The Chemistry of Oil Painting on Symbiartic

The Chemistry of Oil Painting on Symbiartic - Glendon Mellow
Artist and illustrator Glenton Mellow, who writes the Flying Trilobite blog, also co-authors a new blog for Scientific American called Symbiartic, along with scientific illustrator Kalliopi Monoyios.

The tagline for Symbiartic is “The art of science and the science of art”, and topics range freely across that nebulous and fascinating intersection.

In a recent post Mellow gives a nicely succinct overview of The Chemistry of Oil Painting, with a bit of history, discussions of the principal types of oil used and a mention of artistic concerns such as glazing and “fat over lean”.

You can find more of Glendon Mellow’s writing and artwork on The Flying Trilobite and his website.

 
Facebooktwittergoogle_plusredditpinterestlinkedinFacebooktwittergoogle_plusredditpinterestlinkedin

40 Replies to “The Chemistry of Oil Painting on Symbiartic

  1. I found this article, the Chemistry of Oil Painting, to be both poorly written and misinformed. I’m an oil painter and teacher, and have been working in the field for over 40 years. I also regularly consult w/ art conservators, have worked as a conservator, and have read much of the literature in that field. An example of the misinformation: One paints fat over lean in order to compensate for the binder that has been absorbed by the previous layers of paint. It is that absorption that provides strong adhesion, along with molecular cross-linking of the binding medium. If there is too much addition of binder (fat) then the subsequent layers will not adhere well, and cleavage and cracking will occur. This is due to the lack of absorption of the over-saturated layer. If there is too little, then the subsequent layers will ‘powder’, as has happened with some of Van Gogh’s work. Any paint layer, no matter how much binder, is going to impede oxidation of the underlying layers, and competent painters recommend a significant drying period, which varies by medium, between layer applications.

  2. Thanks for your comment highlighting the problems of ‘powder’ or ‘dusting’ in painting, William.

    However, I am having some trouble understanding your comment: “An example of the misinformation: One paints fat over lean in order to compensate for the binder that has been absorbed by the previous layers of paint.”. It’s not a good example of misinformation on my part, because I never said that in my post.

    Binders in paints are different from the oil which is usually referred to as the medium or vehicle. I intentionally left information on binders and driers and other ingredients out of the article since much of the audience (but of course, not all) are new to what happens in the oil painting process. The beauty of blogging though, is some of those discussion continued in the comments below the post, much like they are here. ;-)

    Good point about length of time between layers. Perhaps I should have emphasized that, and that many oil painters have several works on the go in their studio so they can keep producing while watching paint ‘dry’.

  3. Glendon – The oil that is present in oil paint straight out of the tube, and which is added in the form of a medium is precisely what I said, a binder. It’s function is to bind the powdered pigment into a paint laminate. If you peruse the literature, you will discover that an oil painting medium is also often referred to as a ‘binding’ medium. That is its principal function. Driers are sometimes added, but are really not integral to the purpose, and sometimes introduce adverse and unintended reactions. Varnish is also sometimes added, with the intent of affecting the relative ‘sheen’ of the paint laminate, but that is also tangential.

    It is useful to think of a binder in relation to its nearly polar opposite, a diluent. The most common diluent in oil painting is pure gum spirits of turpentine. Mineral spirits are sometimes substituted. I say that the diluent is almost the polar opposite of a binder because it is used as a dispersant. It disperses the pigment and binder, in order to make the paint mixture flow more readily.

    An oil paint laminate consists of powdered pigment bound in oil. It is really that simple. The diluent, for all intents and purposes, evaporates completely, and does not form a component of the resultant laminate. It does however affect the quality of the laminate before it evaporates. Too much diluent will create a poorly bound paint layer, as it separates the oil (the binding medium) from the pigment. That a dispersant is sometimes mixed with oil to create a painting medium does not obviate the separate functions of these two components.

    Understanding the role of the binding oil and the diluent are fundamental to understanding the structure of a painting. Driers, retouch varnishes, etc. are tangential. Beginning students often confuse the separate and distinct roles of binders and diluents in oil painting, and will frequently use the binding medium indiscriminately to ‘dilute’ the paint. This could be a recipe for disaster, as the paint laminate will become saturated with oil, subsequent layers will adhere poorly, causing cleavage between layers, cracking and loss of paint laminate.

    Speaking generally, we live in an era where artists know the least about the materials that they use in our history, in practical terms. At the same time, because of ongoing scientific investigation in the field of conservation, we know more about the technical aspects of painting than at any time in history. Funny, that.

  4. With regard to the misinformation I felt was in your article, I did not intend to accuse you of saying “One paints fat over lean in order to compensate for the binder that has been absorbed by the previous layers of paint.” I said that as counter to your assertion that the use of ‘fat over lean’ in paint application was in order to obviate cracking due to drying differentials. I pointed out the traditional reason for painting fat over lean, which has nothing to do with the speed with which different paint laminates dry, but everything to do with compensating for oil absorption and inter-layer adhesion.

    Indeed, different pigments have different drying properties: Some, like the umbers are natural siccatives. Others, like Zinc white take forever to dry. You are absolutely correct that the rate of drying of layered laminates of paint can cause cracking because of the respective layers’ different rates of oxidation, but relative oil content will not alter this, and is not the traditional rationale for painting ‘fat over lean’.

  5. I certainly don’t disagree with your statements in comment #3, William. And I don’t believe my article mislead anyone on those points, although I did not get into an extensive discussion on all points and ingredients in oil painting. My post was an overview for an audience that is largely unfamiliar at all with painting.

