H I S T O R I C A L P E R S P E C T I V E

Art, Optics and History: New Light on the Hockney Thesis

Michael John Gorman

David Hockney’s recent book Secret Knowl-edge: Rediscovering the Lost Techniques of the Old Masters arguesthat painters from Van Eyck to Ingres achieved a remarkableimitation of nature not through sheer painterly talent but byemploying optical devices. Specifically, Hockney argues froma bewildering array of visual evidence that, from around 1430,many painters employed a concave mirror to project brightlylit subjects onto a canvas, thus allowing them to render figureswith an unprecedented naturalism. Hockney’s claim derivesfrom experiments carried out in his home with a shaving-mirror and the assistance of optical scientist Charles Falco ofthe University of Arizona. Together, Hockney and Falco sug-gest that, in addition to the “natural” depiction of their sub-jects, a number of anomalous features of 15th- and 16th-centurypaintings can easily be explained by assuming that the artistsemployed the “mirror-lens,” Hockney’s slightly confusing termfor the concave mirror [1]. According to Hockney, the pat-terned tablecloth in a 1543 painting by Lorenzo Lotto exhibits“optical artefacts” associated with refocusing, for example,multiple vanishing points and loss of focus on the pattern. Re-garding the famous convex mirror depicted in Van Eyck’sArnolfini Wedding (1434), Hockney remarks, “If you were to re-verse the silvering and then turn it round, this would be allthe optical equipment you would need for the meticulous andnatural looking detail in the picture” [2]. Hockney, however,

does not consistently describe theexact technique by which artistsmoved from the projected image tothe finished painting. Sometimes,he suggests that artists traced theprojected image. On other occa-sions, he states that they merelyused the projection to mark a fewkey points.

According to Hockney’s chronol-ogy, around the end of the 16th cen-tury, painters began to use refractivelenses instead of concave mirrors toproject their tracing images. Unlikeconcave mirrors, convex refractive lenses have the property ofreversing left and right, in addition to inverting the image.Hockney points to a sudden increase in the number of left-handed drinkers in paintings executed after the last decade ofthe 1590s as conspicuous evidence of the shift from mirror tolens, citing paintings by Caravaggio as marking the point oftransition from reflection to refraction [3].

Visual evidence is essential to Hockney’s argument. Playfully,his book is prefaced by a forged document purporting to be aquotation from art historian Roberto Longhi: “Paintings areprimary documents. Archival documents can be faked; criticaljudgements, not.” Perhaps unsurprisingly, not all of Hockney’srevelations have been embraced enthusiastically by the schol-arly community. At the colloquium organized by LawrenceWeschler at the New York Humanities Center in December2001, a number of vociferous critics attacked Hockney’s thesis.

© 2003 ISAST LEONARDO, Vol. 36, No. 4, pp. 295–301, 2003 295

A B S T R A C T

David Hockney’s recent bookSecret Knowledge: Rediscover-ing the Lost Techniques of theOld Masters argues that 15th-century painters employedoptical devices to achieverealistic portraiture. A reexami-nation of the history of opticalprojection techniques raisesproblems for Hockney’s provoc-ative hypothesis.

Michael John Gorman (lecturer), Program in Science, Technology and Society, StanfordUniversity, Stanford, CA 94305-2120, U.S.A. E-mail: !mgorman@stanford.edu". Web:!http://www.stanford.edu/~mgorman/".

Object

Image

Fig. 1. Optical diagram of Girolamo Cardano’s 1550 camera obscura (i.e. Hockney’s “late” device). (© Michael John Gorman) Placing abiconvex lens in the aperture produces a reduced, inverted, but sharp image.

Leonardo_36-4_255_344 7/17/03 12:02 PM Page 295

Art historians noted that Hockney did notgive adequate credit to the significant cor-pus of previous scholarly work on artists’use of optical devices including the cam-era obscura. Perhaps the most scathingcriticisms, however, came not from thehistorians but from two optical scientistsinvited to the colloquium, David Stork ofStanford University and Christopher W.Tyler of the Smith-Kettlewell Eye Re-search Institute [4].

