This fine collection of essays on friendship in classical antiquity emerged from a conference celebrating the sixty-fifth birthday of Géza Alföldi, held on 10-11 June 2000 at the Seminar für Alte Geschichte in Heidelberg. Of course, I have a personal interest in the subject, as the editor, Michael Peachin, notes in the introduction to the volume: “D. Konstan has recently argued against the majority opinion and has tried to inject more (modern-style?) emotion into ancient amicitia. Various articles presented below, on the other hand, point us back to a heavily formalized, even legalized, bond between friends” (7). Peachin adds, however, that “it is difficult to imagine a world utterly devoid of what we would call truly friendly relationships,” and reasonably supposes that there were, among the Greeks and Romans, both bonds such as “we would readily recognize as proper friendships” and also ties that, although they are labelled amicitia or philia, would not seem “particularly friendly” to us (10).
No one can take exception to this statement, but I would observe that modern “friendship” too names a broad range of relations. A social climber claims to be “a good friend” of the local banker; on an institutional level, British barristers refer to their opponents as “my learned friend,” and allied nations are labelled “friendly powers.” These designations do not imply deep affection, but we do not therefore suppose that “friendship” normally signifies a “formalized” or “legalized” affiliation. Thus, although the philoi of a Hellenistic king were his inner council, and philos and amicus were regular terms for a foreign ally, these conventional uses were, in my view, parasitic on a primary or core meaning of philia and amicitia as an affective bond between individuals, which endured even in the stratified and class-conscious world of the Roman Empire. In what follows, I shall take my cue from Peachin’s observation and attend particularly to how the several contributions affect our undertanding of the emotional element in Graeco-Roman friendship.
In the first four chapters, four Italian scholars divide among themselves the task of investigating amicitia in Latin inscriptions. Silvio Panciera takes on honorary dedications to friends found in Rome and its environs (collected in CIL 6). The sum total of these is ten (or nine, if one pertaining to Ostia is excluded); as Panciera remarks, “the first thing that impresses one is their relative scarcity” in comparison with the huge number of inscriptions honoring senators and knights (14). Even so, Panciera extracts much useful information concerning the domestic rather than public location of these inscriptions and the unequal status of the dedicator and honorand. Panciera concludes that the friendships in question are not “born of reciprocal affection,” but are rather to be understood as “a relation of dependency in which there is a superior amicus who assumes the role of patron and an inferior amicus who occupies the position of protected party,” entailing mutual obligations “in terms of beneficia and officia” (16). But this begs the question of why mentions of amico or amico optimo are so rare: if amicus was simply a polite way of designating a client, we should expect it to appear with some regularity in such contexts. Might we not rather infer that the exceptional instances in which a dedicator specified a relationship of amicitia betokened precisely a more personal tie?
M.L. Caldelli investigates 315 out of 475 inscriptions in and around Rome in which amicus or amica occur, leaving aside the honorary dedications studied by Panciera and the 151 military inscriptions to be examined by Cecilia Ricci. The remainder are all funerary inscriptions, and belong chiefly to the first and second centuries A.D. The vast majority involve people of low social status, including slaves and liberti, and the friendships are mostly among equals. Caldelli takes amicitia to be an institutionalized relation, formally established and dissolved, entailing mutual duties, and arising under specified conditions (22), and finds support for the latter two elements in the epigraphical record. Thus, many inscriptions indicate that the place of burial has been granted or acquired by the dedicator, e.g., CIL 6.10236, in which Calupius Daulicius Paetus declares that he has erected the funerary monument for himself and his freedmen and specifies further that he has granted to his amici, Philemon and Hellas, the right to an aedicula. Again, in CIL 6.10240, T. Fuficius Felix says that he bought the sepulchre for his wife, children, and their freedmen, and as a locum ponendi corporis utrorumque for his friend P. Aelius Natalis and his liberta. Caldelli concludes: “It is possible that behind such gifts one ought to see the munera amicitiae which, according to B. Albanese, constitute one aspect of the mutua officia envisioned by the iura amicitiae” (23, citing B. Albanese, “La struttura della manumissio inter amicos,” Annali del Seminario Giuridico dell’Università di Palermo 29  5-103). I see rather a favor freely bestowed on friends who are outside the circle of dependents, whether family or freedmen, with no suggestion of obligation or “laws of friendship.” On the basis of two inscriptions that describe the deceased as amicissimo avi and amici sui filiis ( CIL 6.27386 and AE 1988 nr. 54), Caldelli writes: “there seems to exist the possibility that the bond of friendship between families was transmitted, and this not just on the level of the elites” (24). Is it not equally plausible that Romans then, like people today, occasionally recognized dear friends of their parents and children (in the latter case, the boys died young) without positing anything so formal as hereditary amicitia ? I would also hesitate to equate the terms sodalis and socius with amicus (24), without denying that on occasion the two categories may overlap in the case of a given individual. Of particular interest is the possibility that amicus or amica was used to express a quasi-matrimonial relationship between slaves (26), although when amicus is used in combination with other words such as sodalis, coniux, and mater (26-27), I am inclined to suspect that it is adjectival and equivalent to carus.
