Hostname: page-component-586b7cd67f-t7czq Total loading time: 0 Render date: 2024-11-27T11:19:25.108Z Has data issue: false hasContentIssue false

THE TABLET-MAKERS OF PYLOS: AN EXPERIMENTAL INVESTIGATION INTO THE PRODUCTION OF LINEAR B TABLETS

Published online by Cambridge University Press:  29 June 2023

Anna P. Judson*
Affiliation:
British School at Athens and University of Durham
Rights & Permissions [Opens in a new window]

Abstract

The Linear B administrative texts of Late Bronze Age Greece were written on clay tablets, whose production therefore formed the first stage in the process of document creation, though it generally remains unclear whether the tablets’ writers were also their makers. This study combines experimental archaeology with autopsy of the tablets from Pylos in order to investigate the methods by which the Linear B tablets were created at this site. It thereby sheds light not only on the physical processes involved in shaping the clay, but also on the decisions involved on the part of the tablet-makers, and hence on the relationship between the ‘making’ and ‘writing’ stages of the process of creating the Linear B documents.

Τα διοικητικά κείμενα Γραμμικής Β ́ της Ύστερης Εποχής του Χαλκού στην Ελλάδα γράφονταν πάνω σε πήλινες πινακίδες, των οποίων η παραγωγή ήταν και το πρώτο στάδιο στην διαδικασία δημιουργίας αυτών των εγγράφων. Παραμένει ασαφές κατά το πόσον οι συγγραφείς των πινακίδων ήταν οι ίδιοι με τους παραγωγούς των πινακίδων. Η παρούσα μελέτη συνδυάζει τις πρακτικές της πειραματικής αρχαιολογίας με αυτοψία των πινακίδων από την Πύλο με στόχο την διερεύνηση των μεθόδων δημιουργίας των πινακίδων στην Πύλο. Η μελέτη στοχεύει να διαλευκάνει όχι μόνο την διαδικασία παραγωγής και διαμόρφωσης του πηλού, αλλά και τις αποφάσεις οι οποίες πάρθηκαν από τους παραγωγούς των πινακίδων, και επομένως την σχέση μεταξύ των σταδίων δημιουργίας και της συγγραφής των εγγράφων της Γραμμικής Β ́. Μετάφραση: Χρ. Κωνσταντακοπούλου

Type
Articles
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NCCreative Common License - SA
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike licence (https://creativecommons.org/licenses/by-nc-sa/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the same Creative Commons licence is included and the original work is properly cited. The written permission of Cambridge University Press must be obtained for commercial re-use.
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of The Council, British School at Athens

INTRODUCTION

Documents in the Linear B writing system almost exclusively take the form of clay tablets, which were used for record-keeping in the administrative centres of Late Bronze Age Greece (c. 1400–1200 BCE). The tablets themselves generally receive less attention than the texts written on them, which provide a wealth of information on Late Bronze Age Greek society as well as representing the earliest recorded form of the Greek language. However, the process of shaping clay to form tablets prior to writing on them is a crucial stage in the creation of these texts: focusing on this less-studied aspect of the Linear B documents sheds light not only on their materiality, but also on the operation of the whole administrative recording process.

In this study, I combine experimental tablet production with autopsy of the original tablets from Pylos in south-western mainland Greece. As the majority of the c. 1000 Linear B texts from this palace are securely associated with its final destruction, c. 1200–1180 BCE (early Late Helladic [LH] IIIC),Footnote 1 this site provides an opportunity to investigate the practices of a single, contemporaneous community of tablet-makers and -writers.

Previous experimental work

Many Mycenologists will have made replica Linear B tablets at some point, with students or as a public engagement activity,Footnote 2 if not as part of experimental research. Examples of the latter include creating tablets to test the use of the styli found at Tiryns (Godart Reference Godart1988, 248–50; Reference Godart and Marazzi1994; on styli, see also Steele Reference Steele2020); experiments carried out as part of Sjöquist and Åström's (Reference Sjöquist and Åström1991, 19–25) investigations into tablet production; demonstrations that dry tablets could be re-wetted in order to edit their texts (Pape Reference Pape2002; Pape et al. Reference Pape, Halstead, Bennet, Stangidis, Galanakis, Wilkinson and Bennet2014) and that they could survive being transported long distances (Hallager Reference Hallager, Nosch and Enegren2017); and an investigation of the means of smoothing tablet surfaces and edges (Greco and Flouda Reference Greco and Flouda2017, 149–51). The different methods used for creating tablets – and what reasons a tablet-maker might have to choose one over another – have, however, not previously been the subject of systematic experimental investigation. This study therefore seeks to improve our understanding of the processes involved in shaping clay to form tablets, the impact of the choice of particular methods of doing so, and hence the considerations in play for the tablet-makers during this procedure.

How were tablets made?

The first stage of creating a tablet is, of course, the collection and processing of the clay.Footnote 3 A wide range of different clays appear to have been used at Pylos, including both fine and coarse clays, as observed by autopsy (Palaima Reference Palaima, Sjöquist and Åström1985; Reference Palaima1988), macroscopic petrographic analysis (Nakassis, Pluta and Hruby Reference Nakassis, Pluta, Hruby and Cooper2021, 168; Hruby and Nakassis Reference Hruby, Nakassis, Bennet, Karnava and Meißnerforthcoming), and portable X-ray fluorescence (Wilemon Reference Wilemon2017; Wilemon, Galaty and Nakassis Reference Wilemon, Galaty and Nakassis2020). However, pending full publication of the latter two studies, it is not possible to discuss this aspect of tablet production in detail, and this study focuses on the next stage of the process, shaping the clay to form a tablet.Footnote 4

Linear B tablets are classified into two formats (Fig. 1), ‘palm-leaf’ (long and narrow; usually one or two lines of text recording a single piece of information) or ‘page-shaped’ (rectangular, orientated horizontally or vertically,Footnote 5 with more lines of text and usually containing multiple administrative entries), although this classification obscures the great degree of variation in size and shape within each format, as well as the extent to which they can overlap in size and function (cf. Driessen Reference Driessen2000, 42; Palaima Reference Palaima, Duhoux and Davies2011, 104 n. 134; Tomas Reference Tomas, Nosch and Landenius Enegren2017, 120–1; and ‘Tablet-makers and tablet-writers’, below). Other document types include labels – small pieces of clay attached to baskets or trays containing tablets (see, e.g., PT 3, xliii–xliv, lxxi) – and sealings, which are usually three-sided and formed around a knotted string, bearing a seal impression and sometimes short inscriptions with information about the goods they accompanied from other locations to the palace.Footnote 6 It is often said that palm-leaf tablets were used for preliminary documentation, before the information from a set of palm-leaves was transferred to a page-shaped tablet as the final document (e.g. Palaima Reference Palaima and Cline2010, 360; Del Freo Reference Del Freo, Freo and Perna2019, 172). However, direct evidence for this multi-stage processing of administrative information is limited;Footnote 7 in many cases, a basket of related palm-leaf tablets could have formed the final ‘document’, while a page-shaped tablet could have been the first and only stage of documentation.Footnote 8

Fig. 1. Illustration of tablet shapes: palm-leaf (Fr 1203, top left); vertical page-shaped (An 1, right); horizontal page-shaped (Es 647, bottom left). Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.). Scale bar: adapted from photographic reference scale by Jim Elder, Ottawa, Canada (smallpond.ca/jim/scale), CC-BY-NC-SA 4.0.

The most systematic previous discussions of methods of making tablets are provided by Palaima's (Reference Palaima1988) descriptions of the characteristics of tablets written by different scribes at Pylos, which in some cases include the techniques used to create them, and by Sjöquist and Åström's (Reference Sjöquist and Åström1985; Reference Sjöquist and Åström1991) studies of the palmprints on the Pylos and Knossos tablets, both of which will be further discussed below. General comments on the process of making tablets usually relate to palm-leaves, and describe sheets of clay being folded up to form long narrow tablets (e.g. Palaima Reference Palaima, Sjöquist and Åström1985, 103; Reference Palaima, Duhoux and Davies2011, 105–6; Del Freo Reference Del Freo, Freo and Perna2019, 172–3); however, palm-leaves could also be made by rolling out a cylinder of clay and flattening it with the palms (Sjöquist and Åström Reference Sjöquist and Åström1985, 44; Reference Sjöquist and Åström1991, 13–16, 18; Driessen Reference Driessen2000, 41; Fig. 2) or simply by moulding the clay into shape with the fingers.Footnote 9 Some tablets – usually, but not exclusively, palm-leaves – also had a piece of straw or string inserted longitudinally, identifiable by the channel left through the tablet (see Godart Reference Godart1988, 248–9; Driessen Reference Driessen2000, 40; Palaima Reference Palaima, Duhoux and Davies2011, 105). The process of making page-shaped tablets is more rarely discussed, although Palaima (Reference Palaima1988, 46, 59; Reference Palaima, Duhoux and Davies2011, 105) states that some were made by moulding fine clay over a coarser core (but see ‘Experimental methodology’ below), and Bennett (Reference Bennett, De Miro, Godart and Sacconi1996, 29) refers to the edges of page-shaped tablets being folded over ‘to make nicely rectangular shapes’. The edges of tablets were also neatened by pressing against a flat surface (e.g. Driessen Reference Driessen2000, 41) or by burnishing (Greco and Flouda Reference Greco and Flouda2017, 149–50). Tablets were frequently cut along one or more sides, either to remove unused clay or to create multiple separate records (Palaima Reference Palaima, Sjöquist and Åström1985, 103; Reference Palaima1988; Tomas Reference Tomas, Piquette and Whitehouse2013).

Fig. 2. The author rolling out a cylinder of clay to form a tablet in the Fitch Laboratory. Photo: Evangelia Kiriatzi.

Where were tablets made?

At the time of Pylos’ destruction, c. 80 per cent of the tablets were stored in the ‘Archives Complex’, two rooms next to the palace's main entrance.Footnote 10 The rest are mostly from other areas of the palace where items were processed and/or stored (Palaima Reference Palaima1988, 135–69), including a pottery storeroom (Room 20); three oil storerooms (Rooms 23, 32, and above Room 38: Shelmerdine Reference Shelmerdine1985, ch. 4); the North-Eastern Building, a clearinghouse for receiving and recording goods (Bendall Reference Bendall2003); and the South-Western Building, where ‘taxation’ records were written.Footnote 11 Evidence that at least some of the tablets found in the Archives Complex had been transferred from other locations around the palace is provided by the Sa series of chariot wheel records: these were mostly found in the Archives Complex, but one remained in the North-Eastern Building, where presumably the whole series had been written (Palaima Reference Palaima1988, 179; Bendall Reference Bendall2003, 220). The Sh series, recording armour, were stacked in a labelled basket near the Archives Complex's doorway at the time of the destruction, and may also have just been transferred from the North-Eastern Building (Palaima Reference Palaima, De Miro, Godart and Sacconi1996a; see also Kyriakidis Reference Kyriakidis1996–7, 214–24). Presumably tablets were made in or near all of the areas in which they were found,Footnote 12 as well as potentially other areas in, or just outside, the palace. An outside area might often have been a more convenient place – with more space, better light, and less of a need to clean up clay – to make and write documents, especially compared to the Archives Complex's two small rooms (cf. Palaima and Wright Reference Palaima and Wright1985, 259). It is also increasingly frequently being suggested (though has not been conclusively proven) that some palm-leaf tablets, like sealings, may have been written away from the palace – whether by writers based at another administrative centre or by palace-based writers who had travelled to other locations – and brought back for filing and/or compiling onto summary documents.Footnote 13

Who made the tablets?

