Main Text

pg 89Part II Seneca's Treatise pg 90

pg 914The Date and Addressee

THE DATE

There is general agreement among scholars on the chronological limits for the publication of Seneca's essay. First, the Emperor Claudius must have been dead when Seneca wrote of him at 1. 15. 6, 'Should what was offered by Claudius not have been accepted?' (Non erat accipiendum a Claudio quod dabatur?), since what Seneca goes on to say of that Emperor's judgement is so disparaging. In fact this terminus post quem of 54 can be advanced to 56 by what is said at 2. 21. 6 of the consular Caninius Rebilus, who committed suicide in that year (Ann. 13. 30. 2):1 in a comparison with another dissolute senator, he is called 'a person of similarly bad reputation' (homo eiusdem infamiae). Seneca does not often refer to living contemporaries and never in such insulting terms.2

As for the latest possible date, we have the valuable evidence of Ep. 81. 3. Here Seneca, after answering a purported complaint from Lucilius about the ingratitude he has experienced, first offers some advice on the subject and then adds, 'But we have said enough about this topic in those books which are entitled On Benefits' (sed de isto satis multa in iis libris locuti sumus, qui de beneficiis inscribuntur). The points Seneca has just made about ingratitude can pg 92indeed be found in De beneficiis, notably in Books 1 and 7, but also in Book 4.3 That would suggest that he is referring, in the line quoted, to the whole work, allowing us to place its publication before the 'dramatic date' of the Letter, which is the late spring or early summer of 64, before the end of June.4 However, Seneca goes on to discuss a question that he says has been explained, but not sufficiently explained, namely, whether a person who has benefited us, but and harmed us afterwards, thereby frees us from obligation. This question is touched on in Book 3. 12. 4 of the treatise, and more substantially discussed in Book 6. 4–6. Seneca is presumably referring to these discussions. Hence the idea often advanced by scholars, that it is only the first six books that can be confidently dated on the basis of Letter 81.5 But since the Letters themselves were intended for publication, not just for the eyes of their addressee Lucilius (Ep. 21. 3–5), Seneca's reference to De beneficiis in the sentence before must be pointing his readers towards a work that was published and available at the dramatic date of the letter.6 Now it is unlikely that Book 7 was published separately from Books 1–6, given the symmetry it displays with Book 1, as will be shown in Chapter 6. In fact, some of the differences that induced Préchac to believe that the last book was published later are precisely what makes it an integral part of the work, which it would be essential for the reader to have before him from the start. For, as will be argued in Chapter 7, this book is the culmination of the development of thought in the course of the treatise, a development that implements Seneca's pedagogic strategy.

The chronological limits of 56 and June 64 for the work as a whole allow us to supply only a broad context. As regards larger historical events, the work was written, certainly published, in the reign of Nero, after the death of Britannicus, and before the Great Fire of Rome that broke out on 19 July 64 and lasted nine days. The persecution of the Christians had not yet taken place, nor the Pisonian conspiracy. As for Seneca himself, he had not yet openly retired from public life, and he had not yet contributed a substantial portion of pg 93his wealth to the rebuilding of Rome, both of which he was to do after the Fire.7 His addressee Aebutius Liberalis had not yet seen his town of Lugdunum make a contribution to those repairs, only to be laid waste by fire later that same summer.8 As for its relation to Seneca's other philosophical writings, the treatise can be placed after the Apocolocyntosis, the Consolationes to Marcia, Helvia, and Polybius, De ira, De brevitate vitae, and probably De clementia, though that work could just be contemporaneous with De beneficiis.9 The treatise belongs before the later Letters to Lucilius (at least those after Ep. 80), but its relation to five of the Dialogi, to the Natural Questions, and to the early Letters remains uncertain.10

Scholars have naturally found it frustrating not to have more precise knowledge of the background to the work.11 Was the year 59, when Nero murdered his mother Agrippina, already in the past? Had Tacitus' turning-point of 62, marked by the death of Afranius Burrus, Praetorian Prefect and Seneca's political ally, and by the cruel murder of Nero's wife Octavia, already occurred?12 As for Seneca, had he already written the shameful letter explaining Agrippina's death? Was he already semi-retired, refusing from 62 to hold the usual morning receptions and rarely going out when in Rome, pleading ill-health and devotion to philosophy?13 Was he in Rome or travelling in Campania and to his villas, as we see him in the early Letters?

Given the reluctance of Seneca himself to mention current events or his current public position, or even to give many details of his private life, it is not surprising that scholars have turned to a variety of indirect techniques, in an effort to narrow the chronological limits of 56 and summer 64. The most common is the hunt for (i) allusions to historical events datable through outside evidence and (ii) allusions to Seneca's circumstances known from the historical sources. Both types are usually disputable, as to whether or not they really are pg 94allusions or, if they are, as to what precisely they indicate in terms of chronology. More subtle is (iii) the negative argument that certain people or events could not have been mentioned, or certain things could not have been said, at a particular period of time.14 Finally, (iv) similarities to, or contrasts with, what he says in other works have been used to argue for dates close or distant to them.

The general difficulties involved in dating Seneca's works have been discussed many times. I surveyed at length the speculative techniques just mentioned in the Introduction to Griffin 1976, where I decided to avoid them, because they would have introduced circularity into my investigation there of the relation of Seneca's life to his writings. My own attempt at narrowing the chronological limits for De beneficiis to before his semi-retirement in 62 (Appendix A1.G, 399) was based on the sheer volume of work that we know Seneca composed after that date: the Natural Questions, most of the Letters, and the lost Moralis Philosophia, mentioned in the Letters. Scepticism has been expressed, and rightly.15 It can be objected that in the early 60s Seneca was busy composing some of the tragedies, but also, more fundamentally, that we understand little of Seneca's methods or rates of composition.

To review all of the suggestions advanced with a view to narrowing the chronological limits would be tedious, if not pointless. Instead, I give some examples, to illustrate the difficulties these techniques present, and to support the idea that it is better to rest content with the limits of 56 and summer 64.

(i) Alleged allusions to historical events. In 2. 7. 2–8 Seneca blames Tiberius for his censorious way of dealing with requests for cash by impoverished senators (see ad 2. 7. 2–8. 2). It is not unreasonable to connect what he gives explicitly as his own opinion, 'It is not really proper even for the emperor to give a gift in order to humiliate', with Nero's own generosity in 58, when he made annual grants to some noble senators (Tac. Ann. 13. 34). But how? Friedrich, followed by Münscher, regarded it as a flattering allusion to Nero's generosity;16 but since Seneca was influential in 58 and Nero was probably following his advice then, Seneca could have expressed what is his own view, before or after Nero implemented it. Then again, Grimal, followed by Chaumartin,17 thinks that Nero's generosity after Agrippina's death (Dio 61. 18. 1; Suet. Nero 11. 2) provided the occasion for a treatise on generosity. But Nero 's generosity was lavish on many occasions, as the Suetonian passage shows; and, in any case, De beneficiis is critical of such indiscriminate largesse (p. 76).

pg 95(ii) Alleged allusions to Seneca's personal circumstances. Many scholars have assumed that Seneca's offer in 62, to return some of the riches that he had received from the Emperor, is somehow reflected in De beneficiis.18 But how? In 2. 19. 2 Seneca emphasizes that a benefit must be something one receives willingly: if a tyrant forces one to accept something, it is not a benefit, and one is under no obligation for it. Préchac assumed this was Seneca's answer to Nero's insulting refusal to accept the return of his wealth; Grimal felt that Seneca would not have said this after the refusal, because Nero would have seen it as an insult.19

(iii) This brings us to the negative approach to allusions, which has been widely applied to this work. Thus Abel 1985, 708 found a date before the Great Fire of July 64 confirmed by the praise of Agrippa's buildings in 3. 32. 4, because that shows that Rome was not yet damaged by the Fire. In fact, most of Agrippa's buildings (see ad 3. 32. 4) were on the Campus Martius, which was not damaged, and were used by Nero to provide relief housing for victims of the Fire (Tac. Ann. 15. 39. 2).

The difficulty of allusion-hunting in Seneca is in fact well illustrated by the Great Fire. Seneca did not mention this important event even in those Letters certainly written after that event, though it could have been used to demonstrate the fragility of human life and of tangible possessions,. Indeed in Ep. 91. 13, writing about the later fire at Lugdunum, Seneca allows himself only the most general reference to fires at Rome, saying that the historian Timagenes, who lived in the time of the Emperor Augustus, resented them because he knew that even better buildings would arise. It would be natural to believe that Seneca wrote this Letter before the July 64 fire, if we did not know for certain, from Tacitus Ann. 16. 13. 3, that the fire at Lugdunum postdated that at Rome (see below, p. 97). For teaching purposes, the philosopher apparently thought that a less notorious event, of personal significance to a friend, would be more effective.20 We may safely disregard alleged allusions to the Great Fire in the later books of De beneficiis 6. 37. 3; 7. 19. 8; 7. 31. 5,21 all of which are of even greater generality.

(iv) The relationship to Seneca's other works is hard to fathom. Many scholars have remarked the similarity of discussions in De beneficiis to the concerns in De vita beata, datable to 58, about wealth and its use; to the theological pg 96arguments in the undatable De providentia; and to the themes of the early Letters of 63–4, such as the treatment of slaves (Epp. 31. 11; 44. 4; 47; Ben. 3. 18–28) or the respect owed to teachers (Epp. 64. 9; 73. 4; Ben. 6. 16). Is any one of these affinities to be explained by chronological proximity? In fact, the chronological relation of this work to others by Seneca cannot be established through thematic comparison.22 Since, as he says, philosophy teaches us most of all to be grateful for benefits and to return them (Ep. 73. 9), it is not surprising that the subject occurs in various works (see Ch. 8). Then again, negative remarks about Alexander the Great, common in the treatise (see ad 1. 13. 1–3), occur in many works from De ira on, and Phalaris is castigated in De ira and De clementia, as well as in De beneficiis and the Letters. Closest to Seneca's concerns in De beneficiis is De vita beata, datable to 58, but the attitude to wealth is different (see Ch. 8). It is important to remember that Seneca is not just producing a record of his own thoughts. Like all writers of works intended for publication, he has his readers in mind. What he writes to a particular addressee must strike them as appropriate to his circumstances, whether real or invented. Moreover, Seneca is addressing his readers as a teacher, so that events tend to be paradigms, and people exempla. It follows that, given similar subject-matter, similar opinions can occur in different contexts at different times, while contrasting opinions can emerge in different contexts close in time. Especially in a writer whose extant works belong to his maturity, we cannot make chronological schemes out of such repetitions and differences.

THE ADDRESSEE

We have noted that Seneca depicts his addressee Aebutius Liberalis as a rich and generous benefactor (Ch. 1), and we shall have more to say in Chapter 7 about Seneca's use of him as an exemplum at the start of Book 5.23 Just after this presentation of Liberalis as an exemplum, Seneca draws attention to his name, at 5. 3, by inverting his nomen and cognomen and addressing him 'to you, Liberalis Aebutius, the best of men by nature and prone to benefits' (homini natura optimo et ad beneficia propenso, Liberalis Aebuti), whereas elsewhere he is just 'Liberalis', 'mi Liberalis' or 'Aebuti Liberalis' (ad 5. 1. 3 Liberalis Aebuti). Liberalis' name is certainly deliciously appropriate for the addressee of a work on generosity. Should we assume that Seneca chose Aebutius Liberalis from pg 97among his friends as addressee for the work because his cognomen was so appropriate and his lifestyle suitable,24 or should we suppose that Seneca invented characteristics to go with his friend's cognomen?

One idea we can rule out is that Seneca invented him altogether. Against Liberalis' being a fiction stands the fact that Seneca mentions him again as a friend in Letter 91, where he acquires a more concrete identity as a citizen of Lugdunum, inconsolable at the destruction of his patria by fire.25 Nor is there any reason to doubt that he actually had the suitable lifestyle attributed to him in De beneficiis, where Liberalis is assumed to be not only a rich man, but an educated man who can pick up literary references without identification, and who understands the working of the civil law (e.g. Ben. 6. 5. 4–5). He seems to be an eques, like so many of Seneca's correspondents—Lucilius, Annaeus Serenus, Pompeius Paulinus. That the real Aebutius Liberalis had these characteristics is perfectly credible. Lugdunum was a rich and prosperous Roman colony, presumably well supplied with equites Romani. Seneca describes it as a Roman city located in a province, rather than a provincial town (Ep. 91. 10). According to him, the fire, which he reports as a recent event, completely destroyed the city. Now Lugdunum is known from Tacitus to have made a substantial contribution to the rebuilding of Rome after the Great Fire there in mid-July 64, a gesture for which it was rewarded by the Emperor, who had its contribution repaid for use towards its own reconstruction in 65 (Ann. 16.13. 3). The date for the fire of Lugdunum must therefore be after mid-July 64 and before the end of 65, and is probably before the autumn of 64, according to the dramatic date of Seneca's Letters.26 As a wealthy man, Liberalis may have contributed to his city's benefaction to Rome.

