Skip to navigation – Site map

 « Leave it or take it away »: ethical considerations on the removal of overpaintings

The case of the Ghent Altarpiece
Maximiliaan P.J. Martens


During the current conservation of the Ghent Altarpiece, carried out by a team of KIK-IRPA, a large amount of old overpaint has been discovered. These are studied with a number of innovative analytical techniques. It is argued here that the introduction of these techniques allows for a re-evaluation of the conservation theoretical and ethical dimensions of coping with such phenomena.

Top of page

Full text

1Throughout his career, Roger Marijnissen has been a leading figure in the professional ethics of art conservation. Often his points of view were way ahead of his time and what he advocated decennia ago not seldom against a wall of miscomprehension, is considered as standard practice today. However, Marijnissen usually concentrated on specific case studies and never attempted to write a comprehensive theoretical treatise, like Cesare Brandi or more recently Salvator Muñoz Viñas or Barbara Appelbaum. He often preferred to publish in the local press in Belgium, ‘kicking people consciousness’ as Louis Paul Boon, another controversial Belgian author, used to say. He has a preference of stating his views in powerful one-liners, like ‘the best treatment is the one that never has to take place’, inviting his audience to vivid discussions.

  • 1  Session of the advisory board of 2 December 2013.

2Discussions are at the core of the decision-making process around the conservation-restoration treatment of the Ghent altarpiece, carried-out by a team of the Royal Institute for Cultural Heritage (KIK-IRPA).About a year ago, during a meeting in the conservation studio in the Ghent Museum of Fine Arts where the panels of the Ghent Altarpiece are being treated, the advisory board (that gathers all stakeholders) was informed on the extensive layers of old overpaint and retouching that had been discovered by the conservators throughout the exterior wings of the altarpiece1. While the people present were watching with astonishment the small windows made in the overpainted coats of Joos Vijd and Elizabeth Borluut, Roger asked me ‘afblijven of afnemen’? (leave it or take it away?).

  • 2  The overpainting removal tests and comparisons with eyckian masterpieces was presented by the cons (...)

3It should be mentioned here that by that time the independent international expert commission had advised in a first meeting to remove all varnish layers, even those that remained untouched during the last major conservation campaign in 1950-512. As a result and as the work progressed, much of the surface damage and (discoloured) overpaint dissimulated before by oxidized varnish had become visible. In a second meeting, this same expert committee formulated almost unanimously the advise to uncover the original paint layers by the Van Eyck brothers, on the condition that the whole stratigraphy of overpaint was characterized and dated as precisely as possible and that removal could be effectuated safely.

  • 3  B. APPELBAUM, Conservation Treatment Methodology, Amsterdam (…): Elsevier, 2007.

4The theoretical point of view that comes to mind is what Appelbaum called recently ‘the ideal state of conservation’, and the intricately entangled problems this implies.3 In this article I attempt to evaluate this theoretical notion in the context of the present conservation treatment of the Ghent Altarpiece.

5On the other hand, I would like to argue that good practice based on professional ethics is a dynamic rather than a static phenomenon, that may or even ought to be adapted when parameters change, e.g. through the introduction of new technology. New scientific findings or technological means, e.g. the current possibility of visualising the condition of the original layers underneath the overpaint, contributes effectively to the decision-making process and risk assessment, and thus invites to question previously established practice.

  • 4  APPELBAUM 2007, 171.
  • 5  APPELBAUM 2007, 173.
  • 6  APPELBAUM 2007, 193.
  • 7  APPELBAUM 2007, 179.

6To Appelbaum, establishing the goal of treatment comprises two methodological steps: 1. choosing the ‘ideal state’, and 2. determining a realistic treatment goal.4 The (art) historical meaning is the foremost determining factor in the former, while the state of preservation plays a role only in the latter. On her concept of the ‘ideal state’, the author explains that it is “by definition, the state that best embodies the object’s values (and it) must be one of the object’s actual historical states (…) No one period in an object’s life is self-evidently the most important one (…)”.5 However, she realizes that “… The ideal state is ideal in two ways: it is intrinsically a theoretical construct, and it is often unattainable…”.6 The author further states that “… A treatment goal should come from historical fact, not wishful thinking or personal bias… (it) must be guided by knowledge of the creator’s taste (…), by the aesthetic preferences of the period.”7

  • 8  A. DE SCHRYVER & R. MARIJNISSEN, ‘Histoire matérielle’, in P. COREMANS, L’Agneau mystique au labor (...)

