UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Digital art on the World Wide Web, 1996-1997 Johnson, Mia 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1997-250741.pdf [ 20.44MB ]
JSON: 831-1.0054696.json
JSON-LD: 831-1.0054696-ld.json
RDF/XML (Pretty): 831-1.0054696-rdf.xml
RDF/JSON: 831-1.0054696-rdf.json
Turtle: 831-1.0054696-turtle.txt
N-Triples: 831-1.0054696-rdf-ntriples.txt
Original Record: 831-1.0054696-source.json
Full Text

Full Text

DIGITAL ART ON THE WORLD WIDE WEB, 1996-1997 by MIA JOHNSON B.Ed. Simon Fraser University 1971 M.A. University of British Columbia 1993  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Curriculum Studies)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA October 1997 © Mia Johnson  l^"^  In presenting this thesis  in partial fulfilment  degree at the University of  of  the  requirements  for  an advanced  British Columbia, I agree that the Library shall make it  freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department  or  by  his  or  her  representatives.  It  is  understood  that  copying  or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  /><_^.vv\ r ' j a v u v w ,  The University of British Columbia Vancouver, Canada Date  DE-6 (2/88)  />^r\,  \  f\^  Abstract This study draws upon what Howard Becker (1984) called "conventions" or characteristics that identify an art world, to determine conventions in the use of tools and techniques by digital artists. Two hundred images were sampled from approximately 600 "digital art" sites on the World Wide Web during 1996. A taxonomy of the characteristics or properties of these images was developed, partly by the researcher alone, and partly by the researcher in conjunction with five independent raters. Digital art conventions were identified, defined and analyzed in five categories: formal properties, content, styles, and digital properties unique to 2D and 3D graphics. The frequencies of 98 characteristics in these categories were then compared by gender, types of sites, types of programs and to each other. This study provides field-based groundwork for educators interested in developing curriculum for digital art, using approaches afforded by Discipline-based Art Education (DBAE). Following the summary of data, content analysis and discussion, this study assesses the findings in terms of their implications for discipline-based digital art production, aesthetics, criticism, and digital art history.  n  TABLE OF CONTENTS PAGE Abstract  ii  Table of contents  iii  List of tables  vii  List of figures  ix  Acknowledgments  xi  Dedication  xii  CHAPTER I.  II.  INTRODUCTION TO THE STUDY General statement  1  Background to the problem  4  Purpose of the study  7  Research questions  9  Theoretical perspective  10  Personal grounds  12  Definition of terms  13  Delimitations of the study  14  Limitations  16  Significance of the study  18  REVIEW OF THE LITERATURE Introduction  20  Discipline-Based Art Education  20  Art production.  22  Art history  23  Aesthetics in art  23  Art criticism  24  iii  Research in computer art education Previous research in computer art education  25  Reviews of hardware and software  26  Ethnographic studies and observations  27  Action research studies  29  Summary of research problems  31  Art worlds Definition of art worlds  33  Function of art worlds  34  Conventions used by art worlds  34  Implementing a discipline-based approach to computer art education  35  Summary  38  III. D E S I G N OF T H E STUDY A N D METHODS OF D A T A COLLECTION Introduction  39  Content analysis  41  Research d e s i g n  43  Population Target population  45  Identification of digital art  46  Sample Procedure  47  Selection of sample  48  Variables in the selection  50  iv  Instrumentation Instrument design  51  Selection of categories  52  Identification of category items Formal properties  52  Content  53  Style  54  Digital properties  ".  54  Definition of items  56  Pictorial examples  57  Internal validity Procedures used in conjunction with raters..  80  Selection of raters. ...;..  81  Interrater agreement  82  Variables in interrater agreement  84  Procedures for a stability estimate of interrater-"" agreement  ,  Data analysis IV.  85 85  ANALYSIS AND DISCUSSION OF DATA Introduction  90  Findings and interpretations Gender.  90  Site  106  Type  113  Style  116  Formal properties  128  Content  153 v  2D digital properties  159  3D digital properties  165  Summary V.  169  RECOMMENDATIONS Introduction  173  Recommendations for art production domain  175  Recommendations for art historical domain  182  Recommendations for aesthetics domain. :  189  Recommendations for art criticism domain  196  Recommendations for further research  201  Conclusion  205  EPILOGUE  207  REFERENCES  213  APPENDIX A. Credits for sample images  225  APPENDIX B. Example of directory listing  228  APPENDIX C. Sample of recording taxonomy  22 9  APPENDIX D. Sample of rater taxonomy  23 0  APPENDIX E. Sample of SPSS output  232  APPENDIX F. Values and variables  233  APPENDIX G. Sample of style identification  235  APPENDIX H. Example of home page  236  vi  LIST OF TABLES TABLE I.  Rater profiles  82  2.  Percentages of interrater agreement  83  3.  Percentages of males and females using each formal property  94  Percentages of males and females using each type of content  96  4.  5.  'Percentages of males and females using each 2D digital property  100  Percentages of males and females using each 3D digital property  105  7.  Percentages of sites producing each style  109  8.  Percentages of each 2D digital property used by artists at each type of site  Ill  Percentages of each 3D digital property used by artists at each type of site  112  6.  9.  10. Frequencies of comparable 2D and 3D techniques... 115 II. Preferences for style by gender  119  12. Percentages of styles at each type of site  120  13. Percentages of each style using 2D or 3D graphics  121  14. Percentages of 2D digital properties found in each style  123  15. Percentages of 3D digital properties found in each style  126  16. Percentages of formal properties in 2D and 3D images  12 9  vii  17. Percentages of each format type in 2D and 3D images  13 0  18. Percentages of each format type found by site....131 19. Percentages of each format type found in each style  132  20. Percentages of each content type by site  154  21. Percentages of each content type used by males and females  157  22. Percentages of 2D properties in 2D imagery  159  23. Percentages of 2D properties used by males and females  162  24. Percentages of 3D properties used by males and females  166  25. Percentages of 3D properties found in each style  167  viii  LIST OF FIGURES FIGURES I.  Percentages of males and females at each type of site  91  Percentages of sites by gender  92  3 . Gender preferences for content  97  2.  4.  Percentages of males and females using 2D and 3D graphics  5.  98  Number of males and females using each style  102  6 . '• Percentages of styles by gender  . 102  7.  Gender preferences for 3D digital properties  104  8.  Percentages of 2D and 3D art at each type of  9.  site  108  Percentages of styles  118  10. Percentages of artists at each type of site using realistic and imaginary proportions II. Percentages of realistic and imaginary scale in each style of art 12. Percentages of formats with realistic and imaginary scale  134 135 13 6  13. Percentages of limited and unlimited palettes by format type  13 8  14. Percentages of high and low resolution by type of site  141  15. Percentages of high and low resolution by format type  143  16 . Gender use of field of view  145  17. Percentages of cropped and uncropped views by type of style  146  xx  18. Percentages of 2D and 3D images with image dividers  152  19. Percentages of mechanical imagery by gender  158  20. Percentages of 2D and 3D work by gender  160  21. Percentages of area fills found in styles of art  164  x  Acknowledgments I would like to acknowledge and thank the Social Sciences and Research Council of Canada and the J. Paul Getty Center for Education in the Arts for the fellowships that supported this study. I am indebted to my captain, Dr. Ron MacGregor, who demonstrated great courage and fortitude by embarking into uncharted "virtual" space and steering us safely on course. I would also like to thank Dr. Rita Irwin for introducing me to the work of Howard Becker, for her commitment to my academic development, and for her always generous spirit,and Dr. Dave Forsey of UBC's MAGIC Lab for his infinite patience and enthusiasm. Dave, you are a born teacher, and my own personal exemplar of a computer graphics' instructor. Personal thanks go to the many technicians, instructors, and friends who have succored me throughout. Melenie, Deborah and Moreka, I could not have undertaken graduate studies without your constant support. In particular, I would like to express my gratitude to Ken Stoddart for the "present of his self in our everyday life", as well as his on-going contributions to my understanding of sociology. And yes, my darling daughter Laurel, you may have the computer back now...  xi  To the memory of my mother and father, who would have been proud  xii  Chapter 1 Introduction to the study  General statement One primary purpose of art education in the 20th Century has been to educate students about the visible phenomena or characteristics of artworks. In the spirit of this convention, this study sought to observe characteristics found in computer artworks: specifically, to categorize digital art phenomena, to create terms for analyzing and discussing digital art, and to compare frequencies of observed phenomena. For the purposes of art education, these characteristics are as yet uncategorized and unnamed. This study was delimited to digital art on the World Wide Web Web in order to sample the largest possible target population of computer artists available. During the time of this study (1996), more than 600 sites exhibited changing displays of up to 50 artists: numbers which could not be accommodated by such traditional venues as concrete galleries or print publications. Digital art as well as computer art output in a twodimensional format was found to form the bulk of work on the Web created with computer technology for art and design. It should be noted that this emphasis may have resulted from the development of the World Wide Web as a stimulating new forum for digital artists in particular, as discussed in  1  Chapter 3, and does not preclude the importance of other forms of computer art. At the time of the study, however, the Web was not technologically capable of displaying other such forms of computer art as computer-assisted video or animations. This study examined 200 samples of two-dimensional computer art or "digital art" on the Web. "Two-dimensional" here refers to the image display, rather than to the kind of software used to create the imagery. Two-dimensional computer art has been termed "digital art" in this study to avoid confusion with "2D software", and includes art produced using Draw, Paint, image-processing or 3D software. Ideas about the characteristics of digital artworks in this study were derived from a statistical study of observed phenomena in Web images, and from notes on characteristics as they were found to be shaped by the general conditions of the developing culture of the World Wide Web. This study is an attempt to identify some constituents of an art education curriculum expanded to include digital art. The findings of this study, however, do not in themselves comprise aesthetics theory or a curriculum. They are intended to be used as a starting point for curriculum development in digital art education; as an initial survey for further research in digital art as well as other areas of computer art; or as a historical baseline for future research about computer art on the Web. As Franke put it, "The statistical analysis of works of art is not itself to 2  be regarded as an aesthetic theory; rather, it can count as an aid in theory construction and is of decisive theoretical as well as practical significance" (Franke, 1985, p. 155). There are a number of underlying assumptions at the heart of this study. The first assumption is that verbal cues ("naming" or giving terms to art phenomena) play a vital role in understanding and recognizing art phenomena (Koroscik, 1983). Regardless of the particular purpose and focus of an art program, this study assumes that terminology particular to the art in that program plays a crucial role in comprehension and appreciation. For example, it aids in the processes of description and analysis, articulation of stylistic concepts, and the expression of critical judgments. A second assumption of this study is that terminology used to signify phenomena is socially constructed. This study thus relied on the following for identification and categorization: the application of terms as they are already established in the literature of art education (including curriculum guides); the application of terms as they are already established in the literature of computer graphics (including technical manuals); a comparison of these previously established terms for the perceivable characteristics of traditional and digital artworks (Johnson, 1993); and the ways in which terms were used and defined by independent raters who had previous education and experience in both traditional art and computer graphics. 3  This study acknowledges that the degrees of importance assigned to phenomena are also a social construct. However, it does not seek to establish an agenda for art educators, since even within the curriculum construct that has been utilized as a research framework--that of Discipline-Based Art Education--there are a wide array of aims and practices (Wilson, 1997). In summary, this study sought to observe the visible characteristics of digital artworks; to propose categories of classification for observed phenomena; and to propose terminology rooted in considerations of the practical usage of terms in both the field of computer graphics and in prior traditional art education. It provides correlations of the frequencies of observed phenomena and recommends that educators evaluate the meaningfulness of relatively high or low counts according to their individual agendas. This study, again, intends to serve as basic research on one area of the field of computer art.  Background to the problem There is a need for information about digital art as a new medium in computer art education. Computers have become pervasive in well-funded school systems. Many art educators have added software programs with two-dimensional output to their repertoire of media. Research papers such as those by Ettinger (1988), Chia and Duthie (1992), and Madeja (1993) advocate the inclusion of digital art in elementary and 4  secondary curricula. Respectively, they variously perceive digital art as a new medium, as a means to promote cooperative activity among students, or as a prerequisite for entry into fields of study related to visual art. Computer technologies for art and design, as they have been robustly developed over the past 3 0 years, can conceivably provide a wealth of opportunities for art education. Yet because of the rapid development of these technologies, many educators lack knowledge of their diversity and complexity. In addition, conflicting beliefs about computer art are presented in the literature. For example, computer artworks have been construed as "decidedly machine-1ike ... and aesthetically distasteful" (Youngblood, 1988, p. 23); or, conversely, as images that "cannot be understood outside the complete interactive system in which they occur" (Binkley, 1990, p. 13). On one hand, computer artworks are perceived as evidencing "the embedment [sic] of previous aesthetic theories and reality constructs" (Jones, 1989, p. 37) in "an extension of older art forms" (Roland, 1990, p. 60). On the other, they are portrayed as "new expressive forms" (Freedman, 1991, p. 47), and "uniquely different from what is possible by conventional means" (Hubbard & Greh, 1991, p. 20) . Such generalizations, together with a lack of knowledge, make it difficult for practitioners to formulate curriculum models for computer art. A curriculum model is a framework for structuring the goals and objectives of formal 5  education. Whether that framework be student-centered, subject-centered, society-centered or a combination of such aims, its formulation precedes the development of content and goals (Eisner, 1987, p. 19). More than 80 research studies in the area of computers for art and design have been conducted in art education (Johnson, 1996). Unfortunately, they often have been limited in scope, and have not yet been synthesized into the forms usually taken by curricula. For example, ethnographic studies like those by Bhargava (1992) and Chang (1994) suggest that students can transfer prior art knowledge and skills to computer graphics image-making. Such studies provide important information on developmental growth and children's art-making strategies, as these relate to computer use. Action research studies like those by Chia and Duthie (1992, 1993) and Freedman (1989) have shown that children rely heavily on social interaction when using a computer to make art. These studies demonstrate how art teachers can use the art-making capacities of computers to address children's fears, uncertainties and difficulties with hands-on participation. Studies like those by Kough (1984) and White (1985) have examined specific software or hardware within an educational context, and have demonstrated the potential of computer tools in creative situations. Such studies do provide much-needed information about the technologies themselves.  6  But unlike the bodies of literature that have been created around such parent disciplines as "painting" or "drawing", there is little common knowledge about what is meant by--or involved in--"computer art" as a field. For example, there have been no field-based studies of the diverse content and conventions of computer artworks in general. Nor does the literature of art education appear to recognize that all forms of computer art and graphics involve entire art worlds of adult artists and support personnel, without whom there would be no concepts of computer art. Additionally, with the exception of two case studies (Demaria, 1991 and Morbey, 1992), there has been no basic research on individual computer artists, and what characterizes their practices, their theories or their history.  Purposes of the study The first purpose of this study was to analyze the artistic practices of digital computer artists, so that I might describe them to interested educators. Knowledge of conventions in digital art can enable teachers and their students to deal more meaningfully with both digital art and more sophisticated forms of electronic art. For example, they may expand their understandings of the conventions of Draw, Paint, image-processing and 3D computer art; be more able to "read" or analyze many of the visible features of other forms of electronic art; or be more prepared to take 7  part in the development of new aesthetics for digital art (Brown, 1989; Ettinger, 1988; Manovitch, 1994). The second purpose of the study was to examine the conventions of digital art in terms of a "discipline" or "field"-centered approach to curriculum development in computer art. Discipline-based approaches have been vigorously promoted in some quarters of art education over the past 15 years. Discipline-Based Art Education (DBAE), a curriculum philosophy supported and promoted by the J. Paul Getty Center for Education in the Arts, "seeks an evolution from a naive (untutored) to a sophisticated  (knowledgeable)  understanding of the subject of art" (Greer, 1984, p. 215)-. At the core of this knowledge are the practices of artists, art critics, aestheticians and art historians. Their roles are "established as models for outcomes of a visual arts education program" (Clark & Zimmerman, 1986, p. 34). These need not be conceived of as separate domains of knowledge, but can be integrated in the practice and understanding of art as a form of cultural expression (Hamblen 1986; Chalmers, 1987). This study sought to frame its findings and recommendations in terms of these four perspectives. For example, Clark and Zimmerman maintain that the professionalism of the adult artist manifests itself in self-generated works that demonstrate knowledge of art concepts and experience with materials (pp. 40-41). However, the tools and techniques of digital artists are largely 8  unknown. While the art historian ideally can identify works of art and stylistically categorize them (pp. 43-44), the features of computer artworks, the styles used by computer artists, and the technologies that have prompted their development have not been examined for the purposes of art education. Clark and Zimmerman define an aesthetician as being able to apply aesthetic criteria and terminology to works of art (pp. 44-45). As yet, however, neither the working language of digital artists as revealed in their conventions has been analyzed, nor the terminology of computer art in general categorized. Similarly, Clark and Zimmerman propose that the art critic be able to demonstrate knowledge of objective criteria, judgments and interpretations about art (p. 42). Yet the critical standards of computer artists in general, as might be revealed in part by their use of conventions, have not been analyzed.  Research questions Applying a discipline-based framework to understanding digital art images, this study addressed the following questions. 1. What are the conventions of digital artists in terms of tools, techniques and content? 2. What do the conventions of digital artists, as evidenced in their work, reveal about technological developments in computer art? 9  3. What do conventions of digital artists, as evidenced in their work, reveal about their aesthetic notions of computer art? 4. What do conventions of digital artists, as evidenced in their work, reveal about their critical standards?  Theoretical perspective The inspiration for this study was supplied by Howard Becker's concept of "art worlds", which are identified in part by "conventions" shared by their "members" (1984). The concept of "membership" was originally developed in the fields of sociology and anthropology. American sociologists Erving Goffman (1959) and Harold Garfinkel (1967) contended that there are "taken-for-granted" assumptions by which we frame the world and conduct our interactions. Theorists who continue to work in traditions established by Goffman and Garfinkel believe that members of a culture are uniquely sensitized to tacit knowledges in that culture. Becker provided a unique account of how tacit knowledges in various art worlds emerge and become conventionalized. He proposed that conventions in art manifest themselves in such forms as styles, content, choice of media and tools, formats and unique schemata. He believed that particular knowledges and uses of such conventions define memberships in an art world (wherein membership is  10  the activity of interchanges and recognitions, rather than some form of formal subscription). Becker addressed both photographic and analog conventions in visual art. His theories are especially relevant to the study of computer art, since, as he notes, equipment in particular produces knowledge by embodying conventions (p. 57). Further, he believes that because art worlds "typically devote considerable attention to trying to decide what is and isn't art" (p. 36), we must observe how an art world makes those distinctions before we can understand much of what goes on in that world. His tenets appeared to have direct relevance to a discipline-based art education model: first, in the sense that DBAE is grounded in distinctions about roles primarily made by practitioners in the field or world of art itself, rather than by classroom practitioners; and second, in relation to his portrayal of the idea of conventions in art. However, it should be noted that this study does not seek to deconstruct all aspects of the computer art world in the complexity indicated by Becker's work, since such an undertaking would be beyond the scope of one study. Rather, specific parts of Becker's work, as noted in Chapter 2, have been used as an inspiration for examining persistent or dominating characteristics of digital art as it appeared on the World Wide Web in 1996.  11  Personal grounds The impetus for this study began with my Master's thesis (Johnson, 1993), when I analyzed 900 terms commonly used to describe the visual structures of traditional and digital art images. For my comprehensive examinations, I focused on literature previously written on the topics of computer art and computer art education. For example, I reviewed the history of computer art and explored aesthetic debates about analog and digital media (for example, see Pope, 1988; Binkley, 1989; Nadin, 1989; or Hickman, 1991). I analyzed the kinds of methodologies that have been used by educators to study the use of computers in art classrooms (see Chapter 2 ) . While researching literature on disciplinebased curriculum in general, I became interested in the work of members of the computer art world themselves. Becker suggested that audiences "learn unfamiliar conventions by experiencing them, by interacting with the work and, frequently, with other people in relation to the work" (p. 64) . My preparation for this dissertation included four years of interaction with computer scientists and artists, as well as research and coursework in desktop, mainframe, and network applications for computer art and design. During the analysis of the images sampled for this study, I subsequently relied on the expertise of professional members of the computer art world for agreement about the meanings of conventions found in the artworks, and to substantiate the findings.  12  Definition of terms art world.  . a pattern of collective, cooperative activity among a group of people who recognize each other as members by their participation in conventions  computer art., art produced with electronic media, including digital, interactive, animated, and multimedia art conventions... accepted practices, attitudes or beliefs resulting from implicit consent among a group of people curriculum.... a course of study whose focus may be on the content, the participants, the process and/or on ultimate social aims digital art... computer art that was created using a Paint, Draw, 3D or image-processing program, or from original programming digitized art..analog artworks such as paintings or sculptures that have been reproduced in binary form, so that they may be viewed on a monitor screen discipline  . a field of knowledge  education  ..the formal process by which learning takes place  members. ...  ..cultural participants with specialized, tacit knowledge of its conventions 13  traditional art..analog or material art forms, such as painting, sculpture, printmaking or textiles  Delimitations of the study 1. The sample of artworks was delimited to computer art as it was found on the World Wide Web. The Web appeared to be the most time-efficient and economically-advantageous venue for a sufficient sample of artwork by a. target population that is numerically large and geographically wide-spread.  2. The sample was further delimited to "digital" artworks. These appeared to be the most common type of computer artworks on the World Wide Web during 1996-- possibly by default, since Web scripts for other forms of electronic art had not yet been developed. However, the kinds of software and hardware used to produce the analyzed images appeared to have direct local relevance to classroom practitioners, who used similar software and hardware in the school system at the time of this writing. This match appeared advantageous for a delimited study.  3. The methodology of this study was delimited to content analysis of the art works themselves. Content analysis is a quantitative methodology which emphasizes the frequencies of observed phenomena, either in isolation or in relationship  14  to each other. This delimitation addressed one of a number of emphases in art education at this time: to educate students about the visible phenomena or characteristics of artworks.  4. Primarily, this study took a formal and quantitative approach to analyzing and discussing visual conventions. This delimitation was based on the belief that successful integration of knowledge about a field depends first on an appreciation of familiar components. Visual conventions have been a major focus of North American art education since the early 20th Century, when Arthur Wesley Dow first delineated methodologies for teaching what he termed "the elements and principles of design" (Green, 1990) . A more recent study (Johnson, 1995) found that familiarity with the elements and principles of design continued to be the first or second listed objective in over 70 North American curriculum guides for art education. However, despite the persistence of such formalist orientations in curriculum guides, the field of art education now appears to be in a transitional phase (Moore, 1991; Pearse, 1992; Wolcott, 1996) wherein other perspectives are emphasized. This study has therefore endeavored to place the findings in the broader context of a field in transition.  15  5. This study is delimited to visual rather than appreciative aspects of digital art, on the grounds that we first engage with artwork through perceptual qualities. As claimed by Czikszentmihalyi and Robertson (1990) and Haanstra (1994), an appreciation of visual qualities are, a priori,  indispensable for aesthetic experience; certainly an  erroneous understanding of visual features can lead to misinterpretations of cultural or social aspects (Johnson, 1996). Since no previous work has delineated visual aspects of computer art, this study has given these aspects precedence.  6. This study is delimited to the analysis of digital art images at face value, with no dialogue with the artists responsible for their creation. As substantiated by the independent raters, this study found that it was possible to readily and clearly identify digital conventions by appearance alone.  Limitations of the study 1. This study acknowledges certain problems in the delimitation to computer art that was intended to be output in two-dimensional form (electronically or concretely). The findings do not represent endeavors in the field of computer art as a whole. Many electronic art forms are not intended for transmission on the Internet, and transmission scripts had not yet been developed for others. This study thus 16  represents the status quo of conventionalized digital art, rather than indicates features of what might be considered to be "cutting-edge" work in the field of computer art.  2. The findings of this study have limited relevance to the question of what computer art should be or what it might become. Since computer art is created with technologies that ceaselessly afford new opportunities in terms of the development of tools and interfaces, the effects of history on digital art on the Web--even within the year since the data was collected--have been dramatic. (See Epilogue.) Similarly, classroom teachers have continued to upgrade their hardware and software. Thus this study is, by default, limited in its contemporary relevance.  3. The methodology of content analysis, as used in this study, limits the kinds of information that can be derived about digital art. Ideally, subsequent research would correlate the findings with additional data, using both quantitative and qualitative methodologies, as outlined in the recommendations for future research.  4. The findings of this study are limited in terms of the development of a complex appreciation of aesthetic elements of digital art. The recognition and comprehension of visual elements are only one aspect of the art experience. As noted in recommendations for both curriculum development and 17  future research, an appreciation of visual elements and the procedures used to generate them should subsequently be integrated with such other appreciative aspects as: an understanding of the theories that inform computer art; the effects of history regarding tools and techniques; the context of the Web; critical standards; and social issues such as gender and access (see Chapter 5 ) .  5. Because this study was delimited to an appreciation of visual aspects in completed art works, the intentions of the artists are presumed. This study is limited in terms of understanding the process of creativity, such as how, when and why the images might have been seriated; how "mistakes" might have been incorporated; or the meaning of individual images in the context of bodies of work that may have been generated by artists. This study recognizes the need for interchange and dialogue concerning creative aspects, as suggested in the recommendations for curriculum development.  Significance of the study The findings in this study are intended to be of use to art educators in general. The recommendations are specifically addressed to educators who are interested in the development of discipline-based digital art education, or a curriculum which would integrate digital art production, aesthetics, criticism, and digital art history.  18  This study is significant in that it looks beyond current classroom practices and the field of computer art itself, to a synthesis of digital and discipline-based art education concepts. The study also provides a valuable collection of basic research material, since it is the first research study undertaken of any computer art conventions. Therefore the findings may be of interest to sociologists, art historians and art critics as well as educators.  