Seneca offers the comforting thought that the city may rise from its ruins even better than before, and he predicts that its citizens will contribute to that result (Ep. 91. 14). In the end, the Emperor had to help, and Liberalis' grief, mentioned repeatedly (§§1, 3, 13), may have been compounded by financial loss.27 Bedon 1991 argues that the lack of archaeological evidence for serious damage, and the lateness and small scale of Rome's help show that the fire was not as great as Seneca suggests, but Rome's own need for massive reconstruction counts against the last argument, and individuals could still be pg 98seriously affected, if their property was in the wrong place. Seneca could not have mentioned this mundane fact in a letter exploring, in elevated language, the themes common in consolation literature. But two inscriptions from Dalmatia, ILS 5953 and 5953a, may provide a clue. They reveal a Quintus Aebutius Liberalis serving as a centurion in the primi ordines, that is as one of the centurions of the first cohort, of the eleventh legion. He is in fact the most junior of these, the hastatus posterior. The legate of the province, under whom he is carrying out boundary rectification, is A. Ducenius Geminus, whose term as legate of Dalmatia can be fixed to between 62, when he served on a finance commission composed of consulars, and 19 January 69, when he is attested at Rome.28 Centurions were recruited from the ranks of the legions or the praetorian guard, or they were directly commissioned from men of equestrian rank who had often held municipal office. The latter, described as ex equite Romano (e.g. ILS 2654–6) did not lose their equestrian status, and they had the best chance of promotion to the rank of primuspilus, which was highly lucrative. Promotion was via the primi ordines, and these centurions earned about twice what a praetorian soldier did, and about half of what a primuspilus earned. The primary reason why an eques would wish to become a centurion was money. Frontinus (Strat. 4. 6. 4) speaks of a young man, without military talent, becoming a centurion because of his poor family circumstances, and later Pertinax preferred it to entering the militia equestris (SHA Pert. 5–6), which did not guarantee the same lengthy period of lucrative employment.29

Various scholars have suggested that the centurion in Dalmatia, serving under Ducenius Geminus, is Seneca's addressee. There is no obstacle. If Aebutius Liberalis did indeed become a centurion because of financial losses in the fire at Lugdunum, that would have been no earlier than the latter part of 64 and before 69: in any case, some time after Seneca had written De beneficiis. We can assume that both addressee and author were rich men when that work was composed.

pg 995The Title

Only a few of the scholars who have studied De beneficiis have addressed the question of the title, at least explicitly. Sonntag and Gomoll, however, mention in passing that Seneca renders charis as beneficium in the absence of a Latin word that, like the Greek, could be used to cover both sides of the exchange: giving and returning.1 As we shall see, this is a gross over-simplification.

GREEK WORKS ON THE SUBJECT

The philosophical tradition to which Seneca's treatise belongs is that established by Theophrastus and later Hellenistic philosophers who wrote works wholly and specifically devoted to the topic of charis (see Ch. 2A). Works Peri charitos are attested for Theophrastus, Demetrius of Phalerum, Epicurus, Dionysius the renegade Stoic, Philodemus, and the Stoic Cleanthes.2 It is common to translate Peri charitos as 'On Gratitude', and the full title of Epicurus' work, Peri dōrōn kai charitos, like the title attested for Dionysius, Peri ploutou kai charitos kai timōrias, suggests that this is correct, for that would yield the sense 'On Gifts and Gratitude' (natural correlatives), for the first, and 'On Wealth and Gratitude and Revenge', for the second:3 timōria would be revenge for an injury parallel to gratitude for a benefit, as in Seneca's 'Benefit and injury are the opposites of each other' and 'You conferred a benefit, and afterwards inflicted an injury; gratitude was owed for the benefit, and revenge for pg 100the injury'.4 Moreover, as Moussy shows, of the abstract meanings of charis, gratitude for benefits is not only earlier, but remains more common, than the sense of favour or good will leading to their conferral.5 The translation of the title Peri charitos as 'On Gratitude' is further supported by the emphasis in Seneca's work on ingratitude, especially as the Stoic paradoxes on the subject, which he discusses, suggest that this emphasis was already in the Greek Stoic authors he read (Ben. 5. 12. 3–17. 5).

The titles of works by Stoics and other philosophers on this subject are usually given, as we have seen, in the singular. But the first work that Seneca himself indicates as a forerunner was an exception. Early in De beneficiis, at 1. 3. 8, he alludes to a work by Chrysippus which is clearly that elsewhere attested as Peri charitōn (SVF ii. 1081). Chaumartin may be right to argue that there is philosophically no significant difference between singular and plural titles of this type.6 Chrysippus may have used the plural because he wished to include in the work an elaborate allegory of the three Graces (Charites) representing three aspects of the exchange of benefits—giving, receiving, and returning (Ben. 1. 3. 8)—and thus to stress reciprocity (see below, p. 103). However, when used in the plural, the charis of the title now had, not the abstract sense of gratitude, but its original and concrete sense, in the context of beneficence, of objects or services given or, secondarily, returned: a sense already well established in Homer.7 This could have aroused expectations that he would discuss the practicalities of exchange in more detail than philosophers usually did.8

pg 101The allegory has usually been described as a Stoic invention.9 This is plausible, because the Stoics regularly interpreted anthropomorphic gods as allegories for the benefits of divine providence, and also looked in the traditional poets for the seeds of their philosophical wisdom.10 Moreover, this allegory is found in the Stoic philosopher Cornutus (SVF ii. 1083), and at least one of the two versions of it given by Seneca (Ben. 1. 3. 3), probably the first, should be assigned to the Stoic Chrysippus, as Seneca goes on at 1. 3. 8 to criticize him for including stories of this kind in his work, and thus for 'saying very little about the duty itself of doing, accepting, or returning a benefit'.11 The implication seems to be that his allegory covered these principal aspects of the subject.

It has been argued that this Stoic invention sheds no light on the ordinary Greek use of the term charis.12 However, the allegorical interpretation of the Graces involves understanding charis, as we have noted, not in the abstract sense of 'grace' or 'joy' in which it was originally applied to the Graces, but in its concrete sense of a gift causing joy, a benefit given or returned,13 and this association of the Charites with the exchange of concrete benefits was anything but remote from ordinary Greek usage, at the time when Chrysippus was writing. The Hellenistic polis had shown the way here, while Aristotle had already provided this usage with philosophical credentials. In the course of his discussion of reciprocity in the Nicomachean Ethics 5. 5. 1133a3–5, Aristotle remarks that shrines are set up to the Charites 'to promote the requital of services; for this is characteristic of charis: one should repay a service done one, and should another time take the initiative in doing a service oneself.'14 The epigraphic evidence pg 102bears him out. Not only at Claros and Teos15 were the Charites interpreted as symbols of gratitude and associated with the repayment of the favours of benefactors through decrees that conferred honours and privileges: at Athens itself, where the Charites had long had a special shrine at the entrance to the Acropolis, a new precinct, to Demos and the Charites, was built northwest of the Agora, probably in the 220s bc. In it, honorary decrees for foreign benefactors were displayed, to demonstrate the gratitude of the Athenian people. The precinct, it has been persuasively argued, had been set up to celebrate the liberation of the city in 229 bc.16 Chrysippus would have been in Athens at that time, having become head of the Stoa c.232 bc. The precinct became the place where statues and inscriptions were erected, honouring benefactors to the city.17 Whatever further significance the cult had for the Athenians, the sanctuary was the place where Athenians expressed their gratitude to foreigners.18

In such decrees the usual formulae use euergetēs, euergetein, euergetēma of the benefactor or his service to the city, and charis or charites of the return made by the city. A typical example in one of the municipal decrees reads: ὅπως φανερὸν ἦι πᾶσιν, ὅτι τοῖς εὐεργέταις ἀποδιδῶι ἡ πόλις ἀξίας χάριτας τῶν εὐεργετημάτων‎ ('that it may be clear to all that the city makes worthy returns to its benefactors').19 However, the use of charis for a benefit given is also common in Hellenistic epigrams and funerary inscriptions, while Plutarch speaks of the charites of Flamininus, and imperial grants are called hai tōn Sebastōn charites.20 Chrysippus was therefore following Greek usage familiar in his own day, both in using charis in its concrete sense of benefits given and benefits returned, and in developing his allegory of the Charites on the basis of this interpretation.

In employing charis in connection with taking the initiative in giving, as well as with showing gratitude in receiving, Chrysippus was also within the philosophical tradition. Aristotle in his Rhetoric (2. 7. 1385a17–19) uses the word both in the concrete sense of objects or services given, and in the abstract sense of good will motivating such a gift, while his discussion of the cult of the Charites in the Nicomachean Ethics stresses its use on both sides of an exchange. The Peripatetics recognized charis as the ἀρχὴ τῆς εῦεργετικῆς φιλίας‎ ('the source of beneficent friendship') (Stob. 143. 5–8 W). The Stoics clearly emphasized pg 103the duty of giving, as is shown in the Eurynome part of the allegory, which is not only attributed to Chrysippus (Ben. 1. 3. 9), but found in Seneca's contemporary Cornutus (SVF ii. 1083): the name of the mother of the Graces—Eurynome—was explained by the wide distribution of benefits that is characteristic of an extensive fortune.21 Chrysippus' use both of the plural of charis in his title and of the images of the Graces points to his desire to emphasize the element of reciprocity in beneficence, as is made clear by Philodemus, who says that he represented the Charites as 'our initiations and returns of favours'. Seneca writes of the 'most honourable rivalry in outdoing benefits by benefits, to which Chrysippus thus urges us'.22

WHY NOT DE GRATIA?

Seneca uses the Latin Gratiae to render Χάριτες‎, when discussing Chrysippus' allegory (though he also uses Charites). Though there was no etymological connection between these Greek and Latin terms, the connection in meaning had been made already in the time of Plautus.23 The reinterpretation of the Charites as symbols of gratitude in Aristotle's time will have made the equation easier, for the primary meaning of gratia is 'gratitude', and all meanings of the term in the classical period develop easily from that fundamental meaning. The word is used in the active abstract sense of 'a feeling of gratitude', as in gratiam habere; it is also used in the active concrete sense of attesting one's gratitude: by verbal thanks, as in gratias agere, or by giving a service in return, as in gratiam referre.24

Gomoll, as we said, found the reason for Seneca's irritation with Chrysippus' allegory in his frustration at not being able to reproduce in Latin the Greek philosopher's title, inspired by that allegory: Latin, he said, cannot use the same word to cover both sides of an exchange of favours.25 However, gratia can in fact be used of a benefit without a preceding favour. Both concrete senses, a pg 104benefit given and a benefit returned, are found from the time of Plautus,26 though gratia as a primary gift is not used as the object of verbs of giving: gratiam dare is not a classical Latin expression. The process tends to be seen from the receiving end, as in gratiam accipere ('to receive a favour'), or gratiam remunerare ('to repay a favour'), or gratiam exigere ('to demand a favour').27 Nonetheless, the sense of a benefit given (not in reciprocation), though less common, leads to gratia sometimes being used in a doublet with beneficium.28 Cicero even uses gratia in the plural in this concrete, non-reciprocating sense: non excellentibus gratiis paucorum, sed universi populi iudicio, consulem ita factum ('elected consul, not through the outstanding favours of a clique, but through the esteem of the whole people').29 So what Sonntag and Gomoll and, still later, Schwarzenberg say, is not strictly true:30 Latin does have a word which, like charis, can be used both of what is given and of what is returned. Indeed gratia also has, as its less common active abstract sense, that of the 'goodwill' or 'favour' initiating benefits, just like charis.

It is, of course, true that it is more common in Latin to use beneficium of what is given and gratia of what is returned, and that the most common verbal phrases with gratia, given above, are connected with the notions of recovery and gratitude. Nor is it difficult to find examples of beneficium and gratia used as correlatives of each other, like the Greek euergetēma and charis.31 Seneca himself shows a clear preference for using gratia in its active sense, to mean (in the abstract) 'gratitude' and (in its concrete meaning) 'attestation of gratitude (by words or recompense)'.32

Nonetheless, it is clear that Seneca could have preserved the link with Chrysippus' allegory of the Gratiae by calling his work De gratia. He would pg 105have had to raise the minority active meanings of the word—i.e. (concrete) benefit without a preceding favour and (abstract) goodwill leading to a gift—to equal status with the more common reciprocative meanings, but that is not a serious objection, for he had to stretch beneficium in a similar way, though in the opposite direction. He was in fact prepared to use beneficium of both sides of the exchange, e.g. at Ben. 1. 4. 4 where he renders Chrysippus' exhortation to gratitude with 'to this most honourable rivalry in outdoing benefits by benefits' (ad hanc honestissimam contentionem beneficiis beneficia vincendi sic nos adhortatur Chrysippus), or Ben 4. 18. 1, 'It is only through the interchange of benefits (beneficiorum commercio) that life becomes in some measure equipped and fortified against sudden disasters.'33 In this he was not unique.34 But why did he not instead stretch gratia to cover both sides of the exchange?

