Erwin Panofsky 1939

Iconography and Iconology:

An Introduction to the Study
of Renaissance Art

i Iconography is that branch of the history of art which concerns itself with
the subject matter or meaning of works of art, as opposed to their form. Let
us, then, try to define the distinction between subject matter or meaning on
the one hand, and form on the other.
When an acquaintance greets me on the street by lifting his hat, what I see
from a formal point of view is nothing but the change of certain details within
a configuration forming part of the general pattern of color, lines and volumes
which constitutes my world of vision. When I identify, as I automatically do,
this configuration as an object (gentleman), and the change of detail as an
event (hat-lifting), I have already over-stepped the limits of purely formal
perception and entered a first sphere of subject matter or meaning. The mean-
ing thus perceived is of an elementary and easily understandable nature, and
we shall call it the factual meaning; it is apprehended by simply identifying
certain visible forms with certain objects known to me from practical experi-
ence, and by identifying the change in their relations with certain actions or
events.
Now the objects and events thus identified will naturally produce a certain
reaction within myself. From the way my acquaintance performs his action I
may be able to sense whether he is in a good or bad humor, and whether his
feelings towards me are indifferent, friendly or hostile. These psychological
Copyright © 2009. Oxford University Press. All rights reserved.

nuances will invest the gestures of my acquaintance with a further meaning

which we shall call expressional. It differs from the factual one in that it is
apprehended, not by simple identification, but by ‘empathy.’ To understand
it, I need a certain sensitivity, but this sensitivity is still part of my practical
experience, that is, of my everyday familiarity with objects and events. There-
fore both the factual and the expressional meaning may be classified together:
they constitute the class of primary or natural meanings.
However, my realization that the lifting of the hat stands for a greeting
belongs in an altogether different realm of interpretation. This form of salute
is peculiar to the Western world and is a residue of mediaeval chivalry: armed
men used to remove their helmets to make clear their peaceful intentions and
their confidence in the peaceful intentions of others. Neither an Australian
bushman nor an ancient Greek could be expected to realize that the lifting
of a hat is not only a practical event with certain expressional connotations,
but also a sign of politeness. To understand this significance of the gentle-
man’s action I must not only be familiar with the practical world of objects

220 mechanisms of meaning

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 and events, but also with the more-than-practical world of customs and cul-
tural traditions peculiar to a certain civilization. Conversely, my acquaintance
could not feel impelled to greet me by lifting his hat were he not conscious
of the significance of this act. As for the expressional connotations which
accompany his action, he may or may not be conscious of them. Therefore,
when I interpret the lifting of a hat as a polite greeting, I recognize in it a
meaning which may be called secondary or conventional; it differs from the
primary or natural one in that it is intelligible instead of being sensible, and
in that it has been consciously imparted to the practical action by which it is
conveyed.
And finally: besides constituting a natural event in space and time, besides
naturally indicating moods or feelings, besides conveying a conventional
greeting, the action of my acquaintance can reveal to an experienced observer
all that goes to make up his ‘personality.’This personality is conditioned by his
being a man of the twentieth century, by his national, social and educational
background, by the previous history of his life and by his present surround-
ings; but it is also distinguished by an individual manner of viewing things
and reacting to the world which, if rationalized, would have to be called a
philosophy. In the isolated action of a polite greeting all these factors do not
manifest themselves comprehensively, but nevertheless symptomatically. We
could not construct a mental portrait of the man on the basis of this single
action, but only by co-ordinating a large number of similar observations and
by interpreting them in connection with our general information as to his
period, nationality, class, intellectual traditions and so forth. Yet all the quali-
ties which this mental portrait would show explicitly are implicitly inherent
in every single action; so that, conversely, every single action can be inter-
preted in the light of those qualities.
The meaning thus discovered may be called the intrinsic meaning or con-
tent; it is essential where the two other kinds of meaning, the primary or nat-
ural and the secondary or conventional, are phenomenal. It may be defined as
a unifying principle which underlies and explains both the visible event and
its intelligible significance, and which determines even the form in which the
visible event takes shape. This intrinsic meaning or content is, normally, as
Copyright © 2009. Oxford University Press. All rights reserved.

much above the sphere of conscious volition as the expressional meaning is

beneath this sphere.
Transferring the results of this analysis from everyday life to a work of art,
we can distinguish in its subject matter or meaning the same three strata:
1. Primary or natural subject matter, subdivided into factual and expressional.
It is apprehended by identifying pure forms, that is: certain configurations
of line and color, or certain peculiarly shaped lumps of bronze or stone, as
representations of natural objects such as human beings, animals, plants,
houses, tools and so forth; by identifying their mutual relations as events; and
by perceiving such expressional qualities as the mournful character of a pose
or gesture, or the homelike and peaceful atmosphere of an interior. The world
of pure forms thus recognized as carriers of primary or natural meanings may
be called the world of artistic motifs. An enumeration of these motifs would
be a pre-iconographical description of the work of art.
2. Secondary or conventional subject matter. It is apprehended by realizing
that a male figure with a knife represents St. Bartholomew, that a female

erwin panofsky 221

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 figure with a peach in her hand is a personification of veracity, that a group
of figures seated at a dinner table in a certain arrangement and in certain
poses represents the Last Supper, or that two figures fighting each other in a
certain manner represent the Combat of Vice and Virtue. In doing this we
connect artistic motifs and combinations of artistic motifs (compositions)
with themes or concepts. Motifs thus recognized as carriers of a secondary
or conventional meaning may be called images, and combinations of images
are what the ancient theorists of art called invenzioni; we are wont to call
them stories and allegories.1 The identification of such images, stories and
allegories is the domain of what is normally referred to as ‘iconography.’ In
fact, when we loosely speak of ‘subject matter as opposed to form,’ we chiefly
mean the sphere of secondary or conventional subject matter, viz., the world
of specific themes or concepts manifested in images, stories and allegories,
as opposed to the sphere of primary or natural subject matter manifested
in artistic motifs. ‘Formal analysis’ in Wölfflin’s sense is largely an analysis
of motifs and combinations of motifs (compositions); for a formal analysis
in the strict sense of the word would even have to avoid such expressions
as ‘man,’ ‘horse,’ or ‘column,’ let alone such evaluations as ‘the ugly triangle
between the legs of Michelangelo’s David’ or ‘the admirable clarification of
the joints in a human body.’ It is obvious that a correct iconographical analysis
presupposes a correct identification of the motifs. If the knife that enables us
to identify a St. Bartholomew is not a knife but a corkscrew, the figure is not
a St. Bartholomew. Furthermore, it is important to note that the statement
‘this figure is an image of St. Bartholomew’ implies the conscious intention of
the artist to represent St. Bartholomew, while the expressional qualities of the
figure may well be unintentional.
3. Intrinsic meaning or content. It is apprehended by ascertaining those
underlying principles which reveal the basic attitude of a nation, a period, a
class, a religious or philosophical persuasion—qualified by one personality
and condensed into one work. Needless to say, these principles are mani-
fested by, and therefore throw light on, both ‘compositional methods’ and
‘iconographical significance.’ In the fourteenth and fifteenth centuries, for
instance (the earliest examples can be dated around 1300), the traditional
Copyright © 2009. Oxford University Press. All rights reserved.

type of the Nativity with the Virgin Mary reclining in bed or on a couch was
frequently replaced by a new one which shows the Virgin kneeling before the
Child in adoration. From a compositional point of view this change means,
roughly speaking, the substitution of a triangular scheme for a rectangular
one; from an iconographical point of view, it means the introduction of a new
theme to be formulated in writing by such authors as Pseudo-Bonaventure
and St. Bridget. But at the same time it reveals a new emotional attitude
peculiar to the later phases of the Middle Ages. A really exhaustive interpret-
ation of the intrinsic meaning or content might even show that the technical
procedures characteristic of a certain country, period, or artist, for instance
Michelangelo’s preference for sculpture in stone instead of in bronze, or the
peculiar use of hatchings in his drawings, are symptomatic of the same basic
attitude that is discernible in all the other specific qualities of his style. In thus
conceiving of pure forms, motifs, images, stories and allegories as manifest-
ations of underlying principles, we interpret all these elements as what Ernst
Cassirer has called ‘symbolical’ values. As long as we limit ourselves to stat-

