Janes, 'And Magellan Thought He Had Problems: "Navigation" in a Network Environment', LIBRES v4n01 (February 28, 1994) URL = http://hegel.lib.ncsu.edu/stacks/serials/libres/libres-v4n01-janes-and.txt LIBRES: Library and Information Science Research Electronic Journal __________________________________________________________________ ISSN 1058-6768 February 28, 1994 Volume 4 Issue 1 Quarterly LIBRE4N1 JANES And Magellan Thought He Had Problems: "Navigation" in a Network Environment Joseph W. Janes and Louis B. Rosenfeld School of Information and Library Studies University of Michigan 304 West Engineering 550 E University Ann Arbor MI 48109-1092 Abstract Navigation is a metaphor commonly used to describe the process of finding information among distributed data resources on the Internet. However, a closer examination of the metaphor reveals both similarities and major differences between navigating the open seas and finding information on the Internet. Because numerous standardized tools and methods are employed to convey Internet users across the network seas, the process of "getting there" is generally less difficult. However, good Internet maps are hard to come by: finding one's destination (e.g., a host address) is a much more formidable task than actually getting there. This paper suggests that "seafaring" is a more appropriate metaphor, as it includes both "exploration" of the Internet, and traveling its known paths. We further suggest the notion of "travel agents", information professionals who serve as guides to cyberspace. Introduction In discussions and writings about wide-area information networks such as the Internet, one often hears the process of working with such networks, searching for information, referred to as "navigation". The metaphor implicit in this reference is that moving through cyberspace is similar to moving through real, physical space: that similar tactics and techniques are used, that similar devices can be used to find one's way around and become familiar with the territory, and that maps or other descriptions of the environment exist. This metaphor is a facile one, and is used loosely, but it does have power and intuitive appeal. When people search through networked resources, many feel as though they are moving through space, so a comparison to a system of maneuvering through unknown space or over an ocean does make some sense. In this paper, we will investigate this metaphor to a somewhat greater depth than we have seen before. We will examine the basic concepts and language of "true" navigation to see how closely it parallels what goes on in cyberspace. Elements of Navigation To get a sense of the rudimentary concepts and language of real navigation, we turn to the article on "Navigation" in the Encyclopedia Britannica (Anderson 199X). Here we find that the central concern of navigation is "the planning and execution of safe, timely, and economical operation of water craft, aircraft, and spacecraft." (p. 750). "The earliest navigators probably learned to steer their ships between distant ports by familiarizing themselves with the sequences of intervening landmarks. Keeping these reference points in view required them to stay quite close to shore." (p. 750). The major concepts we find include: [1] Direction finding These earliest sailors used celestial (the Sun and stars) and meteorological (directions of steady winds) clues to aid in determining which direction was which. These cues are all natural (i.e., the occur in nature), but they are also instinctive and intuitive. Sailing instructions Written instructions (the pilot book or "periplus"), probably collections of oral traditions, arose early (examples over 2400 years old exist) and describe "routes, headlands, landmarks, anchorages, currents, and port entrances". Distance and speed measurements "Distances were cited in the early pilot books in units of a day's sail". Speed measures were obtained by dropping floating objects in the water and seeing how long it took to pass them. Compass "It is not known where or when it was discovered that the lodestone...aligns itself in a north-south direction." Dead reckoning "keeping a running record of the distances and directions traveled". Latitude "The first instruments used at sea for...measurements seem to have been the quadrant and the astrolabe" both of which used the positions of celestial bodies to determine latitude. Longitude This was determined by the use of almanacs and tables; one of the earliest was the Ephemerides of Regiomontanus, published in 1474. Water depth "An Egyptian temple decoration dating from about 1600 BC shows a...member of the crew [measuring this] with a long pole." Distance from a cliff "could be estimated by timing the echoes of shouts or drumbeats." Collision avoidance "A ship underway at night displayed running lights by which sailors on nearby vessels could judge its course and speed." Rules "that specified what lights must be shown, what signals must be given, and how ships must navigate in respect of each other were formulated for British mariners in 1862. These rules formed the basis of the International Regulations for Preventing Collisions at Sea...in 1889." The Internet: Similarities and Differences Most of the concepts discussed above (with the exception of rules) are direct results of the physical or geographic nature of travel across the sea and the requirement to overcome distance and time. Knowing where you are, which direction is which, which direction you need to go, how fast you are going, how deep the water