[In the following essay, Farago outlines the method by which Leonardo distinguished painting as superior to poetry, music, and sculpture. She also analyzes Leonardo's treatment of painting as a science, discussing his views on the creation of optical effects.]

AN OVERVIEW OF LEONARDO'S ARGUMENTS

The 46 passages compiled in the Parte Prima originate in various manuscripts of which only two are identified today.¹ The heterogenous nature of the Parte Prima is due in part to its being an anthology of excerpted paragraphs and in part to the many sources of its richly conceived arguments. None of Leonardo's defenses of painting seem to derive wholesale from another source, but most of his individual arguments do have precedents, in some cases so many diverse precedents that it would be difficult to favor one at the expense of others. The passages are assembled thematically and arranged in roughly five subdivisions that reflect a humanist interpretation of the liberal arts. It is unlikely that Leonardo would have arranged his treatise in the same way.² The organization is useful, however, because the range of his comparisons between painting and each of the other arts can be seen at a glance.

Leonardo's central defense of painting is that it is a science, based on perspective and defined as the branch of optics that represents things on a flat surface.³ Painting is the primary focus in all of his polemics: the other arts merely help him to define it by contrast. Thus, his discussion of sculpture revolves around his scientific investigations of relief feigned in painting. Similarly, musical harmony, which he compares to the geometric proportions of perspective, is largely based on an understanding of proportionality as something that can be expressed as a visual configuration, an idea that he probably drew from Alhazen's optical theory or perhaps more immediately from Nicole Oresme.⁴

The comparisons with poetry form a somewhat special case, because such comparisons are so prominent in Leonardo's sources. … Aside from isolated texts like Dio Chrysostom's Twelfth Olympic Oration, there is no established literature for critical comparisons of painting and sculpture or painting and music. In several drafts included in the Parte Prima, Leonardo refers explicitly and repeatedly to the classical tradition of comparative literary criticism, ut pictura poesis, paraphrasing the maxim of Simonides told by Plutarch that painting is a mute poem. But in these passages Leonardo defines painting in literary terms insofar as pictorial ornament is an attribute of descriptive poetry. Thus, in his comparisons of painting to poetry the usual roles are reversed. Here as with all the other arts, however, poetry merely provides differentiating characteristics for defining painting.⁵

The pervasive influence of literary theory is most apparent when Leonardo's critical terms are considered. His central argument concerns the nature of artifice. Leonardo claims that painting is superior because perspective is evident artifice, which ornaments painting with copious variety that delights all viewers. This artifice is due to the inventive powers or ingegno of the painter who represents natural appearances truthfully. The painter's inventions spring from his imagination, which, by means of disegno, depicts the emotions and mental states (“mental accidents”) of figures.

Much of this argument was indebted to Alberti's treatise on painting, which in turn was based on ancient Roman rhetorical theory, but the ut pictura poesis analogy also had an immediate history in artistic practice that valued ornament in a manner directly opposed to Alberti's restrained, neo-classicizing critical stance. Leonardo's coupling of rilievo and ingegno, key terms in his critical vocabulary, was fundamentally indebted to the workshop tradition that Cennini, more than Alberti, records. Leonardo praises painting in conventional terms by comparing it to poetic ornament, the discrezioni of landscape in particular, but he simultaneously devalues poetic artifice by claiming that visual images are better than verbal ones. Leonardo borrows from medieval contrasti and uses related motifs in the tradition of neoplatonic love poetry, but his adaptations are permeated with a scientific rationalism foreign to the original metaphors.⁶

One of the intriguing aspects of Leonardo's polemics on the arts is the manner in which a few motifs suggest so many different issues. The complexity of his ideas can be appreciated best by studying each individual passage, as the commentary notes accompanying the texts here do, but the main arguments against poetry, music, and sculpture can each be reduced to a sentence or two:

1. Painting is superior to poetry because it represents the works of nature, whereas poetry represents the lesser works of man, namely words, which are arbitrary conventions. A number of related issues, such as the scientific status of painted images, criteria for liberal as opposed to mechanical arts, and the reasons for preferring a visual image, stem from this basic argument.

2. Painting is superior to music because, even though both compose a “harmonic proportionality,” painting can be contemplated as a whole at once and enjoyed for a longer period of time. Leonardo also adapted this argument to poetry, which is likewise ephemeral because it is temporal; and to sculpture, which does not present a whole seen [sic] all at one time.

3. Painting is superior to sculpture because it involves more mental effort and less physical exertion. A number of related issues develop from this argument, too, such as the characterization of sculpture as a “natural body” totally lacking in artifice (and therefore inferior to painting, the artifice of which, perspective, requires great ingegno, or mental effort); and praise of difficult artifice that measures the nobility of art by the role of the artist. These passages also present additional arguments about the liberal and mechanical arts, distinguished from one another on the basis of the kind of labor required, physical or mental, with reference to concrete artistic procedures.

THE VISUAL FORCE OF PAINTED IMAGES

Displays of skill and ingenuity that emulate both nature and man-made art are at the center of Leonardo's defense of painting. Now that Leonardo's fragmentary literary remains have been ordered and medieval optics has been shown to be an important basis for the artist's definition of painting, it is finally feasible to study developmental aspects of the extensive manuscript evidence. The rest of this chapter is concerned with the ways in which Leonardo expected to create the optical effects that he associated with pictorial artifice in his theoretical considerations of painting.⁷

As we have already seen in Chapter Two, Leonardo defined painting on the model of a scientia media, a term that was applied to Aristotelian sciences that “mixed” theoretical and practical knowledge. Leonardo defended the primacy of painting over the other arts on this foundation that painting is a physical science grounded in both mathematical principles and experience, like optics. The most mature version of his definition of painting is preserved in the first chapter of the Codex Urbinas, which has direct precedents in Madrid Codex II, defining painting as a mathematical science that belongs to geometry, the investigation of continuous quantities.⁸ In the Madrid Codex drafts, however, Leonardo did not yet distinguish clearly between mathematical and physical entities, as he would in later statements in the Codex Urbinas and elsewhere concerning pictorial composition based on optical principles.⁹

As Leonardo gradually refined his definition of painting as a science capable of achieving mathematical certainty, he drew close connections between sight and imagination (in terms of painted and mental images) grounded in the geometric analysis of light associated with Alhazen's theory of direct vision and medieval, non-Aristotelian psychology of the internal senses.¹⁰ The geometry that explains how light rays strike surfaces, and the conception of the imagination as part of a complex of internal organs with the capacity first to receive external images (like a mirror), then to combine, analyze, and store them, are the foundation for Leonardo's continuing discussions of color and light in relation to the visual force of painted images. A clear schematic diagram of this geometry is preserved in the Parte Prima, Chapter 4, in a passage entitled “Principle of the Science of Painting.”¹¹

Leonardo synthesized literary and scientific theory when he discussed the visual force of painted images in terms of color and light. In many passages, he defended painting both as a science and as the most noble art because its artifice presents images to the imagination in conformity with the conditions of vision.¹² That is, painting was based on a theory of light and vision that accounted for actual appearances. For example, on Ms. G, fol. 23 verso, and CA 277 v-a, ca. 1513-1514, perhaps his latest comparison of painting and sculpture, Leonardo discussed how the artist manipulates oppositions of light and dark by selecting contrasting values for the sake of pictorial harmony. Leonardo's considerations of the bounding surfaces of objects, which developed through his study of mathematics, are the key to his evolving definitions of pictorial relief. His discussions of disegno in late writings such as the opening chapters of the Codex Urbinas identified painting completely with the science of chiaro e scuro, concerned with visible things like the (material) surfaces of bodies covered by (immaterial) color.

