep
looking for a while. The semantic transitions you are likely to experience
are remarkable. Since it was painted by a man, the contents are only finite. Magritte’s Le Seize Septembre (1956). It
Yet, you may feel uncertain to have exhausted the contents, even after a “the blue sky” a single natural part? How
considerable period of scrutiny. is that in your holiday photographs? Is the
pixel-count of your camera of relevance?

Atomicity An atom is an object lacking proper parts. Consider Kasimir

Malevich’s Black Square (1915). Is there a single part? Consider René
Magritte’s Le Seize Septembre (1956). René Franois Ghislain Magritte (1898
– 1967) was a Belgian surrealist artist. How many parts has the blue sky?
Malevich’s Black Square may be considered a single atom, as might the
expanse blue sky in Magritte’s Le Seize Septembre. The blue sky is a whole In painting, the individual touches might be considered atomic, at least if
containing a single part, namely itself, the trivial one-element model. On the they were intended to be visually apparent. An alternative would be to say that
other hand, in a digital photograph of the blue sky one usually treats the pixels there are no atoms, but a blue density, “density” implying that any part has a
as atomic. From the point of visual awareness this is nonsensical, since a one- proper part. This implies a continuum of parts. The structure defined by the

17
regular open sets of the Euclidean plane with set inclusion as parthood relation Goodman’s (1958) sense means: things built up from exactly the same atoms
is the model. One has a choice here, the blue sky as an atom fits the description are identical. As a result identity is trivially defined, and so forth. This has ob-
of distribution of pigments accurately, but the density interpretation appears vious advantages. However, it appears useless from the perspective of visual
to capture the affective tone of a “deep” blue sky. awareness.
In many paintings one meets with cases where there is some atomicity,
though not throughout the work. Roy Lichtenstein (1923–1997) was a promi-
nent American pop artist. His Girl in mirror dates from 1964. The areas
painted with fake Ben-Day dots evidently lend themselves to an atomistic
mereological treatment. However, notice that Liechtenstein also paints flat
areas. One acknowledges the existence of atoms, but does not require every-
thing to have an atomic basis.

Mereology in art and (vision) science

From the above, it should be clear that mereology without ontology is a

nonentity. With regard to pictures one needs to take the distinct ontic strata
into account.
An even more primitive distinction is between the patterns of pigments on
the picture plane, and the pictorial content in visual awareness. Vera Gräfin
von Lehndorff-Steinort or Veruschka von Lehndorff (born 1939) is a Ger-
man model, actress, and artist who was popular during the 1960’s. She is
known professionally as Veruschka. Many will remember her from her part
in Michelangelo Antonioni’s movie “Blow up”. Consider the book cover from
1986.46 In terms of the former ontology (which we perhaps might designate
the “image processing” ontology), there are two major parts, a blue area on
top, and a whitish area below. In terms of pictorial content (which we might
Roy Lichtenstein Girl in mirror (1964). Notice the painted Ben-Day dots. designate the ontology of “awareness”) there are several distinct presenta-
Are these “natural parts”? “Atoms” perhaps? tions, one of which is composed of three major parts, namely a layered order
of the (pictorial) body of the Countess von Lehndorff in front of a (pictorial)
stone wall, in front of the (pictorial) blue sky.
The pixel-array model (including the pixel-value discretization, e.g., as
numbers in the range of zero to two-fifty-five) leads to the trivial mereology 46
Lehndorff , V., Trülzsch, H. (1986). ‘Veruschka’. Trans-figurations. Introduction by
used in “image processing”. It is hyperextensional. “Hyperextensionality” in Susan Sontag. Thames and Hudson, London.

18
 deal to paint a pointillist rendering, and so forth.
Similar configurations can be set up through completely abstract construc-
tions. Consider the one shown earlier in relation with Jastrow’s duck-rabbit
(page 9). In the image processing ontology there are nine distinct parts, the
white page, the four black pac-men, and the four red circular segments. The
black pac-men may be merged on the basis of color, so may the red segments.
Moreover, the pac-men and segments may be merged pairwise on the basis
of good continuation. The four disks, again, may be merged on the basis of
non-accidental location, as they form a perfect square.
The awareness ontology is more complicated. There may be a variety of
presentations, although (for most observers) there are two major ones. One
may see a red square on a pattern of four black disks, painted on the page. In
that case the red square is transparent, and seen to cover the white page at the
center (one actually notices a redness). Alternatively, one may see four holes
in the white page, offering a view of an indefinitely extended black space
behind. Through the holes, one sees a red square. This square is opaque. It
is partly occluded by the white page, one notices no trace of redness at the
center.
Perhaps surprisingly, both “red squares” are complete squares, for nothing
is lacking. Their qualitative properties are quite different. They cannot be seen
at the same time. One is “amodally present”, the other “amodally absent”.
The two ontologies give rise to very different parthood relations. Appar-
ently one needs to make a choice here. It is a simple, tractable example of
two parallel universes.
On a single ontological level, you also meet with extreme differences, not
easily handled in a unitary manner. For instance, consider the Pollock Con-
vergence (1952) and Mondrian Composition with Red, Yellow, Blue and Black
(1921)47 paintings. Even in the image processing ontology they can hardly be
Vera Gräfin von Lehndorff-Steinort (or “Veruschka”) in an unusual pose.
analyzed in a single mereology.