is--these are all absolutely necessary to orient oneself and move purposefully and efficiently through the physical space of the sea. Yet in the cyberspace occupied by information networks, there is no physical space to overcome. Geographic distance is unimportant; one can as easily Telnet to an Internet host thousands of miles away as to one down the hall. There is no sense of "position" in this space, no latitude or longitude lines to locate yourself, no compass or stars to orient you, no sense of depth or distance to use in helping you steer a course. There is no need to overcome the elements or physical distance to travel in cyberspace. You just need to know where to go, give the command, and you're there. And thus it is knowing where to go that is the hard part, not the getting there. If we wish to view use of the Internet in a seafaring mode, then we are currently in its Age of Exploration, searching for new information resources and their locations , all the while improving our mapping technology as our experience grows. (Whether we shall ever leave this Age, given the nature of the Internet, is an open and interesting question.) Thus "navigation" through the Internet is a trifling matter, once you know where your port is. Exploration, though, is much harder, because not only are there few maps to guide the searcher for new resources but because the coastline is constantly changing. New resources appear continually, old ones change or are removed with little or no warning, and certainly with very little fanfare. So a friendly port that you have depended on for some time may one day be gone, and three new sources have been added which would be even more helpful, but you never know about them. The obvious recent example of "mapping" in cyberspace is the dramatic rise of Mosaic as a user-friendly interface to the WorldWideWeb and numerous other Internet tools. It is certainly the case that this easy-to-use interface operates like a map in many ways, pointing users directly to resources and tools. But it is always the case that Mosaic is "better" than its predecessors? We claim not. There is as yet no searching tool comparable to Archie or Veronica (poor as they are) in hypercyberspace. One can move directly to a particular known resource through Mosaic, but only through the entry of a Uniform Resource Locator (URL) which are even more complicated than an Internet address that one could Telnet to. Finally, Mosaic falls victim to the same idiosyncratic design paradigm as Gopher. When a user encounters a Mosaic document, the links incorporated are based on the designer's idea of what should be linked to what and how. The fact that these links are all located on the same page gives the impression that they are somehow related, which may not be how any given user perceives them. Furthermore, what one finds at the other ends of these links is not always apparent or intuitive. It is another case of "Here's how I think things should be organized", which can reduce users' confidence in the utility of both the interface and resources. Exploratory work is also difficult because it can be very hard to predict just where things are located in cyberspace. The concepts of adjacency and context don't really apply. Just because one interesting source on philosophy is located at a given domain or a given institution tells you next to nothing about whether or not there are other good related ones in the same place. In this sense, there are no real neighborhoods on the Internet. Also, there is little consistency in naming these resources, and a tendency to give curious or unique but unhelpful names. To be sure, virtual communities do exist, and they serve as a rich and important aid to exploration. People from around the world can and do gather to discuss or share thoughts on topics of mutual interest. But to find these communities from a naive starting point can be frustrating if not impossible as they could exist within Usenet newsgroups, LISTSERVs, or even private email group names, and again, there is no effective searching tool for discovering them. The process of resource discovery on the net is thus difficult, slow, and laborious. But once a resource is uncovered, the map is created, and travel there becomes trivial, as long as it doesn't vanish or change addresses. One of the concepts of navigation described above does seem to have an analogue in this environment: the rules of sailing. We are no longer concerned about collision navigation, but the guidelines proposed by many people for netiquette and the use of emoticons (such as ;-) ) to convey nonverbal cues through electronic communication are in some sense rules for working in this space. Tools of the Trade The metaphor of navigation, then, is not as powerful as we might have thought, but the generic seafaring notion does seem to still have some appeal, as it incorporates both the notions of exploration and travel, as discussed above. Let us continue this analysis by looking at some of the tools which have been developed to aid in searching, using, and exploring networked resources in this light. Telnet is the ship which we use to carry us across the net. If you know where you're going, Telnet will get you there immediately and with little difficulty. As an exploration tool, it leaves a lot to be desired, since without a specific address to travel to, it is useless. One can imagine attempts to Telnet to random IP addresses in the hope of eventually hitting something, but this would be an extraordinarily bad idea, not to mention