Connoisseurship: The Penalty of Ahistoricism
Author(s): Gary Schwartz
Source: Artibus et Historiae , 1988, Vol. 9, No. 18 (1988), pp. 201-206
Published by: IRSA s.c.
Stable URL: https://www.jstor.org/stable/1483343
REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/1483343?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
IRSA s.c. is collaborating with JSTOR to digitize, preserve and extend access to Artibus et
Historiae
This content downloaded from
147.251.173.95 on Thu, 09 Mar 2023 11:29:27 UTC
All use subject to https://about.jstor.org/terms
GARY SCHWARTZ
Connoisseurship: The Penalty of Ahistoricism
No one who has worked with a gifted connoisseur would be
so foolish or ungrateful as to deny the existence of the gift or
its usefulness.1 One of my own close encounters with the
phenomenon dates back to 1968, when I entered the employ
of the Rijksbureau voor Kunsthistorische Documentatie in The
Hague. At the time, the RKD had 650,000 photographs of
Dutch and Flemish paintings on file, and I was told in quiet awe
that the director, Sturla Gudlaugsson, had recently visited the
Zentralinstitut fur Kunstgeschichte in Munich to check their
100,000 Netherlandish photos for unknown examples and
differing attributions. The ability of those such as Gudlaugsson
to recognize and sort countless images saves the rest of us
years of labor and speeds up the placing of particular works by
decades or even centuries. They provide a framework for the
formation of the mental images we use for our own
attributions. Many of us, myself included, have experienced
the satisfying flash of recognition that convinces us that we
have found the hand of a known master in an unidentified or
wrongly attributed work.
We do this as trained art historians, but there are nonacademics
- at auction houses, for example - who are able
to make determinations of period, place and authorship of old
master paintings, Oceanic artefacts or Coca-Cola bottles with
no more than an elementary knowledge of art history to help
them. The connoisseur's sensitivity to style, at its best, can
be compared to the ear of the musician with absolute pitch, the
feeling for numbers of the mathematical genius, the intuition
of the first-class detective or the practiced diagnostician.
Without being able to tell us how they know certain things, they
are sure that they do, and we believe them.
But we only believe them up to a point. In general,
knowledge of this kind is not considered a definitive source of
authority. One goes through two stages in judging such
performances. Hearing the expert opinion, our sense of
wonder is tinged by bemused skepticism. It is only when the
judgment is corroborated by outside evidence - a tuning fork,
a computer calculation, a smoking gun, a lab report, a signature
- that we truly are impressed, and we relinquish disbelief.
The problem begins when the connoisseur's judgment
cannot be confirmed by outside evidence. In that situation, we
no longer see expert opinion as a welcome shortcut to a more
dependable source of knowledge, but as a value unto itself.
Even then we are apt to accept the judgment of a recognized
authority as long as we have no reason to doubt it. It is only
when another authority challenges an attribution that belief
becomes insufficient. At that point we have no choice but to
challenge the connoisseur, demanding to know the foundations
of his judgments.
Many connoisseurs have met the challenge. Despite
the proverbial laconism of the connoisseur, the literature
is rich in manifestos, such as Bernard Berenson's Rudiments
of Connoisseurship,2 Max Friedlander's On Art and
Connoisseurship,3 and Jakob Rosenberg's On Quality in A
Attempting to account for their convictions, these
connoisseurs insist that their judgments proceed from the
perception, whether or not instantaneous, of specific
201
This content downloaded from
147.251.173.95 on Thu, 09 Mar 2023 11:29:27 UTC
All use subject to https://about.jstor.org/terms
properties - composition, drawing, brushwork, modelling,
expression, technical or anatomical features and so forth which
manifest the hand of the maker. Each writer agrees that,
in Friedlander's words, "the artist [...] at bottom remains the
same, and [...] something which cannot be lost reveals itself
in his every expression."5 That something which cannot be
lost is also, they concur, something which cannot be imitated.
The task of the connoisseur is to attune himself to these unique
and inimitable features.
All connoisseurs share this assumption, whether they
search for the master's fingerprint in his technical habits, the
minutiae of his style or the particulars of his quality. All operate
on the belief that the object itself contains all the information
needed to establish its authorship, and that this information can
be interpreted properly by the observer who knows what to
look for.
