Why Restoring Art Is Scientifically Hard

In August 2012, an octogenarian parishioner named Cecilia Giménez tried to touch up a flaking fresco of Christ in a small church in Borja, Spain, without consulting a conservator. Within days, images of the result, a smeared, furry-looking face that the internet immediately nicknamed "Monkey Christ," had circled the globe (BBC, 2012). The Ecce Homo disaster is often told as a joke about bad art, but it is really a case study in the central problem of restoration: deciding how much to touch a fragile, aging original, and living with the fact that some of those decisions cannot be undone.

What makes restoration hard is not the paint, it's the decision

Restoring art gets treated, in popular imagination, as a technical skill: know the right solvent, apply the right varnish, match the right color. But the technical part is the easier half. The harder problem is epistemic, meaning it is about what a conservator can actually know before acting. A conservator staring at a centuries-old painting has to decide what is original material worth preserving, what is later overpaint worth removing, and what is accumulated grime that is neither, all without a definitive test that settles the question outright.

This is why professional conservation training now runs five or six years, combining chemistry, art history, and studio practice, and why organizations like the American Institute for Conservation (AIC) require members to follow a formal code of ethics rather than simply a set of techniques (AIC, Code of Ethics and Guidelines for Practice). The code exists precisely because the technical steps, cleaning, consolidating, retouching, are the low-risk part. The judgment calls about where to stop are where irreversible mistakes happen.

Every restoration decision is really three decisions bundled together: what to remove, what to leave, and what to add back. Each carries a different risk profile. Removing discolored varnish is largely safe if done slowly with tested solvents. Removing what looks like overpaint is riskier, because "later overpaint" and "original passage the artist revised" can look identical under normal light. Adding new material, retouching paint, a replacement varnish layer, is where the reversibility principle becomes the whole ballgame, because it is the one category of intervention modern ethics insists must be undoable.

The Ecce Homo fresco in Borja, Spain, painted around 1930 by Elías García Martínez, drew roughly 45,000 visitors to the small town of around 5,000 people in the years after Cecilia Giménez's 2012 amateur retouching went viral, according to reporting on the site's transformation into a tourist attraction (BBC, [Spanish fresco restoration botched by amateur](https://www.bbc.com/news/world-europe-19349921), 2012).

The reversibility principle: conservation's foundational rule

Reversibility — the idea that any material a conservator adds during treatment should remain removable later, using a solvent or method that will not damage the original underneath it. This principle is the closest thing the conservation field has to a constitution, and its roots run through more than a century of debate about how much a restorer is allowed to change.

Early conservation theory traces to 19th-century figures like John Ruskin, who argued that restoration was effectively destruction disguised as repair, and Eugène Viollet-le-Duc, who took the opposite view that a restorer could and should complete a damaged work in the style its creator would have used. Neither extreme survived into modern practice. The synthesis came later, most influentially in Italian theorist Cesare Brandi's 1963 book Teoria del Restauro (Theory of Restoration), which argued that any conservation intervention should be both reversible and visually distinguishable from the original to a trained eye, so future viewers and future conservators can always tell what the artist made versus what a restorer added (Brandi, Teoria del Restauro, 1963).

That second half, distinguishability, matters as much as reversibility itself. A retouched area is often deliberately built from tiny vertical hatching strokes (a technique called tratteggio) or a slightly different color saturation, visible up close but blending at normal viewing distance. The goal is not to fool the eye permanently. It is to let today's viewer see a coherent image while leaving an honest, legible record for the conservator working on the same painting in 2075.

Is anything in conservation actually 100% reversible?

No treatment is perfectly reversible in practice, and most conservators will say so openly. Cleaning removes material permanently; even a fully soluble retouching paint sits on a support that keeps chemically changing underneath it, so the same solvent that worked cleanly in 2026 may behave differently on a more degraded surface in 2060. Reversibility functions as a discipline and a goal, a way of choosing materials that keep options open, rather than a literal guarantee (Appelbaum, Conservation Treatment Methodology, 2007).

