Genetics and the History of Latin America
Andrés Reséndez and Brian M. Kemp
Given that we live in the much-vaunted era of the “genomics revolution,” one
cannot help but wonder whether those long strands of DNA will ever wash up
on the historian’s lonely shore. Sure, we are all aware of one or two high-profile
instances when genetic evidence was able to change a historical interpretation.
The case that most readily comes to mind is the Y-chromosome DNA test conducted by a retired pathologist in Charlottesville, Virginia, in 1998—intended
to settle the question of whether Thomas Jefferson had fathered a child by his
slave Sally Hemings.1 Among other things, the ensuing furor and soul-searching
produced a spate of works reexamining slavery, miscegenation, and the connections between the private and public spheres in early America.2
Still, we (professional historians) ordinarily regard DNA work as a mere
curiosity. For one thing, this type of inquiry has been carried out by outsiders:
retired physicians, forensic experts, history buffs, and the like. But more to the
point, DNA’s strong suit is “whodunit” types of questions—such as whether
President Jefferson sired any children with his slave or whether the famous outlaw Jessie James faked his own death, attended his funeral, and went on to live
quietly for many years. (Disappointingly, DNA evidence fails to support this
legend.) Such projects, dreamed up mostly by amateurs and designed to solve
lingering historical mysteries, appear far removed from the more sweeping and
ethereal social, political, economic, and cultural explanations that are the hallmark of our academic discipline today.
At the other end of the spectrum, those of us working on topics pertaining
to the history of Latin America in the last five hundred years may be aware of
We want to thank Emilio Ferrer, Angélica González Oliver, Nara Milanich, Cara Monroe,
Ripan S. Malhi, David Glenn Smith, and Chuck Walker for their valuable comments and
suggestions. We also express our gratitude to the UC Mexus–Conacyt Program and the
Wenner-Gren Foundation for financial support of this research.
1. The original story appeared in the U.S. News & World Report, 9 Nov. 1998.
2. On the many ripples created by the Jefferson-Hemings relationship, see Gordon S.
Wood, “Slaves in the Family,” The New York Times Book Review, 14 Dec. 2003.
Hispanic American Historical Review 85:2
Copyright 2005 by Duke University Press
284
HAHR / May / Reséndez and Kemp
the work being done by geneticists and molecular anthropologists on ancient
America. Since DNA yields information that no historical, archaeological, or
linguistic source can provide, it is not surprising that much of the pioneering
genetic work has been focused on the remote past, broaching subjects such as
the peopling of the Americas, the early process of group formation, and the
assessment of ancestor-descendent relationships among pre-Columbian indigenous groups. Even so, DNA scholarship has occasionally moved squarely into
the historians’ turf. The work of Juan Carlos Martínez Cruzado and his group
of researchers at the University of Puerto Rico is a case in point.3 This team
started out in the 1980s by studying the mitochondrial DNA (mtDNA) of some
pre-Columbian skeletal remains found on the island. To contextualize this data,
the group also conducted mtDNA tests among living Puerto Ricans. Astoundingly, Martínez Cruzado and his collaborators concluded that 53 percent of
islanders have indigenous ancestry through the maternal line. To say the least,
these results contradict the dominant historical narrative of annihilation of the
indigenous population of Puerto Rico by the middle of the sixteenth century.
Can these findings be somehow explained by subsequent indigenous migration
to the island, or is it time to critically reassess the standard historical narrative
and the colonial sources on which it rests?
In this report, we will survey the DNA literature most relevant to historians
of Latin America (especially those working in Mexico and the American Southwest) and discuss how the genetic evidence jibes with our socially constructed
notions of race and ethnicity. We will be selective in our coverage by necessity,
but we also point readers to a fairly comprehensive list of references available at
http://resendez.ucdavis.edu. We hope this piece can serve as a bibliographical
resource for all historians curious about how their own interpretations square
with the DNA evidence.
One prefatory clarification is in order here. We do not take the view that
DNA data should take precedence over other kinds of textual, linguistic, or
archaeological evidence. This is an especially important and sensitive matter in
these heady times, when DNA is accorded so much authority that it can singlehandedly condemn or exonerate individuals accused of serious crimes. Instead,
we advocate a brand of methodological eclecticism in which different lines of
evidence are cross-checked against each other and used to advance interpretations consistent with multiple sources of data.
3. Martínez-Cruzado et al., “Mitochondrial DNA Analysis in Puerto Rico,” Human
Biology 73 (2001): 491–511.
Genetics and the History of Latin America
285
An Emerging Genetic Map of the Continent
DNA may well be the ultimate archival repository, containing clues about largescale population movement, conquest, sexual and reproductive patterns, and
even group identities—after all, with whom we have children constitutes one
important way to define community. Historians may be put off by DNA’s focus
on sample sizes and statistics. But on the positive side, while oral and written
sources limit us to the most literate social groups and to the last five or six hundred years (two thousand or so if we include Maya glyphs), DNA evidence sheds
light on the most numerous and ordinary individuals—representing a bottomup approach as it were—and at the same time extends our scope to the first
inhabitants of the Americas and beyond.