    I stand by my statements about fat over lean though: I’ve included a couple of references on the article Charley kindly points out above. I agree different pigments affect the siccative rate, though I believe that has more to do with pigment dispersal to oil ratios (some colours need more or less pigment in the oil for a consistent colour in each stroke, and if it needs more, the drying may be quicker, less, the drying may be slower die to the volume of oil.

    William, you mention;
    You are absolutely correct that the rate of drying of layered laminates of paint can cause cracking because of the respective layers’ different rates of oxidation, but relative oil content will not alter this, and is not the traditional rationale for painting ‘fat over lean’.

    However, the different amount of oil in the layers is precisely what affects the fat over lean rule. Adding diluents such as turp will, as you say, evaporate quickly. Adding turp to oil+pigment means there is less proportion of oil in that layer. Ergo it will dry faster.

  6. As I said, I have over 40 years experience in the profession. I’ve read most of the historical treatises on painting methods and materials, and keep abreast of advances in the field of painting conversation, as well as consult regularly with conservators. The rationale that you claim for painting ‘fat over lean’ is one that I have never, in my reading or practice, encountered before. Nor is much of what you say in your last post supported by the research. It takes approximately 15 minutes for turpentine to fully evaporate from a paint film, about the same time for whatever moisture in the form of water to evaporate as well. Following that the oil, whether linseed, poppyseed or other, oxidizes at a rate which decreases over time. This rate is fixed to the age of the film, rather than to the quantity of the oil.

    Drying oils do not in fact dry. They undergo molecular transformation over time, and that transformation never stops, even centuries after the painting is completed. Major changes in a paint laminate occur during the first few days of its application, and then the process slows considerably. The best way to avoid cracking and buckling of the paint layers is to wait a decent period between applications of paint. Painting fat over lean has nothing to do with this. Otherwise, subsequent layers will become surface dry and impede the drying rate of the underlying layers, complicating the oxidation differentials between layers.

    As for the siccative qualities of various pigments, it has been well established for decades that certain pigments speed the oxidation of slower drying pigments that they are mixed with, regardless of oil content. What’s more, the aforementioned umbers have higher oil absorption rates and do require more oil in grinding. But the idea that they dry quicker because they contain more oil negates your whole theory about fat over lean. One would want to slow the drying of subsequent layers, rather than speed them up, to match the slower speed of oxidation of the underlying layers. If your theory were correct we would want to paint lean over fat.

    I’m not trying to get into a pissing match here, but I do believe that one should not pass on misinformation. In the end, your theories re. fat over lean are somewhat convoluted and ultimately self-contradictory. I’m not surprised that you were unsure of the meaning of the word ‘binder’ in your previous comment, or that you run into trouble when trying to discuss the relation between natural siccatives and oil content.

    I will state it as clearly as I can: The reason that painter’s paint ‘fat over lean’, i.e. add more oil to each subsequent layer of paint, is to replace the oil that is absorbed by the underlying layers of paint. This absorption, along with molecular cross-linking, is what creates adhesion between layers. If there is insufficient binding medium (oil) then each subsequent layer of paint, due to this absorption, will have a greater deficit of oil, and will be poorly bound because of that deficit. If too much oil is added, then the paint layers become saturated and cannot absorb oil from subsequent layers. This results in a lack of adhesion between layers, which leads to cleaving, cracking, separation and loss of paint.

    Painting conservators deal with such problems every day. Painters don’t, and that is probably why there is so much confusion about the issue. I’m confident in saying that rate of oxidation is not a reason to paint fat over lean.

  7. Likewise I’m not trying to get into a pissing match either William, and this is the second time you’ve used the argument from authority and cited your 40 years of painting with oils.

    So here’s my street-cred then. I have read a lot about oil paints in the 17 years I’ve been painting with them; I managed an art supply store for ten years and kept abreast of product information, and I have an art studio fine arts degree. What’s more, I have actually used bibliographical notes to cite what I am talking about. May I (and readers following this exchange)please see some links that show I’m incorrect about oil drying, or that spell it out the way you do?

    Furthermore, my post on Scientific American actually largely agrees with what you are saying about how oils behave. I’m starting to wonder if you even read the whole thing.

    Let’s compare. In your comment above you say, “Drying oils do not in fact dry. They undergo molecular transformation over time, and that transformation never stops, even centuries after the painting is completed.”

    From my post: “Watercolour and acrylic paints have water as part of their medium – they dry by evaporation. But oil paints don’t. They dry by what’s called a siccative quality. That is they absorb oxygen from the air.”

    Much of what you say about the adhesion between layers is true. But denying that part of the reason for fat-over-lean is related to the oxidation rate as you do in your last sentence is erroneous.

    From one of the sources I cite in the original post (emphasis in bold mine):
    “In practice, it is still advisable to avoid fast-drying colours with high oil contents by volume (such as Umbers, Siennas and Cobalt Blues) in underpainting as cracking may occur in overlaying layers if these do not contain relatively higher oil contents.
    The Artists Handbook. Ray Smith, 2000 Alfred A Knopf. p.182

    Same source, page 181:
    “The oil film may crack due to the superimposition of lean (non-oily) paint layers over oily ones.” I covered this with my cookie analogy.

