Decoding Caravaggio, Part 1

Welcome to Part 1 of a three-part series by Timothy Joseph Allen, MFA, on Caravaggio and bridging the gap between technical analysis, theory, and performance.

Decoding Caravaggio, Part 1

by Timothy Joseph Allen, MFA
www.americanartistinrome.com

Introduction

In January of 2012 I had the distinct pleasure and privilege to copy two paintings by Velázquez in London’s National Gallery, a detail of the Rokeby Venus and a copy of his late portrait of Philip IV. The mission was simple: make a facsimile of each using the commercial paints, brushes and canvas I had at my disposal, all while trying to glean some insight into how he had done it. At the time, I did have a general understanding of the materials and pigments he had used, but I was not overly concerned with adhering to any rigid interpretation. Put simply, I was looking from the outside in.

Timothy Joseph Allen with Dr. Marco Cardinali (left) and Dr. Beatrice De Ruggieri (right)
Timothy Joseph Allen with Dr. Marco Cardinali (left) and Dr. Beatrice De Ruggieri (right)

I mention this experience to better explain the nature of my current collaboration with Dr. Beatrice De Ruggieri and Dr. Marco Cardinali and our investigation into Caravaggio’s “Judith and Holofernes” at the Palazzo Barberini: in contrast to my experience in 2012, our objective has not been to make a superficial facsimile. Instead, we began the project with their technical analyses of the painting to determine the materials and pigments. Then, using their x-rays, infrared, and ultraviolet imagery, in combination with a first-hand observation of the paining itself, we were able to make better informed decisions on where specific pigments were used, in what quantity and proportion, and in what order the layers of paint were applied.

In essence, we have worked from the inside out, and are asking the fundamental question: to what extent can a technical analysis inform a convincing painting process and how close would the result come to the original when analyzed?

To conduct our investigation, we were granted the extraordinary privilege to spend a total of three days working in front of the painting in the museum: Monday, April 15th, Monday, April 29th and Monday, May 20th, 2024. During that time, two copies were executed. The first copy was painted in one day on a middle value(1) greenish gray ground. The second copy was painted over a darker ground (closer to the inherit value of raw umber) and was executed over a two day period: the first day was the block in of the lights and shadows; the second allowed for an exploration of unifying the lights and shadows with halftones made from glazing.(2)

The final step in our project will be a technical analysis of the copies, then a comparison between those results and the analyses of the actual painting. Our goal is to complete the work by the end of the 2024.

(1) For the laymen, ‘value’ indicates how light or dark a color is in relation to a grey scale. A ‘middle value’ means a color that falls in the center of that gray scale. In other words, imagine a range from 0 to 10, with 0 equal to black and 10 equal to white. A middle value is a 5.
(2) In oil painting, glazing is the technique of painting a thin, transparent layer on top of a dry layer.

Initial Observations and Discoveries

In anticipation of the project’s conclusion, there are several intriguing observations and unexpected discoveries that emerged during the process of executing the copies which merit attention:

I. Lost in Translation: ‘Earth pigments’

A technical analysis does not directly result in a list of named pigments. Instead, deductions are necessary to name the color. For example, when mercury is detected, this translates to vermilion; when lead is detected, it is lead white; when manganese is detected, that often means a kind of raw umber. But what happens when iron is detected? Iron means ‘earth pigment’, but for a painter, that could mean many things. Often when reading a technical analysis about a specific part of a painting, I find myself asking, “What exactly is this ‘earth color’? Is it raw sienna? Burnt sienna? Yellow ochre? Venetian red? Or something else?”

Obviously, part of that can be answered by reading the analysis and then looking at the painting to determine if the color of the pigment leans toward a yellow or red. But after making that distinction, would not a rose by any other name smell as sweet? No, it would not. Differences still remain that could have a dramatic impact, as I learned.