The criticisms leveled by Stork andTyler concern the technical details of theoptical techniques that Hockney andFalco suggest were employed by artistsfrom the time of Robert Campin (c. 1375–1444). I do not intend to reviewall their criticisms here. However, one ofStork’s arguments appears potentiallyfatal to Hockney’s hypothesis. Stork notesthat in order to produce an image of ap-propriate scale for a Renaissance paint-ing, a mirror of long focal length wouldbe required. Hockney’s shaving-mirrorhad, according to Stork, no optical equiv-alent in the Renaissance. The mirrors (al-most all convex, rather than concave)depicted in paintings of the 15th and16th centuries can all be seen to havefocal lengths significantly shorter thanthe approximately 55 cm estimated byHockney and Falco to be necessary forprojection of an image of the dimensionsof the final painting. Van Eyck’s Arnolfinimirror is no exception. If we assume thatthe artists actually traced directly fromthe optical images, then we encounter aserious difficulty here. Moreover, as Storkpoints out, a spherical concave mirror of55 cm focal length, if made of glass,would have to be cut from an enormoussphere almost 7 feet in diameter! It wasabsolutely impossible, Stork claims, for aRenaissance glassblower to create asphere of such huge dimensions [5].

Considering the relevance of the his-tory of optical instruments to this debate,it is surprising that historians of sciencehave not had much to say about theHockney thesis thus far. In spite of theprotests of art historians, I do not believethat Hockney’s point is unoriginal. True,many art historians have emphasized thegeometrical investigations of artists andthe instruments that they have used. Thework of Hockney’s correspondent dur-ing the writing of his book, Martin Kemp,is exemplary in this respect [6]. Some his-torians have even pointed to the impor-tance of the camera obscura, particularlyfor Vermeer and other 17th-centuryDutch artists [7]. However, the specificdevice considered by Hockney, at leastfor the earlier period—i.e. the concavemirror used as a projective device—is not

one that has been investigated as a vitaltool for artists by anyone prior to him, tomy knowledge. So, either Hockney iscompletely wrong about the use of con-cave mirrors for projection or he has hiton a device that has been almost com-pletely ignored by historians.

As an artist’s tool, the concave mirrorhas some shortcomings. Assuming forthe moment that one can find a concavemirror of sufficient focal length and op-tical quality, and that there is sufficientillumination on the subject—all assump-tions that have been challenged at onepoint or another by Hockney’s critics—the image projected by the mirror in thedarkened chamber of the artist is still in-verted, undeniably not the ideal situationfor a painter. A further objection toHockney’s hypothesis is the absence of asignificant written record about artistsemploying concave mirrors, an absencethat Hockney ascribes to fear of the In-quisition, adding the spice of a conspir-acy of silence to the story. “Heresy keptthe lenses secret,” Hockney suggests mys-teriously [8].

In this article I intend to shed light onthose parts of the debate relating to thehistory of science and technology. DidHockney’s proposed projective systemexist? If so, when was it created? Whomight have used it? These are the ques-tions that I address here. Surprisingly, thedocuments, when read carefully, do havea great deal to say about the use of con-cave mirrors to project images. By look-ing closely at the documents concerningthe history of the camera obscura, we canmake a fairly precise estimate of the dateof creation of a camera obscura employ-ing a concave mirror. We need not resortto any conspiracy of silence and we canaccess a vivid picture of the technical lim-itations governing image-projection tech-niques in the crucial period of the late16th century, the time at which Hockneydates the transition from reflective to re-fractive projection techniques.

The evidence that I review here relatesto the crucial issue of what type of opti-cal projection was technically achievableat a particular historical moment. It is im-portant to emphasize, however, that thefact that a type of projection was techni-cally feasible does not imply that any par-ticular artist used the technique, in theabsence of further evidence, visual ordocumentary. This article will not enterthe dispute on internal visual evidence inpaintings that is currently raging betweenthe Hockney camp and Stork and Tyler,but the evidence provided should, Ihope, affect the historical and technicalparameters of that debate. Ironically, it

will become clear that concave mirrorsbegan to be used as projective devices inthe mid-16th century and were used incombination with convex lenses fromthat century’s end, precisely the momentthat Hockney associates with the transi-tion from mirrors to lenses.

However, nobody in this debate, to my knowledge, has so far suggested thatmirrors were used both alone and in con-junction with convex lenses. The combi-nation of concave mirror and convexlens produced a magnified, uprightimage in which left and right were not re-versed.

Moreover, the addition of a convexlens had the effect of increasing the ef-fective focal length of the concave mir-ror. Thus, given a suitable convex lens,even a mirror of small focal length couldaid in producing a life-size or magnifiederect image.

