Geographical understanding of communications and transportation usually proceeds macroscopically, from the vantage point of a remote and detached observer. Total minutes of telephone communication between a set of countries, presented as a network map, would be one form such knowledge might take. Frequency of flights between a set of cities would be another. While the macroscopic perspective provides a good sense of the overall degree of interaction between places and how such interaction varies spatially, it obscures the way communication and transportation are incorporated in people’s lives in real places. Much that is of interest from a cultural geography or social geography viewpoint is lost in the process of aggregation.

Time geography’s seminal question "where are the people in regional science" (Hagerstrand 1970) is as relevant as its subsequent focus on activity and authority patterns in time and space. The question "where" must be answered through a focus on individual lives, but benefits greatly from the inclusion of distanciated activities. These surround the physical nucleus of the human body like a flickering flame or, as I have earlier argued, an amoeba (Adams 1995). The interactions of people’s extensions (e.g. telephone conversations) are no less constitutive of social structure than the interactions of co-present individuals. A fax, phone call, letter, memo, video image, or broadcast voice may affect consumption, direct business activities, pursue a romantic encounter, reinforce or challenge political values, provide a connection to the past or future, or build community. Mediated or disembedded interactions can not be understood fully as if they were objects existing at the macro- (societal) scale. They must be understood also from the perspective of individual lives, particular daily routines pursued in and through the virtual places of various media.

This study applies two "lenses" to understand personal accessibility or extensibility in time-space. Both lenses are directed towards the lives of five people who live in the Albany, NY metropolitan area. First, a narrative lens provides a general "feel" of their different lifestyles. These lifestyles range from a quiet retirement schedule to a busy schedule coordinating several professional positions. All five schedules include both work and leisure, with distinctive contrasts and complementarities between the two spheres of activity. Also included in the narrative are intersections of the five life-paths, encounters and communications between the individuals. Through these intersections the five daily routines are integrated as a social network.

Second, a "map" of these daily routines is shown. This map is in fact a three-dimensional dynamic representation of an abstract representational space in which virtual and physical interactions are shown. Stored as a set of objects in a computer aided design (CAD) database, the map can be rotated and examined from various angles, and it can be querried to reveal selected themes. For example, computer network communications can be shown without any other communications displayed. Another thematic selection would show only two-way communications such as telephone, and disregard one-way communication links such as radio and television. The "map" therefore, like GIS, has no given appearance, but is a representation of information selected and organized by the user. Furthermore, like GIS, this manipulable map facilitates analysis of the data because it renders a large amount of data visible in a single image.

This study enhances understanding of how emerging technologies are affecting spatial behavior. It provides a space-time topology which links physical and virtual worlds. From the vantage point of an abstract representational space, it indicates the prevalence of a kind of virtual commuting, the movement people undertake between physical and virtual places, which occurs with increasing frequency. It also explores the task of representing such a multidimensional space in a computer database format.

References:

Adams, Paul. 1995. A Reconsideration of Personal Boundaries in Space-Time. Annals of the Association of American Geographers 85(2):267-285.

Hagerstrand, Torsten. 1970. What About People in Regional Science? Papers of the Regional Science Association 24:7-21.
 

One definition of the Internet is: "a collection of resources that can be reached from those networks"

(Krol & Hoffman 1993, page 1).

This definition by Krol & Hoffman provides the starting point for my conceptualisation of accessibility in the Information Age. I will examine how one can begin to measure and visualise the aspects of accessibility to resources within the Internet. I start my discussion with the assumption that a person has physical access to the Internet, via a networked computer. I also take a broad view of accessibility, going beyond just formal mathematical models.

There are many important dimensions to accessibility beyond simple Internet connectivity that need to be considered and measured. Once you have access to the Internet, how accessible then are the information resources, people and electronic places that are potentially available. For example a person may be seated at their networked Windows98 PC and click on the Internet Explorer icon which will connect the to the ‘net and by default load the Microsoft homepage, but then, taking the Microsoft mantra, "where do you want to go today? ®". What resources are accessible within the Internet and how do you reach them? People and electronic places are not obviously accessible when you start "surfing", things are largely invisible from a distance and can only be located with precise directions.

It is also important to consider how these dimensions of accessibility can be represented, particularly given the abstract nature of information spaces, which are often assumed to have no geography, and the ethereal nature of searching and browsing activities being undertaken. These representations need to be dynamic and interactive, as well as being readily available whilst navigating the Internet.

I will argue that the scope of accessibility needs to be expanded to encompass notion of information accessibility. The growing importance of the Internet, and its layered services, for receiving and sending all manner of information and for personal interaction requires us to consider how concepts of accessibility are played out within these electronic spaces. Gaining access, in a timely fashion, to the right information resources (be that a web page, an email, a video clip or real-time chat) is problematic for a number of human and technical reasons. The Web may well facilitate easy access to vast arrays of information from servers around the world, but this does not mean one can find useful, current, reliable and affordable information at the right time. As Pirolli et al comment "The apparent ease with which users can click from page to page on the World-Wide Web (WWW) belies the real difficulty of understanding the what and where of available information" (Pirolli et al 1996, page 1). The size of the Web has been estimated at some 320 million documents in March 1998 (Lawrence & Giles 1998). The problems of information retrieval through searching and browsing this massive Web space are crucial for understanding accessibility in the Information Age. There is now an increasing awareness of the problem of "information overload", with "a tsunami of data crashing onto the beaches of the civilized world" (Wurman 1997, page 15). Accessibility to too much information is potentially as significant an issue as accessibility to too little information. A great deal of research effort is being directed by scientists in a range of disciplines to cope with the problem of information retrieval and information overload through filtering, structuring, analysing and visualising information to aid the limited human capacity to search for, absorb and comprehend information. Much of which, I will argue, is relevant to broadening the scope of "geographic" accessibility to encompass information spaces.

It is important to note that the Internet contains a number of distinct services, which are often thought of as different "spaces", with differing virtual landscapes. The different spaces support different types of information exchange and levels of social interaction. Therefore, they are likely to require differing measures of accessibility and forms of graphic representation to be able to model their true nature. At a fundamental level, the different information spaces are caused by different network protocols used by software tools to communicate over the Internet, which give rise to the different form and functions apparent to the end-user. Figure 1 shows a sketch map produced by John December showing the principle information spaces and some of the connections between them. The map provide a good way of thinking about the information spaces of the Internet as distinct and self-contained domains, but with fluid, complex boundaries and many interconnections and overlaps. The map was drawn at the end of 1994 and the nature of the Internet has changed markedly since then, with certain spaces dying off (WAIS and Gopher) and the inexorable and exponential growth of Web space. For many end-users the Web and the Web browser representation, is the key information space, although email is still the most widely used information service on the Internet (ITU 1997). More recent work on conceptualising the form and structure of different information spaces includes Michael Batty’s examination of "Virtual Geography" (Batty 1997), Paul Adams’s work on "Network Topologies and Virtual Place" (Adams 1998) and Brian Gaines’s research on the human-factors of Internet information spaces (Grimes 1997).

In this paper I shall consider how we can develop the theme of information accessibility within the Internet by closer examination of the issues measurement and representation in two information spaces - Web space and virtual worlds. In terms of Web space we are particularly concerned with issues of information visualisation and retrieval, while in the 3d virtual worlds issues of navigation and design will be the focus. In addition, consideration will also be given to the measurement and mapping of the performance of underlying Internet infrastructure that is vital to support the "higher-level" information spaces.

