Early Maya Writing at San Bartolo, Guatemala

William A. Saturno, et al.

Science 311, 1281 (2006);
DOI: 10.1126/science.1121745

The following resources related to this article are available online at

www.sciencemag.org (this information is current as of March 24, 2008 ):

Updated information and services, including high-resolution figures, can be found in the online
version of this article at:
http://www.sciencemag.org/cgi/content/full/311/5765/1281

Supporting Online Material can be found at:

Downloaded from www.sciencemag.org on March 24, 2008

http://www.sciencemag.org/cgi/content/full/1121745/DC1
A list of selected additional articles on the Science Web sites related to this article can be
found at:
http://www.sciencemag.org/cgi/content/full/311/5765/1281#related-content
This article cites 6 articles, 3 of which can be accessed for free:
http://www.sciencemag.org/cgi/content/full/311/5765/1281#otherarticles

This article has been cited by 2 article(s) on the ISI Web of Science.

This article has been cited by 1 articles hosted by HighWire Press; see:
http://www.sciencemag.org/cgi/content/full/311/5765/1281#otherarticles

This article appears in the following subject collections:

Anthropology
http://www.sciencemag.org/cgi/collection/anthro

Information about obtaining reprints of this article or about obtaining permission to reproduce
this article in whole or in part can be found at:
http://www.sciencemag.org/about/permissions.dtl

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright
2006 by the American Association for the Advancement of Science; all rights reserved. The title Science is a
registered trademark of AAAS.

12. Y. Ohshima, K. Sato, Y. Sumiyoshi, Y. Endo, J. Am. Chem.
Soc. 127, 1108 (2005).
13. C. S. Brauer et al., Chem. Phys. Lett. 401, 420 (2005). 18.
14. M. I. Lester, B. V. Pond, D. T. Anderson, L. B. Harding,
A. F. Wagner, J. Chem. Phys. 113, 9889 (2000). 19.
15. Y. Sumiyoshi, H. Katsnuma, K. Suma, Y. Endo, J. Chem. 20.
Phys. 123, 054324 (2005). 21.
16. K. Suma, Y. Sumiyoshi, Y. Endo, J. Chem. Phys. 121,
8351 (2004). 22.
17. The Gaussian 03 suite of programs (29) and Molpro 23.
2002.6 suite of programs (30) were used for ab initio
calculations. The centrifugal distortion constants were 24.
obtained by B3LYP/aug-cc-pVTZ [the Becke 3-parameter- 25.
Lee-Yang-Parr density functional method using Dunning’s
correlation-consistent polarized triple split valence 26.
basis set with diffuse functions (18)] of Gaussian 03.
Other ab initio calculations with SCF, B3LYP, and 27.
RCCSD(T) (the restricted single and double excitation 28.
coupled-cluster approach with noniterative perturbational
treatment of triple excitations) were performed using
Molpro 2002.6.
R. A. Kendall, T. H. Dunning Jr., R. J. Harrison, J. Chem.
Phys. 96, 6796 (1992).
H. C. Longuet-Higgins, Mol. Phys. 6, 445 (1963).
The character table of the group G4 is given in (23).
Y. Endo, H. Kohguchi, Y. Ohshima, Faraday Discuss. 97,
341 (1994).
M. Iida, Y. Ohshima, Y. Endo, J. Chem. Phys. 94, 6989 (1991).
Materials and methods are available as supporting
material on Science Online.
J. A. Odutola, T. R. Dyke, J. Chem. Phys. 72, 5062 (1980).
W. Klopper, J. G. C. M. van Duijneveldt-van de Rijdt, F. B.
van Duijneveldt, Phys. Chem. Chem. Phys. 2, 2227 (2000).
A. A. Nanayakkara, G. G. Balintkurti, I. H. Williams,
J. Phys. Chem. 96, 3662 (1992).
K. V. Chance et al., J. Mol. Spectrosc. 183, 418 (1997).
K. Suma, W. Funato, Y. Sumiyoshi, Y. Endo, J. Chem. Phys.
122, 184302 (2005).
REPORTS
29. M. J. Frisch et al., Gaussian 03, Revision C.02, Gaussian,
Inc., (Wallingford, CT, 2004).
30. Molpro is a package of ab initio programs written by H.-J.
Werner et al.
31. This work was supported by a Grant-in-Aid for priority
research entitled ‘‘Radical Chain Reactions’’ (grant
no.13127101). K.S. acknowledges the support of the
Research Fellowships of the Japan Society for the Promotion
of Science for Young Scientists (JSPS No. 16-10895).
Supporting Online Material
www.sciencemag.org/cgi/content/full/311/5765/1278/DC1
Materials and Methods
Tables S1 to S4
Fig. S1
References
19 December 2005; accepted 31 January 2006
10.1126/science.1124022

