229 229

Making There: Methods to Uncover Egocentric Experience in a Dialogic of Natural Places

Nicola J Bidwell James Cook University

Discipline of IT (Cairns) Australia

nic@it.jcu.edu.au

David Browning James Cook University

Discipline of IT (Townsville) Australia

david.browning@jcu.edu.au

ABSTRACT We reflect on two methods that explore personal experi-ence of natural places within a human-centred framework for design. The methods attend to the meanings people make of their experiences in natural places and emotional and intellectual sense of being in a place. We use Egocen-tric Point-of-View video to explore transformations of immediate sensory transactions in natural places and Na-ture Probes to elicit affective qualities of experiences. We apply McCarthy and Wright’s dialogical approach (2005) to uncover relationships between place and self and dis-cuss insights emerging from our methods with respect to belonging to a community through its natural landscape.

Author Keywords Nature, place, embodiment, egocentric video, probes

ACM Classification Keywords H5.m. Information interfaces and presentation (e.g., HCI): Miscellaneous.

INTRODUCTION We seek to elucidate personal experience of natural places to ensure a human-centred approach to technolo-gies for usage contexts related to these experiences. Con-ceptions of place during design processes tend to obscure people’s emotional and intellectual sense of being in a place. Discussion of ubiquitous computing echoes with Weiser’s (1991) exhortation to pursue qualities of interac-tions that are “as refreshing as taking a walk in the woods”. Yet the field emphasizes a utilitarian view of space as equipment and urbanism towards people’s con-nection with place. Certain practical and socio-cultural “infrastructures” (Dourish & Bell, 2006), often linked to micro-managing time and space, are very overt in urban spatiality. We propose this may limit abstracting space in design and that considering people’s situated experiences of natural places may increase sensitivity to sensory and affective phenomena. Here, we describe formative en-deavours to harness a dialogical approach (McCarthy & Wright, 2005) to uncover and depict relationships be-tween place and self in the sense people make of their experiences of natural places.

We are creating a design framework for technologies to enrich community connectivity in tropical Queensland

where the natural landscape is a distinctive regional char-acteristic. We aim to leverage qualities of experiences in nature that contribute to a person’s sense of themselves within a community. In the north it is not only indigenous Australians for whom natural landscape has cultural sig-nificance (e.g. Dourish & Bell, 2006). While, the bush might be a mythical place unadulterated by human activ-ity in suburban imagery (e.g. Tuan, 1974) it dominates how people in “the regions” describe their “life-style”. The landscape is both worked, for agriculture, mining and tourism, and the primary milieu for leisure and self-expression. Currently, we have an undefined design vi-sion of “digital patchworks” which will be compatible with people’s affective and informational experiences of natural places. This is driven by various local socio-economic and ecological needs arising from factors such as population transience and over-visitation of certain nature sites. Whatever digital patchworks may be (e.g. a situated display system or mobile guide) to enable people to “feel in place they need to engage them at the level of their personhood” (McCarthy & Wright, 2005). So, inte-gral to the evolving design we hermeneutically refine our propositions (e.g. Strauss & Corbin, 1990) on the mean-ings of natural places and our methods to collect and ana-lyse the data that shapes these meanings.

McCarthy and Wright’s (2005) dialogics refers to experi-ence in which a place’s identity emerges from spatially-embedded interactions. It is a metaphor which distin-guishes a person’s immediate experience of a place from the meaning of that experience. People’s enacted (i.e. pre-ontological) experience of place is embodied in interac-tions that do not separate mind, body and environment (Dreyfus, 1997). Theories of action (e.g. Suchman, 1987) refer to this in designing technologies for urban places, often supplemented by spatial frameworks (such as Alex-ander’s or Lynch’s). However, a place’s meaning as a construct separate from self emerges as a person reflects upon the immediate experience (i.e. ontologically) and then, personally and communally, appropriates this mean-ing. Thus, focusing on activity does not explicitly refer to the relationships that transform spatial settings to place. Undoubtedly, designs to support learning (Rogers, et al, 2002), navigating (Abderhalden & Krug, 2003) and communicating (Axup & Viller, 2005) in natural settings are helped by theories of action. However, abstracting a natural place as a space (Dourish, 2006b) affording play-ing, relaxing and engaging with ecological and socio-cultural history excludes many aspects of the meaning of the experience. People’s immediate experience and inter-actions with nature are multi-sensory and corporeally replete. These intertwine with emotional, aesthetic, cul-tural, reflective, ideal, and material dimensions of experi-

OZCHI 2006, November 20-24, 2006, Sydney, Australia. Copyright the author(s) and CHISIG Additional copies are available at the ACM Digital Library (http://portal.acm.org/dl.cfm) or ordered from the CHISIG secretary (secretary@chisig.org) OZCHI 2006 Proceedings ISBN: 1-59593-545-2

230 230

ence. McCarthy and Wright’s (2005) dialogical approach to technology accentuates place’s plurality and relation-ships with an individual’s sense of their self.