For historians, DNA scholarship’s most significant contribution comes
in the form of a genetic map. Virtually all indigenous peoples of the Americas have been found to cluster into one of four founding maternal lineages—
haplogroups—determined by mutations in the mtDNA.4 These four founding
lineages are so widespread and ancient that they can be identified among contemporary Native Americans as well as in samples from individuals who lived
thousands of years ago. DNA scholars conveniently—if somewhat unimaginatively—call them haplogroups A, B, C, and D.5 These four founding lineages
can be traced back to Asia, where they coexist with others, but are quite different from those of Europe and Africa.6 This simply indicates that the early
inhabitants of the Americas came originally from Asia but that not all Asian
4. We are hesitant to use “mutation” because the term may have negative connotations.
DNA scholars often use the more neutral but mystifying “polymorphism.” In this paper we
will continue to refer to mutations but in the understanding that they do not generally affect
the performance of the individuals or lineages that have them. Although much of the early
work has been conducted on the mitochondrial genome, it is possible to create other genetic
maps based on Y-chromosome information or on nuclear DNA, as discussed later.
5. Although much of the discussion in the DNA literature revolves around haplogroups
A, B, C, and D, about 3 percent of Native Americans belong to a fifth haplogroup called
X. Interestingly, haplogroup X has also turned up in Europe and in Central Asia, thus
fueling speculation, until quite recently, about whether Europeans may have contributed
to the genetic makeup of the New World before Columbus. The latest work indicates that
haplogroup X, as found among Native Americans, is not closely related to the European
haplogrup X variant. See M. Reidla et al., “Origin and Diffusion of mtDNA Haplogroup X,”
American Journal of Human Genetics 73 (2003): 1178–90.
6. The main haplogroups in Asia are A, B, C, D, F, G, and Z; in Europe, they are H, I
J, K, T, U, V, W, and X; and in Africa, they are L1, L2, and L3.
286
HAHR / May / Reséndez and Kemp
haplogroups made it to America, thereby resulting in reduced genetic diversity
in the New World.7
Two features make mtDNA ideally suited for historical and anthropological
investigations.8 First, it is passed down solely from mother to child and without
recombination. A woman gets an identical copy of her mother’s mtDNA, and,
in turn, she will pass it on to her children. In other words, when we determine
a mtDNA lineage, we are in fact peering at a genetic legacy passed through an
unbroken line of female ancestors going back for millennia. In this sense, we
can conceive of haplogroups A, B, C, and D as the indelible imprints of four pioneering native women whose progenies gave rise to nearly all of the indigenous
peoples of the New World.9 It is important to emphasize that while the time
depth revealed by mtDNA is astonishing, its ability to predict physiognomic or
racial traits or capture genetic mixture is limited at best, because it traces only
one line of descent (from grandmother to mother to daughter, etc.) while ignoring all other genetic contributions to the individual. In other words, a person
may look—and indeed be—decidedly African or European and still possess one
of the four Native American matrilines or vice versa.
The second feature that makes mtDNA especially relevant for historians is that one region of the molecule mutates rapidly.10 As mtDNA is copied
with each succeeding generation, slight copying errors (mutations) are inevitably introduced. Bases can be added, deleted, changed, or simply jumbled. As
mutations accumulate, patterns emerge in branchlike fashion. Due entirely to
chance, certain mutations become associated with certain populations and are
absent in others.
We cannot know for sure when a particular mutation first occurred, and
therefore DNA provides only a crude historical clock. But we do know that
some lineages are more widespread and basic than others, indicating that the
former occurred earlier than the latter. We can thus think of mtDNA as a large,
sprawling tree. At the trunk we will find mutations that are common to all liv7. There is a large literature devoted to the origins and initial peopling of the
Americas. For an insightful and recent survey, see Herbert S. Klein and Daniel C.
Schiffner, “The Current Debate about the Origins of the Paleoindians of America,” Journal
of Social History 37, no. 2 (2003): 483–92.
8. Mitochondria are elliptical-shaped organelles that provide energy to the cell. Each
mitochondrion carries two copies of a small circular strand of DNA.
9. This does not mean that the continent was peopled by four women only. It simply
means that the mtDNA lineages of the other women became subsequently extinct.
10. Called the “hypervariable region,” this feature allows one to make differentiations
between closely related populations, such as Native Americans.
287
Genetics and the History of Latin America
Table 1. Haplogroup Frequencies of Selected Populations in the Hemisphere.
