Geography and the visual image

I Introduction

Images can make us cry, shock us, change our mindsets, and haunt our thoughts and dreams. Yet images also surround us at all times, unnoticed, banal and clichéd. They comprise a large part of the background to our day-to-day lives, informing (knowingly or not) our actions. Thus, images operate at different and changing affective, representational, material and ideological registers. As such, they cannot be studied in a single, prescriptive and unifying way or analysed discretely from their complex contexts. This paper forwards a hauntological approach to the study of visual images. The purpose is not to introduce a revolutionary new approach that supersedes all others, but to re-examine the image’s existing status in recent geographical work, bringing this to focus as a constitutive haunting. An approach is formulated that is general and specific, personal as well as public, posing the undecidable, in-between status of the image not as a problem but as a potential. It contributes new lines of thought and inquiry to the already expanding body of geographical work that is engaging with visual methods in light of the performative turn by stressing the co-constitutive registers of visual images.

The paper begins by examining the uncertain status of geography’s images, particularly through landscape studies and artworks, after which it introduces the figure of the ghost. It adopts Jacques Derrida’s (1994) hauntology together with W.J.T Mitchell’s (2005) notion of the image as dead and alive, to posit a hauntological approach which argues one is always necessarily haunted, and constitutively so, by the other. It engages with images as both representation and presentation, still and animate, dead and alive like the ghost.

With the ontological turn and its focus on material cultures and performance, studying ‘representations’ has become less fashionable. Indeed, non-representational theory suggests representational studies deaden the ‘liveliness’ of the world (Thrift and Dewsbury, 2000), and asks what more there is. In contrast, Castree and Macmillan (2004) call for a ‘finessed’ approach to representation that need not go beyond it. Though geographers are researching and disseminating through less traditional – and more visual – media, little focus is directed towards the broad category of the image itself: picture, photograph, advertising, hallucinations, memories, graffiti. While academic geography has a history entwined with pictorial representations, images continue to have an uncertain – dual, even – status in the discipline, as both evidentiary and deceptive, as lively and deadening. Given current concerns in cultural geography which finds representation to be in tension with precognition (Della Dora, 2011), as a problematic tackled by performance studies (Kraftl, 2006) and needing the antidote of practice (Cresswell, 2003), the paper focuses on the image’s constitutive haunting of its representational register by the rupture of non-signification, but no less so by the signifier, in the performative, the affective. It examines what images can do, their power, ‘agency’ and meaning, and our responses to them. It sets up this binary of representation verses non-representation to highlight their constitutive haunting, but of course the image is a far more slippery notion than this.

Images haunt between the visible and invisible, real and virtual, as material objects and abstract cognitive, embodied, subjective processes. Their seemingly disembodied and spectral nature is one reason they have been thought of as supernatural, as ghostly phantasmagoria (see Castle, 1988; Pile, 2005; Warner, 2006). Images, together with ghosts, are never truly present or absent, bringing what we take as fact (as real) into doubt and raising questions about visual ‘evidence’ and our ‘imagination’. In this paper, I focus on visual images, but I deliberately retain a broad definition of ‘image’ (over ‘picture’, for example) because it is more productive here, encompassing different effects and registers of visual experience as ‘complex assemblages of virtual, material and symbolic elements’ (W.J.T. Mitchell, 2005b: xiii).

Before beginning to discuss geography’s relationship with visual images, I ought to outline the reasons the paper takes this broad direction. Although I will give some specific examples, I largely remain theoretical. As such I hope to avoid a close focus on one type of visual image (such as ‘photograph’) which might eclipse the differing genres, uses, contexts and forms that comprise such images and forget that all visual media operate and work on us in slightly different ways. Instead, I discuss here the haunting which is common between media, and the hauntology I develop is applicable to them all (though the results will be necessarily different because of their differences). Thus the reader may feel haunted by noticeable absences such as discussion of film, cartography, GIS, geopolitical engagement with news media and videography, all current, pertinent ways geographers engage with visual images. Several methodological reviews have been published since the ideas in this paper were first developed, some explicitly hauntological and others resonating with the themes discussed here – see Mulvey (2005) and Donaldson-McHugh and Moore (2006) for hauntologies of film; Bauch (2010) on GIS; Cosgrove (2008) and Jacob (2006) on maps; Garrett (2010) and Simpson (2011) on videography; Hawkins (2010a) and Tolia-Kelly (2012) on geography and art. I ask the reader to bring their own specificity and comparisons from these discrete yet overlapping literatures to the general ideas of the paper. I cannot speak directly to all visual methods; the hauntological approach is, instead, a general theory of the specific, as described in more detail throughout.