222 mechanisms of meaning

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 ing that Leonardo da Vinci’s famous fresco shows a group of thirteen men
around a dinner table, and that this group of men represents the Last Supper,
we deal with the work of art as such, and we interpret its compositional and
iconographical features as its own properties or qualifications. But when we
try to understand it as a document of Leonardo’s personality, or of the civilization
of the Italian High Renaissance, or of a peculiar religious attitude, we deal with
the work of art as a symptom of something else which expresses itself in a countless
variety of other symptoms, and we interpret its compositional and iconographical
features as more particularized evidence of this ‘something else.’ The discovery
and interpretation of these ‘symbolical’ values (which are often unknown to
the artist himself and may even emphatically differ from what he consciously
intended to express) is the object of what we may call ‘iconology’ as opposed to
‘iconography.’
[The suffix ‘graphy’ derives from the Greek verb graphein, ‘to write’; it
implies a purely descriptive, often even statistical, method of procedure.
Iconography is, therefore, a description and classification of images much
as ethnography is a description and classification of human races: it is a lim-
ited and, as it were, ancillary study which informs us as to when and where
specific themes were visualized by which specific motifs. It tells us when and
where the crucified Christ was draped with a loincloth or clad in a long gar-
ment; when and where He was fastened to the Cross with four nails or with
three; how the Virtues and Vices were represented in different centuries and
environments. In doing all this, iconography is an invaluable help for the
establishment of dates, provenance and, occasionally, authenticity; and it
furnishes the necessary basis for all further interpretation. It does not, how-
ever, attempt to work out this interpretation for itself. It collects and classi-
fies the evidence but does not consider itself obliged or entitled to investigate
the genesis and significance of this evidence: the interplay between the vari-
ous ‘types’; the influence of theological, philosophical or political ideas; the
purposes and inclinations of individual artists and patrons; the correlation
between intelligible concepts and the visible form which they assume in
each specific case. In short, iconography considers only a part of all those
elements which enter into the intrinsic content of a work of art and must be
Copyright © 2009. Oxford University Press. All rights reserved.

made explicit if the perception of this content is to become articulate and

communicable.
[It is because of these severe restrictions which common usage, especially
in this country, places upon the term ‘iconography’ that I propose to revive
the good old word ‘iconology’ wherever iconography is taken out of its isola-
tion and integrated with whichever other method, historical, psychological
or critical, we may attempt to use in solving the riddle of the sphinx. For as
the suffix ‘graphy’ denotes something descriptive, so does the suffix ‘logy’—
derived from logos, which means ‘thought’ or ‘reason’—denote something
interpretative. ‘Ethnology,’ for instance, is defined as a ‘science of human races’
by the same Oxford Dictionary that defines ‘ethnography’ as a ‘description of
human races,’ and Webster explicitly warns against a confusion of the two
terms inasmuch as ‘ethnography is properly restricted to the purely descrip-
tive treatment of peoples and races while ethnology denotes their comparative
study.’ So I conceive of iconology as an iconography turned interpretative and
thus becoming an integral part of the study of art instead of being confined to

erwin panofsky 223

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 the role of a preliminary statistical survey. There is, however, admittedly some
danger that iconology will behave, not like ethnology as opposed to ethnog-
raphy, but like astrology as opposed to astrography.]
Iconology, then, is a method of interpretation which arises from synthesis
rather than analysis. And as the correct identification of motifs is the pre-
requisite of their correct iconographical analysis, so is the correct analysis of
images, stories and allegories the prerequisite of their correct iconological
interpretation—unless we deal with works of art in which the whole sphere
of secondary or conventional subject matter is eliminated and a direct transi-
tion from motifs to content is effected, as is the case with European landscape
painting, still life and genre, not to mention ‘non-objective’ art.
Now, how do we achieve ‘correctness’ in operating on these three levels,
pre-iconographical description, iconographical analysis, and iconological
interpretation?
In the case of a pre-iconographical description, which keeps within the
limits of the world of motifs, the matter seems simple enough. The objects
and events whose representation by lines, colors and volumes constitutes the
world of motifs can be identified, as we have seen, on the basis of our practical
experience. Everybody can recognize the shape and behavior of human beings,
animals and plants, and everybody can tell an angry face from a jovial one. It
is, of course, possible that in a given case the range of our personal experience
is not wide enough, for instance when we find ourselves confronted with the
representation of an obsolete or unfamiliar tool, or with the representation of
a plant or animal unknown to us. In such cases we have to widen the range of
our practical experience by consulting a book or an expert; but we do not leave
the sphere of practical experience as such, which informs us, needless to say, as
to what kind of expert to consult.
Yet even in this sphere we encounter a peculiar problem. Setting aside the
fact that the objects, events and expressions depicted in a work of art may be
unrecognizable owing to the incompetence or malice aforethought of the
artist, it is, on principle, impossible to arrive at a correct pre-iconographical
description, or identification of primary subject matter, by indiscriminately
applying our practical experience to the work of art. Our practical experi-
Copyright © 2009. Oxford University Press. All rights reserved.

ence is indispensable, as well as sufficient, as material for a pre-iconographical

description, but it does not guarantee its correctness.
A pre-iconographical description of Rogier van der Weyden’s Three Magi
in the Kaiser Friedrich Museum at Berlin would, of course, have to avoid such
terms as ‘Magi,’‘Infant Jesus,’ etc. But it would have to mention that the appar-
ition of a small child is seen in the sky. How do we know that this child is
meant to be an apparition? That it is surrounded with a halo of golden rays
would not be sufficient proof of this assumption, for similar halos can often be
observed in representations of the Nativity where the Infant Jesus is real. That
the child in Rogier’s picture is meant to be an apparition can only be deduced
from the additional fact that he hovers in mid-air. But how do we know that
he hovers in mid-air? His pose would be no different were he seated on a pil-
low on the ground; in fact, it is highly probable that Rogier used for his paint-
ing a drawing from life of a child seated on a pillow. The only valid reason for
our assumption that the child in the Berlin picture is meant to be an apparition
is the fact that he is depicted in space with no visible means of support.

224 mechanisms of meaning

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 But we can adduce hundreds of representations in which human beings,
animals and inanimate objects seem to hang loose in space in violation
of the law of gravity, without thereby pretending to be apparitions. For
instance, in a miniature in the Gospels of Otto III in the Staatsbibliothek of
Munich, a whole city is represented in the center of an empty space while
the figures taking part in the action stand on solid ground.2 An inexperi-
enced observer may well assume that the town is meant to be suspended in
mid-air by some sort of magic. Yet in this case the lack of support does not
imply a miraculous invalidation of the laws of nature. The city is the real
city of Nain where the resurrection of the youth took place. In a miniature
of around 1000 ‘empty space’ does not count as a real three-dimensional
medium, as it does in a more realistic period, but serves as an abstract, unreal
background. The curious semicircular shape of what should be the base line
of the towers bears witness to the fact that, in the more realistic prototype of
our miniature, the town had been situated on a hilly terrain, but was taken
over into a representation in which space had ceased to be thought of in
terms of perspective realism. Thus, while the unsupported figure in the van
der Weyden picture counts as an apparition, the floating city in the Otto-
nian miniature has no miraculous connotation. These contrasting inter-
pretations are suggested to us by the ‘realistic’ qualities of the painting and
the ‘unrealistic’ qualities of the miniature. But that we grasp these qualities
in the fraction of a second and almost automatically must not induce us to
believe that we could ever give a correct pre-iconographical description of
a work of art without having divined, as it were, its historical ‘locus.’ While
we believe that we are identifying the motifs on the basis of our practical
experience pure and simple, we really are reading ‘what we see’ according to
the manner in which objects and events are expressed by forms under vary-
ing historical conditions. In doing this, we subject our practical experience to
a corrective principle which may be called the history of style.3
Iconographical analysis, dealing with images, stories and allegories instead
of with motifs, presupposes, of course, much more than that familiarity with
objects and events which we acquire by practical experience. It presupposes
a familiarity with specific themes or concepts as transmitted through liter-
Copyright © 2009. Oxford University Press. All rights reserved.

ary sources, whether acquired by purposeful reading or by oral tradition.