Gian Luca Gregori looks at some 23 inscriptions relating to friendship found in Brixia (modern Brescia); of these, 16 are funerary, 7 honorary (the latter constituting about 10 percent of honorary inscriptions in the region). Like Caldelli (28), Gregori finds that most friendships recorded on funerary inscriptions are between social equals (32, 38), especially among people who have the same or similar occupations. With honorary inscriptions, a difference in rank between dedicator and honorand is to be expected; what, then, does friendship mean in this context? The formula ob insignem eius amicitiae fidem et aetern(am) concordiae laudem ( Suppl. It. n.s. 8 [Brixia] nr. 3, discussed on 36, quoted on 39) perhaps indicates some particularly gracious benefaction on the part of the patron, a tribunus cohortis etc. named C. Bellicius Primus ( patrono is written in smaller letters at the end of the inscription), but surely this need not exclude a genuine sentiment of friendship.
Cecilia Ricci concludes the set by discussing the 155 inscriptions by urban troops in Rome, almost all funerary, which mention friendship. Ricci observes that “the very fact that the word amicus appears only in sepulchral inscriptions [apart from one case], and in any case always in a private context, seems to be symptomatic of the desire, or the need, on the part of the soldiers not to turn this affective bond into an official one” (44). Naturally, military life fostered friendships among soldiers, which again were predominantly between equals (45); very few cases of friendship between a soldier and a civilian are recorded (46). The inscriptions give no indication that there were rituals of friendship among soldiers, which may suggest simply that there were none; nevertheless, more than two thirds of these inscriptions also indicate that the friend was a heres (a ratio peculiar to urban troops), and very likely there was some formal act in recognition of this status (47). Was amicus then a term of art for “heir” in military parlance, connoting a quasi-legal status? Or was there simply a tendency for unmarried veterans to nominate those dearest to them, who were naturally their fellow-servicemen, as their beneficiaries? We cannot read the minds and hearts of these soldiers, but surely the fact that the term amicus was largely reserved for heredes suggests that it represented a powerful emotional tie, which was the motive for the legacy.
The next two chapters deal with the element philos in Greek inscriptions, whether in proper names or common nouns and adjectives. Heikki Solin takes up proper names, covering everything from Philagathos to Kharmophilos, in both Greek and Latin scripts. One may be interested to know that the simple form Philôn is far the commonest name, followed by Philippos, Philokratês, and Theophilos (56; based on the Lexicon of Greek Personal Names vols. I-IIIa); this is doubtless because Philôn is short and sweet, and with a change of accent can, as Solin remarks, be taken to mean “Loving” (55). The first female name in order of frequency is Philoumenê, beating out Philoumenos by nearly two to one. Philippos is less common in Attica than in the islands, I imagine because names in – hippos were considered an aristocratic affectation (cf. Aristophanes Clouds 26-27, 62-74). This is all very well, but it seems to have little to do with the theme of friendship.
Chryssoula Veligianni examines philos and its compounds in connection with the new forms, philokaisar and philosebastos. In the pre-Roman period, the most important category of philos -words was family relations, especially in funerary inscriptions, as indicated by compounds such as philadelphos, philandros, and philoteknos. Right after family members on grave inscriptions come friends. Members of the same club or association might also be identified as friends, but “the function of club member and the quality of philos are not synonymous” (64). In hierarchical relationships, a person might be honored as philos kai euergetês, with the element of patronage made explicit by the second term (65). After an examination of terms such as philotimos and philopatris, Veligianni turns to philorhômaios, which, she argues, is like philellên in indicating a pro-Roman disposition rather than a formal title like amicus populi Romani (67); so too, philokaisar and philosebastos are not equivalent to amicus Caesaris and Augusti respectively (68); the latter are represented in Greek by the simple philos plus the relevant genitive, e.g. tou Sebastou. In compounds, the element philos means rather something like “pro-,” and can be applied to groups like the dêmos and to minor client kings. More especially, philosebastos bears a particular relation to the cult of Caesar, as the name itself might suggest.