19  Chapter 2 Review of the literature  Introduction This chapter is designed to review literature on aspects of Discipline-Based Art Education (DBAE) that are relevant to possible adoption of digital art as an area of disciplined classroom study. Selected aspects of research in computer art education are then examined. Thereafter, Becker's concepts of "art worlds" and "conventions" are described, and linked to a discipline-based approach to implementing digital art programs in the classroom. The review of discipline-based art education literature is limited to writing between 1984-1993 and provides an outline of its basic objectives. The review of computer art education literature describes studies done between 1987 and 1994.  Discipline-Based Art Education During the past 2 0 years, the field of art education has moved from an emphasis on expression in studio practice to a conceptual approach based on appreciative aspects of art (Greer, 1993) . Studio practice has become more balanced with the domains of art criticism, art history and aesthetics. These fields of study, or parent disciplines, are marked by "recognized communities of scholars or practitioners, established conceptual structures, and accepted methods of 20  inquiry" (Clark, Day & Greer, 1987, p. 130). The implementation of Discipline-Based Art Education (DBAE), with its emphasis on the integration of the four disciplines, has been supported since 1984 by the Getty Center for Education in the Arts and given a forum for discussion by the National Art Education Association. First proposed as a philosophical rationale (Greer, 1984), this model takes a humanistic approach to the arts (Levi & Smith, 1991). That is, it addresses the needs to communicate and share experiences with others through art making,- to understand the social and historical context of art and art making; and to respond to art with informed engagement. DBAE is inclusive to societal models, self-actualization, and "subject-centered" orientations (Clark, 1991; Dobbs, 1992; and Greer, 1984, 1993). DBAE was initially criticized for such aspects as its non-democratic character in a pluralistic society (see, for example, Blandy & Congdon, 1987); its Eurocentric perspective (see Hamblen, 1987); and its ambiguity regarding emphasis on process or product (see Efland, 1990). However, the ultimate purpose of DBAE has been to help students acquire "lenses" or conceptual structures based on the methods and content of the four discipline areas. The barrage of criticism has therefore been important in instigating numerous theoretical refinements and a more balanced presentation of artworks. For example, DBAE now considers the examination and presentation of artwork from  21  diverse cultures; supports feminist scholarship; and has undergone some curriculum revision to emphasize "accessibility to all parts of the world and knowledge about the art of all times and places" (Moore, 1991, p.39). My intent is to contribute to this movement by providing a foundation for curriculum developers in digital art production, history, aesthetics and criticism. The following descriptions of artists' roles in these areas are adapted from Clark and Zimmerman's original model for visual arts education (1986, pp. 40-45), which drew upon a discipline-based perspective.  Art production Discipline-based art production emphasizes professional practices in the field of art, rather than school practices. The artist is seen as an informed participant in creative and material processes, who uses and applies art concepts and knowledges, and who can, at the same time, express personal style and technique. By independently generating works of art, the artist masters skills and media and solves technical and aesthetic problems. As the computer artist would learn the technical aspects of software programs, for example, the artist would identify the aesthetic features and possibilities of the medium. The computer artist would thus be seen as what Becker calls a "member" of the computer art world, who contributes to the development of the field through reflective involvement (p. 36). 22  Art history DBAE approaches to art history emphasize the role of the art historian in understanding works historically and culturally. The historian must be aware of the historical conventions of the genre and be able to identify stylistic categorizations. An appreciation of socio-cultural interpretations and of theories related to cultural history contributes to an understanding of the "authenticity", or cultural significance of the generated artworks. Referring to computer art in particular, Nadin pointed out that "art is far more than the mere physical presence of an artifact; and this is why the digital approach to art must consider the human being, society and its evolution under new circumstances of life and work" (1989, p. 46). The computer art historian must ask such questions as "What is culturally unique and significant about computer art?" and "How has it come to be so?" Aesthetics in art DBAE approaches can encourage the generation of aesthetic criteria for the analysis of artworks. These may entail written or verbal interpretations of the content of artworks that describe their thematic, symbolic and expressive properties. The aesthetician is able to situate and define artworks according to diverse and congruent theories about art, since there is a plurality of contexts which frame answers to such questions as "What is the nature of art?" or "What does art consist of?" In both aesthetics  23  and art criticism, DBAE particularly emphasizes the ability to understand and apply specialized art terminology. Computer art has evolved during a challenging period in the field of aesthetics, as theorists grapple with transitions from a modernist to a post-modernist paradigm. Hicks has proposed that aesthetics be redefined within the context of an information age to include "an extended vocabulary, a greater focus on process, openness to collaboration and cooperation, a revised definition of originality, greater focus on spontaneity and the unexpected, an expectation for greater diversity of imagery, greater focus on cross-cultural and cross-technical connections, and a look for closer relationships between art and science, mathematics, engineering, and even manufacturing" (Hicks, 1993, p. 44). The aesthetician must currently address both modernist and postmodernist contexts. Art criticism DBAE approaches to art criticism stress understanding of external conditions beyond the work of art. These involve categories of description and analysis used by art critics to interpret and judge artworks. The art critic can appreciate differences between personal interpretation and shared criteria. The purpose of theoretical inquiry, as Dunnahoo put it, is to enable artists "to move beyond natural expressions of preference towards more reflective thinking" (1993, p. 57). The computer art critic could entertain such larger 24  questions as, "What are the theories of computer art?" and "What is computer art for?"  Research in computer art education DBAE offers a useful model for evaluating the potential contributions of the field of computer art, as characterized by the activities of computer art specialists. It is a field characterized by rapid historical developments (Csuri, 1979). It is alive with aesthetic issues, from aspects of "cultural colonization" manifested in recognizable algorithms in widely-used software (Jones, 1989) to issues around appropriation, copyright and access by gender. Studio practice has involved both a mastery of tools and art conventions unique to computers, and stylistic developments. Theoretical issues have included positioning various forms of electronic art in fine art contexts, and developing criteria for these expanded repertoires of expression. Previous research in computer art education Prior to 1988, numerous dissertations and studies proposed strategies for computer-assisted instruction in art education. CAI was generally used to teach art history and design basics. Today, CAI studies for art education are the exception in both dissertations and academic journals. Philosophical debates about the "place" of computers in art education were also more common during the 1980s when, typically, a kind of opinion paper identified "experts" in  25  computer art education and solicited information from them on what a computer art curriculum should entail. Since 1989, only three such studies have been undertaken (Kleim, 1994; Wang, 1992; Weaver, 1989). Quasi-experimental research on the ability of computer tools to simulate traditional art tools has also become less common, with the exception of dissertations by Chumley (1987), McAllister (1990) and Passmore (1991). These studies compared differences in students' use of traditional drawing media and computer drawing tools. However, they did not study phenomena or values that might lead to the development of curriculum. Three qualitative kinds of methodologies have become more prevalent in art education research over the past ten years, each of which has advantages and disadvantages in determining curriculum goals and values. The first, a kind of consumer review, emphasizes the constraints and possibilities of particular hardware and software in educational settings. The second, drawing on ethnographic research methods, emphasizes the interactions of students with the technologies and with each other. The third, action research, emphasizes the learning process itself. Reviews of hardware and software Researchers have examined specific software or hardware within an educational context, usually a computer lab or art classroom in a public school (for examples, see Pogue, 1992, Nicholson, 1993, Dawson (1994). Their reports begin with 26  technical and functional descriptions of electronic components, details of installation, and descriptions of program tutorials and interfaces. These are followed by descriptive studies as students are introduced to the tools. The rationale for using this methodology is to examine the potential of the tools themselves in creative situations. (See, for example, White's 1985 review of the KoalaPad, a pressure-sensitive drawing tablet.) These kinds of studies provide much needed information about the technologies themselves. Unlike such areas as painting or drawing in art education curriculum, there is relatively little common knowledge about what is meant by--or involved in--teaching computer art. One disadvantage of this research methodology, however, is that it is usually very particularized. Researchers seek to identify unique characteristics of specific hardware and software products, rather than describe generic properties or characteristics of that type of software or hardware in general; that is, the common features of Paint or Draw programs, or the common features of digitizing tablets. In addition, the evolution of technologies quickly makes such particularized studies obsolete. Ethnographic studies and observations In ethnographic studies, researchers often propose to examine specific moments in what MacGregor and Hawke called "the ongoing flow of classroom events" (1982, p. 38) . Concentrating on individual behaviors and concerned with 27  understanding from the participant's perspective, researchers direct their observations to how the subjects participate in semi-closed systems. Ethnographic computer art studies have included those by Bhargava, (1991); Chia and Duthie, (1993); Freedman, (1989); McDevitt-Stredney, (1993) ,- Pike, (1988) ; Rhee and Bhavnagri, (1991) ; Stokrocki, (1986) ; Weishampel, (1989) . Ethnography is therefore a methodology flexible enough to accommodate shifting foci between what MacGregor and Hawke call "macrolevels" (where the focus is on structure) and "microlevels" (where the focus is on incidents). For example, in their study of the social interactions and learning strategies of primary children, Chia and Duthie (1993) included both generalized observations of the group as they shared computers, and the verbatim conversations and idiosyncratic behaviors of individual children. Likewise, Freedman (1989) moves between generalized observations of the gender attitudes and collaborative strategies of 40 elementary students in three art classrooms, and their specific utterances and behaviors. Ethnographers do not attempt to demonstrate the certainty of any generalization, but try to "maximize their confidence" (Hammersley, 1992, p. 181). They may do this by triangulating their data, which involves coordinating perspectives obtained from different strategies in what MacGregor and Hawke have called a kind of "navigational fix" (p. 39). For example, in their case study of university 28  students learning to use Paint software, Freedman and Relan (1992) used six kinds of documentation procedures to facilitate triangulation: structured interviews, journals/sketchbooks, survey/questionnaires, unstructured observations, and both photographic and taped records. Researchers may also maximize their confidence by studying subjects over time. In order to analyze the composition, initiation, and form of student interactions, Rhee and Bhavnagri (1991) videotaped 18 preschoolers while they,drew with a computer for eight 45-minute sessions. Pike (1988)' observed students' explorations of the potential of graphics tools, and their attempts to achieve technical dominance with them, over a period of one month. These kinds of studies are useful in that they can give educators insight on those technologies that are most complex and challenging to students. Ethnography is also well-suited for classroom studies where generalization is problematic and replication is not critical. But there remain significant gaps in the literature. In particular, relationships between students and computer art, either perceived as a fine art discipline or as a commercial field, are still largely unexplored. Action research studies Action research involves methods used to directly work with and implement the goals of the participants. Two kinds of research situations have been used to implement and study  29  computer technology for art and design education. The first involves implementing programs with students in preschools, the public education system, or higher education. (For example, see studies by Chia & Duthie, 1992; Edwards, 1985,Greh, 1986; Kough, 1984; O'Connell, 1985.) In these situations, the researcher directly teaches students and observes their progress. The second kind involves teacher training, or how the researcher teaches teachers how to teach computer graphics to other students. (See studies by Bridwell & McCoy, 1991; Campbell, 1991; Chia & Duthie, 1992; Freedman & Relan, 1992; Hubbard, 1985; and Kleim, 1994). Rationales given for action research studies in computer art education have frequently been based on what Bridwell and McCoy called "keeping up with and adjusting to the rapidly changing world of computers" (1991, p. 54). As several of these researchers point out, "The journey to this new world can be fraught with uncertainties and fears" (Campbell, 1991, p. 114). Chia and Duthie (1992), in their action research study of art specialist students working with primary children, described the "new and unfamiliar" as being "received with discomfort" (p. 209). They sought to encourage the children in their case studies by providing them with prescanned files of the children's own pen drawings which they could then elaborate, erase or copy and paste. Greh (1986) noted that fear of the computer may result from its "manipulative, dehumanizing aspects" (p. 5 ) , and provided an account of her search for more user-friendly  30  interfaces for secondary students. In a problem-solving computer art study with intermediate students, Edwards claimed that to many students, "the complex technical problems of computer graphics seem to put it beyond their reach" (p. 30). In essence, action researchers in computer art education have sought to address fears, uncertainties and difficulties with hands-on participation. One advantage of action research that -begins from such a foundation is that these studies can readily chronicle the "enthusiasm" and "success" that become transcendent features of the program. However, research studies proceeding from fear and discomfort have sought to change behavioral responses without setting them in the context of specific curricular goals, or even with reference to experiments in the development of new curricula. As Chia and Duthie found, "There are crucial differences in the process of production as well as in the visual qualities of computer graphics as compared with other, more traditional media" (1992, p. 222) . These "crucial differences" have not, however, been isolated or specifically described. Summary of research problems Several difficulties have confounded the state of research for computer art education. Many studies in computer art education are centered around questions of how children interact with a new art medium in already-familiar art education contexts. The predominant focus is thus on how students accept or transfer previous art experience to a 31  computerized environment, rather than on exploring the technologies in a foundational sense. The conventions of computer art, as Becker would say, embody certain understandings of the medium--not only its classroom potential or disadvantages. The majority of researchers also follow a technological imperative, believing a priori that it is good and necessary for the educational system to perpetuate technological progress, while bypassing critical evaluation of the system itself. Most important to the current study, researchers have been studying themselves. That is, they have been operating within the closed system of the public school, rather than looking to discipline-based links between the fields of computer art and art education, from which classroom practices are derived. Hobbled by a limited vision of what they may be, strategies for deriving curriculum from classroom studies have been neither effective nor productive. To implement a discipline-based curriculum, researchers first need to examine hardware and software as it is developed and applied in industry, science, and the computer art world. For example, although Campbell asked the extremely relevant questions, "What aspects of this technology can be successfully integrated into art education?" and "What do teachers need to know about this technology to teach it effectively?" (p. 115-116), he then 32  turned to art teachers with little or no prior experience of making art on a computer and asked them to determine the requisite knowledge. Studies of computer artists and their art world are essential. Through critical examination of their work, artists can provide, or help generate, models of theoretical and historical development in computer art. A disciplinebased approach, in which professionals serve as exemplars, has become more common in other areas of art education over the past fifteen years. Unfortunately, only two art education studies have been done of professional adult exemplars in the field of computer art: a dissertation on the work of pioneer computer artist Charles Csuri (DeMaria, 1991) , and a dissertation on Harold Cohen, who applied an artifical intelligence system to art production (Morbey, 1992). From studies such as these, we can learn theories and practices of computer artists that may not otherwise be revealed.  Art worlds Definition of art worlds Howard Becker described art worlds as "all the people whose activities are necessary to the production of the characteristic works which that world would define as art" (p. 34). He used the term "world" to mean the cooperative activities of the participants, rather than structures or organizations. Some networks are large, complicated and 33  necessitate numerous kinds of collaboration, while others are small with few specialized personnel. Function of art worlds Works of art, from a sociological viewpoint, are not only the products of individual artists but joint products of all who cooperate to make each genre possible. Extending Becker's model to the computer art world, the participants (or members) would include hardware, software and interface designers; creators and users of educational resources, such as instructors, consultants, authors of manuals and texts as well as their publishers, user group participants, vendors and corporate sponsors; technical assistants for specialized printing and lab processing; the personnel of concrete galleries, trade fairs and conventions; and Web programmers and designers--as well as artists, jurors, critics, curators and writers (see Johnson, 1995/96 for a more detailed analysis). Each of these areas of participation generates further realms of collaborative activity, and at the same time is co-dependent on all other areas. Like other art worlds, the computer art world functions interdependent^ through tacit agreements. Conventions used by art worlds All art worlds have standards that are recognized and standards that are emerging. Becker used the term "conventions" to refer to tacit knowledge: which is to say, knowledge that no longer requires explicit description and support, but instead is implicit to the group. Conventions 34  provide the basis on which art world participants produce work characteristic of that world. The combined activities of computer art world participants would perpetuate such conventions as tools and digital schemata; the degree to which the computer itself is visible or invisible in the artwork; the size, format and resolution of artworks; and the manner in which they come to be exhibited. While acknowledging the current need to identify all aspects -of the computer art world, as they might be characterized by Becker, the primary purpose of this study is the identification and description of one area: the visual and procedural conventions of two-dimensional output or digital art. Thus, Becker's work on "conventions", specifically, is used as an inspiration for this study.  Implementing a discipline-based approach to computer art education Art educators develop discipline-based curriculum for other art media by referring to resource material in the field of art. This may include publications on the history of art or studio techniques, artists' diaries and letters, or theoretical works in aesthetics and culture. Artworks are studied in their original form or through reproductions. Studio production is often learned by witnessing the techniques of professional artists. To learn more about computer art, researchers and educators must likewise turn to resources in the computer 35  art world. Numerous publications are available in the field, from historical surveys such as those by Friedhoff (1989) and Kerlow and Rosebush (1986); to trade publications that center on developments in art technologies, such as Computer Graphics World or Computer Artist; or to journals that publish articles on aesthetics and criticism, such as Wired or Leonardo. More recently, art educators have been turning to the World Wide Web as yet another resource (for example, see Wongse-Sanit, 1997). The World Wide Web refers to the Internet technologies that make possible the display and transmission of graphics, sound and motion from one computer to another via an analog-digital relay, using already existing telephone cables connected to modems on the computers or using new Internet wiring. The Web has made it possible to see increasing numbers of artworks on-line, even since this study began. Although the Internet as a transmission vehicle for text has been in development since the early 1980's, the Web is a relatively recent development. Only three years ago, the National Center for Supercomputer Applications in the United States released Mosaic as a shareware, or free program. It was the first software to enable graphics, photographs, video and sound to be transmitted over the Internet in addition to text. Rheingold noted that within six months of its release, more than a million copies of Mosaic were downloaded from the Net, and thousands of new 36  sites combining text and images went on-line within months (1995, p. 144). These sites, which make up that part of the Internet known as the World Wide Web, are dedicated to the display of all forms of data that are not text-only. Among them are the thousands of "virtual" galleries that showcase art. Mosaic has since been eclipsed by the more sophisticated Netscape program. Both Mosaic and Netscape not only facilitate the development of Internet sites containing text, sound and graphics, but act as "browsers" or retrieval systems for information at these sites. The speed at which this area is growing can also be seen in the fact that four later and more powerful versions of the original Netscape program were developed during 1996 and 1997, as well as a new browser produced by Microsoft. In response, the number of sites and links has likewise grown. (See Epilogue for a comparison of site generation during this time.) Information can be retrieved from the Web by using increasingly sophisticated directories of site addresses for thousands of topics. (As noted in the Epilogue, not only has the number of directories radically increased during 19961997, but "robots" or "spiders" have been designed to infiltrate the entire Web structure and retrieve all data on a subject area from numerous directories, at once.) As I discovered in the course of this study, digital art sites are particularly numerous. The extensive community of computer artists and art writers not only has proved to 37  be a multi-faceted art world in Becker's terms, but a community as active as that from which DBAE prototypes were derived in other areas of art.  Summary This chapter reviewed the basic objectives of Discipline-based Art Education as it emphasizes the practices of artists, art critics, aestheticians and art historians. Strategies for deriving discipline-based computer art curriculum from classroom studies have not been effective in determining conventions in the use of tools and techniques by computer artists, or what these conventions might reveal about technological developments in computer art production, aesthetic notions, or critical standards. Researchers and educators will need to investigate resources in the field of computer art itself, and examine the conventions of the "members" of the computer "art world", as defined by Becker.  38  Chapter 3 Design of the study and methods of data collection  Introduction In this chapter, the research design and methodology are described. The research questions were: 1) What are the conventions of digital artists in terms of tools, techniques and conventions? 2) What do the conventions of digital artists, as evidenced in their work, reveal about technological developments? 3) What do conventions of digital artists, as evidenced in their work, reveal about their aesthetic notions of computer art? 4) What do conventions of digital artists, as revealed in their work, reveal about their critical standards? Using a disciplinebased approach of studying adult exemplars, I examined the visual aspects of digital art. Since this is a ground-breaking study of an emerging field, I originally had planned to apply ethnographic methods to the computer art world, and, in this way, achieve a description of that "world" and its "members." I also planned to analyze digital art in hard copy form at exhibition sites. However, I soon found that opportunities to analyze hard copies were limited to exhibits that had previously been curated, such as SIGGRAPH '96 in New Orleans. The exhibitions were geographically far apart, and selections 39  tended to be based on the preferences or criteria of curators and jurors, rather to reflect applications and conventions in the field of computer art at large. Given these conditions, it was difficult to design an ethnographic study of common practices. At the same time, as I searched for exhibitions and user groups on the Internet during the winter of 1995-1996, I became aware of the larger sampling possibilities of computer art exhibited on the World Wide Web. At that time there appeared to be more than 1000 sites showcasing the work of up to 50 "digital", "electronic" or "interactive" artists at each site. Thus, although the Web was in its infancy, I found there already was a wealth of imagery on sites created throughout the electronic "world" that could be accessed with relative ease through the Internet. The majority of images also appeared to be posted without curation or other intermediary criteria, thus reflecting the common practices of computer artists who had access to the Web. The framework of the study was finalized at this point to focus on a content analysis of 200 digital artworks on the World Wide Web. The sampling and analysis was conducted over a six-month period in 1996, and included procedures to have independent raters test the findings. The methodology entailed multiple stages of electronic data collection. Categories of information were refined as the study proceeded (Strauss & Corbin, 1990). Two primary 40  characteristics of data collection were the constant comparison of collected data with emerging categories, and theoretical sampling of different groups to maximize the similarities and differences of information (Greene & Caracelli, 1989;  Miles & Huberman, 1984). Following is a  description of the research design, population, sampling procedures, instrumentation, variables, and findings.  Content analysis In the first stage of the study, content analysis was used to identify the visual properties of digital imagery. Content analysis is a methodology used to describe and categorize what is commonly known as "material culture" (Bolin, 1995), or what Gall, Borg and Gall referred to as "communication media (1996, p. 357). "Material culture" is a label given to include "all artifacts, past and present, that have been created or modified by humans; therefore, material culture encompasses all non-natural objects" (Bolin, p. 1 ) . "Communication media" refers "both to documents and to materials whose messages are primarily visual or auditory" (Gall, Borg & Gall, p. 357). The various kinds of artifacts may include archeological findings; the contents of newspapers, maps or diaries; or media products. As Feldman put it, "Description is a data-gathering process, a listing of facts. Descriptions are answers to the questions, What is here? What am I looking at?  What do I  know with certainty about this image?" (1982, page 12). He 41  goes on to note that critics derive much of their descriptive information from closely attending to what can be seen. Content analysis has been used as a methodology for more than 40 years, since Berelson defined it as "a research technique for the objective, systemmatic [sic], and quantitative description of the manifest content of communication" (1952, p. 18). Content analysis was used for three reasons. First, little has been written about the nature or appearance of computer art works, or about the terminology used to describe the visual structure of computer images. Content analysis provided a means to expand that knowledge base. Second, the development of the World Wide Web over the past three years made computer artwork more visible and easily accessible. A sufficient amount of material had accumulated, to the point where the timing was opportune for a content analysis. The economy of the method for a large-scale study, as well as the fact that the material for data collection was not geographically constrained, appeared advantageous. Third, the Web enabled me to access and unobtrusively determine actual practices. This study acknowledges, however, that the views and theories of intervening mediators, such as curators or jurors, are equally important and could be the subject of a different study.  42  Research design Preliminary studies Content analysis involves counting the frequency with which various values of a variable occur (Crowl, 1993, p. 127). Determining variables in the content of computer art required three stages of preliminary research. First, I obtained access to the Web and used a multistage procedure for systematically sampling digital art sites for examples of images. Second, I developed categories for analysis and created taxonomies to record the characteristics. Third, I located independent experts in the area of computer art to verify the definitions of the characteristics and to analyze samples. During this research, I connected to the five or six directories that were then available through Netscape (such as "Yahoo" or "Excite"), and searched their lists of subject areas in order to find URLs (Universal Resource Locators or Web site addresses) for "computer art", "electronic art", and "digital art". Graphical user interfaces allowed me to proceed in various ways within the directory, such as to go back to a previous selection; to record sites already "visited"; to file URLs for future reference in personal storage areas known as "bookmarks"; and to save or print information directly. Because my hours of access and printing allowance were limited at the university, I obtained a private Internet account with a local commercial provider (Mindlink) in  43  British Columbia, Canada. The commercial account included a copy of Netscape 1.0 and allowed me to access World Wide Web directories and URLs on an hourly basis at any time of the day for a nominal charge. For the greatest amount of flexibility in the duration and hours of my working time, I collected data using my home computer. In order to capture and print images, I upgraded to Netscape 2.0 and obtained a Canon bubblejet color printer. I spent approximately 120 hours browsing computer art sites and saving images to disk in order to print them out. Although it was possible to print most images directly from Netscape 2.0 connections, the process took three or four times as long. The sample of 2 00 images was printed in hard copy for several reasons: to facilitate the process of analysis without returning to sites to compare and contrast characteristics; to provide pictorial definitions for terms (pp. 50-72); to illustrate points made in the text; to facilitate independent rater checks; and as a general record for validation. They also provide a unique record of the state of digital art imagery in 1996, since the majority of the sites and the artworks themselves no longer exist on the Internet (see Epilogue). Due to difficulties contacting the artists of images used for pictorial definitions (see problems of attrition in Chapter 5), these images are credited by site only (Appendix A ) .  