Whether tablets were made by the same person who wrote on them was a major question in Sjöquist and Åström's (Reference Sjöquist and Åström1985; Reference Sjöquist and Åström1991) studies of the palmprints left by the tablets’ makers. At Pylos, their results were inconclusive due to the limited number of identifiable prints: only four makers were found to have certainly or probably left prints on more than one tablet, while a further six could be distinguished by a single print each (compare the c. 30–40 identified scribal hands at this site).Footnote 14 The complex relationship between these prints and the scribal hands will be discussed below in more detail, but it suggests that in some cases the tablet's maker and writer may have been the same person, while in others different individuals were responsible for each stage of the process; whether the makers in the latter cases were other scribes or assistants is not clear.Footnote 15 In this paper, I use ‘tablet-maker’ to refer to each tablet's creator, and ‘writer’ to refer to the person who inscribed the text, regardless of whether a particular tablet-maker was or was not also a writer (of that or any other text) or vice versa. Combining experimental production of tablets with autopsy of the originals and analysis of the correspondences between their manufacturing technique, format, contents, palmprints, and scribal attributions will produce a fuller understanding of the relationships between the making and writing stages of Linear B tablet production, whether these were the work of one person or two.

INVESTIGATING TABLET PRODUCTION

Experimental methodology

The experiments described here were carried out in the Fitch Laboratory at the British School at Athens using clay from the region of Elis, to the north of Messenia (the clay available in the Fitch's reference collection which was closest both geographically and geologically to the region around Pylos). Although as stated above it was not possible, nor was it intended, to replicate the original clays used, I used three clays to imitate some of the range of clay types found: a fine, silty clay (Fitch Laboratory reference KAVGS17/06), a coarse clay (KAVGS17/15), and a very coarse clay (KAVGS17/04). The process of experimentation was, of course, also a learning process for me, involving a fair amount of trial-and-error in order to find out, for instance, what consistency of clay made it easiest to produce tablets using a particular method – something that an experienced Mycenaean tablet-maker would no doubt have known instinctively through long practice. It is important to bear in mind, particularly in the following discussions of the degree to which different production methods may have required more or less care or effort, that the tablet-maker's training and experience are unreplicable aspects of their practices.

The initial experiments were carried out in parallel to the autopsy of the tablets in the National Archaeological Museum, so that each study could inform the other. One notable impact of the autopsy was that I observed no evidence for the claim that certain page-shaped tablets (Es 644 and Jn 605) were made by covering a coarse clay core with a layer of finer clay (Palaima Reference Palaima1988, 46, 59). On Jn 605 (Fig. 3), the appearance of a smooth layer of clay over a rougher layer on both the top and bottom edges is the result of the common practice of removing clay by cutting partway through the tablet (leaving a smooth surface, on which cut-marks are clearly visible) and then tearing it the rest of the way, leaving a rough surface; note that large inclusions can be seen in the smooth upper part as well as the rougher lower part. On Es 644, a similar effect is due to breakages along ruled lines, both at the bottom and various points in the middle of the tablet, producing the appearance of a finer upper layer. I therefore did not experiment with using a coarse core covered by a layer of finer clay.

Fig. 3. Jn 605 lat. inf. Photo: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

Conversely, the most important impact of the experiments on the autopsy was – unfortunately – the observation that the traces left by creating rolled and folded tablets can be effectively indistinguishable: a folded tablet can appear identical to a rolled one if the seam on the verso has been entirely smoothed over and the ends have been pinched together or smoothed, and even the appearance of small seams on the ends does not necessarily indicate a folded tablet, since the ends of rolled tablets can be shaped in a similar way. Thus, although some tablets show clear signs of folding, many others are uncertain, and it is extremely difficult to be sure that a tablet has not been folded. In practice, therefore, it generally remains unclear how consistent any given series of tablets is in its method of manufacture.

Folded tablets

As stated above, although flattening out sheets of clay and then folding them up is often said to be the usual method of creating palm-leaf tablets, the simpler method of rolling out and flattening a cylinder of clay is also referred to; in his study of the Pylos tablets, Palaima (Reference Palaima1988) only mentions folding in reference to a few groups of tablets (H1's Aa series, and the Ma, Na, and Ta series). Page-shaped tablets are also sometimes folded, probably by folding a flattened sheet in half, since when seams are visible this is along one or more edges rather than in the middle of the verso as on palm-leaves. The folding method evidently has more requirements than the rolling-and-flattening method, needing a means of flattening out a sheet of clay (in my case, a rolling pin; I imagine that a similar implement would have been the easiest way for a Mycenaean tablet-maker to create a flat sheet, but no evidence for this exists), and a large enough surface on which to do this; it also requires more time and effort in neatening the resulting tablet by smoothing over the seam where the sheet was joined. (In practice, tablets made in this way are much more often identifiable from the visible edges of the folded sheet than from the seam, which is normally completely or almost completely smoothed over: Fig. 4.) In addition, experimentation showed that creating a folded tablet is much more dependent on the clay type and consistency – it is easier to do with fairly wet fine clay than with drier and/or coarser clay, though even then the edges tend to crack while being folded, and clay that is too wet tends to stick to the surface beneath it or to the rolling pin. However, both coarse and fine clay can be easily rolled and flattened even while comparatively dry. Creating folded tablets therefore would have required more planning and work at all stages of the process, from clay preparation to neatening the finished tablet, not to mention cleaning up the wetter clay – so why was this method sometimes chosen by the tablet-makers?

Fig. 4. Ma 123, lat. sin. and verso. The fold is clearly visible on the lat. sin. but there is only a trace of the seam at the left-hand end of the verso. Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

In preliminary experiments, I observed that after partly drying, folded tablets were less prone to bending than rolled ones. To test whether this effect lasted after fully drying, and therefore whether folded tablets were ultimately sturdier than rolled ones, I created a series of folded and rolled tablets of fine clay (the former using relatively wet clay, the latter using both wetter and drier clay).Footnote 16 These were left to dry outside until they were too dry to inscribe properly, tested by writing signs on each tablet every hour until the resulting strokes were very shallow (as seen on actual tablets when additions have been made some time after the initial writing; outdoors in Athens in July this took five to six hours: Fig. 5).Footnote 17 Weighing each tablet every hour showed no difference in drying rate – all tablets lost similar proportions of their mass every hour – and by the end of this period there was no observable physical difference between the folded and rolled tablets, none of which were at all flexible. A repetition of this experiment in which I observed the tablets’ flexibility – how much they naturally bent when held by the middle or (when dry enough that this did not happen) how much they could be deliberately bent without a risk of breaking – every hour showed that while initially the folded tablets were significantly less flexible than the rolled ones (Fig. 6), after one to two hours (depending on the clay's original consistency) all of the tablets were equally unable to be bent more than very slightly without cracking or breaking. Further experiments using the two coarser clays and with page-shaped tablets showed the same effect, although this was less pronounced for the very coarse clay (which produced less flexible tablets to begin with) and for page-shaped tablets (whose greater width relative to their length likewise made them less flexible even when wet).

Fig. 5. The author writing on drying tablets to test their consistency. Photo: Emily Sherriff.

Fig. 6. Experimental tablets made of fine clay (left) and coarse clay (right) showing effects of holding by the middle while wet. Top: rolled; bottom: folded. Photos: author.

Thus, the advantage of folding is a short-term one, increasing tablets’ stability only during a relatively short period after their creation. This method therefore seems intended to produce tablets which can more easily be handled while still relatively wet, reducing the risk of them bending or breaking at this stage. Almost all of the series which were inscribed while the clay was still fairly wet contain at least some tablets with clear traces of being made by folding (cf. above on the difficulty of establishing how consistently this method was used),Footnote 18 so that in these cases the makers’ intention could have been to facilitate the tablets’ inscription, whether by themselves or other writers. The same is, however, true of many series inscribed – as appears to have been more usual – after the clay had partially dried, implying that the tablet-makers’ concern may have been for the tablets to retain their shape until this point (for instance, if being moved to another position for drying). In this method of tablet-making, therefore, we may see interactions between multiple different stages of document creation: tablet-makers on some occasions chose a more complicated method, involving more steps and more care, in order to enable themselves and/or others to more easily handle the resulting tablet – whether during the rest of the manufacturing process or while inscribing the text.

Straws/string

Evidence of the inclusion of a piece of straw or string,Footnote 19 usually running horizontally through the centre of the tablet and identified by the holes left at its exit pointsFootnote 20 and/or the channel left through the tablet (Fig. 7), is found in five main series. These are listed below along with the proportion of each series which, based on autopsy, originally contained a straw/string:Footnote 21

  • Sh series (H5), palm-leaf records of suits of armour, found in the Archives Complex: 100 per cent (12/12).

  • Eb series (H41), palm-leaf first-stage landholding records, found in the Archives Complex: 97 per cent (61/63).Footnote 22

  • Sa series (H26), palm-leaf records of chariot wheels, found in the Archives Complex and North-Eastern Building: 94 per cent (33/35).Footnote 23

  • Eo series (H41), palm-leaf and page-shaped first-stage landholding records, found in the Archives Complex: 77–85 per cent (10–11/13).Footnote 24

  • Ad series (H23), palm-leaf personnel records of groups of men and boys (described by their relationship to the women workers recorded in the Aa series by H1 and H4)Footnote 25 at various locations in Pylos’ territory, found in the Archives Complex: 14–16 per cent (5–6/37).Footnote 26

Fig. 7. Eb 149 lat. dex., with straw/string hole (left); Sh 740 recto, with partially exposed straw/string channel (right). Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

In addition, straw/string holes appear in a small number of other tablets, including two of H21's Cc series (Cc 1283 and 1285, sheep records from the North-Eastern Building; it is not clear what administrative relationship, if any, these have to the rest of the Cc series, which lack this feature);Footnote 27 two Va series tablets attributed (tentatively) to H42 (Va 404 and 482, ivory-working records from the Archives Complex; there is no clear relationship to the other Va series records tentatively attributed to this hand);Footnote 28 and a small number of isolated, unattributed tablets.Footnote 29

