Museums and the Web 1999

Best of the Web

Archives & Museum Informatics
2008 Murray Ave.,
Suite D
Pittsburgh, PA
15217 USA

Join our Mailing List.

Published: March 1999.


Exploiting the Potential of 3D Navigable Virtual Exhibition Spaces

Cristina Cerulli, University of Florence, Italy

Museum Spatial organisation

An attempt will be made to define the historical and cultural context in which the development of virtual exhibition space needs to be placed by looking at how gallery and museum spaces changed throughout the years from the sixteenth century's galleries to the virtual gallery on the Web.

Physical Museums

Traditionally the design of a viewing sequence that combined circulation paths with exhibition spaces, has always been the main issue for the curators of a Museum or an Art gallery. The spatial organisation of Museums is also strongly informed by the ways in which collectors, including the Museums themselves as such, are organised and in what way do they organise the space to display their collection. In his essay Museums without walls Malraux [Malraux, 1954] defines the museum, with and without walls, as a spatial relation expression of the ordering of the social as well as the ordering of the works of Art.

Throughout the centuries the Museum spatial layout has evolved reflecting the socio-cultural context of the time. In the Kunstkammer or Cabinet of Curiosity, in the late sixteenth and seventeenth century, the principle behind the arrangement of the early collections was wonder and attractive decoration. Different kinds of artefacts were displayed within the same exhibition space so the visitor was able to float freely within the exhibition space and to make his own connections between artefacts creating personal narratives.

The enlightenment brought ordering, organisation and separation and one of the aspects of the eighteenth century collecting and displaying becomes the tendency to separate the types of items. The classification system became strictly logical based on scientific observation and informed by concordance and order, both key characteristic of the language syntax.

Also most of the neo-classical Museums spaces of the eighteenth and nineteenth century were organised on rigid grid plans that would reflect the taxonomic organisation of knowledge. The circulation system was forced from a grand entrance hall through symmetrically arranged galleries. In the British Museum, opened in 1759, the galleries appeared as a symmetrical enfilade of rooms with a central isle for the circulation and display cases on the sides strictly arranged on the grid plan by taxonomy.

An important element of novelty in the determined spatial layout was introduced by Von Klenze in the Münich Alte Pinakothek (1826-36) where the traditional large interconnected galleries opened on one side into a series of small ones, also interconnected, and on the other side into a long corridor, the loggia. This improved circulation system allowed the visitor to navigate within the space in any sequence.

The task of the curators within such spatial layouts was to structure the exhibits into meaningful sequences and the ordering criterion would vary according to the nature of the displayed objects. A rigid systematic organisation of the material, though, was only possible if the collections were relatively static. Due to the dynamic growth of the collections Museums started to need more flexible spaces to host also temporary exhibit and the ordering of the exhibits started to become less scientifically correct and more expressive.

The Von Klenze Münich corridor represented a significant departure from static, determined, spatial sequences and a shift towards a more and more fluid space, where sub-spaces are devoted to specific aspects of the topic and the navigation paths through the exhibition are at the visitor's will. In the Deutscher Werkbund exhibition in Paris in 1930 Gropius and Bayer, with photographs, forniture and models placed on many planes, explored the possibility of extending the visitor's field of vision placing the exhibits throughout the available space in a flowing, non linear, sequence. A further step forward in that direction was taken, more recently, by Tschumi with his exhibition Architecture and Event at the MOMA in New York in 1994 where not only images and models were floating in the space, but also projected filmic images and mobile lighting suggested a dynamic, flowing, space.

Virtual Museums

In virtual space we move from destination to destination, from piece of information to piece of information. Distance is no longer a relevant measure of travel. [Best, 1993]
In virtual Museums the exhibition space could be a text based environment, a 2D graphical interface or a 3D model accessible through the Internet while digital images, videos, 3D computer models and virtual sculpture are just some of the objects that can be displayed.