Víctor Alonso Troncoso raises the question of the relationship between philia and paideia in the context of Hellenistic palace culture, where the king depended on an inner circle of advisors known as philoi : “Of course, only someone who stands in a close personal relationship to the king can be called a friend” (81), though the relation is founded in the first instance on mutual usefulness (82). More paricularly, Alonso focuses on the so-called basilikoi paides, who were educated together and formed the close ties that characterized suntrophoi. Might the tutors of royal youths later serve as their friends and advisors, as Seneca and Fronto did at Rome (83)? There is some evidence for such a practice, although it is exiguous: Alonso cites, for example, the relationship between the Stoic Persaeus and Antigonos Gonatas’ son Halcyoneus (Diogenes Laertius 7.6-9). It may also be that some of the Alexandrian scholars served in such a pedagogical capacity for the Ptolemies, even if they might not have enjoyed an official status as “friends of the king” (84-86).
What was the composition of the so-called cohors amicorum, or rather philôn ilê (Appian Hisp. 84), apparently consisting of 500 friends, that accompanied Scipio Aemilianus to Spain at the time of the Numantine War (it is plausible that the additional force of 3,500 clients departed slightly later with Scipio’s nephew Buteo)? Was it, as Mommsen supposed, equivalent to the cohors praetoria (cf. Festus 249 Lindsay)? Francisco Pina Polo argues that there never was an official unit called a cohors amicorum; the cohors praetoria, in turn, might include a commander’s friends, or else designated a company of particularly loyal troops (92). Pina points out that ilê in Appian refers normally to a cavalry unit (91), which is presumably the capacity in which Scipio’s friends (as opposed to clients) served; what was unusual in this case was not that Scipio recruited from among his friends, but that their number was so large (93). Historical circumstances explain this exceptional case, according to Pina: with the slave uprising occurring in Sicily, the senate gave Scipio unique permission to levy his own forces for the war in Spain (93-94). Now, by any reckoning, 500 friends exceeds the usual number of a person’s intimate acquaintances; in what relationship did they stand, then, to Scipio? Pina searches the record for the identities of some of the participants, which may have included Lucilius and Polybius (96); might it have been Polybius, I wonder, who first identified the contingent as composed of Scipio’s friends? He had, perhaps, a personal motive for suggesting such a description, and the term might have caught on as a label for the unusual corps of volunteer citizens (or assidui) who signed on to Scipio’s campaign.
John Nicols examines the role of hospitium and its relationship to friendship in Cicero’s Verrine orations, which indeed provide a rich fund of evidence. Nicols distinguishes hospitium from patrocinium, on the basis of the distribution of these terms in the Verrines (99-100); of hospitium, he remarks: “The formalization of the connection typically followed an invitation to dine and/or to reside at the house of one party … and could be transferred to the descendants of both parties” (99), as evidenced by tesserae and tabulae hospitalis, conserved to attest to the relationship. Often the bond of hospitality obtained between equals, but it could extend also to relations of dependency, as indicated by the formula hospes et cliens (101). What relationship did hospitium have to amicitia ? Nicols provides a convenient chart of all cases of hospitium in the Verrines (21 in all, arranged by name of guest; I list them here in order of occurrence: II.2.83, 89, 94, 110, 111 quater, 117 and 118; 3.18; 4.6, 18, 25 bis, 37, 89, 145 bis; 5.109, 110). What emerges from my own review of these passages is that hospitium does not usually coincide with amicitia; rather, they are most often separate and distinct, even when the terms occur together (cf. 2.118: hospites et amici, publice privatimque; Cic. De invent. 1.103: in maiores natu, in hospites, in vicinos, in amicos, etc.). Occasionally, the two categories overlap. Thus, Cicero denounces Verres for betraying, in Sthenius, a friend and a host, the first a sublime ( clarissimum) relation, the latter a sacred ( sanctissimum) one (2.110; cf. 2.18: hospes, inquam, et familiaris; 4.6). Cicero is doubtless being ironical about the friendship; in any case, I do not find support for the claim that Sthenius “is consistently described as the hospes atque amicus of some of the most important Romans of his day” (101). Nicols well describes the iura hospitii,(102-03), but there are no corresponding iura amicitiae (contrast II.3.14, of Verres destroying consuetudinem a maioribus traditam, condicionem amicitiae, ius societatis; cf. also II.3.123). Nor do the tabulae that commemorate acts of hospitality, which Nicols usefully collects and illustrates, seem to mention friendship. The great virtue of Nicols’ chapter, in my view, is just the distinction it enables us to draw between hospitium as a relationship involving reciprocal benefits and obligations, and personal friendship as an elective bond, to which Cicero indeed refers numerous times in the course of the orations, independently of any mention of hospitality (e.g. I.1.3, 18; of Verres’ corrupt friends, I.1.23, etc.). Where the term amicus did indeed have a more formal significance was in the area of international relations, as indicated by the frequent occurrence of the formula socius et amicus as a way of emphasizing the loyalty of allied rulers and populations (I.1.7, 13, 53, II.1.11, 15, 45, 54, 56, 58, 59, 76, 78, 2.121, 122, etc.) This traditional usage, however, does not compromise the sense of the term amicus in the private sphere, at least on the evidence of Cicero’s speeches.
