Two Ambiguities in Object-Oriented Aesthetic Interpretation

The aesthetic theory of Graham Harman’s Object-Oriented Ontology (OOO) revolves around the concept of allure, a nonliteral experience of an object’s displacement from its qualities that draws attention to a deeper reality. But applying allure to aesthetic interpretation is hampered in two ways. First, OOO necessarily moves between the constrained viewpoint of experience and a more global perspective. Yet mixing these “inside” and “outside” views can risk ambiguity. Second, the phenomenological difference between the parts and qualities of an object must be clarified before Harman’s model of wholes and parts can be incorporated into OOO aesthetics. Addressing these two ambiguities will make it possible to further develop OOO’s resources for aesthetic commentary. For instance, one conclusion is that allure itself has two varieties: a tension between the object and its qualities (“allusion”) and a tension between the whole and its parts (“collusion”). These options parallel Harman’s twofold critique of reductionism. Another conclusion is that the literal needs an explanation within the framework of OOO insofar as it is a genuine feature of experience.


Introduction
Graham Harman memorably compares the metaphysical problem of causation, the main theme of Object-Oriented Ontology (OOO, "triple-O"), with aesthetic experience.¹ The analogy makes otherwise abstract ideas about metaphysical problems more accessible, and it counters the prevailing tendency to ally the intellectual underpinnings of philosophy with the achievements of literalist enterprises such as science, law, and economic theory.² But the comparison is also relevant for the arts. In Art and Objects, for instance, Harman connects general OOO insights about form and context with close readings of several influential writers on art. The philosophy thus suggests new ways to think about fundamental themes in art criticism.³ And an object-oriented aesthetic theory might someday take up more detailed topics such as natural beauty, art history, animal cognition, aesthetics and X (commerce, politics, gender, food, shelter, or clothing), taxonomies of media, and outsider art. But this essay will take a step on the intermediate path of establishing OOO's nascent resources for analyzing and interpreting artworks.⁴ In its approach to both causation and aesthetic interpretation, OOO usually focuses on the tense link between an object and its qualities. This develops out of the idea that reality is unavailable to direct contact. When a real object, a "beholder," encounters something else, it touches nothing more than a distorted image called a sensual object. The question then becomes how indirect influence occurs. Harman avoids appealing to any one kind of go-between, whether that be God, primary qualities, minds that synthesize information, or matter. Instead he looks to the way relations are structured. Normally, a sensual object simply is as it literally seems to the beholder; the current state of the world hums along in banal fashion. But when an encountered object enters friction with its qualities, the beholder's bond with it also changes. In terms of everyday human experience, we take most things at face value and are intermittently surprised or moved by others. This interplay between banality and tension, also known as the literal and non-literal, is the basis of aesthetic experience for OOO.
From the above, Harman infers two kinds of objects (real/sensual) and two kinds of qualities (real/sensual), adding up to four possible pairings of an object with qualities. But he sees only one pair as especially important for OOO aesthetics. This is allure, the sporadic tension of a real object with sensual qualities (RO-SQ), in which the default bond between the encountered thing and its features breaks down and a hidden reality becomes at issue.⁵ Out of deference to a tradition of discourse leading back to Kant and beyond, Harman also uses the term beauty: "While the meaning of beauty is often left hopelessly vague, OOO defines it very precisely: as an RO-SQ split, the opening of a fissure between a real thing and its sensual qualities." He adds that "this particular rift is the central topic of all OOO writing on art."⁶ Allure may be a central topic but applying it to writing on art is tricky. New questions arise from trying to integrate the complexities of aesthetic experience with a consistent grasp of the theory.
