TY - JOUR
AU - Davis,, Whitney
AB - Abstract This paper considers the use of the ‘Mondrian Stimulus’, invented by Edwin H. Land of the Polaroid Corporation, in various investigations in the visual neuropsychology, the neuroaesthetics, and the social psychology of aesthetic response to works of visual art (notably, investigations by Semir Zeki and A. Michael Noll). What difference does it make—in the set-up of these investigations and in our interpretation of their putative results—that the Mondrian Stimulus might be taken to be a ‘real’ painting by the actual Dutch artist Piet Mondrian? How does the existence of a set of ‘real’ Mondrians—more or less well known to many people, including those investigated in experiments by Land, Zeki, and Noll—affect the ways in which the Mondrian Stimulus is apprehended? The paper argues that the Mondrian Stimulus is ‘bound’ to the history and visual recollection of ‘real Mondrians’ at the same time as the ‘real Mondrians’—in their historical afterlives—are bound to other creations of ‘modern abstract art’. These proposals enable a revised approach to the relations between invariant visual responses (such as Land and Zeki derived for colour vision) on the one hand, and different visual cultures or visuality on the other; because of its own special history of ‘binding’ and ‘unbinding’, the Mondrian Stimulus works as an interface between—a binding of—both ‘bottom-up’ perceptual input and processing and ‘top-down’ direction of attention. What is usually known in perceptual neuropsychology as a ‘Mondrian scene’ or a ‘Mondrian stimulus’ (MS) is an apparently abstract pattern, whether black-and-white or coloured, that resembles real paintings—or more generally certain styles of painting—by the Dutch artist Piet Mondrian (1872–1944). Under certain circumstances the MS might be taken by a beholder to be a ‘real Mondrian’ (RM). In some respects, the MS might be nearly identical visually with the RM, though many MSs are not. Visually an MS can look like a painting from one or another of the various phases of Mondrian’s career after his turn to abstraction in the mid-1910s—from his earliest abstract work (sometimes only black-and-white, using only horizontal and vertical bars) to his mature Neo-Plasticism (using only the ‘primary’ colours red, yellow, and blue admixed with black and white) to his ‘late’ and more animated abstractions, made in New York City (the Boogie Woogies). This paper’s main focus will be on RMs serving as Mondrian stimuli, most of which are not Mondrians. I will consider experimental and historical investigations in which the MS has been used to explore human subjects’ perceptual and aesthetic experience of their visual worlds. Three investigations will concern me here. As we will see, each partly turns on the way in which elementary and possibly invariant processes of visual perception which are triggered by the MS can be integrated with—or ‘bound’ to—different visualities. That is, to different visual cultures which organize different networks of what I have called the ‘forms of likeness’ of visible things, especially things made to be visible.1 These two registers, however, cannot be cleanly disaggregated, separating the processes which determine visuality (as is often said, ‘bottom-up’) from the processes ordered by visuality (‘top-down’). Rather, they are reciprocally constituted. Developing a middle ground, we might speculate, then, that a particular historical aesthetics is woven equally into both registers. Therefore it acts as the active interface between the visible world and visual culture. In general terms, this might seem to be self-evident. It is surprisingly difficult, however, to find an exemplary case that enables us to observe the interface at work. But the Mondrian Stimulus does indeed seem to be such a case. 1. Designing a Colour Mondrian Stimulus The term ‘Mondrian stimulus’ was originally applied by the inventor of Polaroid photography, Edwin H. Land, and his collaborators when they were investigating what they called ‘retinex’, the interaction of retinal images and cortical processing in visual perception and in particular the phenomenon of colour constancy (Figure 1).2 In 1963, Land had asked his collaborator Lucretia Weed to model the retinex MS as Mondrian-like—a reminder, as he said, of a particular RM he himself had seen in the Tate Gallery in London, almost certainly Mondrian’s No. VI/Composition No. II (1920) (Figure 2), though in fact Land’s MS is (and needs to be) quite different from the Tate RM in important ways. Weed did not have access to a colour photograph of the Tate RM and therefore configured her MS from a black-and-white photograph of it.3 It is no surprise, then, that the colour scheme added—or restored—to Weed’s retinex MS would not have appeared in most RMs and did not appear in the Tate RM, whether or not this was known to Weed. Fig. 1. View largeDownload slide Rendering of a Land ‘Mondrian’. By Julie A. Wolf (2018), after Edwin H. Land, ‘The Retinex Theory of Color Vision’, Scientific American 237 (1977), cover illustration. Fig. 1. View largeDownload slide Rendering of a Land ‘Mondrian’. By Julie A. Wolf (2018), after Edwin H. Land, ‘The Retinex Theory of Color Vision’, Scientific American 237 (1977), cover illustration. Fig. 2. View largeDownload slide Piet Mondrian, No. VI/Composition No. II, (1920). Oil on canvas, 99.7 x 100.3 cm. Courtesy of Tate Modern, London. Fig. 2. View largeDownload slide Piet Mondrian, No. VI/Composition No. II, (1920). Oil on canvas, 99.7 x 100.3 cm. Courtesy of Tate Modern, London. In fact, there are several MSs known to be associated with Land’s experiment as his MSs; we might call them the family of ‘real Land MSs’, the MSs actually produced by Land and his associates and used in their investigations of particular perceptual responses to the configurations. For reasons internal to his hypotheses about retinex, Land actually needed some black-and-white ‘Mondrians’. A colour MS is illustrated in black-and-white in a drawing of Land’s apparatus, and a different colour MS in his photograph of the apparatus in a publication that also illustrated yet another colour MS (Figure 1).4 In his influential synthetic book Eye, Brain, and Vision, the vision scientist David H. Hubel published another colour Mondrian, which he credited to Land; of all these real Land MSs, it is least like an RM in having, for instance, a prominent circular pink patch.5 There are, then, at least four real Land colour MSs more or less distantly resembling the prototype RM (in Land’s memory and Weed’s photograph), and of which they could be said to be transformations guided by the hypothesis of retinex newly introduced by Land into the visual world of RMs—the field of expectations in and hypotheses about the visual world, especially its colours, likely entertained by the real Mondrian. In addition, there are an indefinite number of MSs that are not ‘real Land MSs’ though they have been used in Land-like experiments. The neurologist Semir Zeki has taken the neuroscience of vision to be relevant to aesthetics, art history, and the history of visual art. In illustrating what he took to be one of the most salient scientific findings, Zeki has presented a diagram of a person looking at either an array of coloured patches (such as what could be taken to be an abstract painting, on the left of the diagram) or a