frustrating. FTP (File transfer protocol) is the cargo ship of the net. Once an FTP site is uncovered, FTP can be used to download large amounts of data very quickly and bring it home. Once a site has been found, it's not too difficult to explore it to see what's there, although some file names can be somewhat cryptic, but again, though, you've got to know where it is. Archie is a satellite which provides the network seafarer with updates on the contents of various warehouses along the coast. Archie is a popular Internet tool used for searching indexes of filenames which reside in hundreds of FTP sites. The ability to search through a large number of FTP sites for a given character string can be of enormous assistance in looking for resources or software. WAIS (Wide Area Information Servers) are of help in both exploration and transportation. Searching the WAIS server of servers, you can give a query, get back a ranked list of possible places to find an answer and then go there and get the information. Both Archie and WAIS are limited to the resources which are included, but they both cover broad expanses of the Internet Sea and can be very useful to the explorer. Individuality and the Need for Standards Maps are hard to make. They require time, a large number of meticulous measurements, specialized skills, and, if they are of a sufficiently large area, significant resources. In addition, they are hard to maintain, and minute errors in production can render maps entirely useless. In cyberspace, because getting there is a given, the process of making a map is much easier. Provided one has the required equipment and knowledge, after exploration has uncovered a new resource, map- making simply requires telling people what you've found and where it is. This kind of informing goes on in both structured, broadcast ways (via LISTSERVs, Usenet newsgroups) and unstructured, one-on-one ways (email, word of mouth, responses to direct queries). The cartographic profession must now have a sense of what it's like to keep track of what goes on in cyberspace. In the last few years, old countries have disintegrated into new ones, borders have shifted, names have changed. Yet, in the midst of this radical change, there are standards, developed over thousands of years, to help cartographers and map users. [2] North is usually at the top of the map. Dots are used to represent cities. Different forms of lines represents different kinds of roads, rivers, and borders. Most map-makers use these standards, which make the maps easy to use and accessible to most people, once they learn the code. The continual and sometimes rapid dynamism of the network environment leads one to wonder about the development of standards here. Will they arise at all? Will their development be rapid or slow? There are already important standards in place, largely in the technological domain. The TCP/IP standards make communication over the network possible, and each host has a unique host name which does not change. But what about standards of a higher order: standards of resource description, access, interface? We now make an important distinction between two kinds of maps. Up to now, in following the seafaring metaphor, we have discussed hosts and resources in cyberspace as ports of call, and maps as how to get there: IP addresses, file names in FTP sites, etc. Researchers in psychology and related disciplines speak, though, of a different kind of map: a cognitive map, which is how individuals see and perceive the world. Each of us has a different cognitive map of our world. We often give each other rides to the Detroit Metro Airport. When leaving Joe's driveway, you can turn either left or right. Either direction leads you, relatively quickly, to the interstate. When Joe drives, he turns left. That route, working through neighborhood streets, is slightly shorter but less obvious than turning right, which leads immediately to a major road. Lou always takes that route. Although he has lived in Ann Arbor several years longer than has Joe, he does not know that neighborhood as well, and thus does not immediately think of the shorter but more involved route. Yet we could both read the same map of southeastern Ann Arbor and plot the same route. Our cognitive maps give way to "real", standardized representations of this space. Can any such standard work in the network environment, to ease the way people move through cyberspace, and provide a single way of moving? One already has-- Gopher. Gopher as Standard Gopher, developed at the University of Minnesota, employs a hierarchical, menu-based protocol to provide easy access to a wide variety of sources of information available over a distributed network. In many ways, it has become a de facto standard for moving around in the Internet without having to be an explorer.