To see the difficulty I am referring to, we need to only notice that Harman's book about the style of horror author H. P. Lovecraft associates not just allure but all four object/quality pairs with Lovecraft's artistry.⁷ Similarly, in a recent overview of his philosophy, Harman links one tension to the architecture of Tom Wiscombe (RO-SQ) and another to that of Mark Foster Gage (SO-SQ), though he praises both as equals and leaves intriguing wiggle room on Gage's side.⁸ The latter pair of sensual object with sensual qualities, which I will call simulation (SO-SQ), returns with competing meanings in Art and Objects.⁹ First, it refers to a daydreamy quasi-beauty described by Kant in his third Critique, according to which the hypnotic "material" pleasure of an object, such as the dance of fire in a fireplace, differs from proper  3 Harman, Art and Objects. 4 Examples of this intermediate path include Harman, Weird Realism; Lunning, "The Crush;" Wiscombe, "Discreteness." The three categories listed here should imply a broad continuum of interests. I have passed over commentary along the lines of Arsenjuk, "Impossibility," which denies that an object-oriented film theory is possible, or Behar, "The Other Woman" and Fowles, "The Perfect Subject," both of which aim to exclude OOO from their fields on the basis of critiques of power. Leaving aside the relative merits of such arguments, the present discussion is about cultivating technical resources rather than appraising the larger philosophical groundwork. 5 Harman, Guerrilla Metaphysics, 141-4. Harman has never abandoned this basic definition of allure or its overall significance for his philosophy. Still, as we will see in Section 2, he has substantially honed the concept in a way that takes account of what I will call "inside" and "outside" viewpoints. Another important post-Guerrilla Metaphysics distinction between RO-SQ and SO-RQ is beyond the scope of this essay.  Harman,Weird Realism, I address this ambiguity with respect to SO-SQ below. 9 "Simulation" is Harman's term in Bells and Whistles, 68 (no. 59). It replaces "confrontation" in Quadruple Object. For reasons that would take me too far afield, I will avoid Harman's other term for SO-SQ, "time." formal beauty.¹⁰ Kant is "right" to separate SO-SQ from beauty: "the constant arousal of a difference between the phenomenal object and its qualities is something very close to beauty without quite clearing the hurdle."¹¹ Later in the book, though, Harman associates the same fireplace example with lowinformation media that require us to "add our own reveries" in a sense meant to epitomize beauty.¹² The likely reference to allure here is difficult to square with his previous interpretation of the fireplace example. And in a third passage, Harman writes, "I do not see how any literal object (SO-SQ) can possibly count as art."¹³ Does that which cannot possibly count as art fall under the same banner as the almostbeautiful experience of Kantian charm? Finally, let us grant that literal objects are not art. But since literal content is an unavoidable aspect of all experiences, how could OOO incorporate it in interpretive analysis? In the passage just cited, at least, Harman sees no other route than SO-SQ.
Another challenge arises from the fact that we never encounter artworks in a unitary instant. How might OOO integrate this roving piecemeal experience into its aesthetic theory? OOO does after all have a unique model of wholes and parts. The idea is that every indirect bond, or attachment, forms a new real object that is irreducible to its own parts just as every object is irreducible to relations with it. Socrates grasping an urn is one thing. The urn and Socrates are others. As Harman explains, this independence of an object from its parts implies the object contains an endless descent of attachments and also suggests an entity can appear, persist, and perish due to relations on its interior without itself interacting with other thingshence, the cosmos has a lonely surface of isolated entities.¹⁴ Lastly, Harman argues that the basic unit for the study of ethics, aesthetics, and metaphysics is the attachment-entity itself, not the beholder or the sensual object on the interior of that attachment.¹⁵ Despite this fascinating theory, the relationship between wholes and parts within the aesthetic universe of banality and tension has not received the same scrutiny in OOO as the object/quality pair.¹⁶ I will focus below on two core ambiguities in OOO aesthetics. They will be addressed together since they operate in tandem when interpreting artworks. The first is a question of perspective. Are we considering (1) the overarching model of objects, qualities, the real, and the relational in OOO, or are we looking at (2) the dynamics of tension and its lack on the interior of relational experiencethe beholder's viewpointwhich is where OOO locates the engine of causation?¹⁷ As I will show, this lens influences the theory so it must be closely monitored. The second ambiguity is a question of qualities versus parts. When are they separate categories? A strong handle on this distinction can clear the way for OOO to incorporate its model of wholes and parts in its aesthetic theory. This problem of qualities versus parts and the first problem of outside and inside views support several conclusions, including some that will not be  10 Kant, Critique of Judgment, 94-5. 