Our Australian bushman would be unable to recognize the subject of a Last
Supper; to him, it would only convey the idea of an excited dinner party.
To understand the iconographical meaning of the picture he would have to
familiarize himself with the content of the Gospels. When it comes to rep-
resentations of themes other than Biblical stories or scenes from history and
mythology which happen to be known to the average ‘educated person,’ all of
us are Australian bushmen. In such cases we, too, must try to familiarize our-
selves with what the authors of those representations had read or otherwise
knew. But again, while an acquaintance with specific themes and concepts
transmitted through literary sources is indispensable and sufficient material
for an iconographical analysis, it does not guarantee its correctness. It is just
as impossible for us to give a correct iconographical analysis by indiscrim-
inately applying our literary knowledge to the motifs, as it is for us to give
a correct pre-iconographical description by indiscriminately applying our
practical experience to the forms.

erwin panofsky 225

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 A picture by the Venetian seventeenth-century painter Francesco Maffei,
representing a handsome young woman with a sword in her left hand, and
in her right a charger on which rests the head of a beheaded man, has been
published as a portrayal of Salome with the head of John the Baptist.4 In fact
the Bible states that the head of St. John the Baptist was brought to Salome
on a charger. But what about the sword? Salome did not decapitate St. John
the Baptist with her own hands. Now the Bible tells us about another hand-
some woman in connection with the decapitation of a man, namely Judith. In
this case the situation is exactly reversed. The sword in Maffei’s picture would
be correct because Judith beheaded Holofernes with her own hand, but the
charger would not agree with the Judith theme because the text explicitly
states that the head of Holofernes was put into a sack. Thus we have two
literary sources applicable to our picture with equal right and equal inconsist-
ency. If we should interpret it as a portrayal of Salome the text would account
for the charger, but not for the sword; if we should interpret it as a portrayal
of Judith the text would account for the sword, but not for the charger. We
should be entirely at a loss were we to depend on the literary sources alone.
Fortunately we do not. As we could supplement and correct our practical
experience by inquiring into the manner in which, under varying historical
conditions, objects and events were expressed by forms, viz., into the his-
tory of style, just so can we supplement and correct our knowledge of literary
sources by inquiring into the manner in which, under varying historical con-
ditions, specific themes or concepts were expressed by objects and events, viz.,
into the history of types.
In the case at hand we shall have to ask whether there were, before
Francesco Maffei painted his picture, any unquestionable portrayals of
Judith (unquestionable because they would include, for instance, Judith’s
maid) with unjustifi ed chargers; or any unquestionable portrayals of
Salome (unquestionable because they would include, for instance, Salome’s
parents) with unjustified swords. And lo! while we cannot adduce a single
Salome with a sword, we encounter, in Germany and North Italy, several
sixteenth-century paintings depicting Judith with a charger;5 there was a
‘type’ of ‘Judith with a Charger,’ but there was no ‘type’ of ‘Salome with a
Copyright © 2009. Oxford University Press. All rights reserved.

Sword.’ From this we can safely conclude that Maffei’s picture, too, repre-
sents Judith, and not, as had been assumed, Salome.
We may further ask why artists felt entitled to transfer the motif of the
charger from Salome to Judith, but not the motif of the sword from Judith to
Salome. This question can be answered, again by inquiring into the history of
types, with two reasons. One reason is that the sword was an established and
honorific attribute of Judith, of many martyrs, and of such virtues as Justice,
Fortitude, etc.; thus it could not be transferred with propriety to a lascivious
girl. The other reason is that during the fourteenth and fifteenth centuries the
charger with the head of St. John the Baptist had become an isolated devo-
tional image (Andachtsbild) especially popular in the northern countries and
in North Italy; it had been singled out from a representation of the Salome
story in much the same way as the group of St. John the Evangelist resting
on the bosom of the Lord had come to be singled out from the Last Supper,
or the Virgin in childbed from the Nativity. The existence of this devotional

226 mechanisms of meaning

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 image established a fixed association of ideas between the head of a beheaded
man and a charger, and thus the motif of a charger could more easily be sub-
stituted for the motif of a sack in an image of Judith, than the motif of a sword
could have penetrated into an image of Salome.
Iconological interpretation, finally, requires something more than a famili-
arity with specific themes or concepts as transmitted through literary sources.
When we wish to get hold of those basic principles which underlie the choice
and presentation of motifs, as well as the production and interpretation of
images, stories and allegories, and which give meaning even to the formal
arrangements and technical procedures employed, we cannot hope to find an
individual text which would fit those basic principles as John 13: 21 ff. fits the ico-
nography of the Last Supper.To grasp these principles we need a mental faculty
comparable to that of a diagnostician—a faculty which I cannot describe better
than by the rather discredited term ‘synthetic intuition,’ and which may be bet-
ter developed in a talented layman than in an erudite scholar.
However, the more subjective and irrational this source of interpretation
(for every intuitive approach will be conditioned by the interpreter’s psych-
ology and ‘Weltanschauung’), the more necessary the application of those cor-
rectives and controls which proved indispensable where only iconographical
analysis and pre-iconographical description were concerned. When even our
practical experience and our knowledge of literary sources may mislead us if
indiscriminately applied to works of art, how much more dangerous would it
be to trust our intuition pure and simple! Thus, as our practical experience had
to be corrected by an insight into the manner in which, under varying histori-
cal conditions, objects and events were expressed by forms (history of style);
and as our knowledge of literary sources had to be corrected by an insight into
the manner in which, under varying historical conditions, specific themes and
concepts were expressed by objects and events (history of types); just so, or
even more so, must our synthetic intuition be corrected by an insight into the
manner in which, under varying historical conditions, the general and essen-
tial tendencies of the human mind were expressed by specific themes and
concepts. This means what may be called a history of cultural symptoms—or
‘symbols’ in Ernst Cassirer’s sense—in general. The art historian will have to
Copyright © 2009. Oxford University Press. All rights reserved.

check what he thinks is the intrinsic meaning of the work, or group of works,
to which he devotes his attention, against what he thinks is the intrinsic
meaning of as many other documents of civilization historically related to that
work or group of works, as he can master: of documents bearing witness to
the political, poetical, religious, philosophical, and social tendencies of the
personality, period or country under investigation. Needless to say that, con-
versely, the historian of political life, poetry, religion, philosophy, and social
situations should make analogous use of works of art. It is in the search for
intrinsic meanings or content that the various humanistic disciplines meet on
a common plane instead of serving as handmaidens to each other.
In conclusion: when we wish to express ourselves very strictly (which is of
course not always necessary in our normal talk or writing, where the general
context throws light on the meaning of our words), we have to distinguish
between three strata of subject matter or meaning, the lowest of which is
commonly confused with form, and the second of which is the special prov-
ince of iconography as opposed to iconology. In whichever stratum we move,

erwin panofsky 227

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 object of interpretation act of interpretation

i Primary or natural subject matter— Pre-iconographical description

(A) factual, (B) expressional—constituting (and pseudo-formal analysis).
the world of artistic motifs.
ii Secondary or conventional subject Iconographical analysis.
matter, constituting the world of images,
stories and allegories.
iii Intrinsic meaning or content, constituting Iconological interpretation.
the world of ‘symbolical’ values.

our identifications and interpretations will depend on our subjective equip-

ment, and for this very reason will have to be supplemented and corrected
by an insight into historical processes the sum total of which may be called
tradition.
I have summarized in a synoptical table what I have tried to make clear
thus far. But we must bear in mind that the neatly differentiated categories,
which in this synoptical table seem to indicate three independent spheres of
meaning, refer in reality to aspects of one phenomenon, namely, the work of
art as a whole. So that, in actual work, the methods of approach which here
appear as three unrelated operations of research merge with each other into
one organic and indivisible process.
ii Turning now from the problems of iconography and iconology in gen-
eral to the problems of Renaissance iconography and iconology in particular,
we shall naturally be most interested in that phenomenon from which the
very name of the Renaissance is derived: the rebirth of classical antiquity.
The earlier Italian writers about the history of art, such as Lorenzo
Ghiberti, Leone Battista Alberti, and especially Giorgio Vasari, thought
that classical art was overthrown at the beginning of the Christian era, and
that it did not revive until it served as the foundation of the Renaissance style.
The reasons for this overthrow, as those writers saw it, were the invasions of
barbarous races and the hostility of early Christian priests and scholars.
Copyright © 2009. Oxford University Press. All rights reserved.