The case in point concerns the translation of the color ‘red earth’ into an actual pigment. Micro-destructive investigations of Caravaggio’s paintings indicated that “the flesh tones in Caravaggio’s paintings are produced with red pigments (earths, cinnabar and lake) diluted in lead white and suitably mixed to obtain the correct shade through the addition of lead-tin yellow and/or yellow earths.” (3)

(3) Positano, Matteo et al. “Micro-Destructive Investigations.” Caravaggio, Works in Rome: Technique and Style, vol. I, Silvana Editoriale, 06 July 2016, p. 321.

Based on my own experience, an instinctual interpretation of this name, especially in the context of a skin tone, suggests a color that is brownish in hue and low in chromatic intensity, which in turn points me to a commercially common color like ‘burnt sienna.’ Burnt sienna is a “dark reddish brown iron oxide pigment prepared by burning raw sienna”, and gets its name from the Italian city, Siena. (4)

(4) Cameo Materials Database, Museum of Fine Arts, Boston, accessed 27 May 2024,  https://cameo.mfa.org/wiki/Burnt_sienna.

Furthermore, as a practicing painter, I know burnt sienna has a relatively low tinting strength. This means that when mixed with a lead white, it does not dominate the mixture. This quality is most useful in controlling subtle shifts in value, a necessity in the creation of convincing forms and volumes, like the turn of a cheek or the circumference of a forearm.

But ‘red earth’ could also mean ‘red ochre’, which, again, does not point to one specific pigment. To demonstrate, the Cameo Materials Database lists the following as synonyms for red ochre: “red ochre (Br.); red earth; earth red; iron oxide red; red iron oxide; Indian red; brun rouge; Roter Ocker (Deut.); minium de fer (Fr.); reddle; Spanish brown; Venetian red, English red; Spanish red; caput mortuum; Indian red; light red; burnt sienna.” (5)

(5) Cameo Materials Database, Museum of Fine Arts, Boston, accessed 29 May 2024, https://cameo.mfa.org/wiki/Red_ocher.

So can the technical analysis offer additional clues? Yes. In reference to the pigments used in the ground layers, the analysis revealed the following:

“These earths are mostly made up of iron oxides with a medium to low degree of purification, less frequent use being made of umber, sienna, green earths, and ochres. In the cases examined, Raman spectroscopy identified red ochre, yellow ochre and green earth, whose primary components are respectively hematite, goethite, and celadonite.” (6)

(6) Positano, Matteo et al. “Micro-Destructive Investigations.” Caravaggio, Works in Rome: Technique and Style, vol. I, Silvana Editoriale, 06 July 2016, p. 311.

Critical here is the connection of red ochre to hematite.

This brings us to the part of the narrative where chance and circumstance play a role in the decision making, as I am sure was also true in Caravaggio’s time. In other words, I doubt that Caravaggio always used specific pigments, from a specific location, prepared in specific way, for the entirety of his peripatetic career; instead much depended on geographical location, availability, convenience and budget. Like Caravaggio, I am faced with similar criteria. Unlike Caravaggio, I have not had to live on the run, and have built up over the years a collection of various kinds of unique pigments. As fate would have it, one of these pigments just happened to carry the name “hematite.”

Figure 1: A comparison of mixtures between Lead White and Burnt Sienna (left) with Lead White and Hematite (right).
Figure 1: A comparison of mixtures between Lead White and Burnt Sienna (left) with Lead White and Hematite (right).

A couple of weeks prior to our first day in front of the painting, I decided to hand-grind some of the hematite pigment and put it to the test, comparing a mixture of it with lead white to that of burnt sienna, mixed with the same lead white (Figure 1). From that experiment, I realized two things:

  1. The hematite has even less tinting strength than the burnt sienna, making the subtle shifts in value even more manageable;
  2. The hematite is much less ‘hot’ in terms of color temperature and much less ‘red’ in terms of intensity, making it an even better ingredient in the mixture of Caucasian flesh tones, which are best when balanced between a warm and cool temperature, and are low in chromatic intensity.

As a consequence, I decided to use the hematite, not burnt sienna, as an ingredient to make the base flesh tone mixture for the copies.

Stay tuned! Soon we’ll publish Part 2 of this article, “Quest for the Holy Grail of Flesh Tones: Lead-Tin Yellow Types I and II.”