INTRODUCTION

The main linguistic interest of the four tablets published in this volume (Bowman, Thomas and Tomlin Reference Bowman, Thomas and Tomlin2019) is lexical, but there are a few noteworthy syntactic features. Some additions to the lexica have turned up, and also some established terms display new semantic developments. Several usages are relevant to the Romance languages. One or two terms remain problematic.

The misspellings in the new texts that reflect pronunciation rather than the rules of grammarians are usually banal. It must be stressed that, while a misspelling may be described as a ‘vulgar spelling’ if one wishes to use such terminology, many misspellings in fact reflect educated or general pronunciations, not uneducated speech, and are therefore mere literacy errors rather than pointers to the speech characteristics of a particular social or regional group. For example, the assimilation of two similar vowels in hiatus into a single vowel is a phenomenon found in high poetry, and any misspelling that reflects the assimilation is merely an indication of a general speech development and not of any peculiarity of the poorly educated. There is, however, one misspelling in the tablets that derives from a pronunciation that there is reason to believe was long resisted by the educated (ct > t: see next section).

Some types of deviant or non-standard syntax and morphology will be explained below, but a good deal of the syntax is correct by classical standards. ‘Correctness’ may in part be due to the formulaic character of parts of the letters, and also to the part played by scribes, who were presumably chosen for their education.

Since the corpus published here is small, I have not arranged the material according to linguistic topics, but have discussed each letter in turn.

890 (Inv. no. WT 2017.15)

Cumprimum ‘as soon as’ is written as a single word. The cum in this combination is of course the conjunction cum (rather than the preposition), which has not been dealt with yet by the TLL (the lemma will have the form quom). The preposition cum is sometimes written as a single word with primis (cumprimis: TLL IV.1380, comparing inprimis), and even if cumprimum happened to be otherwise unattested, such a spelling might well have been influenced by the prepositional term.

After mitte there is now read .aṇ. paṛṭem, followed by ueṭụram quam hodie ad uos dimisi. The letters before partem may (the editors suggest) be part of mane. Veturam can only be uecturam, a term that has come up already in Vindolanda tablets (III.600, 649), with a non-classical meaning (‘wagon’) reflecting a change from abstract (in classical Latin the term expresses the act of conveying goods or passengers: OLD s.v. 1) to concrete, and anticipating the sense of the word as it survives in Gallo-Romance (cf. Fr. voiture) (see Adams Reference Adams2003, 559, with bibliography). The assimilation of ct > t seen in this example was markedly substandard (see Adams Reference Adams2013, 166–9).

Partem must form a partitive apposition with ueturam, which would entail giving a slightly different meaning to the latter. It would mean, not ‘wagon’, but ‘wagon-load’. Such a modification would be slight. ‘Send a wagon of vegetables’, for example, is readily interpreted as ‘send a wagon-load’.

Pars is often in (partitive) apposition to a noun in the same case, usually (in the standard manner of partitive appositions) with the other noun placed first. For a collection of examples see TLL X.1.454.17ff. For the usual order see e.g. Livy 26.33.14, senatus iuratus maxima pars, ‘the senate under oath, the greatest part’, > ‘the majority of the senate under oath’. However, the order is sometimes reversed, as is noted at TLL X.1.454.18 and illustrated at some length by Norberg (Reference Norberg1944, 6 n. 1). The earliest example with such reversal is Plaut., Capt. 232, maxuma pars … homines (cf. also Vitr. 9.2. 2, dimidia parte caeli spatio ‘a half part (of) the space of the heavens’; in late Latin examples are more common). A suggested sense of the Vindolanda example would be: ‘as soon as possible send off part of the wagon-load that I have today dispatched to you all.’ Pars might even mean ‘half’. For pars without a specifying adjective = ‘half’, see TLL X.1.452.79ff.

A second person plural (ad uos) used to a single addressee implies ‘you and the others associated with you’: see Tab. Vindol. III.622.b.2, 643.back.3, Adams (Reference Adams2003, 550) (add Brutus ap. Cic., Fam. 11.26, rogo te, uidete quibus hominibus negotium detis). I assume that the writer Verecundus sent a wagon-load to the addressee Audax, which was meant to be shared with others (uos). Audax now has to dispatch part of the load to the others.