field of movement (such as the black-and-white and perhaps televisual stimulus modeled on the right) (Figure 3).6 The abstract-painting-like pattern on the left of the diagram belongs to the family of ‘real Zeki MSs’. Another colour MS (not the one just mentioned) appears in a photograph of the Land apparatus which Zeki apparently used in his experiments on colour vision, and yet another colour MS in one of Zeki’s initial studies of colour vision in monkeys.7 According to Zeki, neuroscientific research based on brain imaging techniques (applied to the MSs) has determined that different groups of cells in different parts of the visual brain—functionally specialized and localized groups—‘fire’ when confronting the coloured and moving stimuli respectively, processing the affordances: colour is processed in the neural topography of the visual brain known as V4 and movement in V5. In turn, V4 processing of colour and V5 processing of motion are integrated in ordinary human perception of a three-dimensional, full-colour, and moving world; indeed, Zeki writes, ‘several visual systems’ act in parallel, ‘the activity in each leading to both seeing and understanding a particular attribute of the visual scene’.8 Fig. 3. View largeDownload slide Semir Zeki’s diagram of functional specialization in the human visual brain: V4 lower left, V5 lower right. Reproduced from Fig. 7.5 in his Inner Vision: An Exploration of Art and the Brain (Oxford: OUP, 1999). Courtesy of Semir Zeki. Fig. 3. View largeDownload slide Semir Zeki’s diagram of functional specialization in the human visual brain: V4 lower left, V5 lower right. Reproduced from Fig. 7.5 in his Inner Vision: An Exploration of Art and the Brain (Oxford: OUP, 1999). Courtesy of Semir Zeki. The neurotopographical ‘dissociation of colour and movement processing’ is one of Zeki’s main findings, along with the further finding that colour processing occurs measurably before (faster than) movement processing in delivering the output(s). For example, we can have an experience of ‘something red’ which, about 80 milliseconds later, is ‘moving from my visual left to my visual right’. As Zeki writes, then, a ‘hallmark of the visual brain is a temporal hierarchy superimposed upon a spatial parallelism’.9 In his studies of ‘art and the brain’ in Inner Vision and elsewhere, Zeki (re)connected the MSs to RMs, (re)identifying the one with the other. He can be taken to say that when someone is looking at an actual painting by Mondrian (modeled by the MS) their V4 is lighting up: that is what is going on ‘aesthetically’, if we take ‘aesthetics’ in its etymological sense simply to denote perception. And indeed Zeki’s descriptions of RMs as beheld in the real artworld—for example, at the Tate—tend to treat them as visual provocations specifically for V4. (It is worth noting, however, that in Zeki’s MS [Figure 3] V4 lights up in the particular way it does partly because there are no black contours between the areas of colour, which can inhibit contrast and consequently comparison between adjacent areas—comparison at the heart of Land’s theory of colour vision [Figure 1; Land’s MS also lacks black contours]. But in the Tate RM [Figure 2] and RMs of its type, there are such black contours—what have often been called ‘grids’. In principle, then, we might expect a difference—albeit a perhaps negligible one—between V4’s responses to the Land and Zeki MSs on the one hand and to the Tate RM on the other. This possibility, however, was not addressed in the investigations I am considering here, and for the sake of argument I will set it aside. In 1917–18, Mondrian himself made a handful of RMs with colour planes and no black contours, though the colour planes in these paintings do not directly abut one another.) For Land’s purposes in parsing retinex and for Zeki’s purposes in differentiating V4 and V5—both models were indisputably brilliant in theory, method and results—all the Land and Zeki MSs I have enumerated would seem to be interchangeable: they were equally viable for the experimentalists’ purposes. None of them, however, could ever be confused visually and formally—let alone socially and semantically—with an RM such as the Tate RM (Figure 2) that had initially provided the actual model created by Weed and Land. Or so we might think. All the MSs enumerated so far would not dupe the present writer (so he thinks)—an art historian who has studied Mondrian and his RMs for many reasons other than the writing of this essay. But what about the nine anonymous test subjects (all male) who were recruited for Zeki’s original brain imaging of V4 as activated by a Zeki MS? Or the even more shadowy subjects who were recruited in Land’s set-ups, which depended heavily—among other things—on normal conditions (among males) of full capacity for red/green colour discrimination? Would some of them have taken the MS to be an RM, whether or not they ‘liked’ it or knew anything about the history of Mondrian’s art? If they did (or not), what difference did that make? 2. ‘Binding’ Unbound? Several assumptions—and guesses—were built in to Land’s and Zeki’s experiments. Above all, the MSs had to be analytically and experimentally isolated from certain features and forms of likeness of the RMs—even as RMs provided the model for certain questions asked of MSs and the answers were used in turn to account for the RMs. Zeki’s ‘main difficulty’, he said, was to deploy a stimulus—the MS—that would activate some cortical locales of visual processing but not others; the hypothesis of such localization, remember, was under investigation. ‘Most visual stimuli’, as Zeki recognised, ‘will contain attributes of possible relevance to many specialised visual areas’.10 For example, the apparent edges and boundaries in an MS, including the fact that RMs were intended to be seen as framed and presented on a wall more or less perpendicular to the beholder’s direct line of sight, will necessarily activate certain ‘orientation selective’ complexes of cells in V3 and V3A which are wholly indifferent to colour. Overall, if Mondrian himself had been able to adopt present-day neuroaesthetic terms—terms perhaps somewhat aligned with his own quest to find putative universals in art and its perception—he would likely have said that he intended RMs to stimulate both colour vision (V4) and orientation selectors (V3 and V3A) in perceptually unusual ways—for example, by using only the ‘primary’ colours and by fashioning peculiar grids of colour planes that manage to preserve ‘rectangularity’ under rotation in perspective, both of which we could take to be infrequent features of the natural visual/visible world outside paintings. We have just encountered an underlying neuropsychological process sometimes described as the ‘binding’ of topographically separate cortical representations of such different features of a single stimulus-object as its colour and its orientation—binding which Zeki needed to dodge so far as possible. To do the dodge, Zeki’s experiments included a contrast between a coloured MS and a gray-scale replica of the very same MS—that is, two MSs that were ‘identical in all respects save the one of interest’, namely, colour. These MSs make up the family of real Zeki MSs. As predicted, the test subjects’ V4 reacts to the former MS (colour) but not to the latter (gray-scale). Zeki concluded, then, that the added value—colour—is