The problematic restoration of a conceptual light artwork
- This article is a translation of:
- La restauration problématique d’une œuvre lumineuse conceptuelle [fr]
Abstracts
The "Erlauf" artwork by Jenny Holzer is a LED sign part of a series made in 1998,and was defective. A preliminary study was carried out before its restoration, which raised the difficulties relating to this kind of object. This case has allowed an international collaboration between the artist's studio, technicians, and museums, where the conservator-restorer is in charge of ethical and technical coordination and managing sometimes opposing points of view.
Index terms
Top of pageOutline
Top of pageFull text
Introduction
1In the recent years, changes have been happening at an ever increasing rate, and we tend to lose our foothold. The obsolescence, far from decelerating, keeps growing as fast as innovation hits the market, not only in the fields of technology but on all stages of everyday consumption. The museum are the most vulnerable in front of those changes, facing unprecedented challenges when their mission of conserving the collections to ensure its permanency become affected. The obsolescence became its Nemesis. As the artworks have recently lost their old conception of “eternal objects”, as they were produced in the last centuries, we tend to ask more frequently the question of an “expiration date” when facing contemporary art.
2The limited lifetime of artworks containing industrial parts is a complex problem, which a quick understanding and recognition could be an easy solution, since preventive conservation of those objects is way easier as restoration choices are reduced as time passes and replacement parts become unavailable.
- 1 LED = Light-Emitting Diodes.
3This paper will give you a glimpse at the intervention process which followed the failure of a luminous artwork made by Jenny Holzer, Erlauf, made in 1998. This blue-LED1 sign is a deposit at the CAPC Museum of Contemporary Art of Bordeaux, and after years of exhibition, its lights were fading, one after the other (Fig. 1).
Fig. 1 : Jenny Holzer, Erlauf, 1998

Legend: Photography of the artwork, such as visible in 2016 at the CAPC of Bordeaux.
Crédit : Steib Olivier, CAPC of Bordeaux.
- 2 STEIB, O., « Restaurer l’art lumino-cinétique », CeROArt [En ligne], EGG 5 | 2016, mis en ligne le (...)
4This restoration follows our previous study2, which concerned the restoration of a lumino-kinetic work made by Nino Calos, where the artist was missing. This time, we will see how the artist, still active, was involved in the decision-making process and how this influenced the choices we have left. We will see how important the notion of scientific documentation is, and what are the consequences when it’s lacking. We do hope that this study can provide an additional incentive to raise the awareness regarding its crucial need, at a time when it’s more than ever threatenead.
Jenny Holzer work
- 3 BERBION P., BERNADAC, M., COUSSEAU, H., HOLZER, J., Jenny Holzer OH : exposition du 1er juin au 2 s (...)
- 4 DOCAM, « Jenny Holzer, UNEX Sign No. 2 (selections from "The Survival Series"), 1983-84 », in Guide (...)
5Jenny Holzer is a conceptual artist, born American, in 1950. She is still very active nowadays on the international artistic scene. Her work consists of series of writings, which can be placed chronologically in her artistic career. Her writings revolve around themes that concerns the artist, thus focus on the human condition: war and its atrocities, sexual crimes, violence against women, acts of racism. During her early career, Jenny Holzer experimented with different ways to show the texts she had written, her goal being to spread them in the public space. Very early she used alternative media to convey her meditations3. That way, the first texts of the Truisms series, made in 1977-79, were printed on T-shirts, or stuck to the walls all around Manhattan. Finally, they were displayed on the giant billboards of Times Square, in New York. The Survival series (1983-85) among others tests, were originally presented on very specific electronic magnetic-tape panels, called UNEX type (the DOCAM alliance carried out a remarkable study4 on the subject).
6It is clear that the technology used at one point is important to understand the chronology of her works, since it anchors them in a timeframe. The texts however remain alive for Jenny Holzer: she reuses them, recombines those years later on other supports, in other situations. If the medium and its presentation modes are of importance to the artist, because they will allow her to reach the greatest number of people, they remains for us the only support of the texts we can transmit. The works produced by Jenny Holzer are the unique combinations of a text and its support, with different scrolling speed, shapes, colors, selected by the artist with rigor and poetry. The example of Lustmord illustrates the richness, and intense semantics of the media chosen by the artist. This text, written in 1994, evoked the violence against women during the war, and Jenny Holzer had chosen to write them on the skin of people, whose photographs were then exposed.
