First-Hand:The VanDerBeek-Knowlton Movies


The VanDerBeek-Knowlton Movies[1]

A. Michael Noll[2]

May 5, 2016

© 2016 AMN

This history will be published in Leonardo, the journal of the International Society for the Arts, Sciences and Technology.

Copyright © 2016 A. Michael Noll


During the second half of the 1960s, artist-filmmaker Stan VanDerBeek collaborated with Bell Labs researcher Kenneth Knowlton in the production of ten computer-animated movies. This paper describes that collaboration and discusses some of the movies that resulted. This was a very early example of collaboration between an artist and a computer technologist. In this collaboration, VanDerBeek advanced his experience to learn computer programming, and Knowlton extended his artistic sensitivities and programming languages – each learned from the other. The paper ends with a discussion of the term “computer artist” during those early days of computer art and animation. In my opinion, VanDerBeek by doing his own computer programming became a computer artist, while Knowlton’s creativity in creating computer-animated sequences made him an artist.


Fig. 1. Knowlton’s 1964 computer-animated movie tells how to create an animated movie. One sequence depicts a ball rolling down an incline. (Frame taken from AT&T Tech Channel – Courtesy of AT&T Archives and History Center.)
Fig. 2. “The End” is nicely animated in Knowlton’s 1964 movie, with the words shadowing in an almost 3D effect. (Frame taken from AT&T Tech Channel – Courtesy of AT&T Archives and History Center.)
Fig. 3. The words “Life Like” are shadowed in an almost 3D effect in this frame from PoemField No. 2. (Frame taken from AT&T Tech Channel – Courtesy of AT&T Archives and History Center. With permission of Estate of Stan VanDerBeek.)
Fig. 4. The ending credits to PoemField No. 2 include a nicely animated sequence. The frame here is the credit to “Ken Knowlton.” (Frame taken from AT&T Tech Channel – Courtesy of AT&T Archives and History Center. With permission of Estate of Stan VanDerBeek.)
Fig. 5. Random patterns appear in PoemField No. 2, as words disintegrate into patterns, as in this example. (Frame taken from AT&T Tech Channel – Courtesy of AT&T Archives and History Center. With permission of Estate of Stan VanDerBeek.)
Fig. 6. Frame from PoemField No.5 showing computer-animated image superimposed on shadowy skydiver and globe. (Courtesy Estate of Stan VanDerBeek and Andrea Rosen Gallery. © Copyright VanDerBeek Estate.)
Fig. 7. Frame from PoemField No. 5, showing computer-animated numbers 123456. (Courtesy Estate of Stan VanDerBeek and Andrea Rosen Gallery. © Copyright VanDerBeek Estate.)
Fig. 8. Frame from PoemField No. 7 showing a geometric pattern that expands with its outer edge in color against the different color of the background. Bob Brown and Frank Olivey performed the colorization, under the direction of VanDerBeek, for the PoemField movies. (Courtesy Estate of Stan VanDerBeek and Andrea Rosen Gallery. © Copyright VanDerBeek Estate.)

The mid 1960s was an exciting period in the application of digital computer animation to artistic purposes. The early 1960s were mostly scientific and engineering animation, with computer animation at Bell Telephone Laboratories, Inc. (aka: Bell Labs) in 1962 by Dr. Joseph B. Kruskal, Jr. and in 1963 by Dr. Edward E. Zajac.

In the artistic domain, John Whitney, Sr. was a visiting artist at an IBM research facility in Los Angeles, starting in 1966, where Dr. Jack P. Citron created programs for Whitney to use.[3] I saw Whitney’s 1966 movie per-mu-ta-tion[4] back then, and was impressed by its dancing circular abstract patterns, nicely timed with the music, and with development to create a story. Original black and white and then colorized – two colors. Patterns enter from the sides into what seems to be a circular stage. It all builds to a visual climax, in synchrony with the percussive music.[5] Even today, I am impressed by what Whitney created. [6]

In a same timeframe, but independently, artist-filmmaker Stan VanDerBeek (also appearing as “Vanderbeek”) came to Bell Labs and collaborated with programmer-scientist Kenneth Knowlton on the use of computer animation for artistic purposes. This paper describes that collaboration and some of the artistic movies that resulted. The movies that VanDerBeek and Knowlton created are examples of the state of the art of computer animation in the 1960s. I hope this paper initiates more interest and study of this topic, including the artistic evolution of the series of movies.

