SOURCE: Jones, Frederick. “Naming in Pliny's Letters.” Symbolae Osleonses 66 (1991): 147-70.

[In the following essay, Jones studies Pliny's Letters as a means of gaining insight into the social conditions and protocols under which Latin name forms were used.]

Language inevitably makes and enacts presuppositions about the social conditions under which communication takes place. Thus Cicero distinguished private and various kinds of public discourse (ad Famm. 9.21; 15.21), Quintilian distinguished persuasive functions (12.10.59), and writing and speaking (12.10.49f), and stressed the importance of gauging the audience and the circumstances (4.1.52; cf. also [Quint.] Decl. Min. 316.2 with Winterbottom ad loc.), attributing an intelligent formulation to Cicero: eius (= iudicis) vultus saepe ipse rector est dicentis (12.10.56). Vocabulary, syntax, thematic material, stylistic ornamentation provide some of the differentia involved; another way of looking at the differences is to consider utterances in the light of their interactive value, interactive, that is, between speaker and hearer.¹ One aspect of this is name-usage. Nomenclature has been widely used for prosopographical purposes, and to some extent for textual purposes, and there is considerable literature on the development of the tria nomina. However, study of the sociolinguistic protocols applying to the use of personal names in Greek and Latin as a matter of interest and importance in itself is less advanced. Conspicuous are the articles of Adams and Vidman on Cicero and Pliny respectively.² Adams provided considerable detail about the social conditions under which particular name forms (single or double; the various possible combinations of the three parts of the tria nomina) were used. Within narrower limits Vidman performed the same service for Pliny, but since the two studies were independent of each other an opportunity was missed for attempting to identify more general linguistic characteristics. I intend here to consider Pliny's practice against the Ciceronian background (on which I hope to publish further material myself) and thereby identify a paradigm of Latin name-usage in the phenomenon of the introduction.

When we come to Pliny's letters two features of great value in assessing Cicero's usage are either lacking or unreliable. Any original distinctions made in letter headings to the various addressees are thoroughly distorted and overlaid, and (apart from the letters from Trajan or his secretariat in Book 10) there are no letters to Pliny from his correspondents. Moreover, Books 1-9 are avowedly (Ep. [Epistulae] 1.1) tidied up for publication and selected with a literary self-consciousness. Nevertheless the epistolary format is the same so that legitimate comparisons may be made, and it will be seen that there is a great deal of consistency.

The fundamental tendency to denominate with greater fullness (chiefly by using more than one name) for a character's first mention in a letter and thereafter to use single names as anaphorics, clearly visible throughout Cicero's letters, is also present very clearly in Pliny. In Cicero the main variations to this pattern are: (i) the stratification according to the status or distance of the addressee whereby letters to intimates contain a higher proportion of single-naming than letters to superiors or persons with whom Cicero was not on good terms; (ii) the tendency to decrease formality of reference as regards the initial mention in a letter of persons who have been introduced into the correspondence in earlier letters between the same writer and addressee; (iii) the special treatment (in varying degree) accorded to the names of the relatives of writer or addressee, and of women; (iv) the avoidance or replacement of the names of those disliked by, or at friction with, the writer, or whose mention could offend or embarass the addressee, and the avoidance of names for other reasons of discretion. To different degrees there are traces of all these variations in Pliny's letters. In addition, as with Cicero, some special comment on direct and indirect address and self-naming is required.

Of the first of these variations there is little sign in Books 1-9. Single names are scattered through the collection in various functions, but there are no perceptible accumulations in letters to any specific individuals. On the other hand the collocation of 7.7 to Saturninus about Priscus noster and 7.8 to Priscus about Saturninus noster is clearly meant to represent or exemplify a certain level of informality. Furthermore, certain individuals in certain letters receive only a single name apparently because they were well known to writer and addressee.³ The general absence of overall informality in letters to particular friends is a direct result, presumably, of the nature of Pliny's collection: selection, polishing, editing and re-writing with an eye to the public readership conspire to raise the level of literary self-awareness shown, and tend to produce a more even level of formality.

However, this does not mean that such variation is absent from Pliny's correspondence, for there is a noticeable difference between the first nine books and the tenth. In the latter initial references to Roman citizens almost universally use double-naming. The one exception is Virdius Gemellinus introduced into the correspondence by Pliny (10.27) as Gemellino, optimo viro, to which Trajan replies with Virdio Gemellino, procuratore meo (10.28; see too 10.84). Perhaps this may be explained as the result of Trajan's prior reference to the man indicated in Pliny's letter by iussisti.⁴ Not only is double-naming the staple,⁵ but very frequent also are appositional phrases (like those found exceptionally often in Cicero's very formal letters to Appius Claudius in ad Famm. 3); see 10.2.1.; 5.2; 6.1; 10.1; 11.1; 21.1; 25; 27; 28; 29.1; 51.1; 56.2; 56.4; 57.1; 58.3; 60.1; 60.2; 61.5; 63; 65.3; 67.1; 74.1; 75.1; 77.1; 81.1; 84; 85; 86a; 87.1; 87.3; 94.1; 106; 107; 112. Furthermore, while references to any particular individual in the first nine books may, after an initial double-naming, be by single-name, especially within the same letter (where this tendency is very regular), the position in Book 10 is somewhat different. Usually here a reference to a character in a letter later than that in which he first appeared retains the double-name (or triple, where relevant) as at 10.10, 22 and 86a, 30, 57, 61 and 62 and 77, 66, 76, 82, 95, 105, 107. Double-naming is replaced by single-naming only by Trajan (10.60, Paulus and Archippus; 10.71, Claudius) and in one of these cases Pliny retains the double-name after Trajan has moved to single-naming (10.81, Flavius Archippus).⁶ Similarly, Pliny is more likely to retain appositives than Trajan (cf. 10.22, 30, 62, 66, 95 where Trajan omits an appositive; 10.61.5 (cf. 42); 77.1 (cf. 62); 86a (cf. 22.1) for cases where Pliny uses an appositive where Trajan had not).⁷ It is very clear that Trajan is allowed more flexibility and initiative than Pliny, a clear status differentiation, and that the whole correspondence is more formal than that of Books 1-9 where the distance between writer and addressee is consistently smaller.