The spatial sequence along the circulation system of physical Museums is replaced by a temporal sequence of the display with a consequent significative reduction of physical space needed to access the collections [Mitchell, 1995]. Instead of the extensive physical space required for the display of physical works of Art, virtual collections will only need the space of the interface through which they are accessed. Like in physical Museums, in virtual ones the organisation of the material is still the main curatorial concern. But the sequencing of the exhibit can become more complex and sophisticated using the technology of hyperlinks that allows jumps between data, creating a whole new set of navigational possibilities. The hyperlink, implying a spatio-temporal discontinuity, can only take place in the virtual realm.

Each item of the collections, instead of being simply physically adjacent to another, can be hyperlinked to any other object in the museum's database so that the visitor of the virtual exhibition can follow an extremely customised navigation path.

Exhibition Curators and Virtual Museums - Levels of interaction

Only with the understanding of the way physical museum works and of the nature of their limits it is possible to understand the real potential of virtual Museums. Sheldon Annis in his essay Museums as a Staging Ground for Symbolic Action [Annis, 1986] gives an interpretation of how Museums work. In first instance, he individuates a discrepancy between the stories that the curators try to tell and the "script" of the museum visitor. This is due to the fact that, while curators are more concerned with the object of the exhibit, visitors build their own "warehouse of symbols" interacting with exhibition spaces at three different levels: dream, pragmatic and cognitive.

The dream space is the locus where the museum is experienced as a "flickering of and amongst symbols", time and space are stretched at the subjects' will, in his mind.

The pragmatic space is the level of interaction where "it does not matter whether the coins were Roman or Chinese" and the visitors experience goes beyond the objects of the exhibit. It is also the level of interaction where the museum-goer ticks a kind of personal checklist, "done the Louvre, done the Hermitage, done the Met".

Cognitive space is where the visitors experience of the exhibition is closer to the "idea ... that the designer writes in physical form across the museum's floors and walls", where the visitor visits the exhibit in the closest way to the one wanted by the curator. But "since fully understanding the curatorial message requires patience, some quiet and no pushing, most museum goers" still interpret the museum in a personalised way and "enter cognitive space selectively".

In virtual exhibitions that really exploited the potential of the WWW as a medium, those interaction levels can be rethought. Dream space for instance, usually out of the curators' interest in physical Museums, could become a crucial part of the design of the virtual exhibit. Annis' "flickering of and amongst symbols" can be easily referred to the jumping between links of the web surfing and the visitor of the gallery can actually make all the connections between object that the visitor of physical exhibitions pictures in his mind can be explicitly made by a hyperlink.

The lack of physical constraints, proper of a virtual exhibition space, should be seen by curators as a stimulus to create on line exhibitions that could not take place in the physical Museums. But curators still don't seem to be too heavily involved in most Museums on-line sites. Galluzzi, Director of the Museo di Storia della Scienza, Florence, Italy, argues that the sites are in most cases implemented and maintained by engineers that pack the pages with contents without shaping the meaning of the objects, which is, by definition, the curatorial task [Galluzzi in Marshall, 1996].

The narrative sense of the curators make the objects of the exhibit acquire much of their meaning. This applies to both physical and virtual Museums although, due to the intrinsic differences of physical space and cyberspace, the modes in which this narration is developed have to be different. In the virtual realm, the move from the objects of a collection to the meanings of an exhibition does not come only from linear juxtapositioning, grouping and sequencing of the objects but from hyperlinks with other objects and from the interactivity between the institution and the public.

Virtual environments design issues

Differences between physical environments and virtual environments

Although virtual and physical environments may be experienced via the same "perceptual process employed for perception of the real world", they present some inherent differences [Charitos and Bridges, 1997:147]. The understanding of these differences is preliminary to a design of virtual environments that exploited the possibilities of virtual reality as a medium.