44  Population Target population In preliminary searches of art sites on the World Wide Web, I identified three different types of on-line art that were all commonly called "computer art" in 1996. The first and most common type consisted of reproductions of traditional art work (such as paintings by Van Gogh or sculpture by Rodin) that had been digitized and stored in public or private museum URLs. These digitized artworks (as opposed to digital artworks) made up the bulk of art online. For example, in 1996, ARTNEWS reported that even a relatively small site like Art Resources in Seattle had 4,300 gallery and artist listings, while a larger site like ArtNet had images of over two million artworks (Franke, 1996, p. 98). A second type of art on the Web, and the least common at the time, was interactive. These conceptual art projects used the Web as a transmission medium by providing art works for other artists to alter, or by providing procedures called "meta-tools" for other artists to carry out productions. The third type of art was called "digital art" or "cyberart" (less reverently, it was found to be referred to in artists' statements as "smart art" or "artificial art"). I selected only the art described as "digital art" for this study because it was produced with software similar to that most commonly used in the public schools (i.e. Paint, Draw,  45  image-processing and 3D programs. See Chapter 4 for definitions), and therefore was expected to have the most local relevance to studio practice. Since it was created using mainly electronic media, it offered possibilities for establishing an aesthetic with characteristics similar to, yet unlike, those of paintings or sculptures. Unlike digitized art, it represented endeavors to work with a medium that is in constant transition. Digital art has an interesting art history and continues to be surrounded by critical issues and theory. As exemplars for curriculum development in computer art, digital artworks were more numerous than were examples of interactive art, during the period of this study. Animations and other dynamic forms of electronic art are also digital in nature. Examples were not included in the sample, however, since "scripting" or programming languages used to exhibit and view these forms of art on the Web were in an early stage of development at the time. Their exclusion from this study was a necessary result. Identification of digital art Like "painting", "digital art" refers to both process and product. It is created from binary code on a computer in one of two ways. An existing image may be scanned into the computer, downloaded from another computer, imported from a CD-ROM, or retrieved from a laser disk, digital film or video and further manipulated with software tools (for the  46  sake of brevity, I refer to all types of "found" art input in this study as "scanned images" or "image-maps"). Digital art also may be created as an original file in several ways. Using a Paint program, the user simply moves a mouse or uses a special pen and digitizing tablet to make concurrent alterations in the content of the image buffer. These alterations appear as marks within a format in a window on the monitor screen. The marks can simulate such traditional art tools as a crayon, pen or paintbrush. Using a Draw program or a 3D graphics program, the user clicks points on the screen for the computer to connect as lines (called "vectors"). Ready-made patterns, text fonts, shading, colors or lighting can also be assigned to the image by the program. Most of the current software programs for art and design (like Fractal Painter 4.0) combine several or all of these capabilities.  Sample Procedure Six search directories on Netscape 2.0 were used to find sites for "digital art". These were Excite, Yahoo, Magellan, Infoseek, Lycos and Alta Vista (this study acknowledges that Alta Vista is technically a search engine rather than a directory). Each directory subdivided digital art sites differently. Yahoo subdivided digital art into names of individual artists, group or gallery listings, and 3D artists. Infoseek had separate listings for 47  "computational art" (such as fractal and other mathematical art), 2D art and 3D art. Magellan produced different lists for the search terms "digital art" and "electronic art", but many of the sites were the same. The results of directory searches also varied in number. Magellan produced 2 03 sites from the "ranked and reviewed database" for "digital art", and 485 matches for "electronic art". Yahoo produced 177 matches for "digital art". Alta Vista matched both search terms, "digital" and "art" for a total of 800,000 "hits", but only the first ten proved useful. Towards the end of the sampling, as site listings began to recur more frequently, the URL  www.msstate.edu/F ineart_Online/art exhibitions/digital.html  -  resources/  provided more than 600 listings of  almost every site I had seen listed on all six directories. By cross-checking the URLs on the first six directories, I determined that I had sampled more than a third of the digital art sites on-line. Selection of sample As my review of sites and artwork proceeded, I also began keeping a record of characteristics of the images. When I had listed approximately 60 characteristics, I delimited the sample to a total of 200 images. The number was originally chosen as a manageable number of images for analysis. As previously mentioned, it later proved to represent approximately 30% of the sites listed by  48  msstate.  The selection was truly random, since I had no idea what the images would look like until I connected to each one. Only site names and image titles, with some limited annotation, were listed. (See Appendix B for an example.) Although most directories "ranked" the databases on their list, they appeared to be rated and reviewed either for their relevancy to the search terms or by the number of "hits" or visitors to the site. The selected images were also geographically random. Although the majority were found to be produced by artists in the United States, directory lists did not categorize by geography. Images were found to be produced by artists in Canada, Japan, England, France, Norway and Australia. I began sampling by intending to take 4 0 images from each of five directories: Excite, Yahoo, Magellan, Infoseek and Lycos. I divided the number of listings in each directory by 40 and selected the resulting number of images from that directory. For example, from Magellan's list of 203 sites, I randomly selected every fifth site. Some sites, however, included more than 50 artists. I therefore selected one-fifth of the listings at each site with the intention of conducting a second sampling to reduce the number, although I was concerned about two possible results: one, that I would ultimately have more than 200 samples, and two, that a random sampling of artwork in a large site would produce similar kinds of imagery.  49  However, neither turned out to be the case. I found that by the time I had selected images from Excite and Yahoo, I had to take increasingly smaller samplings from Magellan, Infoseek and Lycos due to the number of repeated listings. The kinds of imagery also appeared to be quite disparate. The main commonalties were gender and type of site (such as commercial or fine art, schools or independent artists). As these patterns grew stronger, I began recording the gender and type of site for subsequent analysis. I identified gender by information appearing on the home pages for each site, such as names, artists' photographs, biographies, curatorial statements, e-mail information, or reviewers' summaries or references. The types of site were similarly identified (see Chapter 4 for further details). I kept notes throughout the process for reference during the writing stages and for use as those anecdotal examples that appear in Chapters 4 and 5. I recorded the URL and file name for each sampled image, and retained both hard copies and computer files of all images. Variables in the selection Numerous unexpected variables affected the selection of images. The majority were the result of electronic problems. For example, in a few cases, the host for a selected site could not be contacted. After attempting to connect at two different times, I selected the next site on the directory list. I followed the same procedure if the site was deserted or "under construction". A small number of file  50  format problems also occurred while downloading imagery. In some cases, I simply needed to change file "tags" or formats; for example, from "jpe" to "jpg". Where problems persisted after verifying addresses and format tags, I again chose the next site on the list. Eight of the 200 images would not print from disk. I returned to the msstate  listing and selected an additional  eight images by taking images from sites I \had not • previously sampled.  Instrumentation Instrument design The instruments used to analyze characteristics found in the images were two self-designed taxonomies. The first taxonomy, which evolved as my own recording device, underwent dozens of revisions as the study progressed (see Appendix C for a sample of the final taxonomy). The second taxonomy was designed to avoid experimenter bias in the rater checks (see example, Appendix D ) . It was given to five raters, chosen for their experience with both 2D and 3D graphics (see Table 1 ) , as well as analog media. Revisions to the first taxonomy resulted from three procedures. Characteristics were added as the analysis proceeded and more were identified, while other characteristics were dropped due to low frequencies. Ultimately I analyzed 85 characteristics in the 200 images.  51  The raters analyzed the same 85 characteristics in 50 images. Selection of categories The characteristics were grouped into four major categories for three reasons: to organize the definitions and taxonomies, to expedite the process of analysis, and to aid in the development of implications for developing curriculum in computer art. These categories were 1) "formal properties", or the sizes, shapes and colors of images; 2) "content" or subject matter within images; 3) artistic "style"; and 4) "digital properties", or characteristics that are unique to art produced on a computer. The fourth category was later divided into "2D digital properties" and "3D digital properties" for a total of five categories. Following is a description of how items for each category were identified. Identification of category items Formal properties. Formal properties are design features which "glue" the separate parts together. The term "formal properties" is applied in this study to describe underlying structures of design or composition. In traditional art, formal properties are usually called "principles" of art, as opposed to "elements" or sensory properties. In this study, I have referred to elements as "digital" or "local" properties. Principles of art traditionally include such global properties as balance,  52  movement, repetition, emphasis, unity, variety, harmony, contrast and rhythm. I initially derived characteristics for "formal properties" and "digital properties" from the visual schemata I identified in previous research (Johnson, 1993). In that study, I found discrepancies in the literature of art education regarding definitions of terms for the elements and principles of art in traditional art; discrepancies in the application of many traditional art terms to computer artworks; differences in the meaning of terms in traditional art and computer graphics; and that there were no comparable meanings in computer graphics for the traditional art terms of balance, emphasis, unity, variety, harmony and rhythm (Johnson, p. 160). For example, in computer graphics, repetition was perceived as a local element involving the replication of motifs (p. 140) rather than a global property, and contrast was perceived in terms of "tone" (meaning "focus", for shape detection, p. 136) or local color (p. 135). The study revealed other terms that were more applicable to computer art than to traditional art forms. The terms for "formal properties" used in this study are defined on pages 45-50 and discussed in detail in Chapter 4. Content. In attempting to characterize the nature and appearance of digital art, I also examined typical subject matter. I began with a long list of all types of subject matter identified by myself and the independent raters in 53  the 200 images. Items that occurred in fewer than 2% of images were dropped from analysis. Categories of content and their definitions appear on pages 51-57. Style. The category of "artistic style" was defined by Feldman as "a resemblance among diverse images, created by a characteristic handling of content, presentation and schemata" (Feldman, 1982, p. 12). In consultation with the independent raters, I identified eight distinct styles of art in my sample: simulations, visualizations, illustrations, computations, image-processing, collage, drawing and painting. These categories are original to this study. I characterized the first three according to their slightly different portrayals of realistic imagery (p. 58), and the last three by their similarity to traditional art styles. I identified "collage" by the prevalence of composite images; "drawing" by the use of simple lines and hatching; and "painting" by simulations of brushwork. "Computations" were identified by overall recursive or iterated schemata such as fractal patterns. "Imageprocessing" was characterized by the predominance of a photographic image. Definitions and examples of each style appear on pages 63 and 64. Digital properties that were found to differentiate each style are discussed in detail in Chapter 4. Digital properties. In traditional art, the elements of art include such properties as line, shape, space, form, pattern, texture, color and tone. I had previously examined 54  comparable elements in digital art and found that their structure depended on one of three forms of graphics: 2D pixel-based systems such as Paint programs; 2D object-based systems such as Draw programs; or 3D graphics (Johnson, p. 51). These programs are explained in more detail in Chapter 4. As a result, each traditional term for elements of art was shown to be present as more than one term or to have more than one meaning in computer graphics. These terms were not only more numerous, but also more specific and measurable. An aesthetic framework appropriate to a formalist model--with its emphasis on qualitative and subjective properties rather than quantitative, mathematical logic--appeared to have little relevance for computer graphics. •I initially drew on the findings of the same study in categorizing all images as either 2D (produced as a result of 2D programs) or 3D (produced as a result of 3D programs). 2D imagery was identified by the flat, mosaic-like quality of raster graphics or by the use of polygons and twodirectional xy lines in vector graphics. 3D images were identified by their representation of one view of solid objects in a 3 60 environment, by their use of polyhedrals, or by three-directional xyz lines that created a sense of perspective with a vanishing point. In my previous study, subjects had similarly referred to 2D space as being "coplanar" (Johnson, p. 124) and 3D space as having "multiple planes in a void" (p. 125).  55  To identify 2D and 3D digital properties, I began with a list of those properties that were most clearly defined in my earlier study, and added new properties as they were found in the images. Properties which did not occur in more than 2% of the images were dropped from analysis. The resultant choices are by no means inclusive, but represent a range of common properties. They are defined on pages 60-67 and discussed in more detail in Chapter 4. Definition of items Early in the course of assembling the material for content analysis, I began creating a glossary to define characteristics that were apparent in the images. These served as what Crowl termed "operational" definitions (Crowl, p. 85), or temporary definitions created for the purpose of analysis. Wording for many terms and their definitions were subsequently revised according to suggestions made by the raters. Each characteristic is defined in words, as well as by a prototype image selected from the sample. Again, raters made suggestions regarding these choices. The images, or selected areas of them, are shown in reduced form. See Appendix A for credits.  56  Formal properties square format within a format that approximates a square  black ground created on a black background or in a black void  limited palette in a monochromatic or analogous color scheme  unlimited palette using the full spectrum of color  57  Formal properties  black and white 1). without color, or 2) using the gray scale only 3) using black and white only  cropped view shapes or objects are "cut off" by the format; not completely revealed  uncropped view shapes or objects are not cut off by the format; they may be centered or have some space between their edges and the format  2D graphics space found in images produced by Paint, 2D Draw, or imageprocessing programs 1) the image is created on a single-plane surface (one bitmap) or 2) the image is composed of polygons or xy lines  58  Formal properties  3D graphics space found in images produced by 3D programs 1) the image shows one view of objects in a 360 environment or 2) the image contains polyhedrals or xyz lines  realistic scale objects in the image are in natural proportion to each other  imaginary scale objects are not in natural proportion to others  realistic proportions individual parts of an object are in natural proportion to the whole of the object  59  imaginary proportions objects in the image have exaggerated or "mannered" parts, with some parts out of proportion to the rest of the object  high value 1) high contrast, or 2) exhibits a range of light and dark tones  low value  1) low contrast, or 2) exhibits an overall grayness or opacity  high resolution  1) finely rendered, or 2) smooth or detailed surface without pixellation  60  Formal properties  low resolution 1) coarsely rendered, or 2) pixellated  grid lines 1) imagery designed with a grid structure, or 2) imagery on which a grid structure is imposed 3) imagery including structures with vertical and horizontal grid lines frame or border imagery with a "picture" frame or border  other formal partitions imagery that is sectioned or structured into sections other than a grid format or bordered  interactive imagery that is primarily static, but also contains hyperlink or hypertext features  61  Content  human figure a representational or stylized human figure (including realistic toy imagery)  head only a human face, or upper torso and face  other body part a recognizably human feature, such as a hand or lips  combination a combination of human figures, heads and/or body parts  62  Content  humanoid, spirit figure or robot an imaginary figure based on the form and proportions of a human figure  animal a representational, identifiable animal  insect a representational, identifiable insect  bird a representational, identifiable bird  63  Content  creature an imaginary, animal-like form or figure  computer representation of recognizable computer components; hardware or software imagery  machinery an object based on mechanical or engineering structures  technology electronic equipment or devices other than computer hardware or software  64  Content  vehicle an object for human transportation (such as a car or plane) including toy vehicles  weapon an implement for fighting or defense (such as a gun or spear) including toy weapons  clock an analog or digital timepiece  clouds cloud formations in a real, imaginary or reflected sky  65  Content  flora or vegetation natural or imaginarybotanical growth  landscape or exterior scene 1) a view or vista of the earth, or 2) an exterior setting  water depicting wetness, the sea, a lake, waterfall, etc.  religious building an architectural structure built for spiritual purposes such as a church or temple  66  Content  other architecture depicting buildings, parts of the built environment or cityscapes  planet an earthly body or sphere (such as a sun, planet, or meteor) in the sky or outer space  still life objects 1) common everyday objects, or 2) small-scale decorative objects  abstract polygon 1) a flat shape, or 2) a two-dimensional plane  67  Content abstract polyhedral a solid object in a 360 "virtual" environment  text as an aspect of an object words, lettering, alphabet shapes, numbers or narrative used in a denotative manner (such as a label on a jar or a traffic sign)  text as an aesthetic element words, lettering, alphabet shapes, numbers or narrative used 1. metaphorically 2. as design or compositional elements  68  Style  simulation the imagery represents the natural, "real-life" appearance, scale, details and proportions of an object, scene or person  visualization the imagery represents an imagined or proposed object/ scene in a natural, real-life manner  illustration 1. the imagery is fictional or exaggerated, while still retaining elements of the "real" (such as cartoons or poster-art 2. the image is not a simulation, visualization, image-processed work, or collage 3. the style does not intend to simulate traditional drawing or painting  computation 1) imagery resulting from direct computer programming, or 2) an algorithmic display or design  69  Style  image processing the image is largely based on one photograph or other found image, which has been scanned and then altered with computer tools and/or filters  collage the image is made of a composite or montage of two or more images which have been scanned and manipulated to create one fused image  painting the imagery simulates the appearance of traditional painting styles (such as "brushwork", paint "washes", or freehand modeling)  drawing the imagery simulates the appearance of traditional drawing styles (such as sketching or freehand line drawing)  70  2D digital properties  2D segments freehand lines consisting of rows of contiguous pixels, created with raster graphics (Paint programs)  2D vectors straight lines that connect points created with 2D (Draw) programs . . . . . .  2D arcs or splines mathematical curves seen in images produced by 2D (Draw) programs . - . .  2D framing freehand segments that outline 2D forms (called "contour lines" in art education)  71  2D digital properties  mirrored imagery that has been "transformed" by copying and flopping (turning it over) or repositioning  2D iteration equivalent items or images that are repeated by the computer in a serial or patterned manner  2D motion convention suggesting movement through traditional painting and drawing conventions, such as 1) blurring or smudging the appearance of an object, or 2) the use of repeated lines or off-registered images, or 3) the use of diagonal or curving lines/forms  scanned photos photographs or other images that have been digitized into binary code for use with graphics software  72  2D digital properties  block-pixing the use of low resolution to create visible "blocks" or squares of solid-color pixels  2D drop shadow an off-set duplication of the original shape, calculated by the computer, which acts as a "shadow"  2D gradation a gradual blend from one color or tone to another on a flat plane (usually as a background area)  modeling freehand gradations of color or tone to suggest form and depth  73  2D digital properties  filtered "paper surfaces" or areas of a bitmap which have been treated with special computerized effects (such as embossing)  area fill a computer pattern or texture applied to an enclosed area of a 2D image  photo-retouching the use of computer tools to color or blend areas of a scanned image  74  3D digital properties  3D vectors straight lines that connect points to create a 3D framework , „ . - • * •  3D arcs or splines mathematical curves that create a 3D framework -  3D objects forms with height, width and depth created using a 3D graphics program, including 1) polyhedrals 2) 3D wireframes that have been covered with "materials" or surfaces 3) 3D solid objects which have been assigned such physical properties as weight or mass  void or reverse area the virtual environment of 3D graphics, which appears as an infinitely receding space  75  3D digital properties  3D recursion imagery which has been copied increasingly smaller or larger in scale or proportion by the computer (generally seen in fractal or other computational art)  ambient lighting 1. having a computerized light source which creates matte or even lighting and which can be adjusted for contrast level in the scene 2. soft general lighting without cast shadows  illumination 1. having a computerized light source with a cone (or beam) which can be positioned, intensified and targeted to create contrast and cast shadows 2. having omni or spot lights in addition to ambient light  specular highlights highlights produced on high spots as a result of one or more spotlights in 3D graphics  76  3D digital properties  3D cast shadows shadows produced byspotlights or directional lights in 3D graphics  3D motion the suggestion of movement through computerized techniques (such as a motion trail)  continuous shading 1) shading on 3D objects created during a computer procedure called rendering 2) similar to "modeling" in art education  3D transparency a computer procedure that causes the less-detailed or refracted appearance of 3D objects seen through a "transparent" surface  77  3D digital properties  3D atmospherics a computer procedure that causes progressive color and detail changes in objects receding into the distance of a scene or the void  3D image map photographs or other images that have been digitized for "mapping" (wrapping around) onto surfaces  3D reflection 1) a "mirror"-like image on a surface that has been assigned reflective qualities or 2) an image which has been mapped on a 3D surface to simulate a reflection (called a "reflection-map"  texture-map a pattern that is mathematically applied, wrapped, and scaled by the computer onto 3D surfaces  78  T=T  Internal validity In addition to checking and cross-checking my own analysis, I relied on five independent raters for a criterion-referenced evaluation of the terms and their frequency of use in the sampled images. Raters reviewed the operational definitions and drew on a range of professional experience in computer art, from eight to fourteen years, in analyzing the images. Procedures used in conjunction with raters The independent raters validated the analysis in three stages. In the first stage, acting as what Becker called members of art worlds, they reviewed the terms and definitions of all 98 characteristics identified in my analysis. Drawing on their previous experience and expertise, raters suggested some changes in wording or definitions. Each rater also examined the pictorial images accompanying the definitions, and two raters suggested replacement images where necessary for increased clarity. To provide an interrater agreement validation, they then analyzed a total of 50 images or 25% of the sample. I sought to avoid the effects of instrumentation by having each rater analyze only ten images. Each of the five raters randomly chose ten images, without looking, from the 2 00 hard copies. Using the rater taxonomies, they checked the images for each characteristic. They were encouraged to refer back to the definitions as they did so. For stability,  80  they checked each image for all characteristics, and then checked each characteristic in all images. In the third stage, I checked the rater findings against the findings in my own taxonomy, and isolated areas of disagreement. I then discussed these differences with raters and, where necessary, re-directed them to definitions and comparative examples. The raters acted as peer reviewers in establishing common grounds of agreement in their findings. Selection of raters The raters were selected for their previous knowledge in the area according to three criteria: the length of time they had worked in the field; their confidence in differentiating between images created with 2D software and 3D software; and their familiarity with graphics terminology. They represented different kinds of software specializations and different fine art backgrounds, ensuring a breadth of experience in the field. All raters were also affiliated with professional computer companies or institutions. Each rater was paid a stipend of $50.00. They spent up to three hours each, analyzing and discussing their selected images. They all demonstrated extremely high visual acuity, with a marked attention to detail and to the procedures used to create the images. Table 1 is designed to present a profile of their computer specializations and fine art backgrounds.  81  Table 1 Rater- profiles  Job specialization 3D animator and Web programmer architectural drafting (CAD) graphics software designer 3D animator and lecturer computer-generated special effects for film and t.v.; graphics instructor  Software specialization 3D graphics and animation 3D graphics and Paint programs image-processing and computer vision 3D animation and 2D image-processing photo-compositing, 3D animation. Paint programs  Fine art drawing, photography, fine art collector art history drawing, ceramics, electronic sculpture M.A. Fine Art, drawing and painting drawing, painting, film, animation  Interrater agreement Percentages of agreement before and after stability measures (p. 78) are presented in Table 2. The percentages were calculated by multiplying the number of agreements in the 50 selected images by two.  82  Table 2  Percentages of i n t e r r a t e r agreement, before and a f t e r the a p p l i c a t i o n of s t a b i l i t y measures type style format color view scale proportion value resolution dividers people animals mechanical clouds  2D iteration 2D motion scanned photos 2D gradations 2D modeling filtered 2D area fill photo retouch 3D vectors 3D objects  before 96 78 96 88 92 92 98 98 92 98 100 94 94 100  before 100 92 100 96 98 96 94 92 100 96  after 98 88 100 92 96 96 100 98 98 98 100 100 98 100  after 100 100 100 100 100 98 100 96 100 100  before a f t e r buildings 100 100 text 96 98 abstract shapes 82 100 borders/frames 98 100 black ground 96 100 interactive 100 100 planets 96 100 flora/vegetation 94 98 landscape 100 100 water 100 100 still life objects 88 98 segments 92 100 2D arcs 92 100 framing 90 100  before 3D void 100 3D image-map 94 3D lighting 98 94 3D highlights 3D cast shadows 98 3D motion 98 continuous shading 92 transparency 96 atmospherics 94 reflections or 100 reflection maps  83  after 100 98 100 100 100 100 100 100 100 100  Variables in interrater agreement The following contributed to disagreements during the initial assessments. Instrument error. Because each page of the definitions did not have a heading for the categories at the time, raters made some mistakes in matching the taxonomic items. For example, cast shadows in a 3D image were checked as 2D modeling, even though the rater had marked "space" as 3D. Selection of graphic type. Selection of a different type of graphic (for example, 2D rather than 3D) resulted in the selection of different digital properties. Single rater variables. One rater was responsible for 44/110 errors due to less experience with 2D graphics. Rater observation difficulty. Raters did not initially observe the presence of some content, such as "polygons" or "planets". Subjectivity. Six characteristics were found to have a high degree of subjectivity: "simulation", "visualization", "illustration", "limited palette", "unlimited palette", "resolution" and "still life objects". Information overload. Raters were not previously trained in the application of the terminology created for this study, and experienced some exhaustion in the course of their initial assessment.  84  Procedures for a stability estimate of rater reliability The following measures were taken during discussions of the findings to provide chance for raters to adjust their initial choices. Redirection to definitions. Raters were reminded of particular definitions and asked to re-evaluate these characteristics. Comparing two or more images with the same characteristic. Raters were asked to examine several images at once, for characteristics that had uneven results in their check lists. Contextual descriptions. Raters were provided with further explanation of how terms like "modeling" or "limited palette" are used in art education. Following these discussions, raters voluntarily made some changes to their check lists. Comparison to previous ratings. Problematic images were re-examined by all raters, and validated or reclassified.  Data analysis Descriptive or summary statistics were used to organize, reduce and summarize the large numbers of observations, as suggested by Creswell (1994, pp. 121-156). The analysis provided both an estimate of the frequencies of characteristics and demonstrated how the variables were distributed by gender and types of sites. Analysis was done in three stages. 