Read an Excerpt
The Huasteca
Culture, History, and Interregional Exchange
By Katherine A. Faust, Kim N. Richter UNIVERSITY OF OKLAHOMA PRESS
Copyright © 2015 University of Oklahoma Press, Norman, Publishing Division of the University
All rights reserved.
ISBN: 978-0-8061-4956-1
CHAPTER 1
The Huastec Problem
A Linguistic and Archaeological Perspective
John Robertson
Stephen Houston
One of the remaining problems in Mesoamerican studies is determining the origin and historical disposition of Huastec, a Mayan language spoken in three dialects by about 150,000 people in the Huasteca region, which comprises eastern Hidalgo, northern Veracruz, eastern San Luis Potosí, northern Puebla, and southern Tamaulipas, Mexico (Kaufman 1985:473–74, n1). The great diversity of Mayan languages spoken in and around the Yucatán Peninsula makes it virtually certain that this region was the general area of Mayan linguistic genesis (Kaufman 1976; Kroeber 1944); the challenge is to determine how and when Huastec traveled to the northern reaches of Mesoamerica, some 800 km away from the historical distribution of all other Mayan languages (Kroeber 1944). An added complication is that a branch of Huastec, the now-extinct Chicomuseltec, was spoken in Chiapas, geographically close to speakers of Mayan languages such as Mam, Tzeltal, and Tojolabal (Campbell 1988:199; Campbell and Canger 1978; Sapper 1897; Termer 1930).
To date, the dominant interpretation is that Huastec split off from its linguistic family at an early time—indeed, representing the oldest split in Mayan. This interpretation might be termed the "Swadesh/McQuown/Kaufman Model" after its principal proponents, Morris Swadesh, Norman McQuown, and Terrence Kaufman, who have argued in various publications that Huastec and its relatively close relation, Chicomuseltec, differ so radically from other Mayan languages that they must have diverged in remote antiquity (e.g., Campbell and Kaufman 1985:192; Kaufman 1972:13; Swadesh 1953:226). Using glottochronology, McQuown places this date at 1800 B.C., and Swadesh some centuries before (McQuown 1964:69, fig. 62; Ochoa Salas 1989:30). In the most recent estimate, Kaufman asserts that the split took place between 2200 B.C. and 1000 B.C., with a preferred time more to the earlier than later part of this span. Accordingly, Huastec is thought to have reached its current homeland "probably before 1500 BC and surely before 1000 bc" (Campbell and Kaufman 1985:192; Kaufman 1976:106). To account for the historical location of Chicomuseltec, Kaufman (1980:110) posits that it originated from the Huastec area and then moved south toward the Maya heartland around A.D. 1250 as a result of pressures from speakers of intruding languages such as Totonac.
A minority view, developed by archaeologists working in the Huasteca, holds that Mayan speakers inhabited the entire Gulf Coast from an early time and that Mayan languages stretched from "Peten [Guatemala] to Panuco," forming a belt that was later broken up by Totonac speakers (Ekholm 1944:504–505; MacNeish 1954:625). We might call this the "Gulf Coast Origin Model." Leonardo Manrique Castañeda (1989:210) goes even further to suggest that proto-Mayan was found as early as 500 B.C. in the Huastec region (or near that region) and stretched southward to the coastal plain, possibly to the Tuxtla territory. In his view, much like Kaufman's, Chicomuceltec departed from the Huasteca and moved to Chiapas relatively late, at about A.D. 1000 (Manrique Castañeda 1989:map 5). A Huastec homeland for protoMayan is unlikely, however, considering current understandings in linguistic geography. For example, most specialists place the Indo-European homeland between 14 and 45 degrees longitude (corresponding to the Pontic-Caspian steppe) because that is where most such languages are or were spoken; the location of Sanskrit or Tocharian well to the east simply reflects their position as outliers (see Mallory 1989:152–153). Similarly, the Maya "center of gravity," that is, the predominant location of most Mayan languages, points to an area that lies east of the Isthmus of Tehuantepec. It is doubtful, however, that the homeland can be pinned down to the Solóma region of highland Guatemala, as Kaufman (1976:104) suggested, because it was sparsely populated during the Early Formative or Late Archaic periods, the presumed time of Mayan linguistic emergence (Clark and Cheetham 2002).
At face value, dates obtained from glottochronology and lexicostatistics are little more than hypotheses with scanty reliable support. According to many linguists, application of the methods results at best in imprecise estimates and at worst in misleading statements resting on debatable assumptions (Campbell 1984:4; Houston 2000:158). Archaeologists, however, often assign an unwarranted certainty to dates obtained in this way. Rather than conjectures, glottochronological hypotheses are treated as independent variables with which to interpret archaeological material (Ochoa Salas 2001:13). Through a process of epistemological transformation, what began as theories shift to solid data when transferred from one discipline (linguistics) to another (archaeology). Aside from the inherent uncertainty of such dates, there is a second, equally troubling consideration. A considerable body of linguistic information, some of which is discussed below, indicates that, contrary to common perception, Huastec is not so radically different from other Mayan languages and should be grouped with other languages spoken in the area of eastern Chiapas. Though the most parsimonious view is not necessarily the correct view, the statement of one scholar, who has described the proposed movement of Chicomuceltec back to the Maya region as "difficult to imagine," should not be taken lightly (Campbell 1988:209; Campbell and Canger 1978:228–229). It requires that the Huastec subset of languages move away from the Maya region at an early time, and then, for unknowable reasons, one of its components moves back again—3,000 years later, squarely into the same region. A model involving fewer moves—from Mayaland to the Huasteca—would be more economical if supported by other, corroborating data. The mediating indicators of this simpler view are the linguistic similarities and differences discussed below.
