Sunday, November 29, 2015


Brent Davis, Minoan Stone Vessels with Linear A Inscriptions. AEGAEUM, 36. Leuven; Liège: Peeters, 2014. Pp. xxiv, 421. ISBN 9789042930971. $142.00.

Reviewed by Thomas Palaima and Aren Wilson-Wright, University of Texas at Austin (;

Version at BMCR home site


In this ambitious, well-conceived monograph, Brent Davis examines the data for the inscribed stone vessels known as libation tables and ladles found on the island of Kythera and Crete from 1897 to 2000. Henceforth scholars taking different approaches to understanding this class of ritual objects and their inscribed texts will turn to Davis's comprehensive study for the primary evidence and for his ideas about how they relateto the religious beliefs and ritual practices of protohistoric Minoan Crete, its distinctive regions, its writing systems and the language(s) spoken by its inhabitants.

Davis makes it easy for readers to access the evidence for each object and text in order to evaluate his reasoning and interpretations. His catalogue and concordance section (Appendix A pp. 319-383) serve as a 'corpus' of this class of inscriptions. Relying on existing editions, Davis provides high-quality photographs of the objects and their inscriptions; drawings and transcriptions of the texts, both in normalized characters 1 and in conventional phonetic values assigned to Linear A signs that have plausible correlates in the Linear B script.2 Each object is described by: site, find spot, context date, manufacture date, current location and museum number(s), find date, percentage of object preserved, color, material, form of libation table (according to the Muhly and Warren systems, see pp. 67-71), dimensions of the object proper and its receptacle area(s), surface finish and decoration. A concordance of publication histories of the objects and their inscriptions (pp. 384-390) conveniently includes find dates, find spots, and museum numbers.

The rest of the volume is devoted to the meaning and use of these artifacts and their texts as standard instruments of religious practice. In chapter 2 ,(pp. 19-98), Davis carefully characterizes the sites, mainly extra-urban peak sanctuaries, that produced inscribed stone libation vessels and the sites that had uninscribed counterparts, often in sizable quantities. These extra-urban sanctuaries had much in common. Their "shared intervisibility with other sanctuaries"—with the exception of Kato Syme—created a network of common religious belief.

Davis provides computer-generated images of solstices and equinoxes as viewed from Iouktas, Petsophas, Vrysinas, Vouno on Kythera and Karphi (Psykhro cave).3 Yet he argues (Appendix C, pp. 401-419) that the sanctuaries, especially Petsophas and Iouktas (pp. 22 and 33), were situated not primarily for seasonal astronomical—Davis calls them 'calendrical'—observations, but for their intervisibility with related settlements (Roussolakkos and Knossos) and other sanctuaries.

Only nine of 39 peak sanctuaries survive into LM I. Of these, three have no apparent calendrical sightlines. What they have instead is large viewsheds. In Davis's view, the 'consolidation' of sites favored, although not exclusively, sites with calendrical functions, but was driven by what Watrous terms "regionalism rather than political hierarchy" (p. 411 and note 1639).

The sanctuaries are designed for open-air rituals and are able to 'see', i.e., look out upon, and be seen from the areas they served (p. 22). Sacrificial animals could be maintained in nearby mountain pastures. While proceeding through or looking upon these pastures, ceremonial participants could feel a close contact with nature. The larger sanctuaries include built terraces suitable for assembled crowds. Almost all these open-air and cave sanctuaries—again Kato Syme is an exception—use natural cavities or clefts in the rock, for dropping or wedging votives. The architecture of the sanctuaries that have formal structural elements demarcates sacred from non-sacred space and reinforces 'social and political hierarchies' among worshipers according to: a) degrees of access to ritual activities; b) available sight lines; and c) proximity of smell and hearing.