Comparison between the use of address in Books 1-9 and Book 10, and within Book 10, confirms this picture. Excluding quotations there are only two vocative addresses in Books 1-9, Spurinna at 3.10.1 (in a letter whose heading includes two addressees, one of whom is here singled out) and collega carissime at 7.21.1 (to Cornutus; collega has some rhetorical point). Sherwin-White (n. at 7.21.1) implies that the rarity of the vocative in Books 1-9 is connected with the intention or process of publication. That the incidence of the vocative in these books is unnaturally low might be supposed on grounds of comparison with Cicero, a somewhat suspect argument; of more secure significance is the comparative frequency of name vocatives in quotations in these letters, which argues a certain normal level of address by name (in spoken language) between peers and approximate peers. The reciprocal arrangement at 9.19.5 is particularly relevant.

On this basis it is possible to add weight to an already more substantial contrast between Pliny's mode of addressing Trajan, and the latter's way of addressing Pliny in Book 10. Pliny addresses Trajan with a vocative imperator + adjective twice in 10.1 (because of his recent salutation), “in Ep. 14 for his military victories, and in Ep. 4.1 for no clear reason” (Sherwin-White at 10.1.1). Otherwise he uses domine (sometimes twice) in every letter except 10.41. 59, 64, 86b (where the text is insecure), 102 and 116. On the other hand Trajan uses mi Secunde carissime in fourteen letters, Secunde carissime in five, and mi Secunde in one (10.97): there is no address in thirty-one letters. Sherwin-White observes the omission in the three private rescripts (10.3b; 10.7; 10.9), but this may be insignificant (n. at 10.16). He also suggests that the omission of the address “in several rescripts of substance is probably due to the compilers of the collection” (p. 546). A very clear pattern is visible. Whereas Pliny is, apparently, almost bound to use address (and it is very tempting to regard the six omissions as the results of someone's carelessness), Trajan may or may not (and only does so in about two fifths of the cases). Furthermore, whereas Pliny does not use a name, Trajan when he uses address at all uses the name, and with the familiar mi in sixteen out of twenty-one cases at that. Pliny's usage here bears comparison with his references to Emperors in general (details below), for he tends to use a title only for the current Emperor, but a title + name for favoured dead Emperors. It is clear that with regard to the Emperor Pliny is at the lower end of a non-reciprocal system both for address and for name-use; this system parallels the distinction already observed, that the initiative for reducing formality in third person reference rests mainly with Trajan in Book 10. Although the various non-reciprocal address systems in Cicero are not as nakedly differentiated as mi + name vs domine, there is a definite parallel, and domine had been a common form of upward address between individuals of different status in a broad range of social levels, including from children to parents (see Sherwin-White at 10.2.1). It follows that the situation in Book 10 is simply a more extreme version of the normal linguistic differentiations which depend on and support social inequalities and that the use of someone's name is a privilege, confirment of which is dependent upon such inequalities.⁸

It should be noted that this argument and the preceding material do not depend on Trajan's personal authorship of the rescripts, for Pliny consistently writes as though he were dealing directly with the Emperor, whether this is a mere formality or not (see Sherwin-White, p. 541), and the rescripts are consistently written in the persona of the Emperor, however much a secretary may have had to do with the drafting (see Sherwin-White, pp. 536-546, with 545-6 on the formula mi Secunde carissime).

We can, then, see a general system which distinguishes both Pliny's approach to Trajan and the reverse, and Pliny's approach to Trajan and his approach to his other addressees, in terms of name-usage between the correspondents and in third person references to others. Within Books 1-9 such distinctions are negligible (especially in comparison with the variegations of Cicero's usage), a fact which should be attributed to the effects connected with publication.

Of the second variation of the underlying system in Cicero mentioned at the outset, that characters introduced into a sequence of letters between two individuals may in later letters be referred to with less formality, there are signs in Pliny. We have already observed something of such a tendency in Book 10, where the initiative tends to lie with Trajan, and there is more evidence in the other books.

There is a clear tendency for the objects of recommendation to be introduced with considerable formality, particularly with the tria nomina, in Cicero. The position in Pliny is less clear, but worth observing. In the specific case of petitions for citizen rights in letters to Trajan Pliny uses the tria nomina which he does not do elsewhere (see Vidman, p. 586); in other recommendations to Trajan he uses two names (see 10.4, 94). In recommendations or introductions to others Pliny again uses two names (usually with introductory paraphernalia, such as relatives, qualitities, biographical detail etc.), as at 2.13, 4.15, 7.16, 7.31 (cf. 9.28.2); sometimes the combination praenomen + nomen (as 1.14.6; 2.9.1; 4.4.1.; 4.17.1) is used for the object and/or one of his relatives, which certainly suggests formality (see Vidman, pp. 585, 594); although sometimes this combination distinguishes otherwise homonymous fathers and sons (Sherwin-White at 2.9.1),⁹ this necessity could perhaps be satisfied by using the recommendee's name and, for his father, pater. When the double-name does not include the praenomen a degree of formality is still present, but one must presume that this arises from the specific character of these letters on rather thin grounds: the nature of the collection has already raised the level of formality in general, so that double-naming is not necessarily distinctive in itself, nevertheless the absence of single-names in initial references in these letters (but note 1.24.1) and the presence of introductory formulae suggest that double-naming was an original requirement. A sequence of letters later than the original introduction (which is not recorded in the letters) is given at 7.7, 7.8 and 7.15 and in these the single-name is used, just as in Cicero letters subsequent to a recommendation may show less formality.