7Although it is hard to disagree with these guiding principles, they do raise a plethora of questions. Admitting that it would be possible to identify all of an object’s values (historical, aesthetical, symbolic, functional, …), and realizing that all of them have most probably shifted in the course of time, how and on what basis can we attempt to prioritize them? And, assuming that such an endeavour is possible, how are we supposed to validate our conclusions? Moreover, this exercise becomes increasingly complex in the case of a work of art that assumes a prolific place in the history of art, like the Ghent Altarpiece. Its known history is already intricately complicated, and by definition, can be uncovered only partially.8 Created for a prominent childless couple to commemorate their souls in afterlife, during later centuries the Ghent Altarpiece was mostly admired and taken great care of for its (mistaken) status as the first oil painting in history. According to a well-known legend, Jan Van Eyck, one of its creators, was credited with the invention of this painting technique that continues to exercise its influence up to this very day. After Germany returned the side panels to the Belgian state in 1920, as stipulated in the Treaty of Versailles, the altarpiece gained stature as a symbol in geopolitical relationships. Trying to prioritize these many historical ‘values’ is impossible, as it inevitably would imply non-objective judgments in what is important to whom at any given moment in history.

8Other problematic issues in Appelbaum’s presumptions are the knowledge of the creator’s taste or intentions and the aesthetic preferences of the period of origin. With respect to a time when ego-documents and even written witnesses of art theoretical reflection are almost entirely lacking, the only historical sources at our disposal are the works of the artist themselves, and by extension those from the same artistic environment. Again, non-objective judgment comes into play, now in the form of connoisseurship. Although irrefutably crucial in any art historical assessment, certainty –in a scientific epistemological sense– cannot be obtained. Moreover, it is self-evident that the intent of the artist can only be based on interpreting the original state. This is impossible in the case of the Ghent Altarpiece, where different layers of later overpaint obscure the original paint layers. Returning to the original question, whether or not to remove these later additions, the prerequisite to comprehend the artist’s original intent leads to a perfect circular reasoning.

  • 9  APPELBAUM 2007, 183.

9Another problem is the condition that the ideal state “must be one of the object’s actual historical states”. As Appelbaum points out herself, cracks and other irreversible phenomena of ageing prevent by definition to re-establish the original state.9 But there is more. Some aspects of the original appearance, e.g. the degree of gloss of the original varnish, are impossible to reconstruct. A faithful re-enactment of the display of any old work of art, even when still preserved in its original setting, would be inacceptable where lighting conditions are concerned (be it dimmed daylight or candle light), for contemporary viewers and with respect to safety and preventive conservation prescriptions.

10Does this exclude a priori an attempt to return to the original state? Is the only remaining option, then, to “leave it”?

  • 10  P. PHILIPPOT, ‘Restoration from the perspective of the humanities’, in N. S. PRICE, M. KIRBY TALLE (...)

11It turns out that a comprehensive theoretical approach, pretending to be applicable in every case, does not provide enough solid ground in a complex decision-making process. A reasonable degree of certainty and avoiding circular reasoning is the least one may expect from an approach in contemporary conservation practice, which has become inextricably dependent on a scientific paradigm. Nevertheless, some of Appelbaum’s ideas are useful, be it in what I would like to propose here as subsequent steps in decision-making. After all, the core of the problem of restoration remains respecting history while at the same time recovering the artistic integrity of the object, which is an almost impossible task, as Paul Philippot stated two decades ago.10

12Before considering removing later interventions such as overpaint and retouching, it is of utmost importance to understand the reason(s) why they were applied in the first place. Only a very careful and detailed study of the work’s material history allows for a better understanding of earlier interventions, their historical importance, their condition and their impact on the work.