    From manufacturer Winsor & Newton’s The Oil Painting Book p75 (emphasis mine again):
    “Fat over lean. This is the most often-repeated principle when referring
    to “building” the oil painting film. What it really implies is flexible over
    less flexible, for, when increasingly flexible layers are built one on top of
    another, the final paint film will have the greatest possible resiliency, and
    will be more resistant to cracking. Increasing flexibility is accomplished
    by adding more medium or oil (a “fatter” mixture) and less solvent to each
    layer of colour. Contrary to many publications, neither oil absorption, nor
    oil index information is required for observing this rule
    .”

    Ah. Oil absorption by layers beneath is not relevant to fat over lean as you assert, according to one of the leading manufacturer’s best advice.

    The symptoms you cite for not observing fat-over-lean are correct (“This results in a lack of adhesion between layers, which leads to cleaving, cracking, separation and loss of paint.”) but your understanding of why this is so is not. Or at least, differs from mine on emphasis.

  8. If it’s agreeable to all parties, a transcript of relevant posts to date could be published on Symbiartic in conjunction with the original article, and a pointer posted here to direct those following the thread to the continued discussion.

  9. Perhaps that’s a good idea Charley, if William agrees. I only meant I wished the discussion were happening there because we’re a new blog,and it’s the type of discussion I’d like to cultivate.

    Happy to talk about it here too!

  10. As much as I enjoy having these kinds of informative discussions take place on Lines and Colors, it would make sense for this particular conversation to be closer to the context of the original article.

  11. Unfortunately, I don’t have time to continue this discussion. I only wish to point out that the statement from WN does not necessarily lead to the conclusion that absorption of oil by underlying layers is irrelevant, as they are talking about the oil absorption index of specific pigments, not the function of fat over lean. I also want to point out that nothing that you have so far cited supports your view that painters use fat over lean to mitigate the oxidation differentials of paint layers. The flexibility mentioned by WN is certainly an issue, but not related to oxidation differentials.

    Interestingly, the Wikipedia entry muddies the water even further, stating that flexibility is a result of the difference in time that it takes for low and high oil content films to dry. One need only consider the constituents; flexible oil vs. inflexible pigment grain, to realize that drying times are not part of the equation.

    Obviously we both agree that a painter should follow the rule of painting ‘fat over lean’ or of adding more oil w/ each subsequent layer. I questioned your rationale for doing so, that it somehow equalizes the oxidation differential of discreet paint laminates. Though I don’t have the time to continue this discussion, perhaps someone else will want to. If you deem it useful, you have my permission to reprint this on Symbiartic

  12. With regards to Ray Smith’s book, I would not regard him as a reliable source. Considering the passage cited, we can assume that most of the Venetian and much of Ruben’s and Titian’s output would be full of cracks and in a sorry state of disrepair. The fast drying umbers were widely used as underpainting and imprimatura throughout the Renaissance and well into the 19th Century. Though the umbers are fast dryers, they also have high oil absorption indexes, and what one more commonly sees when overpainting them is a pronounced dryness in the subsequent paint layer.

  13. William I understand you are too busy to continue the discussion. Thank you for your permission to put it on Symbiartic as well. I’ll see how it takes shape.

    I would like to point out though, for others reading this exchange, that Ray Smith’s book still has it right: yes, high-oil umbers were often used in imprimatura and under-painting. As he says in what I’ve quoted, this is less of a problem if you continue to add even more oil in the layers above.

    William, you seem to be drawing different conclusions than I do from the information I’m citing. I would like to point out though, that I am able of finding and linking to sources that support my view that fat-over-lean is meant to mitigate the cracking & adhesion problems that come from improper ‘drying’ times between layers, and you have yet to cite any sources that fat-over-lean has anything to do with under-layers absorbing oil from subsequent layers added. Winsor and Newton goes so far as to directly refute that idea.

  14. From Gamblin Artists Colors:
    “Follow the FAT OVER LEAN rule: apply oil rich colors over more matte colors so that the oil rich colors do not break the leaner layers in order to dry.” Again, see my cookie analogy on the Symbiartic post.

    From LeFranc and Bourgeois, p3 (pdf):

    “This is why it is not recommended to paint too thickly with oil paints (not more than 5 mm).
    To improve thorough drying, the process must be reversed by applying increasingly oily paint layers
    one after the other, that is layers with increasing amounts of painting medium.
    This technique is known as LEAN to FAT. It increases the penetration of oxygen to the lower layers to
    harden them.
    This is why it is not recommended to paint too thickly with oil paints (not more than 5 mm).
    “To improve thorough drying, the process must be reversed by applying increasingly oily paint layers one after the other, that is layers with increasing amounts of painting medium.
    This technique is known as LEAN to FAT. It increases the penetration of oxygen to the lower layers to
    harden them.”
    See also their accompanying diagrams.

  15. Glendon – I’m sorry that you’ve taken the defensive posture that you have. I did not intend that. You seem intent on proving me wrong, and yourself right. I suppose that it is a natural response when challenged. As I’ve already stated, the WN quote does not have anything to do with interlayer absorption, they are talking about the oil absorption index of specific pigments. If you were reading for comprehension you would understand this. Instead you are looking for evidence to support a refutation.

    With regard to the Gamblin statement, I don’t know what to say. He’s someone that I’ve met, he’s been in my studio, he’s a bit wonkish re. color. But the statement that oil rich colors won’t break the leaner layers does not make sense in material terms. Nor does the L & B statement. I can think of no mechanism whereby an oil rich layer of paint will convey oxygen to underlying layers in sufficient quantities to mitigate buckling if both layers are relatively fresh. According to the actual research that has been done, the upper layers become surface dry, creating a seal which IMPEDES THE CONVEYANCE OF OXYGEN, and therefore hinders the oxidation of the underlying layers.

