¹ For the date, see Silvae 4.4. 12–28, together with praef. 14–16, and Vollmer's commentary, p. 10. Statius will be cited in this paper according to the second edition (1911) of A. Klotz's Teubner text; Martial according to the second edition (1929) of W. M. Lindsay's Oxford Classical Text; Pliny according to R. A. B. Mynors' text of 1963 in the same series.

² For the date, see Friedlaender's commentary, vol. i, pp. 55–6.

³ It may be helpful to offer a provisional listing of other poems to patrons which must have required presentation on a specific date or occasion during the year: 2.85, 3.6, 5.18, 5.59, 6.21, 6.38, 6.80, 7.21–3, 7.47, 7.74, 7.89, 8.45, 9.23–4, 9.39, 9.52–6, 9.60, 9.90, 10.44, 10.78, 10.87, 11.36, 11.57, 12.67, 12.74. To this group should probably be added many of the epigrams written (often in epitaph form) in honour of the dead, for example, 1.78, 1.93, 1.114, 1.116, 4.73, 6.18, 6.28–9, 6.52, 6.68, 7.40, 7.96, 9.28, 9.30, 9.51, 10.26, 10.50, 10.53, 10.71, 11.13, 11.69, 12.52. For the latter, the slow appearance of a book would not do; as Statius observed (Silvae 2 praef. 17–18), ‘paene supervacua [sunt] tarda solacia.’

⁴ Friedlaender did, as a matter of fact, elicit an addressee for this poem by reading Care for cara in line 5, with one manuscript. But this meant ignoring the use of cara cognatio in its attested technical sense (CIL vi, 2, 10234, 13) as a name for the Caristia. A better refuge would have been to declare that 9.54 can be taken as addressed to the recipient of either the preceding poem (Quintus Ovidius) or the following one (Flaccus). But this proposition has only the strength of any bold assertion.

⁶ Only slightly less mystifying than the total absence of a name would be the use of a praenomen and nothing more, as at 1.109, 2.57, 5.5, and 10.98, or a name (like Maximus, or Priscus, or Rufus) which could point to any one of several known acquaintances.

⁶ Silvae I praef. 5–19. In 4 praef. 36–8, Statius says that the emperor had received some of the Silvae prior to publication: ‘multa ex illis iam domino Caesari dederam, et quanta hoc plus est quam edere ?’

⁷ As Laurens, P. has observed (‘Martial et l'épigramme grecque du I^ersiècle après J.-C.’, REL 43 (1965) 325Google Scholar), ‘Ces poètes opposent à l'idéal artistique exigeant de Callimaque une aesthétique de l'impromptu.’ See also Reitzenstein, R., RE vi (1907) 94 ffGoogle Scholar.

⁸ Cf. 5.16.9 ‘at nunc conviva est comissatorque libellus’, and 2.1.9–10 (to his liber) ‘te conviva leget mixto quincunce, sed ante / incipiat positus quam tepuisse calix.’

⁹ Viz., 1.11 1.18, 1.20, 1.27, 1.43, 1.50, 1.106, 2.1.9–10, 2.19, 2.37, 2.43, 3.12–13, 3.17, 3.23, 3.45, 3.49–50, 3.60, 3.94, 4.68–9, 4.85, 5.16.9, 6.11, 6.74, 7.48, 7.59, 7.78–9, 8.22–3, 8.59, 9.9, 9.87, 9.97.10, 10.48–9, 11.3, 11.35, 11.52, 12.28, and 12.48.

¹⁰ A respite at Anxur is the boon for which Martial thanks his friend Frontinus in 10.58. Though Martial does not say so, he certainly took advantage of this stay with Frontinus to write 10.51 to Faustinus, a poem of the same date, and written from Anxur.

¹¹ Concerning 3.1 he writes in the preface, ‘primum limen [libri] Hercules Surrentinus aperit, quem in litore tuo consecratum, statim ut videram, his versibus adoravi.’ And he seems to imply that he wrote 1.3 under the eye of his host: ‘Manilius certe Vopiscus … solet ultro quoque nomine meo gloriari, villam Tiburtinam suam descriptam a nobis uno die.’ 2.2 presupposes a lengthy inspection of Pollius Felix' villa, but Statius does not say that he wrote it there.