Friendship with Rome was, as Leszek Mrozewicz shows, a risky proposition; while loyalty was expected of an ally, and was often backed up by what was in effect a hostage system in which sons of foreign kings were reared at Rome, Rome interpreted its own obligations according to its interests and might well abandon a friend when circumstances demanded it (109). Such was the experience, for example, of Maroboduus, king of the Marcomanni (Tacitus Annals 2.63), and later of Vannius ( Annals 12.29-30). Mrozewicz examines the role of foreign friendships in the Danube region under the empire, where it formed “the chief instrument of Roman politics” (110). While much of the chapter is devoted to questions of provincial division and adminstration, Mrozewicz concludes that amicitia was employed with particular success in the region by Claudius (117).
The next three chapters take us back to Rome and the world of personal ties of friendship. Edward Champlin takes as his point of departure Syme’s prosopographical analysis of the younger Pliny’s connections with the region around Comum and demonstrates that Pliny’s Tuscan associations, centered on his estate near the town of Tifernum Tiberinum (modern Città di Castello), were at least as dear to him. Among his attachments there figured some of his closest friends, such as Passennus Paulus, Terentius Iunior, and perhaps also Atilius Crescens, whom Pliny “loved … dearly since they were boys” (124), though Syme assigned him to the Comum circle. Champlin concludes that, at Comum, Pliny “repeatedly plays the public man, the eager friend and benefactor,” whereas Tifernum “is the place for otium” and literary friendships (125). His friendships there were intimate, relatively informal, and disinterested.
While Pliny was proud of his many friendships and was doubtless aware that they might “enhance his social status and his political importance,” friendship in Pliny’s letters is, “first of all, an emotional concept,” writes Lukas De Blois (129). de Blois cites such terms as ardentissime diligere (1.14.10) and ardenter amare (2.7.6) as evidence of the intensity of feeling that underlies Pliny’s friendships. To be sure, friendship imposed certain obligations, such as attending a wedding, witnessing a will — when has it not? ” Amicitia was, however, never reduced to a practical exchange of services quid pro quo“; rather it required “moral congeniality ( morum similitudo), reciprocal love, loyalty ( fides), and practical help and support” (130). Of course, letters of recommendation for friends had political significance (131), though I would add that not every such connection was described in the language of amicitia. As de Blois points out (132-33), Pliny, like Dio of Prusa in his third oration on kingship, was also concerned to propagate the image of Trajan as a ruler who fostered and himself depended on ties of friendship, in contrast to Domitian, whom Pliny, not entirely fairly, painted as despotic and hence, according to a tradition going back at least to Xenophon’s Hiero, friendless. I agree that this emphasis on the ruler’s commitment to friendship was peculiar to Trajanic ideology (see my “Friendship and Monarchy: Dio of Prusa’s Third Oration On Kingship,” SO 72  124-43).
