information design and the new media

36 i n t e r a c t i o n s . . . o c t o b e r 1 9 9 5

Photo: Robert Vizzini

37i n t e r a c t i o n s . . . o c t o b e r 1 9 9 5

a r t i c l e

The ways of handling information that work well in

(print and broadcast) do not always translate grace-

fully into new media environments. Although enter-

tainment (in video games and CD-ROMs) and

communication (on the Internet and on-line services)

are finding new forms and new configurations in the

new media, information (news, reference, educa-

tion), which still relies largely on text, usually takes

on forms native to print environments. These can

appear very awkward in new media. This article

compares people’s relationships to digital media and

print media; it also examines information and peo-

ple’s needs for and expectations of information,

with an eye toward adapting information design to

suit new media environments. This concerns not so

much the death of print as it does fitting the inter-

face to both the medium and the message—which

some would say are the same thing.

“Old Media”“Old Media”

InformationDesign and the

New Media


a r t i c l e

InformationDesign and the

New Media

Media A useful basic distinction can be made between active and passive, or between

users and consumers. The old media are massified; they broadcast on a one-to-many

model, and I don’t really get to choose what I want, except on the widest level. The mass

media deliver information on their own terms and on their own schedule, assuming the

existence of a passive consumer on the receiving end. All periodicals and broadcast media

select subsets of information from the vast pools available, and transmit these subsets in

complete packages that they create, in a format of their choosing. The most active thing

that I, as a consumer, can do is go out and choose a package. After I have it, I can’t add to

it. I can’t rearrange its contents. I can’t respond to it except inside my own head.

Melinda McAdams


McLuhan, who did not live to see the prolif-eration of personal computers, set television on apedestal as the “coolest” of the electric technolo-gies, one he referred to as “the mosaic mesh.” Hecredited TV with breaking up the linearity ofhuman lives and thinking, a linearity that hadevolved from print culture—from centuries ofwords set in type, set in lines, set in pages thatfollowed one another in invariable numericalorder. The increasing use of new media contin-ues this breakup; new media, in fact, reveal therelative linearity of television (and all time-basedmedia) and offer far more possibilities for aninformation mosaic. As oral cultures becameprint cultures after Gutenberg [3, 7], print cul-tures are now becoming electronic cultures.

Evidence of this change lies in the way weprefer to receive and process information. Areyou impatient with long articles? Do you turnthe page when you see there are no subheadingsto break up the long columns of type? Do youlike to flip channels on the TV set? Have younoticed how more movie previews take the formof montage, images strung together to a boldmusical score with a lot of cutting from scene toscene and not much dialogue, and afterward youfeel as if you’ve seen the whole movie?

These preferences, and this way of thinking,demand new ways of presenting information sothat people will be able to use it. As we moveaway from the book and from print culture,many questions come up: How will people seekand find information when they are no longer“looking it up’’? How can we search for imagesand sounds without reading text descriptions ofthem? How will education change? How willthe meaning of “news’’ be affected?

When we talk about users, we’re assumingthat they use something. They don’t passively

read or view or consume; they actively use. Theworst kind of CD-ROMs assume a passiveviewer; they provide slide shows in a complete-ly linear format, a perfect example of what hap-pens when a designer does not understand themedium. A disc cannot be called interactivejust because the user clicks a button labeled“next” to go to the next picture. Informationusers expect to be able to explore and discover,to seek and find. Otherwise, they go back tobeing information consumers.

To address this difference, we need to consid-er responsiveness (an effective alternative to theoverused “interaction”). The Internet and on-line services offer an ad hoc form of responsive-ness: Messages can be sent and received in afree-form way between those who produce infor-mation products and those who use them. CD-ROMs can be responsive, too, but in a differentway: By allowing the users to set the agenda, tolook for things that interest them and entirelyskip the things that don’t, even this containedmedium provides more responsiveness than anewspaper, book, or television program.

Media are extensions of ourselves. Media arelike long arms extending out from our bodies.Some media, like highways, enable an extensionof an individual’s world. With old informationmedia (the mass media), the extensions reachout to us, from their point of production, anddrop products onto us, the audience. Because ofthe participatory nature of new media, exten-sions come from both directions—from pro-ducer and user—and meet at a midpoint.