We are interested in ways to capture the immediate expe-rience of a place and its transformation as meaning emerges through the experience. The value of gathering data on momentary interactions in the specific setting in which people, think reason and act is unequivocal (Suchman, 1987). However, varied physical and social contexts accompanying movement thwart conventional methods to situate the researcher within a participant’s experience at different geographic locations (see: Hagen et al, 2005). Theories of perception of natural environ-ments are often derived from laboratory-based question-naires, rating scales and physiological indicators around images of natural environments (see: Sell et al 1984). These limit the immediate multi-sensory experience of nature place. In situ data on aesthetic preferences and landscape and environment assessment is often gathered by evaluating subjective responses as discrete events (e.g. surveys). However, this does not depict the contingent and ephemeral phenomena of “being in the world” (Drey-fus, 1997).

We use two complimentary, self-reporting methods to address the need for empirical data on personal experi-ence of specific natural settings (e.g. Peterson, 1974) and the dynamics of the embedded experience. Participants in both methods express qualities of place arising in per-sonal history and aesthetic and multi-sensory experience of nature sites. We analyse their immediate transactions with “spatial resources” (in: Dourish, 2006b) during ex-periences and the transformations of these experiences. To scope a range of interdependencies we avoid con-straining transactions between a person and the setting by specific activities (e.g. fishing) around a device (e.g. fish-ing rod). Of course, regardless of simplicity, any data instrument will inextricably mediate the meaning of places depicted. For example, a place in a return from a camera probe (Gaver et al, 1999) is characterized by, and reflexively tied, to the instruction for the photo and all proximal objects the photo incidentally records. So, we also attend to relations between our methods and the meanings participants find in their interactions with na-ture.

We report on trialing our methods in separate pilot stud-ies in different natural places. We first reflect upon spa-tio-temporal aspects of experiences of natural places in situ gathered by Egocentric Point-of-View (egoPOV) video and audio. Next, we summarise using Nature Probes to elicit tangible interactions and affective quali-ties of experiences. In describing and reflecting upon our results we refrain from referring to current or imagined technologies. While such speculation makes it easier to convey the relevance of our approach to design, it dis-tracts from the details of the complexly interwoven inter-actions in natural places that we seek to uncover. Instead, our discussion emphasises reconciling place and “person-hood” in a dialogical approach to ubiquitous computing design. We indicate paradoxical and temporal characteris-tics in people’s use of spatial infrastructure. We conclude by suggesting what this means to spatial abstractions of natural places aimed to support belonging to a commu-nity.

EXPLORING THE DESIGN OF EGOCENTRIC VIDEO Potentially, automatic egoPOV recording can depict both fleeting and temporally-ongoing experiences which con-tribute to creating meaning during spatially-embedded interaction. EgoPOV video is likely to support interpret-ing the situated experience, as similarity between speaker’s and listener’s eye-movements in viewing scenes promotes comprehension (Richardson & Dale, 2005). Various wearable cameras systems (e.g. mounted on heads or in spectacles) have been used to capture ac-tivities and spatial resources from an egoPOV while mo-bile. Often these depict proximal objects and people dur-ing activities; for example, as people work (Mark, et al. 2001) or make decisions in urban and natural settings (e.g. Omodei et al, 1998; Bentley et al., 2005). Mausner (2006) has extended the method to gather data on multi-sensory perceptual encounters of natural landscapes. She trailed behind groups of hikers while one, wearing a cam-era, talked about features “which caught their attention in a special way“ and especially “captivate[d]” the senses.

We explored using egoPOV to gather the detail and dy-namics of engagement and intersubjective construction of natural places. We focused on different participants’ ac-cumulated experience of places which, while not in their ordinary routines, have been visited regularly in the past. Natural places that hold memories enable uncovering ways that people recreate and augment place in managing space in subsequent visits (McCarthy & Wright, 2005). We extended this proposition by framing participants’ activity in accounts of the place in their personal history. We also co-located the researcher and participant to pro-voke articulating references to physical resources and capture how deictic communication is reflexively tailored with physical contingencies to construct meaning.

Egocentric Video Method

Capturing Video In Situ Four participants separately recorded egoPOV data in situ while traversing local nature sites and conversing with an accompanying researcher. We asked participants to imag-ine that the researcher was a non-local companion to whom they were showing what made the site personally significant. We explained that, while the researcher would contribute to the conversation, he would aim to be a recipient of information rather than an active interroga-tor.