In thinking as they did the early writers were both right and wrong. They
were wrong in so far as there had not been a complete break of tradition dur-
ing the Middle Ages. Classical conceptions, literary, philosophical, scientific
and artistic, had survived throughout the centuries, particularly after they had
been deliberately revived under Charlemagne and his followers. The early
writers were, however, right in so far as the general attitude towards antiquity
was fundamentally changed when the Renaissance movement set in.
The Middle Ages were by no means blind to the visual values of classical
art, and they were deeply interested in the intellectual and poetic values of
classical literature. But it is significant that, just at the height of the mediaeval
period (thirteenth and fourteenth centuries), classical motifs were not used
for the representation of classical themes while, conversely, classical themes
were not expressed by classical motifs.
For instance, on the façade of St. Mark’s in Venice can be seen two large
reliefs of equal size, one a Roman work of the third century a.d., the other
executed in Venice almost exactly one thousand years later.6 The motifs are so

228 mechanisms of meaning

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 equipment for interpretation corrective principle
of interpretation
(History of Tradition)

Practical experience (familiarity History of style (insight into the

with objects and events). manner in which, under varying
historical conditions, objects and
events were expressed by forms).
Knowledge of literary sources History of types (insight into the
(familiarity with specific themes manner in which, under varying
and concepts). historical conditions, specific themes
or concepts were expressed by objects
and events).
Synthetic intuition (familiarity with History of cultural symptoms or
the essential tendencies of the human ‘symbols’ in general (insight into the
mind), conditioned by personal manner in which, under varying
psychology and ‘Weltanschauung.’ historical conditions, essential
tendencies of the human mind were
expressed by specific themes and
concepts).

similar that we are forced to suppose that the mediaeval stone carver delib-
erately copied the classical work in order to produce a counterpart of it. But
while the Roman relief represents Hercules carrying the Erymanthean boar
to King Euristheus, the mediaeval master, by substituting billowy drapery
for the lion’s skin, a dragon for the frightened king, and a stag for the boar,
transformed the mythological story into an allegory of salvation. In Italian
and French art of the twelfth and thirteenth centuries we find a great number
of similar cases; viz., direct and deliberate borrowings of classical motifs while
the pagan themes were changed into Christian ones. Suffice it to mention
the most famous specimens of this so-called proto-Renaissance movement:
the sculptures of St. Gilles and Arles; the celebrated Visitation at Rheims
Cathedral, which for a long time was held to be a sixteenth-century work; or
Copyright © 2009. Oxford University Press. All rights reserved.

Nicolo Pisano’s Adoration of the Magi, in which the group of the Virgin Mary
and the Infant Jesus shows the influence of a Phaedra Sarcophagus still pre-
served in the Camposanto at Pisa. Even more frequent, however, than such
direct copies are instances of a continuous and traditional survival of classical
motifs, some of which were used in succession for quite a variety of Christian
images.
As a rule such reinterpretations were facilitated or even suggested by a
certain iconographical affinity, for instance when the figure of Orpheus was
employed for the representation of David, or when the type of Hercules drag-
ging Cerberus out of Hades was used to depict Christ pulling Adam out of
Limbo.7 But there are cases in which the relationship between the classical
prototype and its Christian adaptation is a purely compositional one.
On the other hand, when a Gothic illuminator had to illustrate the story
of Laocoön, Laocoön becomes a wild and bald old man in contemporary
costume who attacks the sacrificial bull with what should be an ax, while the
two little boys float around at the bottom of the picture, and the sea snakes

erwin panofsky 229

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 briskly emerge from a pool of water.8 Aeneas and Dido are shown as a fash-
ionable mediaeval couple playing chess, or may appear as a group resembling
the Prophet Nathan before David, rather than as a classical hero before his
paramour. And Thisbe awaits Pyramus on a Gothic tombstone which bears
the inscription ‘Hic situs est Ninus rex,’ preceded by the usual cross.9
When we ask the reason for this curious separation between classical
motifs invested with a nonclassical meaning, and classical themes expressed
by nonclassical figures in a nonclassical setting, the obvious answer seems to
lie in the difference between representational and textual tradition. The art-
ists who used the motif of a Hercules for an image of Christ, or the motif of
an Atlas for the images of the Evangelists,10 acted under the impression of
visual models which they had before their eyes, whether they directly copied
a classical monument or imitated a more recent work derived from a classical
prototype through a series of intermediary transformations. The artists who
represented Medea as a mediaeval princess, or Jupiter as a mediaeval judge,
translated into images a mere description found in literary sources.
This is very true, and the textual tradition through which the knowledge
of classical themes, particularly of classical mythology, was transmitted to
and persisted during the Middle Ages is of the utmost importance, not only
for the mediaevalist but also for the student of Renaissance iconography. For
even in the Italian Quattrocento, it was from this complex and often very cor-
rupt tradition, rather than from genuine classical sources, that many people
drew their notions of classical mythology and related subjects.
Limiting ourselves to classical mythology, the paths of this tradition can
be outlined as follows. The later Greek philosophers had already begun to
interpret the pagan gods and demigods as mere personifications either of
natural forces or moral qualities, and some of them had gone so far as to
explain them as ordinary human beings subsequently deified. In the last
century of the Roman Empire these tendencies greatly increased. While
the Christian Fathers endeavored to prove that the pagan gods were either
illusions or malignant demons (thereby transmitting much valuable infor-
mation about them), the pagan world itself had become so estranged from
its divinities that the educated public had to read up on them in encyclo-
Copyright © 2009. Oxford University Press. All rights reserved.

paedias, in didactic poems or novels, in special treatises on mythology, and

in commentaries on the classic poets. Important among these late-antique
writings in which the mythological characters were interpreted in an alle-
gorical way, or ‘moralized,’ to use the mediaeval expression, were Martianus
Capella’s Nuptiae Mercurii et Philologiae, Fulgentius’ Mitologiae, and, above
all, Servius’ admirable Commentary on Virgil which is three or four times as
long as the text and was perhaps more widely read.
During the Middle Ages these writings and others of their kind were
thoroughly exploited and further developed. The mythographical informa-
tion thus survived, and became accessible to mediaeval poets and artists.
First, in the encyclopaedias, the development of which began with such early
writers as Bede and Isidorus of Seville, was continued by Hrabanus Maurus
(ninth century), and reached a climax in the enormous high-mediaeval works
by Vincentius of Beauvais, Brunetto Latini, Bartholomaeus Anglicus, and so
forth. Second, in the mediaeval commentaries on classical and late-antique
texts, especially on Martianus Capella’s Nuptiae, which was annotated by

230 mechanisms of meaning

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 Irish scholars such as Johannes Scotus Erigena and was authoritatively com-
mented upon by Remigius of Auxerre (ninth century).11 Third, in special
treatises on mythology such as the so-called Mythographi I and II, which are
still rather early in date and are mainly based on Fulgentius and Servius.12
The most important work of this kind, the so-called Mythographus III, has
been tentatively identified with an Englishman, the great scholastic Alex-
ander Neckham (died 1217);13 his treatise, an impressive survey of whatever
information was available around 1200, deserves to be called the conclusive
compendium of high-mediaeval mythography, and was even used by Petrarch
when he described the images of pagan gods in his poem Africa.
Between the times of the Mythographus III and Petrarch a further step in
the moralization of classical divinities had been taken. The figures of ancient
mythology were not only interpreted in a general moralistic way but were
quite definitely related to the Christian faith, so that, for instance, Pyramus
was interpreted as Christ, Thisbe as the human soul, and the lion as Evil defil-
ing its garments; while Saturn served as an example, both in a good and in a
bad sense, for the behavior of clergymen. Instances of this type of writings are
the French Ovide Moralisé,14 John Ridewall’s Fulgentius Metaforalis,15 Rob-
ert Holcott’s Moralitates, the Gesta Romanorum and, above all, the Moralized
Ovid in Latin, written around 1340 by a French theologian called Petrus Ber-
chorius or Pierre Bersuire, who was personally acquainted with Petrarch.16
His work is preceded by a special chapter on the pagan gods, mainly based
on the Mythographus III, but enriched by specifically Christian moraliza-
tions, and this introduction, with the moralizations cut out for brevity’s sake,
attained great popularity under the name of Albricus, Libellus de Imaginibus
Deorum.17
A fresh and highly important start was made by Boccaccio. In his Genea-
logia Deorum18 he not only gave a new survey of the material, greatly enlarged
since about 1200, but also tried consciously to revert to the genuine Antique
sources and carefully collate them with one another. His treatise marks the
beginning of a critical or scientific attitude towards classical antiquity, and
may be called a forerunner of such truly scholarly Renaissance treatises as the
De diis gentium … Syntagmata by L. G. Gyraldus, who, from his point of view,
Copyright © 2009. Oxford University Press. All rights reserved.