Only a sketchy chronology can be established for Leonardo's paintings. Extensive manuscript evidence, however, confirms and provides a rich context for examining Leonardo's developing interest in reflected color, manifested as an increasingly complex tonal structure and progressively lighter palette, as John Shearman discovered by means of formal visual analysis.¹³ As a prime example of Leonardo's late style, Shearman discusses the Louvre Virgin and Child with St. Anne. But the ongoing conservation of the Last Supper demonstrates that Leonardo had lightened his palette at the least by the mid-1490s, about three years after he recorded extensive discussions of reflected color in Ms. A. As the following discussion will clarify, even if the Last Supper was not entirely typical due to the experimental nature of the medium Leonardo employed, its coloristic qualities are in line with trends clearly noticeable in the Louvre Virgin and Child with St. Anne. New evidence arising from the conservation of the Last Supper indicates that Leonardo tried to apply color in transparent layers, probably in imitation of oil techniques indebted to the example of Flemish art, which he emulated even in his earliest Florentine works.¹⁴

But Leonardo's considerations of tonal relief and color cannot be separated in the way that many scholars today assume. Modern distinctions between “hue” and “value” are misleading categories to provide criteria against which to judge Renaissance discussions.¹⁵ In De sensu et sensato (442a), a leading source of these discussions, Aristotle arranged colors on a linear scale as intermediaries between the extremes of light and dark, and compared the mixture of colors to sounds and to the layering of pigments by painters (De sensu et sensato, 442a and 439). This seven-color scale, however, is more complex than a scale based simply on light intensity, such as Plato's color ladder of gradations arranged according to luminosity(Timaeus 67E5-8), or the even older classification of Democritus, who identified white, black, red and green with the four elements. This system apparently dominated ancient literature.¹⁶ Working within the Aristotelian tradition he inherited, Leonardo tried to formulate a theoretical model for painted color consistent with direct visual experience—or at least consistent with certain critical assumptions that he defined in terms of direct experience, such as the beauty of apparent color fully illuminated by light. Leonardo belongs to a scientific artistic tradition that needs to be distinguished from the tradition of modeling with color, most recently glimpsed in the restoration of Michelangelo's Sistine ceiling frescoes, and already discussed in medieval painting manuals as early as the tenth century.¹⁷

If he had differentiated between qualitatively and quantitatively intense light as we do today, Leonardo probably would not have arrived at the conclusion that painting can represent on optical principles the beauty of color fully illuminated by light. Like many of his contemporaries, Leonardo expressed values in relative terms as geometric ratios, not absolute measurements.¹⁸ This habit of thinking in terms of ratios and proportional relationships (a:b=c:d) allowed him to develop principles of pictorial composition that do not differentiate between the quantity of bright light and its quality of brightness. As early as Ms. A, fol. 84 recto, ca. 1492, Leonardo cautioned painters not to be misled by heightened contrasts resulting from the direct juxtaposition of white and black.¹⁹ Yet the same habit of comparing qualitative and quantitative values (expressed as proportional relationships) also allowed Leonardo to claim that the viewer who cannot tolerate a certain quantity of intense, colored light when he looks directly into the path of that light can delight in its quality of splendore, or brilliance, when he sees it from an oblique angle.

Accordingly, Leonardo devised optical principles to guide painters that take into account hue, value, and point of view, even though he never separated these considerations into such modern categories. In his early writings, he followed the same headings derived from optical treatises as Alberti, to devise categories like corpo, figura, and colore (Ms. A, fol. 92 verso); or linear perspective, perspective of color, and the perspective of moving into the distance (or more literally, “expedition” [spedizione] but there is no adequate translation for the term) which describes how things appear less defined when they are further away (Ms. A, fol. 98 recto). These categories are the precedents to Leonardo's eventual distinction between formal optics and pictorial perspective, i.e. between “natural” and “accidental” or “artificial” perspective.

THE “TRATTATO SEQUENCES”: APPARENT AND PICTORIAL COLOR

Leonardo's emerging thoughts on the qualitative aspects of color can be examined in a series of substantially intact notebooks accounting for the thirty years of his literary activity. According to this evidence, Leonardo did not articulate clear distinctions between formal optics and pictorial perspective until the end of his career: not until ca. 1508 did he define a discipline concerned exclusively with graphic representations on a flat surface. Leonardo recognized differences between natural vision and artificial perspective as early as Ms. A, ca. 1490-1492, but these categorical terms first appeared in his writings around 1508. The actual state of affairs has been clarified only very recently by Martin Kemp, who notes that Leonardo differentiated between “natural” and “artificial” (i.e., “artificial” (i.e., pictorial) perspective for the first time in writings datable to the late period of his literary activity.²⁰

For the most part, Leonardo's categories and terminology regarding the nature of perspective remained fluid throughout his career, but we can generalize upon the forty-plus definitions that survive in his notes. In their original context, his formal definitions of painting as perspective record how Leonardo progressively and definitively shifted painting away from the direct imitation of natural appearances and towards a theory of artificial pictorial composition. He consistently counseled painters to choose carefully among appearances in nature. In his early writings, however, Leonardo recommended that painters imitate as closely as possible the most subtle gradations of light and shadow,²¹ whereas in his last recorded statements he advised painters to create strong contrasts of color according to their own ingegni, but founded on a knowledge of optical principles.²²

We can examine the development of Leonardo's ideas by comparing the formal definitions of perspective embedded in discussions of painting preserved in four manuscripts for which we have a firm relative chronology: Ms. A, ca. 1490-1492; Madrid Codex II, 1503-1505; and Mss. E and G, ca. 1510-1515. In writings that span the course of Leonardo's literary career, these definitions recur in connection with loosely associated topics of discussion that are found in a repeated sequence. This recurring sequence of topics may be described as a “trattato sequence” inasmuch as the order and form of discussion follow Alberti's Treatise on Painting of 1435, which Leonardo studied at the time he wrote Ms. A. The pattern of ideas, moreover, does not occur in Leonardo's earlier discussions of perspective, such as the optical treatise Ms. C, ca. 1490, or the series of sheets from a dismembered notebook datable ca. 1489, which was incorporated into the Codex Atlanticus.²³

The “trattato sequence” occurs eighteen times in Leonardo's intact manuscripts. If partial sequences are taken into account, the sequence occurs over forty times. … With some variations, the sequence consists of a formal definition of the parts of perspective, followed by a discussion of problems of rilievo associated with painting. Comparisons of the arts, or of the senses, occur in the midst of these discussions. The sequence ends with precepts addressed to painters on the order of Alberti's prescriptions for figurative decorum in Book 2 of the della Pittura (II.34-45).²⁴ In the course of time, Leonardo progressively conflated the definition of painting as perspective and Albertian prescriptions for figurative decorum, so that pictorial decorum was described in increasingly formal terms. Late statements by Leonardo owe little to Alberti's moralizing theory of pictorial order, although these discussions preserve traces of the “trattato sequence” in both form and content. The earliest sequences are closest to Alberti, and Ms. A includes closely paraphrased passages. On the other hand, Leonardo's discussions of perspective and rilievo composed even earlier than Ms. A reveal a level of knowledge beyond anything Alberti wrote—as one might expect from Leonardo's assimilation of medieval optics by 1489-1492.²⁵

Leonardo's idiosyncratic interest in optics, which seems so peculiar from the standpoint of the history of science, is substantially clarified by the discussions of painting in the “trattato sequences.” We can gain an overview of ongoing issues he investigated in these sequences by singling out his treatment of color in relation to lustre. Leonardo approached the problem as a painter who needs to create the illusion of lustre. He examined color in painting from the viewpoint of a scientia media that reconciles experience with optical principles.

Leonardo considered the topic of lustre as part of the larger problem of reflected color, a subject that accounts for many if not most of his studies of the action of light. Lustre, a point of light which moves as the eye moves, is a special case of the general principle that the reflected image of an object can be seen from all points of a plane mirrored surface placed parallel to that object. Today we restrict the definition of lustre to that gleam or sparkle that changes and scintillates as the viewer moves—distinct highlights, that tend to settle on ridges or protrusions of lit objects.²⁶ For many years, Leonardo investigated the phenomenon of lustre—in ways departing radically from modern understanding of the phenomenon—in broader terms as the most beautiful condition of apparent color.

