07 Luthy & Smets - Towards A History of Scientific Imagery

Early Science and Medicine 14 (2009) 398-439 www.brill.
nl/esm
Words, Lines, Diagrams, Images:

Towards a History of Scientific Imagery
Christoph Lüthy & Alexis Smets*

Radboud University Nijmegen
To John E. Murdoch, eminent iconographer
Abstract
is essay examines the problems encountered in contemporary attempts to establish
a typology of medieval and early modern scientific images, and to associate apparent
types with certain standard meanings. Five particular issues are addressed here: (i) the
unclear boundary between words and images; (ii) the problem of morphologically
similar images possessing incompatible meanings; (iii) the converse problem of com-
parable objects or processes being expressed by extremely dissimilar visual means; (iv)
the impossibility of matching modern with historical iconographical terminologies;
and (v) the fact that the meaning of a given image can only be grasped in the context
of the epistemological, metaphysical and social assumptions within which it is embed-
ded. e essay ends by concluding that no scientific image can ever be understood
apart from its philosophical preconditions, and that these preconditions are often
explained during disputes between the protagonists of different iconographical types.
Keywords
scientific imagery, epistemic images, word and image, taxonomy of images, iconogra-
phy, chymistry, Marsilio Ficino, Giordano Bruno, René Descartes
* Faculty of Philosophy, Radboud University Nijmegen, P.O. Box 9103, NL—6500

HD Nijmegen, e Netherlands ([email protected]; [email protected]). We would
like to thank the editors of this volume, William R. Newman and Edith Sylla, for their
precious comments and observations.
© Koninklijke Brill NV, Leiden, 2009 DOI : 10.1163/157338209X425632

C. Lüthy, A. Smets / Early Science and Medicine 14 (2009) 398-439 399
Introduction
Contemporary science is pictorial and often even picturesque. Hardly
any laboratory press release comes without colorful graphics, no lec-
ture seems to be able to do without gaudy slides, no journal article
without glossy pictures. e multiple reasons behind this prolifer-
ation of images need not concern the historian of medieval and
early modern science. But if Martin Kemp is correct in claiming
that “the modes of representation in twentieth-century science are
very much the heirs of the Renaissance revolution” that led to “the
rise of illustration as a major tool of science,” then it falls to the
historian to elucidate both this alleged revolution and to retrace its
wider implications, also for our own times.1
In fact, much has in the past fifteen years been written about
the visual aspects of science past and present. Historians of art have
begun to study non-artistic visual manifestations, including so-called
‘epistemic images’ appearing in the context of theory formation.2
Historians of science, in turn, have paid increasing attention to the
visual manifestations of scientific theory and practice across the cen-
turies, although they have shown a tendency “to treat scientific pic-
tures only as after-images of verbal ideas.”3 Philosophers of science,
finally, are busy analyzing the function of models and diagrammatic
representations in the logic of scientific discovery and explanation
as well as in the dissemination of knowledge.4
1)
Martin Kemp, “Seeing and Picturing. Visual Representation in Twentieth-Centu-
ry Science,” in Science in the Twentieth Century, eds. John Krige and Dominique Pestre
(Amsterdam, 1997), 361-90, at 363.
2)
In this paper, we use the term ‘epistemic image’ to refer to any image that was made
with the intention of expressing, demonstrating or illustrating a theory. e more fre-
quent term, ‘scientific image’, which we also invoke because of its greater accessibility,
has however an awkwardly anachronistic ring to it when applied to centuries in which
scientia meant something quite different than ‘science’ does today.
3)
e quote is from David Topper, “Towards an Epistemology of Scientific Illustra-
tion,” in Picturing Knowledge. Historical and Philosophical Problems Concerning the Use
of Art in Science, ed. Brian S. Baigrie (Toronto, 1996), 215-49, at 215.
4)
Davis Baird, ing Knowledge. A Philosophy of Scientific Instruments (Berkeley,
2004) views as instruments both two- and three-dimensional ‘models’ (see notably his
400 C. Lüthy, A. Smets / Early Science and Medicine 14 (2009) 398-439
Underlying much of this prolific literature, however, one finds

the assumption that it is legitimate to speak systematically about
images, rather than merely historically and contingently. is view
presumes (1) that there are timeless criteria for distinguishing images
from non-images; (2) that images possess a fairly stable ontological
and epistemic status across the centuries; and (3) that it is possible
to develop a stable classification or taxonomy of images.
All three assumptions seem to us not only doubtful, but also
open to refutation. Indeed, it is the purpose of this essay to docu-
ment a number of complications that arise from any supra-histor-
ical, essentialist approach to epistemic images and to suggest instead
an approach that takes into consideration the epistemological, onto-
logical and pedagogical assumptions that surrounded their produc-
tion. It is our conviction that the investigation of epistemic images
requires a patient study of that endless range of stable or unstable,
but always temporally and geographically bounded iconographical
traditions, on the one hand, and on the other hand of the way in
which each of these traditions was in time extended, subverted,
redefined or simply replaced by another.
e reason why this subject matter deserves to be discussed in
the present context is that this essay originated in a lecture held
during an academic celebration of John Murdoch’s 80th birthday.
With his monographic contribution to I.B. Cohen’s Album of Sci-
ence series, Murdoch has, after all, provided one of the earliest and
most trenchant analyses of the logic of historical epistemic images.
Unlike the other authors in Cohen’s series, who in keeping with
the series title word did produce ‘albums’, Murdoch spent little
energy on the display and discussion of images that merely illus-
trated scientific theories and practices (e.g., images of star-gazers,
surgical instruments, or plant species), but chose instead to focus
on the question of how images attempted to express, prove, organize
or interpret the arguments found in the texts to which they were
ch. 2: “Models: Representing ings”). In a similar direction go recent attempts to

place images in the category of ‘paper tools’; see e.g. Ursula Klein, Experiments, Models,
Paper Tools: Cultures of Organic Chemistry in the Nineteenth Century (Stanford, 2003).
attached (by means of diagrams, geometrical proof, sciagraphiæ,

etc.).5
Problem I: e Unclear Boundary between Words and Images

Particularly praiseworthy seems to us Murdoch’s attention to the
way in which words can pattern themselves out over a page and in
so doing take the first step towards image formation—a sensitivity
that is absent from all those contemporary discussions that either
bluntly treat images as texts, or alternatively presume with equal
bluntness an essential distinction between word and image. A study
of the illustrations included in part II of Murdoch’s book, “Stan-
dard Schemata and Techniques for the Visual Facilitation of Learn-
ing,” will inevitably remind us of the fact that the Greek word
γραφεῖν (grafein) means any gesture that (literally) ‘engraves’ some-
thing on a tablet—irrespective of whether the result is a word (which
might, or might not, violate the rules of orthography), or a diagram,
or indeed a graphic design. When the γραφίς (grafis), the slate pen-
cil, is made to draw its lines, the result can be a ‘drawing’, a ‘let-
ter’, or some other type of ‘text’—but in each case it will be a
γραφή (grafê) or a γράμμα (gramma)—for these two all-embrac-
ing words mean all of these things.6 In the particular case of hiero-
glyphics, the ‘drawing’ and the ‘letter’ may even fall together, and
only the context will tell you whether the drawings should be read
as text or as image.7
5)
John Murdoch, Album of Science. Vol. 1: Antiquity and Middle Ages (New York,
1984). Under the general editorship of I. Bernard Cohen, three more volumes were
published in this series.
6)
Noëlle Batt, “L’expérience diagrammatique: Un nouveau régime de pensée,” in Pen-
ser par le diagramme: De Gilles Deleuze à Gilles Châtelet, ed. Batt (Saint-Denis, 2004),
5-28, suggests that the verb γραφεîν hails from the Indo-European root ‘grbh-mn’,
in which ‘grbh’ means ‘to scratch’ and ‘mn’ ‘image, letter, text’. According to this et-
ymological reconstruction, the manual gesture and its outcome would early on have
merged into a single verbal unity. is etymology does not differ very significantly
from that given by James Elkins in e Domain of Images (Ithaca, NY, 1999), 82.
7)
e image-text division of hieroglyphics is further complicated by the fact that hi-
eroglyphic texts can choose between phonetic and ideogrammatic scripture. On this
In his ground-breaking e Domain of Images, James Elkins chal-

lenges Nelson Goodman’s division of ‘image’ into writing, notation
and picture.8 He demonstrates that, upon close inspection, almost
any image seems to rely on a blend of all three categories, though
examples can be found to document the gradual progression from
‘pure scripture’ to ‘pure picture’.9 Interestingly, Murdoch’s collection
of images appears to corroborate Elkins’ research, although it doc-
uments a different type of progression from word to image. It allows
us to witness how arguments occurring in a running text are ordered
into columns or what we would now call ‘text boxes’ by an assiduous
copyist (see fig. 1); how lines are drawn between such singled-out
words to show logical dependencies, as in the case of dichotomies
(see fig. 2); how dichotomies can be organized into the shape of
trees (see fig. 3); and how these trees end up being decorated with
leaves and apples and sometimes inserted into landscapes or cos-
mological maps (see fig. 4). is succession of examples, which doc-
uments the transformation from a word pattern over a simple
dichotomy to a ‘tree of Porphyry’ (which can in turn be planted
within a fully-fledged cosmological Weltbild), proves that there is
no definite borderline between text and image.10
e ease with which words can become patterns and patterns can
become images is fascinating—and at the same highly problematic:
issue, and for a criticism of the image-text distinction, see W.J.T. Mitchell, Iconology:
Image, Text, Ideology (Chicago, 1986), esp. 24-30.
8)
Elkins’ concept of notation differs from Goodman’s. For the former, a notation is
“an image employing organizational principles other than the formats associated with
pictures or writing systems,” whereas for the latter, notational systems are syntactically
and semantically unambiguous systems that in turn “ensure a one-to-one correspon-
dence between the notation … and real-world denotata” (Domain of Images, 257 and
69, respectively).
9)
Elkins proposes instead a classification into seven categories that triangulate, as it
were, between Goodman’s three categories. For his suggested progression from ‘pure
writing’ through ‘allographs’, ‘subgraphemics’, and ‘hypographics’ to potentially ‘pure
pictures’, and the notion of potentially ‘pure notations’, see his Domain of Images,
82-91 and part II.
10)
On the relation between world-view, visual view of the world, and Weltbild, see Die
Welt als Bild. Interdisziplinäre Beiträge zur Visualität von Weltbildern, eds. Christoph
Markschies and Johannes Zachhuber (Berlin, 2008).
2
1
3 4
Figures 1-4: e gradual evolution of a tabulated text into a landscape.
Fig. 1: To facilitate comprehension, an intelligent fourteenth-century scribe has tab-
ulated Cassiodorus’ description of the nine valid modi of the first formula of categor-
ical syllogisms (MS Bibliothèque nationale, Paris, fonds latin, 8500, fol. 34v.
Reproduced with kind permission of the BNF). Fig. 2: A fifteenth-century student
sets out the then available works on mathematics and physics in a dichotomizing form,
which begins, on the left, with the division of mathematics into ‘discrete’ and ‘con-
tinuous’ quantities. Conceptual relations are indicated by lines (MS Öffentliche Bib-
liothek der Universität Basel, F.II.8, fol. 45r. Reproduced with kind permission of the
→
for this fluid transition from writing to imaging massively complicates

