2.3.2 Units of Meaning – Video Structure Meaning

Movies in a dark theater transport us to realms outside ourselves in a manner rather like the songs
of Homer moved listeners along on a dream journey. Regardless of the mechanism, the appeal lies
in being taken along at the pace and to the place dictated by the text. e technologies of writing
and printing oer scholars an environment in which scholars might review and analyze a verbal
text. Unfortunately, the scholar intent on close scrutiny of a lm text has had few options but to
make the dream journey several times—not infrequently with a ashlight between teeth and a
notebook and pen in hand.
Mechanism has posed a major barrier to user control over locus of insertion, depth of pen-
etration, and time of engagement. For decades, the lm scholar sat in a dark room and watched
the serial unrolling of the text. Sometimes this was in a theater where no control over the text was
available; sometimes this was in a room with a projector that could be stopped, reversed, slowed,
speeded up. In each case, however, the engagement required serial viewing; that is, all frames be-
tween any two points in the text had to be viewed. Even the scholar with access to lm production
equipment or later, videotape could not escape the serial nature of the engagement.
Reviewing Caravagio’s Secrets (Bersani and Dutoit, 1998), a recent work on the baroque
painter Michelangelo Caravaggio, Arthur Danto complained about the teaching of the history of
art, congratulating the authors for focusing our attention on the paintings themselves without lapsing
into formalism (Danto, 1998). It is hard to imagine the same comment being made about lm
studies where both formal analysis and the practice of what Danto calls ekphrasis (from the Greek
for description: collaborative description that is sensitive to how meaning and pictorial detail are
connected) have become a thing of the past. Why? Some reasons are theoretical and do not concern
us here. Others are eminently practical and undermine the lm scholars’ best attempts to come to
terms with the complex operations to produce the lm-eect.
Videodisc technology enabled random access to points within a lm text and a limited form
restructuring the text. Yet two signicant forms of intrusion of mechanism remained. e rst
resembled the long-standing problem of hardware—players need a separate computer (or digital
gymnastics on a multi-button remote) and frequently two monitors. e second is fundamentally
vexing—actually”grabbing sections of a lm text for comparison with others, or for comparison
with words (e.g., a script or a journal article) required photographic copying of individual lm
frames with a still camera or, later, the use of a video printer. Even when this was accomplished one
had only bundles of still pictures.
We can restate our motivation here as an attempt to come to grips with the precarious bal-
ance between stillness and movement that we encounter at every step. Ultimately, it is the transit from
one still image to another which has defeated attempts to freeze the lm-text as it turns into
something else—a still photograph if you stop the machine, or any number of quasi-cinematic
representations if you slow it down or speed it up. Yet it remains, in Bellours term, unattainable
in motion (Figure 2.3).
Figure 2.3: A Bellour text analyzing a segment of e Birds. ere is no access to the motion com-
ponent of the lm, to the sound track, to the color. Frame enlargements are selected and produced
e primary problem is the very nature of the moving image document; that is, the stillness/
motion tango at the heart of the perceptual and technological mechanism. e shortcoming of
most previous tools has been achievement of a small measure of control over the image text at the
expense of the annihilation of the lm eect. e urge to deal with the image/movement issue is
what motivated this project. What the digital environment oers is precisely to have it both ways.
ierry Kuntzels’ close scrutiny of the eect produced by the delement of a series of still images
passing through the gates of a projector, and Raymond Bellour’s analysis of the inection of the
lm-text resulting from the director’s intervention in the dialectics of the look in the diegesis of
the lm, and between the characters and the spectators, delineate a new space of investigation
for lm theory. Both studies deal with what is perhaps the most inaccessible, “unattainable”
among the numerous operations which interact in the production of the cinema-eect, what
Kuntzel also calls, “the lmic”: that delicate balance between stillness and movement, whereby
“the lm-projection is generated by the lm-strip in the denial of this same lm-strip by the
lm-projection, in the rubbing out of the work of signication.” A text which also is unattain-
able, in the sense rst used by Bellour of “introuvable” by being literally and guratively unquot-
able, everlastingly slipping through in the instance of being identied, seized for closer scrutiny .
Augst (1980a)
Over several decades we have used dierent tool environments. We have generated sheets of
paper on which were printed frame blow-ups from segments of lm, as pictured in Figure 1.1. ese
were made by a process of taking the original motion picture lm in hand, then:
eyeballing the lm for an individual frame;
setting the frame on registration pins in a copying device attached to a camera;
exposing the lm in the camera;
locating any subsequent frames and repeating the process;
developing the lm;
generating contact prints—images the size of the copy image frames;
cutting those into individual frame prints;
then gluing them into some order on a sheet of paper; and
duplicating the sheet of paper.
Attach to Copy Camera
Film Gate
Adjusting Screws (3)
Adapter Unit - A
Barrel Unit - B
Focus Screw (1)
Frame Centering Screws (4)
Gate Unit - C
Figure 2.4: Instructions for the Century Duplikin Device for Copying Individual Film Frames.
is was, to say the least, time consuming. In addition, the small size of the images together
with the similarity of any one frame in a sequence to the next led to frequent confusions.
See Figure 2.5 for a model of the lmic representation mechanism. If we were to:
set up a movie camera at a rodeo;
begin recording shortly before the arrival of a horse and rider;
leave the camera running as the horse and rider go by; and
stop recording when the rider has fallen o,
then we would have a representation that would play back the colors and shapes of the horse and
rider in the same sequence and for the same duration as in the original event. Such a representation
depends on sampling a data stream at a given interval and frequency, then playing back the sampled
data at the same interval and frequency. If Homer had had a video camera, there would be little
puzzling over just how the wine dark sea looked or just what a well-greaved Achaean carried for
weaponry (though the issue of Athenas appearance might be more problematic.)
Yet, the Homeric story and the lm text depend on more than mere playback of single event
data streams. Compression, expansion, substitution, and weaving together of dierent story strands
are the stu of story telling. Accomplishing these with an understanding of an audience’s coding
system distinguishes reportage from the engaging tale. e pre-Homeric scholar attempting to do a
structural analysis of a heroic tale would have been in a dicult position. Engagement with the text
would have meant going to a performance, listening for words or metrical patterns that recurred,