Into the Minds of Ancients: Advances in Maya Glyph Studies

Journal of World Prehistory, Vol. 14, No. 2, 2000 Into the Minds of Ancients: Advances in Maya Glyph Studies Stephen D. Houston 1 A decade of Maya glyphic decipherment creates many opportunities for historical, linguistic, cultural, and archaeological interpretation. New evidence points to improvements in understanding decipherment as a discipline and social practice, the origins of Maya script, the use and meaning of glyphs in ancient society, and the language and sociolinguistic implications of Maya texts. The glyphs reveal information about Maya kingship and its relation to supernatural forces along with cues to a synthesis of history during the Classic period (A.D. 250 850). A test case from Piedras Negras, Guatemala, relates such discoveries to the ongoing excavation of a Classic city with abundant inscriptions. KEY WORDS: Maya Lowlands; Classic period; decipherment; Maya archaeology. Nothing will ever be attempted, if all possible objections must be first overcome. Dr. Johnson, Rasselas, 1759 Some years ago, I contributed an essay to the Journal of World Prehistory that described the impact of hieroglyphic decipherment on Maya archaeology (Houston, 1989a). An invitation to contribute another essay on a half-period anniversary (half of the 20-year span beloved of the Maya) is highly welcome. It also invites self-scrutiny: What did I think then that I no longer believe or find useful and important? What is commonly held today that never remotely entered my mind in 1989? The objective of that earlier piece was to highlight contributions that Maya glyph studies might make to archaeology. This article focuses 1 Department of Anthropology, Brigham Young University, Provo, Utah 84602, USA. 121 0892-7537/00/0600-0121$18.00/0 2000 Plenum Publishing Corporation

122 Houston more directly on the rich epigraphic data, and relates them where possible to archaeology, but as a secondary aim. The assertion here is that the epigraphic evidence should first be understood on its own terms, before premature correlations are established with material patterns. The task is large, for the 1990s have been among the most important in the history of Maya glyph studies. Maya epigraphy made monumental advances throughout the 1990s. Texts can no longer be regarded as an incidental instrument in illuminating Classic Maya civilization (ca. A.D. 250 and 850, as found throughout the Yucatan peninsula and adjacent areas of Central America and Mexico). Rather, with decipherment, texts speak from the minds of ancients. Whether the glyphs are lying (the common libel, as we shall see) or whether they reflect only elite thoughts and preoccupations are almost beside the point. The inscriptions crystallize representations of ancient realities from a period in the New World when no other indigenous people left a clearly legible record. It is seldom stated so bluntly, for fear of offending esteemed colleagues, but neither Zapotec texts nor Isthmian writing yield information that can be regarded as completely or even partly deciphered (cf. Marcus, 1992a; Justeson and Kaufman, 1992, 1993). A good illustration is the Zapotec calendar, which remains controversial (Urcid, 1992). Attempts to interpret noncalendrical portions of Zapotec inscriptions demonstrate, despite intensive and insightful efforts, a disquieting inexactitude about appropriate word order, spelling principles, or linkages with Zapotec speech, the presumed but still unproved tongue of the texts (Marcus, 1992a; Masson and Orr, 1998; Urcid, 1992, 1993b; Urcid et al., 1994; Whittaker, 1992). The heavily logographic or word-based nature of this script continues to present obstacles to even the most determined researchers. For example, specialists have found no convincing instances of rebus, the cornerstone of many decipherments (Javier Urcid, personal communication, 2000). More recently, signal advances have been made in understanding other Mesoamerican scripts, to which Maya glyph experts can contribute their comparative expertise in the future (Berlo, 1989; Chinchilla Mazariegos, 1996; Taube, 2000b). What, then, did I (and others) believe in 1988, when the earlier review of Maya writing and archaeology was written? Bloodletting was thought to be a dominant theme of the texts. Polities of the Classic period seemed to be relatively small, with only glimmerings of higher-order hierarchies. Objects and buildings were known to be important in glyphic nomenclature, but details were still lacking. There was little attention to variability in the script. Grammar and syntax were assumed to involve split-ergative languages of varying affiliation, Yukatekan to Ch olan: Ch ol, spoken in the general area of Palenque, Mexico, seemed the most likely

Advances in Maya Glyph Studies 123 language for many of the inscriptions. 2 Historical linguistics was still relatively uninvolved with decipherment, not least because of controversies about readings of particular signs. Precise translations were generally unavailable, and epigraphers relied heavily on paraphrases and loose semantic equivalents in English or Spanish. Some influential monographs of the 1980s presented readings that diverge considerably from recent ones (e.g., Schele and Miller, 1983). Discussions of kingship and religious belief and practice tended to follow evidence from Palenque, where such matters were comparatively well understood. The underlying concepts of architecture and indigenous notions of existence were only beginning to attract attention, building on superb and deep ethnographic research by the Harvard project in Chiapas (Vogt, 1994). As of the year 2000, there is much that is new and different, although resting on solid foundations fashioned since the 1960s. This paper divides its subject into two major headings. The first discusses decipherment as a process and social practice, with the objective of explaining the nature of such issues, ethical and otherwise, to nonspecialists suspicious of epigraphic craft. The attempt here is to improve self-understanding among Mayanists namely, to forge what has been described as sociological literacy, which should, as in any healthy discipline, begin at home (Shapiro, 2000, p. 68). The second heading advances the author s views on improvements in glyphic knowledge since the 1980s. Among these changes are more developed understandings of the origins of Maya script, its role and use in Maya society, the language of the inscriptions and its bearing on vertical and horizontal social bonds (if such can ever be truly defined in the socially asymmetric universe of the Maya), the nature of rulers and their relation to divinity, history and politics as interpretive challenges, and the nature of the ancient political economy. A final section reports on an example of epigraphically informed excavations at a pivotal Maya city, Piedras Negras. The tone of the essay is meant to be direct and provocative, so as to rouse more focused discussion of the issues forced on all Mayanists by glyphic decipherment. 2 A note for the linguistically averse: Ergativity refers to a system of transitivity. In ergative systems, verbs attach prefixed (ergative) or suffixed (absolutive) pronouns for various purposes: the prefixed pronouns mark the subject of transitive verbs ( He scatters incense ) and the possessor of nouns ( his stela ); the suffixed pronouns serve as the subjects of intransitive verbs ( he dies ) and the objects of passive verbs ( he is captured ). Splitergativity shifts this system slightly. Prefixed pronouns now mark the subject of transitives and intransitives in what is called the incompletive aspect (i.e., the state of noncompleted action). In contrast, suffixed pronouns are only used for passive verbs in the incompletive. It is clear from comparative linguistics that split-ergativity came into existence after the breakup of earlier forms of Mayan languages (Robertson, 1992, pp. 53 54).