attempts at a typology or taxonomy of epistemic imagery; renders
the question of the function and role of such images impossible to
answer in any general way; and undermines attempts to define the
relation between words, concepts, theories, and images.
Whereas the transition from the flowing prose text to the table
and from the table to the dichotomy (figs. 1-2) can be handled by
a general theory of semiotics, the introduction of figurative elements
eludes attempts at a timeless interpretation of signs. e tree in fig-
ure 3, for example, requires an understanding of the reference to
the lignum vitæ of Apocalypse 22,2, which is mentioned in the root
of the tree, that is to say, to “the tree of life, bearing twelve fruits,
yielding its fruits every month: the leaves of the tree for the heal-
ing of the nations.” e twelve leaves in this image are each asso-
ciated with a prophet, and each branch with four characterizations
of Jesus. e pelican doesn’t nest for decorative reasons at the top
of this tree, but in reference to Christ’s self-sacrifice at the Cross.
In fact, an inscription on the trunk states that “there should be an
image of a crucifix depicted in the center of the tree.” Now, given
that the Christological pelican is often depicted at the top of a cru-
cifix, this tree can also be read as a twelve-armed cross—a duplica-
tion of both morphology and meaning that undermines attempts
to fit such an image into any unambiguous category.11
11)
Cf. Murdoch, Album, 49.
UB Basel). Fig. 3: When maps and dichotomies grow upwards, rather than from left
to right, they become trees, as in this fifteenth-century “tree of prophets” (arbor prophet-
arum; MS Bibliothèque nationale, Paris, fonds latin, 3473, fol. 80v. Reproduced with
kind permission of the BNF). Fig. 4: e tree has become a pictorial component in a
miniature landscape in this fourteenth-century initial ‘C’, in which a teacher of logic
teaches two students the very fundamentals of this art that are shown in the dichoto-
mizing tree in whose shadow he sits (MS British Library, Burney 275, fol. 166r. Repro-
duced with kind permission of the British Library). (On the four images, see Murdoch,
Album, 37, 42, 49 and 51.)
Problem II: Similar Shape, Different Meaning

is fluid transformation from a dichotomy of words into a fully
formed tree can be documented throughout history. Take that note-
book entry by Charles Darwin, that famous impulse-turned-graph,
the “I think” followed by a drawing, subsequently annotated and
explained (fig. 5). It is a particularly impressive demonstration of
how an idea can first announce itself visually before becoming fully
verbal—a phenomenon trenchantly analyzed in Rudolf Arnheim’s
Figures 5 and 6: Two ‘trees of evolution’? Fig. 5: Charles Darwin’s 1837 sketch of spe-
ciation in Notebook B (Cambridge University Library, Dar. Ms 121, fol. 36; repro-
duced by kind permission of the Syndics of Cambridge University Library). Fig. 6:
Ernst Haeckel’s Stammbaum der Organismen, from his Generelle Morphologie der Orga-
nismen (Berlin, 1866).
study on Visual inking.12 But let us compare these ink lines with
Ernst Haeckel’s tree of organic life (fig. 6). It seems to us evident
that Darwin and Haeckel would have understood each other’s images
correctly, because figures 5 and 6 are both expressions of the same
tradition of representing divisions of conceptual or biological spe-
cies in terms of branching lines. And yet, are the two ‘images’ equiv-
alent? Are they both ‘trees’? Assume that we allow this name also
for figure 5, then, does it matter that it spreads out in various direc-
tions, whereas Haeckel’s tree (fig. 6) only grows upward? We should
like to argue that it does matter; and that it makes a marked log-
ical difference whether a dichotomy is unidirectional or not; whether
its unfolding suggests goal-directedness (as in figs. 3 and 6) or ran-
domness (as in fig. 5); and finally, whether the branching-out fol-
lows logical distinctions or bifurcations in time, or both. In sum,
then, a tree isn’t a tree. Our morphological names often don’t seem
to capture the precise logic of a given image.13
e meaning of the different exemplars of what might appear to
constitute the same morphological type is thus dependent on the
logic of the conceptual relations that they are each meant to visu-
alize. We have just seen how easily a mapped-out list of names can
turn into a biblical tree possessing soteriological overtones, and how
an intuitive drawing of the logic of speciation can turn into an his-
torical taxonomy.
But there even are more radical examples of the diversity of mean-
ings of morphologically similar images. e extremely simple case
of six circles of equal diameter grouped around a seventh is, in this
respect, very striking.14 Structurally speaking, figures 7, 8, 9, and
10 differ but little. And yet, they were intended to explain extremely
different theories, entities and spatial dimensions. Figure 7 hails
from a 1495 edition of omas Bradwardine’s Geometria speculativa,
12)
Rudolf Arnheim, Visual inking (Berkeley, 1969).
13)
On the relation of Darwin’s drawings to other types of ‘trees’, see Horst Brede-
kamp, Darwins Korallen. Die frühen Evolutionsdiagramme und die Tradition der Natur-
geschichte (Berlin, 2006), notably ch. 1.
14)
is example has been analyzed in greater detail in Christoph Lüthy, “e Inven-
tion of Atomist Iconography,” in e Power of Images in Early Modern Science, eds.
Wolfgang Lefèvre, Jürgen Renn and Urs Schoepflin (Basel, 2003), 117-39.
Figures 7-10: How many theories can six circles placed around a seventh circle possi-
bly visualize? Fig. 7 demonstrates the perfection of the number ‘6’ and of the circle
(omas Bradwardine, Geometria speculativa [Paris, 1495], f. 8r). Fig. 8 explains the
numerological logic of the six days of creation (Pietro Bongo, Mysticæ numerorum sig-
nificationis liber [Bergamo, 1585]). Fig. 9 shows us 7 worlds touching (Giordano
Bruno, De l’infinito, universo et mondi [London, 1584], 145). Fig. 10 documents how
atoms aggregate to form larger globules (Giordano Bruno, De triplici minimo et men-
sura [Frankfurt, 1591], 50).
where it appears as the symbol of both the number ‘6’ and of “the
perfection of the circle” (because the six center points of the periph-
eral circles coincide with the extremities of the three diameters going
through the center of the central circle, and because the six sides
formed by the new figure can be inscribed into a new circle). ough
crafted like figure 7 in numerological praise of the number ‘6’, fig-
ure 8 wishes instead to express the logic of the six days of creation
and the seventh of rest: the six days are grouped around the cen-
tral divinity, whose restfulness, quies, is suggested by the inscribed
‘Q’. e other letters are also abbreviations: ‘E’ stands for Elemen-
tale, ‘M’ for Minerale, ‘V’ for Vegetabile, and so forth. is image,
like the previous one, receives its meaning and coherence thus exclu-
sively from the accompanying text, which in this case is Pietro Bon-
go’s Mysticæ numerorum significationis liber. Figure 9, by contrast,
doesn’t purport to reveal any numerological mystery, but intends to
visualize cosmological configurations. Personally engraved by Gior-
dano Bruno, this woodcut is meant to illustrate the way in which
worlds (mundi) touch each other. According to Bruno’s teaching,
each one of the infinite worlds has a sun at its center (which he
rendered visible with an aggrandized compass needle point). Figure
10 is also by Bruno, but this time, what he engraved were not
worlds, but atoms. Importantly, the morphological similarity between
his two woodcuts was not only intended, but was, in Bruno’s eyes,
Figure 11: e same seven circles modelling “dendritic growth” of silver atoms on a
close-packed surface (J.V. Barth & H. Brune, “Atomare Prozesse,” 256).
pregnant with meaning. In fact, the stars placed in the four cor-
ners of figure 10 are a mnemonic device to remind the viewer of
the similarity of the cosmological and the atomic configurations.
Bruno, after all, subscribed to Cusanus’ doctrine of the coincidence
of the opposites (according to which contraries, including minimal
and maximal geometrical forms, coincide) and also appreciated Dem-
ocritus’ atomistic interpretation of the infinite worlds. is is pre-
cisely why he felt that he could depict minimal material atoms by
means of essentially the same image as maximally large worlds, and
why the seven circles represented, in his eyes, an ‘archetype’.
Figure 11, by contrast, is taken from a recent article in a phys-
ics journal. As its authors explain, the seven black disks represent
a “heptamer of 7 silver atoms,” which are “growing in the direc-
tion preferred by snowflakes” on top of a “hexagonally close-packed
surface.”15 Iconographically speaking, the heptamer is the direct his-
torical descendant of figure 10 and thereby adds confirmation to
Martin Kemp’s claim that contemporary conventions in scientific
iconography often have a Renaissance background.16 Nevertheless,
these visually similar atoms refer to entities that belong to incom-
mensurable theories of matter: the (composite and hence divisible)
15)
Johannes V. Barth and Harald Brune, “Atomare Prozesse an Oberflächen,” Physik
in unserer Zeit, 29 (1998), 251-60, at 256.
16)
e iconographical continuity in representing clusters of atoms—from Bruno
through Johannes Kepler’s crystallographic Strena to contemporary imagery produced
by scanning tunneling microscopy—is documented in Christoph Lüthy, De draad van
Ariadne: een pleidooi voor de wetenschapsgeschiedenis (Nijmegen, 2007).
silver atoms represented by the dark disks in figure 11 share no