Why amateur restoration and professional conservation diverge so sharply

The Ecce Homo case is instructive precisely because it shows what happens when someone skips every safeguard the reversibility principle is designed to enforce. Giménez used ordinary paint, not a conservation-grade retouching medium chosen for later solubility, applied directly over the deteriorating original with no documentation, no testing, and no plan for a future conservator to intervene (BU Sequitur, Grotesque Irreverence: The Transformation of "Ecce Homo", 2016).

A professional conservator approaching the same fresco would have started differently. First: diagnostic imaging and small-scale material testing, often including ultraviolet fluorescence photography to map prior overpaint and infrared reflectography to check for underdrawing, before touching the surface at all. Second: a condition report documenting exactly what exists, photographed and written, so the pre-treatment state is preserved as a record even if the physical surface changes. Third: reversible, conservation-specific materials, such as retouching paints in a synthetic resin medium like Paraloid B-72, chosen because they dissolve in solvents (acetone, for instance) that will not touch a cured, aged oil or fresco-bound pigment layer. Fourth: minimal intervention, meaning the conservator does the least necessary to stabilize and clarify the work, rather than the most that could aesthetically satisfy a viewer.

Approach	Goal	Materials used	Reversibility	Typical risk
Minimal intervention / conservation	Stabilize and clean without altering appearance beyond removing degradation	Conservation-grade consolidants and solvents, tested per painting	High	Low, but can leave visible losses unfilled
Compensating restoration	Fill losses and retouch so the image reads coherently at viewing distance	Reversible retouching media (e.g., Paraloid B-72 based paints), tratteggio or similar distinguishable technique	Moderate to high	Moderate, hinges on correct pre-treatment diagnosis of what is original
Full reconstruction	Recreate missing or destroyed sections to complete the artist's presumed intent	New paint layers, sometimes new support material, applied broadly	Low	High, risks substituting the restorer's judgment for the artist's
Amateur/uncontrolled retouching	Cosmetic fix, no formal training or documentation	Whatever is on hand, often non-reversible commercial paint	Effectively none	Severe, can permanently obscure or destroy original material

The further right a restoration philosophy sits on the intervention scale, the more it depends on the restorer correctly guessing the artist's original intent, and the harder any mistake becomes to undo.

The Ecce Homo case sits in the far-right column of that table: cosmetic, undocumented, non-reversible. It became internationally famous because the result was visually absurd, but the underlying failure mode, permanent alteration without a documented, reversible plan, is the same failure mode conservators worry about even when a botched job looks convincing rather than comic. A skillful but undocumented amateur repaint is arguably more dangerous than an obvious one, because nobody notices there is a problem to fix.

Some damage is reversible. Some categorically is not

The reversibility principle only works because different kinds of damage have genuinely different chemistry, and conservators need to know which kind they are looking at before choosing a treatment.

Yellowed varnish is the most forgiving case. Natural resin varnishes like dammar or mastic oxidize and cross-link over years, turning increasingly amber and eventually brown, which mutes and warms the colors underneath. Because the varnish is a separate layer physically and chemically distinct from the paint film beneath it, a properly matched solvent can dissolve the degraded varnish while leaving a cured oil paint layer intact, which is why cleaning is one of conservation's most routine and reliable procedures. Old overpaint from a prior, less careful restoration campaign is a related case: if that overpaint sits in its own layer above the original, cross-section microscopy of a paint sample from an inconspicuous edge can usually reveal the boundary, and the newer layer can often be removed.

Solvent damage to original paint sits at the opposite end of the spectrum. If a cleaning solvent is too strong, or if the original paint's binder has degraded in a way that makes it more soluble than expected, the same solvent action that safely removes varnish can also strip pigment particles or dissolve original glazes, thin, semi-transparent paint layers artists used for subtle color effects. That loss is permanent. There is no chemical process to reconstruct a glaze that has been wiped away, because the specific arrangement of pigment particles the artist created cannot be regenerated from what remains. This asymmetry, some damage is a solvable chemistry problem and some is a one-way door, is the single fact that makes restoration risk management so different from ordinary repair work.

