Generalized Exchange

The concept of a very general reciprocal exchange is reconstructable to proto-Mayan culture through the lexical root *jal (Table 4). This term, and by extension the relationships that it represents, can be posited to proto-Mayan based upon survival of its reflexes in both Eastern and Western Mayan language subfamilies as well as Huastec (Kaufman with Justeson Reference Kaufman and Justeson2003:779–780). While temporally established diagnostic sound innovations associated with the phonology of this root are few, its presence in geographically distant Huastec with virtually identical semantic content more or less rules out a late diffusion. Though exact semantics for the early proto-Mayan period are rather hazy, the term may have denoted what one might call a “quasi-economic” sense of obligatory reciprocity carried out through exchange and substitution. Or it may have simply meant “to change,” with examples of a purely economic significance appearing only later among disparate languages by coincidence.

Table 4.

“Exchange”

The connotations of the later reflex /jel/ clearly became much more political than economic among Central Mayan languages. The jal to jel correspondence between Eastern and Western Mayan subfamilies can be explained rather simply by a shared conditional /a/ to /e/ vowel shift in proto-Western Mayan (see Carrasco [2005:Footnote 58] for a possible second example of the same shift, EM *cha / WM *che to say), with the Yucatecan cognate being lost but subsequently borrowed back from a Western Mayan language. The reflex remained *jal in Eastern Mayan and Huastecan. The term /jel/ has been posited in the Classic hieroglyphic script in a number of locations by various epigraphers (Carrasco Reference Carrasco2005:114–118). In most cases, it has been read in contexts related to succession, or changes in rulership, such as in the tablet from the Temple of the Foliated Cross at Palenque describing the accession of Kan B'ahlam (Carrasco Reference Carrasco2005:117) and the array of lintels from Structure 12 at Yaxchilan that lays out the dynastic sequence from that site (Riese Reference Riese, Justeson and Campbell1984). Again, semantics of the root in the very deep past can only be called “quasi-economic” at best.

The parallel root *k'ex can only be reconstructed to Core Mayan based upon occurrences in both Eastern and Western Mayan languages (Kaufman with Justeson Reference Kaufman and Justeson2003:781–782). The lack of an attestation from Huastecan prevents any secure attribution to proto-Mayan. As one can see from Table 4, the root functions semantically in an almost identical manner to *jal and, like *jal, its reflexes have survived in almost every subgroup of the Mayan language family, implying the continued vitality of its semantic content through time. Also like *jal, *k'ex seems to have represented a quasi-economic reciprocal obligation of exchange, trade, and repayment. The nominalized form of this verb is used in several disparate Mayan languages to denote a “substitute” or “namesake” in political position or kinship roles, suggesting ancient use in that manner as well. The manifestation of this “namesake/substitution” relationship, also represented in some languages by *jal, is perhaps best characterized through the relatively common Mayan practice of naming children after their grandparents, whom they are seen as “replacing” in what is viewed by many as an ancient and deep-seated form of ancestor worship (Carlsen and Prectel Reference Carlsen and Prectel1991; Carrasco Reference Carrasco2005:118–121; McAnany Reference McAnany1995).

Using texts and iconography from Classic period art in conjunction with contemporary ethnographic studies, Taube (Reference Taube and Kerr1994) has demonstrated the religious implications of k'ex. It has a long history connoting acts of substitution or exchange through offerings to deities—apparently extending to the realm of human sacrifice on the behalf of Classic period rulers. As Taube (Reference Taube and Kerr1994:671) describes:

While one could debate the economic content in these contexts, some of the very same highland Mayan languages (K'iche’, Tzotzil) that use the term for such sacrifices today use it for “barter” or “to exchange money.” The combined ethnographic, iconographic, and linguistic evidence then suggests that, cognitively speaking to the ancient Maya, reciprocation as denoted by *k'ex was a relationship to be maintained not only among humans, but between gods and mortals as well. The term serves as a prime example of how economic behavior is often couched in terms of cosmology and ritual in human societies.

A reconstructable verb for the act of borrowing, *majaan, is attested in Eastern and Western Mayan languages as well as Yucatecan (Kaufman with Justeson Reference Kaufman and Justeson2003:772–773) (Table 5). Based upon this distribution, the root can be attributed to Core Mayan at minimum, thereby establishing an ancient presence of delayed reciprocity. While Kaufman and Justeson would reconstruct the root to proto-Mayan, I hesitate to do so because of its lack of attestation from Huastecan languages. The nominalized form of the original verb root may have been mangled a bit among Greater Lowland languages, where the derived form of *maaj was apparently cognitively reanalyzed by some speakers as a verb itself. The meaning also shifted in K'iche'an languages, though the exact semantics are unclear since the Spanish verb quitar, by which the word is glossed, has several meanings. In these languages, the word was replaced as a term for borrowing by Greater K'iche'an *qaj. Incidentally, a very similar shift occurred in Western Mayan with a verb root for “to sell,” *k'aay, discussed at some length below.

Table 5.

“Loan; to borrow”

There are at least four additional reconstructable root lexemes for debt obligations in Mayan languages (Kaufman with Justeson Reference Kaufman and Justeson2003:788–789). Independent terms for this relationship go at least as far back as proto-Eastern and proto-Western Mayan (Table 6). Kaufman (with Justeson Reference Kaufman and Justeson2003:788–789) suggests that the root *b'et appeared among Western Mayan languages and subsequently diffused about the lowlands, perhaps as a part of Classic or Formative period culture. The divergent semantics in Yucatecan languages, however, make its borrowing from Greater Tzeltalan less likely and perhaps better explained as a Core Mayan innovation. A nearly perfectly complementary distribution suggests that reflexes of *b'et may have been pushed out of existence via the rise of the root *k'aas in Eastern Mayan languages. Words related to *k'aas were evidently influential enough to subsequently diffuse into some Greater Q'anjob'alan languages. Even later, more localized terms such as *p'aax and *tx'ooq arose among proto-Yucatecans and proto-Ixilans, respectively. The general picture is one of new words for debt obligation appearing periodically and being regionally or locally confined, with older terms often being replaced and lost among certain subgroups.

Table 6.

“Debt”

Reconstructable individual terms for inheritance and possession are confined to highland Maya languages (Kaufman with Justeson Reference Kaufman and Justeson2003:759–760) (Table 7). The root *e:ty can be traced back to proto-Eastern Mayan, as /ty/ corresponds with /tz/ and /ch/ in Mamean and Kichean languages, respectively. Multiple frozen forms also demonstrate the antiquity of this etymon in these tongues. The limited geographical reach of the root may imply that complex systems of ownership first appeared in the Guatemalan highlands among Eastern Mayan speakers, but any reaching conclusions are impeded by the fact that there are other methods of signifying possession in Maya languages. For instance, certain Classic period artifacts are known to have been “name-tagged” with text through simple stative use of the ergative pronoun (Houston and Taube Reference Houston and Taube1987). In fact, Stuart (Reference Stuart1995:10–22) has suggested that the earliest motivations for phonetic writing in the Maya area may have been the designation of ownership of portable objects.

Table 7.

“Property”

Reciprocation

There is no shortage of terms for gifts and giving in the Mayan languages (Kaufman with Justeson Reference Kaufman and Justeson2003:774–778, 785–787) (Table 8). It is no surprise that at least one of these, *mat/*mahtaan, can be quite confidently traced back to proto-Mayan. While there is some question about the original semantics of this root, the substantive form seems to have clearly taken on the significance of “gift” by at least Core Mayan times. I consider this adequate evidence for the original meaning of the verb. The semantics apparently narrowed to contexts of charity, and then subsequently shifted—perhaps as a result of interference with competing terms like *tz'ah—in Yucatecan. A similar situation seems to have developed in Huastecan with *mat narrowing to contexts of giving requiring repayment. In addition, five other reconstructable roots for gifts and giving are found within the Mayan family more or less confined to individual linguistic subgroups, although the fact that there is a lack of congruence within the various subfamilies may suggest that many of the terms coexisted for a long time in many languages.

Table 8.

“Give/gift”

A reflex of Central Mayan *aq’ is found in the Classic script as yak'aw [ya-k'a-wa]. Upon Stela 32 from Naranjo it directly refers to the giving of tribute “bundles” (Stuart Reference Stuart1995:358), demonstrating that cognate terms could be used for the mandatory transfer of goods as well as the voluntary bestowing of gifts. In essence, one gives tribute as one gives a gift—in contrast to English in which we would “pay” tribute or taxes. Both *mahtaan and *siih are also found intermittently in the hieroglyphic corpus. Various forms of *siih have been identified on a number of pre-Columbian artifacts and stone monuments from sites such as Tamarindito, Itzam, and elsewhere. In many cases, the possessed form of the term, usiih (written u-si), was used to signify portable objects as gifts intended for a specific individual of high rank (Mora-Marín Reference Mora-Marín2001:36, 474). In one example, a slate disk mirror back recovered from an archaeological site in distant Costa Rica has been identified as a gift between rulers of the Classic Maya sites of El Zotz and El Peru (Houston Reference Houston2008). Another term for giving, mataan [ma-ta-na], occurs repeatedly as a part of augury phrases in the Late Postclassic/Contact period Madrid Codex almanacs. The presentation of gifts was indubitably a central aspect of diplomatic relations among elites during the Classic period, and all indications are that this method of exchange was of vital political and economic importance from proto-Mayan times forward.