The subject of lustre occurs only incidentally, along with many other observations, in the early Ms. A but is reformulated with increasing insistence and clarity in later sequences. In early passages, Leonardo associated the uncolored light of lustre reflected off the surface of an object with the study of color reflected onto that object by other, colored objects in its vicinity. On folio 100 recto of Ms. A, he applied the problem of reflected color to painted rilievo, in a passage entitled “Come i corpi bianchi si devono figurare,” (how white bodies ought to be represented) which occurs in the original manuscript between two other passages included in the Parte Prima.²⁷ On folio 113 verso of Ms. A …, Leonardo defined lustre as uncolored right, distinguishing it on this basis from other kinds of light, and compared it to the beauty of colored, reflected light: “Which part of color, reasonably speaking, ought to be the most beautiful? … [with a geometric diagram]. … The illuminated part which we call rose … will be much more beautiful. … The difference between lights and lustres is that lustres are not numbered among the colors, and they are always white, and born from the crests [highlights] of bodies which are bathed [in light], and its light takes its color from the place where it is born, like gold or silver, or similar things.”²⁸

About ten to fifteen years after he wrote Ms. A, when he recorded his thoughts on lustre in Madrid Codex II, Leonardo made the same connection between lustre and reflected color on even stronger terms. On folio 26 recto, he described lustre according to three categories of colored light: the color of the light source, the color of the body that reflects the light source, and the color of the body through which this light passes if the body is transparent:

Lustre will take on much more the color of the light that illuminates the reflecting body than the color of that body. And this occurs in the case of opaque [dense] surfaces. The lustre of many umbrageous bodies is entirely the color of the illuminated body as in the case of burnished gold and silver and other metals, and similar things. The lustre of leaves, glass, and jewels will participate little in the color of the body where it is born, and a lot in the color of the body that illuminates it. Lustre generated in the depth of transparent solids has the first degree of beauty of that color, as is seen in the case of rubies, glass and similar things. This happens when all the natural color of the transparent body comes between the eye and that lustre.²⁹

Leonardo's language makes sense if we understand the phenomenon of lustre on the principle that rays causing lustre travel through a medium (such as air or water), just like other forms of reflected light that strike the surface of a mirror or some other kind of body. Reflected color was a subject traditionally discussed in optical treatises which, even before Leonardo, had entered into scientific discussions of painting.³⁰ Aristotle, as is well known, defined color as visible, sensed light that moves the imagination (De anima 428b). Color resides in the limit of the transparent; it “always inheres in the bounding surface. … Thus the conditions which in air produce light and darkness in bodies produce white and black” (De sensu et sensato 439b). Aristotle also discussed apparent color in terms of its movement through the medium of air, in the Meteorologica, and he ordered the hues sequentially on the model of nature's rainbow (Meteorologica 374b). It seems that Alberti was indebted to this discussion since he evolved the same ideas in the same order in the first book of De pictura.³¹ Alberti reserved a discussion of lustre for his second book, to describe a difficult challenge for pictorial representation: skillful painters can juxtapose light and dark pigment to imitate the appearance of shiny metal surfaces.³² Writing about two decades after Alberti, Lorenzo Ghiberti also included a passage on reflected color entitled “corpo bianco et mondo,” in his third Commentari, copied directly from an Italian translation of Alhazen's De aspectibus, perhaps even the same manuscript that Leonardo later consulted.³³ Leonardo's investigation of reflected color follows the same scientific tradition.

“BELLO RILIEVO” AND THE PROBLEM OF PUPIL DILATION

In the scientific tradition of Alhazen, Leonardo examined the optical laws governing rilievo and investigated the representability of appearances on a flat surface. By considering the difficult cases for the representation of relief, Leonardo also confronted those natural conditions in which the causes of appearances are unknown or of relative uncertainty. This brought the burden of proof to bear on the ingegno of that painter who could discover means of demonstrating nature's laws in painting.

Leonardo judged painted color according to the manner in which its bellezza, chiarezza, splendore, and varietà correspond with direct optical experience.³⁴ Since Leonardo thought about painted color in terms of apparent color, and since he defined apparent color in terms of colored light (that, according to the Aristotelian tradition, inheres in the boundaries of objects), it is not surprising that he described color using the vocabulary of light—color has beauty, clarity, and brilliance (bellezza, chiarezza, splendore), thus producing the conditions that the painter imitates to achieve varietà. The most beautiful color, and the greatest variety of colors, he argued, were produced by the natural conditions that governed lustre, a special case of reflected color, on the principles of geometry.

By demanding a close correspondence between painted images and direct experience, however, Leonardo raised a conflict not raised in the optical literature. The mingling of reflected colors in mirrors had been discussed among the errors or “fallacies” of vision as early as the second century by Ptolemy, one of Alhazen's main sources, who included it as a difficult problem of perception that potentially refutes (but actually verifies) optical theory; Ptolemy also discussed the case of illusionistic painting in his section on fallacies, describing the difference between constructions on a flat surface and actual sculpted relief.³⁵ One interesting variant of this discussion that praises the deceptive appearance of imitated relief over actual relief, occurs in fourteenth-century commentaries on Aristotle's De anima.³⁶

Leonardo, however, investigated the relationship between color fully illuminated by light and the limited ability of the pupil of the eye to endure such brightness. Among his artistic predecessors, Alberti notably avoided all consideration of ocular anatomy, and did not deal with the internal senses. The problem of pupil dilation was discussed by optical theorists, however, and Leonardo's considerations of pupil dilation have been singled out as central to his investigations in optical science, and an area in which he made original contributions.³⁷ In Madrid Codex II, Leonardo discussed at length in relation to painting the problem of how the imagination is to receive the most beautiful color.³⁸ On folio 26 verso, for example, he stated that: “When the object is strongly luminous, the pupil, not being able to tolerate it, contracts until the likeness of the luminous object reaches the pupil with diminished brightness and magnitude. Because of this diminution the sense is capable of tolerating the brightness facing it.”³⁹

Leonardo's considerations of pupil tolerance are closely related to discussions at the beginning of optical treatises in the tradition of Alhazen, who, following Aristotle's theory of the limits of sense discrimination, associated intensely bright light with pain.⁴⁰ Although he rejected the argument by ca. 1490 (in favor of intromission), Leonardo was also familiar with Pecham's discussion of the extramission of light by the eye, which followed Aristotle even further by contending that the natural light emitted by the eye plays a role in vision by moderating excessively bright lights so that they do not overwhelm the power of sight.⁴¹

Leonardo formulated the same problem concerning pupil tolerance that he had raised in the Madrid Codex more clearly about five years later in Ms. E, which was composed after further investigations of pupil dilation and image formation (found in Mss. D and F, both datable ca. 1508).⁴² Related statements on pictorial rilievo, also subsequent to these investigations, probably occur first in Libro A, ca. 1508-1510, where the exact sequence of the passages, however, cannot be reconstructed.⁴³ Leonardo defined the problem in a series of propositions in Ms. E, fol. 17 verso, dated by Pedretti 1513 or slightly later. The conflict that concerned him is that darkness is the best condition for the eye to see and know, but brightness is the best condition for essential, or fully illuminated, color:

Painting. First. The pupil of the eye diminishes as much as the quantity of illumination increases and impresses itself.

Second. The pupil of the eye increases as much as the brightness of the day or other light that impresses itself on the pupil diminishes.

Third. The eye sees and knows the things that are its object so much more intensely as its pupil dilates and this is proved in nocturnal animals like cats and other, flying [animals] like the owl whose pupil makes the greatest variation between large and small in darkness and illumination.

Fourth. The eye placed in illuminated air sees darkness behind the illuminated windows of dwellings.

Fifth. All the colors placed in shadowed places appear to be of equal obscurity among themselves.

Sixth. But all the colors placed in luminous places never vary from their essence.⁴⁴

As this sequence clearly shows, the problem Leonardo saw (taking sides with those scientists who defined categories of dark, unsaturated and bright, saturated colors) is that colors are most essentially themselves in bright light, which the pupil cannot tolerate.⁴⁵ He restated the last two propositions about color in terms of the “natural bellezza,” in a passage on the next page, folio 18 recto, under the heading “pictura”:

Colors placed in shadow participate more or less in their natural beauty as they will be more or less dark; the greatest beauty is in the luminous, great splendore.⁴⁶

Leonardo imagined a scenario in which an adversary argued that the variety of colors visible in shadow is equal to the variety visible in bright light. Leonardo disagreed, maintaining that there is less variety in the darker surfaces (vestite) of painting.