Population
Kuna (Panama)
Nahua-Cuetzalán (central Mexico)
Tohono O’odham (Arizona)
Yanomama (Brazil and Venezuela)
N
%A
%B
%C
%D
%X
89
31
37
110
79.8
61.3
0
2.7
20.2
32.3
56.8
11
0
6.5
37.8
39
0
0
5.4
47.3
0
0
0
0
Source: Compiled from a list of sources available at http://resendez.ucdavis.edu.
ing humans and that must have been present when our ancestors still constituted one single group in Africa, prior to the momentous dispersal to the other
continents. Or we may choose to train our sight on more recent branches, such
as the founding lineages of the indigenous peoples of the New World, associated
with mutations introduced just prior to or immediately after the initial peopling
of our continent. It is further possible to look at subbranches and even twigs on
the basis of even more recent mutational events that may be associated with the
formation of populations or language groups. Indeed, each haplogroup can be
further subdivided to a level of resolution that can give us clues about events in a
time horizon more consistent with that commonly employed by historians.11
What can the geographic distribution of mtDNA variation teach us about
the history of Latin America? Thus far, scholars have studied genetic markers of indigenous groups in all major areas of Latin America, including South
America, Central America, the Caribbean, Mexico, and the American Southwest, and they have found that haplogroups A, B, C, and D are widely dispersed
throughout the hemisphere.12 Even a cursory look at the mtDNA of such diverse
populations as the Yanomama of Brazil and Venezuela, the Kuna of Panama, the
Nahua of central Mexico, and the Tohono O’odham of Arizona (just to name
a handful) shows that each population contains at least three and often all four
founding lineages, although the proportions vary widely from one population
to another as a result of their separate histories (see table 1).
Most immediately, the wide distribution of haplogroups has a bearing on
discussions about the initial peopling of the Americas. Studies in the 1980s and
1990s, based on early mtDNA findings, as well as linguistic and craniomet11. For example, in another context Stephens has established a link between a nuclear
DNA mutation in the CCR5 gene and the bubonic plague in Europe. See J. C. Stephens et
al., “Dating the Origin of the CCR5-Delta-32 Aids-Resistance Allele by the Coalescence of
Haplotypes,” American Journal of Human Genetics 62 (1998): 1507–15.
12. For a comprehensive list of DNA articles of each of these regions, readers can turn
to http://resendez.ucdavis.edu.
288
HAHR / May / Reséndez and Kemp
ric evidence, argued for three distinct migration events from Asia to the New
World.13 More recent DNA scholarship has cast doubts on multiple-migration
models. While no scholarly consensus has been reached, many DNA researchers
now favor explanations that involve a single migration and subsequent expansions or radiations throughout the hemisphere.14 The most obvious objection
to multiple-migration models is the widespread distribution of the four founding lineages throughout the Americas. Moreover, the four haplogroups evince
a similar amount of accumulated diversity, suggesting a common time depth in
the New World for all. In short, it is unlikely that multiple migrations—totally
unrelated to one another, taking place thousands of years apart, and possibly
originating in different places—would have given rise to a scenario in which
four haplogroups were so widely distributed among virtually all of the native
populations of the continent.15
The History of Mexico and the
American Southwest as Revealed by Genes
Studying genetic markers in narrower regions has much to add to our understanding of history. Some tantalizing patterns emerge in the case of Greater
Mexico, for example. In the core regions of Mexico—roughly coterminous with
the extent of Mesoamerica—haplogroup A predominates, with additional but
smaller percentages of the other haplogroups. DNA studies conducted among
Nahua, Maya, Mixe, Zapotec, and Mixtec populations show significant genetic
similarities (see table 2).16 The similarity of haplogroup frequencies across these
13. Joseph H. Greenberg, Christy G. Turner II, and Stephen L. Zegura, “The
Settlement of the Americas: A Comparison of the Linguistic, Dental, and Genetic
Evidence,” Current Anthropology 27, no. 5 (1986): 477–97. Interestingly, Zegura—one of
the three authors who originally launched the “three-wave model”—has come back full
circle and now supports a single migration. S. L. Zegura et al., “High-Resolution Snps
and Microsatellite Haplotypes Point to a Single, Recent Entry of Native American Y
Chromosomes into the Americas,” Molecular Biology and Evolution 21 (2004): 164–75.
14. See especially W. A. Silva et al., “Mitochondrial Genome Diversity of Native
Americans Supports a Single Early Entry of Founder Populations into America,” American
Journal of Human Genetics 71 (2002): 187–92.
15. Thus far, we have largely been concerned with mtDNA-based studies. DNA
scholars have also begun investigating Y-chromosome DNA, as well as autosomal
(i.e., nuclear) DNA markers. Like mitochondrial DNA, the Y chromosome has a
nonrecombining region—that is, a region that is passed down unchanged—this time
through the paternal line. For a review of recent Y-chromosome literature, see Klein and
Schiffner, “The Current Debate,” 485–88.
Genetics and the History of Latin America
289
groups is remarkable if we bear in mind that they represent all four major linguistic families of Mesoamerica—Uto-Aztecan, Mayan, Mixe-Zoquean, and
Otomanguean, respectively. Linguists believe that each of these language families has a time depth that runs in the thousands of years and that they are thus
not closely related. Moreover, comparisons of ancient and modern DNA data
show that haplogroup frequency distributions have remained relatively stable
for at least six hundred years (the demographic catastrophe of the early colonial
period notwithstanding) and perhaps considerably longer. How can we explain
these long-term genetic similarities in the face of profound linguistic differences? One possibility is that all major indigenous civilizations that flourished
in Mesoamerica descend from a single ancient population that already bore a
large frequency of haplogroup A. Recent mtDNA testing of the skeletal remains
of an 11,000-year-old individual found in Mexico City lends some support to
this idea.17 Alternatively, one can theorize that the continuous human interaction within Mesoamerica in the form of warfare, trade, and successive macroregional political arrangements—including those centered at sites like La Venta,
Teotihuacán, Tula, and Tenochtitlán—may have, over time, blurred whatever
genetic differences originally existed.
