The widespread reignited interest recently witnessed in the studies on Bronze Age Aegean scripts, especially the undeciphered writing systems (Cretan Hieroglyphic, Linear A, Cypro-Minoan), has led to the publication of multiple contributions on several aspects of the issue. In particular, for its relation with Linear B (deciphered in 1952 and recording the earliest form of Greek language known to us), Linear A has been a primary focus of interest. Our understanding of Linear A and Linear B as related scripts is based on the considerable graphic similarity between the two syllabaries. However, it is worth highlighting that a shared writing system does not necessarily imply a linguistic connection, as widely shown by the use of the Latin alphabet to write a large number of different, and often unrelated, languages. Thus far, the language that Linear A encodes has been tentatively interpreted as Greek, Etruscan, or Akkadian, among others. Most of these undertakings, however, have often been deemed as based on pseudo-scientific methodologies, spanning from the examination of just a few documents producing at first glance a (more or less) plausible interpretation, to what seems to be sheer imagination.
In this A5-size volume, in which each part is written in both Italian and English, the authors claim to have deciphered Linear A on the basis of "a research work performed in a deliberately and methodically isolate way" (p. 11), as highlighted in the brief Introduction (pp. 11–12), and that it is Greek "beyond any denial" (p. 18). They also inform the reader of the lack of any comparison with Linear B, because of the uncertainties they think still exist in the decipherment of this script. However, the nature of these uncertainties is neither outlined, nor listed, nor briefly mentioned, nor referred to in any of the bibliographical references.
Consisting of three and five pages respectively (p. 17–19; 27–31), Chapter 1 "The Greek before the Greek... again: The Minoic (sic) language as a "'Greek dialect"'" and Chapter 2 "General features of Minoan Writing: Elements for the deciphering of Linear A" cursorily hint at the supposed relationship between Linear A and Greek. Such a comparison is not carried out on the basis of the whole corpus of Linear A inscriptions but rather on the basis of "the near totality" (p. 27; italics are mine). Unfortunately, the authors do not specify what of the corpus they are basing their conclusions on. Chapter 3 "Deciphering: Translation and analysis of Linear A inscriptions", deals with only a few Linear A documents, and again, the criteria for their selection are not clarified. Some bibliographical references are listed at the end of the volume. However, they are hardly exhaustive; the most recent work dates to 2012 and some fundamental references are rather noticeably missing, including several works by Bennet, Davis, Duhoux, Godart, Hallager, Hooker, Packard, Palaima, and Schoep.
From a linguistic perspective, two main issues ought to be pointed out. First, where an etymology is reconstructed, it is not indicated by means of the commonly accepted conventions, that is to say an asterisk followed by the lexical root written in Latin italics, e.g., *sem-. Here, the root is rather expressed by means of the mathematic sign for the square root (√ἀρ, p. 72). Second, Linear A is seen here as the direct ancestor of the Doric dialect, a statement for which two arguments are provided: (i) the Homeric verses (Od. 19, 172–174) describing Crete as a place with "countless men, without end, with ninety cities and mixed languages" (it is not specified whether or not the translation is by the authors), and (ii) "a strong presence of a, like in the Doric dialect spoken in Crete in the subsequent centuries" (p. 30). As regards this latter point, phonetic and chronological observations are worth being highlighted. What is here called a "Doric" feature is actually an Indo-European element. The Indo-European *ā has been preserved as such in some first-millennium dialects (e.g., Dorian: mātēr < *mātēr), whereas in others it has undergone a change *ā > ē (e.g., Ionic and Attic: mētēr < *mātēr), an innovation commonly understood to have taken place in the first millennium BC. Given that Linear A inscriptions date to the second millennium BC (ca. 1800–1450 BC) and at this point in time an "a" would be expected in an Indo-European derived language, the presence of an "a" does not prove its relation to one dialect or another. Whatever the language behind Linear A might be, any tentative interpretation of its phonetic elements must take into account the chronological gap with the following documentation as well as rely on the earliest reconstructable stages of any language it is compared to.
As regards the methodology, even though presented as "original" (p. 12) and "radically innovative" (back cover), the authors either simply attribute phonetic values based on Linear B to Linear A signs, a method that was tried soon after the decipherment of Linear B (in 1956) 1 and is commonly understood to be problematic, or they employ deviations from this pattern without providing any explanation. Likewise, the supposed strong connection linking the inscription and the object on which it appears (p. 12) is hard to call original, given the attention paid in the last few years to the medium as artefact. It is also unusual to look for this kind of relation in objects like tablets, commonly understood to record administrative transactions.
Dealing with an undeciphered script, it is little wonder that the attribution of any phonetic value to Linear A signs is among the most challenging questions to answer. For this reason, it is essential carefully to elucidate the reasons behind any phonetic attribution. Fruitful results have been obtained by a comparison between anthroponyms and toponyms in Linear A and Linear B, which led to establishing that both scripts seem to share the phonetic value of (at least) 16 signs. Further comparisons based on phonetic or declensional alternatives have then led to the identification of an additional group of another 14 signs that seem to have common phonetic values in Linear A and Linear B. Instead, in this book assumptions are made and no explanations are provided. For example, let us take the sign AB 31, palaeographically common to both Linear A and Linear B and phonetically corresponding to sa in the latter script. In this volume, AB 31 is read as la and the authors not only fail to provide any clarification of this supposed identification but they also use it to further speculate on words bearing this sign. As a result, the Linear A sequence 59-31-17 (commonly transcribed, for ease of reference, according to the phonetic values of Linear B, that is to say as TA-SA-ZA) is here read as thalassa ‘sea’ and no explanation is provided for SA = la nor ZA = sa (p. 24). Moreover, AB 04 and AB 54, which belong to the groups of signs whose phonetic value seems to be common to both scripts, respectively correspond to te and wa in Linear B. In this book, though, the former is read as ZE on the inscription KY Za 2 (p. 43) but as TA on the document HT 87 (p. 95; neither of these two attributions is accompanied by any explanation), while AB 54 is attributed the phonetic value of ga. The correspondence wa = ga is vaguely mentioned as trustworthy lexica. However, neither is it specified in which lexicon this equivalence w = g is to be found, nor are examples provided that appear in any Greek text. For example, the authors cite the supposed correspondence between ϝοινέες and γοινέες, but the term ϝοινέες is unattested as such. All the interpretations of the Linear A inscriptions mentioned in this volume are based on these and similar criteria.
Finally, images are given with a caption indicating their source (it is referred to as "font" in the book), but details about copyright, for example from GORILA2 or the British Museum, are nowhere specified. One of those images, namely the document HT Wc 3014, is referred to as "tablet?" (p. 98), whereas its classification as Wc and its picture makes it clear that the object is a roundel.
In dealing with texts in alphabetic Greek, prior to providing a translation, which might be—depending on the circumstances —either the ultimate goal or a middle phase, one focuses on understanding Greek by carrying out tasks such as the analysis of inner structures, textual parallelisms, and cultural context. Identifying the language behind a particular undeciphered script is not so different from a translation, which actually—at least to some extent—it is. Likewise, approaching a yet-to-be deciphered script is a multistep process whose different phases must complement each other. Tentatively identifying its language is not the starting point, but rather, the culmination towards which all the knowledge previously acquired ought to be oriented. The language that Linear A encodes still remains a tantalizing conundrum.
1. Furumark, A. 1956, Linear A und die altkretische Sprache: Entzifferung und Deutung, Berlin.
2. GORILA = Godart, L., Olivier, J.-P., Recueil des inscriptions en Linéaire A, 1-5, Paris, 1976-1985.