Caballus continues to turn up in these tablets as the term for ‘horse’, with equus not yet attested (see Adams Reference Adams2003, 563–4). This is a significant lexical feature, as it shows that by the early second century the word that was to produce the term for ‘horse’ in all Romance languages was already the norm in everyday Latin, and was not pejorative as were the few examples in high literature.

The word with caḅallis duobus has caused problems, but it is now agreed by the editors to be the participle s[o]lutis. Caballis duobus solutịs raises the greatest problems in this letter. It is not clear with what it belongs. It follows dimisi and may go with that. Or it may be at the start of a new sentence, after a full stop at dimisi. Alternatively it may be the complement of the earlier mitte, with the relative clause quam hodie … dimisi intervening. For the sake of argument I make the assumption here that it goes with mitte, and refrain from speculations about what it would mean if either of the other placements were right.

Soluo of horses means ‘unharness’ or the like (OLD s.v. 5b). If solutis were attributive and the phrase a complement of mitte (‘send … by means of two unharnessed horses’), the point might in theory have been that the half-load was now to be transported onwards by saddle bags rather than a wagon (or by two small carts?). Saddle bags, however, seem ruled out by the reference later to a vehicle, uelatura (see below). Or could it be predicative (‘send by means of the two horses, once they have been unharnessed’, i.e. ‘rested’). Mane ‘in the morning’ would fit with this possibility. This translation is to assume that the same two horses were to be used for the next stage. Alternatively one might translate without a definite article, referring to new horses (‘send by means of two horses that have (previously) been unharnessed’, i.e. ‘out of use’, > ‘fresh’?; but is such an extended meaning acceptable?). A fourth possibility would be to take the phrase as an ablative absolute, with one of two possible meanings. First, ‘send on [the half-load], with two of the horses having been unharnessed’. On this view there would originally have been more than two horses pulling the full load, but now the number is to be reduced by two for the part-load. Second, ‘send on the part-load [by some unspecified means], having unharnessed the two horses [that I dispatched]’. The phrase is not straightforward.

Laedatur needs a preceding ne, but the subject is obscure because of the missing line(s). It might possibly be the load to be transported, but that cannot be certain.

A question raised by laedatur ạ uẹḷạtura in qua aḍferetur uiridia hoc esṭ cymas et holorac̣ias et ṇạpic̣ias eaque mitte concerns the sentence division. One possibility would be to take uiridia as a collective feminine singular deriving from the neuter plural,Footnote ¹ and standing as subject of adferetur. That however would necessitate an explanation of why cymas with its associated terms is in the accusative plural when it is in apposition to uiridia. There are late examples of hoc est/id est behaving like prepositions and governing the accusative,Footnote ² but the several complications (the singular use of uiridia, hoc est + accusative) make it more likely that a sentence ends at aḍferetur. Viridia hoc esṭ cymas etc. then become objects of the following mitte (uiridia ‘vegetables’ has been turning up recently in Egyptian tablets: see Adams Reference Adams2016, 310). That interpretation in turn leads to the issue of the -que attached to ea.

Ea(que) is a resumptive pronoun, that is it resumes the list of nouns preceding and summarises them. Resumptive pronouns are often used for clarity (see Adams Reference Adams2016, 700, s.v. ‘pronouns … resumptive’). The use here is unusual, in that the pronoun has what looks like a redundant -que. This however must be the use of -que attached particularly to pronouns that is paraphrased by Kroll (Reference Kroll1929) on two passages of Catullus (31.13, 102.3) as an equivalent of quoque. Note Catull. 102.3, si quicquam tacito commissum est fido ab amico, | cuius sit penitus nota fides animi | meque esse inuenies illorum iure sacratum, ‘if anything has been entrusted by a faithful friend to one given to silence, whose own good faith is absolutely known, me too you will find bound by their rule’ (meque = me quoque). The -que here has tended to be emended away (see Trappes-Lomax Reference Trappes-Lomax2007, 282), for example to (me) aeque, on the grounds that such a use of -que is scarcely attested (see e.g. Thomson Reference Thomson1997, 539). Eaque at Vindolanda is an exact parallel for meque, and the Vindolanda tablets, though they are later, reflect a level of Latinity that scarcely surfaces in the high literary language but was certainly sometimes drawn on by Catullus. The emendation to me aeque produces the same meaning as meque, and is pointless, based as it is on the assumption that usages can be dated precisely just from the literary record. See also Hofmann and Szantyr (Reference Hofmann and Szantyr1965, 475) and Pinkster (Reference Pinksterforthcoming) §22.21 for further examples of -que used thus.