Early Maya Writing at Sub-V platform extends 28 m by 12 m at its

Downloaded from www.sciencemag.org on March 24, 2008

base and supported three separate masonry
rooms. The 2005 excavations established that
San Bartolo, Guatemala its central room was richly decorated and
painted with polychrome murals. The surviving
William A. Saturno,1* David Stuart,2 Boris Beltrán3 doorjamb bears a colorful image of the Maize
God, who is a central character in the mytho-
The ruins of San Bartolo, Guatemala, contain a sample of Maya hieroglyphic writing dating to logical scenes of the later Room 1 murals (4).
the Late Preclassic period (400 B.C. to 200 A.D.). The writing appears on preserved painted walls The line of script was possibly associated with
and plaster fragments buried within the pyramidal structure known as ‘‘Las Pinturas,’’ which was this religiously themed scenery, but its original
constructed in discrete phases over several centuries. Samples of carbonized wood that are closely placement within the room is uncertain.
associated with the writing have calibrated radiocarbon dates of 200 to 300 B.C. This early Maya We obtained accelerator mass spectrometry
writing implies that a developed Maya writing system was in use centuries earlier than previously (AMS) radiocarbon dates on five charcoal sam-
thought, approximating a time when we see the earliest scripts elsewhere in Mesoamerica. ples from sealed deposits in the three archi-
tectural strata (Sub-VI, Sub-V, and Sub-IV) in
esearch on the origins of Maya hiero- One example of this earlier painting comes order to bracket the age of the painted blocks

R glyphic writing has long been hindered by

the paucity of good archaeological con-
texts and reliable dates for inscribed artifacts and
from a block from a dismantled wall of the
building that once stood on the platform of the
Sub-V construction phase (Fig. 2). The Room
(Fig. 3). The first of these—from within the floor
of the Sub-VI platform, the construction phase
that was encapsulated by Sub-V construction—
monuments. With a few exceptions, examples of 1 murals were painted on the Sub-I phase of provides a maximum uncalibrated radiocarbon
archaic Maya script appear on illicitly excavated the pyramid, that is, four construction episodes date of 2260 T 40 years before present (years
objects that can be stylistically dated to no earlier later than the Sub-V phase. The È4-m-high B.P.) E400 to 200 B.C.; 2s (95% probability)
than about 100 B.C. to 100 A.D., when writing
seems to have been already well established
elsewhere in Mesoamerica. Here we provide
new evidence of early Maya writing preserved
in the ruins of San Bartolo, Guatemala.
The ruins of San Bartolo, Guatemala (Fig. 1)
were identified in 2001 and include early wall
paintings buried within a pyramidal structure
today known as BLas Pinturas.[ These had been
partially exposed by illicit digging a few years
previously, and subsequent scientific excavations
in Room 1 (as that location is now designated)
have uncovered most of this important mural,
dating to È100 B.C. (1–4) (figs. S6 to S10).
Tunneling deeper into the Las Pinturas structure
has since led to the discovery of other buildings
with remains of painted decoration that are
substantially older than the Room 1 murals.