The inclusion of these straws/strings has been suggested to have a variety of purposes: giving the tablet greater stability (Godart Reference Godart1988, 248–50; Reference Godart and Marazzi1994;Footnote 30 Palaima Reference Palaima1988, 27; Reference Palaima, Duhoux and Davies2011, 105–6;Footnote 31 Del Freo Reference Del Freo, Freo and Perna2019, 172–3); allowing the tablet to be lifted while still wet without distorting its shape or smudging the inscription (Palaima Reference Palaima, Ferioli, Fiandra and Fissore1996b, 104–5; Reference Palaima, Duhoux and Davies2011, 105); preventing the loss of any pieces broken off in handling or transportation (Bennett Reference Bennett, De Miro, Godart and Sacconi1996, 28–9; Palaima Reference Palaima, De Miro, Godart and Sacconi1996a, 382 n. 10; J.-P. Olivier in Palaima Reference Palaima, Ferioli, Fiandra and Fissore1996b, 105); or attaching sealings to tablets for authentication (E.L. Bennett in Palaima Reference Palaima, Ferioli, Fiandra, Fissore and Frangipane1994, 334–5; Palaima Reference Palaima, Ferioli, Fiandra and Fissore1996b; Flouda Reference Flouda2010, 65–6; Younger Reference Younger and Cline2010, 334; Panagiotopoulos Reference Panagiotopoulos2014, 190–1, 252–4). The last of these hypotheses is not, in my opinion, a plausible one: sealing as an authentication practice seems to have been required only at the administrative stage represented by the sealings – extremely short documents likely to have been written in multiple locations and sent to or from the palace along with the goods they registered – whereas the act of writing a palm-leaf or page-shaped tablet, whether this took place in or away from the palace (as discussed above), seems to have constituted all the authentication required. Otherwise, we would expect far more examples of tablets which could potentially have had sealings attached to them – or, indeed, the impression of seals directly onto the tablets themselves, as was frequently done on Ancient Near Eastern cuneiform documents.Footnote 32 Palaima (Reference Palaima, Ferioli, Fiandra and Fissore1996b) argued that the particular tablet series which show this feature at Pylos could have required special authentication, but also (since relatively few sealings have been found within the Archives Complex, where these series of tablets were all stored and where other records important enough to require authentication of this type would be most likely to have been kept)Footnote 33 that the individuals concerned in the records would have retained these sealings as ‘receipts’. However, if the sealings were not to be retained with the tablets, there would be no need for them to be attached in this way, and Palaima's (Reference Palaima, Duhoux and Davies2011, 105–6) own more recent summary of tablet manufacturing processes favours different suggestions (listed above).Footnote 34 The experiments therefore focused on the hypotheses that straws/strings provided extra stability, a means of moving the wet tablet, and/or a protection against the loss of broken fragments.

The experiments described above already demonstrated that the effect of folding was to give the tablet greater stability in the initial stages of drying. Subsequently, I tested the effect on this of incorporating a straw/string,Footnote 35 by creating rolled and folded tablets of fine and coarse clays with and without straws. Although incorporating the straw made a slight difference, folding made a much larger one – the unfolded tablets with straws were still very prone to bending (the thickness of both the straw and of the tablet also had some effect on this), while the folded tablets with straws, although the most stable option, were only slightly more so than the folded ones without straws (Fig. 8).

Fig. 8. Experimental tablets showing effects of holding by the middle while wet. From top to bottom: rolled, without straw; rolled, with straw; folded, without straw; folded, with straw. Photo: author.

Tablets made with a straw/string frequently have visible channels near one or both ends of the recto or verso,Footnote 36 sometimes even cutting through the beginning or end of the text,Footnote 37 while in the Sa and Sh series, whose clay was especially wet while inscribing, the edges of the straw/string holes are also sometimes distorted (e.g. Sa 487, 682, 794; Sh 739). Both of these features initially implied to me during autopsy that the straw/string had been partly pulled up out of the clay, as also happened in experiments if using the straw/string to lift a tablet that was wet enough to stick to the surface underneath, or if the straw/string was very near the surface of the tablet. However, experiments showed that trying to lift a sticking tablet in this way tended to result in a much longer portion of the straw/string being pulled out of the clay, significantly damaging the tablet; in addition, many tablets were clearly handled while the clay was wet, as shown by the presence of finger-marks on the top and bottom edges (e.g. Sa 753, 758, 766, 769, 790, 791, 834; Sh 743, 744). I therefore think it more likely that these channels were the result either of the tablet being formed in such a way that the straw/string exited the clay slightly before the end of the tablet, or (where the channel cuts through the text) of later accidents – the straw/string catching on something during handling or transportation, or the loss of the thin layer of clay above a straw/string running close to the tablet's surface after drying (as happened in one experimental case – see below).Footnote 38

Experiments in using straws/strings to lift tablets for a few seconds, in order to simulate a use such as transferring the wet tablet to another location to dry, produced mixed results. The end of the first trial example cracked and broke off completely after drying; further experiments demonstrated that it was possible to avoid this result, but that this was very variable and related to a variety of factors. In a second experiment, a tablet whose clay had been rolled around a straw cracked while one that had been folded did not; in a third, none of the range of tablets of various thicknesses (1–2.1 cm) made of relatively dry clay cracked or broke, but a single tablet made with much wetter clay (of medium thickness: 1.5 cm) developed a large crack near the middle. Finally, a series of tablets were made to test more systematically the impact of the consistency of the clay and of folding the clay around the straw/string or rolling it, as well as whether lifting by only one end of the straw/string, rather than both, could prevent breakage. In this case nearly all of the tablets (rolled or folded; containing straw or string; made of drier or wetter clay or fine or coarse clay; carried by one end or both ends of the straw/string) remained undamaged, while a single tablet (made of drier clay, folded, and carried by one end) cracked slightly. Thus, although damage to tablets from this use was not inevitable, it is certainly unreliable and carries the risk of greater damage than would be caused simply by handling the clay itself – as the finger-marks mentioned above imply frequently happened. There is also no correlation between the consistency of the clay while inscribing and the presence of straws/strings: the Sa and Sh series were inscribed while the clay was still relatively wet, but so were the Ad series (with only a minority containing straws/strings) and the Aa, Es, and Qa series (with no straws/strings); conversely, the Eb and Eo series were not particularly wet when inscribed (cf. Palaima Reference Palaima1988, 70, 87, 91, 98–9, 121).

Finally, to test the ability of straws/strings to prevent the loss of broken fragments, eight tablets were created (of fine and coarse clay, rolled and folded around straws and strings), left to dry, and then transported from Athens to the UK and back in hand luggage.Footnote 39 Two were deliberately broken in the lab, a third broke before leaving when accidentally dropped, and two further tablets broke in transit. On returning to Athens, three of the broken tablets – including one which had broken into five pieces – were still held together by their straws/strings (Fig. 9:1–3), one had lost a fragment from the end where the string had been pulled out of the tablet before drying (Fig. 9:5), and only one had lost a piece that had contained the string (Fig. 9:4): as the string had been very close to the tablet's edge, damage to the surface resulted in the string coming loose (or, alternatively, the surface was broken by tension on the string).Footnote 40 While not a fool-proof method, then, using a straw/string would have significantly lessened the risk of losing fragments from any tablets which broke during transportation – whether they were being transported from another part of the palace to the Archives Complex, or to Pylos from another location within the Pylian territory.

Fig. 9. Broken tablets containing straws/strings on their return to Athens. Photo: author.

These experiments suggest that the most likely purpose(s) of the straws/string were to increase the tablets’ stability (in combination with folding) and/or to keep fragments together in case of breakage during short- and/or long-distance transportation. Unfortunately, the series with straw/string holes show no particular patterns in terms of findspot or subject-matter to suggest that they are more likely than others to have been transported either to or within the palace. As said above, the Sa and Sh series may well both have been transferred to the Archives Complex from the North-Eastern Building; however, of the tablets actually found in the North-Eastern Building only two of H21's four Cc series tablets have straw/string holes.Footnote 41 Of the series containing (some) straw/string holes, only the Ad, Eb and Eo series explicitly record activities taking place in locations outside of the palace.Footnote 42 In the Ad series, which records work-groups located in both the ‘Hither Province’ (southern and western Messenia, including Pylos itself) and the ‘Further Province’ (the area to the east of the Aigaleon mountain range),Footnote 43 there is no correlation between the presence or absence of straw/string holes and the location referred to.Footnote 44 Moreover, neither H1's Aa series tablets, which list the related work-groups of women and children in the Hither Province records, nor H4's, which list the same for the Further Province, contained straws/strings (H21's Ab series, which also lacks this feature, contains records of rations only for the Hither Province work-groups). The Eb and Eo series refer to landholdings at a place called pa-ki-ja-ne (perhaps /Sphagianes/), the site of a religious sanctuary near the palace, but whether they were written on a ‘site visit’Footnote 45 or based on information conveyed orally or on other written materials from that location remains unknown. H43's Ea series of palm-leaf tablets, which record similar (though less detailed) landholding information about another, unknown, location (Lejeune Reference Lejeune and Lejeune1997), are made without straws/strings; it is possible that this reflects a difference in writing location, but there is no other positive evidence for this. The hypothesis that an administrative reason (such as the need to transport these preliminary documents from the location of their writing) underlies the use of straws/strings, rather than a purely physical reason relating to tablet manufacture, is supported by the lack of a relationship between the distribution of straws/strings and tablet shape/size in the series in which these are used inconsistently (the Ad and Eo series).Footnote 46 It is similarly supported by the fact that, in the only series to contain straws/strings in page-shaped tablets (the Eo series), these are functionally identical to the palm-leaves as first-stage administrative documents.Footnote 47 However, the lack of a definite correlation between tablets’ likely movements and the presence or absence of straws/strings, along with the probability that more than just these fairly restricted groups of tablets were moved around, make it most likely that incorporating straws/strings was an option for tablet-makers to consider, not a requirement when making tablets that would be moved around within, and perhaps to, the palace.

Tablet shape and finishing

Palm-leaf tablets characteristically have a tapering shape, with one or both ends narrower than the middle in width and/or height, the latter meaning that the ends of the verso frequently curve up (Fig. 10). My experiments showed that this shape naturally occurs when a cylinder of clay is flattened with both hands positioned near the ends, or when a sheet of clay is flattened using a rolling pin before folding up, since in both cases more pressure is put on the edges than the middle. Of course, the exact shape could be adjusted both in the course of rolling/flattening (by shifting hand position or altering the amount of pressure) and once the basic tablet had been formed, by either drawing out or blunting one or both ends (cf. Palaima Reference Palaima, Sjöquist and Åström1985, 103).

Fig. 10. Ad 683 recto and lat. inf. Note the tapering right-hand end and the curvature of the verso at both ends. Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

During an autopsy session, Evangelia Kiriatzi observed that the tablets’ surfaces and top and bottom edges had been not merely flattened, but smoothed in a way that resembled the burnishing of pottery, as already remarked on by Greco and Flouda (Reference Greco and Flouda2017, 149–51) based on their studies of tablets from Knossos. In line with the results of Greco and Flouda's experiments, I found that this effect could be achieved using the rounded edge of a wooden clay-shaping tool: wetting the tool made it easy to draw it across the clay surface, producing a smooth finish on the writing surface and edges. The various shapes of smoothed tablet edges – which can be either rounded or flattened, with rounded or slanted edges – could easily be produced by altering the angle and pressure of the tool. Note that this smoothing can produce an appearance very similar to that of an erasure (since it is, effectively, the same process); it is therefore possible that at least some of the tablets identified as palimpsests due to showing ‘erasure’ marks, but which do not preserve any identifiable traces of previously written text, may in fact be showing traces of this smoothing process.