The sky extends beneath the wall, and the wall runs on behind her body. Her
body is painted, and very artfully (that is singularly) posed. These relations 47
Pieter Cornelis (“Piet”) Mondriaan, after 1906 Mondrian (1872 – 1944), was a Dutch
could easily be captured by very different styles of depiction. It would no big painter, an important contributor to the De Stijl art movement and group.

19
 Jackson Pollock (1952)
Convergence. Are there
“parts” in this painting?
Apart from the individual
drippings, I mean?

Andy Warhol’s Campbell’s Soup II (1969). Here are some nice parts!
Piet Mondrian (1921) Composi-
tion with Red, Yellow, Blue and
Black. Are there parts? How
many? Since there are about twenty
rectangles one could consider about
2.43 . . . 1018 distinct subsets. I con-
sider it unlikely that Mondrian had
all these in mind.

Andy Warhol (1928 – 1987) was an American pop artist . In his Campbell’s
Soup II (1969) the situation looks somewhat more promising for the image
processing ontology. But consider the ontology of awareness. Hans Makart
(1840 – 1884) was a nineteenth century Austrian academic history painter,
designer, and decorator. His Triumph of Ariadne dates from 1874. Both the
Warholand the Makart have a number of similar objects (soup cans, or human Hans Makart’s Triumph of Ariadne (1874), has obvious parts too. Human
bodies as the case may be) in some particular configuration. But you would figures instead of soup cans in this case.
be hard put to describe them in a similar mereology.

20
 Jean-Auguste-Dominique Ingres (1780 – 1867) was a French Neoclassical
painter. His La source (1820) illustrates a related problem. In Pablo Picasso’s
Girl with a Mandolin (Fanny Tellier), painted in Paris, spring 1910, the figure
merges into the background, which does not really runs on behind the fig-
ure. Fluctuating figure-ground distinctions yield a dynamic impression. Both
paintings show a female figure in a relaxed pose. But notice the part-hood
relations between the figure and its environment (the figure-ground relations
of the Gestalt school). They are qualitatively different.

Ingres’ famous La source (1820). The figure is

well delineated and clearly appears in front of
the background.

Picasso’s equally famous Girl with a Man-

dolin (Fanny Tellier) (1910). Picasso knitted
the figure tightly into the background. There
is no gap between the background and the
figure here. Suzanne Unrein Rapid East (2010). An “illegible Rubens”?