Sequences of letters provide further interest. Vidman (p. 586) has noted the tendency in Pliny for two names to be used for a new character, thereafter one, vor allem in demselben Brief (see too p. 591 and cf. Sherwin-White, p. 113), implicitly treating the collection as a single unit as arranged for publication, irrespective of the various addressees.

There are two clear cases of diminishing formality within a particular correspondence in Pliny. The first concerns an introduction and its sequel. In 7.16 Pliny writes to Calpurnius Fabatus, his wife's grandfather, introducing Calestrius Tiro with two names (and biographical details) and promising his assistance to Fabatus. In 7.23 and 7.32 he writes to Fabatus in later stages of the same business, now referring to Tiro with one name only (we might recall here the informality of reference of 7.7, 7.8 and 7.15 referred to above). Somewhat similarly, in 2.11 he writes to Maturus Arrianus about the Priscus trail: at 2.11.23 he mentions a minor matter still hanging over, involving one Hostilius Firminus, and at 2.12.1, to the same addressee, he picks this minor matter up, but now using one name for Firminus. Two other cases show a certain complication. Four letters to Cornelius Ursus and four other letters are involved. In the description of the trial of Julius Bassus Pliny refers to the role of Titius (but the text is not certain) Homullus (4.9.15). In the description of the trial of Varenus Rufus, a sort of follow-on, Homullus receives one name (but Julius Bassus still receives two), at 5.20.6. In the same letter Varenus receives two names (5.20.1), but in two sequels only one (6.5.1; 6.13.1 where Acilius Rufus receives two names for a second time: 5.20.6; 6.13.5). The complication is that after their appearance in letters to Ursus both Homullus and Varenus still receive single-names in letters to other addressees, Homullus at 6.19.3 to Nepos, and Varenus at 6.29.11 to Quadratus, and 7.6.1 and 7.10.1 to Macrinus, more sequels.

This complication becomes rather more troublesome when we see that certain characters appear with two names in a letter to one addressee, and subsequently with one in letters to other addressees. In 1.14 Pliny Pliny writes a letter introducing Minicius Acilianus to Junius Mauricius (Acilianus has two names at 1.14.3). Later Acilianus has one name in a letter to Annius (2.16.1). We might discount this on the grounds that the earlier formality is required by its introductory nature. But this will not apply to Fonteius Magnus who receives two names at his first appearance (5.20.4) to Ursus (unus ex Bithynis explains who he is) and one at 7.6.2 (with ille) and 7.10.1, both to Macrinus. Since these three letters concern stages of the same affair (and the same goes for Homullus and Varenus), and Pliny makes a point of the continuum at 7.10.1, it may begin to look as though the difference in addressees becomes a secondary consideration (in which case ille at 7.6.2 might refer to the man's role in 5.20), and that we are seeing the normal tendency to reduce the number of names after the first mention in the same work. Certainly Pliny was aware of the audience beyond the addressee, and in a letter to Cremutius Ruso (9.19) he notes that Cremutius has read something in quadam epistula mea. Since this refers to a letter in an earlier book (as of course it would have to), 6.10 to Lucceius Albinus, we may see that there is no difficulty in the fact that the sequences we are considering extend over more than one book. And there are more examples. M. Regulus receives two names on his first appearance in the whole collection (1.5.1, to Voconius Romanus) and subsequently only Regulus (1.20.14, to Tacitus; 2.11.22, to Arrianus; 2.20.2, to Calvisius; 4.2.1, to Attius Clemens; 4.7.1, to Catius Lepidus) until 6.2.1 where, in a letter to Arrianus (again) he receives two: but this is an obituary notice and we shall see below that this is a sufficient explanation for the return to formality.¹⁰ Ummidius Quadratus provides another example: two names at 6.11.1 (to Maximus), one at 7.28.6 (to Geminus). So too Julius Frontinus: see 4.8.3; 5.1.5; 9.19.1; all to different addressees, two names only in the first reference. Similarly Julius Servianus is given two names at 7.6.8 to Macrinus and one a 8.23.5 to Marcellinus (the double-name at 10.2.1 is justified by the context of Book 10). Similarly Arulenus Rusticus (two names at 1.5.2 and 1.14.2, one at 3.11.3 and 5.1.8; all to different addressees) and Mettius Carus (two names at 1.5.3 and 7.19.5 (well separated) and one at 7.27.14, all to different addressees).

There are some more puzzling cases. Atilius Crescens receives one name at 1.9.8 (to Fundanus) and 2.14.2 (to Maximus), and two at 6.8.1 (to Priscus). Since in the first two cases the name is accompanied by noster and the last is a letter of commendation we may have a trace of the introductory paradigm so clear in Cicero. Spurinna receives one name at 1.5.8, 3.1.1 and 4.27.5 (all different addressees), and two at 2.7.1 (to a fourth), possibly because of the formal subject matter. Fronto Catius receives two names at 2.11.3 (to Arrianus) and one at 4.9.15 (to Ursus), but then in another letter to Ursus (6.13.2) he is back to two names. This case appears to be a genuine anomaly, but not a serious impediment to the other patterns. Perhaps Fronto was simply not memorable enough to continue to survive on one name (unless the formal context is at work). The double-name references to Licinius Nepos at 6.5.1 (contrast 5.4.2; 5.9.3; 5.13.1; Licinius had been introduced with two names at 4.29.2) and to Corellius Rufus at 7.11.3 (introduced with two names at 1.12.1, but then see 4.17.2; 5.1.5; 7.31.4; 9.13.6) are also hard to explain, but in both cases we should observe the basic pattern of introduction by two names followed by letters to other addresses using single-names which is only slightly disturbed by the anomalous references noted here.