124 Houston DECIPHERMENT AS DISCIPLINE AND SOCIAL PRACTICE First I address Maya epigraphy as an academic field, which, for all its esoteric content, involves more engaged amateurs than professionals: handwringing about a lack of interaction with the public is moot in this branch of archaeology (Fagan, 1993). It also focuses on the sociology of the field and the difficult issues of disciplinary practice and moral obligation. Determined efforts are now being made to involve Maya epigraphy and archaeology in ethnic politics, with unclear results in the long term. Decipherment: Practices, Practitioners, and Products Since 1989, there have been great strides in decipherment, building on techniques pioneered decades before by, among others, Hermann Beyer, Heinrich Berlin, Tatiana Proskouriakoff, and the so-called Mesa Redonda group, especially Linda Schele, Peter Mathews, and Floyd Lounsbury, along with linguistically sophisticated scholars such as Victoria Bricker (1986), John Justeson (1978; Justeson and Campbell, 1997), and Barbara MacLeod (1987). We have come a long way from Gelb s statement that the Maya had no full phonetic writing system (Gelb, 1975, p. 82). Useful treatments of the history of decipherment may be found in Stuart (1992), Coe (1992, 1999), and Houston, Chinchilla Mazariegos, and Stuart (2000); a full-length, cross-referenced bibliography is also under preparation by Houston and Zachary Nelson (for early sources, see Cazes, 1976; for a limited, working bibliography, consult J. Harris, 1994; Bricker, 1995, provides a useful but conservative review of recent decipherments). Stuart s treatment presents unmatched scholarship on the beginnings of Maya epigraphy, and Coe offers a vigorous and entertaining narrative with a cast of heroes (Y. Knorosov) and villains (J. E. S. Thompson), and, after publication, with its own set of critics (Daniels, 1996, pp. 154 155; Hammond, 1993). There is little doubt that Coe correctly perceived the problems in Thompson s career, which, after initial, brilliant work that focused on calendrical and rebus spellings, limited glyphic decoding for at least two decades. To some extent this was attributable to Thompson s unparalleled reputation and force of personality, which brooked little disagreement. However, there was also the effect of his idiosyncratic understanding of writing systems and penchant for mystical exegesis an example of interpretative virtuosity gone sadly awry. Later works by Thompson left little sense of future direction or productive openings, although most epigraphers retain great respect for his erudition, particularly with respect to the Maya calendar. [On scanning Thompson s (1950) opus for the first time, the author can attest first-

Advances in Maya Glyph Studies 125 hand to feelings of futility and abject incomprehension. In contrast, Lounsbury s (1973) classic study of an important glyphic title showed with utmost clarity the reasoning behind a particular reading.] In turn, Knorosov, whose legacy of phonetic research in the 1950s stands untarnished, ended his career in sad denouement, beating a full retreat from his most trenchant conclusions, and with politicized and publicly voiced derision for epigraphic advances made by North Americans (Victoria Bricker, personal communication, 1995; Knorosov and Yershova, n.d.). The next step in epigraphic historiography is plain. We must go beyond review of published sources, the gist of most studies of decipherment, to unplumbed stores of personal correspondence, of which a central repository of letters, interviews, and unfinished manuscripts is urgently needed, perhaps on the model of personal papers of scientists preserved by the American Philosophical Society. David Lebrun of Nightfire Films, which is preparing a two-part documentary on decipherment based on Coe s book Breaking the Maya Code, has compiled dozens of hours of interviews with active epigraphers. The hallmark of current research is its strongly international flavor, with vibrant groups in Germany (Colas, 1998; Graña-Behrens et al., 1999; Grube, 1999; Grube and Nahm, 1990, 1994; Nahm, 1994, 1997; Voss, 1995; Voß and Eberl, 1999; Wagner, 1995), and smaller circles in Belgium (Lefort, 1998; Lefort and Wald, 1995), Canada (Zender, 1999), Denmark (Nielsen, 1998; Wichmann and Lacadena, 1999), France (Davoust, 1995, 1997), Holland (Boot, 1997; Schele et al., 1998), Russia (Beliaev, 1998), Japan (Yasugi, 1999a,b), along with an especially active focus in Spain (García Campillo 1992, 1995a, 1996, 1997, 1998a,b, 1999; García Campillo and Lacadena García-Gallo, 1988, 1989, 1990, 1992; Lacadena, 1992, 1994, 1995, 1997a,b, 1998). The limited job opportunities in Europe are a source of concern in the long term, although there is now a vibrant annual meeting that parallels self-consciously the growth of a pan-european identity (Colas et al., 2000). The European approach tends to be strongly philological, humanistic, and regionally oriented, without, for better or worse, the anthropological and theory-building stance in some North American work. The United States continues at a strong clip, with several recent dissertations and theses, of which the majority come from art history departments (Christy, 1995; Herring, 1999; Looper, 1995, 1999; MacLeod, 1990; Nelson, 1998; Newsome, 1991; Sanchez, 1997; Stuart, 1995; Vail, 1996; Villela, 1993). The recent passing of Linda Schele, a key figure in Maya research, will surely have an impact on epigraphy and its popular appeal in this country (England, 1998); by unhappy coincidence, Knorosov and Lounsbury died within a short time of their respected colleague (Grube, 1998a; Grube and Robb, 1999; Houston, 2000). Unfortunately, the future of epigraphy in North American departments of anthropology is murky because there seems to be compara-