85  I analyzed all images for 98 characteristics and recorded the results horizontally, as a taxonomy. I also analyzed the characteristics vertically on the taxonomy, image by image, as a double-check. For a stability estimate of reliability, as suggested by MacMillan and Schumacher (p. 266), I repeated this process at two different times. Findings were further refined following discussions with raters. During this process, I used the operational definitions as exemplars and discussed borderline occurrences with raters. As a result of these discussions, I refined the wording of some definitions and, where necessary, selected more representative pictorial examples. I then eliminated several characteristics from subsequent analysis either because they proved too difficult to identify, or were rarely found in the images: penumbras, morphs, bump-mapping, recursion, interpolation, and procedural modeling. Second stage of analysis I entered all taxonomic data into SPSS for Windows, Student Version (see Appendix E for an example). The student version allows 50 variables and 500 cases. This study produced 50 variables among the 200 cases, with a total of 88 value labels for the 50 variables. Characteristics that were similar in type but had not been found to occur simultaneously in images were grouped as values under one' variable label. Characteristics that were found in the same  86  image as other characteristics were given separate variable labels. As the data collection proceeded, I had noticed some similarities in imagery collected at different sites. I therefore decided to include the gender of each artist and the type of electronic site in the content analysis. To do so, I returned to each Web address and identified gender either by the artist's name at the Web site or by curatorial descriptions of the work. I could not identify gender for two images. I also found that the imagery was produced at what appeared to be three general kinds of sites, which were labeled or described as follows: a higher education "school" site, where work by college, university or art school students was showcased; an "association", where professional computer artists had grouped to present both commercial and fine art; and "private" or independent sites, where individual artists exhibited their work. The nature of these sites is described in detail in Chapter 4. In the taxonomy and subsequent discussions, I have used the terms "school", "association" and "private" to denote the three types. The first SPSS variable was "gender". The value labels of "male" or "female" were ultimately used, rather than "boy", "man", "girl" or "woman", since some sites indicated the age of artists and others didn't. This study recognizes, however, that the terms "male" and "female" are generally used to denote sex, rather than gender.  87  The second variable was "site," with three value labels of "school", "association", or "private". Four of the SPSS variables had up to eight values each. These were "style" (including simulation, visualization, illustration, computation, image-processing, collage, drawing, or painting); "people" (human figure, head only, other body part, humanoid/robot, figure and head, humanoid/spirit figure, or not applicable (N/A); and "animals" (animal, insect, bird, creature, or N/A). Images of hardware and software, machinery, other technology, vehicles, clocks and weapons were grouped under the variable label "mechanical". I then used SPSS to count the frequencies of each characteristic in the four categories. An additional nine characteristics were eliminated following the data entry because they were found in less than 2% of the artwork. These were 2D vectors, 2D flipped, 2D mirrored, 2D drop shadows, block-pixing, 3D arcs/splines, 3D wireframes, 3D flipped, and 3D iteration. The remaining 50 values were grouped under 44 variable labels (see Appendix F ) . The majority of these characteristics had two values: they were either present or not applicable. Third stage of analysis Correlational statistics were used to determine how the variables related to one another. In order to summarize the findings efficiently in Chapter 4, I have discussed the 44 variables in eight componential categories: gender, site,  88  type (2D or 3D), style, formal properties, content, 2D digital properties, and 3D digital properties. I also examined the relationships between the nominal variables for their expected and actual frequencies. The sample was large enough to accommodate the findings of small differences or slight relationships. As a result of variables in the selection of samples and in interrater agreements, a larger sample may have proved useful in offsetting errors inherent in small, and perhaps atypical, samples.  89  Chapter 4 Analysis and discussion of data Introduction This chapter is designed to discuss the analysis of data and to suggest possible interpretations. I assessed all 200 computer art images and 50 were assessed by five independent raters, to determine what characteristics were commonly found in digital imagery created by adult computer artists. Areas of disagreement are mentioned throughout this chapter. It should be noted that all interpretations are of a conjectural nature. They are intended to serve as a starting point for educators interested in developing curriculum in this area.  Findings and interpretation Gender I analyzed 200 images by 200 artists for the study. As previously noted, the gender of 2 artists could not be identified from data at the gallery site. The remaining 198 images were produced by 59 females and 13 9 males, or 2 9% female and 69% male artists. More than two-thirds of the images were therefore produced by males. Almost half of all images by females were produced at universities or other educational sites, while almost half of all images by males were found at association or gallery sites and one-third of art by males was found at private  90  sites (Fig. 1). Art by females was least represented in private sites and art by males was least represented at educational sites (Fig. 2).  Figure 1. Percentage of males and females at each type of site 120  artist's g e n d e r  c  Ifemale  0)  o  I male  L.  Q.  lunknown school  private association  T y p e of W e b site  91  Figure 2. Percentages of sites by gender  120* 100' 80' 60'  T y p e o f W e b site 40-  c  8 Q.  Ischool 20'  ^association I private female  unknown male  artist's gender  The implications of the findings for gender and site are not encouraging. Statements at association sites in particular indicated an emphasis on promotion or commercial representation. That female computer artists were represented in only 21% of these would suggest either that females are not taking advantage of the type of sites that would promote professionalism or make their work commercially viable, or they do not have equal access to them.  92  To some extent, however, female artists appear to have deliberately avoided public exhibition of their work. As earlier discussed, many female-only association sites were impossible to access without first obtaining membership or other pre-conditions. This obviously limits the extent to which female artists may have their work seen by a wider audience. Artwork by females tended to show some technical differences from that produced by men. Because female artists were twice as likely as male artists to be in schools, they were inclined to depend on Web sites created by their department or instructor, rather than creating their own. In the process of being "batched", these images suffered some degradation. I also had problems downloading eight artworks which later all proved to have been created by female artists who were not in schools. Links to female artwork were likewise not as well established. While associations with high numbers of male artists created extensive links to other sites, associations with high numbers of females were difficult to find and even more difficult to access. Male artists also worked in significantly higher resolution than did female artists (61% compared to 34%, Table 3 ) , which increased the overall clarity and detail of their artwork. I expected to find that male and female artists used formal properties equally. However, slightly more male than female artists used a horizontal format (Table 3 ) . Female 93  artists experimented slightly more with vertical and square formats. More male than female artists applied a realistic scale to their content, and slightly more males than females used realistic proportions. Male imagery overall tended to be more traditional in the use of such formal approaches as format, scale and proportion than did female imagery.  Table 3 Percentages of Males and Females Using Each Identified Formal Property  Formal property no format horizontal format vertical format square format limited palette unlimited palette black and white uncropped field of view cropped field of view realistic scale imaginary scale realistic proportions imaginary proportions high value low value high resolution low resolution dividers borders black ground  % of males >5 59 (high) 24 13 47 53 0 39 61 50 (high) 38 53 33 75 25 61 (high) 39 17 21 53  Note: The expected ratio was 50:50.  94  % of females >5 51 27 19 42 49 8 34 66 36 41 47 29 75 25 34 66 (high) 24 25 49  The frequency of use of human figures in imagery by males and females was equal (Table 4 ) . However, almost all of the nude imagery of female figures was produced by female artists, perhaps as a result of the large number of lesbian artists with private sites. One of the raters also pointed out that female nudes by male artists tend to be traded on anonymous FTP sites, since male artists are concerned about the legal and social repercussions of exhibiting female nudes at private sites. Imagery of the human figure created by female artists was generally more stylized than realistic. The use of abstracted imagery was also noticeable in all imagery by females. For example, female artists used a higher percentage of abstract polygons (Table 4 ) . 20% of female artists used text as a decorative or symbolic feature in their work, as compared to only 10% of males.  95  Table 4 Percentages of Males and Females Using Each Type of Content  Content % of males % of females human figure 22 22 head only 9>5 other body part >5 7 combination >5 >5 humanoid/robot >5 5 all animal imagery 15 12 all mechanical imagery 27 (high) 8 all buildings 18 14 landscape/cityscape 25 19 water 9 (high) >5 flora/vegetation 27 (high) 19 planets 11 (high) >5 clouds 27 (high) 10 still life objects 25 28 text as an aspect 9 (high) >5 text as an aesthetic 10 20 (high) polygons 23 34 (high) polyhedrals 12 (high) >5  Note: The expected percentages were 50:50. A summary of gender preferences for content (Figure 3) shows that male artists generally used a wider range of content than did female artists, and it tended to be more realistic. By comparing numbers in Table 4, I found that more male imagery than female imagery emphasized "nature" content, such as landscape, flora and vegetation, clouds, planets and water. More males than females produced  96  mechanical imagery (27% versus only 8% of all female artists), and imagery of buildings. More males than females used text as an aspect of other content such as building signs or product labels; that is, to convey a denotative message.  Figure 3. Gender preferences for content (percentages of use in images)  I females 1 males  20  30  97  40  50  The most significant gender findings, however, were in the area of program type and the use of 2D and 3D digital properties. (2D and 3D graphics programs are described in more detail on page 88). Figure 4 shows that female artists produced significantly more two-dimensional imagery (90% versus 65% of male artists).  Figure 4. Percentages of males and females using 2D and 3D graphics  120 100  artist's gender  •  •female • male  0 0)  • unknown 2D  3D  Type of image  98  Female two-dimensional imagery was mainly in the styles of collages and paintings, which are further discussed on page 103. These were followed by much lower but similar amounts of visualizations, illustrations, computations, image-processed works and drawings (see Figure 5). While freehand segments were the most common 2D digital property in both male and female imagery (Table 5 ) , female artists used proportionately lower percentages of more technical computer procedures, such as splines, iterations or gradations than would be expected by the 65:90 ratio of male to female artists using 2D graphics. They tended to rely on scanning and freehand drawing for the majority of their 2D images.  99  Table 5 Percentages of Males and Females Using Each 2D Digital Property  2D digital property arcs/splines framing iteration motion gradations scanned photos modeling freehand segments filters area fills  % of 2D images 9 33 23 22 21 42 34 48 19 37  % of 2D by males 9 25 19 17 17 25 22 32 12 23  % of 2D by females >5 (low) 22 (low) 18 (low) 17 (low) 15 (low) 41 32 44 19 37  Note: To determine whether the percentages were high or low, they were compared to the 65:90 ratio of male to female artists using 2D graphics (Fig. 4 ) .  Unlike 2D programs, which provide one window with an xy central axis and contain tools that can simulate intuitive art techniques like sketching and modeling freehand, 3D graphics programs require users to "build" structures in a 3 60 void. This is usually done in four viewports that represent different perspectives of the structures. Once structures are "built" with lines resembling a wire sculpture (called a "polyhedral"), each surface of each structure is "assigned" a material surface. The surfaced structures are called "objects" rather than "polyhedrals".  100  They are lit by placing and adjusting different kinds of lights in the scene. The "void" in which they are created must be handled three-dimensionally as well, rather than treated as a static "background", as it is in 2D programs. The final art image is one selection from a wide range of views. 3D programs are thus quite technically challenging and certainly non-intuitive in the traditional definition of drawing and painting. The findings in Figure 3 revealed that more than a third of male artists produced art using 3D programs, compared to only 10% of female artists. And more than three times as many male artists created visualizations and simulations, which are generally created on 3D programs (Figure 4 ) . Conversely, between 77 and 82% of simulations, visualizations and illustrations were produced by males (Figure 5 ) . These types of images, further discussed under "styles", tend to represent geometry-based content like buildings or vehicles.  101  Figure 5. Number of males and females using each style  style  artist's g e n d e r •female •  male  •  unknown  %.%L % > % %  W*  9,  Figure 6 Percentages of styles by gender 120  style 100  i  80 60 40  c  (D 0  20  i_  (D  a.  0  M  female  male  102  •  simulation  •  visualization  •  illustration  •  computation  •  image processing  •  collage ^drawing  n unknown  painting  The tendency for female artists to produce mainly 2D art may result from less access to 3D programs at educational sites, where females were more frequently found to be situated, or perhaps from the tendency of artists using the more technically challenging 3D programs to exhibit in commercial associations or private sites, where males tended to be situated. It may also result from content preferences among female artists for non-geometric imagery, which may, in turn, be a cultural artifact. Much higher use of all 3D properties was found in artwork by males (Figure 7 ) . Even the smaller amount of 3D work created by females used 3D objects less frequently and applied continuous shading (a result of placing spotlights in the scene) less frequently. Voids in images by female artists using 3D programs were also less well-defined.  103  Figure 7 Gender preferences for 3D digital properties (percentages used in images!  10  15  20  25  30  35  In Table 6 it can be seen that, as in their use of 2D programs, female artists who used 3D programs used a lower percentage of specialized procedures. For example, the applications of image-maps, transparency, atmospherics or reflections in particular were lower for females than a  104  35:10 ratio of male and female 3D artists lead one to expect. Table 6 Percentage of Males and Females Using Each 3D Digital Property  3D property vectors objects void image-map ambient light illumination highlights cast shadows motion continuous shading transparency atmospherics reflection  % of 3D images 20% 98% 80% 36% 20% 73% 25% 55% 24% 76% 36% 45% 31%  % of males 8% 35% 29% (high) 13% 8% 29% 8% 19% 9% 27% (high) 19% (high) 19% (high) 12% (high)  % of females >5% 10% 8% >5% 0 8% >5% >5% >5% 8% >5% >5% >5%  Note: To determine whether the percentages were high or low, they were compared to the 35:10 ratio of male and female artists using 3D graphics (Fig. 4 ) .  These findings are disturbing. The majority of work in computer-related fields (such as architecture, engineering, medical imaging or scientific visualization) is currently done with 3D programs, but female artists--at least in this  105  study--did not demonstrate an interest in the potential of these programs. Nor did they appear to be as advanced in their use of tools and procedures in either 2D or 3D programs. At least in terms of these fields as they have been developed by males, female artists appeared to be limited both technically and aesthetically. They did not appear to be participating to the same degree as male artists in the exploration and evolution of a full spectrum of computer mediums as a fine art form. Nor did they have the same repertoire of choices for style and technique as did male artists. Female artists may not be receiving sufficient education in the full potential of computer tools and programs, or they may not be situating themselves in environments where they can learn. Conversely, male artists may not be encouraged to work with more intuitive or less skill-demanding tools to the same degree. I therefore looked next at the types of sites where artwork was produced. Site I have earlier described what appeared to be three kinds of sites in the on-line computer art world: schools, associations, and private sites. "Schools" included institutions of higher education only, such as art schools and university or college art departments. "Associations" included galleries and collective exhibitions produced outside schools, as well as professional organizations. 106  "Private" sites were those created and maintained by an independent artist, although, interestingly, private sites were frequently presented in the course of this study, in home pages and links, as if they included a group of artists. More often than not, I had to look at all imagery at a private site before determining that it was created by one artist. Associations created the largest number of Web sites (42%) compared to a smaller but almost equal number of school and private sites (28% and 29%, respectively). This finding indicates the importance of collective sites in promoting computer artwork. The greater numbers of artists in each association (frequently up to 50 artists) makes them a powerful force in the computer art world. Artists in associations also offered a greater number of art-related services, such as Web design, illustration or animation. Larger associations in general created more Web links than did schools or independent artists. This increased the potential number of "hits" on associations and therefore the visibility of the artists who exhibited at associations. On the other hand, artists at private sites stood a better chance of having their artwork viewed once a connection had been made, for navigating the larger sites took more time, and called for greater persistence. Schools and private sites produced the greatest percentage of 2D artwork, while associations produced more than twice as much 3D artwork as did schools and private 107  sites (Figure 8). Even more simulations and visualizations, which are created with 3D graphics, were produced byassociations than the 28:42:29 ratio of artists at each type of site would suggest (see Table 7).  Figure 8.  Percentages of 2D and 3D a r t at each type of s i t e 120r 100  T y p e of i m a g e o  Itwo-dimensional Ithree-dimensional school private association T y p e of W e b site  As indicated in Table 7, schools and p r i v a t e s i t e s produced almost no simulations. This would suggest e i t h e r t h a t t h e i r members have l i m i t e d access to the kinds of software t o o l s required to b u i l d extremely n a t u r a l i s t i c imagery, or t h a t t h e i r members do not p r e f e r - - o r p o s s i b l y  108  accept--simulation as an art form. Notes taken from home pages during the data collection indicated that artists at private sites were particularly concerned with "having fun", "using imagination", and "doing nothing serious". They frequently asked visitors to "enjoy" themselves and "celebrate diversity". Unfortunately, most school sites did not provide statements about the art created, or rationales for it.  Table 7 Percentage of Sites Producing Each Style  styles simulation visualization illustration computation image process collage drawing painting  % schools >5 11 23 (high) 16 (high) 14 14 >5 18 (high)  % associations 13 (high) 21 (high) 19 >5 8 (low) 11 (low) 9 15 (low)  % private >5 10 17 9 17 28 (high) >5 14  Note: To determine whether the percentages were high or low, they were compared to the 28:42:29 ratio of each type of site (p. 100).  Schools produced proportionately high percentages of illustrations, computations and paintings, while private artists created the most image-processed work and collages.  109  This may indicate a disproportionate emphasis on found imagery by artists at schools and private sites. While computational art was a significant art form among members of schools and private sites, associations produced almost no computations. This finding suggests either that members of associations do not have experience or education in generating mathematical art, or they do not choose to promote this style. Overall, with the exception of simulations, school and private site artists used a greater range of styles. They also used each one with greater frequency than did members of associations. Some differences were observed in the kinds of 2D properties emphasized by artists at the three kinds of sites. Table 8 shows that artists in schools used higher percentages of arcs and splines and iteration than would be expected (these findings are later discussed in more detail). Artists at private sites emphasized imageprocessing techniques, such as scanned photos, filtering, and re-touching. Members of associations tended to do traditional freehand drawing rather than using either imageprocessing procedures or vectors. For example, they used greater percentages of framing (called "contour lines" or "outlining" in traditional art) and freehand segments than might be expected, in comparison with the output of the other groups. They also tended to simulate motion through freehand sketching. These differences between artwork  110  produced at each site argue for computer art worlds that have possible differences in aesthetic conventions.  Table 8 Percentages of Each 2D Digital Property Used by Artists at Each Type of Site  % 2D property arcs/splines framing iteration 2D motion scanned photos gradations modeling filtered area fill re-touching freehand segments  schools 53% 29% 44% 21% 30% 31% 35% 32% 33% 18% 30%  associations  (high) (high)  27% 46% 28% 48% 26% 28% 35% 14% 37% 35% 44%  (high) (high)  (low)  (high)  private 20% 25% 28% 30% 44% 41% 30% 54% 29% 48% 26%  (low) (high) (high) (high) (high)  Note: To determine whether the percentages were high or low, they were compared to the 32:35:32 ratio of 2D art at each type of site (Fig. 8 ) . Associations not only produced more than twice as much 3D artwork as did schools and private sites, but association artists used the more sophisticated 3D properties to a much greater degree than would be expected by the 18:40:19 ratio of artists at the sites.  In particular, artists in  associations demonstrated a much stronger use of 3D motion,  ill  transparency, atmospherics and reflections in their work (see Table 9 ) . Artists at schools and private sites used even the most basic 3D features less frequently, such as the use of a void to increase spatial depth, or the application of illumination or reflections to increase illusion. Table 9 Percentages of Each 3D Digital Property Used by Artists at Each Type of Site  % 3D property vectors objects void image-map ambient lighting illumination highlights cast shadows 3D motion shading transparency atmospherics reflection  Note:  schools 18% 19% 18% 30% 9% 19% 29% 23% 23% 24% 20% 8% 24%  associations 73% (high) 61% 62% 55% 73% (high) 73% (high) 50% 57% 62% 62% 70% 76% (high) 71% (high)  private 9% 20% 20% 15% (low) 18% 18% 21% 20% 15% 14% 10% (low) 16% 6% (low)  To determine whether the percentages were high or  low, they were compared to the 18:62:20 ratio of 3D art at each type of site (Fig. 8 ) .  112  In general, a profile of artists exhibiting in associations would include an emphasis on realistic imagery created with 3D programs rather than drawn or painted freehand, or scanned from photographs. As for the smaller number of artists at schools or independent sites producing 3D work, it may be that they lack experience with the full potential of 3D programs. This in turn may have resulted from less access to 3D programs with the necessary technical tools (particularly high-end 3D programs); less studio education in their use; or less aesthetic education in the value of 3D tools for the production of naturalism. Certainly the categorization of art by sites provided a useful start on understanding the field itself. Subsequent analyses in this study continue to develop the implications. Type Altogether, 145 images (72%) were two-dimensional and 55 images (27%) were three-dimensional. Almost threequarters of all artwork therefore was created with Paint, Draw or image-processing programs. I did not, however, categorize 2D art into these types of programs because many programs layer features and it would be difficult to do so. However, the use of such properties as arcs and splines, filters and scanned photographs is later analyzed under 2D properties (p. 152), which gave me an idea of their emphasis in 2D imagery generally. The larger number of 2D images was predictable due to the wide-spread availability of 2D programs--particularly 113  Paint programs, which can be found even as an accessory in most word-processing programs. 3D programs at the consumer level (i.e. based on Windows or Macintosh platforms with pull-down or pop-up menus) have been released only in the last few years. Artists in schools or at private sites may have been disadvantaged by time lags in acquiring 3D programs due to such variables as budget constraints or the limited potential of older hardware to accommodate the more spacious 3D programs. The considerably higher resolution of the 3D images in this study, compared to the low resolution of 2D images (Table 10), is one significant factor in support of the possibility of hardware constraints. Artists in schools and private sites might be disadvantaged in a cultural sense as well, since, at least in this study, there were fewer 3D computer art examples to serve as exemplars.  114  Table 10 Frequencies of Comparable 2D and 3D Techniques  technique % of 2D images modeling/shading 34% motion 22% scanned photos/image maps 42% segments/vectors 48% (high) realistic proportions 48% imaginary proportions 34% realistic scale 43% imaginary scale 38% limited palette 46% unlimited palette 51% high resolution 47% low resolution 53% (high) uncropped view 37% cropped view 63% high value 74% low value 26% black ground 45% dividers 21 (high) borders 25 (high)  % of 3D images 76% (high) 24% 36% 20% 58% (high) 27% 55% (high) 40% 45% 55% 71% (high) 29% 40% 60% 75% 25% 69% (high) 12 15  Note: The expected percentages were 50:50. Analysis of the frequencies of comparable artistic approaches in 2D and 3D artworks, as presented in Table 10, revealed a greater emphasis on realism in 3D work.' 3D imagery tended to have more modeling or shading; a more realistic scale between objects in the same scene; and more realistic proportions between the parts of individual objects. 3D artwork appeared more conservative as a result. But since simulations and visualizations intend to provide a  115  view of a naturalistic world, it is likely that the theoretical basis of 2D and 3D art differs. This implication was supported by other differences in the use of formal properties. For example, 3D artwork had a much higher incidence of black grounds. This may result from the fact that black is often the default color for the background in 3D graphics programs, and thus can reduce rendering time considerably. It may also be attributed to a greater desire for illusion among 3D artists, since a black ground on the monitor screen (where the art is created) increases the effects of highlights, color, and color modulations--and thus the effects of spatial depth. Conversely, 2D artwork tended to be the result of characteristics that increased the flat or graphic nature of the imagery. For example, 2D imagery used more dividers such as grids and partitions, and more borders or picture frames around.the artwork. The theoretical basis of 2D artwork may therefore owe more to a conception of art as design than to one based in naturalism. Style Raters found the categorization of images into styles to be their most difficult task, since many of the artists had combined several styles. Raters were asked to rate styles on a continuum; for example, that the style was "more like" a painting than any of the other style possibilities defined on pages 63 and 64. They were also asked to consider the appearance of each image as a finished product rather  116  than as a sequence of procedures. However, it should be noted that the categories of style had the highest incidence of subjectivity (see Table 2 ) , and rater bias toward their own specializations in computer graphics was also a possibility. For example, Appendix G depicts an image that was assessed by one rater as a 3D image created with "Poser", a recent 3D program for modeling human figures. Only by knowing the date of creation could I determine that it was, in fact, a 2D Paint image. The resulting analysis showed almost equal percentages of visualizations, illustrations and collages, followed by paintings and image-processed work, with fewer but a relatively equal number of computations, simulations and drawings (Figure 9 ) .  117  F i g u r e 9. P e r c e n t a g e s of  styles  30  20  19 16  "15*1  12 I  •  o  | 8 |  | 6 | *r %  V  | 15 |  <£  Q^  |6j %  °Q,  %.  %•  style  While male artists created a larger number of works in each style than did females due to their larger numbers in the study, collage was the most preferred art style for both male and female artists (Table 11). The popularity of collage may be due to the ease with which the computer can cut and paste seamlessly, compared to analog collage methods. Visualizations and illustrations, which were overall the most popular styles by male artists, appeared to  118  have in common a view of art as a means to realistically depict the imaginary. Table 11 Preferences for Style by Gender (counted and ranked)  style 1. collage 2. visualization 3. illustration 4. painting 5. image processing 6. computation 7. simulation 8. drawing Total  males 18 24 32 20 16 12 10 7 139  females 15 6 7 11 7 5 3 5  unknown  59  2  total 33 30 39 31 25 17 13 12  2  200  Although the expected percentages of styles per type of site would be 33% each, members of associations created higher percentages of simulations, visualizations, illustrations and drawings (Table 12). The bias here appeared to be pragmatic. Since the first three styles portray recognizable and illusionistic imagery, practitioners at these sites may be interested in displaying such work to visitors from other computer fields where artistic renditions such as simulations, visualizations and illustrations are valued. Indeed, the home pages of many association sites actively promoted the availability of  119  their artists for commercial work, and many of the association artists separated their pages into commercial and fine art portfolios. Notes made during the data collection also indicate that there was an overt emphasis on portraying skill and technique.  Table 12 Percentages of Styles at Each Type of Site  % of style simulation visualization illustration computation image process collage drawing painting  schools 8 20 33 53 (high) 32 24 17 32  associations 85 (high) 60 (high) 41 18 (low) 28 (low) 27 (low) 67 (high) 42  private 8 20 20 (low) 29 40 (high) 48 (high) 17 26  Note: To determine whether the percentages were high or low, they were compared to the 28:42:29 ratio of each type of site (p. 100).  The highest percentage of computational art was found at school sites, possibly as a result of the kind of mathematical education necessary to create such art. Computations were readily found as 3D art as well (Table 13). Both members of schools and artists at private sites appeared to be interested in "pushing" the boundaries of early 2D computational art by exploring computational art as  120  a 3D form. On the other hand, they may simply have less education in the use of "bundled" tools in 3D programs. Table 13 Percentages of each Style Using 2D or 3D Graphics  % of style  simulation visualization illustration computation image process collage drawing painting  2D images 15 37 56 59 100 97 100 100  3D images 85 63 44 41 0 3 0 0  All styles were analyzed to determine the 2D and 3D properties found in each. I was particularly interested in determining the factors that differentiated each style. Since the raters had indicated some subjectivity in their choices, I looked for properties that were particularly evident. Illustrations. I found that illustrations depended heavily on the use of such 2D properties as arcs and splines, framing, gradations and area fills (Table 14). The emphasis of graphic elements clearly differentiated illustrations from simulations and visualizations. Arcs, splines and framing--rather than continuous shading or cast  121  shadows--were used to exaggerate fictional elements. Elements of the "real" tended to be created with gradations and area fills, rather than with modeling or texture-maps. Again, this finding supported the theoretical basis of 2D art as "design" rather than having its roots in naturalism. Raters found it easiest to categorize art that was most graphic--such as cartoons, poster art, or heavily bordered or "framed" art--as "illustration" rather than "drawing" or "visualization".  