The point of the -que in the context may be that, if only a half-load is to be sent on, there is room for additions.

Ạ uẹḷạtura with laedatur has a noteworthy use of a(b). With passive verbs animate agents are usually in classical Latin expressed by ab, but inanimate causes (on the distinction between agent/cause see Pinkster Reference Pinkster2015, 245) by the plain ablative. Here however the inanimate cause, uelatura, has the preposition ab. This prepositional usage is found particularly in more mundane and didactic prose, perhaps most notably that of Vitruvius (see Adams Reference Adams2016, 172–7, with examples showing that with many inanimate nouns Vitruvius prefers ab to the ablative; see too Pinkster Reference Pinkster2015, 245–6; also OLD s.v. ab 21, though citing not a single example from Vitruvius).

In classical Latin uelatura occurs just twice, in Varro (Ling. 5.44; Rust. 1.2.14). It is a verbal noun in both places, of the act of transporting. In one it complements uiuunt, i.e. uelaturis uiuunt, ‘they make a living out of transportation, carriage’, and in the other it is object of facere, ‘to do transportation, carriage, to act as a carrier’. It might just have had this verbal meaning here, ‘(lest) it be damaged by/from the act of transportation’, were it not for the clause following, in qua aḍferetur (‘in which it will be brought’), where in establishes that uelatura is concrete (‘vehicle, conveyance, carriage’), just like uectura elsewhere. In an earlier tablet (Tab. Vind. III.649.13) its adjective (omnem uelatuṛam) is suggestive of the meaning ‘load’, but the context is incomplete and the interpretation problematic. In any case its semantic development is similar to that of uectura, which turns up in the same tablets.

Hoc est and id est were long in rivalry, with for example hoc est preferred by Cicero in his earliest works. For some information about their distribution see Wölfflin and Meader (Reference Wölfflin and Meader1900, 378–80) and the detailed presentation of hoc est at TLL VI.2.2709.1ff.

In the phrase cymas et holorac̣ias et ṇạpic̣ias the first term cymas is a noun and the other two are adjectives of material modifying cymas. There are no feminine plant names holeracia and napicia. The construction is et … et ‘both … and’.

κῦμα according to LSJ (in this botanical sense) is a young sprout of a plant. The OLD gives the meaning for cyma as ‘the spring shoots of a cabbage or similar vegetable’. It is not a plant in its own right, but a shoot of various types of vegetable, the types in this context being specified by the adjectives. Note that, like other Greek loan-words with this ending, the word has changed gender here to feminine (cf. e.g. glaucumam at Plaut., Mil. 148; schema is neuter in tragedy in Latin but feminine in comedy, and a freedman in Petronius uses it in the feminine: see Adams Reference Adams2013, 449). Cyma is attested in Latin both in its original neuter, and as a feminine (see OLD, and also below on Columella). In combination with the two adjectives of material following, the meaning must be ‘young sprouts/spring shoots both of … and of …’. I take the two adjectives in turn.

First, holorac̣ias. The suffix is -aceus (with a commonplace closing of e in hiatus: see below on commiatus). -aceus is an adjectival suffix of material with the meaning ‘consisting of’ et sim., found particularly with plant names as its base (e.g. mustaceus, hederaceus, betaceus, uiolaceus, etc.: see Leumann Reference Leumann1977, 287). The spelling ought to be holeraceus (a form that exists but is attested only once, it seems), with the same base (holer-) as that of other suffixal derivatives of the same root, e.g. holer-arius, holer-ator, holer-osus. The spelling here implies that the writer declined holus not as holeris etc., but as holoris, which would lead him to holor-aceus. Holoris (etc.) does not seem to be attested, but this sort of variation in third declension -us neuters does occur. The genitive of facinus, for example, may be either facineris or facinoris, and both derived adjectival forms, facinerosus and facinorosus, are attested (see TLL VI.1.76.35ff.), with grammarians recommending in both noun and adjective the forms with e (see also TLL VI.1.77.37). The implied holoris here may be an addition to the attested stock of morphological variations in Latin, and it must have been substandard.