Beginning with the idea that Roman friendship may have contained elements “likely to strike us as odd, as rather unfriendly, perhaps even as perverse,” Michael Peachin examines the custom of “entertaining dinner guests with abuse” (135). Peachin recognizes that Romans such as Cicero and Seneca prized intimate conversation over dinner (cf. Ad fam. 9.24.3), but Roman banquets often “reinforced social hierarchies” (136), and selective mockery was among the means for doing so. Slaves or other buffoons and jesters might ridicule their masters’ guests (Seneca Constant. 11.3), and guests might attack one another as well (Petronius Sat. 57-58, cit. Peachin 137). First-time guests were particularly vulnerable to mockery, and the Romans, as Peachin shows, could be quite inventive in the forms of mistreatment they dreamed up for their amusement (as in Elagabalus’ trick of seating fat men together on a small bench). Now, as Peachin acknowledges, “abuse and insult were woven into the fabric of Roman culture” (140), from Fescennine verses at weddings to the railery soldiers aimed at their generals, not to mention forensic oratory and public punishments. In this respect, behavior at convivia“was hardly aberrant; it was typical” (142). Dinner parties were also theatrical occasions, that called for stagey kinds of humor; I suspect too that there may have been an element of ritual, or at least convention, behind such behavior as well (cf. the name Servilius Balatro, “Servile Buffoon,” given to the coarse guest in Horace Satires 2.8.20, with precedents perhaps as far back as Iambe’s jesting in the Homeric Hymn to Demeter 202-04; also Apollonius of Rhodes Argonautica 4.1725-27). Nevertheless, Peachin concludes that the Roman dinner-party scene, “in which friendship was supposed to thrive, is likely to confound our own notions of friendly, convivial intercourse” (143). Perhaps so, and yet, as far as I can see, none of the cases of festive abuse collected by Peachin seems to have occurred between friends, that is, amici, unless we count Caligula’s crude treatment of Valerius Asiaticus, ostensibly “one of his very best friends” (138; cf. Seneca Constant. 18.2). The barbs of professional clowns are hardly such, nor is the demeaning of lowly guests at the table. Roman friends, I venture, did not insult each other at banquets or anywhere else, unless perhaps they were forced to do so by a deranged tyrant.
Some friendships could be dangerous, and in the final chapter Silvia Orlandi concentrates on consulars who were friends of tyranni, that is, usurpers of the Roman throne in the 4th and 5th centuries A.D. Upon the fall of a usurper, his acts might be annulled, but it did not follow that the legitimate ruler, once restored, necessarily persecuted systematically magistrates who served under the previous regime. Constantine, for example, was lenient toward supporters of Maxentius, even though he vigorously carried out the damnatio memoriae of Maxentius himself (147). In particular, mentions of earlier magistracies were not always expunged from inscriptions, and might even be indicated, sometimes obliquely, in new dedications. Orlandi observes in conclusion that careful study of the epigraphical record can shed light on those interludes in imperial history which emperors such as Arcadius and Honorius might wish to have eradicated from the historical record (152; cf. Codex Theod. 15.14.9: tempus tyrannicum, ac si non fuerit, aestimetur).
It should be evident that the several studies in this volume contribute importantly to our understanding of Greek and Roman social relations, in most cases making expert and sophisticated use of epigraphical sources, as befits the honorand Géza Alföldi. Do they, however, “point us back to a heavily formalized, even legalized, bond between friends,” as Peachin affirms? My sense is that the evidence marshalled in them, at least, does not. But it is incumbent on me at least to suggest why scholars of the caliber of the contributors to this volume have mostly drawn a different conclusion. One reason is perhaps methodological. For me, a text is not about friendship unless the parties to the relationship are identified as amici or the like. That they act as hosts, are heirs, or attend the same dinner party does not in itself constitute evidence of amicitia. Again, if two people are identified as amici and something else, for instance dedicator and honorand, patron and client, members of the same sodality, or as sharing the same funerary aedicula, it does not necessarily follow that amicitia is reducible to the latter relation; indeed, one might as reasonably conclude that the precise opposite was the case, particularly if mentions of amici occurred in only a small subset of the relevant documents. To some extent, too, as Ricci observes, the very medium of inscriptions may tend to give an image of a more practical kind of relationship, whereas papyri (and literary texts) “are more obvious evidence of friendship understood as an affective bond as well” (43).
Of course, if one takes what Peachin calls “the standard modern view of Roman friendship, a view that tends to reduce significantly the emotional aspect of the relationship among the Romans [and Greeks], and to make of it a rather pragmatic business” (135 n. 2), it is easier to make the slide from friendship to formalized bonds like clientela, hospitium, and the like. Since I myself believe that philia and amicitia designated a voluntary, affective tie, I incline to a greater philological pedantry in discriminating between explicit mentions of friendship and other social roles. This is not to say that friendship entailed no obligations whatsoever in antiquity, any more than is the case today. Nor is it to deny that in certain formulaic contexts, then as now, the term “friend” might have acquired a more status-like significance: I am thinking especially of the sphere of foreign relations, but the formula “friends of Caesar” and the like may also have designated an official rather than a sentimental connection. If we wish, however, to understand what personal friendship was like in classical antiquity, I still believe that we do well to start by presupposing that the terms amicus, philos, and their congeners meant what Greek and Roman thinkers such as Aristotle and Cicero said they meant and not hastening to assimilate them to other, more formal relationships. Approached in this way, the sources, including inscriptions, may well incline one to conclude, with de Blois, that amicitia was “never reduced to a practical exchange of services.”