    In point of fact, all of the references that you cite recommend a waiting period between application of paint layers. This is so that the major oxidation of the paint laminate can take place. Once that has occurred, subsequent oxidation and change in the paint laminate is not as dramatic, and the surface presents a stable ground for additional paint applications.

    It is when two interfacing layers are still in the early radical stage of oxidation that the push and pull of oxidation differentials comes into play. Adding oil to subsequent layers of paint at this stage will not improve the situation. It is only when a layer of paint has reached a state of relative stability that it can accept and accommodate a newer layer.

    You are correct in suggesting that the ‘fat over lean’ rule is inherently more stable because it builds more flexible layers over less flexible ones. This is true in all construction, such that it almost goes without saying.

    You are incorrect in stating that absorption is irrelevant re. the fat over lean rule. It is a significant reason for it, integral to the principal that a paint laminate be sufficiently bound. This should be obvious, and is clearly stated in the literature of conservation from Helmut Ruhemann to the present. I have noted two possible outcomes, one regarding paint that is insufficiently bound, one re. a paint layer that is over-saturated with binding medium. Both are situations that conservators deal with on a daily basis, both are related to the fat over lean rule, both are directly related to interlayer absorption.

  16. Le sigh.

    It’s not that I’ve taken a defensive posture William: I’ve asked you repeatedly to cite your assertions, and so you have not.

    As for the Winsor and Newton quote, let’s look at it again (emphasis mine):
    “Contrary to many publications, neither oil absorption, nor oil index information is required for observing this rule.”

    It’s about two topics, not one. Neither oil absorption nor oil index information. So it does support what I am saying, and I believe my reading comprehension is intact.

    Again, I agree with many of the techniques you are saying here to prevent cracking and produce a stable film – waiting between layers, for example – I do not agree that my simplified brief overview on Symbiartic contained misinformation however.

  17. The sentence only makes sense if you place the implied ‘information’ in its logical spot “Neither oil absorption ‘information’, nor oil index information is required.” Show me a grammatical construction in which the above sentence can be construed as “Oil absorption is not taking place.” The absorption takes place whether one is observing the rule or not. You think that you’ve found the smoking gun, and that my friend is a laugh.

    As I’ve said, I did not want to get into a pissing contest. Your demand for citations to “prove that I’m right” is absurd. I know that I’m right, and whether or not you agree, is a matter of indifference to me. I opened this can of worms by saying that your article was poorly written and misinformed. It is full of misspellings and grammatical errors. There was no ‘Master of Flemaille’, it was the ‘Master of Flemalle’. I still stand by my assessment. On top of that I pointed out one significant technical error, and you have doubled down on it. If I were being paid to write this, then I might be willing to search for citations, but I don’t really care if you ever develop an understanding of painting structure. Anyone else reading this can draw their own conclusions, or they can compare our respective resumes.

    In the meantime I would recommend that all would-be painters avail themselves of the latest, scientifically tested information on painting technology by familiarizing themselves with the literature of art conservation. You can start by reading Helmut Ruhemann’s “The Cleaning of Paintings and Oil Paintings in Particular” if you can find it. The Journal of the American Institute of Conservation is another good resource, as are the now somewhat dated but still interesting Technical Studies in the Field of the Fine Arts. I would recommend taking works like Ray Smith’s book with a large grain of salt, as it is primarily a compendium of anecdotes and shop-talk, without scientific rigor. Historical treatises are of value for their historical insight, but should also be tested against what is now known. Ruhemann’s critique of Max Doerner’s research method is still spot-on.

  18. 21 comments in, and you’ve finally given sources. Thanks for the new reading material! My ‘demand’ for citations is not an uncommon one on science blogs when there are disputed assertions.

    And you’re correct – I mispelled Flémalle by adding an extra ‘i’ where there shouldn’t be one. As for the rest, well, I’m Canadian. We like our color with a little extra colour.

    I’d also like to add that the quotes from companies like W&N, Gamblin and Lefranc & Bourgeois are conceivably wrong, but they also work with researchers on getting the latest research and results on painting tech.

    I could be wrong about some things I’ve cited, but the fat over lean problem is not one I’m convinced I’m wrong about so far -perhaps I left out a number of circumstances, factors, ingredients, but the broad strokes of what I said, I still stand by.

    William, you’ve done any of Charley’s or Symbiartic’s readers a tremendous service by fleshing out a lot of relevant points about proper care when painting in oil and the problems that can result; points that my original introductory post left out. Which is cool. Thanks.

    I also particularly enjoyed your paintings at Pepper’s Farm. Feels like some of the landscapes here in Ontario that I escape to in the summer.

  19. My apologies for not having read all of this in detail before commenting, however, I have added comments to The Chemistry of Oil Painting article that this discussion relates to and felt it appropriate to say something here as well.