Holus denotes ‘vegetable(s), particularly green vegetables’. Cymas holoracias ought then to mean ‘young sprouts of greens’.

If napicias is the correct reading, it both is an addition to the lexica and may be relevant to the formation of another (Romance) term. In the recent TLL fascicule IX.1.I the adjective napicius is not recorded. There is however a noun napicium, described (IX.1.48.1) as a ‘nomen herbarum’. Those plants that it embraces are explained by André (Reference André1985, 169) as the swede/turnip, or (Fr.) ravenelle. In the Vindolanda tablet napicias is not this neuter noun but an adjective napicius, modifying, in the feminine plural, cymas. It is the first attestation of the adjective napicius. This is clearly a derivative of napus, which is usually rendered ‘swede, turnip’, but by André (Reference André1985, 169) is given five different meanings. There is no point here speculating about what its precise reference is in this passage. What is interesting, however, is the possible relationship of this new adjective to a French term. Napicium (see above) has traditionally been thought to survive in Gallo-Romance (navet ‘turnip’; see TLL IX.1.47.73), but in Latin itself is attested only very late (Dioscorides Latinus: see TLL). The TLL (IX.1.47.73) alludes to an alternative derivation of navet, as follows: ‘francog. navet fort. a *napicius (Barbier), quam coni. tamen neque necessariam neque probabilem esse putat M.-L. 5821; sim. Wartburg VII.11 Schd.’ That is, a derivation from a (then) unattested *napicius has been suggested but widely rejected. We now have the adjective napicius attested about three centuries before the noun napicium, and the origin of Fr. navet should at least be reassessed; could that adjective have been substantivised in the masculine on the analogy of the masculine noun napus?

The expression cymas napicias has a semantic parallel in Columella. The shoots in the tablet ‘belong to/come from napi’. At Col. 12.56.3 napi ‘put forth shoots’: curandum est ut … antequam caulem agant et cymam faciant, dum sunt tenera, componantur, ‘Care, however, must be taken that … they are put in store while they are still tender, before they form a stalk and put forth shoots’ (Loeb, Forster and Heffner). Columella uses the term in the feminine here, and the singular is collective. The collective singular is common in botanical terms in Latin (Löfstedt Reference Löfstedt1956, I.14; cf. caulem in the same clause, in reference to a multiplicity of napi and their caules; the OLD specifically allows a collective meaning for cyma). Columella incidentally also uses cyma as a neuter, but in verse (10.129, frigoribus caules et uerno cymata mittit: note the similar context to that of 12.56.4, but with plurals in this case).

If the reading (suggested by the editors but then rejected) were rapicias (an attested adjective), the meaning would be much the same. The OLD defines rapum as ‘a turnip’, and napus as ‘a turnip, navew’. A distinction is made between the two at Col. 12.56.3. André (Reference André1985, 216) gives the primary meaning of rapum as ‘Rave, Chou-rave (Brassica rapa L.)’, and (185, 169) of napus as ‘Chou-navet, Navet (Brassica napus L.)’.

The rough meaning of the whole phrase in the letter would be ‘shoots/sprouts of greens and turnips’.

Claueṃ quọque aliam misisti cum cista quam debueras seems to mean, ‘you sent too a key with the casket other than (that which) you ought to have sent’. The expected construction would have been clauem aliam quam eam quam debueras mittere, but there has been ellipse of eam quam. Cf. e.g. Amm. 15.5.12, animaduertit … longe alia quam dictarat Siluanus, literally ‘he noticed far other things than (those which, = ea quae) Silvanus had dictated’ (Rolfe, Loeb: ‘and other matter added quite different from what Silvanus had dictated’, rendering the understood ea quae with ‘what’). See further TLL I.1635.28 (and a few other passages cited nearby).

The last four words are read as haec enim ḥọrrioli ḍicitur. The explanatory sentence-opening haec enim would be expected to refer back to the subject of the previous sentence (the key), rather than to a prepositional expression within it. If the idea were haec enim clauis, this (wrong) key would have to be that of something else. Horrioli (horreolum ‘small storehouse’, found at Valerius Maximus 7.1.2) would give good sense (‘for this (key) is said (to be) that of a/the small storehouse’; ellipse of the verb ‘to be’ comes up below). The form horriolum would have mundane closing of e in hiatus. The editors considered but did not prefer the reading haec enim harrioli dicitur, in which case haec would have been taken as referring back to cista. If the sequence is complete, the meaning would then be, ‘for this is called the soothsayer's (casket)’. Harriolus would have a false geminate.