Appropriation and Transfers

Appropriative economic action entails the use of political force or coercion in the movement or transfer of goods or labor (McAnany Reference McAnany, Feinman and Nicholas2004). Customs that fall under this rubric can range from obligatory gifts to taxation and tribute offerings to outright military plunder. Perhaps no other phenomenon better exemplifies the nature of appropriative economies than the direct appropriation of labor—slavery. Unfortunately, very little is known about the nature and extent of this institution among the ancient Maya. Earlier in the last century scholars seemed to find abundant evidence of pre-Columbian slavery in ancient Mayan art, from the stelae of Calakmul (Morley Reference Morley1933) to murals from Chichen Itza (Sharer Reference Sharer1994:503). There are numerous older works that specifically examine the nuances of forced servitude as practiced among Mesoamerican peoples and its contrast with the Old World institution (Izquierdo Reference Izquierdo and Fernández1984), as well as its broader economic ramifications among the diverse indigenous American cultures that adopted it (MacLeod Reference MacLeod1928). But in recent decades there has been virtually no systematic discussion of the possibility that individuals were maintained for use of their labor (but see McAnany [Reference McAnany, Sabloff and Henderson1993] and Sharer [1994:503] for acknowledgement of its probable presence). In fact, it is almost as though the topic has been consciously avoided.

Ethnohistorical documents reveal the more or less ubiquitous presence of indigenous slavery across the Maya area at the time of European contact. Slaves were an important commodity in the markets of Acalan-Tixchel in Tabasco (Scholes and Roys Reference Scholes and Roys1948). Diego de Landa discussed various uses for and means of creating slaves in the Yucatan in his Relación (Tozzer Reference Tozzer1966). The role of female slaves and the ambiguous status of their offspring are addressed in the colonial documents of Cozumel on the Caribbean coast (Roys Reference Roys1943). Among the highland Quiche, slaves were obtained in virtually the same ways described for the lowlands and typically served a period of work service before eventual ritual sacrifice (Carmack Reference Carmack1973, Reference Carmack1981). Warfare, impoverishment, and conviction of certain crimes are all described in primary sources as grounds for enslavement (de Landa Reference de Landa and Gates1978[1566]; de las Casas Reference de las Casas and O'Gorman1967[1561]; López de Cogolludo Reference López de Cogolludo2007[1688]). Unfortunately, very little was written during the Contact period about the labors or uses such persons were put to in daily life, only that many ended their days as sacrificial victims. Perhaps this implies that the uses they were put to were not all that different from what would be expected of European slaves.

Landa (Reference de Landa and Gates1978[1566]:15) specifically reports that slavery was first introduced into the Yuctatan by the Cocom rulers of Mayapan in Late Postclassic times. This can be rather soundly refuted, however, through etymological analysis of the Mayan root *mun (Table 9). Kaufman (with Justeson Reference Kaufman and Justeson2003:60) reconstructs *muuniil to proto-Mayan, *muun to proto-Central Mayan, and *munaty to the Greater Lowland Language group, all dating to well before the Late Postclassic period. That this root was not diffused across the Maya area at some late date is demonstrated by the /t/ to /ch/ correspondence between Tzeltalan languages and Yucatec, and the preservation of vowel length in Teko and Mocho. While the glosses of many cognates suggest the possibility of original semantics related simply to “work,” closer inspection reveals that these glosses are almost entirely confined to Q'anjobalan languages. This makes it much more likely that the distribution simply represents a semantic narrowing in that branch of the language family. Kaufman's reconstruction of the root to proto-Mayan is rendered questionable by the absence of attestation in Huastecan languages. Core Mayan origins still remain a strong possibility, however, with a likely presence by Central Mayan times considering the wide distribution of cognates. It was certainly present among Western Mayan languages by the period of Greater Lowland Mayan influence given the sound correspondences and a reconstructable presence in both Q'anjobalan and Greater Tzeltalan languages.

Table 9.

“Slave/servant”

There is certainly no shortage of captives depicted and discussed in Classic Mayan glyphic texts. Most scholars of late have tended to assume that these high-ranking figures all purely represent sacrificial victims and there has been little discussion of the possibility that some served as slaves for at least some time before being sacrificed. Nor is the probability entertained that there were considerably more captives of lower social status living out the remainder of their days in servitude. At least one term thought to refer to slaves or servants, habtal, has been read in the Classic script (Boot Reference Boot2009:67). Others, such as baak and winak in their possessed forms, may refer to it more indirectly and metaphorically. It is particularly interesting that at least one Yucatec ethnohistoric dictionary specifically refers to individuals captured in warfare but remaining indefinitely in the service of their captor as slaves as bac-zah (Álvarez Reference Álvarez1997:153). The combined evidence overwhelmingly suggests that slavery had a very ancient presence among Mayan speakers, and its economic importance as a source of labor may well be greatly underestimated.

A widely attested term for tax obligations or work debt, *pataan, also shows indications of a Core Mayan origin (Kaufman with Justeson Reference Kaufman and Justeson2003:60) (Table 10), suggesting the presence of vertical economic transfers and/or work obligations to those of higher status by late proto-Mayan times. While the lack of sound innovations associated with the phonology of this term makes it difficult to differentiate genetic transmission from borrowing, the distribution of cognates among Eastern Mayan, Western Mayan, and Yucatec is most typical of words originating among proto-Core Mayan speakers. The elimination of vowel length common to all Greater Tzeltalan languages is clear, proving the existence of this root—and by implication the economic arrangement it signifies—by Classic times at the very least.

Table 10.

“Tribute/tax debt”

The term pataan (phonetically written pa-ta-na) occurs repeatedly on Classic period monuments and ceramic vessels in the context of economic demands placed on subordinate lords by dominant political centers (Stuart Reference Stuart1995:354–356). The iconography presented alongside such phrases suggests that such transfers typically took the form of cloth, feathers, and cacao, and were commonly the result of defeat in warfare. In a number of extant languages, the meaning of *pataan cognates has shifted somewhat, revealing aspects of the economic relationship not apparent in Classic period texts. For instance, the term's repeated semantic association with planting and milpas among disparate language groups may be evidence that agricultural work also formed a significant component of tribute obligations in ancient times. It is perhaps not surprising that most of the Mayan languages that have retained reflexes of *pataan are those known to have been spoken among societies characterized by high levels of social stratification in the past, namely the Yucatecan, Cholan, and K'iche'an languages.

There are those who maintain that *pataan is actually derived of Uto-Aztecan origins and represents Classic period contact between Nahua and Mayan languages (Macri and Looper Reference Macri and Looper2003). They argue this point on the basis of similarity to Uto-Aztecan words for “to trade,” patla in Nahuatl and Pipil, or “to change” *patla in proto-Nahua and *pa:ta for proto-Southern Uto-Aztecan as reconstructed by Dakin (Reference Dakin1982). Yet the similarities appear to be completely superficial, since there is no indication of the complex initial preconsonantal vowel /a:/ in Mayan languages, nor the /l/, and the semantics are decidedly different anyway. Moreover, as discussed above, *pataan otherwise reconstructs quite nicely as a purely Mayan root to proto-Core Mayan, which in any model dates to well before the period of Teotihuacan influence that Macri and Looper (Reference Macri and Looper2003) wish to connect with Nahua contact. Superficial lexical similarities are often coincidental and extremely weak bases for establishing cognacy. There is no reason to believe that the term is of Uto-Aztecan, or Teotihuacano, origin.

The primary reason for including terms for “carry” and “load” in this study is the use of ihkaatz in the hieroglyphic script in apparent reference to bundles of tribute goods (Stuart Reference Stuart1995). Kaufman (with Justeson Reference Kaufman and Justeson2003:892–893) reconstructs *ihqatz to proto-Central Mayan, as the word has clearly undergone the /q/ → /k/ sound shift in Western Mayan languages (Table 11). Stross (Reference Stross1988), however, finds sufficient evidence to reconstruct it further to proto-Mayan, and Dienhart (Reference Dienhart1989) does indeed produce an attested reflex from Huastec that substantiates such an origin. As mentioned above, ihkaatz is fairly common in Classic period texts—usually found in the phonetically transparent form ‘i-ka-tsi or occasionally the possessed form yi-ka-tsi. The term was originally thought to identify bundles of religious goods used in rituals associated with public office (Stross Reference Stross1988). It is now more commonly recognized as appearing in economic contexts identifying bundles of goods used in payment of tribute or as part of “name-tagging” phrases on portable objects, particularly those made of jade. It must be noted, however, that there is absolutely no linguistic basis for reading ihkaatz as a specific reference to jade (see, for example, Stone and Zender Reference Stone and Zender2011). The evidence for doing so is very circumstantial and comes completely from epigraphic contexts that involve the appearance of a possessed form, yikatz, on personal portable objects (Stuart Reference Stuart, Guernsey and Reilly2007). Although these have hitherto been found almost exclusively on jade artifacts, there is no more reason to believe that yikatz inherently refers to the jade raw material than, for instance, that usiih written upon the disk mirror back discussed previously under terms for gifting refers specifically to the slate raw material of that object. Indeed, if there is any parallel, it would suggest that the mode of acquisition or transfer of the objects, rather than the raw material itself, is the basis of reference of yikatz on jade items. Moreover, jade pendants, especially those in the form of “ceremonial celts,” are typically seen in Maya art dangling in sets from the belt of royalty as noise-making tinklers (Stone and Zender Reference Stone and Zender2011:71). They are in that sense physically heavy, direct material representations of the ruler's “burden of office” (compare with Stross Reference Stross1988). Thus it is preferable to understand ihkaatz as a term pertaining to ceremonial objects obtained through tribute, or perhaps a specific measure of objects so obtained, rather than an exclusive reference to jade itself—even if jade happens to be the most archaeologically visible material exchanged in such a manner.