¹² Viz., 1.12, 1.82, 1.109, 1.114, 1.116, 3.35, 3.40, 3.58, 4.32, 4.42, 4.47, 4.57, 4.59, 4.64, 5.55, 6.15, 6.47, 6.73, 6.86, 7.15, 7.17, 7.19, 7.29, 7.38, 7.44–5, 7.50, 8.40, 8.46, 8.68, 9.23–4, 9.43–4, 9.56, 9.74, 9.76, 9.103, 10.30, 10.37, 10.58, 10.99, 11.9, 11.48, and 11.50.

¹³ There are other allusions to recitations at 1.52.3, 1.63, 1.76.10, 2.71.2, 9.81.1, 11.52.16.

¹⁴ It would be disingenuous to represent my interpretation of this incident as the universally accepted view, and all the more so as the passage has very recently attracted fresh discussion: see Laughton, E., CR n.s. xxi (1971) 171–2Google Scholar and Yardley, J. C., CR n.s. xxii (1972) 314–15Google Scholar.

¹⁵ To raise a tangential point, Pliny promises in the same letter to send Falco a copy of Augurinus' book as soon as it is published. On a first impression, this assurance might seem to bear out a widely held theory that in accepting a dedication, the dedicatee of an ancient book also assumed responsibility for its publication: see Birt, Th., Kritik und Hermeneutik nebst Abriss des antiken Buchwesens (Handbuch der klassischen Altertumswissenschaft (Munich 1913) 313–14Google Scholar). This is a theory I distrust, without, however, knowing how to refute it. But in this particular case, it makes a great difference whether Falco was at Rome or abroad when he received Pliny's letter. What is known of his career suggests that he was in fact in the East at this time: see most recently Birley, A. R., Epigraphische Studien iv (1967) 69Google Scholar. If that is the case, there would be nothing unusual about a friend's offer to forward a book which would otherwise be unavailable to him.

¹⁶ See 1 praef. 7, 22, and 34; 2 praef. 21; 3 praef. 7, 16, and 31. For the connotation of libellus, see Birt (supra n. 15) 274 and 292.

¹⁷ The procedure is well described by Guillemin, A.-M., Pline et la vie littéraire de son temps (Paris 1929) 41 ff.Google Scholar, and can be seen exemplified in Epist. 1.2, 1.8, 2.5, 3.63, 4.14, 5.12, 7.20, and 8.19.

¹⁸ In Epist. 2.5 we find Pliny passing around only a part of one of his speeches. As for poetry, Pliny did send the full text of his book of hendecasyllables to Paternus with Epist. 4.14, but he implied in section 6 that it would have been a common procedure to send only samples. In 8.21 he said that at public readings also it was customary for poets to read only selections from their manuscripts. But the best analogy to Martial's policy appears in Epist. 4.18. The eminent senator Arrius Antoninus had written some epigrams in Greek. Pliny took a few, rendered them in Latin verse, and presented the result to Antoninus with a flattering letter.

¹⁹ Friedlaender in his commentary, vol. i, 52, and in the notes to 1.1; Gilbert, W. in WKP v (1888) 1072Google Scholar.

²⁰ Lehmann, E., Antike Martialausgaben, Diss. Jena (Berlin 1931), p. 18Google Scholar.

²¹ See Friedlaender's commentary, vol. i, 134.

²² I allow no weight to the use of the word libellus in the last line of the poem. Though in Martial's parlance libellus sometimes does mean a little book, it also serves as an indifferent synonym for liber. Or rather, the difference is that liber is a more or less objective word, while libellus carries some affective overtone. See Sage, E. T., TAPA 1 (1919), 168Google Scholar.

²³ In his commentary, vol. i, 67, and vol. ii, 218.

²⁴ The evidence is gathered by Birt, Th., Das Antike Buchwesen (Berlin 1882) 296–7Google Scholar.

²⁵ Birt gives statistics from a sampling of books (supra n. 22).

²⁶ Turner, E. G., Greek Papyri: An Introduction (Oxford 1968) 4CrossRefGoogle Scholar.