II Geography and the visual image

Current geographic work relating to visual images appears as a continuation and critique of geography’s historical engagement with them. Geography is a visual discipline, ‘working with the comparative power of vision’ (Matless, 2003: 222), and with a past reliant on visual aids like maps, globes, models, slides and photographic illustrations. While geography has ‘always entailed making and interpreting images’ (Cosgrove, 2008: 15), images continue to be thought as either reliable or deceptive. Visual ‘evidence’ can lend legitimacy and weight to a speaker’s arguments (Rose, 2003a). There is also a sense that prosthetic visual technologies like film and photography can help us access events better than writing or personal observation. Yet visual power is simultaneously expressed as something to be wary of, serving historically as the prime tool of the white, western gaze of academia. Although geographic representations ‘retain an official legitimacy and an enhanced authority’ (Winchester et al., 2003: 6), this runs parallel with increased concern about the power relations, truth-effects and mimetic capacity of representations: ‘the belief that we should strive to produce as accurate a reflection of the world as possible’ (Duncan and Ley, 1993: 2). ‘Scientific’ observation and the idea of visual representations as transparent media or ‘unproblematic reflection of the world’ are questioned (Gregory, 1994: 75, in Crampton, 2001: 236).

As things that are created but appear to be a natural reflection of reality, visual images can be manipulated to produce or maintain certain ideologies and power relations. It follows that the ‘truth’ of the image must be sought through its deceptive surface to the ‘social practices’ beyond it (Cosgrove, 2008; Rose, 2007). Yet Cosgrove (2008: 4) argues that ‘the critical stance that today frames cultural geography’s relationship with pictorial images has itself tended to subvert their expressive authority’. The ‘expressive authority’ of the image to produce its own meanings, effects and realities separate from the intentions of an ‘author’ or from the viewers’ interpretation, to constitute rather than to reflect, distort, mystify and obscure real material relations is also considered. This has predominantly been through considering the image’s materiality (Della Dora, 2009; Tolia-Kelly, 2004; Vasudevan, 2007). However, a ‘distinction between the objects, materiality and biophysical “reality” of the world’ (W.J.T. Mitchell, 1980: 360) remains upheld, though they are differential and part of the same continuum.

The image’s ‘expressive authority’ might also be understood in terms of the role of the imagination: ‘Vision’s meaning incorporates imagination: the ability to create images in the mind’s eye, which exceed in various ways those registered on the retina of the physical eye by light from the external world’ (Cosgrove, 2008: 8). It is the capacity of imagination to create new images ‘that have not previously existed in the material world of their maker’ that give visual images unique emotional power, which Cosgrove notes ‘has always generated anxiety, prompting social control of their production and effects’, like censorship (Cosgrove, 2003: 253). Stephen Daniels (2011: 185) also stresses images’ creative agency when he suggests that the ‘technologies of image-making raise questions of the location and constitution of the imagination beyond its traditional place in the human mind, as pictures within our heads, to a place out there in the world’. It is impossible to talk about vision without talking about what we see and vice versa. Daniels refers to the close relationship between how we understand vision, visual representations and imaging, and their import to the discipline, when he talks about a ‘geographical imagination’. He argues that ‘Geography’s field of vision across its various traditions remains to be fully explored as common ground’ (p. 182). This paper focuses on the visual image as a haunting to usefully encompass its various elements – imaginings ‘out there in the world’ – to provide a set of conceptual ideas, as a starting point rather than a prescriptive toolkit. Before moving on to discuss haunting and hauntology, I concentrate on two ways in which visual images have been discussed in geography explicitly, through landscape and as artworks, to give a clearer sense of the way the theoretical ideas introduced so far developed.