13First of all, one needs to exclude with absolute certainty that what is considered overpaint (or retouching) are not pentimenti, and thus deliberate changes applied by the original artist within the artistic creative process. Although this distinction seems obvious, it isn’t in the case of very old interventions within a complicated stratigraphy as that of the Ghent Altarpiece. Only when microscopic examination clearly shows that age cracks are overpainted, one can be sure that one deals with a layer applied several decades after the original one had already been subjected to a physical aging process.

14When overpaint was applied to cover an older deteriorated state, it is necessary to question whether no other motive provoked its application (see further). Although covering damage offers, of course, sufficient explanation for its presence, one cannot exclude other additional reasons by definition.

15If the overpaint (or reconstruction) is deteriorated and visually disturbing, it will affect the final result of the conservation treatment. Action is needed in this case; removal or improving the integration are the only options. The discoloured and weak reconstruction in the proper left sleeve of the grisaille with Saint John the Evangelist is a good example.

16A following issue is to determine the extent of the damage of the original layer. Especially radiographies and IRR-documentation are helpful in this endeavour. Their interpretation is hampered by features and artefacts characteristic of such technical documents. Interference of a cradle, scattering, areas with low density (absence of heavy metals), etc. limit the information provided by X-radiographs. Retouching shows up in IRR’s only where clear differences in penetration can be observed, usually by the presence of carbon-containing materials.


17However, with the introduction of MA-XRF, the damage of the original layer can be far better estimated than ever before.11 Filtering on specific chemical elements in the dataset provides a clear visual image of their spatial distribution. It compliments the information gathered from traditional XR- or IRR-examination.


18Moreover, recent experimental results provided by crack density maps are promising in determining the extent of overpaint over an undamaged original layer (high density) or the presence of lacunas (low density).12

19In other words, technological innovation has a considerable impact in this phase of the decision-making process. However, this information needs to be interpreted by referring to the careful visual examination of the works of art themselves.

20When the original layer is heavily damaged and the superimposed overpaint in good condition, optimal integration of the latter is to be preferred over extensive new reconstruction. However, if the original layers are in fairly good condition, the question ought to be asked whether they can be revealed safely. In the case of the external wings of the Ghent Altarpiece, much of the overpaint is applied on one or more layers of old varnish, which protect the original layers. If the risk assessment is negative, removal of overpaint should not be attempted at this stage. One should be conscious that future treatment methods might eliminate the risks faced today to affect the original.

21But even when overpaint can be removed safely to uncover the original work, one last question remains: is the added value of the original’s artistic quality considerable enough? To my mind, this is the trickiest question in the decision-making process. On the one hand, it relies on aesthetic judgment, subject to differences in opinion and known to change over time. On the other, also the art historical value, i.c. its uniqueness needs to be estimated. Here, traditional art criticism and art historical connoisseurship, of which the problematic epistemological status in a scientific paradigm has been pointed out already, remains the crucial method. The thorough understanding of an artist’s stylistic idiosyncrasies and paint handling must be the main guidance. By unveiling the original draperies of the patron’s robe, their three-dimensional illusion, subtle light handling and more complicated configuration will be recovered (at the time of writing, this work is in progress). That the overpainted draperies were admired by many generations, mistaken for centuries as original, does not outweigh the stunning recovery of what every trained connoisseur recognizes as characteristic for the master’s proper handling. In the particular case of the Ghent Altarpiece, another argument plays. Within the artistic canon established in historiography over centuries, Jan van Eyck’s artistic superiority is irrefutable. This counters any individual subjective aesthetic judgment.

22As a preliminary conclusion, overpaint can be removed if it can be ascertained that it was meant to cover a (not all too) deteriorated state of an original layer, and when it has deteriorated itself, is visually disturbing, and cannot be sufficiently integrated, or if the original is in a fairly good condition, the superimposed layer can be removed safely, and if the artistic quality of the original sufficiently adds to the value of the final result of the treatment.

  • 13  see forthcoming publications.