7Erlauf is a text written after a stay in Austria, in 1995, in the city of the same name. There, the artist created a work for the memorial celebrating 50 years of peace after the end of the Second World War, inaugurated on May 8th (Fig. 2).
Fig. 2 : Jenny Holzer, Peace Memorial, Erlauf, Austria, 1995

Photography of the memorial, projecting a powerful beam of light after night has fallen.
Crédit : Museum der Friedensgemeinde Erlauf.
8The memorial consisted of a light beam, projected towards the sky, as a symbol of peace, while the text, engraved on the white stone pavement, reminding a funerary symbol (gravestones) and incarnating the permanence of the monuments, addressed the themes of suffering, the rapes, the horrors of war.
9Texts such as Erlauf and other works, were used again several times later by the artist, and especially in installations using LED signs. The Erlauf artwork made in 1998 used a blue LED sign, and was part of the Blue series (1993-1998), shown with other series at the Yvon Lambert Gallery in Paris, and the Monika Sprüth Gallery in Cologne. In this series we could find other text written before, such as Lustmord, Erlauf, Arno, aside from Blue.
Words fly away, writings remain…what about the medium?
10This is the main difficulty we faced during the study for its restoration. The work of Jenny Holzer is singular in its way, because it has a double temporality: on one hand you have the texts, which are timeless, because they can be reused, transferred from one period to another, according to the will of their author. On the other hand, you have their artistic media, which gives them an opposite chronological anchor. Indeed, since the text has no real link in time, apart from its date of writing or its composition, it’s still the support that will allow us to understand when the work is located in the artist’s production. Thus, the UNEX panels, using magnetic tape, were only used for the Survival series realized from 1983 to 1985. Since the years 2010, Jenny’s works are rather sculptural, and the panels tends to occupy more and more space. In this case, even if the texts are timeless for the artist, the object that was acquired by an institution or gallery has a temporality, an identity, and an inventory number.
11Thus this question of temporal anchoring appeared obvious to us, looking at the choices for the restoration of the LED panel in deposit at the CAPC. A fairly large number of diodes were faulty, affecting the reading clarity of the text. After a closer examination, we considered various options for the restoration, following an increasing level of intervention: the partial replacement of diodes and/or other components, the replacement of complete LED modules, and the full replacement of the modules and internal controller (i.e. the recreation of the work by the artist).
Preliminary study
12In order to get a precise idea of what we could do, it was necessary for us to proceed on the full study of this artwork, so we could grasp how it operated. Only then we would be able to provide sufficient, and realistic solutions in order to repair this panel the right way.
Description of the artwork
13The work is an electronic sign with blue LEDs, model EXL-3000, which was produced by Sunrise Systems®. This old model is no longer produced, and after contacting the manufacturer, support was no longer provided. This leads to the total obsolescence of the panel. This device is made of a black metal case, inside it are a display controller (which contains the text written by the artist and the settings of its presentation mode), and a power supply. On the front, the text of Jenny Holzer scrolls indefinitely on 8 LED modules. Each of these board is made of a matrix (an array) of 7x32 or 224 diodes, for a total of 1,792 diodes throughout the whole panel (Fig. 3). These LEDs light on and off intermittently to produce the characters (each diode being a pixel). The non-visible components which control this matrix are some shift registers, a couple of transistors for regulation, some resistors and low-pass filters, which we will not go into detail in this paper.
Fig. 3 : Detailed view of a module

One of the eight modules, with the LED matrix and the components used to drive it.
Crédit : Steib Olivier.
Reverse-engineering
14In order to properly evaluate the state of the device, we made a complete diagnosis of its technology and design, using all possible means to find out how it was working (which is also called reverse engineering). To do this, one can equip himself with a multimeter, datasheets you can find using the labels on most components, then find the path of the PCB traces (when the paint hides them, like in this instance, you can put a projector behind and observe them in transparency). This way, we have been able to find how the circuit was operating, where came the signal, the role of the outputs terminal, etc.
15The two types of electronic failures (Fig. 4) thus had three causes. The ghosting effect, a light artefact that occurs when a line remains half-lit while it should be turned off, is certainly due to a fatigue-related current leakage on a shift register (drain output). Properly functioning, it is supposed to grant or prevent the passage of current. Here it is unstable. The extinction of the diodes is linked to two problems: their age (these are end-of-life components which luminous efficiency has declined), and connection defects (cold soldering, mechanical fatigue, shock, vibration, etc.).
