GEOGRAPHY OF SITE AREAS
The general area with which we are concerned here is the Olmec
"heartland" which lies on the southern Gulf Coast of Mexico    extending
roughly from Laguna de Alvarado on the west to the lower reaches of the
Grijalva River on the east (map 1). Discussions of the geography, geol-
ogy, natural areas, etc., of this region have been presented by Tamayo
(1949), West (1964a, 1964b) and Vivo (1964).
Within the above mentioned limits and for a distance of 60 to 100
kilometers inland, the area may be described as a flat, low-lying coastal
plain made up of alluvial deposits overlying older sedimentary formations.
The north central portion of the plain is dominated by the Tuxtla Mountains,
a volcanic formation which rises gently from the surrounding lowlands and
drops off sharply to the Gulf shore along its northern edge. The most prom-
inent peaks, San Martin Pajapan and San Martin Tuxtla, rise to elevations of
more than 1600 meters, and are part of a chain of recent cones forming the
northern rim of the highlands. To the south is a series of smaller, more
heavily eroded peaks of Plio-Pleistocene age, including Cerro El Vigia and
Cerro Cintepec (Williams and Heizer 1965, map 2), which were important
sources of stone for the Olmecs.   Several rivers transect the area, among
them being the Papaloapan, Coatzacoalcos, Tonala, and Grijalva.
The climate of the region may be described as tropical, Af to Am in
terms of the Koeppen system (Vivo 1964:212-213). Rainfall is heavy, aver-
aging in excess of 2000 mm annually for the area as a whole.    Although
precipitation occurs in measurable amounts throughout the year, it is
heaviest during the period from September to November. Daytime temperatures
during this time average 25-300 C. In the dry season, from January to May,
the temperatures are higher, though they rarely exceed 350 C. Dense tropi-
cal vegetation, broken occasionally by savannah grassland, covers most of
the higher ground, with scrub forest, mangrove swamp, or open marsh the
dominant ground cover in seasonally or perpetually inundated lowland areas.
From the time of the Spanish Conquest to the beginning of the twenti-
eth century the Olmec area was relatively unknown and rarely visted by out-
siders.  The difficult terrain, the oppressive climate, and the often hostile
attitude of local inhabitants were major factors in this situation. Contacts
with civilization were limited to areas on or near the trans-Isthmian rail
line and to activities related to the growing of coffee, the major export
crop.
- 1 -
2
Even as late as 1925, when Blom and La Farge visited the area, it
could still be described as extremely isolated. These authors make partic-
ular reference to the lack of travelers, and describe the region as being
torn by all manner of civil and military strife-a haven for bandits, rebels,
and political revolutionaries (Blom and La Farge 1926:16).   Foster, who
spent some time in the Tuxtlas fifteen years later, remarks that there had
been a measure of population growth and a return to somewhat more peaceful
conditions as the outside world began to have greater influence in the area.
Paved roads connected a few of the more important centers although the horse
still provided the major means of cross-country transportation. Spanish was
more widely spoken than it had been previously, and money was coming into
more general use (Foster 1942:4-15).
More recently the region has undergone a marked change. A network
of paved and graded roads has been built, local industry has developed, and
the population has expanded rapidly, opening previously uncultivated areas
for farming. Minatitlhan and Coatzacoalcos, which Blom and La Farge des-
cribed as small and run down, have grown enormously. Towering black clouds
billow up from the refineries in these cities where a growing petroleum
industry dominates the economy. Even the agriculturally-oriented Tuxtla
highlands have changed. The village of San Andres, where Blom and La Farge
could find only one hotel-which shared quarters with a cigar factory-now
has- three modern hotels, two banks, and a busy bus depot. However, lest
the reader think that civilization has completely engulfed the indigenous
culture, we note that bare breasted Indian women still wash clothes along
the banks of the Coatzacoalcos, almost within sight of Minatitlan's smoke
stacks, just as they did in the days when Cortez first saw the region.
Within the area described, we are concerned with those Olmec ceremo-
nial centers which have produced colossal heads: Tres Zapotes, San Lorenzo,
and La Venta. The westernmost of these is Tres Zapotes, lying near the foot
of the Tuxtla Mountains, on the eastern edge of the San Juan River flood
plain. The area of the site is characterized by Stirling (1943:8) as a low,
upland region consisting of a series of small, abrupt ridges and plateaus.
Clear, swift-running streams flow through the deep gullies and narrow
valleys that separate the ridges. The highly fertile volcanic soils of the
region presently support a sizable farming population, and the large number
of mound groups indicates that a similar situation probably obtained in the
past. Dr. Drucker recently made inquiries and learned that the population
of Tres Zapotes, which numbered 300-400 in 1940, had swelled to 1600 in 1966.
Extensive archaeological investigations were carried out at Tres
Zapotes just prior to World War II, under the direction of Dr. Matthew W.
Stirling. Results of these studies have been reported by Stirling (1943),
Drucker (1943), and Weiant (1943). The site consists of a large number of
3
/WS i TP L P - r
*,,raw
/~~~~~~p 0*.' *
E049/P VA
P     iucvr 001                   A
/           Lt                          n
Map 2
Tres Zapotes, showing location of mound groups
(after Stirling 1943, fig. 2).
mounds located on and about a low ridge which runs SW-NE for a distance of
over 3 km along the west side of the Arroyo do Hueyapan, just opposite the
modern village of Tres Zapotes (map 2). Many of these mounds are clustered
together in irregular assemblages that Stirling (1943:10) referred to as
"groups."  There are six or more of these groups, none of which is arranged,
so far as is known, according to a precise geometrical plan, nor has there
been any apparent attempt at orientation.
The earth mounds are of moderate size, the highest measuring about
12 m. Of the long low mounds, the largest and most impressive is 130 m long,
17.5 m wide, and 7.18 m high (ibid., 11).   The nearby Nestepe group, a sep-
arate locality to which we will refer later, lies just across the arroyo from
the Tres Zapotes site, at a point roughly opposite Mound Group 3.
South and east of Tres Zapotes and the Tuxtla region the land falls
away gradually to the low, hot and humid plain that forms the northern ter-
minus of the Isthmus of Tehuantepec. This area has been described by Coe
(1965a:681) as an "extensive flatland of clayey alluvium, mangrove forests
and swamps."  Six of the colossal heads now known were discovered at the site
..Ju
GooL
,.  .a
-4
0
N
-41j
0
a
'i-
o d
IC
9%.
of
Cs
4i
0
()
co
F4
4-J
Cd
0
cn
s-a
0
a4
0
4-4
co
p
0
al)
0
4-i
CU
0
03
4-i
0)
CU
I"
%4
0
t
0 'd'u
c-   o  .   0
4i E
C     0 X4
X ,1 @
0 4    co*
Ov o 4- * f
Cl)  r4  s-
s4i Q4      00
4  O
,n ,.: z  *a
Ucfl   r-
-    0 4
* d  4-  i-  CU
Cl) .4 .-4 -A   4Ja
CU *CU4-i 0
oo
,~CO)CO0 4
CO o    0 4i
-i--   CO) > 5
44 -4-I     0
0     '-l CU 0
4-4  (0)  -
U 0 4-   0
CO) Oc CUC
u1) 04 X
4-A 4-i- ~  0
0        kes-
s-i 9  -   0
co       a)<  i
C04- 4 44
ON 4-i -
CrY0%'  0
H-1 W Cd 3
J
0
II1
K
5
of San Lorenzo, which lies well within this lowland region.   This is one
of a group of three sites located near the village of Tenochtitlan on the
RIo Chiquito branch of the Coatzacoalcos River, about 80 km inland from
the coast (Stirling 1955).
The population of the area around Tenochtitlan and San Lorenzo has
increased sharply in the past twenty-five years, although occupation is
primarily restricted to the natural levees bordering the river channels
and to the slopes of nearby ridges. Tropical rain forest is the dominant
ground cover, except where it has been burned off in the process of milpa
clearing.
Preliminary archaeological investigations were carried out at San
Lorenzo by Stirling and Drucker in 1946, and were reported in part by
Stirling (1955). More recently M. D. Coe has begun an extensive program
of re-excavation at the site and to date two preliminary reports have been
published on this work (Coe 1966, 1967).
As described by Stirling (1955:9), the site of San Lorenzo lies on
the remains of a "flat topped mesa of gravelly soil, cut through here and
there by steep ravines," about 4 km west of the river.   It consists of a
large complex of long mounds and smaller platform mounds, enclosing several
plazas.  Most of these features, as Stirling indicates, are rather unim-
pressive; for example, the principal mound is only 8 m in height. In
addition to the colossal heads, numerous other examples of Olmec monumental
sculpture have been recovered from the vicinity of San Lorenzo, including
several altars, seated figures, and a basalt sphere.
La Venta, the third Olmec site with which we are concerned, is
located about 85 km northeast of San Lorenzo on a low island which rises
above the back-waters of the Tonala River, 19 km inland from the coast.
Although the site lies in a vast mangrove swamp, it has recently been the
focus of much attention due to the discovery of petroleum in the vicinity
and the subsequent establishment of a base camp for-drilling operations,
which has since been enlarged into a petrochemical plant. As a result, the
site has been largely destroyed.
Archaeological investigations at La Venta were carried out by
Stirling, Drucker, and Wedel during the years 1941-43, and reports on this
work were published by Stirling (1943) and Drucker (1952).   In 1955, a
Smithsonian Institution- University of California expedition undertook an
extensive excavation program, the results of which were reported by Drucker,
Heizer and Squier (1959).
Although there are several clusters of mounds on La Venta island, the
6
most important is that which Drucker (1952:8) calls the Central Group
(map 3). It is dominated by Complex C, a large pyramid rising 33 m above
the surrounding land. Just to the north of the pyramid lies Complex A, a
symmetrical group of low mounds and platforms laid out along the same
centerline as that of the pyramid. The Central Group is oriented 8 degrees
west of north.
Map 3 was taken from the original made at La Venta in 1955, and shows
accurately (as the maps of the La Venta site in Kubler 1962, fig. 16; and
Covarrubias 1946, fig. opp. p. 93, do not) the original positions of the
four colossal heads found at La Venta.   In addition, the A-2 mound at the
north end of Complex A has been shown here as a stepped platform mound,
based upon new information resulting from the investigations at La Venta
in July, 1967, by R. F. Heizer and P. Drucker.   Further, the base plan of
the pyramid and the form of the basal platform of the pyramid (referred to
in Drucker, Heizer and Squier 1959 as Complex C) shown here are the true
ones and are markedly different from those depicted in Drucker, Heizer and
Squier (1959  frontispiece, figs. 4, 5).   The errors of our surveyor in
1955 are responsible for the earlier published shape and ground plan of
the pyramid, and an account (to be published) correcting these has been
written (see Addendum for citation to Heizer and Drucker manuscript).
7
HISTORY OF RESEARCH ON THE COLOSSAL HEADS'
Since Jose Melgar first visited what was then the Hacienda de
Iueyapan in 1862 to view the colossal head which we refer to here as TZ 1,
and returned to Jalapa to write "Notable Escultura Antigua' in 1869, a siz-
able number of reports have been published on the subject of the colossal
heads. This literature is discussed here in its relation to the following
subjects: discovery and description of the heads; racial affiliation of
the heads; sources of stone for the heads; removal of the heads to their
present locations; the heads in relation to the Olmecs; and bibliographies
concerned with the heads.
Discovery and Description
As stated above, the first colossal head to be discovered and des-
cribed was TZ 1.  In his original report, Melgar (1869) recounts the dis-
covery of the head by a peasant clearing a milpa, and presents a long
argument claiming that the Ethiopian features of the head prove there were
prehistoric Negroes on the Gulf coast of Mexico. TZ 1 was briefly described
by the German archaeologists Eduard Seler and C. Seler-Sachs (1922) and the
American amateur Weyerstall (1932) before it was first fully described by
Stirling (1943).
The first note on a La Venta head is that of Blom and La Farge
(1926:85) for LV 1 which they discovered while spending a day at the site
during the Middle American Research Institute's first expedition to Meso-
america in 1925.
However, it was not until the several joint Smithsonian Institution-
National Geographic Society expeditions to Tres Zapotes and La Venta that the
heads at these sites were adequately exposed, studied, and described.   This
work culminated in Stirling's Stone Monuments of Southern Mexico (1943). In
this report, Stirling was able to give measurements, site locations, and
stylistic features for TZ 1, and LV 1, LV 2, LV 3, and LV 4. He noted (ibid.
16, 17, 56-60) the flat backs that characterize the heads, relates TZ 1 to
LV 1 and LV 4, and states that TZ 1 depicts not a Negro but a "broad-nosed,
short-faced" physiognomy that is an early type "found over a considerable
area in Middle America."
Each of the twelve colossal heads is referred to hereafter by the
initials of the site and number of the head. For details on nomenclature
see table on p. 16.
8
It was on the 1945-46 Smithsonian Institution- National Geographic
Society expedition to Mexico that the first five heads at the site of San
Lorenzo were excavated and studied. Although the expedition's work has
never been fully published, the monuments have been described by Stirling.
He gives preliminary measurements for each of the heads, describes them,
and compares the ten stone heads known at that time.   He feels that "stylis-
tically the heads are so similar, and the parallels in detail are such as
to force the conclusion that no great time interval could have elapsed during
which they were made" (Stirling 1955:9-13, 21).
It was not until 1965 that descriptions and photographs of the
eleventh and twelfth heads (NS 1 and SL 6) were published. Although NS 1
had been exhibited on one edge of the principal plaza of Santiago Tuxtla
(Veracruz) since 1951 and had often been seen and photographed, it had never
been described.  Therefore Heizer, Smith, and Williams provided a useful
service when, rather than announcing a new discovery in their note on NS 1,
they described the head, identified its stone source, noted its close resem-
blance to TZ 1, and remarked that all eleven of the heads then known were
"obviously part of the same sculptural complex" (Heizer, Smith and Williams
1965:103). Although they labeled this head Tres Zapotes No. 2, they noted
that this designation would probably have to be changed since it came from
a site two miles north of the spot where TZ 1 was discovered.
The first reference to SL 6 is by Aveleyra Arroyo de Anda (1965).
He briefly describes the head and its provenience, and notes its specific
similarity to the other San Lorenzo heads and its general similarity to all
the other Olmec heads. Michael Coe, the actual discoverer of the head.,
added little to Aveleyra' s remarks concerning SL 6 when he published his
preliminary report on his excavations at San Lorenzo (Coe 1966).
Opinions on Racial Affiliations of Colossal Heads
The question of whether or not the Olmec heads represent Negroes and
thus can be taken as evidence of the existence of a prehistoric black race
in North America has interested commentators ever since Melgar first saw
TZ 1 in 1862. Melgar devotes well over half of his original publication to
"proof" of prehistoric North American Ethiopians. Furthermore, in the
revised edition of his article (1871; given in English translation in
Stirling 1943:17), the only changes apparent are slight expansions and
improvements of his Ethiopian data.
The practice of relating North America to the Old World through an
analysis of TZ l's racial type was continued by Alfred Chavero, who postu-
lates (1883:62) that in Tertiary times all the continents were united and
9
contained Negro inhabitants who disappeared from many localities when the
continents were later separated. Chavero feels that the Tres Zapotes
colossal head, "whose type is clearly Ethiopian," serves as evidence for
the presence of an early Negro population in Mexico (ibid., 63).    In addi-
tion to Chavero's work, the question of the Negroid aspect of the heads
is considered by D. W. Jeffries in his article entitled "Pre-Columbian
Negroes in America." Jeffries accepts (1953:213) the Negroid appearance
of the colossal head found in 1862 in the "Canton of Tuxtla" as "valuable
circumstantial evidence" for his theory that Arab sailors were indulging
in a Negro slave trade with the Indians of Mesoamerica prior to Columbus'
voyages to the New World. Finally, in 1962, a special correspondent for
the newspaper Muhammed Speaks (1:7:23-28, May 1962) claims that LV 1 was
"most definitely carved with a black man sitting as a 'model' more than
1300 years ago." This fact helps the correspondent to prove that black
men were first on the earth and were busy creating the civilization that
greeted the white man who, last to descend from the apes, was spending all
his time living in caves.
Sources of Stone
Williams and Heizer (1965) have made field investigations of the
sources of the stone from which the colossal heads were carved. Within
the context of a general discussion of the geology of the Olmec area, they
note that the source location for the stone of the TZ 1 and NS 1 heads was
Cerro El Vigfa, while the probable source for the La Venta and San Lorenzo
heads then known was the Cerro Cintepec.   They further believe (ibid., 5)
that the Olmecs utilized "detached and rounded boulders" for their sculp-
tures, rather than quarrying the lavas in situ.
Removal to Present Locations
The removal of the heads to their present locations in modern times
has scarcely been noted in the literature. The transportation of the four
La Venta heads to Villahermosa is briefly mentioned by Drucker, Heizer and
Squier (1959:297), while Heizer, Smith and Williams (1965:102) mention that
Medellin Zenil told them something of the circumstances of the removal of
NS 1 to Santiago Tuxtla in 1951. While something is known about the recent
travels of SL 1, SL 2, and SL 5, the circumstances pertaining to the removal of
SL 3 and SL 4 to the Museo de Antropologla de la Universidad de Veracruz in
Jalapa are unknown.   Of the twelve colossal heads, only TZ 1 and SL 6 still
remain at the sites where they were discovered.
SL 5 is by far the most widely traveled of the Olmec heads. As part
10
of a traveling exhibit of Mexican art, it toured museums in fifteen Euro-
pean cities, including London, Paris, Rome, Stockholm, Copenhagen, West
Berlin, Vienna, Warsaw, Leningrad, and Moscow, and resided briefly at the
Los Angeles County Museum (Gamboa 1963:viii) before-returning to its present
location at the Museo Nacional de Antropologfa in Mexico City. SL 1 left
Mexico to be placed on exhibit in the Seagram Plaza in New York City, and
was later returned to the Museo de Antropologia de la Universidad de Vera-
cruz (Artforum 1966).
It is only in connection with colossal head SL 2 that we can learn
something of the problems involved in the transportation of the heads. This
sculpture-which formed part of the exhibit. "The Olmec Tradition"' at the
Museum of Fine Arts in Houston, Texas, before being returned to the Museo
Nacional de Antropologfa in Mexico City-aroused considerable attention
when it was removed from its in situ position at San Lorenzo and transported
to the United States. J. J. Sweeney, Director of the Museum, has written
(1963) a fairly detailed account of the methods utilized and the problems
encountered in the removal and shipping of this colossal head. Included in
his article is a description of the discovery of the exact location of the
head by helicopter survey, an account of how removal of the head was arranged
with the villagers of San Lorenzo Tenochtitlan in exchange for the promises
of a prefabricated schoolhouse, and a series of excellent photographs dealing
with the actual techniques employed in transporting the head. A less detailed
and far more sensational report of the removal of SL 2 was published by F.L.
in Archeologia (1965:2:41-47).
The Colossal Heads in Relation to the Olmecs
Of perhaps the greatest interest to the student are those works which
consider the problem of the colossal heads in relation to Olmec tradition
and its chronology. These studies take their impetus in large part from the
late Mexican artist and art historian, Miguel Covarrubias. Fascinated by
the Olmecs from the beginning of Stirling's work (see Covarrubias 1942), he
wrote continuously about them and sought their artifacts for his private
collection up to the time of his death in 1957. While most of Covarrubias'
work is concerned with the larger problems of placing the Olmecs in time
and space, and their position in Mesoamerican high culture, he did make some
comments pertaining solely to the heads. Covarrubias felt (1957:65) that
the heads depicted "fat youthful personages with Negroid features, wearing
helmets rather like those for football," and that they were the product of
the florescence of the Olmec culture, a culture which probably existed
during the Middle Preclassic period (ibid., 80-81).   Covarrubias (1946:97)
felt that the fact that the heads were "meant to be simply heads without
bodies" and that the La Venta heads "face directly north-south" might imply
11
that "they could have been astronomical sights [i.e. sighting points] or
simply memorial monuments."
Following Covarrubias, a whole series of works concerning the Olmecs
were written, many of which discussed the colossal heads. Kubler (1961:67)'
made the first attempt to seriate the heads. He suggested that they were
like "ideal portraits," that they showed a clear development through two
and possibly three generations of sculptors, and that this development
could be subdivided into three stages. First came "round heads of grim
aspect" (LV 1 and TZ 1); next, heads of intermediate shape characterized
by parted lips (LV 2, LV 3, LV 4, and SL 2); and finally, and latest, "long
heads of majestic expression" (SL 1, SL 3, SL 4, and SL 5).    It was Kubler's
opinion (ibid., 333, fn. 12) that the heads must have been carved between
200-300 A.D., since, he claimed, the earlier radiocarbon dates for La Venta
do "not accord with the sculptural evidence" for "parallels with Maya sculp-
ture require a terminal data c. AD 300 for La Venta carvings." However,
Squier (n.d., 253), in "A Reappraisal of Olmec Chronology," implies a seri-
ation which opposes Kubler's. While noting "the tenuousness of this sugges-
tion," Squier feels that "San Lorenzo may provide a bridge between the
monumental art of La Venta and some of the monuments at Tres Zapotes [in-
cluding TZ 1]."  Furthermore, since Squier believes that the Tres Zapotes
head is the youngest and was probably erected during the first phase of the
Tres Zapotes sequence (ibid., 183), a period which terminated around A.D.
100 (ibid., fig. 7, 258a), he further implies that Kubler's chronology is
short by at least two hundred years.
Two works appeared in 1963 which make reference to the giant heads.
Alfonso Medellin Zenil, writing a section on "The Olmec Culture" for the
unpaginated catalogue (section titled "Man") of the Houston Museum of Fine
Arts exhibit of Olmec culture related art, states that the heads "have
vigorous and precise negroid physical features"; and that they were probably
carved during the "gran Horizonte clasico," which he dates as lasting from
the first century B.C. until the ninth century A.D. (ibid., section titled
"The Culture"). T. Smith (1963), in "The Main Themes of the Olmec Art
Tradition," writes a section on the stone heads that is for the most part
a summary of how the heads relate to the Olmec style in general. She feels
that "although they are different from other Olmec art, they employ the
major features of the art style (ibid., 129), and that "there seems to be a
clear affinity of one head to another at each particular site' (ibid., 128).
The problem of whom the heads depict was considered in two works
which appeared in 1964. In El Pueblo del Jaguar, Roman Pifia Chan and Luis
Covarrabias state (1964, legend of plate facing p. 36) that the heads could
represent either distinguished warriors or champions of the Mesoamerican
ball game. They also note (ibid., 48) that the heads might be evidence for
12
the custom of decapitation which "seems to be a characteristic of the
Olmecs."  Peter Armillas, on the other hand, enlarges (1964:303-304) upon
Stirling's (1955) suggestion that the heads were individual portraits of
chieftains when he states that "the colossal heads may be chieftains'
memorials." Kubler (1962:69) writes that "the colossal heads manifest a
pharaonic desire for eternity, for physical survival beyond all the acci-
dents of time."