23In any case, removing older additions affects the work of art in its function as historical witness. Therefore, akin to archaeological methodology, they need to be documented, materially characterized, and dated as precisely as possible.13 When this historical value is considerably significant, however, removing overpaint (i.c. historical relevant material) is to be avoided.

  • 14  M. J. FRIEDLÄNDER, Altniederländische Malerei, I:  Die Van Eyck – Petrus Christus, Leiden: A. W; S (...)

24The overpainted areas in the Ghent Altarpiece discussed thus far are all applied on original layers that had aged before. It must be stressed that this excludes, as mentioned earlier, that we are dealing with pentimenti, i.e. Jan van Eyck correcting his own work (or that of studio assistants). Such corrections would provide insight in Jan’s creative process. It specifically also excludes the possibility that we would be observing Jan correcting the work of his older brother, Hubert.  As is well known from the famous inscription, Hubert is considered to have started the work, while his younger brother, Jan, completed it. Although the discussion among art historians on the separation of the hands of both brothers has been debated hotly since the inscription was discovered in 1823, not one single irrefutable material trace of Hubert has been found thus far. As Max Friedländer wrote more than 80 years ago, the problem still remains one of the most frustrating issues in art connoisseurship.14 If such a trace is to be found, it would allow to reconstructing (partly) the division of labour in the altarpiece. However, everything discussed in this paper has nothing to do whatsoever with this long-standing art historical debate.

  • 15  COREMANS 1953, 109-112.

25Another instance of a (possible) later addition in the Ghent Altarpiece is the tower of Utrecht Cathedral in the Adoration panel. This has been considered often as an addition that Jan van Scorel applied at the occasion of the restoration he effectuated with Lancelot Blondeel in 1550.15 Although it has not been possible to verify this hypothesis thus far –here the point would be proven if the tower does run over a cracked underlying layer–, it would be inconceivable to remove such an important historical testimony.

  • 16  R. SPRONK in M. P. J. MARTENS (ed.), Bruges et la Renaissance: de Memling à Pourbus, vol. II : Not (...)

26In 1550, the city council of Bruges ordered Pieter Pourbus to overpaint monks and other clerics in a hell scene on a Last Judgment (Bruges, Groeningemuseum), painted 25 years earlier by Jan Provoost for City Hall.16 This command was a result of a decree by the Emperor Charles V of 29 April 1550 that prohibited scandalous representations of clerics. In 1956, Pourbus’ work was removed with the exception of a small detail, although the documents concerning this overpaint were published and known since 1861. Obviously, such an intervention would not be considered anymore. Pourbus’ overpaint testified of religious tensions and changing ideology around the middle of the 16th century.

27However, an overpaint being the result of changing ideology or ethical standards, is not always a sufficient reason to maintain it. ‘Repeints de pudeur’, like leaves or cloths over erogenous zones, are removed systematically. Are witnesses of religious troubles of the 16th century historically more significant than those testifying of Victorian morals? Stated this way, it seems we have reached a deadlock in professional ethics. Yet, a work of art is something else than a written document; it cannot be judged exclusively as an historical testimony. ‘Repeints de pudeur’ may well be documenting changed morality, often they are applied by anonymous, second-rate painters and do not seldom disfigure an original work of art.

28The art historiographical canon offers a reference, be it a tentative one. That it doesn’t allow measuring the relative artistic importance of Jan Provoost or Pieter Pourbus, is reason enough why the latter’s addition should not have been removed. But it does provide sufficient argumentation to unveil the work of a prolific artist like Jan van Eyck, under the conditions enumerated above.

29Roger Marijnissen repeatedly stated that he does not like to discuss with colleagues, but prefers to discuss with the works of art. A comprehensive theoretical framework is useful, but does not substitute the need to consider each work of art individually. Rigorous scientific methodology is indispensible to understand as fully as possible the material characteristics and the build-up of a work of art, as well as its degradation. But science alone cannot help us in deciding to ‘leave it or take it away’. The lab of the humanities is needed to consider carefully artistic, historical, symbolic, aesthetical and ethical values.