The most recent systematic description of connoisseurship
known to me - a chapter by Willibald Sauerlander in an
excellent introduction to art history for German students6 dilutes
its claims to some degree. Determinations of time and
place, he writes, can be made on a firm scholarly basis, but
attributions to individual masters cannot escape the "stigma of
subjectivity and time-boundedness."7 It is his sanguine
opinion that the field has taken account of this, and that
"attribution to individual hands [...] no longer fills the central
place it occupied a few brief decades ago."8 In my field this
is certainly not the case. The major efforts in Rembrandt
studies for the past thirty years have been the attempts of Otto
Benesch, Horst Gerson, Werner Sumowski and the Rembrandt
Research Project to establish corpora of the paintings and
drawings of Rembrandt and his school. In fact, Sauerlander
cites the Rembrandt Research Project as a model of how
connoisseurship can "constantly sharpen and refine its
instruments."9
If it were true, as Sauerlander claims, that art-historical
common sense has rendered connoisseurship unproblematical
and that steady progress is being made in applying it, I do not
think there would be any need to discuss it further. My own
discomfort with Rembrandt connoisseurship has certainly not
been laid to rest by the publications of the Rembrandt Research
Project, as thorough and well documented as they are. At the
same time, I know from experience how difficult it is to
determine the source of that discomfort and, more importantly,
to deal, as an art historian, with the ultimate defense of
connoisseurship: that while you cannot live with it, you cannot
live without it, either. Academic purists may squirm, but
they too practice connoisseurship or accept its findings.
Connoisseurship is still the id of the art-historical ego.
In the following essay, I would like to take on these two
oversize problems: namely to put my finger on what I feel is
most wrong with connoisseurship, and to sketch an art history
which, freed from its flaws, can withstand close intellectual
scrutiny.
Essentially, what the connoisseur does is to define a relation
between an existing work and a historical category. Dealing
with works of uncertain status, the connoisseur treats the
other two elements in the equation as givens: the categories
are formed by works whose authorship is firmly documented,
and defining the relation is an analytic technique whose
intricacies can be explained, although they mysteriously
continue to resist codification. A closer look reveals that the
two "givens" - the categories and the techniques by which
unknown works are matched to them - are actually quite
dubious.
The connoisseur's comparative material consists, in theory,
of existing works whose authorship is documented. This
sample, historically precious as it is, is, however, insufficient
for the stated purposes of connoisseurship. The disappearance
from sight of the entire oeuvres of many documented masters
distorts the record, so that the connoisseur's categories do not
correspond to historical reality.10 It is as if the sorted contents
of a number of containers were dumped on a heap, half the
containers were broken, and one then tried to sort the same
material into the remaining containers. It may be a valuable,
perhaps necessary exercise, but one should not entertain any
illusions concerning its truth to historical fact.
Even more questionable is the process by which the relation
between the work and the category is defined. In 1929,
Berenson said, under oath, "[...] when I see a picture, in most
cases I recognize it at once as being or not being by the master
it is ascribed to; the rest is merely a question of how to try to
fish out the evidence that will make the conviction as plain to
others as it is to me,"11 and, while the academic connoisseur
of today may bend over backwards insisting that his
attributions are analytical rather than intuitive, to my mind
connoisseurship has never laid the ghost of Berenson and his
acts of recognition.12
The methodological consequences of this go far to
undermine whatever theoretical foundations connoisseurship
may be said to possess. Allow me to illustrate this with some
examples from the connoisseurship of Rembrandt paintings.
There is an unsigned, undated, undocumented painting of
Christ in the Hyde Collection which has been praised most
generously as a Rembrandt by Otto Benesch, Seymour Slive
and Horst Gerson, and is unusual in its never having been
doubted by anyone, to my knowledge. It serves as one of the
benchmark works against which the authorship of other
paintings of Christ is judged. In 1956, Benesch called it "the
most monumental painting" of its kind,13 and Valentiner
considered it the "center figure" of a supposed series of
202
This content downloaded from
147.251.173.95 on Thu, 09 Mar 2023 11:29:27 UTC
All use subject to https://about.jstor.org/terms
four three-quarter-length evangelists, including the Louvre
St. Matthew.14 In 1965, Seymour Slive wrote of it, comparing
it with a smaller Christ, formerly in the De Boer Collection:
"Perhaps Rembrandt actually began the painting now at the
Hyde Collection with the thought of making a larger version of
the small panel, but his incredible power of invention led him
to depict quite another aspect of Christ's character."15 In
1969, Horst Gerson called it "a fine interpretation of an ideal
image of Christ."16 In the 1960's, Slive, Gerson and Bauch
illustrated the work after a fine-looking black-and-white
photograph.17 In the color illustration in my book on Rembrandt
of 1985, one notices that there may be something seriously
wrong with the condition of the painting. Recently, the Photo
Archive of the Getty Center yielded some photographs, taken
for the restorer William Suhr during his work on the canvas in
1958, which reveal that the middle of the canvas was cut out
at some point and later reattached, so that the composition of
the painting as we know it is not the work of Rembrandt at all,
but of a restorer. The condition of the surface, moreover, is
so poor that it tells us little about the original quality of the
painting.