891 (Inv. no. WT 2017.16)

The letter begins with the construction rogo te + habeas: rogamus te, domin[e] Verecunde per notissimam iustitiam tuaṃ dignos nos habẹ[a]s. There is now a detailed study of rogo + plain subjunctive (as distinct from rogo ut) in non-literary letters, by Halla-aho (Reference Halla-aho2009, 81–5). At Vindolanda this construction is far more frequent than that with ut (Halla-aho Reference Halla-aho2009, 82). The construction without ut is not, however, to be regarded as non-standard, as it (as well as rogo ut) is used by Pliny in letters to Trajan.Footnote ³ On the other hand, the long separation (of four lines) here between rogamus and habeas is not in line with Halla-aho's suggestion (Reference Halla-aho2009, 83) that the ‘asyndetic construction … was used more in contexts where the subordinate verb was close to the governing verb’.

In these letters, as Coleman's account (Reference Coleman2012) makes clear, the rogo-construction not infrequently is employed in appeals for a favour to be done on behalf of another, as is the case in this letter. Halla-aho (Reference Halla-aho2009, 81 n. 73) and Coleman (Reference Coleman2012, 200–1) also point out that Pliny six times uses a formula rogo (ut) digneris, ‘I ask that you see fit to …’. The construction with digneris looks like a refined variation on the expression with dignus, and ‘it resulted in a more complex structure’ (Halla-aho Reference Halla-aho2009, 81 n. 73). For rogo dignum me habeas at Vindolanda (II.167, 174, 176) see Halla-aho (Reference Halla-aho2009, 83). The expression is attested there in appeals for leave, but that does not mean it would have to be so restricted. Halla-aho (Reference Halla-aho2009, 83) indeed cites, from a petition to the prefect of Egypt (Cugusi Reference Cugusi1992–2002: I, text 149), a version of the formula with a different but equivalent subjunctive verb in a context that is not to do with leave: ṛọg̣ọ domine [dig]num me ịụḍịc̣ẹṣ ut … For the idiom, which is old, cf. Plaut., As. 80, is me dignum quoi concrederet / habuit. The other construction, rogo (ut) digneris, is not restricted to applications for leave.

For pleas/swearing ‘in the name of’ someone or something see TLL X.1.1155.6ff. (but not it seems citing examples of per iustitiam) and OLD s.v. per 10b. Appeals to someone's justice (accusative), of the type deprecor tuam iustitiam, are however attested (TLL VII.2.715.2ff.). For rogo per see Sen., Contr. 10.4.10, tu, inquit, qui oculos non habes, per oculos rogato; tu, inquit, qui manus perdidisti, per manus rogato.

For the form of address (with domine + the vocative of the name) at Vindolanda see Adams (Reference Adams1995, 118). It is used elsewhere particularly to emperors (TLL V.1.1928.56ff.), and must have been highly deferential.

The words following praestes ut are conjectural, but facias is the verb in the ut-clause which contains a request on behalf of Crispus. Thanks to the TLL X.2.914.38, where there are cross references to examples of praesto ut in the article, one can get some idea of the possible meanings of the construction. It refers particularly to offering or undertaking to do something.

Note first Panegyrici Latini 2.12.1, solus inquam omnium qui adhuc imperauerunt ut princeps esses praestitisti, ‘you alone, I say, of all who so far have ruled, offered that you might be emperor’ (as distinct from seizing power). Ovid, Fasti 4.149–50, ut tegat hoc celetque uiros, Fortuna Virilis | praestat, is translated by Frazer (Loeb) as ‘Virile Fortune undertakes to conceal the blemish and to hide it from the men’. ‘Offers to’ would probably fit the context. Cic. fil., Fam. 16.21.2, praestaboque et enitar ut in dies magis magisque de me duplicetur opinio is translated by Shackleton Bailey (Loeb) as ‘I shall make sure and work hard to see that this tiny new image of mine goes on getting bigger and bigger as the days go by’. It is not clear that ‘make sure’ is the right translation, as praestabo is followed by a verb of trying hard. Taking on the task is surely the idea, and ‘undertake’ would fit.