constituted in V4 processing of the MS. Of course, an RM can also have certain distinctive ‘depth’ and ‘motion’ properties in viewers’ actual perceptual and corporeal engagements with it. Somewhat like Jackson Pollock, Mondrian sometimes partly worked on his canvasses laid out horizontally—in Pollock’s case, rolled out on the floor; in Mondrian’s case, laid out flat on a table. Combining, coordinating and calibrating vertical and horizontal presentations (MSs) of the RM would allow him to address the ‘rectangles’, colour planes, and grids under the full range of oblique angles of vision—that is, the full continuum of the potential visibility of the RM as a beholder moves around it. When a beholder approaches an RM in a gallery or museum (that is, as hung upright on a wall) under a highly oblique visual angle, the rectangles might be perceived as trapezoids. (Moreover, in many RMs the colour planes and/or bars extend around the side edges of the canvas, and they can only be seen, then, from the side; Mondrian’s frames often allowed these sides to be inspected.) In general, an oblique approach can afford a greater sensation of virtual depth in the configuration than the straight-on view. Sensitive to all this, in 1978 the art historian Meyer Schapiro gave a compelling account of ‘cue[s] for depth’ in the ostensibly ‘impenetrable plane of the canvas’ in some of Mondrian’s lozenge-shaped abstract paintings.11 Entirely independent of the beholder’s motion, we can readily find RMs that might be thought—in binding—to engage and entertain V5 (or at any rate the feeling of motion) as much as V4 (the processing of colour) (see Figure 3). Because Zeki wanted unbound colour-stimulating MSs, against the grain he had to downplay the role of motion effects in Mondrian’s Boogie Woogies (painted in his last years in New York City), such as Broadway Boogie Woogie of 1942–3. Even these configurations, according to Zeki, are ‘static’. Strictly speaking, this is true. But beholders will routinely cite the jittery, bouncy look of the Boogie Woogies—the animation (as it were the ‘eventfulness’ or mouvementé) of the pattern in the visual field. The art historian Yve-Alain Bois has argued in convincing detail that Mondrian’s explicit intention in these RMs was to ‘prohibit any stasis or fixing of perception in a systematic manner’.12 Beyond finding cortically localized (and temporally staggered) ‘seeing’ of colour, motion, orientation, etc., Zeki’s MS was designed, he hoped, so as to ‘minimise any effect of memory and learning’.13 For example, because the MS is putatively ‘abstract’ it supposedly does not involve perceptual and cognitive routines of ‘object-recognition’. (It is, as it were, just colour + orientation + motion effects.) That is, supposedly it does not involve beholders’ recognition of objects depicted in a virtual (pictorial) space, though the MS, and especially an RM serving as an MS, might itself be recognized as a special kind of real object in visual space—that is, as a configuration somewhat resembling RMs in galleries and museums (such as the Tate RM, Figure 2): ‘This is an abstract painting I see before me’. It seems implausible to hope that some of the test subjects presented with an MS did not recognise ‘something resembling a real Mondrian painting’ as an-object-in-vision for them, however different from RMs. (I will return to the ‘something resembling an RM’—that is, to the ubiquity of MSs in modern visual life.) Indeed, the very visual uniqueness of the MS—its distance from the ordinary stimuli of the natural visual world—might have provoked the test subjects automatically to reel in their ‘memory and learning’ to make sense of the array presented to them. ‘Is this an abstract painting I see before me?’ ‘Yes, it likely is.’ Recursively, then, ‘I am attending to colours, shapes and boundaries as such (sub-routine: and as recognising an abstract painting before me, I do not attend to its many possible representational and figurative suggestions).’ Land had confronted a similar problem. In investigating colour vision, he chose his MS (reminding him of the Tate RM, Figure 2) partly in order, he hoped, to eliminate the effect of ‘memory colours’ in test subjects’ reactions—that is, their existing knowledge that an orange is (ordinarily) orange, blood is (ordinarily) red, etc. (The role of memory colour had been an ongoing topic in the history of research on the constancy of colours ‘regardless of always unpredictable, shifting and uneven illumination’.14) Because the MS/RMs are supposedly abstract, one can have no memory of the colours of the objects they depict—except, perhaps, of the colours of RMs as abstract. In this latter context, we should remember that some beholders of MS might know, for instance, that RMs produced during Mondrian’s mature Neo-Plastic phase (including the Tate RM which suggested the MS to Land) wholly abandoned the use of green—a fact Land only learned long after his initial studies of retinex. In Mondrian’s mature theory, Neo-Plastic RMs must use only red, yellow, and blue (the ‘primary’ colours) and black and white, whether stand-alone or admixed with the primaries. Land’s MS specifically and advisedly does employ green because retinex theory is partly an investigation of red/green colour discrimination. In virtue of such beholders’ memory and learning of the colours of Neo-Plastic RMs, they might recognize, then, that Land’s MS is not an RM: ‘This is not an abstract painting by Mondrian I see before me.’ While Land and Weed had built their colour MSs from a gray-scale representation of an RM in order to investigate colour vision, in order to investigate colour vision Zeki extracted a gray-scale representation from a colour MS. Either way, these were ingenious experimental manoeuvers. Nonetheless the problem of binding is intractable. A neuroaesthetics that is granular at the neuronal level (like Zeki’s) often needs supposedly ‘unbound’ stimuli like the MS (real Zeki MSs), or at least a method of dodging the binding. By the same token, a neuroaesthetics that is system-wide at the neuronal level (like Land’s) also often needs ‘unbound’ stimuli—real Land MSs. By contrast, art history and art criticism usually investigate the same stimuli (such as RMs) as fully ‘bound’. 3. The Mondrian Stimulus Bottom-Up and Top-Down According to Zeki, when someone is ‘looking at’ (perceptually processing) an MS, their V4 is lighting up in a particular way. Indeed, when anyone who is sighted and has decent colour vision is viewing any MS under appropriate illumination and from an appropriate distance and viewing angle (the conditions ordinarily arranged for RMs in galleries and museums), their V4 is lighting up. One of Zeki’s major conclusions, in fact, is simply that ‘at an elementary level what happens in the brain of one individual when he or she looks at works of art is very similar to what happens in the brain of another.’15 This point is widely shared among neuroaestheticians. For this very reason, of course, and still remaining entirely within the framework of neuroaesthetics, one might take the results to be underwhelming in saying much about what’s going on neurophysiologically when someone is ‘looking at’ (aesthetically apprehending) an RM notionally made specifically to be looked at as a stimulus arguably to be called ‘abstract art’. (If one exits the framework of neuroaesthetics