The available information on the archaeology of the southern Gulf
Coast region of Mexico was synthesized by the publication of three articles
on the Olmecs in Handbook of Middle American Indians (1965, vol. 3). These
articles-two by M. D. Coe (1965a; 1965b) and one by M. W. Stirling (1965)-
make frequent reference to the colossal heads. Coe feels that while the
heads are "somewhat divergent in style from the usual Olmec canon, being rather
Negroid in physiognomy" (1965b:741), they are definitely part of the culture,
and, in fact, are part of its classic form which flourished during the
Middle Preclassic of from 800 to 300 B.C. (1965a:681-689). Coe further notes
that if he is correct in assigning the Tres Zapotes sequence a Late Pre-
classic date, then the Tres Zapotes heads were carved elsewhere and only
later brought to this location (1965a:694). In regard to the distinctive
"helmets" of the heads, Coe suggests (1965b:763-764) that perhaps they were
defensive and functioned as protection against clubs of various shapes since
spears and spear throwers are not depicted in Olmec art and arrowpoints are
not found in Olmec middens.
Stirling (1965:733) feels as he did in 1955: that each head "has an
individual quality and was probably the portrait of a prominent leader."
He characterizes the heads (ibid., 721) as being almost completely realistic,
but notes that the Olmec artists used conventionalizations for certain
effects. Such conventionalizations include the "area at the base of the
nose [nasion], which the artist altered to achieve the expression desired"
(ibid.), and perhaps the helmet which could be an "artistic convention for
the purpose of eliminating protruding parts" (ibid. 733-734).    Finally,
Stirling notes (ibid. 734) that all the heads have flat backs, which might
indicate that they were meant to be set up against a wall.
Westheim (1965:131) believes that the U-shaped design on the head-
dress of LV 1 and LV 4 indicates a "relationship with the moon," and that
"the headdress refers, therefore, to fertility, which is the work of the
moon. So we may interpret the colossal heads of La Venta as representat-
tions of the vegetation god."
Two recently written analyses of the Olmec style make further con-
tributions to the discussion of who is depicted in the heads and how long
it took to carve them. Wicke (n.d. 71, 158) believes that the heads were
13
monuments raised in honor of dead chieftains and that they were carved
over a period of at least two hundred and forty years. He further notes
(ibid. 130-133), from his ordering of the width, depth, facial animation,
and iris indication of the heads, based upon Guttman's (1944) scaling,
that the Tres Zapotes heads were made earliest, followed by the La Venta
heads, and that the latest are the San Lorenzo heads.
On the other hand, the Mexican archaeologist Ignacio Bernal, in a
recently completely work on the Olmecs now in press, states that the heads
are not portraits but rather may be depictions of "chieftains or warriors
in a general sense" (Bernal n.d.). He also feels that the heads must have
been sculptured over a fairly short time span, perhaps two or three gener-
ations, for at least some of them are the work of the same sculptor or
family of sculptors. In a footnote, Bernal mentions a letter he received
from M. Coe in which the latter disputes Kubler's seriation, and proposes
instead that the La Venta heads are earliest, due to their "close resem-
blances to the jade and pottery figurine style," then come the San Lorenzo
heads, and finally the Tres Zapotes heads, which are "extremely portrait-
like and lack pseudo-drilling at the corners of the mouth" (Coe, cited in
Bernal n.d., fn. 40). It should be noted, however, that Coe now believes
(1967, figs. 1, 2) that a series of early radiocarbon dates from San Lorenzo
"reverses the usual scheme and puts San Lorenzo at the beginning, not the
end, of Olmec development," followed by La Venta, and finally Tres Zapotes,
where there was "the final flicker of a civilization which could now barely
be called Olmec" (ibid. 1400).
Bibliographies
To conclude this summary of the history of research on the Olmec
heads, reference is made to two bibliographies concerning the Olmecs.
"Bibliography for Olmec Sculpture" (Jones 1963) is very extensive up to its
date of publication. It does, however, lack certain entries of interest to
the archaeologist, such as reviews of site reports. Heizer and Smith's
(1965) "Olmec Sculpture and Stone Working, a Bibliography," while less
extensive, is more archaeologically oriented. Together these two sources
probably contain the most important references to Olmec archaeology since
Melgar's first paper of 1869.
14
DESCRIPTION OF INDIVIDUAL COLOSSAL HEADS
Introduction
The existence of the colossal stone heads of the southern Gulf
Coast of Mexico has been known since 1869, and they have received more or
less constant attention since that time. Much of the interest in the
heads, of which there are now twelve known examples, has arisen from their
inherent artistic quality.   Art historians tend to view the heads as some-
what apart from the Olmec society that produced them. While we would not
disagree that the colossal heads are superb masterpieces of an ancient art,
it is our opinion that they can only be understood as one element-albeit
a prominent one-of Olmec culture as a whole. In this paper, we are mainly
concerned with a detailed description of the dozen colossal heads and the
differences between them. What part these particular sculptures played in
the ceremonial life of the people who built and occupied the sacred centers
of Tres Zapotes, Nestepe, San Lorenzo, and La Venta we do not know, and it
is probable that we will never find the answer. What relationships to
other forms of Olmec sculpture the heads can be shown to have is not dis-
cussed here. We present in this paper that which we believe to be of pri-
mary importance in the study of Olmec sculpture-detailed information on
the statues themselves. Such detail cannot be obtained from the published
literature, and the purpose of this report is to provide this information.
Before Olmec sculpture can be studied properly, and meaningful con-
clusions derived on the nature of the style as it is variably expressed at
the Olmec culture sites, equally detailed records of all the monuments
("altars,"1 "stelae," and other forms) must be secured.  To do this, ironi-
cally enough, we must rely not on the usual data of the archaeologist, but
rather upon the more limited resources of the art historian; namely, the
heads themselves. As Heizer and Smith (1965:71) have pointed out, while
Olmec society certainly constituted a cultural entity, most of the previous
scholarship on the topic has been in terms of its distinctive art style.
This is particularly true of the colossal stone heads, since they all lack
any but the vaguest archaeological associations.   Thus, it is from a care-
ful scrutiny of the heads themselves, and from a comparative and inferential
approach to their details, that we may hope to draw relevant cultural con-
clusions.
In order that our study may be placed in chronological perspective,
we have included a short history of the Olmec colossal heads since their
discovery, and a brief discussion of the larger-than-life heads in pre-
Columbian Mesoamerica which are non-Olmec in origin (Appendix II). Atten-
15
tion has been paid to the geography of the Olmec area from which the
colossal heads have come.
In the detailed descriptions which follow, it will be noted that
we have established our own designation system by which the heads are
numbered according to the order of their discovery at each site. There
are now six heads known from San Lorenzo, four from La Venta, one from
Tres Zapotes, and one from Nestepe. We have abbreviated the designations
of the heads by utilizing the initials of the site from which they came,
and their respective numbers, in the following manner:   San Lorenzo, SL 1
through SL 6; La Venta, LV 1 through LV 4; Tres Zapotes, TZ 1; and Nestepe,
NS 1. Designations by which the heads have previously been known and their
present locations are shown on page 16.
We have attempted to emphasize the descriptive attributes of the
individual heads. In the rather detailed descriptions of each head, it
has been necessary to utilize certain terminological distinctions which
are explained below, with the aid of the three schematized drawings in
Figure 1.
On the tops of the heads is what we have designated the headgear.
This term is used to refer to that portion of the head-covering which
rests upon the dome of the cranium. In most cases the headgear is dis-
tinct and separate from the head band, which is wrapped around the lower
cranial area, passes across the area of the forehead, and covers the area
of the eyebrows. The term chinstrap is used to designate that portion of
the head apparel which extends downward from the lower margin of the head-
band just in front of the ear and beneath the chin. If the chinstrap is
rather short, extending downward only so far as to be roughly between the
ear hole and the cheek, we have referred to it as abbreviated, or as a
cheekstrap.  A full chinstrap extends to the bottom of the jaw and passes
beneath the chin. The chinstrap, in each case where it can be clearly
seen, appears to be a part of the headgear, since it runs underneath the
headband and is therefore presumably attached to the lower edge of the
headgear. The combination of headgear, headband, and chin- or cheekstrap
is what has been referred to by other authors as the "helmet."
Ear ornaments are the decorative elements attached to the ears of
the heads. They appear to cover, pierce, or dangle beneath the lobe of
the ear in each case where they are represented.
In describing facial features, we have used the anatomical terms
that we feel most accurately indicate the actual physiognomic area as it
has been translated into stone sculpture. Nasion refers to that area where
the more fleshy skin of the forehead meets the taut epidermis of the nasal
N
0
54
cU
4)
x
o
uT-
4-i
CU
(/
4-i~~~~~~~~~S
00
o                 >
*  '4  S
N    X)  N    N   N
>    *   >    >   >    'u
co   o        co  coX
co   0   co   CU  cUo ,
CU   .   CU   cU  CU
5%  ~ ~ ~ ~   *'  .' ~~~~~ E-I
CU       CU   CU  CU
1-4   *   -4  1-I  u-I
Cd   C   'd   co  c
Cl              pii  14 < C
z~~~~~
0    *   0    0   0
CD cn      coen X
4.
r"'.  rb  00   CU  rN e0           I   0     <-I
un   n    W'   od   -I                       - l_  0  -
*-   *0    '   X    *-  S4    ..   ..   ..  .
cn  cn   cn         n    a) .114 un  n  Ln un
'4  -4   .4    Q    '4     N 1     Lf,  Lun  L
0%   as   0%   :1   as   *-A 0%O  0%   0%   0%
1-I  u-I  u-I                 r uI  u-I  u-I  u-I
o    uo  Io    C    o    * 0  o u-  u   a    o
C: 0      O    0  u- I   C rr  to uI   b 0   tO0
C'4  e-4  '-4     N<     4.v-4  0'  u-I v-  r4  C',
~~4  p  p  CUC       0-u-I 4   4    4
01  0{   0,1  ,0        N    0    0    0,1  01 co
4i   4    4-  O   bo  "  Z   4 :W           4i
U)  Ut  U)~~~~~~~~4 1 - 00   co  Cd0U  t
la   Q)  Q)Q         )    )   a
cs  n  4      Si ?   4 < CY    <    Nco  C4
.|  4  4   N b0 4    v 3 4     4    4 -4
2    2:   2    aU~       ~   2     2    2t 2: C
3 i a ' 3 X < i S i 3~
C'4      CY)      -4       1-4
:>       >        >          N
-X        H
(/2
cn
x
r I         e l        e ' ,
(/2        (/2      *(/
*-      *
'-4     c'4
Ln t    LO
ON       N      c
'u-    'H        0
60      bO     1?
4.,1    4.1    (<
r.      0       a,%
'4     L4       c
0,-     0       0
43              4 n
2?      2       2
'4           If          P
(/          (E           cn
H          _t  col  (~~~n   Ln   %
C14  cn  z ~ ~~~~~~~~~-     'I  e'   4    '
u-     el  C'   '4   Z                0    0   0    0   0   0
u-       Z    Z   Z   Z    Z   Z
*   *   *   ~ ~*  Co
0      0O  0    0   0        *       0    0   0    0   0    0
Z      Z   Z   Z    4.0               N   N    N   N   N    NQ
CU     CU   CU  CU                    0   0        a0   a0   a0
4.1    4.1  W4.  4.  CU      0                p    p
>   >    >    Co      4
X  0  X   X    S        Q       X    0   X 0   Co
CU     CU  CU  CU                    CU   cU  cU   CU   CU   cU
O   O   0   0~ H                  /   (/2  (/2  (/2  (/  (/2
CU
n
Q
(12
CU
'-4
u-I
,5
Co
0c
CU
1-I
*u-4
0
co
CU
Co
'-4
1-I
'u-
5%
CU
4.)
0
p4
cd
Ct
w-4       0
r-4       N
>         CU
N
cU       >
4.1
0)       Co
>         0
4.1
CU       0
CU
0)      N
0
N        0o
P4       H
16
0n
co
CU
Co
Co
W.
u-I
co
C,
4
0
0
0
0
co
0
4.1
cU
Co
0
5%
co
w
CU
I c
'u-I
4.1
0
co
IO~C
NH
CU
p4
w
a,   C
O 10
v    co
Sp  E-1
Cd
N
LO
00
0%
bo r.-
s0    0
-1. -4
r.2N
o   *4
0   or4
1-I r4
m /2
I"- -v
O 3
(L   Q)
00
:2 2
0
'-I4
4.1
CU
a)
Ut
P4.
0
0
co
.p
bo
0
Pr4
0
.p4
4.1
o to
0
c0
co
Co
0
0
C,
u-I
17
breadth
- headgear
- head band
- fatty pad
- nasion
-pit in corner of mouth
b
I  r  breadth
77
_ ' N
dimpled pit
iris
tear duct
lw
a
-        1+       "n
Ai
A
A
headgear
head band
chin strap
(abbreviated)
ear ornament
c
Figure 1
Descriptive attributes of the individual heads
I
A
-0001-             --T,-.- -
18
bridge, thereby creating a pinched, wrinkled, or folded effect. The terms
subrhomboidal and triangular refer to the approximate shape of this folded
area, as does the term double, which actually describes a double or divided
rectangle. Fatty pads are those fleshy bulges which appear immediately to
either side of the nasion and extend upward to disappear or diminish under-
neath the head band. In speculating as to the physiological interpretation
of the nasion and fatty pads, we suggest the possibility that these are not
normal physical characteristics but represent protruding folds of flesh
caused by the downward pressure of the tightly fitting head band. Jowls
are the puffy areas which hang beneath the cheeks at the lower front sides
of the face, serving in most cases to emphasize the squareness of the head
from a frontal view.
The nares are found at the broadest part of the nose, and are the
fleshy areas immediately above and covering the nostrils, which take the
usual form-drilled and/or pecked holes separated by a septum.
Tear duct refers to a slight though carefully executed extension of
the area enclosed by the eyelids at the corner of the eye.
Dimpled pits, or, p      with dimples, are a form of defacing mark,
each of which consists of a cup shaped depression with a small concavity or
dimple in the bottom and exact center. Further description and discussion
will be found in the section on defacement.
It is important to note that when we refer to the right ear or left
eye, for example, we are referring to the right ear or left eye of the head,
and not to the feature as viewed from the right or left. Reference to head
length conforms to the standard anthropometric measurement of distance be-
tween the frontal area just above the nasion and the rear of the occipital
region. Width, in our usage, refers to the widest part of the head above
the temporal region, measured laterally.
Tables summarizing dimensions, weights, and stylistic elements of
the colossal heads appear in Appendix I.
19
La Venta Colossal Head No. 1
LV 1 (figs. 2, 3a, 3b; pis. 1, 2a, 2b, 6a) was first reported by
Blom and La Farge (1926), but was not fully described until the publication
of Stirling article, "Great Stone Faces of the Mexican Jungle" (1940:310,
328), and his "Stone Monuments of Southern Mexico" (1943:56-57).    Carved of
basalt secured from Cerro Cintepec in the Tuxtla Mountains (H. Williams and
R. Heizer, personal communication), this head was found at the s-uthexn edge
of the main pyramid at La Venta, facing south.< It is fairly large in rela-
tion to other Olmec heads, as it weighs 24 tons and is 2.41 m high and 6.4 m
in circumference.  Thus, only colossal head SL 1 exceeds it in weight, only
SL 1 and SL 2 exceed it in height, and only LV 4 exceeds it in circumference.
LV 1 is now located in the Parque La Venta in Villahermosa, Tabasco.
In common with the other Olmec heads, the lower edge of the headgear
extends to a position immediately over the eyes. However, in contrast to
several of the others, the nasion does not have incised extensions running
up to the helmet, but rather has incised extensions over each eye. The
nasion is subrhomboid and is characterized by the fact that it is only slightly
incised-a lack of detail most probably due to partial erosion. The eyes of
LV 1 are flat and slightly inset within the eyelids, and the right eye is
placed higher up on the face than the left. The iris is depicted by a raised
disc in the middle of the right eye, and a raised disc slightly off-center
toward the nose in the left eye. The eyelids do not overlap but join on the
inside and outside of the eye, and a small, rounded "tear duct" is formed at
the outside juncture of the lids.
The nose is quite wide and noticeably flattened, and the nostrils are
visible in a full face view. In common with the other La Venta heads, the
base of the nose and the edge of the upper lip are placed quite close togeth-
er. This upper lip is bow shaped, while the lower lip is in the shape of the
letter "U". The mouth is closed, and has small rounded pits in the lip cor-
ners. The chin is depicted by a cursive "W", the ends of which join with the
outer ends of the mouth. The chin is further emphasized by the fact that the
middle inverted V of the W is elevated and rises to the middle of the lower
lip's bottom edge. In profile, the ears are placed unnaturally forward in
the face. They are small in relation to those of other heads, and appear to
be more angular than curvilinear. The ear lobes are completely covered by
ear ornaments.
LV 1 has quite prominent jowls. They extend upwards on either side
of the mouth to meet with the cheeks and form a single unit vertically bound-
ing the nose and mouth. This unit along with the flat, broad nose contributes
to LV l's appearance of an extremely flat, square face.    Furthermore, in
profile LV 1 is decidedly prognathous. The face forms practically a single
wide plane which inclines inwards at an angle of approximately 25 degrees from
20
the base of the face to the lower edge of the headband.
In relation to many of the other heads, LV 1 bears relatively little
decoration. The head band is plain in the front, and is comprised of six
sets of double squares on the left side of the head. The right side of the
head band is now plain, but this is most probably a result of erosion of the
incised lines. The headgear is reminiscent of that of SL 1, and is simply a
raised U-shaped design placed directly above the front center of the head.
band. This U-shaped motif extends backwards over three-quarters of the top
of the head, expanding laterally slightly one-third of the way back. Hang-
ing from this U-shaped element over and to the bottom of the front of the
head band are three attachments depicted in relief. The center attachment
is oval, while the outer ones are comma-shaped. These are also to be seen
on the headgear of the central figure in Stela 2 from La Venta (Drucker
1952, fig. 49).
The ear ornaments are large rectangles with rounded corners. The
long edges of the ornaments run vertically, and the ornament extends from
where it covers the earlobe almost to the base of the head. It is probable
that within each rectangle there was a relief diamond whose tips touched
the midpoints of the four lines of the rectangle. In the center of the
diamond was an incised circle. The left ear area, however is now so eroded.
that only the two lower incised lines of the diamond remain. In front of
each ear and extending downward to the rectangle is the characteristic cheek
strap. Although it apparently passed beneath the ear ornament, the strap
does not reappear below the ornament.
The only ancient defacement apparent on LV 1 is a series of grooves
on the dome of the head. One series occurs within the U-motif, and larger.
ones are present toward the back of the dome. These grooves are long and
narrow, being approximately 5 cm in width and depth, and over 20 cm in
length.
LV 1 has apparently suffered some post-Conquest defacement.    Both its
right side and face have recently been sectioned into square areas by very
thin incised lines. Perhaps some modern artist indulged in this lamentable
practice in order to facilitate his rendering of the head. A considerable
amount of natural surface scaling has occurred on the stone. The back,
which apparently was once quite flattened, is now very eroded, as are the
rear section of the right side and the areas of the face mentioned above.
The pattern of scaling on the back follows the natural laminae of
the lava flow which produced the original stone, and it would therefore
seem that the back of LV 1 was part of the original surface of the stone.
This clearly indicates that LV 1 was carved from a large, naturally rounded
boulder.
21
LV 1 is both asymmetrical and minimally modified from the original
cylindrical boulder. The right side of the face is much wider than the left,
the eyes are not aligned, the ears are too far forward, and in full-face all
the features appear to be slightly askew. In profile, the head appears far
too wide in relation to its height to be anatomically correct. Furthermore,
as implied in the discussion of the facial features, LV 1 as a whole is
carved in much lower relief than those from other sites.    Nevertheless, LV 1
is an excellent example of the sculptor's art, and one is impressed by the
naturalism discernible in its over-all appearance.
La Venta Colossal Head No. 2
LV 2 (figs. 4, 5a, 5b; pls. 3, 4a, 4b, 5a, 5b) was discovered during
a joint Smithsonian Institution-National Geographic Society expedition to
the Gulf coast of Mexico in 1939-40.   In 1940, Stirling published (1940:331,
333) two photographs of it in his article, "Great Stone Faces of the Mexican
Jungle," and later (1943:57), in his "Stone Monuments of Southern Mexico,"
described it as follows: "This [head] is the middle one of three big heads
set in a row and facing north, two hundred yards north of the big mound [i.e.
the pyramid]." Carved from the same Cerro Cintepec basalt used in most of the
La Venta heads, LV 2 is the lightest and shortest of the La Venta heads, with
a weight of 11.8 tons and a height of 1.63 m.    It has a circumference of 4.24
m, and thus is slightly wider than LV 3.    LV 2 is presently located in the
Parque La Venta in Villahermosa, Tabasco.
The headgear of LV 2 extends down nearly to the eyes; the nasion is
incised and subrhomboidal in form.   Joining the nasion on both sides, a
deeply incised line curving up to the head band indicates the forehead area.
The eyes of LV 2 are quite sunken, for the eyeball is not shown as it is on
the other heads, that is, filling the eye socket, but rather the center of
each optic cavity is slightly indented and flattened to form an iris. The
upper eyelid overlaps the lower one on the inside of the eye, while the eye-
lids meet on the outside edge to form a small, rounded tear duct. Although
the nose is heavily eroded, it does appear to be wide, with nostrils visible
in full-face perspective. At present the nose appears to be slightly snubbed,
but because of heavy erosion it is impossible to accurately reconstruct this
feature in its original form.
Since the upper lip has completely eroded away, it is impossible to
either reconstruct its form or to estimate the distance between the nose and
the upper edge of the lip. The lower lip is U-shaped, with a concave upper
surface.  Since the lips are parted, and what remains of the upper lip's
surface is convex, the mouth has been termed "smiling" (Stirling 1943:57;
22
1965:733). The presence of large, upward-slanting oval pits in both mouth
corners further contributes to the apparent smiling visage of this head,
but the smiling effect of the mouth would, in our opinion, be markedly
lessened were the upper lip present to frame it.    The four upper incisors
are present in the mouth, and since the two central incisors protrude be-
yond the inner ones, the denture appears to be somewhat buck-toothed.
In profile the chin appears somewhat recessed, although once more
this may be partially due to erosion. The recessed nature of the chin as
now seen is further emphasized by an eroded area in the lower edge of the
stone below the center of the chin. The upper portion of the right ear of
LV 2 is well made, and appears more naturalistic than those of several of
the other heads. The ear ornament covers the ear lobe. The ear lobe is
also covered on the left side, but here the erosion is so severe that only
the barest outlines of the upper parts of the ear remain.
Of all the heads, LV 2 has the most oval face. Furthermore, the
face is slightly wider across the jowls than across the forehead. Thus
these features, plus the apparently smiling mouth, suggest a great similar-
ity between this head and the tradIitional Halloween pumpkin. This similar-
ity would no doubt be much lessened were it not for the extreme facial
erosion. So many of the facial features have disappeared that the face
seems soft and devoid of detailed facial planes. That this is so can be
demonstrated by the prominenL Cheekbones.   These, the least eroded feature,
contribute only to a certain pouchiness in the face. However, were the
other facial features as prominent as the cheeks, the head would have a
more impressive appearance.
Of all the colossal heads, LV 2 is the only one with an asymmetric
head band design. The central element of this design resembles a flattened,
upside down medieval shield (or family crest). This is the element termed
by Stirling (1943:57) a "bowknot." While there is no doubt that this ele-
ment was incised to the left of the head's centerline, it is not possible,
due to erosion, to now ascertain the shape of the element that joins it and
continues around the left side of the head. While we have rendered this
element as a sort of flying wing, it must be confessed that this is at best
a provisional reconstruction. The element that continues around the head
to the right from the central element is roughly oval in shape. Placed at
an oblique angle to the head band, it joins the band in the front cotermin-
ously with the central element.