Pliny, writing to Trajan, uses flowery language at Epist. 10.4.6: rogo ergo, domine, ut … honestis, ut spero, adfectionibus meis praestes, ut non in me tantum uerum et in amico gloriari iudiciis tuis possim, ‘I ask you therefore, lord, that ... you offer up to my honourable (I hope) affections the possibility of my glorying in your judgement not only in my own case but also in that of a friend’. The context of this passage is not unlike that of the Vindolanda letter. Pliny appeals to Trajan to make Voconius a senator. Note rogo ergo domine, and the fact that this is a request to a superior for a favour.

Here finally are three more passages illustrating these senses: Ovid, Pont. 2.9.79 (addressed to a king), tua nunc uicinia praestet, / inuiso possim tutus ut esse loco, ‘let your proximity now offer the possibility that I may be safe in a place I hate’, Sen., Epist. 57.2, nihil illo carcere longius, nihil illis facibus obscurius, quae nobis praestant non ut per tenebras uideamus, sed ut ipsas, ‘nothing could there be longer than that dungeon, nothing more obscure than those torches, which offer us the possibility, not of seeing through the darkness, but of seeing the darkness itself’, Pliny, Epist. 10.51.1, difficile est, domine, exprimere uerbis quantam perceperim laetitiam, quod et mihi et socrui meae praestitisti ut adfinem eius Caelium Clementem in hanc prouinciam transferres, ‘it is difficult to put into words, Lord, the joy I felt in that you undertook to me and my mother-in-law that you would transfer her relative Caelius Clemens to this province’. ‘Undertake’ or ‘offer’ would fit the context of the Vindolanda letter.

There is an ambiguity to the construction C̣rispum mensorem facias ụt possiṭ beneficio tuo ḷẹuius miḷitare. The meaning, first, might be, ‘that you make Crispus a mensor so that he may be able to have through your generosity a lighter service’. A difficulty raised by this interpretation is that surveying was highly technical and required training and it seems odd simply to ‘make’ someone a mensor. There are several other possible meanings of this term (see the editors ad loc.), but whatever the role alluded to, the job is likely to have been demanding. The alternative is to take Crispum mensorem as a proleptic accusative, of which, with facias, it would be a classic type. On this interpretation the meaning becomes, ‘… that you see that Crispus the mensor may be able to have a lighter service’. The essence of a proleptic accusative is that it is in effect the subject of the verb in the dependent clause (here possit), but is expressed as a sort of heading at the start as object of the higher verb. Note for example, in a letter from the Myos Hormos road (Cuvigny Reference Cuvigny2003, II.405, M1107; also Adams Reference Adams2016, 291), eosdem camellos iube adaquentur, ‘order that those same camels are watered’. The order is not given to the camels, but to unnamed persons. The camels by implication are the subject of the dependent verb adaquentur but are expressed not in the nominative, the logical case, but as accusatives with iube. The proleptic accusative, which is well attested in early Latin and lives on (see Hofmann and Szantyr Reference Hofmann and Szantyr1965, 471; Halla-aho Reference Halla-aho2012; Pinkster Reference Pinkster2015, 759–61, the last referring to the accusative as a ‘pseudo-object’), is particularly common with facio (see Bennett Reference Bennett1910–14, II.223 on this point). Note for example Cato, Agr. 312.1, uectes iligneos … facito uti sient parati, cited by Bennett. The meaning is not ‘make levers etc., so that they are ready’, but ‘see that (existing) levers are at the ready’. On this second interpretation Crispus’ duties as a mensor are currently too demanding. The whole sequence from rogamus te in line 2 would mean something like this: ‘we ask you … to undertake to bring it about that … Crispus the mensor be able to have a lighter military service.’ Facio followed by ut would in this corpus be expected to mean ‘see that’ rather than ‘do X, so that’. Cf. Tab. Vindol. 256.i.5 facias ut uenias, and also, without ut, 343.26 fac (denarios) mi mittas, 645.i.9 fac ut certum mihi [r]e[s]cribas.

The adverb leuius is of interest (ụt p̣ossiṭ beneficio tuo ḷẹuius miḷitare). The adverbial use of the neuter (accusative) leue is non-existent in classical Latin and has not been attested hitherto until a long time after the Vindolanda tablets (see TLL VII.2.1216.39ff.: e.g. Prudentius, Claudian, Cyprian, Venantius Fortunatus).