Author Information

Martin Dodge is a researcher in the Centre for Advanced Spatial Analysis (CASA), with a background in social geography and GIS, one of his major research interests is the geography of the Internet. He maintains the Cyber-Geography Research web site at http://www.cybergeography.org/. He can be contacted at the Centre for Advanced Spatial Analysis, University College London, Gower Street, London, WC1E 6BT, United Kingdom. Email : m.dodge@ucl.ac.uk. Telephone : +44 (0)171 391 1782. Fax : +44 (0)171 813 2843. Web : http://www.casa.ucl.ac.uk/.

References

Adams P., 1998, "Network Topologies and Virtual Place", Annals of the Association of American Geographers, Vol. 88, No. 1, pages 88-106.

Batty M., 1997, "Virtual Geography", Futures, Vol. 29, No. 4/5, pages 337-352.

December J., 1995, "A Cybermap Gazetteer", TeleGeography 1995 : Global Communications Traffic Statistics & Commentary, edited by Gregory C. Staple (TeleGeography Inc., Washington DC, USA).

Gaines B.R., Chen L.L., & Shaw M.L., 1997, "Modeling the Human Factors of Scholarly Communities Support Through the Internet and World Wide Web", Journal of the American Society for Information Science, Vol. 48, No. 11 pages 987-1003.

International Telecommunication Union (ITU), 1997, Challenges to the Network : Telecommunications and the Internet, September 1997.

Krol E. & Hoffman E., 1993, "FYI on "What is the Internet?" ", Request for Comments: 1462, May 1993. (Available at ftp://ftp.ripe.net/rfc/rfc1462.txt)

Lawrence S. & Giles C.L., 1998, "Searching the World Wide Web", Science, Vol. 280, 3rd April 1998, pages 98-100.

Pirolli P., Pitkow J. & Rao R., 1996, "Silk from a Sow’s Ear: Extracting Usable Structures from the Web", Computer-Human Interaction (CHI’96) Conference,

3-18th April 1996, Vancouver, Canada.

Wurman R.S., 1997, Information Architects, Graphis Press Corp.

This paper considers the issue of accessibility and human activity scheduling in the context of space adjusting technologies, proposes some concepts for implementing an enhanced view of access and activity, and attempts an application of these in the domain of the geography of tertiary learning.

It is widely accepted that accessibility is a complex and difficult concept to make operational. Traditionally the quantitative measurement of accessibility has been treated as a purely spatial phenomenon, based upon simplistic assumptions of distance, with limited cognisance of either a temporal dimension or process. Time and activities were consistently neglected in these models of accessibility potential and spatial interaction, for a variety of conceptual and computational reasons. However refined models have emerged based upon more realistic measures of spatial separation and interaction modelling, including topological or network travel time, and cost.

Although not conceived of as an accessibility measure by Hägerstrand the field of Time Geography which grew from his seminal work on life-lines, has nevertheless contributed to a wider appreciation of ‘real’ access. Time-geographic studies stress the importance of the temporal dimension by explicitly incorporating time in their models of process, and so in any calculation of accessibility and spatial opportunity. Human activity programmes (sequences of social, economic and recreation events) can be seen as having a given duration, and occupying a given location in time and space. These locations are subject to a set of constraints which inhibit movement and hence participation in activities. The ability to meet these constraints is very much the essence of effective accessibility.

We argue elsewhere that the strengths of Time Geography were not realised in the 1970s due to an inability to scale the concepts into substantial applications in the real world, and to implement better conceptions of 'space' as a separator for physical process. We now witness advances in geographic information technologies and spatial modelling packages which are seizing new opportunities and contributing to a growing volume of literature that applies core concepts of time-geography to specific research problems. Both new application areas and enhanced concepts are features of these developments.

This paper accepts that there is a need to extend these studies into an improved language of access and interaction, and draws on and reports earlier work on a discrete cellular implementation of Hägerstrand’s geometric conceptualisation of space and time by the authors. It also accepts that a major overlooked issue in a complete Time Geography has been the substitution effect of 'virtual' for 'real' meetings or access to functions. In the 1970s this principally reflected the use of the telephone making a face to face meeting unnecessary. Currently, advances in telecommunications make the potential for substitution far more pervasive. This poses new problems in defining the interaction of individual time geographies and structures. In particular, it requires us to consider in greater depth the issues of discretionary sequencing and physical/virtual substitution.

The aim of this paper is to develop concepts in these areas for application in a specific domain. The domain we have chosen is the geography of tertiary learning. We do this for two reasons. Firstly, the student life experience is an excellent context to develop and test concepts in : it is structured with numerous 'markers' in terms of lecture timetables and such, can be well documented. Secondly, it is an area of potentially high physical/virtual substitution. Emerging virtual technologies in education are able to bring about changes that are both complex and inevitable in the delivery of tertiary education. These changes will lead to greater diversity in tertiary learning, and are likely to induce significant structural changes in tertiary education delivery, specifically with regard to the separation of the production and delivery of learning materials, and options for different student lifestyles which are more temporally flexible.

Currently, for students at a metropolitan university, the timing of lectures and the accessibility of campus sites dominates the activity patterns in the daily lives of students, at a time when financial requirements to work put additional pressures on their daily schedules. The development of virtual technologies could allow the modification of spatial and temporal constraints to the benefit of students, and the ability to describe and understand these changes has important practical applications in the context of tertiary learning delivery, as well as more general implications for the geography of tertiary education.

This paper reports on work to model student daily life choices and considers the potential to identify new university structures from this which might offer the best provision of 'accessibility' to a geographically widespread student body. It concludes by reviewing the concepts of Time Geography, sequencing and substitution in the light of our experience with them in the domain of tertiary learning.

The advent of information technology (IT) prompts a reassessment of the conceptualization and measurement of accessibility. I argue here that traditional concepts add measures of accessibility need to be rethought, and their limits acknowledged before we attempt to develop new accessibility measures that incorporate IT. Simply trying to evolve new, IT-relevant access measures uncritically from traditional measures will not engage the most difficult and important access problems and is unlikely to yield useful outcomes.

Traditional measures of accessibility focus on the location of potential destinations relative to an origin zone (for aggregate measures) or to an individual's home location (for individual- or household-level measures). In this sense potential destinations (whether workplaces, stores, hospitals, or parks) are essentially considered as an opportunity surface, centered on an origin. Missing from this concept/measure of access, even before we begin to consider the impact of IT on accessibility, are many elements that are essential to accessibility. First is the ability or willingness of those at the origin to travel. Some access measures begin to handle this by creating separate measures for people with and without a car, for example, but the assumption remains that those within each group experience the same friction of distance. Second, traditional measures of accessibility do not consider time-space constraints on the mobility of individuals; a partial exception is the space-time prism of time-geography, but these were never operationalized for large samples. A third essential element missing from traditional access measures is the role of personal contacts (social networks) in opening or constraining avenues of access. Each individual at the origin is embedded in a network of social interactions that to a greater or lesser degree provide information about, and ease of entry to, potential destinations. Understanding the relationship between an individual and her/his social networks is therefore crucially important if we are to understand accessibility. For all these reasons, simply considering the location of the individual vis a vis a set of potential destinations is inadequate; it is necessary to consider the multiple dimensionality of accessibility as it is situated in everyday life.

