JMW Turner the source Marilyn Laver Lavinco CCI Canadian Conservation Institute Joyce Townsend conservation scientist Tate Gallery Joyce Townsend techniques of JMW Turner Tate Gallery Publications CCI  
                                                 Scientific Report on 'Shipwreck, the Rescue'
     This section possible through the professional efforts of Marilyn Laver, P. Eng               


Shipwreck, the Rescue
Black and White photograph of the full painting


Art World’s Dirty Little Secret
r malcolm setters / graham setters


"In the last few decades art-science has expanded along with our greater reliance on science and technology in general. This empiricism, although clearly crucial for our modern understanding of the world, has usurped much of the emotional basis on which our social fabric was once woven. It is a trend that will never reverse itself short of a second-coming, but will always have to share its present role with people’s deeper needs. Those that truly believe: science and religion, art and sport, Buddhism and Christianity do not have common ground, deceive themselves. It is also folly, to believe that we can make science our religion. Our inflexible commitment to a singular cause can only prove fatal to personal growth, adaptability, and cultural survival.

This need for a diverse outlook is vital when moving among grander ideas and philosophies. When talking about glowing sunsets based on glazing techniques, love of a mother based on the Christian Madonna, or movement in a picture based on its geometric orientation, one is left empty. We must be able to mate illusion with reality, imagination with knowledge and spirituality with the mundane in order to fully appreciate the world around us, and in particular, art.

Imagine yourself as a technician in a laboratory denuding a work of art by exposing its hidden secrets: like dissecting an angel. It is sacrilege, unless done by a ‘conservation scientist’ with at least the will to turn their ghastly earthly deed into something less carnal. Should they confess a deeper emotion than mere curiosity? Some do, some do not, and some pretend. Those that do, are enlightened, while those that do not, will only find ‘lead white’ in the fragile wings of Murillo’s angel or bitumen in the haloed grottos of el Greco.

     Shipwreck, the Rescue c1802 - The Report

1.      Physical Characteristics

1.1       Canvas

"The painting measured 166 x 224 cm (5.45’ x 7.36’). A comparison was done with sizes of canvases given by Butlin and Joll. The work under investigation was not a standard size but examination of the stretcher garland pattern around the edges of the work indicate that it had been cut down slightly in size at some point. There were stretcher garlands, which show up as an undulating pattern in the canvas weave, around the perimeter, but no nail holes. By comparison[2], several seascape works in the listing are slightly larger than this one:


                                         “Calais Pier”       (1803)        172 x 240 cm

                                         “Shipwreck”       (1805)        170.5 x 241.5 cm

                                         "Trafalgar”          (1806)        171 x 239 cm

                                         “Spithead”          (1808)        171.5 x 235 cm

                                         "Transport Ship” (1810)        172.7 x 241.2 cm 


"The original canvas of the painting was roughly spun (approximately 13 x 12 threads per centimeter) and had been joined horizontally across the middle. The rough weave is clearly visible in the raking light photograph; the weave patterns of the two pieces of fabric can be seen in the area of the join.[3]

"Three subsequent lining canvases have been attached to the back of the original canvas over the years, suggesting that the work is not of recent fabrication. The first lining canvas was likely applied when the painting was cut down to its current size and fitted to a new stretcher.[4] This first lining canvas was of finer weave than that of the painting canvas (over 18 threads per centimeter).[5]

 1.2        Stretcher


"The stretcher in “Shipwreck, the Rescue” consisted of four vertical members and three horizontal ones, keyed at the corners and at each join with the perimeter. By comparison, Townsend mentions that in the early years, many of Turner’s paintings had home made stretchers with the diagonal cross-members generally found in the 18th century.[6] However, this stretcher may have been built later when the painting was cut down so the evidence here is inconclusive. ---


" The label on the stretcher reading “Shipwreck – The Rescue   J.M.W. Turner R.A.” may also have been affixed at this time. The paper of the label was not a wood pulp or Japanese paper; hemp or bast (Linen), were possibilities (i.e. rag paper), neither of which are unique to Turner’s period."

eveled inner edges on the inside of the vertical stretcher bar

Author's Note:
It should be noted here; that the present stretcher is probably from Brown of High Holborn, a firm that Turner was known to patronize circa 1830-1850. It is the right age, and has the proper configuration and construction, including beveled inner edges of the stretcher bars. It is, in any case, the type he would have used once he started buying commercial supports after his father’s death in 1829.