Table 11.

“Cargo”

The lexical root *kuch similarly denotes a “burden” or “load” and can also function as a verb for “to carry” such objects. While reconstructable only to Western Mayan, its reflexes apparently diffused about the lowlands as a part of Classic (or Formative) period culture (Kaufman with Justeson Reference Kaufman and Justeson2003:889–890). Likewise, kuch is relatively common in the Classic Maya script, written both phonetically [ku-chu] and as a logogram [KUCH], but it seems to have no direct economic implications there—usually functioning instead as a verb for “to carry.” Nonetheless, the word as it occurs ethnohistorically is tightly related to concepts of rulership, office, and responsibility among Ch'ol, Tzotzil speakers and others as a part of what is known as the “cargo system” (Cancian Reference Cancian1965; Vogt Reference Vogt1969). In this sense, *kuch operates somewhat like the root *ihqatz above, and this connection may reflect the inherently economic aspects of cargo-based leadership institutions both ancient and historic. It has long been understood that local religious leaders in many parts of the Maya area see their heightened economic and political status as a burden to be borne for the benefit of the group. To hold a position of leadership often entailed great expenditures, as well as the entitlement to demand offerings from members of the community in order to undertake such expenditures. As kuch frequently occurs in the dedicatory phrases on stone monuments in the Classic script, it may be that it alludes to economic responsibility in those contexts as well.

The verbal root *toj, signifying “to pay,” enjoys a widespread attestation in Mayan languages (Kaufman with Justeson Reference Kaufman and Justeson2003:783–784) (Table 12). Given its occurrence in both Eastern and Western Mayan branches as well as Yucatecan, the term can be attributed at minimum to a Core Mayan origin. Kaufman (with Justeson Reference Kaufman and Justeson2003:783) reconstructs it all the way back to proto-Mayan; I do not, given the lack of a Huastecan attestation. A nominalized form, /tojol/, with semantics denoting “price” or “value” comes into evidence somewhat later among the Western and Lowland Mayan language groups. To modern Western sensibilities, the semantics of this root may seem to more appropriately fall into the realm of commercial activity; yet tojol [to-jo-li] is found in painted texts on a number of Classic period polychrome vessels and on carved monuments in contexts generally thought to depict payment of tribute (Reents-Budet Reference Reents-Budet1994; Stuart Reference Stuart1995:359; Stone and Zender Reference Stone and Zender2011). Furthermore, the historically attested use of /to:j/ in Q'eqchi’ referring specifically to tribute payment suggests an ancient retention and demonstrates that the semantics remain compatible with use for vertical transfers of wealth. Use of a cognate in Yucatec to refer to debt obligations suggests noncommercial origins as well.

Table 12.

“To pay; price”

It should be further recognized that unidirectional payments occur near universally among simple human societies and typically arise in contexts of dowry payments and bride price, dispute settlements, and so on (Dalton Reference Dalton, Earle and Ericson1977; Earle Reference Earle2002; Sahlins Reference Sahlins1972). In all likelihood, *tohol arose among Mayan languages to denote just these types of payment. The term is considered here under appropriations and transfers because payments in and of themselves are one-way transfers of wealth, because of the widespread epigraphic acceptance of the term's association with tribute payment in Classic texts, and because the remnant etymological evidence supports noncommercial origins. It is acknowledged that the etymon is not necessarily in and of itself prima facie evidence for tributary relationships.

Commerce

The root *k'aay appears to have been an original proto-Mayan verb for the impersonal, purely economic exchange of goods (Kaufman with Justeson Reference Kaufman and Justeson2003:792–794) (Table 13). In the etymological sequence, the */k/ → /ch/ change in Greater Tzeltalan is apparent, as is the */k/ → /tx/ innovation in Huastecan, providing strong support for proto-Mayan origins. In extant Eastern Mayan languages this word can take on a nominalizing suffix (-b'il or -b'al) to produce a word for “market.” If such a form were reconstructable, it would suggest the presence of such institutions in the Guatemalan highlands from very early times. While the lost /b’/ in Q'eqchi’ gives rise to such a possibility, terms derived through simple inflection do not usually provide satisfactory evidence for the antiquity of such derivation. The concept of a market as a “selling thing” could have easily appeared at a relatively late date and spread through the highlands as a calque, or semantic loan, leaving no evidence of borrowing since the suffixation of each receiving language is typically used (Foley Reference Foley1997:385–386).

Table 13.

“Sell”

Intriguingly, the fact that reflexes of *k'aay came to mean “to buy” in Core Mayan, as opposed to “to sell” in Huastecan, and then eventually shifted to “to lose” in Western Mayan suggests that buying and selling may not have been conceptually distinct in proto-Mayan times—as would have been the case if systems of bartering were the dominant mode of exchange. Thus, the original proto-Mayan root may have essentially meant “to exchange” or “to trade” in the economic sense missing from terms such as *k'ex and *jal discussed previously. The ambiguity and lack of semantic marking associated with *k'aay evidently persevered through the divergence of Eastern and Western Mayan languages. Around that time *k'aay apparently underwent a major semantic shift among the Western Mayan languages, wherein it came to mean “to lose” or “to no longer have.” The use of this root in some Q'anjobalan languages in expressions meaning “to spend” may more accurately reflect its ancient economic connotations in proto-Central Mayan, thereby providing a fairly understandable chain of semantic shift. A much later possible Cholan reflex, ch'ay, appears repeatedly in Classic script in the context of death statements, read something to the effect of “it was lost/spent/extinguished his white/pure blossom/thing/spirit” (Freidel et al. Reference Freidel, Schele and Parker1995:183). The fact that the root both appears in the Cholan-based script and is attested in modern Tzeltalan languages allows us to reconstruct it to Greater Tzeltalan and thereby connect it to proto-Mayan *k'aay through Western Mayan.

The semantic shift of *k'aay/ch'aay to mean “to lose” in Western Mayan languages appears to have been roughly coeval with or to shortly follow the innovation of *konh as a word for “to sell” in Mayan languages (see Table 13). Kaufman (with Justeson Reference Kaufman and Justeson2003:795–796) reconstructs *konh to Western Mayan and explains its presence in Yucatecan languages through diffusion as part of Greater Lowland culture. This is presumably due to its absence from Eastern Mayan languages. The phonology, however, is equally consistent with—and perhaps more parsimoniously explained by—the term having been present in Core Mayan. If the latter were the case, the appearance of *konh with its solidly reconstructable semantics related to selling would mark our earliest linguistic evidence of true “commerce” in Mayan languages. Its absence from Eastern Mayan languages is easily explained by an early loss in that group's proto-language where it would have had to compete with *k'aay, which would have been consolidating its meaning as a term for “to sell.” Kaufman and Justeson's more chronologically conservative Western Mayan attribution, however, does securely establish presence of the root—and therefore the act of “selling”—by about 1000 b.c. in their chronology.

Somewhat later, reflexes of *konh proceeded to diffuse about the Maya area and even beyond, perhaps in association with new, more commercial trade practices or a new disposition toward them. In Greater Tzeltalan, the *k → ch innovation is evident, demonstrating a genetic transmission of the etymon from Western Mayan. In Greater Q'anjob'alan languages, however, the word often begins with an unexpected /ch/ or /tch/ that suggests borrowing. Kaufman (Reference Kaufman1976:110) implicates Greater Tzeltalan languages or, more specifically, Cholan ones as the donor group. The appearance of the /nh/ in the Chujan forms, however, indicates that the *nh → n shift in Greater Tzeltalan had not yet occurred at the time of borrowing. This both allows us to impose a relative chronology of the *k → ch and *nh → n sound shifts in Greater Tzeltalan and confirms that the borrowing was not from a later Cholan or Tzeltalan source. To explain the phonological distribution, Kaufman (with Justeson Reference Kaufman and Justeson2003:795–796) suggests multiple borrowings of cognates of *konh into various Q'anjob'alan languages from both Cholan and pre-Cholan sources.

*Konh also seems to have been borrowed into Xincan, a non-Maya language on the southern periphery of Mesoamerica in the southeastern highlands of Guatemala, as kunu (Campbell Reference Campbell, Justeson and Campbell1984). The source of this loan, given the initial /k/, must have been either the Yucatecan languages or a Western Mayan language from before the Greater Tzeltalan /k/ to /ch/ shift. Campbell (Reference Campbell, Justeson and Campbell1984:8) considers the latter more likely, given the time frame of other similar loan words into Xincan (see *k'iwik below). The full range of evidence is most consistent with either proto-Western Mayan or an early form of Greater Tzeltalan being the donor of *konh to a number of disparate languages, both Mayan and non-Mayan.

Furthermore, at some much-later point a nominalized reflex of *konh, *konhob’ became frozen in proto-Q'anjob'alan as a word for “town.” Close examination shows this to be the *konh “to sell” reflex combined with an instrumental suffix /–ob’/, denoting “a place where one sells things.” This inflection apparently became frozen in the proto-Q'anjob'alan vocabulary and developed into an established word for “town”—which was evidently “the place where one sells things”—even after *konh as a verb for “to sell” was lost in Q'anjob'alan languages. Reflexes of *konh may well have been pushed out of proto-Greater Q'anjob'alan in the process of incorporating a new term for selling, perhaps associated with new commercial practices or attitudes toward them, just as *konh was replaced with *txon from neighboring Greater Tzeltalan in which the initial /k/ sound had long before shifted to /ch/. The fact, however, that there is no indication of such a shift in the Kotoke complex suggests that this was a rather late development in Q'anjob'alan languages, and it can thus only be placed in the Postclassic period according to Kaufman's chronology. The initial /k/ and /-ob/ suffix in fact suggest Yucatecan origins, yet this would have had to have taken place before the root final velar nasal was lost in the *nh → n sound shift in proto-Yucatecan. The situation is obviously complicated, and remains to be fully resolved, but the etymon does seem to suggest towns as “the place where things are sold,” by the onset of the Postclassic period on the southern margin of the Maya lowlands.