A CONDENSED HISTORY OFTHE CAMERA OBSCURAIn the 4th century B.C.E., Aristotle no-ticed that during a solar eclipse, the cir-cles of light on the ground under a planetree gradually turned to thin crescents,thus recognizing the fundamental phe-nomenon of image projection throughsmall apertures on which the camera ob-scura is based. The 10th-century Arabscholar Ibn Al-Haytham described howthe image of the sun could be projectedthrough a hole in the wall of a room ontoa wall opposite during an eclipse, and foralmost 600 years, until the mid-16th cen-tury, the camera obscura remained justthis: a small hole in the wall, describedby a great number of scholars, includingFrancesco Maurolico, Gemma Frisius,Leonardo da Vinci and Erasmus Rein-hold [9].

Then, in 1550, the Milanese astrologerand physician Girolamo Cardano sug-gested a dramatic improvement to thecamera obscura in his work On Subtlety:inserting a convex lens made of glass intothe hole. This produced a much clearerimage (Fig. 1). Here are Cardano’s words:

If you want to see the things that go onin the street, at a time when the sunshines brightly place in the window shut-ter a bi-convex lens [orbem e vitro]. If youthen close the window you will see im-ages projected through the aperture onto the opposite wall, but with rather dullcolors; but by placing a piece of verywhite paper in the place where you seethe images, you will attain the eagerlyawaited result in a wonderful manner[10].

Cardano’s discovery, which does notseem to have gone into common use, was

296 Gorman, Art, Optics and History

Leonardo_36-4_255_344 7/17/03 12:02 PM Page 296

later echoed in 1585 by GiambattistaBenedetti, philosopher to Carlo Emanuele, Duke of Savoy, in his DiverseSpeculations on Mathematics and Physics.Benedetti, in a letter to Pirro de Arzoniabout the properties of optical images,writes:

On this subject, I don’t wish to remainsilent on a certain marvellous effect of athing of this kind. This is that a roundhole is made, of the size of a lens, andthe hole is stopped up with one of thoselenses that are made for old people (notfor short-sighted people), I mean oneboth of the sides of which are convex, notconcave. Then place a sheet of whitepaper at a sufficient distance from thehole so that the objects outside appearon it. If these objects are illuminated bythe sun, they will be seen so clearly anddistinctly that nothing more beautiful ordelectable can be imagined, but they willbe inverted. But if you wish to see themupright, this we do very well by means ofreflection in a plane mirror [11].

So, we have two interesting new addi-tions to the camera obscura: the lens byCardano in 1550 and an oblique planemirror by Benedetti in 1585. Cardanoand Benedetti were both among the mostadvanced mathematical practitioners oftheir time, and it is impossible to believethat they would be unaware of a projec-tion technique being used widely byartists. Another version of the camera ob-scura was described for the first time byGiambattista Della Porta in 1558. DellaPorta is mentioned several times inHockney’s book, and an excerpt from hisNatural Magic figures prominently in the“Textual Evidence” section [12]. How-ever, there is a problem with this excerptas published by Hockney: It is completelyincomprehensible. The excerpt is takenfrom an anonymous 17th-century trans-lation of Della Porta’s work that wasclearly carried out by somebody who hadat best a mediocre command of Latin andfailed to understand the technical details,confusing lenses with mirrors to an extentthat the text becomes gibberish.

What did Della Porta really say about thecamera obscura? And, given that Hockneylocates his second crucial transition withCaravaggio at the end of the 16th century,could Della Porta’s camera obscura haveimplications for this transition?

Della Porta was as famous as a play-wright and impresario as for his investi-gations of secret natural processes andamusing mechanical contrivances. Wewill see that his theatrical productionsand camera obscura are intimately con-nected. Indeed, the sophisticated cam-era obscura described in the enlarged1589 version of Natural Magic was de-signed specifically to project theatricalperformances to allow them to be viewedby an audience of nobles sitting insideDella Porta’s dark chamber.

Natural Magic was first published in1558, when Della Porta was 23 years old,in four books. It included a brief descrip-tion of a basic “hole-in-the-wall” cameraobscura—ignoring Cardano’s improve-ment of 8 years previous. However, DellaPorta went on to describe how to use aconcave mirror to project an image ontoa piece of paper—without using a lens—prefacing his account with the very por-tentous words: “Now I will describe thatabout which I have remained silent untilnow, and thought to keep secret: how youcan see all things with their colors if youdesire.” Della Porta continues:

Opposite, place a mirror, not the kindthat disperses the light by dissipating, butthe kind that unites by collecting to-gether. Move it closer and further awayfrom the hole until you see it reach theperfect and true quantity, approachingthe necessary distance from the center.If you look attentively, you will see theface, gestures, movements and clothes ofmen, the blue sky with dispersed clouds,the very distant mountains, and, in asmall circle on a piece of paper, whichyou will attach above the hole, you willsee almost an epitome of the world.These things, when you see them all in-verted, will make you marvel to no smalldegree [13].