Each walk lasted about 20 minutes. We selected two sites, of the seven nominated, based on proximity and familiarity with all participants. Alligator Creek, 28 km south of Townsville CBD, is part of the large Bowling Green National Park. We walked in an area of 2,500 m2, which includes a camping, and picnic area, swimming hole, sections of the creek and adjoining Cockatoo Creek and a path to a waterfall. It is popular with families to swim and picnic at weekends but is otherwise quiet and suited to walking and watching wildlife. The National Parks authority manages the area and facilities (e.g. car park, information booth, barbecues and tables). This con-trasts with the other area we walked which only had a bird watching hide some 500 m from a car park. This was in Townsville Town Common, a 23km2 nature reserve, bounded by a small airport, the Many Peaks Range and the Rowes Bay shoreline. The site is 6 km north of the

231 231

CBD and we accessed it near Bald Rock via a causeway across a saltpan, which is, aside from bird-watchers, a starting-point for walks skirting the Range to the beach.

Participants wore the first prototype of our head-mounted video camera and microphone system. This consisted of a standard surveillance bullet cam (dimensions: 87 mm x 26 mm; resolution: 350 TV lines; lens angle: 60° horizon-tally, 44° vertically), and small microphones. The camera signal was recorded by a NV-GS400 Panasonic digital video camera with a colour ccd sensor (resolution: 582 x 500). Participants’ speech and some ambient sound were picked-up using two unidirectional, electret, condenser microphones. The camera recorded both audio channels and video simultaneously to simplify synchronisation. The audio was amplified to line level by a powered adapter and a custom–built small pre-amplifier powered by 12volt batteries. The batteries, the camera and its 12volt DC supply were put in a backpack (weight: 3.5 kg) worn by the participant. The bullet cam and microphones were mounted on a lightweight cycle helmet to provide a comfortable, adjustable and stable platform. While we showed participants the hat before going to the site, we set it up in situ easily using Velcro patches to set attitude (pitch, yaw and roll angles). The camera’s ccd was 15 cm directly above eye level and we adjusted the bullet cam to align the video image vertically and centrally with the participant’s binocular field of view with gaze ahead. It subtended approximately 33% of the horizontal extent of the binocular visual field (Dorr, 2004) and was tilted slightly downwards so its view corresponded as closely as possible to the participant’s relaxed forward gaze.

Capturing Post-Situ Retrospectives One hour after visiting the site participants viewed their video and verbalized any additional information that came to mind. The egoPOV was displayed on a laptop computer in a university office and we recorded both par-ticipant and researcher using a tripod-mounted video. We asked participants to first view the video without audio to provoke their extending accounts rather than just ‘re-experiencing’ (e.g. Omodei, et al. 1998) the visit. Then they reviewed the material with the audio. We sought to de-emphasise any linearising effects on recall by encour-aging participants to pause the video at any time and sup-plement verbal explanations with illustrations.

Theme Extraction We transcribed all dialogue and open-coded video and audio for verbal content, visual content (e.g. spatial re-sources, gestures) and qualities (e.g. forward or lateral translatory movement) and ambient sounds. We axially extracted themes from the initial coding and used Hyper-RESEARCH (ResearchWare) to index and retrieve ex-amples. Both authors independently viewed and analysed data to validate findings of one who was in situ.

Egocentric Video Themes On average 19 minutes egoPOV was collected by each participant. Of that time, a mean of 64% of movement in the video was forward translations, from walking, and in the remainder there were lateral translations only, from head movements while the participant was standing. Only one person other than the researcher was captured in the

video which reflected the absence of people in situ. Fo-cussed vision is relatively narrow (Clark, 1997) and eye and head movements are often closely tied (e.g. Land, 1992) suggesting that the video approximated to partici-pants’ view. Timelines of video and the semantics of their verbal dialogue (e.g. Fig. 1) suggest that participants’ verbal themes are determined by what they view. Often a feature appears in the video followed by verbal thematic change and overt, reference to the feature. During walk-ing, an average of 25% of the time the video showed the ground. This depended on the relative unevenness of the terrain traversed. For example, the video captured by an older participant walking along a stony, unpaved road and a muddy track often showed ground as he looked down to ensure safe footing. While, another walking mainly on smooth tarred paths and wooden walkways showed ground only 10% of the time.

Axially coded themes related to: familiarity with the place; family activities and concerns; social relationships; affects; aesthetics; motives for visiting; and, the rhythm of a visit. We describe these by relating them to how they draw, and are drawn by, the physical site into the dia-logue. Objects participate in embodying meaning at a personal level in uniting self and place. For example, feelings of familiarity and wistfulness relate personal history to activities around specific landmarks and changes to infrastructure. We relate these themes to the use of egoPOV to enrich insight into in situ dialogue be-tween participant, researcher and spatial resources. We illustrate these with examples that are in close temporal proximity to assist understanding amidst a diverse data set. However, these are a small fraction of the many rep-resentative examples we could use.