Absence of physical constraints

"Architecture begins with the creative response to climatic stress". [Benedikt, 1991:122]

In the virtual realm there are no physical constraints to determine the spatial nature of the environments: there is no micro nor macro climate, and, therefore, no need for shelter, no gravity, no friction. Enclosure, for instance, will not need to exist for shelter reasons, neither wall for structural reasons or partitions for activity separation. There is not an absolute above/below, an inside/out unless expressly implemented. There is no limit for the number of dimensions of virtual environments. Data objects and data point have up to three extrinsic dimensions, and this allows their geometrical model to be constructed in the Cartesian space, but each object can unfold, opening up an infinite number intrinsic dimensions, which can 'unfold ... in principle, nested ad infinitum' ) [Benedikt, 1991:144].

The lack of the physical constraints that typically informed architectural design, though, does not mean that there are not any constraints at all. The virtual realm, as well as the physical, has its own set of constrains which will affect the design of virtual worlds. Number of polygons, image resolution, disk space, bandwidth, memory, hardware platforms and software tools are some of them. Therefore spatial elements in virtual environments need to be rethought and designed to respond to the constraints proper of the virtual realm.

Spatio-temporal discontinuity

In the virtual real time and space do not need to be continuous.

In particular, space is "non-contiguous, multidimensional and self-reflective" [Bridges and Charitos, 1997:147]. Hyperlinks between environments and objects allow movements through the space not based on spatial contiguity. In a virtual environment, like, for instance, a museum, the set of data, the collections, can be sequenced allowing jumps between objects, that can be inter-linked without being juxtaposed.

Time as well can be discontinuous and its pace may be adjusted to the operators will.


In the virtual realm the operator can alter at will the scale of environment, from a "geographical level" to an extremely detailed one. [Charitos, 1996].

Level of realism

World Design Inc. [in Bridges and Charitos, 1997] classified virtual environments in terms of levels of realism:
  • Hyper realities, that aim to represent the complexity of the real world. In these virtual environments the creative input by the designer is mostly confined to the interaction of the operator with the environment rather than focused on the environment itself.

  • Selective realities, simplified representations of the material worlds, where some selected aspects of the environment are treated with high level of realism and others are transformed, cutting unnecessary detailed information.

  • Abstractions, particular virtual environments, where are represented either abstract information about very complex real environments or information that cannot have a physical representation. In this kind of virtual environments becomes crucial the way the designer abstracts the complex information and structures it in an understandable way for the user.


Virtual environments often use metaphors at various levels to represent the complex sets of information. Bryson (1994) indicates three levels of metaphorical abstraction than can be used in the design of a virtual environment:
  1. Overall environment metaphor

  2. Information presentation metaphor

  3. Interaction metaphor
Campbell [1996] suggests that the use of abstract architectural elements and environmental cues can facilitate the navigation in virtual environment, but it can be argumented that "in the metaphorical approach the metaphor becomes the ... principle to which all other aspects of the design must adhere". "The problem is that ... slavish adherence to a metaphor prevents the emergence of things that are genuinely new" [Nelson, 1990]. So using the metaphor of physical architecture to inform the design of virtual environments carries the risk of importing into the virtual realm those constraints proper of the physical one, where there is no need to do so.

In a virtual gallery, e.g., using the overall environment metaphor of the traditional museum building, made of a sequence of rooms where paintings are linearly arranged, would, on one hand, facilitate the orientation and navigation of the visitor, but, on the other hand, would mean to apply to cyberspace those physical constraints it lacks of and of which curators would willingly do without. [Marshall, 1996]. Adhering to the museum building metaphor in the design of a virtual gallery would, therefore, prevent from exploiting the possibilities and potential of the virtual.

Direct spatial meanings of virtual environments (VE)

While Campbell [1996] sustains the general validity of an overall architectural metaphor as a way to represent complex information in abstract environments, Bridges and Charitos [1997] suggest an alternative approach to the design on VE. They suggest that the methods for the creation of space defining Form in VE should move from the architectural theory of physical environments. Because the physical constrains that produce architectural Form cease to exist in the virtual realm, though, VE design should not imitate the forms of architecture and urban design but, instead, be informed by the general principles of composition of space typical of physical architecture.