This is precisely the device that DavidHockney claims was used by artists be-ginning in the 1430s—his “early” device(Fig. 2). Della Porta, in the 1558 editionof his work, goes on to explain “howsomeone who doesn’t know how to paintcan paint the likeness of a man, or an-other thing, if he only knows how tomatch the colors,” by using this device[14]. The second edition of NaturalMagic, published over 30 years later in1589, contained a vastly expanded dis-cussion of the camera obscura. After hisdescription of the old hole-in-the-wall de-vice, Della Porta added a new line: “If youplace at the hole a crystal lens, you willimmediately see things much moreclearly, the faces of those walking in thestreet, the colours of their clothes, theirclothes and all things, just as if you sawthem close up, not without enormouspleasure, so that those who see this can-not marvel enough” [15]. This was sim-ply the convex lens suggested by Cardanoand Benedetti.

In the second edition, however, DellaPorta also added a dramatic revision tohis concave mirror camera obscura.First, he had suggested, following Car-dano and Benedetti, the use of a lens inthe aperture. Now, to turn the imagesupright, he combined the lens with theconcave mirror projection system that hehad proposed in 1558. The extraordi-nary result was a device that projectedmagnified, upright images. In modernterminology, the inverted image formedby the convex lens, falling at a short dis-tance in front of the focal point of theconcave mirror, served as an object forthe mirror, which turned it upright andmagnified it (Fig. 3).

The crucial section begins:

But if you wish for the images to appearupright, this will be a great feat, at-tempted by many but not discovered byanyone until now. Some [perhaps DellaPorta refers here to Benedetti] place flatmirrors at an oblique angle near thehole, which reflect an image onto thescreen opposite that is more or less up-right but dark and confused. We, by plac-ing the white screen at an oblique angleto the hole, and looking towards the partfacing the hole, saw the images almosterect but the pyramid [of light] cutobliquely showed the men without anyproportion and confusedly. But in thefollowing way you will have what you desire. Place an eyeglass made from a bi-convex lens [specillum e convexis fabrica-tum] in front of the hole. From here theimage falls on the concave mirror. Placethe concave mirror far from the center[i.e. from the focal point of the lens] sothat the images which it receives invertedit will show upright, because of the dis-tance from the center. In this way, abovethe hole on the white paper you will see

Gorman, Art, Optics and History 297

Object

Image

f

Concave mirror

Fig. 2. Optical diagram of Della Porta’s 1558 system (i.e. Hockney’s “early” device). (© Michael John Gorman) Note that the concave mirror has a long focal length (i.e. it isquite “flat”), but still produces an image reduced in scale.

Leonardo_36-4_255_344 7/17/03 12:02 PM Page 297

the images of the things that are outsideso clearly and openly that you will never cease to be delighted and amazed.But here I should warn you, so that youdo not waste your efforts, that it is nec-essary for the lens to be proportioned to the concave mirror, but, as you will see, here we will speak of this many times [16].

Despite this promise, Della Portanever spoke of it again, and never gaveaway the proportion required betweenthe focal length of the lens and that ofthe concave mirror. In this way, even aconcave mirror of 20 cm in focal lengthcould be used to project life-sized im-ages, given an appropriate convex lens.Such a mirror, and the appropriate lens,unlike the long focal-length concavemirrors required by Hockney’s thesis,were within the manufacturing capabil-ities of the late 16th century, as describedby Della Porta himself in vivid detail inNatural Magic. This device is never de-scribed in Hockney’s book (apart fromin the garbled translation) and, as far asI know, has not been considered in anyof the ensuing discussion.

Clearly, between writing the first andsecond editions of Natural Magic, Gi-ambattista Della Porta made a startlingdiscovery: combining a convex lens witha concave mirror, placed at the correctdistance with suitably proportioned focallengths, could allow very clear, large, up-right images to be projected on a screenabove the aperture. What happened tothe Neapolitan magician between 1558and 1589 to inspire this new invention?For one thing, he traveled extensively,collecting secrets for publication in hisNatural Magic and directing perfor-mances of his plays for private audiencesthroughout Italy. One such journey in1580 took Della Porta to Venice. The fol-

lowing is the description taken fromLouise Clubb’s account of Della Porta’slife:

Settled in Venice for the time being,Della Porta began work on a parabolicmirror and an “occhiale.” The latter wasprobably a large magnifying or burningglass, or perhaps an experiment in eye-glasses. Campori believed that it was atelescope, hoping to connect the projectat Venice with the widespread belief thatDella Porta was the original inventor ofthe telescope. Cardinal d’Este returnedalone to Rome, but Della Porta kept himinformed of all progress. He was de-lighted to have at his disposal the skilledglassworkers in the environs of Venice,and reported on November 29, 1580 thatwith the help of Giacomo Contarini hehad found an artisan of Murano capableof constructing the delicate mirror. Butif the Venetian craftsmen pleased him,the aria grossa of the lagoons did not,and he blamed it for the new attack offever which now forced him to bed.When the Cardinal returned to Ferrara,he sent for Della Porta, but the eagerlens-designer was loath to leave his Mu-rano project. He used bad weather andhis own illness to excuse the delay. Fi-nally, with rather bad grace, he agreed toleave Venice in December 1580, but hislove of secrecy was outraged by having toreveal his plans to Contarini and to leavethe lens-grinding under his supervision[17].

Although Clubb does not mention thispossibility, there can be absolutely nodoubt that the “parabolic mirror and oc-chiale” that Della Porta worked on so intensely in Venice were the vital com-ponents of the new camera obscura pre-sented in the 1589 edition of NaturalMagic. This edition also included a de-tailed account of the way lenses and mir-rors were made in Venice, anotherby-product of Della Porta’s time on theisland of Murano in 1580.

Why did Della Porta invest so muchtime and money in building his new camera obscura? The principal reasonseems to have been for his extravaganttheatrical productions. The 1589 editionof Natural Magic gives a vivid account ofusing the new device to “see in a darkroom a hunt, a battle, and other won-ders”:

Now, to reach the end of this material, Iwill add a secret that is surely the mostingenious and beautiful for pleasinggreat lords. In a dark room, on whitesheets, you can see hunts, banquets, bat-tles of enemies, games, and finally, everything you like, so clear and lumi-nously, and minutely, as if you had themright before your eyes. Let there be a spa-cious area outside the room where youare going to make these appearances,which can be well illuminated by the sun.In this, you will place trees, houses,woods, mountains, rivers, real beasts oranimals made with skill from wood orother materials, which have children in-side them who move, as we frequentlyuse in the intermissions of comedies,deer, wild boar, rhinoceroses, elephants,lions and other animals that please you.Each of these emerges one by one fromits lair, and comes into the scene, thenthe hunters come with spears, nets andother necessary instruments, and areseen to hunt the animals, playing horns,trumpets and conches, so that those in-side the room see the trees, the animals,and the faces of the hunters, and theother things, so naturally that they can-not tell whether they are real or due totrickery.

Unsheathed swords shining in the sunwill give a great splendor inside, whichwill frighten the audience. Many times Ihave given such spectacles for my friends,who admired them with great wonderand astonishment. Even though I gavethem the explanations of Philosophy andPerspective they did not want to believethat these were natural things, until,opening the door, I showed them thetrick [18].

Della Porta’s description of theatricalstage settings and “unsheathed swords”frightening the audience makes onethink immediately of the dramatic stage-management in the paintings of Car-avaggio in the Contarelli chapel in Rome.If Caravaggio had experimented withDella Porta’s 1589 device, rather than his1558 device (Hockney’s “early” device)or Cardano’s 1550 device (Hockney’s“late” device) he would have benefitedfrom an upright image much larger inscale than the image produced by thesame mirror without a convex lens.

So, to recap, until 1550, the camera ob-scura was just a hole in the wall. ThenCardano and later Benedetti fitted a con-vex lens (a “lens for old people”) into thehole. Benedetti added a further obliquemirror to produce a horizontal image.

298 Gorman, Art, Optics and History

Object

LensConcave mirror

1

2

Fig. 3. Optical diagram of Della Porta’s 1589 system. (© Michael John Gorman) The object(on the far left) at 4 m from the lens; the focal length of the lens is 84 cm. The focal lengthof the concave mirror is 12 cm. A life-size, erect image of the object is produced at circa 1 mfrom mirror (this is the image numbered “2” in the diagram). The focal point of the lens is ashort distance outside the focal point of the mirror, as Della Porta advises. For practicalpurposes, it is necessary to tilt the concave mirror slightly so that the image (2) is formed ona screen above the aperture containing the lens.