The solution that Leonardo worked out to this problem of representing the splendore and varietà of color on optical principles (first in Madrid Codex II and elaborated in the lost Libro A and Mss. E and G with respect to pictorial composition) has already been mentioned: the viewer who cannot tolerate the intense light of splendore when he looks directly into the path of that light can delight in its beauty when he sees it from an oblique angle. Thus, by selecting these conditions for his depiction (which substitute perception of the quality of intense light for its quantity), the painter can present the viewer with the beauty of fully illuminated color on optical principles. Here, then, Leonardo defined painted color in terms of natural light effects. And, consequently, he defined painted color in terms of tonal unity: without light, all colors are perceived as being dark. In this way, Leonardo transformed the quantitative aspects of color into qualitative aspects that had critical values attached to them. And in doing so, he transformed words like splendore, chiarezza, and obscurità from the descriptive vocabulary of optics to the critical language of art.⁴⁷

PRINCIPLES OF THE SCIENCE OF PAINTING: TREATMENT OF BOUNDARIES

That Leonardo intended to provide the art of painting with a set of unified, scientific principles is suggested by the two distinct kinds of writings in which he treated problems of pictorial composition: advice addressed to students in the form of precepts, and discussions of optics, often in the form of propositions like the passage just cited from Ms. E, concerning the appearance of rilievo in painting. The original editor of the Codex Urbinas destroyed the organic relationship of these discussions when he separated passages connected in the original manuscripts to arrange them according to subject matter. Their context must, therefore, be studied in the original manuscripts. In their original order, that we have dubbed the “trattato sequences,” Leonardo recorded problems formulated with reference to optical principles tested against various kinds of experience such as descriptions of direct observations and imagined situations, “thought experiments,” geometric sketches and calculations, and practical demonstrations. In other passages, he reduced his conclusions to the rules of a preceptive “art” suitable, like Alberti's treatise, for training students. These are the passages, most familiar to modern readers from the Trattato, commonly associated with Leonardo's pictorial procedures. Yet the precepts are intimately connected with, and truly inseparable from, the specific scientific discussions they once accompanied.

We can fully understand how the subject of lustre figured into Leonardo's considerations of painting only by turning to a second pictorial problem that he investigated through his study of formal optics. Discussing principles of pictorial composition in Madrid Codex II around 1503-1505, Leonardo was concerned with the problem of representing the most beautiful (i.e., fully illuminated) color, given the limited tolerance of the pupil to let light into the internal sense of the imprensiva. In the later Mss. E and G, the focus of discussion gradually shifted from the relationship between light and the internal senses of the explanation of the causes of observed phenomena. Comparison of the late writings with Madrid Codex II and Ms. A, and related evidence in other notes, suggests that in the course of his investigations of optics, Leonardo synthesized two distinct problems of pictorial representation that he drew from his Albertian inheritance and qualified on the basis of his scientific studies. One problem is what drawn lines can correspond to in nature, given that mathematical lines are not visible. The other is how to represent the greatest beauty of color on optical principles, given that the pupil cannot tolerate splendore. Beginning with his studies of Euclidean geometry around 1497, and during his subsequent investigations of light and shadow ca. 1502-1505 (and until his latest writings), Leonardo developed the basis for a new definition of painting that made the distinction between mathematical and physical line (a commonplace of Euclidean geometry) an important consideration.⁴⁸

The core of a consistent program for the representation of optical relief, the basis of Leonardo's last definitions of painting, appears in Mss. E and G. Here the artist was primarily concerned with light reflected at the boundaries (termini) of colored, curved surfaces, to the variable extents that these boundaries are distinct. In Ms. G, on fol. 23 verso, Leonardo gave a clear definition of artificial, or pictorial, perspective in these terms, corroborated by many other notes of the same period. Seen from the standpoint of his pictorial considerations in Ms. G, Leonardo's investigations of lustre in Ms. A and Madrid Codex II make much more sense:

The primary principal part of painting are the fields (champi) of painted things in which the boundaries (termini) of natural bodies have convex curvature. One always knows the figures of such bodies in the fields even though the colors of the bodies are the same color as the fields. This comes about because the convex boundaries of the bodies are not illuminated in the same mode by the same light that illuminates the field. Because of this, the boundaries are often more bright or more obscure than the field.

But if such a boundary is the same color as the field, without doubt, information about the figure with such a boundary will be prohibited in that part of the painting. And the choice of painting such as this is loathed by the ingiegni of good painters who know that it is the intention of the painter to make his bodies project from these fields. And in the said case the contrary thing happens—not only in the painting, but also in the things in relief.⁴⁹

In his late writings on painting, notably in Libro A and articulated even more fully in this passage from Ms. G, Leonardo seems to have reconciled the nature of line to the problem of representing the beauty of color: light is reflected at the edges of objects (i.e., boundaries seen in oblique view) differently from the way it is reflected from surfaces that face the viewer. His considerations, like Alberti's, belong to the Aristotelian-Euclidean science of optics that defines boundary as a surface which becomes an edge when it is seen in profile view. Lines are contours, and contours define the continuous gradation of surface, that is sculptural relief. But Leonardo's theory of pictorial composition is based on optical principles that, even more than Alberti's, evolved from the physical science of optics. Discussion of the difference between natural and mathematical lines in earlier mathematical theory was also readily available to Leonardo in a number of forms, including the practical handbooks of geometry that he owned at the time he wrote Madrid Codex II⁵⁰ Leonardo “corrected” Alberti, who recognized the difference but nonetheless equated mathematical lines with drawn lines (or marks on a surface), and said that the lines that can be represented in painted relief are the edges of objects visible in conditions related to lustre on the principles of geometry.⁵¹ In other words, continuous points of lustre are joined to make visible, physical “lines.” We might think of the way the edge of an object sometimes glistens or gleams when the light strikes it in a certain way.

Beyond the simple incidence of lustre, Leonardo's main principle of pictorial composition in its most fully developed form, still within the Albertian tradition, was based on the optical principles governing the complex phenomenon of heightened contrast which results from the direct juxtaposition of different intensities of reflected colored light. Painters who do not develop their compositions carefully will produce images lacking in grazia according to a statement in Libro A, ca. 1508-10:

Objects against a bright and illuminated field display much more relief than against a dark one.

The reason for this proposition is that if you wish to show relief in a figure, you must depict it so that the part of the body farthest from the light participates least in that light, and thus will be darker, and since it terminates against a dark field, its boundaries become confused. For this reason, if no reflected lights fall on it, the work remains without grace and from a distance nothing will show except the luminous parts. It is correct, therefore, that the other parts be of a darkness similar to the same field from which the things appear detached and be kept less [dark] than they should be.⁵²

Overwhelming evidence suggests that Leonardo's desire to endow painted images with the greatest beauty of apparent color largely shaped the way he addressed pictorial issues. We are used to thinking about Leonardo's “tonal composition,” but our modern vocabulary mitigates the central role played by colore, defined in Aristotelian terms different from our own, throughout the scientific artistic tradition.

ON LEONARDO'S CONTRIBUTION TO PICTORIAL PERSPECTIVE

The problem of depicting line and color on unified optical principles was central to the artistic tradition Leonardo inherited and tried to improve by providing painters with principles to manipulate pictorial structure. Leonardo's statements about composing paintings on optical principles seem closely related to the pictorial inventions we associate with new styles of optical naturalism, from Raphael's Roman period to painters associated with the reform of Italian painting in the later part of the sixteenth century, like Federico Barocci and the Carracci.⁵³ Even today we tend to think of painting that imitates optical effects as being devoid of artifice, although every image constructed of colors on a surface is, of course, completely artificial.

Leonardo's formulations have significant consequences for later scientific, artistic practices, but he was by no means the only writer to describe painting as a mathematical science. Like Lorenzo Ghiberti, Filarete, and Piero della Francesca, he followed Alberti's precedent.⁵⁴ Beyond this, issues that concerned Leonardo interested many other artists and writers. The scientific investigations of fifteenth-century artists were foreshadowed in Cennino Cennini's remarks that painting is scienza that deals with rilievo.⁵⁵ Less studied but equally significant, as we have seen in Chapte Two, are new versions of an Aristotelian classification of knowledge proposed by leading Quattrocento humanists who aligned painting, sculpture, and architecture with the new productive, applied sciences, as well as the liberal arts. By the mid-sixteenth century, arguments for the liberal (or near-liberal) status of the visual arts were conceived in a fully Aristotelian framework that, like Leonardo's, drew on Galen's discussions of methodology.⁵⁶

Historians of science like David Lindberg have recently suggested that Leonardo was an autodidactic thinker who had a poor conception of the issues facing formal optics. In relation to the history of art, however, his scientific investigations had significant historical consequences. For Leonardo, the primary importance of the geometry that explains the physical action of light is the analogy between mental images, painted images, and images reflected in mirrors. Leonardo championed the nobility of sight in Scholastic terms, and furthermore he investigated vision on the mechanical model of percussive motion. The implications of his theory of vision for the status of art as knowledge are enormous. In his comparison between painted images of God and “any other science of human work,” in Parte Prima Chapter 7, for example, Leonardo followed the tradition of Hugh of St. Victor and Bonaventure to argue eloquently for the power of the artist's ingegno. Leonardo's argument also recalls medieval justifications for images based on their ability to communicate the intelligible. St. Augustine had described the didactic function of “illumination” as throwing “light upon meaning” for those who desire instruction (De doctrina christiana, 4.30). He distinguished between good, functional artifice (which instructs) and bad artifice, which is merely decorative, resting on false doctrines. Writers following Augustine described figures of speech as “images” indispensible to poetic language because, like painting, verbal images communicate by signs and function as language for the illiterate, as in the case of the Biblia pauperum. Thus (borrowing an ancient rhetorical paradox), illusionistic painting is more true the more false it is. Leonardo's conception of painted images, that artifice is praiseworthy if it is substantive, if it not only delights but also moves and instructs the beholder, is in effect an answer to Plato's condemnation of sophistry in the form that it was elaborated by the Christian tradition of St. Augustine.

