1.1.1 Obitum fecit = ‘died’ as a ‘support-verb’ phrase

The use of facio in obitum fecit = ‘died’ as required by the context can be explained with the meaning ‘to do, perform (the action contained or implied in the object)’ given by OLD for facio with verbal nouns and sim. (OLD s.v. 23): obitus is such a verbal noun par excellence, a derivative of obeo ‘to meet (one's death), die’,Footnote ² meaning ‘the act of dying’ (lit. ‘the act of meeting [one's death]’) (see OLD s.v. 2); hence obitum fecit ‘performed the act of dying’ (i.e. ‘died’, = obiit).

Facio used in this way can be described as ‘support verb’ (French ‘verbe [du] support’, Spanish ‘verbo soporte’, German ‘Funktionsverb’), a term now commonly applied to verbs in noun–verb combinations where the idea normally expressed by a related verb is instead expressed by the noun. This is not the place to launch into a detailed discussion of support verbs.Footnote ³ However, in order to confirm the suggestion that obitum fecit = ‘died’ is a support-verb phrase, we need to look into possible criteria for distinguishing ‘support’ uses of facio from other uses of facio + direct object.

One criterion for identifying phrases with support verbs commonly invoked in studies on the subject is the replaceability of the phrase with a simple verb, as for example in Plaut. Cist. 73–4 confidam fore | si medicus ueniat qui huic morbo facere medicinam potest, where facere medicinam is the equivalent of mederi; cf. W. de Melo's (Loeb) translation ‘I trust it will [get better], if the doctor comes who can cure this illness.’Footnote ⁴ A revealing contrast is that between the following two instances of facio, both in exactly the same phrase with odium as object—one of which, I would suggest, is a case of a support verb and the other is not:

In (i), the phrase means ‘to cause hatred (of studies)’; cf. D. Russell's (Loeb) translation: ‘But there must be moderation in holidays: if we refuse them, the boys will hate their work; if there are too many, they will get used to being idle’ (lit. ‘But there must be moderation in holidays lest they cause either hatred of studies if refused, or a habit of idleness if excessive’). Odium facere is a standard way of expressing the idea here, and is not replaceable by a simple verb. In (ii), on the other hand, the same expression is a periphrasis of, and replaceable with, odisse ‘to have aversion to’: ‘[the animal] affected by this illness draws breath frequently and will have aversion to hay’.Footnote ⁵ As E. Löfstedt points out, Vegetius, drawing on the Mulomedicina Chironis, expresses the idea of feni odium faciet with a synonym of odisse, fastidire + direct object (Mul. 1.30 spiritum crebrius agitat faenumque fastidit ).Footnote ⁶ Odium facere in (ii) behaves as a support-verb construction, while in (i) it does not.

A similar contrast of meaning between two cases of facio with the same object is found in the following passages:

In (iii), clamores facere means ‘to cause shouts (of approval)’ (OLD s.v. clamor 1c), i.e. ‘to raise, win applause’. In this expression facio, or its synonym in the meaning ‘to cause to occur, bring about’ (OLD s.v. 2a), efficio, is what one would normally expect: cf. for example Cic. De or. 1.152 haec sunt, quae clamores et admirationes in bonis oratoribus efficiunt (‘These are the things which in good orators produce applause and admiration’, G.P. Goold, Loeb); QFr. 3.1.7 clamores efficiam si eum mecum habuero otiosus (‘I shall do wonders if I have him by me on vacation’, D.R. Shackleton Bailey, Loeb).Footnote ⁹ In (iv), on the other hand, clamorem facere is the equivalent of clamare, ‘to shout’; cf. de Melo's translation above, and note that the phrase here picks up quid clamas? two lines up (872). Thus (iv) exemplifies facio as a support verb, while (iii) does not.

Compare likewise a third pair of passages of the same kind:

In (v), fugas facere means ‘to cause, bring about stampedes’, a phrase not replaceable by the transitive verb fugare ‘to cause (someone) to flee’. In (vi), on the other hand, fugam faciunt is the equivalent of fugiunt; here again facio can be described as a support verb.Footnote ¹²