In reconceptualizing accessibility, I am particularly interested in using access to paid employment as an example precisely because employment relations are embedded in everyday life, including power dynamics within households, gender ideologies and identities, cultural, norms, daily activity patterns, and social interactions. Access to jobs is often contingent on knowing someone who can connect you with a particular employer, in this sense one's job history is important in shaping access to future employment, not in the form of a resume, but in the form of a contact field of previous co-workers. Recognizing the social, cultural, and political dimensions of accessibility-not just the geographic ones-acknowledges the contested nature of access. It also opens the way for assessing the impact of IT on accessibility in a far more complicated fashion than simply asking how IT changes one's relationship to distance.

As everyday life increasingly includes the Internet and other forms of IT (cellular phones, Intelligent Transportation Systems, seamless computing interfaces), how will accessibility in its multiple dimensions be changed? How, for example, does IT affect people's willingness to travel (are women more willing to drive long distances with cellular phones, which can provide access to help in emergencies)? What are the interactions between IT and time-space constraints (with parents and children toting phones and wearing pagers, are household members more readily able to share in running errands at the end of the day)? And what is the role of IT in building, weakening, reinforcing, and undermining webs of social interactions and personal connections? How does IT intersect with face-to-face communication? To what extent can IT help to bridge social worlds that presently have little access to each other? We know that distance is frequently used strategically to prevent or inhibit access to and interaction with certain groups (as in the case of white flight from the city or the Sears headquarters move to exurban Chicago to shed its Black labor force). Can IT help to overcome these geographic schisms? This paper will develop a framework for assessing the relationship between accessibility (concept and measures) and IT. In particular, I shall consider how representations can highlight patterns of lack of access independently of the lack of interaction.

This project conceptualizes and constructs a "personal real-time accessibility index" to measure global remote personal real-time accessibility with special emphasis on the role of daily activity patterns. By personal real-time accessibility is meant availability for direct, immediate responsive interaction among individuals. It is anticipated that the index will vary both spatially and temporally. The potential for interpersonal temporal accessibility in some locations will greatly exceed the potential in other locations. Additionally, for any given location potential will vary over the diurnal cycle as distant locations and peoples become more or less accessible.

Section one examines briefly the nature and significance of interpersonal interaction (social contact), the significance of various types of interaction and the rationale for focusing on personal real-time interaction. Communication can require temporal coincidence or not and geographical coincidence or not (Janelle, 1986). The focus here is on temporal coincidence. Additionally, communication can be formal or informal. Both pose problems. However, informal conversations tend to require greater co-ordination to achieve co-presence and they require regenerating context because of time lags and intervening activity between intermittent interactions (Whittaker, Frohlich and Daly-Jones, 1994). This suggests that the highest potential for informal communication would occur in a routine temporal-behavioral topography that provides easy access.

Section two will identify and examine factors that can be expected to promote or impinge on real-time interpersonal temporal accessibility. Activity constraints (Hagerstrand, 1970) and activity settings (Harvey, 1997) will provide the organizing focus. Hagerstrand posits three sets of constraints, capability, coupling and authority constraints. Each set can be expected to play a role in determining global remote personal real-time accessibility. Technologically, increasing availability and diminishing cost of communications technology is enhancing global personal contact capability via the Internet/ World Wide Web. Currently it is estimated that it connects some 68.69 million persons and that this number may exceed 100 million by the year 2001 (Headcount, 1998). However, the actual potential for real-time personal interaction at any given moment represents a very small subset of connected individuals.

The pool of persons accessible in real-time is diminished by a variety of factors. Firstly, it is argued that equal technological access across space does not exist (Salomon, 1988). There is geographic variability both in the availability and quality of access. Secondly, language differences present a major barrier to interpersonal real-time communication. Thirdly, even where the technological and language barriers are absent, temporal availability represents a major impediment to interpersonal real-time communication. The diurnal routine as it is played out in a dispersed geographic plane provides varying levels of remote real-time access. In terms of Hagerstrand a behavioral coupling constraint comes into play. Individual temporal reach and reachability influence individual access and availability.

Time-use data provides ample evidence that behaviour tends to be structured temporally (Goodchild and Janelle, 1984;Harvey et al.; 1997). Subjectively it has been found that much behaviour is routine (Cullen and Godson, 1975). Given these realities it is possible to develop an index that will reflect spatial-temporal differences in real-time accessibility. Diurnal accessibility is defined in terms of activities and activity settings that preclude or facilitate interaction.

The study utilizes macro time-space data from Canada, several European countries and Japan. Research has shown temporal variation in the city’s social geography due to activity patterns and settings as well as to traditional factors (Goodchild and Janelle, 1984; Janelle, Klinkenberg and Goodchild, 1998). While there is an extreme dearth of time-space data, time-diary data exists for a large and growing number of countries. While most of the existing data lacks detailed geographical coordinates it does provide some spatial detail, defined in terms of generic locations and time zones, with respect to the location of activities. For purposes of the proposed study the geographical dimension of major relevance is time zones. However, north-south and urban-rural orientations may also be significant. In the available data individuals can be minimally identified by time zone, country and generic location (home, workplace, other place or travelling). Generic location would be a major factor in the type of contact made (work to work, home to home, work to home, home to work). The index developed would, in its simplest form, be the effective number of persons reachable from a given location at any given time. Elaboration could be made in terms of role group (i.e. worker, student, other); type of contact and possibly other dimensions.

One output will be a conceptual plan for a GIS database capturing the temporal-spatial dimensions of accessibility. Canadian, European and Japanese data will be used to develop, as a basis for discussion, a preliminary demonstration of its potential.

Cullen, I. and Godson, V. (1975). Urban networks:The structure of activity patterns. In Diamond D. and McLoughlin J.B. Progress in Planning 4(1).

Goodchild, M.F. and Janelle, D.G. (1984). The city around the clock: space-time patterns of urban ecological structure. Environment and Planning A 16:807-820.

Hagerstrand, T. (1970). What about people in regional science? Papers and proceedings of the Regional Science Association, 24:7-24.

Harvey, Andrew S., Taylor, Maria Elana , Ellis, Scott and Aas, Dagfinn (1997). 24-Hour Society and Passenger Travel. Report Commissioned by Transport Research Centre, Ministry of Transport, Publc Works and Water Management. Rotterdam.

Harvey, Andrew S. (1997). From Activities to Activity Settings. In Dick Ettema and Harry Timmermans (co-eds.), Activity-based approaches to travel analysis. Tarrytown, New York: Pergamon.

Headcount (1998).

Janelle, D.G., Klinkenberg, Brian and Goodchild, M.F. (1998). The temporal ordering of urban space and daily activity patterns for population role groups. Geographical Systems.

Janelle, D.G. (1986). Metropolitan expansion and the communications-transportation trade-off. In Hanson, S. The geography of urban transportation. Guildford Press, pp. 357-387.

Salomon, Ilan (1988). Geographical variations in telecommunications systems:the implications for location of activities. Transportation 14:311-327.

Whittaker. Steve; Frohlich, David and Dalt-Jones, Owen (ND). Informal workplace communication:What is it like and how might we support it? Mimeo.