Reconstructable terms for the specific act of buying tend to confirm what was just deduced above about the time frame for the origins of commerce. While the phonology of *man is disappointingly stable with a lack of relevant shared innovations available to differentiate borrowings from genetic transmission, distribution of the reflexes is not inconsistent with Core Mayan origins (Table 14). Its absence from Eastern Mayan languages would have to be explained perhaps by the appearance of *loq’ as a verb for “to buy” in proto-Eastern Mayan, which could have pushed out and replaced a reflex of *man at some point before that language diversified. As with *konh above, however, Kaufman (with Justeson Reference Kaufman and Justeson2003:797–799) more conservatively reconstructs the root only to Western Mayan and explains its Yucatecan attestations through Greater Lowland diffusion. This securely establishes the concept of buying, as distinguished from selling, among Western Mayan speakers.

Table 14.

“Buy”

The root *k'iwik, used to signify both markets and architectural plazas in several Mayan languages, is reconstructable at minimum to the period of Greater Lowland fluorescence based upon its distribution among Yucatecan and Tzeltalan subfamilies (Kaufman with Justeson Reference Kaufman and Justeson2003:799) (Table 15). The /k’/ to /ch’/ correspondence between Yucatecan and Cholan-Tzeltalan languages is clear, confirming that the root existed to undergo the *k’ → ch’ transformation in proto-Greater Tzeltalan. The appearance of the cognate k'wik in Mam, an Eastern Mayan language not generally known for accepting loans from the lowland culture complex, may suggest a still deeper reconstruction to Core Mayan. If it was loaned into Mamean from Greater Tzeltalan, this would have had to have taken place before 200 b.c., given the postulated time frame of the *k’ → ch’ transformation (Kaufman and Justeson Reference Kaufman and Justeson2007:199). Otherwise k'wik would have had to have been borrowed from a Yucatecan language—unlikely given the distance and terrain that separates Mam from the northern lowlands. Given that the root has survived in very few and very disparate Mayan languages, the evidence is much more suggestive of an ancient retention than vigorous later diffusion. Moreover, its early borrowing into and survival in some rather distant non-Maya languages, such as Xincan, where it appears as ki:’wi, and possibly Huave, where it occurs as kiap (Campbell Reference Campbell, Justeson and Campbell1984:8), also tend to confirm the root's great antiquity.

Table 15.

“Market”

Reconstructing the original semantic content of *k'iwik is also problematic given that, like virtually all other reconstructable Mayan words for “market,” it is synonymous with architectural “plaza.” That the original etymon in fact signified “plaza” or “patio” rather than “market” is suggested by its meaning in non-Mayan languages. In both Xincan and Huave it is glossed solely as “patio” (Campbell Reference Campbell, Justeson and Campbell1984:8). The term is thought to have diffused into Xincan from Greater Tzeltalan, along with a reflex of *konh “to sell” discussed above, at sometime during the Middle to Late Formative period when Greater Tzeltalan would have been spoken in the Guatemalan highlands in close enough proximity to explain the transfer (Campbell Reference Campbell, Justeson and Campbell1984; Mora-Marín Reference Mora-Marín2005). Widespread semantic adoption of *k'iwik as a term for “market” apparently occurred at some later point and only in the Maya area, after the word had been loaned into Xincan and Huave.

Entertaining the idea that the early semantics of *k'iwik denoted a patio, platform, or plaza allows some potential calibrating with archaeological data. Some of the earliest known artificial platforms in Mesoamerica are actually found on the southwestern margins of the Maya area along the Pacific coast of Guatemala and Mexico. At sites like Paso de la Amada in the Mazatan region of the Soconusco, large, leveled, open platform spaces that might be considered “patios” have been documented in association with perishable structures in Locona phase contexts between 1700 and 1500 b.c. (Lesure Reference Lesure1997). While there is debate as to the significance of these structures and whether or not they were truly associated with large open “plazas” (Clark Reference Clark, Herndon and Joyce2004; Lesure and Blake Reference Lesure and Blake2002), the fact remains that artificially leveled flat open spaces that might be called “patios” were present by the middle of the Early Formative period on the southwest margins of what is today considered “the Maya area.” Thus, a Core Mayan reconstuction of a term for patio could conceivably be supported by archaeological evidence.

If we accept for a moment that Core Mayans were using *k'iwik to denote the same type of platform or patio that appeared during the Locona phase in Soconusco, then we find that the archaeological and linguistic chronologies are in fairly close agreement. Given that glottochronology is a relatively weak dating method, calibration might simply entail pushing Kaufman's dissolution date for Core Mayan (1800 b.c.) forward several hundred years to correlate with the Locona phase (1700–1500 b.c.) archaeological developments. This would of course also imply that Core Mayan speakers were situated somewhere in close enough proximity to the Soconusco in order to be aware of such architectural innovations. Since the Soconusco is a lowland area, the finding could also be seen as supporting Brown's (Reference Brown, Staller and Carrasco2010:92–94) contention that proto-Mayan was a actually lowland language on the margin of the highlands, rather than a highland language on the margin of the lowlands as Kaufman (Reference Kaufman1976) maintains. Perhaps then those lowland proto-Mayan homelands were along the Pacific coast rather than on the northern slopes of the highlands as Brown implies. Indeed, if the Soconusco was all Mixe-Zoquean during the Early and Middle Formative period, as Kaufman's model of Mesoamerican linguistic dispersal maintains, what was a distant western language isolate like Huave on the Pacific coast of the Isthmus of Tehuantepec doing borrowing words for “patio” from Mayan? Should it not have instead been adopting them from Mixe-Zoquean?

Moving on, there is but one reported occurrence of *k'iwik in the hieroglyphic corpus. It evidently occurs on an unprovenienced Late Classic cylindrical vessel fragment in the form of aj = k'i-wi-ki or aj k'iwik (Dahlin et al. Reference Dahlin, Bair, Beach, Moriarty, Terry, Staller and Carrasco2010:196). Through use of the male proclitic aj-, this phrase could be used to signify “he of the market” and therefore possibly a “merchant.” Of course, it could also mean “he of the plaza” and thereby refer to a courtly official in charge of maintaining or organizing use of public space, in the same way, for instance, that aj tz'ib denoted a court scribe (Reents-Budet Reference Reents-Budet, Costin and Wright1998). The remaining glyphic context and accompanying iconography of this vessel were apparently lost, so little of substance can be said about the text. We might note, however, that the form aj k'iwik is decidedly Yucatecan and not the Cholan/Tzeltalan *ch'iwich that we would expect from the southern lowlands of the Classic period. It is also worth noting that this would be the first, and only, recognized instance of a term for “market” in the Classic period Maya corpus.

In fact, despite the immense corpus of Maya hieroglyphic writing—estimated to consist of over 10,000 texts written on stone, clay, and wood (Houston Reference Houston2000)—there is not a single known unequivocal reference to markets or commercial exchange in Classic texts. Nor are there depictions of markets or commercial exchange in Classic period art. The single known possible exception is a recently discovered set of murals from the Chiik Nahb’ plaza group at Calakmul (Carrasco Vargas et al. Reference Carrasco Vargas, López and Martin2009, Martin Reference Martin, Golden, Houston and Skidmore2012). In scenes from these murals, a range of goods—mostly foodstuffs such as atole, tamales, corn, and salt, but also including ceramics and possibly needles—are apparently being offered for sale by individual ‘vendors,’ many if not most of whom are women and children (Hull Reference Hull, Staller and Carrasco2010). Dating to the Late Classic period, the scenes are thought by many to present a Classic Maya marketplace (Hull Reference Hull, Staller and Carrasco2010, Martin Reference Martin, Golden, Houston and Skidmore2012). As there is no clear explanation of the context in the accompanying text, there are alternative interpretations of the work—for example, that it may instead depict a banquet (Boucher and Quiñones Reference Boucher and Quiñones2007). At several locations within the murals, however, there are indications of goods being provided in exchange for the commodities dispensed. The small objects supplied are small and homogenous and could represent cacao beans, which would be an expected currency of the period (see further discussion below). Exchange of one good for another is exactly the type of quid pro quo we would use as a diacritic of commercial activity, and this strongly supports the proposition that market commerce is indeed what was depicted at Chiik Nahb. It is doubly interesting that these murals appear in the general vicinity of a very large plaza that was long ago inferred to be a marketplace archaeologically (Folan Reference Folan1992). It is also worth noting, however, that the layout and architecture of this plaza generally bear little or no resemblance to other archaeologically postulated marketplaces of the Classic period, such as the East Plaza of Tikal (Jones Reference Jones1996). They do, arguably, show some similarity to Terminal- and Post-Classic ‘commercial’ plazas like the Court of the Thousand Columns at Chich'en Itza (Ruppert Reference Ruppert1952).

