creativity support tools - cs.umd.edu

20 December 2007/Vol. 50, No. 12 COMMUNICATIONS OF THE ACM

Illustration by Brad Yeo

Since scientific discoveries and engineer-ing innovation produce broad benefits,improved tools that advance individual,group, and social creativity are impor-tant contributions. The current andforthcoming generations of program-ming, simulation, information visualiza-tion, and other tools are empoweringengineers and scientists just as animationand music composition tools have invig-

orated filmmakers and musicians (see the sidebar “New Media Arts and theFuture of Technologies”). These and many other creativity support toolsenable discovery and innovation on a broader scale than ever before; eagernovices are performing like seasoned masters and the grandmasters are pro-ducing startling results. The accelerating pace of academic research, engi-neering innovation, and consumer product design is amply documented injournal publications, patents, and customer purchases.

By Ben Shneiderman

How can designersof programming

interfaces, interactivetools, and rich socialenvironments enable

more people to be morecreative more often?

CREATIVITY SUPPORT TOOLSAccelerating Discovery and Innovation

While telescopes and microscopes extended anindividual’s perceptual abilities to make discoveries,modern creativity support tools also enable newforms of expression for individuals, and they are espe-cially potent in supporting group collaboration andsocial creativity (see the table here). Creativityincludes discovery or invention of a significant idea,pattern, method, or device that gains recognitionfrom accepted leaders in a field, while innovationrequires further steps to ensure adoption (see the sec-tion “Defining and Supporting Creative Processes”).For example, manyresearchers extend theirperceptual abilities byapplying general-purposescientific or informationvisualization tools, whichenable them to make dis-coveries about their data(see Figure 1). Otherdomain experts, such asgenomic researchers, usespecialized visual analysistools to discover biologicalpathways. Scientists andengineers draw on power-ful mathematical, design,and simulation tools tosupport their discoveryand innovation (see Figure2). New media artists real-ize their desire for personalexpression with powerfuldevelopment environ-ments that support anima-tion, music, or videoediting tools.Even more remarkable

opportunities have emergedfor group collaboration across time and space, asafforded by programming environments that enabledistributed teams to accelerate development of soft-ware projects. Still broader impacts stem from socialcreativity tools, such as wikis, citizen journalism, andmedia sharing, which enable thousands of cooperatingindividuals to create and share significant new contentand services.Never before has it been possible to arrange rapid

and broad collaboration among numerous contentcreators and service providers. Understanding the pas-sion and persistence required for individual creativityis difficult enough, so designing for social creativityrequires rigorous research, with fresh theories of col-lective efficacy and the motivational impact of

rewards and recognition (see the sidebar “SustainingSocial Creativity”).

A HISTORIC SHIFTDuring the past half-century, computing professionalshave developed potent productivity support tools thatreduced manufacturing costs, tightened supply chains,and strengthened financial management. These busi-ness productivity support tools were designed to meetclear requirements such as improving insurance claimprocessing, reducing costs for airline reservations, or

simplifying order entry.These tools were conve-niently evaluated by stan-dard measures such astime per task, cost pertransaction, and errorsper order.But now, a growing

community of innovativetool designers and userinterface visionaries isaddressing a greater chal-lenge and moving fromthe comparatively safeterritory of productivitysupport tools to the morerisky frontier of creativitysupport tools. The chal-lenges they face stemfrom the vague require-ments for discovery andinnovation, as well asfrom the unorthodox userbehaviors and unclearmeasures of success. The

risks are high, but so are the payoffs for innovativedevelopers, ambitious product managers, and boldresearchers. Creativity support tools extend users’capability to make discoveries or inventions from earlystages of gathering information, hypothesis genera-tion, and initial production, through the later stages ofrefinement, validation, and dissemination.A way forward for research and development on

creativity support tools has been to focus on specifictasks that support discovery in the sciences, explo-ration in design, innovation in engineering, andimagination in the arts. For example, we alreadyknow that an accelerator for creative efforts is thecapacity to locate, study, review, and revise existingprojects and performances, such as open source soft-ware modules, Web page source code, architecturaldrawings, or music scores. The Web has done muchto make existing projects and performances accessible


Information visualization tools

Specialized visualization tools: GIS

Specialized visualization tools: gene expression analysis

Mathematical manipulation

Engineering, architectural,industrial, product design

Simulation

New media development environments

Animation and interaction

Music

Video editing

Concept mapping

Spotfire, SAS JMP, DataDesk,ManyEyes, Digg

Google Maps, ArcInfo

GeneSpring, DNASTAR

MatLab, Mathematica

Autocad Inventor, DataCAD, SolidWorks

SPICE, Tierra

Max/MSP, Pd, processing

Flash, FLEX, OpenLaszlo

Cinescore, Cakewalk Sonar

Premier, Final Cut Pro, Lightworks, iMovie, Windows MovieMaker

Inspiration, MindMapper, MindManager, Axon

Software development

Wikis

Citizen journalism

Media sharing

Music

Eclipse, JDeveloper, Visual Studio

Wikipedia, Wikia

Blogger, Ohmynews, Slashdot

Flickr, YouTube

Garageband, MacJams

Group and Social Creativity Support Tools

Individual and Group Creativity Support Tools

Samples of classes ofcreativity support tools and

examples of products.

