41
IV. THE RECORDING OF MAYA SCULPTURE
John A. Graham and Steven R. Fitch
The beginnings of a useful recording of Maya sculpture may be said
to date to the travels of John L. Stephens and Frederick Catherwood in the
early 1840's. A number of earlier explorers undertook the drawing of Maya
monuments, but the scope of their work was never so extensive and their
renderings of sculptures were almost invariably Europeanized, as, for
example, in the unfortunate neo-classic mannerisms of Waldeck in his
drawings of Palenque stucco. But drawing upon the resources of the camera
lucida and their own remarkable talents, Stephens and Catherwood succeeded
in accumulating a record of extraordinary excellence which embraced an
impressive sampling of Maya art from Copan to Chichen Itza. The wide cir-
culation of Catherwood's drawings deriving from the success of Stephens'
lively narratives was a major stimulus in the re-awakening of interest in
the ancient civilizations of the New World and led eventually to the
beginnings of scientific research. Although of outstanding importance for
the age, Catherwood's drawings are by no means adequate for modern studies,
and today they are to be valued mostly for their antiquarian charm and as
works of art in their own right. The often repeated notion that "Cather-
wood's drawings are so accurate in their details that the hieroglyphs can
be read" is true at times although much more accurate glyphic transcriptions
are required for serious epigraphic investigations.
Four decades subsequent to the explorations of Stephens and Catherwood,
and as a direct result of their stimulus, the beginnings of modern scientific
field archaeology in Mesoamerica were initiated by A. P. Maudslay in the Maya
area. The precise and detailed recording of Maya sculpture was at the fore-
front of Maudslay's plan of campaign, and the excellence with which he
realized his goal has as yet not been surpassed in an endeavor of comparable
magnitude. Relying upon photography and casting, already introduced into
Maya field studies by Desire Charnay, Maudslay published in a lavish format
magnificent photographic plates accompanied by line drawings of the sculp-
tures. The drawings were prepared on the basis of photographs, casts, and
42
field notes, and then were often checked against the originals in the field.
It is difficult to be overly generous in the praise of Maudslay's contribu-
tions which laid the foundations for modern studies of Maya art and epigraphy.
It is therefore in no way depreciative of his outstanding accomplishments to
remark that it is surely inevitable in any great project involving such ex-
tensive drawing that despite every precaution some errors occur in Maudslay's
renderings. However, since Maudslay's drawings are ordinarily paired with
excellent photographic illustrations, it is often possible to correct the
errors by reference to the photographic plates.
The investigations of Teobert Maler largely overlap the period of
Maudslay's field studies in the Maya area. Maler was not the scientist that
Maudslay was, but with often meager resources and under frequently most
difficult field conditions, he produced a photographic record of Maya sculp-
ture that is of inestimable value.  His drawings made little pretention to be
beyond the level of field sketches.
Regretably, the superlative standards of Maudslay were not emulated
in subsequent recording of Maya sculpture. The great contributions to Maya
art and epigraphy of Sylvanus G. Morley and Herbert J. Spinden, while adding
enormously to the corpus--above all in the case of Morley, seldom achieved
the photographic level of excellence set by Maudslay and rarely even distantly
approached the level of reliability of his drawings. Fortunately, however,
recent years witness a renewed effort at careful and painstaking photography
and drawing of Maya monuments', as in the work of the University Museum's
Tikal project and the Maya text recording project of Ian Graham under auspices
of Peabody Museum, Harvard.
The accurate drawing of Maya texts and sculptures is an extraordin-
arily difficult task, partly since few pieces exist in their pristine condi-
tion. While various scientific techniques have been developed to further the
accuracy of drawings, such as tracing from enlargements or projections of
various negatives made under varying lighting controls, the intangible
property of artistic insight is required to interpret line in such a manner
as to successfully evoke the axtistic qualities of the original. Thus
43
literal line drawings, even when attaining a high and commendable level of
accuracy and thus being invaluable for iconographic and similar inquiries,
are seldom useful or even usable for purposes of critical or esthetic
studies. The problem of sensitive and effective translation from the three
dimensions of relief or sculpture to the two of drawing are almost insurr-
mountable in simple, direct line drawing, and thus there is the necessity
of primary documentation by photographic and other means.
Another method of recording Maya sculpture is through rubbings. The
first application of this technique in Maya field studies may have been
undertaken by John H. Denison. At the suggestion of A. M. Tozzer, Denison
undertook in 1932 to make rubbings of some of the sculptures at Chichen
Itza and subsequently at sites visited by the Carnegie Institution of
Washington's Campeche expeditions of 1933, 1934, and 1938 (Ruppert and
Denison 1943:1-2, 99). We have not had the opportunity to examine any of
these rubbings at first hand, but to judge from the published reproduc-
tions, there were significant limitations in Denison's techniques that
prevented a fully adequate exploitation of the method. Although various
students have subsequently utilized the technique of rubbing, it is only
in recent years that a significant and large scale effort to record Maya
sculpture with an advanced rubbing process has been initiated by Merle
Greene Robertson (1). Although time consuming and clearly depending in no
small degree upon the artistic sensitivity of Mrs. Robertson, an invaluable
and accurate record of Maya sculpture is now emerging which provides an
amazingly successful evocation of the original qualities of the sculpture.