But replaceability alone is not always a reliable guide to identifying support verbs. As is pointed out in The Cambridge Grammar of the English Language, in which such verbs in English are discussed under the heading of ‘light verbs’, ‘the use of a light verb and noun tends to yield a significant increase in syntactic versatility over that of the associated verb construction. Most importantly, it generally allows for dependents to be added to the noun, allowing a considerably greater range of elaboration by modifiers and determiners.’Footnote ¹³ Consequently, constructions with support, or light, verbs are not always replaceable by a simple verb in a way that would not alter the sense. Examples from English quoted in The Cambridge Grammar include, for example, She gave him an unusually passionate kiss, which ‘is less awkward than She kissed him unusually passionately three times’; and ‘a special case of such dependents [which] involves quantification’, as for example in She made three very astute comments on his suggestion, which ‘is not the equivalent to She commented three times very astutely on his suggestions’.Footnote ¹⁴

The scenarios illustrated by the two examples above apply to Latin as well: consider, for example, the following cases:

In (vii), the phrase can be rewritten as ei grauiter conuiciati sunt without any detriment to the sense: cf. Shackleton Bailey's translation (his letter 418), ‘(they) booed him savagely’; for conuicior used thus with a dative, cf. for example Quint. Inst. 12.10.24 dum illi conuiciantur. In (viii), however, Pisoni consuli mirifice conuiciatur ‘he scolds Piso in a remarkable manner’ would not be the equivalent of conuicium Pisoni consuli mirificum facit ‘he gives Piso a remarkable scolding’: the adverb might refer to the way in which the reproach had been made (it could mean, for example, ‘in a remarkably skilful manner’), while the adjective places the focus on the magnitude of that reproach (cf. Shackleton Bailey's ‘spectacular’); moreover, the use of the nominal construction here allows for the parenthetical si id est conuicium. Similarly, in (ix) tardius primum … metere oportebit would not express the same idea as tardius messim primam … facere oportebit: it could mean ‘first it will be necessary to cut (it) rather late’.

A special case in Latin is represented by passive constructions where the corresponding verb is deponent, as for example in:

Here, conuicium fieret serves as the passive of conuicior, which is non-existent.Footnote ¹⁸

The examples in (vii)–(x) above demonstrate that not all support-verb constructions are subject to (straightforward) replaceability by a simple verb. It therefore seems necessary to establish an additional criterion for identifying such constructions. I propose the following semantic criterion for distinguishing support-verb phrases with facio of the type discussed here from other cases of facio + direct object. As a support-verb phrase by definition indicates an action that is expressed by the noun rather than by the verb, which is thereby used just to ‘support’ the noun, facio must be used in the weakened (‘supportive’) meaning ‘to perform (the action contained or implied in the noun)’ (OLD s.v. facio 23), and not in its standard meaning as a fully fledged verb = ‘to make’ in the sense of ‘to produce, bring forth, create’ (for example, OLD s.v. 1–3) or ‘bring about, cause’ (OLD s.v. 11).Footnote ¹⁹ Thus in the case studies (i)–(vi) of homonymous phrases with contrasting meaning discussed earlier in this section we have odium facere = ‘to perform the act of hating’, i.e. ‘to hate, shun’, as opposed to odium facere = ‘to cause hatred’; clamorem facere = ‘to perform the act of shouting’, i.e. ‘to shout’, as opposed to clamorem facere = ‘to cause, raise applause’; fugam facere = ‘to perform the act of fleeing’, i.e. ‘to flee’, as opposed to fugam facere = ‘to cause a flight’. We see this also for example in a phrase mentioned by J.M. Baños Baños, uerba facere = ‘to perform the act of talking’, i.e. ‘to speak’ (as, for example, in Sal. Iug. 33.4), as opposed to uerba facere ‘to produce/form words’ (e.g. Cic. De or. 3.154).Footnote ²⁰ Mutatis mutandis, the same criterion could also be applied to homonymous phrases with some other verbs that can act both as ordinary and as support verbs. One can compare here the English phrase with ‘have’ quoted in The Cambridge Grammar as an example of exactly this type of ambiguity, I had a bath, where the light-verb interpretation is I bathed (in a bath) (with the noun–verb phrase replaceable by a simple verb), while the ordinary-verb interpretation is I owned a bath (with ‘have’ in its ordinary meaning as a fully fledged verb).Footnote ²¹