InformationThere are two ways of talking about informa-tion: the engineering view, which is largelyabout the transmission of signals, and the

About the Author:

Melinda McAdamsConsultant,

Information Design121 Dublin St. N.,

Apt. AGuelph, Ontario

N1H 4N5, [email protected]://www.well.com/

user/mmcadams/

arshall McLuhan [5] separated “cool” media, which demand active par-

ticipation and an “involvement in process,” from “hot” media, like

print, which come in complete packages and encourage passive consumption.

In McLuhan’s view, roads and vehicles, money, and weapons are media, just as

movies, books, and radio are. Media act as extensions of the human body, and

electric media are extensions of the human nervous system.

M


meaning view, which concerns how peoplethink about information. New media producersneed to concentrate on the latter.

In the broadest sense, everything is informa-tion: sounds, smells, tastes, and anything else thatenters our sense organs. But most people woulddefine information as something that has mean-ing—that is, it must make sense or have signifi-cance. Meaning is, of course, relative; what hasmeaning to one person may have none to another.Meaning also has nothing to do with importanceor relevance; news of a war in a foreign country hasmeaning, but that news may be neither of use norof interest to any number of people.

An often quoted definition of informationsays it is “a difference which makes a differ-ence’’ [1]. But even “makes a difference” pre-sents us with at least two possible meanings:that some change is brought about, or thatsomeone cares (in the sense opposite to thestatement “It makes no difference to me’’). Thesecond is probably closest to what most peoplehave in mind when they go looking for infor-mation (or when they express dissatisfactionwith the information available tothem); they want something thatmakes a difference to them. In thecontext of one person’s search forsomething that will be of use or ofinterest to him or her, all the end-less gigabytes of information in the world aremeaningless junk except for the bits that fulfillthe purposes of that person’s search.

We can define the standards people use tojudge the information that is available to them:They want information that is either useful orinteresting or both. That is the standard theywill use to measure whether information theyget is satisfactory or not. Although they mayalso apply other standards, such as truth andcompleteness, they will probably do so afterthey have determined whether the informationis of use or of interest—or they will choose theirsource in the first place based on whether theybelieve it to be reliable or complete.

Recognizing that information is somethingthat has meaning, but also that not all informa-tion is of value to all people, we can start think-ing about information overload and whatpeople want from information.

Too much is not enough. “We live in a world

of proliferating information and shrinkingsense,’’ Jean Baudrillard [2] wrote in 1980.There is more information out there—moreTV channels, more magazines, more books—than was available fifty or even ten years ago.Baudrillard called it “bombardment by signs’’and referred to “a state of perennial emulsion.’’

Much of this information is useful or interest-ing to someone, but for each individual, very lit-tle is of use or of interest. In addition, at the sametime that they are being bombarded with morethan they can take in, people also lack informa-tion that they need or want. Sometimes theydon’t even know what they want, but they expe-rience a vague sense of dissatisfaction, as if there’smore they should know, but they don’t knowwhat that “more” is or where to look for it.

Richard Saul Wurman, a graphic designerand creator of innovative guidebooks, notesthat many people waste a great deal of time andeffort learning things that do not really interestthem, out of some social compulsion to feelinformed or up to date [8]. They act, he says, asif they think there will be a test.

Noise not only makes it more difficult tofind what we want, but it also makes us feeloverloaded, numb, and powerless. For anysubject, some people will want more informa-tion about it, but many will want less (ornone). The existing mass media do not pro-vide people with filters to block what they donot want. People find that they have no timeto “keep up with the news”—to read a news-paper or the magazines they subscribe to, or towatch all the TV programs they have taped.

a r t i c l e

e can make a useful distinction between

noise and redundancy in this context.

Both interfere—in different ways—with

people’s ability to get meaningful information [4].

“Noise” is any information that is neither interest-

ing nor useful to us, getting in the way of what we

really want. “Redundancy’’ is too much information

that is too much the same, like 10,000 articles

about how to get skinny.