Stan VanDerBeek, Kenneth C. Knowlton, and BEFLIX

Stan VanDerBeek was an artist and experimental film animator. He received a Certificate of Art in 1952 from Cooper Union, and worked with modern dance, such as with Merce Cunningham and Yvonne Rainer.[7] He experimented with technology for artistic purposes, such as his Movie Drome inside which he projected images.[8] Knowlton remembers “Stan was impetuously energetic, and technically savvy, from the start. Calling himself a tech-art ‘fruitpicker’ he later went on to visit numerous educational and industrial computer-rich laboratories to pick the ‘low-hanging fruit.’” [9]

Kenneth C. Knowlton came to Bell Labs in 1962 after receiving a Ph. D. from the Massachusetts Institute of Technology. He was assigned to the research area in a department that worked in computing techniques in the computing science research center.

In late 1962, Knowlton proposed a programming language and system for the creation of computer-animated movies for “education in science and technology.”[10] In 1963, he developed the BEFLIX (for Bell flicks) programming language and wrote a Technical Memorandum describing BEFLIX.[11] BEFLIX was written in macro FAP (Fortran Assembly Program) and was used to produce imagery in raster mosaic graphics on the Stromberg-Carlson SC-4020 CRT-based microfilm plotter at Bell Labs.[12] BEFLIX was used initially in programs written in FAP, but was later expanded into new versions for use in Fortran programs. [13]

In early 1964, Knowlton used BEFLIX to program a film about BEFLIX being used for a computer-animated movie about animation, working from a storyboard that he developed showing imagery and text for the story.[14] This 17-minute animated movie required four hours of computer time to create the unique frames, which were then expanded at an optical laboratory to avoid using costly computer time for duplicating identical frames.

In 1966, Knowlton made another animated movie about his L6 programming language. It exhibited beetles, with octagonal bodies and six articulated legs, scurrying about the screen. In 1971-72 Knowlton took a sabbatical from Bell Labs to teach computer graphics and computer art at the University of California at Santa Cruz, using a version of his EXPLOR programming language.[15] Later, Knowlton developed and created computer-assisted mosaic art. [16]

The VanDerBeek-Knowlton Collaboration

Peter G. Neumann was a researcher at Bell Labs who knew Knowlton and who also knew artists in New York City because of his involvement with Charlotte Moorman’s Avant Garde Festival. Neumann introduced VanDerBeek to Knowlton, and thus initiated a collaboration that resulted in about ten computer-animated experimental art movies, starting in 1965 and lasting to the beginning of 1969.

When Stan VanDerBeek first came to Bell Labs, Knowlton programmed computer-animated sequences, which he gave to VanDerBeek, who took the sequences away and edited and colored them to create final experimental art movies. VanDerBeek created the poetic words used in the computer animations and also obtained and added music to the final movies. During this initial period, VanDerBeek did not do programming — although they together discussed the computer imagery that Knowlton programmed. Their mutual appearance in the 1968 documentary "Incredible Machine" shows that they both enjoyed the collaboration and each energized the other.[17] In her 1974 book, McCauley quotes VanDerBeek: “Gradually, with the help of Carol Bosche and Ken Knowlton, I was tutored in programming and began to work in the area of computer animation…“ [18]