122  Table  14  Percentages of 2D D i g i t a l P r o p e r t i e s Found in Each Style  styles  % arcs/splines  % framing  simulation visualization illustration computation image-process collage drawing painting  0 0 47 (high) 13 >5 27 >5 0  0 >5 27 (high) >5 5 15 19 23  styles  % gradation  simulation visualization illustration computation image process collage drawing painting  0 >5 31 (high) 19 >5 19 >5 16  styles  % filter  simulation visualization illustration computation image process collage drawing painting  0 >5 14 0 32 (high) 3 9 (high) 0 >5  % scanned  >5 15 >5 0 36 (high) 39 (high) 0 >5 % area  % iteration  0 >5 28 25 11 19 >5 >5 % modeling  >5 8 22 8 8 16 >5 31 (high)  % 2D motion  >5 >5 15 0 >5 15 15 42 (high) % segments  0 >5 23 0 0 14 17 36 (high)  fill  >5 >5 24 (high) >5 13 27 (high) 16 13  Note: Percentages are marked high or low in r e l a t i o n t o other percentages for each c h a r a c t e r i s t i c .  123  Collages. A high percentage of collages also contained arcs and splines, gradations, and area fills. However, these properties were heavily augmented with scanned photographs and filters, which were significant in distinguishing collages from illustrations. Less than 5% of illustrations used scanned photographs. Artists doing collages thus seemed to take particular advantage of 2D computer tools other than freehand techniques. Computations. Raters found computations easiest to categorize. Computations were clearly differentiated by the absence of scanned photographs, motion and filters, as well as by very low percentages of such freehand techniques as framing or modeling. Computations also had the highest percentages of iteration among the eight styles. On the other hand, drawings tended to be entirely freehand, with little use of iteration, gradations, scanned photographs, or filters. Paintings. Paintings could be distinguished from drawings and illustrations in several ways. Paintings had significantly more gradations than did drawings. Paintings also had more freehand segments compared to drawings and illustrations. They contained more modeling than did illustrations or drawings. Paintings could further be differentiated from illustrations by an absence of 2D arcs and splines and a much greater percentage of motion conventions (42% compared to 15%). Notes taken during the data collection indicated that many artists working in the 124  style of "painting" sought to imitate the styles of wellknown exemplars in traditional art (see Appendix H for an example). Thus, this style in computer art was frequently recognized by myself and those raters who had previous education in the history of traditional painting styles. Visualizations. The highest percentages of 3D vectors, 3D objects, and a void were found in visualizations (Table 15). They were differentiated from simulations by greater percentages of spotlights and cast shadows. As one rater noted, illumination--particularly spot lighting--makes imagery look less realistic. Artists doing visualizations may therefore use these properties for increased drama and "presence". Visualizations likewise had significantly more continuous shading than did simulations or illustrations. The most predominant characteristics of visualizations were atmospherics and transparency, which are used to emphasize the illusion of depth and distance. More than half of the visualizations used atmospherics, compared to only 20% of simulations and illustrations. Half of all visualizations used transparency, compared to 35% of simulations and 15% of illustrations. Thus the emphasis in visualizations clearly seemed to be on creating a high impact, almost "staged" setting. This was confirmed by two raters.  125  Table 15. Percentages of 3D Digital Properties Found in Each Style styles  simulation visualization illustration computation image process collage drawing painting styles  % vectors  27 45(high) 18 9 0 0 0 0 % ambient  % 3D objects  20 35 (high) 30 (high) 13 0 >5 0 0  27 27 9 27 0 9 0 0  19 38 (high) 33 (high) >5 0 0 0 0  styles  % motion  % shading  15 46 (high) 38 (high) 0 0 0 0 0  % image map  20 45 (high) 25 0 0 >5 0 0  %illumination % highlights % cast shadows  simulation visualization illustration computation image process collage drawing painting  simulation visualization illustration computation image process collage drawing painting  % void  20 33 (high) 31 (high) 11 0 >5 0 >5  24 33 (high) 31 (high) 10 0 >5 0 0  styles  % atmospherics  simulation visualization illustration computation image process collage drawing painting  20 52 (high) 20 >5 0 >5 0 0  126  0 36 (high) 50 (high) 14 0 0 0 0 %transparency  35 (high) 50 (high) 15 0 0 0 0 0  20 40 (high) 30 (high) 10 0 0 0 0 %reflections  24 35 (high) 35 (high) 0 0 0 0 0  Note: Percentages of each property are marked high or low in relation to other percentages for each property.  Summary of styles. I found that visualizations relied on 3D programs and their properties; simulations and illustrations used similar amounts of most 2D and 3D properties; while computations, image-processed work, collages, drawings and paintings relied almost entirely on 2D programs and their properties. The categorization of all artwork into these eight styles was clearly supported by the frequencies of particular digital properties for each, as well as by differentiations in the use of properties. Overall, raters expressed surprise at the diversity of styles in the sample of computer art. However, since many 2D programs now include 3D features and many 3D programs include 2D features, it would appear either that digital artists create styles by choice rather than necessity, or that they have limited access to programs which offer a combination of features. The finding that only four 2D images and one 3D image contained interactive features  ("hot  spots" to click and reveal other images) similarly suggests that computer artists either have limited access to newer programs, or that they prefer to work within these eight styles of computer art. For example, among the 200 images, there were no examples of morphing sequences, 3D Paint maps, radiosity renderings, reaction-diffusion techniques, or  127  evolutionary art--all of which could theoretical y form additional styles. Thus while conceptions of computer art may be definable, they are also fairly limited in scope. As one of the raters commented, expanded tools in newer software continue to limit the styles even though the choices of tools are wider than they were five years ago. Formal properties All images were assessed for their formal or "global" properties; that is, those characteristics which affected images as a whole. As previously noted, the majority of the terms are commonly used in fine art--particularly "format", "palette", "cropped", "value" and "ground". However, these terms were new to those raters with non-traditional art education backgrounds. Several raters commented that they knew what was "meant" by them, but had not known what they were "called". Table 16 presents the percentages of each formal property found in all 2D and 3D images. Additional rater comments are included in the discussions of each property.  128  Table 16 Percentages of Formal Properties in 2D and 3D Images  % of formal property realistic proportions imaginary proportions realistic scale imaginary scale limited palette unlimited palette high resolution low resolution uncropped view cropped view high value low value black ground  2D images 69 77 67 71 73 71 64 (low) 83 (high) 71 74 72 73 63 (low)  3D images 31 23 33 29 27 29 36 (high) 17 (low) 29 26 28 27 37 (high)  Note: To determine whether the percentages were high or low, they were compared to the 72:27 ratio of 2D and 3D images (p. 106).  Format. The term "format" is used in both traditional and computer art to refer to the size and shape of an image. Since the format of computer monitor screens is horizontal, I expected to find that most 2D and 3D imagery also was horizontal. Raters had the same expectation, and we found that the majority of males used horizontal formats, as did half of all females (Table 3 ) . However, artists creating 2D imagery--as opposed to 3D imagery--used disproportionately high percentages of other  129  format types. For example, 88% of all formats without defined boundaries, as well as 79% of all square formats and 80% of vertical formats, were found in 2D imagery (Table 17) . Table 17 Percentages of Each Format Type in 2D and 3D Images  % of format type no format horizontal format vertical format square format  2D images 88 (high) 67 80 (high) 79 (high)  3D images 14 33 20 21  Note: To determine whether the percentages were high or low, they were compared to the 72:27 ratio of 2D and 3D images (p. 106). More experimentation with formats therefore seemed to be undertaken by 2D artists. This may be a consequence of earlier work with analog mediums (such as drawing, printmaking or painting) or exposure to the wider range of formats in concrete art. I was interested in finding out who they were, in terms of their sites. Although I found that almost half of all artwork with a horizontal format was created by members of associations (see Table 18), association members were also responsible for more than half of all artwork with no format boundaries. It may be that association members prefer images with no format boundaries  130  for certain styles of art only, such as collages. As indicated in Table 19, collages were found to have the highest incidence of no format boundaries. However, it should be noted that it is more difficult to create art that has the appearance of no format boundaries for presentation on the Web. The edges of a white background may still be visible against a slightly grayer home page. Variations such as this are not entirely in the artist's control, since they depend on the calibration and brightness of the monitors of "visitors" to the site. Association members produced smaller percentages of work with vertical or square formats than would be expected by the 28:42:29 ratio of artists at each site. Thus, with the possible exception of collages, they did not seem to experiment with format types to the degree that the artists at private sites or schools did. Table 18 Percentages of Each Format Type Found by Site  % of format type no format horizontal vertical square  school 29 25 32 34  association private 57 (high) 14 (low) 46 (high) 28 38 30 31 (low) 34  Note: To determine whether the percentages were high or low, they were compared to the 28:42:29 ratio of the three types of sites (p. 100).  131  A relatively greater percentage of vertical and square formats were used by artists in schools and private sites. Vertical and square formats were used for computational art, compared to horizontal formats or formats without boundaries (Table 19), and most computational art was produced by artists at schools or private sites (Table 12). Since the range of formats in 2D art generally was wider than in 3D art, format--like arcs, framing, gradations and area fills-appears to be a more important feature of design in 2D art than it is in 3D art. Table 19 Percentages of each Format Type Found in Each Style  styles % simulation visualization illustration computation image process collage drawing painting  no format % horiz 14 9 0 15 29 (high) 18 0 5 0 16 43 (high) 13 14 >5 0 19  % vertical % >5 18 (high) 22 (high) 10 6 22 6 12  square 0 14 17 (high) 21 (high) 14 14 10 10  Note: I expected to find a 1:1:1:1 ratio of format types for each style.  Proportion. The term "proportion" is used variously in traditional art to describe relative measurements of scale,  132  size or shape in artwork. In comparison to "scale", I have delimited "proportion" in this study to mean the relativity of parts within objects, rather than the way whole objects compare to each other. To clarify earlier implications of findings by style and site, I analyzed proportions for their degree of "realism" (i.e. their imitation of proportions in the natural world) or "imagination" (i.e. mannered, or distorted from what would be expected in the natural world). It should be noted that the concept of proportion was not relevant to 17% of images, such as computations. I was surprised to find that artists consistently treated their images as realistic, or as imaginary. That is, few images combined objects with realistic proportions with objects with imaginary proportions. Half of all artwork was found to use realistic proportions, while one-third used imaginary proportions. This means that the concept of naturalism, at least in terms of proportions, was not found to be relevant in half of all imagery sampled in this study. The greatest number of images using realistic proportions was found in art by males; however, almost equal percentages of male and female artists used realistic proportions (Table 3 ) . Imaginary proportions were more common in 2D images than 3D images (Table 10). The greatest percentage of work with realistic proportions was produced by artists in schools, followed by associations and private sites (Figure 10). It would appear that while realistic proportions were an important 133  characteristic for simulations, which were mainly produced by associations, the overall desire or need to make proportions realistic diminished as artists worked in increasingly less institutionalized settings.  Figure 10. Percentages of artists at each type of site using realistic and imaginary proportions 60  Proportions  school  •  realistic  •  imaginary  private association  Type of W e b site  Scale. The term "scale" is used in this study to describe the relative proportion of one object to another. Like proportion, scale was analyzed as either "realistic" (in terms of scale expected in the natural world) or 134  "imaginary". I found that 46% of all imagery made use of realistic scale and 3 8%, imaginary scale. The concept of scale could not be applied to 16% of the images. Male artists tended to favor realistic scale (Table 3 ) , while female artists were almost equally divided on the percentages of realistic and imaginary scale employed. Imaginary scale was used most frequently in visualizations and drawings, while realistic scale was most likely to be used in simulations and image-processed works (Figure 11).  Figure 11. Percentages of realistic and imaginary scale in each style of art 120 style 100  c 0 i_  •  simulation  •  visualization  •  illustration  •  computation  •  image processing  •  collage drawing  Q)  painting  Q.  realistic Scale  135  imaginary  Works with a realistic scale tended to be horizontal rather than vertical or square (Figure 12). Realistic scale tended to be most important to male artists doing horizontal simulations, while imaginary scale tended to be most important to female artists doing horizontal illustrations. Overall, both scale and proportion as defined in this studyare extremely viable characteristics in this emerging art form, and were applicable to the majority of images.  Figure 12. Percentages of formats with realistic and imaginary scale  Scale c o  I realistic  '_  |imaginary  Q.  Image format  136  Color. In traditional art, color is considered to have significant visual and emotional impact. Color is perceived as a characteristic which contributes to a sense of an image's unity, or, conversely, to a sense of discord or harmony between parts of the image. In art education, color theory is usually formulated by means of the "color wheel", which presents red, blue and yellow as triadic or equidistant hues. Palettes are considered "limited" when monochromatic or analogous hues are used, and "unlimited" when the full spectrum is applied. A great deal of work has been done in the field of art education to analyze the impact of color, and to propose "schemes" or palettes based on the color wheel. All computer artwork was therefore analyzed for the use of a limited or unlimited color palette. I was particularly interested in the findings for this property. Theoretically, a greater potential for color discord exists in computer art since, technically, a minimum of 256 colors are readily available in the computer palette. Despite the potential for easily applying a full spectrum of color, almost half of the sampled images used a limited palette in the sense of a color scheme. Slightly more than half used an unlimited palette and 3% were black and white. (Black and white images which had been significantly re-touched with color were analyzed according to a limited or unlimited color palette). Males and females equally used limited and unlimited palettes, although all of 137  the black and white images were done by females, at schools, in low resolution. Works without format boundaries tended to have unlimited palettes, as did works with vertical and square formats (Figure 13). These findings again indicate a tendency for imagery to have an overall consistency by site in terms of conservative or less conservative approaches.  Figure 13. Percentages of limited and unlimited palettes by format type 120 Image format •  no f o r m a t b o u n d a r i e s  •  horizontal  •  vertical  •  square  '<s> %  Color  138  ®  For example, unlimited palettes were used more by artists in schools--who had previously demonstrated less traditional approaches to 3D formats and 2D styles, even though they tended to use more realistic proportions. And although the percentages of limited and unlimited palettes were similar in 2D imagery and 3D imagery (see Table 9 ) , limited palettes were used slightly more by artists in associations, in horizontal format works with high resolution. Unlimited palettes tended to be used most in low resolution vertical or square format imagery created by artists in schools. These findings suggest either that computer artists in schools are receiving insufficient education in the use of traditional color theory, or they tend not to apply it. However, the imagery produced by artists in schools did not suggest any clear patterns of new approaches to color. Resolution. The term "resolution" is used by computer artists to describe the density of pixels in an image. When an image has a resolution that is low in proportion to the size of the image, pixels of the same hue and intensity block together in groups or "pixellate". This results in a decrease of detail and color nuance in the image as a whole. However, low resolution images also use less space in the computer's working and storage memory. This is an important consideration--particularly in 3D art--if drive space is limited due to hardware constraints, since procedures such  139  as texture-map or atmospherics in 3D art are very memoryintensive. Artists speak of these decisions as "trade-offs". Resolution is also relative to the output medium. I had the advantage of seeing the images on the monitor. Raters were asked to assess the same images output as hardcopies from my printer, which had an even lower dot-per-inch resolution than the pixel-per-inch resolution of the image on the monitor. This confused rater agreement to some extent, as did higher rater standards for "high" resolution. For example, one rater considered resolution to be high if the original image had a resolution of 1500 x 800. However, images on the Web currently tend to be presented in the 800 x 600 range or lower to reduce transmission time--and, as explained, were further reduced in resolution by my printer Taking the constraints of the media into consideration, and assessing all images according to the continuum as it was presented to them in the body of hardcopies--rather by external standards--the raters agreed that fewer than half of the 2D images had high resolution, while 3D works had a much higher percentage of high resolution than would be expected (Table 10). Although only 61% of work by males was done in a high resolution, work done in a high resolution was mainly created by males (79%). As a result, 71% of all 3D works had a high resolution. More than half of all works with a high resolution were produced by artists in associations, followed by artists in private sites, then schools (Figure 14). 140  Figure 14. Percentages of high and low resolution  by type of s i t e 120 100  Resolution  <D  • high  Q)  CM low school  private association  Type of W e b site  Artists in schools may be required to work in lower resolution due to course or hardware constraints, but independent artists theoretically have more flexibility. I looked for trade-offs against the use of other properties and found that, overall, works with a high resolution also showed a slight tendency towards high value or contrast-particularly simulations, visualizations and illustrations. Therefore there did not seem to be significant trade-offs of  141  value for resolution, at least in works using these three styles. Works with high resolution did tend to use limited palettes. Again, the pixels in an image originally created in a less limited palette may have been "dithered" or averaged to reduce transmission time on the Web. Artists may thus trade-off palette for resolution. High resolution was found in significantly more works with a horizontal format than in those where the format was vertical or square (Figure 15). It may be that these artists--who were mainly in associations--either were more concerned about the overall clarity of their images, or their style naturally supported a higher resolution. This implication was supported by findings that the highest resolution works overall were visualizations and illustrations, and artists in associations had created the highest percentages of these styles (Table 12). Artists creating visualizations may have been more concerned about the nuances of their images than the transfer time incurred by viewers on the Web. Illusrations, with their large areas of fills, gradations and other 2D properties, tend to compress well, and therefore automatically have higher Web resolution. Raters perceived the positive attribution of credit for high resolution to be a relative judgment. As one pointed out, many images "suit" low resolution. High resolution is not inherently desirable for all styles of computer art. 142  Figure 15. Percentages of high and low resolution by format type  120 100Resolution <D  • high • low  i  Field of view. The term "field of view", while not common in traditional art, is particularly significant in computer graphics. It is mainly used in connection with 3D graphics and animations to describe the width of a virtual camera's "eye". The format may be similarly perceived as a kind of "container" for structural patterns and imagery, and imagery described as "cropped" by the format (Johnson, p. 147). I therefore analyzed the images for an "uncropped" field of view (that is, without shapes or objects cropped by  143  the format) or a "cropped" field of view (formats in which the content is cut off). I expected to find that computer imagery would have a disproportionately high percentage of cropped fields as a result of artists adjusting their format or cropping sections of images as they worked. I also expected to find some unique applications of field of view in computer art since the potential for experimentation with views and formats is greater. Unlike traditional art media used for painting or drawing, the computer facilitates even radical changes of format or field of view at any stage of the image development. In traditional art, decisions about format and field of view are generally made at the commencement of the creative process, and thereafter altered only with difficulty. Though males and females generally preferred an enclosed or cropped field of view, more than a third of the computer imagery had an uncropped view (Figure 16). This finding may be attributed to the high number of free-form collages which did not have format boundaries, but even then the content was frequently presented in the middle of fragments with open space around them. In imagery with vertical and square formats, the content was similarly placed. These findings supported the concept of "balance" as the "centering" of space and objects (Johnson, p. 137), and "clarity" as a term used to include the idea of "symmetry" in composition (p. 139). 144  Figure 16.  Gender use of field of view 120 100  artist's gender 40  <D 0  •female • male  20  •unknown  Q.  open  cropped  Field of view  In many cases, the choice of format and field of view did not even bear much relationship to the size and shape of the content. The exception was horizontal imagery, twothirds of which used a cropped view. Artists doing simulations, which tended to be horizontal, exhibited a definite preference for a cropped view (Figure 17). Thus there appeared to be a trend in 3D styles towards using format less as a "container" and more as a frame in the sense of a camera's eye.  145  Figure 17. Percentages of cropped and uncropped views by type of style  style •  simulation  •  visualization  •  illustration  • computation • image processing c  • collage  <D 0  drawing  L.  painting  Q.  open  cropped  Field of v i e w  The findings for field of view overall were disappointing. Little novelty has yet to be developed in this area. Artists appeared to be working either in a conservative fashion, creating what are called in traditional art "windows to the world", or they placed their content in ways that showed little evidence of traditional compositional knowledge. At the same time, they did not appear to be experimenting to any significant degree with new applications of field of view.  146  Value. "Value" in traditional and computer art refers to the overall range of light and dark, or contrast. Since evaluating this characteristic was arguably the most subjective, I erred on the conservative side by choosing low value in all borderline decisions. I analyzed value both as it appeared in hard copies of the images and according to the appearance of the images on the monitor screen. The quality of the hard copies, despite being downloaded on the same computer and printed on the same printer with the same type of ink and paper, had the potential to confound the analysis of value much more than any other characteristic, since the light properties of the computer screen result in value treatment that is impossible to download successfully onto hard copy. On the other hand, because the calibration and brightness of different monitors can vary significantly, the hard copies were useful as an estimate of stability overall. I expected to find fewer deliberate differences in the use of value as a formal property between images. However, three-quarters of all images had significantly high value. This again may be attributed to the influence of aesthetics in traditional art forms--especially since the highest percentages of images with high value were found to be illustrations and collages, followed by paintings, and the highest percentage of images with low value were the less traditional visualizations and image-processed works.  147  Works with high resolution overall had a higher value (79%) than did works with low resolution (21%) . Since resolution refers to the number of picture elements rather than to their hue or color saturation, this was not necessarily expected. Works that included the human figure, which could be perceived as one of the most traditional types of content in art, also had the highest value and were done in comparatively high resolution. The use of high value was found to be consistent with distributions of gender and site. It was proportionate to the 69:29 distribution of male and female artists: 70% of all images with high value were produced by males and 28% by females. More 2D work (72%) had high value than did 3D work (28%, Table 16), a finding consistent with the 72:27 ratio of 2D and 3D artwork. Similarly, works with high value were created more by associations than by schools or private sites (43:28:30% respectively), a finding consistent with the 42:29:29 ratio of works found at these sites. In summary, while they are subjective to variations of viewing conditions, the properties of light on the monitor screen may technologically contribute to the appearance of value in computer art. Artists who deliberately applied high value to their imagery, however, often worked in traditional ways. Thus, they may have a greater comprehension of the aesthetic value of contrast as a property, and of how that may be applied to computer art works.  148  Black grounds. Black grounds are a characteristic which appears to be unique to computer art. With the exception of some less common types of traditional art like scratchboard drawing, most traditional art begins with a white ground onto which darker values are laid. In computer imagery, the color of the ground could potentially be any of the 256 or more colors created by light. I was therefore struck by the popularity of black grounds when I began sampling art on the Web. I was interested in determining when and why computer artists would choose this color, although I expected that they would use a black ground consistently for all styles regardless of palette, format or gender. The findings bore this out. Black grounds were used almost equally in works with limited and unlimited palettes, and works with horizontal, vertical or square formats. They were used equally by male and female artists. Black grounds were found in every style, with slightly more emphasis in image-processed works. The latter suggests that photographs with black grounds may be chosen more frequently for imageprocessed work and for collages using photographs. Overall, however, black grounds were used less for 2D art than would be expected and much more for 3D art (Table 10). This tendency may result from a slightly higher use of black grounds by members of associations than would be expected, who are responsible for the greatest percentage of 3D artwork. Black is the "default" background color of many 149  3D programs. In this sense, it was the judgment of the original programmers that black should signify the void, or areas where light is not reflected from objects. White is the default background color of most Paint programs, where it may signify the white "paper" or "canvas". However, the slightly higher emphasis found in imageprocessed work and collages, as well as 3D art generally, leads me to suspect that as the experience of artists with tools unique to computer art increases, so may their interest in manipulating light as a property. Work on black grounds has a startling vivacity, as the pixels of other colors appear to glow more intensely. In drawing and painting, it may be used as the most direct way to create a dramatic effect. The pictorial examples at the bottom of pages 52 and 64 demonstrate the appeal of imagery on a black ground, despite the fact that the pictures are extremely simplistic and of low resolution. Borders and frames. Borders and frames appear to be characteristics unique to computer art. 22% of all imagery used simulated picture frames or borders around the content. This treatment of digital art as a physical object could certainly be perceived as a direct influence of the way traditional art forms such as paintings, drawings, photographs, prints and collages are treated. Not only did artists use virtual picture frames, but the metaphors of "gallery", "salon" or "studio", in which the art was "hung",  150  were used at Web sites. Prints for sale were usually done on "archival quality" paper. A significant amount of art, though, was not presented in any kind of gallery format. It was used as home page illustration; commercial promotion for various techniques; as part of a logo; or covered with type for titles and contact links. Thus, there seemed to be discrepancy among computer artists about the purpose and treatment of art. Proportionately, female artists used slightly more borders or frames than did male artists, and artists in schools and at private sites used more borders than did artists in associations. Again, these findings argue for computer art worlds with differences in aesthetic conventions. Dividers. Dividers such as grids and partitions are often used by printmakers or photographers to emphasize the multiplicity of imagery afforded by printmaking. Computer programs, with their tools for cutting, pasting, copying, revising formats or scaling, would appear to be an even more useful art medium for multiplying and duplicating imagery. All imagery was thus analyzed for the frequency of dividers. Almost equal percentages of grid lines were used in 2D and 3D work (Figure 18), as compared to the 72:27 ratio of 2D and 3D work. However, partitions alone were found almost entirely in 2D work, and all of the works using both grids and partitions were 2D. Surprisingly, 100% of work using both grids and partitions was done by male artists. More than three-quarters of work using only grid lines and more 151  than half of the images with partitions were also done by males. Although female artists overwhelmingly produced the most two-dimensional imagery, the significantly lower use of grids and partitions by females again suggests that--like gradations, area fills or filters--2D computer procedures of a highly technical nature are less likely to be used by females.  Figure 18. Percentages of image dividers in 2D and 3D images  120  Dividers • grid lines • partitions • grid & part • n/a 2D  3D  Type of image  152  Content In general, raters found content easy to identify and raters agreements for content characteristics were high with two exceptions. Analysis was most time-consuming for images with a high degree of content or images with subtle or complex features. For example, rater agreement initially was low for abstract shapes such as polygons or polyhedrals as their use was often subtle. Another category, "interactive", was dropped from further analysis because it was difficult for raters to assess without experiencing the kinds of cursor changes that occur when the mouse moves across a "hot spot" or interactive area. All other findings are discussed here by site and gender. Site. Overall, association members used the widest range of content, and to higher degrees than would be expected (Table 20). This again underscores the interest and potential of association members in finding art-related commercial work. These artists, who are mainly male, showed a particular interest in robots, creatures, technology, vehicles, clocks, weapons, and planets (see also Figure 19).  