and search engines like Google have helped innova-tors to quickly find what they want. Future searchengines can be designed far more elegantly to enableusers to find the relevant results with specific features.While current search engines can find sonatas, sin-gling out those in a romantic style with acceleratingtempo written in France during the 20th century maybe more difficult to achieve. Often searchers need helpin discovering the range of possibilities, while learningthe concepts and terminology. Such exploratorysearches may take users weeks or months to complete,requiring note taking, consulting with colleagues, andrefining their goals. Of course, intellectual propertypolicies must be modernized, to let users more easilybuild on previous work while paying fair licensingfees. Diverse proposals for copyright reform, patentmodernization, and the Creative Commons offermodernizations of legal structures that accommodatethese new technologies and new ways of working.A second example of creativity support is the

capacity for users to rapidly generate multiple alterna-tives, explore their implications, or revert to earlierstages when needed. Hypothesis generation for scien-

tists, prototypes for software engineers, models forarchitects, and sketches for artists are well establishedas important steps in their agile creative processes.Certainly well-designed software tools can help cre-ators in generating multiple possibilities, showing theimplications of their choices and tracking their designdecisions [12]. The best tools enable users to save theirhistory, edit it, email it, and replay it thousands oftimes with different parameters.These and other examples distill emerging design

principles, but skeptical business professionals,inspired artists, and diligent academics still worryabout whether creativity support is an achievable goal.They may deride suggestions that creative humanendeavors can be aided by inherently structured user

COMMUNICATIONS OF THE ACM December 2007/Vol. 50, No. 12 23

Figure 1. The IN Cell Analyzer automated microscope was usedto identify proteins influencing the division of human cells. Afterthe images were analyzed, quantitative results were transferred toSpotfire DecisionSite. This screen revealed the previously unknowninvolvement of the retinol binding protein RBP1 in cell cycle control(see Stubbs, S. and Thomas, N., Methods in Enzymology 414 (2006),1–21). Retinol—a form of Vitamin A—plays a crucial role in visionand during embryonic development (courtesy of Nick Thomas,GE Healthcare).

interfaces that inevitably limit exploration. Theseconcerns are taken seriously by creativity supporttool designers, who recognize the difficulties, butalready see grand successes and future opportunities.Just as telescopes, microscopes, and cameras are pow-erful devices that enable discoveries and innovations,they are still only tools; the act of creation is carriedout by the users.As a computer and information science research

topic, creativity is still emerging. The ACMComput-ing Reviews Classification System has more than1,500 entries, but does not include creativity, discov-ery, exploration, or innovation. By contrast, creativ-ity-related topics are currently high in nationalpriorities worldwide, generating calls for supportfrom national science research boards [6, 7]. At thesame time, national legislators and regional plannersare concerned about promoting competitiveness,enhancing workplace innovation, and attracting cre-ative industries. Their emphasis ranges from supportfor game design entrepreneurs and film animationcompanies to pharmaceutical drug discovery teamsand consumer product designers.Traditional descriptions of creativity often sug-

gested that creative personalities—the Einsteins andPicassos of the world—were rare occurrences withspecial talents who came along once in a generation totransform the world. The modern belief, held bymany teachers and researchers, is that creativity can be

taught, and that everyone can be creative. This is aremarkable transformation from 400 years ago, whenscholars devoted much energy to copying or translat-ing the words of Aristotle and other long-deadauthors. While learning from and building on pastwork is important, the Web and the broad use ofinformation and communications technologies hasraised expectations that every student should writepoems/programs, make photos/videos, design inter-faces/games, and then disseminate them to others.The widespread availability of books and then elec-tronic media transformed education so that nowevery student is expected to compose original texts,videos, animations, music, and art. Teachers alsoexpect their students to produce science and engi-neering projects with fresh empirical evidence, origi-nal discoveries, or innovative devices [8].

DEFINING AND SUPPORTING CREATIVE PROCESSESWhile there has been extensive research on creativityin many disciplines, the topic is a relatively new one


Figure 2. Using Autodesk Inventor’s Design Accelerators forengineering tasks such as shaft design, gear design, and bearingselection, engineers at Stork Townsend, Inc. were able to create acustom gearbox with confidence that the unit would perform toexpectations in a harsh environment. Shown here: the Worm GearGenerator is used to create matched sets of paired gears used in thisgearbox design. All mating assembly constraints are automaticallyadded, and an additional benefit to users is that this same interfaceis used for any edits to the gear pairs (courtesy of Autodesk andStork Townsend, Inc.).

in computer and information science. The excellentHandbook of Creativity [11] covers many researchdirections but terms such as “computer” and “userinterface” don’t even appear in the index. The largeamount of literature on creativity, discovery, design,innovation, and composition may be sorted intothree intersecting schools:

• Structuralists believe people can be creative if theyfollow an orderly method, typically described withseveral stages, such as preparation, incubation, illu-mination, and verification. There is ample anecdotalevidence that great breakthroughs happened accord-ing to this generic method, but many variations are

promoted by self-help books, organizational creativ-ity consultants, and systematic discovery methodssuch as TRIZ (Theory of Inventive ProblemSolving; www.triz.org). Systematic approaches toexploratory search include the Arrowsmith(http://arrowsmith.psych.uic.edu) method for find-ing unusual overlaps in distant disciplines and thecombinatorial hypothesis generation (exhaustivesearch) for multiparameter simulations.• Inspirationalists argue that breaking away fromfamiliar structures elicits creative solutions. Theyadvocate working on unrelated problems, gettingaway to scenic locations, and viewing random pho-tos or inkblots. Inspirationalists promote meditation,hypnosis, dreaming, and playful exploration. Theyseek to liberate thinking from old habits so as tobreak through to the Aha! moment of inspiration.This school of thinking advocates sketching toquickly explore possibilities, concept mapping to dis-cover unexpected relationships, and visualizationstrategies to see the big picture.• Situationalists recognize that creative work issocial. They seek to understand the motivation ofcreative people, their family history, and their per-sonal relationships with challenging teachers,empathic peers, or helpful mentors. They under-stand the need for distinctive forms of consultationat early stages when fear of rejection, ridicule, andrip-off are high versus later stages when validation,refinement, and dissemination are prominent. Sit-uationalists seek to understand the motivatingroles of rewards and recognition (for example, does

the Nobel Prize promote creative work?), as well ascompetition vs. collaboration.Each of these three schools offers important lessons

for designers of creativity support tools. Structuralistthinking encourages systematic tools that includeprogress indicators with reminders of what is stillneeded. The inspirationalist view supports develop-ment of image libraries, thesauri, sketching interfaces,and concept-mapping tools. Situationalists broadenthe designer’s view to include email and collaborationtools, as well as the e-science notebooks that guideusers and coordinate groups through scientificprocesses over weeks, months, and years.A prominent situationalist researcher is Mihaly

Csikszentmihalyi, whose in-depth interviews with 91famously creative people [2] led him to make theseuseful, but provocative definitions:

Domain: “consists of a set of symbols, rules, and pro-cedures” that are accepted and used by a well-defined community, such as mathematics or biology.

Field: the respected leaders in a domain: “theindividuals who act as gatekeepers to thedomain...decide whether a new idea, perfor-mance, or product should be included.”

Individual: creativity starts with individual motiva-tions and insights, but requires social confirma-tion. He defines creativity as “when a person...hasa new idea or sees a new pattern, and when thisnovelty is selected by the appropriate field forinclusion in the relevant domain.”

On first reading, Csikszentmihalyi’s definition maybe disturbing, since it implies that contributions arecreative only when recognized by journal editors,patent examiners, symphony directors, and others insimilar roles. Many people feel they are able to judgetheir own contributions, but Csikszentmihalyi’s defi-nition asserts that to gain recognition, contributionsmust be judged by accepted leaders in a field. Hemakes clear that creative people need to respect previ-ous work and to present discoveries and innovationsin a way that clarifies their contributions. Csikszent-mihalyi’s definition stresses context, making creativitya social and political process in which the structuredmethods and Aha! moments are merely middle stages.


Just as telescopes, microscopes, and cameras arepowerful devices that enable discoveries and innovations, they are still

only tools; the act of creation is carried out by the users.

CHANGING MIND-SETSGetting information technology companies and aca-demic researchers to invest resources in creativity-related research and development requires at leastthree significant changes in mind-sets.

• Developers who understand that benchmark taskcompletion is giving way to playful exploration,richer search features, generation of multiple alter-natives, and easy backtracking with rich history-keeping. They also recognize that Web-enabledsocial creativity environments can support innova-tive approaches to software development, contentcreation, and rapid dissemination of new ideas.• Product managers who conceive of their customersas creators, rather than merely users or consumers,are already changing their requirements analysis,feature selections, and marketing strategies. Theyknow creative people want open systems they canextend and that they want an audience, feedback,rewards, and recognition.

• Researchers who study and evaluate software usageare getting past old strategies of controlled studiesand short-term usability testing to embrace ethno-graphic styles of observation, long-term case stud-ies, and data logging to understand patterns ofusage. They know that motivation, empathy, play-fulness, and surprise are part of the creative land-scape. They also know that getting the correctcombination of individual discovery, supportiveconsultation, and community brainstorming gen-erates high user engagement.

These changes to expectations for individuals andtheir institutions are important first steps in enablingmore people to be more creative more often. But evenwith clarity about the goals, there are still numerouschallenges such as developing design guidelines andappropriate research methods.

DESIGN PRINCIPLES FOR CREATIVITY SUPPORT TOOLSWorld-famous architects such as Norman Foster andFrank Gehry claim their innovative buildings wouldnot be possible without computer tools that enabledthem to create their complex structures. Excellentinterfaces, sometimes with rich domain-specific fea-tures, are essential for creativity support, as usersneed to apply their cognitive resources and passionsfully to their discoveries and innovations. Whileexperience across domains is diverse, there areunderlying principles to guide designers [5, 9].These principles include:

Support exploratory search: To be successful atdiscovery and innovation users should be aware ofprevious and related work, but finding relevant

items may prove challenging with traditional key-word search. Google is great for fact finding, and itcan be helpful for exploratory search projects, butthere is much room for improvement. The inspira-tionalist school of creativity encourages viewingmany relevant examples of previous work to engageinnovators in a creative mind-set. Faceted search(simultaneous menus on independent aspects suchas people, geography, and time) helps guide users byproviding compact visual cues about attributes andattributes values. Dynamic queries (changes to slid-ers, selectors, and filters, producing rapid changes todisplays) support rapid incremental and reversibleexploration that enables users to learn about distrib-utions, gaps, and outliers. Improved search servicesprovide rich mechanisms for organizing searchresults by ranking, clustering, and partitioning withample tools for annotation, tagging, and marking.Advanced search services also enable seamless col-laboration with shared views, chat rooms, andemailing of result sets. Since serious discovery andinnovation may require group processes that last forweeks or months, as in legal, patent, or scientificarticle searches, history-keeping facilities are helpful,as are overviews of what has been done and whatremains to be accomplished.