After their initial work together, Knowlton created a programming package (called TARPS and written from BEFLIX as macros) so that VanDerBeek could do some programming himself. TARPS (Two-dimensional Alphanumeric Raster Programming System) enabled VanDerBeek to program words to move through various paths and change their sizes, geometries, and motion sequences. They together debugged and tweaked the programs – a true collaboration. Recalling their collaboration, Knowlton wrote in 2010: “Stan’s mental and physical energy raced ahead of both of us. … The Knowlton approach to things was methodical … The VanDerBeek approach was try just about anything to see what happened…” [19]

Knowlton was a strong believer in collaboration between artists and programmers. At that time, he saw each as being separate though, and he believed that new programming tools would result from such collaboration. In 1972, Knowlton wrote:

“I expect, therefore, that art will continue to come from artists or perhaps artists working with programmers – I do not expect much art to come directly from programmers who have devised clever gimmicks for doing things.” [20]

Knowlton gave thought to his collaboration with VanDerBeek and wrote in 1972: “The first few months of interaction with Vanderbeek were mutually frustrating. … After a few months, Vanderbeek became quite proficient with TARPS; in due time he was programming almost completely on his own, while I served essentially as a debugging consultant.”[21] The collaboration that resulted in the VanDerBeek/Knowlton movies clearly was mutually beneficial. By seeing what visual effects VanDerBeek wished to create, Knowlton expanded on the programming packages, creating TARPS and other language packages that followed.

Since Stan VanDerBeek was not an employee of Bell Labs, the legal division drafted a one-page agreement between Knowlton and VanDerBeek, dated June 4, 1968, which VanDerBeek signed on September 18, 1968. The terms of the agreement specified that the movies be co-credited to VanDerBeek and Knowlton, “in approximately the following form: ‘Realized by Stanley Vanderbeek with Kenneth C. Knowlton.’” VanDerBeek was allowed to copyright the films in his name, and Knowlton was licensed by VanDerBeek to exhibit the films, but not “on television, or other place charging admission.” The agreement permitted VanDerBeek to use the facilities of Bell Labs and to receive the original of the films.


VanDerBeek and Knowlton are credited with collaborating on ten movies between 1965 and 1969. A series of eight movies, called PoemFields, were made, employing poetic text in fantastic computer animation, along with other imagery. The FilmMakers’ Coop describes the PoemFields movies:

“All of these films explore variations of poems, computer graphics, and in some cases combine live action images and animation collage; all are geometric and fast moving and in color. There are eight films in the computer animated art series. As samples of the art of the future all the films explore variations of abstract geometric forms and words. In effect these works could be compared to the illuminated manuscripts of an earlier age.”[22]

In his 1971 book, Gene Youngblood described the PoemFields:

“Whereas most other digital computer films are characterized by linear trajectile figures moving dynamically in simulated three-dimensional space, the VanDerBeek-Knowlton Poem Fields are complex, syncretistic two-dimensional tapestries of geometrical configurations in mosaic patterns. … Variations on the mosaic field became more complex with successive experiments, until simulated three-dimensional depth was achieved in the form of infinitely repeated modular units in perspective. … The Poem Fields are filmed in black-and-white, with color added later through a special optical process that permits color gradations and increments almost as complex as the forms themselves.”[23]

The following is a list of the VanDerBeek-Knowlton movies, based on material at the AT&T Tech Channel, the Filmmakers’ Coop, the Estate of Stan VanDerBeek, and a biography of VanDerBeek:[24]

  • PoemField No. 1 (1965 or 1967)
  • PoemField No. 2 (1966) (with a jazz soundtrack by Paul Motian)
  • PoemField No. 3 (1967)
  • PoemField No. 4 (no date)
  • PoemField No. 5 (1968) (computer-generated soundtrack)
  • PoemField No. 6 (no date)
  • PoemField No. 7 (1967-68) (with a soundtrack by John Cage)
  • PoemField No. 8 (no date)
  • Collide-Oscope (1966) (VanDerBeek, Knowlton, and Bosche) [25]
  • Man and His World, 1967 (shown at Expo ’67)

