The reverse of a painting often reveals elements that remain hidden from view at the front: the canvas, tacking margins, wooden support, old nail holes, repairs and sometimes a secondary supporting canvas. Together, these features can provide important clues about the painting’s original construction and later interventions.
Anyone who looks at a painting on canvas only from the front sees just part of the object. The reverse reveals how the painted surface is physically supported. Here we find the canvas itself, stretched over a wooden framework and secured with nails, tacks or staples. Small wooden keys may be present in the corners. Sometimes there is a second layer of fabric. Other paintings show old holes, repairs, trimmed edges or several generations of fastenings.
These elements can tell us a great deal about what has happened to a painting over time. A seventeenth-century painting may have been fitted with a new stretcher in the nineteenth century and lined in the twentieth. The present appearance of the reverse can therefore reflect several different moments in the history of the same object.
The canvas as the primary support
In a painting on canvas, the textile forms the primary support for the paint layers. Historically, linen was widely used, although other fibres occur depending on period, location and availability.[1] Little of this is usually visible from the front. From the reverse, however, the weave may be much easier to examine, unless the painting was subsequently lined.
There is often a temptation to estimate the age of a canvas from its colour, coarseness or general appearance. It is not that simple. A dark, coarse-looking canvas need not be particularly old, while a comparatively regular weave may have been produced centuries ago. Technical examination therefore focuses on more specific characteristics: weave type, thread count, seams, selvedges, tacking margins and possible evidence of commercial preparation.
In technical art-historical research, the structure of a canvas can even be analysed digitally. So-called weave maps record variations in the weave pattern. When two paintings display closely corresponding characteristics, this may suggest that their canvases were cut from the same larger piece of cloth or roll. This type of research has been applied, among others, to paintings by Johannes Vermeer.[2]
Strainer or stretcher?
The canvas is attached to a wooden framework. In everyday language, such frameworks are often referred to simply as stretchers, but technically there is an important distinction. A strainer is a fixed wooden framework. Its corner joints cannot be expanded to increase its dimensions. A stretcher, by contrast, has adjustable corner joints. Openings in these joints accommodate small wooden wedges or keys. By carefully expanding the construction, the dimensions of the stretcher can be increased very slightly.[1]
This distinction can matter when investigating the construction history of a painting. Expandable stretchers developed later than simple fixed strainers and became particularly common during the nineteenth century.[3] If a seventeenth-century painting is now mounted on a nineteenth-century stretcher, this tells us only that the present stretcher cannot belong to the original construction. At some point in its history, the painting may have been removed from its original support and restretched. This is by no means unusual in older paintings.
What are the wooden keys for?
Traditional adjustable stretchers often have small wooden wedges in their corners. These are generally referred to as keys or stretcher keys. There are commonly two at each corner. When a key is carefully driven further into the joint, the two stretcher bars are pushed slightly apart. The stretcher becomes fractionally larger and the tension of the canvas increases.[4]
This system is highly practical during the initial stretching and preparation of a painting. On an old painting, however, it is not a straightforward way of tightening a canvas that has become slack. When a canvas begins to sag, the obvious solution may seem to be simply to tap the keys further into the corners.
But an old painting does not consist of textile alone. On top of the canvas are sizing or preparatory layers, a ground, paint and possibly varnish. These materials have changed through ageing and possess different mechanical properties. If the tension of the canvas is suddenly increased, the paint layers may not respond without consequence.
There may also be several reasons why a canvas has become slack. The textile itself may have relaxed over time. The tacking margins may have weakened. Fastening nails may be missing. The stretcher may have become distorted. Fluctuations in temperature and relative humidity also affect the mechanical behaviour of the canvas and paint layers.
Why does the tension of a canvas change?
A stretched canvas remains under continuous mechanical tension. At the same time, the different materials within a painting respond to their environment. Wood expands and contracts. Textile responds to moisture. Sizing and ground layers behave differently from oil paint. As materials age, they may also become stiffer and more brittle.
Tension can therefore vary across the surface and change over the years. The way the canvas was originally stretched, the number of fastening points and any subsequent restoration treatments also play a role. A slight deformation is not automatically a serious condition issue. Some irregularities may be old and stable, while others may be associated with active change.