In stark contrast to the mtDNA profile of Mesoamerica, in the American
Southwest haplogroup A is generally absent while haplogroup B predominates
(see table 2).18 Indeed, populations in the Southwest exhibit some of the highest frequencies of haplogroup B to be found anywhere in the New World. For
instance, among River Yuman in southwestern Arizona, fully 63 percent of the
population exhibits haplogroup B, while among the peoples of Jémez Pueblo in
New Mexico an overwhelming 88 percent do so.19 With the limited mtDNA
information at our disposal, it would thus appear that substantial genetic dif16. Admittedly, sampling is still spotty, and sample sizes are generally small. Larger
and more comprehensive studies carried out in the future may qualify these sweeping
results.
17. Description and dating of these ancient remains in Silvia González et al., “Earliest
Humans in the Americas: New Evidence from Mexico,” Journal of Human Evolution 44
(2003): 379–87. Angélica González-Oliver (personal communication) has performed the
mtDNA analysis and concluded that it belongs to haplogroup A.
18. For a comprehensive survey of mtDNA in the Southwest, see R. S. Malhi et al.,
“Native American mtDNA Prehistory in the American Southwest,” American Journal of
Physical Anthropology 120 (2003): 108–24.
19. Joseph G. Lorenz and David G. Smith, “Distribution of Four Founding mtDNA
Haplogroups among Native North Americans,” American Journal of Physical Anthropology
101 (1996): 307–23; David G. Smith et al., “Distribution of mtDNA Haplogroup X among
Native North Americans,” American Journal of Physical Anthropology 110 (1999): 271–84; and
Malhi et al., “Native American mtDNA Prehistory.”
290
HAHR / May / Reséndez and Kemp
Table 2. Haplogroup Frequencies of Selected Populations in North America
Population
N
%A
%B
%C
%D
%X
Aztecs (Tlatelolco, Mexico City)1
Cora (Nayarit)
Huichol (Nayarit)
Ancient Maya (Quintana Roo)1
Contemporary Maya (Yucatán)
Highland Mixe (Ayutla, Oaxaca)
Alta Mixtec (Nochixtlán, Oaxaca)
Baja Mixtec (Huajuapán, Oaxaca)
Atocpan Nahua (Milpa Alta)
Cuetzalán Nahua (Puebla)
Tarahumara (Durango)
Valley Zapotec (Oaxaca)
Akimal O’odham (southeast Arizona)
Anasazi (U.S. Southwest)1
Cochimi (southern Baja California)
Delta Yuman (western Arizona)
Fremont (Great Salt Lake)1
Jémez (northwestern New Mexico)
Kiliwa (northern Baja California)
Pai Yuman (northern Baja California)
River Yuman (western Arizona)
Tohono O’odham (Arizona)
Zuni (northwestern New Mexico)
23
81
52
24
26
16
15
14
49
31
72
15
43
25
13
23
30
36
7
27
22
37
26
65.2
28.4
34.6
87.5
53.8
62.5
73.3
92.9
38.8
61.3
33.3
33.3
4.7
24
7.7
0
0
0
0
7.4
0
0
15.4
13
51.9
53.8
4.2
23.1
31.3
13.3
7.1
40.8
32.3
29.2
33.3
53.5
60
46.2
56.5
80
88.9
100
66.7
63.6
56.8
76.9
4.3
16
11.5
8.3
15.4
6.3
13.3
0
16.3
6.5
31.9
33.3
39.5
16
46.2
43.5
13.3
2.8
0
25.9
36.4
37.8
7.7
17.4
3.7
0
0
7.7
0
0
0
4.1
0
5.6
0
0
0
0
0
6.7
0
0
0
0
5.4
0
0
0
0
0
0
0
0
0
0
0
0
0
2.3
0
0
0
0
8.3
0
0
0
0
0
1Precontact
population
Source: Compiled from a variety of sources found in http://resendez.ucdavis.edu.
ferences exist between the indigenous peoples of Mexico and those of the
Southwest, even though we know that architectural fashions, cultigens, trade
goods, and at least some population groups have historically moved across these
regions. Have the great deserts of northern Mexico constituted a real barrier
through the ages preventing the widespread movement and mixing of peoples
(see graph 1)?