was fully entitled to look down upon his most popular mediaeval predecessor
as a ‘proletarian and unreliable writer.’19
It will be noticed that up to Boccaccio’s Genealogia Deorum the focal point
of mediaeval mythography was a region widely remote from direct Mediter-
ranean tradition: Ireland, Northern France and England. This is also true of
the Trojan Cycle, the most important epic theme transmitted by classical
antiquity to posterity; its first authoritative mediaeval redaction, the Roman
de Troie, which was frequently abridged, summarized and translated into
the other vernacular languages, is due to Benoît de Ste. More, a native of
Brittany. We are in fact entitled to speak of a proto-humanistic movement,
viz., an active interest in classical themes regardless of classical motifs, cen-
tered in the northern region of Europe, as opposed to the proto-Renaissance
movement, viz., an active interest in classical motifs regardless of classical
themes, centered in Provence and Italy. It is a memorable fact which we must
bear in mind in order to understand the Renaissance movement proper, that
Petrarch, when describing the gods of his Roman ancestors, had to consult

erwin panofsky 231

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 a compendium written by an Englishman, and that the Italian illuminators
who illustrated Virgil’s Aeneid in the fifteenth century had to have recourse
to the miniatures in manuscripts of the Roman de Troie and its derivatives.
For these, being a favorite reading matter of noble laymen, had been amply
illustrated long before the Virgil text proper, read by scholars and schoolboys,
and had attracted the attention of professional illuminators.20
It is indeed easy to see that the artists who from the end of the eleventh
century tried to translate into images those proto-humanistic texts could not
but depict them in a manner utterly different from classical traditions. One
of the earliest instances is among the most striking: a miniature of about 1100,
probably executed in the school of Regensburg, depicting the classical divin-
ities according to the descriptions in Remigius’ Commentary on Martianus
Capella.21 Apollo is seen riding in a peasant’s cart and holding in his hand
a kind of nosegay with the busts of the Three Graces. Saturn looks like a
Romanesque jamb-figure rather than like the father of the Olympian gods,
and the raven of Jupiter is equipped with a tiny halo like the eagle of St. John
the Evangelist or the dove of St. Gregory.
Nevertheless, the contrast between representational and textual tradition
alone, important though it is, cannot account for the strange dichotomy of
classical motifs and classical themes characteristic of high-mediaeval art. For
even when there had been a representational tradition in certain fields of clas-
sical imagery, this representational tradition was deliberately relinquished in
favor of representations of an entirely nonclassical character as soon as the
Middle Ages had achieved a style entirely their own.
Instances of this process are found, first, in classical images incidentally
occurring in representations of Christian subjects, such as the personifica-
tions of natural forces in, for example, the Utrecht Psalter, or the sun and the
moon in the Crucifixion. While Carolingian ivories still show the perfectly
classical types of the Quadriga Solis and the Biga Lunae,22 these classical types
are replaced by nonclassical ones in Romanesque and Gothic representations.
The personifications of nature tended to disappear; only the pagan idols fre-
quently found in scenes of martyrdom preserved their classical appearance
longer than other images because they were the symbols par excellence of
Copyright © 2009. Oxford University Press. All rights reserved.

paganism. Secondly, what is much more important, genuine classical images

appear in the illustrations of such texts as had already been illustrated in
lateantique times, so that visual models were available to the Carolingian art-
ists: the Comedies of Terence, the texts incorporated into Hrabanus Maurus’
De Universo, Prudentius’ Psychomachia, and scientific writings, particularly
treatises on astronomy, where mythological images appear both among the
constellations (such as Andromeda, Perseus, Cassiopeia) and as planets
(Saturn, Jupiter, Mars, Sol, Venus, Mercury, Luna).
In all these cases we can observe that the classical images were faithfully
though often clumsily copied in Carolingian manuscripts and lingered on in
their derivatives, but that they were abandoned and replaced by entirely dif-
ferent ones in the thirteenth and fourteenth centuries at the latest.
In the ninth-century illustrations of an astronomical text, such mytho-
logical figures as Boötes, Perseus, Hercules or Mercury are rendered in a per-
fectly classical fashion, and the same is true of the pagan divinities appearing
in Hrabanus Maurus’ Encyclopaedia.23 With all their clumsiness, which is

232 mechanisms of meaning

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 chiefly due to the incompetence of the poor eleventh-century copyist of the
lost Carolingian manuscript, the figures in the Hrabanus illustrations are evi-
dently not concocted from mere textual descriptions but are connected with
Antique prototypes by a representational tradition.
However, some centuries later these genuine images had fallen into obliv-
ion and were replaced by others—partly newly invented, partly derived from
oriental sources—which no modern spectator would ever recognize as clas-
sical divinities. Venus is shown as a fashionable young lady playing the lute
or smelling a rose, Jupiter as a judge with his gloves in his hand, and Mercury
as an old scholar or even as a bishop.24 It was not before the Renaissance
proper that Jupiter reassumed the appearance of the classical Zeus, and that
Mercury reacquired the youthful beauty of the classical Hermes.25
All this shows that the separation of classical themes from classical motifs
took place, not only for want of a representational tradition, but even in
spite of a representational tradition. Wherever a classical image, that is, a
fusion of a classical theme with a classical motif, had been copied during the
Carolingian period of feverish assimilation, this classical image was abandoned
as soon as mediaeval civilization had reached its climax, and was not reinstated
until the Italian Quattrocento. It was the privilege of the Renaissance proper
to reintegrate classical themes with classical motifs after what might be called
a zero hour.
For the mediaeval mind, classical antiquity was too far removed and at
the same time too strongly present to be conceived as an historical phenom-
enon. On the one hand an unbroken continuity of tradition was felt in so far
as, for example, the German emperor was considered the direct successor of
Caesar and Augustus, while the linguists looked upon Cicero and Donatus
as their forefathers, and the mathematicians traced their ancestry back to
Euclid. On the other hand, it was felt that an insurmountable gap existed
between a pagan civilization and a Christian one.26 These two tendencies
could not as yet be balanced so as to permit a feeling of historical distance.
In many minds the classical world assumed a distant, fairy-tale character like
the contemporary pagan East, so that Villard de Honnecourt could call a
Roman tomb ‘la sepouture d’un sarrazin,’ while Alexander the Great and Vir-
Copyright © 2009. Oxford University Press. All rights reserved.

gil came to be thought of as oriental magicians. For others, the classical world
was the ultimate source of highly appreciated knowledge and time-honored
institutions. But no mediaeval man could see the civilization of antiquity as
a phenomenon complete in itself, yet belonging to the past and historically
detached from the contemporary world—as a cultural cosmos to be inves-
tigated and, if possible, to be reintegrated, instead of being a world of living
wonders or a mine of information. The scholastic philosophers could use the
ideas of Aristotle and merge them with their own system, and the mediaeval
poets could borrow freely from the classical authors, but no mediaeval mind
could think of classical philology. The artists could employ, as we have seen,
the motifs of classical reliefs and classical statues, but no mediaeval mind
could think of classical archaeology. Just as it was impossible for the Middle
Ages to elaborate the modern system of perspective, which is based on the
realization of a fixed distance between the eye and the object and thus enables
the artist to build up comprehensive and consistent images of visible things;
so was it impossible for them to evolve the modern idea of history, based on