Fig. 4 : Detail of the different LED failures types

The ghosting effect, the dim-lit LED, and the dead ones.
Crédit : Steib Olivier.
16Reverse engineering allowed us to produce a complete documentation on the panel operating mode, and we made a list of the components used with their datasheets. This allowed us to locate obsolete components, and those that have become unavailable on the market. Almost all components are now obsolete since the late 2000s, and if most components are still available (in a limited stock though), the original blue LEDs have been impossible to find. The fast paced evolution of LED technology, which has made significant leaps in just a few years, has caused this type of diode to be unavailable in a very short time, each technology replacing the previous one. The final result is that we only have second choices available now. The 5 mm diffusing diodes currently produced are not of the highest quality, and they’re not identical anyway.
17The only reference we managed to find which had almost the same frosty effect and dome color, but was sadly ovoid, not round. The comparison of the original LED and this reference shows how much the old ones were worn out. At a same voltage, the brightness difference is astonishing (Fig. 5 and 6).
Fig. 5 and Fig. 6 : Comparison of the original LED (on the left) and almost identical new ones (on the right)


Using the same voltage (3.2V), the original LED struggle to light up, while the brand new one is shining really bright.
Crédit : Steib Olivier.
Contacting the Artist’s Studio
18As the artist was still active, we contacted her Studio in order to get their opinion on the different solutions we were considering.
19This contact was very productive, and we obtained interesting information about the history of the work, filling the voids in the documentation, which was lacking. The Studio Holzer has been working with LED panels for many years, and this was not their first case of a restoration, so we were able to have a fruitful exchange of mails during several weeks to answer our questions about the different ways to repair the work. Our wish was to restore the closest the appearance of the blue diffusing diodes used at the time, without having to modify the controller, which produces the text of the artist, and which was in perfect working condition.
20From that point, things became very complicated. The studio, contrary to our expectations, first advised us to do a full replacement, cheaper and more durable. As a museum, we wanted the less interventionist option for the collection, which was first to intervene only on the diodes level. Since the operation was heavy and uncertain (the replacement of the diodes was not as durable as replacing the whole board), the Studio offered us an intermediate solution, which was to reproduce the LED modules, in full compatibility with the original controller. They put us in contact with a company, Kiboworks®, having worked together for a long time on projects. This company, specializing in LED display solutions, had already designed the replacement circuit, properly following the PCB design of the old modules.
21After much discussion, we finally decided to use this intermediate solution, which preserves the controller and the original frame. The Studio supervised the production and quality control of the replacement parts, in order to ensure that it complies with the artist’s wishes. Moreover, this solution was economically more advantageous and logically more durable, since the replacement components are not worn out, which would have been the case if we had chosen to repair the old circuit (desoldering and replacing 1792 diodes and more than 40 other components isn’t exactly a safe and easy solution, and very long).
22The main drawback of this solution was that, as the LED technology had evolved, the diffusing diodes became unavailable and were replaced by the new transparent blue LEDs. Those are more powerful and durable (Fig. 7). The aesthetics of the latter being substantially different, being the color of the diode or the shape of its light beam, we were puzzled as to use them or not. After questioning the Studio on this subject, we were told that the artist was in full agreement with this replacement using different diodes, as her artistic intention was precisely the use of the latest light technologies to spread her texts. On the other hand, they still had the right to reject the work that we would have modified without their control and consent, if it did not collide with the quality and brightness that they wished.
Fig. 7 : Comparison of both shapes and intensity of two LED types

The diffusing diode emits a steady halo, while the transparent diode has been designed to maximize its light output, producing an intense main beam in front of it, and a secondary beam to light the sides
Crédit : Steib Olivier.
23In order to respect the homogeneity of the collection, restored this way and historically by the Studio through several institutions, and in order to respect the intention of the artist, who expresses the wish for this restoration, we were therefore forced in some way to choose a solution that was ultimately far away from our original wish, which was the respect of the aesthetics of the work, in its historical dimension.
24As a conceptual work, where language prevails over its form, the restoration of the reading clarity was certainly a priority. However, it was with a certain bitterness that we had to give up before obsolescence, and had to settle for the good, rather than the best.
Conclusion
25The restoration of the reading clarity and the brightness of the panel made it possible to re-establish the artist's intention, which uses in her work the latest technologies to show her texts, deliberately very bright.