common trait with Bruno’s (indivisible and ensouled) minima.17
Figures 7 through 11, then, place us before the problem that
even where morphologically speaking, one seems to be confronted
with a single type of image that survived the centuries quite intact,
it is impossible to give it a name, let alone a unitary meaning. e
different geometrical, numerological, theological, cosmological, and
atomistic meanings are just too diverse to allow for a unitary descrip-
tion, either as a geometrical construction, as a symbol or structural
model. In fact, just like the tree in figure 3, which also invoked a
cross, figures 8 to 10 quite explicity rejoice in their multiple icon-
ographic meanings. For Bongo, the geometrical and numerological
properties enhanced the power of his image of creation (fig. 8), and
for Bruno, the ubiquitous applicability of the same growth pattern
to the smallest and largest dimensions added force to his general
archetype of generation (figs. 9 and 10).
Possibly the most striking evidence for the divergence between
shape and meaning is provided by what is defined as a ‘diagram’
by some and as ‘graph’ by others, and which consists in a line
between two coordinates, as in figure 12.18
17)
Aby Warburg would have been fascinated by this particularly powerful example of
pictorial ‘Nachleben’—which is obviously quite counterintuitive, given that it seems
to transcend the various theoretical revolutions occurring in the domain of matter the-
ory between Bruno and today’s physical theories. On Warburg, see Georges Didi-Hu-
berman, L’image survivante: Histoire de l’art et temps des fantômes selon Aby Warburg
(Paris, 2002), notably 103.
18)
See, for example, Jacques Bertin, Sémiologie graphique: Les diagrammes, les réseaux,
les cartes (Paris, 1973), 50: “quand toutes les correspondances dans le plan peuvent être
établies entre toutes les divisions d’une ordonnée et toutes les divisions d’une autre or-
donnée, alors la construction est un diagramme.” John J. Roche, “e Semantics of
Graphics in Mathematical Natural Philosophy,” in Non-verbal Communication in Sci-
ence Prior to 1900, ed. Renato G. Mazzolini (Florence, 1993), 197-233, at 212: “e
invention of coordinate geometry by René Descartes made possible and stimulated
the representation of the quantitative laws of physics by diagrams, today commonly
called graphs.” e mutual relation between the terms ‘diagram’, ‘graph’ and ‘schema’
remains problematic, as becomes evident in Batt, “L’expérience diagrammatique,” and
in François Dagognet, Écriture et iconographie (Paris, 1973).
Figures 12a and b: e transformation of a diagram (12a) into a graph (12b). When
the diagonal becomes the curve in a graph, the bottom left angle becomes the origin
and the left vertical and bottom horizontal sides of the former rectangle become direc-
tional axes, which can be indefinitely extended in the direction of the arrows.
Although the ‘graph’ is, iconographically speaking, an indirect

descendant of the mathematical diagram of Greek extraction and
may still resemble its ancestor (as in figs. 12a and 12b), any such
similarity is deceitful. For there are two abysses that separate the
graph from the traditional geometrical proof. e first is consti-
tuted by the definition of the horizontal and vertical lines as axes
carrying a numerical scale, with the line drawn between the two to
express a functional relation to both. Despite all visual resemblances,
this numerical functionality (in which the line can either be the
visual expression of a given function, or conversely, in which the
function can be derived from the line as the combination of empir-
ical measurement points) has a radically different status than the
old geometrical drawing, which formed part of a formal proof. e
second abyss is constituted by the fact that these lines need not
express any spatial extension. Whereas the traditional geometrical
figure not only represents space, but gives us the space of which the
mathematical proof speaks, the horizontal axis of the graph can rep-
resent any kind of magnitude, such as time, pressure, a probability
density, or any other magnitude that can be meaningfully corre-
lated with another magnitude. Having crossed these two abysses,
the diagram turns into a graph which functions, in Michael Mahon-
ey’s words, “in a mathematical space wholly divorced from the phys-
ical space,” a mathematical space moreover that is intended for the
“mind’s eye … peering into the structural relations among quantities
belonging to many different conceptual (rather than perceptual)

spaces.”19
e seven circles in figures 7-11 as much as figure 12 forcefully
demand answers to the following questions: (1) Which iconograph-
ical similarities are essential (in the sense of pointing to an inher-
ent logical or ontological similarity in a theory), and which ones
are accidental? (2) Is the historical ancestry of an image of any
importance, or is the only aspect that counts its function within
the system in which it is assigned its current role? (3) What does
the fact that similar types of representations are used in allegedly
incompatible scientific paradigms across the ages tell us about these
paradigms, our mental structures, or the conventionality of the visual
language being used across numerous scientific generations?
Problem III: Identical Signifiers, Different Types of

Representation
We have just seen why visually similar images do not necessarily
carry identical meanings. But what about the inverse relation? Do
things that are denoted by the same name necessarily require rep-
resentation by the same type of image? To see that the answer is
once again a ‘no’, it will suffice to take a look at the astonishing
range of visual references to mercury in chymical texts across the
centuries.20 In fact, the observed iconographical latitude echoes the
definitional latitude: ‘mercury’ was, after all, not only a specific
metallic substance, but for a long while also one of the three
philosophical principles of chymistry, and as such less specific and
often more spiritual than the quicksilver obtained by reduction from
cinnabar. But the observed iconographical latitude is also due to
the even more fascinating issue of how to represent chemical properties
19)
Michael Mahoney, “Diagrams and Dynamics: Mathematical Perspectives on Edg-
erton’s esis,” in Science and the Arts in the Renaissance, eds. J.W. Shirley and F.D.
Hoeniger (London, 1985), 198-220, at 209.
20)
On the term ‘chymistry’, see William R. Newman and Lawrence M. Principe, “Al-
chemy vs. Chemistry: e Etymological Origins of a Historiographic Mistake,” Early
Science and Medicine 3 (1998), 32-65.
as manifested both in different aggregate states and during opera-

tional interaction with other substances or principles.
Figure 13 is the oldest of the chemical images presented here.21
While clearly a representation of mercury, it also stands in an older
and long-lasting tradition of religious alchemical imagery. Christ
and the Virgin Mary are shown here in mystical, hermaphroditic
union.22 As Barbara Obrist has shown, mercury is omnipresent in
this image. Relying on a lexicon of alchemy of 1612, she points to
the dragon at the bottom as a conventional manner of denoting
“quick-silver or mercurial vapor.”23 But the dragon also denotes Christ
(because of Psalm 22,7, in which David, prefiguring Jesus, claims
to be a ‘worm’). Furthermore, mercury is also represented more
kaleidoscopically in the rest of figure 13, notably in the recoiling
and raising snakes in the hermaphrodite’s hands. Rolled up, the sin-
gle snake refers to the properties of fixed mercury, while the three
snakes emerging from the vessel designate its volatility. But the same
metal is concomitantly identified with the philosopher’s stone, because
of its essential contribution to the generation of gold and silver.
Visually, this special power is represented by the entire image (minus
the two trees). is leads us to our first general observation: given
that chymistry was less a science of individual materials than an
operational science, it was the relational properties of mercury and
its specific agency during interaction with other substances and prin-
ciples that had to be visually documented. Nothing could be fur-
ther from such an understanding than a representation of mercury’s
inner material structure.
21)
is illumination is reproduced and analyzed at length in Barbara Obrist, Les dé-
buts de l’imagerie alchimique (XIV e-XV e siècles) (Paris, 1982), 152 ff.
22)
e process of sexual metaphorization and its usefulness for chymistry is studied in
Allison B. Kavey, “Mercury Falling: Gender Malleability and Sexual Fluidity in Early
Modern Popular Alchemy,” in Chymists and Chymistry: Studies in the History of Alche-
my and Early Modern Chemistry, ed. Lawrence M. Principe (Sagamore Beach, 2007),
125-136.
23)
However, the attribution of ‘symbols’ to specific substances was subject to great
variations throughout the history of chymistry. See Obrist, Les débuts, 253; Maurice P.
Crosland, Historical Studies in the Language of Chemistry (London, 1962).
Figures 13 to 20: Diverse representations of mercury.

Fig. 13: Mercury of the philosophers (Das Buch der heiligen Dreifaltigkeit, ca. 1420;
Germanisches Nationalmuseum Nürnberg, MS 80061. Reproduced with kind per-
mission by the Nationalmuseum). Fig. 14: Mercury dissolving gold and silver (Aurora
consurgens, Zentralbibliothek Zürich, MS Rh. 172, fol. 27v. Reproduced with kind
permission by the Zentralbibliothek). Fig. 15: Mercury protected by iron against the
effect of fire (Michael Maier, Atalanta fugiens [Oppenheim, 1618], Emblem 20). Fig.
16: Mercury represented as the messenger god Hermes-Mercury, with caduceus and
winged helmet (Johann Joachim Becher, Parnassi illustrati pars tertia: Mineralogia, Das
ist: Deß erläuterten medicinalischen Parnassi dritter eil, nemlich das Berg-Buch [Ulm,
1663], 40).
Fig. 17: Mercury immobilized (Giovanni Battista Nazari, Della transmutatione metal-
lica [Brescia, 1572], 11). Fig. 18: Mercury dissolving gold (Nicolaas Hartsoeker, Con-
jectures physiques [Paris, 1706], 130). Fig. 19: e symbol of mercury (Nicaise
Lefebvre, Traité de la Chymie [Paris, 1660], 153). Fig. 20: ‘Combinatorics’ with alchem-
ical symbols (David de Planis Campy, Bouquet composé des plus belles fleurs chimiques
[Paris, 1629], 991).
In figure 14, we are once more confronted with a representation of

mercury, gold, and silver.24 is time, however, the viewer cannot
observe the birth of gold and silver out of mercury, but rather their
demise in the process of calcination. According to Obrist, the pictorial
24)
Reproduced and commented in Barbara Obrist, “Visualization in Medieval Alche-
my,” Hyle 9 (2003), 131-70.
elements of the snake and the spiky crown might either refer to
Egyptian mythology or to the first decan of the Lion.25 While fig-
ures 13 and 14 seem to share a common ‘style’ of metaphorization
and allegorization of the same three metals, they represent their
object differently and according to the canons of different tradi-
tions and methods. Figure 13 is Christian by inspiration and attempts
to describe the physical properties of mercury. Figure 14, by con-
trast, looks to Egyptian and astrological sources. Moreover, apart
maybe from the ax that ‘Mercury’ bears on his left shoulder, this
personification does not seem to indicate any physical property of
the metal, unless one interprets the color and the crown as such
indications, in analogy to the snakes and the dragon of figure 13.
From a relational or operational viewpoint, however, Mercury decap-
itating Sun and Moon may be a reference to its putrefaction or cal-
cination in the amalgamation of the three metals; yet there is no
visible element announcing its further transformation. As is mani-
fest, the two images also represent gold and silver quite differently.
While figure 13 explains chemical change in a biological mode, rep-
resenting silver and gold as the fruit of trees that are nourished by
mercurial rivers, figure 14 personifies the two metals, allowing their
identification only because of their specific coloring. e duplica-
tions at work in these images are also quite noteworthy: in both
cases, the precious metals are indicated by their colors and by sun
and moon symbols (a moon-face is recognizable in figure 14).
We have just now spoken, however provisionally, of a common
style of ‘metaphorization’ and ‘allegorization’. For the iconographer,
the question arises as to whether these two terms can be clearly
defined and applied as classificatory terms to epistemic images. On
the one hand, our dictionaries would suggest that a metaphor is a
figure, and thus a single element, whereas an allegory connotes a
longer process, a story, poem or picture in which several elements
are at play.26 On the other hand, such a categorical distinction raises
25)
Obrist, Les débuts, 236 ff.
26)
is distinction between process and element is already drawn in early modern dic-
tionaries, as in the first edition of the Dictionnaire de l’Académie française (Paris, 1694),
s.v.: “Discours par lequel …” (allégorie); “Figure de discours …” (métaphore).
various problems, as contemporary studies on metaphorical and alle-