The status of the visual image follows a similar trajectory in geographical literature across many media dependent on changing theoretical developments, outlined next with regards to theories of landscape and art. While rarely thought about in terms of the ‘image’ itself, they are always entangled in discourses of the visual and the representational. Put crudely, this trajectory can be conceived as the separating out of art and science as two different types of representation, the first more subjective, creative and imaginative and the second more objective, realistic and reliable (Cosgrove, 2008: 161; Moran, 2006: 673). Landscape, which forms the first half of the following review, has been interpreted variously as painting and scientific observation, as cultural or natural and real. This art-science distinction has also been noted in relation to cartographic design. J.B. Harley’s deconstruction of maps was especially influential for work on geography’s visual images, pointing out that they always involve flattening and abstraction, reifying and legitimating, producing a particular perspective (Harley, 1989: 7; Wainwright and Bryan, 2009: 154, 155), and particularly illustrating the capacity of representations to construct new realities through the redrawing of lines on a map.

The experiential, creative and participatory aspects of landscape and mapping have been of great interest to geographers too, through community or collaborative mapping and landscape-art projects. These offer more democratic representations, moving power from the elite to the public, studying process over product. The second notable trajectory then is a shift in attention and significance from the representational product as a finished, fixed form to the process and practice through which this is achieved, where visual image is ongoing, transitory and performative (Cosgrove, 2008: 2). The geographical literature on artworks will form a second brief review.

III Ghosts, hauntology and dead- and alive-ness

While there is no one singular notion of haunting, ghosts and spectrality are critical tropes across a number of disciplines. Spectres of the past rudely erupt everywhere into Modernity’s linear narratives, cities and technologies. Ghosts are understood variously as vengeful figures, as uneasy presences reminding us of social injustices, or as animating forces intruding into the present. Within geography this ghostliness inhabits spaces and ruptures practices within the city (Comaroff, 2007; Edensor, 2005; Pile, 2005). Ghosts also figure as voices of the marginalized, subaltern, indigenous (Cameron, 2008; Lloyd, 2005; Maddern, 2008), resisting, disrupting or haunting official commemorative places and dominant (western) narratives and modes of thought, making other voices, stories and interpretations heard, enabling different politics to come into being (Gordon, 1997; McEwan, 2007: 43).

Haunting is experienced as disorientating, unsettling, confrontational, chilling, spooky, or even mundane. It can surprise us, implicating us bodily, affectively and ethically (Derrida, 1994; Gordon, 1997) and it is performative in the sense that each haunting (if the ghost is doomed to incessantly return) is different, iterative, as it ‘enters into a new and distinctive assemblage’ (Holloway and Kneale, 2008: 300). The ghost, then, ‘makes things happen’: it transforms (p. 300). A hauntological position is open to this. It is one of deliberate indeterminacy, enforced hesitancy or uncertainty over presupposed givens and operations involving visibility and invisibility that constitute our reality, throwing ‘a condon sanitaire around that which we claim to be real, material and truthful’ (Dixon, 2007: 206). Like images, ghosts occupy ‘liminal’ states, transgressing or oscillating between thresholds, clear-cut categories and binary oppositions, blurring the distinction between natural and supernatural, objective or subjective, fact and fiction.

Derrida’s (1994) hauntology, by which, along with Freud (1959), much ghost-work is influenced, critiques traditional scholarship for its desire to ontologize and rationalize, and its refusal to deal with anything that exceeds the realms of ‘objective scientificity’ (Kaplan, 2008). He argues that scholars are not in a position to speak with spectres or let them speak because they draw sharp distinctions between the real and unreal, the living and the non-living, the being and non-being. Haunting brings this either/or logic into question. The ghost eludes full presence, truth, certainty, and therefore (perhaps quite obviously) most academics do not believe in ghosts, or in what Derrida (1994: 11) calls ‘a virtual space of spectrality’ because ‘scholars believe that looking is sufficient’. This space is not immediately available to their gaze. It is undecidable, troubling, creative, transformative, blurring reassuring oppositions. It asks us to ‘rethink the “there” as soon as we open our mouths’ (p. 176). Hauntology, then, does not replace other concepts but instead inhabits (or revisits) them, introducing ‘haunting into the very construction of a concept’ (p. 161). Haunting is deconstructive and deconstruction itself follows a logic of haunting where terms are contaminated and mutually constitutive of each other, understood through the trace, an absent-presence gesturing to other traces in infinite deferral (Derrida, 1994; Royle, 2000). Yet, while deconstruction has been critiqued for its textual basis, Derrida made clear hauntology is concerned with the ontic (haunting it). As art historian Keith Moxey (2008: 131) posits, ‘the “life” of the world, materially manifest, once exorcised in the name of readability and rationality, has returned to haunt us’.