Particular tours de force are the five narrative scenarios that Davis presents (pp. 125-141) for the enactment of ritual events and deposition of the following inscribed objects—a carefully chosen representative sample: ladle IO(uktas) Za 1 ; libation tables PK (Palaikastro-Petsophas) Za 8 and SY (Kato Syme-Building U, Room 8) Za 2; tiny petaliform bowl IO Za 6; and PS (Psykhro) Za 2, a libation table with, unusually, three receptacles. Davis's reconstructions are exercises in vivid historical imagination, à la Emily Vermeule or Nanno Marinatos, that help us understand the social processes and the effects of movement, placement, light, smell, sound, sight and belief. Davis cites particular evidence for each facet of his reenactments and relies wisely on the insightful work of Yannis Hamilakis in understanding ritual practice as practice.4

In order to provide a broader context for interpreting the texts inscribed on these stone vessels, Davis discusses the possible functions of the objects themselves by looking at similar vessels in Hittite, Akkadian, Egyptian, Levantine, Israelite, Mesopotamian and Indo-Iranian mythologies and rituals (pp. 99-141). He suggests (pp. 99-107) that "non-draining receptacles on or near benches and altars" were used to hold poured libations that could be placed with food as meals for the gods. The Akkadian and Hittite terms for 'pour', i.e., 'make a libation', also mean 'sacrifice' in the literal sense of 'to make holy'. We should note that the related Greek vocabulary for pouring, libation and ritual killing (literally throat-'clefting') of animals does not shift into the semantic sphere of thereby 'making holy'. χέω 'pour', σπένδω 'pour a libation', and σφάζω 'cut, slit' are not used in place of θύω, which itself shifted semantically from 'make to smoke' (with incense or animal flesh) to literally 'sacri-ficio', or ἱερεύω 'acting as a hiereus to make hieros'.

The vessel in the outstretched hands of the male figures in procession on the serpentine relief rhyton fragment from Knossos is convincingly reinterpreted by Davis as a stone ladle (pp. 113-115) rather than a shallow bowl. The organic shape of these objects is suited for holding in cupped hands. Davis compares the ladle shaped as two cupped hands from Shaft Grave III at Mycenae (p. 117 and n. 578). The ladles cannot be used for scooping or receiving poured liquids without getting the hands of the ritual practitioner wet, so Davis proposes that they were intended for use with water (pp. 117-118) for ritual purification, even of the practitioner's hands, and for ritual pouring upon trees (date palms) common in Minoan and Sumerian iconography.

Within these well-defined contexts, Davis speculates (chapters 4 and 5, pp. 143-278) about the language(s) and language family of the Linear A texts on these vessels in relation to: a)other Aegean pre-alphabetic scripts (Cretan Hieroglyphic, Linear B, Cypro- Minoan and Cypriote Syllabic,and b) language groups in the 2nd-millennium eastern Mediterranean and Near East. Davis acknowledges (p. 157) that the whole corpus of Linear A still falls short of the minimal critical mass needed for decipherment, roughly 8100 signs—well short if we consider the peculiar brevity of entries on Linear A tablets, inscribed sealings and roundels. The inscriptions on libation tables offer hope because many have longer sequences than one or two sign-groups.

Careful analysis and hypothesizing is valuable. Davis's work stands alongside Duhoux's, Finkelberg's and Packard's as clear, plausible and rich in useful data.5 Still, as Alice Kober understood in her pre-decipherment research on Aegean scripts, every wrong assumption or error of fact introduces noise into the decipherment process.

Even positing that these texts are ritually dedicatory and formulaic and therefore can be broken down into canonical categories of information, the linguistic analyses of these texts by Yves Duhoux6 and his conclusion still hold true: "La plupart de ces interpretations ont cependant un gros défaut: elles sont trop hypothétiques, parce qu'invérifiables." Duhoux's own earlier analysis makes this clear. The whole Linear A corpus has approximately 800 'words' and of these a mere six may have 'roots' of four signs or more and 15 others have roots of three signs.7 This does not provide a firm basis for Davis's claim (p. 167) that the Minoan language taken as a whole "employed a significant number of prefixes." What he calls prefixes, Finkelberg (2001, p. 89) associates with "chains of introductory particles characteristic of Hittite-Luwian."

A major question is whether one or many languages were spoken in 2nd-millennium Crete. Davis ingeniously tracks ideograms, single-sign phonetic ideograms and single phonetic signs functioning as 'transaction signs' in Linear A texts at different Cretan sites. He concludes: "the fact that several of these abbreviations are attested at multiple sites suggests that the same words are attested everywhere" and therefore that perhaps a single language was serving island-wide "as a lingua franca on both administrative and ritual documents," whether or not many languages were spoken on the island (pp. 180-181). It might suggest no such thing, if we consider how Sumerograms and Akkadograms were used in Hittite documents to 'represent' underlying Hittite words. Analogically, in Linear B the ideogram NI (undoubtedly from Minoan nikuleon) was used as the ideogram for figs, but scribes would have written out the sign as sukon. Likewise, the ideogram for goat is identical to the peculiar phonogram *22 that has a Minoan-derived value /mbi/. But 'goat' in Linear B compound nouns is written ai-ki-.