126 Houston tively little institutional interest in supporting this research at the doctoral or professional level. Harvard, Tulane, the University of California at Riverside, and Brigham Young are notable exceptions to a trend that sees epigraphy as regionally important but intellectually narrow and insufficiently engaged in broader issues of social theory (see below). An example of this might be a recent statement that Mayanists and their funding sources are suspect because they largely address the literate ruling classes of buffer races, a memorable if inaccurate summation (Patterson, 1995, p. 116). From Mexico and Guatemala comes high-quality work from a few energetic researchers (Arrellano 1998; Ayala Falcón, 1995, 1997; Chinchilla Mazariegos and Escobedo Ayala, 1999; Escobedo, 1991, 1992, 1997a; Escobedo Ayala and Fahsen, 1995; Grube et al., 1992; Morales Guos, 1995; Valdés and Fahsen, 1995, 1998; Valdés et al., 1994, 1999), yet general or even academic interest in those countries seems, after a promising start, to be limited, despite the presence of expert epigraphic instruction at the Universidad Nacional Autónoma de México, the Universidad Autónoma de Yucatán, and the Universidad del Valle de Guatemala. A few basic texts on Maya glyphs exist in Spanish, but they are out of date, insufficiently disseminated, or prepared by nonspecialists (Alaniz Serrano, 1999; Berlin, 1977). The large number of glyph workshops in the United States and, increasingly, in Europe make the need for a textbook on ancient Maya writing somewhat less pressing. A challenge in preparing such a volume is our present difficulty in achieving consensus on the most basic of matters, including spelling rules and grammar (see below). Unfortunately, the few books on the market are now out of date or inaccurate (Harris and Stearns, 1997; Houston, 1989b). In the past, one could speak of schools of interpretation, such as that of Tübingen or Yale and Leningrad (Barthel, 1977; Houston et al., 2000). A sign of health in Maya epigraphy is a growing, collective awareness that some directions have proved more fruitful than others, and that scholars have now forged a common basis for discussion. In-bred schools can no longer exist in our era of internet communication and a populous community of independent-minded scholars. The new synthesis accepts syllabic glyphs of consonant vowel (CV) value (a sore point with some scholars until the 1970s), promotes a clear understanding of the relation of language to orthography, examines the biographies of historical actors, conceptualizes indigenous categories where possible, builds on a large number of signs with commonly accepted readings, and operates according to shared methods and principles of decipherment. Decipherment is held to be a full explanation for a sign in terms of its iconic referents, its meaning and sound, and its development through time. Of necessity, it must concern the minutia of signs and their behavior. Not all specialists agree on this

Advances in Maya Glyph Studies 127 point: one feels that such details can be bypassed in favor of a more anthropological framework pioneered by... Thompson (Marcus, 1992a, p. xix), as though attention to spellings and language represent an optional detour for students of ancient script. This is surely not the case, because precise knowledge of glyphs allows us to mine anthropological meaning. Peter Daniels distinguishes between primary decipherment, which proposes entirely new principles, and minor filling in (Daniels 1996, p. 142), as in the Maya case. Nothing could be less applicable to Maya glyphs. Every new decipherment of a logographic or syllabic sign is as vital and incrementally progressive as the last. Syllables have to be teased out one at a time, as do logographs, which depend on individual semantic and phonetic constraints for their elucidation. Thus, code breaking for Maya glyphs comes, not rapidly, but in fits and starts. For some signs there are insufficient contexts in which to test readings (the problem of hapax legomena, things said only once ), or they may spell opaque terms that seem not to have survived in daughter languages. This means that certain signs will not easily deliver their sound or sense. Complete decipherment is unlikely, given our tattered, if slowly expanding, epigraphic record (see below), and scholars must, for reasons of truth in advertising, curb triumphalist declarations about the state of glyphic readings. Aside from the celebrated alphabet of Bishop Diego de Landa (Coe, 1999, pp. 145 166), the key to decipherment has proved in large part to be internal, consisting of distributional patterns with apparent substitutable signs. The epigrapher can presume with a high margin of certainty that such substitutions constitute equivalent glyphs or, at least, signs in close, alternating relationship (Lounsbury, 1984, 1989); the most incisive demonstration of this approach was Stuart s detection of large numbers of substitutable u signs in a verbal expression meaning it happened, uti:y (1990b). A distinction between logographs (word signs) and syllabic glyphs leads to a more incisive understanding of patterns in such substitutions. Moreover, by inserting signs of known or suspected reading, phonic or semantic values can be tested repeatedly in different contexts. If there is a large set of documented texts, defective readings become rapidly untenable, whereas those with merit yield insights into hitherto unexplained sign clusters. Still a milestone and the best example of this approach is Stuart (1987), a study of enduring insight with interlocking conclusions that remain basically unmodified despite 15 years of subsequent research (cf. criticism in Hopkins, 1997). It is important to understand that broad consensus is not the same as total agreement. An appropriate model for understanding Maya epigraphy or any other rigorous discipline is the tension between core and frontier knowledge (Cole, 1992, pp. 229 230). The core comprises a