In sum, there are two issues. First, current lexicostatistical methods cannot reliably establish a date for the departure of Huastec from the Maya region in the second millennium B.C. Second, linguistic evidence presented here contradicts the accepted view that Huastec and Chicomuceltec are distant outliers of the Mayan linguistic family. Specifically, the Huastec grouping shares many features with the Ch'olanTzeltalan subgroup. This chapter develops the latter argument in more detail, with an aim to opening new possibilities for interpreting the linguistic prehistory of Mesoamerica. We are skeptical of unwarranted attempts to correlate archaeological patterns with language or any effort to establish precise dates via lexicostatistics for these correlations (cf. Josserand 1975; Kaufman 1976). Speculations about the dating, homeland, and subsequent movements of Indo-European show that even well-controlled data are subject to intense debate and disagreement (Mallory 1989:7, 22–23). We argue, however, that a reasonable case can be made that Huastec speakers moved beyond the Maya region at a time much later than proposed by Kaufman, and as justified by modern, colonial, and hieroglyphic data. The move may have occurred at any period after the first millennium B.C.
Linguistic Arguments
In what follows, we identify several linguistic attributes that provide evidence against Kaufman's proposed Huastec departure from mainstream Mayan some 4,000 years ago. Each of these attributes rests upon certain characteristics of Huastec that denote either genetic or geographic contact with the Ch'olan-Tzeltalan language group, which for reasons of linguistic change could not have existed at such an early period.
Calvert Watkins (1973) has pointed out that, when there are similarities among languages, those similarities have four possible sources: chance, universal tendency, language contact, or genetic filiation. Because, as we discuss below, the similarities among the Huastec and Ch'olan/Tzeltalan languages are significant and numerous, it would seem difficult to argue that the similarities result from chance alone; nor, because of their narrow, delimited character, could they arise from some universal tendency in language change. This leaves proximity, either through geographic or genetic relatedness, either of which would establish a late Huastec homeland in the area of Tzeltal/Ch'olan. Either factor would refute Kaufman's proposal for a separation from mainstream Mayan in the late third or early second millennium B.C.
Phonology
The debate about the relationship between Huastec and Ch'olan/Tzeltalan rests in part on a perception of a series of shared changes that occurred in the inventory of speech sounds (i.e., inventory of phonemes) of these two language groups. Given the Common Mayan (or proto-Mayan) inventory of speech sounds, Campbell (1988:211) lists several phonological changes shared by Huastec and Ch'olan/Tzeltalan: *r > y, *k(') > ch(') / ..., *q > k, *ty > t. In contrast, Kaufman (1976:106) states, "Huastecan has no known special relationship with any other group in the Mayan family. Two phonological changes in Huastecan found also in other Mayan languages (*r > y and *q > k) I consider to be independently motivated." In fact, these changes are probably not independently motivated, since they are found stretching across areas that are geographically contiguous, starting with the locus of the Cho'lan-Tzeltalan subgroup, spreading north and east to the Yukatekan territory and southwest to Huehuetenango, Guatemala. Furthermore, Kaufman makes no comment regarding *ty > t and *k > ch / ... and, consequently, one might conclude that he also sees these phonological changes as independent innovations. Any one of these changes could have occurred independently in Huastec, but it is less likely that they, along with the other changes discussed below, could have happened independently of Ch'olanTzeltalan. As Campbell (1988:211) points out, even if the changes are due to borrowing and not to common ancestry, they suggest that there was considerable overlap between Huastec and the Ch'olan-Tzeltalan languages.