This record is invaluable to the study of Maya sculpture, and the enthu-
siastic reception of exhibitions of her rubbings at leading museums of the
country testifies to the success of her efforts (2).
While the casting of Maya sculptures is clearly next best to having
the original at hand for study, this method of reproduction is time con-
suming and is usually expensive. The introduction of liquid latex for mold
making was a great advance, but latex has more recently been eclipsed by
the vastly superior silastic materials which can be designed with a remark-
ably extensive range of qualities and potentialities.   Unfortunately,
44
silastic continues to be very expensive although no doubt more economical
techniques of its use can be devised while the cost of the material will
decrease with its wider use in industry. A major remaining problem with
casts and molds is the museum space required for their storage. There exists,
however, still another method of easily recording Maya sculpture in such a
manner that casts can always be made when needed and which has negligible
requirements of space for storage--namely, photogrammetry.
Some years ago the photograxmmetric recording and replication of some
Egyptian sculptures was undertaken in Aswan Dam salvage archaeology, and the
potentialities and capabilities of photogrammetry are being rapidly and
increasingly realized and expanded today in such projects as NASA's program
of lunar mapping and recording. Since photogrammetry is still not very fami-
liar to many archaeologists, we venture here to sketch briefly the method as
it might be applied to the recording of Maya sculpture (3). It is not our
intention to propose or design a detailed program; thus far, we have carried
out minimal field experimentation, and we assume that colleagues considering
employing photogranmetry will carry out their own experiments as to equipment
and procedures most suited to their own project requirements.
Basically, the process of photogrammetry involves simultaneously making
two photographs from two cameras separated by a small distance. From this
stereoscopic pair of photographs all the three dimensional information of the
subject can be obtained. The principle is identical to the way in which the
human eyes operate. In order to successfully record Maya sculpture photogram-
metrically, minimal equipment additional to the usual field photographic
inventory is required, and only two measurements need be made.
Although special and often prohibitively expensive camera apparatuses
are manufactured for making a stereoscopic pair of photographs, the considera-
tions of portability, as well as economy, suggest the use of a single camera
probably to be best suited to Maya field work. In order to employ a single
camera, a specially designed base must be used which consists of a precisely
calibrated bar or tract which is mounted horizontally upon a sturdy, profes-
sional quality tripod. The base can be cheaply machined from a material such
45
as aluminum and should probably calibrated in half centimeters. The camera
is mounted upon the base and is moved along the calibrated track to make the
two pictures. The base serves to keep the camera back in exactly the same
plane while the photographs are taken, and the calibrated track allows a
simple reading of the distance between the two camera positions of the two
photographs.
The base distance separating the positions from which the two photo-
graphs are made is one of the crucial measurements that must be recorded in
the photogrammetric process. The second and final measurement that must be
made in the field at the time of making the photogrammetric record is the
focal length of the camera lens at the moment the photographs are taken.
If the camera is focused upon infinity, the focal length is the given focal
length of the lens. If the lens is focused at less than infinity, a table
can be prepared, or obtained from the manufacturer of the camera, which will
convert the focused distance (say 7 feet or 2 meters) into the corresponding
focal length of the lens (in millimeters).
The camera employed for making the photographs should have a larger
format than 35mm. The 2 1/4" x 2 1/4"11 format may be adequate since Maruyasu
and Oshima (1964:11) report an accuracy of .3mm to 1.5mm attained in the
recording of giant Buddha sculptures using the techniques described here with
their 2 1/4" x 2 1/4" format stereo photography.   If this format is adopted,
then the Hasselblad 500C with an 80mm planar lens would seem to be an excel-
lent choice for the camera.
Unfortunately, however, the square format is not an entirely efficient
utilization of space. Partly for this reason, serious consideration should
be given to employing the 4" x 5" format. The still larger negative area,
of course, means greater detail and resolution and consequently more
accuracy in the final construction of a contour map or replica of the
sculpture. Naturally, to further guarantee high resolution in the negatives,
whatever the size employed, a fine grain film such as Kodak Panatomic-X and
fine grain developer such as Ediwal's FG7 should be used. Similarly, film
should be stored in a can or similar air-tight cannister with a supply of
desiccated silica gel to protect it from the tropical moisture (Life Library
1971:9-42).
Once the pair of photographs have been made and processed, the two
prints are dry mounted onto cardboard separated by a distance somewhat less
than between the human eyes (Manual 1960:106). The stereo pair can now be
viewed through a stereoscope to give a three dimensional image.
While our concern here has been merely to illustrate the ease with
which a photogrammetric record of Maya sculpture can be made at the time of
regular photographic recording, it may be of interest to briefly indicate
the use of the stereoscopic pair for the making of a contour map or a replica
of the sculpture, and for this purpose the reader is referred to the appendix
of this paper.
