What Technical Art History Can and Cannot Prove
Infrared, X-ray, dendrochronology and pigment analysis are powerful and routinely oversold. Here is what each actually establishes.
In short
- Every technique is stronger at exclusion than at confirmation.
- Dendrochronology dates the tree, not the painting.
- Underdrawing distinguishes invention from copying better than it identifies a hand.
- Restoration contaminates almost every sample taken from a picture's surface.
Dendrochronology
Tree-ring dating measures the growth rings in a wooden panel and matches them against regional chronologies. On Baltic oak, the standard support for Northern European panels, it can be very precise. What it produces is a felling date range for the tree, from which a plausible earliest date for the painting is inferred, allowing for seasoning.
The inference is the weak point. A panel could be used decades after felling, and old wood was sometimes reused. Dendrochronology therefore sets a floor, not a date: it can tell you a picture cannot be earlier than a certain year, and almost nothing about how much later it might be.
Infrared reflectography
Infrared penetrates upper paint layers and reveals carbon-based underdrawing. Its evidential value is often misunderstood. It is only moderately useful for identifying a hand, because underdrawing habits are shared within a workshop.
It is extremely useful for a different question: whether the composition was invented or transferred. A drawing full of changes, hesitations and abandoned contours indicates a mind working out a picture. A drawing that is confident, continuous and exactly matches the paint above it suggests a design being copied from something that already existed. That distinction separates an original from a version more reliably than style.
Pigment and binder analysis
Sampling identifies the materials present, which can be checked against the date those materials became available. This is where forgeries usually fail, and it is why the technique is the backbone of exclusion.
The complication is restoration. A picture four hundred years old has usually been cleaned, retouched, relined and varnished, sometimes repeatedly. Modern material in a surface sample may mean a forgery, or may mean a competent conservator in 1963. Sampling therefore has to be located, not just taken, and the results read against the conservation record.
X-radiography
X-rays show dense materials, principally lead white, and therefore reveal structure: earlier compositions beneath the visible one, changes of mind, the weave and tacking edges of the canvas, and the pattern of losses and repairs.
Its strongest use is in detecting reuse. An abandoned composition underneath is very good evidence of a working artist rather than a copyist, because a forger has no reason to paint a picture he intends to hide.
Questions
Pigment analysis, because it can produce an outright anachronism. Everything else narrows possibilities without closing them.
Rarely on its own. It establishes date range, structure, materials and working method. Assigning those to a named individual is still an argument built with connoisseurship.
Not necessarily. Analysis is expensive and invasive, and is usually commissioned when there is a question worth the cost.