Despite this, one continues to "recognize" the hand of
Rembrandt in the work. Even the absence of the most essential
qualities the connoisseur says he needs to analyze in order to
arrive at an attribution or recognize the hand of a master composition,
drawing, paint surface - does not diminish one's
certainty that this is a work by the master. This painting is
fortunately untypical. But how many of the works we study
would not be disavowed by their makers, on account of their
condition, as caricatures of their style and technique?
On the other hand, no amount of positive evidence, however
solid, is capable of forcing a recognition. An unprepossessing
painting of a man's head, monogrammed and dated "RHL
1629," is rejected by the Rembrandt Research Project,18
although it displays dendrochronological and technical
characteristics virtually identical to those of an accepted
painting also dated 1629.19 The rejected Rembrandt in the
Corcoran Art Gallery in Washington,20 signed "Rembrandt
f 1632" and traceable in documents as a Rembrandt from
the eighteenth century on, is painted in a technique
indistinguishable from the accepted portrait of Johannes
Uytenbogaert of 1633.21
My argument is not with the attributions themselves, with
which I happen to agree, whatever that information and the
opinions behind it are worth. The point I wish to make
concerns the priorities of connoisseurship. Once the hand of
Rembrandt has been recognized or rejected, the Rembrandt
Research Project is sometimes obliged to dismiss as irrelevant
technical evidence of the kind which it otherwise considers
corroboratory. The same unfortunate methodology is applied
to the signatures. Those on the rejected works I mentioned
correspond in type to the signatures on unquestioned works of
the period. The authors of the Corpus, however, applying the
standards of quality connoisseurship to the signatures,
conclude: "The shaping of the letters does not seem
spontaneous, and does not carry conviction,"22 and "Because
of the absence of spontaneity, the letters and figures do not
give the impression of authenticity."23
These examples, which are not untypical, show that a
"good" painting retains its power of conviction, even in the
absence of the qualities which are supposed to make it good,
while no surfeit of technical evidence will make a "bad" one
look any better. They make me doubt whether the further
refinement of the "instruments of connoisseurship" which
Sauerlander praises is going to be of much use. Given their
priorities, connoisseurs are unlikely ever to define the rules for
using those instruments.
If, then, the connoisseur establishes the relation between a
work and a category by ad hoc means and if the categories
themselves are demonstrably inaccurate, of what value can a
connoisseur's attribution be?
This question, if I am not mistaken, is being asked, in var
forms, by many if not all connoisseurs. Their collective
reaction, to date, has been to change not the substance
but the style of connoisseurship, in an attempt to neutralize th
effects of what Sauerlander calls "subjectivity and timeboundedness."
The encyclopedic connoisseur who depends
on his individual sensitivity is being replaced by interlocking
circles of specialists. More importantly, present-day
connoisseurs make eager use of old and new scientific
techniques for peering below the surface of their objects of
study. These developments are intended to objectify
attributions.
If my analysis of the logical defect in connoisseurship is
correct, however, then subjectivity is not a root problem. How
else but subjectively can one pair, by whatever means the
occasion seems to call for, an object of uncertain status to an
historically untenable category? To do this by committee
rather than individually, to consult autoradiographs and thread
counts in addition to qualitative criteria, to average out the
results by a process of consensus cannot change this. In fact,
such developments will probably only aggravate things, as art
historians enlist the aid of scientists in the search for chimeric
proofs of authenticity, defined by stricter and stricter
standards.