128 Houston body of ideas and methods that are held by all members of the epigraphic community; gains in the positivist sense of confident decipherment are impossible without these joint ideas. The frontier embraces those ideas still being probed and evaluated, with only a relatively small number ever likely to pass from this tentative state to the security of the core; the core itself may, however, undergo change as a result of some probing on the frontier. Some scholars seem to believe that epigraphers do not distinguish between the two, and that changing views reflect badly on the whole enterprise, being little more than a reflection of dubious practices, self-delusion, and reciprocal back-scratching (e.g., Baudez, 1999, p. 948; Porter, 1999). Such criticisms are illogical, for what discipline, other than a doctrinal one, is expected to validate itself by remaining the same? Maya epigraphy will always be interpretive on some levels, as we organize glyphic evidence and discern therefrom the society and underlying ideas that engendered it. It would be wrong to deny the existence of what Brannigan calls mediators, dominant figures who socialize younger scholars and determine the nature and importance of discoveries (Brannigan, 1981, p. 167). This essay is my attempt at mediation in the special, hegemonic sense of this term. However, certain readings are now set, regardless of personal whim and inclination, as reconstructed confidently by us and as intended explicitly by the Maya. Present-day decipherment, then, has passed its physical. The representational context is becoming well known (Miller, 1999). The syllabic grid has been partly filled in by intensive work that peaked in the late 1980s and early 1990s (e.g., Grube and Nahm, 1991), although certain glottalized consonants and Ce syllables are still unattested or insecurely identified. Common logographs continue to reveal their secrets, albeit at slower pace than before: A recent advance is David Stuart s important proposal of the ch e:n or cave, rocky outcrop glyph (Vogt and Stuart, 1999), now shown to express a ubiquitous, indigenous concept of place (Stuart and Houston, 1994, Fig. 9; see also Boot, 1999) and the probable locus of pilgrimages attested hieroglyphically (Stone, 1995). Spelling rules are coming into clearer focus (Justeson, 1989b), although prompting further controversy. For example, Bricker (1989) suggests that orthographic irregularities of Yukatek Maya written in Roman script reveal hieroglyphic antecedents. The difficulty with this intriguing hypothesis is that few of the proposed patterns accord with present understandings of glyphic orthography: Colonial ahau uob and mul lil, thought to be reflections of logosyllabic rules for word division (Bricker, 1989, p. 41), correspond to no known pattern in the hieroglyphs, which would have spelled such words AJAW-O:B, mu lu IL, MUL-IL, and so forth. Nonetheless, the main and fully valid point is that hieroglyphic and European literacies could overlap in the same person. This was undoubtedly so in early Colonial Yucatan. More

Advances in Maya Glyph Studies 129 debatable is the proposition that Colonial documents represent direct translations of hieroglyphic originals (Tedlock, 1992, pp. 236 237, 1996, pp. 28 29). This idea has a strong allure to those who wish to see particular documents as unadulterated, pre-columbian productions that have shifted into a new code, contents intact (see Miram and Bricker, 1996, for effective counterarguments). Another proposal concerns the syllabic hypothesis itself. Long ago, Knorosov demonstrated the presence of synharmony, in which appended syllables contain the same vowel as the syllable or logograph they modify. What he could not explain, however, were instances of disharmony, in which the vowels differ. Houston et al. (1998) make a case that disharmony was a device signaling complex vowels (long, glottalized, or accompanied by h) in preceding logographs or in words spelled by two or more syllabic signs. There are suggestions that these rules can be further refined and that certain instances of h may not have been spelled by Maya scribes (Alfonso Lacadena and Søren Wichmann, personal communication, 1999). These proposals are still being debated and require sophisticated explorations in historical linguistics before they can be resolved. Some issues may never be resolved because of gaps in our data. One development is Houston, Stuart, and Robertson s (2000b) proposal of a third category of sign, the morphosyllable, namely, a set of postfixes of -VC form that convey both sound and morphological meaning; their exact pronunciation or surface expression remains in doubt, but they seem to trigger vowel harmony or vowel variation with fixed consonants, rather like most signs in Egyptian or Semitic scripts. The vocalic flexibility in these morphosyllables may have accommodated Mayan dialects that handled vowels differently, a feature also used to explain the vocalic imprecision in Egyptian script (Parkinson, 1999, p. 52). The presence of morphosyllables implies a rigorous grammarian tradition among the ancient Maya, perhaps along the lines of the morphophonemic classes compiled by Pāṅini in his renowned grammar of Sanskrit (Bright, 1990, p. 143). A crucial methodological advance has been the palaeographic study of sign use through time and space, although such studies are still incomplete (Lacadena, 1995, 1999; see also Satterthwaite, 1938). It has become apparent that such research will lead to refined understandings of grammatical elements and their spread (Hruby and Child, 1999) and the politically conditioned nature of sign use (Brewer, 1996). In contrast, palaeographic attempts to revise the correlation between European and Mayan calendars presumed, continuous rates of stylistic change supposedly point to an abbreviated Postclassic period have generated disagreement (Lacadena, 1995, pp. 413 424) because such changes are likely to involve many variables. The value of the palaeographic approach is methodological in that it helps to combat what might be called the