Video Records Scanning & Brief Glances Different qualities of the views of objects aid interpreting what the participant is looking at. For example, when a participant says, “Never see much in the way of birds”, the video shows a scanning of the horizon as if on the look out for some. Momentary views suggest a partici-pant glances in the direction of objects in a way that would be difficult to determine and record using third person POV. For example, while at the car park, a par-ticipant talks about using a road to “go round to Shelley Beach, a few times”. The view in the video suggests he indicates the direction by glancing at a particular end of the Many Peaks range. Sometimes repeated glances into-nate or emphasise. For example, while a participant re-counts her children showing their parents parts of the Creek they had discovered, the video very briefly shows fruit on a tree which 3 seconds later she begins to talk about. The fruit remains in the participant’s verbalisation and in the video except for two quick, blurred views of the walkway. Hardly pausing, she returns to a previous topic relating to improved site infrastructure, mentioning the walkway and glancing at remnants of a path.

Video for Disambiguation & Richer Interpretation Capturing the setting on which the participant gazes en-riches interpretation. The comments below take on a quite different quality when accompanied by the tranquil scene of the almost still water of the swimming hole:

232 232

Figure 1. A 1.5 minute timeline of visual and verbal content captured egocentrically by one participant

Participant But this is, I mean this is just lovely here, and it's lovely at the moment because there are not many people here, but it can, it's pretty busy at times... Researcher I can imagine. Participant...on a Sunday... ...lots of noise. Researcher School holidays.… Participant Yeah. Researcher Whatever. Yeah, it's just beautiful. Participant Well this is the nicest time, because you can hear the cicadas and the birds in the background. You wouldn't hear it if there were lots of people, I suppose, would you? Now, it's gorgeous.

The video assists interpreting relations between the phys-ical site and the mutual, ongoing shaping of references by participant and researcher. For example, a participant talks about the water level in the creek and wildlife asso-ciated with the creek. As the video turns from the creek to a first bench, the creek, the walkway, a second bench, and, briefly, a sign (see: Fig 1a - f.), her remarks on ”tourist-friendly…” seem to refer to the site generally, rather than wildlife (e.g. “turtles”) or the immediate vi-cinity specifically. The video also indicated when both participant and researcher refer to the same physical cue. For example, when the participant says “...definitely, and obviously safer, with lots of danger signs” the video turns from the creek toward vegetation, a walkway and then back to the creek, suggesting the comment is about the creek, rather than other safety features (e.g. the walkway).

Indexicality & Space as Infrastructure The video indicates that gestural eye or head movements enhance interpretation. For example, over a period of a few minutes, six sets of hills are clearly visible before the view rests on some particular hills and the participant talks about “these hills”. His head orientation, shown by the video, clarifies which hills he refers to even though he never names or uses hand gestures towards them. This, reciprocally, shapes the interpretation of comments he makes later about a crumbling, disused road running along the foot of these hills. At one point the old access road is visible and at another point he distinguishes it

from the track that he is walking along. Interestingly, this participant had said that it was unacceptable to point with his hand and instead oriented or gestured with his head, such as in referring to the airport. However, for other par-ticipants the video captures larger hand gestures, such as pointing upstream in the direction of a creek or at large rocks which were salient to verbal comments.

Participants’ comments related to specific spatial con-texts, even when they did not name or describe salient referents. For example, when referring to increased water level after a recent rain a participant said “Well, the water just down there...” and pointed to a steep, vegetation-covered bank. The creek is not visible from that specific location and had not yet appeared in the video. At times such indexicality (Garfinkel, 1967) suggested participants assumed the researcher’s knowledge of landscape ecol-ogy or local geography. For example, one participant re-ferred specifically to the river and its catchment without naming them when neither was in view. Careful inspec-tion of the area later suggests the only cues would be slight differences in the relative density of trees and mar-ginally disparate light and temperature. Thus, the shared context assumed by the indexicality does not reside in a built physical infrastructure (Dourish & Bell, 2006) such as a relationship between the creek and park facilities.

Participation of Space in Recall Many instances of temporal relations between video and conversation suggested that information in the landscape prompted participants’ recollections. For example, as the creek comes into view, a participant shifts from talking about the walkway, on which she is walking, to recalling the creek’s water levels on previous visits. “I mean there was nothing like this here, there was no raised platform; you just had to walk on the creek bed and go down. And, of course, we've been at times when the creek's been full and when the creek's been empty.” Another participant, when looking at the Range, describes how he used to cycle along the foothills and an incident to dispose of a friend’s girlfriend’s “pretty objectionable cat” with a .303 rifle. He said the cat ran around the inside of the vehicle, on the “terrible trip” to the site, as if it knew “It was on

233 233

its ninth life.”

Spatial resources are appropriated to retain information. It appears that participants associated a memory with a par-ticular feature and that feature, reciprocally, became a cue for recall. For example, when describing activities during past visits a participant refers ’out there’ to some promi-nent rocks on the far side of the Creek (Fig. 1h). She draws on the reference 20 seconds later, when she has a clearer view of the same rocks. As she points at the rock with her hand she weaves them into previously described themes (e.g. walking across the creek, the children’s ac-tivities).