Additional research on the biological connections between Mesoamerica
and the American Southwest adds new twists. Given the apparent lack of shared
matrilines, scholars have studied other genetic markers to confirm or challenge
these results. David Glenn Smith and his team have studied a unique type of
human serum known as albumin Mexico. Although exceedingly rare, albumin
Mexico is found in several linguistically unrelated groups throughout Mexico
and also in various communities of the Southwest, thus providing clear evidence
291
Genetics and the History of Latin America
Graph 1. Principal
=DSRWHF
7DUDKXPDUD
Component Analysis
&RFKLPL
of Haplogroup
$NLPDO2 RGKDP
QG3ULQFLSOH&RPRSRQHQW
'HOWD<XPDQ
7DRQR
2 RGKDP
5LYHU<XPVQ
0D\D&RQWHPSRUDU\
0L[WHF$OWD
Frequencies of
Selected Groups
from Mexico and the
American Southwest.
3DL<XPDQ
0D\D;FDUHW
$]WHFV
1DKXD$WRFSDQ
0L[WHF%DMD
0L[H
)UHPRQW
1DKXD
&XHW]DODQ
$QDVD]L
-HPH]
&RUD
.LOLZD
=XQL
+XLFKRO
VW3ULQFLSOH&RPRSRQHQW
of a biological connection. Smith and his collaborators believe that albumin
Mexico may have originated more than three thousand years ago somewhere
in central Mexico and subsequently moved northward along the Tepiman corridor, spreading to various groups associated with the Hohokam culture in what
is now the American Southwest.20 If this analysis is correct, then how can we
reconcile the cross-regional distribution of albumin Mexico with the apparent
lack of shared matrilines between Mexico and the Southwest? One possible
explanation is that much of the human interaction between the two macroregions has occurred via males—such as long-distance merchants (like the Aztec
pochteca), all-male armies, or predominantly male migrations.21 It is also possible
that more thorough sampling of Mexican and Southwestern indigenous groups
and investigation of mtDNA on a finer scale will eventually reveal the elusive
shared matrilines.
These results are especially relevant in light of the recent discussion among
ethnohistorians, linguists, and archaeologists concerning the origins and spread
20. David G. Smith et al., “Implications of the Distribution of Albumin Naskapi and
Abumin Mexico for New World Prehistory,” American Journal of Physical Anthropology 110
(1999): 271–84.
21. Precisely this scenario is advanced in R. S. Malhi et al., “Native American mtDNA
Prehistory.”
292
HAHR / May / Reséndez and Kemp
of the Uto-Aztecan language family. When Spanish conquistadors made their
way into Central Mexico in the early sixteenth century, they found that Nahuatl
(a member of the Uto-Aztecan family) was the most common language within
the Aztec Empire, functioning as a lingua franca in the core region of Mesoamerica. But while contemporary scholars agree on the preeminence of Nahuatl
at the time of contact, they disagree about its origins. Most scholars subscribe
to the notion that Proto-Uto-Aztecan (and therefore Proto-Nahuatl) speakers
originated outside of Mesoamerica—in northwestern Mexico or farther north—
and moved south, populating central Mexico in successive migratory waves that
took place roughly over the thousand-year period prior to the Spanish arrival.22
Linguist Jane Hill has called this interpretation the northern origin model. Yet
some scholars, including Hill, have recently proposed an alternative southern
origin model, arguing that Proto-Uto-Aztecan peoples lived within Mesoamerica at least since the beginnings of maize domestication more than five thousand
years ago.23 The supporters of this interpretation reverse completely the direction of the language spread and contend instead that it started out in central
Mexico and subsequently expanded to the northwest and beyond on the basis
of agricultural technology. These scholars, who have not overturned current
scholarly consensus favoring a northern origin, have nonetheless produced relevant linguistic evidence that merits further investigation. More DNA evidence
would help elucidate this ongoing debate that is so fundamental to the overall
history of early Mexico.
The efforts of our own research team at the University of California,
Davis, have also been geared toward further exploring the genetic connections
between Mesoamerica and the Southwest. Specifically, we have studied the
mtDNA of geographically intermediate populations, including Coras, Huicholes, and Tarahumaras (see table 2). We have found that these groups have high
frequencies of both haplogroups A and B, as befits “frontier” or “contact” populations. Yet, our preliminary findings indicate that the transition of haplogroup
frequencies from the Southwest to Mesoamerica does not occur uniformly and
smoothly following a north-south gradient. Instead, we have discovered nota22. Christopher S. Beekman and Alexander F. Christensen have cogently and
concretely laid out this interpretation in “A Synthetic Analysis of the Nahua Migrations”
(ms).
23. Jane H. Hill, “Proto-Uto-Aztecan: A Community of Cultivators in Central
Mexico?” American Anthropologist 103, no. 4 (2001): 913–34. The basic notion is that large
linguistic families such as Indo-European, Austronesian, and Uto-Aztecan must have
grown at the expense of other languages and on the basis of technological advances such as
plant domestication.