The ghost requires thinking about visibility, presence and viewing positions, requiring ‘a particular kind of seeing’ (Pile, 2005: 139). Bearing in mind the long history that has intertwined vision, representation and hauntings, the visual image is a place where spectral motifs are a prime concern and a notion of haunting might provide a productive approach:

When visual culture tells stories, they are ghost stories … The ghost is one place among many from which to interpellate the networks of visibility that have constructed, destroyed and deconstructed the modern visual subject. (Mirzoeff, 2002: 239)

Mitchell (2005b) suggests humans create a ‘second nature’ composed of images which does not just reflect our values but creates and threatens them, meaning images appear to have a life of their own, as ‘phantasmatic, immaterial entities that, when incarnated in the world, seem to possess agency, aura, “a mind of their own”’ (p. 105). Identifying the paradox of the image as simultaneous dead- and alive-ness, it seems only natural Mitchell would use ghostly vocabulary here, echoing the projection of fear and desire Freud (1959) attributes to occult beliefs. He determines the history of images as one of iconoclasm (a type of exorcism in the wish to destroy images) and idolatry (belief in their ‘presence’, magical or divine).

Mitchell gives an animistic analogy suggesting images embody a vernacular vocabulary of vitality, agency and reproductive powers, and, like species, are able to mutate and vary in different specimens (media). The purpose of the metaphor (and the anthropomorphism of asking what images want) is to deliberately focus on the ‘wildness and nonsensical obduracy’ (Mitchell, 2005b: 2) of the image that ‘has legs’. We might ask whether an image flourishes, reproduces itself, thrives and circulates, and, if so, why it persists in having significance, its obduracy a haunting presence meaning something to an individual or society. Mitchell’s example is the way dinosaurs continue to capture our imaginations. Another might be the way Eduard Munch’s The Scream (Figures 1–3) or the Second World War ‘Keep Calm and Carry On’ poster have been reinterpreted and reworked in popular culture and different contexts, their meanings changing beyond original purpose and use, seemingly taking a life of their own. Munch’s painting is enigmatic precisely because of its polyvalence, evident from its initial reception. The painting flourishes and reproduces itself, the content morphing into an alien, a stressed office worker, a politician, often for comic effect. The material object itself is one of the most valuable paintings in the world, appreciating in worth and notoriety every time it is stolen or auctioned. The painting, perhaps because of its supplementary nature, lending itself to reinterpretation, obdures.

OPEN IN VIEWER

Figure 1. ‘The Alien’s Scream’, by Alan Gursz (reproduced with permission). Figure 2. ‘Scream’ cartoon (reproduced with permission). Figure 3. ‘Scream’ mask (reproduced with permission).

What does this mean for a theory equating image and ghost? Well, hauntologically, where one term inhabits another constitutively, Mitchell (2005b: 55) asserts that the ‘birth of an image cannot be separated from its deadness’. Rather than attribute too much life to the image, he calls them coevolutionary entities, ‘quasi life-forms (like viruses) that depend on a host organism (ourselves), and cannot reproduce themselves without human participation’ (p. 87). The visual image as ‘undead’ goes some way to explain its uncanniness in everyday language and gestures to its spectral and corporeal presence. As theorized, the uncanny overlaps with haunting very closely, and haunting involves aspects of the uncanny (see Freud, 1959; Hook, 2005; Royle, 2003). Mitchell asks us to think of images as animated from stasis; they are ‘pregnant with unfulfilled possibility’ conjured or irrupting in different moments and contexts (Gordon, 1997: 183). To think of images as only alive and autonomous or only dead, inert and powerless is to ignore this.