130 Houston synoptic fallacy in studies of Maya script the presumption that we are dealing with a single, frozen script that fails to change over space and time (see below). Historical, linguistic, and cultural conclusions are also described below, but suffice it to say that the dynastic reconstructions that characterized much of the 1970s and 1980s (e.g., Houston, 1993; Mathews and Willey, 1991) have been revitalized by masterful historical syntheses from Martin and Grube (1995, 1999) and close and effective correlations between epigraphic and archaeological finds in large-scale archaeological projects (e.g., Demarest et al., 1997; Escobedo, 1997a; Sharer, 1999). Work on the four Mayan codices or screenfold books builds on a long and distinguished foundation in glyphic studies, going back to the early years of decipherment. For many years, Maya epigraphy was the study of these books. Notable contributions throughout the 1990s include Love s (1994) treatise and facsimile edition of the Paris codex, a commentary on the Dresden (Davoust, 1997), a series of edited papers on the Madrid Codex (Bricker and Vail, 1997; see also 1994, and Lacadena, 1997a), studies of evident Ch olan and Yukatek bilingualism in the Dresden and Madrid codices (Lacadena, 1997a; Wald, 1994a), and a plethora of articles on selected topics (Bricker and Bill, 1994), many archaeoastronomical [e.g., Justeson, 1989a; Aveni, 1992; Bricker and Bricker, 1992, 1997; Love, 1995; see also Malmström s (1997) attempt to link the calendar with an origin in specific sites]. This work, although addressing Postclassic materials, has proved immensely useful in refining more ancient Maya concepts of ethnoastronomy (Tedlock, 1999). It suggests new ways of understanding the layout of buildings in Classic Maya cities. Architects at Yaxchilan, for example, appear to have attended to solstitial alignments (Tate, 1992, pp. 111 115). An influential volume on creation and cosmology as documented in Classic Maya art and writing rests in part on zodiacal information in the Paris Codex (Freidel et al., 1993, Fig. 2.33). Nonetheless, the opacity of many texts in the codices, both in their reading and deeper, inferential meaning, makes this work at once challenging and incomplete. A focus on numbers and their properties cannot replace closer study of hieroglyphic captions and glosses (Houston, 1997c). Related to Maya decipherment is the recent work on Isthmian writing. It must be addressed for two reasons: First, because Isthmian has been regarded by some as a precursor to Maya script (see below); and second, because the decipherment has been described as the most rigorous and compelling work ever done on a New World writing system (Kelley, 1993, p. 29). However, published reports raise concerns about the proposed decipherment (Justeson and Kaufman, 1993, p. 1711; see also Justeson and Kaufman, 1992). The decipherment rests on a number of premises: that crucial signs are iconically transparent ( hide, skin, boss, lord, in-

Advances in Maya Glyph Studies 131 cense, water, moon, liquid,drops, sky ) andthus relatableto words in the target language for decipherment ( pre-proto-sokean ); that concepts and signs from Maya writing also apply to this script ( accede, let blood, pierce ); that quotative expressions and other aspects of speech should occur in a monumental text, often with a puzzling absence of references to the agent; and that the resulting readings are inherently complete, coherent, and grammatical (Justeson and Kaufman, 1997, p. 210). Whether this is so can be judged against a recent translation: Behold, there/he was for 12 years a (title). And then a garment got folded... What I chopped is a planting and a good harvest. (A) shape-shifter(s) appeared divinely in his body (Justeson and Kaufman, 1997, p. 210). It is fair to say from this translation and the nonsequiturs it contains that there are at least some grounds for doubt that progress has been made. Moreover, such statements are unlike others from monuments in Mesoamerica. The Maya themes introduced in the decipherment are arresting, for they appear to reflect readings and emphases of the mid- to late-1980s, when the authors were more intensively involved in Maya epigraphy. (For example, bloodletting signs are now thought to be rare in the inscriptions; a glyph for shapeshifter was deciphered in the late 1980s but appears infrequently in monumental texts, except as a root for an architectural term, Houston and Stuart, 1989.) A few years ago, Michael Coe called into question an attempted decipherment of Indus Valley script, noting the absence in that writing system of a large database, bilingual texts, well-understood cultural context, and pictorial clues from images accompanying texts (Coe, 1995, p. 393). None of these are present in the Indus Valley, nor, unfortunately, do they characterize the archaeology of the Isthmian zone. Not even the local artistic style the iconographic crucible from which script emerged is abundantly documented or well explained. It is all the more striking that Justeson and Kaufman have produced 21 syllables and dozens of logographs from a total of 10 inscriptions, of which only 2 are long or clear enough to be readily usable. The Maya corpus is vastly larger as a testing ground for readings. A reasonable estimate of the total number of texts, including those on fragmented pottery from excavations, would exceed 15,000. The Maya corpus can be augmented by three means: additional fieldwork and excavation, devising tools for accessing data, and new imaging techniques. An important epigraphic find is marked by three things: the preservation, length, and content of the text. The longer the text, the more likely that it will contain new information. Throughout the 1990s major finds have been made at Palenque, in stuccos and a lengthy throne text found in Temple XIX (Stuart, 2000b); finds from concentrated parts of the small site of Pomona, Tabasco; a throne, panel, and hieroglyphic stairway from Dos