Shipwreck, the Rescue - verso Arrows indicate shadows from the photography

 For two reasons, it could be concluded that the artist himself requested this painting to be put on a commercial stretcher only 50 years after it was painted. The first being: it is standard knowledge that Turner let his paintings lay waste in the damp moldy environment of his studio until he resurrected one at a later date, to sell, or exhibit after the proper repairs were carried out. Under normal circumstances, as with other artists, there would have been no need, so soon, for such a dramatic restoration on such a young painting. Secondly, the vintage label was surely titled and signed during Turner's lifetime. This might even have been done by the artist himself.[7]   


In keeping with this timing, Turner’s actions have been paraphrased in many ways: for example; “Following the publication of the first volume of Ruskin’s Modern Painters, with its spirited championship of his art, in 1843, his pictures were again in keen demand, and he was encouraged to put some earlier canvases into repair, and keep by him some finished work, to tempt the new purchasers who now came to the gallery.”[i]

      An observation should also be made regarding the ‘title label’ and its relationship to the stretcher. Even the seasoned wood of quality furniture, and in this case the stretcher, would naturally shrink further given enough time. The top on a small Georgian tea table can shrink as much as ¾ inch across one plain, leaving the top slightly oval. This happened with the stretcher wood also; the wood shrunk in one direction—across the grain. The reason this is relevant here, is that while the stretcher wood has shrunk over one and a half centuries, the paper label has not. The result is that the label has wrinkled, and in places has separated from the wood and chipped. This is significant because it almost certainly guarantees that the label is vintage to the stretcher and that both are from Turner's era.                                           
Another description of this label from 1990 by David Audet, a qualified Senior Appraiser with the International Society of Fine Art Appraisers, describes this stretcher and label as follows: it “retains patina that is very definite to the early eighteen hundreds,” and “The linen paper labels on the stretcher are written in quill, fixed on the frame with glue. In other words both the frame and labels are of the same time frame. Circa 1830.” [ii] He uses the plural (labels) because there is also, what is probably an inventory-number label attached to the stretcher, number B 514.

David Audet felt that the painting should be attributed to J.M.W. Turner and noted, “Turner has signed and named (the) painting… on a piece of linen paper size 3 x 5 inches.”

The hand lettering has not been compared forensically because sufficient suitable reference samples of Turner's work which has been done under similar circumstances have not yet been identified.

1.3 Technique 


Pentimenti:  design change seen in one of the x-radiographs from Shipwreck, the Rescue

Laver's report continued:

A detailed analysis of the composition and technique is properly the province of a curator. However some observations were made during the examination, which can be briefly discussed. The work was painted very thinly, as an unfinished work would be; some details are painted so thinly as to be transparent, revealing the underlying composition and suggesting pentimenti or design changes."

x-radiographic image. Note the angular sail: bottom left and extending half way up the image. In the finished picture it has been painted out and replaced with one that is already half collapsed. Here the front mast is prominently seen running through the original design at an appropriate angle (see the complimentary sketch from the Wilson Sketchbook.

           "Any original impasto has possibly been flattened and lost during the sequence of linings, with the result that the main surface contours are now due to the threads of the coarse canvas.

    "Wide brush strokes were used in some features, for example the pier structure, with no subsequent detailing; this also suggests an unfinished work or sketch. Hairs from a brush were identified as camel hairs. Townsend notes that Turner used sable or camel brushes to finish his works, following the initial lay-in with a stiff brush such as hog’s hair.

    "In this work the ground is visible in some areas of the composition, giving the work an unfinished appearance. This is consistent with Townsend’s observation that patches of ground were generally not left exposed in his exhibited works.