153  Table 2 0 Percentacres of Each Content Type by Site  % of content schools associations private human f i gu r e 41 (high) 23 36 head only 17 50 (high) 33 other body part 25 38 38 (high) combination 29 29 43 (high) humanoid/robot 20 60 (high) 20 animal 14 (low) 50 36 insect 33 0 67 (high) bird 60 (high) 40 0 creature 20 80- (high) 0 • hardware/software 33 0 67 (high) machinery 40 (high) 20 40 (high) technology 11 56 (high) 33 vehicles 25 75 (high) 0 clocks 0 100 (high) 0 weapons 29 57 (high) 14 technology/vehicles 33 67 (high) 0 vehicles/weapons 0 100 (high) 0 religious buildings 0 50 50 (high) . other buildings 27 53 (high) 20 landscape/cityscape 23 (low) 34 43 water 27 27 47 (high) flora/vegetation 28 36 36 (high) clouds 32 39 (high) 29 planets 22 61 (high) 17 still life objects 30 43 28 text as an aspect 13 67 (high) 20 text as aesthetic 35 42 23 polygons 33 33 35 21 polyhedrals 37 (high) 42  Note: To determine whether the percentages were high or low, they were compared to the 28:42:29 ratio of the three types of sites (p. 100).  154  Artists at schools generally were more consistent in the use of content and the degree to which they used it. They showed slight preferences for imagery of birds and the use of text as an aesthetic element, but there was no discernible pattern to their choices overall. This may •result from the relatively equal numbers of males and females in schools (Figure 2 ) , or more culturally diverse populations at school sites. The profile of artwork found at private sites was quite skewed. While 12 kinds of content were found (more than expected), 6 kinds of content were not found at all. These artists demonstrated a disproportionate preference for imagery of people, nature content such as vegetation or clouds, religious buildings, insects, and content representing computer hardware and software. The use of the latter showed a certain consciousness of their tools in much the same way as representations of artists' studios or materials do in traditional art. Abstract shapes--both polyhedrals and polygons--were also used disproportionately more by artists at private sites. In general, their content seemed to look to nature and spirituality for its inspiration. Gender. The percentages of content by gender (Table 21) were compared to the 69:29 ratio of males to females. Despite this ratio, it can be seen that males used higher percentages of most types of content. For example, although almost one-quarter of all artwork included mechanical 155  imagery, it was mostly produced by males (Figure 19). In particular, male artists created more than 90% of the images of hardware and software, clocks, technology and vehicles, vehicles alone, vehicles and weapons, and religious buildings. They created proportionately higher amounts of machinery, technology alone, planets, nature imagery and polyhedrals. Significantly more imagery of animals was produced by males. Thus male artists demonstrated a bias towards realistic or representational imagery, and themes of industry versus nature.  156  Table 21 Percentages of Each Content Type Used by Males and Females % of content human figure head only other body part combination humanoid/robot animal insect bird creature hardware/software machinery technology vehicles clocks weapons technology and vehicles vehicles and weapons religious buildings other buildings landscape/cityscape water flora/vegetation clouds planets still life objects text as an aspect text as aesthetic polygons polyhedrals  males  70 72 50 71 60 79  30 22 50  (high)  (high) (high) (high) (high) (high) (high) (high) (high)  73 74 ' 87 (high) 76 (high) 86 (high) 83 (high)  72 87  (high)  54 62 89  (high)  30 0 20 22 8 0 29 0 0 0 27 23 13 22 14 17 28 13 46 38  (high)  (high)  21 33 40  71 100 100 100  (high)  29 40  67 60 70 100 80 78 92 100  females  (high) (high)  11  Note: To determine whether the percentages were high or low, they were compared to the 69:29 ratio of male and female artists  (p. 8 3 ) .  157  Figure 19. Percentages of mechanical imagery by gender 120 i • hardware/software • machinery •technology • vehicles •  clocks  •weapons c  Tftech & vehicle  09 O  •vehic & w e a p o n  i_  <D Q.  in/a female  unknown male  artist's gender  Female artists demonstrated a preference for abstract or symbolic imagery in their use of geometric and freeform polygons, text as an aesthetic element, humanoid or spirit figures and isolated body parts, One-third of female mechanical imagery consisted of weapons (Figure 19). The weapons of choice for females were spears and knives (which also are easier to draw symbolically), while the weapons of choice for males were guns. In conclusion, female artists not only tended to use less technical computer procedures, but they also preferred  158  to create less skill-demanding content.- The latter may be a result of their relative inexperience with computer tools other than those for freehand drawing. 2D digital properties Table 22 presents ranked frequencies of ten properties found in the 145 samples of 2D art: freehand segments, scanned photographs, area fills, modeling, framing, iteration, motion, gradations, filters, and arcs and splines. I also calculated the extent to which these properties were used by male and female artists, as compared to the ratio of 2D artworks by male and female artists (Figure 20). Table 22 Percentages of 2D Properties in 2D Imagery (ranked by frequency)  2D property freehand segments scanned photographs area fills modeling framing iteration motion gradations filters arcs and splines  48 42 37 34 33 23 22 21 19 9  159  Figure 20. Percentages of 2D and 3D work by gender  I20r  Type of image  • 2D • 3D female  unknown male  artist's gender  Traditional manual approaches of freehand segments, framing or contour lines and modeling, along with two relatively simple computer procedures for area fills and scanned photographs, were the most common 2D properties in the sample. An analysis of these properties indicated that preferences for their use may be based on limited experience overall. Freehand segments.  2D artwork included more freehand  segments than any other property. 3 6% were found in paintings (see Table 14). This traditional approach involves  160  simply moving a mouse with a color and line width selected. More freehand segments occurred in images with a cropped field of view, suggesting that as computer artists draw, they tend to carry their strokes "off the page" as traditional artists might. Almost half of the female artists used freehand segments. They were disproportionately used by artists at school sites, and in lower resolution than might be expected (62%). Half of all 2D segments-were found in work with framing, and occurred more often (63%) in work with an unlimited palette. Segments were mainly used to draw the human figure and'isolated body parts, animals, and abstract polygons. It appeared that computer artists with less art experience as well as less computer experience relied on simple sketching to create imagery. Framing.  Framing was found in one-third of the 2D  imagery (see Table 22) and used more by males than females (Table 23). The distribution of framing among drawings, paintings and collages was relatively uniform, with slightly less framing used in image-processed work and slightly more in illustrations. Almost half was found in works on a black ground, suggesting that these artists may have felt a need to further define the edges of the lighter shapes against the black.  161  Table 23 Percentages of 2D Properties Used by Males and Females  % of arcs and splines framing iteration motion gradations scanned photographs modeling freehand segments filters area fills  males females 20 80 (high) 73 (high) 27 69 (high) 31 70 (high) 30 72 (high) 28 39 57 61 39 63 37 61 39 40 58(high)  Note: To determine whether the percentages were high or low, they were compared to the 62:37% ratio of 2D work by males and females (Fig. 20).  Modeling. Raters experienced some difficulty with the term "modeling" since it has a different meaning in computer graphics, where it is used to describe the process of creating a 3D solid object. Modeling in traditional art refers to free-hand gradations of color to suggest form and depth. Applying the traditional art meaning, modeling was found in one-third of the 2D images, with a significantly higher percentage in paintings. As would be expected from the definition employed in this study, most images with modeling also had a high value (75%) . Only one-third of work 162  with freehand segments used modeling as well, which suggested that computer artists do differentiate between drawing styles and painting styles. Imagery of the human figure, still life objects, and imagery with realistic proportions did not use much modeling, possibly because they tended to be scanned from photographs and therefore could be considered already modeled. Likewise, modeling was almost non-existent in all types of animal imagery and architecture. These findings again suggested that artists with less experience tend to rely on scanning or simple line drawing. Area fills. "Area fills" are tools found in even the simplest 2D programs. Flat, enclosed areas in Paint and Draw images can be filled with patterns and textures provided by the software simply by selecting them and clicking with the mouse.-Area fills were found in more than one-third of 2D artwork, although, like freehand segments, they tended to be used significantly more by artists at school sites and by female artists generally. Female artists used area fills in more than half of their 2D work (Table 23). They were used in three-quarters of the drawings and almost half of the collages overall (Figure 21), indicating that artists are less likely to see them as a painting technique. This was supported by the finding that half of the area fills were computer-generated gradations.  163  Figure 21. Percentages of area fills found in styles of art  120 style  100.  c  s  •  simulation  •  visualization  •  illustration  •  computation  •  image processing  •  collage drawing painting  Q.  a r e a fill  n/a  A r e a fil  Scanned photographs. After freehand segments, scanned photographs were the most popular digital property. Since the computer facilitates this relatively easy form of imagemaking, I was interested in determining how they were used. Area fills were present in almost one-third of the works with scanned photographs, although the use of modeling in works with scanned photographs was less than 5%. Almost two-thirds of the scanned photographs were simply retouched; more than one-third were simply filtered. They were not even manipulated to any significant degree with other  164  computer procedures such as partitions or grids, or iteration. To my disappointment, images appeared to be treated rather superficially post-scanning. This was also noticed by one of the raters, who was particularly interested in re-touching. As he put it, the majority of scanned images looked simply "captured" rather than "created". Filters. 44% of the filtered images were found in collages and 24% in image-processed work. In addition to drawing freehand and using area fills, scanning and tinting or filtering photographs may be one of the earliest skills mastered in making art on the computer. In this study, almost all filtered work was done at schools and private sites, and a disproportionate amount by female artists. 3D digital properties I analyzed 13 properties of 3D computer artwork (Table 24) and found that the percentages of five properties was higher for males than would be expected from the 89:11 ratio of male and female computer artists. Not only are male artists creating a disproportionate amount of 3D work, but they are using almost half of the properties to a much greater degree. As earlier noted, art by females showed particularly low usage of the more specialized procedures for transparency, atmospherics and reflections.  165  Table 24 Percentages of 3D Properties Used by Males and Females % of vectors 3D objects void image-map ambient lighting illumination specular highlights cast shadows motion continuous shading transparency atmospherics reflections or reflection map  males females 100 (high) 0 89 11 91 9 90 10 100 (high) 0 88 12 79 21 (high) 90 10 92 8 90 10 95 (high) 5 96 (high) 4 94 (high) 6  Note: To determine whether the percentages were high or low, they were compared to the 89:11 ratio of 3D work by males and females (Fig. 20).  I therefore focused first on the application of 3D properties to the three main styles of art created with 3D programs: simulations, visualizations and illustrations. All 3D properties were compared in Table 25 to the 6:15:19 ratio of these styles. I found that simulations disproportionately tended to rely on higher percentages of 3D vectors, objects, continuous shading and cast shadows, and they used ambient lighting more than would be expected. These artists tended  166  to concern themselves with the presentation of simple geometric forms--even wireforms and simple polyhedrals. Illusions in simulations therefore relied on the use of properties most basic to 3D programs. Table 25 Percentages of 3D Properties Found in Each Style  % Of  simulations  vectors 3D objects void image-map ambient lighting illumination specular highlights cast shadows motion continuous shading transparency atmospherics reflection or map  27 20 20 20 27 (high) 19 0 20 15 24 (high) 30 20 24  visualizations  45 35 33 45 27 38 36 40 46 33 50 52 35  (high)  (high)  (high) (high) (high)  illustrations  18 30 31 25 9 33 50 (high) 30 38 31 15 20 35  Note: To determine whether the percentages were high or low, they were compared to the 6:15:19 ratio of all artworks in the styles of simulations, visualizations and illustrations (Fig. 9 ) .  A greater sense of illusion may have been sought by artists creating visualizations, since they applied higher  167  percentages of image-mapping, transparency, atmospherics and reflections. Artists creating visualizations may have been most interested in the surface qualities of the objects and the illusion of depth in the void, while artists creating illustrations appeared more interested in the dramatic use of lighting and the appearance of motion. However, it should also be noted that such "bundled" procedures as transparency or atmospherics are relatively easy to apply, for a quick and dramatic effect. There were fewer discernible patterns of use by site. Compared to the 18:40:19 ratio of 3D artists at school, association, and private sites respectively (Figure 8 ) , association artists used disproportionately high percentages of 3D vectors, 3D objects, and a void. They also used proportionately more motion and atmospherics. As in their use of 2D image-processing techniques, artists in schools used significantly more 3D scanned image-maps than might be expected. Thus the most significant finding was that artists in associations were not only the strongest creators of 3D art but the most active users of the features of 3D programs.  168  Summary This section provides a review of the major findings. They are summarized by gender, site and style. Gender Female artists produced 90% of the 2D graphics and only 10% of the 3D graphics. Half of their images were produced at educational sites. Female-only associations were difficult to access, and female private sites were much lower in number. Female artists tended to rely on such simple and direct tools as freehand drawing, rather than on such computerized techniques as gradations, filters or dividers. The lower percentage of female 3D artists also used much lower percentages of specialized 3D procedures overall. Although their use of most formal properties was generally equal to that of male artists, the content of imagery by females was more abstract and less skilldemanding. Male artists were mainly represented by associations, which were found to have mainly male members. Two-thirds of male artists produced 2D work, which differed from that by females in its higher use of non-intuitive software procedures, resolution, and a representational approach to content. They also used more horizontal formats, realistic scale and proportions, and tended to use freehand drawing rather than scanning. Their 3D work emphasized a much higher interest in all 3D properties--particularly those used for illusion, depth and realistic surface qualities.  169  Site Association artists and those in other large group sites such as galleries appeared particularly concerned about the overall clarity of their work in terms of value and resolution. This was further emphasized by a higher use of black grounds. They appeared interested in portrayals of technique and created the majority of simulations, visualizations and illustrations. Associations actively promoted both their fine art and their art-related services, and tended to build extensive links to other sites. Artists in schools experimented less with such design features as format, palette or borders and more with content. They also produced the most computational art. However, their work overall tended to be representational. Like artists at private sites, they used disproportionate amounts of scanned input. It was difficult to assess theories of art at school sites because they rarely included curatorial or theoretical statements, promoted services or had hyperlinks to other sites. Artists with private sites were unique in that illustrations and statements by this group on private home pages tended to be dramatic and personalized, even though the artists frequently indicated that they represented groups of artists. They had fewer links to other sites than association and gallery artists. They tended to do less 3D work, and their 3D work used even basic 3D features less often than was the case for the other two groups. Their 170  content preferences were people and images of nature. They frequently used text aesthetically, as well as to incorporate metaphors and poetry within the images. Like artists in schools, they demonstrated a disproportionate emphasis on found imagery and image-processing. Style Eight general types of style were identified, and the conventions of each were unexpectedly quantifiable. All formal and digital characteristics were "namable" and terms were very specific. This finding was supported by the high percentages of rater agreements about the kinds of characteristics found in the samples, as well as by their tacit comprehension of the majority of terms and applications. The styles of computer images (other than computational art) were found to depend on program capabilities, which, because they depend on sequence, have particular suitabilities. All choices must be made within the continuous environment of the computer. The work of computer artists appears to be more task-oriented than conceptuallyoriented. In addition to program capabilities, the theoretical bases of 2D and 3D graphics appeared to differ. 2D art tended to use characteristics that increased the flat or graphic nature of the images, such as grids, partitions, borders or collaged photos. Formats, arcs, gradations and area fills were also found to play a more meaningful role in 171  design in 2D than in 3D art; formats acted like a "container" of graphic features. Resolution did not appear as important--or perhaps not as possible, since pixel-based tools and techniques are necessarily cruder. 3D images exhibited a higher use of realistic scale and proportions, and more limited and realistic uses of palette. 3D styles like simulations and visualizations emphasized a view of a naturalistic world. That is, they acted as "windows to the world", where content was often cropped to increase the sense of a view of an environment. Format therefore was in this case used less as a "container" and more as a frame, in the sense of a camera's eye. In terms of the accumulative complexity of tools and parameters, computer art exhibited a developmental trend from manual, intuitive tools like freehand segments to simple computerized procedures like area fills and scanned input; followed by "building" objects in simple environments; and, ultimately, to using 3D tools for lighting, mapping and increasing a sense of atmosphere and density. Likewise, there appeared to be a developmental trend in styles, beginning with freehand drawing, followed by painting, image-processing, collage, illustration, visualization, then simulation.  172  Chapter 5 Recommendations  Introduction The final stage of a content analysis is to interpret the meaning of the results. Traditionally, recommendations in art education are addressed both to practitioners and to researchers working in that area. In this study, the recommendations are addressed to educators interested in developing discipline-based curriculum for digital art. They are organized and discussed in four sections: art production, art history, aesthetics of computer art, and art criticism. Following these sections, recommendations for future discipline-based research in digital art have been developed from the findings of this study. The relevant research questions asked were: 1) What are the conventions of digital artists in terms of tools and techniques? 2) What do the conventions of digital artists, as evidenced in their work, reveal about technological developments in computer art? 3) What do conventions of digital artists as evidenced in their work reveal about their notions of art? 4) What do conventions of digital artists as evidenced in their work reveal about their critical standards? Three general points regarding professionalism, gender, and paradigm should be kept in mind throughout. First, although many advocates of DBAE recommend that curriculum  173  goals in the area of studio production be based on the "professional" practices of "mature" artists, this study did not seek to assess the degrees to which artists responsible for works sampled in this study were "professional", despite the availability of related commercial services at some digital art links, or the size or credits of some organizations. Nor did it seek to establish their "maturity", a concept which links age with a host of valued attributes. The ages and experience of the artists in this sample were unknown, and it is just as likely that the artists who created the sampled imagery were young and inexperienced. In addition, the traditional concepts of "professional" and "maturity" may not be applicable to the current generation of artists working with computers. This study indicates a need for further refinements in such theoretical tenets. It should also be understood that, although the findings in this study were analyzed according to the presumed gender of the artists, there is no such thing as a "universal" male or female--nor a "universal" man,  woman,  boy or girl. Digital artists appeared to be as affected by class, elitism, sexual preferences, and disadvantage as any other artists. This study has revealed a need for more indepth research about the causes of gender differences, as noted throughout. Third, this study does not intend to argue for one paradigm over another (such as modernism over post174  modernism, or vice-versa). It argues for the critical appreciation and analysis of the features of computer artworks and the conditions of their surrounding culture on the Web. The apparent impact of previous analog art forms on computer art, as well as evidence of emerging features, has created a need to critique computer art in poststructuralist terms. In post-structuralist terms, the features of paradigms are treated as artifacts of social consensus at a particular time in history--as they are in this study. However, there is an obvious need for theoretical development in this area, again as indicated on pages 189-195. Production of digital art Art production as a discipline-based component emphasizes the process of creating art. Recommendations for curriculum developers regarding digital art production are based on the evidence of conventions as they were found to be embedded in tools and techniques, and the ways these conventions appeared to define the limits and the potential of digital software for art production. Recommendations have also been based on findings about preferences by sex (termed "gender" in this study) for different tools, techniques and styles; on observations about the critical significance of interfacing technologies; and from observations made during the data collection.  175  Recommendations for art production domain 1. That education in the production of digital art introduce conventions unique to digital media. These include the properties described and pictured in Chapter 3, and which were discussed in detail in Chapter 4.  2. That curriculum developers emphasize knowledge of the generic features of software programs, since turnover in commercial software appears to be rapid and increasinglysophisticated. This study found that tools and materials used by digital artists in 83% of the sample took the form of commercially-developed programs. Images therefore demonstrated the role and limitations of the technologies in an unexpectedly straight-forward way, and made them relatively easy to analyze for digital properties or schemata. Imagery could be recognized as resulting from either a 2D or 3D program, and most program tools could be readily identified.  3. That curriculum should emphasize the generic features of input and output devices. This study did not find that it was possible to distinguish between brands of hardware that may have been used to produce the art. However, screen imagery and output imagery were found to have significant differences in color, value, and the size of images. "What you see" on the monitor 176  screen was not found to be "what you get" in hard copy form, as some Windows products have advertised.  4. That curriculum developers emphasize certain differences between 2D and 3D graphics, and the ways they are like or unlike analog media. For example, computer graphics appeared to be more "task-oriented" than "concept-oriented" (Gbugh, 1989, p. 238), in the same sense that sculpture may involve more tools and armatures than painting. Work in 3D graphics, in particular, can involve simultaneously juggling up to ten parameters--such as lighting, the properties of assigned surfaces, the scale of textures, the positioning of shadows and highlights, viewpoint, and eye level. In this respect, 3D computer art significantly differs from 2D art, and from more intuitive analog media. Although programs and tools have become complex and sophisticated, certain digitized features, or features that result from digital means, were still predominant. While it may be possible to make a watercolor painting look like an oil painting, or an oil painting look like a pastel drawing, it would be much more difficult to make an image appear scanned or iterated if it wasn't.  5. That curriculum include an emphasis on seriation. The process of creation in digital art, unlike that of most traditional forms, can be recorded by saving the image 177  at intervals in separate files. This could afford an ideal opportunity for critical, reflective discussions of studio choices; of how "accidents" might be incorporated into images; or of problem-solving approaches. Discussions could be in classrooms with individuals or groups, or in Web dialogues with students at schools elsewhere.  6. That curriculum introduce ideas of spatial depiction, properties of light, and the symbiotic nature of resolution in digital art production. Ideas about space and spatial relationships in 3D graphics may well be linked to earlier understandings and uses of space in areas like sculpture or perspective drawing. At the same time, they offer new potential in art education. By understanding two-dimensional conventions for representing forms and shapes as an array of polygons or pixel areas on a bitmap, the conventions for portraying forms and shapes in a three-dimensional environment could more easily be recognized. By understanding conventions in both 2D and 3D graphics, this study presumes that other forms of computer art that are based on 2D and 3D graphics can more readily be appreciated, such as the sequences of static frames with altered viewpoints that result in "animations". New parameters for the application of "lights" in digital imagery offer potential for curriculum development in art production. The properties of light in the physical 178  sense are not usually taught in traditional art, since analog forms depend on light reflected from the concrete environment. In the more advanced forms of 2D and 3D graphics, "spotlights" were found to have been used to reorganize the global arrangement of pixel values, or positioned as 3D objects to target specific areas images. Lighting in computer imagery was closely linked to degrees of value or contrast, the hue and saturation of colors, and resolution. Curriculum could be developed to help students understand the symbiotic relationship of parameters that affect resolution. In conversations with raters, it was impossible to talk about resolution without reference to studio conditions for input and output. The subject of resolution in studio production was also found to be integrated with aesthetics, in terms of how it affects appreciation of the image; with criticism, in terms of image use or function; and with computer art history, in terms of how capacities for higher resolution develop according to advances in hardware and software.  7. That curriculum address issues of "style" in digital art. Despite evidencing a much wider range of styles than has been indicated in the literature of computer art education, the sample represented a more limited range of styles than indicated in the literature of computer art itself. 179  The majority of images in the sample imitated traditional art forms, such as painting, drawing, collage or photography. Artists at schools and private sites produced almost no simulations, and, in general, artists at school sites tended to use older conventions of computer graphics. These findings were significant to studio production for two reasons. If the development of skills for the workplace is important, curriculum developers will need to address these discrepancies. On the other hand, as a result of the high degrees of imitation and appropriation, curriculum developers may need to reassess the meaning and the relative importance of the modernist concept of "style" as it would be applied to digital art; that is, what might be implied by "a characteristic handling of content, presentation and schemata" (Feldman, 1982, p. 12). "Characteristic handling" in digital art appeared to result as much from the characteristics of software tools as from'. the ways they might have been handled by the artists.  8. That curriculum address the ultimate aims of computer art education, as it affects gender. As a result of their relative inexperience with 3D programs, female artists may be hampered technically, aesthetically and professionally. At the same time, the tendency among male artists to focus on 3D programs may have blinded them to the advantages of working with 2D programs and such related input devices as digitizing tablets. Male 180  artists may not be encouraged or motivated to work with fewer parameters or in more intuitive ways.  9. That curriculum address discrepancies in gender use of tools, styles and techniques. The findings of this study indicated that there may be inequalities of access or education. Because 90% of female artists in the study produced 2D graphics only, it may be that females are disadvantaged in their studio repertoire. Female artists in general did not use a wide variety of content or the more technical and less intuitive tools, and tended to create less skill-demanding content. They did not appear to participate in as wide a range of styles, to promote their artwork, or to develop links to other sites to the same degree as did male artists. Female artists may receive less education in designing images for the Web and creating sites themselves, or may not have the same motivations to participate as do male artists. Or, it may be that a pluralist perspective, or the "legitimacy" of variety, should be valued over a "unity of logical consistency" (Garber, 1990, p. 22).  10. That curriculum emphasize specifications for the kind of output that is to be used in computer art. Numerous technological problems faced artists using the Web as a transmission vehicle during 1996. These included the degradation of images due to compression; limitations on 181  file formats for Web exhibition; and problems connecting or downloading that resulted from large and complex amounts of material included on home pages and linked pages. This study found that a file intended to be output to the Web should be sampled as a low resolution file, with a transmittable tag. Conversely, this study found that a file intended for output in hard copy should be in as high resolution as possible and ideally created in a palette of "printableonly" colors; that is, colors selected from the palette of computer colors that printer inks can best simulate. Most programs for art and design offer this choice.  11. That curriculum be developed around opportunities to view and exhibit computer art in the Web. Opportunities for critical dialogue might increase student interest and participation in computer art production and its conventions. This study found that many artists felt a strong sense of isolation, as indicated by frequent invitations to "let me know what you think" or "give feedback" regarding the images. Internet resources such as user groups, news groups, and Web sites appeared to be vital to interactivity and the development of conventions.  