My paper will argue that the mathematics of fuzzy set theory can operationalize the concept of accessibility in a way that encompasses both its geographic and virtual dimensions. As discussed below, fuzzy set theory is a branch of mathematics introduced by Zadeh (1965) that recognizes degrees of memberships in sets that we shall call "clubs", in reference to a branch of the public goods literature. Club theory argues that individuals voluntarily form groups or clusters in order to derive tangible or intangible benefits through mutual association. Cornes and Sandler (1986) summarize this literature well, tracing its antecedents through Buchanan (1965) and Tiebout (1956).

The proposition I will advance builds on recent papers by Roger Bolton, myself, and others:

• Bolton (1997) establishes a compelling correspondence between places and networks, arguing that, "Recent theory on communications networks, from microeconomics, geography, and sociology, has relevance to planning, because one can model a well functioning local community, especially one with a strong sense of place, as a network of social interactions."

• Heikkila (1996), in turn, demonstrates empirically that municipalities (or "places") within complex urban settings such as Los Angeles can be usefully understood as "clubs". ANOVA tests for sixteen broadly based socioeconomic indicators reveals that a "club-like" clustering within municipal boundaries explains a substantial proportion of the overall spatial variation in those indicators.

• Richardson et al (1990) and Heikkila et al (1988), in their empirical studies of residential land values in Los Angeles group their explanatory variables in three sets, corresponding to (i) property characteristics, (ii) neighborhood attributes, and (iii) geographic accessibility. The first two components account for ninety percent of the overall spatial variance in land values.

What is called for, then, is a more general theory of membership association that encompasses geographic proximity, transactions-based interactions, socio-political influence and other dimensions of association. From this perspective, the relevant question is not so much how to adapt geocentric concepts such as spatial interaction and accessibility to a virtual context. Instead, we may be better advised to ask whether membership association can be defined in a sufficiently robust way that it may be extended to encompass both geographic and virtual domains.

Fuzzy set theory may be a logical place to begin this inquiry. Introduced by Zadeh (1965), fuzzy set theory allows for degrees of membership in so-called fuzzy sets, where degree of membership is defined by a membership function. According to John (1996), "For any set A the function m_A indicates the degree of membership that x, of the universal set X, belongs to set A, and is usually expressed as a number between 0 and 1:

m_A(x): X --> [0,1] "

Consider, for example, how this concept may apply to the traditional geographic concept of an urban rent gradient. Instead of conceiving of a particular parcel of land as being "accessible" to the central business district, we may instead think of that parcel as having a certain degree of membership (as defined by m_A) in the CBD "fuzzy node". By an extension of the same line of reasoning, the entire urban space can be conceived of as the union of a set of overlapping fuzzy urban nodes. Like clubs, these nodes need not be geographic in nature.

It would appear that fuzzy set theory has considerable potential for integrating geographic, social, virtual, and other elements of accessibility or association. Fuzzy set theory provides an array of logical operators that can be used to define the union and intersection of fuzzy sets. Geographic space is not lost in this formulation, however, it is operational primarily within the confines of the membership function, m_A. From this perspective the advent of communication technologies may alter the membership structure of certain fuzzy nodes or clubs, but not in a way that impacts the logical or analytical foundations by which we may understand the relationships between clubs. It is this perspective that I would like to pursue for the Varenius project.

Bolton, Roger (1997) ‘Place’ as ‘Network’: Applications of Network Theory to Local Communities; paper submitted for Volume in Honor of T.R. Lakshmanan; Department of Economics, Williams College, Williamstown, MA 01267; December.

Buchanan, J. M. (1965) An Economic Theory of Clubs, Economica, vol. 32, 1-14.

Cornes and Sandler (1986) The Theory of Externalities, Public Goods, and Club Goods; Cambridge University Press.

Heikkila, Eric J. (1996) Are Municipalities Tieboutian Clubs?; Regional Science and Urban Economics, vol. 26, 203-226.

Heikkila, Eric J. et al (1988) What Happened to the CBD-Distance Gradient? Land Values in a Polycentric City; Environment and Planning A, vol. 21, 221-232.

John, Robert I. (1995) Fuzzy Inferencing Systems: Problems and Some Solutions; CMS Working Paper No. 62; School of Computing Science, De Montfort University, The Gateway, Leicester, LE1 9BH, U.K.

Richardson, Harry W. et al (1990) Residential Property Values, the CBD and Multiple Nodes: Further Analysis; Environment and Planning A, vol. 22, 829-833

Tiebout, Charles (1956) A Pure Theory of Local Expenditures, Journal of Political Economy, vol. 64, 416-24.

Zadeh, L.A. (1965) Fuzzy Sets, Information and Control, vol. 8, 338-353.
 

In the physical world, the problem of accessibility is basically a problem of overcoming the impedance of physical separation between locations of demand and supply. Accessibility in physical space therefore depends largely on the spatial distribution of urban opportunities, available means of transport, and travel mobility. Its foundation is the place- or location-boundedness of opportunities where access to facilities and services is predicated on meeting the space-time co-location and co-presence requirements for spatial interaction. With distance between locations as the major impedance of movement, and with a given space-time distribution of opportunities in the urban environment, individual accessibility is specifiable in terms of these fundamental elements. In such an environment, considerations of the geometry or topology of the transportation network, and the space-time constraints faced by individuals in their everyday lives are crucial for evaluating accessibility.

In the virtual world enabled and created by telecommunication technologies and the Internet, however, the nature of access to opportunities or resources is drastically different from that in the physical world. In cyberspace, access to resources and interaction between different individuals are media-based (that is, depending on the communication resources available rather than the time or costs necessary for overcoming physical separation). Except in the cases where the cost of access or interaction still depends on the physical separation of locations (e.g. phone charges), distance between locations or individuals has virtually no effect on access and spatial interaction in cyberspace. Further, there are reasons to believe that access in cyberspace is increasingly characterized by such a de-coupling of distance-dependent access cost and physical separation between locations.

If physical separation between locations is playing a less important role in determining access and spatial interaction (as in the cases where the Internet is used for a wide variety of activities), many fundamental determinants of accessibility in conventional models are no longer important. For instance, the "information superhighway" does not have many similarities with conventional transportation networks. Since locations are connected through worldwide computer networks which enable multiple access paths and operate on optical fibers at electronic speed, the effect of physical separation between locations and topology of the transportation network is obliterated. Further, since tele-communication technologies and the Internet provide various means for moderating the space-time constraints of many activities (i.e. space-time adjustments), the space-time co-location and co-presence requirements for access or spatial interaction for many activities are also obliterated (e.g. voice mail, electronic bulletin boards, the World Wide Web). It has now become possible for one individual to be at several (cyber)locations at the same time, and thus violating the constraint that "one individual cannot exist in two places at one time" (Carlstein 1982), This means that the role of space-time constraints in the conventional sense may also become less important in cyberspace. The spatial-time configuration of resources or opportunities in cyberspace is therefore drastically different from what are available in the physical world.

With these complexities introduced by "cyber-accessibility", the problem of accessibility in the new hybrid physical-virtual world is far more complex and difficult to deal with. All of these suggest that conventional models of accessibility are inadequate for representing and measuring individual accessibility in such new hybrid spaces. Further, the disjuncture between accessibility in the physical and virtual worlds requires new conceptual and analytical models since we are dealing with two drastically different realms and their interface.

To overcome the limitations of conventional accessibility concepts and measures for representing and measuring "hybrid-accessibility" in the new physical-virtual space, I propose to conceptualize individual accessibility as consisting of two components: accessibility in the physical world and accessibility in cyberspace. Any measures of individual hybrid-accessibility therefore need to incorporate the effect of three elements: (a) access in the physical world; (b) access in cyberspace; and (c) the effect on accessibility benefit due to the interaction between activities in these two realms (for example, substitution and displacement).