   "Examination of the cross-sections showed the presence of 1-5 paint layers over the ground, and most often 2 or 3. Many of the pigments are coarsely ground, a characteristic observed in Turner’s works (see below: electron-microscopic cross-section of paint layer).


   "The x-radiographs and examination in normal light showed little or no evidence of cracking in the paint layers, which might be expected on lengthy storage of an unfinished work (particularly if the work includes components such as egg-based priming or bitumen in dark paint areas)."[9]


The stability of Turner’s earliest oil paintings has been noted by one of Turner’s chief biographers, Philip H. Hamerton: “Turner began by painting soundly in oil; his unsound work belongs to his full maturity. We know on excellent authority—the authority of the owner—that the portrait, which Turner painted of himself, at the age of seventeen, is still in a perfect state of preservation. The pigments used in that portrait, and the manner of applying them, were alike in accordance with the Academic tradition of the time.”[iii]


James Orrock, a notable artist and Turner expert from Hamerton’s era gives the same evidence from a more personal perspective, “Turner’s early oil pictures as a rule are not cracked, and therefore as a rule the desire to produce the brilliancy and aerial qualities of water-colour in oil induced him to prepare a white ground to ‘opalesce’ his colours on, but unfortunately, as we know, the pigments did not always adhere to the ground. This I have proved from careful observation, especially of a picture of the kind possessing the defects in question which is in my own collection.”[iv]

In fact more than a century ago Orrock gave a rather explicit—almost scientific description of what was going on beneath the painted surface. On the critique of Barnes Terrace painted 1827, he writes: “Turner is now resolved to rival in oil the brilliancy and light of his watercolours. This he feels assured can only be affected upon a thickly-prepared white ground, as pure as paper or as porcelain. He prepared the grounds himself, and frequently painted on them in water-colour impasto. He painted, as in watercolour, chiefly with body-colour, and finished up with oil-colour and medium. Several of these delicious pictures have suffered from being exposed to dry-heat galleries; detached pieces here and there falling from the canvas.”[v]


Shipwreck, the Rescue is certainly very early and therefore is expected to be in better condition than Turner’s mature work at a point when he was experimenting outside of the classic tradition of material usage. And particularly when experimenting with those white grounds that he would use for the rest of his oil painting career. 


Ultra violet photograph of Shipwreck, the Rescue.
The milky fluorescence indicates the predominance of unadulterated antique varnish

 "The ultra violet photograph shows the condition of the varnish layer(s). There is evidence of an overall coating, which shows a milky fluorescence, consistent with the presence of an aged varnish. Dark areas in the photograph indicate alterations to the varnish, including re-touching and some cleaning. It is to be expected that a painting from the 19th century would likely have undergone conservation treatment. It is unlikely, however, that Turner himself would have applied a varnish coat to an unfinished work; any varnishes present probably postdate the artist.”

The level of finish of Turner’s work generally, is of course a contentious issue. It appears that in some cases he meant to leave works permanently unfinished. If it is true that Turner had given this painting to his executor; re-stretched, touched up, lined, with a tile label and ready to hang, this would indicate that under certain circumstances he considered works such as this, complete. While highlighting Turner’s popular yet very thinly painted, The Evening Star, John Golding supported this same line of reasoning: “Traditionally regarded as unfinished, the painting has about it a perfection that suggests Turner was satisfied with the canvas as he left it.”[vi]

The suspected sequence of events at the time of refurbishing should be mentioned here. If Turner refurbished Shipwreck, the Rescue in the 1840’s (as much of the evidence suggests), it is likely that the varnishing was done shortly thereafter. Bearing this in mind it is very unlikely that anyone but Turner would have left a painting in such a state for so many years without a varnish coat  in the first place, and ultimately, when it was finally varnished, from all accounts, no one but Turner would have considered it an acceptably finished work of art.
     It was likely  presented to Thos. Griffith at this point, c1844.
A reasonable thesis would be that the re-stretching and lining were carried out shortly before, and if done by the artist himself, it would substantiate the 1844 correspondence between he and Thos. Griffith, “The stormy Picture you said in the Parlour for Mr. Foords (sic) Hero to advise with about both cleaning and lining but can not find out who was employ’d.” Given that, John Gage in his Collected Correspondence of J. M. W. Turner has dismissed Hero and Leander as being the ‘stormy Picture’; it is certainly possible that Turner is speaking of Shipwreck, the Rescue. The timing would surely be right and he was known to have called The Shipwreck a "Stormy picture" in earlier correspondence.