To return to obitum fecit. We have already seen that the construction in the inscriptions discussed in this section is replaceable by a simple verb, obiit;Footnote ²² noun and verb share the same root, as do medicina and medeor in the Plautine example (pp. 526–7 above). Obitum fecit therefore meets the first criterion considered in this section. With its meaning ‘performed the act of dying’, it also fulfils the second, semantic, criterion. This latter criterion can also be applied to some other verbs that may behave as support verbs. Note for example the use of habeo in the parallel phrase mortem habere at Propertius 3.15.39–40, Dirce | ducitur in multis mortem habitura locis (‘Dirce is drawn along to find death in many places’).Footnote ²³ Habeo here does not have its ordinary full force of ‘to have’, i.e. ‘to possess’, but like facio in obitum fecit is instead weakened to a mere support of the noun, with the phrase replaceable by a verb of the same root as mors, moritura = ‘to die’.Footnote ²⁴

Just like the phrases odium facere, clamorem facere and fugam facere, obitum facere in a different context could potentially mean ‘to cause death’, i.e. ‘to kill’ (= obitum facere as a non-support-verb phrase); however, this use, to the best of my knowledge, is not attested in either literary or inscriptional sources.

1.1.2 Obitum fecit and other facio-phrases: register, chronology, use

The use of periphrastic phrases with facio of the type discussed above as support-verb constructions has traditionally been described as belonging to the Umgangssprache.Footnote ²⁵ Outside texts displaying colloquial features, it is also found in didactic works, for example medical texts, where the passive of facio, fio, is commonly used thus (as, for example, in purgatio fiat for purgetur).Footnote ²⁶ D.R. Langslow looks at this feature in medical writings from an angle different from that found in discussions of support verbs. He relates it to the phenomenon of ‘alteration of a finite verb-form with its nominalization’, i.e. ‘the alteration (or replacement) of the structure [subject (or object) + finite verb] (e.g. nerui resoluuntur) with [genitive + verbal noun] (e.g. neruorum resolutio)’. The ‘nominalization’ ‘behaves syntactically, of course, as a noun, and appears either with a semantically uninteresting verb (est, fit, oritur, nascitur) or in a prepositional phrase (ad neruorum resolutionem)’.Footnote ²⁷ In considering the frequency of ‘nominalizations’ in medical and other didactic works of different periods and register, Langslow expresses doubts as to the validity of any general statements about the register of this phenomenon: ‘Probably, the question is misplaced and it is quite misguided to attempt to characterize nominalization as in and of itself either literary or colloquial’,Footnote ²⁸ and concludes: ‘I am inclined to apply to our medical authors the judgement of Hofmann and Szantyr (1965: 742) [(n. 25 here)] on the great number of nominalizations in agricultural works, from Cato onwards, and in the anonymous Ad Herennium and Vitruvius, and to conclude that their prominence is a feature “more technical than vulgar”’:Footnote ²⁹ the popularity of ‘nominalizations’, he points out, fits in with certain pragmatic tendencies of technical writing, namely ‘syntactic simplicity’, ‘listability’ and ‘concreteness’.Footnote ³⁰

Whether described as ‘support-verb’ phrases or ‘nominalizations’ with an auxiliary verb, the use of facio-phrases of the type discussed here goes back to Early Latin, where they are particularly common in comedy (especially Plautus), but are found also in more ‘serious’ genres such as didactic literature and historiography.Footnote ³¹ During the Classical period, there is clearly a negative attitude to such phrases, as seen at Cic. Phil. 3.21–2, where Cicero comments scornfully on the use by Antony of a phrase of this type, contumeliam facere ‘to make/cause an insult’.Footnote ³² The same stance is later taken also by Quintilian, who at Inst. 9.3.13 refers back to this comment of Cicero.Footnote ³³ And yet such phrases have crept into the writings of the critic himself: cf. for example Cic. Fam.13.29.1 ut primum per aetatem iudicium facere potueris quanti quisque tibi faciendus esset (= iudicare); Sen. 63 (Lysander) cuius modo feci mentionem (= quem … commemoraui); Nat. D. 1.63 Protagoras, cuius a te modo mentio facta est (= qui a te … commemoratus est); Div. 2.65 de passerculis coniecturam facit (= coniecit); Att. 16.8.2 ei conuicium graue fecerunt (see [vii] above, with discussion; for conuicium facere = conuiciare in Cicero cf. also Att. 1.14.5, Clu. 74, Fam. 10.16.1 [in the passive]).Footnote ³⁴ Cicero was clearly being either inconsistent, or otherwise biased, in mocking his arch-enemy Antony for the use of a turn of phrase to which he himself was not immune. No one can be trusted in making judgements about his native language. People often have a negative attitude to a usage without being conscious that they employ it themselves.