The praise that critics accorded painted rilievo constructed on a model of perception that Leonardo inherited from his predecessors (and bequeathed to his successors) is well documented by seventeenth-century writers such as Mosini, Agucchi, Félibien, Bellori, and even by less theoretical writers like Gerolamo Tezio, the author of the guide to Urban VIII's “Museum” in the Palazzo Barberini, mentioned in Chapter Two. Leonardo's writings on light and shadow, color, and atmosphere took on a new importance in the seventeenth century. Mathematical treatises on perspective did not furnish academically trained painters with the information they needed to compose bello rilievo. Writings devoted to shadow projection, which also dealt with the tonality of surfaces (i.e., rilievo), come closest and artists utilized these texts.⁵⁷ But Leonardo's writings on optics in the tradition of natural science, and other treatises in the same vein (which included Zaccolini's commentary on Leonardo's optics, treatises by Accolti, Cardano, Cigoli, Lomazzo, and others who are less well studied, that have been mentioned in Chapter One), met artists' needs for the graphic treatment of rilievo, and especially for a systematic explanation of color based on optical principles.

Growing interest in Leonardo's writings since the mid-sixteenth century points to a complex situation, but from the information presently available it is already obvious that of all his writings, his statements on optics in relation to painting most interested, and indeed preoccupied, artists and patrons by the opening decades of the seventeenth century. Sixteenth- and early seventeenth-century artists had access to Leonardo's optical writings from the original notebooks and especially from abridged manuscripts of the Trattato.⁵⁸ While these artists would not have had the overview of Leonardo's writings that we have today, their own familiarity with the same basic sources in Euclid, Galen, and the optical literature would have enabled them to grasp the pictorial issues he addressed. Indeed, from the critical literature as well as the visual evidence, it might be suggested that the artist's inventive powers, his license to invent, were redefined in terms of the optical science concerned with the representation of surfaces, a science that was exclusively the province of painters in the late sixteenth and early seventeenth centuries, when the modern separation of art and science did not yet exist.⁵⁹

Counter-Reformation writers and their seventeenth-century successors who wrote about artistic invention were aware that the faculties of the artist's imagination remain free to combine and analyze images, as poets and painters traditionally had license to do, but they added the new stipulation that their inventive powers be subordinated to the decorum of the subject matter defined in conformity with the conditions of vision. Thus, optical effects were appropriate for the representation of supra-natural events such as the celestial visions favored by post-Tridentine painters of devotional subjects. Images like these supposedly called attention away from the artist's skill by presenting an image so real that it could be taken for the event itself. The critical dimensions of these developments have a long history.

Notes

1. The two known texts are Ms. A and the lost Libro A, reconstructed by Pedretti (1964). As the commentary notes to the Parte Prima texts reveal, because of the nature of this anthology, each passage must be established independently of the others in the context of Leonardo's surviving original notes.

2. See the introduction to the Parte Prima texts.

3. There is a long history of scholarly debate concerning Leonardo's views on perspective; for a recent review, see Veltman, Studies on Leonardo da Vinci I, Introduction, and CN 3-6 here.

4. See discussion at CN 31.

5. On the other hand, when he compares painting to poetry with reference to compositional procedures, he probably depends on the theory of music rather than directly on literary theory. See CN 25.

6. Echoes by his contemporaries like Lorenzo de' Medici, Ficino, and Poliziano are discussed in CNs 22 and 24.

7. In addition to the recent ordering of Leonardo's literary remains by Carlo Pedretti and Anna Maria Brizio, significant specialized studies have contributed to a clearer understanding of Leonardo's optics; see Lindberg, Theories of Vision, 154-168; M. Kemp, “Leonardo and the Visual Pyramid”; Ackerman, “Leonardo's Eye”; Strong, Leonardo on the Eye; and most recently Veltman, in collaboration with Keele, Studies in Leonardo da Vinci I, 30-142 on the long history of scholarly debate concerning Leonardo's views on perspective. M. Kemp, review of Veltman, Studies in Leonardo da Vinci I justifiably takes Veltman's methodology to task. Lindberg, Theories of Vision, 168, writes that Leonardo frequently expressed a “confused and garbled form of traditional theory” that showed no understanding of the central issue of traditional optics, namely the problem of the multiplication of rays influencing all parts of the eye, and he tended to treat radiation in “an unsatisfactory holistic manner.” Like the majority of studies to date, the analysis of Leonardo's optics by Eastwood, “Alhazen, Leonardo, and Late-Medieval Speculation,” refers Leonardo's innovations to the history of science. The art historical scholarship has also emphasized the progressive aspects of Leonardo's optics at the expense of the Euclidean foundation of his views. As James Elkins recently noted in “Did Leonardo Develop a Theory of Curvilinear Perspective?,” the Euclidean basis of Leonardo's knowledge is a significant factor even in his late writings, for example in his considerations of image size dependent on the viewer's angle of vision. The importance of the Euclidean tradition to Renaissance artists in general deserves further study, but see now, M. Kemp, The Science of Art. Euclid's optics was available in Leonardo's day in the recension of Theon of Alexandria: on the manuscript tradition, see Lindberg, A Catalogue of Medieval and Renaissance Optical Manuscripts, especially “version 1” of De visu, discussed on pp. 50-52, the text to which Leonardo most likely had access. For direct references to Euclid's Optica in Leonardo's manuscripts, see Hart, The Mechanical Investigations of Leonardo da Vinci, 62-63; and Pedretti, Commentary, passim.

8. The definitions in Madrid Codex II, are found on fols. 62bis/r, 66r, and 67v. Another draft is preserved in the Codex Urbinas, Parte Prima, Chapter 33. Pedretti, Commentary, 1: 122-123, discusses Leonardo's definitions of painting, with a transcription and translation of Madrid Codex II, fol. 67r.

9. Among the earliest indications of a changed attitude is a passage on CA 68 v-a: see Appendix 2.

10. A study of Alhazen that, unfortunately, appeared too late to be used here is Sabra, The Optics of Ibn-al-Haytham.

11. Compare Pecham, Perspectiva communis, I.29 ｛32｝ (ed. Lindberg, 111, with a diagram): “The eye would be unsuited for perception of size if it were not round.”

12. A majority of these passages are collected in the Parte Prima of the Codex Urbinas. A guide to related passages in Leonardo's extant notes is given in Appendix 2.

13. J. Shearman, “Leonardo's Color and Chiaroscuro.” For a historiographical overview, see Gage, “Color in Western Art.”

14. On the restoration of the Last Supper, see Brambilla Barcilon, Il Cenacolo di Leonardo, especially p. 12; Brown, Leonardo's Last Supper. On Leonardo's color, see Matteini and Moles, “A Preliminary Investigation,” and “Il Cenacolo.” I am grateful to Jennifer Rashleigh, to whom I owe these references, for discussing the issues with me. See her Master's thesis, “The Science of Leonardo's Art,” 57-62.

15. For example, Cropper, “Poussin and Leonardo,” 570-582, who states (578) that Pietro Accolti's joint interests in relief and color “seem to reflect his criticism of the continuous quality of Leonardo's sfumato, where color is often sacrificed to shadow, or where relief melts away into an infinity of reflections”; and (581) that Poussin's ordering of light and darks in Rebecca and Eliezer at the Well, Paris, Louvre, painted in the 1640s, “is quite contrary to the sfumato of Leonardo's practice.” A similar misconception underlies Dempsey's Annibale Carracci, in which he superbly analyzes (30-36) how sixteenth-century painters handled color, a subject of considerable complexity that deserves even further attention, yet Dempsey also claims that “Leonardo had all but banished color from his art through his infinitely minute divisions of the half-tones of reflected light” (31). The basis of all Renaissance discussions of color is Aristotle, who also defined the kinds of colors as originating from the ontologically essential white and black, which are contrary extremes between which every color comes to be or passes away (Generatione et Corruptione 328b). According to this principle of arranging colors on the ontological model of opposing contraries, paired colors had opposing values. White and black were one accepted pair, followed either by yellow (or red) and blue, and red and green (or gray). This organization could suggest other, similar patterns such as musical notes to name an analogy offered by Aristotle, or patterns of painted color, as Alberti suggested (On Painting, Book II.48). For a useful overview of Renaissance terminology, see Gavel, Colour. On the Aristotelian tradition of color inherited by Renaissance artists including Leonardo, and its later association with musical theory, see now M. Kemp, The Science of Painting, 264-284 ff., and my remarks in the Acknowledgments, n. 4; here, n. 36.