Also, one can reasonably assume that the present varnish layer is predominantly the original one applied to the painting, and is on top of all the paint layers. When the additional sweeps of paint were added to the water around the rescue boat during refurbishing, the fresh paint would necessarily have been applied to an unvarnished surface in order for proper adherence. This is why Turner was able to return to many of his canvases after decades of neglect, and thus, do improvements without worrying about stripping back old varnish. As for the later restoration: "Dark areas in the photograph indicate alterations to the varnish, including re-touching and some cleaning," this would likely have been done contemporary with the strip-lining. If there is in fact another varnish layer present as Laver's report suggests, this re-enforces further the early age of the first varnish and the likely-hood that it was applied shortly after the refurbishing took place c1844. The milky florescence would indicate that even the subsequent varnish, if there is one, is also very old.

report cont': “The infrared photograph shows most of the features of the composition quite clearly and there is no indication of underlying grid lines or a sketch, or any other pentimenti not already visible in the work in normal light. Even though the paint layers, carbon-based materials such as charcoal or graphite would have been made visible using this method. Townsend notes that pencil under-drawing would be unusual, except for complex areas such as detailed architecture or battleships.[10]

"X-radiographs were taken of some details. They generally show the locations of lead-based pigments, particularly lead white (as well as any others which contain heavy elements). The canvas weave is revealed, as well as highlights of the composition; fine details are generally not visible. According to Townsend, from her letter of 1997, “Thin x-radiographs are typical of Turners which have not been repainted by the artist.”


2.  Materials

          2.1  Canvas

    "The plant fibers of the original canvas were identified as Linum usitatissimum or flax. When flax is manufactured into a textile, the product is called linen. Flax based textiles have been available prior to and during Turner’s lifetime."

         2.2  Ground Layer                



Electron-microscopic cross-section of paint layer, Shipwreck, the Rescue (brown ground)


"The ground layer was revealed in many of the cross-section samples and consisted of brown priming on the canvas. Townsend notes that Turner’s father prepared most of the canvases until his death in 1829. His primings typically had a rough texture and were prepared from whole egg and used consistently without a coating of size in order to be more absorbent; the principal filler was usually lead white, although chalk and gypsum were tested. It was only after 1829 that Turner bought ready stretched, ready primed canvases, with a commercial priming of lead with chalk in oil. Although coloured grounds were noted by Townsend (e.g. A red one in Coniston Falls), none of them were brown. In the early years they were cream coloured, pinkish, or sometimes white; by 1810, white or off-white primings were most frequently used. According to Townsend’s recent comments, all of the early seascapes were on white ground, so as to render the light areas more luminous."[vii]

That Turner used a brown coloured ground for his earliest paintings is extensively discussed on the Art/Science page. Townsend reasoned that by using a white ground Turner was able “to render light areas more luminous.” This is true but limited in scope. Kenneth Clark also recognized Turner’s need for dramatic tonal contrasts but explains it in a broader context. He includes Turner’s sea-pieces from an earlier date when he was almost exclusively using coloured grounds. Clark is therefore able to clarify an important evolution. “With Turner’s dark, early sea-pieces… effect of light is not achieved by contrast of tone (as it is in the Shipwreck) but by a most subtle alteration of colour.”[viii] There should be no argument with Townsend that by 1805 when Shipwreck was exhibited or even earlier that Turner was looking for a way to brighten his sea-pieces, and that the solution of using a white ground came from his experience as a watercolourist.