The issue at hand has been addressed in slightly differen
form by the philosopher Nelson Goodman in an essay entit
"Art and Authenticity," first published in 1968 and reprin
in 1983. Asking how we are to distinguish between
"Rembrandt's original painting Lucretia and [...] a superlative
203
This content downloaded from
147.251.173.95 on Thu, 09 Mar 2023 11:29:27 UTC
All use subject to https://about.jstor.org/terms
imitation of it,"24 Goodman establishes that sense perception
cannot do the trick,25 and concludes: "the only way of
ascertaining that the Lucretia before us is genuine is thus to
establish the historical fact that it is the actual object made by
Rembrandt."26 This leads us, I believe, to the crux of the
matter. The connoisseur's close examination of an object can
only produce an attribution on the model: "This painting is a
Rembrandt because it displays such and such aesthetic
qualities and/or such and such chemical and physical
properties." This is not only an ahistorical statement, it is
actually at odds with the historical statement: "This painting
is a Rembrandt because certain evidence proves that it was
made by Rembrandt van Rijn," which implies nothing whatever
about aesthetic or physical properties. In judging the
competing claims implicit in those two approaches to
authenticity, I follow Goodman in concluding that the only way
of ascertaining the authorship of an object is to establish the
history of its production. The other claim, which seeks proof
of authorship in the object itself, is simply unfounded. It
assumes that the artist - whether original creator, studio
assistant, copyist, forger or restorer - leaves a fingerprint in
every work, a fingerprint whose whorls can be properly
interpreted by the later observer who knows what to look for.
The Morellian part of this assumption has never been proved.
Despite valiant attempts by critics, psychologists and
physiologists, we still do not know whether the sum total of
stylistic and technical features of a work amount to a
fingerprint.27 To go by the record of connoisseurship as a
means of attribution, one would sooner conclude that nothing
in art is inimitable.
The second part of the assumption - that the fingerprint
is interpretable to the good observer - pays a thoroughly
undeserved compliment to the perspicacity of the connoisseur.
The annals of forgery, let alone the infinitely vaster ones of
restoration, should be enough to undermine one's faith that the
examination of a work of art, by even the most sophisticated
observer, puts one in touch with the psyche or handwriting of
the maker. The link provided by the work of art between the
creative personality of the artist and the response of the
modern connoisseur is too tenuous to justify the ahistorical
attribution. What is said of forgeries - that they seldom retain
their power of conviction for longer than a generation - is
equally true of attributions, for much the same reasons. The
study of technical and stylistic properties may lead to a
presumption of period and place strong enough to withstand
reasonable doubt - and in that sense I agree with Sauerlander
- but to employ them as indications of personal authorship is
to court self-deception.
The consequences of this go far beyond the specific area of
204
attributions. In general art history, as well, the works of
individual masters are still described as embodiments of
personal stylistic and technical attributes recognizable to the
knowing eye. Art historians, trained from the start to toggle
back and forth between historical and ahistorical approaches
to art, never seem to notice the basic contradiction between
them.
The application of Goodman's conclusion makes it possible
to define relations between works and categories more
consistently. To begin with, the works and categories
themselves will be different from those of the connoisseur.
The hard core of material will be much larger than the
connoisseur's, namely all the artists and works known from
sources and documents. These will be studied in the context
in which they are encountered, embedded in the treatises,
poetry, government records, contracts, inventories, testaments
- in the lives - of the times which produced them, not in the
museum world of today. (There is no compelling need to
despair of studying material from prehistorical or non-verbal
cultures in the same way. Their archaeological or
anthropological context can be every bit as revealing as written
records.) The relationship which we will be called on to define
will be given to a great degree by that context. It will be far
more varied than the brush-and-canvas relation sought by the
connoisseur, and will touch on all aspects of the family,
religion, politics, finances and social life of the artist, his
patrons and their ambience.