LV 2 parallels TZ 1 in that the dome is absolutely plain and undeco-
rated. The ear ornament is likewise plain, being simply a flat disc with a
hanging tassel. The disc covers the ear lobe, while the tassel hangs down
to the lower edge of the head. This ornament is reminiscent of the small
23
discs with tassels that form the ear ornaments of SL 4 and SL 5. The cheek
strap hangs down in front of the ear to a level slightly below the top of
the ear ornament.
LV 2 is perhaps the most altered of all the heads. Pits without
dimples are present on the top of the head, on the head band, and in vari-
ous locations on the face. A few dimpled pits are found on the top of the
head. The entire back of the head has been made concave by the many narrow
grooves, wide, deep gouges, and grooves within gouges that are present there.
These defacements appear to follow a random pattern. There are a few more
deep gouges at the base of the back.
The combination of erosion and defacement has made LV 2 the most
difficult of the heads to visualize in its pristine state. It would appear,
however, that this head was similar in its general proportions to the tall,
narrow San Lorenzo heads, such as SL 1 and SL 4.   While LV 2 no doubt has a
more cheerful visage than do the San Lorenzo heads, which have been des-
cribed by Kubler (1962:67) as "majestic," nevertheless, in its original
state, it must have approached some of the majesty of the San Lorenzo heads.
It is our opinion that to label LV 2 as the head of a youth or of a smiling
person is to misread the effects of erosion and defacement on a head that
must have appeared considerably different in its original condition two and
a half millennia ago.
La Venta Colossal Head No. 3
As in the case of LV 2, LV 3 (figs. 6, 7a, 7b; pls. 7, 8a, 8b) was
discovered by the 1939-1940 joint Smithsonian Institution-National Geo-
graphic Society expedition to the southern Veracruz-northern Tabasco region
of Mexico, and was first described and illustrated by M. W. Stirling (1943:
57-58, pi. 42b). LV 3 was originally located to the north of the great
pyramid, facing north. It was the easternmost of the three colossal heads
lined up here in an east-west row.   Like most of the La Venta heads, LV 3
is made of basalt from the Cerro Cintepec.   It is slightly above average in
size, weighing 12.3 tons, standing 1.98 m high, and is 4.05 m in circumfer-
ence. It is presently located at the Parque Museo de La Venta in Villa-
hermosa.
As with the other heads, LV 3's headgear comes down to a position
just above the eyes. The nasion is a double one, whose outside edges curve
out and up under the headgear.  It is impossible to ascertain what features
were sculptured on the left side of the face since this side is severely
eroded. The right eye does not bulge, but appears to be flush with the face,
and the left eye is so eroded that little can be seen. Facial erosion in
24
the area of both eyes is so great that no additional observations can be
made beyond noting that the eyelids of the right eye apparently met at the
inner corner and formed a point at the outside corner. Erosion on the
nose is also great. However, it does appear that it was narrower across
the bridge than those of the other La Venta heads, and that in profile, it
was slightly snubbed.
Since the upper lip has almost disappeared, it is impossible to
ascertain its original shape, or the distance between the upper edge of
the lip and the nose. It would seem probable that the mouth of LV 3 was
open, and that the lower lip was U-shaped. There is a definite pit in the
right corner of the mouth, but this feature does not appear on the left
side. The chin in full-face view appears similar to the W-shaped chin of
LV 1; however, in profile LV-3's chin is straight while LV l's chin pro-
trudes. The ears of LV 3 are quite unobtrusive. The left ear is entirely
covered by the ear ornament, while of the right ear only the ear pit and a
portion of the central part of the external structure are not covered by
the headgear or the ear ornament.
In sum, the face of LV 3 has been eroded in such a manner as to
nearly obliterate all of the features. Nevertheless, sufficient detail
exists to enable us to have some idea of the relationships of the various
features and to give us a fairly clear picture of how this face must have
looked originally. LV 3 apparently presented a rectangular visage, and
lacked prominent facial contours such as high cheekbones and pronounced
jowls.  In profile, also, this face probably had no one outstanding feature,
such as a jutting jaw, prognathism, or enlarged brow-pads.    In general
facial alignment, then, LV 3 must have appeared quite similar to such even-
featured heads as SL 2.
Due either to the effects of erosion or to a particular convention
of decoration, it is not possible to note any differentiation between the
head band area and the dome of LV 3. The headgear is basically a raised,
undecorated area which covers about seven-eighths of the top of the head.
It ends toward the rear of each side of the head and extends downward on
each side to about the level of the middle ear. There is no indication of
the headgear on the back of the head or at the extreme rear of the sides.
Thus the headgear represents either a plain helmet or scarf, covering most
of the head. The ear ornament of LV 3 is basically a rounded square with a
hole in its center, and bears a striking similarity to the jade ear spools
found in the Olmec region (Drucker 1952, pls. 52, 56 right; 1955, pls. 42-
48; Drucker, Heizer and Squier 1959, pls. 37, 39, 40).    It would appear
that the ear ornaments are placed approximately over the ear lobes.
The chin straps oi'LV 3 extend from the lower edge of the headgear
25
down to the base of the head and give the appearance of continuing under
the chin. The left chin strap differs from the right in that it is wider,
and covers the whole ear before passing beneath the ear spool. It is con-
siderably narrower below the ear ornament. The right chin strap is of the
same narrow width throughout, permitting the central and rear portions of
the ear to be shown.
There is considerable defacement of LV 3. Long, narrow grooves are
present all over the top of this head and under the right eye. Pits with-
out dimples are found on the headgear, on the nose, under the right eye,
near the left orbital area, and around the mouth.    Pits with dimples are
found on the head band, and one of the three pits inside the central hole
of the left ear spool is dimpled. This dimpled pit is particularly inter-
esting for there exists the distinct possibility that pits with dimples
were part of the original construction of LV 3. The central area of this
ear spool is apparently unfinished. and the fact that a dimpled pit occurs
w the hole in the ear spool should occur implies that it was
used  at least in this instance, as part of the technique of Olmec stone
carving.
As mentioned above, notwithstanding defacement and erosion, this
head bears certain resemblances to the general San Lorenzo head type.
LV 3 is relatively tall and thin, with a quite flattened back, and there is
a slight tendency toward constriction from the bottom of the head to the
top. All this indicates that were LV 3 in its original form, it would
probably appear much like a San Lorenzo head both in its proportions and
in the form and relationships of its features.
La Venta Colossal Head No. 4
LV 4 (figs. 8, 9a, 9b; pls. 6b, 9, lOa, lOb) was also discovered
during the 1939-40 joint Smithsonian Institution-National Geographic
Society expedition and was first described by Stirling (1940:315, 329, 332;
1943:58). LV 4 was found facing north, and was the westernmost of the three
heads standing in a row north of the great pyramid. It weighs 19.8 tons,
is 2.26 m in height, and 6.53 m in circumference. LV 4 is now located at
the Parque La Venta in Villahermosa.
LV 4 is made of a basalt which is recognizably different from that
used to fashion LV 1, LV 2, and LV 3. It is a "basaltic andesite" and is a
porphyritic, pilotaxitic, augite andesite. The stone is characterized by
abundant phenocrysts of plagioclase from ca. 0.2 to 2.0 mm long, all having
murky cores crowded with specks of glass and with thin, clear rims. The
26
cores are of medium to calcic labradorite; the rims are of calcic andesine.
The small plagioclase laths in the groundmass also consist of calcic andes-
ine. Next in abundance, forming about 15 per cent of the bulk, are pheno-
crysts and microphenocrysts of pale yellowish-green diopsidic augite, few
exceeding 1.0 mm in length. Prisms of hypersthene, some of them discrete
crystals, but most of them enclosed in jackets of augite, make up approxi-
mately 1.0 per cent. The same quantity obtains for roundish granules of
olivine which have been largely altered to iddingsite. This rock is much
finer-grained than the basalts of El VigLa and Cerro Cintepec, and differs,
from these last in lacking the conspicuous phenocrysts of augite and
olivine.  San Lorenzo Monument 20 is made of essentially the same rock as
LV 4.  The exact source of the stone from which LV 4 and San Lorenzo Monu-
ment 20 are made is not known, but it is most probable that it comes from
the Tuxtla Mountains. Almost certainly its source is not La Union.
Although much of the right side of the head band of LV 4 has scaled
off, it would appear that, in common with the other heads, this feature
came down nearly to the eyes. The nasion is a deeply incised version of
the subrhomboid type noted for LV 1 and LV 2. On either side of the nasion
the ends of the incised lines are intersected by deeply incised curved
lines which go up and under the head band. The deep incisions of the nasion
and these curved lines give LV 4 the appearance of having a deeply furrowed
brow, or perhaps a frown.
The eyes of LV 4 are sculptured flush with the face, and the right
eye is lower than the left. The iris is depicted by a circular raised
relief area which is intersected by the lids and disappears under them.
The upper lid overlaps the lower at both the inner and outer edges, and
the eyes are pointed rather than rounded at their outer corners. The nose
is extremely eroded, especially on the right side, but it would seem that
it was originally quite broad and flat, and had nostrils visible in full-
face perspective. Since the upper lip is eroded away and the nose also is
quite eroded, it is impossible to gauge the original distance between the
nose and the top edge of the upper lip. The lips are parted and the four
upper incisors are present. What remains of the upper lip indicates that
it was originally bow-shaped; the lower lip also is bow-shaped and is out-
lined with a raised ridge. The mouth has circular pits in both corners.
The details of the chin of LV 4 are completely eroded away. The ears also
have almost totally disappeared. However, since there is space for them
and a few traces of incised lines remain, it cannot be doubted but that
they were originally present.
In profile LV 4 is quite prognathous. It resembles LV 1 in that the
face forms practically a single oblique plane, slanting inward from its
base to the head band.   However, the parted lips and overhung upper lip of
27
LV 4 tend to break this plance slightly more than it is broken in LV 1. In
full-face view, the features of LV 4's face are slightly askew, and the area
of the right eye and cheek is lower than that of the left side. This gives
LV 4's face a superficial resemblance to LV 1.   However, differences in the
eyes and mouth, and the prominent cheekbones serve to distinguish this head
from LV 1.
The head band of LV 4 is nearly totally eroded away, but it is possi-
ble to partially reconstruct it. In front, it is comprised of a series of
five incised horizontal parallel lines. The three upper lines and the two
lower ones are placed close together, and these two groups are slightly
separated. On the left side of the head, the head band is first comprised of
a rounded square. Proceeding from this element is a curvilinear meander
which runs backward almost to the back of the head. It would seem that the
rounded square and the meander depict designs appliqued on the head band.
On the right side of the head the head band design is nearly eroded away.
The lines that remain, however, hint that the motifs on the right side were
similar to those of the left.
The headdress design is comprised of a three claw motif in the front
and a series of six plain incised rectangles with rounded corners which pro-
ceed in parallel fashion back from it over the head. The three claws touch
the top edge of the head band. This claw motif is similar to the claw head-
dress of SL 5. However the design on SL 5 has two series of three claw
motifs and the claws overhang the head band. Furthermore, the designs of
the back parts of the headdresses of these two heads are in no way similar,
since the latticework of SL 5 implies an openwork cap while the parallel
rectangles of LV 4 hint at a solid helmet.
The ear ornament of LV 4 is a large square with rounded corners. On
the left side it is so eroded that it is impossible to discern what design,
if any, was present within it. The central design of the right ear ornament
is a relief cross whose ends meet the center of each side of the square.
The outer edges of the ear ornament are notable for the fact that they con-
sist of a double, rather than a single, line. The right ear ornament as a
whole is reminiscent of a motif on the famous Humboldt celt (Coe 1965b:748,
fig. 18). LV 4 is notable for the fact that it may not have had chin or
cheek straps. No trace of a strap can be detected on the left side of the
head, and the right side is too eroded to settle the question.
LV 4 is defaced only on the headdress area. Here there are many
grooves all over the top, and a few pits without dimples toward the left
front and right middle of the top. On the other hand, it has been severely
damaged by erosion, particularly scaling.   As noted above, the chin, ear
areas, and portions of the face have been obliterated by natural causes.
28
In over-all view, LV 4 appears to be a minimally dressed boulder
with a flattened back. The relief of its carving is low in comparison to
many of the other heads, and in profile the head appears to be too wide in
relation to its height to be completely naturalistic. In these and other
features noted above, LV 4 bears a distinct similarity to LV 1.    However,
in the use of the claw motif and in the modeling of the lips, LV 4 parallels
SL 5. Thus, while LV 1 and LV 4 are distinctive of a La Venta type, there
is a relationship of these heads to those of San Lorenzo in certain artistic
motifs.
Tres Zapotes Colossal Head No. 1
TZ 1 (pls. 11, 12a, 12b, 35) has been known in the literature for
almost a century, having been first described by Melgar (1869) who called it
"La Cabeza Colosal de Hueyapan." It is presently located in an open field
about one kilometer northwest of the village of Tres Zapotes. The head lies
on its left side, the result of a recent attempt to move it with a bulldozer
to another location. Stirling (1943:17; 1965:734) refers to it as Tres
Zapotes Monument A and states that excavation showed it to have originally
been placed in front of the south mound of Group 1, facing north and resting
upon a foundation consisting of a layer of unworked stone.
TZ 1 is relatively small in comparison with the other colossal heads,
weighing about 7.8 tons. It stands 1.47 m in height and measures 5.49 m in
circumference. The stone from which the head is fashioned is a coarse
grained olivine and augite-rich basalt, much darker in appearance than that
used for the La Venta and San Lorenzo heads. Large spheroidal boulders of
this material may be found not more than 8 km distant from Tres Zapotes, on
the slopes of Cerro El Vigla (Williams and Heizer 1965; Heizer, Smith and
Williams 1965:102). Several of the monuments from Tres Zapotes are also made
of this particular basalt, among these Monuments F., C, and 9. The Nestepe
No. 1 colossal head (Williams and Heizer 1965:4) and the El Meson Stela
(Covarrubias 1957, fig. 68) are also made of the El Vigla rock.
In terms of physical features, TZ 1 differs from the colossal heads
at both La Venta and San Lorenzo. The nasion is triangular in shape and
quite flat. The fat pads on either side of the nasion are sharply defined,
flat-, and angular.  The eyes areseti      ell-dfinedsockets and are smaller,
relatively, than those of the other heads. No iris is indicated. The-apper
eyelid overlaps the lower at both the inside and outside corners of the
right eye, the latter corner being sharply pointed.    We are uncertain about
this feature on the left eye since it could not be seen at the time of our
visit.
29
The nose is short and broad but not markedly flattened, and the tip
is broken off. The upper lip also is somewhat damaged. It seems likely,
however, that TZ 1 resembled NS 1 in lip form and that the upper lip was
bow-shaped. The lower lip is straight. There is a slight ridge along the
outer edge of the lower lip. The mouth is closed, and the chin is short
and broad, with a deep indentation between the lower lip and the point of
the chin.
In contrast to the well modeled character of the facial features,
the ear of TZ 1 is quite stylized. It is outlined in low relief with only
its most salient external features defined. The ear hole is represented
by a short horizontal groove. The ear lobe is elongated and pierced by an
ear ornament.
In general, the facial features of TZ 1 are well modeled. The
cheekbones are clearly defined, as are the lines about the nose and mouth.
In profile the face is orthognathous, in contrast to the prognathous character
of several of the other heads, especially NS 1, which is geographically the
nearest.
Turning to decorative elements, we first note the headgear.     A plain
cap, lacking in ornamentation, covers the dome of the skull.     Two small
notches are present on the lower front edge of the cap, on a line with the
corners of the eyes. (A third notch, just to the right of the notch over the
right eye, appears to be the result of recent damage as it is not present in
earlier photos of the head.) A shallow U-shaped groove, varying from 13-15
cm in width and 2 cm deep, separates the head band from the cap. This groove
narrows slightly at the front of the head. The head band is plain, uniform
in width, and completely encircles the head, with the exception of the flat-
tened area at the back.
The chin strap extends from the lower edge of the head band, passes in
front of the ear and between the ear ornament and the cheek, and ends at the
bottom edge of the head. The ear ornaments are cylindrical plugs or spools
which pass through the ear lobes. They are roughly rectangular in shape,
with slightly expanded forward ends.    The leading edges are concave.    On the
right ear ornament there is an incised vertical line just ahead of the chin
strap. Whether this feature is also present on the left ear plug we cannot
say.
The back of TZ l's head is flattened and bears a series of eight
shallow, vertical, incised grooves.    These extend from the top of the rear
of the head to its base.   They may be a stylized representation of hair, and
if so, are similar to the braided strands of hair on the back of NS 1.
30
There has been little defacement of TZ 1. As noted above, the tip
of the nose and part of the upper lip have been broken off. Stirling
(1943:17) states that the left eye was mutilated, but is not specific as
to the character of such damage. There are two V-shaped grooves on the
head-a short one on the lower edge of the head band above the right chin
strap, and a longer groove (79 cm long, 2 cm wide) running vertically up
the back of the head from the right bottom edge. Finally, there are sev-
eral nicks and scratches on the top of the headdress. These appear to be
quite recent, and may have resulted from attempts to move the head.
Nestepe Colossal Head No. 1
The colossal head which is designated NS 1 (fig. 10, 11, 12a, 12b;
pls. 13. 14, 15a, 15b) was first described and illustrated by Heizer, Smith
and Williams (1965), who referred to it as Tres Zapotes No. 2. More recent-
ly Stirling (1965:733) has called it Monument Q. Tres Zapotes. Heizer and
his associates (ibid., 102) note that the head was not found in the immedi-
ate area of the Tres Zapotes site and that designations implying this to be
the fact were probably in error. Although some confusion still exists as
to its exact original location, we were informed in February, 1967, by
residents of the village of Tres Zapotes, that the head was uncovered during
bulldozing operations in a grove of royal palms on the west edge of a large
flat-topped mound known locally as the Cerro Nestepe. This mound lies just
east of the Arroyo de Hueyapan, about 3 km to the north of the village of
Tres Zapotes. The Nestepe mound group apparently does not appear in the
Stirling-Drucker sketch of the Tres Zapotes site (Stirling 1943, fig. 2).
The distance separating the Nestepe from the Tres Zapote site would seem to
indicate that the former is probably a separate site and should be so desig-
nated. For this reason we have now renamed the former Tres Zapotes No. 2
head as Nestepe No. 1 (NS 1).
It was reported to us by a Tres Zapotes villager that when discovered,
NS 1 lay at the east edge of a large flat area, perhaps a court, almost com-
pletely buried. The head faced north. There is no information concerning
the presence or absence of a base or platform such as that on which TZ 1
rested.  Heizer, Smith and Williams (1965:102) state that the head was moved
to its present location-the plaza of the village of Santiago Tuxtla -by the
Presidenteofthe municipality in 1951, although the plaque at the base of
head says (probably correctly) 1950. NS 1 is 1.4 m high, 4.9 m in circum-
ference, and weighs about 8.5 gons. The stone from which the head is sculp-
tured is similar to that of TZ 1 and probably came from Cerro El Vigfa.
The physical features of NS 1 are similar to those of TZ 1. The nasion
31
is flat and triangular.   The fatty pads above the eyes are also flat, al-
though they arch more than do those of TZ 1.   The relatively small eyes
are set in well defined sockets and appear to bulge slightly; the iris is
not represented. The upper eyelid overlaps the lower at the outer corners
of the eyes, and the lids meet evenly at the inner corners. As in TZ 1,
the outer corners are pointed.
The nose of NS 1 is straighter in profile, and more pointed, than
those of the other heads. The nostrils are drilled on a horizontal plane
rather than being inclined. The upper lip is bow-shaped, and the lower
lip is straight and markedly protruding. There is a raised ridge along
the outside edge of the lower lip. The mouth is closed. There are two
small circular pits at the corners of the mouth. The chin is not shown,
perhaps because of lack of space on the boulder below the mouth.
The ears of NS 1 are not as stylized as those of TZ 1. The modeling
of the exterior ear features is more realistic, and the rear edges are
smoothly concave in contrast to the rectilinear quality of the ears of TZ 1.
The ear holes are represented by small drilled pits. The lobes are pierced
by the ear ornaments.
In regard to over-all facial features, NS 1 differs quite sharply
from the other colossal heads. The protruding lower lip and the strongly
prognathous character of the face are the major distinguishing features.
The cheek bones are well-defined, and the lines about the nose and mouth
are somewhat less accentuated than on TZ 1.
The headgear resembles that of TZ 1 in form, although differing in
detail. The top of the headgear is made up of two distinct elements (see
sketch on following page). The topmost element may be described as roughly
rectangular, with a tongue flanked by two small projections extending from
the front edge. At the rear corners of the rectangle are two notched pro-
jections. The second major element is ovate in shape, with the narrow end
toward the front of the headgear. A pinched or waisted effect gives the
narrow end of the element a tongue-like appearance.
Seven braided strands, probably representing hair, extend from be-
neath the rear edge of the headdress element, over the lower element, and
down the back of the head to a point midway between the base of the head
band and the bottom edge of the head. Each braid terminates in a button-
and-tassel element. The head band is plain and narrows slightly toward the
front of the head. There is a small rectangular projection on the lower
edge of the band, directly over the nasion.   As on TZ 1, the chin strap ex-
tends from the lower edge of the head band to the base of the head, passing
under the ear ornament, next to the cheek. The ear ornaments are similar
32
Top of NS 1 (schematic). Head faces toward bottom of page.
to those on TZ 1, although the upper edge of the forward portion of the NS 1
ornament is slightly convex whereas the upper edge of the TZ 1 ear spool is
straight.
There has been no apparent intentional defacement of NS 1. A large
slab has been knocked off the lower left rear edge of the head, but it is
impossible to determine whether this took place before or after the head was
fashioned.
In summary, NS 1 is much like TZ 1 in several specific features.
These shared traits include nasion, eye form, style of ear ornament, head
band, and stone source. It differs in its marked degree of prognathism,
nose form, and specific features of the headgear. On the basis of these
similarities, and in spite of the differences, we feel justified in placing
the NS 1 and TZ 1 heads in a single stylistic group, separate from the La
Venta and San Lorenzo heads. Both are made of stone secured from the same
locality about five miles distant.
33
San Lorenzo Colossal Head No. 1
'SL lu(figs. 13, 14, 15a, 15b; pls. 16, 17, 18a, 18b), one of the most
impressive and the heaviest of all the heads, was known at the time of its
discovery by Stirling (1947:159) as "El Rey." It was found tying face u.qn
the sid     asteeap     ..a.ine some 300 yards to the southeast of the principal
mound group (ibid. 1955:9-11).
The head weighs 25.3 tons, is 2.85 m high, has a circumference of 5.9
m, and a frontal breadth of 2.11 m. The stone from which the piece was
carved was from the Cerro Cintepec basalt source (Williams and Heizer 1965:
16-17).
In terms of physical features, SL 1 seems to bear a subjective simi-
larity to the other heads from San Lorenzo. The nasion is of the subrhom-
boidal type, and exhibits pronounced fatty pads on either side. The eyes are
relatively large, and, being set at more of an angle, thus appear to be more
slanted than those of the other heads. Both eyes exhibit an iris in raised
relief.  Both inner eye corners have tear ducts, while the outer eye corners
come together at a point and do not overlap. The nose is flattish and well
executed, and is interesting because of the drilled pits which are present
on the upper side of the nares (see below). The mouth is open, with no
teeth showing, and the upper and lower lips are exedcut'ed' as bow-shaped, with
narrow raised ridges on upper and lower borders (pl. 34). A round pit is
drilled into either mouth corner at a slightly offset angle, aiming, as it
were, into the fleshy portions of the lower face to the sides of the mouth.