There are, then, traces of progressive decreases in formality in name-usage between correspondents, but they are heavily overlaid with a pattern of decreasing formality irrespective of addressee. To a large extent this justifies Vidman's implication that the collection is in some degree a single literary work (p. 586, cited above). This suggests literary self-awareness and editorial planning and activity. Furthermore, this allows us to see that, in a sense, the reader replaces the various addressees with the result that the new pattern of name-usage may be regarded as a good parallel to the Ciceronian (and thus provides evidence which suggests that in Pliny's original correspondence he followed the same principles).¹¹

It would follow from the tenor of this discussion that the use of a single-name for the first appearance of a character should be distinctive, as indeed it proves to be. But since this raises the question of the hierarchy of name-forms, discussion had better take place in that context. Adams produced the following schema for Cicero (p. 164):—

a) highly formal	three name reference
b) formal	two names
c) informal	one name
d) intimate	mi + one name (voc.) praenomen with or without mi

Adams further observed the greater aristocratic exclusivity of praenomen + cognomen as against nomen + cognomen and (even more) praenomen + nomen. Arguably, Pliny preserves the high formality of the triple name reference in the pleas for citizen rights in letters to Trajan. We have certainly seen evidence that the distinction between (b) and (c) still obtains, and mi + one name (voc.) is characteristic of Trajan's rescripts to Pliny: not intimacy, but the familiarity which the great allow themselves. The praenomen is exceptional in Pliny and needs treatment below. As to the distinct strata of name forms in (b), Pliny shows a considerable difference. The praenomen + cognomen (aristocratic for Cicero) is characteristic of references to historical characters (Sherwin-White, p. 113; Vidman, pp. 593ff); the praenomen + nomen is formal, but not associated with a particular social level; rather it is used for introductions, recommendations and other formal contexts (Sherwin-White, p. 113; Vidman, pp. 594ff). The usual form of reference is nomen + cognomen, but senatorials may be treated to two cognomina instead (Vidman, pp. 588f). This may be seen as a result of the shift in the values of the nomen and the cognomen at the expense of the former which becomes visible after Cicero's time. Vidman's study shows Cornutus Tertullus whose official name, as it were, was C. Iulius Cornutus (p. 588), but also Arulenus Rusticus (two cognomina) cited in the Fasti Ostienses and Potentini as Q. Arulenus Rusticus. Vidman provides further examples (p. 589) showing that the cognomen (or rather the first cognomen) has achieved something of the status of the nomen. Moreover, whereas in Cicero the single nomen is used for men of lower status than the single cognomen, in Pliny the cognomen is the usual name for single-name references. Although that distinction has virtually disappeared in Pliny (see Vidman, p. 593 for single nomina), it may still be the case that the difference between nomen + cognomen and cognomen + cognomen was felt and that the Ciceronian social distinction has shifted rather than disappeared.

As to the single-name there are distinct occasions for its use, some well known. Non-citizens (Vidman, p. 587), Greeks, even some with citizenship (Vidman, p. 589f), and, less markedly and with other peculiarities, women tend to be designated by single-name (women need further discussion below). Otherwise very well known contemporaries, especially senatorials (Vidman, p. 590), or those well known to Pliny and his addressee (Vidman, p. 591f), are so designated as well, the latter with meus or noster. Both usages are reflections of the introductory paradigm where formality is greatest when the person referred to is not known to the addressee/reader or not well known to the writer.

A noticeable sub-category of the very well known is that of certain poets. The Editor of Liverpool Classical Monthly cited in one editorial (LCM 13.10 (1988), 130-131) “Macniece's Eclogue for Christmas (death and eminence have earned the author the right to drop the forename).” Clearly dead people do not necessarily achieve single-names, even famous ones, as we may see from Vidman's list of historical characters who are designated by praenomen + cognomen (p. 594). Indeed the recently dead receive in their obituary notices two names (even if they have been referred to by either double- or single-name already in the collection, and even if other characters are single-named in the obituary letters), unless they are women (4.21; 5.16; 8.5; none named; but even here see 7.24, Ummidia Quadratilla). The only exceptions are more apparent than real, a pupil of the addressee in a consolatio to Julius Genitor (7.30) and a history of the Elder Pliny's death for Tacitus (6.16): neither are named. See in this regard: 1.12; 3.7; 3.21; 5.5; 6.2; 7.24; 8.18; 8.23; cf. also 9.9. Turning to dead poets we see that almost invariably they do receive a single name (I omit Greek poets who have only one name anyway). Thus Ennius and Accius at 5.3.6; Catullus at 1.16.5, 4.14.5 and 4.27.4; Horace at 9.22.2; Lucretius at 4.18.1; Nonianus (apparently) at 1.13.3; Plautus and Terence at 1.16.6 (cf. 6.21.4): Propertius at 6.15.1 and 9.22.1. Calvus is single-named as a poet at 1.16.5 and 4.27.4, but also as an orator at 1.2.2 (with meum), and at 5.3.5 he is double-named as an orator (who also wrote light verse) in a list of orators and public figures. Virgil is anonymously quoted at 5.8.3, 6.20.1, 6.33.1 and 7.20.4. But at 3.7.8 and 5.6.43 he is single-named as a poet. Oddly he is treated to two names (praenomen + nomen) at 5.3.6. The anomaly is curious and one notices the proximity of Ennius Acciusque; but also of Cornelius Nepos and a long list of orators and public figures (5.3.5) double-named. Both Senecas only use the nomen for Virgil, but Quintilian uses the nomen some 45 times and once the same combination as Pliny, P. Vergilius, at 8.3.24. This may have something to do with Virgil's exceptional status and we shall see a similar curiosity in Pliny's treatment of Cicero. Silius Italicus (3.7) and Valerius Martialis (3.21) are double-named, but this is a case of obituary formality (perhaps the same may be said of Valerius Flaccus and Caesius Bassus at Quint. 10.1.90 and 10.1.96). Pomponius Secundus is double-named at 3.5.3 in the title of the Elder Pliny's biography and in a slightly curious way at 7.17.11 (Pomponius Secundus (hic scriptor tragoediarum) …). An element of quality judgement could perhaps be present in that Pomponius was an amateur whose plays and poems were unpopular. Quintilian too uses the double-name in his review of literature (10.1.98), perhaps somewhat unusually in the poets' section). At any rate, Pliny also double-names Seneca in a list of public figures at 5.3.5, while writers as such are separately listed at 5.3.6. Living poets (see 1.16; 6.15; 6.21; 9.22) have double names, but in these cases we are clearly dealing with contemporary figures whose writing is only a part of their life, or whose status as poet is unconfirmed. When a person who writes poem dies, he may become a poet, a historical figure who also wrote, or a non-entity: until then he is, as it were, sub judice, only such a person as his status and character make him, and subject to the normal etiquette of naming. A poet, on the other is distinctive, everybody's property: he may be treated as informally as friends or possessions.