W


Excessive redundancy makes people try toshut things out—maybe the things they mostwant (or need) to know. In the mass media,the two extremes are superimposed: much ofthe noise is redundant, and much of theredundancy is noise.

Time-based media require a lot of redun-dancy; the eleven o’clock news always repeatsmuch of what was on the six o’clock news, andtoday’s newspaper has to assume that you didnot read yesterday’s. With hypertext links, newmedia can avoid at least redundancy (if notnoise); links that change in appearance (as insome World Wide Web browsers) toshow that you have already followedthem can help.

Within a contained product, suchas a CD-ROM, redundancy can beavoided. But in any dynamic form,redundancy is a danger.

Noise can also be avoided in new media moreeasily than in the old, because of the differencebetween a user and a consumer. Users can deter-mine quickly whether a particular site or seg-ment will be interesting or useful, and if neitheris the case, the user can just as quickly move on.Channel surfing is the mode of operation nativeto new media. Users are cruisers at high speed.

Use of InformationNot all users who enter an information envi-ronment have the same goal. Some come inhoping to find a particular thing, to answer aspecific question, to research a certain topic.Others come in to enjoy themselves, to browsewhat’s available, to see what’s new. These twotypes of activity in an information space are notexclusive, and any user who comes in for onetype of activity may want to switch to the otherat some point. Moreover, both searching andbrowsing consist of various activities.

A searcher wants, first, to find. But findingincludes several steps. The searcher needs searchtools—not only a text search for words andphrases, but also ways to find sounds andimages, subject categories, and material relatedto the primary topic. Then, when the set ofsearch results is in hand, the searcher may needto evaluate which items deserve further atten-tion; for example, if the search returns a largeset of documents, the searcher usually does not

want to read every one of them. What if the setis not satisfactory? Can the user modify thesearch from this point and go deeper, or doesshe have to start over?

With the finding stage completed, thesearcher wants to study the found items. Are theitems presented in a form that’s good for thispurpose? Can the searcher save or store the set offound items? Can she make notes for herself asshe’s examining them? Can she call up relateditems without losing her place? Can she connectthose in this set to other items—so that she canrefer to them later, or pass them on to a friend?

Some browsers may not want to browseeverything, but may want to limit their infor-mation space to certain subjects, or to a certaintime frame. Others may want to have a widefield, but will prefer to eliminate, or screen out,certain subjects that do not interest them.These filters, then, are of two kinds: one letsthings in, and the other keeps things out.

In some cases, the browser will become asearcher in the same session. Something hecomes upon may make him want to initiate asearch. A searcher may also wish to suspendsearching temporarily and browse, perhaps inan exploratory way, to determine whether thisparticular space contains suitable or sufficientinformation.

For both kinds of users, an information spacemust do more than provide access to informa-tion; it must be organized and navigable, and itmust provide comfortable ways to handle theinformation once it is called up. If users cannot

rowsing too comprises several activities.

Being casual, the browser doesn’t want to

spend a lot of time configuring the system

or evaluating items he comes across. A browser’s

environment requires few words, appropriate

images, and easy movement. However, once the

browser sees something of interest, he wants to

stop and look at it more closely. He wants to

switch modes—but this does not necessarily

transform him into a searcher.

B


meet their goals—if, either as searchers or asbrowsers, they cannot find what is of use or ofinterest to them—then the information spacefails. It is a mistake for developers to complainthat users are at fault if they can’t figure out howto function effectively in the environment.

A well-designed information space is onethat (like any well-designed environment) isconsistent, predictable, and transparent.

Don Norman [6] identifies four principles ofgood design that can be applied to an informa-tion environment to make it self-explanatoryand not frustrating: (1) visibility, so that userscan see clearly what state the system is in andwhat it is possible to do next; (2) a good con-ceptual model—that is, the metaphor thedesigner chooses to present to users; (3) good“mappings,” so that users can figure out andanticipate relationships between their actionsand the results they will get; and (4) feed-back, so that users always know what theiractions have caused to happen.

OrganizationAn information space, whether dynamic(on-line) or static (nonrewritable media) beginswith a large collection of information. The firststep must be to organize the information in alogical and useful way.