However, Knowlton does not recall the movie Collide-Oscope, and it is thus possible that VanDerBeek worked solely with Bosche on this movie. Knowlton recalls that VanDerBeek was working on PoemField No. 9 and No. 10. [26]

The VanDerBeek-Knowlton movie PoemField No. 2 (1966) is posted at the AT&T Tech Channel.[27] There is much use of a flashing effect, as images and sequences flash on the screen. There are colors, added optically by filmmakers Robert Brown and Frank Olvey, in collaboration with VanDerBeek and under his instructions.[28] According to Chelsea Spengemann, “This type of collaboration persists through VanDerBeek’s work. He definitely depended and thrived on it. He did not shy away from it.” [29]

The words LIFE LIKE appear, superimposed upon each other, but shifted, creating an almost three-dimensional effect. There are also the various words that morph into patterns, including a computer-animated THE END at the very end, along with computer-animated credits to VanDerBeek and Knowlton. Patterns and words shift and move across the screen. Multiple colors are used in many of the sequences. Zabet Patterson describes PoemField No. 2 in some detail in her recent book. [30]

The VanDerBeek Estate allowed me to view privately many of the PoemField movies, other than PoemField No. 2, which is posted at the ATT Tech Channel. My descriptions of PoemField No. 5 and No. 7 follow, but to be fully appreciated the movies really need to be seen.

PoemField No. 5 (1968) animates the words FREE FALL and superimposes shadowy human shapes skydiving in space and spheres upon computer-animated sequences, all in color (mostly strong reds and blues), with computer-generated ascending and descending tones. The visual effect is almost three-dimensional and is quite exciting artistically, in my opinion, developing to visual climaxes. There seems to be less flashing than in PoemField No. 2 and more expanding textures and patterns of words and shapes. PoemField No.5 seems more sophisticated than PoemField No. 2. THE END is computer animated and extends as if the movie does not want to end.

PoemField No. 7 (1967-68) opens with an image that looks like a large plus sign and the words LOVES OR. Patterns expand, with one eating away and dissolving into another, sometimes with different colors for the expanding edges. The patterns are quite geometric, in contrast to the more fluid imagery in PoemField no. 5. Multiple colors of red, blue, and green appear, one color morphing into another. A nice after-image effect occurs at one segment.

ChromaDepth® glasses (manufactured by American Paper Optics) shift colors spatially, thereby creating a 3D stereoscopic effect.[31] These glasses utilize flat film prisms that translate colors thereby giving the stereo parallax shifts that the brain transforms into the perception of stereoscopic depth. Richard A. Steenblik invented and patented the idea for this stereoscopic viewing in the early 1990s. [32]

It is fascinating to look at PoemField No.5 with the ChromaDepth glasses, as the colors in the movie are transformed into three-dimensional depth, sometimes consisting of more than one plane. This becomes an analysis technique to help understand how VanDerBeek created the movie by superimposing together separate portions, each on top of the other. It is not known whether VanDerBeek was interested in actually creating 3D when he made the movies.

All the PoemField movies have an impressive flow and development, and seem to tell a story from beginning to end, sometimes reaching a climax in the middle section. Perhaps the use of words helped to focus the flow and development, but so too did VanDerBeek’s prior experience as a filmmaker.

I compared the animation programmed by VanDerBeek with what Knowlton had earlier created to see if what VanDerBeek did was different from what Knowlton had earlier programmed. The words “THE END” evolve at the very end of Knowlton’s 1964 computer-animated movie into an almost three-dimensional shadow form. This type of animation is then expanded and developed in the animation later done for the VanDerBeek-Knowlton movies. I also notice a progression of computer-animation style from Knowlton’s 1964 computer-animated film to the animations initially done for the VanDerBeek-Knowlton movies. There is also a similarity that might be because all the animation utilized raster graphics and animated textures. The later PoemFields seem to have a more sophisticated artistic computer animation, probably because VanDerBeek by then was doing his own programming using TARPS.