Both ears are present, and are probably the most realistically executed
ears of the twelve colossal heads. A prominent chin is shown between heavy
jowls which sag at either side.
The head of SL 1 is decorated with a sectioned head band which is
tipped back slightly. A curvilinear subrectangle appears in relief on the
front, directly over the nasion. This element is apparently in geometric
and decorative relation to the U-shaped element, with a quadripartite bottom
division which appears-directly above it on the headgear. At the back of the
head band, on the flat part of the rear of the head, is a knot like the one
on SL 2.
The head band is separated from the remainder of the headgear above
it by a shallow groove. The decoration of the headgear is weighted toward
the front, with the lines of the U-shaped element running parallel over the
top of the head and terminating in a subtrapezoid toward the rear. An
abbreviated chin strap appears on each side of the head, extending from the
head band and ending below an ear plug shown in profile.
34
Purposeful defacement of SL 1 consists of a number of pits with
dimples which have been ground into the face to depths up to 3 cm. A pair
of these pits appear to have been symmetrically placed, one on the side of
either nostril. A less visible, and perhaps modified, pit with a dimple
is present in the center cavity of each ear. The placement of these pits
in the ears is consistent enough with the anatomy depicted to suggest that,
at least in this case, such pits were used as part of the original sculp-
tural fashioning of the stone.
In general appearance, SL 1 is rather broad in frontal view, and
slightly more long and domed in profile appearance than the other San
Lorenzo heads. This effect is due to the extensive flattened area on the
back of the piece, which, although it has been damaged and recently repaired,
still shows signs of originally having been horizontally scored or striated
in the flattening process.   The heavy jowls, somewhat protruding at the sides
of the mouth, tend to subordinate the chin, giving SL 1 the appearance of
having an underslung jaw. Such physical variations as this, within the
sample of twelve heads, lend themselves to the interpretation of the heads
as individualized portraits.
San Lorenzo Colossal Head No. 2
When discovered, SL 2 (figs. 16, 17a, 17b, 18a, 18b; pls. 19, 20, 21a,
21b) was lying on its back, totally buried, on a trail leading south from
Tenochtitlan (Stirling 1955:10).   The stone from which it is made, like that
of the other San Lorenzo pieces, comes from Cerro Cintepec.
SL 2 weighs 20 tons, is 2.69 m in height, and has a circumference of
5.44 m.  The frontal breadth of the piece is roughly 1.83 m, while the thick-
ness is approximately 1.5 m.
Physically, with its heavy, sagging jowls and distinct, somewhat
pointed chin, the piece bears a rough resemblance to SL 1. The nasion is
of the subrhomboidal type and seems to have large fatty pads on either side.
However, these have been somewhat obscured by weathering of the specimen.
Both eyes have the iris depicted by flattening of the appropriate surface
area, the right eye displaying this feature much more distinctly than the
left. The corners of both eyes are pointed; in the inner corners, the upper
eyelid overlaps the lower, while at the outer corners no overlap is apparent.
The nose is broad and flat, and the openings of each nostril appear
as ground out bean-shaped depressions on either side of the rather wide
septum. The mouth is open and shows four faintly distinguishable teeth.
35
Both upper and lower lips are eroded and damaged, but it would appear that
the upper lip was probably sculptured in the shape of a shallow, inverted-U
or a bow, while the lower lip was probably straight and without a great deal
of modeling. There is no evidence at the corners of the mouth of the distinc-
tive circular drilled pits which characterize a number of the heads, but
there are bean-shaped depressions which were obviously made by grinding.
The head band is placed on the head at a slightly oblique angle, and
is decorated with three zoomorphic elements in applique. These are probably
stylized parrot heads (see figure) and are positioned directly in the front
center and on the two front temporal sections of the head band. Parrot, or
possibly owl, heads have been found on sculptures in other areas of Mesoamerica,
especially Teotihuacan (Bernal 1963:48). Moreover, bird motifs, mainly of
raptorial species, are not unconmmon to Olmec art (Drucker 1952:194-195; Coe
1965c:14). In fact, there is no reason to assume that the three-clawed elements
in the SL 5 and LV 4 headdresses are not bird claws rather than jaguar claws,
as is usually assumed.
Reconstruction of parrot head on center of SL 2 head band.
A knot, very similar to the one on
SL 2. It is placed slightly to the left
SL 1 appears
of center.
on the flat back of
The head band appears to be directly attached to the remainder of the
headgear, which appears as a lattice of cross-hatched bands, evenly distrib-
uted over the entire head above the head band. The chin straps are abbrevia-
ted and end slightly above the ornaments which decorate each ear. These ear
36
plugs appear as squares with holes, the left square divided by a number of
indistinct straight lines and the right square apparently undecorated.
Several types of purposeful defacements may be seen on SL 2.    On the
front of the face there are a number of the curious drilled pits with dim-
ples.  These appear on the face itself, as well as on the front of the head
band. Some of these pits appear to have been arranged in a crude sort of
bilateral symmetry, with one member of each pair appearing in the same
position on either side of the face.
An L-shaped gouge some 4.8 cm in depth, and apparently purposefully
inflicted, appears on the head band above the right ear of SL 2. Most
peculiar of all the defacements are the seven large rectangular cavities or
squarish pits which have been excavated in the flat back of the head.
These appear to have been carved out as units, and range from 2.5 to 14.4 cm
in depth. There has been some speculation concerning their possible func-
tion, but in themselves they seem to afford no clues. Similar defacements
appear on San Lorenzo Monument 14 and La Venta Altar 4.
In frontal view, SL 2 is rather broad-faced and resembles SL 1.
From the side, SL 2 does not appear as long as SL 1, but it exhibits an
impression of squareness or depth. The back of the piece is remarkably
flat, and horizontal striations are plainly visible. It would appear that
the flattening was done after the excavation of the seven large rectangular
cavities referred to above. If anything, this discounts the possibility of
their ritual or religious use during the carving of the heads, since the
flattening of the back was the final stage in the completion of all the
heads.
San Lorenzo Colossal Head No. 3
SL 3 (figs. 19, 20, 21a, 21b; pls. 22, 23, 24a, 24b, 33a) was discovered
lying face down on the side of a ravine not far from the other San Lorenzo
eads (Stirling 1955:11).    Stirling (1965:735) points out that this, as well
as the other five San Lorenzo colossal heads, had anciently been r~emoved
from its original position and been dumped down into the ravine.    It is
smaller than the first two San Lorenzo heads reported here, weighing 9.4
short tons.  It is 1.78 m in height and 4.02 m in circumference.    It has a
frontal breadth of 1.63 m, and is 0.95 m thick. The basalt for this head
also came from Cerro Cintepec.
Like the other colossal heads, SL 3 has no eyebrows.    The nasion,
unlike SL 1 and SL 2, is of the double variety, and the fatty pads present
37
on the latter two heads are absent on SL 3.
Irises are visible in both eyes; they are flattened by pecking and
surrounded by an incised line. The outer eye corners show tear ducts and
the eyelids meet at a point, while in the inner corners of both eyes the
upper eyelid overlaps the lower. The mouth is open, with slightly parted
lips, but no teeth are shown.  The upper lip is bow-shaped; the shape of
the lower lip is uncertain due to its having been damaged. No circular
drilled pits are now apparent in the corners of the mouth nor in the area
of the nostril holes. SL 3 displays a prominent chin and cheeks, but some-
what subdued jowls.
The head band is composed of four horizontal, parallel bands of equal
diameter (pi. 33a). The bands are plain and set on the head at a slightly
tilted angle, the front higher than the back.   They are sectioned, or
divided, by a number of short, straight, diagonal incisions which give the
whole head-piece the appearance of being twined. No decorative elements
which show applique are present. The back of the head is rather worn but
enough remains to determine that no indication of a knot, such as occurs on
'L 1 an1d ,SL 2jipresents   The upper portion of the headgear appears to be
a continuation of the head band, with a cap of twined cord or rope apparent
toward the top of the head. A geometric design, consisting of three or four
diagonally continuous small triangles, appears on the headdress. Four paral-
lel vertical bands hang from either side of the headdress, overlapping the
head band, covering the ear area, and possibly serving as a complete chin
strap. Due to weathering, it is impossible to be certain of this last point.
In general, the headgear elements are evenly distributed and symmetrical in
relation to the entire figure.
Defacement of SL 3 consists of a number of circular drilled pits with
a dimple in the center of each. These are placed above the lower limit of
the head band and on the front and top of the head. No symmetry is apparent
in the placing of these pits.
In general appearance, SL 3 is unusual in that, from a frontal view,
it appears to be almost pointed at the bottom and heart-shaped in general as
a result of the pointed chin and lack of imposing jowls. In profile view,
the head appears to be small and flat. This is due to the reduced facial
relief and the extreme flatness of the back. As with the other San Lorenzo
heads, horizontal smoothing g'-ooves are visible in the back of the sculpture.
38
San Lorenzo Colossal Head No. 4
(j4 (figs. 22, 23, 24a, 24b; pls. 25, 26, 27a, 27b) was discovered
by Stirling during his field work in 1946. It was lying buried on its
right side in the jungle some "600 yards northwest of the main mound'
(Stirling 1955:11, pl. 9).   Weighing six short tons, this piece is the
lightest of all the known colossal heads.   Its stand 1.78 m high, has a
circumference of 3.8 m, a frontal breadth of 1.17 m, and a depth of 0.95 m,
The source of the stone used for this head was Cerro Cintepec.
The nasion of SL 4 is subrhomboidal, with an incised subtriangle
prominently visible in the center.   An iris is present in both eyes, made
by flattening of the iris area and circumscribing with a shallow incised
line. Each eye has relatively rounded inner and outer corners. The upper
and lower eyelids are flat and wide, making the face appear to have pouches
under the eyes. /The nose is massive and well executed, the nostril-s-having
been made by drilling round pits into the stone and then modifying them-by
pecking. /The mouth is open and there are no teeth showing. The lips are
quite well preserved, the upper being curled into a bow-shape, and the lower
being thick and relatively straight. Round pits modified by grinding appear
in both corners of the mouth. /The chin is small -yet noticeable, and is sur-
rounded by the rather puffy, or fleshy, portions of the jowls. One ear, the
right one, is present on the sculpture.    It has suffered a good deal of
weathering, but is nonetheless interesting as it contains a circular drilled
pit in its center, obviously intended as part of the general sculptural
treatment of the stone.
The head band in the left front and right side areas of SL 4 appears
as four parallel horizontal bands, sectioned with diagonal incisions and
giving an over-all twined appearance. On the right front section the head
band is overlapped by another portion of the headgear consisting of eight
vertical parallel bands, divided into three groups, each of which hangs from
a circular element near the top of the head in a manner suggesting that the
bands represent hair. The chin strap on the right side of the face consists
of two vertical bands with diagonal incising, similar to the elements which
make up the head band and probably representing a downward extension of it.
On the left side of the head, where the ear is covered, three clip-like
elements occur, suggesting, perhaps, some sort of decoration worn in the
hair. These are almost identical to the elements on the front of a carved
stone "celt" now in the Villahermosa Museum (illustrated on page 39).
Westheim (1965, pl. 47) erroneously ascribes this specimen to the La Venta
site and incorrectly labels it a "ceremonial knife." From the right ear of
SL 4 hangs an ornament consisting of an undecorated disc, below which is a
tassel very similar to those on either ear of SL 5.
39
Elaborately carved "celt" from Ejido Ojoshal, Municipio Cardenas,
Tabasco (near Sanchez Magallanes). Length ca. 9 inches; material,
a very dense, hard black stone. Now in Museo Regional, Villahermosa.
Deliberate defacement of SL 4
pits with dimples placed unevenly on
of the head. It is of interest that
portion of the piece.
consists of some twenty-five drilled
the flattened portion of the back part
there is no defacement on any other
In general appearance, SL 4 is broad-faced and exhibits a rather high
domed head. The face is very finely sculptured, with a delicate and natural
approach to detail. The headgear is executed in a much more angular style
than in the other San Lorenzo heads, with detail decreasing on the sides as
one moves toward the rear. It would therefore appear that the head was made
to be viewed full-face, or nearly so, as tht profiles present a somewhat
incomplete appearance.
40
San Lorenzo Colossal Head No. 5
ASLA (figs. 25, 26, 27, 28; pls. 28, 29, 30, 31) was discovered in
a ravine south of the main         in    on its face (Stirling 1955:12).
The headlth     l.6 short tons and is 1.86 m high.    (When on exhibit at
the Los Angeles County Museum of Art in the Fall of 1964, it was erroneously
reported as weighing only 6 tons.) It has a circumference of 4.6 m, a
frontal breadth of 1.47 m, and a thickness of 1.15 m. The basalt from which
the head is carved is from the Cerro Cintepec source.
The nasion is of the subrhomboidal variety and displays massive fatty
pads on either side. These pads are much more angular than those of the
other heads with such pads, especially SL 1, whose pads give the appearance
of having been molded much more softly. Both of the eyes of SL 5 display
irises depicted by having been incised and flattened in the same manner as
those of SL 2, SL 3, and SL 4. The outer corners of the eyes are rather
pointed and show no overlap of either upper or lower lid. The inner corners
show tear ducts. The upper eyelids appear more fleshy and swollen than those
of the other heads. The nose is large and flat, and the nostrils are circu-
lar pits without dimples. The mouth is open but no teeth are displayed. The
upper lip is bow-shaped and slightly damaged in the center. The lower lip is
also bow-shaped. Drilled circular pits are present in the mouth corners.
These pits are without dimples and are positioned at oblique angles to the
lips, pointing, as it were, slightly outward. The chin is pointed and round,
and has a drilled pit with a dimple directly in its center. This positioning
suggests that it is a decorative feature. Both ears are present, although
neither one is done with a delicacy and modeling equal to that of the facial
features. The jowls are massive, giving the face a very square and harsh
appearance.
The head band, although somewhat eroded, appears as a series of curvi-
linear sections.  On the front, there is an inverted V-cleft directly over
the nasion. The band is seated on the head at an oblique angle, tipped
slightly above the horizontal toward the front of the head. Overlapping the
front of the head band from the upper part of the headgear are two three-
clawed elements which are thought by some to be jaguar claws (Stirling 1955:
12), but which, in our opinion, may be bird claws. These claw elements con-
stitute the front portion of the headdress. The remainder consists of a
pattern of cross-hatching or latticework, perhaps intended to represent
pieces attached by sewing to the body of the upper headgear. Geometric design
elements are present in the form of circular incisions on the sides and back
of the headdress. In general, the decoration is weighted toward the front.
The head band of SL 5 appears to be wider than that of any of the other
heads, and is probably an integral part of the more decorative portion of
the headgear above it. The chin straps are abbreviated and undecorated.
41
Covering the lobe of each ear is a disc with a tassel trailing beneath it.
This same type of lobe plug is present on one ear of SL 4.
The only marks of purposeful defacement on SL 5 are two drilled
circular pits with dimples on the face. One, as noted above, is placed in
the center of the chin; the other is located just below the outside corner
of the right eye. The back of the head is wholly free of defacement.
Considering its general appearance, it would seem that SL 5 was
intended to be viewed from all angles. This is indicated by the fact that
the head band continues all the way around the head, and also by the presence
of decoration over the entire area covered by the helmet or headgear, includ-
ing the back and sides. The head is angular and stern in appearance. The
large jowls, squared off near the bottom, contribute to the massive effect of
the full-face view.  In profile, the back of the head appears flat, and,
indeed, has been deliberately flattened by horizontal striations. The chin
appears to protrude beyond the jowls and is slightly forward of the prominent
cheek bones, giving the appearance that the head tends toward prognathism.
San Lorenzo Colossal Head No. 6
The designation  j6has been assigned to the most recently dis-
covered of the colossal heads at San Lorenzo (fig. 29; pls. 32, 36). This
head was first described by Aveleyra (1965:12-14), and was found buried
face down on the 5sQpgof a ravine to the west of the ceremonial zone.
M. Coe (1966:3) designates the head as Monument 17.
SL 6 still lies half-buried in the ravine where it was first found.
All the other San Lorenzo heads have been removed from the site. It weighs
between eight and ten short tons (Aveleyra 1965:14). It is 1.67 m high,
1.26 m wide, and 1.41 m long. It is believed that the basalt for this
sculpture, like that of the other San Lorenzo heads, came from Cerro Cin-
tepec (H. Williams, personal communication).
SL 6 displays a double nasion, roughly resembling a divided rect-
angle. Both eyes exhibit irises which are indicated by incised grooves.
It is interesting to note that in this head the iris does not consist of a
full circle; instead the eyelids truncate top and bottom of the iris, giving
the impression that the eye is partially closed. The pupil does not seem to
be as flattened as on the other heads from this site. The upper eyelid over-
laps the lower one at both the inner and outer corners. Both corners are
slightly pointed.
42
Truncated iris in eye of San Lorenzo No. 6
The nose is wide and very flat in profile. A drilled circular pit
with a dimple appears directly in the center of the bridge. The nostrils
consist of ground circular pits without dimples. The mouth is open, with
no teeth showing. It is difficult to determine the precise shape of the
lips as Aveleyra's photographs (1965, photos 18, 19) do not stress this
detail, and when the present authors saw the head in February, 1967, it was
still half buried. Circular drilled pits are present in the corners of the
mouth. They are drilled straight back into the mouth. The ear on the right
side is relatively well executed. The cheeks are unusual in that they are
extremely flattened and thus exhibit a peculiar stylization not present in
the cheeks of any of the other colossal heads.
The head band is placed horizontally on the head and is character-
ized by an unusual flatness in front. The frontal portion is divided into
two sections by a vertically incised line directly over the nasion. The
portion of the head band beginning above the right ear and extending to the
rear of the head is sectioned into four parallel corded or twined bands.
These are quite similar to the quadripartite bands on the headdress of SL 4.
The remainder of the head band is not elaborately adorned, and there is no
knot present at the back. The rest of the headgear of SL 6 appears to be a
series of cross-hatched lattices with some large circular elements contain-
ing small depressed interior circles present. The chin strap is undecorated
and extends the full length of the face. An ear ornament of roughly cylin-
drical shape is attached to the lobe of the right ear. It is very similar
to the ear spools on TZ 1.
There is very little purposeful defacement present on the exposed
portion of SL 6. The only apparent defacements are three drilled pits with
dimples on the face.   As noted above, one of these is in the center of the
bridge of the nose, and the other two are placed next to each other just
under the right eye.   It is of interest that a single similar pit appears in
roughly the same position on SL 5.
The most interesting aspect of the general appearance of SL 6 is the
use of flattening to achieve sculptural effect. The front of the head band,
the nose, the cheeks, and to some degree, the lips, appear as very flat.
43
This is striking in full front view as well as in profile. Of all the
heads, SL 6 appears to be closest to the stylized anthropomorphic jaguars
common in portable Olmec art.   The everted lips, the flat face, and the
narrowness of the integumentary upper lip (in this case 3.5 cm, the
narrowest of all the colossal heads) all contribute to this effect.
44
STYLISTIC COMPARISON OF THE TWELVE COLOSSAL HEADS
Interest has been aroused in recent years in the possibility of
seriating the Olmec colossal heads, with the intention of arriving at
some sort of chronological scale upon which they could be relatively
placed. This interest may have been generated by the fact that until
recently only one of the sites at which the heads occur (La Venta) has
been adequately dated (Berger, Graham and Heizer 1967).    Since these dates
are earlier than a number of scholars had expected, the relative dating of
two other important Olmec sites has become a matter of prime interest, and
seriation of the colossal heads provides one means of guessing at such a
relative dating. The heads have the added advantage-especially for art
historians-of inherently implying certain developmental traits, so that
not only could sites be relatively dated but the whole ontogeny of Olmec
art could be hinted at through their typology (Kubler 1962:65-69).
Another factor which may have prompted attempts at seriation was
the obvious similarity which the heads at any given site bear to each
other. So striking are these resemblances that even a comparatively un-
trained eye is quickly able to identify the site of origin of a head on
subjective impression alone. This subjective impression has a basis in
empirical fact. For example, one need not be an anthropologist or art
historian to immediately recognize that the La Venta heads are much more
weathered and eroded than the others, or that the Tres Zapotes heads appear
to have been sculptured from a different, more darkly colored stone than
the heads from La Venta and San Lorenzo.   Unfortunately, such observation,
albeit grounded in fact, imparts little of the chronological, cultural, and
technical information that we would like to have for the Olmec heads.
George Kubler (1962:65-69) was the first scholar to attempt a
serious seriation of the heads.   Before him, Miguel Covarrubias (1957:79-83)
had fitted them as a group into a general developmental scheme of Olmec art.
Kubler's seriation is based upon a rough grouping of the heads by shape.
He classes TZ 1 and LV 1 as round and early.   LV 2, LV 3, LV 4, and SL 2,
as a group, occupy a middle position in the sequence. Although LV 3 and
SL 2 are long-headed and LV 2 and LV 4 are round-headed, the group is given
cohesion by the fact that all four exhibit "parted lips" (Kubler 1962:67).
Kubler's third group, dating latest in the series, includes one round head
(SL 3) and three long heads (SL 1, SL 4, and SL 5).
Kubler's categories are based upon a subjective impression of each
head. As the following grouping of the individual colossal heads makes
apparent, subjective evaluations of "round-headed" as opposed to "long-
45
headed" are often influenced by such factors as the position or angle from
which the observer views the sculpture. In fact, any seriation based upon
such an impressionistic approach is likely to measure nothing more than the
positions from which the observations were made. In an attempt to quantify
Kubler's approach to the seriation problem, we have calculated the length
of each individual head, divided by the height. The results are shown in
Table 1 and in Graph 1, and express head shape in mathematical ratios.
Table 1
Profile Ratios of Head Shapes
(in cm)
* La Venta      || NS    TZ              San Lorenzo
II  1  1  2    3  I  4                 II  1  II  1  II  1  2  3  I  4  I  5  6
Lengthl| 1951   981 1001 186 |126 1145       1631  1501   951   951 115    141
Height |241 | 163    198 j 226 || 145  147   285 | 269 | 178 | 178 | 186 j 167
Ratio  11.809 1.601 1.505 1.823 11.869 11.980 11.572 1.558 1.534 1.534 1.618 1.844
The figures in Table 1 and Graph 1 indicate that, from a profile view,
there are two groupings of heads. LV 1, LV 4, NS 1, TZ ,. and SL 6 are all
relatively "thick," or round  heads. Their ratios range between .809 and .980.
LV 2, LV 3 SL 1, SL 2, SL 3, SL 4   and SL 5 are all relatively "thin   or
long, heads. with ratios ranging between .505 and .618. Considering that
Kubler was unaware of the existence of NS 1 and SL 6, it is of interest that
in only one case, that of LV 2, is there a disagreement between Kubler's
estimate and our figures. He feels that LV 2 is round-headed, while we con-
sider that it clearly falls into the "thin," or long-headed, category. Our
groupings are compared with Kubler's in Table 2.
Table 2
Comparison of Kubler's Head Types with Those of Present Authors
11l         Round            II                  Long
Kubler || LV 1, LV 2, LV 4, TZ 1, SL 3 || LV 3, SL 1, SL 2, SL 4, SL 5
This   | LV 1, LV 4,TZ l, NS 1, SL 6 I LV 2, LV 3, SL 1, SL 2,5 SL 3, SL 4, SL 5
paper ||                           II
, TZ 1    T
* NS 1
SL 6,
, LV 4
* LV 1
. SL 5
. LV 2
.SL 1
*SL 2
, SL 4   . SL 3
LV 3
Thick
G)
60
co
Ad
Thin
Graph 1. Plot of length/height ratio
46
100
951
90
851
801
75.