gorical styles amply document. Furthermore, the reduction of images
such as figures 13 and 14 (together with the texts in which they
are embedded) to metaphors and allegories is also historically inac-
curate, because the rhetorical vocabulary at the disposal of chymi-
cal authors was both larger and more subtle than that. In the
fourteenth century, Petrus Bonus conceived of no fewer than a dozen
rhetorical means of referring to a chymical substance, of which only
the first was by its ‘proper name’.27
Such ‘improper’ visual means of representing chymical substances,
principles and operations as those that we have encountered in fig-
ures 13 and 14—whether they be allegorical, metaphorical, hyper-
bolical, or enigmatic—were enduring enough to be in use as late
as the beginning of the seventeenth century, the period in which
our figure 15 was crafted.28 is figure, however, carried the offi-
cial title of ‘emblem’, a classification that refers to a Renaissance
conceptualization of image making.29 We see a girl being protected
by a knight against the deleterious effects of fire. According to Hel-
ena de Jong, the “vulnerable girl … is the personification of the
volatile Mercury.”30 To those acquainted with this particular mercurial
27)
Petrus Bonus, Margarita pretiosa novella correctissima, in eatrum chemicum (Stras-
burg, 1659-1661), 6 vols., 4:507-713, at 516: “Et est sciendum, quod ars ista qua-
si sola inter omnes mundi, in sui doctrina utitur nominibus propriis, et extraneis, et
inusitatis, et allegoriis, et ænigmatibus, et metaphoris, et æquivocationibus, et trans-
sumptionibus, et involucris, et prosopopoeis, et hyperbolis, et ironiis.” Quoted from
Obrist, Les débuts, 48.
28)
is engraving is reproduced and commented in Helena M.E. de Jong, Michael
Maier’s Atalanta fugiens: Sources of an Alchemical Book of Emblems (York Beach, 2002),
162-166.
29)
e concept of ‘emblem’ originated in Andrea Alciati’s Emblemata of 1531. e
expression ‘emblema vermiculatum’, meaning ‘mosaic work’, is indicated as the ori-
gin of the concept of ‘emblem’ by one of Alciati’s numerous editors: “Quare et ipse Al-
ciatus, sua hæc epigrammata appellatione convientissima inscribi voluit Emblemata.
Sunt eum Emblemata vermiculata opera extessellis insititiis apta et composita, inter-
prete Budæo, quod et ipsa vocabuli Greci origo ostendit.” (Andrea Alciati, Emblema-
ta, eds. Macé Bonhomme and Guillaume Rouillé [Lyon, 1551], 4; see also Mario Praz,
Studies in Seventeenth-Century Imagery [Rome, 1964], 23).
30)
De Jong, Maier’s Atalanta fugiens, 166.
personification, figure 15 may indeed have illustrated the limited

resistance of this metal to the action of fire. Similarly, the knight
would have alluded to the fact that when allied with another sub-
stance, mercury could avoid evaporation in the face of fire. But
although the textual sources informing this image don’t differ much
from those used for our previous two figures, this image is strik-
ingly different.31 One of the differences lies in the fact that not all
elements in this copper engraving seem to be part of the allegory.
Only the initiated will understand whether the landscape surround-
ing the woman, the knight and the fire has an emblematic role or
is purely decorative. If the latter, figure 15 would possess the dou-
ble nature of showing an emblem within a landscape or a chymi-
cal image within a non-chymical one. is doubleness reminds us,
in some strangely significant way, of the transition from text to
image in figures 1 to 4, above. In both cases, particular and informed
habits of reading and of viewing seem to be required to recognize
or ‘read’ the significant image in the image, or the visual pattern
in the text, respectively.32
In figure 16, we encounter yet another incarnation of mercury.
Instead of the dragon, the snake, and the naked woman, we have
here the classical representation of the mythological messenger-god
Hermes, or Mercury to the Romans, with wings on helmet and
sandals and the caduceus with its two snakes, which was to become
the pharmacists’ emblem. e text surrounding the image leaves no
doubt: this emblematic figure means the metal mercury. However,
given that it might with equal right refer to the planet of the same
name, the question arises as to whether his portrait tells us any-
thing specific about the metal for which it here made to stand. If
there were no chemical ‘symbols’ inscribed in Mercury’s helmet and
foot, the answer would be ‘no’. But by means of the location of
31)
e subsequent emblem XXI of Maier’s Atalanta fugiens depicts mercury also as a
naked woman, but this time in company of a naked man, who expresses sulfur, the
principal chemical counterpart (cf. De Jong, Maier’s Atalanta fugiens, 169).
32)
is raises the question of the image in relation to its frame. See on this Johannes
Grave, “On the Aesthetics of Scientific Objects. ree Case Studies,” in Wandering
Seminar on Scientific Objects, eds. Sophia Vackimes and Konstanze Weltersbach (Max
Planck Institute for the History of Science, preprint 339, 2007), 35-48.
these ‘symbols’, Becher distinguishes between the volatile and fixed

forms of mercury: pure mercury and mercury sublimate are assigned
to the air, and mercury precipitate and cinnabar are assigned to the
earth.33 e style of representation has also changed. Figure 16 is
heterogeneous, in that it superimposes letter-like signs on a person-
ification, as if on a map.34 e accompanying text is, as in some
emblem books, divided into three parts: there is a title that carries
the different names of the substance (Mercurius, Argentum vivum,
and Quecksilber); there is an allegorizing poem; and finally, there is
concrete, sober and non-allusive prose in which an account of mer-
cury is given.35
In figure 17, which is taken from Giovanni Battista Nazari’s Della
transmutatione metallica (1572) and is thus a century older than
figure 16, Mercury is recognizable only thanks to his winged hel-
met, given that his hands and feet are missing. In contradistinction
to figure 16, however, figure 17 has strong allegorical overtones, as
several of its pictorial elements (the basin with aquatic plants, the
33)
e circle horizontally cut into two equal parts and surmounted by a cross is giv-
en as a denotation of cinnabar in Nicaise Lefebvre, Traité de la chimie (Paris, 1660),
“l’explication des caracteres chymiques,” 152-53. is ‘sign’ minus the line that cuts
the circle horizontally is also presented as cinnabar in Nicolas Lemery, Cours de chy-
mie (Paris, 1757), s.p. [789].
34)
Different types of heterogeneity and duplication are found in the earlier astrolog-
ical representations of metals; cf. Françoise Cannella, “Alchemical Iconography at the
Dawn of the Modern Age: e Splendor Solis of Salomon Trismosin,” in e Power of
Images, eds. Lefèvre et al., 107-16; and Guy de Tervarent, De la méthode iconologique
(Bruxelles, 1961).
35)
In the first, playful verses of the poem, the god Mercury is described as being se-
rious with the chymists, but playful with the alchemists: “Es komt Mercurius, der
schnell geflüglet Gott / Er brauchet einen Ernst / und lässet seinen Spott; Den er zu
treiben sonst mit Alchymisten pflegt / hier aber sich bey ihm / ein andrer Ernst er-
regt.” e textual description of mercury is, by contrast, of empirical sobriety: “Das
Queck-Silber Erß wird gewürket in seinen eigenen Bergsteinen / von seiner Natur der
Sals-Erden / und behendiger flüchtigen Erden / einer feuchten schwierichten wässeri-
gen Olität / die vermenget wird mit der allersubtilesten rothschwefelichten gekochten
Erden / mit der allerschwächesten gemachsamen Verbindung / als eine ohnzeitige an-
genehme Frucht aller besonder Metallen.” Cf. J.J. Becher, Parnassi illustrati pars tertia,
Mineralogia, das ist: Deß erläuterten medicinalischen Parnassi dritter eil, nemlich das
Berg-Buch (Ulm, 1663), 40 f.
podium, and notably the lack of hands and feet) refer to chemical
processes, which are discussed in the accompanying text. In fact,
the entire image can be understood as fixed mercury. Curiously
enough, figure 17 displays fewer pictorial elements than those indi-
cated in the text (which mentions, for example, a village in the
background and an infinity of canes), and is thus the exact coun-
terpart to figure 15 (where the image is pictorially richer than the
text).36
Figure 18 constitutes a clear break with all the images of mercury
we have so far considered. All human-like figures have disappeared
in favor of a circle surrounded by two pentagons. But this image,
too, attempts to describe some properties of mercury, although it
does so in a radically different, geometrical manner. Nicolaas Hart-
soeker, who wished to walk in the footsteps of Descartes’ reduction
of all physical qualities to micro-particles, was convinced that we
could induce the invisible corpuscular structure of matter from the
empirically recorded properties. Wishing to explain how mercury
could dissolve gold, he sought to locate this property in the ‘inti-
mate structure’ or geometrical make-up of the mercury particle.
Being of spherical shape (and, pace Descartes, also heavy), such a
particle can force its way into the pores of gold molecules and sep-
arate these into ‘parcels’ (which Hartsoeker sometimes calls ‘atoms’).
What figure 18 purportedly shows is the soft amalgam that results
from the intermixture of such gold ‘parcels’ with mercury parti-
cles.37
36)
is translation of the textual ‘infinity’ of canes into a pictorial handful demon-
strates one of the inherent peculiarities of the image-text relation: it shows one of the
inherent limitations of imaging, which must rely on the imaginable and can therefore
not do visual justice to certain concepts. (See on this issue Elkins, e Domain of
Images, 40-1.) at this has implications regarding the difference between conceptual-
ization and imagination became clear, for instance, in the controversy between Leibniz
and Stahl (cf. Gottfried Wilhelm Leibniz, “G.G. Leibnitii animadversiones circa as-
sertiones aliquas eoriæ Medicæ veræ Clar. Stahlii. Cum ejusdem Leibnitii ad Stahl-
ianas observationes Responsionibus,” in Leibniz, Opera omnia, ed. Ludovicus Dutens,
6 vols. [Geneva, 1768], 2-ii:131-61, at 151: “Contendit Responsio actualem cujusli-
bet partis subdivisionem esse supra omnem conceptibilitatem; quia scilicet conceptum
cum imaginatione confundit.”)
37)
Nicolaas Hartsoeker, Conjectures physiques (Paris, 1706), 130 f.
With figure 19, we have arrived in the domain of what are alter-
natively called ‘symbols’, ‘signs’ or ‘characters’. A circle topped by
a semi-circle, with a cross underneath: what had originated as a
sign referring to the planet Mercury had long become also a sym-
bol of the homonymous metal.38 But what is the relation between
‘sign’, ‘symbol’, and ‘image’?39 e question is unanswerable, for it
leads us once more into that swampy frontier area between letters,
words, and images that we have entered from a different direction
earlier on in this article. What is clear is that figure 19 is more
than a simple abbreviation such as the modern ‘Hg’, which stands
for ‘hydrargyrum’, because it can once again become an iconograph-
ical element of a more complicated, combinatorial figure. is is
precisely what happened in figure 16, in which the signs for sub-
limed and precipitated mercury are developed out of the sign for
pure mercury.40 And it is also what happens in figure 20, in which
a symbol of the philosophical stone (monas hieroglyphica) is com-
posed of the signs for gold (the character’s head), silver (its hat, as
it were), and mercury. But lo and behold: the abstract symbols have
been crafted into a human-like figure, so that, much to our bewil-
derment, we find ourselves somehow back in the figurative section
of our little iconography of mercury.
Problem IV: Typological Names of ‘Images’