26The appearance before and after intervention was significantly altered (Fig. 8), especially due to the change in technology, but also to the fact that the original diodes had reached an almost complete extinction point. Of all the colored LED produced and commercialized, blue LED technology has always been the most intense in terms of light output. Regarding the effect, these LED panels in their prime, optimal performance, should have been close to the appearance they have today. The difficulty lies in the eyes, accustoming to the reduced luminosity: as the diodes lose their brightness by only a few % per year, it is not easy to appreciate their evolution without putting them in contrast with fresh ones. Most lamps have a higher brightness for a few months, before beginning to slowly fade over time.
Fig. 8 : The artwork after intervention

The artwork is bright again, which allows us to read the text more clearly. However, the sign emit a lot more light in front of it than on its sides, which modify its perception.
Crédit : Steib Olivier, CAPC of Bordeaux.
27The lessons we can learn from this experience are numerous. We have been able to measure the weight of the artist's influence in the decision-making process, when this one is still alive and active. The restorer is usually only an intermediary, only there to interpret the acts of the artist, not to replace him. In this case, the restoration, which produced a highly contrasted result, remains legitimate as it honors the artistic intention. Unfortunately, the authenticity of the tangible part is affected, at the expense of the intangible part.
28Maybe the difficulties related to the involvement of the artist in the restoration process lie in a series of shortcomings? The lack of information concerning the physical identity of the work, for example: where does the artistic intention reside? Is it in its brightness, its shape, its speed, or the assembly of those three? Without any kind of document providing proper explanation and attesting to the importance or non-importance of the tangible parts in the definition of the artwork, the words of the artist himself remains the most legitimate support to rely on. In the present case, the brightness of the Erlauf panel have to be strong, thus the fading, defective panel, was in fact the undesirable appearance of the artist's text. However, there is no clear mention of it anywhere. Perhaps it would have been necessary to replace the LEDs when their luminosity was decreasing, in order to maintain a high level of brightness, which would have avoided to create violent contrasts between two restorations? Thus, it raises the question of the financial (and ethical) capacity of this kind maintenance, which can be a very significant cost.
29We also found out that the Studio kept the right to disapprove the restoration of the artwork, if it wasn’t done according to their recommendation. Thus, the non-conforming restoration would in fact have been a fake, a new creation, as the restorer would have replaced the artist and modified the work against his will. This is not a position that can be assumed in our profession. This also mean there is some sort of cultural barrier between the artists and the museums. An artist can be in a constant process of creation, and has a right to recognize the work he signed. This is a moral right, which can collide with the patrimonial aspect given to an object that has joined a museum collection. Can this unique characteristic, defended by the people in charge of the collections, be enough to override the artist's right of recognition?
30The possibility to alter the work of an artist against his will could perhaps be defensible if the documentation concerning the work’s design was sufficient to reproduce the object identically, with a scientific thoroughness. Obviously, this would imply that its production was originally outsourced (ready-made), and only imaginable if the original component were still available to reproduce it.
31In this situation, if the artist were to wish to replace parts of his work with better materials (i.e. altering the history of his work), this modification would be unjustified, as the original technology would still be available. There would be no reason to modify the original appearance of the work as it was acquired. Therefore, this logic would favor respecting the cultural heritage dimension of the artwork, something which isn’t necessarily understood by the artists. In order to distinguish the notion of restoration from re-creation, it is required to have enough elements to prove that one can fully restore a known state, not to interpret it. If this is possible in theory, the reality is less obvious.
32In this instance, these elements were not known: the original materials, obsolete, couldn’t be found anymore. The scientific documentation concerning the lamps (the light output, angle of illumination, diffusion, colorimetric data, etc.) was lacking, but the artist was still active. Therefore we cannot go against her point of view, because we do not have any documents that can prove otherwise when the artist is defending his intent.
33This study highlighted once again the importance of an adequate, scientific documentation, and possibly the need to review the method of acquisition, in order to give museum institutions the ability to protect the cultural heritage by providing them with scientific legitimacy to defend the works. It also shown the importance of a certificate that describe clearly what the intent is, and how it’s shared to the public.
34If many questions have been asked and a large part of them still remains unanswered, raising them has allowed us to see the paradox of contemporary works at the time of their restoration, and to think about possible solutions to overcome it. Only time and hindsight will allow us to appreciate the value of our reflections and our actions. Meanwhile, being cautious and meticulous in our decisions-making process and interventions, remains the safest way to continue protecting our collections.