Seneca's choice appears even more curious when we consider that gratia should have had great advantages for him over beneficium. For one thing, Seneca shows great interest in the subject of ingratitude, in treating all three of his chosen topics given in 1. 4. 3: the discussion of giving benefits (1. 1. 1–3; 1. 10. 4–5; 2. 1–13) places a lot of emphasis on not incurring ingratitude, the advice on receiving benefits (2. 26–30) analyses the vice, while the discussion of returning benefits (3. 1–17. 4) largely takes the form of a discourse on ingratitude (see pp. 116–18).35

More important is the fact, unlike the case of beneficium, the abstract sense of gratia is the predominant one; or, as Saller puts it, gratia, unlike beneficium (and officium and meritum), primarily signified an attitude, rather than an action.36 In the formulation by Hellegouarc'h, 'la gratia est d'abord une disposition de l'esprit créée par le beneficium et qui conduit à se comporter d'une certain manière'.37 Seneca regards the real beneficium as a virtuous act, whose pg 106essence is its intention: 'A benefit is the actual goodwill of whoever bestows it' (<beneficium est> ipsa tribuentis voluntas, 1. 5. 2). It can only be repaid by a grateful attitude: 'If he (the donor) accomplishes what he sought, if his state of mind is conveyed to me and inspires me with a joy that we both share, he has gained what he wanted' (si quod voluit effecit pervenitque ad me animus eius ac mutuo gaudio adfecit, tulit quod petit) (2. 31. 2). He is at pains to explain that the true beneficium is incorporeal (6. 2), even for the Stoics.38

Thus Seneca's beneficium is a matter of the correct attitude, whereas he speaks of the things given (7. 13) as 'the means employed to give a benefit' (ea per quae beneficium datur), the true beneficium being here equivalent to benevolentia. Sometimes, it is true, he expresses the distinction as one between the actio and 'that which is given' (ipsum quod datur, Ben. 2. 34. 5): the latter can be a material gift or a service such as ransoming someone from the pirates (1. 5. 4). But it is an action that must be performed with the right intention. At 6. 10 he explains his meaning clearly: voluntas is a necessary (though not a sufficient) condition for a beneficium. One is only obliged to a person for giving or performing a service intentionally (ex destinato). Hence his definition: 'What then is a benefit? An act of benevolence bestowing joy and deriving joy from bestowing it, with an inclination and spontaneous readiness to do so' (quid est beneficium? benevola actio tribuens gaudium capiensque tribuendo in id quod facit, prona et sponte sua parata) (1. 6. 1, cf. 2. 34. 5 actio benefica).

It is important to realize that Seneca does not construe beneficium as an 'action' in the sense contrasted with attitude by Saller: Saller's 'action', which is inseparable from what is given, is that present in the etymology of the word beneficium, i.e. the giving of a material thing or a service.39 Intrinsic to Saller's action, unlike Seneca's benevola actio, is the handling of things that are morally indifferent. Seneca's beneficium is construed as a benevola actio, which is a good in Stoic terms, and is contrasted with the thing given, which is an indifferent: it has attitude as its essential element. This is in fact a paradoxical use of beneficium, a term which in common usage clearly operated at the level of action in Saller's sense, i.e. the giving of tangible things, as the dedicatory inscriptions to benefactors show.40 Seneca himself often uses the word of the favours and services bestowed (ad 1. 5. 1–9. 1 fin.), and he respects the etymology of the word in switching at 2. 34. 5 from the definition benevola actio (1. 6) to actio benefica. The common idiom beneficio with the genitive, as in the expression beneficio legis, shows the importance of the result of the action for the receiver. By contrast, the idiom gratia after the genitive, like charin after the pg 107genitive, signifies intention. One might think that it would have been easier for Seneca to insist on the mental state as the essence of a favour if he had used gratia rather than beneficium for both sides of the transaction.

POSSIBLE REASONS FOR THE CHOICE OF TITLE

We must therefore conclude that Seneca had definite reasons for preferring beneficium despite its obvious disadvantages. Several reasons can be adduced:

(i) The influence of the similar title De officiis. Though there is no actual evidence that Seneca had read Cicero's work, which covered the topic of beneficentia, it is likely that he at least knew of it (see p. 7). Moreover, he himself had written a work with that title, which may or may not have handled the subject matter of beneficence.41 He may also have thought that, like this plural title and Chrysippus' plural title in Greek, the plural De beneficiis would convey the nature of his treatise as a work of practical ethics (above, p. 100 and n. 8).

(ii) The fact that gratia, in the less common of its active abstract meanings, i.e. 'goodwill' or 'favour', can be pejorative in meaning, indicating partiality.42 Even in the less common of its concrete active senses, it often means a favour done with a political end in view, as in the passage of Leg. ag. 2. 7 cited above (p. 104). In its passive sense, especially in the Ciceronian period and the early empire, it means political influence, often undue influence in legal contexts where it interferes with justice (Ben. 4. 12. 2), and it can be associated with potentia (Sen. Vit. Beat. 2. 4; Epp. 21. 6; 94. 72).43 In Latin writers generally, it has been calculated that the use of gratia to mean political credit is 14.85% of the total.44

These reasons might not seem sufficient to outweigh the advantage of using gratia. We must look further.

(iii) Beneficium focuses attention mostly on the donor, whereas gratia focuses it on the receiver. Chrysippus could remedy this by using charis in the plural, but gratia was far less common in the plural, and its plural meaning is restricted to manifestations of gratitude, and its use to verbal phrases like gratias agere, habere, and referre.45 Now, the subject traditionally was concerned with both giving and receiving, and Seneca, as we saw, spends a lot of time on the ingratitude of receivers. We think of the paradoxes at the end of book 2, and of his concern that recipients should not be oppressed by favours. Then again, in 6. 43 he says that receiving gifts can be more difficult than giving them. Nonetheless, the primary emphasis in De beneficiis is on the donor and, pg 108as Inwood says, the 'major message of the treatise' is, indeed, that 'man's ingratitude should never incite (and cannot justify) the abandonment of giving'.46 Thus the work begins and ends with an account of the faults of donors, and the giving of advice to donors, as the people principally responsible for ingratitude (see ad 1. 1–3. 1 fin.).

In De Officiis Cicero devotes more space to giving than to receiving or returning benefits. It is natural to connect this with the fact that he is concerned only with his (at least in theory) social equals. Seneca too is primarily interested in this class.47 Seneca, as we said, is more concerned than Cicero with the social relations existing between those who are not equals: not only masters and slaves, but also parents and children, and those in absolute power (including the Emperor) and their subjects.48 The last two topics focus his attention on problems of receiving and returning benefits, because they affect his own class. But the primary emphasis is still overall on the donors, with the result that, as we have seen, there are remarkable continuities between Seneca's picture of the beneficence of the governing class in his day and Cicero's picture of this activity under the Republic (see Ch. 3C).

(iv) Finally, the apparent disadvantage of running counter to common usage and common conceptions could have been regarded by Seneca as an advantage for the philosophical task he has set himself. In the spirit of the Stoic paradoxes, Seneca may have wished to challenge directly the common conceptions of material exchange as the basis of friendship and society.

Beneficium was an important word for the members of the Roman governing class. It was widely used in connection with public life, where it could be synonymous with high office and privileges.49 Gratia too merits substantial space in Hellegouarc'h's study of the vocabulary of Roman politics.50 But beneficium, apart from the advantage that it lacks the pejorative meaning of gratia (above, ii), highlights by its very concreteness the misconceptions that Seneca wants to correct.51 Thus there was a liber beneficiorum recording the Emperor's favours,52 and, at least from the time of Trajan, epigraphic evidence for a post of a commentariis beneficiorum, terminology that fits Trajan's reference, in a letter to Pliny, to having entered in his commentarii the beneficium of the ius trium liberorum granted to Suetonius.53 Moreover, the jurists regularly use beneficium with the genitive of the gerund to denote specific imperial pg 109privileges.54 Indeed, in the military sphere, soldiers who were released from routine duties to serve a senior officer were called beneficarii: they are attested in Caesar's Bellum Civile at the end of the Republic, and on inscriptions, from at least the mid-first century ad.55 The connection of the term beneficium with imperial beneficence must also have appealed to Seneca, as one of his aims was to minimize the differences between the generosity to individuals shown by the Princeps and that shown by other members of the governing class, and to treat them as subject to the same code (Ch. 3C). In this respect, beneficia was the Latin equivalent of charites, used, as we have seen, of imperial privileges.

Seneca knows that he is trying to gain acceptance for the use of terms not in their usual sense, as is shown by his explicit discussion at 2. 34. 2–5 of the meaning of beneficium as primarily a beneficent act (rather than an object given) in the context of the Stoic paradox that feeling gratitude is a sufficient return of a benefit.56 In fact, the revisionist direction of his theme is underlined by the important place accorded to paradox throughout the treatise.57 But how far does Seneca's insistence on mental attitudes as the essence of the social exchange of favours amount to an attempt to 'change the currency'?58 As we have seen (Ch. 3B), his advice is not really as unrealistic as it looks. Not only do we find such sentiments in the popular morality of Publilius Syrus, e.g. 'He who has given to a worthy man has received a benefit in giving' (beneficium dando accepit qui digno dedit, 68, cf. 683): Seneca makes it clear that his most high-minded teaching is actually a form of exhortation and not meant to be taken literally, as he explains towards the end of the work (7. 22. 1–23. 2). Thus, when discussing this very paradox, that 'He who receives a benefit gladly has returned it', he explains how his conception of the exchange of benefits as a transaction between minds (2. 34. 1) accords with the social bond that the Stoics take so seriously. Seneca solves the problem brilliantly, by his use of two powerful images: the first (2. 32) likens the exchange of benefits to a ball game and makes a distinction (a) between the morality of the 'players', for which throwing with the correct intention, and trying to return the ball, is enough, and (b) the success of the game, which requires that the throw actually be returned (see Ch. 3A). The second (2. 33) likens giving a benefit to the work of the artist, for whom the creative process is the reward from his art (ars); the returns he may receive in reputation and material advantage are the rewards of pg 110his work of art (artificium).59 So the first fruit of a benefit is the awareness of having conferred the gift as he intended; glory, and what is received in exchange, are secondary rewards (Ch. 3B).

Returning to the paradox itself, he tells his readers how to read it: 'Although we say that he who receives a benefit gladly has repaid it, we, nevertheless, also bid him return some gift similar to the one he received. Some of the things we say seem abhorrent to our normal way of speaking, but then they come back round to it by an indirect path' (2. 35. 1–2). In the same spirit, he will later make a similar adjustment to the other side of the exchange, explaining that though intention is a necessary condition for a beneficium, it is not a sufficient condition: the actio that constitutes the beneficium actually comprises res et animus, so some favour or service is also required (ad 6. 10. 2; 6. 11. 3).

That is to say, Seneca's hyperbolic exhortation is here, as elsewhere, not to be taken literally. He is really stressing the importance of benevolence in the exercise of beneficence, and helping his readers to fulfil their own highest ideals of generosity. His adoption of the title De beneficiis, using a term whose most common conventional meaning he would go on to challenge, was itself a part of that exhortation.

pg 1116The Structure of De beneficiis

De beneficiis is the longest surviving treatise by Seneca on one subject. It is also the only treatment of that subject to come down to us from antiquity. An understanding of its structure is therefore of great potential importance, as a key to Seneca's literary and philosophical methods, and as a clue to the treatment of this important and once popular subject by earlier philosophers.

SCHOLARLY INTERPRETATIONS

The structure of De beneficiis, except in its broadest outlines, is hardly perspicuous. Like Erasmus before him, Justus Lipsius, despite his deep knowledge of the Senecan corpus and his detailed annotation of this work, pronounced himself baffled: 'The books are good but notably confused in order and treatment, which it is scarcely possible to unravel, even for one making an effort. Still, I am doing the best I can' (Libri boni sunt sed mehercule in ordine et tractatu confusi: quem vix est vel adnitentem expedire. Tamen ut possumus). Many commentators have followed his general condemnation. Thomas, for example, in 1918 regarded it as worse with regard to structure than Seneca's other works and gave up: 'Cet ouvrage est le plus mal composé, le plus décousu, de tous ceux de Sénèque: il serait aussi difficile que peu utile d'en donner une analyse.'1 His view was deemed too extreme by Albertini in 1923, who described the treatise as lacking a methodical composition but did not regard it as unusual among Seneca's works, all of which, he aimed to show, were loosely composed.2 As late as 1975, Sandbach wrote: 'The work lacks structure; even within a single paragraph Seneca leaps from one thought to another; indeed in search of epigrams he sometimes transcends thought.'3 More recently and more pg 112sympathetically, Bellincioni traces its lack of a perceptible design, as well as its size and complexity, to the ambitious scope of its programme: she quotes with approval the dictum of Diderot, 'Le style de Sénèque est coupé mais les idées sont liées'.4

Not all scholars have been so ready to deny the work any methodological arrangement. Sonntag in 1913 discerned in the treatise a regular plan: but, in accordance with the view of philosophical works in Latin prevailing at the time, he ascribed the plan to Seneca's Greek source, which he identified as an unattested work Peri charitos by Hecato, who is frequently quoted in the treatise.5 According to Sonntag, Seneca obscured Hecato's methodical arrangement, not only by reducing the scale of Hecato's work, while amplifying or adding purple passages on such topics as human wickedness and divine generosity, but, more drastically, by separating the main exposition of doctrine from the smaller casuistical arguments, with the result that the division of books no longer coincided with the division of subject-matter.6 Sonntag took it for granted that any Greek philosophical treatise would be tightly organized, and that any Roman philosopher would take all his ideas from the Greek and then adorn them rhetorically, having diminished their philosophical value by a combination of poor comprehension and wilful oversimplification.