1 Fig. 1. Map. San Bartolo in relation to other Maya

Department of Anthropology, University of New Hamp-
shire, Durham, NH 03824, USA. 2Department of Art and
archaeological sites. [Drawing by J. Kowan and W.
Art History, University of Texas at Austin, Austin, TX 78712, Saturno]
USA. 3Escuela de Historia, Universidad de San Carlos,
Guatemala City, Guatemala.
*To whom correspondence should be addressed. E-mail:
wsaturno@unh.edu

www.sciencemag.org SCIENCE VOL 311 3 MARCH 2006 1281

 REPORTS
Fig. 2. Glyph block. The Sub-V painted block in
situ. [Photograph by B. Beltrán]
calibrated range^ (fig. S1). A sample from within
the floor of Sub-V dates the construction of the
room at 2200 T 60 years B.P. uncalibrated E390
to 80 B.C.; 2s (95% probability) calibrated
range^ (fig. S2). The final three samples with
dates of 2260 T 40 years B.P. uncalibrated E400
to 200 B.C.; 2s (95% probability) calibrated
range^, 2180 T 40 years B.P. uncalibrated E370
to 100 B.C.; 2s (95% probability) calibrated
range^, and 2150 T 40 years B.P. uncalibrated
E360 to 60 B.C.; 2s (95% probability) calibrated
range^ (figs. S3 to S5) surround the painted
blocks and relate contextually to both the
destruction of the Sub-V painted room and
the subsequent construction of the Sub-IV
platform above it. Taken together, these samples
and those analyzed in association with the final
two phases of construction imply that the text
was painted between 300 and 200 B.C.
The painted block bears a column of 10
hieroglyphs (Fig. 4). The text appears to be the
end of a longer sequence of signs that con-
tinued above. All are painted in a thick black
line on white plaster, apparently along a subtle
pinkish-orange stripe that served as a guideline
for the scribe. As with later examples of Maya
writing discovered at San Bartolo, its decipher-
ment remains a challenge (4). Later texts from
the Room 1 murals are only partially readable,
because sign forms appear considerably differ-
ent from the familiar elements of later Maya
script. The San Bartolo Room 1 paintings date
centuries before the first fully legible Maya
writing from around 250 to 300 A.D., and the
signs of the Sub-V block are older still, con-
taining archaic forms.
The one fully recognizable glyph (pA7) is
an early version of the sign that reads AJAW,
a ubiquitous title in Maya texts that means
Blord,[ Bnoble,[ or Bruler.[ It evidently
formed part of a more extended title phase
in reference to some person, either historical or
mythical. Some signs have qualities that might
be vaguely pictorial, such as pA2 with its
suggestion of a hand holding a brush or
alternatively a sharp bloodletter. Other signs
are more abstract-looking forms, probably
ancestral to components of later Maya script.
In their overall appearance, the text bears
1282 3 MARCH 2006
Fig. 3. Las Pinturas. Ar-
chitectural profile illus-
trating AMS radiocarbon
dates for the construction
sequence of the structure,
the location of the Sub-V
building phase, the painted
glyph block, and the
Room 1 mural. Scale in
meters. [Drawing by J.
Kowan and W. Saturno]
some resemblance to the so-called Epi-Olmec
script used by neighboring peoples to the
west during the Late Preclassic and Early
Classic periods (5, 6). All examples of that
script postdate the San Bartolo block, how-
ever, raising the question of the direction in
which any influence may have flowed.
Preclassic writing from the Maya area is
scarce and has been difficult to date accurately.
Most other examples are known from stone
monuments found in surface or near-surface
contexts or from illicitly excavated portable
objects. One notable early inscription from El
Mirador probably dates to no earlier than 100
BC on the basis of stylistic comparisons (7).
Another carved monument with glyphs from
El PortFn, Guatemala, may date to the first