Periphrastic phrases, or support-verb constructions, with facio are commonplace in Late Latin, as discussed at length by Löfstedt.Footnote ³⁵ Some salient examples cited by Löfstedt include aquam facere = aquari (Per. Aetheriae 6.1 alia die facientes aquam),Footnote ³⁶ facere mansionem = manere (Per. Aetheriae 4.8 fecimus ibi mansionem, cf. 3.1 ibi ergo mansimus), moram facere = morari (Vulg. Matt. 24:48 moram facit dominus meus uenire) and odium facere in Mul. Chir. 121, p. 39.12 Oder (discussed above, 1.1.1). In this regard, we can note in passing that ILCV 2907 and ILCV 2907a, the only inscriptions that I have been able to trace with the phrase obitum fecit in the meaning ‘died’, are both late (Christian), with 2907 dated precisely to 394 by the consular date (cons(ulibus) dd(ominis) nn(ostris) Arcadio III et Honorio II AAuu[gg(ustis)]), and so fit into the picture for Late Latin as presented by Löfstedt.Footnote ³⁷

1.2.1 The meaning and date of the formula

The two inscriptions with obitum fecerunt from Mauretania Caesariensis quoted by TLL for obitus = ‘tomb’ read:

The context in both, with cui referring to the deceased as the indirect object (or dative of advantage) dependent on the verb fecerunt = ‘made for him/her [i.e. the deceased]’, rules out the meaning ‘died’ for obitum fecerunt. One might be tempted to translate cui … obitum fecerunt as ‘caused the death of’, i.e. ‘killed’ (taking cui as dative of disadvantage). However, such a meaning is impossible in a (commemorative) funerary inscription.

The meaning of the phrase here is ‘set up a tomb’ (or rather, in the context, ‘arranged for a tomb to be set up’), with obitus metonymically used for sepulcrum, ‘tomb’ or ‘funerary monument’ (TLL s.v. II.3).Footnote ³⁹ This meaning is established by the epithet bene merenti in an inscription with obitum fecerunt which is not in TLL, also from Mauretania Caesariensis:

The brothers Donatus and Boinus can only have ‘arranged for a tomb to be set up’ for their ‘well-deserving’ (ben(e) mer(enti)) father in his epitaph.

Alternatively, OLD tentatively suggests that obitum facere might mean ‘(app. [= apparently]) to perform obsequies’ (OLD s.v. obitus 2b, my emphasis). Such a meaning is not impossible, but I have not been able to find parallels for it in Latin epitaphs; on the other hand, phrases such as monumentum/sepulchrum/memoriam sim. fecit/posuit = ‘set up a tomb’ (i.e. ‘arranged for a tomb to be set up’) are commonplace in funerary inscriptions.

Obitus = ‘tomb’ would be a case of ‘abstract for concrete’, a category of semantic change in Latin notably discussed by Löfstedt.Footnote ⁴⁰ As one special type, verbal nouns sometimes come to designate a place, especially in Late Latin: for example, oratio ‘prayer place’, i.e. ‘oratory, chapel’,Footnote ⁴¹ mansio ‘staying place’, i.e. ‘lodgings, house’; so also obitus ‘meeting(-death) place’, alongside dormitio ‘resting place’—hence in inscriptions ‘grave’, exitus ‘departing place’, i.e. the grave, tomb. As ‘a striking parallel to these’, Löfstedt quotes quies, which is sometimes used to mean ‘grave’, i.e. ‘resting place’ (for example, in ILVC 7961 quies Claudiae Priscae).Footnote ⁴²

Several features of the three inscriptions with obitum fecerunt discussed in this section suggest that they are late. Obitus = ‘tomb’ is described by Liana Lupas as Christian,Footnote ⁴³ something that in CIL VIII 21540 (included by Diehl in his collection of Christian inscriptions as 3052b) is confirmed by the use of in pace.Footnote ⁴⁴ The grammar of the epitaph, with the form filios clearly intended as a subject alongside nepotes and therefore explicable as a nominative plural in -os, points further to a late date (sixth century on).Footnote ⁴⁵ The name of the deceased too might point in that direction: misspellings showing s(i) or z(i) for di or ti as the result of palatalisation such as here Crescensa = Crescentia (cf. for example CIL VIII 21751 Inocensa = Innocentia)Footnote ⁴⁶ tend to be late.Footnote ⁴⁷ Better still, the anno prouinciae formula in BCTH 1954, p. 74 provides a precise date for this inscription: anno prouinciae CCCL, i.e. a.d. 389.