Later scholars have been more sympathetic to Seneca's organization, but their very lack of unanimity demonstrates how difficult it is to understand the structure of the treatise in detail. The most significant contribution has come from Karlhans Abel,7 who applies to De beneficiis his experience of analysing the shorter dialogi, which he showed had an internal logic even when the structure, viewed from outside, was not obvious.8 After all, De beneficiis was clearly assigned to the same category of Seneca's works by Quintilian.9 In the case of De beneficiis, however, Abel also succeeds in showing that there is a clear overall structure which is carefully worked out, and that many of the repetitions, which have been seen as signs of artistic failure, are in fact either deliberate pedagogic ploys or thematic recapitulations, designed to modify and deepen pg 113the original argument.10 He also shows that the last three books are not just a ragbag of puzzles but are crucial to raising the level on which Seneca wishes the whole question of interchanges of benefit to be understood and discussed. Though, as will emerge in the following discussion, I do not find everything in his analysis convincing, he has certainly indicated a more fruitful line of approach to the treatise, and one truer to Seneca's own indications.

SENECA'S AWARENESS OF STRUCTURE

De beneficiis contains many signs of the author's self-consciousness about structure. He announces topics for the future (1. 4. 2–3, cf. 1. 1; 1. 11. 1; 4. 1), indicates when he has reached one of the divisions previously announced (2. 1; 2. 18. 1; 3. 1; 4. 3; 4. 16), and sometimes alludes back to previous topics he has treated (5. 1; 7. 14. 1). He is often careful to mark the end of a digression and its relevance to the subject in hand (1. 10. 1; 3. 29; 4. 3; 4. 9. 1; 6. 33), and to signal when he is re-using an argument (4. 22, cf. 4. 11. 4–6; 7. 14. 1, cf. 2. 34; 4. 40; 5. 2; 5. 4. 1; 6. 43. 2). He also compares his order of discussion with an order that might have been more popular with the reader (7. 1. 2) and comments on the appropriateness of his conclusion (7. 26. 1).

As we shall see, Seneca indicates, by an announcement of the main subjects to be covered, what is to be the general shape of the whole work. It must be remembered, however, that in other works, where he gives clues as to structure, they are often not reliable.11 Thus, in De clementia, not all that is promised is treated, probably because the work is incomplete, possibly left unfinished by the author;12 in the part of De ira dealing with therapy, Seneca indicates his programme of two topics at 2. 18 (the nature of anger and the remedies for anger, of which the first has already been treated and the second has two parts: not to become angry, and not to misbehave when angry). Of these two parts, he treats the first in Book 2 but only adumbrates the second at the end of Book 2 (2. 36. 4), and then says at the start of Book 3 that he will take this part up, only to proceed at 3. 5. 2 to announce three parts, the two parts of the topic of remedies for anger already given, and a new one: how to cure another person of anger; and he does not clearly follow that division either.13 In De vita beata pg 114he announces two topics at the outset (1. 1), viz. what happiness is, and how to attain it; but when he reaches the second at 16. 3 he fails to mark it clearly and proceeds in the next chapter to handle the question of attainment so indirectly, via criticisms of the struggling imperfectus, that Fuhrmann thinks he never treated the second topic at all.14 Finally, he appears to revert in 21 to the question of what happiness comprises.15

THE EXPLICIT PROGRAMME

In De beneficiis, however, despite its being a far longer work, Seneca does by and large fulfil his predictions.

The author's own indications divide De beneficiis into two sections. The first section comprises Books 1–4, which, to follow Seneca's own retrospective description at 5. 1. 1, treat the whole subject of giving and receiving benefits. Of the four books in this first section, Books 1–3, as Seneca indicates en route, cover the three topics of giving (1. 11. 1), receiving (2. 18. 1) and returning (3. 1) benefits: the topics announced in Book 1 (1. 4. 3) and already foreshadowed in the allegorical discussion of the Three Graces attributed to Chrysippus (1. 3. 2–3, 8). Book 4 is introduced as a treatment of the most essential question: whether conferring a benefit, and doing a favour in return, are things worthy of choice in themselves (4. 1. 1). The second section, Books 5–7, covers questions which Seneca repeatedly characterizes as an appendix on related but subsidiary topics,16 unnecessary but possibly useful.17 He excuses the pg 115enterprise as undertaken at his addressee's request (5. 1. 2), or at least for his delectation (7. 1. 1), and subject to his consent (6. 1).

That is the bare outline of Seneca's conception as he reveals it, but the structure is not as clear as this suggests. To list only the most striking oddities: Book 3 treats the expected subject of returning benefits rather indirectly; the remaining books (4–7) have not been prepared for in the programme announced in Book 1; and the last three books (5–7) are each announced separately and, as we have just seen, characterized as a collection of rather trivial miscellaneous topics.

BOOKS 1 AND 2

Before addressing these oddities, it will be worth studying how the composition unfolds in the early books. After introductory remarks about ingratitude being the result of incorrect giving and receiving (with the emphasis on giving), and about the pointlessness of the allegorical mode of treatment adopted by Chrysippus, Seneca gives as his aim to provide a lex vitae, teaching 'to give gladly, to receive gladly, to return gladly' (libenter dare, libenter accipere, libenter reddere, 1. 4. 3). But first, he says (1. 5. 1), he must define what a beneficium really is, which he proceeds to do in 1. 5–9. 2. At 1. 11. 1 he announces that the next things to be discussed will be what beneficia should be given and how they should be given (a topic anticipated at 1. 7 and 1. 9. 2). These are clearly subdivisions of the first of the three topics promised at 1. 4. 3: namely, proper conduct in giving benefits. An apparent lacuna in the text prevents us from seeing how the earlier invective against vice, at 1. 9. 3–10. 4, fits into the plan; but 1. 9. 2 appears to relate to the question of who should give benefits, while the question to whom to give is at least mentioned in 1. 10. 5.18 Half of the programme announced in 1. 11 is fulfilled in the rest of Book 1, which treats what should be given, with special emphasis being placed towards the end on the need for discrimination (1. 14–15); it is completed in the first seventeen chapters of Book 2, which discuss how benefits should be given.19

The second topic announced in 1. 4. 3, the receiving of benefits, is reached and clearly marked at 2. 18. 1, the discussion being divided into 'from whom' (2. 18. 3–21) and 'how to receive' (2. 22–35). A discussion of the Stoic paradox pg 116'the person who receives a benefit gladly has returned it' (qui libenter beneficium accipit, reddidit) in the last five chapters (31-end) forms a neat bridge to Book 3 which appears to announce the third topic of 1. 4. 3, the returning of benefits, in the opening phrase: 'it is shameful, and everyone knows it, not to return a favour for benefits conferred' (non referre beneficiis gratiam et est turpe et apud omnes habetur).20

THE PROBLEM OF BOOK 3

That Book 3 is regarded as discharging the third topic is suggested by the start of Book 4, which announces a new topic 'Whether conferring a benefit and doing a favour in return are things to be chosen for their own sake' (an beneficium dare et in vicem gratiam referre per se res expetendae sint). The inference is confirmed by the start of Book 5, where Seneca says that he has now covered the ground of the subject from end to end (5. 1. 1–2).21 However, Book 3 itself, which was supposed to deal with returning benefits, immediately turns into a discussion of ingratitude. Indeed, this negative slant is maintained, even when the topic of returning reappears at the end of the book, in the closing discussion of whether or not children can give parents more than they received (29-end): for the emphasis is placed on the necessity of answering this in the affirmative, if children are not to have an excuse for ingratitude (3. 36. 2, cf. 3. 17. 4). The result is that one scholar does not recognize the presence of the third topic in Book 3;22 and another thinks that the second topic, that of receiving, is continued to chapter 17 of Book 3.23

pg 117In order to solve the problem of Book 3,24 some attention must be given to the precise descriptions of Seneca's programme in the introductions to Books 4 and 5. The latter book actually starts: 'In the preceding books, I thought I had discharged my task, having treated how a benefit should be given and how it should be received; for these define the limits of this responsibility' (In prioribus libris videbar consummasse propositum, cum tractassem quemadmodum dandum esset beneficium, quemadmodum accipiendum; hi enim sunt huius officii fines). Neither the formulation of the subject to be treated in Book 4, nor the statement of accomplishment in Book 5, explicitly mentions the original three topics: rather, they elide receiving and returning, the former one mentioning only returning benefits, the latter only receiving them. In fact Abel, while identifying the subject of Book 3 as returning benefits, remarks that the opening of Book 5 appears to ignore Books 3, 4, 6, and 7,25 all of which concern themselves with beneficium reddere. He rightly marks this as deliberate, and he comes close to solving the problem, when he notes that the original triadic division of the subject-matter is gradually shown to be superficial: to regard reception and reciprocation as separate operations belongs to the vulgar conception of beneficium as a material thing; in the sphere of the moral idea, the lasting relation of friendship is the outcome of giving and receiving and takes over the area of returning.26

In fact, there are important clues throughout the work to a solution. On the one hand, Seneca initially (1. 1. 1) points to our ignorance of how to give and how to receive benefits (beneficia nec dare scimus nec accipere) before announcing the threefold division of topics at 1. 4. 3; and, when he reverts to this original twofold division at 5. 1, he characterizes the two activities of giving and receiving as defining the limits of this responsibility (huius officii fines), i.e. as marking out the whole territory of handling benefits, from end to end. Moreover, at 2. 18 he mentions giving and receiving benefits as the two sides of an ex duobus officium, i.e. one that makes equal demands on two parties: father and son, husband and wife, or, here, giver and receiver; and the twofold division is reinforced by the symmetrical structure of Book 2, with how to give discussed at the beginning, and how to receive at the end; and by the corresponding treatment the two topics receive in Books 1 and 2 (see ad 2. 22–5 and 2. 26–30). On the other hand, at the beginning of Book 4, he frames what he says is the most essential question pg 118in terms of the twin obligations of giving and returning, instead of giving and receiving. The solution is clear: receiving and returning a benefit cannot really be separated, once it is understood that, just as a beneficium is the intentional act that confers the benefit willingly and rationally—not the material thing given or the practical service rendered—so the appropriate return to that act is the good will of grateful acceptance. As Seneca says at 2. 34. 1, the transaction occurs between our minds, and, at the end of Book 4, we are told that the only obligation that applies without qualification to every recipient of every beneficium is 'to offer a grateful heart' (animum praestare gratum).

The topic of how to return a beneficium collapses into that of how to receive one: if we receive it properly, we have returned it. The discussion of the paradox that begins at 2. 31 is not just a bridge to Book 3: it effectively removes the gulf in subject-matter between the two books. Therefore the simplest solution to the problem of Book 3 is to say that the book is not really treating a separate topic from the topic of receiving benefits, introduced in the second half of Book 2 (18–35); rather, Book 3 continues the diagnosis of ingratitude that began in 2. 26 as part of that discussion. Not that the argument fails to progress. For as Abel notes, the emphasis changes, from ingratitude as a failure to appraise rightly what we receive (2. 26–8) to ingratitude as a failure to remember what we have received (3. 1. 2–5), remembering being essential to the return of the benefit.27 The difficulty of returning benefits appropriately is, in fact, highly complex, and most of the problems or quaestiones that occupy the remaining books explore it: the last problem of Book 4, all those of Books 5 and 6, and the last problem of Book 7. There is much justice in the idea that the main theme of the treatise is ingratitude.28

BOOKS 4–7 AND THE SYMMETRY OF THE TREATISE

That the last four books deal with topics not initially announced, does not necessarily point, as Albertini thought, to loose organization. For it is possible to uncover organizational principles not specifically avowed: notably, a large scale symmetry over all.

The symmetry is clear. Sonntag already pointed to the way the end of the work echoes the beginning, and Abel supports the idea.29 Many of the points pg 119made in Book 1. 1–4, where Seneca is concerned with the mistakes of givers of beneficia, are taken up again in 7. 26–32, where he is considering how donors should deal with ingrates: the idea that benefits are already lost at the moment of giving (1. 1. 1, cf. 7. 29. 1; 7. 30. 1); the image of sowing on sterile soil (1. 1. 2, cf. 7. 32); the point that we make recipients ungrateful by harsh reproaches (1. 1. 4, cf. 7. 26. 2; 7. 28. 3; 7. 30. 1); the idea of imitating the gods in not letting ingratitude stop beneficence (1. 1. 9, cf. 7. 31. 2–5); the notion of imitating human optimism in trying again (1. 1. 10, cf. 7. 31. 5); the intrinsic value of giving (1. 1. 12; cf. 7. 32. 1); persistent kindness as a means of winning over the ungrateful (1. 2. 5; 1. 3. 1, cf. 7. 29. 2; 7. 30. 1; 7. 31. 1). But, as Abel points out (1987a, 19 = 1995, 59), there is more here than repetition: the level of discussion has also been raised, for in the meantime the nature of mutual human obligation has been developed. In keeping with the intervening discussion in Book 5. 17. 3–5, Seneca now reminds the giver that he too has probably been ungrateful when in receipt of benefits (7. 28), and whereas, in Book 1, he accused donors of sowing without discrimination on barren soil (1. 1. 2), now he says that the good farmer can overcome that barrenness by care and cultivation (7. 32).