Page-shaped tablets, whether simply flattened or folded, require shaping to form a rectangular tablet rather than one with the curved edges that are naturally produced by flattening a lump of clay. From autopsy observations, page-shaped tablets’ edges were frequently either folded over onto the verso (cf. Bennett Reference Bennett, De Miro, Godart and Sacconi1996, 29) and/or flattened, often sloping onto the verso with a raised area of clay behind them (Fig. 11). Although the page-shaped tablets are often cut on one or two sides to trim off unused clay, like the palm-leaves, and occasionally larger tablets are cut to create two smaller ones,Footnote 48 they are never cut on all four sides to the desired shape.Footnote 49 In experiments I found that the flattened and sloping edges could be produced by squashing the tablet's edge with a flat surface; using a hand-held object such as the side of the rolling pin or a plastic block allowed the angle and pressure to be more easily adjusted than when pressing the clay against the surface it was sitting on. Pressing this object down on the edge of the clay at an angle produced the characteristic straight edge sloping towards the verso, with a ridge of clay pushed up behind it. Both of these methods served to produce rectangular tablets with neater, straighter edges, which could then be further neatened by smoothing the edges and recto as described above.

Fig. 11. Jn 415 verso showing folds and ridges of clay along the bottom edge and sides (the top has been cut). Photo: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

This experimental investigation has thus shed new light not only on the processes involved in creating tablets, but also on the decision-making of the tablet-makers in choosing between those processes, and the ways in which this relates to the next stages of document production and use. In the final section of this paper I shall investigate this last issue further by using a series of case-studies to explore the relationship between the tablets’ makers and writers.

TABLET-MAKERS AND TABLET-WRITERS

No clear connection can be seen between a tablets’ format and/or production method and the identity of their makers, where this is known. As mentioned above, Sjöquist and Åström (Reference Sjöquist and Åström1985, 47–56 and table 1) were able to identify prints (probably) belonging to the same maker on more than one tablet in only four cases:

  • ‘Dokimastikos’: Qa 1292, 1294, and 1311 (all H15).

  • ‘Energetikos’: at least one Ab series tablet (H21),Footnote 50 at least five Ea series tablets (H43),Footnote 51 at least three Eb series tablets (H41),Footnote 52 and perhaps Eo 268 (H41).

  • ‘Mikros’: Ea 801 and 823 (H43), perhaps also Ea 305 (H43) and Aa 94 (H4).

  • ‘Anon I(?)’: perhaps Ea 922 (H43) and Eb 477 (H41).Footnote 53

We thus have one instance of the same maker's prints being found on multiple tablets inscribed exclusively by a single writer – ‘Dokimastikos’ in H15's Qa series – where the simplest assumption is therefore that these may be the same person; unfortunately there are no identifiable prints on the three Qa series tablets inscribed by a different writer, H33.Footnote 54 All tablets within this series are similar in format and probably also in production process (they are relatively small and thin palm-leaf tablets, with flat rather than curved versos; at least some of both the H15 and H33 tablets have been made by folding). It is not possible to say whether this is because ‘Dokimastikos’ made all of the tablets regardless of their eventual writer, but looking at the instances of prints being found in multiple series by different writers implies that this assumption is not necessary. No significant differences are visible between the Ea series tablets made by ‘Energetikos’, ‘Mikros’, or ‘Anon I(?)’, or the Eb series of ‘Energetikos’ and ‘Anon I(?)’; nor do tablets from different series showing prints of these makers appear more similar to each other than to other members of the same series.Footnote 55 The Eb series, for instance, consistently contained straws/strings, and although the ‘Anon I(?)’ tablet has a more blunted left-hand end than those attributed to ‘Energetikos’ (which, when not broken, are either rounded or left without neatening), this is a feature which varies considerably both in general and within this series. As far as can be seen, then, variation in tablet format/manufacture does not relate to individual preferences of the tablets’ makers.

Of the writers of tablets with straws/strings, H5 (Sh series), H23 (Ad series), and H26 (Sa series) are not known to have written any other tablets,Footnote 56 but H21, H41 and H42 have all written palm-leaf tablets made both with and without this feature.Footnote 57 As already discussed, it is harder to be certain how consistently tablets have been rolled or folded, but comparing tablets’ size and shape is more straightforward, and again there is no evidence for any preference by writers beyond the needs of the individual texts. For instance, as a fairly crude measure of this, H1's complete palm-leaf tablets range in surface area from 9.9 cm2 (Na 530) to 68.4 cm2 (Ed 236), and page-shaped tablets from 27.7 cm2 (An 199) to 412.8 cm2 (En 74); note the considerable overlap in size between the largest palm-leaves and smallest page-shaped tablets, as well as variation in other aspects of formatting such as the orientation of page-shaped tablets (e.g. the En and Ep landholding series contain both vertical and horizontal tablets). Thus, there is also no clear connection between writer and tablet format or manufacturing method: we are not dealing with consistent preferences on the part of writers any more than that of makers (remembering again that these may often have been the same person).

In fact, even within broad groups of tablets relating to similar topics and written by a single person there is considerable variation. The Fr series, for instance, consists of records of small amounts of olive oil being issued, mostly for religious purposes (to deities or sanctuaries), written by at least six people (H2, H4, H17, H18, H19, H41). Although the texts display a very similar structure – brief descriptions of the type of oil and recipient (e.g. ‘In the territory of Lousos: to the gods: sage-scented: c. 5 litres of olive oil’ [Fr 1226, H2]) – their format varies widely. The writer responsible for the largest number of these texts, H2, has used both one- and two-line palm-leaves, the latter of which may contain one or two entries, of very different sizes (length 8.9–21.5 cm, width 1.9–3.3 cm, thickness 0.9–1.6 cm) to write distribution records, in addition to one small page-shaped tablet for the single record relating to perfume production (Fr 1184). Other writers have mostly also used palm-leaves, but again these vary considerably in size and shape (e.g. H4: tapering; H17: narrow rectangular; H19: rounded, one- and two-line); H18 has the widest variation in this respect (Fig. 12), with tablets including a palm-leaf with a two-line single entry (Fr 1225), a small horizontal page-shaped tablet (Fr 1218), and a rounded three-line tablet which in shape resembles an enlarged palm-leaf but is formatted like a horizontal page-shaped tablet (Fr 1217).Footnote 58 The findspots of these tablets, almost all of which were found in the palace's oil storerooms (Rooms 23, 32, and above Room 38),Footnote 59 suggest that they were created and written on the spot (or in some cases selected from existing tablets which could be re-used)Footnote 60 whenever an issue of oil had to be made: this variation in size, shape and format therefore represents the results of a series of separate decisions made by the tablets’ makers and/or writers to fulfil the recording needs of each separate transaction.

Fig. 12. H18's Fr series tablets. From top to bottom: Fr 1225, 1218, 1217. Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

Where some level of consistency can be seen is within groups of tablets that are not just on the same topic but are closely related in administrative terms – ones that were clearly written as components of a single administrative action and, arguably, collectively form a single ‘file’ of information; this applies to some tablet series and in other cases to sub-groups of series known as ‘sets’. The creation of tablets to specific sizes and formats for specific (sets of) texts, as mentioned above, has frequently been pointed out, and it is often stressed that many tablets appear carefully designed in size and shape for the needs of their particular text, implying at the very least a close collaboration between maker and writer, if not a shared identity (e.g. Palaima Reference Palaima, Sjöquist and Åström1985, 101; Reference Palaima, Duhoux and Davies2011, 84–5; Del Freo Reference Del Freo, Freo and Perna2019, 173; LSP, 15). An examination of the Pylos personnel tablets shows, however, that the details of tablet design can go beyond the demands of the particular text and speak to the individual preferences or choices of the tablets’ maker and/or writer.

We have already seen that the Aa, Ab and Ad series, while representing separate administrative recording actions, are closely linked by their references to the same or related groups of workers: work-groups of women and children in Pylos and the Hither Province, designated by their occupation or place of origin, are recorded by H1's Aa series and by H21's Ab series, which also records their monthly rations (as the numbers differ slightly these two series were presumably not written at the same time); similar work-groups in the Further Province are recorded by H4's Aa series, without corresponding ration records; and H23's Ad series records men and boys in both provinces who are related to the women of the Aa and Ab series. The following example of three related Hither Province texts, plus one Further Province Aa series text, shows their similarity in structure and content:

  • ‘21 Knidian women, 12 girls, 10 boys, one male supervisor(?), one female supervisor(?)’Footnote 61 (Aa 792, H1, Hither Province)

  • ‘At Pylos: 20 Knidian women, 10 girls, 10 boys: 640 l. figs, 640 l. grain; female supervisor(?), male supervisor(?)’ (Ab 189, H21, Hither Province)

  • ‘At Pylos: sons of Knidian women: five men, four boys’ (Ad 683, H23, Hither Province)

  • ‘textile decorators: 12 women, 16 girls, eight boys, one male supervisor(?), one female supervisor(?)’ (Aa 85, H4, Further Province)Footnote 62

The textual variations which occur do so for both administrative reasons (the Ab series contains additional ration information; the Ad series work-groups do not include supervisors) and due to writers’ differing choices about how to present the information; the latter are comparatively minor (H1 does not specify the workers’ location when they are at Pylos, unlike H21 and H23; H1 and H4 include numerals for the supervisors while H21 does not). However, in format they are entirely different (Fig. 13) – even H1's and H4's Aa series tablets are strikingly dissimilar, with H4 using extremely long tablets (mean length 23 cm, mean height 2.5 cm), while H1's are much shorter and wider (mean length 14.6 cm, mean height 2.8 cm; in both cases, at least some of the tablets have been folded). That these differences reflect preferences of the writers rather than the tablet-makers (or, if those are the same people, preferences based on writing practices rather than on the process of tablet production) is shown by the way they correspond to differences in the writing of the text – H4's signs are written much larger and are more spaced-out than H1's – as well as the fact that nearly all of H1's tablets are cut after writing to remove excess clay; this writer therefore had extra space they chose not to fill. Meanwhile, the Ab series (at least some of which, again, were folded) falls mid-way between the two Aa sets in terms of length (mean 17.4 cm), despite having more content to include (the details of rations); H21 chose to do this by ruling a short line at the right-hand end of each tablet to create a two-line space to record the rations (and supervisors), a peculiarity of this series which makes what is presumably the most important part of these records stand out clearly (an effect generally achieved by other writers through spacing out the ideograms and numerals, e.g. frequently in both H1's and H4's Aa series). As discussed above, the Ad series is the only one to include straws/strings (intermittently, and with no clear pattern relating to content or size); these tablets are nearly as long as H4's Aa series (mean length 21.7 cm) and even taller than H1's (mean height 3.3 cm).Footnote 63 This group of related texts therefore suggests a highly individual, even idiosyncratic set of preferences for tablet manufacture and formatting on the parts of their makers and/or writers.