²⁷ After Friedlaender's discussion, O. Immisch [Hermes xlii (1911) 482 n. 1] proposed a new interpretation of maior … charta minorque which is grammatically possible but not enticing. According to him, Martial does not have in mind two separate texts at all. Charta maior means the earlier part of Book I, containing the first two epigrams on the hare and lion theme: maior has the sense of longior. That is, these epigrams stand further away from the umbilicus of the roll than the later ones. Charta minor refers to the latter half of Book I, containing the last four epigrams on the theme: they are a shorter distance away from the end of the roll than those in the first section. As for bis idem facimus, Martial can say that because 1.44 divides the six epigrams into two distinct groups. I think that this explication subjects simple language to unnecessary strain, and that the interpretation offered of bis idem facimus seems lame. A couple of other explications of this poem are considered by Helm, R. in Lustrum ii (1957) 190–91 and 202Google Scholar.

²⁸ Proponents of the multiple-edition theory naturally take a different approach to this poem and the one preceding. They contend that 1.101 and 2.91 were inserted when Books I and II were issued for the second time, but that the libelli referred to are those books as they appeared in the first edition.

²⁹ See Friedlaender's commentary, vol. i, 134.

³⁰ There are two other ways in which one might account for this group of epigrams, without postulating libelli. One is that suggested by Sage (supra n. 20, p. 170): each poem stood both in the body of the book where we read it and again at the front of the copy sent privately to the friend or patron named. But this would involve an awkward and unnecessary repetition. It would be better to suppose that the presentation pieces which stand in Book IV, say, had been composed as Martial was circulating copies of Book III, and that he had then simply added them to the store he was accumulating for publication in Book IV. Against this it must be said, however, that whenever he mentions a book number (2.93, 5.15, 6.1, 6.85, 7.17, 8.1, 8.3, and 10.2), it is always the number of the book in which the poem appears, not the number of a book preceding.

³¹ Some may prefer to think that Book XII was published by Martial's executors after his death, rather than by the poet himself; and nothing can be proved against this theory.

³² In 5.60 and 12.61, Martial pretends that even adverse publicity would be welcome to some. Again, the context must be that of published work.

³³ My observations are mostly based on, and illustrated from, the store of examples compiled by Graefenhain, R., De more libros dedicandi apud scriptores Graecos et Romanos obvio (diss. Marburg 1892Google Scholar) and Ruppert, J., Quaestiones ad historiam dedicationis librorum pertinentes (diss. Leipzig 1911Google Scholar). The examples cited to support my contentions are not meant to be exhaustive but illustrative. It is in order to simplify my work that I have selected them almost entirely from Latin writers; my conclusions, therefore, may not adequately describe the Greek practice. The not inconsiderable data provided by Martial's and Statius' works have been excluded because I wished at the beginning to establish the general standards in the light of which these two can then be examined.

³⁴ The Oxford English Dictionary cites a text of the year 1598, the dedicatory epistle to John Florio's dictionary, as the first passage in which the noun ‘dedication’ is used to mean the dedication of a book; it is clear, however, that by this time the word must already have been accepted as the ordinary name for the convention. At about the same time, ‘dedicator’ is established as the proper term for one who dedicates a book. ‘Dedicatee’ is not attested until the late eighteenth century. The word appears to have acquired a generalized value by a similar evolution in France. According to Leiner, W. [Der Widmungsbrief in der französischen Literatur (1580–1715) (Heidelberg 1965), 15–16Google Scholar], ‘Die verschiedenen Varianten des Wortes [dédicace], die in altfranzösischen Texten belegt sind, bezeichnen allerdings zunächst nur die religiöse Weihe einer kultischen Gebäudes oder eines Teiles solcher Bauwerke (‘dediement de l'autel’). In der Bedeutung von Buchzuschrift taucht das Wort erst später auf. Beidem Begriff ‘Dédication’, für den im 16. Jahrhundert zahlreiche Beispiele belegt sind, handelt es sich wohl urn ein Wort, das als Neu-Schöpfung der Entdecker der lateinischen Literatur und nicht etwa als Nachfolger des altfranzösischen ‘dedicacion’ zu betrachten ist. Mit der ausklingenden Epoche der Plejade verschwindet das Wort wieder, um der ‘dédicace’ das Feld zu räumen, das diese dann auch bis heute … behauptet hat.’

³⁵ See TLL v, 260, 60–69 and 966, 69–73.