Genetics and the History of Latin America
293
ble frequency differences and have even detected the presence of haplogroup
enclaves. For instance, while haplogroup A tends to predominate in Nahuaspeaking populations, such as those of precontact Tlatelolco and modern-day
Cuetzalán, Puebla, we have been able to identify at least one Nahua community—San Pedro Atocpan, right outside Mexico City in the district of Milpa
Alta—where haplogroup B constitutes the largest percentage. Our long-term
goal is to add sufficient detail to this rough genetic map to enable scholars to
pinpoint instances in which haplogroup-A–bearing Mesoamerican groups made
inroads into the Southwest and, conversely, identify cases where haplogroupB–bearing peoples of the Southwest (conceivably Chichimecs, Teochichimecs,
and their forebears) migrated south into the core regions of Mesoamerica.
One final way to tackle this problem is by focusing on specific mutations
that arose recently and can be credibly linked to historical phenomena. We have
identified what we believe is a particularly promising genetic marker—a derived
form of haplogroup A that carries a nine base-pair deletion in the mitochondrial
genome. Our interest in this marker was sparked after studying the remains of
27 “Aztec” individuals from precontact Tlatelolco. Three of these individuals
carried this derived form of haplogroup A.24 A literature search reveals that this
particular mutation has been reported only for populations located within the
present-day Mexican states of Chihuahua, Puebla, and Yucatán, as well as in
Costa Rica and Puerto Rico.25 Although the 9-bp deletion has occurred previously in Asia and Africa, its association with haplogroup A constitutes an independent New World mutational event that most probably arose in Mesoamerica
relatively recently.26 Given that this mutation appears at a significant frequency
among the Aztecs of Tlatelolco and that its general distribution corresponds
roughly with that of Nahuatl-speakers—from northern through central Mexico
with additional pockets in Central America—it is tempting to theorize that this
derived form may be related to the spread of Nahuatl, even though considerably
more data would be required to substantiate this interpretation.
In addition to tracking large-scale population movements, the distribution
of mutations will also serve to address some vexing problems of colonial fron-
24. B. M. Kemp et al., “An Analysis of Ancient Aztec mtDNA from Tlatelolco:
Pre-Columbian Relations and the Spread of Uto-Aztecan,” in Biomolecular Archaeology:
Genetic Approaches to the Past, ed. David M. Reed. Forthcoming.
25. Relevant citations can be found in http://resendez.ucdavis.edu.
26. A. Torroni et al., “Asian Affinities and Continental Radiation of the Four
Founding Native American mtDNAs,” American Journal of Human Genetics 53 (1993): 580;
and B. M. Kemp et al., “An Analysis of Ancient Aztec mtDNA from Tlatelolco.”
294
HAHR / May / Reséndez and Kemp
tier history. The case of New Spain’s northwest illustrates well the formidable
challenge that historians have faced. This area was first described by Spanish
conquerors in the 1530s and 1540s. These early entradas provide glimpses of the
numerous and varied peoples that inhabited the region and give us a sense of
the social world as it existed prior to contact. But it is only with the advent of the
mission system, well into the seventeenth century, that we are able to count on
more substantial and regular sources. Unfortunately, it is precisely between the
conquest and the spread of the mission system—a period for which we lack adequate sources—when indigenous communities underwent enormous upheavals and precipitous population declines, resulting in displacements, amalgamations, and disappearances, as well as remarkable survivals.27 As historians come
to grips with these events, DNA evidence will be instrumental in addressing
some specific problems. First, genetic markers will help archeologists and historians establish links and sort out relationships between pre-Columbian sites
and their colonial descendants. For instance, Phil C. Weigand has proposed
that the ancestors of the Tepecanos once inhabited the Cerro de Colotlán in the
Bolaños Valley of Jalisco. As he wrote in the mid-1980s, “It seems that the dozen
old Tepecanos who still use the Cerro de Colotlán circle may be among the last
remnants of the entire Teuchitlan Tradition which dominated much of western
Mesoamerica during the Classic period.”28 Such concrete lines of inquiry can
now be pursued through DNA analysis. More generally, the slow but continuous accumulation of archeological data from northern Mexico now puts us in a
better position to connect the pre- and postcontact pasts in more credible ways;
DNA evidence will be a crucial tool in this monumental task.29
Second, DNA work can also help us revisit the issue of the magnitude of
population decline among different indigenous groups. This is a subject vigorously debated since colonial times and more recently in the classic works
of Sherburne F. Cook and Woodrow Borah. The controversy arises largely
27. For an overview, see Susan M. Deeds, “Northwest Mexico,” in The Cambridge
History of the Native Peoples of the Americas, ed. Richard E. W. Adams and Murdo J. MacLeod
(Cambridge: Cambridge Univ. Press, 2000) 2:52–53; and Cynthia Radding, Entre el desierto
y la sierra: Las naciones O’Odham y Tegüima de Sonora, 1530–1840 (Mexico City: CIESASINI, 1995), chaps. 2 and 3.
28. C. Weigand, “Considerations on the Archaeology and Ethnohistory of the
Mexicaneros, Tequales, Coras, Huicholes, and Caxcanes of Nayarit, Jalisco, and Zacatecas,”
in Contributions to the Archaeology and Ethnohistory of Greater Mesoamerica, ed. William J.
Follan (Carbondale: Southern Illinois Univ. Press, 1985), 165.