erwin panofsky 233

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 the realization of an intellectual distance between the present and the past
which enables the scholar to build up comprehensive and consistent concepts
of bygone periods.
We can easily see that a period unable and unwilling to realize that clas-
sical motifs and classical themes structurally belonged together, actually
avoided preserving the union of these two. Once the Middle Ages had estab-
lished their own standards of civilization and found their own methods of
artistic expression, it became impossible to enjoy or even to understand any
phenomenon which had no common denominator with the phenomena of
the contemporary world. The high-mediaeval beholder could appreciate a
beautiful classical figure when presented to him as a Virgin Mary, and he
could appreciate a Thisbe depicted as a girl of the thirteenth century sitting
by a Gothic tombstone. But a Venus or Juno classical in form as well as sig-
nificance would have been an execrable, pagan idol while a Thisbe attired in
classical costume and sitting by a classical mausoleum would have been an
archaeological reconstruction entirely beyond his possibilities of approach.
In the thirteenth century even classical script was felt as something utterly
‘foreign’: the explanatory inscriptions in the Carolingian Cod. Leydensis Voss.
lat. 79, written in a beautiful Capitalis Rustica, were copied, for the benefit of
less erudite readers, in angular High Gothic script.
However, this failure to realize the intrinsic ‘oneness’ of classical themes
and classical motifs can be explained, not only by a lack of historical feeling,
but also by the emotional disparity between the Christian Middle Ages and
pagan antiquity. Where Hellenic paganism—at least as reflected in classical
art—considered man as an integral unity of body and soul, the Jewish-Christian
conception of man was based on the idea of the ‘clod of earth’ forcibly, or
even miraculously, united with an immortal soul. From this point of view, the
admirable artistic formulae which in Greek and Roman art had expressed
organic beauty and animal passions, seemed admissible only when invested
with a more-than-organic and more-than-natural meaning; that is, when
made subservient to Biblical or theological themes. In secular scenes, on
the contrary, these formulae had to be replaced by others, conforming to the
mediaeval atmosphere of courtly manners and conventionalized sentiments,
Copyright © 2009. Oxford University Press. All rights reserved.

so that heathen divinities and heroes mad with love or cruelty appeared as
fashionable princes and damsels whose looks and behavior were in harmony
with the canons of mediaeval social life.
In a miniature from a fourteenth-century Ovide Moralisé, the Rape of
Europa is enacted by figures which certainly express little passionate agita-
tion.27 Europa, clad in late-mediaeval costume, sits on her inoffensive little
bull like a young lady taking a morning ride, and her companions, similarly
attired, form a quiet little group of spectators. Of course, they are meant to
be anguished and to cry out, but they don’t, or at least they don’t convince us
that they do, because the illuminator was neither able nor inclined to visualize
animal passions.
A drawing by Dürer, copied from an Italian prototype probably during his
first stay in Venice, emphasizes the emotional vitality which was absent in the
mediaeval representation. The literary source of Dürer’s Rape of Europa is no
longer a prosy text where the bull was compared to Christ, and Europa to the
human soul, but the pagan verses of Ovid himself as revived in two delight-

234 mechanisms of meaning

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 ful stanzas by Angelo Poliziano: ‘You can admire Jupiter transformed into a
beautiful bull by the power of love. He dashes away with his sweet, terrified
load, and she turns back her face to the lost shore, her beautiful golden hair
fluttering in the wind which blows back her gown. With one hand she grasps
the horn of the bull, while the other clings to his back. She draws up her feet
as if afraid that the sea might wet them, and thus crouching down with pain
and fear, she cries for help in vain. For her sweet companions remain on the
flowery shore, each of them crying ‘Europa, come back.’ The whole seashore
resounds with ‘Europa, come back,’ and the bull looks round [or ‘swims on’]
and kisses her feet.’28
Dürer’s drawing actually gives life to this sensual description. The crouch-
ing position of Europa, her fluttering hair, her clothes blown back by the
wind and thus revealing her graceful body, the gestures of her hands, the fur-
tive movement of the bull’s head, the seashore scattered with the lamenting
companions: all this is faithfully and vividly depicted; and, even more, the
beach itself rustles with the life of aquatici monstriculi, to speak in the terms of
another Quattrocento writer, while satyrs hail the abductor.
This comparison illustrates the fact that the reintegration of classical
themes with classical motifs which seems to be characteristic of the Italian
Renaissance as opposed to the numerous sporadic revivals of classical ten-
dencies during the Middle Ages, is not only a humanistic but also a human
occurrence. It is a most important element of what Burckhardt and Michelet
called ‘the discovery both of the world and of man.’
On the other hand, it is self-evident that this reintegration could not be
a simple reversion to the classical past. The intervening period had changed
the minds of men, so that they could not turn into pagans again; and it had
changed their tastes and productive tendencies, so that their art could not
simply renew the art of the Greeks and Romans. They had to strive for a new
form of expression, stylistically and iconographically different from the clas-
sical, as well as from the mediaeval, yet related and indebted to both.
Copyright © 2009. Oxford University Press. All rights reserved.

erwin panofsky 235

Preziosi, D. (2009). Art of art history : A critical anthology. Retrieved from http://ebookcentral.proquest.com
Created from soas-ebooks on 2019-10-04 06:48:04.
 4. An extended discussion of the proto-semiotician, the latter
historiography of ‘social history’ in a general overview of visual-verbal relations
academic art history during its most fruitful from the perspective of traditional literary
period, the early fourth quarter of the criticism.
twentieth century, may be found in 7. It does not mention the work of Jan
D. Preziosi, Rethinking Art History, ch. 6. Mukarovssky (1891–1975), an eminent
5. The literature on the subject is Czech aesthetic philosopher and a
enormous. A few of those relevant to visual semiotician of art and architecture who
culture studies and art history include was a member of the ‘Prague School’ group
the following. The best compendium of of the 1930s. Several of his essays became
information about all aspects of semiotics more widely known in Western Europe
is Winifried Noeth, Handbook of Semiotics and America only after the Second World
(Bloomington, Ind., 1990), in which all War. One of the most important of his
aspects of visual semiotics are discussed books (Aesthetic Function, Norm and Value as
(pp. 481–550). A good survey of visual Social Facts (1936) ) was republished in 1979
semiotics is Goeran Harry Sonessdon, in English by the University of Michigan
Pictorial Concepts: Inquiries into the Semiotic Press; his 1934 essay ‘Art as Semiological
Heritage and its Relevance for the Analysis Fact’ was republished in English in 1976 in
of the Visual World in the series Ars Nova L. Matejka and I. R. Titunik (eds.), The
published by the Institute of Art History, Semiotics of Art (Cambridge, Mass.).
University of Lund, Sweden (Lund, 1989). 8. According to the authors, one entire
Other useful texts are Fernande Martin, section of the originally submitted essay,
Semiotics of Visual Language (Bloomington, dealing with what they had argued was
Ind., 1990); Meyer Shapiro, ‘On Some an essential component of semiotic study,
Problems in the Semiotics of Visual Art: namely gender, had to be removed if the
Field and Vehicle in Image-Signs’, Semiotica, essay was to be accepted for publication;
1 (1969), 223–42; id., Words and Pictures: On ‘feminism’ was deemed separate from
the Literal and the Symbolic in the Illustration of studies of social systems of signs, despite
a Text (The Hague, 1973). A good overview its essential connection to semiology in
of the philosophical interest in systems literature in art history and literary and
of signification from the 17th to the early other studies elsewhere.
20th centuries is Hans Aarsleff, From Locke
to Saussure: Essays on the Study of Language Erwin Panofsky: Iconography and
and Intellectual History (Minneapolis, 1982), Iconology: An Introduction to the study of
in which two essays are of interest here: Renaissance Art
‘Taine and Saussure’, pp. 356–71, illustrating 1. Images conveying the idea, not of
the ways in which key notions of Ferdinand concrete and individual persons or
de Saussure, the Swiss linguist credited objects (such as St. Bartholomew, Venus,
(along with the 19th-century American Mrs. Jones, or Windsor Castle), but of
Copyright © 2009. Oxford University Press. All rights reserved.

philosopher Charles Sanders Peirce) with abstract and general notions such as Faith,
being the progenitor of modern semiology, Luxury, Wisdom, etc., are called either
were derived from Hippolyte Taine’s personifications or symbols (not in the
lectures on art in Paris when Saussure was Cassirerian, but in the ordinary sense, e.g.,
a student there (summarized in Preziosi, the Cross, or the Tower of Chastity). Thus
Rethinking Art History, ch. 4., pp. 80–121, allegories, as opposed to stories, may be
and ‘Condillac’s Speechless Statue’, pp. defined as combinations of personifications
210–24). and/or symbols. There are, of course, many
6. Among many useful writings on the intermediary possibilities. A person A. may
subject of relations between iconology be portrayed in the guise of the person B.
and semiology is Christine Hasenmueller, (Bronzino’s Andrea Doria as Neptune:
‘Panofsky, Iconography, and Semiotics’, Dürer’s Lucas Paumgärtner as St.
Journal of Aesthetics and Art Criticism, George), or in the customary array of a
36 (1978), 289–301. Related book-length personification (Joshua Reynolds’ Mrs.
studies include Michael A. Holly, Panofsky Stanhope as ‘Contemplation’); portrayals
and the Foundations of Art History (Ithaca, NY, of concrete and individual persons, both
1984), and W. J. T. Mitchell, Iconology: Image, human or mythological, may be combined
Text, Ideology (Chicago, 1986), the former with personifications, as is the case in
an attempt at understanding Panofsky as a countless representations of a eulogistic