    The painterly concept of ‘fat over lean’ predates anything that Lefranc & Bourgeois, or Winsor and Newton, have to say on the subject by several hundred years. To fully comprehend its meaning you have to first understand what ‘fat’ and ‘lean’ mean in this context. These terms reflect the observed characteristics of the materials used by Master Painters in their studios. Essentially ‘lean’ means any of the following: absorbent or a tendency to be absorbent of oil (which is not necessarily the same thing), containing less binder or a binder that creates ‘lean’ properties (again not necessarily the same thing), thin, reduced by diluents, or less ‘fat’ in relative terms than something else. The term ‘fat’ references physical properties as well as behavioural ones. What it is not, is a simple reflection of the oil content. This means any of the following, though generally a combination of several of these things: having ‘body’ of its own (physical presence independent of pigment), thickly fluid and sticky, thixotropic, resistant to ‘sinking in’, or by association—glassy and transparent, durable, elastic, or more ‘fat’ in relative terms than something else.

    Here are some examples of these terms in use:

    A true gesso ground (not acrylic gesso) is ‘lean’. A gesso ground with an oil priming over it, in the context of oil painting is still ‘lean’, however, for tempera painting it is too ‘fat’ to paint on.

    ‘Dead colouring’ should be ‘lean’ either because some of its oil content has been pulled down into a yet ‘leaner’ priming or lower layer, or because diluent has thinned its binder somewhat. The finishing layers that sit over it should be ‘fat’, not simply oil rich—capable of wetting the lower layer to create good adhesion, whilst presenting and maintaining saturated colour effects and luminosity.

    And here are some practical examples:

    In a Rubens painting, a relatively direct execution sits over what is essentially a simple toned drawing in paint on a tinted priming, or pale ground. Effects are achieved by varying the paint thickness and opacity and by preserving transparency where necessary, using a medium that is too ‘fat’ to ‘sink in’.

    There is no evidence of turpentine or other diluents being used in early Flemish oil painting. The likelihood is that part thickened oils, including both sun treated and boiled ones, are constituents of the paint layers. The ‘fat over lean’ principle is particularly evident if this technique is attempted, as is the role of ‘oiling out’, since it is not possible to progress the upper layers unless they are ‘fatter’ than the ones below.

    Yes, ‘fat over lean’ produces durable paint structures. There is considerably more to it than that though. The selective use of particular colours, or painting methods, to account for predicted behaviours within a single work, is part of good technique too. Badly cracked paintings appear more often from the 18th century onwards as the technical mastery of painters declines. Some of the worst examples are relatively recent.

    An oil that is not ‘fat’ enough to perform in a ‘fat over lean’ sequence will cause a paint layer to wrinkle as it dries and may darken the painting excessively. It is not the quantity, it is the quality of the oil or medium that makes it ‘fat’.

    There is more on this in The Materials and Techniques of Painting, published in 1989, which provides a detailed overview of the subject.

    This comment is intended to clarify what is actually meant by ‘fat over lean’ and does not necessarily contradict any or all of what has been said above.

  20. It should be noted that there are paintings which date from the 18th Century to the present that are not badly cracked. It’s entirely possible that the severely damaged paintings that precede this period simply did not survive or were discarded. Poorness of execution is often accompanied by poverty of spirit, and people tend to preserve what is deemed valuable to the neglect of what is not.

    Many of the early Flemish paintings were begun in egg tempura, a continuation of medieval practice. As noted in “The Science of Painting,” the “earliest known European oil paintings are a group of Norwegian alter frontals” that were probably painted with a mixture of drying oils and egg yolk. The usual Flemish practice was to create an elaborate drawing, probably done in charcoal then fixed in ink. The painting would be layed in w/ egg tempera, and the final modeling done in oil, probably as Mr. Stephenson suggests with sun-thickened or other partially polymerized linseed. The execution and finish followed the lines of the original drawing to the letter. It was not until Georgione and Titian that the kind of improvisational approach that oil painting opens up came into its own.

    I’m not sure if I could agree with Jonathan regarding the use of glazes in old master paintings. Titian bragged, in a letter to a patron, of having used over 40 glazes in a work. Conservators often must take exceptional caution to avoid removing thin glazes that were intended to adjust a color or value, a problem made more difficult by the addition of resins that make the paint laminate more soluble and easy to abrade.

  21. Glendon – Thanks for the comments re. “Pepper’s Farm”. I painted there for a period of ten years, often taking my students there as well. The owner was an eccentric old man that I met when he was in his late ’70s, and he was thrilled to have people on his property enjoying the sights. I moved farther away and regret not being able to continue working at the Farm after that. Bob Pepper died a few years later, but I did do a commission for his daughter that had old Bob taking his daily ‘constitutional’ around the full perimeter of the farm.

    On one occasion I went painting there with the Portland OR painter Henk Pander. There was a dead cow in the middle of a field, with a skirt consisting of the New York City of maggots. Henk loves that sort of thing (“great still life material!”) And I was afraid he was going to try to stick the whole thing in his van and haul it back to his studio.

  22. Hello William

    Yes, the use of oil colour over tempera does occur and almost certainly derives from the earlier practice of rubbing ‘fat’ oil thinly over tempera paintings as a protective final varnish. That is mentioned by Cennini, circa 1390, as is oil painting as such, if I recall correctly—and earlier references and similar examples are not difficult to find. There is a late medieval screen in an English church that comes to mind, where the glassy bead at the edge of a particular colour is a tell-tale sign of it having been placed in a processed oil. ‘Lute-makers’’ varnish is sometimes referred to, hence the common belief that amber is present.