70
65
60 4
55,
50,
100t,   . TZ 1    -
, LV 4
oNS 1
, SL 3
* LV 3
. LV
, SL
* SL
. SL
1LV
5
1
6
70  a SL 2
, SL 4
r
0)
co
L
Round
2
*1"*
a)
C4
Narrow
Graph 2. Plot of breadth/height ratio
47
95.
85.
80.
75.
65
48
Unfortunately, the correspondence between our grouping and Kubler's
holds only so long as we are considering a profile view of the heads. If
we take a frontal view of the sculptures, and divide the frontal breadth,
measured between temporals, by the height, we are presented with ratios
that give us an entirely different grouping. As Table 3 and Graph 2 indi-
cate, LV 3, LV 4, NS 1, TZ 1, and SL 3 fall into the extreme group of round
heads, while the remaining heads can be categorized as more or less long.
The groupings are summarized in Table 4.
Table 3
Frontal View
Ratios of Head Shapes
(in cm)
La Venta         NS    TZ|            San Lorenzo
Al  1  1  2  3  I  4  Il  1  1  Al  1  I  2  j  3  I  4    5  j  6
Frontal   208   135   160 | 198   134 || 150  217 j 190  160 | 117 j 147 j 126
width    |    l   l     l                 Il Il            I     I     I
Height    241 | 163 | 198 j 208 || 145 jj 147  285 | 269 | 178 | 178 | 186 j 167
Ratio   11.821 1.822 1.869 1.952 11.924 111.02 11.761 1.706 1.899 1.657 1.790 1.754
Table 4
Groupings by Head Shape, Frontal View
Round            l                 Long
LV 3, LV 4, NS 1, TZ 1, SL 3 11 LV 1, LV 2, SL 1, SL 2, SL 4, SL 5, SL 6
Thus we can justify quantitatively our earlier statement that subjec-
tive impressions are often a result of the observer's viewing angle. This
is particularly evident in the wild fluctuation presented in the two ratios
of SL 3. As Tables 1 and 3 indicate, SL 3 varies between .534 in profile
to .899 in frontal view, the difference being entirely accounted for by the
change in the position of observation. It is probable that Kubler's desig-
nation of "long heads" was derived from a consideration of the heads in
profile. The same description might obtain from a bird's-eye or top view
of the heads, but certainly not from a full-face assessment.
49
Even if we were not faced with this type of difficulty in a subjec-
tive approach to grouping the heads, such categories-head shape in par-
ticular -are ambivalent as to the types of cultural inferences they allow
the investigator to make.
Let us assume for a moment that the groupings of heads by shape did
not vary according to the position from which they were viewed. That is.
let us consider the twelve heads as neatly segragated into two groups -one
round and the other relatively thin in appearance. What, then, would this
mean in terms of cultural differences? Is it to be assumed that we are
dealing with two distinct ethnic types? Or is the grouping due to a differ-
ence in the ideal standard of beauty, or in the stylistic preference of
different groups of artists? Could the apparent differences in head shape
be nothing more than a function of the size and shape of the original
boulders available to a given group of sculptors? Do these differences,
and the reasons for them, actually register temporal trends? Or are they
a reflection of varied technological or artistic ability on a synchronic
level? Obviously, a satisfactory answer to any of these questions would
require use of evidence other than that provided in a subjective segregation
of the heads by shape; a number of the decorative and stylistic elements
appearing on the heads should also be considered. In addition, certain
assumptions might be made dealing with the types of inferences one would
hope to gain in a consideration of the various decorative elements repre-
sented in the heads.
In the discussion which follows we have assumed that there are two
basic types of inferences which can be made regarding Olmec culture on the
basis of the various elements, and groupings thereof, which are represented
in the heads themselves. The first level of inference is derived immedi-
ately from empirical description, and is concerned with objective identi-
fication of features which relate directly to the physical appearance of
the entity or individual portrayed in the sculpture. This would include
such elements as type of headgear, shape of nasion, ear ornament, etc.
In other words, this level of inference is one which assumes that the
sculpture constitutes a relatively accurate portrayal of a real cultural
element, and that the element as depicted in the sculpture is used in the
same way as it was in the real life model from which it was copied. For
example, we would assume that at least one person in Olmec society actually
wore ear lobe plugs such as those which we see carved on SL 1.
The second type of inference is one which allows us to make state-
ments which reflect not so much upon the particular head or heads under
consideration as upon the artist or artists responsible for the production
of the heads.  For example, as we explain below, there are three different
ways in which the iris within the eyeball is depicted in the twelve indi-
50
vidual heads. Since we can assume that all eyeballs of living bearers of
Olmec culture were, for all practical purposes, anatomically identical, it
follows that these three modes of representation are a result of at least
three different sculptural techniques for portraying irises available to
Olmec sculptors during the period in which the colossal heads were carved.
We first group the head,; according to elements which afford us a
direct inference about the object portrayed:
Ear ornaments. As Table 5 shows in summary form, the heads may be
divided into five rough groups on the basis of ear ornaments. LV 1 is
unique in having a diamond-shaped ear ornament, the exact details of which
are difficult to discern because of weathering. It is in all likelihood a
severely stylized Kan cross, an element which occurs on LV 4 and elswhere
in Olmec art (Coe 1965b:761, fig. 43g). The ears of SL 3 are covered by
extensions from the headgear, and the same is true of the left ear of SL 4.
The three remaining groups, however, are of greater significance and inter-
est. Tapered cylindrical lobe plugs are found in TZ 1, NS 1, SL 1, and
SL 6. It is worth noting that all of these sculptures, with the single
exception of SL 1, fall into the category of "thick" heads in profile, as
shown in Graph 1. This fact implies that there was more horizontal space
available on the lower part of the head for use in depicting ear ornaments,
which tend to be horizontally rather than vertically oriented. It is tempt-
ing to postulate in this regard that the type of ear ornament portrayed on
at least some of the heads was dictated not so much by ritual or rank of the
personage portrayed as it was a correlate of the space available on the vari-
ous boulders utilized.   Circular ear plugs are found on LV 3, LV 4, and S    2,
while SL 4, SL 5, and LV 2 display the ear button and tassel element.
Table 5
Ear Ornaments
Oblong     I Circular   I Disc and    I Diamond-shaped I  Ears Covered by
Cylinders   I Spools     j   Tassel    I    Spool      I Headgear Extensions
TZ 1, NS 1, I LV 3, LV 4, | SL 4, SL 5, |    LV 1      | SL 3, SL 4 (left)
SL 1, SL 6 I  SL 2      | LV 2        |               |
There are two additional points worth noting about the ear ornaments
as we have grouped them. First, observe that the groups may contain members
from Tres Zapotes, Nestepe, and San Lorenzo, or from San Lorenzo and La
Venta.  However, in this particular set of groupings, we find no sharing of
51
ear ornament features between Tres Zapotes, Nestepe, and La Venta.
Second, note that the ear ornaments which are represented are not
in any way unusual for Mesoamerican ear ornaments. This fact may provide
us with some insights when we consider the problem of exactly what kind of
personage is represented by each head.
Decorative elements on headdresses. Next we shall consider decora-
tive elements on the headdress or helmet.   As shown in Table 6, there are
five decorative elements by which we are able to group most of the heads.
Some of the heads appear in more than one group, while others do not appear
at all. LV 2 and LV 3 are not listed as they are too badly defaced to per-
mit identification of design elements in the headgear. TZ 1 and NS 1 are
almost identical in having plain cap-like helmets, and these are not noted
here. We have grouped LV 1 and SL 1 together on the basis of the possession
by both of a U-shaped element on the front of the headdress. This element
is a relatively well known Olmec motif and has been commented upon elsewhere
(Drucker 1952:204; Coe 1965b:759, fig. 43d).   LV 4 and SL 5 have been grouped
together because they both exhibit a claw element, LV 4 displaying one large
three-clawed motif on the frontal section of the headband, and SL 5 showing
two smaller three-clawed designs, one over each temporal area of the headdress
above the head band. These have been referred to by Stirling (1955:12) as
jaguar claw motifs. It is of interest that the U-shaped elements in the N
previous group (i.e. LV 1 and SL 1) are considered by Drucker (1952:204) and)
Smith (1963:138) to be stylized jaguar mouth forms. Knowing the high impor-
tance of the jaguar in Olmec cosmology (Drucker 1952; Coe 1965c:14), we do
not find it surprising that jaguar-related elements occur in some of the
headdresses of the Olmec colossal heads.
Table 6
Headgear Decorative Elements
U-shaped Element I Claw Element I Knots on I Cross-  I Four-part
on Head     I   on Head   I   Back  I hatching j  Braid
LV 1, SL I   I LV 4, SL 5 I SL 1,    I SL 2,    | SL 3, SL 4, SL 6
I             I  SL  2  |  SL  5,  l
l             1         1  SL   6  l
SL 1 and SL 2 are grouped together since they share a distinctive
knot element at the rear of their head bands. The knot gives the impression
52
that the head band is a strip of cloth tied in the back. A similar knot
appears on a breechclout on Monument 23, La Venta (Drucker, Heizer and
Squier 1959, pl. 52). Crosshatching is present on the headdresses of SL 2,
SL 5, and SL 6, and on this account they are grouped together in Table 6.
A four-part braided element, possibly hair or some type of twisted rope or
cord, appears in one form or another on SL 3, SL 4, and SL 6.   If this
element does represent rope or cord, we would say that SL 6 exhibits a
left-hand twist, SL 3 a right-hand twist, and SL 4 depicts both types of
twist. These are the smallest of the heads from San Lorenzo and, as fol-
lowing sections reveal, they tend to draw together repeatedly through
related characteristics.
Chin straps. Chin straps on the headgear may be divided into two
types: full, and abbreviated. TZ 1, NS 1, SL 4, and SL 6 show full chin
straps; while LV 1, LV 2, LV 3, SL 1, SL 2, and SL 5 have abbreviated ones.
Chin straps appear to be altogether absent on SL 3, while LV 4 is too
eroded to make designation possible. Notice that here again neither TZ 1
nor NS 1 is grouped with any of the heads from La Venta; and also, that
SL 4 and SL 6 are grouped together. These results are summarized in
Table 7.
Table 7
Chin Straps
Full       I       Abbreviated       I Absent  I Indeterminate
TZ 1, NS 1, SL 4,  | LV 1, LV 2, LV 3, SL 1,  I  SL 3  |    LV 4
SL 6            |   SL 2, SL 5            l         I
Nasions. In addition to the decorative elements discussed above, we
have chosen to assume that the nasion feature is one which provides a more
or less accurate rendition of what the artist actually saw as he carved the
large heads. We realize that a good deal of stylization may be present in
this element; if so, then we need merely to add it to the categories dis-
cussed below to obtain a different sort of inference than the one we pres-
ently choose to draw. Our conclusions will not be altered if the nasions
should be regarded as artistic conventions rather than faithful representa-
tions of live Olmec models.
The nasions fall into three categories, as shown in Table 8. LV 1,
LV 2, LV 4, SL 1, SL 2, SL 4, and SL 5 have roughly subrhomboidal nasions.
53
LV 3, SL 3, and SL 6 each have a double nasion.    Triangular nasions are
found on TZ 1 and NS 1. These groupings show a continuing empirical ten-
dency for the Tres Zapotes and Nestepe heads to associate as a unit.
Moreover, the La Venta heads again appear to group with the San Lorenzo,
but not with the Tres Zapotes or Nestepe, sculptures.
Table 8
Nasions
Subrhomboidal             I      Double      I Triangular
LV l, LV 2, LV 4, SL 1, SL 2, SL 4, SL 5 | LV 3, SL 3, SL 6 | TZ 1, NS 1
When we examine the heads for possible groupings on the basis of
categories which provide us with our second type of inference-about the
artist, his view of society, and his handling of his materials -we note
that some of the trends found in the first series of element groupings are
also present here. Our consideration of the sources from which the stone
for the heads was obtained is an excellent example.
Stone source. As may be noted in Table 9, there were three sources
for the stone used in the Olmec colossal heads. One of these was the "peak
formerly called Cerro Santiago but now known as Cerro El Vigfa" (Williams
and Heizer 1965:4). This extinct Pliocene volcano lies about eight kilo-
meters east-southeast of the site of Tres Zapotes and was the source of the
stone used in both the Nestepe and Tres Zapotes heads (ibid.).    To the
south, along the southern edge of the Tuxtla Mountains and near the locality
of Soteapan, lies the second source of the stone, another extinct Pliocene
volcano known as Cerro Cintepec (ibid. 5, 11, map 2).    We now know that the
stone for three of the La Venta heads (LV 1 - LV 3) and all the San Lorenzo
heads came from this source. (In addition, the so-called Monument de Ahosh
at San Lorenzo is made of Cerro Cintepec stone-H. Williams, personal commun-
ication.) Once again, we call attention to the fact that TZ 1 and NS 1 form
a group by themselves, and that the La Venta and San Lorenzo heads cluster
together on yet another feature.
Weight.  When grouped together by weight-measured by short tons.-the
association of the La Venta and San Lorenzo heads repeats itself; as Table 10
demonstrates, none of the four La Venta heads is grouped with TZ 1 or NS 1.
The latter two heads are grouped with SL 3, SL 4, and SL 6 since all five
weigh less than ten tons. We have already noted that SL 3, SL 4, and SL 6
54
Table 9
Stone Source
Cerro El Vigfa I         Cerro Cintepec         I Unknown Source in Tuxtla Mtns.
TZ 1, NS 1   | LV 1, LV 2, LV 3, SL 1, SL2, j               LV 4
|    SL 3, SL 4, SL 5, SL 6     l
share a number of interesting features (see tables 6 and 7), and, moreover,
that SL 6 shares a number of features with NS 1 and TZ 1 (see tables 2, 5,
and 7).  LV 2, LV 3, LV 4, and SL 5 are categorized together, all weighing
between ten and twenty tons. The final group consists of LV 1, SL 1, and
SL 2, all weighing over twenty tons. Implications of these weight groupings
are discussed in our general conclusions below.
Table 10
Weight in Short Tons*
Under 10 Tons            Between 10-20 Tons        Over 20 Tons
TZ 1, NS 1, SL 3, SL 4, SL6 6   LV 2, LV 3, LV 4, SL5 | LV 1, SL 1, SL 2
*
Exact weight may be found in Table 18.
Iris form. The category of iris form is of particular value because
it allows us to group the heads on the basis of sculptural technique. Such
a category is of importance in considering whether various heads were made
by the same sculptor or group of sculptors. The assumption behind this
category is that, since the different categories of iris form do not repre-
sent a great variation in technical ability of execution (i.e., it is no
more artistically difficult to depict an iris in raised relief than by
flattening and incision), then they must represent a conscious choice by
the artists as to how they wished the irises to be executed. Even if there
were a time difference reflected in the various categories, we would still
argue that the artists, with the exception of those who sculptured the first
heads, would have exercised choice since the earlier heads in any given site
would presumably still be visible for inspection and would reveal the vari-
ous possibilities for mode of execution.
55
As summarized in Table 11, there are four categories of iris form
evidenced among the colossal heads (LV 3 is too eroded to be classified).
SL 2 displays irises which are merely flattened, probably by a process of
grinding. LV 1, LV 4, and SL 1 are grouped together as the irises are all
executed in raised relief. The largest group, in which the process of
flattening and incision was used, consists of LV 2, SL 3, SL 4, SL 5,
and SL 6.  In the group in which irises are totally absent, we find TZ l and
NS 1 once again grouped together.
Table 11
Iris Form
Flattened   Raised Relief    Flattened and Incised j Absent    Indeterminate
SL 2      LV 1, LV 4,      LV2,SL3,     SL,       | TZl,    j     LV 3
SL I               SL 5, SL 6            | NS 1|
Mouth form. A second category in which is reflected a conscious
choice by the artist is that of the execution of the mouths. Choice in
this case is afforded by the fact that any living Olmec model probably had
teeth, and certainly would have been able to pose with mouth open or closed.
However, artistic ability may enter this category to some extent, at least
more so than in the category of iris forms.
As Table 12 demonstrates, we can distinguish three groups of mouths.
In the group with open mouths and teeth are LV 2, LV 3, LV 4, and SL 2.
With mouths open and no teeth showing there are SL 1, SL 3, SL 4, SL 5, and
SL 6. Finally, we have those heads on which the mouths are completely
closed, LV 1, TZ 1, and NS 1.   We have come to expect that TZ 1 and NS 1
will group together; however, in this case, they have been grouped with a
head from La Venta. If one assumes that some differential of technical
ability is involved in the execution of closed versus open mouths, then it
can be said that the artists who worked on TZ 1 and NS 1 were of the same
caliber as some of the artists at La Venta; namely, those who worked on
LV 1. Although the differences in competence between these groupings may
be slight, it is nonetheless interesting to note that at least two modes
are available at San Lorenzo and La Venta, while at Tres Zapotes and Nestepe
the same level of competence in representing mouths is displayed. Thus,
this maintains the trend of TZ 1 and NS 1 to form an empirically valid
grouping, since at La Venta as well as San Lorenzo two techniques of carving
the mouths are present while the Nestepe and Tres Zapotes heads show only
56
one mode of execution. It also presents the possibility that sculptors
with differential abilities were operating at two of the sites-La Venta
and San Lorenzo-where different mouth forms are presented.
Table 12
Mouth Forms
Mouth Open, with Teeth       Mouth Open, No Teeth          Mouth Closed
LV 2, LV 3, LV 4, SL 2| SL 1, SL 3, SL 4, SL 5, SL6 j LVl, TZ l, NS l
Narrow ridges outlining mucous lips.   As was noted in the descrip-
tions of the individual heads, a raised ridge appears on some of the lip
borders. We have grouped the heads on the presence or absence of this
trait, and the results are summarized in Table 13. TZ 1, NS 1, and SL 1
share the presence of these ridges, while they are absent on LV 1, SL 2,
SL 3, SL 4, SL 5, and SL 6.  We were unable to determine the presence of
ridges on LV 2, LV 3, and LV 4, although LV 4 may have had such a ridge on
the lower lip alone. As may be seen, these groupings are in line with
previous ones and present no new material for discussion. The element may
be unimportant, although we feel it to be worth noting that TZ 1, NS 1,
and SL 1 have similar ear plugs as well as having lip ridges in common.
Table 13
Narrow Ridges Outlining Mucous Lips
Present      I         Absent           I       Undetermined
TZ 1, NS 1, SL 1 I LV 1, SL 2, SL 3, SL 4, | LV 2, LV 3, LV 4 (lower?)
SL 5, SL 6
Purposeful defacement. The last element by which we attempt a group-
ing of the heads is that of defacement. Since a separate discussion appears
in this report on the defacement of the heads, and since most of the de-
facement probably took place after the heads were completed, we need not go
into great detail.   Suffice it to note, as we indicate in Table 14, that the
characteristic pits with dimples appear in one form or another on all of the
heads from San Lorenzo. LV 3 exhibits pits and dimples similar to those of
57
the San Lorenzo heads, as does LV 2. While an occasional or random groove
or gouge may appear on some of the San Lorenzo heads, it is the four La
Venta pieces which are characterized by an abundant an      itinct variety of
tD-e   ;  cWFements.  Square, block-like holes are found only on SL 2.   TZ 1
and NS 1, interestingly enough, form a group showing almost no defacements.
TZ I shows slight damage , while NS 1 is in no way defaced.
Table 14
Predominance of Defacement Types
I                    I Square I None or
Pits with Dimples       I Gouges and Grooves I Pits  j   Slight
SLi, 5SL 2, SL 3, SL 4,5L5, I LV 1, LV 2, LV 3,   | SL 2   | TZ 1, NS l
SL 6, LV 2 (few), LV 3 (few) I LV 4              1        I
Conclusions
As the above discussion has attempted to point out in quantitative
terms, it is apparent that the twelve colossal heads may be grouped into
several clusters on the basis of a number of combinations of elements. In
general, we might say that the TZ 1 and NS 1 heads combine to form a dis-
tinct group.  The sxSan Lorenzo heads      while generally constituing a
*~~~~ *v Ws
distinct unit ma be said to break down into two subgroups. SL 3 SL 4
and    6     4,J   grouped together on the basis o  a num er of sh     elements
and, with some regularity, these three heads share traits with the heads of
Nestepe and Tres Zapotes.   SL 1   SL 2  and SL 5 while all maintaining a
certain statistical distinction as San Lorenzo heads, seem to group more
often, although somewhat randomly, with La Venta, and less frequently with
Tres Zapotes and Nestepe, as opposed to a more constant and less random
clustering of the smaller San Lorenzo pieces.
The four La Venta heads   while easily distinRuishable as a separate
group, seem also to be divisible into two subgroups.     LV    an        rein
our opinion, rather different in many          an     2 and LV. This may be
due in part to the heavy erosion of LV 2 and LV 3, but as the tables above
indicate, it is also due to a number of empirically valid differences.
Thus a sort of chain may be visualized, with La Venta a distinct
group but sharing a significant number of ties with San Lorenzo.      San
Lorenzo is distinct as a group in certain respects, but in addition to the
58
ties with La Venta also shares a number of similarities with Nestepe and
Tres Zapotes.  These two latter heads form a third distinct unit, and
share only the most general features with the heads from La Venta, such
as similarities in mouth form.
In terms of these groupings, it is possible for us to separate the
heads into six subgroups on the basis of the above mentioned similarities
between certain of the heads. These subgroups are:
Group A |Subgroup    I - LV 2, LV 3
jSubgroup  II - LV 1, LV 4
Group B |Subgroup III - SL 1, SL 2, SL 5
|Subgroup  IV - SL 3, SL 4, SL 6
Group C jSubgroup   V - TZ 1
JSubgroup VI - NS 1
The groupings may be interpreted to mean that within the Group A
heads from La Venta we find two subgroups. The first of these (subgroup
I), LV 2 and LV 3, stands relatively alone and apart.   The second subgroup
(II), LV 1 and LV 4, shares certain similarities with the first subgroup
(III) in Group B, namely, SL 1, SL 2, and SL 5.   The second of the San
Lorenzo subgroups (IV), while primarily allied with the first (III), none-
theless is more similar to Group C's subgroups V and VI than are any of the
other subgroups. The heads from subgroup I are less like those of subgroups
V and VI than are any of the other groups.   In other words, our groupings
are intended to imply that the closer together any two subgroups are, the
more similarities the heads within each subgroup share with those of the
other subgroup. The farther apart the groups, the fewer the similarities.
Thus, subgroup I of Group A is least like subgroup VI of Group C. We must
stress that we intend these groupings to reflect only similarities of ele-
ments and their execution, and to have absolutely no temporal implications.
We are, of course, aware that with a scheme such as this the problem
of temporal or chronological relationships between the heads must receive
consideration. We have discussed Kubler's grouping and, by implication,
his chronology. Squier (n.d. 183, 255, 258a), writing before NS 1 and SL 6
were reported, guessed at a three period chronology which would place the
site of La Venta oldest, San Lorenzo intermediate, and Tres Zapotes youngest
in the sequence. If we assume that the heads are also in such a sequence,
then Squier's guess would correspond with that of M. Coe who has proposed
a similar chronology in which "the sequence would begin with the La Venta
heads, which show close resemblances to the jade and pottery figurine style;
then the San Lorenzo heads; and finally the Tres Zapotes ones, which are
extremely portrait-like and lack the pseudo-drilling at the corners of the
59
mouth"' (Coe, quoted in Bernal n.d., fn. 40).