[3] Explorers report back their findings to Gopher designers, who may then add these resources to their menus, thus providing access. At this highest level, we may see Gopher as a standard approach, but no two Gophers are alike. Individual designers may choose to structure their menus in any way they like: organizing resources by topic, geographic location, institution, broad categories, and so forth. So though Gophers provide a standard protocol for moving through cyberspace, they are in many ways reflections of the cognitive maps of the individuals who develop them. Some Gophers' menu structures are quite intuitive and easy to use; others are unobvious and, eventually, somewhat frustrating and therefore less likely to be used effectively. Here though, at last, we find an appropriate use for the metaphor navigation. The ways in which people move through the hierarchical menus of Gophers can be seen to resemble real, spatial navigation. In this setting, even if you know where you're going, you can't get there directly; you must follow a particular path, moving down through the menus, until you reach your objective. You also may pursue unknown or uncertain destinations, but rather than stabbing blindly in the dark as with exploration as described above, you follow prescribed paths. As such, perhaps here some of the concepts from real navigation may be of use. As menu structures become more complicated and involve more layers, it becomes easier for people to get lost when moving through them. [4] The devices and tools used for orientation on ships-- compasses, lines of latitude and longitude, dead reckoning--may be applicable. One can imagine a Gopher organized in such a way; perhaps an indicator of where you are in the menu tree, how many layers down from the top you have traveled, how many to the bottom, how far away you are from major Gophers and other landmarks, a log of your moves so far, or a representation showing the contents of levels above, around, and below. As of this writing, Gopher is an attractive way to search because of its ease of use and power, but it clearly could stand improvement on many levels. A closer examination of the "navigational" aspects of searching through menus might help. Conclusions We have examined the extent to which the metaphor of "navigation" for moving through network space has validity. We find that in large part it does not hold, but a more generalized metaphor of "seafaring" does have some interesting parallels; namely, the notions of exploration and map making in this environment. A further look at the questions of individuality and standards led us to Gopher and its use to address the problem of access to distributed networked resources. The bottom line, of course, regardless of the metaphor we select, is to provide access to these resources for all users. An emphasis on "navigating the Internet" clouds the issue. "Navigation" is not the point; access is. Gopher is interesting but has problems of its own. Some other tool--an enhanced Gopher, perhaps, or WorldWideWeb, or something else--is required. What we really need, though, are underlying travel agents to work with explorers and map makers to find, describe, and direct users towards valuable resources. These might well be people who "live" on the Net, and give or sell their services to users to aid them in their search for information. The automatic tools already in use (Archie, WAIS, FTP, etc.) would be of use to these people, but their skills and experience would be difficult to replicate or replace by purely automated systems. References Anderson, E. W. "Navigation", Encyclopedia Britannica, Vol. 24, p. 750- 760, 1987. Footnotes 1. This section is, in large part, paraphrase of pages 750-753 of Anderson's Britannica article. Specific page references will not be given for the sake of clarity. 2. It could be argued that some of these standards may actually impede the cartographer's work, since adhering to them may sometimes make their lives more difficult, but we leave that question for another day. 3. There is another candidate standard, WorldWideWeb (WWW), a hypertext-based protocol for organizing and structuring access to networked resources. Although it shares some features with Gopher, it is neither as popular nor as widely used as Gopher, and so we will focus on Gopher in this discussion. 4. At this writing, a new tool, Veronica, which does for Gopher menus what does for FTP file names, shows promise for giving users a better idea of what exists in Gopherspace and how to get there. _____ Articles and Sections of this issue of _LIBRES: Library and Information Science Research Electronic Journal_ may be retrieved via anonymous ftp to cc.curtin.edu.au or via e-mail message addressed to LISTSERV@KENTVM or LISTSERV@KENTVM.KENT.EDU (instructions below) Papers may be submitted at anytime by email or send/file to: Diane K. Kovacs- Editor-in-Chief, _LIBRES: Library and Information Science Research Electronic Journal, EDITORS@KENTVM.KENT.EDU _________________________________ *Copyright Declaration* Copyright of articles published by LIBRES: Library and Information Science Electronic Journal is held by the author of a given article. If an article is re-published elsewhere it must include a statement that it was originally published in LIBRES. The LIBRES Editors reserve the right to maintain permanent archival copies of all submissions and to provide print copies to appropriate indexing services for for indexing and microforming. _________________________________ ____________________________ GOPHER Instructions ____________________________ GOPHER will be available for future Issues. ____________________________ Anonymous FTP Instructions ____________________________ ftp cc.curtin.edu.au login: anonymous password: guest cd LIB-RESEARCH get filename (where filename = exact name of file in Table of Contents quit LISTSERV Retrieval Instructions _______________________________ Send e-mail addressed to LISTSERV@KENTVM (Bitnet) or LISTSERV@KENTVM.KENT.EDU Leave the subject line empty. The message must read: GET LIBRE4N1 CONTENTS Use this file to identify particular articles or sections then send e-mail to LISTSERV@KENTVM or LISTSERV@KENTVM.KENT.EDU with the command: GET where is LIBRE#N# name) and is the name of the author