11 Harman, Art and Objects, 39. 12 Ibid., 178. 13 Ibid., 91. 14 Harman, Quadruple Object, 111-3 (endless descent), 122 (surface of isolated objects). 15 Harman, Dante's Broken Hammer. 16 Some of the above account of wholes and parts is reflected in Timothy Morton's comments on mereology (Realist Magic, 42-9). But I should reiterate fn. 2: Morton's interest in aesthetics is for the sake of insight into causality. When he does briefly discuss the phenomenology of wholes and parts (e.g., 100-1), he identifies parts with qualities. Another candidate for comparison is the assemblage theory of Manuel DeLanda. However, his theory's partial intersection with OOO has little to do with aesthetic experience. Assemblage theory is "all about objective processes of assembly" rather than the structure of experience (DeLanda, New Philosophy of Society, 3). In OOO terms, DeLanda describes a genetic process by which "a loose collection" of objects becomes "a whole with emergent properties" (ibid., 48). Nevertheless, assemblage theory would become an important point of comparison if the concept of "collusion" described in this essay were extended to causation between real objects. 17 The two perspectives listed here might seem to be a variant of the classic subject/object dichotomy. But that verdict assumes, first, that the experiencing agent or "beholder" is necessarily human, and second, that the only alternative to the subject/object correlate is direct access to the real. By contrast, OOO claims any object can be a beholder and reality is indirectly accessible. The experiencing entity indirectly reaches the real when it couples with SO to form a new real object (an attachment). See the summary of wholes and parts, above. See also Harman, Object-Oriented Ontology, 87-9. On OOO and phenomenology, see Harman, Guerrilla Metaphysics. discussed here.¹⁸ I will just focus on those pertaining to simulation (SO-SQ). First, Harman sometimes identifies cases of SO-SQ in aesthetic experience, but he is actually referring to a unique variant of allure that has not yet entered the OOO canon. This variant, which I will call "collusion," makes it possible to apply Harman's negative critique of knowledge to the positive interpretation of aesthetic encounters. Second, literal content is a worthier topic for OOO than one might expect, given OOO's anti-literalist stanceand that is where we find the true meaning of SO-SQ.

First ambiguity: inside and outside vantage points
Food for thought: an object is both irreducible to interactions and composed of them. This peculiar feature challenges the OOO thinker to keep close track of two vantage points. One focuses on the elements of causation itself, while the other studies the dynamics of interactions from inside a beholder's experience. To see the workings and stakes of these outside and inside perspectives, consider how they come into play in the very groundwork of OOO, a critique of Martin Heidegger's analysis of equipment.¹⁹ The argument goes in two directions depending on whether Harman operates from outside or inside an experiential viewpoint. I will start with the outside view, the main subject of his first book, Tool-Being.²⁰ The rift between an observed or "broken" tool and everyday practical coping has a deeper ontological significance than Heidegger and many of his followers want it to have. It indicates a rift between presence and absence in any relation at all. One thing will always miss the mark of another's being, translating it into a distortion called a sensual object (SO). Hence, both human and non-human beholders fail to grasp the real as such. (This perceptual framing of interaction makes "beholder" a convenient shorthand for any real object that encounters something else.) As Harman writes, "To sit in a chair does not exhaust its reality any more than visual observation of the chair ever does," and "the same holds for inanimate entities, since the chair and floor distort one another no less than humans distort the chair."²¹ A simple statement sums up the thesis: any relation entails a link between a real object and a sensual object (RO-SO).²² Once more, the meaning of the Heideggerian broken tool from the outside view is that objects have two sides as reality (tool, RO) and relation (broken tool, SO). But Harman sees another kind of significance in this dichotomy when viewing it from the vantage point of relational experience. Tool and broken tool not only refer to reality and relation but also refer to two modes of relation. That is, even within a relational framework, tools can break. The sensual chair normally blends into the beholder's usual paradigm for chairs, leaving the status quo unbroken. But the chair might surprise us and so allude to the surplus of the chair-in-itself: "our attention is drawn to it, yet it is still a withdrawn enigma inaccessible to us, incommensurable with any possible relation we might have to it."²³ An inside view of the tool/broken tool dichotomy allows Harman to distinguish the banal chair from the alluring one. Again, the beholder usually encounters an "unbroken" literal chair (banality) except in special circumstances, such as when the wobbliness of the chair makes a hidden reality at stake (tension, RO-SQ). It should now be apparent that the outside and inside views involve very different frames of reference and lead to two distinct results at the heart of OOO.