two or thee centuries B.C., on the basis of a
single radiocarbon date not in direct associ-
ation with the stone (8). The newly discov-
ered San Bartolo text can now be firmly dated
to the same general period, and its fine pres-
ervation offers an unusual look at the form
that Maya script assumed in its early history.
Fig. 4. Painted hiero-
glyphs. Scale drawing of
Sub-V painted glyph block.
Glyphs assigned prelim-
inary column and row
designations. Scale in cen-
timeters (pA 1 to 10).
[Drawing by D. Stuart]
VOL 311 SCIENCE www.sciencemag.org
The San Bartolo text raises the question of
Downloaded from www.sciencemag.org on March 24, 2008
the relation between Maya writing and other
early script traditions in Mesoamerica. In the
Preclassic era, writing systems were firmly
established by about 400 B.C. among complex
cultures in what is now Oaxaca and perhaps in
the Isthmus of Tehuantepec (9–12), although
the dating of evidence for this remains contro-
versial (13–15). It now appears that the Maya
also participated in the Preclassic cultures of
literacy, and at a much earlier date than pre-
viously believed.
References and Notes
1. W. Saturno, Symbols (Peabody Museum of Archaeology
and Ethnology, Harvard University, Fall 2002), p. 3.
2. W. Saturno, D. Stuart, H. Escobedo, I. Graham,
‘‘Reonocimiento Arqueológico y Conservación de San
Bartolo, Guatemala’’ (Instituto de Antropologı́a e Historia,
Guatemala City, Guatemala, 2001).
3. T. O’Neill, Natl. Geogr. Mag. 201, 70 (2002).
4. W. Saturno, K. Taube, D. Stuart, The Murals of San
Bartolo, Guatemala, Part 1, The North Wall (Publ. 7,
Ancient America, Barnardsville, NC, 2005).
5. F. Winfield Capitaine, La Estela 1 de La Mojarra, Veracruz
(Publ. 16, Research Reports on Ancient Maya Writing,
Washington, DC, 1988).
6. J. S. Justeson, T. Kaufman, Science 259, 1703 (1993).
7. R. D. Hansen, An Early Maya Text from El Mirador,
Guatemala (Research Reports on Ancient Maya Writing,
Washington, DC, 1991).
8. R. J. Sharer, D. W. Sedat, Archaeological Investigations in
the Northern Maya Highlands (University Museum, Univ.
of Pennsylvania, 1987), pp. 49–73.
9. A. Caso, Handb. Mid. Am. Indians 3, 931 (1965).
10. J. Marcus, Annu. Rev. of Anthropol. 5, 35 (1976).
11. J. Marcus, Mesoamerican Writing Systems:
Propaganda, Myth and History in Four Ancient
Civilizations (Princeton Univ. Press, Princeton,
NJ, 1992), pp. 41–42.
12. M. E. D. Pohl, K. O. Pope, C. von Nagy, Science 298,
1984 (2002).
13. R. Cahn, M. Winter, Indiana 13, 39 (1993).
14. S. D. Houston, in First Writing: Script Invention as History
and Process (Cambridge Univ. Press, Cambridge, 2005),
p. 293.
15. K. V. Flannery, J. Marcus, Proc. Natl. Acad. Sci. U.S.A.
100, 11801 (2003).
16. Excavations by the Proyecto San Bartolo between 2002
and 2005 were supported by grants from the National
Endowment for the Humanities (grant RZ- 50086); National
Geographic Society, Committee for Research and Exploration
(grants 7065-01, 7222-02, 7393-03, 7601-04, and
 REPORTS
7721-04); Foundation for the Advancement of Mesoamerican Ministerio de Cultura y Deportes, Instituto de Antropologı́a Figs. S1 to S10
Studies, Inc. (grants 01038 and 02039); Peabody Museum, e Historia, and Departamento de Monumentos Prehispánicos References and Notes
Harvard University; Dumbarton Oaks Research Library and for their support.
Collections; American Philosophical Society; Annenberg 24 October 2005; accepted 21 December 2005
Foundation; International Communities Foundation; Supporting Online Material Published online 5 January 2006;
Brigham Young University; and most importantly, the www.sciencemag.org/cgi/content/full/1121745/DC1 10.1126/science.1121745
Reinhart Foundation. We also thank the Guatemalan Materials and Methods Include this information when citing this paper.