Why can't conservators just test a cleaning method safely first, every time?

They do, but small-scale testing only reduces risk, it does not eliminate it. A solvent ratio that behaves safely on a spot test in one corner of a painting can still misbehave elsewhere if pigment density, underlying ground material, or the degree of paint degradation varies across the surface, which is common in large or heavily reworked paintings, so conservators proceed in small, monitored stages rather than trusting a single test to certify an entire treatment.

The Sistine Chapel and the limits of "successful" restoration

Not every controversial restoration involves an amateur. When conservators cleaned Michelangelo's Sistine Chapel ceiling frescoes between 1980 and 1994, removing centuries of soot, candle smoke residue, and old animal-glue varnish applied by prior restorers, the colors that emerged were far brighter and more saturated than anyone alive had ever seen them (Vatican Museums, Restoration of the Sistine Chapel, project documentation, 1980–1994). Some art historians and critics argued publicly that the cleaning had gone too far, stripping away shading that Michelangelo himself had added on top of the fresco once it dried, a technique called a secco retouching, rather than removing only later grime (Beck and Daley, Art Restoration: The Culture, the Business and the Scandal, 1993).

The dispute has never been fully resolved to universal satisfaction, and that is itself the point. Even a technically well-documented, professionally executed cleaning campaign, backed by chemical testing and a large team, could not produce a result that settled the underlying epistemic question: exactly how much of the pre-cleaning darkness was Michelangelo's intent versus four and a half centuries of soot. The project followed the reversibility principle rigorously for its added materials. It could not follow the principle for its removal decisions, because removing soot is not reversible in either direction: you cannot put the soot back to check your work, and you cannot prove after the fact that no original glazing left with it.

This is the deeper lesson the Sistine Chapel case adds to the Ecce Homo case. Amateur restoration fails because it skips documentation, testing, and reversible materials. But even flawless professional practice cannot make every decision safely reversible, because cleaning itself, the single most common conservation procedure there is, permanently changes the object. The reversibility principle governs what conservators add. It cannot fully govern what they take away.

What conservators do differently to manage irreversible risk

Because cleaning and other subtractive treatments cannot be made truly reversible, the field has built a set of practices around managing that risk rather than eliminating it.

• Documentation before, during, and after treatment. Detailed photography, written condition reports, and material analysis create a permanent record of the object's state, so even an irreversible physical change leaves a reversible informational record for future scholars.
• Staged, incremental treatment. Rather than committing to a full cleaning in one pass, conservators typically clean in small test areas first, then expand gradually while monitoring how the paint responds.
• Peer review and committee oversight for major works. High-profile restorations, the Sistine Chapel among them, are typically reviewed by outside scientific and art-historical committees precisely because a single conservator's judgment on an irreversible step is considered too much risk to leave unchecked.
• Minimal intervention as a default posture. When in doubt, the field's ethical guidance favors doing less rather than more, since an under-treated painting can always receive further conservation later, while an over-treated one cannot be walked back (AIC, Code of Ethics, guideline II).
• Non-invasive analysis before invasive treatment. Techniques like infrared reflectography, X-radiography, and multispectral imaging let conservators map what lies beneath the surface without removing any material at all, reducing how much irreversible sampling is needed.

None of these practices make restoration risk-free. What they do is shift the profession's error rate from "occasionally catastrophic and undocumented," the Ecce Homo pattern, toward "occasionally debatable but fully documented," the Sistine Chapel pattern. That shift, from unrecoverable mistakes to recorded, arguable judgment calls, is what separates conservation science from a paint-and-hope repair job.