Wurman’s “five ways of organizing informa-tion’’ are by category, by location (for example,maps), by continuum (that is, from best toworst, largest to smallest, and so on), alphabet-ically, and chronologically. In a new media envi-ronment, some kinds of information can beorganized in all five ways simultaneously, andusers can choose which organizational schemethey prefer. It’s important to remember that inan electronic space, organization can be virtual;that is, the way the information appears to beorganized does not need to have any relation tothe way it is stored in digital form.

Ways of organizing information are alsoways of understanding, as Wurman points out.When people can move easily from one topic toanother, or from one source to another, theybuild mental connections that reinforce learn-ing and comprehension.

Some new media products take feeble stepsin this direction. CD-ROM encyclopedias, forexample, often offer a time-line view as an

option to a subject-oriented view. But the timeline is not integrated across all the entries; userscannot switch from every entry to a precisepoint on the time line, and choosing a point onthe time line does not take them to an array ofall concurrent articles.

Great promise for information environmentslies in new media’s ability to show connectionsand relations among multiple, simultaneoussets. Hypertext naturally plays a part in theorganization of a new media information envi-ronment. Much has been written about usingmachine intelligence to construct these connec-tions, but it may be that human experience willalways provide a better framework for buildinginformation sets that will be interesting anduseful to other humans—especially given themyriad possibilities presented by large sets con-taining overlapping information.

Whether this kind of selectivity is welcomedepends on the users’ preferences. Searchers willoften want the complete list of relevant itemsand will want to evaluate all of them. Morecasual users may prefer to have the best or theclearest or the most recent selected for them.

The less repetition, or useless redundancy, auser finds in a set of information, the morevaluable it must be. The same is true of noise.It is probably too much to expect thatmachine intelligence will be able to configuresets with low factors of noise and redundancy.Thus organizing information for use is likelyto remain a highly skilled task, even more so

a r t i c l e

In searching for relevant texts, software systems

can use word occurrence to bring back only texts

that are primarily about a requested topic,

rather than texts that merely mention the topic. But

software is unlikely to be able to tell which of five,

or ten, or one hundred articles about a topic pre-

sents the best introduction or overview, and soft-

ware also seems an inadequate judge of which

additional articles will complement, rather than reit-

erate, the information found in the overview.

I


than the work of determining subject labelsfor library card catalogues.

Whenever whole texts originally produced inprint environments are linked together, redun-dancy is inevitable. If the texts can be brokendown into smaller segments, some of the redun-dancy can be avoided. If the texts are createdspecifically for the new media environment,then they should be developed with linking inmind, and also with consideration for browsingusers, who will want to move quickly from textto text. Even though some texts do not lendthemselves to incorporation of embeddedhypertext, most can be broken up into smallersegments that will retain their meaning.

Although use of hypertext can cut down onredundancy, much of the hypertext found onthe World Wide Web introduces a lot of noise,and thus offers little improvement from oldmedia environments. Gratuitous or excessiveuse of links does no user a favor; rather, it willdiscourage many users from following links.Designers of an information space need tofind a balance between allowing for serendipi-tous discovery and providing a frustratinglylarge number of options. One way to do so isto evaluate the entire set of links in each seg-ment and eliminate those that are redundant.The elimination of such links does a great ser-vice to the user.

Organizing the whole of the informationspace constitutes only one level of an informa-tion designer’s task. Each element in the spacemust have its own internal organization; ele-ments that belong to one set must be organizedas a set; sets that are related to one another musthave appropriate connections. Finally, the usersmust be able to come into the full space, or col-lection, and find a form of organization thatmakes sense to them so that they can movearound comfortably within it.

NavigationAlthough organization has some bearing onnavigation, it is not the only factor affectingnavigation. Because users have different goalsand desires when they enter and while they arewithin an information space, they need a vari-ety of navigational tools and options.

The word navigation invokes an image ofan explorer or a traveler. Some insight into

navigational options may be derived from atravel metaphor:

• If you know exactly where you’re going,or what you want to see, you want to getthere quickly and directly, with as littletrouble as possible.