Toward Automatic Reconstruction Theoretical and practical limits to reconstructing a

tree of life have been put forward, such as the
insufficient amount of discriminating characters
of a Highly Resolved Tree of Life available, even in information-rich genomic data
sets (4), and the computing resources required to
Francesca D. Ciccarelli,1,2,3* Tobias Doerks,1* Christian von Mering,1 Christopher J. Creevey,1 cope with large numbers of species (1). Further-
Berend Snel,4 Peer Bork1,5† more, there are factors that hamper accurate
reconstruction of phylogenetic trees regardless
We have developed an automatable procedure for reconstructing the tree of life with branch of the methods used, such as sampling biases of
lengths comparable across all three domains. The tree has its basis in a concatenation of 31 species included (5) and dilution of phylogenetic
orthologs occurring in 191 species with sequenced genomes. It revealed interdomain discrepancies signal by horizontal gene transfer (HGT) (6), the

Downloaded from www.sciencemag.org on March 24, 2008

in taxonomic classification. Systematic detection and subsequent exclusion of products of
1
horizontal gene transfer increased phylogenetic resolution, allowing us to confirm accepted European Molecular Biology Laboratory, Meyerhofstrasse
1, 69012 Heidelberg, Germany. 2Department of Experi-
relationships and resolve disputed and preliminary classifications. For example, we place the mental Oncology, European Institute of Oncology, Via
phylum Acidobacteria as a sister group of d-Proteobacteria, support a Gram-positive origin of Ripamonti 435, 20141 Milan, Italy. 3Institute of Molecular
Bacteria, and suggest a thermophilic last universal common ancestor. Oncology Foundation, Fondazione Italiana per la Ricerca
sul Cancro, Via Adamello 16, 20139 Milan, Italy. 4Center
for Molecular and Biomolecular Informatics, Nijmegen
econstructing the phylogenetic relation- been questioned recently (2, 3). Moreover, even Center for Molecular Life Sciences, Radboud University

R ships among all living organisms is one

of the fundamental challenges in biolo-
gy. Numerous attempts to derive a tree of life
using various methods have been published Efor
a review, see (1)^, and its principal existence has
under the assumption of a tree of life, numerous
groupings and taxonomic entities still remain
heavily debated, and the advent of molecular and
genomic data has increased the variety of clas-
sifications rather than reducing the problem (1).
Nijmegen Medical Center, Toernooiveld 1, 6525 ED,
Nijmegen, Netherlands. 5Max Delbrück Centrum, 13092
Berlin-Buch, Germany.
*These authors contributed equally to this work.
†To whom correspondence should be addressed. E-mail:
peer.bork@embl.de

Fig. 1. Overview of the procedure. The white boxes represent the major steps
for building the pan-domain phylogeny presented here. Steps in gray represent
automatable parts of the procedure that need to be carried out for including
further species. For the 31 clusters of orthologous groups (COGs) used in the
analysis, we manually derived 1:1 orthologs by removing mitochondrial and
chloroplast paralogs from corresponding multiple alignments. We built domain-
specific alignments by using corresponding proteins encoded by the 31
orthologs and aligned the resulting profiles. With this procedure, we maximized
the number of positions of the global alignment and reduced the number of
misaligned residues. For a detailed description of the methods, see (8).

www.sciencemag.org SCIENCE VOL 311 3 MARCH 2006 1283