Spatial resources may flexibly extend a verbal theme or provoke recall of a similar feature. For example, a par-ticipant referred to a path when returning to a theme about the site in the past. Simultaneously, the concrete structures of a picnic table besides a path appeared behind some foliage in the video (Fig. 1o). Much later it became clear that this path is not the part of the ‘old path’ (Fig. 1p); however, the participant did not correct herself and seemed committed to using it to describe the site histori-cally. Paradoxically, she had previously noted scrambling down banks and over pebbles because there was no path or walkway. This may relate to recalling changes to the site that accumulate for visits over several years. It also shows that spatial resources as retainers for memories and indexicality are dynamic.

Visual spatial resources seem effective in cueing recall of less intellectual aspects, such as affects. A participant said quite wistfully that whilst she enjoyed seeing the site when it was so peaceful, she would especially enjoy visit-ing it when families were there to remind her of visits with her own children.

Temporal Relations in Spatial Dialogics The participant’s verbal theme interchanges occur fre-quently yet coherently. For example, in a period of about 2.5 minutes, mostly while moving forward a participant talks about: Her children’s history of visiting Cockatoo Creek; A sighting of a specific seasonal vegetation and her family’s past interaction with it; The site as it was before and is now; The state of the creek; Wildlife seen; Improved tourism friendliness; The state of the creek and its seasonality; Her children’s activities relating to the creek and wildlife; The state of the creek (again); Her children’s wildlife collecting activities; and, The site as it was before and is now (again).

Temporal relations between visual and verbal content suggested that objects in the landscape contributed to extending verbal threads. Momentarily, verbal themes significantly differ from the visual content. They retain a sequential fluency while the participant encounters the landscape visually along several different themes. How-ever, verbal fragments appear to result from the visual interaction. For example, when the creek appears in the video for the first time, the participant’s dialogue smoothly but rapidly transitions from referring abstractly to ‘the creek’ to more direct, descriptions relating to ‘it’. As she walks along the walkway in view of the creek she remarks on the creek’s wild-life and depth relative to the walkway, Then the video briefly shows two benches (Fig. 1a & c) and the participant returns to noting changes to

make the site more “tourist-friendly. A glance at a sign (Fig. 1e & f) with information about local wildlife may have provoked her to refer to safety and danger signs in adding to her comments on tourist oriented enhance-ments. Potentially, it may also relate to two warning signs (crocodiles, slippery rocks, strong currents) that she passed some minutes before but did not, at that time, re-mark on.

Temporal relations between the participant’s comments and the video during review indicate that visual resources extend verbal threads in multiple ways and cue recall of a ‘holistic-intuitive’ (Omodei et al., 1998). For example, a participant notices a main information board when view-ing video without audio (Fig. 2). She talks specifically about the board in the context of changes to the site and its lack of facilities in bygone times provokes musing on how the family learnt about it. She assumes someone else told them and then recollects that they met people and found out about Alligator Creek via parents associated with her children’s schools.

Figure 2. A participant notes community relationships a)

when seeing an information board in her b) egoPOV video

Future of EgoPOV Recording The range of data collected in EgoPOV suggests benefits from the basic method. However, concurrently exploring the design space in evolving propositions enables us to respond to participants’ comments on methods. One felt that, provided there were few people about, she might feel freer and less constrained to 'entertain' the researcher by recording her EgoPOV alone. She also said that she might articulate different details if families had been present or she had been accompanied by her husband who shared her past visits and would have supplemented recollec-tions.

We are currently trialing a prototype panoramic camera system and wide-screen video review. All participants remarked that their 3-D in situ field of view seemed com-pressed in the video. One participant was struck by the size of a prominent landmark in the video during the post-activity review. He looked towards Castle Hill frequently in situ without commenting on its size; but remarked that it appeared to be 'not much more than a pimple' in the video (Fig 3a). It seemed he had a particular view of it as a very large rocky structure so afterwards we asked him to take a photo of it as he might have it in his 'mind's eye' (Fig 3b). He regularly sees it from a variety of angles and sent us a view as he normally sees it whenever he leaves his residence at its foot. Wide-screen video seems likely to address this. Additionally, we collect data in the pe-ripheral field and more ambient qualities using 3 simul-taneous cameras to encompass the horizontal extent of the field of view. “Excessive footage” of the ground (Maus-ner, 2006) represents the embodied experience of a place

234 234

with uneven terrain. However, to capture the view that a participant might glance at even when their head remains lowered we have added a 4th camera on top of the hat that extends the vertical sub-tense.