Fig. 13. Personnel tablets. From top to bottom: Aa 792 (H1), Aa 85 (H4), Ab 189 (H21), and Ad 683 (H23). Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

However, we can also see consistency within a single administrative ‘file’ where that includes texts written by more than one person. The Qa series, produced by two different writers whose fairly distinctive tablets are physically indistinguishable, has already been discussed above; the Jn series of page-shaped bronze allocation records, written by H2 and H21, similarly shows no significant differences in tablet manufacture or format between the two writers (cf. LSP, 15–16), or between the different stages of the administrative process (finished tablets, works-in-progress, and discarded tablets) which can be identified within this series.Footnote 64 Whether this means in each case that all of the tablets were made by the same person, that two (or more) makers followed a common set of instructions from the writer(s), or that the two writers made their own tablets in accordance with a shared idea of the appropriate way to create them for this particular purpose remains unknown.

Conversely, in the landholding tablets referring to the site of pa-ki-ja-ne (including H41's preliminary Eb and Eo series; H1's summary En and Ep series; and the totalling Ed series, with contributions by both writers: see Footnote n. 7), we see notable differences in format and production between and even within series of tablets referring to the same location and written by the same person.Footnote 65 Sometimes this is clearly due to administrative reasons: H41's use of a mixture of palm-leaf and page-shaped tablets in the Eo series, compared to their consistent use of palm-leaf tablets in the Eb series, is due to the former including several cases of the same person leasing several plots, each recorded on a single page-shaped tablet (H43 has made a similar administrative choice in writing a single page-shaped tablet, Ea 59, to record multiple landholdings held by the same individual, in a series otherwise consisting entirely of palm-leaves). At other times, this variation may provide evidence for the process followed by the writer, including changes to their preferences or decisions in the course of compiling the ‘file’ of documents. The variation in size and format of H1's page-shaped tablets (whose surface areas range from 92.6 cm2 to 412.8 cm2) is evidently partly related to the number and length of entries to be recorded on each tablet, but may also reflect a changing process of recording: Bennett (Reference Bennett, Heubeck and Neumann1983, 44–7) argued that the variation between vertical and horizontal orientations in the Ep series is due to the writer switching from the former (Ep 301, 613) to the latter (Ep 705, 212, 539, 704 – in the order suggested by Bennett) in order to better accommodate this series’ relatively long entries. Similarly, we have already seen that the inconsistent use of straws/strings in the Eo series is difficult to explain on the basis of the tablets’ clay, size and shape, or potential place of writing; could this be due to the situation changing between making the tablets with straws/strings and those without (or vice versa) – which might have been done at different times – or simply to the maker or writer changing their mind about this series’ requirements?Footnote 66 Although reconstructing the circumstances in which particular tablet series were created and written, and the precise factors underlying the makers’ and writers’ decisions in doing so, is generally not possible, considering questions like this reminds us of the potential complexity of different, possibly changing situations in which even those groups of tablets we now regard as relatively consistent series might have been produced.

CONCLUSIONS

This experimental investigation into the various methods used to produce the Linear B tablets at Pylos has shed significant light on the effects of these methods on the tablets, and thus on the probable reasons behind tablet-makers’ choices of different methods in different circumstances. In particular, creating a tablet by folding up a sheet of clay has been shown to provide increased stability while the clay remains relatively wet – an advantage if the tablet is being handled and/or written on at this stage – while the primary benefit of using a straw/string is in keeping fragments together if the tablet is broken in transit. These results demonstrate that consideration for both the experience of the tablet's future writer (whether that writer was also its maker or not) and for the further administrative processes the tablet would undergo (such as transportation to or within the palace) played a part in the tablet-makers’ decisions as to what process(es) to use when creating a tablet.

However, investigating the relationship between tablets’ production processes and contents in the light of these results has also shown that these decisions are frequently very difficult to reconstruct for any given tablet or group of tablets. Tablet-makers’ and/or writers’ preferences for how tablets were made (and written on) are highly individual, and not always based purely on clearly reconstructable administrative reasons. Although the administrative unity of groups of texts which collectively record a single operation is frequently also shown in their format and/or production process – regardless of how many people were involved in their making and/or writing – these individual idiosyncrasies, together with the potential for circumstances to alter or for makers or writers simply to change their minds, frequently create a more complicated situation. What this reinforces, though, is that the relationship between the making and writing stages of the process was a very close one, as demonstrated by the consistency in tablet production in cases where palm-print evidence shows the involvement of multiple makers, compared to instances where this process changes across a series of documents in accordance with the writer's administrative needs. Regardless of the identity of a tablet's maker in any given case – most likely in some instances writers made their own tablets, while in others this was done by another writer or an assistant – careful consideration of the needs of the writer was inherent in the production of a tablet as the first stage of the Mycenaean administrative recording process.

ACKNOWLEDGEMENTS

This research was carried out during a Marie Skłodowska-Curie Actions fellowship at the British School at Athens, as part of the project ‘Writing at Pylos (WRAP): palaeography, tablet production, and the work of the Mycenaean scribes’. This project has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No. 885977. I would like to thank the staff of the National Archaeological Museum, Athens, in particular Katerina Voutsa, for facilitating access to the Pylos tablets; the staff of the Fitch Laboratory, especially Evangelia Kiriatzi, for providing experimental facilities and advice; and John Bennet and the two anonymous reviewers for their comments.

Footnotes

1 Vitale Reference Vitale2006; Vitale, Stocker and Davis Reference Vitale, Stocker, Davis, Jung and Kardamaki2022. For a refutation of the recent proposal to redate this destruction to an earlier period (ARN, xvii–xix; LSP, 85–91), see Davis et al. Reference Davis, Stocker, Vitale, Bennet, Brecoulaki, Judson, Bennet, Karnava and Meißnerforthcoming.

The following abbreviations are used in this article: ARN = L. Godart and A. Sacconi, Les archives du roi Nestor. Corpus des inscriptions en linéaire B de Pylos, 2 vols (Pisa and Rome, 2019–20); LSP = L. Godart, Les scribes de Pylos (Pisa and Rome, 2021); PT 3 = J.L. Melena, The Pylos Tablets, 3rd edn with R.J. Firth (Leoia, 2021); PTT 2 = J.-P. Olivier and M. Del Freo, The Pylos Tablets Transcribed, 2nd edn (Padua, 2020).

3 Existing tablets could also be either reused by erasing the text and writing a new, palimpsestic text, or recycled by using the clay to form new tablets. However, on recycling see Hruby and Nakassis (Reference Hruby, Nakassis, Bennet, Karnava and Meißnerforthcoming); on palimpsests see section entitled ‘Tablet shape and finishing’, below.

4 As the tablets were originally only air-dried, and then fired when the palace burned down, the exact composition of the clay is less crucial for the purposes of this study than for the experimental recreation of fired pottery.

5 Page-shaped tablets are sometimes further sub-divided by size or orientation (e.g., Driessen Reference Driessen2000, 42; Tomas Reference Tomas, Nosch and Landenius Enegren2017); I prefer not to further sub-divide this already somewhat arbitrary modern classification.

6 On Mycenaean sealing practices, see Younger Reference Younger and Cline2010; Panagiotopoulos Reference Panagiotopoulos2014; on Pylian sealings, see Flouda Reference Flouda2000; Reference Flouda2010; Shelmerdine Reference Shelmerdine, Carlier, de Lamberterie, Egetmeyer, Guilleux, Rougemont and Zurbach2012b. There are also small numbers of ‘noduli’ – small lumps of clay without strings which bear inscriptions and seal impressions – and inscribed but not sealed ‘nodules’. Experimental work on document types other than tablets was outside of the scope of this study; for an experimental investigation of Cretan sealing practices (in a Linear A, rather than Linear B, context), see Finlayson et al. Reference Finlayson, Bogacz, Mara and Panagiotopoulos2021. Some inscriptions also exist on media other than unfired clay: on inscriptions painted before firing on transport stirrup jars, see Judson Reference Judson2013; on the very rare other types of inscriptions, see Pluta Reference Pluta2011, 95–118.

7 The only instance where both shorter preliminary texts (mostly, but not entirely, palm-leaves) and page-shaped summary documents are certainly known is a group of landholding texts from Pylos: Hand (H) 41's Eb and Eo series were compiled in H1's En and Ep series, with both writers contributing to the related totalling Ed series (on the relationships between these series and the administrative process involved, see, e.g., Bennett Reference Bennett1956; Del Freo Reference Del Freo2005, 88–93, 104, 108–10, 120–2, 131–5; Salgarella Reference Salgarella2019; Judson Reference Judson2020b, 538–9); see further below. The page-shaped tablet MY Ue 611, found in the ‘House of Sphinxes’, is a record either of the same delivery of vases represented by the sealings found in the doorway of the same house (Wt 501–507), or of a similar delivery (Müller, Olivier and Pini Reference Müller, Olivier and Pini1998, 15–16; Sacconi Reference Sacconi, Deger-Jalkotzy, Hiller and Panagl1999, 545–6; Shelmerdine Reference Shelmerdine, Deger-Jalkotzy, Hiller and Panagl1999, 571–2; Panagiotopoulos Reference Panagiotopoulos2014, 170–2).

8 The Pylos Jn series, recording bronze allocations, contains page-shaped tablets functioning as both preliminary and final records (Smith Reference Smith1992–3).

9 Driessen Reference Driessen2000, 40. During my autopsy, I observed only a single Pylos tablet, Va 1324, which had been roughly made in this way (which is more characteristic of the labels).

10 This includes tablets found in the Archives Complex during excavations and others which are likely to have been displaced from there (Palaima Reference Palaima1988, 162–9; Shelmerdine Reference Shelmerdine, Bennet and Driessen1998–9, 309 n. 4).

11 Shelmerdine Reference Shelmerdine, Bennet and Driessen1998–9. Inscribed sealings were also found in the Wine Magazine (Shelmerdine Reference Shelmerdine, Carlier, de Lamberterie, Egetmeyer, Guilleux, Rougemont and Zurbach2012b). The tablets from the Megaron have been variously interpreted as evidence for an upper-storey textile workshop or storeroom (e.g. Jasink Reference Jasink1990–1, 228; Kyriakidis Reference Kyriakidis1996–7, 217) and as earlier fragments from wall-fill (e.g. Melena Reference Melena2000–1, 367; Skelton Reference Skelton2010); see Davis et al. Reference Davis, Stocker, Vitale, Bennet, Brecoulaki, Judson, Bennet, Karnava and Meißnerforthcoming.

12 Scraps of clay which may have been intended for making tablets were found in the Archives Complex and Room 23 (Blegen and Rawson Reference Blegen and Rawson1966, 99, 136–7, pl. 276).

13 See, e.g., Hallager Reference Hallager, Nosch and Enegren2017; Wilemon Reference Wilemon2017; Wilemon, Galaty and Nakassis Reference Wilemon, Galaty and Nakassis2020; Hruby and Nakassis Reference Hruby, Nakassis, Bennet, Karnava and Meißnerforthcoming; but cf. LSP, 82. Only one other Messenian site, Iklaina, has produced any Linear B documents. Since the single tablet from this site is dated much earlier than those from Pylos – probably no later than the early/mid-14th century BCE (early LH IIIA2) – it is not clear whether it is the product of an independent administration or of a second-order centre under Pylian control (Shelmerdine Reference Shelmerdine, Carlier, de Lamberterie, Egetmeyer, Guilleux, Rougemont and Zurbach2012a).