The two vantage points have also supported Harman's recent realization about the "theatrical" basis of aesthetics. In early writings, Harman notes that for one of his philosophical heroes, José Ortega y  18 For example, the aesthetic status of allure's half-sibling, SO-RQ ("eidos" or "theory") is of great interest for future work. 19 See especially Heidegger, Being and Time, Sections 14-24. 20 Harman, Tool-Being. 21 Harman,Weird Realism, This gap between real and relational objects points to further conclusions. For instance, the causal influence of RO-SO is non-reciprocal by analogy with the beholder who is moved by a painting. And an attachment is distinct from its components, meaning that it forms a new irreducible object like a chemical compound. 23 Harman, "Materialism is Not the Solution," 108. Gasset, alluring contact (RO-SQ) seems to deliver the "lived" being of the object in the flesh. Yet reality withdraws from access, Harman argues, so it would be more accurate to claim allure makes the object seem to retreat into the shadows: "what it really does is make the visible seem withdrawn."²⁴ Since at least 2014, though, he has recognized a missing piece to this puzzle.²⁵ His early argument overlooks that the reality of an alluring tree for the beholder is never that of the genuine tree-in-itself. After all, an encountered object is always a sensual one. While an alluring tree does "become real" for the beholder, then, this experience of tension does not negate the basic fact that the beholder touches a sensual rather than a real object. Where, then, is the RO that pairs with SQ? (A taxonomic distinction between "imaginary" and "actual" trees would miss the point, since the question is about the source of reality in allure, not about the epistemic accuracy of SO.) Whether the encountered tree seems to unveil its lived being (Ortega) or seems to withdraw (early Harman), it just seems to do so. This strange situation leads Harman to his theatrical conclusion. A haunting chord progression by Radiohead or the antics of a movie mobster played by Joe Pesci may be fictions, but if we want to take seriously the central claim of OOO that they indirectly bring reality into play, then we must identify the location of that reality. And the only real entity present is myself, the beholder. Like a method actor, I theatrically live the musical passage or the film scene as the music or movie. In a sense, Ortega was right: allure does reach the real, but only by inaugurating the creation of a new object, a new attachment, through the involvement of the beholder.²⁶ This sketched evolution of the concept of allure can be encapsulated in terms of the inside/outside schema. Early Harman examines RO-SQ from the experiential or inside view, according to which the aesthetic object becomes real by withdrawing from immediate access. Recent Harman agrees with this claim but also realizes that it does not address the ostensible paradox at the heart of indirect causation. The paradox is only apparent from the outside view, a wide-angle lens that elides the experience of banality and tension on the interior of a relation. Namely, allure provides contact with the real, yet at the same time all encountered things are sensual objects. The "RO" in RO-SQ cannot be the real counterpart of SO because that would amount to the impossible claim that the beholder encounters a real object. A better explanation instead identifies RO as the entity that invests its own being in the encounter and thus enters the attachment-object as one of its parts. Though Harman does not explicitly refer to outside and inside views, he brings new precision to the concept of allure by attending to both.
It should be apparent now that differing outside and inside views will influence how we understand the four object/quality pairs in OOO. And as Harman has shown with the concept of allure, keeping track of these viewpoints is useful. I will only discuss simulation (SO-SQ) and allure (RO-SQ) below, but it might already be apparent that for another of the object/quality pairs, RO-RQ, there simply cannot be an inside view. Both sides are real, withheld from any beholder.²⁷ And I have just indicated that for at least one of the tensions with one real pole (RO-SQ), the lack of identity between outside and inside viewpoints has led to a mixture of complications and insights. Finally, the outside and inside views of simulation (SO-SQ) are most surprising. An outside view gives simulation full credit for upholding the irreducibility of the object to a literal list of qualities. Yet the distinction between banality and tension within experience (the inside view) suggests that something like "literal experience" is possible, and of the four object/quality pairs, only simulation fully embraces the banal level of appearances. It seems that SO-SQ embodies both the non-literal and the literal. 27 One might suppose the beholder "experiences" their own real qualities. But a causal event itself (outside view) should not be confused with the experience of object/quality tension inside such an event (inside view). As we have seen, any relational experience necessarily entails sensual translation by the beholder.