In problem sections I and II, we have simply spoken of ‘images’,
applying this term indiscriminately to Darwin’s hasty sketch in figure
38)
On the relation between planetary signs and metals, see for instance Diderot and
D’Alembert, Encyclopédie, s.v. “Talisman.”
39)
e mutual relation between ‘sign’ and ‘symbol’ constitutes a battlefield on its own.
e respective entries in D’Alembert and Diderot’s Encyclopédie, for example, strong-
ly overlap. By contrast, Barbara Obrist, La cosmologie médiévale. Textes et images. Vol.
I. Les fondements antiques (Florence, 2004), 301 ff., clearly distinguishes between the
two terms in her discussion of an image from Isidore of Seville in which the seasons
are linked (symbolon = link) by the four qualities and elements.
40)
Cf. Marco Beretta’s account of the combinatorial hopes of new chemical symbol-
ism in the last decades of eighteenth century, in “e Role of Symbolism from Alche-
my to Chemistry,” in Non-Verbal Communication, ed. Mazzolini, 279-319.
5 and to the artistic composition of figure 4, whereas in section

III, we have begun to speak of allegorical, metaphorical and emblem-
atic images as well as of signs and symbols. is leads us inevita-
bly to our next problem, that pertaining to typological names.
Much recent literature has relied on a single, all-inclusive term
with respect to all graphic non-textual appearances. Nelson Good-
man has sought to include all images, artistic and scientific alike,
in his theory of ‘symbols’, while James Elkins uses the term ‘image’
for the “entire field of meaningful marks,” not unlike Gottfried
Boehm, whose term ‘Bild’—whose meaning includes both the ‘image’
and ‘picture’—refers to all static visual manifestations.41 But this
all-encompassing use of ‘image’ is undermined by more specific def-
initions of this word in the technical literature. Jacques Bertin, for
example, defines ‘image’ as “a significant visual form that is per-
ceived in one perceptual instant.”42 Whatever requires more than a
single coup d’œil would thus no longer qualify as an ‘image’. is
definition, though seemingly dictated by the professional needs of
a technical image-maker, happens to coincide with the original sig-
nification of ‘imago’, which, in contradistinction to ‘pictura’, referred
to the mentally graspable components of a given visual manifesta-
tion.43
41)
Nelson Goodman, Languages of Art. An Approach to a General eory of Symbols
(Oxford, 1968); Elkins, Domain of Images; Gottfried Boehm, ed., Was ist ein Bild?
(Munich, 1994); id., Wie Bilder Sinn erzeugen (Berlin, 2008).
42)
Jacques Bertin, Sémiologie graphique, 50: “Forme visuelle significative perçue dans
un instant de perception (dans un seul coup d’œil). C’est l’unité temporelle de per-
ception significative.”
43)
On this, see Marion G. Müller, “What is Visual Communication? Past and Fu-
ture of an Emerging Field of Communication Research,” Journal of the Swiss Associa-
tion of Communication and Media Research 7 (2007), 7-34. Müller distinguishes not
only between the mental and material in words such as ‘imago’ and ‘pictura’, but also
between the dominant meaning of the English term ‘image’ and the dual meaning of
the German ‘Bild’. She recognizes a relation between the divergent meaning of these
terms and the divergent focus of Anglo-Saxon and German research on visuality. e
problem of idiomatic diversity in nomenclature can already be recognized in the ear-
ly modern period. Note the German translation of the Latin title in: Icones mortis sex-
aginta imaginibus, totidemque inscriptionibus insignitæ, versibus quoque latinis et novis
By contrast, historians have been more hesitant in their typol-

ogy. John Murdoch begins by distinguishing between “pictorial and
diagrammatic materials,” which he subsequently divides further into
‘distinct types’ possessing different functional and morphological
properties, such a ‘dichotomies’, ‘rotæ’, or ‘squares of opposition’.44
Willem Hackmann, in turn, speaks with respect to early modern
physics of “four types of textbook illustrations which are distin-
guished by their function,” namely (i) “allegorical titlepages and
frontispieces,” (ii) “illustrations of the actual experimental configu-
rations described in the text,” (iii) “engravings depicting the actual
observed phenomena produced by the instruments,” and (iv) “dia-
grams of the supposed underlying structure of these phenomena.”
A fifth type, “mapping of data in the form of graphs,” is briefly
mentioned.45 Barbara Obrist, finally, in her careful enquiries into
the evolution of alchemical and cosmological imagery in the Mid-
dle Ages, wields an immensely rich vocabulary, which includes, in
alphabetical order, the following terms: ‘allegories’; ‘analogies’; ‘func-
tional’, ‘mnemonic’ and ‘synoptic diagrams’; ‘enigmata’; ‘graphs’,
‘hieroglyphics’; ‘icons’; ‘ideograms’; ‘pictorial metaphors’; ‘cosmolog-
ical’ and ‘geometrical schemata’; and ‘symbolic signs’.46 Obrist defines
a number of these terms, albeit in a partly overlapping manner,
while limiting herself in many other cases to mere descriptions or
to a simple illustration of the meaning of the term by means of
the very image to which it refers.
From this small range of representative examples, it becomes clear
that there is neither a standard vocabulary nor an agreed-on typology
with respect to ‘epistemic images’ past and present. e problem is
aggravated by the fact that most descriptive terms in circulation
don’t coincide with historical terms. In other words, however refined
or precise our own typology might be, it will never match earlier
typologies.
germanicis illustratæ. Vorbildungen deß Todtes in sechtzig Figuren durch alle Stände und
Geschlechte derselbigen nichtige Sterblichkeit fürzuweisen… (Nuremberg, 1648).
44)
Murdoch, Album, x.
45)
Willem D. Hackmann, “Natural Philosophy Textbook Illustrations 1600-1800,”
in Non-verbal Communication, ed. Mazzolini, 169-96, at 170-72.
46)
Obrist, Les débuts, passim; and eadem, La cosmologie médiévale, passim.
is inevitable mismatch is due to three principal factors. e

first is the lack of terminological precision in the historical sources
themselves. ‘Figura’, for example, frequently possessed the same catch-
all quality as our own term ‘image’. Indeed, the phrase sicut hæc
figura docet (“as this figure shows”) is found to refer to the entire
spectrum of mathematical, diagrammatic or figurative ‘images’ that
may accompany a text. Interestingly, this nondescript use of figura
lives on in our modern references to visual text-inserts as ‘figures’
(as is testified by the present essay).
e second factor responsible for the inevitable mismatch between
our tentative terminologies and historical ones is the exact contrary
of the first factor. One encounters, notably in the Renaissance,
authors who draw distinctions between types of images that are very
hard to capture by means of our own terminology. e most extreme
case is possibly Giordano Bruno, who in his treatise On the Com-
position of Images of 1591 distinguished between “1. idea, 2. vesti-
gium, 3. umbra, 4. nota, 5. character, 6. signum, 7. sigillum, 8.
indicium, 9. figura, 10. similitudo, 11. proportio, 12. imago.”47 And
as if this were not enough, Bruno—who may have been the only
natural philosopher to illustrate his work with woodcuts of his own
making—employed even more terms in his other works, including
archetypus, forum, atrium, rota, area, etc.48 Sure enough, it is pos-
sible to translate Bruno’s terms into English, or into any other mod-
ern language. But apart from the fact that his idea is not the same
as our ‘idea’ nor his vestigium the same as our ‘trace’, his (trans-
lated) vocabulary would still not constitute a reliable historical typol-
ogy, but only Bruno’s own, and nobody else’s. e reasons for this
extreme limitation will be explained shortly.
e third complicating factor is a mixture of the previous two.
In the scholastic tradition, it was customary to study the opinions
of previous authors before arriving at one’s own conclusion. A late
sixteenth-century author working on vision, perception, imagination
47)
Giordano Bruno, De imaginum, signorum et idearum compositione, ad omnia inven-
tionum, dispositionum et memoriæ genera libri III (Frankfurt, 1591), 97.
48)
See the beautiful facsimile reprints of Bruno’s woodcuts in Mino Gabriele, Giorda-
no Bruno: Corpus iconographicum. Le incisioni nelle opere a stampa (Milan, 2001).
and understanding was, however, confronted with a truly irrecon-

cilable range of views on how the things perceived ended up in our
mind. Intromissionist and extramissionist theories, the multiplicatio
specierum described by the so-called perspectivist authors, Avicen-
na’s and Averroes’ theories concerning the species intelligibiles and
intentionales, and simulacra theories of atomist extraction, they all
presented themselves as possible modes of description, and each
with its own technical vocabulary.49 e consequences of this situ-
ation can be beautifully witnessed in the following hair-raising descrip-
tion of what it is that light carries to the eye. “Light takes on
nothing else than the imago of things, which, whether you call it
forma, or simulachrum, or idolum, or species, or spectrum, does not
matter, if you understand but that alone, that it represents the
thing.”50 Here we have an author (Fabrizio di Acquapendente in his
treatise De visione, voce, auditu of 1600) who clearly has lost his
way in the forest of possible, but mutually contradictory terms.
Problem V: No Iconography without Epistemology and