16. Democritus identified white, black, red and green with the four elements. This classification is rudimentary in its understanding of hue, as compared with Aristotle's classification. Democritus differentiated light intensities alone; Aristotle ordered hues sequentially on the model of nature's rainbows. Pliny, Galen, and the anonymous author of De coloribus demonstrate the tenacity of Democritus's views (see de Santillana, The Origins of Scientific Thought, 165).

17. See Bulatkin, “The Spanish Word ‘Matiz.’” On related Renaissance uses of color as evident artifice, see Summers, “Contrapposto.”

18. On the practical applications of the “Rule of Three” and other useful kinds of computation based on proportional relationships, like gauging, see Baxandall, Painting and Experience in Renaissance Italy, 86-108.

19. Leonardo, Ms. A, fol. 84r: “On painting.” The varied comparisons of various qualities of shadows and lights oftentimes make the painter who wants to imitate and counterfeit things that he sees hesitant and confused. The reason is this: if you see a white drapery compared with a black one, the part of the white drapery that is juxtaposed to the black one will appear much brighter than the [same] part juxtaposed to the greatest whiteness, and the reason for this is proved in my perspective. (De pittura. Il vari paragoni delle varie qualità, d'ombre e lumi fanno spesse volte, ambiguo e chonfuso il pittore che vuole imitare e chontraffare le cose che vede. La ragione è questa: se tu vedi un panno biancho, al pari ad uno nero, cierto, quella parte d'esso panno biancho, che chon finerà chol nero, apparirà molto più chandida che quella che confinasse con maggiore bianchezza. E lla ragion di questo si prova nella mia Prospettiva.) Transcription based on Ms. A., ed. de Toni/Corbeau.

20. M. Kemp, “Leonardo's Visual Pyramid,” 147-148, citing Codex Arundel 62r (R. 109, c. 1508), one of the earliest recorded statements of this distinction, along with discussions in Libro A; see especially Cartas 16, 34, 64, and 68). We usually (mistakenly) assume that these categories were clearly differentiated by the time of Leonardo. Alberti, notably, never used the word optica or perspettiva in his treatise on painting (1435/1436), despite the long discussion he devoted to the subject we call pictorial perspective. Leonardo's reluctance to define pictorial perspective has caused a great deal of confusion among modern students of his writings, who have transferred his late terminology to discussions of his early writings. This confusion has been perpetuated by the most comprehensive ordering of Leonardo's manuscripts to date: see Pedretti, Libro A, 171, and Commentary, 1: 119; Strong, Leonardo on the Eye, xxxii, expresses agreement with Kemp, but his study depends on Pedretti's widely accepted views (compare Strong, 208).

21. For example, on Ms. A, fol. 100v: “Dell'elezione dell'aria, / che dia grazia ai volti. / Se avrai una corte, da potere a tua volontà coprire con tenda di lino, questo lume sarà buono. Ovvero, quando vuoi ritrarre uno, lo ritrai a cattivo tempo, sul fare della sera, … ai volti d'uomini e donne quando è cattivo / tempo; quanta grazia e dolcezza si vede in loro! … E questa è / perfetta aria.” Transcription based on Ms. A, ed. de Toni/Corbeau.

22. For example, in Ms. G, fol. 23 verso.

23. The Codex Atlanticus sheets are related to Anatomical Ms. B, fol. 42r, dated April 2, 1489, according to Brizio, “Correlazioni e rispondenze”; the texts are collected in Brizio, “Fogli d'anatomia e di ottica,” in Scritti scelti, 153-172. See further Pedretti, Commentary, to R. 60 (CA 138v-b). On Leonardo's study of Alberti, see Clark, “Leon Battista Alberti on Painting”; Zoubov, “Leon Battista Alberti et Leonardo da Vinci”; Pedretti, Libro A, 59-61, 117, and passim.

24. Alberti, On Painting, 2: 35-45. Furthermore, Leonardo's scientific discussions of painting correspond to the first book of Alberti's treatise, while his comparisons of the arts correspond to Alberti's praise of painting at the beginning of Book 2. Many of Leonardo's arguments in defense of painting repeat points made by Alberti. On the history of the term, see Freedman, “‘Rilievo’ as an Artistic Term in Renaissance Art Theory.”

25. Problems of interpretation concerning Leonardo's understanding of optics around 1490 are discussed most fully by M. Kemp, “Leonardo's Visual Pyramid,” and Ackerman, “Leonardo's Eye.”

26. As the following discussion elaborates, Leonardo defined lustre in terms never intimated by Gombrich in his two classic essays reconstructing historical notions of lustre from both visual and textual evidence, “The Heritage of Apelles,” and “Light, Form and Texture in Fifteenth Century Painting,” both reprinted in The Heritage of Apelles. Leonardo would not have disagreed with Gombrich's characterization, but his conception led him to incorporate other phenomena as well in his definition.

27. This paragraph continues a discussion on the adjacent page, folio 99, a polemical comparison of painting and poetry included in the Parte Prima as Chapter 19, by adding to that discussion that surfaces such as faces are “animated” by shadow and light, which are the components of aria. It also precedes another passage included in the Parte Prima, as Chapter 12, so its relationship to Leonardo's polemical comparisons of the arts could not be clearer. See also, CN 19.

28. In Ms. A, fols. 112-113, Leonardo discussed light, including lustre. The full text of the passage on folio 113 recto is: “Dico che il lustro, perche è tutto per tutto e tutto nella parte, che, stando nel punto. d., il lustro sembrerà nel punto. c., e tanto quanto l'occhio si tra muterà dal. d. all'.a., tanto il lustro si tramuterà da.c. ad.n. [fol. 113v:] Quale parte del colore, ragionevolmente, deve essere / più bella? [Figure]. Se.a. sarà il lume,.b. sarà illuminato per linea da esso lume.c., che non puo vedere esso lume, vede solo la parte illuminata, la quale parte diciamo che sia rossa. Essendo così il lume ch'essa la faccia.c.. E se.c. sara ancor esso rosso, vedrai essere molto più bello che.b.. E se .c. fosse giallo, vedrai li crearsi un colore cangiante fra giallo e rosso. Che differenza è da lumi a lustri è come i lustri non sono nel numero dei colori, ed è sempre bianco, e nasce nei colmi dei bagnati corpi, ed il lume è / del colore della cose dove nasce, come oro od argento o simili cose. Transcription based on Ms. A., ed. de Toni/Corbeau.

29. Madrid Codex II, fol. 26 recto: “Il lusstro participa assai più del color del lume che allumina il corpo che lusstra che del colore d'esso corpo, e cquesto nasscie in superfitie dense. Il lustro di molti corpi onbrosi e integralmente del colore del corpo alluminato, come quello dell'oro brunito e argiento e latri metalli e simili corpi. I' lusstro di foglie, vetri e gioie poco participerà del colore del corpo ove nasscie e assai del color del corpo che llo allumina. I' lustro fatto nella profondita de' densi trasparenti, sono in primo grado della belleza di tale color, come si vede dentro al rubino balasscio, vetri e ssimil cose. Quessto acade che' infra ll' ochio e esso lusstro s'interpone tutta il colore natural del corpo transparente.” Emphasis added. Transcription based on Madrid Codex II, ed. Reti, 5: 39.

30. For the closest parallel in optical treatises to Leonardo's formulation of the problem of reflected color, see Pecham, Perspectiva communis, 88-89, Proposition 1.14 ｛29｝ (based on Alhazen, Opticae thesaurus, Bk. 1, sec. 17; and cited by Witelo, Opticae, [bound as part of Alhazen, Opticae Thesaurus], Bk. 4.156. The problem of reflected color was also discussed by Robert Grosseteste and Roger Bacon; see Lindberg, Theories of Vision, 245, n. 43. This and other correspondences between Witelo and Leonardo were first discussed by Tea, “Witelo prospettico del secolo XIII.” The problem of reflected color can be traced to Aristotle, De sensu et sensato 439a30 ff.; and the anonymous author of De coloribus, IV.793b24. The medieval transformations point to a widely diffused tradition. Leonardo's discussions related to Ms. A, fol. 100r (cited in n. 27), have been assembled by Maltese, “Leonardo e la teoria dei colori.”