Second ambiguity: parts and qualities
What is the difference between the parts of an object and its qualities? This distinction may suggest new interpretive possibilities for OOO aesthetics. But before opening that door, let us see where the ambiguity lies. First, it is not just the problem of the one and the many since every object contains many parts and many qualities. Nor is it a foregone conclusion that parts qua objects differ in kind from qualities, at least not in an obvious way. Indeed, we might never find "qualities" when we turn our attention to them. As Harman puts it: One option is that the qualities adrift in the world are made of an entirely different stuff from objects, which would leave us with a dualistic cosmos of objects and properties, perhaps along the traditional lines of form and matter. Another option is that the qualities are themselves objects […] This would give us a world filled with a single genre of reality known as objects, unaccompanied by any second, foreign principle.²⁸ Harman takes up the latter option insofar as he agrees with the Husserlian tenet that every profile is the target of an object-giving act. When our attention shifts from a centaur to its features, from its brownness to brown itself, it shifts to other objects. "If we ask what these adumbrations of a centaur might be, we find that they are still not just raw sense data, but are composed of a number of familiar centaur-elements: human head with somber features, equestrian torso, miniature tail, a deliberate and purposeful gait."²⁹ The underlying point here is that there are no free-floating properties prior to objects. Qualities always belong to an object and are individualized by it. "Even if the exact hue of red in my apple can also be found in a nearby shirt or can of spraypaint, the colors will have a different feel in each of these cases, since they are bonded to the thing to which they belong."³⁰ But this idea of a "spiraling interplay of objects within objects" partly obscures the distinction between a thing's parts and qualities.³¹ After all, OOO does separate these categories, so the question becomes how. How do the shifting adumbrations of a building differ from the various pieces encountered on the tour? It comes down to the way each one relates to the larger object. A parta door in a hallway or phrase in a symphonyis already a separate thing, a container for its own features and ingredients. But the creepy mood of the hallway or the bright atmosphere of the music will form a smooth continuum with other profiles of its object through that host's unifying mediation. The object with its qualities is not unlike a "latent space" in machine learning, where each face, font, or other such choice produced by an AI is just one interpolation on a smooth manifold of possibilities.³² Surely a texture, atmosphere, mood, or style can seem to get unglued as part of the single genre of reality known as objects. Yet this strikes me as a potentially misleading claim for the simple reason that it cannot happen in the framework of OOO. A quality with an ostensible life of its own does not wander from one object to the next like a lonely samurai. Rather, the traveling "quality" is already an object for the beholder, one that connotes object-to-object similarity. This exact point is one of Noah Roderick's memorable contributions to OOO.³³ When a designer or chef transfers what we might think of as a "quality" from one chair or dish to another, they are already dealing with a color-object, a flavor-object, and novelty emerges from its interaction with other things.

Allusion and collusion
The relationship between qualities and parts becomes less ambiguousand more relevant to OOO aestheticswhen viewed from the outside and inside perspectives. From the outside view, the two categories differ in how they relate to the object. First, any given experience has an "object position" and a "qualities position" where the object holds a qualitative continuum.³⁴ A beige chair in a certain room at a certain time is just a passing glimpse of a spectrum tied to the one chair. Now, as I have noted, qualities themselves do not travel like everyday language implies. With the statement, "My chair is the same beige as your toilet," we might assume beige moves between chair and toilet. But for OOO, the statement is not grounded in an encounter with self-same "quality particles" that the beholder then objectifies. Rather, it is grounded in objects that individualize their beige-ness. New things emerge, too: a chair-plus-toilet entity, a "beige" unit of knowledge. This is pertinent to the realm of parts because separate objects are also at stake here. The distinction between a qualitative continuum and discrete parts is enough for us to add "whole position" and "parts position" to regular discourse on OOO aesthetics.
But what about the inside view? How does the experience of qualities compare with that of parts? Since the beholder always connects qualities to an object, the more precise question is whether parts can have a banal or alluring link to their whole. I have previously noted that banal experience holds no surprises in reserve. As far as qualities are concerned, the object simply is its outward presentation. But banality could also refer to a dearth of experienced tension between objects and whatever unifies them. It might seem that one way to characterize this missing tension would be the holistic system. Yet what is the difference between parts and qualities in such a system? What distinguishes the seamless merging of parts into their functions from a seamless continuum of qualities? Nothing, it seems: a holistically integrated doorknob is not different than an adumbration of the door qua its opening mechanism. To satisfy the outside view's distinction between parts and qualities, even a banal part must have boundaries separating it from other things. A better characterization of "banal parts," then, is that they stand alongside one another in bland contiguity (SO-SO).³⁵ Such an environment would lack emergence. Insofar as the whole and its parts are both present, the banal whole is a neutral container for its contents. The jug and its water, the story and its characters, the musical phrase and its gestures: they flatten into an aggregate of neighbors. This form of literalism is a common appeal of the critic who insinuates that the artist, author, or musician failed to do something "on purpose." Our critic may be less interested in the psychological motivations of their target than in portraying the artwork as a list of haphazard coincidences that lack a unifying principle.