A special place in this body of basic material will be occupied
by the connoisseur's central examples: the existing objects
whose authorship and provenance are certain. Now, however,
we will understand them in relationship - even in quantitative
relationship - to the total production of art in the society in
which they originated, including lost works. This would
provide a greatly needed framework for the scientific
examination and stylistic analysis of these works, which in turn
would give structure to the more general study of the physical
and stylistic attributes of works of art. The attributions and
"de-attributions" one might make at that point, armed with so
much more powerful weapons than out-of-context stylistic
comparison alone, should be far better founded than the
ahistorical ones of the connoisseur. But that would not be its
main purpose. Even if most undocumented works end up with
what one might call generic as opposed to brand-name labels,
our understanding of those works as well will have been put
on a firmer footing than ever before. To admit to our
uncertainty with regard to the authorship of the great majori
of undocumented works will cause pain and frustration,
especially to collectors, but it will clarify for art historians and
their public the reality of the case. No one - least of all
This content downloaded from
147.251.173.95 on Thu, 09 Mar 2023 11:29:27 UTC
All use subject to https://about.jstor.org/terms
institutional or individual buyers of art - benefits in the long
run from the pretense that the author of a work can be identified
by the "close-reading" techniques of connoisseurship.
I have presented here the sharpest concise critique of
connoisseurship I am presently capable of formulating. I have
done this not for the sake of argument, but out of the deep
conviction that connoisseurship incorporates ahistorical values
which are irreconcilable with the historical ones that have
grown increasingly important to me in recent years - and not
to me alone, if I may judge from the tone of many recent
essays. To continue practicing connoisseurship as if the only
problems it presents are technical limitations and the need to
dispel the misunderstandings of the ill-informed will, I fear, put
our field out of step with the other humanities and sciences and
rob it of the enviable position it has won in our century.
At the same time, I am equally convinced that proclaiming
the abolition of connoisseurship is no answer to the problem,
even if such a thing were possible. The cultivation of the
connoisseur's sensibility and the insights it offers is of great
benefit to our appreciation of art as individuals and to the
scholarly study of art objects, which after all come to us with
an ahistorical dimension. Moreover, we have a responsibility
as museum officials to account as well as we can for the
authenticity and quality of the works we purchase or care for.
Even if we stop short of attaching names to undocumented
This article was written during my tenure as a Getty Scholar of the
Getty Center for the History of Art and the Humanities, Santa Monica,
1986-87. The Getty Center provided an environment for lively critical
debate, giving me the exceptional opportunity to discuss the issues
presented here with distinguished art historians on the Getty staff and
with visiting scholars of the Getty Center and Getty Museum
programs. Their reactions were of great benefit to me in putting the
article into final form. I would like to express my warmest thanks to
the Center for its furtherance of my work and to the colleagues there,
in particular Ben Lifson, who took the time to listen to and to comment
on my text.
It was also presented as a lecture on February 14, 1987, at the 75th
Annual Meeting of the College Art Association in Boston, at a seminar
entitled "The Theory and Practice of Connoisseurship at the End of the
Twentieth Century," chaired by Konrad Oberhuber and Henri Zerner.
The stimulus to participate in the session was provided by John Walsh,
who thrust the call for papers under my nose and urged me to work
out the position I had begun to formulate in "Rembrandt bij de
tandarts", Hollands Maandblad, November 1973, pp. 3-9.
works, we will still need connoisseurship for sorting them into
broader, more defensible categories, and for explaining our
value judgments, however tentatively.
The conclusion which I would attach to the foregoing
arguments is that connoisseurship occupies the wrong position
within art history. As the central clearing house for the objects
we study as well as those we buy, it introduces into the heart
of our field a self-perpetuating system of institutionalized value
judgments, a system which is coming under heavier external
criticism and internal strain all the time.
I would argue for the subjugation of connoisseurship to a
more structured historical approach to art on the lines sketched
above. Within such confines, connoisseurship would be freed
from the perennial demand to promise too much - the
certainty of truth - only to deliver too little: an all too
hypothetical opinion. Connoisseurship in the historical mold
would no longer be called upon to make distinctions of the
classical "Yes-Maybe-No" variety. (The ill-advised A, B and C
categories of the Rembrandt Research Project distort the vast
and valuable information in its Corpus). Its judgments would
be guided by the more graduated scale of historical
relationships in the sources and documents, and would be
capable of fine adjustment.
In this way, I feel, a place of honor can be preserved for
connoisseurship, as art history attempts to hold its own in what
is shaping up to be a highly competitive twenty-first century.
1 The word "connoisseurship" is used in this paper in its technical
art-historical sense: namely the determination of the authorship, date
or place of origin of an art object on the basis of close examination and
comparison.
2 B. Berenson, Rudiments of Connoiseurship: Study and Criticism
of Italian Art, New York, 1962 (first published 1902).