Other kinds of writer seem to be differentiated. Dead orators (apart from Cicero) may be double-named (a list at 5.3.5 including Calvus) or single-named (Calvus at 1.2.2); dead historians are double-named (Cornelius Nepos at 5.3.6; Titus Livius at 2.3.8 and 6.20.5). In Quintillian's review of literature we find Sallustius and Cremutius, but Titus Livius, Servilius Nonianus and Bassus Aufidius (10.1.101-104): under orators (excluding Cicero) we find 7 double-named and 4 single-named. Under poets the proportions are more or less reversed: 8 double-named (a number of these may be explained as obituary or as distinguishing homonymous poets) and 16 different single-named poets. Arguably orators and historians are more likely to retain double-names postumously than certified poets, possibly in part because they tended to be more engaged in public life. Cicero deserves special mention. Quintilian calls him M. Tullius and Cicero (10.1.105f) and Pliny calls him M. Tullius (1.20.4; 3.15.1; 4.8.4; 5.3.5; 9.2.2; 9.26.8), Cicero (1.5.11; 7.4.3, 6), M. Cicero (7.17.13) and Marcus noster (1.2.4., one of the only two single-name praenomina in the collection). The extreme variation is a sign of a special relationship Pliny seems to have felt existed between himself and his model.

Death, eminence and the single-name are found combined also in the case of some dead Emperors. Generally in Books 1-9 dead Emperors are designated by single-name or by name + title (imperator, Caesar, princeps or divus) whereas the current Emperor is designated by title alone (usually imperator, princeps or Caesar, with or without epithets). The only significant difference appearing in Book 10 is that Nerva too is not designated by name, but as Trajan's (adoptive) father (10.1; 10.4; 10.8; 10.58). As to the selection of single-name or name + title for dead Emperors, a pattern emerges, albeit with slightly indeterminate edges: disfavoured Emperors are unlikely to receive the title. Thus Domitian is referred to in thirteen letters by name alone (including two, to and from Trajan), but the title is only used as 4.11.7 (Caesar in a contemporary quotation), 4.11.8 (princeps, where the office is at point)¹² and 10.60 (a letter by Trajan where princeps is used as an anaphoric after the single-name). Conspicuously the name alone is used in 10.65 (by Pliny) in company with Augustus, Vespasian and Titus, each of whom has divus attached to the name. (At 1.12 we find isti latroni in a contemporary quote.) Similarly, Nero is designated by name alone in six letters (1.5; 3.5; 3.7; 5.3; 5.5; 6.31 in a quote), in one of which we find divum Iulium, divum Augustum, divum Nervam, Tiberium Caesarem followed by Neronem enim transeo (5.3.5-6). Galba (2.20) and Vitellius (3.7) are referred to once each and only the single name is used (the context is neutral), but since such a clearly favoured Emperor as Nerva is designated by single name at 4.9.2 (and anaphorically at 4.22.4) against Pliny's usual practice for him, Galba and Vitellius must remain in a shadowy interstice. By contrast, Augustus is always designate by name + divus¹³ (5.3; 8.8; 10.65; 10.79; 10.80; 10.84; the last from Trajan); Vespasian by name + divus (1.14; 10.65) or imperator (3.5) except in a letter which begins with a dense collocation of dead Emperors designated by name alone (4.9: Nerva, Domitian, Vespasian Titus); Nerva has name + divus three times (4.11.14; 5.3; 7.33), name + imperator three times (4.17; 4.22; 7.31, name alone in the letter (4.9) just mentioned. Nerva is also designated by Caesar alone (with aspectual point) at 9.13.22, by princeps alone in a quotation (9.13.23) and by name alone as an anaphoric after name + imperator at 4.22 (similarly ipse imperator in the same letter). Titus is designated by name + divus (10.65); otherwise only at 1.18.3 (unless this is Domitian; see Sherwin-White ad loc.) as part of the set phrase Caesaris amici. Caesar is attached to Tiberius at 5.3.5-6 (in company with Julius, Augustus and Nerva, all with divus, and Nero designated by name alone), and to Claudius at 1.13.3 and 10.70. Caesar and princeps are both used on their own for Claudius in 8.6 where the intention may be to highlight Pallas' outrageous behaviour by concentrating attention on the office. Only at 10.71 is the name alone used, and that by Trajan in a rescript to a letter of Pliny which used Claudius Caesar (10.70).

To sum up: the current Emperor is not to be named (either in direct address or in third person reference), but is designated by title, usually princeps or Caesar. Dead Emperors' names are available (and may have been needed for identification sometimes), but only single-names are used (sometimes with a title too), partly through the public accessibility of fame, partly through complications with praenomina and imperial cognomina. Domitian and Nero are conspicuously designated almost always by name alone, Augustus, Vespasian and Nerva almost always by name + title, except that Nerva becomes Trajan's pater in letters to Trajan. Like poets, Emperors belong to everybody, as it were (unless they are also someone's relative, like Nerva) and therefore are treated with something resembling informality.