Following Norberg-Shulz' studies on existential space [Norberg-Shulz, 1971:20], and Deleuze's on image and movement in time based media [Deleuze, 1983], Bridges and Charitos, introduce a taxonomy of generic space-establishing elements in virtual environments [Bridges and Charitos, 1997:147-149]. They define the virtual environment as a closed system within which solid and void spatial elements are framed to Form the image of the virtual environment. The void spatial elements Form the 'sets', while the solid space-establishing objects Form the 'props'.

Sets: void spatial elements

  1. "Places are the spaces where activities are carried out". The particular activities that take place within a place are usually 'static actions' and have a special meaning in relation to those spaces. A place is defined by its boundaries and its should be designed in such a way that allows the user to discern its 'inside' and 'outside' as well as identify the activities that take place in it and feeling secure enough to engage in them. It has been said by Norberg-Shultz that "a place ... experienced as an inside should generate a spatial sense of proximity, centralization and enclosure".

  2. Paths are kind of spaces which implies dynamic action. They consist of a starting point, a end destination and a direction to follow, which is always evident for its particular spatial layout.

  3. Domains are "relatively unstructured grounds, on which places and paths appear as more pronounced figures" [Bridges and Charitos, 1997:148]. The participant takes 'possession of the environment' [Norberg-Shulz, 1971:20] in organizing it into domains defined by paths and places. Domains cannot be directly experienced as whole and can rather be seen as a mental abstraction of the environment experienced by the subject through paths and places.
  4. Thresholds are transition or intersection spaces between any other spatial element. All the kind of openings, the intersections between paths and between paths and places are examples of thresholds. As interfaces between different elements, thresholds can be seen as the locus of intersection of many possible routes the operator can follow and for this they become a crucial loci for the navigation within the virtual environment.
Props: solid space-establishing elements:

Solid space-establishing elements can be grouped and classified according to their form and function. As far as it is concerned with the form they can be defined as two dimensional surfaces or planar objects or three dimensional objects, whereas according to their function, the space defining elements can be divided into:

  • Bounding objects, define the space, by binding it somehow.

  • Landmarks are distinctive objects that serve as a point of reference within the environment.

Spatial knowledge

One of the key issues in the design of virtual environments is how to improve and assist the participants in the process of orientation and wayfinding within the environment. Spatial orientation can be defined as the participant's ability to mentally individuate his position within the environment on a representation of the same environment [Passini, 1992], while wayfinding as the dynamic process in which the participant uses his spatial awareness of an environment and his navigational ability to reach the desired destination [Satalich, 1995].

5. Cognitive mapping

Passini [Passini, 1992:35] describes the cognitive processes of wayfinding and orientation as relying on the different abilities of cognitive mapping, decision making and decision executing. Amongst these abilities, cognitive mapping, is the most important for the acquisition of spatial knowledge. During the cognitive mapping process the participant acquires, codes, stores, recalls and decodes information about the environment, generating new information that allows him to understand the space [Moore, 1976]. The development of cognitive maps, as product of the cognitive mapping ability, is particularly "important in enhancing the experience of people in places where they are not frequent visitors" [Lang, 1987] and, in such unfamiliar, informational spatial environments, to generate a cognitive map the participant needs to be able to navigate and orientate intuitively [Benedikt, 1991].

6. Acquisition of spatial knowledge

The acquisition of spatial knowledge for orientation and wayfinding within a space is mostly acquired by the direct spatial experience of the environment. The participant needs to move actively through the space, without being passively transported, and to be able to make his own navigational decisions. Spatial knowledge can also be acquired, or improved, indirectly my means of tools like 2D and 3D maps, information points and other signs [Charitos and Rutherford, 1997].

The process of spatial knowledge acquisition can be broken down into three stages [Charitos and Rutherford, 1997]:

  1. Landmark knowledge. When experiencing a new environment we first individuate those elements that we perceive as predominant, landmarks, without acknowledging the rest of the environment.