³⁶ The Thesaurus cites one example of the noun used in this sense, in the heading of a work by Isidore, ‘Dedicatio Historiarum Isidori ad Sisenandum’. The title is quoted from Mommsen's publication of the text in Monumenta Germaniae Historica xi, p. 304; the inquisitive reader will there find that the title is Mommsen's, not Isidore's. Dicatio, on the other hand, has been found with the desired sense in one text which is ancient. The poems of Publilius Optatianus Porphyrius are accompanied by two flowery letters, one from the poet to Constantine in which he presents his poems, the other a reply from emperor to poet. At the close of the latter, Constantine writes ‘Gratum igitur hoc mihi dicationis tuaemunus fuit’ (p. 4 Mueller; p. 41 Kluge). It is not impossible that Constantine used dicatio in an otherwise unattested sense and meant ‘the gift consisting of the dedication of your book’. But it is more likely that he is referring to Optatianus' assertion in the preceding letter, ‘[in scribendo] plus mihi sincerae devotionis studium quam ingenii mei parvitas praestitit’, and that he means ‘the gift inspired by your devotion’. Munus ought to refer to a material object of some kind (in this case, the book), not merely to a complimentary inscription.

³⁷ There are three exceptions, though one is not significant. When Jerome in a letter to Augustine (Epist. 134, vol. viii, 70 of the Budé edition) alludes to ‘libellis tuis quos meo nomini dedicasti’, he is in fact or effect quoting Augustine, acknowledging the complimentary language which Augustine had used. The elder Pliny, however, does seem to generalize the use of dicare: ‘neque enim similis est condicio publicantium et nominatim tibi dicantium’ (HN praef. 6); ‘Sabinus Tiro in libro Cepuricon quem Maecenati dicavit’ (HN 19.177).

³⁸ ‘Iam Varro mihi denuntiavit magnam sane et gravem προσφώνησιν’ (ad Att. 13.12.3); ‘ … Bruto, cui te auctore προσφωνῶ’ (ad Att. 13.21a. 1). At the turn of the first century the word clearly has a general value in Plutarch's vocabulary. Speaking of Blossius of Cumae, who had several works dedicated to him by Antipater of Tarsus, he writes τετιμημένος ὑπ΄ αὐτοῦ προσφωνήσεσι γραμμάτων φιλοσόφων (Ti. Gracchus 8.6). Compare also Lucullus 1.4.

³⁹ ‘An mihi potuit … esse aut gratior ulla salutatio aut ad hoc tempus aptior quam illius libri quome hic adfatus quasi iacentem excitavit?’ Brutus 13; also Brutus 253 and De Senectute I.

⁴⁰ For example, ‘in priore libro has res ad tescriptas, Luci, misimus, Aeli’, quoted from Coelius Antipater in the Ad Herennium (4.18, p. 306 Marx); ‘Apion quidem grammaticus … immortalitate donari a se scripsit ad quos aliqua componebat’ Pliny, HN praef. 25; ‘libros quos ad Marcellum meum de institutione oratoria scripseram’ Quintilian, Epist. ad Tryphonem I.

⁴¹ Lucretius 1.26; Horace, Satires 1.1.1; Seneca, De Vita Beata 1; Tacitus, Dialogus I.

⁴² Ad Herennium 1.1; Vitruvius, De Architectura 1.1; Quintilian, Institutio oratoria I proem. 6; (in the form of letters) Pliny, HN praef.; Pliny, Epist. 1.1.

⁴³ Vergil, Georgics 1.24–42; Ovid, Fasti 1.3–6; Germanicus, Aratea 2–16; Valerius Flaccus, Argonautica 1.7–21. Valerius Maximus prefaces his work with a prose version of such an invocation.

⁴⁴ Ad. Fam. 9.18, which tells Varro that the text of he Academica is on its way to him, is sometimes printed with that work as the dedicatory epistle. Carefully phrased as it is, however, it is not a dedication. The wording of the letter implies that it was sent separately from the text of the Academica; it exhibits the character of an ordinary private letter at the close, where Cicero wishes Varro success in the purchase of a new property; and most important, it did not appear with the ancient text of the Academica. As in the Brutus the ‘dedication’ of the dialogue consisted simply in leading off with the person whom it was intended to honour.