A special use of single names deserves mention. Whereas there is a high proportion of double-naming in Books 1-9 (even higher in Book 10), out of some 40 persons (excluding Emperors and Greeks and direct address) designated by name in quotations only four are designated by double-name (Satrius Rufus and Mettius Modestus in 1.5; Domitius Afer in 2.14.10; Publicius Certus in 9.13.17). This seems to suggest a difference between spoken language and written language of a certain level of formality. Somewhat similarly, vocatives are almost lacking in Books 1-9 and in Book 10 name-vocatives are addressed down the social scale not up it: but name-vocatives are comparatively frequent in quotations in Books 1-9 and always use the single-name (Secunde, 1.5.5; Paete, 3.16.6, 13; Curiane, 5.1.6; Suberane, 7.6.11; Secunde, 9.13.8; vir clarissime Veiento, 9.13.19; Vergini, 9.19.5; Cluvi, 9.19.5) even in formal contexts.

That fame, or certain kinds of fame, enable a name to become public property is to some extent tangential to the decreasing formality of reference under investigation. Nevertheless it leads conveniently to the third deviation from the basic system visible in Cicero, viz. a certain reticence about the name of (some) relatives and (some) women, for which it may be regarded as a corollary.

Pliny refers to a number of his relatives in letters within and without his family circle, including letters to Trajan. About his mother he writes to Calpurnius Fabatus (7.11.3), but also to Tacitus (6.16.4, 21; 6.20.4, 5, 12) and Romatius Firmus (1.19.1). About his second wife only to Quadratus (9.13.4). About his second wife's mother only to non-relatives (1.18.3; 3.19.8; 6.10.1; 10.51.1). About his uncle similarly (1.19.1, to Romatius, 3.5.1, 12; 5.8.5; 6.16.1, 12; 6.20.1, 2 to Tacitus). Calpurnius Fabatus is referred to in letters to his daughter, Hispulla (4.19.1; 8.11.3), and to outsiders (5.14.8; 8.20.3; 10.120.2). Calpurnia's father to his father, Calpurnia's grandfather, Fabatus (5.11.2; 6.12.3; 6.30.4), and to his sister, Hispulla (4.19.1; 8.11.1), but not to outsiders (he was dead and hardly news). Calpurnia is referred to in letters to her grandfather and her aunt, but also Pontius (as was her grandfather), Maximus and (official) Trajan. The aunt is referred to in letters to her father, Calpurnia's grandfather, but also Pontius (again) and (official) Trajan. In all of these cases kin-terms are used, never names. On the other hand, Pliny refers to a relative by marriage of his ex-mother-in-law by name (adfinem eius, Caelium Clementem) in a plea to Trajan (10.51.1), and his ex-wife's step-father also by name at 9.13.13 (Bittius Proculus … uxoris meae, quem amiseram, vitricus) in a letter to Quadratus.

The two exceptions are remote and both need to be named in the context. Apart from them there is complete consistency. Pliny's side of the family is never mentioned to members of his wife's, except his mother at 7.11.3, but this is clearly not a policy of exclusion (no letters to members of his own side were published) and Pliny does not seem to have particularly restricted mention of his family to others (naturally mutual relatives will turn up fairly often in correspondence). That his father is never mentioned is not to do with him being a relative, but probably because Pliny lived with his mother, and his father died while Pliny was young. Similar reasons explain the restriction on references to Calpurnia's father. In Pliny's letters, then, we meet a much more specific name-avoidance than is shown in the partial restrictions visible in Cicero, but it is not possible to tell whether this stems from a respect for his relatives in general (as addressees and as subjects), or for his elder relatives (Calpurnia excepted, but other factors enter here), nor can we, therefore, be confident about assessing the importance of the intent to publish as a factor in the final state of the use of kin-terms in the published letters. Nevertheless, the name-avoidance is significant irrespective of this question, for the reading public can be seen as an extension of the class of non-family to which of Pliny's addressees belong.

The relatives of Pliny's addressees, other than the shared relatives already discussed, are treated in very much the same way. Relatives of the following are referred to by kin-terms: Romatius Firmus (who appeared as an addressee in the preceding paragraph as well) at 1.19.1; Calvina at 2.4; Corellia Hispulla at 3.3; Spurinna and Cottia at 3.10.1; Calpurnius Macer at 5.18.1; Servianus at 6.26.1; Quintilian (not the rhetorician) at 6.32.1; Maximus at 6.34.1; Praesens 7.3.1; Genialis at 8.13.1. There are, perhaps, four special cases. Pliny consistently refers to Junius Mauricius' nephews and niece by kin-terms in 1.14 and 2.18. Junius' brother is called frater four times in those letters. However, at 1.14.2 he explains that it is an honour to choose a husband (for Junius' niece) ex quo nasci nepotes Aruleno Rustico deceat. It may be relevant that the name is in an oblique case, but it is certainly of the greatest moment that the use of the name is pregnant, suggesting that the name and all it means must be continued properly (cf. nomina at 3.3.7 and 8.10.3 and cf. Cic. ad Famm. 2.16.5). It is also significant that Pliny usually refers to Avienus as Junius' frater in these two letters. A similar case appears at 6.32.1 to Quintilian: quamvis et ipse sis continentissimus et filiam tuam ita institueris, ut decebat tuam filiam, Tutili neptem … (at 6.32.2 Pliny calls the girl puellae nostrae as a gesture of warmth). At 3.8.1 to Suetonius, Pliny refers to Caesennium Silvanum, propinquum tuum in a response to a letter of recommendation. The name would have been necessary in Suetonius' letter, perhaps here too, and the closeness of the relationship is not great. Finally, in a letter to Statius Sabinus Pliny writes, scribis mihi Sabinam, quae nos reliquit heredes, Modestum servum suum, nusquam liberum esse iussisse (4.10.1; the name, Sabina, recurs at 4.10.2). Such is the pattern developed so far, that despite the homonymity and the fact that Statius was one of Sabina's heirs, one must suppose that if the two were related, they were not closely related. In general we can see in the matter of relatives a much greater strictness about names in Pliny than in Cicero, but, or rather therefore, one misses the distinction that allows Cicero some degree of greater freedom over his own relatives' names than over his addressees' relatives' names. One may also see the virtual disappearance of the praenomen used on its own (or with a kin-term). In Pliny it occurs only at 1.2.4 (Marci nostri = Cicero) and 2.9.2 as an anaphoric, picking up Sexti Eruci mei (2.9.1) where it distinguishes an otherwise homonymous son from his father (see Vidman, p. 595). Neither of these, be it noted, is a relative (note also the cognomen used in address by Thrasea Paetus' wife at 3.16.6, 13).