31. Another important early record attesting to the acquaintance of artists with Aristotle's discussion of color in the Meteorologica is Giovanni da Fontana's (lost) treatise on perspective, which treated gradations of color and light (not linear mathematical perspective), known through a publication of 1544. Fontana was a leading personality in Venetian scientific culture around the middle of the fifteenth century; his treatise cited Biagio Pelacani as his “master,” and the work was dedicated to Jacopo Bellini. See Canova, “Riflessioni su Jacopo Bellini.” Perhaps the earliest indication that Aristotle's scientific discussion of color was useful to artists is Francesco della Lana's description of drapery folds in terms of advancing and receding color, in his commentary on Dante's Paradiso, Book 24.25-27 (first noted by von Schlosser, La letteratura artistica, 91; see most recently Summers, The Judgment of Sense, 16, n. 8).

32. In his discussion of the science of painting, Alberti, On Painting 1.11, observed that: “Reflected rays assume the colour they find on the surface from which they are reflected. We see this happen when the faces of people walking about in the meadows appear to have a greenish tinge.” Alberti discussed lustre in a separate passage concerning pictorial composition (Book 2.47): “This composition with white and black has such power that, when skillfully carried out, it can express in painting brilliant surfaces of gold and silver and glass.”

33. Both texts are cited by Federici-Vescovini, “Il problema delle fonti ottiche medievali del Commentario terzo di Lorenzo Ghiberti,” 370 (see also p. 369, on the contraction of the pupil in bright light). There is another fifteenth-century treatise on perspective in the vernacular that is indebted to Alhazen: variously attributed to Alberti, Leonardo, and the mathematician Paolo Toscanelli, it is reprinted by Parronchi, who attributes it to Toscanelli (Studi su la dolce Prospettiva, 599-641).

34. Leonardo's use of splendore and chiarezza corresponds closely to his sources in optical treatises, while his valuation of varietà is also indebted to Alberti; see further discussion at CNs 20 and 21. On the critical value of optical metaphors pervasive in rhetorical theory and literary criticism, such as chiarezza, lustre, splendore, obscurità, which were returned to their original context of the visual arts by Renaissance writers, see Summers, “Contrapposto”; and Baxandall, Giotto and the Orators. A study that deals with Leonardo's optics in relation to his pictorial conventions but does not account for the mitigating role played by critical metaphors is Gombrich, “The Trattato della Pittura: Some Questions and Desiderata.”

35. Ptolemy included imitated relief among the errors of vision, where he also discussed the manner in which apparent color mingles with color reflected from other things in mirrors (L'Optique de Claude Ptolémé, Book 2.108, 67 ff. on the mixture of colors reverberating in a mirror; Book 2. 124, 74 ff. on illusionistic painting); Book 4, devoted to the mingling of colors in reflected light. On the contribution of the fallacies of vision to Renaissance art, see Summers, The Judgment of Sense, 36.

36. Marshall, “Two Scholastic Discussions of the Perception of Depth.” The problem of reflected color can be traced to Aristotle, De sensu et sensato, 439a30 ff., and the anonymous author of De coloribus, a text attributed to Aristotle and widely read during the Renaissance period (Aristotle, Minor Works, iv.793b24). Medieval transformations of the problem point to a widely diffused tradition which became conflated with certain Platonic ideas. As early as De coloribus, a distinctly Platonic list of simple colors is combined with observations of nature dependent on Aristotle. Compare Bartolomaeus Anglicus (On the Properties of Things, p. 1268). For studies of Renaissance color, see Ackerman, “On Early Renaissance Color Theory and Practice”; Edgerton, “Alberti's Color Theory”; Hall, ed., Color and Technique.

37. Lindberg, Theories of Vision, 162, 168; see also Strong, Leonardo on the Eye, 343-345.

38. On fols. 24 verso and 26 verso, the pages immediately preceding and following his discussion of lustre in the Madrid Codex, Leonardo considered how the internal organ of the imprensiva will respond to intense light. On Leonardo's term imprensiva, a physical organ of perception which receives the images of objects, is apparently an original coinage that corresponds roughly to Avicenna's imaginatio; see further discussion in CN 2. All the passages on painting in Madrid Codex II concern the relationship between the imagination and perceived color; see Appendix 1 for a summary of the subjects.

39. Madrid Codex II, fol. 26 verso: “Quando il predetto obbietto sarà forte luminoso, la popilla, non la potendo soportare, si fa ttanto minore, che la similitudine di tale luminoso obbietto viene alla popilla non manco diminuità de ssple[n]dore che di magnitudine. Per qual diminuitione il senso po suportare l'antiposto splendore.” Transcription based on Madrid Codex II, ed. Reti, 5:41. A closely related discussion occurs on fol. 27r: see the Appendix. As Martin Kemp notes in “Leonardo's Visual Pyramid,” 137, one of the earliest coherent discussions of pupil contraction in bright light, in relation to the brain, is recorded in the Codex Forster II; 2.2, fol. 158 verso, c. 1495 (Il Codice Forster, 280-281). Kemp also cites a group of related passages in the Codex Atlanticus, of uncertain date (on the dating of these passages, compare Pedretti Commentary to R. 111), but he did not discuss Leonardo's treatment of this problem in Madrid Codex II.

40. Sense discrimination was described in terms of pain and pleasure by Aristotle throughout De anima. The tradition developed through Galen whose De usu partium, Book 10 (on the eye), became another textual source for the association of sight with pain. Leonardo's knowledge of Galen is discussed in Chapter Two. On the history of the textual tradition, see Lindberg, “Alhazen's Theory of Vision and Its Reception in the West,” especially 322-323.

41. Lindberg, Introduction to Pecham, Perspectiva communis, 34-35, on proposition 1.46 ｛49｝. Leonardo was aware of Pecham's argument, which he rejected ca. 1490 (see further discussion by Ackerman, “Leonardo's Eye,” 127 ff.). But as late as ca. 1510 Leonardo referred to Pecham's examples of the nocturnal vision of animals, for example in Ms. E, fol. 17v.

42. My dating of these manuscripts follows the persuasive argument by Strong, Leonardo on the Eye, 212 ff.

43. Pedretti, Libro A, 25. These passages are included in Appendix 1.

44. Ms. E, fol. 17v: “Prima / Pictura. La popilla dell'ochio diminuisscie tanto la sua quantità quanto ecresscie illuminoso che in lei s'inpreme. / Seconda / Tanto cresscie la popilla dell ochio quanto diminuisscie la chiarezza del giorno od altra lucie che in lui s'inprema. / Terza / Tanto più intensivamente vede e chonosscie l'ochio le chose ch'elli stanno per obbietto quanto la sua popilla più si dilata e questo proviamo mediante li animali nocturni chome nelle ghatte e altri volatili chome il ghufo e ssimili li quali. La popilla fa grandisima variatione da grande appichola ec nelle tenebre o nella luminato. / Quarta / L'ochio possto nell'aria illuminato vede tenebre dentro alle finesste delle abitationi alluminate. / Quinta / Tutti li cholori posti in llochi onbrosi paiano essere de quale osschurità infra lloro. / Sesta / Ma ttutti li cholori possti in lochi luminosinosi varian mai della loro essentia.” Transcription based on Ms. E, ed. Ravaisson-Mollien. On lustre, see also Ms. E, fol. 31 verso, cited in Appendix 1.

45. On these categories of apparent color, See Gavel, Colour. A Study of its Position, 27, who notes that lucidus means distinct or saturated, just as obscurus means indistinct or unsaturated colors in Ptolemy's Optica, although saturation was elsewhere connected with dark colors. See further discussion of Leonardo's study of saturated color in M. Kemp, The Science of Art, 267—269, connecting Leonardo's definitions with his knowledge of De coloribus around 1506 and his reading of Roger Bacon's Opus majus, which reinforced his fascination with colored lights and related phenomena.