Though the scenes and texts are badly degraded, the range of goods being exchanged in the Chiik Nahb’ murals, and the commodities emphasized, appears to be rather modest. If there is any emphasis on a particular good or class of good in the murals as a whole, it seems to have been upon foodstuffs and atole—depicted twice in the work at the apparent beginning and end of the text. The most exotic commodity presented is salt, which would probably have been brought in from the coast, though there are ways that even this could have been produced locally. There is also a persistent theme in the murals of consumption of the goods on site rather than, for instance, procurement for removal in the interest of household provisioning. Was this a defining characteristic of early Maya markets? And we can rather confidently say, based upon archaeological evidence for a much broader range of exotic materials available to the Classic Maya, that Landa's (Reference de Landa and Gates1978[1566]:37–38) bold assertion that the full range of goods produced in the land were traded in local markets is not at all borne out for the Classic period in these murals. Aztec Tlatelolco, this was apparently not. If the hyper-commercialist scholars finally have their victory in obtaining some concrete evidence for at least one Classic period marketplace, it has to be bittersweet in the rather modest picture it presents. This is supposed to be the great marketplace at one of the most prominent Classic period political centers? And the cornerstone of it is old women selling atole for on-the-spot consumption? Though such activity may well have served to help redistribute some surplus agricultural production across a small region, it is difficult to see it as the basis of an intensively integrated market economy providing the foundation of Classic peiod polities (see Chase and Chase Reference Chase and Chase1996, Masson and Friedel 2013).

That the historically attested use of all reconstructable terms for commercial “market”—*k'aayb'al, txonb'al, and chonlab’, as well as *k'iwik—is almost invariably synonymous with architectural “plaza” suggests a deep-seated and widespread cognitive association between these two elements among Mayan speakers. Only among some late Yucatecan languages, such as Itzaj and Mopan, does some partial semantic separation and specificity begin to appear. These languages can use the term konol—albeit often requiring some modifiers—to specifically denote a market independent from the plaza in which it is held. The obvious implication, cognitively speaking, is that all plazas were potential markets to the ancient Maya and that markets were inherently expected to be held at some form of plaza. This may actually help to explain the widely recognized absence of archaeological evidence for Classic Maya markets. They may have simply occurred on whatever open plaza space was available and, in most cases, were not formally designated or architecturally designed by the ruling classes.

Moving on to our next etymon, the verb root *p'ol carries the simple meaning of “to increase.” In many glosses, however, it seems to specifically connote an increase associated with commerce in the form of selling or merchandise (Table 16). In that sense, it apparently represented a general Maya conception of “profit,” and such semantics are widespread enough to suggest fair time depth of use. Yet the distribution of the root itself is rather limited, being found in general only among languages of the lowland group (Kaufman with Justeson Reference Kaufman and Justeson2003:797). There is also no solid phonological basis for reconstructing it to any great antiquity; not even to proto-Greater Tzeltalan. The only sound change it can be shown to have undergone is in Ch'orti’, and that is a very late, highly irregular one. Nonetheless, the attestation of cognates in languages as disparate as Tzeltal and Ch'olti’ suggests either a Greater Tzeltalan presence or substantial later diffusion. The fact that it appears in Tojolob'al indicates at least some borrowing and makes later diffusion the preferable explanation. The primary time frame for lowland interaction was, of course, the Classic period. The subset of attesting languages therefore implies a relatively widespread conception of profit associated with commerce in the lowlands of the Late Classic.

Table 16.

“Increase”

Valuables

Perhaps the most familiar ancient Mesoamerican valuable is the cacao bean—widely known to have served as a currency, or commodity money, at the time of Spanish contact (Díaz del Castillo Reference Díaz del Castillo and Garcia1928; Feldman Reference Feldman1985; Hassig Reference Hassig1985; Rojas Reference Rojas1998; Tozzer Reference Tozzer1966). There has already been much discussion and debate over the history of the word for cacao in Mesoamerican languages (Beekman et al. Reference Beekman, Cowgill, Dakin, Hopkins, Rosenswig, Sampeck, Swanton, de Ávila, van Doesburg and Wichmann2010; Dakin and Wichmann Reference Dakin and Wichmann2000; Kaufman and Justeson Reference Kaufman and Justeson2007). To summarize, many of the preeminent linguists working in the region adhere to the established, orthodox view maintaining the word *kakaw diffused about ancient Mesoamerica from a Mixe-Sokean language (Campbell and Kaufman Reference Campbell and Kaufman1976; Justeson et al. Reference Justeson, Norman, Campbell and Kaufman1985; Kaufman and Justeson Reference Kaufman and Justeson2007). Some others, however, subscribe to a more recent proposition that it is proto-Nahua in origin, being a reduplicated innovation on a Southern Yuta-Nawan word for “egg,” *kava (Brown Reference Brown, Staller and Carrasco2010; Dakin and Wichmann Reference Dakin and Wichmann2000; Macri and Looper Reference Macri and Looper2003). The present author considers the Nahua origin hypothesis quite thoroughly refuted. If that language had been involved, at least some of the many Mesoamerican languages that adopted the word would have done so with a word final /t/ to correspond with a proto-Nahua *kakawatl. But this was never the case. When we add to this the facts that the cacao plant would have been utterly foreign from an ecological perspective to the high and dry climate of central Mexico where Southern Yuta-Nawan languages were originally spoken, and that cacao use was already well established in southern Mesoamerica by the Early Formative period (Powis et al. Reference Powis, Hurst, del Carmen Rodríguez, Ortíz C., Blake, Cheetham, Coe and Hodgson2007), no viable reason exists for southern regions to have adopted a term for cacao from Uto-Aztecans. The Nahua hypothesis just seems unbearably far-fetched.

Nevertheless, what is important here for present purposes is the uncontested fact that the preeminent Mesoamerican word for cacao was at some point borrowed into Mayan languages from without. The real question for us then is the time frame of its borrowing and various semantic associations, not its original source. The fact of its non-Maya origin is demonstrated by virtue of the relatively few internal phonological sound changes the word has undergone in any Mayan languages (Kaufman with Justeson Reference Kaufman and Justeson2003:1104–1105) (Table 17). It cannot possibly have been present in proto-Mayan, or historical sound developments would have caused the reflex to occur as chachaw in modern Huastec rather than kakaw. Nor did the word undergo the Late Formative Greater Tzeltalan *k → ch sound shift, despite satisfying the few albeit complex conditions that allowed it to avoid being blocked (Kaufman and Norman Reference Kaufman, Norman, Campbell and Justeson1984:84), or it would likewise occur as chachaw in Cholan and Tzeltalan languages. The earliest changes it can be shown to have undergone are post proto-Cholan, in which complex vowels are simplified in certain Cholan languages like Ch'orti’. There are, however, plenty of depictions of cacao in Maya art and direct references to cacao, written as ka-ka-wa, appearing by Early Classic times in the hieroglyphic script (Stuart Reference Stuart1988, Reference Stuart and Kerr1989). These become particularly common in dedicatory texts written upon cylindrical drinking vessels of the Late Classic. Based upon the combined evidence, one can deduce that the word entered the Maya area somewhere between 200 b.c. and a.d. 400 (Kaufman and Justeson Reference Kaufman and Justeson2007). This roughly conforms with previous linguistic studies suggesting a Late Formative introduction for the word *kakaw into Mayan languages (Justeson et al. Reference Justeson, Norman, Campbell and Kaufman1985).

Table 17.

“Cacao”

Residue analyses, however, have more recently shown cacao to be present archaeologically in the northeast Peten of the lowland Maya area by the late Middle Formative period (600–400 b.c.) (Hurst et al. Reference Hurst, Tarka, Powis, Valdez and Hester2002; Powis et al. Reference Powis, Valdez, Hester, Hurst and Tarka2002). Indeed, the spouted vessels in which such residues were found date back even further in the Ulua Valley of Honduras to the southeast of the Maya area (Henderson and Joyce Reference Henderson, Joyce and McNeil2006). These developments clearly predate the widespread adoption of the term *kakaw in Mayan languages and require explanation. Kaufman and Justeson (Reference Kaufman and Justeson2007:201) have recently proposed that there was an earlier, native Mayan word for wild cacao, *pe:q, which survives only in a few Eastern Mayan languages, such as K'iche’ and Cakchiquel, as a term for cacao. Cognates may exist in several other Mayan languages with shifted meanings that include “skirt” and the quantity “8,000” (Kaufman with Justeson Reference Kaufman and Justeson2003:1105). It is particularly interesting and increasingly suggestive that the earliest presently known archaeological cacao residues have been documented in Early Formative Barra phase contexts from Paso del la Amada on the Soconusco (Powis et al. Reference Powis, Hurst, del Carmen Rodríguez, Ortíz C., Blake, Cheetham, Coe and Hodgson2007), dating to between 1900 and 1700 b.c. Still, there is insufficient linguistic evidence to reconstruct the native Mayan root etymon *pe:q beyond proto-Central Mayan.

It is difficult to reconcile all the conflicting evidence and explain the sequence of linguistic development related to terms for cacao. It is, of course, conceivable that new methods of preparing cacao-based drinks resulted in a relatively late, widespread adoption of a foreign word. There is certainly archaeological evidence for such culinary changes in the rather abrupt disappearance of spouted vessels from ceramic assemblages of the Classic period (Henderson and Joyce Reference Henderson, Joyce and McNeil2006), but it is hard to believe terms for a new drink would universally eclipse old terms for the plant itself. Questions also remain as to whether the appearance of a newly domesticated variety of cacao would sufficiently warrant universal adoption of a completely new term and abandonment of the old one. The most viable explanation may be that the new term *kakaw diffused as part of a widespread recognition of the pods as a useful intercultural commodity unit of exchange above and beyond their use in preparing a flavorful beverage (see Sampeck commentary in Beekman et al. [2010:431]). Increasing familiarity with a foreign term for cacao as a currency could contribute greatly to the demise of an old Mayan word for a generally nonlocal plant used primarily to prepare a drink mostly consumed by a restricted elite.