 During his early years Turner appears to have been doing serious experiments with ground colour. According to the Turner colour-expert, John Gage, “Too few of Turner’s early oils survive in an unfinished state for us to be sure how far he was experimenting in this direction; although it is clear in the small Caernarvon pair; and a scrap of oil landscape around 1798 in the British Museum seem to be on a red ground.”[ix] It should be recognized that when Townsend made her statement about the brown ground colour of Shipwreck, the Rescue, it might justly have been mentioned that a complete survey of this technical area was far from finished.

    It has been noted by several authors that in 1800, about when Shipwreck, the Rescue was likely painted; Turner was being influenced by both the Dutch masters and Titian, and that both these influences would have directed him toward a dark ground. The resulting unsatisfactory review by colleagues and the public who were distressed by the lack of hope in the dark foreboding storm scenes jolted Turner into adding more light and opening up the tumultuous clouds further as seen in Calais Pier, circa 1803. It is fair to assume that condemnation by his peers over some of his early dark grounded experiments forced Turner to move toward white grounds. In the case of his sea-pieces it appears that the move was a quick and dramatic one. It is important to note that other than for his sea-pieces he was still using coloured and off-white grounds well into the same decade. By 1879 Philip Hamerton in his biography on Turner abhors the early dark paintings: “In the year 1800 Turner seems to have thought it necessary, as an Associate of the Academy, to send something of a higher character than usual to its exhibition, so he exhibited The Fifth Plague of Egypt, a tiresome brown picture of a class which would soon become intolerable.”[x]

 The era from which Hamerton speaks is of course when the bright palette of French Impressionism was at its height, and even well into the 1900s dark ominous paintings such as Shipwreck, the Rescue were out of favour. 

Report continued:

   Infrared spectrometry of the ground layer showed the presence of lead white; chalk, if present, was below the detection limit of the method. The binding medium was not determined...”

Shortly, it was decided to test the ground for egg medium, and Laver’s correspondence at the time mentioned, “if the results were negative, it would still be possible that the dark ground was an experiment in both colour and medium.”[xi]

The report again wandered off at this point with lengthy discussions about pigments, paint medium and the like. I will spare the reader the detail because the implications of the data are explained in the conclusion. 

    “A sample from a retouched area of the painting showed at least two-pigmented varnish layers in cross-section. In a sample from the area of milky fluorescence, a proteinaceous material was confirmed using infrared spectrometry - this may be glue (possibly related to lining). As remarked earlier the unfinished state of the work suggests that any varnish layers were not applied by the artist[11], but by a subsequent restorer." [12]

The quality of work that Marilyn Laver did under the auspices of the CCI, and the insight into Turner’s working methods she was able to glean within such a short period of time was remarkable. This is accounting for the limited time given for the research both in 1982, then again at the impatient urging in 1997. The arguments that are presented contrary to the conclusions in her report were developed later as a result of further study, study that had the liberty of ample time.






Text Box: As a general statement: there are certainly absolute negative characteristic that would lead to an immediate rejection, but there are few absolute    positive factors completely guaranteeing authorship.

                                           Might Probability and Bayesian Analysis be essential to art scholarship?.

The question still remains, “is there any way to determine without a doubt, authorship of a painting using science or a statistical method?” It would take a substantial amount of data gathering and the terminology would need framing to accommodate paintings in particular, but I believe the answer is yes. The phraseology of the question would have to be ‘likely to be done’ or ‘unlikely to be done’ by a particular artist at roughly a certain point in his or her artistic production.

One requirement would be to develop a series of ‘fundamental’ binary questions that could be used to determine the probability of a paintings authorship. Simplicity of course is imperative. Right off, there could not be a single characteristic that would automatically exclude an attribution; for example, if the support, medium, pigments or iconography were impossible during the lifetime of an artist, an assessment vis-à-vis that particular artist would, thus, be immediately rejected. As a general statement: there are certainly absolute negative characteristics that would lead to an immediate rejection, but there are few absolute positive factors completely guaranteeing authorship (fingerprint or DNA evidence are two that are normally considered conclusive and indisputable).