14 The identification of Pylos scribal hands is currently the subject of debate: a series of new publications (PTT 2; PT 3; ARN; LSP) each makes various changes to the identifications of Palaima (Reference Palaima1988); for details see Judson Reference Judson2020a. Unless otherwise stated, I use the attribution system of PTT 2 (whose system is also used by PT 3, with only minor differences), as this is more compatible with earlier publications and therefore more user-friendly than that of ARN/LSP; significant differences in attribution between these various works will be discussed where relevant to my argument. ‘-’ indicates that a tablet is unattributed.

15 See Palaima Reference Palaima, Sjöquist and Åström1985; Kyriakidis Reference Kyriakidis1996–7, 203–5. At Knossos, comparisons of the palmprints and scribal hands similarly found a mixture of one-to-one correspondences, associations of multiple prints with a single hand, and vice versa; some prints were also identified as belonging to children and to adults who had carried out rough labour, suggested to be apprentices and assistants respectively (Sjöquist and Åström Reference Sjöquist and Åström1991; Kyriakidis Reference Kyriakidis, Bennet and Driessen1998–9).

16 In general for each experiment I used approximately the same quantity of clay to make tablets of similar size and shape, but with some variation, as is seen in the actual tablets. These tablets’ sizes, for instance, ranged between 19.5–21.1 cm x 3–3.9 cm x 1.2–1.7 cm.

17 During these experiments, writing was done with a pointed metal tool, as I did not yet have access to a replica Mycenaean stylus (extant examples of which are made of bone, and have a flat blade with a curved end: see, e.g., Godart Reference Godart1988, 248–50; Reference Godart and Marazzi1994). A wooden replica of a Mycenaean stylus made for me by Philip Boyes was later used to write the replica tablets shown in my video at www.bsa.ac.uk/videos/how-to-make-a-linear-b-tablet. For a discussion of the impact on writing of using different shapes of styli, see Steele Reference Steele2020.

18 The Aa, Ad, Es, Qa, Sa, and Sh series were inscribed while still fairly wet (Palaima Reference Palaima1988, 46, 70, 79, 87, 91, 121; confirmed by my own autopsy). Of these, the Sh series was the only one in which I could not see clear traces of at least some tablets being folded (though several had traces which might be due to folding, and identification is particularly difficult in this series as several tablets have damaged ends and/or glue covering the verso).

19 These are variously referred to as straws, strings, cords, etc. In some cases, during autopsy I observed horizontal striations along the channel, implying a ridged stalk like a straw (e.g. Eb 364; Sa 1267; Sh 743); in the Eo series, the channels often curve through the tablet, implying a more flexible string. The remains of fibres are also occasionally visible inside the holes (e.g. Sh 734, 736?, 739; Ad 668?). However, as in most cases I could not identify the material used, I refer throughout to ‘straws/strings’. See Palaima Reference Palaima1988, 121; PT 3, xxxix.

20 Most often at both ends of the tablet, occasionally at only one.

21 Cf. Palaima Reference Palaima1988, 87, 91, 98, 121 (though note that the presence of holes is not always consistently noted).

22 Holes observed on all but Eb 886 (damaged) and 976 (fragmentary).

23 Holes observed on all but Sa 1313 (from the North-Eastern Building; complete enough that this feature should be visible if present) and 1561 (fragmentary).

24 Holes observed on all but Eo 160 (page-shaped) and 351 (palm-leaf); 444 (page-shaped) was not certain due to its state of preservation. See further below.

25 On the relationship between these two series, as well as H21's Ab series recording the women's rations, see Chadwick Reference Chadwick, Olivier and Palaima1988, and below.

26 Holes observed only on Ad 664, 668, 679, 683, and 921, and perhaps 357 (the apparent hole on one end may be due to a straw/string or to folding).

27 Cc 660 and 664, records of goats and of sheep and pigs, from the Archives Complex; Cc 1258 and 1284, goat records from the North-Eastern Building.

28 Only 482 is universally attributed to H42; Va 1323 and 1324, records of weapons from the North-Eastern Building, are attributed to H42, and 404 to H42?, only by LSP/PT 3.

29 These include the palm-leaf tablets Ua 407 and Xa 1558, and the page-shaped tablets Un 1482 and Xn 1466.

30 Pace Godart, however, they are certainly not essential to create viable tablets (cf. Driessen Reference Driessen2000, 40).

31 Palaima (Reference Palaima1988, 27) also suggests that they may have ‘facilitated the manufacture of the tablets’; it is unclear to me how incorporating the additional element of a straw/string would do this.

32 For an overview of this practice in the Near East, see ‘Use of seals’, cdliwiki: Educational Pages of the Cuneiform Digital Library Initiative, last modified 20th August 2015 (available online <https://cdli.ox.ac.uk/wiki/doku.php?id=use_of_seals> accessed 19th September 2022); for details of the periods in which this practice is attested, see ‘Diachronic overview of the use of seals in the ANE’, cdliwiki: Educational Pages of the Cuneiform Digital Library Initiative, last modified 16th October 2015 (available online <https://cdli.ox.ac.uk/wiki/doku.php?id=seals_diachronic> accessed 19th September 2022).

33 Between 8 and 13 ‘hanging nodules’ (sealings attached to an object with a string) were found in the Archives Complex, of which only one (Wr 1457) is inscribed (two more, Wp 1415 and Wr 1416, found in the South-Western Area, may have been displaced from the Archives Complex: Shelmerdine Reference Shelmerdine, Bennet and Driessen1998–9, 309 n. 4), while 35–56 (12 inscribed) hanging nodules were found in the North-Eastern Building and 35–42 (four inscribed) in the Wine Magazine (Shelmerdine Reference Shelmerdine, Carlier, de Lamberterie, Egetmeyer, Guilleux, Rougemont and Zurbach2012b, 399–402, table 2; Panagiotopoulos Reference Panagiotopoulos2014, 296–302). The concentration of this type of sealing in two outer buildings used for the receipt and recording of goods, and the fact that many are found deliberately broken and discarded in doorways, supports the view that they were used to authenticate the movement of goods to the palace, rather than at later stages of the administrative process (see Flouda Reference Flouda2000; Shelmerdine Reference Shelmerdine, Carlier, de Lamberterie, Egetmeyer, Guilleux, Rougemont and Zurbach2012b; Panagiotopoulos Reference Panagiotopoulos2014, 189–210).

34 Panagiotopoulos (Reference Panagiotopoulos2014, 59–60, 190–1, 253–4) points out that one hanging nodule in the Archives Complex – which was, unusually, not broken, implying it may still have been an ‘active’ document at the time of the destruction – was found in close proximity to a tablet with a straw/string-hole (Eb 169; sealing 38A), while Eb 1176 and sealing 30 (also unbroken) were both found in the Propylon. However, the latter, at least, have certainly been displaced from their original locations within the Archives Complex, and the other tablets cited as being found close to unbroken sealings do not have straw/string holes. Panagiotopoulos’ (Reference Panagiotopoulos2014, 253–4) alternative suggestion that straws/strings were used to attach tablets such as the Sa and Sh series to the objects they referred to is contradicted by the evidence for the latter being transported in a basket to the Archives Complex while still very wet (see ‘Where were tablets made?’, above).

35 The straws used in my experiments were stalks of einkorn wheat (Triticum monoccocum), of varying thicknesses (where using a thicker or thinner straw had a significant impact on results, this is noted); the string was commercially available twisted string.

36 Ad 683; Eb 169, 321, 339, 347, 369, 416, 473, 477, 495, 498, 566, 818, 846?, 871, 874, 890, 893, 895, 905, 915, 985, 993, 1176, 1440; Eo 211, 268, 269, 471; Sa 751, 755, 758, 760, 763, 766, 768, 787, 843, 1265, 1266, 1267; Sh 735, 739, 743, 744; Va 482.

37 Eb 566, 893, 993; Eo 211, 268, 269; Sa 751, 755, 760, 763, 766, 768, 787, 1267; Sh 743.

38 Cf. cases where the channel is exposed in the middle of the tablet due to clay loss (e.g. Cc 1285 verso; Eo 224.2, 278 lat. inf., 371.B; Sa 403, 483, 1267; Sh 740: Fig. 7) or where a crack has developed in the clay on top of the channel (e.g. Sh 737, 740, 741; Eb 818 verso).

39 Cf. Hallager's (Reference Hallager, Nosch and Enegren2017) similar experiment, though this did not incorporate straws/strings and was intended to test the ability of tablets to survive transportation without breaking.

40 Several weeks later, on taking the tablets to the lab to recycle them, I noticed that the two straws in the broken tablets had snapped, perhaps because by this point they had more completely dried out in the increasingly hot weather – another factor which tablet-makers might have needed to take into account.

41 As noted above, the North-Eastern Building's single Sa series tablet and two Va series tablets attributed (tentatively) to H42 all lack straw/string holes, in contrast to the other members of each series found in the Archives Complex.

42 The Sa and Sh series and Va 404 and 482 presumably refer to items and activities in or around the palace, since they contain no indications of their location. Cc 1283 is fragmentary; while Cc 1285 refers to sheep going ma-se-de ‘to ma-se’, their original location is not specified.

43 For a summary of Pylian geography, see Bennet Reference Bennet, Duhoux and Davies2011, 151–5.

44 Twenty-four Ad series tablets refer to locations in the Hither Province, of which 679, 684 and perhaps 357 had straws/strings; 12 tablets refer to Further Province sites, of which 664, 668 and 921 had straws/strings. See Chadwick Reference Chadwick, Olivier and Palaima1988, 47–59.

45 As recorded on Eq 213, which refers to the official Alksoitas – possibly to be identified as the scribe H1 – travelling to collect landholding information (Kyriakidis Reference Kyriakidis1996–7, 220–4; Bennet Reference Bennet, Voutsaki and Killen2001, 31).

46 The Ad series’ length ranges from 16.8–27.3 cm; all the tablets with string/straw holes (see Footnote n. 26) are in the upper half of this range (22.3–27.3 cm), but so are many without holes (e.g. Ad 326, 26.2 cm long). In the Eo series, in addition to there being both page-shaped and palm-leaf tablets with and without strings (see Footnote n. 24), there is no clear correlation with size in either category. The one page-shaped tablet which certainly lacks a hole, Eo 160, is the shortest and narrowest of this format (17.7 x 5.1 cm) – the others are 18.5–21 cm long (excluding broken tablets) and 6.6–8.6 cm wide – but the one palm-leaf tablet without a string, Eo 351, is the widest of the palm-leaves at 3.9 cm (the others are 2.9–3.5 cm wide; as almost all the palm-leaves are broken the lengths cannot be compared).

47 Compare, for instance, Eo 351 – a palm-leaf tablet recording the amount of land held by a-da-ma-o and a single lease of part of this land – with the structurally identical, but longer, page-shaped text Eo 247, recording the amount of land held by a3-ti-jo-qo and six leases of this land.

48 E.g. Jn 389 and 415 (H2) are two halves of a large tablet cut before writing, as observed by Melena (Reference Melena1992–3, 82 n. 8), and confirmed by my own autopsy (pace ARN q.vv.).