  2. Route knowledge. Once recognized the landmarks within an environment we start to think about possible routes to link them. The proceduralization of route knowledge is affected by the degree of interaction that the operator can have with the environment. When the user can actively engage in the navigation, moving freely through the environment, the development of route knowledge is significantly enhanced compared to when only a lesser level of interaction, like passive 'viewing', is possible.

  3. Configuration knowledge. After the acknowledgment of landmark objects and of the routes that connect them, including other objects placed in intermediate position between landmarks, the final step of the spatial knowledge acquisition is to organize the acquired information. The way this information is structured is a "non-sequential, two dimensional holistic organization of object location and route interrelations in space" [Charitos and Rutherford, 1997].


In the real world physical environments are experienced by continuous movement through space over a period of time. In the virtual realm, instead, the hyperlink technology allows to move through space without moving through time, jumping instantaneously between not adjacent data.

The 'character' of the hyperlinks between different elements of the virtual environment depends on whether or not the linked data are geometrically adjacent and on the way the data on each side of the link are related [Campbell, 1996]. Four kinds of hyperlinks can be defined:

  1. Hyperlinks that connect spatially contiguous environments of a virtual world and in which the informational relationship between the linked set of data is strong. Links of the first kind are usually employed to allow a free navigation through the space optimizing the rendering and download time. For instance they can have the form of a transparent door to another room, with a low level of details and only once the participant expresses the intention to go in that room, by clicking on the door, a more detailed model of the target room is downloaded.

  2. Hyperlinks that connect environments which are not adjacent but contain related information. If a metaphor is used for the object to which the link is attached it needs to be a different metaphor from the one used for links of the first kind. Therefore for this kind of link a texture mapped frame it is to prefer to the metaphor of a door, which would be not appropriate and confusing for the participant.

  3. Hyperlinks in which the elements connected are geometrically contiguous but the information they contain are unrelated. This kind of links, thus technically possible, is practically not used because making explicit links between unrelated objects in adjacent environments it is seldom required.

  4. Hyperlinks that connect spatially non contiguous environments and in which there is not direct informational relationship between the linked set of data.
Links of the second kind are perhaps the ones that allow to exploit more the potential of virtual environments against physical environments.

Reconfigurability of Virtual Galleries Spaces

Allowing the visitor to reconfigure the virtual exhibition space seems the ultimate outcome from interactivity between the museum and the public.

With 2D environments, like the Virtual Curator software, developed as an educational tool by the University of Brighton, UK, [Worden, 1997] the chance to structure the exhibition at the user's will does not present particular problems for the user's orientation within the space.

As far as is concerned with three-dimensional gallery spaces, though, the possible approaches to address the issue of the reconfigurability of the space are at least two. One approach is to carefully design the exhibition layout placing the object within the 3D environment without allowing the visitor to reorder the content of the exhibit. If the spatial organisation cannot be modified the virtual gallery inherits from the physical one the incapability to provide different contexts in which to view the objects of the collection. Another approach would be to allow the visitor to reconfigure the exhibition space according to personal criteria. For instance the user could decide to retrieve and display in a certain spatial sequence all the still life painting or architectural photos. Every time the object are reshuffled, though, the cognitive maps priory developed become of no use [Shum, 1990] so the peculiar "memorability of spatial organisation", that is one of the major advantages of 3D galleries over databases, is lost. Since virtual 3D environment are more costly to implement than databases, this loss has to be taken into account when embarking in the design and production of a reconfigurable 3D exhibition space [Boyd Davis et al, 1997: 70].

The design of a 3D Art Gallery


The project, partly initiated by the Department of Education of the Gallery of Modern Art in Glasgow, is concerned with the design of an abstract virtual gallery which does not intend to reproduce or mimic a physical gallery in its architectural form and modes of exhibition. The aim was, rather, to create an environment, existing independently in its own right, where to exploit new possibilities of exhibition modes based on the peculiar spatiality of virtual environments. Adhering to the overall metaphor of a traditional museum building, with a series of interlocking rooms with pictures hanging on the walls, seemed to be limiting in such intent. Therefore the abstraction used to represent the data set of the exhibit is the one of a virtual environment defined by space-establishing elements like places, paths and landmarks.