Mitchell (2005b: 10) advises we need to grasp both sides of the paradox of the image: that it is ‘powerful – but also weak; meaningful – but also meaningless’. He suggests the question of desire is ideally suited for this inquiry because it builds in at the outset a critical ambiguity. This is because asking what pictures want also raises the question of ‘what it is they lack, what they do not possess, what cannot be attributed to them’ (p. 10). The image is dynamic; it ‘wants to hold, to arrest, to mummify an image in stillness and slow time. Once it has achieved its desire, however, it is driven to move, to speak, to dissolve, to repeat itself’ (p. 72). This psychoanalytic inspired reasoning is an effort to allow images to speak for themselves and delay interpretation. Mitchell’s critical ambiguity is akin to the enforced hesitancy of haunting described earlier. In Mitchell’s theorization of pictures ‘not as sovereign subjects or disembodied spirits but as subalterns’, as ‘go-betweens’ (p. 46), the visual image haunts between arrest and movement, powerlessness and proliferation, as inert and transformative.

In a photograph dated 1912 (Figure 4), spiritualist medium Eva C is centre stage with a band of glowing light between her outstretched hands. She was repeatedly tested and photographed by scientist Albert von Schrenk-Notzing to capture proof of spirit life. With no prior knowledge of its context, one would be uncertain of what exactly is happening. The photograph features in an edited hardback of essays on Spirit Photography. The glossy pages have an ‘arty feel’, giving the images inside gravitas. There are more dramatic, disturbing images in the book of table-lifting, mediums painfully producing ectoplasm from various orifices, and bizarre, exotic-looking cut-out spirit guides. This image is stuffy and still, melancholy and genteel in comparison. The girl is plain and her pose subdued. It is unique in that the mysterious band of light looks like electricity, but also because it features the scientist himself, who is to one side in the foreground, peripheral and slightly blurred. I spent time sifting through spirit photographs on the web, in books and in the archive and how I felt about them and what I wanted from them changed with each material encounter and context of looking.

OPEN IN VIEWER

Figure 4.

Eva C and Schrenk-Notzing (reproduced with permission from Institut für Grenzgebeite der Psychologie and Psychohygiene).

The photograph invites us to look through the way the medium’s face is lowered, and the way the scientist’s gaze directs our own. A gendered way of looking can be read into this scenario as we adopt the male gaze directed at feminine submissiveness. Our view of Eva C’s demeanour might alter when we know that she was actually in a trance-like state that was more conducive for spirits to enter the material world through her. Mediums like Eva C were considered objects of scientific investigation, an apparatus little different to the camera itself (Blum, 2007). Though scientific evidence (or lack of), the photograph is theatrical and classically composed, which feeds into its affective resonance: it has an aesthetically pleasing quality because of this. The sepia gives a romantic tinge to the seemingly naïve notion of capturing spirits ‘on film’, but such photographs persist in being curiosities, having an emotive power of their own. But what does it mean to ask what the image wants? Mitchell suggests this is akin to asking what it lacks. Without viewers it lacks values: it is only meaningful and significant through its relationship with us and other images. This ethical relationship requires critical ambiguity and delayed interpretation. We might first question how an image affects us, how it speaks to us, examine its ‘expressive authority’, before we fix what it means.

How does it affect me? There is a textural quality to the print, a sumptuous depth in the contrasting sepia tones and faded softness of fabric. The darkness is weighty and moody – literally weighing me down, as though I could sink into the depth behind the figures. Yet the glowing aura streaked across the centre adds energy. This starkly contrasting detail captures our attention and imprints most sharply on the eye, but our glance is less settled than this and is simultaneously directed towards the light through the scientist’s and then the medium’s gaze, shuttling back and forth in a dynamic exchange. The image has harnessed movement, not only in the blurred motion of Eva C’s hands but in the tautness of the right-hand curtain’s pleats and the defined tension of the muscles in Schrenk-Notzing’s hand. These details gesture to the image’s desire to arrest movement and its drive to move, speak, repeat.

As both dated anomalies and pertinent reminders of the history of ideas relating to materialism and spiritualism, clearly the image and others like it have ‘legs’. Through examining the differing manifestations of spirit on camera we find that ‘the ghost is nothing without you’ (Gordon, 1997: 179), but that they can speak to us a great deal about the periods during which these photographs were produced. The camera provided a powerful medium on which to project belief, curiosity or pleasure, and the image mutated accordingly. There was an increase in popularity of spirit photographs following war when grief was pervasive (Matheson, 2006) and they have endured numerous resurgences in interest. Most recently, the phenomenon resulted in new exhibitions and shows like Living TV’s Most Haunted (Antix Productions, 2002–2010).