Abel points to another symmetrical feature (1987a, 24–5 = 1995, 65–6). Book 4 stands in the middle of the work, between the first three books in which the basic topics indicated in Book 1 are covered, and the last three books in which particular conundrums are resolved. This book is marked out by Seneca as treating the most important issues: 'Of all the subjects we have treated, Aebutius Liberalis, none is as essential, or, to quote Sallust, more "in need of careful discussion", than the one we have in hand' (ex omnibus quae tractavimus, Aebuti Liberalis, potest videri nihil tam necessarium, aut magis, ut ait Sallustius, cum cura dicendum, quam quod in manibus est, 4. 1. 1). Here the discussion reaches a new level, treating the subject of benefits in the context of the fundamental tenets of Stoic ethics. Thus it is in Book 4 that the imitation of the gods is most pervasive, as a model both for giving benefits and for showing gratitude: 4. 25 on imitatio dei is picked out by Abel as the 'Gipfel des Gipfels' (1987a, 19 = 1995, 59–60).30 In fact, three of the five passages which Abel points to, as showing that gratitude to the gods is the theological basis of the love and friendship manifested in the exchange of benefits (2. 29–30; 4. 4–8, 17–19, 31–2; 6. 20–4), occur in this book, and this is also where the Epicurean view of the gods as unconcerned with us is attacked. It is also in Book 4 that, partly again through contrast with Epicurean doctrine, the connection of giving benefits with the Stoic conception of virtue in general (4. 9. 2–11) and with the Stoic notion of true joy (4. 13–14) is developed.31 Here too the relation of the pg 120Stoic conception of the wicked man and the ordinary one are explored (4. 26–7), and the way in which the Stoic Sage adjusts to circumstances (35). But above all, it is in Book 4 (see ad 4. 18) that we are told that what enables the ultimate function of the exchange of beneficia in binding human society together to be fulfilled, is the work of providence in providing this social instinct, which. along with reason, compensates for human physical weakness (see Ch. 2, pp. 26–7). This idea is a conceptual advance on the praise of the general blessings the gods confer on man in 2. 29, and indeed in 4. 3–9. 1.32 Seneca here indicates that the social phenomenon he describes in De beneficiis is actually the mechanism by which oikeiōsis extends beyond the family.

BOOKS 5–7

In addition to the role they play in the large-scale symmetry of the treatise, two important functions can be discerned for the last three books, which make it absolutely clear that the author's repeated description of them as unimportant and spontaneous (5. 1. 1–2; 6. 1; 7. 1. 1–2) is nothing but disingenuous coquetry, as has often been remarked.33 Even the relatively trivial fact that Books 4–7 close with problems about returning benefits, as seen alternately from the perspective of the beneficiary (in Books 4 and 6) and the benefactor (in Books 5 and 7), suggests that their arrangement is not as casual as Seneca's introductions to the last three books might suggest.34

The last three books, in fact, consist of a series of dialectical exercises, answering hard questions and solving hard cases. This new style of argument had been anticipated in Book 3 and, particularly, in Book 4 which ends with a series of problematic cases that suggest refinements of the notion that one should always return a benefit (4. 40): these two books thus offer a glimpse in microcosm of the structure of the work as a whole.

Some scholars, not without reason, have called the final books of the work casuistical.35 Strictly speaking, of course, the term casuistry is reserved for that part of ethics that resolves particular cases of conscience to which it is difficult to apply the general rules of religion or morality, because the circumstances are pg 121abnormal, or because there appears to be a conflict of duties. The activity is taken to have appeared after ad 1000 in a Christian context; the term 'casuistry' is pejorative and is first attested in the eighteenth century, applied to the activity of the Jesuits, which had reached its peak in about 1550–1650. However, its roots can be found in antiquity.36 For it was recognized by many ancient philosophers that the proof of moral doctrines is in their application, and that requires practice and imagination. As Cicero says at De officiis 1. 59, speaking of the different duties owed to different groups of people, 'precepts on observing duty certainly have been handed down, as I myself am handing them down, but a matter of such importance also demands experience and practice'.

It may not be mere coincidence that Cicero's De officiis also ended with a book of cases. If Sonntag were right, this arrangement would be due to Seneca himself, in contravention of his source (Hecato).37 In that case, we might be tempted to adduce imitation of Cicero by Seneca. However, there is a notable difference in structural conception, for De officiis does not include anything equivalent to Seneca's Book 4. Indeed, in 1. 7 Cicero clearly distinguishes theoretical questions about the nature of officium from the praecepta that he intends to offer.38

It may be the case, however, that many ethical treatises ended with a section designed to give the reader practice in using the doctrines developed earlier, by presenting him with hard cases and moral dilemmas. One only needs to recall that the discussion of philia (what we inaccurately call 'friendship') in the Nicomachean Ethics, which contains a substantial discussion of reciprocation of benefits (notably 8. 13. 1162b23–1163a21), is divided into exposition in Book 8, and consideration of a series of conundrums (including one about benefactions at 9. 7. 1167b17–1168a27) at the end of Book 8 and in Book 9.39 Moreover, Panaetius himself must have intended such a discussion as his third topic of his Peri tou kathēkontos, for it dealt with the question 'if that which has the appearance of honourableness conflicts with that which seems beneficial, how should one decide between them?' Cicero then gives an example of a particular hard case as typical of what Panaetius would have included.40 He also cites in De officiis 3. 89–92 a series of hard cases from his disciple Hecato's book Peri Kathēkontos. Pohlenz plausibly deduced from the fact that they come from the sixth book that this portion of his treatise (fr. 11 Gomoll) comprised zētēmata pg 122('questions') following upon the systematic presentation of the subject.41 Seneca may have been following a well-established tradition. De beneficiis is not the only work in which he used this pattern. The Letters to Lucilius also exhibit a series of complex logical conundrums in the later books, often supposedly requested by Seneca's addressee, while in the earlier books there is more straightforward instruction, Letters 2–33 closing with maxims often drawn from Epicurus, that most dogmatic of teachers.42

A related function of the last three books of De beneficiis is to continue the work of Book 4 in raising the level of philosophical argument, so that detailed moral discussion becomes something more morally profound, and also more complex, than the rudimentary and straightforward instruction of Books 1–3. 17.43 As to profundity, not only is the theme of divine providence maintained (6. 20–4; 7. 31. 2–4), but the sapiens, who is only mentioned in passing before Book 4 (e.g. 2. 18. 4; 2. 35. 2), now features prominently. Thus the Sage, who is shown acting subject to reservation in 4. 34. 3–5, and who then appears in the paradox 'There is no such thing as an ungrateful person' in 5. 12–14, becomes a major player in Book 7, where he receives a full-scale portrait (7. 1, 7; 7. 2. 4; 7. 3. 2–3), is the subject of the paradoxical question 'How can the Sage be given anything, since all things are his?' (7. 4–12), and features in a puzzle about benefactors who become unworthy (7. 17–19. 6).

As to complexity, many of the problems considered in these books probe the meaning of a benefit (e.g. 6. 2–5. 2; 6. 11. 3–4; 6. 19; 6. 26. 1) and of ingratitude (e.g. 4. 26–7; 5. 12. 3–14; 6. 27. 1–4) further than had been done in the earlier books, making important distinctions and continuing the process of refining and correcting vulgar notions that had started with the (re)definitions of beneficium in the first two books (1. 5. 1–6; 2. 19–21. 2; 2. 34. 5–35. 3). Mazzoli urges that the apparent contradictions, when Seneca returns to the earlier themes in these later books, are not explicable as the result of desultory composition, as earlier scholars thought (pp. 111–13), or of opportunistic pedagogy, as argued by Chaumartin (1985, 297–9), but that the casuistry of the later books modifies the definite assertions of the early books, adjusting good actions to particular circumstances.44 Thus in 3. 37. 1–2 Aeneas and two Sicilian youths are given as examples of sons outdoing their father's benefit of giving them life by rescuing him; but in 6. 36. 1 Seneca adduces circumstances in which these rescues would no longer count as examples of pietas: if the sons had wished for the disasters that gave them the opportunity of rescue, that is, pg 123had wished to achieve good ends through evil means. Again, in 3. 1. 1–5 Seneca insists that forgetting a benefit is itself a form of ingratitude; but at 7. 28 he admits that not all forgetting is voluntary, and that we ourselves forget past benefits, so we should pardon others for doing so.45 In 2. 18. 4 Sages are said always to be pleased to honour their obligations as recipients, whereas at 5. 25. 3 even Sages should be reminded, if their benefactors need returns. In Books 1 and 2 we are firmly told that the benefactor must not ask for a gift in exchange (1. 1. 3–4) or even remind the recipient of what he received: all we can do is repeat the benefit (2. 10. 4; 2. 11. 2, 5). But in Book 5 that opposition to reminding is assigned to a fictive interlocutor (5. 20. 6), while the author maintains that, if the benefactor needs repayment, he can ask for a return (5. 20. 7), provided he simply reminds the recipient without reproaching him (5. 21. 2–22). Seneca now distinguishes between the hardened ingrate and the recipient who is not a hardened ingrate (5. 21. 3) but a person who can benefit from such an act of friendship (5. 22. 1–23): in the latter case, preventing ingratitude is itself a second benefit (5. 22. 2). Then, in Book 7, the advice about a promising recipient is repeated (7. 25); but now even the hardened ingrate is said to be reformable through pardon and through patient and persistent goodness (7. 28), and we are told to imitate the gods in continuing to shower benefits on the ungrateful (7. 31. 2–5). Along with the process of refining the argument, there is also a tendency for Seneca's advice to become more and more humane, as we already saw, in the contrasting use of the agricultural image in the first and last chapters of the work (above, p. 119). The process of raising the level of argument, and the level of compassion, was to continue in Letter 81 (pp. 155–61)

These developments run parallel with Seneca's continuous insistence on decoding his Stoic discourse to reveal realistic and practical advice. Thus he unravels the Stoic paradoxes and shows that, by making the right verbal distinctions, what they say can be shown not to contradict common sense (2. 35. 1–2, on which see pp. 109–10; 5. 13. 1–14. 4; 7. 4. 2–6). Finally, in the last book, as we have seen (pp. 47, 110), we are given a strong hint that we must allow for the hyperbolic style of discourse in interpreting the high-minded message that is a dominant theme in the whole work, namely, the asymmetrical demand that the giver should demand no return while the receiver must forever feel his obligation (see ad 1. 4. 5; 2. 10. 4 cum inter prima).46 Seneca spells out what this really means (7. 22. 2): 'We overstate some rules in order that in the end they may reach their true value. When we say, "He must not remember", we really mean, "He must not trumpet it, nor boast, nor give offence." '

pg 124No reader of Seneca could claim that he is a completely systematic writer whose line of argument is always perspicuous and whose division of material is always systematic. Surprises in the form of unexpected points and ingenious examples, rhetorical tours de force, and changes of direction, are part of what makes his writing so absorbing. But De beneficiis is a long and complex work, and it has an intelligible structure. What the pedagogic rationale for that structure is, requires some further investigation. It will be treated in the next chapter.

pg 1257The Pedagogic Strategy of De beneficiis*

Whether or not the structure of De beneficiis fitted into a pre-existing tradition of composition for ethical treatises, there can be no doubt of its relation to Seneca's own conception of philosophical education. For the shape of the work as a whole reflects the way in which Seneca sees the relationship of the praeceptiva pars of philosophy—the translation he offers for the Greek parainetikē—to the dogmatic.

A. TEACHING BY PRAECEPTA

In Letters 94 and 95 Seneca discusses at length the question of the contribution made by praecepta and decreta to virtuous conduct and the development of virtue. In the first letter, he argues against the heterodox Stoic philosopher Aristo of Chios, who held that praecepta, i.e. specific recommendations and advice as to particular courses of action, are unnecessary.1 Seneca defends the pars praeceptiva of philosophy on account of its educational and practical value; but in the second letter, he argues that this part is not sufficient, and that pg 126it is also necessary to have an adequate grasp of decreta, the basic doctrines of the system that alone can provide the rationale for the praecepta.2

Seneca explains in Letter 94 that praecepta have a role in preparing us for the understanding and adoption of decreta, but that they also give us practice, once we have that grasp, in applying that knowledge. For praecepta, as a form of exhortation, can prepare us for the full acceptance of decreta (Ep. 94. 18, 28–31, 34, 36, 49–52); then when our grasp of the decreta has removed the obstacles to natural behaviour, precepts can teach us the details of our duties (Ep. 94. 18–19, 22–3, 32, 50) and give us reminders (94. 21, 25–6) and practice in conducting ourselves virtuously (Ep. 94. 55).3 It is clear that praecepta belong to what Seneca has described earlier, in Letter 89. 14–15, as the third division of the moralis pars philosophiae: the part concerned with actions (de actionibus) and their proper occasions (in ipsa rerum actione tempora), with the 'when, where, and how, each action should be carried out' (quando quidque et ubi et quemadmodum agi debeat). This is the aspect of philosophy that Seneca says elsewhere is particularly open to perpetual development: just as medical remedies constantly need adaptation to particular diseases, and to the particular stage of the disease at which they should be applied, so, in the case of the discoveries of philosophy, which are the remedies for the soul, work must be done on how and when they should be applied (Ep. 64. 8–9).