Conversely, in a binary set of questions, if the likelihood of a certain characteristic is negligible but exists, it would still have to be considered but would represent a negative point against attribution. The age of materials such as canvas, panel, pigments, etc., factor strongly; and their statistical importance might be weighted accordingly on a point basis. If it is commonly understood that an artist used a certain material or working method, but not so frequently as another, but the likelihood is not entirely clear or relevant, this could represent a neutral point and not be included.

 For example, with Turner, although he did many times more watercolours than oils, it is accepted without dispute that he also painted in oil. A mute relationship like this might not be considered at all and thus be excluded from the study. Also, diversity in some instances may support rather than detract; if an artist painted on canvas and wooden panel in equal proportions, but on occasion used copper or other unusual or even unorthodox supports contrary to most other contemporaries, this point could support an authorship and be included as a plus.

Say for example, 40 key questions were posed and the result was 35, likely, and 5, unlikely, then what would be the statistical odds that the painting is ‘right’. And, what relationship would have to exist between the likely and unlikely in order to have an indisputable result. In the end it may be that the only result acceptable would be extremely probable. We have to remember here that when several key points are linked together, a progressive and not just linear relationship applies.

A list of key points might relate to physical factors on the one hand and subject and style on the other. The physical realm could include; support: oak, pine, walnut, lime-wood, cotton, linen, copper, card, paper, etc.; medium: linseed oil, walnut oil, water, etc.; type of absorbent or non-absorbent ground: oil, gesso, polymer primer, emulsion, egg-based, some of which are rare and could be weighted differently.

The stylistic realm could include: impasto, glazing, outlining, separate or blended colours, attenuated or lengthy brushwork, incisive or tentative application of paint, tonal values, colours (for example, some impressionists opted to avoid using black during certain periods of their career), perspective, point of observation (high, low, multiple).

Subject material could include: figurative, landscape, still life, religious, secular, allegorical, topographical, sporting, interiors, or portraits.

It is important to establish the appropriate key points, or framework by which to judge a painter statistically, and these lists must carry on to the point that the statisticians would be satisfied with the numbers. In order to do this and assure the success of such a project, the help of already established specialists on the various artists is key. Once the artist is chosen, the determination of appropriate questions would fall to them. Assuredly, in the case of Rembrandt and Turner, these advisory groups already exist both formally and informally.

Once the lists of binary questions are established and the process is in place, determining authorship should become easier for everyone. One might hazard to say that the imputing of data and interpretation of the results would ultimately remain in the hands of those same experts, or committees of experts, but there ought to be less likelihood of malice or self-interests involved in making judgments. With this sort of backup there will also be less worry over the legal implications of poor decisions.

A rudimentary process along this vein was begun more than a century ago by Morelli but has since gone little further. It is true that the accumulation of data, sampling, and testing has been done diligently in recent years, but these efforts have not been wrapped in with expertise as Morelli might have done in his day. He evaluated idiosyncrasies peculiar to the individual artist, including details of the nostrils, mouth, eyelid, finger nails and then processed them as valuable clues. He proceeded to categorize them succinctly, although certainly not to the level that could be done with modern tools. After divining even those scant few “peculiar idiosyncrasies” Morelli unabashedly carried forth by:

“Dividing works belonging to one ‘family’ into three classes. First, those in which all the peculiarities of the master were present, as well as the emotional qualities, which we look for in his work – these he considered indubitably from his hand. Next, works which presented some of his technical characteristics, together with his forms and types but in which the power of infection which a genuine work of his should possess, was missing; these he classed as simply works from his cartoons, executed probably in his workshop. Lastly, examples which, whilst bearing a certain family likeness to his genuine productions, lacked his technical peculiarities and the higher attributions of conception and design which would be looked for; verdict, not authentic, probably the work of pupils or imitators.”[xii]