While our evidence would force us to disagree with Coe on some of
his specific assumptions, we must at the same time admit that our group-
ings of the heads would not preclude the possibility of a general sequence
such as he proposes.  Wicke, on the other hand, has seriated the heads in
a different manner, one with which we find no agreement whatsoever. He
feels (Wicke n.d., 146) that the Tres Zapotes heads (our TZ 1 and NS 1)
are earliest, the La Venta heads next in the sequence, and the San Lorenzo
pieces latest. He bases this sequence on the opinion that the heads show
"the evolution of facial expression from a lack of animation to a smiling
countenance to an expression of serenity. In proportion to height, the
heads become steadily more narrow and more shallow. The eyes show a
development from no iris to a well defined one." As we have demonstrated
above, head shape is an inconclusive category for seriating the heads.
We feel it unwise, furthermore, to even consider such a subjective cate-
gory as "facial expression." Our Table 11 considers iris form in greater
detail than does Wicke, and for his statement that the full iris depicted
in the San Lorenzo heads "is a subtle refinement at the end of a tradition
within which artists strive to depict the eye with realism" (ibid., 133),
we find no evidence whatsoever. Wicke also dates the Olmec sites, feeling
that the "series should indicate that the center of political power in the
Olmec region shifted respectively from Tres Zapotes to La Venta to San
Lorenzo (ibid., 142). Moreover, he states that "Olmec culture probably
persisted at San Lorenzo long after La Venta was abandoned." We now know,
from Coe's recent radiocarbon dates (1967:1399-1401), that San Lorenzo is
at least as old as La Venta. Thus, on a number of grounds, we are forced
to disagree strongly with Wicke's conclusions.
Our groupings are based upon a number of differences which we have
shown to occur in a relatively patterned manner within the heads. In
aaitio-a to thaese viwtltes, whc me 't.ia"e oi course st-essed, me ShoUI&
mate that  nt %t-ovxt    also tak&e i\iciUAt accv0ount o'i a tnmab  oiD eT.Aat-
ities which prevail among the heads.   Moreover, the heads share a number
of general features which we have not referred to in the tables. For
example, all the heads have headgear of a similar type, often referred to
as "helmets" (Stirling 1965:733). Moreover, all the heads are character-
ized by flattened backs (ibid.). All have a prominent nasion, as well as
a distinctly fashioned chin. Cheek bones and facial modeling, while de-
picted differently on each head, are nonetheless present on all. None of
the heads exhibit eyebrows, probably because they are covered by the head
bands.
Despite the differences between heads from the three sites, we feel
that all show sufficient similarities, considering the present state of
60
the Olmec archaeological record, so that it is not unreasonable to argue
that all twelve heads are roughly contemporaneous. We cannot agree on a
time period, but suggest a span of a century or two at most may be in-
volved (cf. Kubler 1962:67). This implies that San Lorenzo, Tres Zapotes,
and La Venta were occupied at the same time and that the heads were sculp-
tured and emplaced at each site during the same period. While our evidence
does not deny the possibility of such a sequence as Coe has proposed (some
scholars may feel that it enhances such a sequence), we feel that there is
insufficient evidence to warrant a definite commitment to any chronological
ordering of the Olmec colossal heads. Since the archaeological sites where
the heads were found have not been demonstrated to be mutually exclusive
temporally, and since the heads are not firmly tied into the known archaeo-
logical sequence at any of the sites, it seems to us that chronological
guesses would have to be based upon artistic traits alone rather than arch-
aeological fact. As we have stated earlier, we do not consider that the
segregation of heads by artistic traits lends itself to a temporal arrange-
ment.
For ease in comparison, we have included Table 15 which relates in
summary form the chronological distinctions of Kubler, Wicke, and Coe, all
of whom have directly discussed temporal relationships of the heads. In
terms of our groupings by element similarities, we find most agreement with
Coe, who would class our subgroups I and II as early , our subgroups III and IV
as intermediate, and our subgroups V and VI as latest in his sequence.    As may
be seen, this seriation tends to agree with our conclusions on stylistic
similarities within the heads.   Wicke, with whom we have little agreement,
would almost reverse the order of the heads, classing our subgroups V and
VI as earliest, our subgroups I and II as intermediate, and our subgroups
III and IV as latest. Relying upon a subjective approach, Kubler has done
most violence to the groupings as we see them, and his categories bear little
resemblance to ours.   He feels, in our terms, that subgroup V and part of
subgroup II are early, parts of subgroups II and III and all of subgroup I
are intermediate, and parts of subgroups III and IV are latest.
Artistic traits and decorative elements do not, in our opinion, serve
exclusively to define chronological distinctions but lend themselves to
other types of interpretation. Not only have the groupings of elements
which we have employed provided us with empirical reasons for grouping the
heads by site, but they have also shown which heads within a site maintain
close similarities to each other and to particular heads from other sites.
To account for some of these similarities, we suggest that groups or schools
of sculptors were at work, and that within these schools there were artists
of varying abilities.
A few tentative suggestions as to what these artists were portraying
61
Table 15
Summary of Other Opinions on Relationships of Olmec Heads
|      Oldest       |   Intermediate    j   Most Recent
Kubler 1962  j TZ l, LV I          LV 2, LV 3, LV4, j SL I, SL 3, 5SL 4,
SL 2                SL 5
Wicke (n.d.)   TZ 1, NS 1          LV 1, LV 2, LV 3, j SL 1, SL 2, SL 3,
LV 4                SL 4, SL 5
Coe (n.d.)     LV 1, LV 2, LV 3, 5  SL 1, SL 2, SL 3, ITZ 1, NS l
LV 4                   SL 4, 5SL 5
as they carved the colossal heads are offered here. We feel that the
sculptors were depicting human figures, not stylized deities.  Although
some stylization is apparent, there is more than enough individuality of
expression, musculature, etc., to assume that the heads were modeled after
living persons. Whether or not these are actual portraits, we do not know.
A number of writers have suggested that they are (Stirling 1955:20-23; 1965:
721; Kubler 1962:67; Smith 1963:128). Others (Armillas 1964:304; Bernal
n.d., 62) have felt that they are not portraits but are idealized represen-
tations of living persons, possibly chieftains. We feel that the question
is probably unanswerable, and our data could be used to substantiate either
claim. The important point, in our opinion, is that people, not gods were
represented in the heads.
The ornamental elements upon the heads may have been chosen for
their purely decorative value, perhaps even by the fancy of the artists.
We have shown that the shape of the ear plugs, for example, was possibly
related to the amount of space available beneath the ear on which the orna-
ment was to be imposed. None of the other decorative elements seem to be
overburdened with sacred connotations, and since no serious opposition to
the idea exists in the literature on the topic, we venture to guess that
the ornamention and execution of the heads was primarily secular. Any
stylization which exists probably reflects more on conceptions of ideal
physical type than on ritual connected with the sculpture. The fac that
all of the purposeful defacement on the heads seems to have taken place
after their coimpletion also air>es against ritual or religious connection
with their execution.
Since our conclusions grant that the heads could all be relatively
contemporaneous, we must assume that the technology of stone transport did
62
not vary greatly during the time the heads were being carved. We conclude,
then, that the element of weight variation within the heads has absolutely
no bearing on the technological ability of the Olmec transport experts.
That is, we feel that the weight of the heads, as well as the shape of the
boulders on which they were carved, was dictated by artistic preference
rather than varied technological ability in the movement of heavy pieces.
We are, in fact, surprised by the wide choice which the artists who
carved the colossal heads exercised in the selection and execution of a
great number of elements which, in the more rigidly stylized types of Olmec
art apart from the heads, would have been subject to almost no choice in
their rendition.
63
SCULPTURING TECHNIQUES
It is appropriate at this point to include a brief description of
the method used by Olmec artisans in sculpturing the colossal heads. There
are but few published references to this aspect of the problem, most writers
being more concerned either with describing the heads or with analyses in
terms of art style, function, dating, or seriation.   There are some excep-
tions, among them Kubler (1962:69), who refers tog the possibility that pos-
tures and expressions first depicted in clay figurines may have been trans-
ated Tto jade a     "U g after techniques for working the latter had been
de'e`~di. However, he is not specific as to the nature of these techniques
nor to relationships between their application to jade and stone.    Although
we have concerned ourselves solely with the colossal heads, Kubler raises a
point that is worthy of consideration.   It does inde     =        some of the
marks of manufacturing techniques applied to the miniature figurines were
carried over to the colossal heads. Specifically, Covarrubias (1957:56,
ig. 2   suggests a method of making stone figurines which includes the use
of drilled pits at the corners of the mouth. Other authors (Coe 1965b:757;
Drucker 1952:189; Smith 1963:129) have noted the presence of similar pits
at the corners of the mouths of the colossal heads. This may be evidence
that an attempt was made to indicate pits similar to those of the figurines.
The parallel is, however, more stylistic than technological, for in
the jade figurines the drilled holes marking the corners of the mouth seem
to have served as "saw pits" marking the termini of the sawed cuts made to
block out the mouth, and in the colossal heads such pits would not have
been required since the mouths were sculptured by the slow and laborious
pecking process. However, it is not at all certain that the drilled pits
in the mouth corners of the jade figurines are present because they are
ineradicable remnants of the manufacturing technique. Figurines made of
soft serpentine may also have such pits, and pieces of such soft material
were probably not made by the sawing technique. Furthermore, a common
feature in Olmec pottery figurines is the punched pits made with a stick
when the clay at the corners of the mouth was soft. Taken all in all, we
are of the opinion that the drilled pits which occur in the mouth corners
of the jade figurines and colossal heads (as well as some other large
Olmec sculptures) are primarily a perseveration of the method of depicting
human mouth corners which originated with pottery figurines. There are,
of course, certain additional similarities between the figurines and the
heads, but these are more a function of the fact that they belong to the
same artistic complex than of any specific affinity in sculptural technique.
A somewhat more imaginative, if less accurate, approach to the
64
problem is presented by Ayala (1966). One of his figures shows a group
of three burly Olmec sculptors, all with shaved heads, hard at work on
the nearly completed LV 1 head. Each man holds an unhafted green stone
celt in his left hand, while in his right he grasps a wooden mallet or
a fist-sized hammerstone (ibid., 25).    The suggestion   at least in     ,
is that the jade and serpentine celts found at many Olmec sites were used
as chisels in the sculpture of basalt monuments, an idea which seems un-
likely in view of the fact that no known celts bear evidence of such use
in the form of edge abrasion or battering.
Given the scant and rather cursory nature of comment on the sculp-
tural process, it may be worth-while to place our thoughts on the subject
on record.   Much of what follows falls within the realm of conjecture,
because a complete understanding of all steps involved in the creation of
these works cannot be gained in view of the lack of written records or
other forms of graphic representation. Furthermore, there are no known
colossal heads which are incomplete or partially finished. If one of
these should be discovered, we would surely learn a very great deal about
the manufacturing process. However, some inferences can be drawn with a
fair degree of certainty.
The first-te-in the process must have involved the ac
of a suita l   block go; basalt and its transportation to the site at wh ch
it was to be sculptuRe. _depl      e     Williams and Heizer (1965:4) believe
t altt1tleocks were not actually quarried, but rather that detached,
rounded boulders of suitable size and shape were selected.     Such boulders
may be found today on the slopes of both the Cerro El Vigia and Cerro
Cintepec localities. Once a selection was made, the boulders were moved
by land and water to the given sites. It is not known whether boulders
were roughly blocked out prior to transport or if they were brought unmod-
ified to the site.   The presence of a large uncarved boulder of basalt at
La Venta (ibid., 19), as well as a probable area of basalt debitage at
the same site (Heizer, personal communication),           icate that most of
the   sculR,~~~~~~~~~                   t4  Thg on     air~4
the sculpturing was done at the site after transprt.      D
amount of scoutin in    the Tuxtlas no signs         ne
exposures have been observed or reported.
With regard to the sculpturing process, we have little information
on the actual steps taken. However, we suggest the following sequence of
events.  The first step must-have    nvo lve  t   phe             desi   n
the head.  This planning may have been carried out after the basalt b
was brought to the site    but it seems more likely that it tQ Qr
to the selection of a boulder.    Aiternatively, a rather general design
lay-out may have been devised, subject to modification once a suitable
stone was chosen.   In any event, once the design was decided upon, the
65
features of the face and decorative elements were roughed out, presumably
by means of pecking with heavy picks or hammerstones.    It seems probabe
that the heavily pecked surfaces on the sides of SL 4 ad         are the
result of such initial blocklng-o    activity.             -
The next ste                          JNmmer-dres        folowe   b
smoot~ing of the pecked surfaces by means of grinding. It would have been
at this point that the flat areas on the backs of SL 1    SL 2^S     TS   5
LV 3 and Lvor          ound down.  These surfaces are now smooth planes,
with minimum areas of about one square meter characterized by long, parallel
striations which track from one edge of the smoothed surface to the other.
It would seem impossible that such large surfaces were ground     so evenly
through the use of small tools such as hand-held abrasive stones. We
suggest an alternative method (page 66) which involves drawing a long flat
plank  charged with sand or some other abrasive (quartz?     back and forth
acro    the back of the head.  Such a technique would account not only for
the uniform surfaces but also for the long unidirectional striations.     The
stroke of the heavy abraslve-loaded log or plank must have been a long one,
as can be inferred from the length of the striae. A method of smoothing
and polishing stone described by A. Kennedy (1821:50-51) for India is
quoted here as an example of a technique which is somewhat similar to the
one we propose.
A block of granite of considerable size, is rudely
fashioned into the shape of the end of a large pestle.
The lower face of this is hollowed out into a cavity,
and this is filled with a mass composed of pounded
corundum-stone, mixed with melted bees-wax. This block
is moved by means of two sticks, or pieces of bamboo,
placed one on each side of its neck, and bound together
by cords, twisted and tightened by sticks. The weight
of the whole is as much as two workmen can easily manage.
They seat themselves upon, or close to, the stone they
are to polish, and, by moving the block backwards and
forwards between them, the polish is given by the
friction of the mass of wax and corundum.
The flattening on the backs of the Olmec heads was carried out for
unknown reasons. One opinion holds that the heads were meant to be set
flush against some structure and to viewed only from the front (Stirling
1955:20).  However, there is no indication of structures of any kind (walls,
houses, posts, etc., etc.) at the rear of any of the heads found in situ.
While impermanent structures may have been present, there is no archaeolog-
ical evidence to indicate that this was the case. We prefer to view the
66
,~~~~~~~~~~~~~~* 1b Il
#Ar~w V/Jloef
Wiry~~~~~~~~~~~~~~~~~~~~~~~~~i
uggetedmethod of flattening backs of heads (schematic).
rrows indicate direction of motion of plank across back
f head.
67
flat back of many of the heads as primarily a stylistic rather than a
functional feature.  The colossal heads, then, become merely one kind
("stelae" and "table-top altars" are other) of a multiformity of free-
standing stone monuments which were placed in full view in the Olmec
ceremonial centers.
Following the grinding and smoothing of the heads, remaining fea-
tures were-,i  ~fE  n re1   . t                                     spes of th sp ess  i
t  ouan the ba  of S   an d SL  the braids on the back of NS 1, and
the vertical grooves on the back of TZ 1.   Once the heads were completed,
they were moved from the workshop and set in place, perhaps on smal    mounds
or platforms such as that on which TZ 1 rested.    There is some evidence that
the heads may have been painted.   Stirling (1943:58; 1955:20) observes that
a portion of LV 4 was covered with a white slip and painted a dark purplish
red.
The above outlines what we believe to have been the main procedure
followed in sculpturing the colossal heads. In addition, a number of the
heads have been the focus of further attention, some of which has been
discussed under the category of defacement. The question of the rectangular
niches found on the back of SL 2 is considered below. We believe the niches
were carved at some time after the completion of the head and prior to the
t6TfTing of the monument `ito the ravine.   Numeous other pits and'scorings
off-t-ffea-es were executed both during and after construction.   Pits at the
corners of the mouth are in      former category, as are the remnants of
circular pits found in the center of the left ear ornament of LV 3, the pits
which form both ear holes of SL 1 and NS 1, and the right ear hole of SL 4
and the left ear hole of LV 1.
As noted in the discussion of intentional defacement below, many of
the other circular pits, often with dimples in the center, fall clearly into
the category of post-sculptural defacement, as do the numerous gouges and
axe-sharpening grooves present on many of the heads. However, pits with
'U-st8>bX  -,e,  'r- .<..J-;>7. ,  .   V
dimples are not always randomly scattered, and some appear to have been
placed symmetricall   on tin pairs.                   The pits above the nares
on SL 1 and the apparent pairing of c    aEtan pits on the face of SL 2 are
notable examples of this feature.   These may be indicative of a form of
ceremonial defacement, although we are uncertain as to whether this occurred
before or after the heads were completed.
68
WEATHERING AND MUTILATION
In addition to several types of intentional defacement, all Olmec
colossal heads exhibit some degree of natural erosion. For the purposes
of this report, defacement or mutilation is held to mean any purposeful
violence inflicted upon the surfaces of the heads.   It is not always possi-
ble to ascertain whether the disfiguration of a part of a head is due to
vandalism or is the result of natural forces acting upon the rock. To make
an exact division between natural and purposeful or accidental defacement
will, of course remain, in some instances, subjective.   Thus, for example,
the deterioration of LV 4 through natural scaling may have been hastened
by the purposeful prying off of rock fragments along the lines of cleavage
by vandals, and through the action of chain or cable friction during its
transport in 1956 or 1957 from the original site to the Parque La Venta in
Villahermosa.
As a result of prolonged exposure to the elements, all of the Olmec
heads have suffered some degree of erosion. The variation exhibited in
this respect is a function of several factors.
Duration of Exposure to the Elements
Although all of the heads are believed to have been fashioned within
a relatively short time span, it is possible that the generally better con-
dition of the San Lorenzo heads is due to the fact that they were thrown.
into ravines where they were quickly buried by slope wash, while the La
Venta heads remained in situ after the abandonment of the site and were
slowly covered by an accummulation of earth, which raised the general level
of the area on which they stood.   In fact, the La Venta heads may never
have been completely buried; we know that they were periodically disin-
terred by succeeding inhabitants of the area.   This fact has been docu-
mented by Blom and La Farge (1926:85) who reported being guided by natives
to the La Venta site where they photographed the dome of LV 1 where it pro-
truded from the surface.  Unfortunately, Stirling (1943) does not tell us
whether LV 2, LV 3, or LV 4 were partially exposed, or at least known to
the natives, when he excavated them.   If they were, this might account for
the considerably greater degree of erosion apparent on these heads, partic-
ularly the almost complete deterioration of the face of LV 3.    As noted
below, the San Lorenzo heads show the greatest natural deterioration and
purposeful defacement on that area which lay exposed or nearest to the sur-
face, but at the same time these alterations are not limited to the exposed
surfaces.
69
Differences in Rock Composition
The considerations mentioned above do not explain why NS 1 and TZ 1,
which were also found in situ and buried by an apparently similar long pro-
cess of earth accumulation as the La Venta heads, are by far the best pre-
served specimens. The difference is probably due to variations in rock
composition.  LV 1, LV 2, LV 3, and all the San Lorenzo heads are carved
from large boulders of coarsely porphyritic olivine-augite basalts obtained
on the slopes of Cerro Cintepec. These boulders are in all probability of
inferior quality and less resistant to erosion than are the smooth-faced,
spheroidally-weathered olivine-augit basalt boulders from the upper slopes
of Cerro El Vigfa where the people of Tres Zapotes and Nestepe obtained
their lithic material (Williams and Heizer 1965:4).
Surface exfoliation is a feature especially apparent on LV 1, where
rock layers from 3 to 5 cm thick have scaled off from the back on both sides
and below the left cheek, and on LV 4 which shows this characteristic on the
entire right side and lower part of the right cheek. Another characteristic
which is particularly pronounced on the heads fashioned of Cerro Cintepec
boulders is the tendency to form small vesicular pits, which may be due to
the washing out of softer materials included in the matrix. Stirling (1943:
57) agrees with this interpretation and points out that the pits could not
be bubbles such as often occur in lava. These concavities may accumulate
water which acts as a solution agent and erodes even deeper into the basalt.
Although these vesicular pits occur on all the heads, only a few sizable
ones occur on TZ 1 and NS 1. The La Venta heads, particularly LV 2 and LV 3,
show a large number on the tops of the heads with numerous roughly circular
holes ranging from 4 to 5 cm in depth. These naturally occurring vesicular
pits are not to be confused with the artificially ground pits with dimples
discussed below. The former are distinguished not only by the lack of dim-
ples but also by their irregular circular shape.
Faults and cracks are found on basalt heads from both Cerro El Viglfa
and Cintepec. TZ 1 has some superficial cracks on the top of the head,
which according to Stirling (1943:17) may have been "caused by repeated
exposure to fires made in burning grass or clearing land for milpas."
Similar cracks run over the entire face, top, and sides of NS 1,
which also lacks a large piece at the rear left base, probably broken off
after the head had been completed. All the San Lorenzo heads except SL 1
have minor cracks and faults. The latter exhibits an extensive fault on
the back which has been recently patched with concrete. A section which
had scaled off from the head band above the right eye was cemented back in
70
place by Stirling and Drucker, who found the pieces lying on the ground at
the base of the head at the time of excavation (Stirling 1943:10). The
apparent "fault" on SL 2 is discussed below in the section on rectangular
niches.
Since climatic conditions (which are not conducive to preservation
in this area) are generally the same at all three sites, their weathering
effect upon the heads should presumably be the same.   However, a comparison
of photographs published by Stirling (1939, 1940, 1943) with recent ones
shows a marked further deterioration of the La Venta monuments, while the
TZ 1 and NS 1 heads have remained unchanged. This difference may be ascrib-
able to the superior quality of the Cerro El Vigfa basalts of which the
latter heads are manufactured.
Intra-site differences, such as the place of exposure or location of
the head, could not have caused a great disparity in the relative degree of
weathering exhibited by the sculptures since all were imbedded in similarly
humid earth. However, the fact that SL 3 was found lying face down in a
spring, and consequently was continually waterlogged, may have contributed
to the damage exhibited on the lower lip and the formation of a diagonal
crack across the face, running from the right eye to the left cheek, which
coincides with the corner along which the tip of the nose has broken off.
Intentional Mutilation
With the exception of NS 1, all of the Olmec colossal heads present
one or more of the following intentionally inflicted types of defacement:
sharpening grooves; rectangular niches; gouges; and pits with dimples.
Each of these categories is discussed below. Suffice it to say here that
aside from the colossal heads, the extent of intentional mutilation upon
all types of stone monuments of the three Olmec sites discussed here has
been noted by previous authors.   Drucker, Heizer and Squier (1955:229)
counted twenty-four clear-cut cases of intentional mutilation on forty
monuments at La Venta. Stirling (1943:11) notes some degree of mutilation
on all of the monuments from Tres Zapotes. (NS 1 had not been unearthed
at the time of his publication.) The fact that until their recent discovery
most monuments were completely buried may mean that defacement was carried
out shortly after the abandonment of the sites, but it is not possible to
ascertain whether such destruction took place during one or several suc-
ceeding periodic reoccupations. Interestingly enough, the purposefully
overthrown monuments of San Lorenzo are less mutilated than the in situ
ones at other Olmec sites. Stirling (1955:9) concludes from this that
"the displacing of the stones was apparently done by the later aboriginal
occupants of the site, as represented by the upper occupation level. Since
the two levels are separated by a considerable time interval, it does not
seem likely that a conquest took place." Possibly overthrowing of the
71
monuments at San Lorenzo should be considered as an alternative to deface-
ment.