Top of page


1  Session of the advisory board of 2 December 2013.

2  The overpainting removal tests and comparisons with eyckian masterpieces was presented by the conservators to the International Commission for the Conservation of the Ghent Altarpiece  Museum voor Schone Kunsten, Ghent, 17 March 2014.

3  B. APPELBAUM, Conservation Treatment Methodology, Amsterdam (…): Elsevier, 2007.

4  APPELBAUM 2007, 171.

5  APPELBAUM 2007, 173.

6  APPELBAUM 2007, 193.

7  APPELBAUM 2007, 179.

8  A. DE SCHRYVER & R. MARIJNISSEN, ‘Histoire matérielle’, in P. COREMANS, L’Agneau mystique au laboratoire. Examens et traitements, Les Primitifs flamands. III. Contributions, II, Antwerp: De Sikkel, 1953, 21-68; H. DUBOIS forthcoming PhD, UGent.

9  APPELBAUM 2007, 183.

10  P. PHILIPPOT, ‘Restoration from the perspective of the humanities’, in N. S. PRICE, M. KIRBY TALLEY, A. M. VACCARO (eds.), Historical and Philosophical Issues in the Conservation of Cultural Heritage, Los Angeles: Getty Conservation Institute ,1996 (original ed.: ‘La restauration dans la perspectivedes sciences humaines’, Pénétrer l’art, restaurer l’œuvre. Hommage en forme de florilège, Bruxelles: Éd. Groeninghe, 1990); also quoted in S. MUÑOZ VIÑAS, Contemporary Theory of Conservation, Amsterdam (…): Elsevier, 2005, 67-68.

11  M. ALFELD, J. VAZ PEDROSO, M. VAN EIKEMA HOMMES, G. VAN DER SNICKT, G. TAUBER, J. BLAAS, M. HASCHKE, K. ERLER, J. DIK, K. JANSSENS, ‘A mobile instrument for in situ scanning macro-XRF investigation of historical paintings’, Journal of Analytical Atomic Spectrometry, 28/5, 2013, 760-767. The MA-XRF examination of the Ghent Altrapiece is carried out by the AXES group of Antwerp University , under supervision of  Koen Janssens and Geert Van der Snickt).

12  A. PIŽURICA, L. PLATIŠA, T. RUŽIĆ, B. CORNELIS, A. DOOMS, M. MARTENS, H. DUBOIS, B. DEVOLDER, M. DE MEY & I. DAUBECHIES, ‘Digital Image Processing of the Ghent Altarpiece’, Signal Processing Magazine, 2015 (in press).

13  see forthcoming publications.

14  M. J. FRIEDLÄNDER, Altniederländische Malerei, I:  Die Van Eyck – Petrus Christus, Leiden: A. W; Sijthoff, 1934, 87.

15  COREMANS 1953, 109-112.

16  R. SPRONK in M. P. J. MARTENS (ed.), Bruges et la Renaissance: de Memling à Pourbus, vol. II : Notices, Bruges: Hopital St. Jean 1998,   45.

Top of page


Electronic reference

Maximiliaan P.J. Martens, «  « Leave it or take it away »: ethical considerations on the removal of overpaintings », CeROArt [Online], HS | Juin 2015, Online since 27 May 2015, connection on 22 August 2017. URL :

Top of page

About the author

Maximiliaan P.J. Martens

Maximiliaan P.J. Martens is Professor of Art History at the University of Ghent since 2003 and member of the Royal Flemish Academy for Science and the Arts of Belgium (2010-). Before he was Associate Professor of Medieval Art History at the University of Groningen (1992-2003). He received his PhD from the University of California, Santa Barbara (1992). He is also the Director of the Ghent Interdisciplinary Centre for Art and Science and co-chair of the UGent-sponsored research project ‘Archeometry of the Ghent Altarpiece’. He published widely on Flemish Art of the 15th and 16th century and more recently co-authored papers on mathematical image processing techniques and analytical chemistry applied to the investigation of works of art.

Top of page


Licence Creative Commons
CeROArt – Conservation, exposition, restauration d'objets d'arts est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Top of page
  • Logo DOAJ – Directory of Open Access Journals
  • Les cahiers de