The computer animations made before 1968 were programmed on an IBM 7094 computer and plotted on a Stromberg-Carlson SC-4020 microfilm plotter. Computer animations after 1968 were made on a Stromberg-DatagraphiX SD-4360 plotter, which had replaced the 4020 at Bell Labs. [33]

What Defined a Computer Artist in the 1960s?

What made someone a computer artist in the 1960s? Clearly the artist who learned and knew how to program and used a digital computer to create visual imagery was a “computer artist.” There was discussion about whether “patterns” created with a computer were art, and also whether the person programming the patterns was an artist. These definitions were further complicated in the case of collaboration between an artist and a programmer where the artist used material made by the programmer. Today, artists use computers as a medium and do not, themselves, necessarily program.

If someone photographs an oil painting and then manipulates the image digitally, that person does not become an oil painter. Hence if an artist simply used material made by a programmer, the artist was not a computer artist – but the programmer was an artist in creating the visual imagery.

Knowlton was reluctant to consider his computer animation to be art and felt he therefore had to collaborate with an artist. But in my opinion, without Knowlton, VanDerBeek would not have had any access to the computer at Bell Labs, or any computer animation to edit and color. VanDerBeek needed Knowlton; but Knowlton needed VanDerBeek for his editing and coloring and turning the computer-animated sequences into final artistic movies. By seeing what visual effects VanDerBeek wished to create, Knowlton expanded on the programming packages, creating TARPS and other language packages that followed. This clearly was a true collaboration and mutually beneficial.

In the world of art and sculpture, the master sometimes instructs apprentices on what to create – and the master takes all the credit. In other cases, an artist might instruct a technician on what to do. This opens the question of whether VanDerBeek was instructing Knowlton on what to program – VanDerBeek was the master and Knowlton just a simple technician programmer. But it seems clear that what started as collaboration developed over time until VanDerBeek became his own programmer with direct control over what he wanted for the final animations. Furthermore, in the initial stages of the collaboration, Knowlton showed creativity in his computer-animated sequences that he programmed and created – he was far more than just a technician.

In the end, VanDerBeek and Knowlton both were artists, collaborating and working together. They learned from each other. Knowlton expanded his artistic judgment; VanDerBeek expanded on his editing and also learned how to program. Their relationship was truly collaborative, with credit to both.

Discussion and Opinion

I believe that there is artistic creativity and innovation in the computer animation that Knowlton himself programmed, starting with his early 1964 movie and the sequences he later initially programmed for VanDerBeek. There is a computer animation style that emerges that is Knowlton’s as a computer artist, in addition to being a computer technologist. This does not diminish the artistic contributions of editing and coloring that VanDerBeek made to the final movies in their initial collaborations.

Later, VanDerBeek’s computer animations advanced, as he too became a computer animator, and computer artist as defined in the 1960s. The editing and coloring performed by artist-filmmaker VanDerBeek, who was something of a technologist, also developed in sophistication over time. The distinction between artist and technologist clearly was blurred and blended for VanDerBeek and Knowlton.

The PoemField movies created a half-century ago are, in my opinion, amazing, even by today’s standards.