My approach will proceed first from a multi-scale three-dimensional representation of the individual space-time path in hybrid space using GIS. As in the case of Hagerstrand’s original work, this visualization exercise will be helpful for identifying potentially useful constructs for evaluating hybrid-accessibility (perhaps, constructs similar to the space-time prism). As hybrid-accessibility involves opportunities and resources located at various spatial scale relative to the home location of an individual, a multi-scale or multi-level representation is needed to handle the spatial extent of accessible locations. Based on this initial visualization exercise, I will attempt to identify the important elements and parameters for formal specifications of hybrid-accessibility measures. Operational access measures may then be formulated using GIS methods (perhaps, following the logic of the feasible opportunity set specified in Kwan and Hong 1998). The measure derived will be individual-based and person-specific, and will be useful for detecting patterns of differential access to resources in the hybrid space of the physical-virtual world for individuals of various gender/ethnic subgroups.

The nature of modern societies produces a myriad of evolving relations between differing spatial locations. Understanding these complex interactions and their evolving organizational patterns are fundamental to the discipline of geography. The ability to measure, identify and explain patterns in these interactions allows us to both better understand individual behavior and provide for the pro-active design of infrastructures that allow communities to grow.

The effective analysis of complex spatial and temporal interactions has traditionally been hampered by several factors including the lack of detailed data representing the these interactions, the inherent spatial-temporal c omplexity of the processes, and a lack of tools necessary to support effective analysis. Recent technological advances have provided some relief to the barrier of data availability. Simultaneously, however, these technological advances are continuously altering the very patterns of spatial organization and interaction that we wish to understand. The steady proliferation of networked information communities across geographically dispersed locations has engendered a breakdown in the historical barriers of organizational hierarchies, in professional and industry groups, promoting a remarkable shift from centralized toward distributed architectures of network communications. These technological advances provide clear indications that both detailed data representing spatial interactions will become increasingly available and that the corresponding spatial interactions will become increasingly complex. These indications emphasize the growing requirements for both formal models of spatial and temporal relationships and effective analytical tools to help in understanding the processes.

Formal models regarding measurements of interactions and accessibility are traditionally based on physical notions of distance and connectivity. In this research, however, interaction and accessibility are represented in terms of their relative nature rather than on physical distances. This research proposes a framework for better understanding interactions predicated by concepts in exploratory data analysis and the representation of interactions in terms of relative space and time. Relative space and time, in this context of this research, are defined as a flexible, topological structure in which the objects are defined in terms of their corresponding relationships. This framework is instantiated in the design of tools to support visualization, analysis and effective identification of patterns of relative interactions. The utility of these concepts and corresponding tools will be shown in the exploratory analysis and visualization of how users traverse information spaces. In particular, how users travel over time searching and browsing for information onthe web. Data reflecting user traversal paths in conjunction with the harvested topology of the web are used to visualize user navigationpatterns over time. These patterns reflect the individuals explored regions in information landscapes. It is the anticipated hope that a foundation based on the visual exploration of relative spatial interactions provides the opportunity for gaining new insight into factors influencing spatial interactions and accessibility
 

Since 1993, when the first graphical web browser, Mosaic, was released into the public domain, the Internet has evolved from an obscure academic and military research network into an international agglomeration of public, private, local, and international telecommunications systems. Much of the academic and popular literature has emphasized the "distance-shrinking" implications and "placelessness" inherent in these rapidly developing networks. However, the relationship between the physical and political geography of cities and regions and the virtual (or logical) geography of the Internet lacks a strong body of empirical evidence upon which to base such speculation. In an attempt to bridge this gap, we will approach this subject by exploring the following questions; 1) is the Internet an urban or metropolitan phenomenon, or does it truly represent "the death of distance"?, 2) how does this rapidly growing and increasingly important telecommunications medium influence spatial patterns of urban development - the relative advantage of cities to each other and the city-hinterland relationship, and 3) what historical, socioeconomic, and political characteristics decisions play a role in determining these variations in telecommunications development.

This paper builds upon the results of several studies we have conducted over the last two years, which indicate a strongly urban and overwhelmingly metropolitan dominance of Internet development by a handful of cities and regions. We identify and describe an emerging structure of "virtual" hubs and pathways which are inter-connecting a set of major cities in the United States, suggesting that there is a new urban network that goes beyond Castells’ "informational mode of development".

More importantly, this paper provides an empirical approach to measuring inter-urban and inter-regional variations in Internet utilization, diffusion, and accessibility. As the telecommunications industry becomes increasingly competitive, the difficulties in obtaining timely and accurate data will become more pronounced. We hope that our experience identifying consistent, credible, and cooperative sources for data and formulating theoretical arguments for their application to urban research will be useful to other scholars.

Specifically, we analyze the utility and relevance of three measurements of Internet infrastructure for urban research, geography, and public policy. The first measure is the number of host computers connected to the Internet on a full-time basis, based on data collected by Matrix Information and Demography Services of Austin, Texas. We find this measure to be of limited usefulness, as it only measures location-specific hardware installations with little reference to their purpose or function. The second measurement we employ is based on the Internet’s addressing scheme, known as the domain name system. Unique to individual organizations (business, education, non-profit, government), domain names are registered with InterNIC, an administrative clearinghouse contracted by the National Science Foundation. Associated with each domain name is a unique billing address, which permits the localization of the organization using that name. We have found this measure to be the most informative for assessing variations in Internet utilization across a variety of geographic units, from states to individual ZIP codes. The third and final measurement discussed is the capacity of inter-metropolitan "backbone" networks. While the host count measurement describes a completely physical and real phenomenon, and the domain count measurement a thoroughly virtual and logical one, the backbone capacity measurement is a hybrid. Although, some networks operate on isolated fiber optic cables, many are "virtual networks" operated over lines leased from national and regional telephone companies. Finally, we analyze the work of others who have examined geographic patterns of Internet access and utilization.

Based on these three measurements, we find that a small number of cities and metropolitan areas dominate the rapidly emerging telecommunications landscape of the United States, leading the development of increasingly sophisticated applications and technologies. Accessibility to the most advanced and robust real and virtual Internet infrastructure is a metropolitan phenomenon, and highly stratified among regions and cities. Furthermore, we describe each of the three measurements utilized - host counts, domain counts, and backbone network capacity - and their unique advantages and disadvantages for research. Successfully applying them to urban analysis requires understanding not just the fundamental principles of geography and urban dynamics, but also the function, purpose, and design of the array of technical systems of which they are a part and those from which they have evolved.