Like cacao, the use of cloth as an exchange medium was documented both among the Maya and across Mesoamerica at the time of Spanish contact (Feldman Reference Feldman1985; Rojas Reference Rojas1998). Its presentation to elites as payment or gifts is frequently depicted in Classic Maya art (Schele and Mathews Reference Schele and Mathews1998:89–91; Stone and Zender Reference Stone and Zender2011:81). There are numerous terms for cloth in Mayan languages, often shared by different subfamilies with subtle semantic variations (Kaufman with Justeson Reference Kaufman and Justeson2003:1011–1018) (Table 18). While none of the etyma can be confidently reconstructed back to proto-Mayan proper, no fewer than three terms, *tyiinham, *buhq, and *nooq’—plus ancillary terms for things like cotton seed (*tuhx) and spindle whorl (*peteht)—were present among Core Mayan speakers (Kaufman with Justeson Reference Kaufman and Justeson2003:1011–1018). Indeed, substantially more lexical roots for cloth were found and examined for this study than could be presented here. Many details concerning the history of weaving among the Maya could probably be gained by closer examination of all these terms and the relationships between them but, for present purposes, suffice it to say that there is no doubt that the Maya were active producers of cotton cloth by at least the time of Core Mayan dispersal.

Table 18.

“Cloth/cotton”

One reflex of the lexical root *b'uhq can be found in the prophetic texts of the Postclassic codices as bu-ku. The first-person possessed form also occurs in a mythological scene painted upon a Late Classic cylindrical vase (K1398; Hull et al. Reference Hull, Carrasco and Wald2009) treating a thieving rabbit and God L as nibuku [ni-bu-ku], “my clothes” (Hull et al. Reference Hull, Carrasco and Wald2009). Yet another lexical root, /yub/, is found twice in the text of another Classic period vessel (K1728; Stuart Reference Stuart1995), written yu-bu(-te’). The scene on this vessel commemorates some type of “payment” of goods—including cloth—to a lord of Motul de San Jose (Stuart Reference Stuart1995:358–359). The operative verb in the text is a reflex of *tojol, discussed above, providing evidence for cloth being used as a form of payment in ancient times. The title of an attending personage in the scene refers to him as aj yub. As an etymological root *yub is unfortunately a dead end, since it is documented only in Yucatec—in which yubte’ survives as “manta de tribute” or “tribute cloth” (Barrera Vásquez Reference Barrera Vásquez1995). The attendant in the scene might therefore be inferred to be a “keeper of the tribute cloth.”

Both greenstone (jade, jadeite, and serpentine) and marine shell are also known to have served as key valuables in Contact period Mesoamerica (Feldman Reference Feldman1985; Hassig Reference Hassig1985; Rojas Reference Rojas1998; Tozzer Reference Tozzer1966) and are thought by many to have performed similar functions in ancient times (Freidel et al. Reference Freidel, Reese-Taylor, Mora-Marín, Masson and Freidel2002). Workshop areas of marine shell bead production are archaeologically known to date to at least as early as Middle Formative times in several parts of the Maya lowlands (Aizpuruia and McAnany Reference Aizpuruia and McAnany1999; Hohmann Reference Hohmann2002). From a linguistic perspective, separate roots for both land snail and conch as living creatures can be traced back to proto-Mayan in the form of *puur and *pemech, respectively (Kaufman with Justeson Reference Kaufman and Justeson2003:654–657). Proto-Mayan *t'oot’ probably referred to the shell carapace itself, which is our main point of interest here as an exchange commodity. However, the fact that proto-Mayans were aware of the marine-dwelling conch as reflected by *pemech is also important information.

While less is known about the extent and time depth of jade production, certain quarry and production sites are known from the northern margins of the Guatemalan highlands (Kovacevich Reference Kovacevich2006; Rochette Reference Rochette2009). Jade and jadeite objects are not uncommon archaeological finds throughout most of the Maya area, with small beads and ceremonial celts being among the most typical forms encountered. Jade items also figure prominently in Maya art of the Classic period (Freidel et al. Reference Freidel, Reese-Taylor, Mora-Marín, Masson and Freidel2002; Stone and Zender Reference Stone and Zender2011:71; Stuart Reference Stuart, Guernsey and Reilly2007) and were a central part of Olmec cosmology and value systems long before that (Fields Reference Fields1989). Unfortunately, no etyma were found in this study that could specifically illuminate the temporal development of jade use through linguistic reconstruction. This does not mean that ancient Mayans were not aware of the material, nor that they did not speak of it. Rather, it suggests a preference for morphological compounds in reference to the stone in everyday speech. In all likelihood, jade, jadeite, and serpentine were referred to collectively in ancient Mayan languages in much the same way as archaeologists sometimes do in English—as “greenstone,” or yaxtun in Mayan.

There may, however, be a methodological way around the absence of reconstructable terms for jade through terminology for forms into which the material was fashioned, such as decorative beads. One root for the latter, *uuh, can be reconstructed rather soundly to proto-Mayan based upon distribution of its cognates throughout all major subgroups of the Mayan language family (Kaufman with Justeson Reference Kaufman and Justeson2003:1030) (Table 19). Given that shell and greenstone are far and away the most common raw material for beads found archaeologically in the Maya area (Willey Reference Willey1972; Masson Reference Masson, Masson and Freidel2002), there is every reason to believe that the items proto-Mayans were referring to with *uuh included drilled pieces of jade and marine shell. Jade and Spondylus beads certainly figure prominently enough in Maya art of the Formative and Classic periods (Freidel et al. Reference Freidel, Reese-Taylor, Mora-Marín, Masson and Freidel2002; Stuart Reference Stuart1995; Taube Reference Taube and Houston1998), though they are rarely referred to by specific material in texts. Beads are nearly ubiquitous archaeological finds from the lowland Maya area by Middle Formative times, and it is also useful to note recovery of jade beads from disparate Early Formative Locona phase contexts along the Soconusco (Blake et al. Reference Blake, Lesure, Feddema, Hill, Gosser, Clark and Lowe1993; Lesure and Blake Reference Lesure and Blake2002; Rosenswig Reference Rosenswig2010).

Table 19.

“Bead/necklace”

Feathers are yet another type of valuable known from the period of Spanish contact (Feldman Reference Feldman1985; Rojas Reference Rojas1998). The most common and only deeply reconstructable etymon for “feather” in Mayan languages, *k'uk'um, appears to be the instrumentalized form (-Vm) of the noun root for quetzal bird *q'uq’, dating from after the root had undergone the *q’ → k’ phonological change in either a Western Mayan language or proto-Yucatecan. Thus, the term as a pan-Maya word for feathers seems to originate and diffuse relatively late as a Greater Lowland Maya culture word (Kaufman with Justeson Reference Kaufman and Justeson2003:614–615) (Table 20). This does not imply that there was no earlier term for bird feathers in Mayan languages, only that none were so widespread and durable so as to be deeply reconstructable today. Most lexical roots may have simply been lost over the millennia, and it would have been possible to communicate the idea of “feathers” with compound terms and euphemisms that often do not reconstruct using historical methods. Other Mayan words used to denote the external covering of a bird, for instance, such as *xiik, *’ism, and *tzo'tz, reconstruct better as wing, fur and hair, respectively, thus representing the idea of feathers cognitively as the “hair” or “fur” of a bird. What can be said is that only one term, *k'uk'um, apparently acquired sufficient importance at some point to diffuse across the Maya lowlands and that it diffused from either Greater Tzeltalan or Yucatecan languages. This had to have taken place post-Western Mayan dispersal and after Greater Tzeltalan underwent the *q’ → k’ sound change—otherwise it would appear as q'uq'um in Q'anjob'alan languages—yet before the end of the period of Greater Lowland diffusion. Greater Tzeltalan is the most likely innovator given its proximity to the quetzal bird's highland habitat. Thus, Greater Tzeltalan was evidently still close enough to the highlands to recognize *k'uk’ as a word for the quetzal bird and derive a word for feathers from this term. Greater Q'anjob'alan languages subsequently experienced a shift causing *k'uk'um to signify “sparrow hawk” while Q'anjob'al itself at some later point borrowed q'uq’ from an Eastern Mayan language.

Table 20.

“Feather”

Development of a new intercultural term for feathers during the Late Formative or Early Classic would not be terribly surprising, given the rising importance of feathers apparent in Classic period art and ritual (Freidel et al. Reference Freidel, Schele and Parker1995:236–239; Schele and Mathews Reference Schele and Mathews1998; Stuart Reference Stuart1995). They were a commonly depicted element of tribute and prestation packages, often accompanied by stacks of folded cloth and other bundles of goods in painted scenes on cylindrical vases and murals. Feathers are also explicitly referred to in the hieroglyphic script as k'uk'um, written as K'UK’-ma, for instance, on the face of Naranjo Stela 12 in reference to an extraction of wealth following military conquest. The late pervasiveness of *k'uk'um may specifically reflect the increasing importance of quetzal feathers in Mesoamerican exchange systems. In many ways, its wide distribution—yet fairly shallow time depth—bears some resemblance to the etymology of *kakaw. This could be a product of the similar functional roles the two commodities played in pre-Columbian Mayan economies and the similar time frame at which they became necessary components in interregional exchange systems.