132 Houston Pilas (Palka et al., 1991; Houston et al., 1991a,b); Tikal Stela 40, a long but slightly eroded text from the latter half of the Early Classic period (Valdés and Fahsen, 1998); a variety of Terminal Classic altars from Caracol (Chase et al., 1991); fragmented but historically valuable stuccos and miscellaneous texts from the same site (Grube, 1994b); a new stela from Nimli Punit with an unambiguous, local Emblem glyph (see also Wanyerka, 1996, and Grube, 1994a); full publication of the painted (and now partly destroyed) glyphs from the Naj Tunich ritual cave in Guatemala (Stone, 1995); and a trove of stelae, monuments, and stuccoed texts from Tonina, Chiapas, Mexico, that are both pivotal to decipherment and, to judge from available evidence, inadequately documented (Yadeun, 1993). The Tonina work receives a medal of lead or tin; the gold must go to the sustained and meticulous excavations at Copan, which have systematically recovered, reassembled, and recorded texts, many now placed in the sculpture museum at the site (Fash and Fash, 1996). The monuments at Copan, so long and recondite in content, sparked major advances in understanding the names and titles of monuments and the rituals used in dedicating them. A smattering of new texts from La Corona have indicated the probable, original location of unprovenanced monuments from the northern Peten (Graham, 1997; Mayer, 1995, pl. 161; Schuster, 1997b). Challenging but informative inscriptions from Calakmul have augmented in small measure the large but lamentably eroded corpus of this, debatably the largest Maya city of the Classic period (Martin, 1997, 1998; Pincemin et al., 1998). A stunning altar in pristine condition has been excavated at the site of El Cayo, Chiapas, and has just as quickly become mired in the chaotic and ugly politics of that zone (Schuster, 1997a). Nearby, probably from Guatemalan territory, comes a large set of looted lintels showing the subsidiary figures and their overlord from Yaxchilan; they have been published in part, but most still seem to be languishing in a bank vault in Europe (Schele, 1990). More such sculptures are likely to appear as settlement burgeons in the upper Usumacinta drainage. The discovery in 1999 of a new monument at the Guatemalan border village of La Técnica suggests that many more may yet be found. Even well-known sites such as Piedras Negras have yielded new texts, as in the recent excavation of Panel 15 in the stairway debris of Pyramid J-4: this inscription contains more than 150 glyphs and refers largely to battles in the reign of Ruler 2 of that city. Another panel, of supreme important to the history of the Pasión region, has been looted from the area of Cancuen (Mayer 1995, pp. 167 169). Nearby, Aguateca has delivered, as of 1999, several new stelae of unusually early and late date for that city (Takeshi Inomata, personal communication, 1999). From the Yucatan there are new texts, of great linguistic interest and replete with secure Emblem Glyph, from the city of Ek Balam (Vargas de

Advances in Maya Glyph Studies 133 la Peña et al., 1998; Vargas de la Peña and Castillo Borges, 1999; Voss and Eberl, 1999); in February 2000, the excavator of the site, Vargas de la Peña, reported the discovery of a royal burial with pottery texts, along with a buried mural containing approximately 160 glyphs (Alfonso Lacadena, personal communication, 2000). The Spanish project at Oxkintok uncovered important new texts from this Puuc site and clinched its connection to the so-called Chocholá style of carved vessels, as have discoveries of new ceramics at the ruins in 1998 (García Campillo, 1992; Grube, 1990c; Alfonso Lacadena, personal communication, 2000); other inscriptions of that zone are finding their way to scholarly notice (García Campillo, 1995a, 1998a). In the meantime, Justin Kerr continues to photograph pots passing through his studio in New York City, organizing them in a relational database that can be consulted freely (www.famsi.org). It is impossible to predict the precise location of future finds, but we can be sure that they will arise from three general contexts: large-scale excavations in site cores; illicit plunderings of unprotected sites, either in remote areas or those in close proximity to new settlements; and the discovery of new sites with monuments on the surface. Financial circumstances will make large excavations rarer, and diminishing areas without settlement will gradually do away with unexplored zones, leaving the second the more likely, and lamentable, source for new texts. At the Corpus of Maya Hieroglyphic Inscriptions Project, housed at the Peabody Museum, Harvard University, Ian Graham, David Stuart, and their former colleagues, Peter Mathews and Eric von Euw, have worked indefatigably to photograph and draw as many monuments as possible to the highest standard. Looting lends urgency to their efforts. Plans are afoot to increase computer access to these images, although this is made more difficult by limited funds and staff time. So far, 15 definitive folios have appeared. Most epigraphers follow Corpus drafting techniques, not least because line drawings make clean photocopies a great boon to epigraphers doing structural studies of texts. The technique is even more useful because it requires no exceptional artistic skill and can be used by most draftsmen with a firm hand. The impact of the Corpus on decipherment can hardly be overstated, nor can Graham s generosity in distributing drawings well before their release in Corpus volumes (e.g., Tate, 1992, p. 194, Fig. 89). One Corpus project in preparation targets a Maya glyphic dictionary that is lexeme- rather than glyph-based; until recently, such a project would have been inconceivable, yet recent decipherments have made it both doable and necessary (Stuart and Houston, 2000). A federally funded project by Martha Macri at the University of California, Davis, is about to release an exhaustive relational database inventory of keyed texts and glyphic