Practitioners and researchers alike increasingly recognize that the emerging division of "haves" and "have-nots" is no longer between geographically first and third or fourth "worlds" divided according to degree of modernization-industrialization, but instead between groups with and without access to the new information technologies and the power these bring. These new "dual societies" often are found side by side geographically: in Washington, D.C. as well as in Mexico City. (e.g. Castells, Mollenkopf, Robson, 1998; Sanyal, 1997). In addition, much of the enthusiasm for the spread of the latest technology comes from vendors and service providers who see opportunities for new sales to government and non-profit agencies, but who have little concern for the final result of the projects in the daily lives of the disadvantaged poor, homeless, veterans, migrant farmworkers, et al who are "provided access" (Forsher, 1998). In this regard, even the well-meaning concepts behind the Varenius Project need to be critically examined: to consider the

needs and possible aid in terms of "concepts," that "reconceptualize, measure, represent, monitor, and plan for the new emergent geographies" almost inevitably casts the project in the very terms of the dominant "imperialistic" educational process in difference from which already marginalized groups constitute their identity and perversely requires these "have-nots" -- the learners--unconsciously to conform and internalize the "normative" concepts, maps, and images of the dominating groups-- the teachers and fund-providers (Freire, 1970 & 1974; Heidegger, 1961). In its current terms, it would appear that the project is reconceptualizing the issue in the very rationalistic terms that will perpetuate the inequality of accessibility opportunities, insofar as the latter have any substantial economic or political force.

In contrast, as argued elsewhere, adequate and democratically appropriate pedagogy, reciprocal giver-receiver relationships, and autonomy of/responsibility for one's life-world can only result from the practice where "teachers" and technology allow the learners to pass over into the dominant realm, then successfully pass back into their own world too. This alternative is not only possible, but already a proven model in many areas of the world. It may be implemented in several democratic, non-ideological way which provide means for autonomous groups to name their own world (Freire; Donne, 1970; Gadamer, 1972; Mugerauer, 1995).

In the realm of GIS, an adequate alternative technology would provide methods and equipment/systems that allow groups to non- representationally articulate their worlds, by unfolding them in the process of building and using qualitative GIS systems of their own, and then the interfaces with standardize, dominate quantitatively-based systems. As an outline (only) of such a process: the groups to be helped record images of the dimensions important to them from their worlds, critically examines the connections among these, tell the stories of how their worlds work with these active dimensions, then from the resulting images and wordings map out the spatial extensions experienced among these, perhaps competing, dimensions. In the process it is not assumed that the mapping of (to beg the question of what is actually important to them) the groups' own politically, religiously, culturally, and pragmatically important spaces are the same as, or even use the same logic as, those of the West's analytic, quantitatively generated GIS representations. Once the groups have named, visualized, and brought to critical consciousness their own worlds, then these can be systematically compared to the standardized GIS images in a process in which the groups themselves are initially able to choose how to evaluate and react to differences and similarities. In short: the outcome of such a mediating pedagogy and technological exchange is that the learning groups are able, on their own, to choose how to interact politically, economically, and technologically with the dominant societies; and, the latter can better understand and decide whether to contribute to the goals of the previously marginalized groups.

The current proposal follows up on the work described in the Geographical Analysis paper. The full project is expected to run in four key stages. Stage I has been successfully completed. Stages Ib and II of the project would be carried out in preparation for the meeting in November, with the results distributed according to the stated time-table. The final two stages are still to be defined in detail since they depend on the success of stage II, but would involve the development of Internet census information and the fitting of spatial interaction models to WWW information flows.

Stage I: Initial exploration of information and geographical space.

Questions: Is there space in cyberspace? If space exists how do we measure it in a way that is relevant to Internet information transactions? Are tools currently available that we might use to measure cyberspace or must we develop new tools' If tools exist what, limitations or obstacles prohibit their use? Does this space influence the way information is accessed on the WWW? Are WWW based services regional or global in nature? How might spatial variations in cyberspace effect e-commerce for developing (Internet) nations?

Results: Stage Ia of the project is completed. The results will appear in an article in Geographical Analysis Journal and also as a chapter in a new book GeoComputation.

Stage Ib: Quality of service vs. service footprint

Questions: The majority of information delivered over the WWW consists of low volume static HTML documents with perhaps a few inline images. The results of the previous stage show that the number of users visiting an on-line service diminishes with Internet distance. On-line services are gradually moving to higher quality output with associated higher volumes e.g. software applications (including on-line GIS), video services, shockwave animations etc. If the volume of information delivery necessary to use a service increases does the use of these services in distant regions decrease? Will increasing output qualify result in the contraction of service catchment areas?

Methods: Part of the information database built up in stage I included information on the volumes in bytes delivered to a large number of locations from University Web Servers in the UK. By analysing the flow of information to other regions we should be able to provide answers to these questions.

Stage II: Relating cvberspace to geographical Space.

Questions: How does geographical space relate to cyberspace? Can we map locations in cyberspace space to geographical locations and vice-versa?

Methods: I propose to attempt to locate individual Internet addresses in cyberspace using simple triangulation techniques. A number of distance measuring stations will be set up, distributed around the Internet. The distance between a location in cyberspace and each of these stations will be measured simultaneously using methods developed in stage I. Each of the stations will also monitor the cyberspace distance from their location to the other monitoring stations. In this way it should be possible to allow for temporal fluctuations in the network latencies. Using three or more stations it should be possible to define the unique location in cyberspace of an Internet address. By measuring the position of a large number of internet addresses for which geographical locations are available then it may be possible using neural computing techniques to develop a model that can map cyberspace locations as defined by the monitoring stations to geographical locations and vice verse.

If this methodology is successful it should be possible to monitor the temporal changes in cyberspace and its changing relationship with geographical space. Analysis of short-term variations throughout the day may reveal interesting patterns of WWW usage in different parts of the globe and might provide insights into the daily cycles of Internet use. Long term trends may reveal how the quality of the Internet is developing worldwide. Of particular interest is whether or not the disparity between developing and developed nations is increasing or decreasing with time. An automatic gazetteer of this type should also allow us to provide reasonably detailed maps of the density of Internet locations and also of the extents of the current and future Internet domains. Most of the measurement and analysis for stage II will be carried out by the Department of Geography at the University of Portsmouth, however a number of collaborating sites will contribute facilities and resources to the project. These institutions include the NCGIA at UCSB (Mike Goodchild), the Curtin University of Technology in Western Australia (Mark Gahegan), and the Departmentof Geography at the University of Helsinki (Tom Blom).

References:

Murnion S. (1998).Cyber-Spatial Analysis: Appropriate methods and metrics for a new geography. In GeoComputation, edited by S. Openshaw and R. Abrahart, London, Gordon and Breach, Wi11 be published within the next few months.

Murnion, S, and Munroe G.(1994) GIS and the World Wide Web Information Resources for GIS Users. GIS Europe, March 1994.

Murnion S. (1996). Navigating the GTS Web. Geolnfo Systems, August 1996.

In order to cope with the kind of time-space changes occurring in modern cities, we need to extend our notion of accessibility. On a substantive ground, a major extension which have been suggested relates to the new role of the temporal dimension resulting from the impact of the New Information Technologies. As a result, the concept of accessibility is endowed of new potentialities, i.e., accessibility is not simply a time-space opportunity, but also a resource. This requires to broaden the conventional perspective of analysis of accessibility.

In this paper an effort is made to identify some major aspects involved in extending the notion of accessibility. One major focus of the discussion is to show how, within the analytical framework of a formalising process (i.e., in the encoding/decoding process of a modelling activity), the suggested extensions would also affect the definition of policy measures for 'accessibility improvement'.

The paper is divided in two parts. In the first part of the paper we address some main questions which are raised by considering the impact of the New Information Technologies on the spatio-temporal dimension of accessibility, i.e., a) the need to recognise the various ‘field of definition’ of accessibility; b) the necessity of setting up an evaluation framework against which assessing accessibility changes (i.e. the increasing urban interactions relative to their effects in term of the sustainability issue, and quality of life); c) the relationships between accessibility, mobility and urban form (the spatial distribution of activities and pattern of land use) and d) the representation of accessibility and the recognition that the kind of 'knowledge' individuals and decision-makers have about accessibility is a fundamental determinant in its provision.