Suzanne Unrein is a contemporary American painter. In her painting Rapid

East (2010) the figure-ground distinctions are vague and ambiguous, thus the

21
painting remains interesting on prolonged looking. Although on first blush task in bringing the mere presentation into cognition. Conceptually important,
one believes to see something alike a Rubensian rape scene, it is actually un- the taxonomic action actually creates what it investigates.
clear how many objects there are in the scene if you want to see it like that. Notice that Adam did a laudable job with the taxonomy of the organic
The painting is interesting, exactly because it keeps microgenesis forever on world, even without the help of DNA analysis, or fossile records. He must
its toes, so to speak. The part-whole relationships are continually fluctuat- have used methods of selection, analogy, etc, essentially relying on inborn
ing. This is an important property of pictures, always present to some degree. Gestalt principles. Our ancestors used these principles as a matter of course,
Unfortunately, this important property is commonly ignored. without ever having been made explicit. The first to draw attention to this
TAXONOMIES remarkable fact was perhaps Goethe48 , the first attempt at a scientific ac-
In view of the enormous differences between various styles an attempt at a count perhaps Remane49 An understanding from psychology came only later,
uniform description, universally applicable to pictures, is certainly useless. At with the work of Rosch on the principle of “family resemblances” 50 . Mod-
one extreme one might consider single pictures as closed universes. It should ern methods have hardly changed this venerable taxonomy, they merely con-
be feasible to find a mereological description of any given picture. However, firmed it, with (very) few minor corrections. Unfortunately, modern science
one might well object that such an extreme is useless in principle. Perhaps saw it fit to condemn authors as Goethe or Remane as (perhaps commendable,
more reasonably, one might attempt to find homogeneity through selection but essentially proto-scientific) amateurs. It remains an embarrassing fact in
according to historical, geographical, and cultural criteria, perhaps concen- the history of science. Apparently human intuition often manages to carve
trating on “styles”, or artistic careers. Here I make a start by considering nature at its joints.
single works. Even the case of the “single work” is non-trivial, for two ob- Paul Klee (1879 – 1940) is considered both a German and a Swiss painter.
servers confronted with the same physical object may well experience distinct His painting Brother and sister (1930) neatly illustrates how multiple tax-
works of art. Their mereological description will differ. This applies equally onomies intersect. The siblings hold a single heart in common. The drawing
to a “single” observer at different times (technically these are two observers!). has a very tight structure, only a supersurgeon might separate these Siamese
Just consider the duck–rabbit. twins successfully. It contains “obvious wholes” on one ontic stratum that do
Indeed, any picture has to be considered a “world”. By drawing and paint- and do not mesh with objects on another stratum. Notice that the picture con-
ing the artist creates possible worlds, observers do the same by looking. A tains a number (the exact number remains ambiguous, perhaps half a dozen)
natural initial task of a novel inhabitant of the world (the observer) is to name of “rounded rectangles”. These rectangles have ambiguous mutual relations,
everything in sight. This is much like the task set to Adam in the Garden of and are themselves ambiguous due to the treatment of the “lines”. One of
Eden (Genesis 2:19–20): these objects contains three “eyes”. Thus, it must be part of both siblings. But
the brother and sister are composed of the rectangles. The rectangles are in
And out of the ground the Lord God formed every beast of the field, one ontic stratum, the brother and sister on another one. Observers find it next
and every fowl of the air; and brought them unto Adam to see what
48
he would call them: and whatsoever Adam called every living crea- Goethe, J.W. von (1790). Versuch die Metamorphose der Pflanzen zu erklären. Ettinger-
sche Buchhandlung, Gotha.
ture, that was the name thereof. And Adam gave names to all cattle, 49
Remane, A. (1952). Die Grundlagen des natürlichen Systems, der vergleichenden
and to the fowl of the air, and to every beast of the field; . . . Anatomie und der Phylogenetik, Geest & Portig K.-G., Leipzig; Riedl, R. (1975). Die Ord-
nung des Lebendigen: Systembedingungen der Evolution, Parey, Hamburg/Berlin.
. Here “name” should be understood as “notice”, where the noticing is en- 50
Rosch E., Mervis, C. (1975) Family resemblances: studies in the internal structure of
abled by the creative act of microgenesis. To establish a taxonomy is the first categories, Cognitive Psychology 7, 573–605.

22
to impossible not to mix up these strata. The mix forever remains a doubtful As microgenesis progresses, a world comes into being through progressive
one. differentiation, controlled through constraints. Among these constraints is
the art work towards which the agent’s interest is directed. The work helps
“sculpt” the developing world. The world is like a taxonomic tree, but it is a
dynamic taxonomy.
Wholes are not determined, but exist as potential sources of parts. Parts are
not determined either, they have only a floating existence. The whole may
change and make the part irrelevant, or a part may take on the role of a whole.
The duck-rabbit serves as an example. Notice how the parts and the wholes
come and go together.
At the end of the day (a full evolution of a visual presentation, say a tenth of
second) a world, that is a pictorial reality comes to life. The next presentation
is often similar, but might be very different, and — of course — “similar” is
by no means the same as “identical”.
Microgenesis is a fractioning of wholes into parts (Brown, 1999), “wholes”
and “parts” both being “pictorial objects”. These pictorial objects derive from
Paul Klee. more basic objects, and are thus both “parts” and “wholes”. If there are places
where microgenesis starts, then these are the core-self of the observer, and
perhaps the gist of the art work. Eventually, the observer sees itself, through
the mirror held up by the artist. It is the double-intentionality in reverse.