History of digital art The development of discipline-based computer art history would emphasize the identification of works of art, 182  their stylistic categorizations, and their cultural contexts. The conventions of digital artists, as evidenced in their artwork, clearly revealed developments in technology which could form the beginning of new curriculum development in computer art history. Imagery was found to result from a wide range of input, from original algorithmic programming to such applications and tools in commercial art software as maps or imageprocessing filters. The emergence of conventions in digital art also appeared to result from advances in art-related hardware peripherals, such as digitizing tablets. Digital artists seemed to quickly appropriate technologies as they become available. Recommendations are based on evidence of both older and newer conventions in the artwork, and the discovery of new areas for development in sociological, political and cultural art studies. As further indicated by the emergence of listservs and discussion groups, such as Arts Ed Net and Arts Edge, the social and cultural implications of new technologies are fascinating territories for discussion of historical and cultural dimensions of curriculum. They take on especial significance when we realize that our students are the first generation to witness their development.  Recommendations for art historical domain 1. That curriculum for digital art history be concerned with developments in software tools and in hardware. 183  Just as the introduction of the painter's easel in the 1500s or acrylic paint in the 1950s changed the historical course of painting conventions, so the production of commercial, user-friendly software and a breadth of artrelated peripherals--such as the digitizing tablet for input, or printers that can transfer output files to canvas or watercolor paper--have affected the development of computer styles.  2. That curriculum for digital art history be integrated with art production. Studio assignments might be organized in the form of a historical survey. As the new technologies have emerged and developed, there appears to be less evidence of older ones-at least, on-line. For example, computational art, which formed the bulk of computer art until the 1980s, accounted for only 17% of the artwork in this study. This study also found dramatic differences in image resolution. Eight years ago, 200 x 300 pixel resolution was quoted by Andris (1989, p. 17) as being the state of the art. Although resolution of imagery on the Web during the course of this study was limited by the current Web technologies, some of the images were said to have been created originally in resolution as high as 1200 x 1500 pixels. Students might begin with small programming tasks, followed by 2D Draw and Paint assignments, and move on to work with more recent developments in 3D art, animation, 184  sound and video. The artworks in this study evidenced the use of technologies not available even five years ago; they could be called "cutting edge" art mediums. During the course of this study of digital art, I also encountered 3D animation, virtual reality, user touch-screens, interactive art on the Web, sound and video--to name only a few of the recent advances in technology that contemporary artists are currently appropriating.  3. That curriculum include computer art exhibited on the World Wide Web. This study did not seek to answer the question, "What is the future of computer art?" However, it is clear that an analysis of art on the Web is imperative for establishing future goals for integrating computer art history with art production. It is no longer true that "computer art does not generally have the forum for ideas or the visibility made possible for other media" (Csuri, 1979, p. 423) .  4. That curriculum take a sociological approach to understanding the nature of computer art worlds as they emerge, develop, integrate, or vanish. Concepts of "established roles" (Clark & Zimmerman, 1986) may no longer be relevant. For example, changes in the on-line digital art world appeared to be based in part on the changing activities of numerous participants. These were found to include Web designers, sponsors, site developers, 185  and directory list compilers. At the same time, it appeared that anyone with sufficient interest and some technical knowledge (which, it should be noted, is readily available to the public) might take on any or all of these roles, in addition to the role of the artist as a producer of images.  5. That curriculum address the emergence and evolutionary aspects of Web sites. Projects in art history might examine answers to such questions as, How have sites evolved? What factors--such as access, education or theories of art--have influenced their development? Similarly, why were many sites established by all-male or all-female artists? This study revealed many differences according to sites, but left basic questions about the causes unanswered. For example, the causes of differences in "style" of art by site have not been examined in this study. Sites were found to have divergent theoretical perspectives, which ranged from feminist to commercial. There was a wealth of exemplars on the Net for different trends and styles, but a dearth of material accounting for those differences.  6. That political contexts, rather than geographic or ethnic differences, form a critical area of study in the history of art on the Web. This study found that art on the Web is currently in an a-geographic state. Sites were listed in directories by 186  names, such as "Alicia's Computer Art Gallery" or "Anti-Art Productions", rather than by their country of origin. As a result, this study unintentionally sampled images from many different countries, including Canada, the United States, Australia, France, Japan, England and Norway. However, it did not answer the question of possible differences in art produced by artists in different countries. There were certain obstacles to doing so. The spirit of many Web sites was one of camaraderie. Terms such as "global perspective" and "free market" were often used at annotated sites. At the same time, other images were heavily copyrighted and accompanied with threats. It would seem that accepting digital art as part of a school program carries with it the necessity of're-evaluating traditional notions of ownership, as well as exchanging local and national contexts for something evolutionary and still fluid.  7. That cultural studies include an evaluation of the social and commercial functions of sites. Many sites were found to offer related services for Web design, illustration and animation. As Campbell stated, "The history of the development of computer technology is associated with science, mathematics, bureaucracy [and] warfare" (Campbell, 1991, p. 118). These associations might be further explored in digital art history.  187  8. That curriculum in digital art history attend to both differences and similarities with analog media. The inclusive and collaborative nature of digital art as sampled in this study appears to support a societal model grounded in an appreciation of diversity and "augmented perspectives", rather than in "originality", "authorship", or the "authority" of an artwork (Parker & Pollack, 1981). On the other hand, the imitation of older styles of art like Impressionism may reflect the limits of digital artists' prior art knowledge, rather than signify a conscious "appropriation".  9. That curriculum in digital art history involve a sociological study of both public and private concepts of the artist. This idea has been supported by Feldman (1982), in his discussion of creative people and their reasons for making art. Benedikt (1991), Rheingold (1993), Penley and Ross (1992) and Negroponte (1995) have initiated dialogues concerning the impact of computer and Internet technologies on the artist of the future. In this study, artists with private sites in particular seemed to be determinedly self-directed. Statements on home pages were found to accent "outsider" tendencies described by Becker as characteristic of "mavericks, folk artists or naive artists" rather than what he referred to as "integrated professionals" (Ch. 8, pp. 226-271). 188  10. That curriculum address changing notions of "history". The notion of history as a series of pivotal events, interspersed with pauses for consolidation and development, may have to be exchanged for one in which events accumulate at speed, and their combined weight continually forces the existing structure into new directions.  Aesthetics of digital art Discipline-based art education emphasizes the acquisition and application of specialized art terminology, and includes teaching the visual, cultural and social aesthetics of artwork. The findings of this study imply that "computer art" is many things to many artists. Analysis of the conventions of digital art revealed not only that it takes many stylistic forms, but that it appears to combine both modernist and post-modernist notions. Digital art was presented as a "dialogue" with viewers as frequently as it was presented as a "product". Recommendations for curriculum development in aesthetics were derived from the analysis of specialized terminology; from evidence of paradigmatic features; from findings in this study as compared to certain assumptions in the literature of computer art education (as described in Chapter 2 ) ; and from observations made while exploring digital art sites on the Web. 189  Recommendations for aesthetics domain 1. That curriculum emphasize the acquisition of terminology for "reading" digital art and recognizing schemata. Knowledge of terminology was found to be critical to the identification of conventions in computer graphics--as well as to studio practice and historical advances. Terminology in digital art, as identified in this study, could be categorized into at least three types: terminology specific to software, terminology specific to hardware, and terminology specific to the Web.  2. That post-structuralist curriculum be developed to help students understand both modernist and post-modernist characteristics in digital art. This might involve an analysis of the social constructs of paradigms. Features of modernism found in the images included an emphasis on "pictures" within formats, many of them further emphasized as "objects" with borders or frames. Artwork was presented in virtual "galleries", "salons" or "portfolios", or offered as archival-quality prints. These metaphors indicate an emphasis on notions of "value" and conservation. Some site statements positioned the digital artist in modernist terms as a "creator", as "original", and as involved in "self-expression." Other statements positioned the artist in post-modernist terms as a "user", "collaborator" and "consumer". 190  Modernist features were combined with such post-modern features as a new language based on the digital domain; the decline of sole authorship; the unapologetic use of appropriated images; the accumulation of "little narratives" in collage; the use of semiotics; the chimerical ability of computer graphics to change shape and size--in short, what appeared to be a dynamic flux of appropriation and discard. While digital art still looks back in - time for many of its ideas, it appears that they are becoming increasingly ill-suited to its direction. As Moore put it, "Although it is not one of the avowed goals of DBAE to make our postmodernist world more comprehensible, it nevertheless appears that this could be one of its most important consequences" (Moore, 1991, p. 39).  3. That curriculum be developed to assess the degree to which contemporary digital art continues to be "decidedly machine-like" (Youngblood, 1988, p. 23). Characteristics for the most part were found to be "seamlessly" integrated; that is, all areas of images sampled in this study appeared to be created digitally, with uniform surfaces. Concepts about what constitutes "machinelike" qualities in computer graphics may entail older ideas about "jaggies" (staircase-like edges) and regularized or geometric compositions, but they may also entail notions of art as having irregular, less uniform surfaces.  191  4. That curriculum be developed to assess the degrees to which conventions in digital art are extensions of older art forms or, conversely, represent "new expressive ones" (Freedman, 1991, p.47). In particular, curriculum in aesthetics for digital art might be developed to explore a wider use of different kinds of formats; types of content; work that combines imaginary and realistic proportions; color applications; fields of view; and color grounds other than black  5. That curriculum address aesthetic conventions differentiate 2D and 3D graphics from other art forms, as well as those that appear to be co-dependent on other conventions, such as color palettes and image resolution. Certain properties of digital art in particular can provide a basis for further aesthetic inquiry in other areas of electronic art. These included the use of multiplication, replication, interpolation or seriation (the ability to keep • an image in storage, and, at the same time, transform it).  6. That curriculum incorporate education in the use of art as semiotics. This study found a high use of such graphic devices as borders, frames, grids and partitions as communicative vehicles, where juxtaposition and placement influenced the impact of the message. This indicated a strong trend towards semiotics, rather than to the kinds of criteria 192  traditionally used to differentiate the graphic arts from fine art. At the same time, curriculum might be developed to explore what appeared to be the "design"-oriented applications of 2D programs as opposed to the naturalistic applications of 3D programs. This study has indicated typical ingredients of digital art and noted ,the frequencies with which they can be found. This study has not, however, suggested what digital art is as a combination of these ingredients; that is, it has not fully revealed the ways in which formal and digital characteristics are organized, or how the images might make an impact on viewers.  7. That curriculum for studying or engaging in computer aesthetics acknowledge a new sense of community in the online digital art world. For example, the majority of home pages for samples selected in this study invited all viewers--whoever they might be and whatever "qualifications" they might have--to contact the artists with feedback. This may indicate a developing trend towards a less elitist and more informal conception of an art world. In addition, it provides an unexpected educational venue, since artists can access the kinds of sites which offer the best opportunities for discussion and appreciation of their particular programs and styles. Given the more technically-challenging nature of 3D graphics, a need for "feedback" in the form of critiques or 193  information may be another reason why 3D artists were found to predominate in associations.  8. That new curriculum in aesthetics explore notions of "creativity", "originality", or "self-expression". Such modernist concepts were found to be questionable in terms of their significance as features of highly consumable and disposable artworks that can be easily accessed, copied or downloaded from the Web. Pope characterized art as the interaction of "three essential agents": the artist, the artwork and the viewer (Pope, 1988, p. 326). One of the unique features of the emerging on-line digital art world has been the emphasis on keeping track of "visitors" or "hits". For example, many sites have digital counters on their home pages that keep track of the number of visitors. Two to three thousand visitors per month per site appeared to be common, although larger sites had counts upwards of 20,000. This would appear to indicate a significant increase in typical art "consumption". There is no way to assess when, where or whether digital art on-line is copied or downloaded,  9. That curriculum be developed to explore digital art on the Web in terms of the context of Web architecture. Web architecture positions art works in contexts unlike concrete galleries. For example, digital art was "read" on  194  "pages", and experienced in tree-branching hierarchies subsequent to "home pages". Upon connecting to a site, viewers are first confronted with the site home page. This acts like a book cover to the hyperlinked contents. From aggressive, combative statements to self-effacing soliloquies, from violently-colored bloodand-guts front pages to serene and elegant typesetting, home pages create a strong context for viewer response. The "in your face" quality of many site pages is quite unlike the "arm's length" curatorial essays that accompany gallery exhibits in concrete catalogues. The viewer is immediately placed in an intimate relationship with the artist's own personality.  10. That curriculum developers consider moral values and goals when including material found on the Web. For example, many sites have recently developed "guest books". Unlike the guest books often found in concrete galleries, which tend to be for the artist's benefit, Web guest pages seek to foster dialogues between viewers. New "drawing spaces" on Web guest pages seek interaction with the art itself. Although the engagement is certainly active, it may not always be "sophisticated", as Greer put it (1993, p. 3 ) , nor constitute exemplars as envisioned by DBAE. On one hand, the immediate accessibility of decentrally-controlled discussions on the Web affords a new venue for aesthetic education. On the other, the discussions 195  and content of some sites may not be desirable as "models for outcomes of a visual arts program" (Clark and Zimmerman, 1986, p. 34) .  Criticism of digital art Art criticism traditionally depends on an appeal to established art theories and standards, in order to make a point or an argument. While this study found indications of disparities in theory between sites as evidenced in types and styles of art, the work of digital artists in itself did not reveal the existence of common critical standards. Numerous theoretical questions therefore remain unanswered by this study. However, I do not agree with Mok, who claimed that, "It's too early to judge quality. Criticism will be needed in this field, but so far it is too early to make judgments because that may stifle the kind of exploration that is needed" (1995, p.26.) Criticism in discipline-based art education goes hand-in-hand with studio production and the development of aesthetic criteria, and even the small number of findings in the area of digital art on the Web in 1996 may provide some initial groundwork for curriculum development in this area. Recommendations for curriculum in digital art criticism were derived from findings about discrepancies of theory; as a result of problems encountered during attempts to analyze textual statements; from evidence of technological  196  imperatives in digital art; and from observations on external conditions.  Recommendations for art criticism domain 1. That curriculum development in digital art criticism include an analysis of standards that are used for 2D and 3D art. For example, it might address different forms of evidence of standards; questions of whether or not there should be standards; and, if so, to what ends.  2. That curriculum examine the current modernist orientation of digital art on the Web. The appearance of digital art as static formations should be appraised in terms of the ability of art as twodimensional, formatted "pictures" to break free from modernist orientations. In particular, curriculum could be developed to examine the kinds of things that are signified, and that appear to be valued, by current content (such as figures and landscapes) and conventions (such as scanning and filtering).  3. That curriculum examine the usefulness of traditional terms for differentiating types of digital art. This study found that the terms "commercial", "personal", "fine art" or "illustrations" were commonly used by digital artists at Web sites to differentiate groups of 197  work. Images were subdivided into separate, hyperlinked files which were frequently called "portfolios". Given the ease with which digital images may be transformed, it is questionable whether such distinctions continue to be useful.  4. That curriculum development in criticism consider the degrees to which the materials, process, and display of digital art are linked. In order for critical values to emerge, an integrated approach to computer art criticism necessitates that certain links be examined--such as the links between available tools and production, the development of Web sites and their stated functions, and links between aesthetic elements and social factors.  5. That curriculum evaluate the necessity of "text" to express critical standards. In traditional art, critical standards are often expressed in such text as artists' statements, curatorial essays, or promotional material. It was not possible to analyze text found at Web sites for several reasons. High attrition rate and an uneven distribution of artists' statements at association, school, or private sites made it impossible to conduct a comprehensive sample of textual data. Reasons for the subdivision of some sites into such categories as "fine art" or "commercial work" were not made 198  explicit, other than to indicate the availability of some artists for related commercial work. Male and female digital artists appeared to have different concepts of art--as evidenced in their choices of abstraction or realism, in their styles, or their content--but these were not found to be articulated in statements on-line. Likewise, image selection criteria at most sites were not specified. Nor was there sufficient data to distinguish the particular critical standards of digital artists from other kinds of electronic artists, such as those doing conceptual art, multimedia, or collaborative projects. This study has found that there may be other ways to express standards, such as by the complexity of parameters in computer work, or by the ability of digital images to withstand degradations of output to the Web or in hard copy.  6. That curriculum be reflexive in its evaluation of technological imperatives. Technological imperatives of rapid development and evolution demand constant re-evaluation, not only in terms of their impact on the school system and their ultimate social aims, but in terms of their impact on future computer artists. Imperatives might be evaluated in terms of an apparent drive for greater image complexity, as evidenced in the rapid accumulation of tools and conventions; a drive towards ideals of greater "realism", as evidenced in simulations, 199  visualizations and processed photographs (see Jones, 1989); and the support of some members of the current generation (particularly male artists) for increased technological know-how.  7.. That curriculum address the elitist nature of. computer art. For example, this study found that research on computer art was expensive and time-consuming. The actual production of digital art and the use of the Web as an exhibition vehicle would appear to involve costs beyond the reach of many, in what may be a perpetuation of art as "cash-culture literacy" (Hamblen, 1990).  8. That curriculum in digital art criticism examine other external conditions beyond the work of art, such as the motivations of Web and Web site developers. Research projects might investigate such questions as, Why do individual artists post their work on the Web? Why do schools and galleries create sites? What political and economic factors contribute to the formation of associations?  9. That curriculum address paradoxes between the public nature of the Internet and the often intimate settings and statements created by digital artists at their sites.  200  Art educators have explored similar issues arising from the work of contemporary artists who create sociallycontroversial art in public settings. As educators and students increasingly access the Web for content, such paradoxes will become increasingly important.  10. That curriculum include the an appreciation of the roles and contributions of professional computer art critics and aestheticians. As indicated in Chapter 2, there appears to be an increasing wealth of external resource materials that might be incorporated in curriculum for digital art criticism.  Recommendations for further research As a result of limitations in this study, it has not been possible to say conclusively what the "nature" of computer art is, only what conventions can be found in 2D and 3D digital art. This has primarily been a descriptive study. There are numerous areas which deserve further analysis, and which may more advantageously be approached using different methodologies. This study acknowledges the place and appropriateness of both applied and evaluation research for curriculum development in computer art generally. As a result of this study's commitment to fieldbased research, however, recommendations for further research are proposed here with methodologies oriented to 201  collecting additional data about digital artists and artworks.  1. That demographic research on the digital artists be conducted. This might include such items as age, sex, use of platforms, types of hardware and software, output peripherals, and previous computer and fine art education. Demographic data may reveal developmental trends in the use of tools and techniques, behavioral trends, or other correlations with age and sex. Research could be accomplished relatively quickly by using the Internet as a venue for questionnaires addressed to computer artists as they are identified by Web sites, or as they are found in user groups on the Internet.  2. That reasons for affiliations with particular user groups or institutions be examined. Such research might reveal technological standards and aesthetic standards of sub-groups of digital artists. Again, electronic surveys of artists in user groups could be conducted.  3. That the aesthetic theories of digital artists be analyzed in more depth through qualitative methodologies. Qualitative research might reveal the origins and meanings of preferences as they are based in experiences of 202  individuals. Due to geographic distances, this might be accomplished through interview research, using electronic CHAT rooms.  4. That literature reviews of resource materials be conducted, particularly in the domains of computer art aesthetics and criticism, that would be appropriate for education at different levels of the public school system and higher education.  5. That the prior education (formal or informal) of both digital artists and computer artists in general be assessed, in order to determine the kinds of education that they have found to be the most useful. This may include education in fine art or computer technology; prior experience in related computer fields,exposure to the work of other computer artists; and the use of information obtained from commercial publications such as trade magazines or computer newspapers.  6. That surveys be conducted of the applications of digital art in industry, science or medicine. There appears to be a need for a survey of the stated needs for art by such fields, as well as an investigation into their notions of "art".  203  7. That more in-depth research on types of sites be conducted, as well as rationales for creating sites. Researchers could access additional information using the email addresses posted at sites.  8. That in-depth studies be conducted of gender preferences or differences in digital artwork. While such studies have been undertaken in classrooms, no formal research has focused on members of the computer art world.  9. That reviews be undertaken of the generic features of hardware and software currently available for elementary and secondary school curricula, and the relative costs of each type of program and peripheral.  10. That historical research be conducted on how computer art has been established as a field, from the work of early pioneers coming from the fields of mathematics, science and artificial intelligence, to the relatively de-skilled work of artists currently applying the features of commercial programs and using the Web as a transmission vehicle. In particular, more focused research is needed on the collaborative activities of support personnel.  204  Conclusion Ultimately, educators and curriculum developers must answer the most important question of all: Should digital art curriculum be reactive or proactive? Curriculum developers have traditionally been concerned with values and goals for analog forms of visual art that, while experiencing small incremental changes in tools and materials, have stayed more or less the same for hundreds or thousands of years. They have therefore had the luxury of setting reactive curriculum goals in areas like painting, drawing or printmaking. Previous research studies in computer art education, as discussed in Chapter 2, have been similarly reactive--and consequently ineffective in identifying curriculum areas for development. The results of this study, however, have indicated that the on-line field of digital art is both dynamic and unpredictable. While formalist characteristics and modernist styles persist in digital art, post-modernist characteristics were equally present. Art on the Net was found to be varied in content, style and type; influenced by technologies; a-geographic, decentrally controlled, and randomly accessed by viewers; and to have a high site attrition rate. In these circumstances, only a proactive curriculum is possible. The world of digital artists, as it appeared online, has an unprecedented and rapidly evolving nature. 205  Paradigms for a new curriculum must respond accordingly. Educators will need to take initiative in developing new criteria for aesthetics and criticism, and engaging students in participatory venues emerging in the field. To do so, the application of a discipline-based perspective has never been more important. By focusing on the field of computer art itself, rather than by deriving information from classroom practices, educators will be able to explore computer art as a series of linked disciplines which has changed--and doubtless will continue to change--in ways more rapid than any of its analog precedents.  206  EPILOGUE  It was beyond the scope of this study to formally examine the attrition rate of sites on the Web, as well as the development of new sites. However, the effects of time and history on the digital art world cannot easily be ignored. All artworks sampled in this study were created during or prior to 1996. As I re-visited the sites six months later, it appeared that more than half of the sites no longer existed. Many images had been deleted from their original sites and others had been moved to new sites. Reasons for high attrition rates were not entirely clear, although "expense" was cited in some cases as the reason for closing a Web space. An awareness of "quality", perhaps as a result of artistic development, was presented as one reason for "cleaning up" older sites. Listings under directories also had changed considerably after six months. Typically, there was a stronger emphasis on subdivision into different categories of art, such as "pixel art" or "vector art." For example, the msstate  listing now had only 218 sites, for "art that  involves computers, primarily in the realm of computer graphics". The list included more collaborative, interactive and electronic art, and listed fractal or "purely math-based art" separately.  207  In 1997, a year after the initial data collection, the number of directories had grown from less than ten (including the five used for this study) to 226 directories and 141 search engines. (As earlier noted, a directory is a compilation of site addresses or URLs, and a search engine is a program that searches directories for information on a specified topic.) Alta Vista, Excite, Infoseek and Lycos were now search engines, while Yahoo and Magellan were directories that accessed search tools. This tremendous increase reflected the increase in new sites and links on the Web generally. Returning to Yahoo, I found that "digital art" no longer had its own category as a search term. "Computer generated" was now the umbrella term for "digital art", and included 16,190 site matches. These were subcategorized as: Fractals, Graphics, Institutes, Organizations, 3D Art, Forums, ASCII Art, Artists, Landscapes, Genetic Art and Stereograms. There were also 844 site matches for "digital art" under Arts/Visual Arts/Computer Generated. The apparently disparate nature of these subdivision titles resulted from the emergence of the sites themselves, rather from a "top-down" classification system. At the realworld homebase of the Yahoo directory, the URLs of new sites were collected in 1996 and 1997 both from submissions by site owners and by Yahoo's "spider"--a simple program that scans the Web as a whole, crawling from link to link to identify new sites as they emerge. Human classifiers added 208  new categories as necessary, and, at the same time, reorganized their previous hierarchies of classification. Consistency was addressed with the creation of cross-links, so that sites could be reached via numerous search terms. (See Steinberg (1996) for a history of Web classification and indexing during the 1990s.) As a result of new subdivisions, the category termed "digital art" in 1996 was found on five directories in 1997 to no longer include sites displaying art created in the majority of styles sampled for this study. Yahoo now classified the following styles separately: digital art; fractal art, computational art and stereograms; illustrations; visualizations, simulations and 3D art generally; ASCII or art made from rows of text; genetic or evolutionary art, including all recursive forms; and landscapes created with image-processing and Paint programs. Similarly, "drawing" was no longer listed as "digital art", but subdivided into Digital Typography, Calligraphy, Drawing and Sketching, and Graffiti. In other words, sites showing each of these styles of art have grown too numerous to simply be considered "digital": in 1997, these styles were all subdivisions of the overriding terms "computer art" and "computer-generated art"--depending on the directory. There also appeared to be an interesting change in the kind of metaphors used to name sites in 1997. Whereas in 1996 sites had frequently been named after traditional art 209  venues like "galleries", "showcases" or "exhibitions", computer art sites now adopted metaphors more to do with places or buildings in the real world, such as "House", "Centre", "Palace", "Store", "Dominion" or even "Jungle". Comments from visitors were often placed under "graffiti" (as on buildings) rather than in the former "guest books". Likewise, visitors to sites were encouraged to "enter" or "look around". These metaphors confirmed predictions made by Mitchell (1995) that the two-dimensional, hypertexted ethos of the Internet would change to one of a 3D environment based on concepts of a virtual "world" or "city" as the Web developed. I found fewer private sites maintained by individual artists than I had during the original data collection. While 29% of all digital art sites were found to be private in 1996, the current Yahoo directory for all computer art listings included only 14% in 1997. Commercial sites and organizations, however, were represented to a much higher degree than previously, and offered a different type of service. New commercial computer art services focused on such areas as multi-format on-line editing, CGI scripting, Java aplets, and logo design--rather than animation, illustration or Web design. Such services are basically technological in nature, and indicated a market need for assistance with translation into Web "languages" and with transmission, rather than for additional creative products.  