The terms that came to be used for “money” in Mayan languages historically are almost without exception a compound of words for “sun” (*q'iin) and “excrement” (*tzaa’)—a colloquialism for “metal” denoting gold and silver in particular. This same semantic compound was apparently diffused throughout Mesoamerica in pre-Columbian times as a calque for shiny, malleable materials—including copper, silver, gold, and even bronze—that were fashioned into precious items such as ornaments, bells, and small nonfunctional axe-heads (Díaz del Castillo Reference Díaz del Castillo and Garcia1928; Feldman Reference Feldman1985; Rojas Reference Rojas1998; Tozzer Reference Tozzer1966). Such metals were much more common in other parts of Mesoamerica than they were in the Maya area, in which they appear archaeologically only in relatively late Terminal and Postclassic period contexts. The linguistic data tend to reaffirm this fact by demonstrating that the *tzaa’ = q'iin compound was borrowed widely into the Maya area at a very late date through one of the lowland languages (Kaufman with Justeson Reference Kaufman and Justeson2003:400) (Table 21). This is reflected in the initial /t/ found in many languages where /ch/ or /tz/ would be found if the words had been transmitted genetically in earlier times. The earliest secure phonological evidence traces it to proto-Yucatecan, generally agreed to have dispersed at around a.d. 1100, from which it may well have diffused about the rest of the Maya area. The appearance of tumin in Ch'orti’ is actually of Spanish derivation (Wisdom Reference Wisdom and Stross1950). It is a corruption of tomin, a term for a small, low denomination Spanish coin that was also borrowed into Nahuatl in central Mexico (Berdan Reference Berdan and Isaac1986:297). From the fact that there are no archaeologically occurring forms of gold or silver currency known from the Maya area prior to Spanish contact, one can deduce that the use of the word for metal in reference to “money” was purely a result of contact with European coinage.

Table 21.

“Money/metal”

Furthermore, no linguistic evidence could be found to indicate that the pre-Columbian Maya had any deep-seated abstract conceptions of “money” as a commercial technology at all. Each form of currency—be it metal, cacao, shell, cloth, or feathers—was simply a valuable commodity that happened to serve exchange purposes, and the notion of “money” as an item that specifically served as object for exchange or unit of value was apparently never lexicalized. Nahuatl, in contrast, had a general inclusive term, tlacoualoni, that meant “el instrumento para comprar” (Rojas Reference Rojas1998:193), making it cognitively analogous to Western terms for money. The presence or absence of true “money”—formalized objects of stable value that stand independent of the variable intrinsic worth of any particular consumable good—has potentially important implications for the degree of commercialism and, by extension, certain structural exchange capacities that should be afforded to an economic system (Dalton Reference Dalton, Earle and Ericson1977; Einzig Reference Einzig1948; Freidel et al. Reference Freidel, Reese-Taylor, Mora-Marín, Masson and Freidel2002; Smith Reference Smith2004; Snell Reference Snell, Sasson, Baines, Beckham and Rubinson1995; Speal Reference Speal2005; Weatherford Reference Weatherford1997). Commodity monies, particularly perishable ones, face limitations, at least in theory, given inherent issues of fungibility, and periodic and geographic scarcity.

While obsidian and chert may not be generally recognized elements of the Mesoamerican complex of pre-Colombian preciosities, archaeological evidence indicates that—above and beyond their use in the manufacture of everyday utilitarian tools—these materials were also widely traded as prestige goods in certain forms among the ancient Maya (Hammond Reference Hammond1972; Hester and Shafer Reference Hester and Shafer1984; Shafer and Hester Reference Shafer and Hester1991). Both obsidian and chert were sometimes fashioned into elaborate eccentrics and finely crafted bifaces that in many cases appear more ceremonial than functional (Hester Reference Hester, Plew, Woods and Pavesic1985; Iannone Reference Iannone1993; Meadows Reference Meadows2001). Obsidian was also precious enough to be occasionally ground into beads. There is strong evidence that these lithic materials circulated in fairly distinct alternative exchange networks of varying geographical reach according to quality of raw material and manufactured form (Braswell and Glascock Reference Braswell, Glascock and Glascock2002; Hester and Shafer Reference Hester, Shafer, Schwartz and Falconer1994; Santone Reference Santone1997; Scarborough and Valdez Reference Scarborough and Valdez2009; Speal Reference Speal2009). Moreover, the durability, high archaeological visibility, and great antiquity of lithic materials exploitation makes it particularly useful in the calibration of linguistic chronology with archaeological data.

Kaufman (with Justeson Reference Kaufman and Justeson2003:442) reconstructs two terms for silicious stone back to proto-Mayan, *tyaah and *tyooq’ (Table 22). We can be fairly confident in attributing the first, *tyaah, to proto-Mayan given its attestation in Eastern and Western Mayan languages and Huastec with the expected proto-Mayan sound correspondences among them. The second, *tyooq’, is somewhat less secure given that its purported cognate in Huastec occurs only as part of the name of an insect with an unclear significance. Still, *tyooq’ can be minimally reconstructed to Core Mayan given its occurrence in Eastern and Western Mayan languages and Yucatec, with the expected /t/ to /tz/ word initial sound correspondence between Mam and Ch'olan tongues. Perhaps the most interesting contrast between *tyaah and *tyooq’ is that reflexes of the former are unmarked in meaning with regard to flint versus obsidian in most Mayan languages. Cognates of the latter, on the other hand, all specifically denote flint.

Table 22.

“Obsidian/flint”

Incidentally, this contrast in meaning is exactly what one might expect from a dispersing language group situated on the margins of the Sierra Madre del Sur along the Soconusco. It implies an earlier unmarked term that would apply to all flakeable silicious stone (though probably based by default on volcanic highland obsidian through familiarity with that material) and a later, more marked word applying specifically to the flint/chert deposits encountered as some speakers began moving into the karstic Maya lowlands. Proto-Core Mayan times would be a logical point for novel, more specific terms for flint to arise given that proto-Huastecan had theoretically just diverged northward toward the Gulf Coast several generations before and Yucatecan was presently in the process of breaking off on a similar path toward the Yucatan. With these splintering linguistic groups expanding the range of Mayan speakers, it is easy to explain a widespread familiarity with both highland and lowland lithic raw materials through both verbal interaction and trade in material goods.

In Classic times, flint became a very common iconographic element in lowland Maya area art (Stone and Zender Reference Stone and Zender2011:83). It frequently appears in the hieroglyphic script as to:k’, written both syllabically as to-k'a and logographically as TOK’. Perhaps its most common glyphic representation is as part of “flint-shield” statements made in reference to warfare (Stone and Zender Reference Stone and Zender2011:83). Both hand-wielded eccentrics and lanceolate bifaces serving as spear points are depicted on numerous stone monuments found from Yaxchilan to Caracol. There are also apparent references to obsidian as tah in the Classic script, in one case directly upon an obsidian core (Freidel et al. Reference Freidel, Reese-Taylor, Mora-Marín, Masson and Freidel2002:58). Cha'y in Ch'orti’, on the other hand, seems to be a late borrowing from a K'iche'an language, possibly reflecting later lithic trade patterns between those populations in Postclassic period or historic times.

Like lithics, ceramics are known to have served as both high-value prestige goods and as common utilitarian items in ancient times. They are likewise included in this analysis predominantly because of their premium value in calibrating archaeological with linguistic chronology. For unlike lithics, which were exploited in Mesoamerica from the most ancient Paleoindian times, ceramics appear at a very specific time and place in the Mesoamerican archaeological record. That point of origin appears to have been the Soconusco region on the Pacific Coast of extreme southern Mexico, just to the southwest of what is generally considered the Maya area today. Until recently, the time frame of innovation was thought to be about 1500 b.c. based on radiocarbon assays (Blake et al. Reference Blake, Clark, Voorhies, Michaels, Love, Pye, Demarest and Arroyo1995).

Kaufman (with Justeson Reference Kaufman and Justeson2003:986) posits ceramic technology to proto-Mayan speakers through reconstruction of the etymon *laq (see also Kaufman and Justeson Reference Kaufman and Justeson2009:230). The etymological data strongly support this conclusion, with the /q/ to /k/ correspondence between Eastern and Western Mayan languages in place and a cognate attested in both Huastecan languages (Table 23). One is given some pause by the fact that several very old Chicomuseltec dictionaries gloss lak as “tol” or “gourd” (Dienhart Reference Dienhart1989). This gives rise to the possibility that the original root *laq referred to a container made from plant matter. Such origins are much less likely, however, given that the semantics in all other attesting languages so consistently indicate specifically ceramic forms. In order for the original root to have signified “gourd container,” the same semantic shift from gourd to ceramic would have had to occur independently in Core Mayan and Huastec at minimum, and at radically different points in time. In Core Mayan it would need to have taken place before the diversification of that proto-language, and then in Huastec at some much later date after Chicomuseltec and Huastec had diverged—quite a coincidence. If we posit that the Core Mayan shift occurred any later, even more separate proto-languages would have had to independently innovate the same change, making it increasingly unlikely with each stage in time. While the original semantics of *laq may have denoted a rather generalized ceramic form, potentially applying to either a pot, bowl, dish, or plate, it seems rather clear from lowland Maya art and text that the reflex lak, written la-ka, referred to a large flaring ceramic dish—and not any shape easily produced from a gourd—at least by Classic period times (Houston et al. Reference Houston, Stuart and Taube1989). One late semantic shift to “gourd” in the single language of Chicomuseltec is easily the most parsimonious explanation for the overall distribution of glosses, and the soundest conclusion is that proto-Mayans did indeed possess ceramic technology referred to with an overarching, fairly general term.