This brief hauntological discussion is not meant to be a comprehensive example of the approach, but it does begin with the singularity of the image itself and is attentive, hesitant. We can frame this photograph through a feminist critique of the gazes therein, consider the theatricality and staging of the composition and what each signifier means (what is that on her head?), and ask at an ontological level how this photograph affects us and how it circulates materially. In reality, it is difficult to do this in isolation as these approaches slide in and out of each other. It is presentation and representation. With each viewing or haunting the photograph mutates, transforms, performing as part of an assemblage of signification, material objects, affects, multisensory elements and context.

IV Afterimage: hauntological implications

Images and visual experience have a history written through with haunting qualities. Haunting, as a discourse and as a modern condition, is likewise bound up in the visible and the representational. Through focusing on Derrida’s constitutive haunting and Mitchell’s paradoxical or simultaneous dead- and alive-ness, of fixity and movement, reproduction and mutability or plurality, a hauntological approach to images begins to take shape. It stresses that ‘whether moving or still, the image is never stable, fixed, or truly arrested, but rather always in between movement and rest, and always capable of affecting or being affected’ (Rio, 2005: 73). The approach lends a hauntological aspect to existing frameworks for visual analysis, and attempts to ‘refine and complicate our estimate of their power and the way it works’ (p. 73). A hauntological approach takes into account the capacity of the image to take us by surprise, to resist interpretation, to appear to mean nothing (naturalized) and deceive us, examining our ability to invest images with meaning.

So what are the implications for scholars taking a hauntological approach, writing and encountering images in this way? I suggest three central possibilities or hesitancies to take forward as critical to the hauntological approach to visual images. They are intrinsically interlinked: the potentiality of the image; its constitutive haunting; its dependency. Through these, the paper contributes a nuanced understanding of images that speaks back to wider debates about representation, vision and the geographical imagination, and explores common ground.

First, it begins with the image, its specificity and (potential) haunting. The pictorial image draws a bounding line, capable of fixing and stabilizing, but within it is the potential for movement as images ‘leap and pulsate with life’ (Rio, 2005: 74). Moments of haunting bring this point of tension, this potentiality, to the fore, experienced through encounters with material forms. Potential is expressed through performativity, obduracy and their capacity to trouble and disrupt. Experiencing the ‘energies, attunements, arrangements and intensities of differing texture, temporality, velocity and spatiality, that act on bodies, are produced through bodies and transmitted by bodies’ (Lorimer, 2008: 552) is said to be out of the grasp of representation or prior to representational meaning. But representations have affects and images act on bodies and are produced through them. ‘Everyday’ representations such as the way people routinely represent, even if ‘they rarely pause to consider the complexities of the act they are engaged in’ (Castree and Macmillan, 2004: 476), contribute to a ‘performative milieu’. Visual images are ‘lively’, mediating, creating and transforming reality. They ‘take their place alongside a series of other performative practices that conjure the world into being again and again’ (Rycroft, 2005: 355). They perform alongside but are performative in themselves, having affective, synaesthetic and bodily registers. This more compositional and mutually constitutive understanding of representations and performative practices is one way to consider visual images. It is sympathetic to the critique that thinking about ‘pure, blank spaces of social encounter’, seemingly divorced from the overall performative milieu of such encounter, risks erasing subject positions associated with existing representational conceptions of power, identity and agency (Tolia-Kelly, 2006: 215, in Lorimer, 2008: 553). Future work would examine how meaning emerges from relations between, or assemblages of, representational and non-representational registers.

Significantly in haunting literatures, disruptive moments of haunting are possibly opened up; they are a making visible of what is already there, giving pause to rethink what we see unquestionably. To transfer this directly to a hauntology of the image would be to recognize that haunting is not only in images that have emotional power or affective agency, but is latent within every image, residing in what Tolia-Kelly (2012: 136) calls the ‘visual stuff that always already surrounds people’. These have different affects and resonances. While some graffiti are passed daily unnoticed on the commute, some populate a ‘spectral cast’ when encountered in a derelict building (Edensor, 2005). The beyond, the in-between and the duality of re-presentation – medium-mediation, presentation-representation, object-subject, individual-social, referent-signifier, material-immaterial – is the space for haunting: ethical and transformative, even political. Yusoff (2010: 77) calls this space ‘where things are made, both materially and semiotically’ aesthetics, and argues it is part of the practice of politics and a space that ‘configures the realm of what is possible in that politics’. This space is ‘the real arena of dramatic confrontation, things come into being, incorporeal, unexpected and unpredictable’ (Rycroft, 2005: 355). Its potentiality is born out of this constitutive haunting.