Can a painting be restretched?
Yes, although “restretching” can involve several different procedures. In some cases, the canvas can be reattached to an existing stretcher. In others, damaged tacking margins first need to be reinforced. An inadequate or distorted stretcher may need replacing. Particularly fragile paintings may require an entirely different support system. The objective is to provide stable support while placing as little mechanical stress as possible on the canvas and paint layers.
When two canvases are visible
One of the most striking situations on the reverse of a painting is the presence of a second canvas. This may indicate that the painting has been lined. In a traditional lining treatment, a second textile support is adhered to the reverse of the original canvas. The purpose was generally to provide structural support to an original canvas that had become weakened or damaged.[1]
A wide variety of methods has been used over the centuries. Paste lining was an important historical technique. Later treatments included wax-resin mixtures and synthetic adhesives.[5] Once a painting has been lined, something fundamental has changed in what we see from the reverse. The visible textile may no longer be the canvas on which the artist actually painted.
How can you recognise a lined painting?
In some paintings the answer is immediately apparent. A complete second fabric layer is visible on the reverse, and around the edges it may be possible to see how the different layers relate to one another. In other cases, the situation is far less straightforward. Not every secondary textile layer is adhered to the original canvas. Paintings may also have loose auxiliary canvases.
With strip-lining, only strips of material are attached to damaged edges so that the painting can be stretched again.[1] A complete old lining can, conversely, conceal a considerable amount of original information. The reverse of the original canvas is covered, and old inscriptions, repairs, seams and original fastening traces may no longer be visible.
A lining is not automatically a bad thing
Lining was a standard restoration practice for a long period. Many paintings were structurally reinforced in this way and may in part have survived damage and repeated movement because of it. At the same time, it has become clear that some historical techniques could also produce undesirable effects. The use of heat and pressure, for example, could flatten the relief of impasto.[5]
From the second half of the twentieth century onwards, conservators therefore became increasingly cautious about complete lining treatments. Modern conservation generally seeks less invasive solutions where possible. This does not mean, however, that every old lining should be removed.
The tacking margins can be more informative than the centre of the canvas
For technical examination, the edges of a painting are often particularly informative. They may reveal how the canvas was originally folded around the support. Old holes can mark previous fastening points. A pronounced fold may correspond to an earlier stretcher size. Damaged edges may explain why a strip-lining or complete lining was later required.
It may also become apparent that a painting was once reduced in size. If the canvas was cut down, the original tacking margins may have disappeared completely. In other cases, former folded edges were flattened out in order to enlarge the visible image area. The present dimensions of a painting therefore do not necessarily correspond to its original format.
Not original does not mean unimportant
Canvas, paint and ground belong to the painting’s original production phase, but the wooden support may have been replaced later. A painting may have been taken off its stretcher and reattached several times. Some of what we see today may therefore be centuries younger than the painted image itself. A lining may document a restoration campaign. Old fastening holes may preserve evidence of an earlier stretching arrangement.
When several physical clues correspond, the reverse can provide a remarkable amount of information about how a painting was made, supported, altered and conserved. Anyone who examines a painting seriously as an object therefore looks not only at what has been painted on its surface, but also at what physically holds the painting together.
Do you have a particular interest in this artist or in this subject?
Do you follow this artist, or have you come across a work you would like to know more about? We regularly research works appearing on the market and are always happy to exchange views on quality, provenance, art-historical context and previous auction results. You can contact us directly via info@antonello.art.
Notes and sources
[1] J. Paul Getty Trust, Conserving Canvas, “Glossary”, entries including canvas, key, lining, loose-lining, strainer and stretcher.
[2] C. Richard Johnson et al. (eds.), Counting Vermeer: Using Weave Maps to Study Vermeer’s Canvases, RKD Studies, The Hague: RKD – Netherlands Institute for Art History, 2017.
[3] Rijksmuseum, “The Back of the Night Watch”, Operation Night Watch.
[4] Canadian Conservation Institute, “Keying Out of Paintings”.
[5] Cynthia Schwarz, Ian McClure and Jim Coddington (eds.), Conserving Canvas, J. Paul Getty Trust.