Metaphysics
It is precisely because of this unmanageable multiplicity and ambiv-
alence of the available terminology that Giordano Bruno attempted
to clarify his use of terms by clear definitions. But contrary to most
authors, both contemporary to him and to us, Bruno understood
the direct dependence of any iconographical terminology on its phil-
osophical framework, and notably on metaphysical and epistemo-
logical assumptions. After all, each epistemic image is relational. As
49)
On the proliferation of models of perception and intellection in the early modern
period, see notably Leen Spruit, ‘Species intelligibilis’. From Perception to Knowledge, 2
vols. (Leiden, 1994, 1995).
50)
Fabrizio di Acquapendente, De visione, voce, auditu (Venice, 1600): “[lux] non ali-
ud quam rerum imaginem assumit, quam sive formam, sive simulachrum, sive idol-
um, sive speciem aut spectrum appelles, nihil interest, si modo id solum, quod rem
repræsentat, intelligas.” Quoted from Isabelle Pantin, “Simulachrum, species, forma,
imago: What Was Transported by Light into the Camera Obscura? Divergent Concep-
tions of Realism Revealed by Lexical Ambiguities at the Beginning of the Seventeenth
Century,” Early Science and Medicine 13 (2008), 245-69, at 259n45.
Aristotle states in De memoria (450b23-29), it has a double nature,

referring to itself as an image (eikon), and also to something else,
to that which it represents. But that ‘something else’ will possess a
very different status depending not just on the type of ‘image’ one
is dealing with, but also on the type of philosophy within which
it was created.
For someone who assumes, as Aristotle roughly does, that under-
standing ideally consists in grasping the essence and causal embed-
ding of an object or event by the mind’s higher faculties, there won’t
be much need for an additional visualization of natural phenomena.
is explains, as Murdoch puts it, why “the illustrations are noto-
riously few in the manuscript copies of Aristotle and Galen and of
their medieval translations and seemingly endless commentaries.”51
For the typical Aristotelian, the five senses are adequate to the per-
ception of the things that are. is is why he may illustrate rare
things or events, simply because not everyone has seen them, but
will have no need for abstractions or models of things that are well
known. Aristotle, for example, excludes structural models of micro-
levels by means of his Lynceus argument (De generatione et corrup-
tione, 328a13ff), that is, by the argument that the atomists’
philosophy entails the absurd idea that a sharp-eyed Lynceus would
see a different reality than we commonly do, namely one made up
of atomic particles interrupted by little void spaces. On the other
hand, the forceful presence of quaternities in the Aristotelico-Galenic
tradition (four primary qualities, four elements, four humors, four
seasons, four directions, etc.) allowed for the well-known maps of
logical relations and squares of opposites, although these were not
inherently necessary to a comprehension of reality.52
By contrast, if one assumes, as René Descartes does, that the
senses are deceitful and that the world must be explained on the
basis of a hypothetical reconstruction from the most clear and evident
51)
Murdoch, Album, x.
52)
For the Lynceus-argument and its implications for the visualization of matter, see
Christoph Lüthy, “Atomism, Lynceus, and the Fate of Seventeenth-Century Micros-
copy,” Early Science and Medicine 1 (1997), 1-27; on the evolution of squares of op-
posites and maps of quaternities, see Murdoch, Album, chs. 6 and 7; and Obrist, La
cosmologie médiévale, passim.
ideas encountered in the mind, then the metaphysical and episte-

mological framework shifts considerably. Mental representations now
come to take center stage. Indeed, there are moments when Des-
cartes appears to be incapable of thinking and explaining without
the use of images and objects to which these relate. For example,
in one of his conversations with Frans Burman, there is a moment
in which it appears that Descartes,
despite the fact that he has accustomed his mind to imagining, was scarcely able
to conceive of [the effect of vortex motion] without the balls. So others will find
it much more difficult. For these things depend on mathematics and mechanics,
and can be demonstrated better in a visual demonstration than they can in a ver-
bal demonstration.53
Famously, for Descartes, these mental representations ought to coin-

cide with the spatio-material particles that are held to be responsi-
ble for the secondary qualities provided by our senses. Colors are,
for example, only perceived qualities that are provoked by the way
in which the pressure of sunrays affects our organs of sight. e
pressure of these rays can be represented by lines G, which in the
case of figure 21 (which explains the generation of heat) are seen
affecting the globular air particles inhabiting the pores of a collec-
tion of earth particles. Descartes’ own question of “How we may
arrive at knowledge of the shapes and motions of particles that can-
not be perceived by the senses” raises indeed one of the central
epistemological issues of his pictorial physics.54 e answers this
philosopher provided in the course of his life were partially of an
inductive, probabilistic nature; they partly relied on a physiological
translation mechanism (as in fig. 22); and partly on analogies with
imprinting (fig. 23); and in part they yearned for the possibility of
deductive proof. His statement, in a letter to Mersenne, that “I
imagine or rather find [a particular type of particle] by demonstration,”
53)
Descartes’ Conversation with Burman, transl. with introduction by John Cotting-
ham (Oxford, 1976), 67.
54)
René Descartes, Principia philosophiæ (Amsterdam, 1644), part IV, marginal title
of §203.
Figures 21-24: Cartesian images and their self-justification. Fig. 21: Sunrays G heat-
ing up a body (composed of the hatched particles) by affecting the air globules within
it and bringing AB to shake (Principia philosophiæ [Amsterdam, 1644], 204). Fig. 22:
Optical fibres being affected by a star, the nerve end of fibre no. 1 being affected most
strongly, those of no. 2 less strongly, etc., thereby translating a pressure pattern into
visual patterns (La dioptrique, in Discours de la méthode [Leiden, 1637], 68). Fig. 23:
Comparing memorization to the imprinting of a pattern on a linen cloth; the more
often you imprint the pattern, the clearer it will be recognizable on the cloth (De
l’homme [Paris, 1664], 75). Fig. 24: Looking at ‘looking’ without seeing ‘seeing’ (La
dioptrique, Discours 5, 36).
is a telling indication of the multiple origins of his corpuscular

imagery.55
Descartes’ philosophy allows for the perplexing situation in which
we can thus look at the eye (fig. 24), understand its function in
terms of a camera obscura, and yet at the same time know that this
looking at the eye doesn’t tell us anything about seeing as such, let
alone about the truth of what we see. is truth can only be estab-
lished by an inner eye whose relation to the outer eye remains to
this day a question of controversy.56
It is obvious that Descartes’ images can only be understood within
their unique philosophical context. And the same will hold true for
any other maker of ‘epistemic images’. It must, for example, be evi-
dent that neither an Aristotelian nor a Cartesian framework will
allow us to understand images fabricated in that Neoplatonist-
Hermetic tradition that developed notably in the wake of Marsilio
Ficino, and for which certain ‘images’ (imagines) possessed magic
powers. To understand such images, one must first look to chap-
ters 15 to 20 of Ficino’s De vita cœlitus comparanda (finished by
1489).57 ere, we will encounter a cosmos in which celestial bod-
ies are alive, just as their rays are “living and perceiving,” carrying
“marvelous gifts from the imaginations and minds of the celestials.”
ere, we will also hear why ‘images’ engraved on stones under the
right astrological circumstances will bundle, as it were, and reinforce
55)
Descartes to Mersenne, 9 January 1638, in Œuvres de Descartes, eds. Charles Adam
and Paul Tannery, 12 vols. (Paris, 1897-1957), 2:483.
56)
See on the issue of Descartes’ images Klaus Zittel, “Abbilden und Überzeugen bei
Descartes,” in Cognition and the Book: Typologies of Formal Organisation of Knowledge
in the Printed Book of the Early Modern Period, eds. Karl A.E. Enenkel and Wolfgang
Neuber (Leiden, 2005), 535-601; Christoph Lüthy, “Where Logical Necessity Be-
comes Visual Persuasion: Descartes’s Clear and Distinct Illustrations,” in Transmitting
Knowledge. Words, Images, and Instruments in Early Modern Europe, eds. Sachiko Kusu-
kawa and Ian Maclean (Oxford, 2006), 97-133.
57)
Ficino, in turn, could rely on medieval traditions of magical practices; see Nicolas
Weill-Parot, Les “images astrologiques” au Moyen Âge et à la Renaissance: Spéculations in-
tellectuelles et pratiques magiques (XII e-XV e siècle) (Paris, 2002); Richard A. Kieckhefer,
Magic in the Middle Ages (Cambridge, 2000); and Michael Camille, “Visual Art in
Two Manuscripts of the Ars Notoria,” in Conjuring Spirits. Texts and Traditions of Me-
dieval Ritual Magic, ed. Claire Fanger (Stroud, 1998), 110-39.
Figure 25: A magically efficacious ‘image’ of Saturn following Ficino (from Giordano
Bruno, De umbris idearum [Paris, 1582]).
these gifts—a notion that Ficino took from the Hermetic Picatrix.58
ese engraved ‘images’—which notably in the guise of amulets can
provoke strong curative or deleterious reactions—may carry either
geometrical shapes like a circle or a cross, or else allegorical images
of planets (like that of Saturn as “an old man sitting on a rather
high throne or on a dragon, his head covered with a dark linen
cloth, raising his hands above his head, holding in his hand a sickle
or some fish, and clothed in a dusky robe,” cf. fig. 25).59 eir effi-
cacy is due to the fact that the “figures and numbers” observable
in the skies “have the greatest affinity with the Ideas in Mind, the
Queen of the World”—an affinity that allows for a reverberation
like that observed between two lutes, of which one resonates as the
other plays.60
Obviously, this concept of ‘image’ is incompatible with both Aris-
totle’s or Descartes’. Ficino himself is quite aware of abandoning
the prevalent theory concerning images, and he cites with feigned
approval omas Aquinas’ injunction against the belief in their
efficacy.61 is injunction was based on solid Aristotelian reasoning:
“Natural matter is not in any way disposed towards [a new form]
58)
Marsilio Ficino, ree Books on Life. A Critical Edition and Translation with Intro-
duction and Notes by Carol V. Kaske and John R. Clark (Binghamton, NY, 1989),
323ff. (quote from Ficino, De vita cœlitus comparanda, ch. 16).
59)
Ibid., 335 (quote from Ficino, De vita, ch. 18).
60)
Ibid., 328-29 (quote from Ficino, De vita, ch. 17).
61)
Ibid., 341-2 (quote from Ficino, De vita, ch. 18).
by figures. [...] So the bodies on which these figures are put have
the same readiness to receive the celestial influence as any other
bodies of the same species.”62 is argument is worth reflecting on
in our present context, because it states that the notion of an effi-
cacious figura makes no sense in a philosophy that attributes effi-
cacy to forma, and in which forma has nothing to do whatsoever
with figura, because it designates a thing’s essential nature, not its
shape! omas Aquinas’ opposition of forma to figura is important
because of that gradual and drawn-out transformation, so exqui-
sitely documented by Norma Emerton, of the concept of forma into
that of figura in the seventeenth century—a transformation that was
to culminate precisely in Descartes’ corpuscular images of which we
have just spoken.63
By using one term only, namely imago, to denote things as dif-
ferent as the constellations of stars, the natural conformation of
stones as well as the geometrical and emblematic shapes and fig-
ures that we might engrave on them, Ficino deliberately or other-
wise rendered the connection between these various phenomena
both close and obscure. With his meticulous list of terms, Gior-
dano Bruno, who stood very much in the Ficinian tradition, attempted
to bring clarity into the relations that “images, signs and ideas” held
among themselves and vis-à-vis the world of objects. If we return
to Bruno once more at the end of this section, it is because he was
more aware than anyone at the time that one’s iconographic taxon-
omy depended on one’s metaphysics and epistemology. With meth-
odological acumen, he opened his On the Composition of Images,
Signs and Ideas (1591) with a characterization of his philosophical
system on which the relational functions of his various types of
images depended. His text begins with a Neoplatonic distinction
between the three worlds. ere is, first of all, the divine world
(mundus metaphysicus) of ideas (ideæ). It is responsible for the ori-
gin of the mundus physicus, the created, natural world, which con-
tains the traces (vestigia) of the divine ideas. e third world
62)
omas Aquinas, Summa contra Gentiles, 105, in Opera omnia, Leonine edition
(Rome, 1882-1971), 331a.12-17; cited in Ficino, ree Books, ed. Kaske and Clark, 445.
63)
Norma Emerton, e Scientific Reinterpretation of Form (Ithaca, NY, 1984).
(mundus rationalis) comes to exist in our soul, which attempts, by