49 Ep 212, 539, and 705 (H1) are, unusually, cut on three sides after writing.

50 Ab 558; a further five possible Ab series examples.

51 Ea 421, 803, 811, 817, 825; a further eight possible Ea series examples.

52 Eb 149, 842, 1188; a further 10 possible Eb series examples.

53 Prints distinct from the above but identified on only one tablet each: ‘Anon II’: Fr 1205 (H2), ‘Anon III’: Fr 1229 (-), ‘Anon IV’: An 1281 (H12), ‘Anon V’: Fr 1217 (H18), ‘Anon VI:’ Va 1323 (H42?: see Footnote n. 28), ‘Anon VII’: Ad 318 (H23).

54 Note that LSP merges H15 with H14 and H33 with H23 (PT 3 also follows the latter), but following this, two different writers would still be responsible for the Qa series.

55 For photographs of tablets not illustrated in this article, see ARN or Judson, Meißner and Thompson Reference Judson, Meißner and Thompson2016.

56 Unless LSP/PT 3's merger of H23 with H33 (Qa 1289, 1300, 1305) and additions of Fa 16 (otherwise H42), Fa 1195 (-), and Ua 434 (H42) are followed.

57 Tablets without straws/strings: H21: Ab series, some Cc series (as listed above), Fg 368 (and Fr 374?, H21(?) in LSP/PT 3 only). H41: Ed 411 (.1 and probably [[.2]] by this hand), Fr 1207. H42: Ae series, Fa 16? (see n. 56), Ja 1288? (H42 in LSP/PT 3 only), Ua series, Va 1323 and 1324 (see Footnote n. 28).

58 Fr 1240 and 1242 are broken, but the former resembles 1217 while the latter appears to be a single-entry palm-leaf. Fr 1223, a two-line two-entry palm-leaf, is also attributed to this hand by LSP/PT 3.

59 Exceptions: Fr 1184 (H2), Archives Complex (a perfume-manufacture record); Fr 1251 (H2) in Court 63 (dropped between the oil storerooms and the Archives Complex?: Palaima Reference Palaima1988, 169); Fr 1338 (H19/34?, LSP/PT 3; - in others) and 1355 (H30, PT 3; - in others), near Room 103; Fr 1255 (H30, PT 3; - in others), Rooms 71–2.

60 Many of the Fr series tablets are identified in all recent editions as possible palimpsests, including several on which one of the two lines has been left blank (Fr 1216, 1219, 1226, 1232, 1236). Fr 1218, however, which has two inscribed and four blank lines, is not identified as such. Such partial uses of a tablet's writing space might be due to circumstances changing in the course of its inscription as much as to selection of an overly large existing tablet.

61 On the interpretation of the abbreviations DA and TA as standing for male and female supervisors, see Chadwick Reference Chadwick, Olivier and Palaima1988, 71–3 (with further references).

62 The corresponding Ad 290 shows that these women are located in the Further Province site of Leuktron.

63 Ad 684 is often said to have originally been made for H4's Aa series on the basis of its size and shape (Palaima Reference Palaima1988, 87, 89; ARN q.v.; PT 3, xxxv). Presumably this refers to 684's height (2.3 cm), which is shorter than any of the other Ad series (2.6–4.4 cm) but only just outside the range of H4's Aa series (2.4–2.7 cm); in length (19.4 cm) and thickness (1.4 cm), 684 falls within or very close to the ranges of both series (Ad: length 16.8–27.3 cm, thickness 1.0–2.9 cm; H4 Aa: length 21.2–24.5 cm, thickness 1.3–1.8 cm [note that these measurements exclude Aa 96, whose attribution to H1 or H4 is debated]). However, the Ad series shows a very wide variation in all three dimensions, as well as in shape, while H4's Aa series is more consistent. I therefore do not find this argument convincing.

64 Smith Reference Smith1992–3. The Jn series tablets do vary in size and shape, largely due to some originally larger tablets being cut to produce smaller ones, but not in a way that correlates with scribal hand. A far bigger contrast is seen between this series and Jo 438, also a page-shaped tablet containing a large number of entries relating to metal (in this case gold), than within the Jn series. Jo 438 is a very tall, narrow, and thick vertical tablet measuring 23.2 x 8.3 x 2.5 cm; the Jn series’ largest tablet, Jn 829, is 22 x 11.2 x 2.5 cm, and many have a much squarer shape and a thickness of 1.1–2.1 cm.

65 Ed 411, which was begun by H41 and continued by H1 (whereas the rest of the series is entirely by H1), has been suggested to have been made by a different person – usually assumed to be H41 – based on its distinctly different height (4.3 cm, compared to the others’ 3.3–3.6 cm: Palaima Reference Palaima1988, 38; PT 3, xxxv).

66 It has previously been suggested that the surviving Eo series represents two phases of writing, with the page-shaped tablets being ‘secondary’ texts copied from now non-existent ‘primary’ palm-leaf tablets similar to the surviving Eo series palm-leaves (Bennett Reference Bennett, Heubeck and Neumann1983, 42; Salgarella Reference Salgarella2019, 75, 80–1). As noted above, however, this division by tablet format does not correspond to the presence or absence of straws/strings.