and enters, say, the framework of cultural-historical hermeneutics, such recent writers as Matthew Rampley do take the Zeki-style results to be underwhelming in very virtue of their supposedly reductive generality.16) But present-day neuroperceptual psychology does not confine itself either in theory or in practice to the ‘elementary level’—that is, to the early visual system where my V4 lights up just like your V4 in looking at an MS. In the later visual system, ‘selection and recognition of objects in connection to memories and meaning’ are processed.17 Overall, a good deal of contemporary neuropsychology supposes—perhaps obviously enough—that ‘aesthetic experience is derived from the interaction between the top-down orienting of attention and bottom-up perceptual input’.18 Indeed, we might say that the ‘interaction between the top-down . . . and bottom-up’ simply is aesthetic experience—the mutually binding network. Still, aesthetic experience in this sense might be analytically visible—observable—neither to present-day techniques of brain imaging (for example, to fMRI imaging) nor to conventional methods of cultural-historical analysis (for example, to reading the contemporary record of art criticism). Special circumstances, methods and manoeuvers might be needed to spot it. 4. A Real Mondrian In Different Visualities What difference does it make that a particular beholder is ‘looking at’ an MS that is a particular RM, such as the Tate RM, which is analogous for that beholder to other configurations in modernist art? In this beholder’s ‘avant-garde’ visuality or culture of looking, abstract paintings have a particular ‘form of likeness’: as works of art, they explore the sensory fluxes of modern life. In Zeki-style neuroaesthetics, V4 lights up in the same way whether the configuration is one such RM (the Tate RM)—that is, it bears the forms of likeness just mentioned, belonging to that avant-garde visuality for one beholder—or, alternately, is another MS (though it is still the real artifact that is Tate RM) with different likenesses, likenesses presumably constituted for a different beholder looking through the lens of a different visuality. In this visuality, perhaps the painting belongs to a world of commercial experiments with peoples’ likes and dislikes with respect to colour samples arrayed in various ways. Some beholders in London looking at the Tate RM might activate this form of likeness as the salient analogical matrix for what they are seeing, though their way of seeing might be described (by the avant-garde beholder) as a relatively uneducated layperson’s or even a ‘philistine’ visuality—one in which the RM is a quasi-decorative experiment with our colour hedonics, not ‘art’. For the moment, let us assume that both beholders—the avant-gardista and the philistine—are relatively untrained and unpracticed in the actual making of drawings, paintings, etc., though they have very different dispositions, beliefs and judgments regarding the products of such making as presented to them. That is, for the moment we will assume that neither beholder is an actual artist like the real Mondrian. It seems reasonable to suppose that the nine human subjects investigated in Zeki’s original experiments on the cortical localisation and specialisation of colour vision (V4) contained some people to whom MS (let us suppose it was the actual Tate RM) would be like ‘Modern Art’ in the valued avant-garde sense and other people to whom it would be like ‘Modern Art’ in a philistine sense (that is, a charlatan ‘art’ that is not ‘art’). In other words, the test population was likely differentiated internally by visualities though not by the invariant visual activity of V4 across the group. After all, in his seminal 1919 essay ‘Natural Reality and Abstract Reality’ Mondrian himself had set up an imaginary trialogue between interlocutors Z, Y and X—respectively an ‘Abstract-Real Painter’ (Mondrian himself in 1919), a ‘Naturalistic Painter’ (such as Mondrian before his turn to abstraction a few years earlier), and a ‘Layman’ who represents the bemusement of the general and relatively inexpert public when confronted with abstract art.19 Implicitly, therefore, Mondrian imagined three identical MSs, one of which is a full RM for Z.20 In addressing this question—why (arguably) does V4 light up in pretty much the same way for the avant-gardista and the philistine?—a conventionally culturalist art historian, it seems to me, has two options. The first option (more or less pursued by Rampley) is simply to ignore the material phenomenon of V4’s neural activation, however substantive it might be as a neurophysiological event of interest to neuroscientists; it is irrelevant, simply not salient in the description, historicisation, and critique of avant-garde and philistine visualities as brought to bear on the products of modern art, such as the Tate RM. It is not salient, of course, precisely because V4 lights up in pretty much the same way for both the avant-garde beholders and the philistine beholders, and therefore as an invariant correlation it says nothing one way or another about that avant-gardism or philistinism, though this ‘cultural difference’ is, presumably, a salient parameter—what the Tate RM is differently like visually for the avant-gardistas and the philistines respectively. This first option, however, is contradicted by the second option. In virtue of the discriminably and even dramatically different visualities of avant-garde and philistine viewers of ‘Modern Art’, should not V4 indeed light up differently?—definitely and measurably so? Since it does not, should one conclude, in turn, that whatever is going on perceptually and cognitively and visual-culturally is not correlated at all with the neural activity of V4? Both of these options are unpalatable, to me at any rate. The first option is unpalatable because I think it is wrong: it is not irrelevant analytically that V4 lights up in pretty much the same way for the avant-gardista and the philistine. Indeed, it could be extremely significant. And the second option is unpalatable because it is implausible. In what is usually called the strict ‘modularist’ approach to perception and cognition, it assumes the absolute mutual impenetrability of the visibility of colours to someone, on the one hand, and, on the other hand, that same person’s ideology of art (the RM is ‘art’), knowledge of pictorial representation (the MS is ‘abstract’), and colour theory (such as Land’s and Zeki’s). Paradoxically enough, strict modularism is the very psychology that a high culturalist historian of the potential effects of MS (especially RM) on V4 presumably should not want to adopt. 