• When you get there, you want to stay fora while and to have the chance to godeeply into the local culture and history.

• If you’re more interested in the scenery, inseeing the country, you want to keepmoving. You don’t want to get stuck inthe station. But you also want to be ableto get out at will and linger if the urgestrikes you.

• No one likes to be really lost. A little bitlost, temporarily, is all right.

• A traveler is different from a tourist.Tourists want everything to be arrangedin advance, never to deviate from theplan. Tourists want someone else to showthem the sights—even to decide whichsights are worth seeing. Travelers like toguide themselves and make their owndecisions. That does not mean they shunadvice. But they don’t want to followanother person’s itinerary.

Because much of our information is stillconveyed most efficiently in the form of text, anumber of print conventions for navigationappear in new media environments. Some ofthese work fine; a table of contents, for exam-ple, can serve as a kind of map to let users knowwhat to expect in this space. If users can jumpfrom any information segment back to theappropriate section of the table of contents,they will always be able to orient themselves.

Few things are less inviting, however, than avery long, multilevel table of contents in outlineform. Some sites on the World Wide Web usethese, making each line a hypertext link, but theform of presentation overwhelms the user. Abetter strategy would be to present the top lev-els only (Roman numerals I, II, III) and haveeach of those link to a short overview, followedby the next level (A, B, C), which would lead toan overview of the sublevel (1, 2, 3), and so on,eventually taking the user to the individual seg-ments. In this way, the user gets to choose how

PERMISSION TO COPY WITHOUT

FEE, ALL OR PART OF THIS

MATERIAL IS GRANTED

PROVIDED THAT THE COPIES ARE

NOT MADE OR DISTRIBUTED FOR

DIRECT COMMERCIAL

ADVANTAGE, THE ACM

COPYRIGHT NOTICE AND THE

TITLE OF THE PUBLICATION AND

ITS DATE APPEAR, AND NOTICE IS

GIVEN THAT COPYING IS BY

PERMISSION OF THE

ASSOCIATION FOR COMPUTING

MACHINERY. TO COPY

OTHERWISE, OR PUBLISH,

REQUIRES A FEE/AND OR

SPECIFIC PERMISSION

© ACM - ⁄ ⁄ .


deep to go and also gets to take in and evaluateinformation quickly, actively.

A full index is equally tedious on a comput-er screen. Indexes require a great deal ofscrolling or paging even if the user can jump toa selected letter of the alphabet. Instead, theuser could initiate a search for a word or phrasethat would bring up the information directly.Although serious researchers may want someform of index, the print model is inadequatewhen we consider the power of electronic text:the user’s entry of a search term could trigger asearch for all related terms and synonyms, if thesearch tool had been configured appropriately.

Browsing users do not necessarily want anyform of map or index. As discussedearlier here, they are more likely towant to set parameters—if the mecha-nism for setting them can be made rea-sonably straightforward—to constraintheir browsing to subjects in theirrange of interests. Even so, when browsers movefrom one place to another, they will want tohave some feeling of control; that is, they willseldom want to skip about completely at ran-dom. Some environments force randomness onthe user by giving inadequate informationabout the options for the next step. For exam-ple, some hypertexts highlight a single word asthe link to another segment, and that wordalone is often insufficient to provide a mean-ingful expectation of what it may lead to. Otherenvironments not conducive to browsing arethose that provide only an option to go to the“next” segment or to return to the “top,” andthose that don’t offer any expansion of the cur-rent segment, but lead only to related topics.This last flaw is typical of CD-ROM encyclo-pedias, which many users find to be enticinglybroad, but disappointingly shallow.

The navigation needs of a browser, then, arevery different from those of a searcher, and yetany user may become a browser or a searcher atany time. The challenge in a robust informationenvironment is to provide useful tools for bothand to make all the tools accessible all the time.