Notes
1 LED = Light-Emitting Diodes.
2 STEIB, O., « Restaurer l’art lumino-cinétique », CeROArt [En ligne], EGG 5 | 2016, mis en ligne le 04 mars 2016, consulté le 08 janvier 2019. http://journals.openedition.org/ceroart/4956
3 BERBION P., BERNADAC, M., COUSSEAU, H., HOLZER, J., Jenny Holzer OH : exposition du 1er juin au 2 septembre 2001, Bordeaux, capcMusée d'art contemporain, 2001.
4 DOCAM, « Jenny Holzer, UNEX Sign No. 2 (selections from "The Survival Series"), 1983-84 », in Guide de catalogage des collections nouveaux médias [En ligne], 08 January 2019. http://www.docam.ca/fr/introduction-aux-etudes-de-cas/unex-sign-no-2-selections-from-qthe-survival-seriesq.html
Top of pageList of illustrations
![]() |
|
---|---|
Title | Fig. 1 : Jenny Holzer, Erlauf, 1998 |
Caption | Legend: Photography of the artwork, such as visible in 2016 at the CAPC of Bordeaux. |
Credits | Crédit : Steib Olivier, CAPC of Bordeaux. |
URL | http://journals.openedition.org/ceroart/docannexe/image/5969/img-1.jpg |
File | image/jpeg, 324k |
![]() |
|
Title | Fig. 2 : Jenny Holzer, Peace Memorial, Erlauf, Austria, 1995 |
Caption | Photography of the memorial, projecting a powerful beam of light after night has fallen. |
Credits | Crédit : Museum der Friedensgemeinde Erlauf. |
URL | http://journals.openedition.org/ceroart/docannexe/image/5969/img-2.jpg |
File | image/jpeg, 56k |
![]() |
|
Title | Fig. 3 : Detailed view of a module |
Caption | One of the eight modules, with the LED matrix and the components used to drive it. |
Credits | Crédit : Steib Olivier. |
URL | http://journals.openedition.org/ceroart/docannexe/image/5969/img-3.jpg |
File | image/jpeg, 120k |
![]() |
|
Title | Fig. 4 : Detail of the different LED failures types |
Caption | The ghosting effect, the dim-lit LED, and the dead ones. |
Credits | Crédit : Steib Olivier. |
URL | http://journals.openedition.org/ceroart/docannexe/image/5969/img-4.jpg |
File | image/jpeg, 260k |
![]() |
|
Title | Fig. 5 and Fig. 6 : Comparison of the original LED (on the left) and almost identical new ones (on the right) |
URL | http://journals.openedition.org/ceroart/docannexe/image/5969/img-5.jpg |
File | image/jpeg, 96k |
![]() |
|
Caption | Using the same voltage (3.2V), the original LED struggle to light up, while the brand new one is shining really bright. |
Credits | Crédit : Steib Olivier. |
URL | http://journals.openedition.org/ceroart/docannexe/image/5969/img-6.jpg |
File | image/jpeg, 80k |
![]() |
|
Title | Fig. 7 : Comparison of both shapes and intensity of two LED types |
Caption | The diffusing diode emits a steady halo, while the transparent diode has been designed to maximize its light output, producing an intense main beam in front of it, and a secondary beam to light the sides |
Credits | Crédit : Steib Olivier. |
URL | http://journals.openedition.org/ceroart/docannexe/image/5969/img-7.jpg |
File | image/jpeg, 80k |
![]() |
|
Title | Fig. 8 : The artwork after intervention |
Caption | The artwork is bright again, which allows us to read the text more clearly. However, the sign emit a lot more light in front of it than on its sides, which modify its perception. |
Credits | Crédit : Steib Olivier, CAPC of Bordeaux. |
URL | http://journals.openedition.org/ceroart/docannexe/image/5969/img-8.jpg |
File | image/jpeg, 65k |
References
Electronic reference
Olivier Steib, “The problematic restoration of a conceptual light artwork ”, CeROArt [Online], HS | 2018, Online since 14 January 2019, connection on 15 June 2025. URL: http://journals.openedition.org/ceroart/5969; DOI: https://doi.org/10.4000/ceroart.5969
Top of pageCopyright
The text only may be used under licence CC BY-NC-ND 4.0. All other elements (illustrations, imported files) are “All rights reserved”, unless otherwise stated.
Top of page