Enable collaboration: While the Aha! momentsof discovery and innovation are very personal, theprocesses that lead to them are often highly collabo-rative. Inspirationalists and situationalists claim col-laborations at early stages revolve around problemdefinition and setting goals, so consultations must behandled carefully because innovators fear rejection,ridicule, and rip-off. Communications systems thatlet users expose their uncertainties in a safe environ-ment could help build trust, and designs that recordwho said what can document contributions to emerg-ing ideas. Trust, accurate records, and safe exchangesare also needed in the middle stages when informa-tion gathering, idea refinement, and knowledgeablepartners are important. In later stages, when valida-tion and dissemination become dominant, findingappropriate test situations, preview audiences, andmedia partners is helpful. These processes are wellunderstood for individuals and small groups, buttechnology support for them is marginal. For thelarger communities engaged in social creativity,wholly new forms of collaboration are emerging.Wikipedia and its support environment, Wikimedia,have proven to be remarkable and surprising successstories, defying expectations by finding a good bal-ance between freewheeling individual effort and well-enforced administrative principles. Each individualcontribution to Wikipedia may be small, but the


Wikimedia environment produces an intense collabo-rative effort that leads to an impressive and originalproduct. Thomas Edison famously quipped thatgenius is 1% inspiration and 99% perspiration, butnow he might remark that innovation is 1% inspira-tion and 99% collaboration.

Provide rich history-keeping:Many people believediscovery and innovation processes take many forms,so it is difficult to provide precise guidance in a step-by-step manner. But after casting aside rigid and doc-trinaire strategies, semi-structured methods or at leastan orderly process has repeatedly been shown to bebeneficial. The structuralist school embraces system-atic approaches; sometimes around the traditionalphases of preparation, incubation, illumination, andverification, but often around more carefully definedmethods such as the 40 potential phases of TRIZ (seewww.triz.org). While Thomas Edison tested morethan 4,000 filament variations for his light bulb,newer forms of structured discovery apply computer-based exhaustive search of millions of cases to under-stand optimal conditions and relationships amongparameters. Whether discoverers and innovators usedstructured or free-form thinking, the benefits of richhistory-keeping are apparent. Users have a record ofwhich alternatives they have tried, they can comparethe many alternatives, and they can go back to earlieralternatives to make modifications. History-keepingon computers has still more benefits, such as sendinginteresting cases to colleagues for comments and cre-ating macro processes that can be run repeatedly onnew data.

Design with low thresholds, high ceilings, andwide walls: This metaphoric description of desirableattributes for creativity support tools suggests thattools should be easy for novices to begin using, yetprovide ambitious functionality that experts need.Good tools should also have a wide range of func-tionality so that many different services are provided,from data input and statistical analyses to report gen-eration. A single tool with a uniform user interfacereduces frustrating file conversions and enables usersto concentrate on their problems. Of course, there arelimits to what one tool can do and also good argu-ments for modular designs, as well as domain-specificvariations. Still, when users can import datasets easily,handle missing data, transform values, try multiplevisualizations, run statistical tests, include annota-tions, and export subsets of data in desired formats,then they are free to concentrate on their explorationrather than cleaning data, recording comments, andtransforming file formats. One strategy for satisfyingthis principle is to use a multilayer interface designthat allows novices to begin a first layer and move up

as their experience increases and needs require. Manyvideo games have dozens of layers, most searchengines have novice and advanced layers (Google,Yahoo), many art and video tools have three or moreworkspaces (Apple Final Cut Pro, Adobe Premiere),and some tools have as many as eight layers to accom-modate a wide range of expertise and ambition.

RIGOROUS RESEARCH METHODS

These design principles for creativity support toolsand the tools themselves would be difficult to vali-date with controlled studies that measured time tocorrect completion of benchmark tasks. Approxi-mately 300 years of scientific methods based on areductionist model and controlled experimentalstudies have produced huge benefits, but the com-plex nature of human discovery and innovation can-not be studied like pendulums or solid-statematerials [1, 3].Researchers are beginning to understand that

design of discovery and innovation tools is a worthysubject of study, but they are often torn by devotionto traditional controlled studies. They also facepressure from many journal and conference review-ers, who favor statistically significant results, evenwhen laboratory-controlled studies with many par-ticipants are inappropriate. The emphasis on closestudy of domain experts as they make discoverieshas led many researchers to adopt case study, obser-vational, and interview methods with small num-bers of users over weeks and months. Their goal isto capture the processes that precede breakthroughincidents and to collect evidence that supportshypotheses about how software design features pro-mote creative moments.The intense desire for validity that comes from

close observations has led many researchers to takefresh approaches to other research goals like replica-bility and generalizability. Until many more case stud-ies are collected and many related problems arestudied, carefully documented methods are needed toanswer critics who are legitimately concerned aboutmisleading interpretations based on experimenter bias[4]. Individual case studies are meant to provoke mul-tiple case studies that replicate findings with diverseusers and problems. As multiple case studies replicateresults, researchers gain confidence in the replicabilityand generalizability of cause-and-effect conjectures.Many researchers have already demonstrated highpayoffs in understanding how powerful tools can sup-port creative people. These researchers also argue thatcreative work in science, design, or the arts evolves sorapidly that replicability has a different meaning thanin physical sciences research where the properties of



Although creative individuals are often thought of as workingin isolation, much of our intelligence and creativity resultsfrom interaction with tools and artifacts and from collaborat-ing with other individuals [4]. Many traditional approaches tocreativity, however, have disregarded the importance of socialand material surroundings. A richer framework would enablesoftware developers, interaction designers, content man-agers, and researchers to understand the opportunities, chal-lenges, and principles of social creativity.