conceptual and logical means and by abstracting from the vestigia,
to fathom the ideæ, of which it can, however, only capture the shad-
ows (umbræ). With respect to the things in the natural world, “the
ideas are the causes of things before these things; the vestiges of
the ideas are the things themselves or in the things; and the shad-
ows of the ideas are from these things or after them.”64 Our minds
are thus derivative, “like some living mirror, in which there is the
image (imago) of the natural things and the shadow (umbra) of the
divine ones.”65 But whereas a mirror might only receive the images
and figures (imagines atque figuræ) of things, Bruno is convinced
that the human mind is able not only to recognize the causally
active substantial forms (formæ) and the species (species) of the things
perceived, but is able to process these images in such a way as to
render them useful both in theory (e.g., in the intellection of the
divine unity) and in practice (e.g., in the ars inveniendi). Having
received the imagines of the natural world, the mind is able to con-
struct more meaningful images, signs and ideas, which, however,
need to be carefully distinguished. It is at this point that Bruno
introduces his classification into the twelve types that we have cited
above (p. 423).
In circumstances where the visible world is itself considered an
image, however faint or distorted, of a divine world, of which in
turn the mind, based on the visual perception of physical reality,
attempts to reconstruct the ideal idea and image, images proliferate,
crisscross, and threaten to become second- or third-order mirror
images of one another. is is why one must be careful to define
each type in terms of its relation to the perceived world, on the
one hand, and to the ideal world, on the other. Bruno therefore
64)
Bruno, De imaginum, signorum et idearum compositione, 1-2: “Ideæ sunt causa re-
rum ante res, idearum vestigia sunt ipsæ res seu quæ in rebus, idearum umbræ sunt ab
ipsis rebus seu post res …” Because of its oftentimes anachronistic translations, we are
not relying here on Charles Doria’s translation in Bruno, On the Composition of Imag-
es, Signs and Ideas (New York, 1991).
65)
Bruno, De imaginum compositione, 3: “veluti speculum quoddam vivens, in quo est
imago rerum naturalium et umbra divinarum.”
dedicates an entire little chapter to the “definitions of various [visual]

denotations” (variorum notaminum rationes). Let us paraphrase his
definitions in order to document the strong dependence of the
vocabulary on the general philosophical parameters that constitutes
the theme of our essay.66
Nota: anything that by direct or indirect reasoning manages to demonstrate some-

thing.
Character: organized lines or dots signifying something else, e.g., elements.
Signum: a generic term denoting anything capable of signifying, either qua idea
(idea), trace (vestigium), or shadow (umbra).
Sigillum: a diminutive form of the signum, which represents an element or a con-
tracted form of the signum, as when we represent ‘man’ by a hand or a head.
Indicium: something the function of which is not to represent or to signify, but
to show, just as the index finger doesn’t express in itself that to which it points,
but simply invites us to look at the item to which it points.
Figura: whereas the previous types of notamina can refer to both the inside and
the outside of things, the figura only refers to the outside. (Pace Ficino, Bruno thus
returns to the Aristotelian usage of the term figura.) Furthermore, a figura must
enclose a space, whereas the above-mentioned types of images needn’t do that.
Similitudo: in contradistinction to all of the above, each similitude must be of the
same type as that which it represents; as a painting (pictura), a statue or the spe-
cies that is captured by the sense of vision and stored in the phantasia.
Proportio: designates relations between more than two things (unlike similitudo,
which is limited to two). Mathematical relations like 2:4 = 4:8 belong to this
type.
Imago: possesses “more energy, emphasis, and universality” than the above, because
it univocally links two things of the same genus.
is emphatic definition of imago would be fairly incomprehensi-

ble, were it not for the striking clue that Bruno’s discussion of the
difference between similitude and image contains: “Such as an arti-
fact is said to be similar to some artificer, it is yet not said that it
is to, or in, his image, nor in the proximate genus or in the same
species.”67 Ad eius imaginem—to any of his readers, that must have
evoked Genesis 1,26: “And God said: ‘Let us make man in our
66)
Ibid., 5-6.
67)
Ibid., 6: “Sicut et artificium simile dicitur quodammodo artifici, non tamen ad
eius imaginem vel in eius imagine dicitur, nisi vel in proximo genere vel in eadem sit
specie.”
image, after our likeness.’” (et ait faciamus hominem ad imaginem

et similitudinem nostram). But if we humans are ourselves images
of the Deity, there is no end to the mutual mirroring between the
divine macrocosm and our microcosmic selves.
e Evidence from Clashes

In our five problem-raising sections, and by means of a number of
examples, we have attempted to document the lack of precision and
the fluidity of the historical vocabulary. We have seen, for exam-
ple, that neither figura nor imago mean the same thing for Aqui-
nas, Ficino and Descartes, and that none of these authors uses the
large iconographical terminology that Bruno proposes. We have also
been at pains to demonstrate that not just the names themselves,
but also the very meaning and status of scientific images depend
on the philosophical framework within which they are employed.
But if the meaning of images was really as historically contin-
gent as has just been suggested, should one then not be able to
find, in the historical sources, disputes about the standing of images?
Yes, one should; and in fact, one does. ere are numerous such
disputes, although they seldom take central stage. In fact, they always
accompany clashes over incompatible scientific and philosophical
theories. is is why the history of theoretical changes in any given
discipline can profitably be approached through a study of its visual
conventions and the disputes that erupted around them.68
Let us mention just a small number of such disputes. Marsilio
Ficino, for example, had to defend himself against those who ac-
cused him of an idolatrous usage of images: “Marsilio is a priest,
isn’t he? … What does a Christian have to do with magic or images?”69
More detailed and fascinating is that famous dispute between Robert
68)
One of the co-authors of this article, Smets, is currently working on a doctoral dis-
sertation devoted to the evolution of images expressing chemical matter theories from
the Renaissance to the eighteenth century. In this project, the clash between rivaling
modes of visualization is a recurrent theme.
69)
is quote from Marsilio Ficino’s Apologia quædam is taken from Kaske and Clark’s
translation in Ficino, ree Books, 395.
Fludd and Johannes Kepler. Fludd had attempted to demonstrate

the truth of his macro-microcosmic philosophy by means of very
elaborate and expensive copper plates (notably in his multi-volume
Utriusque cosmi … metaphysica, physica atque technica historia of
1617-21). As Robert Westman has rightly stressed, these engravings
must not be viewed “as illustrations but rather as ways of knowing,
demonstrating, and remembering.”70 In fact, Fludd believed that his
expensive prints, though emblematic in nature, possessed demon-
strative force. His famous foldout depicting the macrocosm, for
example, was meant to be an “Emblematic mirror demonstrating”
how “the more liberal arts” (artes liberaliores), which significantly
enough also included the art of painting, led to knowledge of God,
demons, and the creation.71 But in Kepler’s eyes, Fludd merely
indulged in “pictures forged from air,” and he pitted his own dia-
grammata, which he took to provide geometrical proof, against Fludd’s
airy picturæ, figuræ, and hieroglyphica.72 His disagreement with Fludd’s
images was an important, though integral, part of a more compre-
hensive disagreement about the nature of numbers, the meaning of
symmetries, and the structure of the universe.73
Every time that images express world-views, as is clearly the case
with Kepler and Fludd, the charge is inevitably brought up that a
given image is not understood because the theory informing it is
not understood, and vice versa. In this vein, omas Browne, when
chiding certain implausible depictions of animals, was in turn accused
of not understanding their theoretical underpinnings: “e doctor
70)
Robert S. Westman: “Nature, Art, and Psyche: Jung, Pauli, and the Kepler-Fludd
Polemic,” in Occult and Scientific Mentalities in the Renaissance, ed. Brian Vickers
(Cambridge, 1984), 177-229, at 181, original emphases.
71)
Ibid.
72)
See, for example, Johannes Kepler’s statement: “Tuis picturis mea comparavi dia-
grammata; fassus librum meum non æque atque tuum ornatum esse, nec futurum ad
gustum lectoris cuiuslibet: excusavi hunc defectum a professione, cum ego mathe-
maticam agam.” (“Apologia,” Gesammelte Werke, vol. 6, ed. Max Caspar [Munich,
1940], 396).
73)
On the Fludd-Kepler controversy and its implications for iconography, see besides
Westman, “Nature, Art, and Psyche,” also Frances A. Yates, Giordano Bruno and the
Hermetic Tradition (Chicago, 1964), ch. 22, and Judith V. Field, “Kepler’s Rejection of
Numerology,” in Vickers, Occult and Scientific Mentalities, 273-96.
quarrels with some pictures ... but for this he hath not great rea-
son; for [the author of the animal images] tels him that this and
divers other pictures are rather Hieroglyphical and Emblematic then
truly historicall.”74 Or take, as a further example, Andreas Libavius’
displeasure, in 1615, with various schemata, picturæ, tabulæ and icones
found in Heinrich Khunrath’s lavishly illustrated Amphitheatrum
sapientiæ eternæ. Libavius took particular exception to one specific
but recurrent iconographical element, namely the Latin word “Omnia”
descending from the heavens. Like all “Panspermic doctors,” so
Libavius charged, Khunrath with this image willfully suggested a
conflation of Anaxagoras’ “all is in all” (ἐν παντί πάντα) with the
identical “Omnia in omnibus” of 1 Corinthians 12,6. In Libavius’
eyes, no biblical source could legitimately be used for the pagan
notion that the seeds of all things are in everything—a notion that
for him carried obvious magical and talismanic implications.75
A final example: an entire book could be written about the very
diverse reactions to Descartes’ innovative corpuscular images. We
encounter one type of critic that only expressed bewilderment at
the logic informing one particular image. is was the case of Henry
More, who politely enquired why the particulæ striatæ responsible
for magnetism managed to maintain their nice screw structure and
74)
Alexander Ross, Arcana Microcosmi (London, 1652), 156. e quote is from Kevin
Killeen, “‘e Doctor Quarrels with some Pictures’: Exegesis and Animals in omas
Browne’s Pseudodoxia Epidemica,” Early Science and Medicine 12 (2007), 1-27, at 6.
75)
Andreas Libavius, De philosophia vivente seu vitali Paracelsi iuxta Petrum Severinum
Danum ex repetitione I. Hartmanni chymiatri Marburgensis (Frankfurt, 1615), Examen,
101: “Henricus Kunrath amphitheatro suo præfixit schema, cuius apex triangularis
triplicatum nomen Dei tetragrammaton habet Magica forma scriptum intra nubem
ex qua manus porrecta coronam capiti Kunrathi imponit, cum inscriptione: OMNIA.
In pictura graduum septem portæ amphitheatri profani, item iubentur procul abesse,
subscribiturque: omnia in omnibus. In tabula sequente quæ est globus cum decalogo
nominibus Dei, & Icone Christi repetitur triangulus Tetragrammati cum subscrip-
tione OMNIA.” I owe this quote to Peter J. Forshaw; cf., by Forshaw, “Paradoxes, Ab-
surdities, and Madness: Conflict over Alchemy, Magic and Medicine in the Works
of Andreas Libavius and Heinrich Khunrath, Early Science and Medicine 13 (2008),
53-81.
did not break up, as did other particles described by Descartes.76