Moving from those premises, one of the aims of the designed virtual gallery, is to look at how virtual exhibition spaces, providing features that are specific of the virtual realm, can enhance the experience of the exhibits of traditional, tactile, works of Art like, for instance, paintings.

Another aim of this project is to experiment on how the spatial discontinuity of a virtual environment can affect the modes of narration of an exhibition. Space defining elements as well as works of Art will be hyperlinked to other, not geometrically contiguous, objects. But in the design of such connections particular attention has to be paid to minimise the sense of disorientation that might occur in the participant due to the difficulty to develop cognitive maps in a discontinuous space.

The gallery is realised in VRML, the standard developed to integrate the Internet with the virtual reality technology, that allows 3-dimensional worlds navigable on-line in real time, opening up a whole new set of possible applications for virtual exhibition spaces.

Because VRML was created as a standard for the Internet and because potential viewers of VRML world might be using less powerful machines, some VRML specific size issues have been addressed at the language design stage. One of the language features resulting from the size concern is the in-lining, that allows to break large VRML scenes into smaller files that are only loaded when required. Another feature to improve the viewing of VRML files from slower machines is the level of detail (LOD). If the object is far away from the participant's position in the world, it displays a simplified version of the model and dynamically updates the display with more detailed versions of the same model as the operator gets closer to the object.

Both these features were used during the implementation of the model.

The potential of the VRLM 2.0 for the development of virtual exhibition space is to provide an environment where the distance between the collections and the public is minimised by the possibility to interact with the space and manipulate the objects exhibited.

The way the operator can manipulate the exhibition space, though, needs to be carefully designed because, as mentioned earlier, moving and modifying space defining elements as well as exhibited objects will affect the perception and cognition of the environment, making all the priory developed cognitive maps unuseful.

Spatial layout

The gallery space is composed by a series of space defining elements variously interconnected with each other either by means of geometrical contiguity or proximity or by hyperlinks. Given the absence of physical constraints the need to provide a solid enclosure that separated the inside from the outside of the gallery was not felt Instead the overall environment is articulated in gallery space and non-gallery space, dynamically flowing into each other, without a continuous boundary between the two.

This articulation of space is enhanced by the use of colour. The non-gallery space, fades into the background and has no lights and a black colour, to signify the absence of colour, while the gallery elements are lit and in bright colours.

The starting point, where the user finds himself as entering the world, is contained in a cubic cell with no walls, from were it is possible to have a view of the overall space. The entrance cell element is of a bright orange colour and it is lit with a warm yellow light. The idea is to attach to this cubic element a home feeling so, whenever the participant incurs in an element with same shape and colour, he will automatically feel secure. The attributes of the entrance cell element, cubic Form, bright orange colour and warm light, will be attached to a scaled down copy, placed throughout the gallery that will function as a help cell, where participants can retrieve orientation information. Both entrance cell and help cells will become a landmark within the gallery environment.

In the entrance cell it is displayed a small model of the gallery itself that works as a 3D map, hyperlinked to the various parts of the gallery. The user, entering the gallery world, can then decide whether to navigate autonomously through the space following the spatial cues or go directly to a section by clicking on the 3D map. Then, once the operator wants to move on from the section explored he can either go back to the entrance cell, by entering any help cell or continuing his navigation through the gallery space in the geometrically adjacent sections.

The fact that to choose a new destination within the gallery space the operator has to go back to the starting point, may be seen as an obstacle to a personalised navigation route and it would definitely be such in a physical environment, where the movement in space would require also a movement in time. But, since with the hyperlink to the entrance point this movement in space is instantaneous, sending the participant back to the entrance cell, connotated as a secure place, every time he needs navigation and orientation assistance seemed more advantageous than creating a new help element.