SSoocciiaall CCrreeaattiivviittyy.. As demonstrated by leading scientists,movie directors, and influential politicians, an individual’sskills and experience can significantly influence the success ofa project. Individual and social creativitycan and must complement each other. Incomplex design problems such as urbanplanning, for example, the difference inknowledge, expertise, and perspectivesthat exist among individuals provides theground to collaborate toward more creativeand sustainable solutions [1]. The same istrue for scientific collaboratories [8], inwhich computer scientists work withdomain specialists to accelerate knowledgedevelopment and dissemination, and for creative practices suchas art and technology collaboration, the results of which super-sede what a single artist or computer scientist could haveachieved in isolation [3, 9]. Environments supporting mass col-laboration and social production such as annotated collections(GenBank), media sharing (Flickr, YouTube), wikis (Wikipedia),folksonomies (del.icio.us), and virtual worlds (Second Life) areother examples of social creativity. The diverse and collectivestock of scientific content and artistic or stylistic ideas thatindividuals and communities share, reinterpret, and use as abasis for new ideas and visions constitutes the vital source ofinvention and creativity.

FFooccuussiinngg oonn CCoommmmuunniittiieess.. Socio-technical environmentsare necessary for communities to collaborate and bring socialcreativity alive: to express themselves, combine different per-spectives, and generate new understandings. In large and het-erogeneous groups working together for long periods of timeover complex design problems, as well as in communitiesincluding individuals with diverse but converging goals andintentions, distances and diversity between contributing indi-viduals can enhance creativity rather than hinder it. The chal-lenge is not to reduce heterogeneity and specialization but to

support it and manage it at both the technological and sociallevel by finding ways to build bridges between individuals andexploiting conceptual collisions and breakdowns to stimulateimagination and invention. The distances are distributed inmultiple dimensions: spatially (across physical distance), tem-porally (across time), technologically (across artifacts), andconceptually (across individuals, communities, and cultures)[5]. This distribution provides a foundation for social creativityby making all voices heard, harnessing diversity, and enablingpeople to be aware of and access each other’s work and ideas,relate them to their own, and contribute the results back to thecommunity.

CCoonnssttrruuccttiinngg SSoocciioo--TTeecchhnniiccaall EEnnvviirroonn--mmeennttss.. Because social creativity has “nohead,” externalizations are vital to socio-technical environments. Externalizationssupport creativity by: producing a record ofour mental efforts; causing us to movefrom vague mental conceptualizations ofan idea to a more concrete representationof it; making thoughts and intentions moreaccessible for personal reflection; and pro-viding a means by which other individuals

can interact with, react to, negotiate around, and build upon anidea. One example of this is when sketching is used to shareand negotiate design concepts. Externalizations support cre-ativity by enabling individuals also to: sense new aspects ofthe environment experienced by other individuals; interactwith it in new ways; and socially unfold and interpret emotionsand behaviors [7]. One example is Flickr, where people candevelop new photographic styles by looking at how other peo-ple have photographed or “sensed” their environment.Another example is our collaborative mapping system TheSilence of the Lands (www.thesilence.org), in which geo-located information is overlapped with a person’s individualperceptions of specific aspects of the sonic environment. Fromthis enriched perspective, externalizations for social creativity(shared representations, pictures, sounds, and so forth) mustbe thought of as an instrument for creative conversations,rather than a tool for categorization.

HHaarrnneessssiinngg tthhee SSyynneerrggyy ooff MMaannyy bbyy MMeettaa--ddeessiiggnn.. Creativ-ity needs the “synergy of many” [2], and this kind of synergycan be facilitated by meta-design. Meta-design is a socio-technical approach that characterizes objectives, techniques,and processes that allow users to act as designers and be cre-

SUSTAINING SOCIAL CREATIVITY

By Gerhard Fischer and Elisa Giaccardi

Creativity needs the “synergy ofmany.”�

�

electrons or tensile strength of steel can be studiedagain and again under diverse conditions. In medi-cine, business, and other research domains, when casestudy methods are based on established procedures tolimit bias, they are accepted as valuable contributions.Year-long studies of artist-technologist collabora-

tions (as described in the sidebar “New Media Artsand the Future of Technologies”) and a long traditionof ethnographic research have influenced the multi -dimensional in-depth long-term case studies, whichare emerging as an accepted research method for sci-entific discovery and design innovation [10]. The keyidea is to closely study domain experts who are work-ing on their own problems over a period of weeks ormonths. This is a form of hypothesis testing, in whichthe goal is to collect quantitative and qualitative evi-dence about how a creativity support tool benefits itsusers. When the focus is on documenting and under-standing how specific features contribute to successfuloutcomes, the researchers often produce insights thathave substantial and broad value.Researchers become more than participant