But there were also those, like Henricus Regius in his Fundamenta
physices of 1646, who reproduced Descartes’ corpuscular physics
with all its imagery, while rejecting the very metaphysics that, in
Descartes’ eyes, provided these images with their legitimacy. ere
were yet others who did not understand that Descartes’ particles
should be viewed with the inner eye only, and sought for them
with the microscope. Nathaniel Highmore even announced in 1651
that magnetic effluvia “by the help of Glasses” had been observed—
those very particulæ striatæ that More had found so difficult to
imagine!77 And finally, there were those who laughed about the cor-
puscular imagery, such as the Newtonian John Keill, who ridiculed
the allegedly ‘mechanical’ philosophers’ picturesque explanations of
natural phenomena by means of “Figures, Ways, Pores and Inter-
stices of Corpuscles, which they never saw.”78
e question of the legitimacy of certain types of images is explic-
itly raised in all of these debates. What is it that can possibly endow
a given epistemic image with any sort of power? And what power
would that be? Is it a magical power, a power of proof or one of
persuasion? And does the picture derive its alleged power from the
text (of which it offers ocular proof, demonstration, or elucidation),
or does the text instead derive its force from the visual proof in
the way in which in Euclidean geometry, the proof is only com-
pleted when the figure is fully drawn? Quod erat demonstrandum—
but “to be proven” verbally or pictorially, or both?
Only in exceptional, but all the more fascinating cases does nature
draw her own images. In the world of diagrammatic representations,
one such case is represented by the near-parabolic line generated
by an ink-dripping ball that Guidobaldo del Monte let roll over on
an inclined sheet. e other case, which belongs to the opposite
extreme, namely to the realm of figurative realism, is represented
76)
Henry More to Descartes, 5 March 1649, Œuvres de Descartes, eds. Adam and Tan-
nery, V: 346-47.
77)
Nathaniel Highmore, e History of Generation, Examining the Several Opinions of
Divers Authors, (London, 1651), 117.
78)
John Keill, An Introduction to Natural Philosophy: or, Philosophical Lectures Read in
the University of Oxford, Anno Dom. 1700 (London, 1726), iii.
by Johannes Kepler’s camera obscura drawings, which were simply

the retraced outlines of objects projected by a lens onto a screen.
Importantly, his experiments with the camera obscura forced Kepler
to change his iconographical vocabulary. He concluded that the
image appearing on the screen and a fortiori on the camera-like
human retina had to be called pictura, whereas what appeared in
our perception had to be called imago. Although his new nomen-
clature did not persuade others, this is probably the first case in
which ‘machine-generated’ images led to a marked shift in the con-
ception of what it is that an ‘image’ does.79
Conclusion
Generations of scientists have quarreled over the exact definition of
such key concepts as ‘force’, ‘vacuum’, ‘organism’, ‘attraction’, or
‘entropy’. Without an exact definition, no exact science seemed pos-
sible. e contrast with the domain of epistemic imagery could not
be any more extreme: there, utter chaos reigns. e iconographical
and semiotic categories available to us are underdeveloped and ill
defined. e term ‘diagram’, for example, is applied to everything
ranging from a construction plan of an airplane to a statistical dis-
tribution curve. e word ‘model’ fares no better, being used for a
miniature train as much as for a hypothetical description of quan-
tum mechanical interactions.
With respect to historical modes of visualization, the termino-
logical situation is probably even worse. Admittedly, there are fewer
types of epistemic images to be reckoned with. After all, the com-
plicating factor of machine-generated images is absent from the
medieval and early modern landscape, if we disregard the exceptions
just mentioned. By recompense, the epistemological and metaphysical
premises within which a given image functioned, the practices that
generated it, and the meanings and allusions that it could carry, are
difficult to fathom and, to make things worse, appear to be of lit-
tle interest to most scholars. e bizarre asymmetry with which
79)
See on this the collection of essays in Kepler, Optical Imagery, and the Camera Ob-
scura, ed. Alan E. Shapiro (= Early Science and Medicine 13.3/2008).
editors tend to treat historical texts and images is indicative of this

situation. While text variants are obsessively registered in multiple
footnote systems, the images accompanying these very same texts
are either entirely suppressed or are ‘updated’ to be ‘legible to the
modern reader’, as it were. Obviously, if an editor applied the same
procedure to the text itself, he or she would at once be sacked.80
One reason for this recklesness is this behavior of the historical
actors themselves. As we have seen, most ‘producers’ and ‘users’ left
the status of their figuræ undefined and vague. is situation must
give pause for reflection. For it suggests that, like other parapher-
nalia of scientific practice, images were routinely regarded as instru-
ments and as subsidiary to scientific theory itself, and as such as
unworthy of scientific attention. Yet, like all instruments, images
create scientific entities. As a vacuum pump creates the void of
which it is at the same time the principal scientific investigator, so
a taxonomic tree is the mental instrument that guides research, for
example, into missing branches and links. And, incidentally, just as
the vacuum pump could become one of the visual emblems of the
new science, so the taxonomic tree could become the icon of an
evolutionary worldview.81
It is important to realize that it is always in the years that a new
type of epistemic image is being introduced that the awareness of
its status, function and role within a given theory are explicitly dis-
cussed. In case the new type is accepted and becomes embedded in
a shared scientific paradigm, it will inevitably become—not unlike
language—an integral part of the scientific practice, and the awareness
of its specific philosophical premises will disappear. is phenomenon
can be observed across the centuries, and it is as true of the quaternary
patterns of interwoven primary qualities, elements, humors, seasons,
etc., which characterized natural philosophy and medicine until the
seventeenth century, as it is true of twentieth-century Feynman
80)
Cf. Murdoch, Album, 113: “variations in diagrams are often every bit as important
as variations in the text. ey, too, are part of the transmission [of a given work] and
as such must be considered fully in the editing of such works, a desideratum that has,
until recently, been too often ignored.”
81)
On the vacuum pump becoming the emblem of the new science, see Hackmann,
“Natural Philosophy Textbook Illustrations,” 178.
diagrams. In both cases, visual structures can be seen to have con-

stituted the unquestioned starting point of scientific theory forma-
tion and practice.82
is often entirely unquestioned use of visualizing tools in the
sciences puts the historian of epistemic imagery in a complex situ-
ation. In those cases where new types of imagery were created to
express new theories, and in which both theory and images were
explicitly discussed or contested, the historian is in a comfortable
situation. But in that large majority of cases where types of images
figure as the unquestioned visual backbone of ‘normal science’, the
historian’s wish to decode these images will in all likelihood exceed
the level of awareness and the intentions of their makers.
Fortunately enough, historians of epistemic imagery find them-
selves, in this respect, in the same position as historians of science
and intellectual historians, who are accustomed to dealing with the
phenomenon of the internalization of concepts, ideas, and practices.
An understanding of analogous processes is thus present. What is
still lacking, however, is a philosophical history of epistemic images,
which attempts to draw up, for each style of thought, an analytic
taxonomy that reflects the actors’—the authors’ and the artists’—
understanding of the epistemological value and the functionality of
the images that they produced, and within which they and their
disciples ended up thinking.
82)
See Gernot and Hartmut Böhme, Feuer, Wasser, Erde, Luft. Eine Kulturgeschichte
der Elemente (Munich, 1996); David Kaiser, Drawing eories Apart: e Dispersion of
Feynman Diagrams in Postwar Physics (Chicago, 2005).

07 Luthy & Smets - Towards A History of Scientific Imagery

Uploaded by

Copyright:

Available Formats

07 Luthy & Smets - Towards A History of Scientific Imagery

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

07 Luthy & Smets - Towards A History of Scientific Imagery

Uploaded by

Copyright:

Available Formats

Early Science and Medicine 14 (2009) 398-439 www.brill.

Words, Lines, Diagrams, Images:

Christoph Lüthy & Alexis Smets*

To John E. Murdoch, eminent iconographer

* Faculty of Philosophy, Radboud University Nijmegen, P.O. Box 9103, NL—6500

© Koninklijke Brill NV, Leiden, 2009 DOI : 10.1163/157338209X425632

Underlying much of this proliﬁc literature, however, one ﬁnds

ch. 2: “Models: Representing ings”). In a similar direction go recent attempts to

attached (by means of diagrams, geometrical proof, sciagraphiæ,

Problem I: e Unclear Boundary between Words and Images

In his ground-breaking e Domain of Images, James Elkins chal-

for this ﬂuid transition from writing to imaging massively complicates

Problem II: Similar Shape, Diﬀerent Meaning

silver atoms represented by the dark disks in ﬁgure 11 share no

Although the ‘graph’ is, iconographically speaking, an indirect

belonging to many diﬀerent conceptual (rather than perceptual)

Problem III: Identical Signiﬁers, Diﬀerent Types of

as manifested both in diﬀerent aggregate states and during opera-

Figures 13 to 20: Diverse representations of mercury.

In ﬁgure 14, we are once more confronted with a representation of

various problems, as contemporary studies on metaphorical and alle-

personiﬁcation, ﬁgure 15 may indeed have illustrated the limited

these ‘symbols’, Becher distinguishes between the volatile and ﬁxed

Problem IV: Typological Names of ‘Images’

5 and to the artistic composition of ﬁgure 4, whereas in section

By contrast, historians have been more hesitant in their typol-

is inevitable mismatch is due to three principal factors. e

and understanding was, however, confronted with a truly irrecon-

Problem V: No Iconography without Epistemology and

Aristotle states in De memoria (450b23-29), it has a double nature,

ideas encountered in the mind, then the metaphysical and episte-

Famously, for Descartes, these mental representations ought to coin-

is a telling indication of the multiple origins of his corpuscular

(mundus rationalis) comes to exist in our soul, which attempts, by

dedicates an entire little chapter to the “deﬁnitions of various [visual]

Nota: anything that by direct or indirect reasoning manages to demonstrate some-

is emphatic deﬁnition of imago would be fairly incomprehensi-

image, after our likeness.’” (et ait faciamus hominem ad imaginem

e Evidence from Clashes

Fludd and Johannes Kepler. Fludd had attempted to demonstrate

did not break up, as did other particles described by Descartes.76

by Johannes Kepler’s camera obscura drawings, which were simply

editors tend to treat historical texts and images is indicative of this

diagrams. In both cases, visual structures can be seen to have con-

You might also like