Leonardo_36-4_255_344 7/17/03 12:02 PM Page 298

Della Porta, in 1558, suggested, for thefirst time, using the “early” Hockney tech-nique: a concave mirror placed in frontof the hole in the wall, without a lens.After his visit to Venice in 1580, DellaPorta returned to Naples with a new in-strument: a long focal length bi-convexlens and a good concave mirror, com-bined to produce extraordinary quasi-cinematic experiences for his friends andnoble patrons. He published the de-scription clearly in the 1589 edition ofNatural Magic, with an important warn-ing: that the device would not work un-less the curvatures of lens and mirrorwere in suitable proportions. His new de-vice projected magnified, bright, uprightimages. This device was not describedelsewhere beforehand, and we have noreason to believe that it existed beforeDella Porta made it, with great effort, inVenice in 1580.

Could artists have used Della Porta’slatter device, which has not even figuredin the debate on Hockney’s book so far?It certainly would have solved some of theproblems of Hockney’s devices as toolsfor artists: inverted images, the necessityfor implausibly long focal lengths and soon. Caravaggio would seem to be an ob-vious candidate, as he arrived in Rome,as has been noted before, shortly afterthe publication of the second edition ofNatural Magic. Della Porta’s second de-vice, however, would not produce the left-handed drinkers that Hockney hasnoticed in some of Caravaggio’s paint-ings. Caravaggio’s patron Cardinal delMonte and his brother the mathemati-cian Guidobaldo del Monte would cer-tainly have known Della Porta’s work, ifnot the man himself, and could probablyhave provided Caravaggio with a suitablemirror and lens, although there is no evi-dence that they did so. Guidobaldo waseven a correspondent of Contarini, whopersonally supervised the final work onDella Porta’s new camera obscura inVenice.

As he wrote to Galileo in 1611, Cardi-nal del Monte had contacts in Rome whowere skilled at grinding rock-glass lenses,still used for more expensive pairs ofspectacles in the 17th century. One doesnot need to invoke a conspiracy of silenceto suggest that Caravaggio may have beenfamiliar with the new instrument: the de-scription was published, and the devicehad in all likelihood been demonstratedtheatrically by Della Porta himself in thefashionable salons of Rome and Naplesin the manner that he describes in the1589 Natural Magic. Incidentally, one per-son who was deeply interested in DellaPorta’s description of his camera obscura

was the astronomer Johannes Kepler,who used it as the basis of his account ofimage formation in the eye and used hisown rotating camera obscura to sketchlandscapes [19].

It is likely that an artist like Caravaggiowould have taken a great interest in op-tical experimentation with the cameraobscura, given his interest in dramaticchiaroscuro effects and the “theatrical-ity” of scenes such as The Martyrdom of St.Matthew (Fig. 4). The frescoes in the Con-tarelli chapel, in particular, represent astriking use of lighting effects that hasbeen the subject of much commentary.Suggesting that Caravaggio may have ex-perimented with the images produced bythe new camera obscura does not neces-sarily imply that he used it to produce“tracing images,” a hypothesis that is veryproblematic. In any case, any painting is,as Michael Baxandall puts it, “the depositof a social relationship,” bearing thetraces of a variety of constraints that in-clude available technical equipment, thenature of the commission, the space forwhich the work is destined, painting ma-terials and the artist’s training, so to posita simple relationship between the imageproduced by Della Porta’s later cameraobscura and Caravaggio’s works would be

naïve, to say the least [20]. Additionally,the use of even Della Porta’s improvedcamera obscura would imply the carefulmanipulation of natural lighting. Artifi-cial lighting, as David Stork has argued,would not provide sufficient illuminationto create a usable image [21]. In the caseof the Contarelli chapel frescoes, the useof natural light is plausible. The fact thatthe window depicted in The Calling of St.Matthew is not a source of light suggeststhat the scene was staged outdoors (Fig. 5). A large flat mirror found in Car-avaggio’s studio may have been used toreflect sunlight onto his figures.

Where does all this leave the earlierpaintings discussed in Hockney’s book?The device that he claims was used byartists including Robert Campin and Janvan Eyck in the 1430s was, in all likeli-hood, invented by Giambattista DellaPorta just before 1558 and then renderedobsolete by Della Porta himself in 1580.At the very best, it had perhaps a 35-yearworking life as an artist’s instrument, as-suming that it was not just an amusing toyfor those unable to draw, as Della Portahimself suggests. If Caravaggio used acamera obscura, he would have had everyopportunity to avail himself of the latesttechnology—Della Porta’s combination

Gorman, Art, Optics and History 299

Fig. 4. Michelangelo Merisi da Caravaggio, The Martyrdom of St. Matthew, oil on canvas, 127 #135 in, 1599–1600. Contarelli Chapel, San Luigi dei Francesi, Rome. (Photo © Scala/ArtResource, New York)

Leonardo_36-4_255_344 7/17/03 12:02 PM Page 299

convex lens–concave mirror instrument.If 15th-century painters ever used thecamera obscura, however, they used a sim-ple hole in the wall: no mirror, no lens.

