This volume (hereafter KT ⁶) is the latest in a series of publications of the Linear B tablets from the site of Knossos, Crete, following John T. Killen and Jean-Pierre Olivier's fifth edition (KT ⁵ = Killen & Olivier, Reference Killen and Olivier1989). Together with the tablets from other sites on Crete and the Greek mainland, these documents—which represent the earliest written attestations of the Greek language—provide a wealth of information on the administrative and economic concerns of the ‘Mycenaean’ palaces of Late Bronze Age Greece (c. 1400–1200 BCE). The accurate and up-to-date publication of these texts is, thus, of great importance both to Greek historical linguists and to archaeologists of Aegean prehistory.

The Knossos Tablets volumes present the texts in Roman transcription, organised by ‘series’ (groupings of texts on similar topics), together with information such as the texts’ attributions to scribal hands. These transcribed publications are designed to allow more convenient access to the tablets’ contents than a full corpus publication including photographs and drawings (which, for Knossos, is provided by CoMIK I-IV = Chadwick et al., Reference Chadwick, Godart, Killen, Olivier, Sacconi and Sakellarakis1986–1998). KT ⁶ is well-produced, with the texts clearly presented; the decision to remove information on joins from tablet headings in the main text and include it only in the ‘reconstruction’ section (highlighted on p. xxi) certainly produces a more reader-friendly layout. As well as updates to the scribal attributions, on which a significant amount of research has been done since KT⁵ by the authors and others, a key innovation in KT ⁶ is the inclusion of information on the tablets’ findspots, again based on recent work by the authors (on both of these aspects, see further below). At £55, the volume (which is also available as an e-book on JSTOR, but only to subscribing institutions) is reasonably priced compared to other similar publications.

New texts in KT ⁶ which have not previously appeared in a corpus include 115 fragments which have been identified since the publication of KT ⁵ and assigned the serial numbers 10013–10127 (for their original publications, see pp. 499–501). Of these fragments, twenty-seven are joined (or quasi-joined) to existing pieces, while the majority of the remainder belong to the X-series of fragmentary tablets with no preserved ideograms. In order to see how far the readings presented in this volume differ from previous editions, I compared a representative sample of texts—two tablets from each of the eighty-six series in KT ⁶—against the texts given in the most recent previous corpus (KT ⁵ or CoMIK, whose second, third, and fourth volumes post-date KT ⁵). Based on this sample, as well as on texts whose readings I have checked in the course of my research, the majority of texts show no changes from previous editions. Of those which do, the changes frequently relate to the apparatus—for instance, adding more details on the existence and location of erasures (examples include Cg 1039, F(2) 844, and Gv 864). Some transcriptions show improvements to their representation of the tablet's layout—such as the addition of a / to indicate a decrease in text size on Ai(3) 825.1, or of quotation marks to indicate that on Oa 734 the ‘bronze’ ideogram is placed above the ‘ingot’ ideogram (the new transcription reads ‘aes’ *167 + PE[). New suggestions are also occasionally offered as to the interpretation of particular signs’ functions, for instance, that on Ak(1) 5741, the isolated sign me may represent me<-zo> /medzōs/ ‘larger, older’, with an accidental omission of <-zo>, rather than functioning as an abbreviation (for this word or for its antonym me-u-jo/me-wi-jo /meiwyōs/ ‘smaller, younger’).

Changes to the actual readings of texts, apart from those which are due to recent joins, are mostly fairly minor. Where it has been possible to evaluate these changes based on the images published in CoMIK, or where I happen to have conducted my own autopsy of the tablets in question, most new readings or new suggestions in the notes seem well-supported. For instance, the reading j̣ọ-[ on X9213, given as an alternative suggestion to the reading *8̣3̣[ in previous corpora, is a much more plausible reading of this fragment; the reconstruction of the end of Sd 4416.b as a-ra-ro-mo-to-me-ṇạ[ , a-ja-]me-na, which differs from previous reconstructions in including a word-divider, seems reasonable given the high degree of consistency with which word-dividers are used in this series; and the certain reading of a]-na-to caps[ on Sg 7939 (a reading given as ‘not impossible’ in KT ⁵, while CoMIK has a]-na-to c̣ạp̣ṣ[) likewise seems reasonably secure. In other cases, although the suggested readings seem plausible, their presentation in the notes leaves their status slightly unclear. For instance, the term ko-ru-[ on Bk 803.3 has a note ‘ko-ru-j̣ạ[ likely’, and ta-ṭạ-ṛọ on Pp 498.1 has a note that certain features visible in the photograph in SM II (= Evans, Reference Evans and Myres1952) ‘suggest better ta-ẉọ-ṛọ’. It is not obvious why the apparently ‘likely’ or ‘better’ reading is included in the notes rather than the main text. In a few cases, I prefer previous readings to those given in KT ⁶. For instance, the form of the sign read as j̣ẹ on Sc(5) 251 (]-j̣ẹ-u) appears more characteristic of *4̣7̣ (as read by KT ⁵ and CoMIK), although the similarity between some forms of these two signs means that neither reading can be entirely excluded.