3 M. Friedlander, On Art and Connoisseurship, London, 1942.
4 J. Rosenberg, On Quality in Art: Criteria of Excellence, Past and
Present (The A. W. Mellon Lectures in the Fine Arts, 1964), London,
1967.
5 Friedlander, op. cit. (note 3), chap. XXIX, "On Personality and its
Development," p. 200.
6 W. Sauerlander, "Alterssicherung, Ortssicherung und Individualsicherung,"
in H. Belting et al., Kunstgeschichte: eine Einfuhrung,
Berlin, 1986, pp. 116-44.
7 Ibidem, p. 138.
8 "Im ganzem aber nimmt die Individualsicherung bei der kunsthistorischen
Beschaftigung mit den vielen namenlosen oder durchschnittlichen
Werken aus der Vergangenheit heute nicht mehr die
205
This content downloaded from
147.251.173.95 on Thu, 09 Mar 2023 11:29:27 UTC
All use subject to https://about.jstor.org/terms
gleiche zentrale Stellung ein wie noch vor wenigen Jahrzenten."
Ibidem, pp. 139-140.
9 lbidem, p. 138.
10 M. J. Bok, in his essay "Artisans or Gentleman Painters?: the
Social Background of Utrecht Painters in the Early Seventeenth
Century," presented at the 75th Annual Meeting of the College Art
Association in Boston (1987), mentions that in Utrecht, between 1611
and 1625, 36 masters were enlisted in the Guild of St. Luke as painters,
only 16 of whom are known from existing works. My guess would be
that this figure is typical, and that about half the documented artists
living in Western Europe between 1400 and 1900 are artistic
unknowns. This ratio is in strong contrast with that between named
and unnamed surviving works from that period, as encountered in
museum and auction catalogues. That ratio is closer to ten assigned
works for each anonymous one.
11 H. Hahn, The Rape of La Belle, Kansas City (Frank Glen Publishing
Company), 1946, p. 103.
12 In his essay in the same issue of this publication, entitled
"Connoisseurship as Practice," David Ebitz actually embraces the
ghost, praising connoisseurship as an instance of irrational knowing by
recognition rather than analysis. I would be not as content as he seems
to be to regard scholarly work as a universal physiological process
attended by rationalization after the fact.
13 0. Benesch, "Worldly and Religious Portraits in Rembrandt's Late
Art," Art Quarterly, 19, 1956, pp. 335-56, p. 348.
14 W. R. Valentiner, "The Rembrandt Exhibitions in Holland,
1956," Art Quarterly, 19, 1956, pp. 390-403, p. 400.
15 S. Slive, "An Unpublished Head of Christ by Rembrandt," Art
Bulletin, 47, 1965, pp. 407-417, p. 415.
16 A. Bredius, Rembrandt: the Complete Edition of the Paintings,
revised by H. Gerson, London, 1969, p. 614, no. 628.
17 K. Bauch, Rembrandt Gemalde, Berlin, 1966, no. 229.
18 J. Bruyn et al., A Corpus of Rembrandt Paintings (Rembrandt
Research Project), vol. 1, The Hague etc., 1982, no. C 23.
19 Ibidem, no. A 21.
20 Ibidem, vol. 2, 1986, no. C 76.
21 Ibidem, no. A 80.
22 Ibidem, vol. 1, 1982, p. 584.
23 Ibidem, vol. 2, 1986, p. 794.
24 N. Goodman, "Art and Authenticity," The Forger's Art: Forgery
and the Philosophy of Art, Berkeley etc., 1983, pp. 93-114, pp. 93-94.
25 Ibidem, p. 98: "Nothing depends here upon my ever actually
perceiving or being able to perceive a difference between the two
pictures." In Goodman's example, the knowledge that one painting is
an original and the other a copy is given. His observations, however,
apply mutatis mutandis to a situation in which the status of both works
remains to be established.
26 Ibidem, p. 105.
27 Attempts to describe systematically the characteristics of a
master's style and the criteria by which to measure them, such as
Maurits van Dantzig's "pictology," are avoided by practicing
connoisseurs. The main reason, I believe, is that any such
systematization inevitably leads either to attributions which are
unconvincing to the eye, or to the introduction of compromises in the
system itself.
206
This content downloaded from
147.251.173.95 on Thu, 09 Mar 2023 11:29:27 UTC
All use subject to https://about.jstor.org/terms