210  From an informal perspective, computer-generated art continued to appear very much rooted in traditional art styles. In 1997, the difference appeared to be in the decrree to which computer-generated art became increasingly more specialized in its emphasis on specific tools and techniques. For example, the ubiquitous collage, with its melange of image-processing, paint and draw features as analyzed in Chapter 4, was rarely found at sites in 1997. Artwork appeared much more limited and fine-tuned in the use of tools; in general, it appeared "slicker", with a high resolution and a higher quality of technical skill. At the same time, computer art in 1997, also appeared more consumable and disposable. As one artist wrote on the homepage for her Web site, she did not intend her computer paintings to resemble analog paintings in terms of scale, materials or technique, but, instead, sought to have her work resemble "the familiar Web reproduction". There also seemed to be a new differentiation in Web "rules" for appropriation. A year previously, as noted in Chapter 5, images were either generally available or completely offlimits. The more recent standard seemed to be that appropriations in the form of hard copies required copyright permission from the artists, while digital copies did not. One may derive from even this limited snapshot of the world of digital art the sense that digital art history proceeds at an accelerated pace. What is here today may literally be gone tomorrow. 211  However, it would also appear that such changes validate the basic findings and recommendations of this study: that computer art has become well-established as a field or discipline; that the Web has grown immensely in importance as a venue for the communication of digital art; that basic conventions of digital art (as identified in Chapter 3) continue to be foundational; that the "styles" of digital art on the Web (as categorized in Chapter 3 and assessed in detail in Chapter 4) continue to be significant-and, in fact, now have their own categories; and that the Web has become an even more collaborative and connected force in the computer art world. Most of all, the unprecedented growth of Web sites, directories, search engines, display technologies and scripting languages reaffirm that art educators will need to be proactive in their response to computer technology.  212  References  Andris, J. (1989). Exquisite tools, exquisite art: Using digitizers and Paint programs. School Arts, 89(4), 17-19. Becker, H. (1984). Art worlds. Los Angeles: University-of California Press. Benedikt, M. (1991). Cyberspace. Cambridge, Mass: MIT Press. Berelson, B. (1952). Content analysis in communication research. Glencoe, IL: Free Press. Bhargava, A. (1991). A microethnography of children's computer-generated graphics. Unpublished doctoral  dissertation, University of Texas at Austin.  Binkley, T. (1990). Digital dilemmas. Leonardo, Computer Art in Context, Supplemental Issue, 13-19. Blandy, D. & Congdon, K. (Eds.). (1987). Art in a democracy. New York: Teachers College Press. Bolin, P. (1995). Investigating artifacts: Material culture studies and art education. NAEA Advisory. Bridwell, G. & McCoy, M. (1991). Dissemination of computer graphics in Indiana: Teachers teaching teachers. Art Education, 44(3), 54-60.  213  Brown, P. (1989). Art and the information revolution. Leonardo, Computer Art in Context, Supplemental Issue, 63-65. Campbell, R. (1991). Implementing computer graphics in the classroom: A descriptive study of art teachers' perceptions. Canadian Review of Art Education, 18(2), 114-125. Chalmers, F. G. (1987). Beyond current conceptions of discipline-based art education. Art Education, 40(5), 58-61. Chang, Y. (1994). Transferring prior art knowledge and skills to computer graphics image-making. Unpublished doctoral dissertation, Ohio State University. Chia, J. & Duthie, B. (1992) . An encounter with computer art: Primary children's reactions. Journal of Art and Design Education, 11(2), 209-222. Chia, J. & Duthie, B. (1993) . Primary children and computer-based art work: Their learning strategies and context. Art Education, 46(6). 23-41. Chumley, S. (1987). A comparison of computer drawings and paper and pencil drawings. Unpublished doctoral  dissertation, University of Houston.  214  Clark, G. A. (1991). Examining discipline-based art education as a curriculum construct.  Los Angeles: Getty-  Center for Education in the Arts. Clark, G.A., Day, M., & Greer, D. (1987). DisciplineBased Art Education: Becoming students of art. Journal of Aesthetic Education, 21(2), 129-193. Clark, G.A. & Zimmerman, E. (1986). A walk in the right direction: A model for visual arts education. Studies in Art Education, 27(3), 34-49. Creswell, J. (1994). Research design. Thousand Oaks, CA: Sage. Crowl, T.K. (1993). Fundamentals of educational research. Dubuque, IA: Csuri, C. (1979). Computer graphics and art. In Booth, K. (Ed.), Computer graphics. New York: IEEE. Czikszentmihalyi, M. and Robertson, M.E. (1990). The art of seeing. Malibu, CA: J.P Getty Trust. Dawson, J. (1994). Art projects: Art Center. MacHome Journal, 2(8), 41-42. DeMaria, J. (1991). A study of the work of Charles Csuri. computer artist and computer art educator. Unpublished doctoral dissertation, New York University. Dobbs, S. (1992). The DBAE handbook: An overview of discipline-based art education. Los Angeles: Getty Center for Education in the Arts.  215  Dunnahoo, D. (1993). Re-thinking creativity: A discipline-based perspective. Art Education, 46(4). 53-60. Efland, A.D. (1990). Curricular fictions and the discipline orientation in art education. Journal of Aesthetic Education, 24(3). 67-81. Edwards, G. (1985). Compu-Art. School Arts. 84 (6) . 30-31. Eisner, E. (1987). The role of Discipline-based Art Education in America's schools. Art Education, 40(5), 6-26, 43-45. Ettinger, L. (1988). Art education and computing: Building a perspective. Studies in Art Education, 30(1), 53-62. Feldman, E.B. (1982). The artist. Englewood Cliffs, NJ: Prentice-Hall. Franke, P. (1996). Art and the computer. ARTNEWS, 95 (3), 96-105. Freedman, K. (1989). Microcomputers and the dynamics of image making and social life in three art classrooms. Journal of Research on Computing in Education, 21(3), 290-298.  216  Freedman, K. (1991). Possibilities of interactive computer graphics for art instruction: A summary of research. Art Education, 44(3), 41-47. Freedman, K. & Relan, A. (1992). Computer graphics, artistic production, and social processes. Studies in Art Education,33(2), 98-109. Friedhoff, R. M. (1989) . Visualization. New York: Abrams. Gall, M.D., Borg, W.R., &Gall, J.P. (1996). Educational research. New York: Longman. Garber, E. (1990) . Implications of feminist art criticism for art education. Studies in Art Education, 32 (1), 17-26. Garfinkel, H. (1967) . Studies in ethnomethodology. Englewood Cliffs, N.J.: Prentice Hall. Goffman, E. (1959). The presentation of self in everyday life. Woodstock, N.Y.: Overlook Press. Gough, N. (1989). From epistemology to ecopolitics: renewing a paradigm for curriculum. Journal of Curriculum Studies, 21 (3), 225-241. Greene, J.C. & Caracelli (1989). Toward a conceptual framework for mixed-method evaluation designs. Educational evaluation and policy analysis, 11(3), 255-275.  217  Green, N. (1990) . Arthur Wesley Dow and his influences. New York: Cornell University. Greer, D. (1984). Developments in Discipline-Based Art Education (DBAE): From art education towards arts education. Studies in Art Education, 34(2), 91-101. Greer, D. (1993). DBAE: Approaching art as a subject of study. Studies in Art Education, 25(4), 212-218. Greh, D. (1986) . Using computers in secondary art education. Art Education, 39(6), 4-9. Hamblen, K. (1987). An examination of DBAE issues. Studies in Art Education, 28(2), 68-78. Hamblen, K. (1990). Beyond the asethetics of cashcultural literacy. Studies in Art Education, 31 (4), 216225. Hammersley, M. (1992) . What' s wrong- with ethnography? New York: Routledge. Haanstra, F. (1994). Effects of art education on visual- spatial abiltity and aesthetic perception: Two meta-analyses. Unpublished dissertation. Rijkuniversiteit Groningen, The Netherlands. Hickman, C. (1991). Why artists should program. Leonardo, 24(1), 49-51. Hicks, J. (1993) . Technology and aesthetic education. Art Education, 46(6), 42-47.  218  Hodder, I. (1994). The interpretation of documents and material culture. In N.K. Denzin & Y.S. Lincoln (Eds.), Handbook of qualitative research, (393-402). Thousand Oaks, CA: Sage. Hubbard, G. (1985) . Computer literacy and the art program. Art Education, 38(2), 15-18. Hubbard, G. and Greh, D. (1991). Integrating computing into art education: A progress report. Art Education, 48(2), 44-51. Johnson, M. (1996) . A comparative study of terms used in art education and computer graphics to describe the visual properties of images. Studies in Art Education, 37 (3) , 184-192. Johnson, M. (1995/96). Portrait of the computer artist: Between worlds. Journal of Social Theory in Art Education, 15/16, 30-45. Johnson, M. (1995). The elements and principles of design: Written in finger jello? Art Education, 48(1), 57-61. Johnson, M. (1993). A comparative study of visual elements in traditional art and computer graphics. Unpublished Master's thesis, University of British Columbia. Jones, B. (1989). Computer imagery: Imitation and representation of realities. Leonardo: Computer Art in Context Supplemental Issue. 31-38.  219  Kerlow, I.v. & Rosebush, J. (1986). Computer graphics for designers and artists. New York: Van Nostrand Reinhold. Kleim, B. (1994). An exploration of integrating the computer into the art room. Unpublished doctoral dissertation, University of Massachusetts. Koroscik, J.S. and Blinn, L. (1983). The effect of verbalization on visual art processing and retention. Studies in Art Education, 25(1), 23-8. Kough, 'B. (1984). Blazing palettes: A pilot computer graphics project. Teachers and Writers Magazine, 15(4), 8-9. Levi, A. & Smith, R. (1991). Art education: A Critical necessity. Volume I, Disciplines in art education: Contexts in understanding. University of Illinois Press. MacGregor, R.N. & Hawke, D. (1982). Ethnographic method and its application to a school art setting. Visual Arts Research, 16, 38-44. McAllister, L. (1990). Pencils or computers as drawing media. Unpublished doctoral dissertation, University of Minnesota.  220  McDevitt-Stredney, M. (1993) . A descriptive case study of students' perceptions in an introductory computer art course. Unpublished doctoral dissertation, Ohio State Univerisity. McMillan, J. H. & Schumacher, S. (1989). Research in education: A conceptual introduction. New York: Harper Collins. Madeja, S. (1993). The age of the electronic image: The effect on art education. Art Education, 46(6), 8-14. Manovich, L. (1994). The engineering of vision and the aesthetics of computer art. Computer Graphics, 2 8 (4), 259-263. Mok, C. (1995). The digital backlash. Graphis 296, 26. Miles, M.B. & Huberman, A.M. (1984). Qualitative data analysis: A sourcebook of new methods. Beverly Hills, CA: Sage. Mitchell, W.J. (1993). Citv of bits. Cambridge, MIT Press. Moore, J. (1991). Post-modernism and DBAE: A contextual analysis. Art Education, 44(6). 34-39. Morbey, M. (1992).  From canvas to computer: Harold  Cohen's artificial intelligence paradigm for art making. Unpublished doctoral dissertation, Ohio University.  221  State  Nadin, M. (1989). Emergent aesthetics: Aesthetic issues in computer arts. Leonardo: Computer Art in Context Supplemental Issue, 43-48. Negroponte, N. (1995). Being digital. New York: Alfred A. Knopf. Nicholson, M. (1993). Lightning Paint, Easy Color Paint, Paint It!, Expert Paint. The Computing Teacher, 20 (7) , 41-46. O'Connell, K. (1985). Microcomputer graphics workshop. School Arts, 84 (6), 25-27. Passmore, K. (1991). The effect of microcomputers on secondary art students' two-dimensional designs. Unpublished doctoral dissertation, Boston University. . Pearse, H. (1992). Beyond paradigms: Art education theory and practice in a post-paradigmatic world. Studies in Art Education, 33(4), 244-252. Penley, C. and Ross, A. (1991). Technoculture. Minneapolis: University of Minnesota Press. Pike, W. (1988) . Students' perceptions of computer use in art education. Unpublished Ph.D. dissertation, University of Alberta. Pogue, L. (1992) . The animation studio. Electronic Learning, 11 (5) , 35. 222  Pope, E.R. (1988). The significance of the computer in art. In F.H. Farley & R.W. Neperud (Eds.), The foundations of aesthetics, art, and art education (pp. 3 21-33 6 ) . New York: Praeger. Rhee, M.C. & Bhavnagri, N. (1991). 4-year old children's peer interactions when playing with a computer. Rheingold, H. (1993). The virtual community. Reading, Mass: Addison-Wesley. Rheingold, H. (1995). Endpaper. Graphis 2 96, 144. Roland, C. (1990). Our love affair with the new technology:. Is the honeymoon over? Art Education, 43(3), 54-60. Stokrocki, M. (1986). Qualitative interpretation of a microcomputer graphics course for gifted and talented adolescents. Art Education, 39(1), 44-47. Strauss, A. & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage. Wang, T. (1992). Present and potential instructional use of computers in art: A Delphi study. Unpublished doctoral dissertation, University of Maryland College Park.  223  Weishampel, C. (1989). A longitudinal study of six preschool children's comprehension of a computerized graphics system used as artistic medium. Unpublished doctoral dissertation, University of Houston. White, D. (1985). Creating microcomputer graphics with the KoalaPad. Art Education, 38(2), 10-14. Wolcott, A. (1996). Is what you see what you get? A postmodern approach to understanding works of art. Studies in Art Education, 37(2), 69-79. Wongse-Sanit, N. (1997). Inquiry-based teaching using the World Wide Web. Art Education, 50(2), 19-24. Youngblood, M. (1988). The ant and the grasshopper: A program for adopting micrcomputer graphics in the arts. Art Education, 41 (3) , 23-24, 41-45.  224  Appendix A. C r e d i t s f o r sample images Characteristic  Sample #  ht.tp:// Site  no format  156  horizontal format vertical format  161 112  square format black ground  135 189  limited palette  139  unlimited palette  113  black and white cropped view  183 150  uncropped view 2D graphics space 3D graphics space realistic scale  165 192 107 106  imaginary scale realistic proportions imaginary proportions high value low value  37 174  www.ylem.org/ylem/artists/rstanley/ practicm.jpg wahoo.netrunner.net/~jones77/3D/fish.jpg www.fractal.com/gallery/1995Expo/ impossib.jpg www.giraffe.com/gr.see/gr_pic_l.gif www.artnet.org/iamfree/vision/parmley/ seeing2.jpg www. newt'ek. com/cool/users/allardice/ cruise.jpg www.fractal.com/gallery/1995Expo/Kool_Kat. jpg www.wam.umd.edu/~concepts/digthree.jpg turing.pacss.binghamton.edu/philbo/www storage/art/planswir.gif www.zapcom.net/~elliot/pics/elliot.gif www.oz.net/1world/images/dsp01.jpg www.sfgate.com/~heavens/lizardli-jpg www.fractal.com/gallery/1995Expo/ cedar_ke.jpg nttad.com/asci/mage/corrugat.gif www.logicnet.com/ted.warnell/ali.col.jpg  180  www.leonardo.net/eztv/cyber/drev.gif  151 138  high resolution  166  low resolution grid lines frame or border partitioned  167 159 159 9  interactive human figure head only other body part combination humanoid/spirit/ or robot animal  6 186 125 38 26 4  insect bird  160 118  www.tmn.com/iop/map.jpg vassun.vassar.edu/~alwei/html/gallery/ mansea.jpg www.tbyte.com/people/stein/work_006/ chairOOl.jpg www.isc.rit.edu/~raa3101/Portfolio/car.gif www.arteque.com/dmg/memory.jpg www.arteque.com/dmg/memory.jpg www.lm.com/users/Kristen/gallery/ LunchtimeHorror.jpg www.imagic/image/illustration/body.jpg www.proport ions.com/rvf_lg.j pg walden.mo.net/~ms/files/lo_panth.jpg nttad.com/asci/image/electeye.gif home.prolog.net/~angersun/imagel3.gif www.best.com/~jantypas/aicap/augustl996/ taos_pra.jpg www.fractal.com/gallery/1995Expo/ reign_in.jpg www.insects.org/graphics/rdd/locusts.jpg www.fractal.com/gallery/1995Expo/peace.jpg  120  225  creature  136  computer  44  machinery-  137  technology vehicle  153 88  weapon clock clouds  14 71 111  flora/vegetation landscape/exterior  131 57  water religious building  193 54  other architecture  143  planet  177  still life objects abstract polygon abstract polyhedral  76 155 157  text as an aspect text as aesthetic simulation visualization illustration  8 35 79 30 110  computation  52  image processing collage  87 59  painting  100  drawing  141  line segments 2D vectors 2D arcs/splines 2D framing mirrored 2D iteration 2D motion conventions  24 169 169 123 23 133 21  www.csusm.edu/public/guests/fm/Art_spri. jpg ourworld.compuserv.com/homepages/Douglas_ Armand/pix2.gif www.synapt ic.ch/mapweb/map/arkham/ deathpar.gif www.sva.edu/salon/gallery/anderson.gif home.sol.no/edlund/portfolio/probably/ l_train.jpg www.easyway.net/~artek/guncon.gif cnj.digex.net www.fractal.com/gallery/1995Expo/nostalgi. jpg www.csusm.eu/public/guests/fm/art_cat.jpg www.bradley.edu/exhibit96/space/untitled. jpg www.galactica.it/on the way/pic/xl.gif www.ozemail.com.au/~dacesag/bigpics/ pretty_church_goin'_do/pretty_c.jpg vassun.vassar.edu/~alwei/html/gallery/ museum.jpg www.nau.edu/801/~cca/bit_by_bit/ wanielis.jpg cnj.digex.net/~rent/dak/lwirc/spamx.jpg www.teleport.com/~reardon/port/techman.gif www.tribal.com/staff/hAND/gifs/ wireball.gif www.imagic/image/illustration/car.jpg nttad.com/asci/image/redshoes.gif cnj.digex.net/~rent/dak/lwirc/pizzpete.jpg www-syntim.inria.fr/syntim/personnes/ shival.jpg www.fractal.com/gallery/1995Expo/jpg/ diety_co.jpg www.vanderbilt.edu/vucc/Misc/Art/weaves/ tapestry.jpg www.best.com/~stevos/cow.gif www.bradley.edu/exhibit96/senses/ moods1.jpg www.cnet.com/Content/Features/Dlife/ Digart/csuri_sp.gif www.magic.ca/magicmedia/hypervision/ hotel.gif www.cyberagenz.com/museum/hist.gif www.sva.edu/salon/gallery/roca.gif www.sva.edu/salon/gallery/roca.gif www.leonardo.net/eztv/cyberbehr/guitar.gif www.cybart.com/~rexbruce/muralsm.gif www.sva.edu/salon/gallery/debrini.gif www.ccc-infonet.edu/~chris/swhitt/ wscyber.jpg  226  2D scanned photos block-pixing  149 67  2D 2D 2D 2D  191 108 181 127  drop shadows gradation modeling filtered  2D area fill photo-retouching  187 48  3D vectors 3D arcs/splines 3D objects  194 194 39  void 3D recursion 3D image map ambient lighting illumination  82 64 5 11 142  specular highlights 3D cast shadows  13 126  3D motion continuous shading transparency-  31 152 32  atmospherics  28  3D reflection texture-map  5 68  www.sonic.net/patald/topsy.gif www.worldserver.pipex.com/nynex/pix/ mart4.gif members.gnn.com/scheers/folio31.jpg www.socool.com/michael/complex.jpg www.printspace.com/billcurr/gozzip.gif www.synaptic.ch/pulsart/phylacter/images/ aubes-jpg www.fractal.si/spiller/compgraf/nagica.jpg www.Kodak.com/digitalimages/sample/images/ jpeg/1295.jpg www.flsig.org/os/artwork/ABST/demo 33.gif www.flsig.org/os/artwork/ABST/demo_3 3.gif www-syntim.inria.fr/syntim/personnes/ mzl.jpg cnj.digex.net/~rent/dak/lwirc/gossx.jpg imalchemy.com/soft vor.gif www.imagic/image/illustration/beer.jpg www.imagic/image/illustration/-tek.jpg www.newtek.com/cool/users/fryeallen/ cathedral.jpg www.imagic/image/illustration/cc-jpg www.newtek.com/cool/users/alainlacroix/ temple.jpg www-syntim.inria.fr/syntim/personnes/ kjetill.jpg www.ghgcorp.com/mzambetti/art/cand02 ..jpg www.syntim.inria.fr/syntim/personnes/ Coreyl.jpg www-syntim.inria.fr/syntim/personnes/ benedet/hiro.jpg www.imagic/image/illustration/beer.jpg cnj.digex.net/~rent/dak/lwirc/ivx.jpg  227  Appendix B Example of directory listing  ambitious bitch A multimedia Dive into Femininity Magical multitudes of femininity: a turn of the millennium update Bad taste & high culture, Humour & science, Art & media. Ecstasy, critique, with BANG! ANIMA: Arts Network for Integrated Media Applications. ANIMA is an Arts Information Service on the Web, including galleries, resource directories, and libraries of information. Anti-Art Productions Anti-Art Productions, an independent artistic endeavor founded by Robert Dada, is an outlet, a forum, seeking to expose the truth which lies underneath the complacent,.mediocre facade of the contemporary human condition. Withtft this home page, you witi find samples of cyber-electronic music, collages (various media), digital paintings/collages and various written works created by me over the past 6 years. I am also currently featuring the works of like minds-who have penetrated my circle. The intent is to entertain, yes; but also to educate, provoke thought and/or incur the scorn of those of the conservative mentality. AnimAIu Productions Images and animations by 3D graphic artist and animator Jeff Alu. Ars Electronica Center - Museum of the Future fifiWJ The Ars Electronica Center is a museum of the future, a "knowledge machine " whose mission is to help-visitors attain that degree of " fitness" in relation to new digital media which is essential in today'sinformation-based society. The Ars Electronica Center is a digital media centre serving its various users as an interdisciplinary interface between technology, culture and society. It provides a dynamic environment in which science, art and the business world can address questions requiring a networked solution. The Ars Electronica Center derives its conceptual and substantive starting-point from the Ars Electronica Festival and the Prix Ars Electronica, which has been ran since 1987jointly with the Upper Austrian Studio of Austrian Radio. Its unusual architecture and ultra-modern technical equipment make the Ars Electronica Center the very model of a "intelligent environment". ART+COM The Artistic Realm This is a newly born site for artists. Right now there's just me. But the monthly rate is not bad, and assistance is available for those who are still learning about setting up web pages. ArtKiln provocative realism reflecting life past present and future. arttexnet.com H8W' An fine arts collective from Houston, Texas, featuring the work of painters, sculptors, writers, photographers, musicans, videographers, mixed-media and digital artists. (English) Art of David Walters / Image Gallery H8W: Paintings and digitally manipulated photographs. Biomorphic abstraction and fantastic art. (English) ArtUniverse ArtUniverse features the artwork of Larry Lovett and other artists living on Oahu's rural North Shore. This is Hawaii's avante-guarde gallery, featuring very affordable originals, prints and collectables. From Waialua to the Smithsonian and back...Lovett is a seasoned painter and computer art pioneer as well. This online gallery is unique! (Art)An Laboratories The inventors of PHSColograms (Virtual Photography), a unique, full color, 3D art medium. There is an extensive gallery of (Art)An's collaboration with artists, scientists and mathemeticians228  Appendix C  Sample of recording taxonomy-  c:\spsssv\tax1.sav  •V-J  .V*  II  2-  color  no format b  limited palet  closed  unlimited ipa  closed  vertical  unlimited pa  closed  imaginary  illustration  square  unlimited pa  closed  realistic  vertical  unlimited pa  closed  type  male  association  two-dimensi  collage  female  private  two -dimefisi  image pn  square  female  association  two-dimensi  drawing  female  association  two-dimensi  style  ~V  4  format  site  gender  ^  it  y t  scale imaginary ^/  ^  realiotio  tf{\*i  imaginary v -  male  association  three-dimen  visualization  male  association  two-dimensi  illustration I  horizontal  limited palet  closed  realistic  male  association  two-dimensi  visualization  horizontal  limited palet  closed  realistic  male  association  three-dimen  simulation  horizontal  unlimited pa  closed  realistic  female  private  two-dimensi  painting  horizontal  unlimited pa  closed  female  private  two-dimensi  illustration  horizontal  limited palet  open  imaginary  male  association  three-dimen  visualization  horizontal  unlimited pa  closed  realistic} r.  female  school  two-dimensi  simulation  no format b  black & whit  closed  realistic  male  association  three-dimen  illustration  vertical  unlimited pa  open  fealistie ,-./•*.  male  private  two-dimensi  collage  l squafe ve-rl tfmited palet  closed  imaginary  male  private  two-dimensi  collage  vertical  limited palet  closed  imaginary  male  private  two-dimensi  painting  horizontal  limited palet  open  realistic  male  private  two-dimensi  image proc  square  unlimited pa  open  realistic  female  private  two-dimensi  image proc  horizontal  limited palet  open  male  private  two-dimensi  collage  vertical  unlimited pa  open  imaginary  male  school  two-dimensi  painting  vertical  unlimited pa  closed  realistic  male  school  three-dimen  illustration  horizontal  unlimited pa  closed  realistic  v' unlimited pa open y  realistic realistic  p.'t i>  y  male  private  three-dimen  simulation  horizontal  male  private  two-dimensi  illustration  horizontal  unlimited pa  closed  female  association  two-dimensi  painting  horizontal  unlimited pa  open  female  association  two-dim e^si  computatio  horizontal y  unlimited\pa  open  male  private  three-dimen  illustration  horizontal  unlimited pa  closed  male  association  three-dimen  simulation'  horizontal V  limited pal eT closed  229  t**''  -if'  \f realistic  1/  realistic  [^  Appendix D Rater sheet for content analysis  INSTRUCTIONS FOR RATERS  First, read through the definitions. If you disagree with any definition or picture example, please make a note with comments. Next, examine one computer image at a time. Check off only one characteristic in each group of boxes. If the characteristic has its own box, check the box if it is present in the image. Leave the box blank if it isn't. If you're not sure, put a question mark. Try to finish checking each image for all characteritics before going on to the next. However, you may go back and change your mind at any time. Example:  (0 ^ <0 0  i  l)  i  0 £  #  •  0  S  >  format  3  1  ~>  cT  v/J  color  black ground  1  A-  i  i  o  y/ l /  230  0 >  0 field of view  v/  v/  y/  -3 s  /  io \2_  >  "  0  4  1  V^  l/  \S  All  images  4 V •^ HI  0  i u  #  format  7-T  <PV  iC / _7  1  +  1  s  •  "  '  -  '  .  V  V'  \,y  y  kZ  V  \f'~  v-"  c-'V-  L-'-"  'y  <y  \-'' '--'"  \S  ,,- i  n  V  y  Y  0  y  y-  • ' • '  0  X  y  ?7  0  s  s/  # -4''  Q)  f i e l d of viet  y • • /  £-_  A—  .  color  black ground  Of  "1  (0  y  7^  3  1-V  y y  \s  y if-  v"  0 ^ •s  ^ 3 c4 fQ space  #  7<  y  y V  V  •  y >/'  y  &  y  -7  y  V  K'  ,y  v  y  y y  /  \y  y I  >/ <_^-  */  6"^  U--'  • , - •  \.--  L- "  V  '-"'  231  n \-  v/  v^  V  oV-  resolution »/  •y"  -  y V  value  J1 J  y  • **  V  ,  Z.i,  /  V W  0"? v-  proportions  y  v  > .  -i'i *•*••*  scale  y  v/  -70 r! £>?  'J: _3  «  i-:' l--"'  Appendix E Sample of SPSS output  gender  site  type  style  format  color  view  scale  1  male  association  two-dimensi  collage  no format b  limited palet  cropped  2  female  private  two-dimensi  image proc  square  unlimited pa  cropped  3 female  association  two-dimensi  drawing  vertical  unlimited pa  cropped  imaginary  4  female  association  two-dimensi  illustration  square  unlimited pa  cropped  realistic  5  male  association  three-dimen  visualization  vertical  unlimited pa  cropped  imaginary  6  male  association  two-dimensi  painting  horizontal  limited palet  cropped  realistic  7  male  association  two-dimensi  visualization  horizontal  limited palet  cropped  realistic  8  male  association  three-dimen  simulation  horizontal  unlimited pa  cropped  realistic  9  female  private  two-dimensi  painting  horizontal  unlimited pa  cropped  10  female  private  two-dimensi  illustration  horizontal  limited palet  open  imaginary  11  male  association  three-dimen  visualization  horizontal  unlimited pa  cropped  imaginary  12  female  school  two-dimensi  simulation  no format b  black & whit  cropped  realistic  13  male  association  three-dimen  illustration  vertical  unlimited pa  open  14  male  private  two-dimensi  collage  vertical  unlimited pa  cropped  imaginary  15  male  private  two-dimensi  collage  vertical  limited palet  cropped  imaginary  16  male  private  two-dimensi  painting  horizontal  limited palet  open  realistic  17  male  private  two-dimensi  image proc  square  unlimited pa  open  realistic  18  female  private  two-dimensi  image proc  horizontal  limited palet  open  19  male  private  two-dimensi  collage  vertical  unlimited pa  open  imaginary  20  male  school  two-dimensi  painting  vertical  unlimited pa  cropped  realistic  21  male  school  three-dimen  illustration  horizontal  unlimited pa  cropped  realistic  22  male  private  three-dimen  visualization  horizontal  unlimited pa  open  realistic  23  male  private  two-dimensi  illustration  horizontal  unlimited pa  cropped  realistic  24  female  association  two-dimensi  painting  horizontal  unlimited pa  open  25  female  association  two-dimensi  computatio  horizontal  unlimited pa  open  26  male  private  three-dimen  illustration  horizontal  unlimited pa  cropped  realistic  27  male  association  three-dimen  simulation  horizontal  limited palet  cropped  realistic  232  imaginary  Appendix F Values and variables  Variable  Value 1  Value 2  Value 3  type format color  2 dimen no format limited palette open  3 dim horizontal unlimited palette closed  vertical black and white  realistic realistic high high grid lines  imaginary imaginary low low partitions  clouds religious buildings as an aspect  n/a other buildings as an aesthetic element polyhedrals  field of view scale proportion value resolution dividers clouds architecture text  abstract shapes border/frame black ground interactive planets flora/ vegetation landscape/ cityscape water still life objects freehand segments 2D arcs/splines 2D framing 2D iteration  polygons border black interactiv planets flora/ vegetation landscape  n/a n/a n/a n/a n/a  water still life objects segments  n/a n/a  2D arcs /splines framing iteration  n/a  n/a  n/a  n/a n/a  233  Value 4 square  -  grid and partitions  n/a  Variable 2D motion blur 2D scanned photos 2D gradation 2D modeling 2D filtered 2D area fill 2D photoretouching 3D vectors 3D solids 3D void 3D imagemapping 3D lighting 3D specular highlights 3D cast shadows 3D motion blur continuous shading transparencyatmospherics reflectionmapping  Value 1 motion blur scanned photos gradation modeling filtered fill retouching  Value 2 n/a  vectors solids void imagemapping ambient highlights  n/a n/a n/a n/a  cast shadows motion blur shading  n/a  transparen atmospher reflection -mapping  n/a n/a n/a  Value 3  n/a n/a n/a n/a n/a n/a  illumination n/a  n/a n/a  234  -  n/a  Value 4  Appendix G Style identification  235  Appendix H Example of home page  Greetings. This site will serve as a gallery for traditional landscape paintings and digitally manipulated photos. Impressionist landscapes will be the main focus here, since understanding Claude Monet's work and learning to see the world as he did are among my main artistic concerns. Also, look for examples of the Photoshop techniques used to prepare these images, as well as links to my favorite art sites featuring impressionism and landscape paintings. (Warning: These screens use tables and have only been tested on Netscape. If your browser doesn't support tables, the layout will be scrambled). Full screen images will be kept small for quicker imaging. Most will be 64 color GIF files, so expect some loss of quality from the originals. If you have any suggestions (or just want to say hello), e-mail me.  Image gallery  In studio landscapes(12)  October Light paintings (6)  Mt. Ranier paintings (6)  Open air Foothills (6)  Open air Bean Hollow (6)  Digital photo landscapes(11)  Recent assorted paintings (8)  236  Open air Palo Alto (7)  Digital photos and illustrations (19)  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items