Figure 3. Views of a hill a) in the egocentric video; and b) in

a participant’s mind’s eye

EXPLORING THE DESIGN OF NATURE PROBES We were inspired to pilot nature probes to provoke re-search propositions by issues arising in EgoPOV record-ing. Participants capturing EgoPOV spent more time on tracks and walkways, rather than trundle across or sit on plant growth or rocks. They were not as bodily involved with nature’s materiality as they recalled in their personal histories of the site. Thus, Participant’s use of the built physical infrastructure obscured other possible tangible, visual-kinesthetic interactions. The nature probe aims to increase the scope for depicting qualities of the multi-sensory experience and specific, self-reported meanings. Probes accumulate diverse qualities of places across geo-graphically dispersed participants. For example, video clips can gather remote data on ties between people, ac-tivities and localities (Isomursu, et al. 2004). Probes offer glimpses into spatial resources for verbalised and non-verbalised interactions and experience; such as a photo of a summit used as a rendezvous by friends might appear in a rock-climber’s scrap-book (Rouncefield et al 2005). The nature probes simply involved participants returning an object which conveyed a significant feeling from a nature experience.

Probe Methods We explored nature probes in two different contexts that arose, serendipitously, in pre-arranged activities. In both, participants returned from walks in nature with an object that in some way expressed emotion. This could be any-thing at all about their feelings before or during their walk and did not have to be connected to the environment or the activity. We asked school of IT staff members to bring back an object from a walk on a Friday afternoon. The participants, two females and two males including academics and administrators, had arranged the informal activity to mark semester’s end. Their 2-hour walk from the university’s rural Cairns campus followed a path up a rain-forested hill to a dilapidated building (locally known as “the castle”). Upon their return, they individually showed the researcher their object and explained what it was about it that represented their feelings. Over two con-secutive weekends one of the authors participated in an Arts Workshop with 9 others in the Daintree World Heri-tage Rainforest. Participants were women, aged between 20 and 70 who were interested in meeting others in the area and learning new skills (e.g. painting, pottery, crea-tive writing). At the beginning of both days each woman spent 15 minutes finding an object in the natural setting; then, sitting in a circle, in turn they explained to the oth-ers how the object signified their feelings.

Probe Returns & Themes The activities influenced probe returns and participants’ descriptions in several ways. Participants returned from the Friday Walk with a fragment of glass, a breezeblock brick, a flat stone, and a rusted iron rod; and in the Arts Workshop, pods, earth, a branch and a variety of leaves, creepers and flowers.

Tangible Aesthetics & Tactile Interactions Participants tended to speak appreciatively of the objects even if the associated feelings were not positive. They indicated particular, often subtle, lustres, colours, shapes and textures. For example, participants noted with par-ticular pleasure the form factors of the spiraled iron rod and a creeper. Participants’ tactile engagement with ob-jects during their explanations depended on the context of the discussion and their proximity to each other or the researcher effected. With only a researcher present, par-ticipants often placed the object on a desk or handed it to the researcher and pointed towards it. In contrast, when seated around other participants they generally held the object in one hand and moved the other hand and fingers around its features (Fig. 4).

Relationship of the Probe to Land, Affect & Self All participants in the Friday Walk explicitly referred their objects to specific aspects of their destination. The fragment of glass was referred to as “finding something from the ground” and the stone conveyed sadness that the hill should “belong to everybody” but now the formerly publicly-accessible property was private. Two partici-pants returned rubble of the building; the brick indicated its state as “unfinished” and “not pristine” in a similar manner to the iron which, skeletally, “sticks out from the top all around the building”. Several participants in the Arts Workshop also described the context of their return, for example saying a large fern “stands out against the other foliage” and demonstrating the way a creeper hung from a tree. When the researcher asked Friday Walk par-ticipants why they had returned a man-made object all indicated that their destination had particularly character-ized their walk.

Affective responses to physical objects have both poten-tial for expressing individuality and sharing understand-ing. Over both days of the Arts Workshop four people used Poinciana pods to describe quite different emotional states (Fig. 4). For example, one described being full of potential, while others described “cleaning all the dirt that has built up”, or being “exhausted”. These related to different features of pods, such as torsion in a twisted husk, smooth lustre inside the husk and the presence or absence of seeds. Some interactions referred to personal-ity, for example a participant compared the unobtrusive slender leaf to herself “during in the week” but its daz-zling, coloured tip to herself “at a party”. Despite such individuality the objects seem to support empathizing with associations. For example, when the researcher had not been in situ participants’ references to objects were evocative of the “spooky” aspect of the castle suggested by the iron rod or a “discovery when you go on a trip” by the tiny fragment of glass.

235 235

Figure 4. Similar objects returned in the nature probe elicited different affective qualities

Reflections & Future of Nature Probes The probe returns depicted fragmentary snapshots of af-fective aspects of momentary interactions with nature. However, later description transforms these affects, onto-logically, and accumulates with subsequent and present experience. Our current methods use more specific probes to gain insight into how meaning might be transformed. We request individual natural objects with particular characteristics (e.g. “look pleasing”, “smell hopeful“, “is relaxing to touch”, “awaken curiosity”; “denotes free-dom”, “has stability”). We use probes to supplement the enhanced EgoPOV method rather than to access data re-motely as is usual probes. This reveals participants’ tac-tile interactions with objects and also satisfies ethical and legal constraints requiring us to return all collected ob-jects to traditional lands and national parks. We have cre-ated a lightweight “natural treasure” bag for use at a par-ticular site containing labelled, opaque zip-lock bags and tubs of different types. Its form factor encourages main-tained physical contact with collected objects when the participant reflects upon it.