Seneca's tripartite division of the moral part of philosophy into theoretical (inspectio), hormetic (de impetu), and practical (de actionibus), is no more his own invention than the basic division of philosophy itself into moral, rational, and natural, which he discusses earlier on in Letter 89. 9–13. Stobaeus (Ecl. 2. 42. 7–45. 6 W) preserves a division of moral philosophy into the three parts Seneca uses, which was made by the late-first-century-bc Academic philosopher Eudorus.4 The first is theōrētikon, concerned with ends and what pg 127contributes to their attainment, including knowledge of virtue and vice, good and bad, and the indifferents; Seneca calls this 'the speculative part assigning to each thing its own function and assessing the worth of each' (inspectio suum cuique distribuens et aestimans quanto quidque dignum sit, Ep. 89. 14). Eudorus' second division is hormētikon, the way to deal with the passions and impulses to action; for Seneca this is the section de impetu and deals with 'curbing impulses and proceeding, instead of rushing, towards action' (impetus refrenare et ad agenda ire, non ruere, 89. 15). Eudorus' third division is praktikon and corresponds to Seneca's de actionibus, 'making your impulse and your action harmonize, so that under all these conditions you may be consistent with yourself' (89. 14), and knowing the correct circumstances of each action, as noted above. In Eudorus this section includes hypothetikon, protreptikon, and paramuthētikon, that is, presumably, suasio, exhortatio, and consolatio in Latin,5 forms of teaching that Seneca says in Letter 95. 65 were advocated by Posidonius in addition to giving precepts (praeceptio).6 In Eudorus this section deals with both 'correctly performed functions' (katorthōmata), recte facta in Latin, and more ordinary 'proper functions' (kathēkonta), officia in Latin, one category of which concerns our relations with our neighbours and gives rise to the topic peri tōn charitōn, Seneca's De beneficiis. Eudorus' division might have reached Seneca, as it was to reach Stobaeus, through Arius Didymus, but it is not in any case surprising to find Seneca's conception of moral philosophy in Ep. 89 in tune with Eudorus' scheme, since that may reflect the treatment of practical ethics by Posidonius indicated in Ep. 95. 65.7

As Inwood has emphasized, Seneca's De beneficiis is a 'treatise intensely concerned with the practicalities of ethical reasoning'. He cites it as a clear example of Stoic moral reasoning 'which mediates between the need for situational sensitivity and the demand for stable general principles'—what is now often called pg 128'situation ethics'.8 What we have in the first three books is clearly a treatment of the subject in the manner characteristic, according to Letter 89, of the third section of moral philosophy, de actionibus. Seneca tells us en route that he is teaching 'what benefits should be given and how' (quae beneficia danda sint et quemadmodum, 1. 11. 1); 'how people should conduct themselves in receiving benefits' (quemadmodum se gerere homines in accipiendis beneficiis debeant, 2. 18. 1), and, at the start of Book 5, he says he has now treated 'how a benefit should be given and how it should be received, for these define the limits of this responsibility' (quemadmodum dandum esset beneficium, quemadmodum accipiendum; hi enim sunt huius officii fines). This clearly accords with the 'when, where, and how each action should be carried out' (quando quidque et ubi et quemadmodum agi debeat) of Ep. 89. 15, and with the link between praecepta and officium made in Epp. 94. 32–4 and 95. 45.9

In fact, the language of the teaching in these early books shows us that we are dealing with praecepta.10 It is not only that Seneca uses the words praecepta and praecipere, as when he places the crucial instruction 'never to reproach, not even to remind anyone of a benefit' among the 'first and most essential precepts' (inter prima praecepta ac maxime necessaria, 2. 10. 4), and alludes to it later (7. 22. 1) in the phrase 'some things we teach in exaggerated form' (quaedam praecipimus ultra modum); or when he uses the phrase 'when we are teaching these things' (cum ista praecipimus, 1. 15. 2). It is not only that that he uses the words moneo and admoneo, which in Letters 94 and 95 characterize this style of teaching.11 More significant is the fact that the grammatical forms in which he couches his advice in these books conform to those used to illustrate praecepta in these letters. There, it is clear that, whereas decreta are given in the form of statements, in direct or indirect speech, and answer questions of the form, 'what are' (quid sint) or 'what opinion we ought to hold concerning anything' (qualem de quacumque habere debeamus opinionem, 95. 54), praecepta answer questions of the form, 'how we should make use of things' pg 129(quomodo rebus sit utendum) and characteristically take the form of orders or prescriptive utterances in the imperative, or the gerund/gerundive, or the jussive subjunctive, or the future.12 In fact in Ep. 95. 60 Seneca makes the difference in form explicit, when he argues that the proposition that decreta are unnecessary (for which he uses indirect statement) is itself a decretum, whereas instructions to avoid issuing praecepta (which he expresses with the gerund/gerundive) would themselves be praecepta.13 An examination of De beneficiis Books 1–2 in particular turns up a number of such gerunds/gerundives,14 many jussive subjunctives and futures,15 'one should' (oportet) and 'it is fitting' (decet) with the infinitive;16 there are even imperatives.17 Of course, Seneca is pg 130much too skilful a writer to produce a straightforward list of precepts in their standard form. He uses all the other types of teaching that he lists in Letters 94 and 95 as 'kinds of advice' (genera monitionum), notably 'adjurations' (adhortationes), 'rebukes' (obiurgationes), 'encomia' (laudationes), and he dresses up the precept form in many guises.18

When Seneca says at the beginning of Book 5 that he has covered the territory of this officium, or when he says early in Book 7, 'I do not think there is much point in pursuing other points after the things that regulate our conduct have been said' (nec … nimis ad rem existimo pertinere, ubi dicta sunt, quae regunt mores, prosequi cetera, 7. 1. 2), he means that Books 1–3, and to a lesser extent Book 4, have already provided the praecepta that one needs to perform the appropriate officia involving beneficence. That teaching is firmly at the level of imperfecti, as Seneca explains in Letter 94. 50–1.19 In fact, throughout De beneficiis he frequently makes clear his paramount concern with them (e.g. 2. 18. 4; 5. 14. 5; 7. 2. 1; 7. 20. 5).20 Indeed, in the latter part of the work he spells out the obligation of benefactors themselves, when faced with ingratitude, to advise and instruct those people who are imperfect but curable: 'the second-best form of virtue is to be willing and able to take advice … Few men follow reason as their best guide; next best are those who return to the right path when they are admonished; these must not be deprived of their guide' (5. 22–3, 25. 4–5). And near the end of the last book, he again explains, 'There is sometimes room for a reminder (admonitio), but a gentle one, one that does not demand pg 131for trial or summon to court' (7. 23. 3), though he then goes on at 26–32 to explain what to do with those who are not receptive to subtle reminders.

Chaumartin has rightly pointed to the light shed on Seneca's teaching in De beneficiis by the two Letters that set out the theory of the paraenetic approach. He shows how this approach makes sense of the length of the treatise, for the function of refreshing the memory, mentioned in Letter 94. 21, clearly requires time.21 He goes on to argue that the apparent inconsistencies in Seneca's views are to be explained by the paramount importance, in determining what he says, of the practical effect he is pursuing at any point—what Mazzoli termed opportunistic pedagogy (p. 122). Paul Veyne had already written of De clementia, 'Latin thinkers were not immune to self-contradiction, but Seneca had a philosopher's mind … and in his De clementia, as in the rest of his work, only the necessities and psychagogic longueurs of an exhortation to virtue (one needs time in order to convince) appear to drown the clarity of the concept.' However, Veyne went on to suggest that the striking differences between the concept of clementia in Book 1, where it is used as a synonym for misericordia and venia, and that in Book 2, where it is sharply distinguished from them, is to be explained by the need to woo the young Nero at first with the ordinary confused commonsensical concept of clementia, something he can be expected to understand.22 This might suggest that Book 2 presupposes some progress on the part of the imperial addressee. Then again, Cancik, in her perceptive study of the Letters to Lucilius, noted that one can see an increasing complexity in Seneca's teaching during the correspondence.23 The first doxographic letter is 58 and the longer letters start with Book 8 and Ep. 70 (p. 7). The series of dialectic letters in Book 5, Epp. 45, 48, 49, take dialectic less seriously than the later series, Epp. 82, 83, 85, 87 (pp. 36–42, 138–9), while with Epp. 89–95 we have a series handling the various aspects of philosophy itself (p. 42). She also realized that the situation of the addressee Lucilius, who is represented as a learner, determines the form of the epistolary message (pp. 42; 72). Though she rejected the idea that the literary presentation of Lucilius as making progress constituted a consistent spiritual development (pp. 42 n. 70; 72–5), Maurach rightly insisted that Seneca intends the reader to see the more theoretical and dialectical discussion in the later letters as the response of a skilful teacher to the different level of teaching required by a pupil ready for more demanding instruction; these later letters offer the pupil a fuller insight into the fundamentals of the Stoic system.24

pg 132These ideas, in combination, have implications for the structure of De beneficiis, which need to be explored. Whereas Chaumartin seems to think that the key to the whole work is admonitio, the preceptive style of instruction, it can be argued that, in the course of this long work, as in De clementia and the Letters to Lucilius, an evolution takes place in the type of teaching offered, as was already suggested at pp. 121–3. And whereas repetitions can be used to help the memory to retain precepts, and inconsistencies do find some explanation in the pedagogical needs of the context, it is also relevant that the recipients (the constructed addressee and readership) are envisaged as making some moral progress, the different stages of which require different levels and a greater refinement of argument on the same topics. Maurach has in fact pointed out that the later books serve the purposes of intellectual training and recreation.25 In what follows, I propose to treat first the evolution of teaching style, then the evolution of the addressee.

B. THE EVOLUTION OF TEACHING STYLES

In defending the necessity for decreta as well as praecepta in Letter 95. 45, Seneca writes, 'It is not enough, when a person is arranging his existence as a whole, to give detailed advice. Marcus Brutus, in the book which he wrote Peri kathēkontos, gives many precepts to parents, children, and brothers; but no one will do his duty as he ought, unless he has some point of reference. We must set before our eyes the goal of the Supreme Good, towards which we may strive, and to which all our acts and words may have reference.' Similarly, in his De officiis, Cicero remarked that the officiorum praecepta, which he was about to treat, have relevance to the goal of good things, but that this is less obvious (than in theoretical works) because praecepta are more concerned with the instruction of ordinary life.26

It is in Book 4 of De beneficiis that the praecepta of the early books are clearly related to the decreta of Stoicism, as has already been shown (pp. 119–20). The question Seneca poses at the very start of Book 4, 'whether conferring a benefit and doing a favour in return are things worthy of choice in themselves' (an beneficium dare et in vicem gratiam referre per se res expetendae sint) pg 133requires an affirmative answer that is a decretum. For in Letter 94. 11 we are told that the proposition 'fairness is desirable in itself' (aequitatem per se expetendam) belongs to the 'topic of justice' (de iustitia locus), and that it is the notion of justice, and the decreta concerned with it, that provide the proofs for the particular precepts about how to treat a friend, a citizen, or an associate. In the same way, the proposition beneficium dare et in vicem gratiam referre per se res expetendae also follows from the notion of virtue, which has intrinsic value (4. 16. 2): for a true beneficium is a virtuous action, a katorthōma or recte factum (1. 5. 3.).27 And it is the character of beneficium as a virtuous act defined by the intention and will of the actor that supports the particular precepts Seneca gives about how to give, receive, and return. The Sage begins to feature more prominently in Book 4 because his performance of the perfect right action, or correctly performed function, issues from his complete grasp of these decreta. He does not need praecepta, in order to turn into a reality that imitation of the gods that is here recommended.

In the three books that follow Book 4, the effects of that book are still felt, as we have seen, in the continued concern with the gods and the Sage. But, starting with the second half of Book 4, we are involved in a series of dialectical problems. Cancik in fact noted that dialectic is often the type of theoretical argument that Seneca opposes to paraenetic teaching and unflatteringly compares to verba vs. res (Epp. 83. 27; 87. 40).28 So here the introductions to Books 5–7 deprecate what lies ahead, though, as we have seen in the previous chapter (pp. 114, 120), these remarks are clearly not to be taken literally.

Seneca's expressed attitude to logic is in fact quite complex. These belittling descriptions of the subject-matter that Seneca gives in De beneficiis are in keeping with the descriptions in the Letters to Lucilius of cavillationes (45. 5; 82. 8; 111. 1), interrogationes, argutiae, quaestiunculae (48. 5, 9; 85. 1; 87. 11; 111. 2), which are dismissed as 'sophistical argumentation' (captiosae disputationes, 45. 5) or 'Greek absurdities' (Graecae ineptiae, 82. 8), beloved of dialecticians (45. 13). In the Letters to Lucilius, as in De beneficiis, these puzzles are described as useless,29 though entertaining and seductive: games fit only for exercising the intellect.30 Erasmus complained bitterly about Seneca's inconsistency in pg 134practising what he condemned, naming among others two puzzles that resemble those in De beneficiis.31

Jonathan Barnes has inferred from Seneca's attacks on logic in the Letters that Roman youth were very interested in logic, and that Seneca was attacking 'the preferred practice of the bright young things'.32 Should we therefore read as deeply ironic Seneca's avowal at 7. 1. 2, 'If I had wished to curry favour for myself, this work should have formed a gradual crescendo, with that part held back which any reader, however surfeited, would relish. But all that is most essential I have collected at the beginning; now I am merely recovering what escaped me'? Did Seneca know that, in fact, nothing could please at least his younger readers more than what he presents as mere 'remnants'? Such a consideration might well have weighed with Seneca, who was a popular author with the young and not one to deny his public, as Quintilian later remarked.33

But, in fact, Seneca's attitude to logical questions is not uniformly hostile nor incompatible with his having a serious purpose in treating them. Thus in Letter 89 he was to show that the rationalis pars of philosophy cannot altogether be dispensed with, even by schools like the Epicureans and Cyrenaics, who would like to do without it: in fact they have to reinstate some form of it, in order to be able to sort out ambiguities and to detect falseness hiding under the guise of truth (11–12, cf. 9). Though, as Barnes points out, the target of Seneca's attacks is narrower than the whole logical division of philosophy, which embraced the broad range of subjects that the ancients understood by rhetoric and dialectic,34 Seneca also makes it clear that he thinks that even the dialectical puzzles he attacks have their uses, provided they are enlisted in the service of ethics; for he notes that certain logical questions are intertwined with ethical ones.35 Barnes regards him as a 'logical utilitarian', prepared to allow that logic is an instrument to be used in ethical conduct. As Seneca himself says in Letter pg 13589. 18, 'I do not forbid you to read these things—provided that, whatever you read, you at once bring it to bear on conduct (mores).'36

What we find in the second half of Book 4 and in Books 5–7 are a series of questions (referred to as interrogationes or quaestiones), two paradoxes (5. 12–17),37 and syllogisms that involve paradoxes as conclusions (5. 12; 5. 15. 1) or as premises (7. 4. 7–7. 4). Though the earlier books contain paradoxes, notably Book 2. 31–5, and the odd question, notably in Book 3 (3. 6, 18, 29), it is only in the last books that they form the principal subject-matter, and the introductions to Books 5–7, as we have seen, mark the change.