observers, as they may help the users to apply the tooleffectively while recording their reactions. In a grow-ing number of studies, once-a-week visits for one totwo hours over a one- to four-month period enabledparticipant observers to gather evidence about whatworked and what did not. The users benefit by hav-ing access to novel technologies and the participationof sympathetic researchers, who are eager to see theusers succeed. Careful logs of tool usage and audio orvideo recordings document critical incidents andreveal problems with the tool design or usage. Theclose linkage of researchers and users violates tradi-tional experimental design principles, but it seemsnecessary to understand creative processes that involveindividual, group, and social environments. Comple-mentary analyses from usage logs, interviews, surveys,or focus groups—usually referred to as triangula-tion—can contribute additional insights and increaseperceived validity.

CONCLUSIONCreativity support tools have been around for aslong as people have been creative. However, design-ers of modern computer-based environments areenabling new discovery and innovation processesfor individuals, groups, and communities. In orderto improve their design, they need refined theoriesand rigorous empirical studies based on newresearch methods. The close collaboration requiredby multidimensional in-depth long-term case stud-ies can produce breakthrough insights about howdiscovery and innovation occur.


ative in personally meaningful activities [7]. However, a ten-sion exists between creativity and organization. A definingcharacteristic of social creativity is that it transcends individualcreativity and thus requires some form of organization; butelements of organization can and frequently do stifle creativity[6]. Advocates of meta-design address such a challenge andpromote the importance of keeping environments open tousers’ modifications and adaptations by technical and socialmeans that empower participation. This serves a double pur-pose: to provide a potential source for new insights, newknowledge, and new understandings; and to provide a higherdegree of synergy and self-organization. Meta-design canfacilitate social creativity by shifting the focus from finishedproducts or complete solutions to conditions for many toexplore mismatches and embrace new emerging opportunitiesduring use. How to generate and sustain these conditionseffectively is the next challenge for software developers, inter-action designers, content managers, and researchers.

References1. Arias, E.G. et al. Transcending the individual human mind—Creatingshared understanding through collaborative design. ACM Transactionson Computer Human-Interaction 7, 1 (2000), 84–113.

2. Benkler, Y. The Wealth of Networks: How Social Production TransformsMarkets and Freedom. Yale University Press, New Haven, 2006.

3. Candy, L. and Edmonds, E.A. Explorations in Art and Technology.Springer-Verlag, London, 2002.

4. Csikszentmihalyi, M. Creativity—Flow and the Psychology of Discoveryand Invention. HarperCollins Publishers, New York, NY, 1996.

5. Fischer, G. Distances and diversity: Sources for social creativity. In Pro-ceedings of Creativity and Cognition, (London, April 2005), 128–136.

6. Florida, R. The Rise of the Creative Class and How It’s Transforming Work,Leisure, Community and Everyday Life. Basic Books, New York, NY, 2002.

7. Giaccardi, E. and Fischer, G. Creativity and Evolution: A Metadesign Per-spective. Digital Creativity, (forthcoming).

8. Kouzes, R.T., Myers, J.D., and Wulf, W.A. Collaboratories: Doing sci-ence on the Internet. IEEE Computer 29, 8 (1996), 40–46.

9. National Research Council. Beyond Productivity: Information Technology,Innovation, and Creativity. National Academy Press, Washington, D.C.,2003.

Gerhard Fischer ([email protected]) is a professor in theDepartment of Computer Science, a Fellow of the Institute of Cognitive Science, and the director of the Center for LifeLongLearning and Design (L3D) at the University of Colorado in Boulder, CO. Elisa Giaccardi ([email protected]) is a research associ-ate at the University of Colorado Center for LifeLong Learning andDesign in Boulder, CO.

© 2007 ACM 0001-0782/07/1200 $5.00

c


New technology is increasingly becoming an integral part ofcreative practice in the arts. This offers an exciting and chal-lenging opportunity for computer scientists, software engi-neers, and interface designers. To guide their efforts for theaudience of artistic end users, developers will need to knowmore about innovative uses of technology by new mediaartists.

One of the ACM Education Board’s Great Principles of Com-puting project’s four core practices is innovating [4]. Innovationis the process of creating something new that is then adopted inpractice. We need to find ways of drawing more attention to thisaspect of our work and placing concerns for innovation and cre-ativity more centrally in thecanon of methods andterms we use. For innova-tion to take place, however,complex social, organiza-tional, and cultural factorsmust be harnessed; on theother hand, strategies forsupporting creativity aredirectly within the provinceof the computer sciencecommunity via the designand development of cre-ativity support technolo-gies. This is the primaryfocus of this sidebar.

In advancing this ideawe could benefit fromapplying the principle of“Hitching our research to someone else’s driving problems,and solving those problems on the owners’ terms,” which canlead to “richer computer science research” [1]. What betterway forward than to begin by focusing our attention on indis-putably creative application domains and how they meet theneeds of creative professionals and their ever-expandingdemand for more effective computer systems? As an example,consider image manipulation and well-developed applicationssuch as Photoshop, which have achieved widespread successand adoption. The use of a multi-layered approach [11] is pow-erful in the way it enables the user to start with a basic set offacilities and go deeper in as and when more functionality isneeded. This enables the user to focus on the essentials of thecreative task without the hindrance of having to grapple with

unhelpful complexities imposed by the system. Creativity sup-port systems can support the user by encouraging explorationand enhancing collaboration and if technology developersdesign with creative cognition in mind and recognize the vitalrole of the social context of creativity, the tools they developwill be more effective.