The gallery has been organised into generic domains, each pertaining to different kinds of artefacts, but the layout of such domains is not rigidly structured and allows for adjustments to the nature of the collections exhibited. Ideally the Form of the gallery should adjust to the kind of exhibition it is going to host.

Orientation and navigation

The gallery space was designed to facilitate the spatial awareness of the operator and particular attention was paid to clearly define all the space-establishing elements. In particular:
  • Landmarks like the entrance cell and the help cells are made clearly identifiable amongst the other space defining elements. In the case of exhibited objects with a particularly striking shape or distinctive colour some expedients had to be used to avoid to be misread as a landmark objects that not necessarily convey that space defining meaning. Virtual sculptures, for instance, are displayed in a scaled down copy that functions as a 3D icon hyperlinked to another world where the sculpture is displayed in a especially designed environment or enclosure, maybe following, where possible, the directions of the artist himself.

  • Paths are clearly identifiable due to spatial qualities of the elements that define them. Deriving from the sixteenth century's gallery model where in the galleria -a place to walk through- works of Art were displayed and dynamically viewed moving along the path, in the paths of the designed


The gallery was modeled bearing in mind the interactivity features of VRML-2 worlds, although they haven't at the moment been implemented.

In particular, regarding the possibility to modify the environment, VRML 2.0 allows users to interact with other participants on-line (like in a 3D chat). Users can also be allowed to move objects within the environment. This feature could be easily used, e.g., to let the visitors, provided with some store material, curate an exhibition.

It is advisable, though, to allow users reconfigure the space only in confined sections of the whole virtual gallery, specially dedicated to this function. The need to constrain the reconfigurability feature to determined spaces is felt in order to avoid the uncontrolled continuous re-organization of the gallery by web visitors that would create disorientation especially in regular visitors.

As far as it is concerned with the interaction between visitors, no particular area of the gallery has been specifically dedicated to this function. Devoting a space to the social interaction it would, in fact, be seen as a restrictive imposition over the modes of socializing within a virtual environment. Besides interaction features in general can be thought of and designed a priori, but they then need to be adjusted to the specific characteristics of the on-line participants.

A painting exhibition in the Virtual Gallery

A small exhibition was curated to experiment with the possibilities of virtual reality as a medium for the delivery of exhibitions.

The category of works of art chosen for the exhibition is painting. The choice was determined by the fact that paintings are traditionally exhibited in physical art galleries, so it would be possible to compare the two exhibition modes. Painting is also, possibly, the art form the general public is more familiar with and that familiarity would compensate the novelty of the technology the participant would have to use to view the gallery.

A small selection of digital reproduction of paintings by Mondrian was chosen to experiment with new modes of display in the virtual gallery. This painter was chosen because of his interest in experimenting with the meaning of abstract forms as well as with the fourth dimension of time. Besides, being the general public familiar with Mondrian's work, the participant would understand more clearly that what he is experiencing is a new way of exhibiting a physical painting.

The way of displaying the reproductions of the painting aims to explore the possibilities and limits of the discontinuity of space in virtual environments as well as seeking for a new way to experience the work of art.

The digital images of the painting were mapped onto a rectangular extruded object, placed in a thick red frame. These were then positioned in the painting gallery, a space clearly defined as a path. As the participant moves along the path he collides with the framed reproductions of the painting floating within the path space and can experience the front and the rear of the painting object.

Some of the frames, thus with the same shape and size as all the other, have a particular feature that functions as an attractor. Such feature, for instance, could be the being self-luminous or blinking. The painting reproductions contained in these distinctive frames are portals to 3D virtual worlds modeled on the paintings extruding them in three dimension. The image of the painting becomes like a projection on a plane of a three dimensional world. The participant clicking on the painting enters the 3D world that experiences navigating through it.

The 3D environment based on the paintings' formal qualities can function either as just a spatialization of the formal content of the painting, to allow the visitor to experience the shift from a 2D pictorial representation to a 3D dynamic space, or could turn into another exhibition space where extra information about the painting itself, such as its archaeology, its connections to other works by the same artist or other artists or with architecture, can be displayed.