29. For a recent appraisal of the field see Beatriz Braniff C., ed., La Gran Chichimeca: El
lugar de las rocas secas (Mexico City: CONACULTA, 2001), passim.
Genetics and the History of Latin America
295
because works of historical demography have to rely on different population
data sets that are not always readily comparable with one another. Moreover,
they need to make certain assumptions about the scope and reliability of these
data sets.30 DNA can now offer an independent way to corroborate such analyses without having to depend on the vagaries of tribute lists and colonial censuses. In short, by comparing present-day genetic variation with that found in
their precontact ancestors, DNA scholars can estimate the loss of lineages and
reduction of overall diversity. One can expect a population that experienced a 95
percent reduction to be more homogeneous and exhibit fewer surviving lineages
than a population that declined by only 50 percent. Population geneticists call
this phenomenon a “bottleneck effect.” Assessing the magnitude of this effect
for different populations will throw additional light on what is arguably Latin
America’s most contentious demographic debate.31
Finally, DNA studies will prove very valuable in assessing long-term population dynamics and specifying more precisely the biological impact of various
factors, such as the presence of missions or the proximity of mines, on indigenous communities in the immediate surroundings.32 Since the pioneering
works of Edward Spicer and Gonzalo Aguirre Beltrán, much historical work
has been devoted to determining why various ethnic groups disappeared from
the historical record, while others endured as viable cultural, social, and political entities to the present time.33 Scholarly concern with the “enduring peoples”
30. Especially relevant for our specific discussion is Cook and Borah, “The Population
of West-Central Mexico (Nueva Galicia and Adjacent New Spain), 1548–1960,” in Essays
in Population History: Mexico and the Caribbean, ed. Sherburne F. Cook and Woodrow Borah
(Berkeley: Univ. of California Press, 1971), 1:300–75.
31. For instance, Rangel-Villalobos has concluded that present-day Huicholes and
Tarahumaras show a reduced level of genetic variation, presumably due to population
decline after contact. Rangel-Villalobos H et al., “Genetic Variation among Four Mexican
Populations (Huichol, Purépecha, Tarahumara, and Mestizo) Revealed by Two VNTRs
and Four STRs,” Human Biology 72 (2000): 983–95. Interestingly, DNA work to date has
shown a surprising stability of haplogroup frequencies before and after the conquest.
D. H. O’Rourke et al., “Spatial and Temporal Stability of mtDNA Haplogroup Frequencies
in Native North America,” Human Biology 72 (2000): 15–34.
32. Scholars have already noted that Purépechas have more European admixture than
do either Huicholes or Tarahumaras. This is attributed to the relative isolation of the latter
groups. Rangel-Villalobos et al., “Genetic Variation among Four Mexican Populations.”
33. Gonzalo Aguirre Beltrán, Regiones de refugio: El desarrollo de la comunidad y el proceso
dominical en mestizo America (Mexico City: Instituto Indigenista Interamericano, 1967); and
Edward H. Spicer, Cycles of Conquest: The Impact of Spain, Mexico, and the United States on the
Indians of the Southwest, 1533–1960 (Tucson: Univ. of Arizona Press, 1962).
296
HAHR / May / Reséndez and Kemp
and their survival strategies, including forming ethnic enclaves and “zones of
refuge,” continues to play a crucial role in recent scholarship, including that
of Phil C. Weigand, Cynthia Radding, and Susan M. Deeds.34 DNA research
will inevitably engage this historiography as it provides evidence of contraction,
expansion, disappearance, and endurance of lineages in the region.
Our discussion all but makes clear that DNA research will increasingly
address some of our most revered categories of historical analysis: race and ethnicity. Modern DNA scholars have made strenuous efforts to distance themselves from the kind of reductionism that took place during the heyday of the
eugenics movement. But notwithstanding the vast differences in sophistication,
the question remains: if races and ethnicities are indeed rooted, to some extent,
in biological difference, what will be the consequences of pinpointing and identifying human variation at the level of the DNA sequence? Some historians
may reject the DNA research enterprise altogether on the grounds that it will
ultimately serve to reify racial and ethnic categorizations. This serious issue is
best addressed not by total rejection of DNA methodology but by enlightened
discussion and awareness of the dangers of allowing the public at large to read
too much into founding lineages and (worse still) of letting interested parties
use genetic markers to further their ethnic or racial agendas.
One way to address this profound political, epistemological, and ethical
issue is by placing genetic variation in a larger context. In the first place, while
DNA scholars try mightily to find variation among populations, the most obvious insight generally remains unstated: namely, that we humans are practically
identical when it comes to our genetic makeup. Physical traits that we recognize
at a quick glance, such as skin color, eye shape, and body size, may precondition us to believe that there exist significant genetic differences—and perhaps
even commensurate cognitive and temperamental differences—between us. In
fact, these physical traits are rooted in insignificant variations at the level of our
DNA, which underscores the fact that we are all closely related to one another.35
34. N. Ross Crumine and Phil C. Weigand, eds., Ejidos and Regions of Refuge in
Northwestern Mexico (Tucson: Univ. of Arizona Press, 1987); Cynthia Radding, Wandering
Peoples: Colonialism, Ethnic Spaces, and Ecological Frontiers in Northwestern Mexico, 1700–1850
(Durham, NC: Duke Univ. Press, 1997); Susan M. Deeds, Defiance and Deference in Mexico’s
Colonial North: Indians under Spanish Rule in Nueva Yizcaya (Austin: Univ. of Texas Press,
2003).