528 notes
Preziosi, Donald. <i>Art of Art History : A Critical Anthology</i>, Oxford University Press, 2009. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/soas-ebooks/detail.action?docID=453635.
Created from soas-ebooks on 2019-10-04 09:12:06.
 character. A story may convey, in addition, 5. One of the North Italian pictures is
an allegorical idea, as is the case with ascribed to Romanino and is preserved in
the illustrations of the Ovide Moralisé, or the Berlin Museum, where it was formerly
may be conceived as the ‘prefiguration’ of listed as ‘Salome’ in spite of the maid, a
another story, as in the Biblia Pauperum or sleeping soldier, and the city of Jerusalem
in the Speculum Humanae Salvationis. Such in the background (No. 155); another is
superimposed meanings either do not enter ascribed to Romanino’s pupil Francesco
into the content of the work at all, as is the Prato da Caravaggio (listed in the Berlin
case with the Ovide Moralisé illustrations, Catalogue), and a third is by Bernardo
which are visually indistinguishable from Strozzi, who was a native of Genoa but
non-allegorical miniatures illustrating the active at Venice about the same time
same Ovidian subjects; or they cause an as Francesco Maffei. It is very possible
ambiguity of content, which can, however, that the type of ‘Judith with a Charger’
be overcome or even turned into an added originated in Germany. One of the earliest
value if known instances (by an anonymous master
the conflicting ingredients are molten of around 1530 related to Hans Baldung
in the heat of a fervent artistic Grien) has been published by G. Poensgen,
temperament as in Rubens’ ‘Galerie ‘Beiträge zu Baldung und seinem Kreis,’
de Médicis.’ Zeitschrift für Kunstgeschichte, VI, 1937, p. 36 ff.
2. G. Leidinger, Das sogenannte Evangeliar 6. Illustrated in E. Panofsky and F. Saxl,
Ottos III, Munich, 1912, Pl. 36. ‘Classical Mythology in Mediaeval Art,’
3. To correct the interpretation of an Metropolitan Museum Studies, IV, 2, 1933,
individual work of art by a ‘history of p. 228 ff., p. 231.
style,’ which in turn can only be built up 7. See K. Weitzmann, ‘Das Evangelion
by interpreting individual works, may look im Skevophylakion zu Lawra,’ Seminarium
like a vicious circle. It is, indeed, a circle, Kondakovianum, VIII, 1936, p. 83 ff.
though not a vicious, but a methodical 8. Cod. Vat. lat. 2761, illustrated in
one (cf. E. Wind, Das Experiment und die Panofsky and Saxl, op. cit., p. 259.
Metaphysik, cited above, p. 6; idem, ‘Some 9. Paris, Bibliothèque Nationale, MS. lat.
Points of Contact between History and 15158, dated 1289, illustrated in Panofsky and
Science,’ cited ibidem). Whether we deal Saxl, op. cit., p. 272.
with historical or natural phenomena, 10. C. Tolnay, ‘The Visionary Evangelists of
the individual observation assumes the the Reichenau School,’ Burlington Magazine,
character of a ‘fact’ only when it can be LXIX, 1936, p. 257 ff., has made the
related to other, analogous observations in important discovery that the impressive
such a way that the whole series ‘makes images of the Evangelists seated on a globe
sense.’ This ‘sense’ is, therefore, fully and supporting a heavenly glory (occurring
capable of being applied, as a control, to for the first time in Cod. Vat. Barb. lat. 711),
the interpretation of a new individual combine the features of Christ in Majesty
Copyright © 2009. Oxford University Press. All rights reserved.

observation within the same range with those of a Graeco-Roman celestial

of phenomena. If, however, this new divinity. However, as Tolnay himself
individual observation definitely refuses points out, the Evangelists in Cod. Barb.
to be interpreted according to the ‘sense’ 711 ‘support with obvious effort a mass of
of the series, and if an error proves to be clouds which does not in the least look like
impossible, the ‘sense’ of the series will a spiritual aura but like a material weight
have to be reformulated to include the consisting of several segments of circles,
new individual observation. This circulus alternately blue and green, the outline
methodicus applies, of course, not only to the of the whole forming a circle … It is a
relationship between the interpretation of misunderstood representation of heaven in
motifs and the history of style, but also to the form of spheres’ (italics mine). From this
the relationship between the interpretation we can infer that the classical prototype
of images, stories and allegories and the of these images was not Coelus who holds
history of types, and to the relationship without effort a billowing drapery (the
between the interpretation of intrinsic Weltenmantel) but Atlas who labors under
meanings and the history of cultural the weight of the heavens (cf. G. Thiele,
symptoms in general. Antike Himmelsbilder, Berlin, 1898,
4. G. Fiocco, Venetian Painting of the Seicento p. 19 ff.). The St. Matthew in Cod. Barb.
and the Settecento, Florence and New York, 711 (Tolnay, PI. I, a), with his head bowed
1929, Pl. 29. down under the weight of the sphere and

notes 529
Preziosi, Donald. <i>Art of Art History : A Critical Anthology</i>, Oxford University Press, 2009. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/soas-ebooks/detail.action?docID=453635.
Created from soas-ebooks on 2019-10-04 09:12:06.
 his left hand still placed near his left hip, 13. Bode, ibidem, p. 152 ff. As to the question
is particularly reminiscent of the classical of authorship, see H. Liebeschütz, op. cit.,
type of Atlas, and another striking example p. 16 f. and passim.
of the characteristic Atlas pose applied 14. Ed. by C. de Boer, ‘Ovide Moralisé,’
to an Evangelist is found in Clm. 4454, Verhandelingen der kon. Akademie van
fol. 86, v. (illustrated in A. Goldschmidt, Wetenschapen, Afd. Letterkunde, new ser., XV,
German Illumination, Florence and New 1915; XXI, 1920; XXX, 1931–32.
York, 1928, Vol. II, Pl. 40). Tolnay (Notes 15. Ed. H. Liebeschütz, op. cit.
13 and 14) has not failed to notice this 16. ‘Thomas Walleys’ (or Valeys),
similarity and cites the representations Metamorphosis Ovidiana moraliter explanata,
of Atlas and Nimrod in Cod. Vat. Pal. lat. here used in the Paris edition of 1515.
1417, fol. 1 (illustrated in F. Saxl, Verzeichnis 17. Cod. Vat. Reg. 1290, ed. H. Liebeschütz,
astrologischer and mythologischer Handschriften op. cit., p. 117 ff. with the complete set of
des lateinischen Mittelalters in römischen illustrations.
Bibliotheken [Sitzungsberichte der Heidelberger 18. Here used in the Venice edition of 1511.
Akademie der Wissenschaften, phil.-hist. Klasse, 19. L. G. Gyraldus, Opera Omnia, Leyden,
VI, 1915, PI. XX, Fig. 42]); but he seems to 1696, Vol. I, col. 153: ‘Ut scribit Albricus, qui
consider the Atlas type as a mere derivative auctor mihi proletarius est, nec fidus satis.’
of the Coelus type. Yet even in ancient 20. The same applies to Ovid: there
art the representations of Coelus seem to are hardly any illustrated Latin Ovid
have developed from those of Atlas, and manuscripts in the Middle Ages. As to
in Carolingian, Ottonian and Byzantine Virgil’s Aeneid, I know only two really
art (particularly in the Reichenau school) ‘illustrated’ Latin manuscripts between
the figure of Atlas, in its genuine classical the sixth-century codex in the Vatican
form, is infinitely more frequent than Library and the fifteenth-century
that of Coelus, both as a personification Riccardianus: Naples, Bibl. Nazionale,
of cosmological character and as a kind of Cod. olim Vienna 58 (brought to my
caryatid. From an iconographical point of attention by Professor Kurt Weitzmann) of
view, too, the Evangelists are comparable the tenth century; and Cod. Vat. Ilat. 2761
to Atlas, rather than to Coelus. Coelus (cf. R. Förster, ‘Laocoön im Mittelalter
was believed to rule the heavens. Atlas und in der Renaissance,’ Jahrbuch der
was believed to support them and, in an Königlich Preussischen Kunstsammlungen,
allegorical sense, to ‘know’ them; he was XXVII, 1906, p. 149 ff.) of the fourteenth.
held to have been a great astronomer who [Another fourteenth-century manuscript
transmitted the scientia coeli to Hercules (Oxford, Bodleian Library, MS. Can. Class.
(Servius, Comm. in Aen., VI, 395; later lat. 52, described in F. Saxl and H. Meier,
on, e.g., Isidorus, Etymologiae, III, 24, 1; Catalogue of Astrological and Mythological
Mythographus III, 13, 4, in G. H. Bode, Manuscripts of the Latin Middle Ages, III,
Scriptorum rerum mythicarum tres Romae nuper Manuscripts in English Libraries, London,
Copyright © 2009. Oxford University Press. All rights reserved.