All of these conundrums feature in Aristotle's discussion in the Topica of problēmata dialektika, most of which, he says, were also called theseis. Their use, he explains, is to help with choice and avoidance (action) or with truth (knowledge).38 Aristotle also divides the problēmata into universal and particular problems.39 It was the rhetorician Hermagoras, in the mid-second century bc, who, appropriating this area for rhetoric, divided its sphere of investigation in the way we find it in Cicero and Quintilian: into the treatments of theseis (general or non-specific questions), rendered in Latin as quaestiones infinitae or as proposita,40 and hypotheseis (specific questions), rendered as quaestiones definitae (or finitae) or as causae.41 Their inclusion in discussions of rhetoric places these authors on the inclusive side of the rhetoric vs. philosophy debate.42 It is true that Cicero, in his youth, disapproved of Hermagoras' claim that the general quaestiones belong to the province of the orator.43 But, probably under the influence of the Academic Philo of Larissa, from whom his schematic analyses of the thesis probably derive,44 he came to attach more importance than did most teachers of rhetoric in Republican Rome, to such philosophical exercises, which he describes as characteristic of Academic and pg 136Peripatetic philosophers45 and as part of Academic rhetorical theory.46 Quintilian subscribed to his views.47 Cicero had in fact given an oratorical treatment to the theseis that he claims presents the greatest challenge of all to the orator, the paradoxes of the Stoics.48 Seneca is at home in this area, common to philosophy and rhetoric. He had been a successful orator and was able to deploy the tricks of that trade in the service of philosophy.

What we meet in De beneficiis are the theseis or quaestiones infinitae, problems that do not involve particular people, places, times, or incidents, either real or fictional. There is, however, as Quintilian demonstrates, a close connection between the universal and the particular, both in that the resolution of a particular question such as 'Should Cato marry?' involves the general question 'Should anyone marry?', and in that the general and the specific are often a matter of degree: 'Should one participate in politics?' is general; 'Should one participate under a tyranny?' is more specific, but not yet nailed to a particular time and place.49 Some, he says, preferred to distinguish theseis from causae, not by their lack of particularity, but by their concern with establishing the truth, rather than with guiding conduct.50 Seneca's discussion at times illustrates precisely the artificiality of this demarcation between the general and the particular. So at 4. 34. 3 Seneca considers, 'Must you keep a promise to give a benefit, if you discover later that the recipient is ungrateful?' The moral dilemma arises from the obligation to keep a promise, which here comes into conflict with the obligation not to give to the ungrateful.51 Seneca goes on (4. 38) to give an exemplum involving Philip of Macedon, in which he puts Philip's dilemma in the form of a quaestio finita: 'Will Philip give to you (a soldier who has proved ungrateful to a third party) because he promised, even if he ought not to, even if it means inflicting an injury and committing a crime?' Another specific question, to be answered in the opposite sense, follows concerning Zeno (4. 39). Seneca here exploits his familiarity with declamation, for many of the controversiae and suasoriae recorded by his father have philosophical theseis implicit in them or were even inspired by them, and the study of theseis formed part of early rhetorical training.52 In fact, the Elder pg 137Seneca's controversiae include two that are treatments, in the form of specific cases, of the question of ingratitude and the nature of benefit.53

Seneca's purpose in De beneficiis, however, is not rhetorical training, but rather exploration of the general question. His arguments here are hard to distinguish from the later practice of casuistry, in the proper sense of the resolution of particular hard cases of conscience, though the immediate point of posing the hard cases here is not the practical one of advising someone or deciding oneself what to do, but the more theoretical one of exploring the general question and explaining, for example, that promises involve unexpressed reservations (4. 39. 4). The effort of applying general principles to particular hard cases helps to refine those principles and our understanding of them.54 As Sandbach (1975, 157) put it, 'There is a good deal of casuistry in an attempt to determine what does or does not constitute a service, and where the line is to be drawn between gratitude and ingratitude.'

The problems that Seneca treats illustrate the two types of quaestiones infinitae distinguished by Cicero and Quintilian in the wake of Philo of Larissa: one type concerned with knowledge, the other concerned with action. The first type has three subdivisions: an sit (questions of existence, cause), quid sit (questions of definition), and quale sit (questions of moral status).55 They are not always easy to distinguish.56 The second type, action, has two subdivisions, questions of how to behave (officium) and questions of how to control the emotions. In this work, Seneca is naturally concerned only with the first subdivision, officium.

The two questions explored in the early part of Book 4, 'whether conferring a benefit and doing a favour in return are things worthy of choice in themselves' (an beneficium dare et in vicem gratiam referre per se res expetendae sint) belong to the type of quaestiones infinitae concerned with cognitio and, within it, to the subdivision quale sit, to which Cicero assigns questions 'about the desirable and undesirable, the fair and the unfair, the honourable and the shameful' (de expetendo fugiendoque, de aequo et iniquo, de honesto et turpi, Cic. Top. 84). Other examples of quale sit are 'whether it is shameful to be outdone in benefits' (an sit turpe beneficiis vinci, Ben. 5. 7. 1), 'whether they are right in doing this (sc. praying that a misfortune may befall their benefactor so they can prove their gratitude) and act from a dutiful desire' (an hoc recte faciant et pia voluntate, 6. 25. 2).

pg 138Probably to be assigned to the subdivision quid sit are a series of questions that serve to define beneficium more precisely:

already in Book 3. 18. 1, 'The question is raised … whether it is possible for a slave to give a benefit to his master' (Quaeritur … an beneficium dare servus domino possit)57

    in Book 5. 7. 2: 'Whether it is possible for someone to confer a benefit on himself' (An possit aliquis sibi beneficium dare)58

    5. 20. 1: 'If two brothers are at variance, and I save the life of one, do I give a benefit to the other, who will probably regret that the brother he hated did not die?' (Fratres duo dissident; si alterum servo an dem beneficium ei qui fratrem invisum non perisse moleste laturus est?)59

    in Book 6. 2: 'Whether it is possible for a benefit to be taken away' (an beneficium eripi posset)60

Questions in Book 7 concern the different senses in which things can belong to, or be possessed by, people:

7. 4. 1: 'How can anyone give anything to a Sage if all things are his?' (Quemadmodum potest aliquis donare sapienti, si omnia sapientis sunt?)

and (implicitly)

How can anyone give anything to his friend, if friends have all things in common? (Quemadmodum potest aliquis donare amico, si omnia amicis communia sunt?)

Or they concern the meaning of 'returning':

7. 14. 1: 'Whether someone who has done everything to return a benefit has in fact returned it' (An qui omnia fecit ut beneficium redderet, reddiderit).

The question about promises in Book 4. 34. 3, discussed above, clearly belongs to the category of quaestiones infinitae concerned with actio and, in particular, with officium, under the subheading quomodo utamur, as Quintilian puts it (3. 5. 6).61 Others occur in Book 4:

4. 26. 1: 'Will the good man give a benefit to someone he knows to be an ungrateful person?' (an vir bonus daturus sit beneficium ingrato sciens ingratum esse?)

pg 139    4. 40. 1: 'Should a favour be returned in every circumstance and should a benefit in all cases be repaid?' (an omni modo referenda sit gratia, et an beneficium utique reddendum sit?)

    in Book 5. 20. 6: 'Should one ever ask for the return of a benefit?' (an repetendum est beneficium?)

    in Book 6. 5. 4: (of a benefactor who later injures us) 'Ought I to return to him the benefit and nonetheless avenge myself on him … or ought I to combine the two into one and take no action at all, leaving the benefit to be wiped out by the injury, and the injury by the benefit?' (utrum et beneficium illi reddere debeam et me ab illo nihilo minus vindicare … an alterum alteri contribuere, et nihil negotii habere, ut beneficium iniuria tollatur, beneficio iniuria?)

    in Book 7. 16. 5: 'Ought a person to return the benefit that he has received from a Sage if he has ceased to be wise and has turned into a bad man?' (quod beneficium quis a sapiente accepit, reddere debeat si ille desinit esse sapiens et in malum versus est?)

    7. 26. 1: 'How are we to deal with the ungrateful?' (quemadmodum ingrati ferendi sint?)

It was already remarked by Theon, a rhetorician contemporary with Seneca, that the officium subdivision of quaestiones infinitae relating to actio, is often hard to distinguish from questions dealing with cognitio.62 It is not difficult to find examples of such ambiguity in Seneca's questions. So in Book 5. 18. 1, questions of conduct arise in determining to whom gratitude is owed, 'How far am I to pursue the list of relevant persons?' (quousque personarum seriem sequar?), and 'Should a benefit be reclaimed from the (recipient's) father?' (repeti a patre beneficium debet?, 5. 19. 8). But these questions of conduct are really designed to raise the question 'Where does a benefit stop?' and to make us distinguish between a benefit conferred knowingly and intentionally on someone, and the profit that person may derive from a benefit given to another. Again, in 6. 7. 1, 'Whether a person who has beneftted us without that intention, imposes any obligation on us' (an ei debeatur aliquid, qui nobis invitus profuit) turns into a discussion of the intention intrinsic to the notion of a benefit, for, as Seneca says, 'Both this question and any similar one that can be raised will be easily settled, if on every occasion we focus our attention on this idea: nothing is a benefit unless, first, some intention directs it at us; second, that intention is friendly and well-disposed' (6. 7. 2).63

pg 140The relation between these quaestiones infinitae about action and those about philosophical truth is somewhat analogous to the intimate connection which Seneca made in Letter 95, between praecepta and decreta. That is not surprising, given the close relationship between the action questions and praecepta, on the one hand, and between the truth questions and decreta, on the other. The problems raised by the action questions concerned with officium, are clearly a way of exploring the realm of the praecepta given earlier in the work, and of refining them: indeed, Cicero calls this category of question the praecipiendi genus (Part. or. 67).64 For though Seneca can speak, in Letter 94, of praecepta as giving, not universal advice but instructions appropriate to each persona, the personae meant are social roles, such as those of husband, father, master (§1), and he soon makes it clear that they are not geared to individuals and specific occasions, i.e. to 'token actions' (§§14–15).65 For they would have to be infinite in number. Precepts cover categories of action: they are praecepta generalia, in the formulation of §35. Thus praecepta give the answer to the nonspecific questions (quaestiones infinitae) of action, but without the examination that they undergo, when posed as questions.

As for the quaestiones infinitae relating to cognitio, they explore the areas covered by decreta. This type of logical puzzle in the later part of the work brings out the issues that need to be explored, if we are to understand the rationale behind the precepts offered in the early part of the work. However, Seneca is eager to keep even this level of instruction on quite a practical level, the early part of Book 4 being the closest he gets to metaphysics and the highest level of theory, while many of the questions considered, as we have seen, are closely related to questions of conduct. Indeed, in the last book he endorses the argument of Demetrius the Cynic, that the basic principles to be learnt are few and simple. Seneca reduces them to a regula (7. 2. 2), a rule for deciding easily whether what he has done is right, as he defines it in Letter 95. 39. This rule, 'There is no evil except what is shameful, and no good except what is honourable' is similar to the rule of thumb that Cicero had offered in De officiis 3. 81, viz. 'Either the thing that seems beneficial must not be dishonourable, or if it is pg 141dishonourable, it must not seem beneficial.'66 It is significant that Seneca puts the recital of basic essential principles there in the mouth of the Cynic Demetrius, who confines them to ethics and is made to refer to them as praecepta sapientiae (7. 1. 3). This looks like a kind of oxymoron in Stoic terms, for, in the language of Letters 94 and 95, what Demetrius gives is the substance of decreta, not praecepta: general statements of philosophical truth, not counsels to action (Billerbeck 1979, 34–5). However, even Seneca, speaking in his own voice in Letter 94. 31, where he is defending the effectiveness of instruction (utraque res praecipit) by precept, in comparison to that by doctrine, says that decreta are nothing but generalia praecepta, whereas praecepta are specialia, and goes on to blur the distinction by showing (§35) that even the praecepta are general, in not applying to specific situations.67 Moreover, Demetrius' mode of teaching here is to sketch the mind of the sapiens, and Seneca says in Ep. 95. 66 that sketching the exemplar of a virtue, here the sapiens possessing sapientia, is equivalent to giving precepts for that virtue (see ad 7. 1. 3)

One is tempted to argue that, of the two basic functions that Seneca attributes to praecepta in this pair of Letters (p. 126), the first—that of awakening the natural inclination to virtue and preparing one to receive decreta—is covered by Books 1–3 of De beneficiis, which prepare one for Book 4. The logical puzzles, starting in the second half of that book, give us both the means to refine the precepts and the practice in applying them correctly in complex situations, thus enabling them to fulfil their second function.68 Even a person who has the right decreta, he says in Letter 94. 32,

pg 142has indeed learnt to do things which he ought to do; but he does not see with sufficient clearness what these things are. For we are hindered from accomplishing praiseworthy deeds, not only by our emotions, but also by want of practice in discovering the demands of a particular situation. Our minds are often under good control, and yet at the same time they are inactive and untrained in finding the path of advice.

