From Alan Bowman and David Thomas, The Vindolanda Writing Tablets
(Tabulae Vindolandenses II), London: British Museum Press,
1994. pp. 47-49
we devoted a chapter of the introduction (Vol.
I Ch. 4) to a detailed discussion of various aspects of the
palaeography of the tablets there published. Our main purpose was
to indicate all features of palaeographical interest (especially
in the letter-forms). In addition we attempted to relate the new
information which the tablets provided to the known palaeographical
background, in particular to the vexed question of the development
of Old Roman Cursive (ORC) into New Roman Cursive (NRC).1
We do not propose to consider this problem further here, other than
to refer to Tjäder (1979). In this important article Tjäder
suggests that Roman script had developed by at any rate the first
century AD two different variants, an official script (the one used
in military documents) and a private script. He believes that the
official script did not lead to any further developments, and that
it is to the private script that we must look for those features
which were to be significant for the change to NRC in the third
century. He further suggests that alongside these two scripts there
also existed a "popular" script, written by semi-literates, which
influenced the private script and so contributed to the development
Since Tab.Vindol.I was published a new handbook of Latin palaeography has appeared with a brief discussion of Latin handwriting in the classical period.3 There have been sporadic publications of individual Latin papyri and ostraca from Egypt written in ORC,4 as well as an edition of fragments of papyri and ostraca in Latin recovered at Masada, which are close in date to our tablets.5
Special mention should be made of two publications of extensive collections of material written, mostly or entirely, in ORC and which are not from places in the eastern half of the Empire: the ostraca from Bu Njem and the curse tablets from Bath. In both cases the editors have accompanied their editions with very detailed discussions of the palaeography, especially the letter-forms.6 Unfortunately neither of these collections offers close parallels for the Vindolanda material. It is obvious that the writers of the curse tablets (which in any case are somewhat later in date than our tablets) will have used a different technique for inscribing lead from that used for writing on wood (even though they demonstrably were attempting to use the same basic letter-forms). Writers in ink on ostraca will no doubt have used much the same technique as the writers of our tablets, but the Bu Njem collection dates from the middle of the third century, by which time ORC was noticeably different from its various forms around AD 100. Also worth mentioning is the edition of painted inscriptions found in SE Spain, since these inscriptions are mostly in ORC and here also the editors give a thorough discussion of the palaeography of their material.7 Again, however, the technique used must have differed somewhat from that used for writing with a pen on wood.8
The best example of this in the tablets is the line of Virgil (118; Plate I). The letter-forms resemble those in the script usually known as Capitalis Rustica.9 This is especially true of A, with the elaborate thick stroke at the foot of the left-hand diagonal. Note also the forms of R, V and D. P is not in the usual capital form, though such a form, which is much closer to that of ORC and lacks the bow, is sometimes found in capital scripts of this period. For the very remarkable form of E see the introduction to this tablet.
Capital letter-forms are to be seen in the few letters preserved in 119, in 121 (here mixed with cursive forms), in 162 and 163, certainly military documents of some kind, in the consular date of 186 and, much the most surprisingly, in the lines in the margin of 206. This text appears to be a normal account and yet has these two lines in capital script added. It was of course common for military documents to be written in a mixture of capital and cursive scripts (a recently published example is Doc.Masada 722), the capital script being used for the most part in headings. We need not therefore be surprised that capital script is used for the date in 186. But these instances are in no way a reasonable parallel for 206.
The above documents are all more or less bilinear and, apart from 118, the individual letter forms need little comment. Note the long tail to Q in 119, the rather crude forms in 186 (where R is close to the cursive form), the tiny O in 163, and the use of serifs here and in 162; this last document shows forms approaching those of Capitalis Rustica. The form of S in 206 is remarkable: it is little more than a straight diagonal stroke.
We have not attempted to classify the various types of ORC which appear in the tablets in the way that others have done.10 We would, however, emphasise the great variety of scripts occurring. The military documents are on the whole written in rather similar versions of ORC, competent hands of what may be called a normal type. This is only what we should expect. But even here there are surprises, notably the remarkably elegant 152.11 The letters, again as would be expected, show a much greater variety of scripts. Some are remarkably elegant, whereas others are much cruder (contrast, for example, the elegance of 291 with the clumsiness of 310). One notable feature is the way some scripts give an impression of bilinearity (e.g. 302), whereas many others, though using what are essentially the same letter-forms, have exaggerated ascenders and descenders (e.g. 233, 248, 265). Most are written in an upright fashion, but others slope to the right (e.g. 213 and 215). Some writers seem to eschew ligatures altogether; most use ligature very occasionally, but in a few texts it is common (e.g. 213, 225, 313, 343). Often the scripts which are palaeographically the most interesting, since the most idiosyncratic, are those in which the sender of a letter has appended the final greeting in his or her own hand; or those which are mere drafts of which a professional scribe would no doubt later have produced a fair version.
The palaeographical information provided by the new tablets has not caused us to modify our view that the evidence from Vindolanda strongly suggests that Latin writing throughout the empire in about AD 100 was all much the same; indeed the new information merely serves to strengthen our previous impression.
In our earlier publication we examined in detail the handwriting found in those tablets written in ORC (the vast majority), considering in particular all the various letter-forms and the use of ligatures. We have not felt it necessary to repeat here all that was said before. The comments which follow are to be understood as complementary to those made in Tab.Vindol.I, superseding them and adding to them where the new material gives significant new information or causes us to modify our earlier views.