Conclusion

Notice that this makes it impossible to arrive at a “natural” mereology, ex- The conclusion has to be that the concept of a general mereology of pictures
actly one reason why the picture is such a strong one. Things fit together is an oxymoron. In pictures we meet with the creative powers of an artist, or
perfectly, straddling distinct ontic strata. The picture perhaps mimics the com- scientist (it makes no essential difference), mingled with the creative powers
plexity of biological organisms. It will amply repay endless study. of one or more observers. Typically, the artist/scientist and the observers will
The Klee example is a good one, because it illustrates my key point. The never have met as persons. They meet in the artwork, or the visual proof,
core of microgenesis is polyvalent. It has the potential to generate many (of- as the case may be. Both the artist and the observers have experiences that
ten surprising) possible worlds. At the core is the agent, not the art work.51 fluctuate between various parallel pictorial worlds. These by no means need
51
Brown, J. (1999). Microgenesis and buddhism: The concept of momentariness. Philos- be the same, or even similar, although they often mesh. A detached empathy
ophy East and West 49(3), 261–277. acts as a glue.

23
 Thus, a mereological approach is not going to work. This is, perhaps, some- ordered (“nothing can be changed”). This rules out formal methods of de-
what of a surprise, because Gestalt properties are all about parthood relations. scription of order (mereologies, information theory, . . . ). It implies that works
But attempts at defining a mereological description fail because of a number of art are closed systems that do not admit of formal descriptions that rely on
of reasons, among more: external, established frameworks. It is exactly the conclusion reached here.
Similar conclusions have been reached by others53 .
— pictures need to be approached on different ontological levels simultane-
ously;
— these levels may interact in unexpected ways;
— identity of pictorial objects does not recognize Leibniz’s Law;
— pictorial objects are not necessarily self-identical;
— failures of transitivity are common, due to scale issues, and bottlenecks
of structural information processing;
— proper parthood fails for the basic figure-ground split;
— amodal elements often pop-up unexpectedly;
— a presentation is a construction, rather than an analysis;
— a presentation lasts only for a moment. Different presentations are par-
allel worlds, they all have to be taken into account;
— the pictorial worlds of artist and observer are generally distinct.
One is in a very wild and dense jungle here. It is unexplored territory for the
scientist, who at every advance hits on traces of artists who passed through
the region before.
I suggest that — for the purposes of experimental phenomenology — pic-
tures should be approached as singular instances of “possible worlds”, and
that an early task of the observer is essentially taxonomic. This task is mainly
handled by microgenesis, but additionally in early cognition, and even in ret-
rospective, reflective thought. In retrospect this is not unlike Bergson’s52 no-
tion. A work of art is by definition both unpredictable as well as perfectly
52 53
Henry Bergson, L’Evolution créatrice, 1907 (many modern editions and translations Lorand, R. (1999). Bergson’s concept of art. British Journal of Aesthetics 39(4), 400–
available). 415.

24
 “eye measure” proof of the parallelogram of forces.
OTHER E B OOKS FROM T HE C LOOTCRANS P RESS :
The key argument is
1. Awareness (2012) de cloten sullen uyt haer selven een eeuwich roersel maken, t’welck
2. MultipleWorlds (2012) valsch is.
3. ChronoGeometry (2012)
4. Graph Spaces (2012) Simon Stevin was a Dutch genius, not only a mathematician, but also an
5. Pictorial Shape (2012) engineer with remarkable horse sense. I consider his “clootcrans bewijs” one
6. Shadows of Shape (2012) of the jewels of sixteenth century science. It is “natural philosophy” at its
7. Through the Looking Glass: on Viewing Aids (2012) best.
8. Painting to Marble (2012)
9. Experimental Phenomenology: Art & Science (2012)
10. The Spirit of the New Style (2013)
11. World, Environment, Umwelt and Innerworld (2013).

(Available for download here.)

A BOUT T HE C LOOTCRANS P RESS

The Clootcrans Press is a selfpublishing initiative of Jan Koenderink. No-

tice that the publisher takes no responsibility for the contents, except that he
gave it an honest try—as he always does. Since the books are free you should
have no reason to complain.

T HE “C LOOTCRANS ” appears on the front page of Simon Stevin’s

(Brugge, 1548–1620, Den Haag) De Beghinselen der Weeghconst, published
1586 at Christoffel Plantijn’s Press at Leyden in one volume with De Weegh-
daet, De Beghinselen des Waterwichts, and a Anhang. In 1605 there appeared
a supplement Byvough der Weeghconst in the Wisconstige Gedachtenissen.
The text reads “Wonder en is gheen wonder”. The figure gives an intuitive

25
26
27