5. Will the Real Mondrian Please Stand Up? There could be a third way, however—what I have already referred to as the interface of aesthetic experience. It needs to be explored in further intersections of art and cultural history, empirical and sociological aesthetics, and visual neuropsychology. But it is predicted by a general theory of visual culture—that is, by an integrated model of what is cultural about vision and what is visual about culture.21 On this hypothesis, V4 lights up in pretty much the same way for the avant-gardista and the philistine not because it has nothing to do with avant-gardism and philistinism—with their supposedly distinctive dispositions, beliefs and purposes. It lights up in pretty much the same way because there are no such things as fully totalised or ‘pure’ avant-garde and philistine visualities that could light up in different ways: they are chimaeras of art and cultural history in part precisely because—though not only because—they have no neural correlate. Rather, a garland of cross-cutting forms of likeness—successions of analogy to Mondrian’s colours, shapes, and lines and successions from them—might be substantially woven into both of these putatively opposed visualities, binding both of them to certain shared forms of likeness which seem to organise visual and aesthetic responses to MSs—binding them. Yet another episode in the history of the MS—and in the visual reception of RMs and the production of MSs partly bound to it—provides uniquely intriguing evidence that bears on this possibility. In 1962, at the same time as Land was working on retinex at Polaroid, a young engineer at Bell Telephone Laboratories in New Jersey, A. Michael Noll, used an IBM 7090 mainframe digital computer linked to a Stromberg-Carlson 4020 Microfilm Printer to produce ‘a series of interesting and novel patterns’—arrays of dots and of straight lines constructed between points semi-randomly distributed in a two-dimensional field (an XY plane), creating ‘haphazard’ zig-zag configurations with some apparent three-dimensionality.22 Of course, a number of decisions—for example, about the number, the position, and the darkness of the dots—were built into Noll’s programme, written in Fortran. Noll did not regard his decision-making activity as the programmer to be ‘artistic’, and he gave no ‘forethought to the artistic merit’ of the results. Retrospectively he saw, however, that this experiment and subsequent experiments by him and by other pioneers of ‘computer graphics’ had inevitably developed partly ‘in relation to the aesthetics of the era’.23 Because Noll’s initial results raised these very questions, soon thereafter he programmed an IBM 7094 to produce semi-random configurations that could ‘appear similar in composition’ to Mondrian’s well-known black-and-white Composition in Line of 1916–17 (Figure 4, illustrating its second and final state), often regarded by art historians as a breakthrough on the painter’s path toward absolute abstraction. Noll’s configurations can be called the set of ‘real Noll MSs’ (Figure 5).24 Fig. 4. View largeDownload slide Piet Mondrian, Composition in Line (second state) (1916–17). Oil on canvas, 108 x 108 cm. Courtesy of Kröller-Müller Museum, Otterloo, The Netherlands. Fig. 4. View largeDownload slide Piet Mondrian, Composition in Line (second state) (1916–17). Oil on canvas, 108 x 108 cm. Courtesy of Kröller-Müller Museum, Otterloo, The Netherlands. Fig. 5. View largeDownload slide A. Michael Noll, Computer Composition with Lines (1964), © 1965 AMN. Courtesy of A. Michael Noll. Fig. 5. View largeDownload slide A. Michael Noll, Computer Composition with Lines (1964), © 1965 AMN. Courtesy of A. Michael Noll. In studying the prototype RM and in writing his programme for the IBM 7094, Noll concluded that Mondrian had not operated randomly though Composition in Line has the visual feel of an apparently ‘random distribution’. Rather, Noll thought, the artist had carefully ‘planned’ the placement and length of each horizontal and vertical bar and their single and multiple crossings. Noll’s task was to get his programme to produce relatively unbound configurations with ‘reasonably similar’ visual effects to the RM, more or less letting the bound semantic and symbolic chips fall where they would. Still, if Mondrian had used ‘some scheme or program’—an ‘exact algorithm’—to produce Composition in Line, it was unknown to Noll, and to most art historians of real RMs. In inputting his variables and their parameters, then, Noll therefore worked by ‘trial-and-error’ (as Mondrian himself conceivably might have done) to arrive at his suite of real Noll MSs.25 Whether his computer-generated MSs were produced for analytic purposes, creative purposes, or both, at some point in Noll’s experiments—given its stated parameters—they had to be compared by real beholders with RMs. As the title of Noll’s initial publication suggests (‘Human or Machine’), these comparisons were variants of a classic ‘Turing Test’.26 Does a real Noll MS look sufficiently indiscernible from, even identical to, an RM (specifically, Composition in Line) such that a beholder would take it to be an RM? And therefore (on one argument at any rate) does the machine and/or the semi-randomizing programme that generated the MS have the ‘intelligence’ of the human painter who made the RM, whether or not they used a semi-randomizing procedure? To do the Turing Test, Noll showed MS versions of Mondrian’s Composition in Line and a real Noll MS to one hundred test subjects at Bell Labs, asking them ‘which of the two do you think was done by the computer?’ Sixty-nine subjects occupied ‘technical’ jobs (physicists, engineers, programmers, etc.) and thirty-one occupied non-technical jobs (clerks, typists, etc.); there were fifty-six men (mostly technical) and forty-four women (non-technical) (remember, this was the 1960s). In addition, the subjects had recorded their subjective assessments of ‘abstract art’ in advance: 34% ‘strongly liked or liked’ it, 46% were ‘indifferent’, and 20% ‘disliked or strongly disliked’ it. As the MSs for this experiment, Noll used black-and-white photographs of the RM, which has an off-white ground, and the Noll MSs printed by the SC 4020—photographs which had then been ‘copied xerographically to be identical in quality’. This mediation presumably would have had the effect of eliminating the palpable differences between Mondrian’s painted surfaces (partly captured in good photographs of Composition in Line), which sometimes have visible pentimenti, and the surfaces of the Noll MSs. The latter are even, homogeneous and flat, like the Land and Zeki MSs. (Moreover, Noll’s photograph could not show that Mondrian had mounted the RM in a recessed frame which allowed the side edges of the painting to become part of the experience.) Effectively, then, Noll was requiring his subjects to ‘look’ only at the placement and ordering of the black vertical and horizontal bars in all his MSs. Noll used various methods and tests to filter out such effects as the subjects’ advance knowledge that one of the patterns was produced by computer: in the event, this knowledge had been transmitted to the subjects in the very questions they answered, potentially binding the visibility of the MS. But supposedly it had no effect. (Neither did gender, though it dramatically differentiated the technical/male and non-technical/female groups.) We might speculate, of course, that some of the subjects, whether technical or non-technical, might have had some degree of formal exposure to art and art history (say, a course in college) and that some of them might even have been amateur artists. But this possibility was not investigated by Noll, and I will take it that the test subjects at Bell Labs represented a spectrum of the ‘general public’ for ‘Modern Art’—not a specialised population of educated artists and art historians/critics. Noll’s results were unexpected, at least from the points of view that we might intuitively attribute to computer experts on the one hand and art experts on the other. Only 