On the formative professional front one might point to the youthful Bernard Berenson to find the ultimate blend of both impassioned connoisseur and logical analyst. Before losing to the whimsy of overly critical introspection, his approach mocked all others, and particularly the overzealous romantic critic, Ruskin. Science was not developed enough in the field of art for him to commandeer it fully, but the spectrum of other devices including a particularly well seasoned ‘magical eye’ plus Morellian principles proved essential for him. These were applied with concision to divine authenticity, and thus, create a historical record where none before existed. According to Harold Acton, “B.B. concentrated on synthesis and analysis and resisted the temptation to romanticize his impression. Always he stuck to essentials.” On the other hand, in his own words: “I fear I am dropping into literature and science much more than I care to….Paintings hate people that come to them with anything but perfect abandon.” [xiii]

And to add one more important device to the mix: what if Morelli, who was both well ahead of his time, and unfettered by repressive politics, had the modern understanding of statistical probability? Surely expertise today would be more solidly grounded. And what if this device were embraced in the form of Bayesian Analysis?

     Reverend Thomas Bayes 1702-1761

With action in this direction by adroit functionaries connected to the Decision Analysis Society of INFORMS (Institute for Operations Research and the Management Sciences), the result might be revolutionary. It appears that the analytical basis for statistical analysis for nearly all decision making needs is already established using Bayesian Analysis: Not only that—but each year there are awards presented for the best application of decision analysis. The following is a past DAS panel discussion, which suggests this might be the right track. All that is required is to get the bright minds from DAS or some other interested statistically minded think-tank together with our topic, and voila!

SC02.2 What's Cooking at the Lab? Resource Allocation Decisions in Basic Science

“The federal government will spend over $20 billion on basic science research in 2001. These resources are typically allocated via informal processes that allow emotion, politics and personal biases to play large roles.” A decision analytic approach adds structure, insight and clarity to the process.

      Other relevant groups are: ISAB (International Society for Bayesian Analysis), and ISDS (Institute of Statistics and Decision Sciences, Duke University). Art authenticators are normally risk averse, even when considering lower value items, but when it comes to those of very high value, risk tolerance approaches zero.  For the sake of preserving those, yet unrecognized great artworks, this approach may present an impartial salvation.

In March 1950, the father of modern art history, Bernard Berenson was haunted by a premonition, one which unduly pitted science against connoisseurship. It was likely no more than a fear of the unknown, but nonetheless, it inspired him to attack the empirical as a heartless beast. He wrote in his diary, “I try to get some idea of what contemporary science is up to, and find myself balked by its being ever more and more mathematical, and I being without any mathematics…I never approach matters without a feeling that the experts in these matters will govern us, forming an elite with abstract ideals, more arrogantly entertained than by any previous governing clan in the whole of history.”[xiv] This, coming from the pen of the most influential art historian of the modern era has surely injured cooperation between Art and Science. Ironically, what made Berenson the father of modern art history in the first place was his Morellian-style analysis and empirical approach to categorizing art works and their creators. Science is no more than careful close scrutiny taken one step further by the use of ‘tools’. Please permit the analogy comparing our more civilized activities with those of our ancient ancestors—tools, are important.

                                                                                                                             Art World’s Dirty Little Secret

                                                                                                   Terms and conditions J.M.W. Turner
                                                                                         Site criticism is appreciated at Penticton
 © setters 2003, Rescuing Turner: The Art Project &

                  Turner's Style                     Forensics                   Art/Science 

[1] 161 X 222 cm is an accurate measurement for Shipwreck, the Rescue.

[2] The Bridgewater Sea-piece was likely painted the same year as Shipwreck, The Rescue, and is almost identical in size.

[3] Weave patterns and thread variances have been studied for artists such as Rembrandt. Even the thickness of each thread varies on hand loomed canvas, and thus, the consecutive pictures from each roll of canvas can be matched. The ability to sequence canvases of an artist such as Turner will eventually add to the knowledge of his progress and help to identify his work more readily as to where it fits into his early series production.

[4] The entire painted surface is still present. In this case it can be assumed that the term ‘cut down’ refers to the trimming off of the degraded wrap-around part of the original canvas. This would likely have been done at the time the painting was lined.