⁴⁶ Catullus 1; Lygdamus [Tibullus] 3.1; Ciris 9 and 46; Panegyricus Messallae 7–16; Censorinus, De die natali I.

⁴⁶ I assume that these verbs are virtual synonyms, since almost every usage can be illustrated as well from the one as from the other.

⁴⁷ As in Pliny, HN praef. 11 and 19; Sidonius, Epist. 3.12.5; Apuleius, Flor. 18 (p. 39 Helm); Statius, Silvae I praef. 35 (where dedicare must be understood as referring to the language of Silvae 1.4.31–7); and Silvae 4 praef. 6 (where the verb is connected with pietas toward the emperor). In the only passage where ὰνατίθημι is used of ‘dedicating’ a book, Plutarch, De Fraterno Amore 478 B, it is given a very religious colouring. (At Sulla 6.10 ὰνατίθημι probably does not mean ‘dedicate’ but ‘ascribe, make responsible for’, pace Liddell and Scott.)

⁴⁸ Silvae I praef. 35, when taken with 1.4.31–7; and Pliny, HN praef. 11 and 19.

⁴⁹ Sacrare is used once by Ovid, Tristia 2.549, of a dedication to the emperor.

⁵⁰ ‘Dire solennellement, proclamer’, Meillet-Vendryes, Dictionnaire Etymologique de la langue latine⁴, s.v. dix, p. 172.

⁵¹ ‘Ii qui propter veteres inimicitias nullo modo cum Haeduis coniungi poterant, se Remis in clientelam dicabant’ Caesar, BG 6.12.7; ‘non vidit eorum ipsorum, qui tum adulescentes Crasso sedicarant, horribilis miserosque casus’ Cicero, De Oratore 3.11; ‘Balbum quanti faciam quamque ei metotum dicaverim, ex ipso scies’ Cicero, Epist. ad Caesarem frag. 1.2; ‘ego tibi de patrono meo M. Porcio gratias ago, quod eum crebro lectitas … nam uni M. Porcio me dedicavi atque despondi atque delegavi’ Fronto, Epist. 2.14 (p. 34 van den Hout); ‘ita institui omne vitae meae tempus vobis probare, quibus me in perpetuum firmiter dedicavi’ Apuleius, Flor. 16 (p. 23 Helm).

⁵² For the nuance of permanence, compare Donatus' note (in his commentary on the Phormio of Terence, line 62) concerning the difference between ‘operam dicare’ and the more common expression ‘operam dare’: ‘plus est ‘dico’ quam ‘do’: dicatur perpetuo, datur ad tempus.’

⁵³ Cicero, Orator 42; Seneca, Epist. Mor. 65.17; Tacitus, Ann. 15.40.1; Pliny, HN 20.113; Paulus, Dig. 15.3.19.

⁵⁴ Pliny, HN 5.16; 10.190; 18.75; Quintilian, Inst. or. 9.3.89.

⁵⁵ Ovid, Fasti I. 19–20; Phaedrus, 3 prol. 62–3; Balbus, Expositio ad Celsum (Blume-Lachmann-Rudorf, Schriften der römischen Feldmesser i, p. 91); Solinus, Collectanea, epist. ad Adventum; Justinus, Epitome Trogi praef. 6. Jerome's famous remark that Cicero ‘emended’ Lucretius' poems may ultimately depend on a letter which made use of this polite convention.

⁵⁶ ‘Sed quid ego plura? Nam longa praefatione vel excusare vel commendare ineptias ineptissimum est. Unum illud praedicendum videtur, cogitare me has meas nugas ita inscribere “hendecasyllabi” … ’ Epist. 1.8.3–4 also makes plain that Pliny is speaking of unfinished work.

⁵⁷ One cannot point to any certain instance in which two or more letters introduce the same piece of work; that, of course, would be the simplest and most obvious argument against ‘dedication’. But Sherwin-White (in his commentary, p. 450) suggested that 8.3, to Sparsus, accompanied the same speech which had gone with 6.33 to Romanus. And Pliny at least promised to send the same book (his verses) to Arrianus in 8.21 and to Mamilianus in 9.16.

⁵⁸ So the Auctor ad Herennium 1.1; Vergil to Maecenas, Georgics 3.41; Hirtius to Balbus, Caesar, BG 8 praef.; Pliny to Septicius Claras, Epist. 1. 1.