reperti, Celle, 1834, p. 248). It was therefore 1953, p. 320 ff.) has only some historiated
consistent to use the type of Coelus for initials.]
the representation of God (see Tolnay, PI. 21. Clm. 14271, illustrated in Panofsky and
I, c), and it was equally consistent to use Saxl, op. cit., p. 260.
the type of Atlas for the Evangelists who, 22. A. GoIdschmidt, Die Elfenbeinskulpturen
like him, ‘knew’ the heavens but did not aus der Zeit der karolingischen und sächsischen
rule them. While Hibernus Exul says of Kaiser, Berlin, 1914–26, Vol. I, Pl. XX,
Atlas Sidera quem coeli cuncta notasse volunt No. 40, illustrated in Panofsky and Saxl,
(Monumenta Germaniae, Poetarum latinorum op. cit., p. 257.
medii aevi, Berlin, 1881–1923, Vol. I, p. 410), 23. Cf. A. M. Amelli, Miniature sacre e profane
Alcuin thus apostrophizes St. John the dell’anno 1023, illustranti l’enciclopedia medioevale di
Evangelist: Scribendo penetras caelum tu, mente, Rabano Mauro, Montecassino, 1896.
Johannes (ibidem, p. 293). 24. Clm. 10268 (fourteenth century),
11. See H. Liebeschütz, Fulgentius illustrated in Panofsky and Saxl, op. cit.,
Metaforalis … (Studien der Bibliothek p. 251, and the whole group of other
Warburg, IV), Leipzig, 1926, p. 15 and illustrations based on the text by Michael
p. 44 ff.; cf. also Panofsky and Saxl, op. cit., Scotus. For the oriental sources of these
especially p. 253 ff. new types, see ibidem, p. 239 ff., and
12. Bode, op. cit., p. 1 ff. F. Saxl, ‘Beiträge zu einer Geschichte

530 notes
Preziosi, Donald. <i>Art of Art History : A Critical Anthology</i>, Oxford University Press, 2009. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/soas-ebooks/detail.action?docID=453635.
Created from soas-ebooks on 2019-10-04 09:12:06.
 der Planetendarstellungen in Orient und an apparently ‘natural’ device like linear
Occident,’ Der Islam, III, 1912, p. 151 ff. perspective is masterfully demonstrated in
25. For an interesting prelude of this Hubert Damisch’s seminal study, L’Origine
reinstatement (resumption of Carolingian de la perspective, (Paris, 1988).
and archaic Greek models), see Panofsky 3. See e.g. M. Schapiro, ‘On Some Problems
and Saxl, op. cit., pp. 247 and 258. in the Semiotics of Visual Art: Field and
26. A similar dualism is characteristic of Vehicle in Image-Signs’, Semiotica, 1, (1969),
the mediaeval attitude towards the aera 223–42.
sub lege: on the one hand the Synagogue 4. The clearest and most convincing
was represented as blind and associated overview of epistemological currents in
with Night, Death, the devil and impure the 19th and 20th centuries is Habermas’s
animals; and on the other hand the Jewish Erkenntnis und Interesse of 1968 (Knowledge
prophets were considered as inspired by the and Human Interests, trans. J. Shapiro,
Holy Ghost, and the personages of the Old (London, 1972). Habermas’s work has been
Testament were venerated as the ancestors challenged by psychoanalysts who believe
of Christ. that his idealized view of psychoanalytic
27. Lyons, Bibl. de la Ville, MS. 742, fol. practice as a constraint-free communication
40; illustrated in Saxl and Panofsky, op. cit., misunderstands their discipline. See e.g. J.
p. 274. Rose, Sexuality in the Field of Vision, (London,
28. L.456, also illustrated in Saxl and 1986). Habermas’s œuvre is also under
Panofsky, op. cit., p. 275. Angelo Poliziano’s pressure from the side of postmodern
stanzas (Giostra I, 105, l06) read as follows: philosophy, most pertinently by
‘Nell’altra in un formoso e bianco tauro J.-F. Lyotard, in e.g. The Postmodern Condition,
Si vede Giove per amor converso (New York, 1980). These challenges do not,
Portarne il dolce suo ricco tesauro, however, address Habermas’s argument
E lei volgere il viso al lito perso against positive knowledge, but his hope for
In atto paventoso: e i be’ crin d’auro a rational society. If anything, the authors
Scherzon nel petto per lo vento avverso: are more skeptical than Habermas.
La veste ondeggia e in drieto fa ritorno: 5. For the ‘linguistic’ or, rather, rhetorical
L’una man tien al dorso, e l’altra al corno. turn in history, see H. White, Metahistory:
The Historical Imagination in Nineteenth-Century
‘Le ignude piante a se ristrette accoglie
Europe (Baltimore, 1973), and especially,
Quasi temendo il mar che lei non bagne:
for a brief and convincing account of the
Tale atteggiata di paura e doglie
fundamental rhetorical and semiotic nature
Par chiami in van le sue dolci compagne;
of historiography, id., ‘Interpretation in
Le qual rimase tra fioretti e foglie
History’, in Tropics of Discourse (Baltimore,
Dolenti “Europa” ciascheduna piagne.
1978). The most detailed and incisive
“Europa,” sona il lito, “Europa, riedi”—
analysis of the rhetoric of historiography
E’l tor nota, e talor gli bacia i piedi.’
remains S. Bann’s remarkable The Clothing of
Copyright © 2009. Oxford University Press. All rights reserved.

Clio, (Cambridge and New York, 1984).

Hubert Damisch: Semiotics and
6. See e.g. the Rembrandt Research
Iconography
Project, in J. Bruyn, B. Haak, S. H. Levie
1. ‘On some Problems in the Semiotics
et al., A Corpus of Rembrandt Paintings, (The
of Visual Art: Field and Vehicle in Image
Hague, Boston, London, 1982, 1987, 1989),
Signs’, Semiotica, 1/3, 1969.
review by L. J. Slatkes in the Art Bulletin, 71
(1989), 139–44.
Mieke Bal and Norman Bryson: Semiotics
7. Culler, xiv.
and Art History
8. Similar arguments within the social
1. We would like to thank Michael Ann
history of art, explicitly articulating art
Holly for her very pertinent comments on
history with semiotics, have been put
this paper.
forward in a number of places by Keith
2. See C. Hasenmueller, ‘Panofsky,
Moxey. See ‘Interpreting Pieter Aertsen:
Iconography, and Semiotics’, Journal
The Problem of Hidden Symbolism’,
of Aesthetics and Art Criticism, 36, (1978),
Nederlands Kunsthistorisch Jaarboek, (1989), 42
289–301; M. Iversen, ‘Style as Structure:
ff.; ‘Pieter Bruegel and Popular Culture’,
Alois Riegl’s Historiography’, Art History,
The Complete Prints of Pieter Bruegel the Elder,
2(1979), 66–7; and M. A. Holly, Panofsky
ed. D. Freedberg, (Tokyo, 1989), 42 ff.;
and the Foundations of Art History (Ithaca,
‘Semiotics and the Social History of Art’,
NY, 1984), 42–5. The semiotic nature of

notes 531
Preziosi, Donald. <i>Art of Art History : A Critical Anthology</i>, Oxford University Press, 2009. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/soas-ebooks/detail.action?docID=453635.
Created from soas-ebooks on 2019-10-04 09:12:06.