We can nonetheless be grateful toDavid Hockney and Charles Falco forbringing to our attention a forgotten de-vice that had a short-lived vogue in the16th century: the camera obscura em-ploying a concave mirror described inthe 1558 edition of Giambattista DellaPorta’s Natural Magic.

POSTSCRIPT: THE ROMANCE OFTHE ROSEAfter the first draft of this article wassubmitted to Leonardo, Charles Falcomade the suggestion that The Romance ofthe Rose, composed by Guillaume de Lor-ris and Jean de Meun in the 13th century,contained a reference to a concave-mirror optical projection system. Falcocites the following passage from the alle-gorical poem as evidence that the con-cave-mirror camera obscura was knownin the 13th century:

Those who are masters of mirrors makeone image give birth to several: if theyhave the right form ready, they createfour eyes in one head, and they make

phantoms appear to those who lookwithin. They even make them appear,quite alive, outside the mirror, either inwater or the air [22].

While the statement that “masters ofmirrors” make images “appear quitealive, outside the mirror” could be a ref-erence to the properties of concave mir-rors, the phenomenon described couldsimply be the three-dimensional realimage seen when one places an object(e.g. a finger) close to twice the focal dis-tance away from the mirror, a phenome-non that was known in the Middle Agesand is familiar to us through optical-illusion toys available from science stores.This phenomenon does not require asurface or screen, and de Meun’s textmakes no reference to the kind of set-updescribed by Hockney and Falco. Ibn al-Haytham, whose extensive work on op-tics was becoming widely available duringthis period, in Latin translation as De As-pectibus, is a likely source of much of deMeun’s description of the properties ofmirrors. According to A.I. Sabra, an ac-knowledged expert on Ibn al-Haythamand translator of his optical works, “Ibnal-Haytham does of course treat of ‘im-ages’ perceived by reflection from spher-ical, cylindrical and conical concave

mirrors. However, I am not aware of anydescription, in any of his writings, of aprojective system involving an aperture,a concave mirror and a screen” [23]. Toconclude, The Romance of the Rose does notprovide support for the existence of theconcave-mirror camera obscura prior to1558.

Acknowledgments

A short summary of the present article appeared as“Projecting an Image: Did the Old Masters PaintUsing Optical Projection Techniques?” Nature 417(20 June 2002) p. 794. The research for this projectwas sponsored by the Stanford Humanities Labora-tory. My students Mo Khalil, Steve Compton andAdam Vollmer helped build a reconstruction of DellaPorta’s 1589 camera obscura. I am very grateful toMartin Kemp for suggestions on a previous draft, to A.I. Sabra for advice about Ibn al-Haytham, and toCharles Falco, David Stork, David Graves, Amy Ioneand Christopher Tyler for stimulating exchanges onoptics, art and the nature of historical evidence.

References and Notes

1. David Hockney, Secret Knowledge: Rediscovering theLost Techniques of the Old Masters (London: PenguinPutnam, 2001); David Hockney and Charles Falco,“Optical Insights into Renaissance Art,” in Optics andPhotonics News (July 2000). For further discussions ofHockney’s thesis, see !http://www.artandoptics.com".

2. Hockney [1] p. 82.

3. Hockney [1] p. 118.

4. David Stork, “Response to Hockney and Falco,”and Christopher W. Tyler, “What Perspective Con-structions Reveals about Optical Aids,” articlesposted on web site !http://www.artandoptics.com".

5. Stork [4].

6. Martin Kemp, The Science of Art: Optical Themes inWestern Art from Brunelleschi to Seurat (New Haven, CT:Yale Univ. Press, 1990).

7. See, for example, Svetlana Alpers, The Art of De-scribing: Dutch Art in the Seventeenth Century (Chicago:University of Chicago Press, 1984); Philip Steadman,Vermeer’s Camera: Uncovering the Truth behind the Mas-terpieces (Oxford: Oxford Univ. Press, 2001); but seealso Walter Liedtke’s critical review in Burlington Mag-azine CXLIII, No. 1183, 642–643 (October 2001);and Walter Liedtke, A View of Delft: Vermeer and HisContemporaries (Zwolle, the Netherlands: WaandersPublishers, 2000).

8. Hockney [1] p. 246.

9. Helmut Gernsheim, The History of Photography fromthe Camera Obscura to the Beginning of the Modern Era(New York: McGraw-Hill, 1969) pp. 17–42; M.S.Hammond, The Camera Obscura: A Chapter in the Pre-History of Photography (Ph.D. dissertation), Ohio StateUniversity, 1986 (Ann Arbor, MI: UMI facsimile,1990).