Swearing by or addressing someone else's genius was a servile act, or at least a mark of subordinate status (see the examples at Sen., Epist. 12.2 and Petron. 18.5, with Adams Reference Adams2016, 230–1, 239, and bibliography). Here the genius is to be thanked. For thanking someone's genius (attested mainly in inscriptions) see TLL VI.2.1836.23ff.: e.g. Petron. 57.10, tamen – genio illius gratias – enataui; cf. too Plin., Pan. 52.6, simili reuerentia, Caesar, non apud genium tuum bonitati tuae gratias agi, sed apud numen Iouis optimi maximi pateris.

At the end of the letter there is a suggestive use of debemus: ẹ[t] g̣enio tuo gratias ag̣e[re d]ẹḅeṃụṣ. If the reading is right the literal meaning would be ‘and we ought to/must offer thanks to your genius’. But the expression of thanks would be in the future (when the request is granted). The use of the present tense in reference to the future gives the verb a transitional sense, from ‘ought’ to ‘will’. The auxiliary verb with infinitive would be getting close to a future periphrasis, = ‘and we will offer thanks to your genius’. In Sardinian the future may be expressed by dévere + infinitive (see Adams Reference Adams2013, 653; Reference Adams2016, 309 with bibliography). Hofmann and Szantyr (Reference Hofmann and Szantyr1965, 314) find an instance of this future nuance at Petron. 67.7 (but, against, see Schmeling Reference Schmeling2011, 278; Pinkster Reference Pinkster2015, 440). Stotz (Reference Stotz1998, 325) quotes some late examples of the future periphrasis, but doubts about the earlier existence of the usage are expressed by Pinkster (Reference Pinkster2015, 440).

892 (Inv. no. WT 2017.24 + 26)

For the formula peterem … commiatum see TLL III.1826.24ff. It is common in literary texts too. Only one example is cited there for a dependent genitive referring to those for/by whom leave is sought.

The spelling commiatum for commeatum has closing of e in hiatus, a banal phenomenon: for this change (already at Vindolanda) and its chronology see Adams (Reference Adams2013, 102–8). The misspelling here is not of the same kind as penis for penes at ii.10, where the original short e is not in hiatus but is in the final syllable. The spelling with i in the latter is linguistically unusual, as the normal front-vowel misspelling in writing-tablets and inscriptions shows e written for short i rather than i written for short e: see the statistics at Adams Reference Adams2013, 5162. However, this type of misspelling, i for short e, is attested (see Gordon and Gordon Reference Gordon and Gordon2005, 270–1), and sometimes turns up in final syllables (271 (6)) (e.g. siui CIL VI.11645).

Abs is well attested both as a prefix (e.g. abstergeo) and preposition, usually before t in both uses (so abs te) and sometimes before other voiceless stops. Abs te is the norm in Plautus, over a te and ab te, though manuscripts are not entirely to be trusted (see Lodge Reference Lodge1924, 2 col. 1 for a te (infrequent), 2 col. 2 for ab (no examples of ab te listed) and 2 col. 2 for abs/aps te (common)). According to Dyck (Reference Dyck2010, 14), abs te was by Cicero ‘banished from the speeches after Rab. Post. 30’ in favour of a te (though again one cannot be too dogmatic on the basis of late manuscripts). Landgraf (Reference Landgraf1914, 35) has a few further details of Cicero's practice. Abs te does seem to have belonged to Cicero's early period, and in the letters is far outnumbered by a te. Indices verborum show that abs is absent from e.g. Vitruvius and Tacitus, though the full significance of such absences would only be revealed by statistics showing the frequency of a before words beginning with t: a(bs) te is unlikely to occur in a technical or even a historical text. This is the first instance at Vindolanda of abs. A te is not frequent, but does occur, at II.213, 250 and 305. Abs te looks like an old-fashioned relic.

The genitives Litucci, Vitalis [crossed out] et Victoris are in the expected case, in apposition to Raetorum. The other three names, Augustanum Cusium Bellicum (separated by interpuncta) are components of the genitival ‘five’, and should also be in the genitive, dependent as they are ultimately on commiatum. There are sometimes illogical uses of the accusative in lists, and indeed at Vindolanda (see Adams Reference Adams2013, 229, 233), but this is a particularly clear case of an unmotivated ‘list accusative’, because in other passages it is sometimes possible to account for an accusative by understanding a verb of which it could be object. That is not so here. It is not uncommon for switches of case to occur haphazardly within a single list (Adams Reference Adams2013, 227–31).