    However, I would disagree with you about its prevalence: The hybrid approach of oil over tempera that you describe in such detail is not the foundation of the Flemish technique to the extent that you assert. The belief that it is, derives substantially from 19th and early 20th century theories, often repeated since in artists’ manuals and handed on as if the truth. Such ideas are not always based on reliable research—and came about without the benefit of modern scientific investigation to support the conclusions. From Eastlake to Doerner, despite the thoroughness of their observations and earnestness of their intentions, the ‘Old Master’ methods that they describe are substantially speculative. Though it is still valuable in many ways, Mayer’s work originally dates from 1940, so, with all due respect, it is in the same position as a source. Whilst there certainly are occurrences of this type of construction within paintings (oil over tempera), recent analysis of works has shown it not to be as widespread as supposed. The same is true of egg-oil emulsions. There is probable evidence of them, though it is doubtful that they represent a major transitional technique.

    The appeal of oil colour in Northern Europe and in Venice, historically its launch pads, apart from its obvious artistic merits and handling qualities, is its durability in the context of the surrounding climate. There is no practical incentive to maintain a hybrid technique therefore. Moreover some of the earliest accounts describe sealing the ground in a way that would make underpainting in tempera difficult.

    With regard to Titian’s 40 glazes: This is conceivable in his late works perhaps, which are quite different in manner to his earlier pieces—and may accord with Palma Giovane’s account from around the time of his death. However, that was not the case for most of his career—say for the first 40 or 50 years of it. In Titian’s portrait of the Vendramin Family and his Bacchus and Ariadne, for example, the very specific, purposeful and selective use of glazing colours; malachite, blue bice, ultramarine, verdigris, madder, Indian lake and the like; is to be observed. I am not sure if there are any at all in the probable self-portrait from circa 1512, unless the blue of the sleeve is glazed.

    Much depends I suppose on what you consider a glaze to be. Never-the-less I stand by my assertion in my comment on Glendon’s original article that there is a tendency to overstate the historical use of glazes and to confuse them with other finishing methods that serve a different purpose. Yes, it is correct that conservators have to take care not to damage delicate finishing touches on old works, though often the damage has been done by over aggressive cleaning at an earlier date, or by a superimposed ‘gallery varnish’ from the 19th century that has to be taken off; and the risk is arguably greater to say a Turner than to a Titian because of the methods and materials in each case. Perhaps we should also occasionally accept that a particular venerable painting is just not that good, instead of crediting it with supposed lost layers. My point here is that the frequent repetition of the Titian quote raises a presumption that all ‘Old Masters’ paintings are finished with glazes, or that all finishing touches are executed with glazes—neither of which is correct.

    I do though agree with your point that some earlier work has been lost and that this conditions our view of art history. So, works in glue-size on canvas, as exemplified by Dieric Bouts in the early part of the Renaissance, are now rare, though there is circumstantial evidence for them having been quite common. The difference being that the artists of the time knew these materials to be more transient in nature, or less durable—and generally used them accordingly. So, their survival is circumstantial rather than intentional, unlike Durer’s panel paintings in contrast. From the 18th century onwards much of the damage to paintings is just incompetence on the part of the artist and the result of spurious nonsense like the variants of megilp and ‘black oil’, or the well intentioned Maroger’s mediums, being employed; which brings me back to the point of handed down ‘belief’ among painters that originates with 19th and early 20th century attempts to rediscover lost materials and techniques. On this point, the technical decline of painting since the 18th century, I would expect most restoration studios in major galleries to concur. As you rightly say William, not all painting from that time on is poor; however, there is a definite watershed.

    Much as I would like to continue the discussion, I am afraid I must now devote time to other things. I will look back at a later date to see if there is anything further I should respond to.

    If any of the contributors, or any chance reader of this commentary, want to follow up any of what I have said for their own investigations, they will find many of my sources listed at the back of The Materials and Techniques of Painting and in my subsequent works on the methods and materials of the Impressionists and the Watercolour Masters. For the most part these are original sources, or secondary sources that quote the originals. The interpretation is based on an extensive practical knowledge of painting, informed by modern conservation science, a detailed examination of many works of art, as well as an understanding of the underlying chemistry and physics; and the benefit of having been a specialist artists’ colourman for many years. It is though I accept, entirely possible to arrive at different conclusions about what some aspects of the evidence may mean.

    All of this is intended to be helpful, so I hope that it is in some way. And, it is good to see that there are still some painters that recognise the breadth of oil painting techniques and the possibilities of taking what it does beyond just taking it out of the tube.

  23. Though the Titian quote is often cited, it is not really clear what he meant by it; 40 superimposed glazes? 40 patches of glaze here and there on the painting? No one really knows, but we do know, as mentioned before, that Titian’s method was quite improvisational, and he would compose directly on the canvas. Often figures and groups would be reworked or eliminated altogether, with revised passages underpainted with white in semi-chiaroscuro, then overpainted w/ glazes and scumbles.

    It would be a mistake to look at the old masters as implementing a single methodology. Clearly the historical evidence says otherwise, with general principles being observed while serving as a launchpad for experiment and innovation. It should also be noted that the treatises we have on the materials and methods of the old masters were not written, for the most part, by the great masters themselves, but by their lesser counterparts, who may have been more fastidious about the observation of tradition than their mentors (a common fault.) That artists would decry the debasement of their art is not a completely modern phenomena: “Alas my dear Willibald, there are many men with a happy talent for the arts pictorial, running amok like unpruned trees…” – Albrecht Dürer

  24. Since turpentine and other diluents are volatile, they would leave little physical evidence, if any, in a painting that has cured for a few centuries. It is difficult for me to believe that the Flemish masters did not use some diluent to make the stiffer, partially polymerized, sun-thickened or boiled oil flow more readily, especially when considering the extreme delicacy of the paintings.