Only the slightest distinction between male and female relatives can be discerned in the preceding material (because of the almost total ban on names), and that only in the idea that men have a name to pass on and be lived up to. In other references to women there is a distinction, however. The bulk of references to women does not use names. Most often they are denominated uxor, filia or mater and defined in terms of a male relative. Sometimes the genitive of a man's name is attached (3.9.17; 3.16.9; 6.32.1; 9.13.16), most often it is easily supplied from the context. A few are defined in terms of a female relative or patron (4.11.11; 7.19.9; 7.24.2), and municeps nostra (in a quotation at 6.24.2) and amiculam quandam (3.9.13) are quite exceptional. About a third of the references to women in Books 1-9 use names, and of these, references to 8 locate them in relationships to men (1.5.5; 1.12.9; 1.14.6; 2.20.2; 3.9.29; 6.16.8 (on Tascus see Syme, JRS [Journal of Roman Studies] 58 (1968), 140); 9.13.3; 9.13.4), two define them as Vestals (4.11.6; 7.19.1), and two place them in relationships to the remarkable Arria (3.16.1-2; 9.13.3). Eleven individual women are referred to in their own right by name: Aurelia, a distinguished lady signing her will, at 2.20.10; Gratilla, Arria and Fannia, notable exiles, at 3.11.3; Gratilla again at 5.1.8, Arria at 6.24.5 and Fannia at 7.19.1; Sabina, a lady whose will is discussed, 4.10.1, 2; Corellia at 4.17.1 and 7.11.1 (in both letters she is later located in relationships, chiefly to her father); Pomponia Galla, who disinherited her son, at 5.1.1; Gallita, on a charge of adultery, 6.31.4; Attia Viriola, disinherited and taking matters into her own hands, 6.33.2; Ummidia Quadratilla, subject of an obituary character piece (including testamentary details), 7.24.1; Plotina, Trajan's wife (but this is not mentioned), 9.28.1.

The women referred to by name alone and those referred to by name with an indication of their relationship to a male relative form overlapping groups: Corellia, Fannia, Arria are in both. Although there are differences of degree they are all independent, heroic or (Gallita, 6.31.4; Casta, 3.9.29) outrageous. The role of most women would be that of someone's female relative and we have seen that Pliny uses almost only kin-terms for all relatives of his own or of his addressees, whether male or female. In other cases it seems to be presupposed that the woman is an adjunct of her male relative, and to use the name would suggest something unusual or notable about her. Since this could well be ambiguous one may presume that name-avoidance was practised (if one knew the name in the first place). It is certainly striking that in letters of commendation Pliny may fill in the family background of his candidate and may mention a mother or grandmother, but only once does Pliny use the woman's name (1.14.6) and he stresses here probity anyway. In Book 10 the proportion of names increases because Pliny's pleas for his clients need the specification, but the location in relationship to a man remains constant. Only Antonia Maximilla, ornatissima femina (10.5.2) and Furia Prima, Flavius Archippus' accuser (10.60) stand out. As with relatives, so with women, Pliny is in some respects stricter than Cicero. Although prominent women like Arria, Fannia and Quadratilla can be named without any slur and with a freedom not visible in Cicero (Clodia is quite differently treated), the restriction on the names of the bulk of women seems greater and may indicate a polarization at least partly due to conspicuous examples of emancipation.

The fourth variation programmed at the beginning of this section concerned name-avoidance. Quite often Cicero seems ill at ease with using a particular name, either through uncertainty about the social dynamics (see Adams, pp. 163ff) or through distaste for the person (e.g. Clodia, Antony and, at times, Dolabella, Terentia and Pompey) or to avoid giving offense to his addressee (as with Dolabella's name in the letters to Appius Claudius: Dolabella's engagement to Tullia was embarrassing). This last category seems to be absent in Pliny, possibly because sensitive letters of this kind were weeded out before publication. As to the first two types, there is hardly a sign. Isti latroni for Domitian (1.12.8) against Pliny's usual practice for that person is a clear case, but is a quotation of an utterance of Corellius Rufus. Perhaps Regulus' Stoicorum simiam (of Rusticus Arulenus) is another case, but again quoted (1.5.2). The avoidance of Publicius Certus' name until 9.13.13 is in Pliny's own words, but is an imitation in the letter of a rhetorical ploy used by Pliny at the time of the events described. Apart from these insignificant cases Pliny criticizes dead people by name (see Sherwin-White, pp. 54-55), but in these cases the lapse of time since the events is a complicating factor. A letter written at the time might later be considerably clarified by the insertion of names (if, of course, any had been omitted in the first place); a letter written later would be less subject to the emotions that cause avoidance (and would again be the clearer for using names). There is, however, a certain amount of anonymous criticism in Pliny (2.6, quendam; 3.11.2, quibusdam; 6.8.3, cuiusdam … ; 6.17, duo aut tres; 8.22, quidam (8.22.4); 9.12 quidam; 9.27, amici cuiusdam; cf. 9.26.1, quodam oratore … parum grandi et ornato). In some cases it might look as though a name could have been (rather than definitely was) suppressed for publication as a measure of discretion, and this seems possible also in non-critical cases (see 1.24.1, amicus tuus; or was this just a name thatwould mean nothing to the wider audience? 3.9.25, reo gratiosissimo; 3.11.2; 4.17.2, isto; 6.17.1, cuiusdam amici; 7.31.1, discipulum; 8.3.1, cuiusdam eruditissumi; 9.27.1, cuiusdam). Sometimes a moral exemplum for good or ill is constructed and the individual is not terribly important (2.6; 7.26; 8.22; 9.12), but an overlapping category is the anecdotal usage where the characters do not require names, or where names would obtrude too much detail or individuality (1.20.1, quodam docto; 2.3.8, Gaditanum quendam; 2.6.3, qui mihi proximus recumbebat; 3.5.12, quendam ex amicis; 4.9.9; 4.12)—or where the name was simply forgotten. In general, then, Pliny's non-naming (except for slaves and official titles, such as consul etc.) does not suggest the variety or directness of Ciceronian emotions, but rather an eye on the wider audience.