DISCUSSION Our pilot studies yield some analytic considerations in depicting place in natural settings. The main value of this, as for any embodied phenomena, is not measurable by a precise set of implications for design or translation into a final product (Dourish, 2006a). Rather, its significance for creativity and intervention is as an account of how meaning is constructed in experiences of natural places and how participants create that understanding across space and time. Appropriating a natural setting as a place is a physically, temporally and socially unfolding experi-ence. Spatial resources are powerful prompts for memo-ries: they are customised to retain information and cue recall. Visual features in situ provoke a memory that, by thematic association (e.g. changes to a place over time), is exchanged with other features. Thus, in natural places past experiences and immediate interactions with physical features combine and, together, condition the place’s on-going augmentation. So, temporal boundaries around fea-tures are elastic (McCarthy & Wright, 2006) in a way comparable to a “trajectory projection” (Strauss, 1993). For example, momentary, affective interactions with na-ture are accumulated when reflecting on probe returns and the meaning in this cumulative reflection retrospec-tively interacts with the in situ experience. Such a spatial dialogic occurs repeatedly across scale, temporally and socially. Indexicality reveals the participant involving the researcher in the evolving “interaction trajectory” (Fitz-patrick et al., 1998) drawing upon different spatial re-sources and infrastructures. The mutual management does not cease after data collection as the researcher continues to draw on spatial features to disambiguate and interpret

data.

We need to observe the consequence of the ambiguities of place and their relation to dichotomies of self in ubiqui-tous computing design. People’s pragmatic, fictive stabil-ity of a place relates to their own identity. They treat a place as finalized despite significant physical transitions and their own continuous, reciprocal shaping of past, cur-rent and future interactions. A person recognizes a famil-iar place through their personal engagement with spatial resources. If they encounter it extra-ordinarily (e.g. on a weekday not a weekend) it is less indicative of past inter-actions. It is less indicative of past self. People know themselves only by their “inherence in time and in the world” that is, “only in ambiguity” (Merleau-Ponty, 1945), such that a facet of human-centred spatial abstrac-tion of place is its provisionality.

We conclude by briefly indicating paradoxes in depicting natural places that will not diminish the “space of possi-bility” (McCarthy & Wright, 2005) for self and commu-nity. People use spatial resources in unfamiliar places to access and supplement their existing knowledge of many different infrastructures. Using these spatial resources depends on the person’s momentary disposition. People may recognize the meanings that others embed in re-sources intellectually without the same emotional reso-nance. For example, the feelings expressed by Nature Probe objects may be highly individual even if their ver-bal comments are evocative of the in situ experience. It is often tempting to marginalise these differences as an aes-thetic that is intractably inaccessible or non-generalisable for the purposes of design. However, an abstraction of space (Dourish, 2006b) that neglects such personhood has limited scope in helping people “feel in place”.

For “digital patchworks” of natural places we are alert to the consequences of the fictive constancy of place for community integration through natural places. Consider the changing emphasis on different cultural infrastruc-tures in landscape as a newcomer is initiated into, and relates their self to, a community. Obvious physical infra-structure, such as the Parks Authority’s tracks, may reas-sure or disappoint the newcomer when contrasted with their imaginings of the bush. But as place and self trans-form such infrastructure recedes in awareness and the sense of being in the place relates to spatial resources tuned by other infrastructures. For example, a sense of belonging may be embedded into rocks, metaphorically “worn” smooth, by remembering children using them to slide into pools. Thus, our methods have revealed to us a temporal characteristic in a person’s use of the layers of spatial infrastructure that contribute to a sense of being in natural places. We believe our challenge is discover how to design a spatial abstraction that enables a user to move fluidly through this bricolage.

ACKNOWLEDGMENTS We thank all participants and especially, Ian Atkinson (DART & James Cook University) for continued support and John McCarthy (University of Cork) for helpful comments.

REFERENCES Abderhalden, W. & Krug, K. (2003) Visitor monitoring

in the Swiss National Park: towards appropriate infor-mation for the wireless consumer Proc. 10th Int. Conf.

236 236

on IT, Travel & Tourism, 2003, Helsinki. Axup, J., Viller, S.: (2005) Augmenting Travel Gossip:

Contextual Interviews and Social Network Analysis of Backpackers. Proc. Australasian Conf. on Computer-Human Interaction (OzCHI05) Canberra

Bentley, T., Johnston, L., & Baggo, K.V. (2005) Evalu-

ation Using Cued-recall Debrief to Elicit Information about a User's Affective Experiences. Proc. Austral-asian Conf. on Computer-Human Interaction (Oz-CHI05) Canberra

Clark, A. (1997) Being There. Putting Brain, Body, & the World Together Again. MIT Press.