Palatalization
The most complex of the above list of sound changes is palatalization, whereby a historical *k becomes ch in the daughter languages, (*k > ch ). According to Kaufman and Norman (1984:83), "The single phonological innovation that sets off the Greater Tzeltalan (Ch'olan-Tzeltalan) subgroup from all other subgroups of Mayan is the shift of the proto-Mayan velar stops *k and *k' to the affricates ch and ch'." The question of when the sound change occurred bears directly on when Ch'olan-Tzeltalan became a separate subgroup, as well as on when speakers of Huastec migrated. Kaufman (1976:106) estimates that Ch'olan-Tzeltalan diverged by about 1100 B.C. and that the Huastec split took place somewhere between 2200 and 1000 B.C. The Maya hieroglyphs, however, contain material evidence that brings Kaufman's early dating into question. The evidence indicates that proto-Mayan *k had not changed to ch when the syllabary for Classic Maya writing was initially formulated (roughly around the time of Christ), which means at the very least that Kaufman missed the shift from *k to ch in the Ch'olan-Tzeltalan subgroup by a thousand years. It also brings into question his assertion that *k >ch is the "single phonological innovation that sets off the Greater Tzeltalan subgroup from all other subgroups of Mayan subgroups," because the division between the Ch'olan and Tzeltalan would have occurred sometime significantly before the time of Christ—based on what we know about the language of the hieroglyphs. It also casts doubt upon Kaufman's proposed date for the Huastec departure (between 2200 and 1000B.C.). Houston (2000:114) emphasizes that the syllabogram ka is acrophonic, meaning that it takes its sound value from the first two phonemes of the word "kay" 'fish.' Assuming that Ch'olan speakers invented the syllabary, the attested kay should have been chay by Kaufman's reckoning: the Ch'olan and Tzeltalan shift *kay > chay would already have occurred at least a millennium earlier. How is it possible, given the assumption of comprehensive palatalization, that *kay became chay at 1100 B.C., and that sometime around the time of Christ chay reverted to kay, thereby forming the basis for the syllabogram ka? It is significant that one of the earliest syllabograms, ka , retains the original k, whereas another syllable, cho (historically from *koh 'jawbone'), was recruited in the Late Classic. An Early Classic retention (*k >k ) and a late innovative (*k >ch ) would make sense if the shift from *k >ch were ongoing during the Classic period.
That the shift was ongoing is attested by the fact that certain words underwent palatalization while other words—words subject to conservative retention—did not. For example, there are instances of nonlogographic spellings of k'u-huk'uh for "god," and many instances for yu-k'iy-uuk' 'drink,' where we do not see the expected palatalization that would render the words as ch'uj and uuch', as documented in the colonial and modern attestations of Ch'olan-Tzeltalan. Consistent with this pattern are several transparent, atavistic spellings, such as ka-bakaab 'earth,' instead of cha-ba chab. In Ch'olan, the form is kab (e.g., Ch'olti' [Moran 1695] paskab lit. open-earth 'morning'; sah-kab lit. white-earth 'for spinning thread, also earth that is eaten'; Acalan Chontal cab 'earth, village'; Chontal [Knowles 1984] kaba 'agricultural plot'), whereas in Tzotzil we find *k > ch (Colonial Tzotzil [Laughlin and Haviland 1988] chob 'cultivated land'; chabaj 'cultivate'). The most general sense of kab seems to be 'land that is subject to human or even superhuman conditioning.' The word for 'earth, land' in Common Mayan was most likely *kab, which was accorded the special meaning of human or superhuman involvement. In K'iche'an, the last vestige of kab can be found in the Popol Wuj: lord of earthquakes was kab-r-aqän lit. 'earth-his-leg.' In modern K'iche'an languages we find kabraqän "earthquake." All of these examples suggest that the shift was ongoing and had not yet run its course. Nonetheless, following Kaufman (Kaufman and Justeson 2003), these examples may be borrowings from the nonpalatalizing language, Yukatekan. Yet, from Kaufman's (1984) point of view, the words for 'god,' 'drink,' 'earth,' and the like would have already been ch'uj, uuch', chab, ... a millennium before the script, as they are attested today in modern languages. If these words were already present in palatalized form, it seems unlikely that they would have been borrowed as k'uh, uk', and kaab from Yukatekan. In light of the formation of the syllabary as discussed above, and in light of the different outcomes in Ch'olan-Tzeltalan, the better explanation is that the process of palatalization was a continuing process during the Classic period. Therefore these words were not Yukatekan.
The k ~ ch alternation in hieroglyphic spelling invites other considerations. It is well established linguistically that the language of orally fixed phraseology tends to be highly conservative: "Archaic linguistic forms are a typical trait in the language used in religion, magic and in ceremonial language in general" (Sørensen 2007:89). It is possible, therefore, that because the language of the hieroglyph script was religious, ceremonial, and highly scripted the words that today follow the expected *k(') > ch(') might well have preserved their archaic k ('); thus, terms like k'uh 'god/holy,' uuch' 'drink,' kab 'earth.' For instance, of particular interest is the word for mask, k'ooj, which is important in Maya cultural ceremonies to this day. The colonial Ch'olti' lexicon registers two distinct words for 'mask,' one a palatalized version, ch'oj, and a second unpalatalized version, k'oj. Because the Ch'olti' lexicon was written in different hands ( [k'oj] was a later addendum), it is almost surely a case of uneven dialectal variation. Likely, because of its conservative, ritual nature, the word k'oj endures in all modern Ch'olan-Tzeltalan languages, including modern Ch'orti', where the word was recorded as ch'oj and k'oj in its antecedent Ch'olti'.
(Continues...)
Excerpted from The Huasteca by Katherine A. Faust, Kim N. Richter. Copyright © 2015 University of Oklahoma Press, Norman, Publishing Division of the University. Excerpted by permission of UNIVERSITY OF OKLAHOMA PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