The next task after considering the banality of parts is to ask how their allure works. A comparison with qualities is once more a useful place to start. Harman notes that the alluring pairing of an object with its qualities (RO-SQ) "cannot be either too convincing or too unconvincing."³⁶ There must be some superficial plausibility that hooks the beholder to the unlikely object, but not so much plausibility as to render the experience banal. An alluring part/whole pair would work in a similar way. That is, the parts cannot be either too unified or too separate. The trivial kinship of discrete pieces hooks the beholder to the whole while their individual charms push whatever unifies them into the shadows as a real object. Just as we saw with RO-SQ, the outside view only recognizes sensual parts and a sensual whole in the beholder's relation to the object. Hence, the beholder must theatrically substitute for a missing RO. This is an important point because it means the structure of attachment for parts and wholes is similar to the structure of attachment for qualities and objects. Indeed, it would be convenient to house both forms of allure under "RO-SQ." But if we are aiming for accuracy, a better term for the whole/part tension might be RO-SP, where SP means "sensual part." Readers familiar with Harman's work may be reminded here of his critique of "duomining," the pair of methods that reduce an object to literal knowledge about its genetic components (undermining) and impact on its environment (overmining). When taken as the end point of cognition, both kinds of mining are hostile to the object that is "irreducible to its own pieces, and equally irreducible to its outward  34 Precedent for the use of "object position" and "qualities position" can be found in Harman,Dante's Broken Hammer, The use of "contiguity" for a collection of sensual objects that do not form a larger whole is consistent with Harman's use of the term in The Quadruple Object. For example, "sensual objects do not touch other sensual objects, but exist only as contiguous in a single experience that serves as their bridge" (75). 36 Harman,Dante's Broken Hammer,192. relations with other things."³⁷ It then stands to reason that there should be two forms of anti-mining, two ways for a real object to become at issue for a beholder, though to my knowledge no one has yet explained the distinction. The first kind of anti-mining, already familiar to OOO, is allusion, in which a withheld distinctiveness disrupts the object's identification with its current effects (RO-SQ). Let us dub the other kind of anti-mining collusion, where the individuality of an object's parts subverts their reduction to the unifying whole (RO-SP). I will not draw explicit parallels with causation itself, but notice that collusion has a certain symmetry, a point of shared (if hidden) contact among separate units, which further distinguishes the experience of coordinated pieces from allusive surprise.
5 Implication for the analysis of allure I have suggested that allure has two modes: the head-on experience of allusion (against overmining) and the bottom-up experience of collusion (against undermining). Either way, allure attaches a beholder to an object. It works by disrupting literal experience, such that the beholder fuses improbable features to the object or unites disparate pieces on behalf of the missing real ground. This difficult marriage of the object to its qualities or parts is key. Harman's exemplar for such an eventand a good match for allusion, the first mode of allureis metaphor. Take this one, commonly attributed to Groucho Marx: "A hospital bed is a parked taxi with the meter running." Hospital beds and parked taxis are only trivially similar to immobile and expensive containers for the human body (at least in America). But if the metaphor resonates, meaning that we accept the combination, the imperfect fit between a stay at the hospital and its taxi features turns the bed into a surplus reality (RO) on behalf of which the beholder enacts the metered car's grungy, frustrating, extortionate atmosphere (SQ).
What about collusion? Writing about the style of horror writer H. P. Lovecraft, Harman associates what he calls a "cubistic" technique with SO-SQ. This technique involves "sensual or non-hidden objects and their sensual qualities that pile up in disturbing profusion."³⁸ Notably, the cubistic technique illustrates Lovecraft's artistry. Unlike a banal bundle of qualities, "we sense that beyond its mass of qualities, there must be some vile underlying unit holding all these grisly features together."³⁹ In fact, the cubistic technique best fits the variant of allure I have called collusion, and for two reasons. First, the qualities position actually contains a group of discrete object-parts. Second, the object position may in one sense be visible like the hospital bed in the above metaphor, but its coordination of independently engaging parts makes it a hidden "underlying unit" from the inside view of the beholder.