Formatting InformationOne last consideration about the design of aninformation space concerns the basic units ofthe space, or the building blocks—the individ-

ual blocks of information themselves. In somecases these are sounds and in other cases, images,either still or moving. But in many cases they aretext, and one of the biggest differences betweenprinted text on a page and electronic text on ascreen is that it does not scan the same way.Scrolling up and down differs greatly from turn-ing pages, and the screen provides less spacethan most pages (if the text is displayed at a sizethat will not strain the eye, as it should be). Inthis transition period, as we are poised betweenthe print era and the electronic era, many of theelectronic texts available have been dumped infrom print environments, and their format is illsuited to electronic reading.

Contextual use of display type aids a browserof electronic texts. Use of headings should beincreased over print—but not randomly. Eachheading should say something meaningful aboutthe text below it, so that the user can make aninformed decision about whether to read it.There is more reason to make every word countin the subheadings of a scrolling document thanthere is on the front page of a newspaper. Eventhe use of italics and boldface and, of course,color, within the body of the text can be increasedfar more than any print designer would ever per-mit, because this is not print, even though it istext, and the reading patterns are different.

Graphics can serve a wide variety of iconic uses.Small graphics inserted at the beginning of a sec-tion or paragraph can tip off the user to a type of

a r t i c l e

The use of “white space,’’ or empty space,

has become more and more popular in print

since the 1960s, and now even newspapers

(which used to cram the columns so close to each

other, they almost touched) recognize the

improved readability that results. In electronic

environments, a full line of space between para-

graphs makes a huge difference between a read-

able text and an unreadable one. A narrower

line-width, too, is easier to read (and scan) than a

longer one—that also holds true in print.

T


content—a globe could mean geography, a pair ofhuman profiles could mean “related thinkers.’’The larger and more diverse an information spaceis, the more useful such cues could be.

Even a focused searcher may begin scanning atany moment, and visual cues that aid in scanningincrease the value of the information space to itsusers. The implication is not that no one will wantto read deeply, but, rather, that the informationdesigner should strive to help the user avoid read-ing extraneous material and to make it easier tolocate material that is of real value to that user.

SummaryNew media information environments are verydifferent from print and broadcast environ-ments, even when much of the content of the

new media environment is text. For this reason,designers of information spaces can benefitfrom thinking about the capabilities of newmedia, the qualities of information, and thegoals and desires of users of information.

There is excessive information available tomost people; not only is it available, it is intru-sive. Merely providing information or access toinformation is, therefore, not enough. The infor-mation space must be well organized and easilynavigable, and the form of the information (notonly its content) must be useful to those whofind it. Because different users of informationwill have different goals, and any one user mayhave different goals at different times, it’s impor-tant to provide adequate, flexible tools to enablethe users to accomplish what they want.

If we resist the urge to equate text with printmedia, and recognize the full potential of elec-tronic media—for searching, for providing nav-igation options, for setting filters, for linkingsets of information, for offering hierarchicalpaths through large collections of relateddata—we will design information spaces thatare effective and satisfying.

ReferencesBateson, G. Steps to an Ecology of Mind. Ballantine

Books, New York, 1972.

Baudrillard, J. The Implosion of Meaning in the

Media and the Implosion of the Social in the Mass-

es. In Questioning Technology: Tool, Toy or Tyrant?

(Zerzan, J., and Carnes, A., eds.). New Society Pub-

lishers, Philadelphia, 1991. [Article originally pub-

lished in The Myths of Information, 1980.]

Eisenstein, E. L. The Printing Revolution in Early

Modern Europe. Cambridge University Press, Cam-

bridge, U.K., 1983.

Klapp, O. Overload and Boredom: Essays on the

Quality of Life in the Information Society. Greenwood

Press, New York, 1986.

McLuhan, M. Understanding Media: The Extensions

of Man. McGraw-Hill, New York, 1964.

Norman, D. A. The Design of Everyday Things. Dou-

bleday, New York, 1988. [Originally published as

The Psychology of Everyday Things].

Ong, W. J. Orality and Literacy: The Technologizing

of the World. Routledge, London, 1982.

Wurman, R. S. Information Anxiety: What to Do

When Information Doesn’t Tell You What You Need to

Know. Bantam Books, New York, 1989.

[]

[]

[]

[]

[]

[]

[]

[]

information design and the new media

Documents