Second, the images’ constitutive haunting can be analytically productive. Keith Moxey (2008) forwards the stance that considering the ‘interior’ terms of an image – its capacity to affect us, its aesthetic appeal, its status as presentation – should be a valuable addition giving power and complexity to current understandings and interpretative agendas in visual studies. In practice, he finds this relation is examined through a spectrum of interpretative positions in which the iconic/material and the representational/heuristic approaches slide in and out of one another. ‘Ontological’ studies tend to favour the materiality or ‘presence’ of the image over its semiotic, discursive or ‘readable’ elements, whereas one necessarily haunts the other as a noticeable absence: ‘the texts themselves suggest the ideological “supplements” that haunt them’ (p. 133). Perhaps the most famous study of the affective haunting of the image is Roland Barthes’ Camera Lucida (1993) in which he equates being photographed with ‘becoming spectre’ and defines the non-signifying element of the photograph, the punctum, as the cause of his haunting. In his eulogy to Barthes, Derrida re-articulates this haunting as co-constitutive. The punctum haunts the studium (language, representation, universality) because it lends itself to metonymy, allowing itself to be drawn into a network of substitutions: ‘This singularity that is nowhere in the field mobilizes everything everywhere; it pluralizes itself’ (Derrida, 2007: 286). Visual images, then, operate on different experiential registers where the signifier becomes an intensity, a trigger point for movement (O’Sullivan, 2006: 20).

Finally, the image’s dependency and the viewers’ responsibility create an ethical relation, because even seeing ‘makes us responsible in all kinds of ways’ (Yusoff, 2010: 89). Mitchell’s proposed method is ‘pragmatic, localized, heterogenous and improvisatory’ (Mitchell, 2005b: 419) whether this is interpreted as auto-ethnography, participatory or collaborative projects, post-phenomenology or any number of possible responses. He insists scholars should be responsible and conscious that they are contributing to representations in the form of theory, and forwards use of visual images themselves as a mode of theory. This is pertinent as geographers seek alternative, visual methods for research and dissemination. Geographers are responding to the suggestion that ‘we might use visual methods to unpack visuality’ (Crang, 2003: 242), with increasing interest in using photography, film and the visual arts as a method of research and dissemination. Geographers are presenting their work in more creative ways to foster different types of knowledge from those accessible through traditionally ‘representational’ modes like writing, by producing photo-diaries (Bissell, 2009; Watts, 2008), photo-essays (Wylie, 2006b) and cartoon strips (Yusoff, 2007), creating visual narratives through GIS technologies (Kwan, 2008) and graphic mapping (Barnes, 2007). The academic journal is perhaps not the most appropriate and/or desirable format and, while Garrett (2010) notes the ‘reluctance of journal editors to “go digital”’, more experimental approaches are being accommodated by journals online, while research projects more frequently encompass an exhibition/visual display element. It seems important that geographers continue to ‘finesse’ (Castree and Macmillan, 2004) understandings of the visual images being created. Videographers, for example, are advised to be wary of treating video material as ‘a representation of [the videographer’s] experience of reality’ rather than as ‘visual facts’ (Pink, 2007: 112 and 88, in Garrett, 2010: 534). Reflexivity and continued self-critique is what is demanded by the image.

The hauntological approach responds to the quasi- or parasitic status of the image – ‘the ghost is nothing without you’ (Gordon, 1997: 179) – with an attentive type of looking, which questions our assumed truths about images and habitual modes of looking, haunting between conceptual motifs, looking for occlusions and exclusions, traces and gazes, ‘fuzzy borders’ and ‘aporia’ (Kaplan, 2008), elevating the personal or the subjective. Adopting lessons from the scholar who is attentive to ghosts, we might assume a position of hesitancy, undecidability and ‘conscious ambivalence’ (W.J.T. Mitchell, 2005b) to the value of images. Establishing cross-disciplinary links between visual studies and geography, this paper provides a nuanced approach to encountering images.