In the second part of the paper we outline the research project on accessibility which Ires is carrying out in the Turin Metropolitan area. The main idea at the basis of the project is that extending the consolidated vision of accessibility puts new claims on the notions of accessibility improvements as traditionally conceived in urban policies, at least in the Italian context. Besides illustrating the general aims of the project, the questionnaire used in the survey will also be described. This latter tries to take into account some aspects of the proposed extensions of accessibility. In particular, two main levels of definition of accessibility are considered: 1) a systemic level, which reflects the functional and spatial organisation of activities (mix of opportunities, travel times, transport services, opening times of services, etc.) observed at an aggregate level (i.e. as provided by the various transport agencies and service authorities); 2) an individual level, which considers the action-space within which an individual (i.e. a person belonging to an household or collective actor belonging to an organisation) currently lives.
 

The balance in interests supported by current intellectual property laws in the US has made conventional libraries highly successful and valued social institutions. That balance has been disturbed by the shift to electronic publication and dissemination. The ongoing technological shift has placed the interests of the scientific and academic communities, entrepreneurial businesses, and the general public at a distinct disadvantage relative to the interests of publishers and large corporations with the ability to amass intellectual content. In order to support both fairness in access to data sets among all sectors of US society and continued economic vitality for the nation, the imbalance needs to be addressed. Solutions are not simple.

In the traditional publishing world, authors and publishers receive a level of protection which is sufficient to provide incentives for them to make their works available for purchase by individuals and libraries. Libraries, of course, lend out the books, maps, magazines, films and other intellectual works they purchase. Any man, woman or child is typically free to borrow material from the library without paying a per usage fee and use the material for whatever purpose they see fit. No inquiry is made regarding their intended use. Upon returning the material, no record of who borrowed the material is maintained. The anonymity provided by the purposeful failure of the library to ask about use and failure to track the identity of users promotes free exploration of ideas and concepts without fear that one's exploration may be viewed as socially or politically unacceptable.

The ability to freely browse, study, explore, and borrow library materials provides a highly valuable common good upon which anyone may draw knowledge for any purpose. Communities financially support public libraries because all in the community may use the works contained in the library for educational, scientific, business, entertainment and other socially constructive purposes. After a library has purchased an intellectual work such as a book or map or has been given a copy that was legally obtained, neither the author nor the publisher can prevent the work from being placed in a library and lent. We do not support public libraries in order to reward authors and publishers. Indeed, many may object to their works being lent by libraries. Thus, the goal of copyright and the library as an institution has been to strike a balance between giving authors sufficient incentive to make their works available and supporting the rights of users to use the intellectual works of others for socially constructive purposes. A similar balance of interests has not yet been achieved in the on-line world.

Many in the digital library research community as well as the geographic database community appear to believe that the marketplace will "provide the answer" in resolving the relationships between public access and private rights in data sets. In this marketplace approach the library patron is treated primarily as a consumer with relationships primarily controlled through contract law principles. In acting on our own behalf we often do act as consumers. However, in acting on behalf of others we act as citizens (Sagoff 1992, 373). As citizens we decide through representative government "the right thing to do" rather than allowing economics solely to decide our fate. Our relationships are defined by responsibility as well as by efficiency (Gilroy and Wade 1992, 519) Considerations such as equity, justice and fairness come into play through the political process. We admittedly live in an imperfect and unfair world. However, depending on economics and freedom to contract as the sole or even the primary means for ensuring access to knowledge in future digital library environments will exacerbate current inequities, not lessen them.

This paper explains why we do not have on-line the equivalent of a traditional library from a legal rights perspective and why the public commons in intellectual property traditionally provided by libraries is in jeopardy for both the scientific community and the community at large. The evolving social situation is primarily will want access to bodies of intellectual work from future digital libraries as opposed to "correct" or "reliable" answers automatically fed to them from the system (Murr and Williams 19xx). Some of the options to correct the current imbalance in private versus public rights in digital library works are explored and the argument is made that a pressing need exists to explore a full range of economic, legal, and institutional models and prototypes for geolibraries. The research community's goal should be to develop models that will support the public goods benefits of traditional libraries while providing sufficient incentives for private individuals, private publishers, and government publishers to make their spatial data products available through networked library environments.

Note: The finished article will draw from and expand upon short opinion pieces submitted to Proceedings of the Fédération Internationale des Geometres (FIG), July, 1998, Brighton, U.K. under the title "Balancing Intellectual Property Rights and Public Goods Interests in Geolibraries" and submitted to the Distributed Geolibraries Meeting, June 15-16, 1998, Washington, DC under the title "Producer Rights and Public Rights in Spatial Data Sets in Geolibraries."

Explosive suburban employment growth, declining residential densities, economic restructuring, and rapid changes in telecommunications and transportation technologies have fundamentally altered the spatial and organizational composition of where we work and live. How are these broad spatial processes impacting intra-metropolitan accessibility? How are the impacts expressed physically on the urban landscape? Who, in terms of both geographic location and socioeconomic groups, is affected? What are the implications for urban development and planning policy?

A first step in answering these questions is identification of an effective approach for modeling intra-metropolitan accessibility. This proposal addresses the Varenius call for participation in the specialist meeting on "Measuring and Representing Accessibility in the Information Age" by (1) examining definitional issues of accessibility, (2) identifying methodological difficulties associated with measuring accessibility, (3) sketching additional challenges posed by both evolving technologies and activity patterns, (4) outlining fundamental requirements for an effective aggregate measure of accessibility, and (5) proposing a new approach to measuring intra-metropolitan accessibility which utilizes the Getis/Ord Gi* statistical framework within a GIS environment.

A large body of formal urban theory contends accessibility is at the very core of processes shaping urban spatial structure. Concepts of accessibility provide the basis for a variety of urban policy and transportation planning decisions; they serve as a common focus for geographic research relating to economic growth, metropolitan form, and social equity. Nonetheless, accessibility indicators are rarely utilized to monitor system performance, construct regional profiles, or evaluate planning strategies. While a large variety of accessibility measures can be found in the research literature, these present an almost equally large variety of definitions (e.g., physical proximity; the freedom of individuals to participate in urban activities; the set of spatial opportunities associated with specific locations).

The proposed research reviews contemporary accessibility literature to determine how accessibility has been defined, measured, and interpreted. It suggests a working definition for accessibility encompassing three broad dimensions: (1) the distribution, intensity, and character of interaction opportunities in both time and space, (2) the integrity and effectiveness of urban transportation and information networks, and (3) the ability and desire of people or firms to overcome spatial separation and/or temporal constraints. Through the literature review process, a number of definitional criteria for effectively measuring intra-metropolitan accessibility are identified. An effective measure: (1) incorporates all three dimensions above, allowing each to influence measure results; (2) presents accessibility as a function of both time and space (as opposed to a single index); and (3) defines accessibility in terms of "level of service": how well a given location serves surrounding populations, for example.