Table 23.

“Bowl/plate/dish”

Kaufman (with Justeson Reference Kaufman and Justeson2003:982–984) also reconstructs various permutations of another term for ceramic vessel, *q'ib’, to Central Mayan (Table 24). In semantics it appears to more specifically denote a tall pitcher or drinking cup. A potentially related term, uk'ib’, is relatively common in the hieroglyphic script. The latter is often found in the dedicatory texts of Classic period cylindrical drinking vessels in its possessed form, yuk'ib’ [yu-k'i-bi and yu-UK’-b'i among others] (Houston et al. Reference Houston, Stuart and Taube1989). It is interesting that neither the epigraphic form nor the attested Tzeltalan cognate k'ib’ present the expected Greater Tzeltalan *k’→ ch’ sound shift. Though there were phonetic conditions that blocked this change, they are not apparent in this context, preceding /i/, and a nongenetic transmission of the root is thereby suggested. That the written form is nonetheless related to *uk, “to drink,” is clearly shown by the depiction of a person drinking in the glyph. The expected Cholan form would be *ch'ib or *ch'ab, in accord with the Tzeltalan cognate uch’ “to drink.” The attested glyphic forms could reflect Yucatecan influence, or perhaps an archaism (a retention, at least in formal, elite speech, of a preshift form) (David Mora-Marín, personal communication 2012).

Table 24.

“Pitcher/jar”

The main obstacle to reconstruction here is that so many Mayan roots signifying cup, jar, or pot are apparently morphologically derived forms of the proto-Mayan verb *uk’ meaning “to drink,” which bear strong phonological similarity and can be analyzed as “drinker” or “drinking thing” in one language or another. The whole convoluted sequence may reflect a long-standing Mayan tradition of calquing by simply referring to ceramic jars, pitchers, and large pots as “drinkers.” The apparent noncorrespondence between proto-Mayan /k’/ in *uk’ and the /q’/ present in many Eastern Mayan forms, however, as well as a /k’/ based term in those same language groups, remains a major problem in confirming a relationship between *q'ib’ and uk'ib’. While it is entirely possible that there were two completely separate roots operating in Mayan languages and that their similarity in semantics and phonology is purely coincidental, it may also be that all these terms are derived from the same ancient root but became convoluted in ways that are not presently understood. What is clear and most relevant to the present study is that, in addition to an original, generalized word for ceramic container in *laaq, Mayan languages had developed at least one additional, more specialized term for ceramic drinking vessels by at least the time of Central Mayan dispersal, and possibly earlier.

If the original proto-Mayan semantics of *laq did then indeed denote ceramic vessels, as the evidence suggests, then Kaufman's standing chronological framework appears to be undermined by his own reconstructive data. As mentioned above, there are no archaeologically known ceramics from the Maya area—or from Mesoamerica as a whole—during the period Kaufman proposes proto-Mayan to have been in existence (prior to 2200 b.c.). The earliest known ceramics in the Maya area proper do not in fact appear until the Swasey phase of the Middle Formative period lowlands around 1000 b.c. or so (Andrews and Hammond Reference Andrews and Hammond1990). The only place in Mesoamerica that ceramic technology is known to have existed anywhere near the time frame that Kaufman proposes proto-Mayan was spoken was along the Soconusco. Incidentally, recent calibration and additional radiocarbon dates from this region have pushed the beginning of the earliest ceramic-bearing Barra phase back to about 1900 b.c. (Clark and Cheetham Reference Clark, Cheetham, Carrión and Cook2005; Rosenswig Reference Rosenswig2008). This falls tantalizingly close to Kaufman's 2200 b.c. proto-Mayan divergence estimate and might be taken not only as a point of chronological calibration, but perhaps as evidence of a cultural connection. Presuming archaeological dating to be the more reliable of the two methods, it only makes sense to adjust the Mayan linguistic dispersal chronology with the earliest known ceramic technology in southern Mesoamerica, which appears at about 1900 b.c. This should remain the case at least until such a time as a more ancient pottery assemblage is discovered in proximity to the Maya area. Perhaps it is also best to reset the estimated proto-language dispersal date forward a century from this point to account for internal familiarization with the new technology within proto-Mayan—thus setting 1800 b.c. as a date for complete separation with Proto-Huastecan.

An alternative, of course, would be to accept Brown's model of a very late proto-Mayan divergence at 400 b.c. and to reject Kaufman's chronology altogether, as well as any correlations between it, the present etymological reconstructions, and the archaeology of the Soconusco as purely coincidental. The appearance of ceramic terminology by 400 b.c. would, in that case, not be an issue at all, nor even very informative, as ceramics would have been present in the Maya area proper for some 600 years. The bigger question, however, would then be why there was only one reconstructable proto-Mayan root for pottery and not a wider range of ceramic vocabulary in evidence from the beginning. Far more devastating to Brown's model, however, is the fact that hierarchical “lowland Maya” archaeological culture was already well established in the central Peten at sites like Nakbe’ by 400 b.c. (Hansen Reference Hansen2002). To accommodate archaeological facts then, Brown's model would either have to argue that the Middle Formative residents of the southern lowlands were not Mayan or that proto-Mayan was spoken from the lowlands of the Yucatan through the Guatemalan highlands during the Late Formative period, for there is early writing clearly in the Maya tradition at Kaminaljuyu by 200 b.c. The real problem then is that the Kaninaljuyu writing appears to employ a Cholan-Tzeltalan (Greater Tzeltalan) base language (Mora-Marín Reference Mora-Marín2005). This is some 600 years before Brown's model even proposes that Eastern Mayan had diverged to leave a Western Mayan proto-language, let alone the appearance of Greater Tzeltalan. Moreover, similar early glyphs are also now known from San Bartolo in the lowland Peten from about the same time period, or perhaps even earlier at 200–300 b.c. (Saturno et al. Reference Saturno, Stuart and Beltran2006). In order to accommodate these facts, Brown would have to argue that Mayan languages evolved from proto-Mayan to Greater Tzeltalan in a span of 200–300 years—quite an incredible rate of change. In sum, Brown's model simply does too much violence to everything that is currently known about Maya prehistory to accept. The linguistic data examined here make no sense in light of it either.

Overall then, the reconstructed linguistic evidence is most supportive of a ceramic bearing Terminal Archaic/Early Formative proto-Mayan population situated in relative proximity to the ocean followed by the fairly proximate subsequent development of specialized drinking technology and knowledge of “patio” architecture. This linguistic picture, in fact, bears remarkable similarity to that inferred from archaeological facts by investigators working with Early Formative assemblages in the Soconusco (Clark Reference Clark1997, Reference Clark, Herndon and Joyce2004; Clark and Blake Reference Clark, Blake, Brumfiel and Fox1994; Lesure Reference Lesure2011). The earliest Barra phase (1900–1700 b.c.) assemblages, the only known ceramic-bearing deposits at this early time in Mesoamerica, are mostly tecomates accompanied by a few bowls. A more elaborate range of vessels with ceremonial serving applications appear in the immediately subsequent Locona phase (1700–1500 b.c.). As described earlier in this article, elevated platforms that might be considered “patios,” and other forms of complex architecture, such as ballcourts and perhaps even rudimentary “plazas,” have also been documented in these same Locona phase contexts (Clark Reference Clark, Herndon and Joyce2004).

Could this assortment of coincidences be evidence that proto-Mayan speakers were among the Early Formative residents of the Soconusco? Mixe-Zoquean is not a particularly good candidate for a Late Archaic language of the region given that proto-Mixe-Zoquean is only thought to have dispersed at about 1500 b.c. or even later, with its origins on the Isthumus of Tehuantepec (Campbell and Kaufman Reference Campbell and Kaufman1976; Wichmann Reference Wichmann1995). Accumulating evidence linking Mixe-Zoquean with Totonacan continues to suggest an even more northerly highland and Gulf Coast starting point as well (Brown Reference Brown, Staller and Carrasco2010). Of course, one cannot rule out some other extinct or unreconstructed language group as responsible for Early Formative developments in the Soconusco. Yet, given the fairly robust agreement of archaeological and linguistic facts and chronology, it is tempting to pursue the possibility that proto-Mayan speakers were residents of the Soconusco, at around 1900–1800 b.c.

The broader implications of this “Over the Mountain” hypothesis would be that Core Mayan speakers eventually made their way off of the Soconusco, into the Guatemalan highlands, and through the Sierra Madre del Sur by the late Middle Formative period. There is, again, archaeological evidence for just such steady northward movement of population off of the Soconusco in the form of settlement densities, which rise sharply at the end of the Early Formative and dissipate abruptly during the mid-Middle Formative period. Rossenswig (Reference Rosenswig and Lesure2011) in particular has adeptly pointed out indications that populations of the Soconusco progressively shifted eastward and northward off the coast, contributing to the growth at inland sites like La Blanca, Takalik Abaj, and Izapa, in turn. These movements seem to have been accompanied by steady increases in social and political complexity and artistic development (Love and Guernsey Reference Love, Guernsey and Lesure2011; Rosenswig Reference Rosenswig, Burger and Rosenswig2012). The possibility that the populations involved were Mayan speaking, and that portions of them eventually spread out across the Peten and Yucatan in turn, remains to be fully explored.