35. To put it in hard numbers, out of the nearly 17,000 base pairs constituting the
mitochondrial genome, the average number of base-pair differences between two human
sequences is 38.5 among non-Africans, 76.6 among Africans, and 61.6 among all humans.
M. Ingman et al., “Mitochondrial Genome Variation and the Origin of Modern Humans,”
Genetics and the History of Latin America
297
To be sure, different human populations have adapted extraordinarily rapidly
to different environments, much of this driven by culture. But this does not
change the central fact that we are an extremely young species, having arisen in
Africa only within the last 100–200,000 years.36 Even if we compare the most
different-looking human groups (for example, Native Americans, West Africans, and Northern Europeans), we don’t need to go back too many generations
to find common ancestors among them. By contrast, we will find considerably
more genetic variation among gorillas or chimpanzees (even though one chimp,
to our eye, looks essentially like any other chimp). These, our close cousins,
have lived as distinct species for 5 million years or more, and compared to them,
we humans are practically genetic carbon copies of one another.37
Another way to contextualize DNA findings is by emphasizing that there
are no pure races or ethnic/national groupings. The entire eugenics edifice
rested on the perception that humans came in a few unadulterated varieties—
most commonly Africans, Asians, and Caucasians—as well as a range of mixed
or mongrel populations between them. It went without saying that these pure
races were tangible, stable, and easily ascertained. Modern DNA research has
shown the wrongheadedness of such discrete groupings. For one thing, DNA
researchers have found heterogeneity within any given population.38 Pure races
simply do not exist anywhere on the planet. Moreover, racial and ethnic groups
are not stable but extraordinarily dynamic. As a rule of thumb, DNA researchers have found that people tend to be more closely related to those living around
Nature 408 (2000): 708–12. Interestingly, the greater diversity within African
populations—as opposed to less diversity when between African and non-African
populations—may be explained by a significantly longer genetic history for humans in
Africa, something that is consistent with the archaeological evidence. See also Luigi
Luca Cavalli-Sforza, Genes, Peoples, and Languages (New York: North Point Press, 2000),
chap. 1.
36. In a pioneering article, Rebecca L. Cann and her coinvestigators concluded that a
common ancestor to all humans existed 200,000 years ago. Cann, “Mitochondrial DNA and
Human Evolution,” Nature 325 (1987): 31–36.
37. For an entertaining discussion of these issues see Jonathan Marks, What It Means to
Be 98% Chimpanzee: Apes, People, and Their Genes (Berkeley: Univ. of California Press, 2002),
passim.
38. For example, a group of researchers has shown that a small indigenous community
of the Pacific Northwest, the Nuu-Chah-Nulth, contains no less than 28 different
lineages—equivalent to more than 60 percent of the genetic diversity observed in large
ethnic groups such as the Japanese or sub-Saharan Africans. R. H. Ward et al., “Extensive
Mitochondrial Diversity within a Single Amerindian Tribe,” Proceedings of the National
Academy of Sciences of the United States of America 88 (1991): 8720–24.
298
HAHR / May / Reséndez and Kemp
them than to populations that are located farther away: in other words, that
genetic variation is distributed in a clinal fashion. These two simple findings
flatly contradict the existence of discrete and stable human groupings as imagined by eugenicists. All of this also indicates that an open dialogue between
genetics and social scientists about the significance of biological variation among
humans is the best way to prevent the kind of biological reductionism that has
been so pernicious in the past.
The field of modern genetics has opened an exciting new window into our
past that we’ve just begun to explore. Buried within our genes lies a wealth
of evidence about humanity’s long evolution in Africa, our momentous exodus and advance over the rest of the planet, long-distance migration patterns,
warfare, disease, commerce, conquest, marriage patterns, the emergence of
social groups, kinship, and a myriad of other complicated interactions in which
humans engage. Historians have long been interested in these matters. New
genetic tools provide an independent line of evidence, enabling us to revisit
these questions and obtain more satisfactory answers. This research also opens
the possibility of a dialogue between historians and geneticists, a dialogue that
is desirable and badly needed, as we bring to the table complementary strengths.
Historians are in a unique position to formulate sound hypotheses that fully
take advantage of our accumulated knowledge about the relationships between
different human groups, thereby minimizing technically competent but ultimately wasteful research. For their part, geneticists will need to educate us
about what hypotheses can be feasibly tested and about the validity and scope
of the results. It is up to us, historians, to become involved if we want to have a
say in how genetic tools are used to investigate the past and to make sure that
these tools are deployed in ways that will ultimately benefit the peoples whose
trajectories we study.