Take as an example one of Lovecraft's well-known creature-characters, Brown Jenkin, a kind of witch's familiar.⁴⁰ Lovecraft describes Brown Jenkin as having the shape of a rat, an "evilly human" bearded face, sharp teeth, and hand-like paws. All of these are "juxtaposed qualities that do not exceed our cognitive grasp individually but that fail to come together jointly."⁴¹ Recall the ambiguity between parts and qualities. If qualities are juxtaposed in the way Harman describes, their lack of continuity with one another makes them objects or parts. Sure enough, the aesthetic success of this description follows precisely from the parts failing to come together while still somehow doing so, just as metaphor succeeds from a similar improbability. One might call Brown Jenkin a sensual object from the outside viewbut the same could be said about any metaphor, joke, painting, or symphony. From the perspective of aesthetic experience, Brown Jenkin becomes real through the strange collusion of his parts.
 37 Harman,Quadruple Object,47. 38 Harman,Weird Realism,34. 39 Ibid.,35. 40 Lovecraft,"The Dreams in the Witch House." 41 Harman,Weird Realism, The tension between an object and its parts raises another point of interest for OOO aesthetics. We rarely experience an artwork all at once. Rather, we shift between its constituent segments and their context. More to the point, different kinds and sizes of components play by different rules, and their meaningful relation with one anotheras with allure in generalis a special occurrence. The editing or camera angles in a single film scene do one kind of work, while the patterning of these elements over the course of the movie does another. As music scholars have recognized, phrase-level gestures by Mozart and his contemporaries have their own logic of formulaic convention and subversion, while the music's noteby-note counterpoint and large-scale formal construction have theirs. Indeed, the musicologist Robert Gjerdingen has made the object-oriented contention that the partial autonomy of differently scaled parts generalizes to all musical cognition.⁴² What I mean to suggest is that the part/whole tension does more than add a new mode of allure to OOO's interpretive framework (collusion, RO-SP). It also makes it possible to more vividly map the sequential drama of an aesthetic experience within any framework that identifies the familiar objects of that experience (its stereotypes, norms, customs, schemes, scripts, and formulas).
Specifically, a single recognizable element of an aesthetic experience can participate in any of four basic flavors of allure (Table 1). This element, a sensual object when taken alone, can be in the qualities position or object position of RO-SQ. Or it can occupy the whole position or parts position of RO-SP. Take allusion: an element in the qualities position is a kind of style that its host object imitates, like the figuration of a metaphor. In "the hospital bed is like a parked taxi," the taxi fills this slot. But this pair can be reversed into a new metaphor.⁴³ That is, the same element can occupy the object position and so participate in a new effect. In "a parked taxi is like a hospital bed," the taxi fuses with connotations of stillness, comfort, and healing. Now consider collusion. First, the taxi object as a whole can struggle to contain its discrete ingredients: "Witnesses said the taxi had a long body and the shape of a car, but that it had dentist's seats for chairs while its smell was that of a dingy bar and its halting gait an uninsured carnival ride." Finally, the taxi can become one of the elements in such a list: "The young giant's hospital room was fitted with well-wishing adornments and toys: a bundle of fern trees in a water tower, a fresh yellow taxi cab with the meter still running, stuffed humans in playful poses, a plastic wormhole generator that the Old Ones had assured him looked just like the genuine article." 6 A place for the literal Simulation (SO-SQ) is not only a source of uncertainty about part/whole dynamics in OOO but is also at the center of a tug of war over the literal. For Harman, the belief that an object can be paraphrased without loss amounts to dogmatism. "The dogmatist holds that truth is legible on the surface of the world."⁴⁴ Yet the literal remains an important part of the OOO model of causation. The beholder will encounter an interplay of literal banality and non-literal tension, and banality is the norm: "Usually an object does not seem very different from the sum of qualities it presents to us; this is the grain of truth in Hume's 'bundle' theory."⁴⁵ In short, then, literalism both misses the real and is genuinely embedded in experience. It needs an explanation within the framework of OOO, and simulation turns out to be the best candidate for the job. From the outside viewpoint, SO is already partly distinct from SQ. Indeed, one of the lessons Harman gathers from Edmund Husserl is that the object is not a synthesis of pre-existing qualities. "Instead, according to Husserl we encounter the intentional object directly in experience from the start."