[5] If the first lining was done at the time it was stretched in the 1840s, and if it were stretched using Brown of Holborn materials (the main firm Turner was buying art materials from at the time), this thread-count would make perfect sense. For their pre-stretched canvases Brown of Holborn was using finer canvas with a similar thread count during those years. As elucidated in 1.2: the stretcher conformation is also the same as those made by Brown of Holborn.

[6] Martin Butlin later observed latent indications of the painting having had a ‘diagonal’ stretcher/strainer configuration previous to its present 19th c stretcher (refer to The Butlin Tapes).

[7] An expert in handwriting has not yet determined if this label is in Turner’s own hand, but there are certainly similarities to other examples of his writing, particularly the flourishes. There could be difficulties finding other such purposely articulate hand-lettering to make a qualified appraisal. In any case the age appears to be right, the physical stress and shrinkage characteristics of the stretcher in relation to the label indicate that the label was affixed to the stretcher when the stretcher was new.

[8] It was the contention of Peter Paul Biro that, The relining was done with animal glue and not a hot iron and wax. There is no evidence that the impasto was compressed or the colors darkened by the lining process.”

[9] It was later determined by McCrone Associates (Illinois) that the priming was oil-based and was possibly walnut or poppy seed oil. Walnut was an oil that Turner was known to have used. The suspected early date of Shipwreck, the Rescue assures one that he was still using, with the help of people like Sebastian Grandi, the stable old master techniques for canvas preparation.

[10] There is in fact no carbon-based under-drawing present. The under-drawing that is visible, and in particular that of the figures in the rescue boat, appears to be done with a fine brush, and is done in keeping with Turner’s other early figure drawings. The erroneous suggestion that the drawing was pencil was that of the author’s and not Marilyn Laver’s. Blame for the ensuing confusion has thus been properly allocated.

[11] As will be seen later it is likely that Turner was satisfied with the level of finish and was probably involved with the varnishing. This would have been nearly half a century after the painting was first painted and would have coincided with the proposed refurbishing date circa 1844, and the presentation to Thomas Griffith.

[12] Many reference numbers were omitted because of insufficient space for corresponding tables, but the context and content remain intact. Also, a small number of inconsequential changes were made in order to have the report tie directly with other content in this chapter.


[i] David Blayney Brown, From Turner’s Studio -- Paintings and Oil Sketches from the Turner Bequest, (Tate Gallery Publishing, 1991), 19.

[ii] David MacMartin Audet, senior accredited member of the International Society of Fine Arts Appraisers (River Forest, IL) from his appraisal of the painting while owned by Collin G. Grant in Feb 1990, p. 9.

[iii] Philip Gilbert Hamerton, Life of J.M.W. Turner R.A., (University Press: John Wilson & Son, Cambridge, 1879), 27.

[iv] Byron Webber, James Orrock R.I., Painter, Connoisseur, Collector; (Chatto & Windus, London, in two volumes 1903), Vol. 1, 58.

[v] Ibid. 95.

[vi] Michael Lloyd, Turner, essay by John Golding, (National Gallery of Australia, Canberra, 1996), 175.

[vii] From Townsend's letter to Laver, October 1998.

[viii] Kenneth Clark, Looking at Pictures, (John Murray, London, 1960), 146.

[ix] John Gage, Colour in Turner: poetry and truth, (Studio Vista Limited, London, 1969), 30.

[x] Philip Gilbert Hamerton, Life of J.M.W. Turner R.A., (University Press: John Wilson & Son, Cambridge, 1879), 65.

[xi] M. Laver’ s correspondence to the author, March 9, 2000

[xii] Percy Moore Turner, Appreciation of Painting, (Charles Scribner’s Sons, New York, 1921), 226-27.

[xiii] Hanna Kiel, The Bernard Berenson Treasury, (Methuen & Co. Ltd., London, 1964), 14.

[xiv] Hanna Kiel, The Bernard Berenson Treasury, (Methuen & Co. Ltd., London, 1964), 291.