There are two words planta, one (= ‘sole of foot’) irrelevant. The other means primarily ‘a young shoot detached from the parent plant for propagation’ (OLD), i.e. ‘a cutting’. Greater detail may be found in a note by Zainaldin (Reference Zainaldinforthcoming) on Gargilius 2.1.3, which I quote here with the author's generous permission:

The presence of both planta and culter in the same passage makes it likely that planta has the earliest meaning above (see further below).

A culter scissorius could presumably be used for removing slips for propagation (see further below). A relevant literary passage is Virgil, G. 2.23, where plantas is object of a compound of scindo (plantas … abscindens). In Virgil the reference is to a method of propagation whereby ‘shoots coming from below ground … can often be separated with some root attached’ (Mynors Reference Mynors1990, 103). The passage is also cited by Zainaldin (Reference Zainaldinforthcoming) in his note on Garg. 2.5.7, commenting on the method of propagation referred to. Again I quote for the degree of detail that Zainaldin provides:

Presumably a cleaving instrument might be used as an aid in this process of tearing off, and scissorius in the letter compared with abscindens in Virgil points to the precise meaning of plantas in the letter. The numeral ‘two’ that follows plantas is however an oddity.

Scissorius has hitherto been unattested except in medieval Latin. I have consulted Manfred Flieger of the Thesaurus Linguae Latinae and there is no instance of the term in the files.

Qui penis Talampum is without est. Ellipse of the verb ‘to be’ in relative clauses is illustrated in detail from Cicero's letters by Heidemann (Reference Heidemann1893, 44–8).

Missi (ii.12) is not a spelling error but is etymologically correct, and must reflect some sort of training that some scribes had received (in 890 however dimisi is written: there is little uniformity to Latin spelling). The double s spelling is common in the Vindolanda tablets (see e.g. Adams Reference Adams1995, 88–9). It is also found for example several times in the Lex de Gallia Cisalpina (e.g. 12 repromeisserit) (see Crawford Reference Crawford1996, I, p. 464).

893 (Inv. no. WT 2017.18)

An iracundia could be an ‘outburst of anger’, as at Tac., Ann. 14.4.1, ferendas parentium iracundias et placandum animum dictitans, translated by Jackson (Loeb) as: ‘observing from time to time that outbreaks of parental anger had to be tolerated, and that he must show a forgiving spirit’. Iracundiolae are ‘little outbursts of anger’ that deserve to be castigated by the referent's elders. Iracundiola seems unattested before the medieval period (not in the TLL), and it might even have been coined by the writer, who has a distinctive style. For this type of diminutive see Leumann (Reference Leumann1977, 305), at 282A. Many (feminine) examples are listed by Gradenwitz (Reference Gradenwitz1904, 294), and such diminutives could no doubt be coined at will.

For ‘deserving castigation’ see Plin., Pan. 46.6, castigari merebantur. At Att. 3.10.3 Cicero hopes not to be ‘deserving of castigation’: quam ut castigatione aut obiurgatione dignum putares. The bits of this letter surviving in full seem to be correct and rather pompous. The TLL s.v. castigatio has a section (III.530.80ff.) where the castigation is (as here) of a vice, not directly of the person, as at Livy 30.37.1, cum multa castigatione perfidiae.

At first glance praesens in praesentia (ut tecum praesens ạ[ga]m in praesentia scito) may look like an emphatic pleonasm, but it is certain that in praesentia begins a new sentence and that its meaning is temporal. There are two uses of the phrase in praesentia. For the first, which means ‘in (someone's) presence’, see now TLL X.2.856.52ff. The second use has been variously interpreted as consisting of in + neuter plural of praesens, or of in + abl. praesentia. The OLD s.v. praesens 16b takes it in the first way, whereas the TLL X.2.857.48ff. classifies it under praesentia. The meaning is ‘for the moment, for the time being’: see e.g. Quint. 5.7.15, fingamus in praesentia scire, translated by Russell (Loeb) as ‘For the moment let us suppose that he knows’. The writer of the letter thus says that he will discuss a certain matter with the addressee face to face, when he sees him, but ‘for the moment be aware that …’.