A growing number of researchers propose using GIS for the evaluation and measurement of intra-metropolitan accessibility. The GIS environment, however, promotes distinct conceptualizations of space (absolute, static, "container view") and has not, as yet, been effective with respect to representing time. In an age of rapid technological developments, however, a Newtonian vision of space, time, and distance is limiting. Recent commentary by Sheppard_ regarding site and situation, and a series of articles by Couclelis_ dealing with absolute, relative, and proximal space, provide a basis for extending the representation of space in accessibility measures. Review of this literature supports the following additional criteria. An effective accessibility measure: (1) incorporates attributes of site as well as the complex spatial relationships of situation; (2) allows the notion of "connection" or linkage to extend beyond physical contiguity or distance decay; (3) operates at multiple spatial scales; and (4) reflects a rich and flexible vision of space that may, in fact, change depending on the individuals involved, the context of the spatial interaction undertaken, and/or the mode of access utilized.

In order to address methodological issues associated with implementation, the proposed research applies some commonly used accessibility measures to hypothetical data, evaluating the impact variations in model design have on analysis results. I compare job to worker ratios, spatial interaction models, total travel cost indicators, and the local Gi* statistic (the G statistic framework is similar in structure to widely used isochronic indicators). A number of methodological criteria for effectively measuring accessibility emerge from this exercise. An effective measure: (1) avoids reliance on fixed model parameters; (2) minimizes use of actual travel behavior; (3) facilitates comparison among multiple geographic regions; and (4) utilizes readily obtainable data.

The proposed research utilizes the Gi* statistical framework to implement a measure of intra-metropolitan accessibility to employment opportunities in the Greater Los Angeles region. It demonstrates how this model meets the criteria outlined above, presenting an effective approach to "measuring and representing accessibility in the Information Age".

Accessibility is a key concept for defining the strength and extensiveness of spatial relationships among people, places, and activities. However, virtually all existing accessibility measures incorporate only transportation without considering other means of spatial interaction. It is obvious that these measures are unable to depict the full range of spatial relationships in contemporary society, where digital information and communications technologies are deployed on a massive scale. Therefore, to understand the technologically reconfigured geography and to plan our cities and regions, new analytical approaches must be identified.

In a recent study, I developed some modified gravity accessibility measures and used them to analyze the effect of telecommunications on future evolution of metropolitan spatial structure. I demonstrated that, by carefully examining the complementary relationship between commuting and telecommuting, it is feasible to estimate the average daily commuting time of teleworkers. Consequently, it is feasible to modify the traditional gravity measures to represent employment accessibility in the hybrid space defined jointly by transportation and telecommunications.

This proposed research paper will attempt to find out whether or not the above-mentioned approach to measuring accessibility is generally useful for representing spatial relationships in an information society. The research objective is important because, by achieving it, I will be able to define the scope of application for a pragmatic but potentially valuable method of spatial representation and measurement.

The research methodology (procedure) includes the following five components (steps):

First, opportunities (i.e. jobs, services, and others) in an information society will be categorized, and each category will be further divided into two subcategories: (1) those located in the physical space, and (2) those located in the virtual space. This classification scheme will enable the distribution of accessible opportunities – in both the physical and the virtual spaces – to be defined and quantified.

Second, means of accessing available opportunities will be identified. Conceptually, they can be classified into three groups: (1) transportation, (2) communication, and (3) combination of the two means.

Third, impedance functions will be specified, and accessibility will be measured, for individual subcategories of opportunities and means of access. Obviously, it is quite unrealistic to use travel time to calculate the impedance for communication. Therefore, accessibility will be measured for only some of the basic categories of opportunities and means of access.

Fourth, relationships among individual subcategories will be analyzed. The key question is whether or not it is meaningful to combine each category of accessibility existing in the virtual space with one or more categories of accessibility existing in the physical space. Wherever such combination is meaningful, a hybrid space is defined and an operational measure of accessibility can be derived.

Finally, accessibility measures will be specified at several levels of aggregation (i.e. levels of combination of categories of opportunities and means of access) for each locationally, socially, and economically defined population group.

Not only this research will generate some useful conclusions with regard to the use of a modified traditional approach to defining spatial relationships in the information age, but also it will serve as a practical tool for examining some important social policy issues. I plan to develop a number of scenarios and apply the resulting accessibility measures to the simulations. This simulation exercise will help understand some likely social consequences of the technological reconfiguration of geography.

One interesting scenario is that footloose information-processing jobs will increase substantially. This change in the occupational structure of the economy will cause rapid growth of employment opportunities in the virtual space and corresponding decline of employment opportunities in the physical space. What will the change mean to those people who do not have the necessary education background and technological tool to compete for the increasing number of opportunities in the virtual space? Another likely future phenomenon is rapid growth of internet-based commercial activities. This will imply redistribution of commercial services from the physical space to the virtual space. What will this mean to those people who do not have the necessary technological tool to access the commercial services in the virtual space? A third likely scenario is that teleworkers/teleshoppers will take advantage of the increased level of accessibility and relocate their residence. What will be the resulting spatial pattern of urban growth? What will be the likely social impacts?

These are just a few examples of possible scenarios. The simulation can be run either in the context of a real metropolitan area or simply with an artificially created metropolitan area. In this research, I plan to take the latter approach because it will make the simulation more transparent.

Theoretical speculations on the impacts of the ongoing Internet-led telematic revolution are rampant in the literature, with arguments ranging from the space-time compression/distanciation, the death of distance, to the anywhere-anytime-anything paradigm for various social, economic, and cultural aspects in society. And yet, rarely are these assertions supported by convincing empirical evidences because of the lack of relevant data, on the one hand, and the enormous conceptual complexities of measuring access and accessibility in the information age. My paper for this workshop will report an empirical study on the access and usage of the Internet my colleagues and I have been conducting at Texas A& M during the past three years. Our goal is to examine the spatial-temporal patterns of the Internet access at three different scales: the global (international information flows between different countries), the national (interurban information flows between cities within a country), and the metropolitan (intraurban information flows within a particular city) levels. In this study, we aim to go beyond mere theoretical speculations and groundless assertions on the impact of the Internet by gathering and using as much real-world data as possible to test various arguments reported in the literature. A variety of the Internet access data collected by various corporations and government agencies are used to address the accessibility issues at the global and the national level. Interurban analyses are being conducted using data for the U.S., China, and Indonesia. The metropolitan level (for cities in Houston, Jakarta, and Beijing) analyses are conducted using survey data we collected by ourselves during the past three years. This paper will shed light on all the seven issues raised in the call for papers for this workshop, especially issues 4, 6, and 7.

This paper will also address the implications of a more philosophical issue on measuring the accessibility in the information age, namely the tools we use are increasingly becoming part of what is being studied - a phenomena I call it "the technological self-referentiality." Echoing the similar phenomena in language and mathematical logic, technological self-referentiality is one of the main reasons for many of the paradoxes and difficulties we are experiencing right now in measuring accessibility in the information age. One possible way to transcend this technological self-referentiality is to radically reconceptualize our traditional definition of individuals and cities from a holographic perspective. Instead of viewing the world from a reductionistic Newtonian perspective [similar to looking at a photograph which has a one-to- one correspondence between parts of an object and pixels on the photograph (locality)], the holistic quantum view of reality treats reality as a hologram: any portion of a hologram contains (enfolds) information of the whole object imaged (non-locality). I will use this holographic perspective to resolve some of the existing paradoxes and demonstrate how the results I obtained from the three different scales in this empirical study can be tied together to provide us a better understanding of accessibility in the information age.