  25. I did not mean to suggest that glazing was intended for ‘finishing touches’, rather it was an integral part of the painting strategy from the outset. You are correct in saying that only certain pigments were appropriate for glazes, and these are pigments that create fine translucent layers when applied thinly and with sufficient medium. Other pigments, such as the ochres, cadmiums (which would not have been available before the 19th century,) etc. The counterpart of glazing is scumbling, or the application of an inherently opaque paint in such a way that it seems transparent, i.e. it is applied in a discontiguous layer where the brush hits the peaks of the canvas weave and the underlying paint texture, leaving the underpainting showing through the valleys. This is, of course a rather simplistic way of describing the process. You noted that Ruben’s varied the thickness of his scumbled paint over a ‘toned drawing’. According to Ruhemann, this is how he achieved the beautiful pearly flesh tones that we associate with him, due to the turbid medium effect. This same effect can be seen in other painters as well, Strozzi for instance.

    The painters of that period had a much more constrained palette than we have now, and developed indirect methods of application as a way of expanding the repertoire of effects available to them. I would suggest that, since he was taking advantage of the variable effects achieved by adjusting the thickness of his application over the darker ground, that Ruben’s was not painting directly at all, in the sense that we would understand the term.

    Many of the contemporary quests for the ‘secrets of the old masters’ begin with a questionable assumption that leads to a flawed investigative method. Maroger arrived at his theories after studying old carriage decorations. I’m old enough to have been a student when his theories were still popular. The old masters were cognizant of the available knowledge, and conscientious about putting it to use. We should follow their example.

  26. I actually do maintain several blogs. None of them deal specifically w/ art methods and materials, but most are art related. The two that are not are a personal blog that has political rants, music etc. and one for the website of the Seattle branch of the style of karate that I train in.

  27. Yes, much of the source material is from lesser workshops—by no means all of it though; and there is significant content in eyewitness accounts, or the observations of interested parties at the time. It is not all written documentation that is open to interpretation either. Painters painted their own studios as backdrops to their work occasionally and included their equipment and materials in still lives and showed half-finished paintings in their self-portraits. The fact that this body of evidence so often accords with what modern technical examination and scientific investigation tells us, is proof of its worth. As an indication of how detailed this knowledge now is, I would point out that the firm attribution of a Rembrandt today, may rest upon how it was originally painted—what materials were used and how they were employed. In his case the enhancement of earth colours with brighter pigments is a notable trait among several factors that might be expected in his work.

    Given that such knowledge is available, it is the continued repetition of speculative guesses from the past that are just plain wrong, which is unhelpful to contemporary painters wishing to extend their own technical repertoire, by learning from masterful examples. How many nonsense accounts are there in circulation of how Titian’s works are painted on a red bole ground, for example? Which they are not! Or, that use the 40 glazes comment as the basis for a fanciful account of his technique, when it only makes sense in a specific context and in relation to a particular period of his work?

    The evidence for the absence of diluents in early oil painting is that they are not mentioned at all, or illustrated, or demonstrated in any way (with the possible exception of a reference to Leonardo distilling some herbs, which could mean anything) until well into the 16th century and then they are presented as something novel and unusual among painters, as if a newly discovered secret. They are not used for cleaning brushes either until much later—and when it becomes a feature of the artists’ palette, only a single dipper for oil is present. The double dipper for oil and turpentine is a much more recent introduction. This fits with the history of distillation, which although it goes back to ancient times and was evident in Europe in the 12th and 13th centuries, doesn’t seem to have been widespread until after 1500. Even when diluents do appear in the studio their use seems to be limited to keeping the brush free, or they are in their balsamic form.

    Two other factors also have to be taken into account: The first is that fresh, hand-ground, oil colour is nothing like the stuff that comes out of tubes today, which is a descendant of the 19th century industrialised product. The paint used by the Flemish Masters and indeed by most working painters up until the latter part of the 19th century, was much more responsive to the artist’s touch. The second is that a part-thickened oil can be quite fluid. It is not necessarily stiff and sticky, which is the assumption often made; and may be let down with plain oil if it is. The ‘jewel like’ effect that early oil paintings are associated with, is easily achieved by following the oil only method that the evidence points to; and the fact that this is so, is itself corroborative evidence.

    With regard to Rubens, here again a piece of comparatively modern folklore is potentially misleading. The alternative to a glaze is not necessarily a scumble. Rubens is working against a luminous tinted ground and unsullied shading, by varying his paint thickness through the manner of the stroke, to achieve tonal value and perceived colour at the same time. His medium holds these translucent touches in place. The effect produced is because of the relative thickness, not because the layer is broken, as it is in the case of a scumble. His colours too, which are bright, relatively pure and not always obvious choices, are organised in simple tonal arrangements (this is more apparent in his pupils and assistants work where it is more crudely done); and quite a lot of the time he doesn’t bother to paint at all, or uses the same effect to represent lots of different things. Something he uses a lot is not actually on his palette—that is the observer’s tendency to make sense of visual clues and to assume something is what it is not. Velasquez and Hals are rather good at that too. The method is surprisingly fast if practiced and much more direct than say that of a Monet.

  28. I’m curious, Jonathan. Although a google search lists many of your books, and some of the blurbs accompanying those listings describe you as “an accomplished painter,” I have had no success in finding any of your paintings represented online. Is there a source where we can see your work?