The design proposed is to be read as an hypothesis for a truly interactive Internet based exhibition, but the assumptions there made need to be tested with users.

Although the interaction features were limited to the design stage, most of the geometry of the Virtual Gallery was implemented in VRML.

As a future development of the project an open network of institutions (museum and educational) willing to experiment with the latest multiuser 3D Internet technology is being established.


Annis, S. (1986) "The Museum as a staging ground for symbolic action". Museum 38,3:168-171

Benedikt, M. (1991). Cyberspace: First Steps. MIT Press, Cambridge, MA

Best, K. (1993). The Idiot's Guide to Virtual World Design. Seattle, Little Star

Boyd Davis S., J. Lansdown, A. Huxor. (1997) The Design of Virtual Environments. Centre for Electronic Arts, Middlesex University

Bridges, H.A. and Charitos, D. (1997). "The architectural design in virtual environments", R. Junge (ed). CAAD Futures ’97: pp. 719 -732. Kluwer Academic Publishers, Dordrecht

Bridges, H.A. and Charitos, D. (1997). "On architectural design in virtual environments", Design Studies 18:143-154, Elsevier, Great Britain

Brown, R. D. (1997). Virtual Unreality: The Modulation of Time Space and Energy. (accessed June 1997)

Bryson, S. (1994). Approaches to the Successful design and implemantation of VR Applications, Proceedings of the Virtual Reality Application Conference, June 94, British Computer Society, Leeds

Campbell, D. (1996). Design in Virtual Environments Using Architectural Metaphor: a HIT La Gallery .MArch Dissertation, University of Washington

Charitos, D. (1996). Defining existential space in virtual environments, Proceedings of the Virtual Reality Worls '96 Conference, February '96, IDG Magazines, Stuttgart

Charitos, D., Rutherford, P. (1997). Ways of Aiding Navigation within VRML Worlds, Proceedings of the Sixth International EuropIA Conference, 2-3 April 1997, Edimburgh

Cerulli, C. (1997), Museums without walls, M.Sc. Computer Aided Building Design Dissertation, University of Strathclyde, Glasgow

Cerulli, C. (1998), Developing 3D exhibitions online, Cultural Grid MDA Conference Proceeedings, Cambridge

Galluzzi in Marshall, L. (1996). "Code For a Grecian Urn". Wired. 2.09: 75-79, 102, 104.

Lang, J. (1987). "Cognitive Maps and Spatial Behavior". Creating Architectural Theory. Van Nostrand Reinhold, New York

Malreaux, A. (1953). Museums without walls. The voices of silence. Martin Secker and Warburg, London

Marshall, L. (1996). "Code For a Grecian Urn". Wired. 2.09: 75-79, 102, 104.

Mitchell, W. J. (1995) City of bits: space, place and the infobahn. MIT Press, Cambridge MA, London

Moore, G. T. (1976). "The Developement of Environmental Knowing; an Overview of an Interactional-Constructivist Theory and Some Data on Within-Individual Developmental Variations". Psycology and the Built Environment. Architectural, London

Nelson, T. (1990). "The right way to think about software design" in Laurel Brenda (ed), The art of human-computer inerface design, Addison-Wesley

Norberg-Shultz, C. (1971). Existence, Space and Architecture, Praeger Ltd. New Y

Passini, R. (1992). Wayfinding in Architecture, Van Nostrand Reinhold, New York

Satalich, G. (1995). Navigation and Wayfinding in Virtual Reality: Finding the Proper Tools and Cues to Enhance Navigational Awareness. University of Washington, Seattle

Shaw, J. (1994) "The new role of the artist as public figure", Art & Design, 39

Shum, S. (1990). "Real and Virtual Spaces: Mapping from spatial cognition to Hypertext". Hypermedia, 2: 133-158

Worden, S. (1997). "Thinking critically about virtual museums". Proceedings of the Museums and the Web Conference, Los Angeles, CA, March 16-19, 1997