In trying to narrow down the language family to which the language of the libation-table and other Linear A inscriptions belongs, Davis (p. 182) deduces from a group of inventories from Mari (palace of Zimri-Lim, ca. 1780-1760 BCE)— wherein "the 'Chief Merchant of the Caphtorians [Minoans]' required an interpreter while in Ugarit" —that "Minoan and Ugaritic were not mutually intelligible." But it is hard to say what the Mari evidence means for the linguistic identity of Minoan.

Ugaritic and Akkadian were both spoken in the Old Babylonian period, but were not mutually intelligible. So the Mari reference does not necessarily imply that Minoan was non-Semitic. It is also possible that the Caphtorians at Ugarit spoke Ugaritic and the interpreter was needed to translate between Ugaritic and the language of Zimri-Lim's agents. Zimri-Lim's archives are in the Old Babylonian dialect of Akkadian, but Zimri-Lim and his court may have spoken a different West Semitic language (many of their names are Amorite, a catch-all term for a group of West Semitic languages attested primarily in personal names). In either case, the language of Zimri-Lim's agents would not have been intelligible to speakers of Ugaritic. If this is the case, then the Mari reference does not have any implications for the linguistic identity of Minoan in the 18th century BCE.

Among many small corrections to be made, we give the following sample:

p. 123:the etymology of Demeter is not thoroughly Indo-European;
p. 145: the Linear A and B systems of numerals are stricto sensu not identical;
p. 147 figure 93: there is no neutral Linear B ideogram for ungendered person;
p. 148 figure 94: ideogram *129 does not stand generically for 'grain';
p. 148:ke-se-ni-wi-ja is not a woman's name;
p. 155: there is no cat's-head sign in the official CHIC Cretan Hieroglyphic signary.

Even Minos occasionally nods.


1.   As in Louis Godart and Jean-Pierre Olivier, Recueils des inscriptions en linéaire A volumes 4 and 5 (Paris: Libraire Orientaliste Paul Geuthner 1976 and 1982).
2.   For these "correspondances formelles," see Jacques Raison and Maurice Pope in Y. Duhoux, Études minoennes I (1978) 188 and Godart-Olivier (1982) xxii.
3.   See Mary Blomberg and Göran Henriksson, "Minoan Astronomy," in C.L.N. Ruggles ed., Handbook of Archaeoastronomy and Ethnoastronomy (New York: Springer 2015) 1432-1435. DOI 10.1007/978-1-4614-6141-8_141.
4.   Yannis Hamilakis in OJA 15 (1996) 1-32; World Archaeology 31:1 (1999) 38-54; and DAIS: The Aegean Feast, Aegaeum 29 (Liège and Austin: 2008) 3-18.
5.   Margalit Finkelberg, "The Language of Linear A: Greek, Semitic or Anatolian?" in JIES Monograph Series 38 (Washington 2001) 81-105, non citatum. David Packard, Minoan Linear A (Berkeley: University of California Press 1974).
6.   Yves Duhoux, "Le linéaire A: problèmes de déchiffrement," in Yves Duhoux, Thomas G. Palaima and John Bennet eds., Problems in Decipherment (Louvain-la-Neuve: Peeters 1989) 80-90.
7.   Yves Duhoux, Études minoennes I (Louvain: Peeters 1978) 96-109.


  1. A small remark on a small correction: "there is no cat's-head sign in the official CHIC Cretan Hieroglyphic signary." This is certainly true of CHIC, but the removal of the cat sign in its various guises is perhaps still best left an open question (e.g. Jasink 2009; Civitillo 2007).

  2. Thanks, Judith, as always. Vassilis Petrakis also pointed this out in a personal email. In a review of a monumental book that takes on so much so well--one really could have written four 3,000-word reviews of the book's main sections—every word counted. Those small 'corrections' Erin and I listed were a kind of tachygraphy.

    In this case, the 'official' CHIC syllabary is what it is right now with regard to the 'cat's head'. Our mentioning it is a two-headed point of criticism, as it were. Or a side swipe (as with a cat's paw) nudging us all to reconsider whether in fact the sign should be recognized. Indeed, raise it as and keep it for now an open question.