28% of the subjects were able correctly to identify the computer pattern, while at the same time 59% of the subjects preferred the computer pattern. Indeed, subjects ‘strongly liking and liking’ abstract art preferred the computer pattern by more than three to one—perhaps, Noll speculated, because they were ‘more accustomed to the randomness found in many abstract paintings’ and therefore visually (and culturally?) preferred the ‘more random’ of the two MSs, namely, the real Noll MS. By contrast, non-technical subjects were more often ‘fooled’ into taking the RM to be the computer pattern—perhaps, Noll speculated, because they expected such a pattern to have the more ‘mechanical, orderly’ look of the RM. Stated the other way around, they too ‘associated randomness with human creativity’. The technical subjects were supposedly more likely to understand that a computer-generated pattern could actually fall anywhere on a spectrum from visibly ‘orderly’ to visibly ‘random’. Therefore, Noll proposed, they tended to disregard the visual ‘differences in randomness’, however visible, between the two MSs—differences that seemingly were perceived by some subjects in various groups. And therefore the technical subjects tended toward ‘pure guessing’ in identifying the real Noll MS, which they would understand had been generated by the programme’s incorporation of chance—of pure guesswork, as it were. Nonetheless, even these subjects had significantly greater preference for the real Noll MS—the ‘more random’ MS. There is (and at the time there was) much more to say in interpreting Noll’s results and the results of subsequent empirical and quasi-sociological studies along the same lines. For my purposes, and overall, the real Noll MS emerged both as the RM and as preferred, especially by the subjects who valued abstract art. Let us assume that an aesthetics of art (which all the subjects identified as such) was represented in the test population’s various visualities, ranging from a stated avant-garde-like liking of abstract art to a philistine-like disliking. Nonetheless Noll’s results indicated that a substantially shared form of likeness operated across and was distributed among these discursively defined—self-acknowledged—subset groups of visualities in the test population, rendering the stimulus similarly visible to them. To be sure, this shared form of likeness did not wholly flatten their differences. Still, it emerged as an intersection of all the groups in relation to which one can more exactly define the specifically non-commensurate areas of visual judgment, aesthetic disposition, and background knowledge that was particular to each group, distinguishing its ‘visuality’ as identified not so much in brain imaging as in the sociology of a Turing Test with respect to ‘computer’ and ‘abstract’ art. And this is what we should expect, whether or not on the basis of neuroaesthetic findings that perhaps index the same phenomenon—namely, invariant visual responses (to orientation, pattern, colour, etc.) as bound to a seemingly shared form of likeness creating another register of invariance. 6. Forms of Likeness: The Afterlives of the Stimuli In responding to Noll’s results, Schapiro endorsed Noll’s explanation of the test subjects’ marked ‘taste for the relatively shapeless [MS] computer version [of the RM Composition in Line]’ as due to their awareness of ‘the frequency of randomness in recent abstract paintings and the association of that feature with an idea of the creative in general’—specifically, as Schapiro rightly added, with the ‘vogue of Abstract Expressionist painting’ and its ‘impulsive, spontaneous-looking products’.27 This form of likeness was derived from a specifically American visual art of the postwar period, notably Pollock’s. By the early 1960s, seemingly it had constituted a widespread visual culture—what we might call the afterlife of the ‘Pollock Stimulus’ (PS)—distributed to the employees of Bell Labs. And it was readily bound to the stimuli proffered in Noll’s experiment, which were, of course, afterlives of the RM that is Composition in Line. Noll did not explicitly intend his Computer Composition with Lines to look like a real Pollock (RP). And it does not. It is far closer visually and formally to an RM than to such RPs as Lavender Mist: Number 1, 1950 (1950) in the National Gallery of Art in Washington, DC, and One: Number 30, 1950 (1950) in the Museum of Modern Art in New York—paintings which some of Noll’s test subjects could have seen in the flesh as well as in many reproductions. Nonetheless, in programming the IBM 7094 Noll helped his MS to become like a Pollock—like it analogically though not like it visually—by introducing semi-randomising operations that were popularly perceived to be ‘arty’ Pollock-style procedures, whether they were culturally endorsed by avant-gardista visuality as ‘heroic formlessness’ or culturally disdained by philistine visuality as ‘the daubings of a monkey in the zoo’, to use Schapiro’s phrasings.28 And so it happens that the afterlife of RPs among Noll’s test subjects seemingly countermanded the afterlife of RMs—the actual prototype for Noll’s MSs. Perhaps this eventuality seems predictable. Historians are used to the idea that Abstract Expressionism defined many postwar Americans’ picture of ‘Modern Art’—a form of likeness binding their visual reactions across the board, perhaps because Abstract Expressionism itself, as has often been argued, drilled deeply into certain perceptual-existential realities widely shared among mid-century Americans regardless of their inclinations respecting abstract art. But actually the result is art-historically unexpected. Two decades after Mondrian’s death in 1944, as the art historian Nancy J. Troy has written, Mondrian’s ‘signature abstract style had become instantly recognizable as a result of its widespread appropriation, adaptation, and reproduction in architecture, graphic design, fashion, advertisements, and reproductions in mass-circulation magazines’.29 For example, in 1965—in the very years of Land’s and Noll’s experiments—the French couturier Yves Saint-Laurent unveiled his Mondrian Dress, a smash hit in Europe and America that far eclipsed any artworld exhibition of RMs.30 When functioning as an MS the RM in Noll’s population of beholders seemingly was partly bound across the board to PSs—as it were inhabiting the visual common ground (partly created by the PSs) shared between what could otherwise have been wholly polarized cultures of visual response to modern abstract art. 7. What Makes the Difference? I have tried to pinpoint an area of common ground between disjunct visualities that could help make sense of seemingly invariant ‘unbound’ perceptual reactions to the MS—as they are actually bound in the history of the MS/RM. Now we can briefly look at the other side of the coin: the disjunct reactions as disjunct, picked up empirically as a variation in (neurophysiological) response which is possibly correlated with a real variance in the subjects’ visual judgment, aesthetic disposition, and background knowledge. As Noll recognized, one of the drawbacks of his initial study was its lack of data on the art education of the test subjects. In a finely tuned book-length inquiry into ‘the aesthetic encounter’ with visual art as experienced and interpreted by expert art historians and curators, the social