⁴⁶ The SO is thus irreducible to the content of its presentation. However, notice that this must be true of every relation, including banal ones. Just as qualities depend on their object, no sensual object persists without a specific silhouette. Even in its banal state, the object "must always be accompanied by a swirling patina of sensual qualities," as Harman writes in The Quadruple Object.⁴⁷ He does add in the same passage that SO-SQ is somehow amenable to more fission beyond this point.⁴⁸ But I do not see how that is possible without a real object or real qualities becoming at issue (RO-SQ and SO-RQ). A sensual object's banal endurance through its patina of features is the closest SO-SQ can get to tension. Or rather, from the inside view that focuses on the dynamics of experienced tension, the constant pairing of SO with SQ lacks tension. This unbroken continuum is what makes it possible to identify the object with its adumbrations in the manner of Hume's bundle of qualities.⁴⁹ Following up on my previous claim that there are four flavors of allure, I would make a similar interpretive distinction for the literal. We must first recognize that literal experience by definition falls easily into paraphrase. Such an experience leans on content and context through the object's strong ties to its parts or effects (or most likely both). And from "parts" and "effects," the two modes of literal experience should be apparent. One kind of literal element simply is its outward profile. In being continuous with itself, it approaches the status of a "quality," an adumbration of some other object. It fails to surprise as it fulfills its function with smooth efficiency. In the other kind of literal experience, there is no whole (SO-SO). There is just a group of contiguous elements beneath the false object, like the brain that supposedly reduces to a clump of neurons, the character who consists only of a sob story, or the building that is allegedly nothing but bricks and socioeconomic forces.

Conclusion
The OOO interpreter of artworks faces at least two major challenges. The first is to track what I have called inside and outside vantage points. For example, allure (RO-SQ) in one sense refers to the experience of a tense emanation of features from an inaccessible source (inside view) and in another sense refers to the formation of an attachment (outside view). The second challenge is to distinguish the qualities of encountered objects from their parts. Ultimately, these issues make it possible to conclude that aesthetic experiences previously classified as SO-SQ are actually variants of allure. Indeed, allure has not one but two modes: it counters overmining through object/quality allusion (RO-SQ) and it counters undermining through whole/part collusion (RO-SP). How do these forms of allure mix at different levels of experience? I am not certain, but as a step forward we have seen that any element in an alluring experience can occupy the object or qualities position of allusion as well as the whole or parts position of collusion. Yet in both cases the tension between the object-pole and the quality-pole requires a sort of fission between the two." Harman introduces his distinction between a banal version of SO-SQ ("time") and a tense SO-SQ ("confrontation") earlier in the same chapter (102-3). He again suggests that SO-SQ as "regular perception" differs from SO-SQ as "fission" in Chapter 9 (126). 49 A closely related claim concerns the paradigm, which in this context refers to the real qualities that influence how the beholder engages with other things. The paradigm sets limits, including the point at which the SO-SQ continuum breaks down.
The place of aesthetic honor once given to simulation (SO-SQ) now belongs to collusion. A better interpretation of SO-SQ, I have argued, is that it and contiguity (SO-SO) account for literal experience in OOO. The literal is continuous with its effects or contiguous with its pieces (or both). But the dependence of the literal on context also makes it as flexible as the environment around and within it is vast. Indeed, the mobility of the literal is one way to read Harman when he refers to "the revenge of the surface."⁵⁰ In Guerrilla Metaphysics, he writes that concrete sensual elements "are the glue of the world, the vicarious cause that holds reality together, the trade secret of the carpentry of things."⁵¹ As we have seen, he has gotten more specific in recent years, explaining that relations between such elements have three possible outcomes from the inside view. They can be too convincing ("a crow is like a magpie"), too unconvincing ("a hat is like a dolphin"), or alluring ("wine-dark sea").⁵² Comparing similar things risks a literal tally of their features, and implausibly grouping them risks a calcification of their boundaries. As for the third option of allure, notice the sheer banality of "sea" and "wine" alone. In this way, we can begin to see how the literal sets the rules and possibilities for all aesthetic experience.