The most consistently comparable feature in the two name-systems is that a character tends to be named more formally on his first appearance in a letter than later in the same letter. At the outset of this discussion I called this a fundamental tendency, and did so on the grounds that it is so common (it is widespread throughout Latin literature). But in fact it is itself only one aspect of a more general tendency, unevenly reflected in literature, that an introduction requires a formal transfer of information which need not be repeated, or fully repeated, thereafter. Material obviously relevant to this tendency was designated as a variation (the second) at the beginning of this discussion simply because of its spasmodic appearance in literature: in Cicero it requires a good sequence of letters between the same correspondents, and in Pliny the blurring of the boundaries between different addressees further reduces detail. Nevertheless, it can still be seen that the introduction provides a paradigm of name-usage: the use of a name is the assumption of a privilege which, once granted, may be continued with less formality. Status, gender and family relationships have clear effects on the degree of restriction there is on the privilege (thus variations (i) and (iii) too fall into the area covered by the paradigm). We may see supporting evidence or parallels for this analysis in the description of an “introduction” in Horace (Sat. 1.6.54ff), in greetings scenes in Roman Comedy and in the role of the nomenclator as intermediary. It may now be suggested that the use of a name is felt to be a privilege because of its special bonding with its bearer. A vast amount of evidence is available in Greek and Latin for this idea, but it is intractible in a confined space. In this context we may at least recall certain pregnant uses of names in Cicero, for example ad Famm. 2.4.1, quid est quod possit graviter a Cicerone scribi ad Curionem, nisi de Republica? (in a letter to Curio); ad Famm. 4.5.5, noli te oblivisci Ciceronem esse; and ad Famm. 9.26.2, in eo, inquis, convivio Cicero ille! Such pregnant uses are found in Pliny and two have already been observed (1.14.2, Aruleno Rustico; 6.32.1, Tutili) and parallels might be added (4.17.2, Corelli, and probably 7.11.3, Corelli Rufi). For such a use in self-reference there is ego Plinius ille quem nosti (1.6.1). Finally, we might consider the anecdote at 9.23.2-3. Some intelligent, so the story goes, bystander asked Tacitus ‘Tacitus es an Plinius?’ The question may help to suggest an aspect of names which has some role in establishing the bond between name and name-bearer, for the speaker is not asking if his neighbour is called Tacitus or Pliny, but if he “is” Tacitus or Pliny. Although this use of esse is hardly unusual, we might, however, look also at the pleonastic use of εĩναι e.g. Hdts 4.33.¹⁴

Notes

1. See e.g. N. E. Collinge, “Thoughts on the pragmatics of Ancient Greek” PCPhS [Proceedings, Cambridge Philological Society] 34 (1988), 1ff.

2. J. N. Adams, CQ 28 (1978), 145f; L. Vidman, Klio 63 (1981), 585f. Cf. also D. Schaps, CQ [Classical Quarterly] 27 (1977), 323ff on women in Greek oratory; A. H. Sommerstein, Quaderni di Storia 11 (1980), 393ff on women in Greek and Roman Comedy. Hereafter Adams and Vidman are cited by name only.

3. See Vidman, p. 590, 591-2.

4. See Sherwin-White, p. 534 on iussisti; see Vidman, p. 587 on the question of the single name.

5. This is also true for persons such as Pompeius Planta, Suetonius Tranquillus and Voconius Romanus, single-named in Books 1-9.

6. Only Pliny replaces a double name with other forms of reference (10.6, for Antonia Maximilla), but this is one of his friends, and the earlier letter was his own, not Trajan's.

7. In three places Trajan or his secretariat uses fuller nomenclature than Pliny's earlier letter: 10.28, responding to Pliny's unique Gemellinus (10.27): Trajan uses the double-name again at 10.84. For this case see Vidman, p. 587. Perhaps there is something still to be said for supplmenting the text. Secondly, 10.57 for which see Vidman, p. 586; thirdly, 10.60.2 (cf. 10.59) where Pliny's appendix must have supplied the name.

8. Pliny's letters to Trajan commonly use such phrases as tua pietas (10.1.1), bonitas tua (10.2.3) etc., which have not yet reached the status of third person substitutes for direct address (ex hypothesi an extremely unpresuming and humble mode in terms of address and of name use), but are related. On this matter see J. Svennung, Anredeformen (Lund, 1958).

9. Praenomen + cognomen is generally used for historical characters: see below.

10. Formal context in an obvious sense explains the later formality at 10.94-95 (cf. 1.24) and 10.2.1 (cf. 8.23.5); but see also 9.13.13 (cf. 4.22.4) and perhaps 4.22.3 (cf. earlier references to Iunius Mauricus at 1.5.10f; 3.11.3; all three letters are to different persons).

11. Certain persons are always, or virtually always, referred to by double-name: Cornutus Tertullus is always so named for his first appearance in a letter except in a quote in 4.17.9; see too Verginius Rufus and Marius Priscus. Herennius Senecio is double-named at 1.5.3 and 4.11.12, but single named in a list of single-named persons at 3.11.3.

12. If 1.18.3 (Caesar) refers to Domitian (see Sherwin-White, ad loc., arguing for Titus) the same may be said, for Caesaris amici is practically a technical term.

13. Julius Caesar is also designated by single-name (Iulius) + divus: 5.3; 8.6.

14. A final observation: there is a very high degree of consistency between the system observed in Pliny's letters and that shown in Tacitus' Dialogus.