psychologists Mihaly Csikszentmihalyi and Rick E. Robinson confirmed what we might expect: the experts highly valued (and could discursively describe) the contribution of their knowledge of art and its history to their ‘raw’ perceptions and feelings. Therefore they were painfully aware, as they thought at any rate, that non-experts—to quote one art expert—‘aren’t ready to look’.31 Recent research in neuroaesthetics and empirical aesthetics has suggested that, broadly speaking, subjects with substantial training tend to approach visual art and to react to it in distinctive ways compared to less-expert beholders. For example, they recognize objects depicted in somewhat indeterminate pictures more quickly, using more time to scan structural and abstract features of the configuration. On a moment’s thought this finding makes sense. The artist, like other kinds of art expert, has probably learned to look with self-monitoring purpose. Precisely because these subjects can readily process a visual stimulus qua visual, they are substantially ‘thinking it’ (extra- or post-visually) in ways not yet acquired by the novice—the non-expert—who is, as it were, still simply struggling to see the stimulus well enough to complete a given task (such as copying it).32 In taking Noll’s test, we might guess these experts—and unlike Noll’s subjects at Bell Labs—would often have successfully spotted the computer-generated MS precisely because they would already have successfully spotted the RM. 8. Why Mondrian? I have explored certain strands in the history of the Mondrian Stimulus: from real Mondrian paintings, to red/green Land-Mondrians, to Zeki-Mondrians-for-colour-vision-only, to Mondrians for avant-garde visualities and philistine visualities, to computer-generated Noll-Mondrians, to the ‘Mondrian Brand’ and back around again to real Mondrian paintings. (For economy of space, I have not addressed psychoanalytic MSs and shamanistic MSs, though they exist.) It is striking that some of the deepest themes of the psychology, the philosophy, and the history of art have been engaged specifically by way of the MS. Why Mondrian? Surely it is not simply because of Mondrian’s own documented universalist and spiritual concerns, shared by other nineteenth- and twentieth-century artists but likely to be unknown to all but the most historically informed beholders. Only experts in RMs and other denizens of the modern artworld would care about the ‘meaning’ of Mondrians for Mondrian—and even among these experts there has been room for clashing interpretations of Mondrian’s intuitions, ideas and intentions. And it is not merely because of the ‘look’ of Mondrian’s paintings as they have been conveniently fitted in, it seems, to diverse empirical, theoretical and historical inquiries that have little to do with the history and context of RMs. The whole point of many MSs is to alter the look of RMs—even to unbind them so far as possible—in order to explore the parameters of a wider class of visual and aesthetic phenomena which the RMs might exemplify. Is it, then, because of a mere historical accident—the fact that Land came up with a pattern which happened to remind him of the real Mondrian he had seen in the Tate Gallery (Figure 2)? After all, he could have been reminded of a real painting by followers of Mondrian or of real paintings by other members of De Stijl—or, if the timing had been right, of a real painting by Kasimir Malevich, Alexander Liberman, Josef Albers, Victor Vasarely, or Ellsworth Kelly, any one of which could have worked for Land. Or is it because of the historical accident in which Noll might have just happened to like the orderly semi-randomness of Mondrian’s Composition in Line (Figure 4)? Noll never explicitly tells us why he picked that painting for his Turing Test—and other pioneers of computer graphics chose their models in works by Paul Klee and Bridget Riley instead. Is it, then, because real Mondrians can be—and historically were—something for everybody? Possibly. But so too was Pollock, RPs, and PSs. We have seen that the PS could actually bind the MS all the way to the ground, even as the RPs themselves came into being—in endless historical recursion—as an un- and rebinding of the MS (that is, in Pollock’s perception of the forms and legacies of preceding European abstract arts). Probably part of the answer has to do with the very feat Mondrian himself hoped to accomplish, as he stated in the words of Painter Z in Natural Reality and Abstract Reality—namely, to create the fullest possible distancing of his painted abstractions from all of the conditions of natural visual perception of the naturally visible world.33 He tried to be maximally unbound in the most elementary registers—an extreme and possibly unique visual proposition which was rightly sensed as informatively extreme, and then imitated and reconstructed, by his close students in the persons of Land, Noll, Zeki, and their collaborators when delving into the conditions of the binding and unbinding of Mondrian’s free radicals. And so we find that if RMs are bound all the way to the ground by RPs, MSs are bound all the way to the ground by RMs. Footnotes 1 See Whitney Davis, A General Theory of Visual Culture (Princeton, NJ: Princeton University Press, 2011). 2 Edwin H. Land and John J. McCann, ‘Lightness and Retinex Theory’, Journal of the Optical Society of America 61 (1971), 1–11. I will not describe retinex theory; for Land’s most non-technical account, see Edwin H. Land, ‘The Retinex Theory of Color Vision’, Scientific American 237 (1977), 108–128. 3 According to the recollection of John McCann, ‘Lessons Learned from Mondrians Applied to Real Images and Color Gamuts’, in Todd Newman and Jack Holm (eds), Color Science, Systems, and Applications: Proceedings of the Seventh Color Imaging Conference (Scottsdale, AZ: Society for Imaging Science and Technology, 1999), 1–8. 4 For these Land MSs, see Edwin H. Land, ‘The Retinex Theory of Colour Vision’, Proceedings of the Royal Institute of Great Britain 47 (1974), 23–58, at 29, Figure 4; Land, ‘The Retinex Theory of Color Vision’, cover, 5. 5 David H. Hubel, Eye, Brain, and Vision (New York: Scientific American Library, 1988), 86. 6 Semir Zeki, Inner Vision: An Exploration of Art and the Brain (Oxford: OUP, 1999), 64, Figure 7.5. 7 Ibid., 187, Figures 18.1, 18.2 (first published in 1983). 8 Zeki, Inner Vision, 63. 9 Semir Zeki, et al., ‘A Direct Demonstration of Functional Specialization in Human Visual Cortex’, Journal of Neuroscience 11 (1991), 641–649, at 643; Semir Zeki and K. Moutossis, ‘Temporal Hierarchy of the Visual Perceptive Systems in the Mondrian World’, Proceedings of the Royal Society of London B, 264 (1997), 1415–1419, at 1415. 10 Zeki, et al., ‘A Direct Demonstration’, 645 (my italics). 11 Meyer Schapiro, ‘Mondrian: Order and Randomness in Abstract Painting’ [1978], in his Modern Art, 19thand 20thCenturies: Selected Papers (New York: Braziller, 1979), 233–260, at 235–238. For the orientations both of Mondrian’s acts of painting and of one’s viewing of RMs, see the superb close analysis by Hans Janssen, ‘Working at a Table’, in Francesco Manacorda and Michael White (eds), Mondrian and His Studios: Colour in Space (London: Tate, 2014), 112–124. In his studios, Mondrian sometimes hung his paintings quite high, foreclosing a straight-on view. To be sure, the conformation of the apparent virtual shape of things in a picture can remain remarkably constant over different viewpoints. But greater obliquity of viewpoint generally reshapes the appearance of things under the ‘head-on’ view, and transforms the perceived dimensionality of depicted objects and space. For a detailed exploration, see Whitney Davis, Visuality and Virtuality: Images and Pictures from Prehistory to Perspective (Princeton, NJ: Princeton University Press, 2017). 