  29. Also, I didn’t mean to imply that scumbling was the only method of painting with opaque pigments, and gave as a caveat that my description was necessarily simplistic. The reference to Ruhemann’s remarks on Rubens was that he varied the thickness of the opaque pigment in order to take advantage of the turbid medium effect, with the thinner application appearing cooler, modulating to warmer tones w/ the thicker applications.

  30. That is a fair question William.

    All of the demonstration pieces that illustrate the books are by me. Some of them are obviously exactly that: deliberate demonstration pieces with layers of working exposed. Others are shown through to completion and I suppose these qualify as my paintings. All of this was done for a purpose though, which was to explore other painters’ methods, so that is a debatable point.

    My personal work is reserved for occasional exhibition and none of it is currently online.

    I should explain that I have another field of expertise, which now takes up most of my time. That is creative digital technology and its use in innovative projects.

    The book ‘blurb’ may also mention that I was an artists’ colourman for many years.

  31. Jonathan, you have obviously put a great deal of research into these subjects, certainly much more than that of the average practicing painter. I would have to defer to your opinion on matters such as the use of diluents used by the early Netherlandish painters. And I have to admit that simply being contrarian has to some extent motivated my comments regarding the alleged decline of technical competence since the 18th Century. I tend to agree w/ you, there has been a decline, coupled with a lack of interest on the part of the larger public. At one time a painter would specify, in detailed contracts, not only the exact composition of the materials used in a painting, but which parts would be painted by him and which by assistants, how many eyes would be visible in a portrait or a figure, the exact amount of a costly pigment like ultramarine or vermillion. The contract was often the result of negotiation between artist and patron, who took an active interest in getting his money’s worth.

    But I also believe that there are a number of painters still working with sound principles, and who take an active interest in the ongoing investigation of methods and materials, with an eye to insuring the longevity of their works. It may be true that they do not paint like the Flemish or the Venetians, but why should they? They have a wealth of new materials, which necessarily engenders a corresponding adjustment in methods.

    At the highest levels of the art of painting a painter is on his own. He must improvise solutions to expressive problems not encountered before, in order to remain true to himself and to his time. To do this, he must rely on a substantial repertoire of techniques and an understanding of materials that he’s acquired during a lifetime of study and practice. Such is the nature of expression.

    I recall the criticism that a friend of mine, and a student of R. H. Ives Gammell, once leveled against his mentor: “Gammell was a painter that could never forget his lessons.” One can see exactly what he meant when looking at a the work of many of Gammell’s students. They look like set pieces done in homage to a time since passed. The aesthetic is determined by the technique, rather than the other way around.

    Aside from being a painter I’m also a practitioner and instructor of Seido Juku karate. There are many correspondences between traditional painting and the martial arts. Both begin with a somewhat rigid study of basics of increasing complexity which eventually yield, at the highest level, to improvisation and creativity. The extreme and uncomfortable stances and formal techniques that are the foundation of karate practice would be awkward and cumbersome to use in sparring or, worse yet, a street fight. However they provide the foundation of strength, balance, speed and dynamics that inform the fighter facing an aggressive opponent.

    The mature painter is in the same position w/ regard to the blank canvas, onto which he wants to represent his experience, feelings and reflections. You’ve mentioned Titian’s mature vs his earlier works, and Rembrandt’s as well. One could hold up the later works of Homer and Kensett as well. To many these are their best works, even tho’ they deviate from their standard practice. They bring a lifetime of experience to their task, and put it to work in new and innovative ways.

    I say this because nowadays there are a number of young painters that think it sufficient to study the past in order to paint just like Bouguereau or Sargent. It’s as if the world didn’t have enough paintings of languid, slender-fingered ladies in white gowns and parasols. Or worse, that an enormous national tragedy could be reduced to an excuse for another studio set-piece of statically posed naked figures and rose petals. I’d be the first to excuse youth their enthusiasms. I was young once myself. But I think it important that the investigation of technique, historical and contemporary, be placed in its proper context w/ regards to studio practice.

  32. Agreed William, the point of knowing any of this is not simply to repeat what someone else has already done, or to work in a taught manner, followed as an unquestioned process. It is, in my opinion, so that the painter’s own creativity and imagination may be supported by the widest possible technical repertoire. When this is present the artist is enabled, when it is not, their ability to realise their ideas is impaired. Knowing how to paint does not make a great painter. Not knowing how to paint prevents an artist from ever becoming one.

    Yes, it is proper that a painter should be of their own place and time; and I would agree that the experience of life gives a mature painter something to express that a younger one may not even begin to understand. Equally a painter may gain a knowledge and understanding of methods and materials through a lifetime of practice. However, Titian and Rembrandt were already highly accomplished painters whilst they were young. The technical side was taken care of by their training, so they began with a lifetime’s knowledge to build on – more in fact: Cennini speaks of his Master and his Master’s Master, back to Giotto – and before him ‘the Greeks’ of Byzantium. How much better is it for a painter to begin painting in that position?

    Interesting that you mention Homer – and what a shame he was so closed about how he worked, although the 1880 interview tells us something, as do the two surviving watercolour sets together with the brushes and burnisher that he used. He is undoubtedly an American painter, though his methods and materials, not just his subject matter, change after his time in Cullercoats, so it is interesting to speculate on what aspects of British watercolour technique he may have come into contact with during his time there.

Leave a Reply

Your email address will not be published. Required fields are marked *