⁵⁹ Varro, Rerum rusticarum 2 praef. 6; Parthenius, epist. ad Cornelium Gallum; Cicero, De Officiis 1.1–2; Quintilian, Institutio oratorio I proem. 6.

⁶⁰ In his admirable article in Pauly-Wissowa on the ancient book, Dziatzko had correctly assessed the status of the dedication: ‘Selbst die Widmung und Übersendung eines Buches an einen Freund oder Gönner ist zunächst nur ein privater Act, von der buchhändlerischen, dem Autor vielleicht gar nicht erwüschten Veröffentlichung verschieden und auch nicht notwendig von dem Wunsche privater Verbreitung begleitet' (RE iii (1897) 966–7Google Scholar). That the dedication was unrelated to publication is evident also from the appearance of ‘dedicatory’ passages in small poems which could never have been published in isolation, as for example Vergil's sixth or eighth Eclogue.

⁶¹ Dziatzko (supra n. 58) 968–9.

⁶² For example Pliny, HN praef. 6 ‘neque enim similis est condicio publicantium et nominatim tibi dicantium. turn possem dicere: “quid ista legis, imperator? humili vulgo scripta sunt …” ’; Quintilian, Epist. ad Tryphonem ‘efflagitasti cotidiano convicio, ut libros, quos ad Marcellum meum de institutione oratoria scripseram, iam emittere inciperem'; Pliny, Epist. I.I.I ‘frequenter hortatus esut epistulas, si quas paulo curatius scripsissem, colligerem publicaremque.’

⁶³ It will be well to enumerate here all those pieces in Martial's books which I think certainly or probably performed the function of presentation. The following list includes the few which do stand at the head of a book as well as those which do not: 1.4, 1.52, 1.70, 1.in, 2 pr., 2.93, 3.2, 3.5, 4.8, 4.10, 4.14, 4.29, 4.82, 4.86, 5.1, 5.5, 5.15, 5.18, 5.30, 5.80, 6.1, 7.17, 7.26, 7.28, 7.29, 7.68, 7.72, 7.80, 7.84, 7.97, 7.99, 8 pr. and 1, 8.72, 9 pr., 9.26, 9.58, 9.84, 9.99, 10.3, 10.18, 10.33, 10.64, 10.93, 11.1, 11.15, 11.17, 11.57, 11.106, 12 pr. and 1, 12.2, 12.11.

⁶⁴ 1.111, 3.2, 4.10, 5.18, 7.17, 7.84, 9.58, 9.99, 10.18.

⁶⁵ 4.86, 5.80, 6.1, 7.28, 9.26, 12 pr.

⁶⁶ 1.4, 4.14, 5.30, 5.80, 7.97, 10.64, 11.15, 11.106, 12.1.

⁶⁷ 4.82, 5.6, 7.68, 7.80, 10.93, 12.11.

⁶⁸ 1.52, 7.72, 10.33.

⁶⁹ 3.2, 3.5, 7.26. The personification of the libellus probably derives ultimately from Horace, Epist. 1.20, although Horace did not use it in connection with a dedication.

⁷⁰ The closest analogy I know is the Catullan fragment 14^b, which evidently formed part of the introductory poem of a libellus: ‘Si qui forte mearum ineptiarum / lectores eritis manusque vestras / non horrebitis admovere nobis.’ But this introduction does not carry a ‘dedication’ to any particular person.

⁷¹ The peculiar direction of the epigram to Martialis seems less peculiar when the poem is compared with the openings of other books: eachone, from IV to IX, begins with reference to Domitian.

⁷² Janson, T.(Latin Prose Prefaces: Studies in Literary Conventions, Studia Latina Stockholmensia xiii (Stockholm 1964) 106–12)Google Scholar offers some good observations on these epistles (and also on those of Statius) which partly agree with my own.

⁷³ It is indicative that, of the four privati who gain mention in Martial's ‘dedications’, two (Julius Martialis and Toranius) are obscure characters probably belonging to Martial's own social class, and, conversely, that with the one exception of Terentius Priscus, Martial's most distinguished and frequently named patrons do not figure in dedications at all.

⁷⁴ The letter heading the fifth book: see Vollmer's commentary, p. 3.