Since a stereoscopic pair of photographs carries so much more informa-
tion than an ordinary single photograph, it seems urgent to develop techniques
of photogrammetric recording especially in this age in which we witness the
apalling destruction of Maya sculpture on a ghastly scale. Although it may
not be necessary or feasible for archaeological projects or museums at pre-
sent to realize all the potentials inherent in photogrammetry, the essential
record can easily and inexpensively be made and will be available for future
exploitation (4).
47
Appendix
In order to make a contour map of a photogrammetrically recorded
sculpture an autograph machine is employed. With the autograph the con-
tours of the three dimensional image seen while viewing the stereo pair
through the stereoscope are traced.
To understand the basic geometry of the operation and why a stereo-
scopic pair contains three dimensional information, the procedure is
outlined for finding the three coordinates (X, Y, and Z) of a chosen point
on the image.  On the pair of prints the x and L coordinate axis are drawn
and the point A is located on each print.   The x ', y,, x     andy   measure-
a    a     a       a
ments of point A are taken from the stereoscopic pair, the parallax, L, is
computed, and the three formulae listed are used to compute Xa , Ya , and Za
a a a~
In these three formulae B and f were measured when the photograph was taken;
the other measurements were made directly from the two prints. It is the
parallax, or shifting of the position of a on the two negatives, that gives
the three dimensional property to stereoscopic photogrammetry (Moffitt 1967).
Thus the final measurements theoretically required are the coordinates
of a point A on each of the two photographs of a stereo pair. Without the
autograph, it would be necessary to plot these measurements for a very large
number of points in order to construct an accurate contour map. With the
autograph the operator traces contours directly without having to plot each
point and to interpolate between them with connecting lines. Thus the
actual numerical determination by hand of these last measurements is unneces-
sary, and the only measurements actually made are the initial ones of
camera position separation and focal length made at the time the photo-
graphs are taken.
In order to obtain a numerically contoured plaster replica of a
photographed sculpture, a somewhat different machine is employed. The
stereo pair is again viewed through a stereoscope and the various contours
are followed. However, instead of actually drawing a contour map the
48
operator punches a button at regular intervals as he follows the different
contours of the sculpture's image. Depressing the button automatically
results in a computer card recording the X, Y, and Z coordinates for each of
the points selected. Thus, a set of data cards is obtained recording a very
large number of points on the sculpture; the greater the accuracy desired,
the more points are digitalized. The data obtained is transferred to tape
in accordance with an appropriate computer program, and the tape is fed
into a milling machine to carve the replica in any material from plastic to
plaster.
While milling machines exist capable of carving an actual size replica
of a large Maya sculpture (such as are used in the automotive industry),
common university mechanical engineering shops will probably possess the
resources for carving replicas at about 50% actual scale. Although it may
neither be feasible or desirable in the near future to produce any replicas
of Maya sculpture by these means, having the properly made and recorded stereo-
scopic pair of photographs means that all the information necessary to produce
an accurate copy of a Maya sculpture is on record for potential future needs.
And simply having the stereo pair allows the scholar the very substantial
advantage of being able to view the sculpture three dimensionally.
49
Notes
1. The first volume of rubbings resulting from Mrs. Robertson's
work was published by the Museum of Primitive Art (1967) as Ancient Maya
Relief Sculpture. A second and extensive collection of rubbings will be
published shortly. Individual examples of Mrs. Robertson's work have
appeared in numerous articles, including the present volume. The tech-
nique employed by Mrs. Robertson has been described fully (Robertson 1966).
2. The first public exhibition of Mrs. Robertson's rubbings of Maya
monuments was held at the Lowie Museum of Anthropology, University of Cali-
fornia, Berkeley, in 1965. Subsequent exhibitions include the following
institutions: Chicago Natural History Museum, Museum of Primitive Art,
University Museum (University of Pennsylvania), California Palace of the
Legion of Honor, New Orleans Museum of Art.
3. The potentials of photogrammetry as applied to field archaeology
are enormous, and it is surprising that so little use has thus far been made.
4. The Center for Latin American Studies, University of California,
Berkeley, has supported this study.
Where:   1) p (parallax) = Xa - Xa,
2) f (focal length) = 80 mm
3) X and Y are taken from LH negative
4) B = base
Figure 1. The basic geometry involved in photogrammetric recording of a stela.
50
Camera 0
at
A
Xa=X
fA.
X
-  B
a =
y _ B
a - P
z     B
a     P
f
y
51
Bibliography
Life Library of Photography: Special Problems
1971    The Harsh Environment.  Time-Life Books, New York.
Manual of
1960
Maruyasu,
1964
Photographic Interpretation
Fundamentals of Photo Interpretation.  American Society of
Photogrammetry, Washington.
Takakazu and Taichi Oshima
Some Applications of Photogrammetry to Precision Measurement
of Sculptures. Report of the Institute of Industrial Science,
University of Tokyo, Tokyo.
Moffitt, F. H.
1967    Photogrammetry.   International Textbook Co., Scranton.
Robertson,
1966
Merle Greene
Classic Maya Rubbings. Expedition, vol. 9, no. 1. University
Museum, Philadelphia.
Ruppert, Karl and John H. Denison
1943    Archaeological Reconnaissance in Campeche, Quintana Roo, and
Peten. Carnegie Institution of Washington, Washington.