AARRTTIISSTTSS AASS PPOOWWEERR UUSSEERRSS:: AA SSYYSSTTEEMMSS PPEERRSSPPEECCTTIIVVEETaking the focus on creativity support further, there aretoday many examples of creativity enhancing systems beingdeveloped by artists working with new media [2]. Someartists are creative power users who need to access the full

capability of the com-puter: they either performprogramming tasks orcollaborate closely withprogrammers. This kindof artist user is a specialkind of programmer need-ing dynamic and flexiblesoftware environments.

Consider the exampleof Andrew Johnston, asoftware engineer andtrombonist, who in col-laboration with BenMarks, a musician com-

poser, has developed a software instrument that responds tosomeone playing or singing by creating sound and visualimages that capture and reflect the human performance. Musi-cians have described it as rather like playing with a partner.This collaboration, initially driven by creative impulses ratherthan scientific ones, is an example of how creative work acrossdifferent disciplines can not only lead to new tools for creativeexpression but also yield new insights into principles that caninform the design of innovative digital tools. Johnston’s sys-tem is written in Pd (Pure Data). One aspect of his methodinvolves developing the code during sessions in which themusicians use the system. This process involves developingthe code (or parts of it) with the musicians as they explore.Johnston likes to call his software constructs toys, things withwhich people constructively play. Making art systems mightseem a little esoteric and making games rather limited, butmaking creative toys opens up new opportunities for explo-ration and collaboration through playfulness [7].

By Linda Candy

A trombonist plays with an interactive musical instrument from

“Partial Reflections” by Andrew Johnston and Benjamin Marks.

NEW MEDIA ARTS AND THE FUTURE OF TECHNOLOGIES


Art making is as much (or more) about creating new formsas creating new content. In the former case, the artist requiresmore than application-level software to fully support the work.Pd is an environment in the Max/MSP family, which Lyon calledan instrument design language [9]. Artists in many media usesuch systems to build instruments, or software environments,that need to be extended, personalized, or integrated. What isoften needed is a context-dependent environment. Systemslike Max/MSP are often used to build specific developmentenvironments that then enable the artist to explore their partic-ular concerns. In collaboration with Prix Nica winner YasunaoTone, a new audiovisual instrument was constructed usingcomputational representations of the concepts at the center ofTone’s creative practice [5].

John Maeda has argued for more emphasis on the use of“software sketching,” where a loose and initially poorly definedpiece of code can be constructed in order to help move the idea-generation process forward [10]. Creative practitioners spendmuch important effort on exploration and problem finding. Thisrequires a fluid engagement with the materials at hand—muchof this is software. Maeda has drawn upon very successfulmodels of software environment styles, such as Max/MSP [3],which are both visual and dynamic and hence supportive of therapid exploration and feedback needed in the idea-generationprocess. The need for dynamic, flexible, and complete softwareenvironments is central to creative users. Because these envi-ronments must be situated in or adaptable to specific domains(such as music), and these domains are themselves shifting,this poses key challenges for future technologies.

TTAALLKKIINNGG AABBOOUUTT CCRREEAATTIIVVIITTYY:: AANN EEXXTTEENNDDEEDD VVOOCCAABBUULLAARRYYAs well as pointing to a specific focus in software developmentapproaches, the study of creative digital practitioners leads to aneed to develop and extend the vocabulary we use to discussinteractive systems. Jonas Lowgren suggests we need to regu-larly use terms that do not necessarily derive from a work-oriented view (including relevance and usefulness). Instead, hesuggests playability, seductivity, and fluency, for example. Healso points out that ambiguity can be a positive property inrelation to creativity [8]. For the artist, the term engagementmay be a more significant than usability, and is already used byresearchers and software developers. Because building systemsfor creative engagement is still mostly confined to new mediaarts we must make an effort to extend it to other kinds of inter-active systems. With an extended vocabulary, software designcan grapple with the issues of creativity in a much more com-prehensive way.

The ever-expanding demand for more effective computersystems by creative practitioners in the arts, music, literature,and new media, can guide designers in meeting the needs of allcreative professionals. The arrival of complex art systems hasopened up new avenues that are impacting our thinking aboutwhat is possible and desirable. We are already familiar with howcomputing can impact the creative arts, but now we see theemergence of creative communities that can influence comput-ing. This is a lively two-way process that benefits both com-munities. Art is impacting technology and technology isimpacting art. On a wide front exciting changes are in the air.The emergence of communities of research and practice thatare exploring how the arts can influence computing is an excit-ing prospect. The role of aesthetics in programming and othercore elements of computer science is a challenging develop-ment for the future of computing. The recent work of artists aspower users follows the evolutionary trend of artists beginningwith their use of the computer as a tool, and more recently asraw material that can be easily modified for creative applica-tions. The next step beyond software as material is to viewsoftware as the subject material [6] and to use creativity as ameans to redesign software itself.

REFERENCES1. Brooks, F. The computer scientist as toolsmith II. Commun. ACM 39,3 (Mar. 1996), 61–68.