On the other hand, Drucker, Heizer and Squier (1955:230) hold that
the great amount of energy expended by the mutilators of the La Venta monu-
ments, carried out with heavy tools while.working upon "very hard stone,"1
"suggests strongly that its perpetrators were intent on destroying the works
of art of their predecessors."   In this respect, it should be recognized
that what we interpret as intentional mutilation or casual tool-sharpening
marks may in fac1.be signs of some conscious and deliberate ritual or cere-
monial act (ibid., 197).
Sharpening grooves.   Longitudinal grooves ranging between 2 cm and
50 cm in length occur on all La Venta heads} as well as on TZ 1, SL 1, and
SL 2.. They appear to have been ground into the stone with tools such as
axes or celts, and have been interpreted as axe-sharpening channels. Numer-
ous similar grooves have been found on other Olmec stone monuments, such as
Stela 4 at La Venta (Stirling 1943:52, p1. 33d). Drucker, Heizer and Squier
(1959:196-197) illustrate a fragment of a limestone slab (probably imported
from Chinameca, the source of most La Venta limestone according to Williams
and Heizer 1965:12) which shows thirteen very similar celt-polishing grooves
on one flat surface. They speculate that the slab may date either from
Phase IV or post-Phase IV of La Venta, basing this assumption on the presence
of limestone slabs being "undoubtedly late in the site's history."
We discovered a large number of these sharpening grooves on rocky
outcrops along the sea near the village of Punta Roca Partida during the
winter, 1967, expedition (see page 72).   These grinding benches closely
resemble others described for various parts of the world, such as those at
Fernando Po, Nigeria (R. Kennedy 1962:129), axe polishing grooves in France
(Dechelette 1908:525), and a cliff with polishing grooves at Pueblo Bonito,
New Mexico (Judd 1959, p1. 23). The grinding technique as actually performed
by modern primitives is best described by Harrer (1964, p1. 18) for New
Guinea tribesmen, who employ a technique which must not differ greatly from
that utilized by the above mentioned axe polishers, including the Olmecs.
It is not certain whether these grooves constituted a purposeful
defacement of the heads or were, at least in some instances (e.g. LV 1),
part of their decorative makeup. They may simply have been tool sharpening
grooves utilized by post-Olmec peoples who saw in the heads no more than a
convenient sharpening stone in an area otherwise devoid of available lithic
surfaces.  In the rock-poor La-Ventia_''region, all heads are heavily scarred
with these grooves, but at Tres Zapotes and Nestepe, where natural basalt
outcrops occur nearby (Stirling 1943:16), only TZ 1 exhibits a few, very
small grooves.  At San Lorenzo, only SL 2 shows this type of defacement, and
72
Boat-shaped axe sharpening grooves in basalt outcrop
near Punta Roca Partida, Ver.
it is possible that the San Lorenzo heads were not used as a sharpening
surface, since the site is located on an extensive gravel plateau and
abrasive surfaces for axe polishing must have been readily available.
Another possible purpose for these grooves may have been as part of some
ritual axe-connected ceremonies, carried out from time to time, and which
in some specific way may have involved the heads.
It seems unlikely that the grooves were produced by the sculptors
of the heads while sharpening stone tools used in carving the monuments.
If that were the case, the grooves certainly would have been cut on some
surface of the rock which was not meant to be visible, such as the back
or base of the head. Stela 4 of La Venta shows sharpening grooves at the
base, and Stirling (1943:58) suggests the possibility that they were used
by the carvers of the stela "as the surface of the base would not show
when the stela stood up." The fact remains that in the instances where
these grooves occur on the colossal heads, they usually cut across the
sculptured surfaces, partially destroying the decorative elements.   If we
are correct in interpreting these V-shaped grooves as channels in which
the bits of stone axes were honed, some support for one last statement ca
73
be deduced. Sharp-edged stone axes were probabl used for wood-cutting and
not stone carvng.- There is not the slightest indication that chisels or
axes were used as scutos      tools rJJather that hammer-dressing and
stone-pecking were the main stone seh-,  g ,-ques.
In evaluating the probability of any of these hypotheses, it is
useful to study these grooves in the light of their relative position on the
head and the position of the head when discovered.
LV 1, first described by Blom and La Farge (1926:85), was found in
situ in a- normal' upright position   the to  of the head protrudinr from the
0;W1~~~~~~~h  grove   int--/esstn -
soft ground.  The grooves   interestin 1  enou h  occur only on top of the
head,--a--ve the encircling head band.  Consequently, it may be assumed that
they were cut at such a time as the head was protruding at ground level.
That period, of course, must have been later than the occupation of the La
Venta site by its builders, and at any event would not have been contempora-
neous with the sculptors of the head. An alternate possibility, that the
sculptors of the head may have intended them as "part of the decorative
scheme,1" is suggested by Stirling (1943:56). His observation may rest upon
the fact that the grooves are evenly and radially distributed over the top
of the head, pointing to the front and sides. However, this seemingly pur-
poseful symmetrical distribution may be due simply to the fact that over-
lapping or crossing grooves would make bit-polishing of a tool impossible,
since it would have to skip over existing ridges. Only on the right side
of LV 1, just above the head band, does such an intersection of two grooves
occur. This intersection is at right angles, forming a small cross. We do
not think that it was made by stone tools since the grooves do not present
the characteristic deeper center, or concavity and shallow ends, which
characterize the other 44 grooves on the head. Possibly it represents a
Christian cross cut in by modern local natives as part of a spirit-exorcis-
ing act. That local natives did cause some of the defacements on the La
Venta monuments is discussed by Blom and La Farge (1926:89), who mention
"incisions made in recent days with steel tools" on another La Venta sculp-
ture.
LV 2 also found in situ, was standing upright on a stone founda
facing north and in line with LV 3 and LV 4.   We counted 15 grooves on. he
top and 56 grooves on the back of the head. Besides these clearly defined
sharpenfing -g-tove , smaller, rather shallow grooves of a different type-
some of which may be simply part of the erosion suffered by the head, and
others grooves in which grinding was started but then abandoned-occur on
the same areas. The longest grooves found on ancolossal head are on the
back of LV 2; these run toO05 m in length.     The b                      _ed
by these grooves, some of which run inside the four large gouges descri    d
below.
%44tw'9fir:
74
The grooves on the top of LV 2 are in roughly the same symmetrical
distribution as those of LV 1; that is, six on the left and six on the right
side. The frontal area contains only one groove; presumably it was not
utilized for grinding purposes due to the unsuitability of the surface at
this point, since it is heavily eroded and covered with the vesicular pits
described above. Three grooves occur on the center rear of the top of the
head.
While it may be assumed that the axe sharpening grooves on LV 1 are
located on the top of the head because this was the only surface exposed at
the time the groove-makers were active, this is not true for LV 2, on which
grooves occur from the bottom to the top of the rear surface of the head.
Consequently, the entire rear of the head must have been exposed when the
grinding of the grooves situated on the lower part of the back was begun.
This fact presents several interesting problems.
1.  Why were only the topand rear parts of LV 2;.kikized as        inding
0@_____qq                                                  t-' itene*_'?,r td;'?@ ,15,-S
surfaces andT              tadsideswhich also have convenienty fla            d
are   Cold the                ome tabtero                         wuants of
the site whpc revented them from               e             (We assume that
sTttred-ttie rear was not yet buried, neither were the face or sides.) Is it
possible that the head was cleared by later occupants of the area, seeking
some type of grinding surface? If so, why were the other sides of the head
not utilized?
2. Assuming that LV 2 was upright at the time of the cutting of
these axe sharpening grooves, the ones situated on the lower part of the
head could only have been ground by a person lying flat on the ground or
stooping down. In either case, the pressure applied to the tool being
sharpened would have been very slight, and the process of grinding propor-
tionally ineffective. This ineffectiveness would have been reduced if an
abrasive substance had been placed in the grooves while grinding. We have
considered the possibility that some such substance-such as quartz sand or
stone dust-might have been used when sharpening stone tools in these
grooves, but obviously it would have been impossible to place an abrasive
substance in a groove cut into a perpendicular surface such as the rear of
LV 2.
3. As a consequence of the above, it can be suggested that LV 2
was standing on an elevated surface at the time when the grooves at the
bottom of the back were cut. There is no archaeological information for
the La Venta heads to support this proposition, except for the fact that
LV 2 LV 3 and LV 4 we e found "resting on a stone foundation" (Stirling
1940:331).   At the Tres Zapotes site   TZ 1 was found on a stone       rm,
and this same   r             hlnta.
75
4. If LV 2 was standing on an elevated surface when the lower
grooves were cut, the upper grooves must have been ground at a later time,
when the surrounding earth had reached a higher level and sinkage of the
head had begun. It would have been difficult for someone sitting on the
top of the head to grind these grooves downward, and it seems improbable
that a sort of ladder would have been used to reach the required height.
The shape of these grooves provides us with some clues as to the
position assumed by their fabricators. Whereas most of the sharpening
grooves found on grinding benches (such as Punta                   and on top
or_____o_____LV 2 run in straight lines     most of the long
grooves cut into the back of LV 2 have an arced shape.      In the case of LV 1
and LV ATtats, dierence may have been caused by the fact that the grinder,
able to assume a sitting position, would push and pull the tool in a straight
line toward and away from his body, in which case the resultant groove would
be relatively short and straight. However, the long, inward curved sharpen-
ing grooves on the back of LV 2 indicat e to   us tnat the grinder,         grasp-
ing the tool (probably with both hands, in order to increase pressure and for
greater control), moved it from side to side, his arms constituting the
radius for the resultant semicircular sharpening groove.
LV 3   the most eroded and mutilated of all the Olmec colossal heads,
is covered with sharpening grooves    with at least a few on every side.     We
couned t  c-eatry oele  axe sharpening grooves on the top of the head
alone.  Stirling (1943:59) speaks of "many striated grooves" on the back of
LV 3.  On inspection of the head, now in the Parque La Venta, Villahermosa,
we could identify only two of these, located on the upper part of the back
surface, as "typical" axe sharpening grooves. The remainder of these "scored
striations" do not present the typical concave depression, and are rather
closely grouped, parallel, shallow, and narrow grooves. Furthermore, three
sharpening grooves ground into the surface of the face of LV 3 seem to dis-
prove the hypothesis presented     tV 2:     that superstitious fear prevented
the makers of the grooves from utilizing      gfacial surface.
LV 4  which has suffered extensive damage through scaling on its
right side   presents'ewet iharpening            ttidiistyof t            a
es These a concentrated onthe ceteror     te    eaddress    We counted
23 grooves on the cente r     the top, but this number is not absolutely cer-
tain since the scaling on the left top side has destroyed an area which may
have contained additional shallow grooves at some time prior to the detach-
ment of the rock surface. The thickness of the layer which has scaled off
ranges from 2 to 4 cm to judge by the height of the ridge left along the
line of detachment.   Since the depth of these sharpening grooves rarely
exceeds 2.5 cm, it is improbable that there would be remanent depressions
on the head where the grooves penetrated deeper than the scaled-off layer.
76
In any event, this means that the grooves were made prior to the onset of
natural scaling.
The radial distribution of the grooves on LV 4 bears an immediately
apparent resemblance to that of LV 1, sharing the same pattern only on the
top of the headdress. Since the top of the head exhibits these grooves-
possibly because it was the only exposed surface upon which the groove-
makers could work-and since grooves appear in a radial distribution around
the entire top center of the head, except in the scaled-off area, we assume
that groove-making, for whatever purpose, was restricted to a relatively
short span of time. If it had occurred throughout the existence of the
head, such grooves would be found on the freshly exposed surface which en-
sued after scaling. Their absence in the area tends to confirm the former
hypothesis that, at least on this head, groove-making was restricted to the
period between the time of completion of the head and the beginning of the
scaling.
TZ 1, like the La Venta heads, was found in situ, placed upright
upon a tirrd&'r   rou           stnsw7                                 nindicated
that the head hdebeenby              prc4ess of accumulation of earth.
This head shows little defacement and exhibits only three      rooves   all on
the rear of the sculpture.    The back of TZ 1 is crossed by nine shallow,
perpendicular, incised lines which may have been cut as part of the decora-
tive element of the head. According to Stirling (ye- cit., 16), they
indicate hair. Toward the top, the first three lines are crossed trans-
versely by two long, shallow, sharpening grooves which run parallel to each
other. Another long, smooth, and concave sharpening groove appears between
the third and fourth incised lines, running between them from the base line
to about one-fourth of the height of the head. The position of this last
groove presents the same problem as that discussed for LV 2:      it w u     ave
been difficult for the m koThT5he groove       .assume a comfortable position,
and the wrist of his working hand would have continually struck the ground
while moving the tool downwards.    A possi j<,planation is that this groove
cut before the head was placed upright     and while it was in the position
it is today; that 4s, lying on its slie. In other words, it could be con-
temporaneous with the sculpturing of the head.
SL 2 is one of the two colossal heads not found in situ which pre-
sents sharpening grooves.    According to Stirling (1955 :10),   t was fou
lying face up.   We have noted only two grooves on this head      One   on the
right side above- the ear,    is 4.5 cm deep at the4owst point.    This makes
it the deepest of all sharpening grooves thus far discussed. We are reluc-
tant to classify this groove in the general group for several reasons:
(1) it is considerably deeper than all others; (2) rather than being broadly
semicircular, it forms almost a right angle at the deepest point; and (3)
77
only one end of the groove becomes shallow and runs out onto th6 rock
surface level; the other end remains deep, as if the operator had stopped
the movement suddenly and reversed the direction of the tool at the same
point each time. Possibly this groove was intended as part of the decor-
ation. This suggestion is supported by the fact that the groove was prob-
ably cut beforethe head was toppled into the ravine where it was found by
Stirling (o. cit.).    This hypothesis is based upon the observation that
it would have been difficult, though not impossible, to cut the groove
while the head was lying on its back. It may be that this angular incision
was meant to be continued into one of the rectangular niches which appear
on the back of the head. A third possibility is that the groove represents
nothing more than the smoothing out of a fault in the rock.
A vertical roove, 30 cm in length, gouged into the right cheek of
SL 2 is o great interest.    It is 1.5 cm deep, if measured from the ridge
formed on the right side down to the deepest part of the convexity it
forms at the bottom.   Measured from the left side ridge, it is more shallow.
This is due to the curvature of the rock at this point of the cheek. We
hold that this particular groove was cut after the head was toppled into the
position in which it was found, since the groove occurs on the side of the
head which was uppermost as it lay on its back, and the angle of incidence
shows that it was probably cut by someone positioned (sitting?) on the top
of the face, grinding downward (see figure, left). If the groove had been
cut while the head was standing upright, its angle of incidence should have
been as diagrammed in the right figure. We conclude that this groove was
cut after the head assumed its position in the ravine, following its dumping
there.
V / me 4. / r-, # -w
Left                                      Right
SL 1 exhibits sharpening grooves which present a special problem.
Stirling (1955:9-11) states, "When discovered it was lying on its back, head
down on the slope of a small but steep arroyo about 300 yards southeast of
78
the principal mound." All four of the relatively shallow grooves on this
head appear in the same general area, two immediately behind the left ear
and two smaller ones behind and slightly above the left ear. As this area
coincides with the curvature of the left side just before it runs into the
flat back surface, the presence of axe sharpening grooves at this particu-
lar point indicates to us that they were fashioned before the head assumed
the position in which it was discovered.   The possibility exists that,
after being removed from its original location and thrown down the slope
of the steep arroyo, it repeatedly shifted its position as the water of
the arroyo undercut the surface upon which it lay. The grooves could then
have been cut after the head had been toppled and before it had assumed
its final position.
Rectangular niches. The six large rectangular cubical holes which
appear on the back surface of SL 2 are unique among the colossal heads.
The only other instances in Olmec sculpture where holes of this type appear
are on Altar 4 of La Venta and Monument 14 of San Lorenzo (pl. 33b).
Rectangular niches on SL 2.
Niche 1 has a maximum depth of 5.2 cm as measured from the left side.
The bottom of the niche slopes upward to the right from its low point on the
left side until it is level with the surface of the head band. This charac-
teristic distinguishes this niche from the others on this head, and from
those on La Venta Altar 4 and San Lorenzo Monument 14.
79
Niche 2 is 9 cm deep and cut within the knot of the head band, par-
tially obliterating a line which had been cut by the sculptors of the head.
It shares this characteristic with Niche 5, and our assumption that the
niches represent a defacement-type that post-dates the sculpturing of the
head rests upon this fact. Altar 4 of La Venta and Monument 14 of San
Lorenzo also "have had the carvings at one end carefully defaced, and in
the area thus produced, deep rectangular niches were excavated" (Stirling
1955:16).
Niche 5, on the upper part of the back, forms almost a straight line
with Niche 3 in the center and Niche 6 on the bottom. These three niches
also share a characteristic in that they are the shallowest (depths: Niche
5, 4.5 cm; Niche 3, 3 cm; Niche 6, 4.8 cm), and are roughly equal in that
respect.  The deepest niche on SL 2 is Niche 4 (15.5 crm) which is also the
largest. It is probable that the lower surface in which Niche 4 occurs
is due to an original imperfection in the stone, but the possibility remains
that it may have been purposefully carved, since in Monument 14 from San
Lorenzo niches were carved in a new surface of the head created after care-
fully chipping away earlier decoration to a depth of 2.5 cm (Stirling 1955:
16).
The main difference between the niches in SL 2 and in the aforemen-
tioned monuments is thitR TtaV'hfa AltarI4 and San Lorenzo Monument 14
they are almost twice as long as they are wide (ranging from 20.3 to 27.4
cm in length and from 10 to 17.5 cm in width), while in the SL 2 head they
are almost invariably square.   The variation in size, shape. depth,
and relative position of these niches does not suggest what their use may
have been.
We believe that the holes were carved after the head had been com-
pletely sculptured, and that they must have been finished before it was
removed from the original vertical position, since Stirling found the head
lying on its back. This means, therefore, that at least SL 2 was used in
some type of ritual activity involving the/cutting of niches which must have
served a ceremonial or 7tiDtTar`1i purpose.   It is unlikely that the amount
of effort and skill that was expended in carving the niches was directed
solely to "defacing" the head.   Possibly the niches were utilized for de-
positing offerings or for burning incense.   They may have contained figurines
since the emergence of a figure from a niche (particularly at the La Venta
and San Lorenzo altars) is a frequent theme of Olmec sculptural art. La
Venta Stela 1 shows a figure standing fully erect in a large rectangular
opening (Stirling 1943, pi. 33a).   In this respect, it may be remembered
that the large niche on the back surface of San Lorenzo Monument 14 alsp
contains a stone "plug."
80
Gouges. Large, shallow, concave, gouged-out depressions, present
only on the backs of heads LV 2, LV 3, and SL 4, are here referred to as
"gouges."   LV 3 and SL 4 each have one of these defacement gouges; LV 2
has'ou-r on the back surface.    These are smoQl.thed-out concavities  and
appear to have been gro,,und.  The deepest gouge is 6.5 cm, with a length of
38 cm, and is 19 cm across. Three of these contain sharpening grooves
which run longitudinally within them anincrease their depth.        Cn 3uy
this type of defacement-,antedates sharpening grooves and postdates flatten-
ing of the back, at least-sop1 LV 2.  We postu]aM. that-tle.se gouge wre
ground rhidie1Thoning the sides of stone tools, possibly celts or stone axes.
II
P&A1
/1~~~~~
/1/
6,0s, J:r,.W OW qmj
Gouge on back surface of LV 2
Ground pits with dimples. We have labeled ground circular depressions
containing sml.er- depreasions in the center of the concavity as grinding
Rswith dimples in order to distinguish them from other ground pits which
lack the central depression. These pits average a depth of 3 cm if the dim-
ple is included, but this depth varies greatly as a function of differential
weathering. Diameters range between 7.5 to 10 cm. The largest pit is on
81
SL 3, and is 12.5 cm in diameter.
This type of defacement occurs on all six of the San Lorenzo heads,
on LV 2, and, in somewhat different form, on LV 3.   In contrast to the
sharpening grooves and rectangular niches which appear on other Olmec
monuments, these pits with dimples are restricted to the above specified
colossal heads.
At least some of these drilled pits with dimples appear to have been
used-by the sculptors. of the monuments as a sculpturing technique, while the
mjority are superfluous and appear to us..tQ have been purposefully inflicted
defacement.  Thus, for example, onQSt ..lwe counted eleven of these pits (six
on the kface, four on the right side, and one on the left).  Xwo of these pits
occur immediately_! above   ur~s a     hvbeen ground into each nostril
from above. Each ear hole presents a pit with a dimple. Thus, at least these
Pits with dimples on the nares of SL 1
four pits could have been used as a sculpturing technique.   Two pits appear
op._4he cheekbones, diametrically opposed, at the same point under each eye.
The remainder are distributed at random and may represent true defacement.
Two cup-shaped depressions appear in the corner of each mouth, but these lack
the characteristic inner dimple depression and are part of the original sculp-
turing of the head. They obviously represent a different type of drilling
technique. We hold that the secondary depressions are a by-product of a
drilling technique which we have been unable to reconstruct.
SL 2 which was found lying on its back, exhibits pits with di      e
only on the face.  Some of these overlap and present different degrees of
deterioration through weathering, suggesting that they were made over an ex-
tended period of time.  They do not have any symmetrical distribution, but
are scattered at random, principally over the chin, mouth, and both cheeks.
A few appear on the head band, and there is one under the left eye and one
on the nose.
82
SL 3, which was discovered lying face down, exhibits pits with
dimples on the head band and headdress. This indicates that the deface-
ment must have occurred before the head assumed the position in which it
was found.   As on SL 2, many of the pits occur in pairs, their outer edges
sometimes overlapping each other. They are all of roughly the same diam-
eter (between 8-10 cm), except for a larger one on the head band above the
left eye, which is 12.5 cm in diameter.   Also as on SL 2, the pits appear
to be differentially weathered, possibly indicating that they were fashioned
at different times.
This latter characteristic is also apparent on SL 4, which was found
lying on its right side and, interestingly enough, shows 25 pits with dim-
ples, but only on its flattened, otherwise undecorated, back surface.
Although some of these pits are more weathered than others, in general they
appear to have been exposed to weathering for a considerably longer period
than those on SL 2 and SL 3. The majority of the pits are about 10 cm in
diameter; smaller ones measure about 7.5 cm. They are dispersed at random
over the right side of the back, only two of them overlapping each other.
The right ear shows a drilled pit which lacks a dimple. Since SL 4 was
discovered lying left side up, these pits could only have been ground when
the entire back surface was exposed, possibly even before the head was
toppled from its upright position.
SL 5, found lying face up, exhibits two of these defacement pits
with dimples. One of these appears beneath the right eye, and the other
is centered on the chin.
SL 6, recently discovered, was lying on its left side, and since
it is still half buried, we were able to study only the right side, and part
of the face. Three drilled pits with dimples appear on the face, one on
the center of the nose bridge and two-with overlapping outer circles
immediately below the right eye.
Only two other colossal heads exhibit drilled pits with dimples.