12 Zeki, Inner Vision, 186; Yve-Alain Bois, ‘Piet Mondrian: New York City’, Critical Inquiry 14 (1988), 244–277, at 252. Admittedly the cortical localization of this startling effect might not be straightforwardly V4 on the one hand or V5 on the other. Margaret Livingstone has tried to show that the animatedness of Broadway Boogie Woogie is partly an effect (presumably all-V4?) of Mondrian’s subtle crafting of the near-equiluminance of the colours of the little yellow and gray squares on the one hand and the off-white background on the other hand—an equiluminance which Livingstone has strikingly demonstrated by reducing the colour RM, yet again, to its commensurate gray-scale MS (colour and gray-scale MSs illustrated in Margaret Livingstone, Vision and Art: The Biology of Seeing [New York: Abrams, 2002], 154–155), again discovering the added value of animatedness in Mondrian’s arrangement of colours. 13 Zeki, et al., ‘A Direct Demonstration’, 72. 14 Land, ‘The Retinex Theory of Color Vision’, 2. 15 Zeki, Inner Vision, 218. 16 Matthew Rampley, The Seductions of Darwin: Art, Evolution, Neuroscience (University Park, PA: Penn State University Press, 2017); although compare David Hulks, ‘Art History and the Biological View’, Art History 40 (2017), 209–211. 17 Jon O. Lauring, ‘Visual Art’, in his An Introduction to Neuroaesthetics (Chicago: University of Chicago Press, 2014), 128–149, at 131. 18 Ibid., 133, 149. In an influential synthesis, Robert L. Solso proposed to model the crossroads between ‘nativistic perception’ (bottom-up) and ‘directed perception’ (top-down') (see The Psychology of Art and the Evolution of the Conscious Brain [Cambridge, MA: MIT Press]). In ‘Visual Art’, Lauring summarises a growing set of empirical investigations which follow Solso’s roadmap. 19 Piet Mondrian, Natural Reality and Abstract Reality: An Essay in Trialogue Form 1919–1920 (New York: George Braziller, 1995). 20 Piet Mondrian, The New Art—The New Life: Collected Writings, ed. and trans. Harry Holtzman and Martin S. James (London: Thames and Hudson, 1987); for sustained discussion of the text, see Mark Cheetham, The Rhetoric of Purity: Essentialist Theory and the Advent of Abstract Painting (New York: Cambridge University Press, 1991), 45–60. 21 See Davis, A General Theory of Visual Culture, 277–340. 22 A. Michael Noll, ‘Patterns by 7090’, Bell Telephone Laboratories Technical Memorandum (1962), 62-1234-14. 23 A. Michael Noll, ‘Early Digital Computer Art at Bell Telephone Laboratories, Incorporated, 1962–1968’, Leonardo 49 (2014), 55–65, at 55. 24 A. Michael Noll, ‘Human or Machine: A Subjective Comparison of Piet Mondrian’s Composition with Lines (1917) and a Computer-Generated Picture’, The Psychological Record 16 (1966), 1–10. There were at least three real Noll MSs generated for Noll’s initial study, and it seems that additional ones were generated in subsequent refinements of the programme. (The best published illustration of the first three real Noll MSs can be found in Herbert W. Franke, Computer Graphics Computer Art (London: Phaidon, 1971), Figure 91.) And obviously the programme itself logically enables an indefinitely large population of Noll MSs that were not made—actually printed out. In Noll’s initial publication, probably not inadvertently, he illustrated the one real Noll MS that to his eyes, to others’ eyes, and to my own eyes looks more like the prototype RM Composition in Line than the other real Noll MSs that were also generated and printed. Copyrighted by Noll in 1965 (after controversy with the US Patent Office), in Noll’s publication it was captioned Computer Composition with Lines, dated ‘1964’, and attributed to ‘the author in association with an IBM 7094 digital computer and a General Dynamics SC-4020 microfilm plotter’ (Noll, ‘Human or Machine’, Figure 3). For the context and history of Noll’s remarkable investigations, see Grant Taylor, When the Machine Made Art: The Troubled History of Computer Art (New York: Bloomsbury Academic, 2014) and Zabet Patterson, Peripheral Vision: Bell Labs, the S-C 4020, and the Origins of Computer Art (Cambridge, MA: MIT Press, 2015). 25 Ironically, a number of projects subsequent to Noll have assumed that Noll’s results showed not only that a computer programme can generate ‘Mondrians’ (MSs) resembling RMs or even indiscernible from RMs. They also might have shown that in making his RMs, Mondrian himself had used a rule-system of some kind, such as a ‘shape grammar’ or a ‘genetic algorithm’. (For examples, see discussion in Allen W. Wolach and Maureen A. McHale, ‘Line Spacing in Mondrian Paintings and Computer-Generated Modifications’, Journal of General Psychology 132 (2005), 281–291.) But these investigations go against what Mondrian himself actually said about his work and against what we know about the way in which he actually painted. Mondrian’s working methods have been explored in detail in a masterful exhibition of Mondrian’s late work by the art historian and curator Harry Cooper and the conservator Ron Spronk, who among other things closely analysed the pentimenti in RMs as evidence of the painter’s juggling of placements and properties of lines, shapes and colour planes (Harry Cooper and Roy Spronk (eds), Mondrian: The Transatlantic Paintings (Cambridge, MA: Harvard University Art Museum, 2001)). Of course, none of this in itself is to say that a rule-system could not produce MSs closely resembling RMs. And none of it is to say that a beholder of the Noll MS might think that it was done by rule. 26 Alan Turing, ‘Computing Machinery and Intelligence’, Mind 59 (1950), 433–460. 27 Schapiro, ‘Mondrian’, 253. 28 Ibid.; Meyer Schapiro, ‘On the Humanity of Abstract Painting’ (1960), in his Modern Art, 231. 29 Nancy J. Troy, The Afterlife of Piet Mondrian (Chicago: University of Chicago Press, 2013), 4 (my italics); for Mondrian’s impact on American artists, see Nancy J. Troy, Mondrian and Neo-Plasticism in America (New Haven, CT: Yale University Art Museum, 1979). 30 See Troy, The Afterlife of Piet Mondrian, 169–228. 31 Mihaly Csikszentmihalyi and Rick E. Robinson, The Art of Seeing: An Interpretation of the Aesthetic Encounter (Malibu, CA: J. Paul Getty Museum, 1990), 112. 32 A brief review of the literature is provided by Lauring, ‘Visual Art’, 140–142. I adopt the interpretation of the results proposed by John Onians, European Art: A Neuroarthistory (New Haven, CT: Yale University Press, 2017), 7. 33 Mondrian, Natural Reality and Abstract Reality, 89–92. © British Society of Aesthetics 2018. Published by Oxford University Press on behalf of the British Society of Aesthetics. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Binding and Unbinding the Mondrian Stimulus
JF - The British Journal of Aesthetics
DO - 10.1093/aesthj/ayy038
DA - 2018-12-18
UR - https://www.deepdyve.com/lp/oxford-university-press/binding-and-unbinding-the-mondrian-stimulus-74AFXvo06H
SP - 449
VL - 58
IS - 4
DP - DeepDyve
ER -