46. Ms. E, fol. 18r: “Pictura / Li cholori possti nelle onbre participeranno tanto più o meno della lor natural belleza quanto essi saranno i minore o in maggiore osschurità. / Ma se lli cholori saran situati in ispatio luminoso allora essi dimossteran di tanta magiore belleza quanto iluminoso sia di maggiore splendore” (“chiarezza” preceding “splendore” and cancelled). Transcription based on Ms. E., ed. Ravaisson-Mollien.

47. Optical splendore refers to highlights and lustre, but as a metaphysical principle, splendore signifies radiance, a reflection of supra-natural light; see Barasch, Light and Color in the Italian Renaissance, 171-179.

48. See discussion in CNs 1, 3-6, 9; and see passages cited in Appendix 2.

49. Ms. G, fol. 23v: “De pictura / Principalissima parte della pittura son li chanpi delle chose dipincte nella qu li chanpi li termini delli chorpi naturali cheanno in lor churvita chonvessa senpre si chognosschano le figure di tali chorpi in essi chanpi anchora chelli cholor de chorpi sieno del medesimo cholor del predecto chanpo ec questo nasscie chelli termini convessi de chorpi non sono alluminati nel medesimo modo che dal medesimo lume è alluminato il chanpo, perché tal termine molte volte sarà più chairo oppiù osschuro che esso chanpo. / Ma sse ttal termine è del cholore di tal chanpo sanza dubbio tal parte di pittura proibiera la notitia della figura di tal termine, ecquesta tale eletione di pictura he da essere sciftata dalli ingiegni de buoni pictori chonciossia che lla intentione del pictore è ddi fare parere li sua chorpi di qua de canpi, e innel sopra decto chaso achade in chontrario, non che in pictura ma nelle chose di rilievo.” Transcription based on Ms. G, ed. Ravaisson-Mollien. See Appendix 1 for other related notes of this period.

50. On Leonardo's collection of geometry books known as abbachi, see The Madrid Codices, ed. Reti, 3:97, n. 34. Leonardo's commitment to the mathematical nature of line has even been singled out as a fundamental difference between his theoretical views of pictorial art and those of Alberti: Marinoni, “L'Essere del Nulla.”

51. Copare Alberti, On Painting 1.2: “The first thing to know is that a point is a sign (signum) which one might say is not divisible into parts. I call a sign anything which exists on a surface so that it is visible to the eye.” Alberti, Book 1: 48, also defined line “according to the philosophers” using the terminology of optical theory (derived from Galen, De usu partium 10.12; see Edgerton, “Alberti's Color Theory,” 125). Compare Leonardo, Parte Prima, Chapter 37.

52. Libro A Carta 26.38: “Molto più rilievo dimostreranno le cose nel campo chiaro e alluminato che nell'oscuro. / La ragione di quel che si propone è che se tu vuoi dare rilievo alla tua figura tu le fai che quella parte del corpo ch'è più remota dal lume, manco partecipa d'esso lume, onde viene a rimanere più oscura; e terminando poi in campo oscuro viene a cadere in confusi termini, per la qual cosa, se non vi accade riflesso, l'opera resta senza grazia e da lontano non apparisce se non le parti luminose, onde conviene che l'oscure paino essere del campo medesimo onde le cose paiono tagliate, e rimanere tanto men che 'l suo dovere quant'è l'oscuro.” (Emphasis added.) Transcription from Pedretti, 1964, 49 (see his n. 45 for discussion of the translation), translating the last problematic phrase as follows: “It is correct, therefore, that the other parts be of a darkness similar to the background from which the figure has to appear detached, but keep them less dark than they should be on that background.” The principle of heightened contrast through direct juxtaposition recurs frequently in Libro A on Cartas 25.35, 27.42, 30.53, 30.54, 35.77, and 43.93. On the relationship of these statements to the “trattato sequences,” see Appendix 1. An early statement on the subject of heightened contrast occurs in Ms. A, fol. 84 recto, ca. 1492.

53. On Leonardo's contribution to Raphael's style, see Weil-Garris Posner, Leonardo and Central Italian Art. For an introduction to recent issues concerning the late sixteenth-century reform of painting, see Dempsey, Annibale Carracci. Barocci may have known Leonardo's Codex Urbinas from the Castle Durante near Urbino where he was trained: See Gary Walters, “Federico Barocci,” 43-44 ff. I thank Angela Lafferty for this reference.

54. F recent surveys of Leonardo's precedents, see Elkins, “Piero della Francesca and the Renaissance Proof of Linear Perspective”; and Klaus Bergoldt, ed., Der dritte Kommentar Lorenzo Ghibertis: Naturwissenschaften und Medizin in her Kusttheorie der Frührenaissance (Weinheim, c. 1988), which became available only after the present study was completed.

55. Cenini, Il libro dell'arte, Chapter 1. Most recently on the scientific investigations of fifteenth-century artists, which are relatively well studied, see M. Kemp, The Science of Art, Chapter 1.

56. Fo the personalities, see Mendelsohn, Paragoni. Benedetto Varchi and Vincenzo Borghini set traditional issues of the debate between painters and sculptors firmly into the Aristotelian context of the productive sciences. See Chapter Four.

57. For the history of writings on shadow projection, see most recently Bauer, “Experimental Shadow Casting”; see also Kaufmann, “The Perspective of Shadows”; M. Kemp, “Geometrical Perspective from Brunelleschi to Désargues,” who assembles massive evidence to argue, p. 89, that “the development of perspective science away from the concerns of artists had become pronounced by 1600.” Field, “Giovanni Battista Benedetti” characterizes Benedetti as a “natural philosopher employing the methods of geometry” (p. 95). This turn of events nonetheless suggests, in fact corroborates the likelihood, that Leonardo's writings on chiaro e scuro and other treatises in the same tradition filled a much-needed gap in the education of artists during the same period. Brown University, Children of Mercury, includes an overview of teaching practices in early art academies, with a thorough bibliography.

58. While the Parte Prima of Leonardo's Trattato evidently did not circulate, four of the passages could have been known from Ms. A, and perhaps others from Madrid Codex II, Libro A (now lost), and other notes, such as several sheets compiled into the Codex Atlanticus and the lost Codex Sforza mentioned by Gian Paolo Lomazzo: see Steinitz, Leonardo da Vinci's Trattato della Pittura, 21-26; Pedretti, Libro A, 121-128, and Commentary 1: 76-86. On the history of the Codex Urbinas, see Steinitz, 39-44, and Pedretti, Commentary 1: 12-14. An overview of the evidence, with references, is found in Chapter One. Most of these activities centered around Cassiano dal Pozzo in Rome; see Haskell, Patrons and Painters, 97-114. On Cassiano's use of Leonardo's manuscripts, see most recently Bell, “Cassiano dal Pozzo's Copy of the Zaccolini Manuscripts.” On Poussin's study of Leonardo's writings, see Elizabeth Cropper, “Poussin and Leonardo.”

59. Bell, “Cassiano dal Pozzo's Copy of the Zaccolini Manuscripts,” draws the same conclusion from her evidence. The forthcoming study of Federico Borromeo by Pamela Jones will clarify the relationship of Counter-Reformation art theory to earlier writings and later artistic practices.

Abbreviations

AB: The Art Bulletin

DHI: Dictionary of the History of Ideas. Ed. Philip Wiener. 5 vols. New York, 1973.

JHI: Journal of the History of Ideas

JWCI: Journal of the Warburg and Courtauld Institutes

Scritti: Scritti d'arte del cinquecento. Ed. Paola Barocchi. 9 vols. Turin, 1977.

Trattati: Trattati d'arte del cinquecento. Ed. Paola Barocchi. 3 vols. Bari, 1961.

References to Leonardo's manuscripts follow standard forms of abbreviation:

CA = Codex Atlanticus (I follow the folio numbers of Il Codex Atlantico di Leonardo da Vinci nella Biblioteca Ambrosiana di Milano, Reale Accademia dei Lincei, transcribed by G. Piumati, 35 vols., Milan, 1894-1904, in order to permit concordance with the existing anthologies and literature; the current critical edition edited by A. Marinoni has been consulted).

R.1 = Richter, J.P., The Literary Works of Leonardo da Vinci, 2 vols., 2d rev. ed., London, 1939, passage number 1.

All references to the last 24 folios of Ms. A, which have been bound separately as Ashburnham II, are cited as Ms. A, following the recent edition by A. Corbeau and N. De Toni.…

CN = Commentary Notes numbered 1 through 46, accompanying the passages in the Parte Prima of the Codex Urbinas.

Chapters One through Four = introductory chapters of the present study.

Chapters 1 through 46 = numbered passages of the Parte Prima.