What people get wrong about restoration risk

A common assumption is that a "successful" restoration is one nobody notices, meaning the painting simply looks clean and original again with no visible trace of intervention. Professional conservation ethics actually pushes against that instinct. Brandi's distinguishability principle and techniques like tratteggio exist specifically so a trained eye, not necessarily a casual viewer, can identify what was added, which means a technically excellent restoration is not one that erases all evidence of itself but one that is honest about where it intervened.

A second common misconception is that older materials are inherently more stable and newer restoration materials are the risky variable. In reality, the opposite is often true within a single object: 400-year-old paint that has already finished its major chemical changes can be more predictable to work with than a well-intentioned restoration from the 1950s, when conservators commonly used non-reversible synthetic varnishes that are now themselves degrading and difficult to remove without risking the original paint beneath. Some of the hardest conservation problems facing museums today are not damage from the original artist's era at all, but damage introduced by earlier, well-meaning 20th-century restoration campaigns that predate the modern reversibility standard.

A third misconception is that restoration decisions are primarily aesthetic judgment calls made by connoisseurs. Modern conservation is a materials science discipline first. Decisions about what solvent to use, how long to expose a surface, and which retouching medium to select are grounded in the chemistry of specific pigments, binders, and degradation products, tested case by case, because the same treatment that is safe on one painting can be destructive on another with a chemically different paint film.

The takeaway

Restoration is hard not because conservators lack good tools, but because the job requires making permanent, high-stakes judgment calls, what is original, what is grime, what is safe to remove, using a discipline built specifically to keep the added parts reversible even when the subtracted parts cannot be. The Ecce Homo fresco shows what happens when someone skips documentation, testing, and reversible materials entirely. The Sistine Chapel shows that even rigorous, well-reviewed professional practice cannot make every decision reversible, because the single most common conservation act, cleaning, is a one-way door by nature. The honest measure of good conservation was never "you can't tell it was restored." It is "there is a clear, documented, reversible record of exactly what was done, so the next conservator inherits a legible history rather than a mystery."

---

Read next:

• Why old paintings crack over time
• Why oil paint dries slowly
• Why fresco painting is so durable

FAQ

Q: What does "reversibility" mean in art conservation?

It means any material a conservator adds during treatment, adhesive, retouching paint, or varnish, should be removable later with a solvent that does not damage the original artwork underneath. It is a guiding standard formalized through 20th-century conservation theory, not an absolute guarantee, since irreversible actions like cleaning cannot truly be undone.

Q: What happened with the Ecce Homo "Monkey Christ" restoration?

In 2012, an elderly parishioner in Borja, Spain, attempted to fix a deteriorating fresco of Christ without conservation training, testing, or reversible materials, permanently altering it into an image widely mocked online as "Monkey Christ." The case became a global example of what amateur restoration risks compared to documented, reversible professional conservation.

Q: Why did the Sistine Chapel restoration prove controversial even though it was done by professionals?

Removing centuries of soot and old varnish from Michelangelo's frescoes revealed far brighter colors than viewers expected, leading some art historians to argue the cleaning removed original shading along with the grime. The dispute shows that even careful, well-documented professional cleaning cannot fully prove, after the fact, exactly where the boundary between dirt and original paint sat.

Q: Why is cleaning a painting considered irreversible even when the materials used are reversible?

The materials a conservator might apply afterward, like a new varnish, can be dissolved away later, but the discolored varnish or grime that was removed during cleaning is gone permanently. There is no chemical process that restores removed material, which is why cleaning decisions are treated as higher-risk than adding reversible retouching on top of an already-cleaned surface.

Q: How does a professional conservator's approach differ from amateur restoration?

Professionals begin with non-invasive diagnostic imaging and small material tests, document the object's condition in detail, and use conservation-grade materials such as Paraloid B-72 based retouching paints chosen specifically for later solubility. Amateur restoration typically skips testing, documentation, and reversible materials entirely, which is what turned the Ecce Homo fresco into an irreversible loss rather than a stabilized repair.