References

REFERENCES

Bendall, L.M. 2003. ‘A reconsideration of the Northeastern Building at Pylos: evidence for a Mycenaean redistributive center’, AJA 107.2, 181231.CrossRefGoogle Scholar
Bennet, J. 2001. ‘Agency and bureaucracy: thoughts on the nature and extent of administration in Bronze Age Pylos’, in Voutsaki, S. and Killen, J.T. (eds), Economy and Politics in the Mycenaean Palace States: Proceedings of a Conference Held on 1–3 July 1999 in the Faculty of Classics, Cambridge (Cambridge), 2537.Google Scholar
Bennet, J. 2011. ‘The geography of the Mycenaean kingdoms’, in Duhoux, Y. and Davies, A. Morpurgo (eds), A Companion to Linear B. Mycenaean Greek Texts and their World (Louvain-la-Neuve and Walpole, MA), 137–68.Google Scholar
Bennett, E.L. 1956. ‘The landholders of Pylos’, AJA 60.2, 103–33.CrossRefGoogle Scholar
Bennett, E.L. 1983. ‘Pylian landholding jots and tittles’, in Heubeck, A. and Neumann, G. (eds), Res Mycenaeae. Akten des VII. internationalen mykenologischen Colloquiums in Nürnberg vom 6.–10. April 1981 (Göttingen), 4154.Google Scholar
Bennett, E.L. 1996. ‘The Linear B script: some subliminal elements’, in De Miro, E., Godart, L. and Sacconi, A. (eds), Atti e memorie del secondo congresso internazionale di micenologia. Roma–Napoli, 14–20 ottobre 1991 (Rome), 2532.Google Scholar
Blegen, C.W. and Rawson, M. 1966. The Palace of Nestor at Pylos in Western Messenia, vol. 1: The Buildings and their Contents. Part 1: Text (Princeton, NJ).Google Scholar
Chadwick, J. 1988. ‘The women of Pylos’, in Olivier, J.-P. and Palaima, T.G. (eds), Texts, Tablets and Scribes. Studies in Mycenaean Epigraphy and Economy Offered to Emmett L. Bennett, Jr. (Salamanca), 4395.Google Scholar
Davis, J.L., Stocker, S.R., Vitale, S., Bennet, J., Brecoulaki, H. and Judson, A.P. forthcoming. ‘The date of the final destruction of the Palace of Nestor at Pylos’, in Bennet, J., Karnava, A. and Meißner, T. (eds), KO-RO-NO-WE-SA: Proceedings of the 15th Mycenological Colloquium 21–24 September 2021, Athens (Crete).Google Scholar
Del Freo, M. 2005. I censimenti di terra (Pisa and Rome).Google Scholar
Del Freo, M. 2019. ‘I documenti in lineare B’, in Freo, M. Del and Perna, M. (eds), Manuale di epigrafia micenea. Introduzione allo studio dei testi in lineare B, 2nd edn (Padua), 169–84.Google Scholar
Driessen, J. 2000. The Scribes of the Room of the Chariot Tablets at Knossos. Interdisciplinary Approach to the Study of a Linear B Deposit (Salamanca).Google Scholar
Finlayson, S., Bogacz, B., Mara, H. and Panagiotopoulos, D. 2021. ‘Searching for ancient Aegean administrators: computational experiments on identical seal impressions’, JAS 136, no. 105490.Google Scholar
Flouda, G. 2000. ‘Inscribed Pylian nodules: their use in the administration of the storerooms of the Pylian palace’, SMEA 42.2, 213–45.Google Scholar
Flouda, G. 2010. ‘Agency matters: seal-use in Pylian administration’, OJA 29.1, 5788.Google Scholar
Godart, L. 1988. ‘Autour des textes en linéaire B de Tirynthe’, AA, 245–51.Google Scholar
Godart, L. 1994. ‘Gli stili degli scribi di Tirinto’, in Marazzi, M. (ed.), La società micenea (Rome), 296–9.Google Scholar
Greco, A. and Flouda, G. 2017. ‘The Linear B pa-i-to Epigraphic Project’, ASAtene 95, 143–60.Google Scholar
Hallager, E. 2017. ‘LM IIIB Linear B from Knossos and Khania’, in Nosch, M.-L. and Enegren, H. Landenius (eds), Aegean Scripts. Proceedings of the 14th International Colloquium on Mycenaean Studies, Copenhagen, 2–5 September 2015 (Rome), 275–84.Google Scholar
Hruby, J.A. and Nakassis, D. forthcoming. ‘Reduce, reuse, recycle? The clay of the Pylos tablets’, in Bennet, J., Karnava, A. and Meißner, T. (eds), KO-RO-NO-WE-SA: Proceedings of the 15th Mycenological Colloquium 21–24 September 2021, Athens (Crete).Google Scholar
Jasink, A.M. 1990–1. ‘Funzionari e lavoranti nel palazzo di Pilo’, Minos 25–6, 203–43.Google Scholar
Judson, A.P. 2013. ‘The Linear B inscribed stirrup jars’, Kadmos 52.1, 69110.CrossRefGoogle Scholar
Judson, A.P. 2020a. ‘Concordance of Pylos scribal attributions’, last modified 12 July 2021 (available online <https://hcommons.org/docs/judson-copy/> accessed April 2022).+accessed+April+2022).>Google Scholar
Judson, A.P. 2020b. ‘Scribes as editors: tracking changes in the Linear B documents’, AJA 124.4, 523–49.CrossRefGoogle Scholar
Judson, A.P., Meißner, T. and Thompson, R.J.E. 2016. ‘CaLIBRA: Cambridge Linear B Research Archive’, last modified 2015 (available online <http://calibra.classics.cam.ac.uk/> accessed April 2022).+accessed+April+2022).>Google Scholar
Kyriakidis, E. 1996–7. ‘Some aspects of the role of scribes in Pylian palace administration’, Minos 31–2, 201–29.Google Scholar
Kyriakidis, E. 1998–9. ‘Scribes treated as criminals: a note on the study of palm and fingerprints on the Linear B tablets of Knossos’, in Bennet, J. and Driessen, J. (eds), A-NA-QO-TA: Studies Presented to J.T. Killen (Salamanca), 197205.Google Scholar
Lejeune, M. 1997. ‘Analyse du dossier pylien Ea’, in Lejeune, M. (ed.), Mémoires de philologie mycénienne. Quatrième série (1969–1996) (Rome), 117–51. Originally published in Minos 15 (1977), 81115.Google Scholar
Melena, J.L. 1992–3. ‘167 joins of fragments in the Linear B tablets from Pylos’, Minos 27–8, 7182.Google Scholar
Melena, J.L. 2000–1. ‘24 joins and quasi-joins of fragments in the Linear B tablets from Pylos’, Minos 35–6, 357–69.Google Scholar
Müller, W., Olivier, J.-P. and Pini, I. 1998. ‘Die Tonplomben aus Mykene’, AA, 555.Google Scholar
Nakassis, D., Pluta, K. and Hruby, J.A. 2021. ‘The Pylos Tablets Digital Project: prehistoric scripts in the 21st century’, in Cooper, C.L. (ed.), New Approaches to Ancient Material Culture in the Greek & Roman World. 21st-Century Methods and Classical Antiquity (Leiden and Boston, MA), 161–71.Google Scholar
Palaima, T.G. 1985. ‘Appendix’, in Sjöquist, K.-E. and Åström, P. (eds), Pylos: Palmprints and Palmleaves (Gothenburg), 99107.Google Scholar
Palaima, T.G. 1988. The Scribes of Pylos (Rome).Google Scholar
Palaima, T.G. 1994. ‘Seal-users and script-users / nodules and tablets at LM I B Haghia Triada’, in Ferioli, P., Fiandra, E., Fissore, G.G. and Frangipane, M. (eds), Archives before Writing. Proceedings of the International Colloquium. Oriolo Romano, October 23–25, 1991 (Rome), 307–37.Google Scholar
Palaima, T.G. 1996a. ‘“Contiguities” in the Linear B tablets from Pylos’, in De Miro, E., Godart, L. and Sacconi, A. (eds), Atti e memorie del secondo congresso internazionale di micenologia. Roma–Napoli, 14–20 ottobre 1991 (Rome), 379–96.Google Scholar
Palaima, T.G. 1996b. ‘Sealings as links in an administrative chain’, in Ferioli, P., Fiandra, E. and Fissore, G.G. (eds), Administration in Ancient Societies. Proceedings of Session 218 of the 13th International Congress of Anthropological and Ethnological Sciences. Mexico City, July 29–August 5, 1993 (Turin), 37–66, discussion 103–6.Google Scholar
Palaima, T.G. 2010. ‘Linear B’, in Cline, E.H. (ed.), The Oxford Handbook of the Bronze Age Aegean (Oxford), 356–72.Google Scholar
Palaima, T.G. 2011. ‘Scribes, scribal hands and palaeography’, in Duhoux, Y. and Davies, A. Morpurgo (eds), A Companion to Linear B. Mycenaean Greek Texts and their World (Louvain-la-Neuve and Walpole, MA), 33136.Google Scholar
Palaima, T.G. and Wright, J.C. 1985. ‘Ins and outs of the Archives Rooms at Pylos: form and function in a Mycenaean palace’, AJA 89.2, 251–63.CrossRefGoogle Scholar
Panagiotopoulos, D. 2014. Mykenische Siegelpraxis. Funktion, Kontext und administrative Verwendung mykenischer Tonplomben auf dem griechischen Festland und Kreta (Munich).Google Scholar
Pape, T. 2002. ‘Setting the records straight: what implications will the working properties of clay tablets have upon our understanding of the erasures and corrections found within the Linear B administrative records of the Mycenaean palaces?’ (unpublished BSc thesis, University of Sheffield).Google Scholar
Pape, T., Halstead, P., Bennet, J. and Stangidis, Y. 2014. ‘“For it is written”: an experimental approach to the materiality and temporality of clay documents inscribed in Linear B’, in Galanakis, Y., Wilkinson, T. and Bennet, J. (eds), ΑΘΥΡΜΑΤΑ: Critical Essays on the Archaeology of the Eastern Mediterranean in Honour of E. Susan Sherratt (Oxford), 177–86.CrossRefGoogle Scholar
Pluta, K.M. 2011. ‘Aegean Bronze Age literacy and its consequences’ (unpublished PhD thesis, University of Texas at Austin).Google Scholar
Sacconi, A. 1999. ‘Quelques remarques sur les séries W- des textes en linéaire B’, in Deger-Jalkotzy, S., Hiller, S. and Panagl, O. (eds), Floreant Studia Mycenaea. Akten des X. internationalen mykenologischen Colloquiums in Salzburg vom 1.–5. Mai 1995 (Vienna), 543–8.Google Scholar
Salgarella, E. 2019. ‘Non-connective behaviour of the particle -qe in the Linear B documents from Pylos’, SMEA n.s. 5, 7184.Google Scholar
Shelmerdine, C.W. 1985. The Perfume Industry of Mycenaean Pylos (Gothenburg).Google Scholar
Shelmerdine, C.W. 1998–9. ‘The Southwestern Department at Pylos’, in Bennet, J. and Driessen, J. (eds), A-NA-QO-TA: Studies Presented to J.T. Killen (Salamanca), 309–37.Google Scholar
Shelmerdine, C.W. 1999. ‘A comparative look at Mycenaean administration(s)’, in Deger-Jalkotzy, S., Hiller, S. and Panagl, O. (eds), Floreant Studia Mycenaea. Akten des X. internationalen mykenologischen Colloquiums in Salzburg vom 1.–5. Mai 1995 (Vienna), 555–76.Google Scholar
Shelmerdine, C.W. 2012a. ‘Iklaina tablet IK X 1’, in Carlier, P., de Lamberterie, C., Egetmeyer, M., Guilleux, N., Rougemont, F. and Zurbach, J. (eds), Études mycéniennes 2010. Actes du XIIIe colloque international sur les textes égéens. Sèvres, Paris, Nanterre, 20–23 septembre 2010 (Pisa and Rome), 75–7.Google Scholar
Shelmerdine, C.W. 2012b. ‘Pylos sealings and sealers’, in Carlier, P., de Lamberterie, C., Egetmeyer, M., Guilleux, N., Rougemont, F. and Zurbach, J. (eds), Études mycéniennes 2010. Actes du XIIIe colloque international sur les textes égéens. Sèvres, Paris, Nanterre, 20–23 septembre 2010 (Pisa and Rome), 383402.Google Scholar
Sjöquist, K.-E. and Åström, P. 1985. Pylos: Palmprints and Palmleaves (Gothenburg).Google Scholar
Sjöquist, K.-E. and Åström, P. 1991. Knossos: Keepers and Kneaders (Gothenburg).Google Scholar
Skelton, C. 2010. ‘Re-examining the Pylos Megaron tablets’, Kadmos 48, 107–23.CrossRefGoogle Scholar
Smith, J.S. 1992–3. ‘The Pylos Jn series’, Minos 27–8, 167259.Google Scholar
Steele, P.M. 2020. ‘Material entanglements of writing practices in the Bronze Age Aegean and Cyprus’, Sustainability 12, no. 10671.CrossRefGoogle Scholar
Tomas, H. 2013. ‘Saving on clay: the Linear B practice of cutting tablets’, in Piquette, K.E. and Whitehouse, R.D. (eds), Writing as Material Practice: Substance, Surface and Medium (London), 175–91.CrossRefGoogle Scholar
Tomas, H. 2017. ‘From Minoan to Mycenaean elongated tablets: defining the shape of Aegean tablets’, in Nosch, M.-L. and Landenius Enegren, H. (eds), Aegean Scripts. Proceedings of the 14th International Colloquium on Mycenaean Studies, Copenhagen, 2–5 September 2015 (Rome), 115–26.Google Scholar
Vitale, S. 2006. ‘The LH IIIB–LH IIIC transition on the Mycenaean mainland: ceramic phases and terminology’, Hesperia 75.2, 177204.CrossRefGoogle Scholar
Vitale, S., Stocker, S.R. and Davis, J.L. 2022. ‘The destructions of the Palace of Nestor at Pylos and its LH IIIA predecessor as a methodological case study’, in Jung, R. and Kardamaki, E. (eds), Synchronizing the Destructions of the Mycenaean Palaces (Vienna), 121–48.Google Scholar
Wilemon, B.B. 2017. ‘Portable x-ray fluorescence spectrometer analysis of the Pylos Linear B tablets’ (unpublished MA thesis, Mississippi State University).Google Scholar
Wilemon, B.B., Galaty, M.L. and Nakassis, D. 2020. ‘Portable x-ray fluorescence spectrometer analysis of the Pylos Linear B tablets’, paper delivered at the Archaeological Institute of America Annual Meeting, 3 January, Washington, DC.Google Scholar
Younger, J.G. 2010. ‘Mycenaean seals and sealings’, in Cline, E.H. (ed.), The Oxford Handbook of the Bronze Age Aegean (Oxford), 329–39.Google Scholar
Figure 0

Fig. 1. Illustration of tablet shapes: palm-leaf (Fr 1203, top left); vertical page-shaped (An 1, right); horizontal page-shaped (Es 647, bottom left). Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.). Scale bar: adapted from photographic reference scale by Jim Elder, Ottawa, Canada (smallpond.ca/jim/scale), CC-BY-NC-SA 4.0.

Figure 1

Fig. 2. The author rolling out a cylinder of clay to form a tablet in the Fitch Laboratory. Photo: Evangelia Kiriatzi.

Figure 2

Fig. 3. Jn 605 lat. inf. Photo: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

Figure 3

Fig. 4. Ma 123, lat. sin. and verso. The fold is clearly visible on the lat. sin. but there is only a trace of the seam at the left-hand end of the verso. Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

Figure 4

Fig. 5. The author writing on drying tablets to test their consistency. Photo: Emily Sherriff.

Figure 5

Fig. 6. Experimental tablets made of fine clay (left) and coarse clay (right) showing effects of holding by the middle while wet. Top: rolled; bottom: folded. Photos: author.

Figure 6

Fig. 7. Eb 149 lat. dex., with straw/string hole (left); Sh 740 recto, with partially exposed straw/string channel (right). Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

Figure 7

Fig. 8. Experimental tablets showing effects of holding by the middle while wet. From top to bottom: rolled, without straw; rolled, with straw; folded, without straw; folded, with straw. Photo: author.

Figure 8

Fig. 9. Broken tablets containing straws/strings on their return to Athens. Photo: author.

Figure 9

Fig. 10. Ad 683 recto and lat. inf. Note the tapering right-hand end and the curvature of the verso at both ends. Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

Figure 10

Fig. 11. Jn 415 verso showing folds and ridges of clay along the bottom edge and sides (the top has been cut). Photo: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

Figure 11

Fig. 12. H18's Fr series tablets. From top to bottom: Fr 1225, 1218, 1217. Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).

Figure 12

Fig. 13. Personnel tablets. From top to bottom: Aa 792 (H1), Aa 85 (H4), Ab 189 (H21), and Ad 683 (H23). Photos: National Archaeological Museum, Athens/ Department of Collections for Prehistoric, Egyptian, Cypriot and Near Eastern Antiquities. © Hellenic Ministry of Culture and Sports/ Organization of Cultural Resources Development (H.O.C.RE.D.).