134 Houston images (www.cougar.ucdavis.edu/nalc/saa.html), although, in all candor, the proposed indexing of Maya glyphs is unlikely to replace Thompson s compilation of signs (1962). There has also been an effort to compile glyph readings in a single source (Kurbjuhn, 1989), a handy effort because it helps establish authorship for certain readings. Nonetheless, it is safe to say that Maya epigraphy and its tools of access are not yet near the quality or refinement of those for Classical studies (Gordon, 1983; Keppie, 1991; Tracy, 1990; Woodhead, 1981). In comparison, Maya epigraphy is still immature: We need the indexes and prosopographies that Classicists take for granted (e.g., Jones et al., 1971). There have been some improvements. Merle Greene Robertson s rubbings are now available on CD-ROM from the University of Oklahoma Press (although expensively), and Karl Herbert Mayer (1995) has produced a useful series of volumes recording monuments with or without provenance (although opaquely organized). Mayer in particular has been assiduous in tracking down nearly inaccessible sculptures and making them available to scholars; he has led and sponsored fieldwork under the auspices of the journal Mexicon, a valuable source of fresh information on Maya glyphs. American journals such as Latin American Antiquity have tended to report on calendrical and codical research because of their firm restrictions on the use of unprovenanced materials, a category to which, regrettably, many texts belong (Anonymous, 1992, p. 261; but see Palka, 1996, Fig. 2). Another way of extracting glyphic evidence is through new imaging technologies of painted texts. The Bonampak Documentation Project and engineers at Brigham Young University (BYU) have pioneered field use of high-resolution digital cameras, infrared videography and image capture on Super-VHS tapes, computer enhancement of the resulting images, and multispectral roll-outs of Maya vessels (Kamal et al., 1999; Miller, 1997; Miller and Houston, 1998). At Bonampak, Mary Miller has led a team that has looked comprehensively at documenting and interpreting this vastly informative set of Pre-Columbian paintings (Miller, 1986). Many new texts have been detected, included minutely sized hieroglyphs and crisp images of glyphs that had been murky: within minutes of setting up videographic equipment, glyphs spelling quantities of cacao beads were found on tribute bundles under a painted throne in Room 1 of the murals building (Houston, 1997a; Ware et al., 1996). To date, the murals project has retrieved one of the largest sets of names known from any Pre-Columbian site and determined the dynastic relations between figures in the paintings. Enigmatic microglyphs, no more than 1 2 cm in height, embellish parts of the painting (Fig. 1): only close-up viewing would detect such signs, hinting that the murals were not meant to be seen from one vantage point, but from many. The same team from BYU has also worked with James Brady in the painted

Advances in Maya Glyph Studies 135 Fig. 1. Microglyphs from the Bonampak Murals, Room 3 (courtesy Mary Miller, Bonampak Documentation Project). caves of Naj Tunich and Cueva de las Pinturas (Brady et al., 1997; Ware and Brady, 1999), and improvements in equipment have led to markedly superior images. Carbon underpainting, the usual outline of glyphs, come out particularly well, but digital imaging is also discriminating between slight shifts in red. Clearly, painted texts can no longer be viewed solely by the naked eye, restricted as it is by inherent spectral limitations (e.g., Chase and Chase, 1987, Figs. 10 and 37). The technique should be extended to all painted surfaces with hieroglyphs, such as the capstones and murals from the Yucatan, Peninsula (Fettweiss-Vienot, 1981, 1984; Mayer, 1987, 1990) and the unparalleled richness and iconographic diversity of tomb paintings from Río Azul, Guatemala (Adams, 1999, Figs. 3-14, 3-17, and 3-44). The addition of generous funding from a new charity, the Foundation for the Advancement of Mesoamerican Studies, Inc. (FAMSI), has begun to have a profound impact on the direction and practice of Maya archaeology and epigraphy because of its emphasis on ancient images and writing (www.famsi.org); to an extent unacknowledged by many, research follows funding. For the moment, the large amounts injected into Maya research will play foil to the scientific and processual style of research associated

136 Houston with funding from the National Science Foundation. FAMSI has also been instrumental in providing a digital platform for a joint bibliographic effort with the library of the University Museum at the University of Pennsylvania. Ethics, Standards, and the Public Since the early 1980s, before her untimely death, Linda Schele organized and promoted an annual series of conferences and workshops at the University of Texas, Austin (e.g., Schele and Mathews, 1993; Schele and Grube, 1994). These have had an enormous impact on the general public and on professionals, who could glean the latest decipherments and interpretations within months (or days) of their formulation. Her own presentations were vivacious and at times non-academic, yet highly effective in communicating the excitement of decipherment. Through Schele, the subject of Maya decipherment, inherently an esoteric discipline, achieved broad and inclusive appeal, with the implicit promise that amateurs too could make breakthroughs, just as Schele herself did soon after entering the field in the early 1970s (Coe, 1999, pp. 201 203). Understanding the general attraction of the Maya comes easily. The visually stunning nature of Maya glyphs and their architectural setting, along with the glamour of modern decipherment, contribute to popular interest through books, television specials, and magazine articles (e.g., Lemonick, 1993; G. Stuart, 1997). More to the point, Maya ruins lie relatively close to the United States. Tourists can travel inexpensively and in tolerable comfort to savor the exotic and mysterious in photogenic settings. However, in reaching out to the public (Fagan, 1993), we must be wary of sensationalizing the otherness of the Maya, accentuating the peculiarity of their glyphically attested rituals, or most crucial, glossing over interpretive difficulties or disagreements. Public interest can even lead to odd situations, so that, for example, scholars may find themselves conducting simultaneous popular and academic discussions on the same topic. One of my articles (1986) was prompted by seeing an Emblem Glyph I identified but had not yet published in a map from a popular magazine. We have many responsibilities to a public that craves accurate information; to colleagues that deserve a grace period to publish before their readings filter into common knowledge (see below); and to the ancient Maya themselves, through portraits and narratives that show them comprehensively and contextually, to the best of current abilities. This raises the matter of ethics and standards. Accountability should be our byword to fellow specialists, the modern public, the evidence, and the texts as artifacts. In many cases, later viewers will not see the monuments