Dreyfus, H. (1991) Being-in-the-World. A Commentary on Heidegger's Being and Time MIT Press, Cambridge, Mas.

Dorr, M. (2004). Effects of Gaze-Contingent Stimuli on Eye Movements Diplomarbeit.: www.inb.uniluebeck-.de/~dorr/dipl-thesis/Dorr04.pdf. Retrieved 12/06/06

Dourish, P. (2006a). Implications for Design. Proc. ACM Conf. Human Factors in Computing Systems CHI 2006, Montreal

Dourish, P. (2006b). Re-Spacing Place: Place and Space Ten Years On. Proc. ACM Conf. Computer-Supported Co-operative Work CSCW 2006 Banff, Alberta

Dourish P, & Bell, G (2006) The Infrastructure of Ex-perience and the Experience of Infrastructure: Meaning and Structure in Everyday Encounters with Space. En-vironment & Planning B: Planning & Design (in press)

Fitzpatrick, G., Kaplan, S., & Mansfield, T. (1998) Ap-plying the Locales Framework to Understanding & De-signing. Proc. Australasian Conf. on Computer-Human Interaction (OzCHI) 122

Garfinkel, H (1967) Studies in Ethnomethodology, En-glewood Cliffs: Prentice-Hall

Gaver, B, Dunne T & Pacenti E (1999) Cultural Probes: Design Interactions 6(1) 21-29

Hagen P, Robertson T, Kan, M & Sadler K (2005) Emer-ging Research Methods for Understanding Mobile Technology Use Proc. Australasian Conf. on Com-puter-Human Interaction (OzCHI05) Canberra,

Isomursu, Kuutti, Vanamo (2004) Experience Clip: Method for user participation an evaluation of mobile concepts Proc. 8th Conf. on Participatory Design, Toronto

Land, MF (1992) Predictable eye hand coordination dur-ing driving Nature 359:218-320

Mark, G., Christensen, U., & Shafae, M. (2001) A Meth-odology Using a Microcamera for Studying Mobile IT Usage and Person Mobility. In CHI 2001. Seattle, Washington,.

McCarthy J & Wright P (2005) Technology in Place: Dialogics of Technology, Place & Self Proc. Xth Int. Conf. Human-Computer Interaction (INTERACT05) IFIP & Springer-Verlag: Lecture Notes in Computer Science

Merleau-Ponty, M (1945) Phenomenology of Perception 1958 Translation, Routledge Classics London & NY 2002

Mausner, C. (2006) Capturing the Hike Experience on Video: a New Methodology for Studying Human Transactions with Nature. Proc. North-eastern Recrea-tion Res. Symp. April 10-12 2005;, NY.

Omodei, M., Wearing, A. & McLennan, J.(1998) Head-mounted video recording: A methodology for studying naturalistic decision making. In: R Flin, E Salas, M, Strub & Martin, L. (Eds), Decision Making Under Stress: Emerging Themes and Applications. Aldershot : Ashgate

Peterson, GL (1974) Evaluating the quality of wilderness environment: Congruence between perception and aspi-ration Environment & Behaviour (692) 169-193

Richardson D C & Dale R (2005) Looking to Understand: The Coupling Between Speaker’s and Listener’s Eye Movement and its Relationship to Discourse Compre-hension Cog. Sci 29 1045-1060

Rogers, Y., Price, S., Harris, E., Phelps, T., Underwood, M., & Wilde, D., (2002) Learning through digitally-augmented physical experiences: reflections on the am-bient wood project. Equator IRC tech. report.

Rouncefield, M., Saslis-Laougdakis, G, Chevrst, K, Dix A, Fitton, D, Kray, C & Graham, C (2005) Probing De-signs: Designing Probes Appropriate Methods for De-sign in Complex & Sensitive Settings Workshop at Australasian Conf. on Computer-Human Interaction (OzCHI05) Canberra

Sell, J. L., Taylor, J. G., & Zube, E. H. (1984) Toward a Theoretical Framework for Landscape Perception. In Environmental Perception and Behavior: An Inventory and Prospect, 61 - 83.

Strauss, A. (1993) The Continual Permutations of Action. NY: Aldine de Gruyter

Strauss, A., & Corbin, J. (1990) Basics of Qualitative Research. Newbury Park, London, New Delhi: Sage.

Suchman, L. (1987) Plans & Situated Actions: The prob-lem of human-machine communication. CUP

Tuan, Y-F (1974) Topophilia: A Study of Environmental Perception Attitudes & Values Prentice Hall,

Weiser. M (1991) The computer for the 21st century. Sci. Am, 265(3):94-104, Sep.