Some tablet series have been systematically reclassified in KT ⁶, as listed on p. xxii. These reclassifications are largely helpful, as they more clearly distinguish relatively coherent groups of tablets—e.g. personnel records by H104 and H106 (previously B(5) and B(3), respectively) are given new classifications as Bk- and Bo- to separate them from the miscellaneous B-series personnel tablets. The authors have also reclassified some groups of tablets linked by hand and/or findspot into ‘sets’ within the same series—for instance, the Dl-series of sheep tablets has been split into Dl(1a) and Dl(1b) (both by H118; see Firth & Melena, Reference Firth2016a: 256), Dl(2) (by H215) and Dl(3) (by H218). A reference list of these reclassifications would also have been useful. Readers may wish to consult Firth & Melena (Reference Firth2016a: 256) for a list including both series and set reclassifications, although note that this is not entirely complete (e.g. the creation of the Ag-series for H124-A's personnel tablets is not included) and that the Fh(2)-series included in this list is not used in KT ⁶. More one-off examples of individual tablets being moved from one series to another can be identified via the concordance, which lists each tablet by its unique (original) number, allowing readers to locate those which have been reclassified or renumbered.

As referred to above, however, the most potentially significant new feature of KT ⁶ is its incorporation of the considerable amount of work since the publication of previous corpora on the scribal attributions and findspots of the Knossos tablets. Changes to previous scribal attributions are discussed in detail by Firth & Melena (Reference Firth2016a; Reference Firth2016b; Reference Firth2016c). One of the most significant of these is the addition of H226-H234 as new ‘secondary hands’ (groups of tablets which are judged to be by the same hand, but which cannot certainly be either distinguished from or attributed to known scribal hands). The majority of these have previously been identified as possible groupings by various researchers but not formally assigned a number (references in Firth & Melena, Reference Firth2016b: 372–76). In addition, this is the first corpus to incorporate Jan Driessen's work on the scribes of the “Room of the Chariot Tablets” (those numbered H124-A, etc.: Driessen, Reference Driessen2000). The most significant difference from Driessen's work is KT ⁶'s assignation of numbers to the ‘additional tablet groupings’ (124-α, etc.), groups of tablets which may possibly be by the same hand but whose status is too uncertain to designate as a ‘stylus group’, the Room of the Chariot Tablets equivalent of a ‘secondary hand’ (Driessen, Reference Driessen2000: 93–94). Note, however, that the inclusion of these groups in the corpus should not necessarily be taken to imply a relatively certain identification, since the authors have previously expressed reservations as to how secure the attribution of many of these groups to a single writer really is (Firth & Melena, Reference Firth2016b: 341–43, 348–50).

The authors of KT ⁶ have previously published a series of articles reconstructing the tablets’ findspots and their post-excavation history, the latter being particularly important for those numbered 5000+, which generally lack original excavation records (for references, see pp. xix-xxix and pp. 499–501). Including information about findspots in the corpus should make the results of this work far more easily accessible to researchers for whom knowing the locations in which tablets were found is crucial. This is particularly true for those working on the Knossian administrative system and on the (controversial) administrative and chronological relationships between different deposits of tablets, which also have important implications for the chronology of the site as a whole. There are, however, some inconsistencies in this information, both compared to the authors’ previous work, and in the presentation of the information in different parts of the corpus. To cite one example of the former, Fh 462 (H141?) is attributed to its original recorded location, F3 (Magazine IV), despite the authors (amongst others) having plausibly argued, as part of wider discussions of possible links between different deposits and their implications for the tablets’ relative chronology, that this tablet was mis-recorded and probably originates from E1 (the Room of Column Bases: Driessen, Reference Driessen, Driessen and Farnoux1997: 120; Firth & Melena, Reference Firth, Melena, Bennet and Driessen1998–1999: 123, n.30; Reference Firth2016a: 314, n.109).

Systematic inconsistencies are also present relating to tablets numbered 5000+ whose findspots have been reconstructed based on their attribution to a particular scribal hand and/or series. Such tablets are presented in the main text and the ‘reconstruction’ section without any findspot, but listed under their reconstructed findspots in the ‘scribal hand’ and ‘classification’ sections, which lack any indication of the different status of recorded and reconstructed findspots. Although these reconstructions are generally entirely plausible, this inconsistent presentation is potentially misleading, especially to readers unfamiliar with the detailed history of work on the tablets. There are also occasional inconsistencies between the main text and the list of findspots and maps (pp. 681–90): findspots K1, K2, and K3 have been merged as a single location but are shown separately on the map; I3bis and J2bis are not shown on the map; and findspot I2 is variously referred to as ‘SCR. Deposit of “Great Seal”’ (SCR = Spiral Cornice Room: p. 502 and passim in ‘reconstruction’) and ‘Deposit of Great Seal (North Entrance Passage)’ (pp. 682, 687).

Although the inclusion of findspot information is a very positive step, its effectiveness in this volume is unfortunately limited, and readers requiring information on tablets’ (probable) findspots and their differing levels of certainty should still consult the authors’ previous works on the subject. Such consultations would have been made easier if the ‘reconstruction’ section—which includes references to the publications of joins—had similarly included references to discussions of findspots. This might, understandably, not have been possible in a print publication (this section is already nearly 200 pages long)—but given that this is the volume's most technical section, containing the most complex set of information, it would have been worth considering making this available as an online database as well as/instead of including it in the print volume (and eBook). Such a format would have enabled the important information it contains to be presented in a more accessible manner (the current shorthand system referring to publications of joins within the table saves space, but is not particularly user-friendly) as well as enabling the wider provision of references (e.g. to discussions of tablets’ findspots) and allowing this information to be more easily updated as further work takes place on these tablets. Although, of course, there are many advantages to print publication—not least the ease of studying texts from the same series together, which has always been a major purpose of the KT volumes—I would encourage all future authors of Mycenaean corpora to consider how the greater flexibility offered by online resources may complement a traditional print volume.