2. Candy, L. and Edmonds, E.A. Explorations in Art and Technology.Springer-Verlag, 2002.

3. Cycling 74. Tools for new media; www.cycling74.com/.4. Denning, P.J. Great principles of computing, Commun. ACM 46, 11(Nov. 2003), 15–20.

5. Edmonds, E.A. et al. Developing interactive art using visual program-ming. In C. Stephanis and J. Jacko, Eds., Proceedings of Human-Com-puter Interaction 2003. Lawrence Erlbaum, London, 2003, 1183–1187.

6. Fishwick, P. An introduction to aesthetic computing. Aesthetic Com-puting. The MIT Press, Cambridge, MA, 2006, 3–27.

7. Johnston, A. and Marks, B. Partial reflections. Leonardo Transactions40, 5 (Oct. 2007), 510–511.

8. Löwgren, J. Articulating the use qualities of digital designs. In P. Fish-wick, Ed., Aesthetic Computing. The MIT Press, Cambridge, MA,2006, 383–404.

9. Lyon, E. Dartmouth symposium on the future of computer music software:A panel discussion. Computer Music Journal 26, 4 (Apr. 2002), 13–30.

10. Reas, C. and Fry, B. Processing: A Programming Handbook for VisualDesigners and Artists. The MIT Press, Cambridge, MA, 2007.

11. Shneiderman, B. Promoting universal usability with multi-layer inter-face design. In Proceedings of the Conference on Universal Usability(CUU’03), ACM Press, 2003.

Linda Candy ([email protected]) is Senior Associate andHonorary Research Fellow in the Creativity and Cognition Studios,Faculty of IT, at the University of Technology, Sydney, Australia.

© 2007 ACM 0001-0782/07/1200 $5.00

c

The growth of interest in creativity support tools in recent years is gratifying. The June 2005 U.S.National Science Foundation-sponsored workshopon the topic [9] inspired research under the CreativeIT program, a strong commitment to discov-ery and innovation research in the NSF five-yearstrategic plan [7], and the ambitious billion-dollarvision for cyber-enabled discovery and innovationresearch. The risks are high and the scientific methodsnovel, but the payoffs are substantial in bringingabout thrilling moments of scientific discovery andengineering innovation.

References1. Basili, V.R., Shull, F., and Lanubile, F. Building knowledge throughfamilies of experiments. IEEE Trans. Software Engineering 25, 4 (Apr.1999), 456–473.

2. Csikszentmihalyi, M. Creativity: Flow and the Psychology of Discoveryand Invention. HarperCollins, New York, 1996.

3. Fjermestad, J. and Hiltz S.R. Group support systems: A descriptiveevaluation of case and field studies. Journal of Management InformationSystems 17, 3 (Mar. 2000), 113–157.

4. Hewett, T. Informing the design of computer-based environments tosupport creativity. International Journal of Human-Computer Studies63, 4–5, Special Issue on Computer Support for Creativity, E.Edmonds and L. Candy, Eds., (2005), 383–409.

5. Myers B.A., Hudson, S.E., and Pausch, R. Past, present and future ofuser interface software tools. ACM Transactions on Computer HumanInteraction 7, 1 (Mar. 2000), 3–28.

6. National Academy of Sciences. Beyond Productivity: Information Tech-nology, Innovation and Creativity. NAS Press, Washington, D.C., 2003.

7. National Science Foundation. Investing in America’s Future: StrategicPlan 2006–2011. Arlington, VA, 2006.

8. Shneiderman, B. Leonardo’s Laptop: Human Needs and the New Com-puting Technologies. MIT Press, Cambridge, MA, 2002.

9. Shneiderman, B. et al. Creativity support tools: Report from a U.S.National Science Foundation sponsored workshop. International Jour-nal of Human–Computer Interaction 20, 2 (2006), 61–77;www.cs.umd. edu/hcil/CST.

10. Shneiderman, B. and Plaisant, C. Strategies for evaluating infor-mation visualization tools: Multi-dimensional in-depth long-termcase studies. In Proceedings of Beyond Time and Errors: Novel Eval-uation Methods for Information Visualization, Workshop of theAdvanced Visual Interfaces Conference (2006); http://portal.acm.org/citation.cfm?id=1168149.1168158.

11. Sternberg, R., Ed. Handbook of Creativity. Cambridge University Press,Cambridge, U.K., 1999.

12. Terry, M., Mynatt, E.D., Nakakoji, K., and Yamamoto, Y. Variationin element and action: Supporting simultaneous development of alter-native solutions. In Proceedings of CHI 2004 Conference on Human Fac-tors in Computing Systems, ACM Press, New York (2004), 711–718.

Ben Shneiderman (www.cs.umd.edu/~ben) is a professor in theDepartment of Computer Science, the founding director (1983–2000)of the Human-Computer Interaction Laboratory (www.cs.umd.edu/hcil/), and a member of the Institute for Advanced ComputerStudies at the University of Maryland at College Park.

Permission to make digital or hard copies of all or part of this work for personal or class-room use is granted without fee provided that copies are not made or distributed forprofit or commercial advantage and that copies bear this notice and the full citation onthe first page. To copy otherwise, to republish, to post on servers or to redistribute tolists, requires prior specific permission and/or a fee.

© 2007 ACM 0001-0782/07/1200 $5.00

c


creativity support tools - cs.umd.edu

Documents