One of these, LV 2, shows a large number of the natural circular vesicular
pits discussed above. Some of these contain secondary depressions at the
bottom which may easily be confused with the artificial defacement pits
with dimples, particularly since the heavy weathering exhibited by this
head makes it difficult to ascertain whether some of the pits were concen-
tric, and if the dimples were centered at the bottom. We have, however,
identified three ground pits with dimples on the top of LV 2. As with the
three pits with dimples near the top of the right side of the forehead of
LV 3, they do not have the same form as those found on the San Lorenzo
heads, being of greater diameter, more cup-shaped and deeper.    The
inner dimples of the pits on LV 3 are 3 cm deep and 4 cm in diameter, while
83
at San Lorenzo the dimples contained in the pits average 1 cm in depth
and 1 cm in diameter.
Most of the pits appearing on the front and top of both LV 2 and
LV 3 do not contain dimples. The circular left ear ornament of LV 3 con-
tains three pits, two without dimples and one with a vertical-walled
dimple which does not resemble those from San Lorenzo. Apparently we are
here dealing -with the same grinding technique as was used at San Lotenzo
but executed with a different tool.
The possibility sugstVdhby S.tirling.(1955::11),,that the pits
"werev b    id to h   taczh som~e  extraneous decoration, must be discarded since
there is no archaeological evidence to support this assumption, and most
appear to postdate the sculpturing by a considerable period of time. None
ofithe pits bear any trace of paint or adhesive, and we believe they were
not elaborated in any way whatsoever.   The only indication that the heads
may have been painted is a reference by Stirling (1955:20) to a painted
fragment of LV 4.
The possibility remains that, as with sharpening grooves, ihe pits
marked a periodical ritual defacement activity by the Olmecs or later
native occupants of the area.   This assumption is supported by the fact
that this type of defacement is restricted to the colossal heads. However,
the\y Daysimply be the result of a tool grinding technique which we are
unable to reconstruct.
Summary
This section has reviewed the several types-and possible causes -
of defacement exhibited by the Olmec colossal heads. We have attempted to
determine the time when mutilation was carried out in relation to the time
of sculpturing by relating the position of each head to the area where the
defacement appears. The data are summarized in Table 16 which follows.
84
TABLE 16
Summary of Defacement
Position at Time                                                   lWeathering
No.        of Discovery                Grooves           Pits and Dimples   [(degree)*
LV I1  Upright in situ          On top only; numerous   None                |    3
LV 2   Upright in situ        115 on top              13 on top             |    4
I                       | 56 on back             | (divergent type)    I
LV 3 j Upright in situ        | 43 on top             | (divergent type)    |    5
|  3 on face              3 right side fore-
2 on back             |   head
a few on both sides     1 in left ear orna-
ment (divergent)   I
LV 4 | Upright in situ .        23 on top               None                I    3
SL 1   On back, head down     14 on left side         |6 on face                 I 1
|   slope of steep      |                         4 on right side
arroyo                                       j 1 on left side
SL 2'T Lying on back,         j 1 on right side         On face only             2
|   face up               1 on face
SL 3   Upside down on bottom | None                   I On head band and    |    2
of deep ravine, in   |                           headdress above
a spring             |                           face
SL 4   Lying on right side      None                    25, on back only         1
SL 5 | Lying face up            None                    2 on face           j    1
SL 6 1 Lying on left side     | None (possibly on       3 on face (possibly '    1
unexposed surface)      more, unexposed)
TZ 1  Upright in situ        13 on back                None                j    1
NS 11 Upright in situ         j None                  INone                 |    1
* Number indicates degree of weathering; highest indicated by figure 5, lowest
by figure 1. Although purely subjective, this range gives an idea of the relative
variation exhibited by the heads in this respect.
T Also presents 6 rectangular niches.
85
TABLE 17
Dimensions and Source Localities of Olmec Heads
Circumference I
(m)     I.
6.40
4.24
4.05
6.53
5.49
4.90
5.90
5.44
4.02
3.80
4.60
Frontal
Width (m)
2.08
1.35
1.60
1.98
1.50
1.34
2.17
1.90
1.60
1.17
1.47
1.26
Length
(m)
1.95
0.98
1.00
1.86
1.45
1.26
'I.68
1.50
0.95
0.95
1.15
1.41
l1
Stone Source
Cerro Cintepec
Cerro Cintepec
Cerro Cintepec
Tuxtla Mountains
Cerro El Vigifa
Cerro El Vigfa
Cerro Cintepec
Cerro Cintepec
Cerro Cintepec
Cerro Cintepec
Cerro Cintepec
Cerro Cintepec
Weight I
(sh.tns) I'
24.0
11.8
12.8
19.8
7.8
8.5
25.3
20.0
9.4
6.0
11.6
No. |
LV 1
LV 2
LV 3
LV 4
TZ 1
NS 1
St A
SL 2
SL 3
SL 4
SL 5
SL 6
Height
(m)
2.41
1.63
1.98
2.26
1.47
1.45
2.t5
2.69
1.78
1.78
1.86
1.67
I
I
I
. .
I
I
I  I  X  I
I  Xi  I
I  Xi ,
, i ,   U
,  Xi ,
,  X  ,  I
_ _)
C,) ~ _
XI  ,
_)~ *-4 r4 0
,- --  vi 4-
.  ~ _UQ  -
a)s
I I X I
I  I  I  X
I I .X I
I I   XI
IXi  I I
XII  I  U
Xz  I  I  I
IXi II
*  I  I  I
CU  Q  Q-  ~
a)n
86
z
Pei
00
v-I
Cl)
CU
co
a)
'-I
CU
C)
0
0
C)
a)
0
a1)
i-I
a)
0
a)
0
a)
a)
%l)
Cl)
cn
CL)
C0
-It
Cl)
CY)
cn
T-
cn
z
E-4
Nt
t4-
C)
as
v-I
X I
I X
X I
I X
IX
I X
XI
X I
X I
X I
X I
_-
X I
_-
X I
r.
a
ns
CU
4~4 4-i
bt~
0 )
bO N C4
-4 O4 n-
U)
* X I
_-
_          _4
* IX
'1-4    :    4
0       CU
-i     -i0     CU1
z
*   I
I   I
I   I
I   I
__
I  X I
I   I
I   I
I X
w I
I   I
I   X
4-4
0)
-r4
rI-
ro Q)
, I  a )'
04aia
sp  Q) 1
o   4J    Q
4-4 4-i U)
co 4
C   U-4 Pk -I
H
I
I
I
x
x
x
x
U)
o U)
was
110C
O CU
a) Q
OU 4
Oa 0
X . i
Xi *  .X  .
.~   ~
* X    *    X   * *Xi
*~~~~~~~~ -C_-
M   I I    I   I  X  I  X
_ I I
I I N
I      I
X I I
IX   .I X .
IX
I I X
I _ I
- I
M I
I XI
I   I    I                     I        I             I       X         I
X II
I I   X I X
I   X   I              i    X    I   X
U)        U)(      )   1
Cu        Cu      -1 1
'-4       '-4r        :
a ) 0 4 -   a C    0 0Q
S C  4)   S'  c   0  0
co   0  .4     05 4  4 Oi
54    a4)'       a)  U 5 .4 s
X   S  >    S4  >  =1 (1  4
0a)  0 a4  a) 0   Po~0 0
O$-I4         -~     C u O C .) 4   4   4
o    04 0 o  a ) o0  p o   so o
V a                  HZ E-  43 Z  X  u
rz~~~~~~~~~~~~~-
I X I
* X i
Xi,
X I I
Xi_
.* Xi
* ix
I
*X I
I I X
410
JJ4
4-IJJ
4-i
a) a) 0
5    0 0 0
;)
_   _
I X
I      I      I      I
I      * XM
I     X         I    I.   I   X
I   X            *   I    I    I
I    I           I   I    I   M
X _
_ _
- _
. M
X I X
I X I I
I         I         I         I
IX
I     X    I  I
a1)           a)
'r4           *-A
a)    ,. :
'0            '0~
a)            a)cu       r4
E   4  r Dr    4-        t    $4b
0   a)  cu  0     a   C   i   0  4-
oT in Oa Po       ar od    I Po   c
04   ~     0
-A.
1        :m        >
87
U I I
0)
ri)
.1-i
c)
-.4
a)
4.Ja
C)   a
Cd   co
FZ4
4-i
O
co
co VO
44 a)
0
4-i
Q) i-i a)
5-Cu  .4)
X   I  W
-A
ci  XI  I
X  I  I  I
.r4
I  X  I  I
I X I
-   -
0) -
4 -  0
Cl  ,f l04  l
*d H <  z
o~~~~~
88
-j
4-i
0
C)
00
H-
Cl)
Lrn
-It
'-4
rz-4
14.
En
'--4
r.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I    IX
CO 4
Xii
4C      C
Xii
0)
0
C4)
i<   C) 4-
Cl) *H 0) ,.
4.Jc) m
p.1
_ I
I X
, X
I I
I I
- I
XI
_i
I X
I   I
0)
CO)
Cd
4i4
4-i
0)  ;J4Jl
'r
Cd
swJ
9   ,4
4-i:
r4 po
0) 'H
*3 -1-*
4-J
0)
Cd
Cd VO
0
4-i
0 In
4Ji b0O4-J
0) *.4 0i)
Cl) Cl)
5- CdN
XI1-   0
89
APPENDIX II
OTHER LARGE STONE HEADS IN SOUTHERN MESOAMERICA
Although, as noted by Covarrubias (1957:65), Stirling (1965:733),
and Coe (1965b:741), the classic Olmec colossal heads are restricted to
the sites of Tres Zapotes (including Nestepe), La Venta, and San Lorenzo,
monumental stone heads of fairly large size and vague resemblance to the
twelve heads reported here have been found at several other localities.
Outsized carved stone heads also occur elsewhere in the Olmec heartland:
at Copan, Honduras; and on the Pacific slopes of El Salvador and Guatemala,
particularly at the site of Monte Alto, Guatemala.
At least six heads are known in the Olmec heartland: two are
located at the site of Cerro de las Mesas; at least one at Laguna de los
Cerros; two just east of San Lorenzo in the Municipio of Sayula; and one
at San Migxel.
Cerros de las Mesas is located on the south bank of the Rio Blanco,
fifteen miles east of the Bay of Alvarado. The heads are designated Monu-
ment 2 and Monument 8 by Stirling (1943:31, 45-47). Monument 2 depicts a
large masked face, 1.3 m high, 1.0 m wide, and 70 cm thick. The facial
features are very stylized, and the face has decorations on the forehead,
over and under the eyes, below the nose, and on the jowls. These decora-
tions are considered by some to be hieroglyphic in nature (Pinia Chan and
Covarrubias 1964, ilustracione 1). This head is illustrated by Pina Chan
and Covarrubias (o. cit.), Stirling (1943, pls. 26, 27b, 31a), and Gamboa
(1963, cat. 237).
Monument 8 at Cerro de las Mesas is 75 cm in height, and is termed
by Stirling (1943:46) an "anthropomorphic monkey." It has round eyes wit>-
out irises, eyebrows, a wide nose, and an open mouth showing six teeth. It
appears to have a rudimentary head band and chin strap. Monument 8 is
illustrated by Stirling (oi. cit., pl. 30c).
Laguna de los Cerros is located in the Rio San Juan drainage, approx-
imately halfway between the sites of Tres Zapotes and San Lorenzo. Here
Alfonso Medellin Zenil discovered Monolito Numero 1, a basalt head 75 cm
high, 70.5 cm. wide, and 70.5 cm thick, with a circular depression on the
top of the head to "hold water or blood" (Medellin Zenil 1960:87). This
head wears a stylized jaguar mask notable for its rectangular mouth, snub
nose, square eyes with irises depicted by X's, and protruding nasion. The
hair is described by Medellin (oj. cit., 86) as kinky or curly ("crespo o
90
rizado"). Taking the features as a whole, he feels that the face is
Negroid, but probably represents a Tlaloc mask, and that the hair, through
analogy with the Mixtec and Axtec codices, relates Monolito Numero 1 to
the curly haired gods of the earth and of death. This head is illustrated
by Medellin (O. cit., pls. 14, 15; 1963:7), Pifia Chan and Covarrubias
(1964), Hasler (1959), and Gamboa (1963, cat. 239).
There is a second Laguna de los Cerros head, designated Monolito
Numero 2 (Medellin 1963:6), which is reported to be very similar to Mono-
lito Numero 1 (ibid. 1960:86; Hasler 1959:31-32), but it has neither been
illustrated nor fully described.
The two heads from the Municipio of Sayula come from the areas of
Estero Rabon and Medias Aguas, which are respectively east and west of the
Tehuantepec railroad, near the town of Almagres. The Estero Rabon head is
45 cm high and is carved from gray andesite. This head has angular slits
for eyes, a pub nose, and an extreme jaguar mouth.   It also has a head
band and, seemingly, chin straps.   The Estero Rabon head is illustrated by
Medellin (1960, pl. 1; 1963, pl. 4), who terms it "de un purismo estilo
olmeca" (ibid. 1960:76).
The Medias Aguas head is in the form of a mask, 90 cm high, 63 cm
wide, and 47 cm thick. The features are very simply depicted. The fore-
head is plain, the eyes are depressions with pits in their centers, the
nose is barely modeled, and the ears are shown only by pits. The main
feature is the jaguar-like mouth which is wide and notable for its four
large canine teeth. Pits also occur under the nose and on either side of
the top of this head. This head is illustrated by Medellin (1960, p1s. 4
and 5).
A possible additional head from the Olmec heartland is from the
site of San Miguel, located in the State of Tabasco near the headwaters
of the Rio Blasillo, perhaps forty miles to the east of La Venta. This
head is "incomplete and somewhat atypical" in relation to the classic
heads (Stirling'.1965:733), and is broken off at the level of the eyes.
The extant upper portion of the San Miguel head is 1.05 m high, with a
characteristic flat back, and a headdress comprised of a series of round
faces evenly spaced over the dome of the head. The remaining facial
features are eroded beyond recognition. The San Miguel head is illus-
trated by Stirling (1957, pl. 50).
While not exactly in the same category as the monolithic heads just
discussed, two large tenoned heads from the Olmec heartland should be men-
tioned. Monument 6 from San Lorenzo (Stirling 1955, pls. 14a, 14b) and
Monument F from Tres Zapotes (ibid. 1943, pl. 5a) are different in the
91
treatment of their facial features from the colossal heads (ibid. 1955:13),
and probably form part of a sculptural subclass distinct from them. Never-
theless, since these tenoned heads are over sized, show careful sculptural
attention only to facial details, and were fashioned in the same geographi-
cal area as the colossal heads, they and the monolithic heads should be
considered as components of a single Gulf coast artistic tradition which
has as two of its attributes a tendency toward gigantism and an emphasis
on the head as opposed to the rest of the body.
Together, the heads from Cerro de las Mesas, Laguna de los Cerros,
Sayula, and San Miguel stand quite apart from the twelve colossal heads
reported here. With the possible exception of Monument 8 from Cerro de las
Mesas, which bears a superficial resemblance to LV 2, the heads just dis-
cussed are more stylized, much less purposely human, and smaller than the
heads from Tres Zapotes, La Venta, and San Lorenzo.    What the heads just
discussed do emphasize, however, is that in the Olmec heartland the tradi-
tion of carving heads without bodies was not limited to the naturalistic
giants. Rather it would appear that there was a larger Gulf coast artistic
tradition featuring the carving of big heads, which may have persisted in
time well past the demise of the pure Olmec style (see Coe 1965a, table 1;
Stirling 1965:735; Medellin 1960:78).
Another location which has yielded a large carved stone head is the
Mayan site of Copan in Honduras.   The Copan head was found at the northeast
corner of Temple 11 ("Temple of the Inscriptions") and is thought by
Stromsvik (1947:67) to be one monolithic "part of a human figure of heroic
size." The face is angular, with a jutting, pointed chin, and what was
probably a pointed nose before it was broken. The cheeks and jowls are
quite lined and protuberant, framing the mouth and chin. The mouth is open
and lacks the central incisors, while the eyes are large and tear-shaped.
The head wears a head band knotted in front, appears to lack ears, and seems
to be sculptured only in the three-quarter round. This head from Copan is
illustrated by Stromsvik (1947).
All in all, the Copan head seems to depict an old man, and to be
quite distinct stylistically from the Olmec heads.   Nevertheless, it does
resemble the old fire god figurines found in Trench 34 at Cerro de las
Mesas (Drucker 1943, pl. 8) and in Trench 3 at Laguna de los Cerros
(Medellin 1960, pl. 20). If there are any stylistic similarities between
the two areas, however, they are most likely the result of both locations
having either a primary or secondary influence from the Teotihuacan "X-T"
horizon (see Lathrap 1957:59-61; Coe 1965a:702; Longyear 1940:269-270)
rather than their influencing each other.   There are several other large
stone heads from Copan (John A. Graham, personal communication), but they
seem to be even less related stylistically to the twelve colossal heads of
92
the heartland than the Copan head just discussed.
The final group of large stone heads in Middle America is concen-
trated on the Pacific slopes of Guatemala and El Salvador. These heads
occur at the sites of El Baul and Monte Alto in Guatemala, and in the
Departments of Santa Ana and Ahuachapan in El Salvador. The El Baul head
has been described by Thompson (1948:22), who designates it Monument 3.
Noting its resemblance to the old fire god, he emphasizes its aquiline
nose, wrinkles, and large eyes.   Furthermore, Monument 3 has a smiling
mouth, is bearded, and wears a distinctive head band whose central element
is a human face. This head is 1.72 m in height, and is illustrated by
Thompson (oR. cit., fig. lOd) and Dieseldorff (1926, table 28, No. 155).
Whether the similarity of Monument 3 to the Copan head and to the Cerro de
las Mesas and Laguna de los Cerros ceramic figures is due to influences on
El Baul from Cerro de las Mesas, to Teotihuacan influence, or to influence
on the Pacific slopes of Guatemala from Copan is a subject that must await
further research (for a discussion of these possibilities see Miles 1965
268-269).
There is another "large head of the old fire god from the Pacific
slopes (exact provenience unknown), very like [the] one at Copan" (Miles
1965:269). This head is illustrated by Miles (o._. cit., pl. 19d), and,
although it lacks a head band, parallels the Copan head in its angular
features, large eyes, lined face, lack of ears, and three-quarter round
sculpturing.
The colossal heads of Monte Alto more closely resemble the classic
Olmec heads than any of the others we have so far discussed. These heads,
designated Monument 1 and Monument 2 by Parsons and Jenson (1965) are so
similar in size (Monument 1, 1.4 m high; Monument 2, 1.38 m high) and
method of depiction that they may be considered together. The Monte Alto
heads are bald, with closed eyes underscored by incised lines, wide, flat
noses, and thick closed lips. The nose and mouth areas are marked off from
the cheeks by an "inverted V-groove" (ibid. 135), and the ears are stylized
and adorned with perforated discs. While there is a basic similarity of
design and motif between these heads and the Olmec examples, the Monte Alto
heads differ in their treatment of the eyes, their lack of helmets and chin
straps, and in the fact that they are flatter-faced and sculptured in lower
relief and less in the round than the Olmec heads. Monument 1 is illus-
trated by Richardson (1940, pl. XVIIIa) and Parsons and Jenson (1965, illus.
3 and 4); Monument 2 is illustrated by Parsons and Jenson (OR. cit., illus.
5 and 6), Richardson (y. cit., p1. XVIIb), and Villacorta (1938, photo on
p. 27).
Whether the Monte Alto heads predate or postdate the Olmec heads is
93
still an open question. On the basis of their pottery collection from
Monte Alto, Parsons and Jenson (1965:141-143) believe the heads to be from
the middle transitional to late Preclassic periods and thus to slightly
postdate the Olmec heads. Miles (1965:244, 252), on the other hand, on the
basis of her pottery collection from the site, believes the heads to belong
to the early Preclassic period and thus to predate the Olmec examples.
Miles further feels (. cit. 252) that the Monte Alto heads are '"related,
possibly lineally, possibly collaterally" to the Olmec style, while Parsons
and Jenson (oR. cit. 133) believe that there is a "definite relationship
[of the heads] to the Olmec art style, although they are in a class apart."
A third head from Monte Alto of interest to this study is a large
(1.65 m in height) Olmec-like jaguar mask which Parsons and Jenson (1965)
have designated Monument 3. This head is characterized by its arched upper
lip, the scrolls at its mouth corners and on its cheeks, and the angular
appearance of its facial features. While obviously differing from the
classic Olmec heads, it shows a fair degree of similarity to the heads from
Cerro de las Mesas, in the form of-the mask of Monument 2, and to Monument
1 from Laguna de los Cerros, which are discussed above. Monument 3 from
Monte Alto is illustrated by Richardson (1940, pl. XVIIIc) and Parsons and
Jenson (1965, ill. 7 and 8). It is interesting to note that this head also
shows stylistic relationship to such widely diverse sculptures as the Mira-
flores or Arenal phase large stone incensarios from Kaminaljuyu (Monuments
16, 17, 18, for which see Miles 1965, pl. 16a; Villacorta 1938:179), the
undoubtedly chronologically early low relief jaguar head boulder sculptures
from the Departments of Santa Ana and Ahuachapan in El Salvador (Richardson
1940, figs. 33, 34; Spinden 1915, fig. 77), and the coastal slopes of
Guatemala (Miles 1965:247-248, 257). The latter two examples, while tech-
nically not giant heads, are so closely related in concept as to demonstrate
a long standing tradition of head carving on the Pacific slopes as well as
on the southern Gulf coast.
Another group of heads that have been considered by some to be related
to the giant heads are the "cabecitas colosales" (colossal headlets) found in
the Mixteca Baja of the upper Rio Balsas near the town of Acatlan, Puebla
(Paddock 1966:178). According to Paddock, Covarrubias (personal communica-
tion) considered these "rather uncommon" pottery heads to have "a conceptual
similarity to the true cabezas colosales." However, in our opinion, they
appear to be quite different stylistically from the Olmec examples. They are
made of clay, employ applique work for details, and have an entirely differ-
ent style of facial depiction from the Olmec heads, for their oval eyes and
open oval mouths are quite distinct from the Olmec method of mouth and eye
sculpturing. If the fashioning of heads without bodies implies an artistic
connection, then the colossal headlets are related to the Olmec giants;
otherwise, in size, shape, facial features, and artistic style, they are
94
distinct.  The headlets are illustrated by Paddock (o. cit., pis. 199-207).
To conclude, it seems likely that there were Preclassic traditions
of carving naturalistic stone heads both in the Olmec heartland and on the
Pacific slopes of Guatemala and El Salvador, and that these traditions were
stylistically connected. It would further appear that these traditions
persisted over time in each area, even after the cultures that had origin-
ated the carving of the heads had disappeared. That the later Olmec style
of head-carving in the heartland influenced the sculpture of the Pacific
slopes seems probable, and a parallel influence on the Olmec from the- Pacific
slopes is also a fair possibility, at least in late Preclassic times (see
Coe 1965b:769-774). Furthermore, the tradition of head-carving in both areas
persisted even when seemingly modified by art traditions from the Mexican
highlands which did not include monumental heads as part of their artistic
inventory. The major work, however, is yet to be done, for while the
stylistic relationships of the heads parallel the suppositions of many
archaeologists concerning the cultures that made them (see, for example, Coe
1965b; Longyear 1940; Lathrap 1957; Miles 1965; Richardson 1940), they have
not yet been placed in strict chronological relationship, either to their
cultures or to themselves, nor have their cultures been adequately anchored
in time or space. Only when this is done can the suggestions we have made
concerning the relationships of the heads be taken as more than mere possi-
bilities.