  1. This paper is based on discussions with Knowlton, and has been reviewed by Knowlton for accuracy and completeness. Chelsea Spengemann of the VanDerBeek Estate contributed comments. However, the final authorship and responsibility for the paper is Noll’s alone.
  2. A. Michael Noll is Professor Emeritus of Communications at the Annenberg School for Communication and Journalism at the University of Southern California.
  3. Moritz, William, “Digital Harmony: The Life of John Whitney, Computer Animation Pioneer,” posted at:
  4. John Whitney’s film per-mu-ta-tion is posted on YouTube at:
  5. John Whitney’s notes about per-mu-ta-tion are at:
  6. As a published classical music critic, and former computer animator myself in the mid 1960s, I feel a need to express my opinion on the artistic merits of these early movies, and hope that this does not disturb those used to a more objective approach in an academic journal paper.
  7. Wikipedia entry for Stan VanDerBeek and VanDerBeek website:
  8. Claus, Jürgen, “Stan VanDerBeek: An Early Space Art Pioneer,” Leonardo, Vol. 36, No. 3 (2003), p. 229.
  9. E-mail from Knowlton, dated April 26, 2016.
  10. Knowlton, K. C., “A Proposal for Computer-Produced Animated Movies,” Bell Telephone Laboratories Memorandum for File, November 5, 1962.
  11. "BEFLIX, A Programming Language for Producing Animated Diagram Movies," TM63-1271-6 (1963), Bell Telephone Laboratories; and "BEFLIX-II, An Improved Programming Language for the Production of Animated Movies," TM63-1271-9 (1963), Bell Telephone Laboratories.
  12. Knowlton, Kenneth, “The Use of Computers in the Visual Arts,” Proceedings of the Seventh National Sculpture Conference 1972, Elden C. Tefft (Editor), University of Kansas, 1973, pp. 105-112.
  13. Knowlton, Kenneth C. and Lorinda L. Cherry, "FORTRAN IV BEFLIX", Proceedings of the Eighth Annual Meeting of the Users of Automatic Information Display Equipment (UAIDE), November 1969, pp. 411-431
  14. “Animated Movies by Computer,” Science News Letter, Vol. 85 (May 2, 1964), p. 287.
  15. Knowlton, Kenneth C., "EXPLOR — A Generator of Images from EXplicit Patterns, Local Operations & Randomness," Proc. 9th Annual UAIDE Mtg., pp. 544-583 (1970) and Knowlton, Kenneth C., “MINI-EXPLOR: a FORTRAN-coded version of the EXPLOR language for mini (and larger) computers,” ACM SIGGRAPH Computer Graphics, Vol. 9, No. 3 (Fall 1975), pp. 31-42.
  16. See Knowlton’s website at:
  17. See AT&T Tech Channel:
  18. McCauley, Carole Spearin, Computers and Creativity, Praeger Publishers (New York), 1974, p. 67.
  19. Private Essay by Knowlton, “Recollections of Collaborations with Stan VanDerBeek.”
  20. Knowlton, Ken, “Collaborations With Artists – A Programmer’s Reflections,” GRAPHIC LANGUAGES, F. Nake and A. Rosenfeld (Editors), North Holland Pub. Co. (Amsterdam), 1972, pp. 399.
  21. Knowlton, Ken, “Collaborations With Artists – A Programmer’s Reflections,” GRAPHIC LANGUAGES, F. Nake and A. Rosenfeld (Editors), North Holland Pub. Co. (Amsterdam), 1972, pp. 399-418
  22. See:
  23. Youngblood, Gene, Expanded Cinema, Clarke, Irwin & Company Limited (Toronto and Vancouver), 1971, pp. 246 & 249.
  24. The VanDerBeek biography is at:
  25. The AT&T Tech Channel credits this movie to VanDerBeek, Knowlton, and Bosche. However, Knowlton does not recall it (February 2016 interview).
  26. Telephone conversation with Knowlton on April 28, 2016.
  27. See AT&T Tech Channel:
  28. See Andrea Rosen Gallery:
  29. E-mail dated May 5, 2016 from Chelsea Spengemenan (who manages the VanDerBeek estate) to Noll.
  30. Patterson, Zabet, Peripheral Vision: Bell Labs, the S-C 4020, and the Origins of Computer Art, The MIT Press (Cambridge, MA), 2015, pp. 70-74.
  31. See: and
  32. US patent 5002364 filed January 22, 1990, granted March 26, 1991 to Richard A. Steenblik.
  33. Interview of Jerry White (who was the SC-4020 operator at Bell Labs) conducted July 25, 2015 in Stirling, NJ.