10. Girolamo Cardano, De subtilitate (Paris: Ex offic-ina Michaelis Fezandas, 1550), book iv, p. 107, mytranslation.

11. Giovanni Battista Benedetti, Diversarum specula-tionum mathematicarum & physicarum liber (Turin,Italy: apud haeredem Nicolai Bevilaquae, 1585) pp. 270–271, my translation.

12. Hockney [1] pp. 208–209.

13. Giambattista Della Porta, Magiae naturalis, sive demiraculis rerum naturalium Libri IIII (Naples, Italy:Apud Matthiam Cancer, 1558) book iv, my transla-tion.

300 Gorman, Art, Optics and History

Fig. 5. Michelangelo Merisi da Caravaggio, The Calling of St. Matthew, oil on canvas, 127 # 134 in, 1599–1600. Contarelli Chapel, San Luigi dei Francesi, Rome. (Photo © Scala/Art Resource, New York)

Leonardo_36-4_255_344 7/17/03 12:02 PM Page 300

14. Della Porta [13].

15. Giambattista Della Porta, Magiae naturalis libriXX. Ab ipso authore expurgati, & superaucti, 2nd Ed.(Naples, Italy: Apud H. Saluianum, 1589), bookXVII, chapter VI, my translation.

16. Della Porta [13], book iv.

17. Louise Clubb, Giambattista della Porta, Dramatist(Princeton, NJ: Princeton Univ. Press, 1965) pp. 21–22.

18. Della Porta [13].

19. Johannes Kepler, Optics: Paralipomena to Witeloand Optical Part of Astronomy, William H. Donahue,trans. (Santa Fe, NM: Green Lion Press, 2000) pp.224–226; Alpers [7].

20. Michael Baxandall, Painting and Experience in Fif-teenth Century Italy: A Primer in the Social History of Pic-torial Style, 2nd Ed. (Oxford, U.K.: Oxford Univ. Press,1988).

21. Stork [4].

22. Guillaume de Lorris and Jean de Meun, The Ro-mance of the Rose, Charles Dahlberg, trans., 3rd Ed.(Princeton, NJ: Princeton Univ. Press, 1995); line18163; quoted by Charles Falco at !http://www.optics.arizona.edu/ssd/FAQ.html".

23. A.I. Sabra, e-mail message to the author, 30 June2002. See The Optics of Ibn Al-Haytham: Books I-III: OnDirect Vision, trans. with introduction and commen-tary by A.I. Sabra (London: Warburg Institute, 1989).No English translation of the parts of Ibn al-

Haytham’s work dealing with the properties ofcurved mirrors is currently available.

Manuscript received 30 April 2002.

Michael John Gorman teaches the history ofscience and technology at Stanford University.He directs the project “Artificial Eyes: Instru-ments, Observations and Art 1500–1700,” aproject on the history of optical instruments atthe Stanford Humanities Laboratory. He iscurrently developing an exhibition on the de-signs of R. Buckminster Fuller.

Gorman, Art, Optics and History 301

FORTHCOMING

Leonardo Music Journal Volume 13

Groove, Pit and Wave—Recording, Transmission and MusicPublication Date: December 2003

Despite Thomas Edison’s assumption that the gramophone was nothing more than a sonic autographalbum, suitable only for playing back the speeches of famous people, over the last 100 years recording hasradically transformed the composition, dissemination and consumption of music. Similarly, the business-like dots and dashes of Morse and Marconi have evolved into a music-laden web of radio masts, dishes,satellites, cables and servers. Sound is encoded in grooves on vinyl, particles on tape and pits in plastic; ittravels as acoustic pressure, electromagnetic waves and pulses of light.

The rise of the DJ in the last two decades has signaled the arrival of the medium as the instrument—thecrowning achievement of a generation for whom tapping the remote control is as instinctive as tapping twosticks together. Turntables, CD players, radios, tape recorders (and their digital emulations) are played, notmerely heard; scratching, groove noise, CD glitches, tape hiss and radio interference are the sound ofmusic, not sound effects. John Cage’s 1960 “Cartridge Music” has yet to enter the charts, but its sounds aregrowing more familiar.

For Leonardo Music Journal Vol. 13 we consider the role of recording and/or transmission in the creation,performance and distribution of music.

For more information, visit !http://mitpress.mit.edu/Leonardo/lmj".

Leonardo_36-4_255_344 7/17/03 12:02 PM Page 301