M.L. Leschi suggests that obitum fecit in this sense was ‘calquée sur l'expression exitum facere qui signifie aussi faire un tombeau, donner un sépulture’.Footnote ⁴⁸ He was discussing another inscription from Mauretania Caesariensis that allegedly contains a related formula, obitum reddere meaning ‘to set up a tomb’ (but see below). He did not quote any examples of exitum facere = ‘to set up a tomb’, but one can note, for example:

It is impossible to say definitely whether exitum fecit was modelled on obitum fecit or vice versa, especially given the limited surviving examples of both formulae. In any case, with the subject(s) of exitum fecit in all the inscriptions quoted above being clearly the dedicator(s), the phrase as found in them is a precise parallel to our obitum fecit = ‘set up a tomb’: we have here (a) exactly the same development from abstract (‘the act of meeting death/departing from life, i.e. of dying’) to concrete (‘tomb’) (or to ‘obsequies’, if one accepts OLD’s interpretation of obitum facere) that we see in obitus, and (b) a combination with the same verb, facere, to mean ‘to set up a tomb’ (or ‘to perform obsequies’, as in OLD [?]). Moreover, at least one of the epitaphs with exitum fecit (ICUR I 2948 = ILCV 3053) is Christian, and another one (BCTH 1936–7, p. 44 no. 26) has been dated to as late as the third century a.d.;Footnote ⁴⁹ like obitum fecit, this formula too is clearly late.

1.2.2 Obitum reddidit: a synonymous formula in Mauretania Caesariensis?

Another epitaph from Mauretania Caesariensis (Altava, modern Ouled Mimoun) was first mentioned by P. Courtot in Revue Africaine 79 (1936), 416, where he takes OBIT as a complete word, presumably for obi(i)t (commenting ‘Et pour la première fois à Altava, on a trouvé la formule “Obit”’). This was later interpreted by M.L. Leschi (BCTH [1938–40], 209–10) as Dis Manibus s(acrum) Iulia Magdira obit(um) nat(i) red(diderunt), with obit(um) nat(i) red(diderunt) explained by referring to the expression under discussion here, obitum fecit. Given the provenance of the inscription, it is tempting to accept this interpretation, and treat obitum reddere as a variant of obitum facere in Mauretania Caesariensis, meaning ‘her children (?) set this up’. However, in a later publication (Les inscriptions d'Altava [Aix-en-Provence, 1968]), Jean Marcillet-Jaubert prints the text as:

He points out that the stone has NATO and not nat(i), ‘which would have been very unusual’ (‘qui eût été bien insolite’). To judge from the photo published in the Clauss–Slaby Databank, there is indeed what looks like an O after NAT in the last line, but it is smaller than the rest of the letters—which might explain why Leschi, as pointed out by Marcillet-Jaubert (and so also Courtot before him), failed to see the o. As the (emended) text stands, a straightforward interpretation such as obit(um) nato red(didit) = ‘made a tombstone for the son’ would make no sense here as the deceased is a woman; and there is also no dedicator mentioned to provide a subject for red(didit) in this meaning. Far more plausible is Marcillet-Jaubert's interpretation, according to which NATORE forms one word, with the following D standing for (RE)D by haplography through confusion with the last syllable of the preceding noun; he therefore suggests the following reading for the inscription:

Natore is then explained as an error, with o for ū, so standing for nature (= naturae). The issue of whether this is a ‘phonetic fault’ as suggested by Marcillet-Jaubert is not altogether straightforward. While the confusion between u and o is very common in ‘Vulgar’ Latin, it is normally one that affects short u, reflecting a development that would eventually lead to the merger in Romance languages of long o and short u as close o.Footnote ⁵⁰ However, the spelling -ora for -ūra has some possible parallels in inscriptions.Footnote ⁵¹ Whatever the explanation for the form, the interpretation of natore = naturae here makes good sense: it produces the formula synonymous with obitum fecit = ‘died’, obitum naturae reddidit, as included in TLL s.v. obitus II.A.3, to which we will turn shortly (section 2 below). The fact that specifying this after uix(it) an(nos) XLV ‘lived 45 years’ is superfluous is not an obstacle to this interpretation, as inscriptions are often pleonastic: one may compare for example the symbol Θ to indicate that the dedicatee was dead at the time of setting up of the inscription, found sometimes in inscriptions in which the age of death has already been expressed by a formula such as uixit annos/annis or similar.Footnote ⁵²