Advances in Maya Glyph Studies 137 in the same state of preservation as we do (Bell, 1987, p. 43). The texts deserve the greatest attention. In recording and studying them, specialists are morally enjoined to be accurate, to operate as physicians who, above all, do no harm, and to disseminate their records as rapidly as possible. Accuracy is necessary because monuments may be destroyed, stolen, or effaced by forces of nature after the record is made, as has been the case with an important altar recently found at Zacpeten, Guatemala (Don Rice, personal communication, 1997). Incomplete or inaccurate documentation actively prevents others from understanding the text. This means that inscriptions must be photographed and drawn according to the highest standards (Jones and Satterthwaite, 1982; Graham, 1975), not by an artist, but by a trained epigrapher who can identify relevant information, and in such a fashion as to anticipate epigraphic needs of the future (Bell, 1987, p. 44). Unfortunately, drafting is a difficult and time-consuming skill that is rarely taught formally in archaeological programs. It will soon undergo rapid shifts as computer drafting and digital imaging grow to dominate epigraphy. The time spent on renderings is rarely valued as a legitimate use of professorial energy, and many if not most professional archaeologists leave their drafting needs to specialists. A final aspect of accuracy is that monuments should not be moved before adequate account is taken of their archaeological context. Above all, an inscription is an artifact as well as a text (Houston, 1993, p. 69). Some accounts of new finds have, despite studio-lit photography, proved unsatisfactory as archaeological records (Yadeun, 1993), with little sign that further information is to come, as in the case of finds at Yaxchilan (García Moll, 1996, p. 39). This becomes a highly sensitive issue when North Americans, the unwilling heirs of problematic political relations with Latin America, criticize archaeological work by neighbors to the south. Quite simply, the matter cannot be discussed without triggering historical antagonisms, nor is prior work by North American archaeologists without flaws or ethical lapses (Black, 1990). Justifiably, some Mexican archaeologists have questioned the use of archaeology in their country for professional advancement in another (Lorenzo, 1981). Tact and more contact, not less, is the solution to resolving such tensions because all share similar goals of elucidating inscriptions, as facilitated by talents and resources from many countries. The second ethical duty is to do no damage. It is grossly negligent to record a text at the expense of its long-term viability, through abrasive rubbings, molds, and other methods of direct reproduction, or to fail in preserving the text after discovery or documentation. Panel 7 at Piedras Negras, Guatemala, has a coating of indelible epoxy left by a botched attempt to cast the sculpture. To be sure, direct reproduction has its advo-

138 Houston cates (Gordon, 1983, pp. 30 31; Keshishian, 1988, p. 29; Woodhead, 1981, pp. 78 83, 135 136; see also Greene, 1966) because it avoids the intercessions, interpretations, and errors of draftsmen. It has also served a vital role in preserving information through, for example, the casts made by Alfred P. Maudslay in the late 19th century, sometimes of monuments that have subsequently eroded (Joyce, 1938). Nonetheless, preserving a text after discovery and recording can be surprisingly difficult to accomplish, and a good case can be made that most monuments, particularly stuccoed or painted ones, should be reinterred or removed from contact with the public and the elements (Valdés, 1993, p. 96). Every country with Maya remains must walk a difficult path as populations increase and open communication erupts in complaints about historical inequities. For example, in the aftermath of the Guatemalan peace accords, local villagers have begun to lay claim to archaeological sites and monuments in ways that countermand national law. The results, especially in the Pasión drainage of Guatemala, have been calamitous, as villagers hack monuments into pieces for sale and, in aggressive assertions of local sovereignty, prevent government authorities from patrolling sites or protecting inscriptions. Political realities make any countermoves impractical at present. In this light, Ian Graham s intrepid efforts at removing sculptures from danger deserve the highest commendation, as have subsequent efforts, including mold-making, by the Institute of Anthropology and History in Guatemala. The creation of the new sculpture museums at Tikal and Copan (Fash and Fash, 1996) is probably the only solution to protecting important sculptures from long-term damage. Yet, these buildings too will require sturdy construction, especially in earthquake zones, along with government commitments to continued maintenance and museum improvement. As ethnopolitics intrude (see below), the removal of sacred objects from their original settings may prove increasingly controversial. The final injunction, rapidity of dissemination, poses a joint responsibility. As epigraphers, we are mere stewards of information, with no rights to hoard evidence (Chippindale and Pendergast, 1995, pp. 45 46). The socalled Copan Notes and Texas Notes, a series of photocopied manuscripts distributed by the late Linda Schele, represented the most rapid if uneven communications in Maya epigraphy; more finely crafted reports have come from the editorial hand of George Stuart and his Center for Maya Research in a series that is beginning again after a hiatus of some 8 years (e.g., Hammond et al., 1999). By the same token, those receiving unpublished drawings or photographs must be sensitive to rights of first publication and to legal sensitivities resulting from government permits in countries of origin. The growing problems of gray literature, government reports filed in limited-access archives, and of nonacademic, restoration-oriented