Precepts help, but they are not the whole solution; for, as Seneca had pointed out earlier in the correspondence (Ep. 22. 2), it is difficult to apply general rules of conduct, i.e. precepts, in particular circumstances. Cicero similarly had stressed that the precepts of any art need to be supplemented by experience and practice in fitting them to particular cases, so that we become 'good calculators of our duties', and he does this precisely in the context of beneficence, though he extends the relevance of what he says to all duties (Off. 1. 59–60). Accordingly, Seneca can often be seen, in these last books of De beneficiis, exploring the precise character of the obligations enshrined in familiar precepts (e.g. 5. 18; 6. 12). The last section of the treatise is comparable to a graduate level course in officia for the advanced profectus, and Demetrius' advice in the last book is aimed at the proficiens who is on the verge of becoming a Sage.69

C. THE PROGRESS OF AEBUTIUS LIBERALIS

Having considered the evolution of teaching style, we can now raise the question of its relevance to the level of progress made by the imagined recipient (above, pp. 131–2).

Long ago, Sonntag suggested that Aebutius Liberalis, attested in Seneca's Letter 91 as a real person, was selected as the addressee merely because of the appropriateness of his cognomen. He adduced in support the fact that Liberalis is not addressed at length (nor, we may add, characterized) until the additamenta in 5. 1 and 6. 41–2.70 No doubt the cognomen did play an important part in Seneca's choice, as did his choice of Serenus as the dedicatee of De Tranquillitate Animi,71 but that does not mean that the persona of the addressee is not developed. In fact, this emergence from the shadows in the pg 143later books, is precisely what suggests that Seneca wishes the reader to notice an increase in the role of the addressee, as he progresses in the course of the work towards parity with the author.

As is customary in Seneca, Aebutius Liberalis is addressed in the preface to each book.72 He is just a name in Books 1 and 2, although he is also invoked in the body of Book 2 (2. 6. 1; 2. 30. 1). In 3. 1. 2 reference is made to a previous debate he has had with Seneca over forgetfulness and ingratitude, but the distance between Seneca and his addressee is made clear: 'You said such persons were ungrateful, but forgetful; just as if that which makes a man ungrateful could be any excuse for his being so, or as if the fact that a man had this misfortune meant that he is not ungrateful, when this misfortune only befalls the ingrate.' Seneca simply puts him right, without acknowledging any force in his argument.

The opening of Book 4—'Of all the question that we have discussed, Aebutius Liberalis'—is perhaps meant to give a hint of collaborative effort in the work so far, an idea reinforced in 4. 3. 1, where Seneca justifies a digression to Liberalis, and in 4. 31. 1, where Seneca appears to address his defence of the gods specifically to him. But it is really in Books 5–7 that Liberalis is depicted as becoming more active. With 5. 1. 2, 'Since, however, such is your wish' (quia ita vis), the fiction is that Liberalis wants the author to continue writing or, specifically, to elaborate particular puzzles. Then, at 5. 1. 3–5, the focus is clearly on Liberalis himself, whose nomen and gentilicium are reversed for emphasis (see ad 5. 1. 3 Liberalis Aebuti). We are shown Liberalis in action as a kind of exemplum of liberality, 'the best of men by nature and one prone to benefits' (homo natura optimus et ad beneficia propensus). He has 'goodness' (bonitas), he is an 'excellent man' (optimus vir), and 'a great soul' (ingens animus). The conduct Seneca ascribes to him fits very well with Seneca's teaching so far (especially, to give the earlier passages in the order of the points in the Book 5 portrait, 2. 11. 3; 2. 17. 6; 1. 7. 1; 2. 17. 7; 2. 11. 4–5; 1. 1. 10; 2. 9; 1. 2. 4–5). Yet at 5. 2. 1–2 one of the maxims Liberalis most admires, 'It is shameful to be outdone in benefits', is challenged.

At 5. 12. 1–2 Liberalis is first represented as finding the preceding discussion, about not being able to do oneself a benefit, useless and a waste of effort for Seneca; then Seneca imagines him becoming even more annoyed with the puzzles to come, and makes him say the kind of thing that Seneca himself says in some of the Letters to Lucilius, 'What is the good of laboriously untying knots which you yourself have made, in order that you might untie pg 144them?' (cf. Ep. 45. 4; 82. 19). But here Seneca defends the practice of dialectic in a way that makes clear the distance between an unskilled person, an imperitus (like Liberalis), who would find untying such knots difficult, and the person who has tied the knots and can untie them easily (clearly Seneca himself), for whom they provide pleasure and a mental challenge that drives away complacence and sloth.

However, at the start of Book 6 Liberalis is assigned the task of guiding Seneca, indicating by his facial expression how long he should dwell on certain topics and whether or not to dismiss some at once. The pleasure and profit to be derived from these puzzles now seem to be accessible to them both. At 6. 5. 3–7. 1, Liberalis is shown actually performing this task, and indeed going beyond it. He first steers Seneca away from a theoretical question ('whether a benefit has been removed, if we are not under obligation to repay') to a more practical one about our obligations to someone who first confers a benefit, then inflicts an injury (6. 5. 4), which he represents as a move from the kind of useless questions of definitions beloved of iuris consulti to real cases that are treated in the courts, whose practice Stoic philosophers should consider and either adopt or reject. Seneca accepts the challenge on his terms (6. 6). Then at 6. 7. 1 Liberalis ends the discussion altogether: 'Your face, by which I have agreed to be governed, is wrinkled and frowning, as though I were straying too far from the point.' At 6. 12. 1 Liberalis' rather plastic face suggests a question following on from whether we are obliged to an unwilling benefactor: that is, are we obliged to a self-interested one? Seneca addresses himself to this complaint, which he claims often to have heard Liberalis make about such people. At 6. 41. 1 Liberalis is told not to be anxious about returning benefits, and at 6. 42 Seneca says Liberalis is overanxious about requiting benefits and should not try to do it too soon. At the start of Book 7 Seneca again attributes to Liberalis his decision to continue writing. In 7. 17. 1, perhaps our last glimpse of the addressee, Seneca associates himself with Liberalis as imperiti in contrast to the Sage, and, in contrast with the perfect benefit, he speaks of 'the everyday common benefit, which we ignorant men exchange' (see ad 7. 17. 1).73

The degree to which Seneca characterizes his addressees (and himself), in building up an impression of dialogue in his prose works, varies considerably, as Mazzoli points out in a perceptive discussion.74 In the consolations, characterization is at its most dense, because the advice is geared to the specific situation of the bereaved addressee. So Seneca continuously evokes Marcia's particular circumstances while consoling her on the death of her son; he supplies considerable biographical detail about his mother Helvia as he pg 145consoles her for his own exile (Cons. Helv. 2. 4–5; 14–16. 5; 17. 3–4; 18–19); and Polybius' relationship to his brother and the Emperor, along with his literary tastes, is sketched and exploited (Cons. Polyb. expecially, 2. 2–3; 6. 2–5; 8. 3–4; 18. 1–2). In De brevitate vitae Seneca narrows down his rather general advice in chapters 18–19 with a description of Paulinus' duties as prefect of the corn supply. Mazzoli thinks that the other treatises are to be sharply distinguished, in respect of their dialogic character, from the twelve works actually called dialogi in the index to the Codex Ambrosianus, and, in particular, that they are at the opposite extreme from the consolations, where the relationship of author and addressee is almost comparable to the Letters. But he seems to forget that Lucilius is quite fully characterized in the Naturales quaestiones, particularly in the preface to Book 4 (4. pref. 1, 3; 4. 14–18, cf. 3. 1. 1).75 In De beneficiis Seneca has not provided his addressee with any concrete biographical details: we only learn that Aebutius Liberalis came from Lugdunum in Letter 91. But one thing Seneca has managed to do is to build up a consistent picture of Aebutius Liberalis as a benefactor. The advice in Book 6. 41–2 fits with the portrait in Book 5 of Liberalis as someone who is very generous, and who is perhaps too willing to overestimate what he has been given: Seneca has never seen anyone 'so generous in valuing even the most trivial services' (tam benignum etiam levissimorum officiorum aestimatorem, 5. 1. 3). In Book 5 Liberalis is anxious about being outdone as a benefactor; in Book 6 he worries about not repaying quickly enough.

Of greater importance to understanding the structure and strategy of De beneficiis, however, is the fact that Liberalis' role as addressee develops in parallel with the change in the style of teaching. He is a recipient of precepts in the first three books, and clearly in need of straightforward admonition at 3. 1. 2. But after the precepts have been succeeded by decreta in Book 4, and a new level of instruction has been reached, he becomes more active. He is now a more advanced profectus, but still not at the level of his mentor Seneca. In Book 6, he is guiding the enquiry, though still receiving advice, and by Book 7. 17, no distinction is drawn between mentor and addressee as imperfecti. The role of Liberalis as addressee in fact develops in parallel with the change in the style of teaching. The fact that mentor and addressee, like the readership of De beneficiis, are not Sages is underlined by the realistic advice that follows about tactful requests for repayment being acceptable, and even more by the characterization of Seneca's earlier and sterner teaching as hyperbole (7. 22–3), and by the indication that we are all ungrateful (7. 28).

The idea that Seneca's educational strategy involves matching the moral progress of his addressee to the evolution during De beneficiis of his teaching pg 146technique, can be supported by Seneca's practice in other works. Veyne's thesis (see above, pp. 131–2) about De clementia could be interpreted in this sense: the contrast in treatment between the two books could arise from the fact that the advice is geared to the addressee Nero, young and morally green in Book 1, but judged ready for more rigorous training in Book 2, where the necessary distinctions are made between clemency and the related terms. But De clementia is incomplete, perhaps unfinished, and its strategy is not unambiguous.76 Then again, the three dialogues addressed to Annaeus Serenus, De constantia sapientis, De tranquillitate animi, and De otio (if we trust the entry in rasura in the index to the Codex Ambrosianus), certainly show the addressee undergoing a spiritual development, from an avowed Epicurean scoffing at the Stoics (Const. sap. 3. 2; 15. 4), to an imperfectus striving to follow Stoic doctrine (Tranq. An. 1. 10–12; 1. 17), to, finally, an orthodox Stoic championing those doctrines against the apparent heresy of the author (Ot. 1. 4; 7. 2).77 But the fact that this happens across separate works makes it harder to discern a correlation between Serenus' development and changes in Seneca's teaching technique, though De tranquillitate animi is clearly more therapeutic and the others more polemical, with De otio advancing more complicated and subtler arguments. Easier to establish is the theory that the Letters to Lucilius attribute to the addressee a fictional spiritual development that parallels the change in the style and content of the Letters. I have sketched Lucilius' spiritual development elsewhere and need not repeat the details.78 Here it is the parallel between the structure and content of the Letters and of De beneficiis that principally concerns us.

Foucault described Seneca's correspondence with Lucilius as a kind of adult education course, beneficial to the teacher as well. 'His correspondence with Lucilius deepens a preexisting relationship between the two men, who are not separated by a very great difference in age, and tends little by little to transform this spiritual guidance into a shared experience, from which each derives a benefit for himself.'79 The method of teaching moves with the progress of Lucilius. In the early letters of Books 1–3 (1–29), Lucilius is given, at the ends of his mentor's letters, little moral tags to remember, in line with the advice in Letter 2 to pick out one precept every day from his reading and digest it. This practice is illustrated with a citation from Epicurus (2. 4–6), appropriate because Lucilius is depicted as having Epicurean sympathies when the correspondence opens (Ep. 23. 9, cf. 20. 9). At Letter 6. 5, Seneca makes it clear that pg 147Lucilius is being taught through praecepta and exempla: he should come to visit Seneca in order to have first-hand experience of an exemplum, 'because the road is long via praecepta, short and effective through exempla'.80 Then in Letter 13, Lucilius is said to be more in need of admonitio than of exhortatio (15) and to be training himself (1) in 'precepts that are health-giving and effective at overwhelming hardships' (praeceptis salutaribus et dura vincentibus), i.e. the more uncompromising Stoic precepts. Clearly, all the techniques associated with the preceptive part of philosophy are being used. After Letter 29, at the end of Book 3 of the correspondence, there is a change in the form of the letters, to which Seneca draws attention in Letter 33. He explains why he now refuses to give moral tags at the end of his letters, especially the Stoic tags that Lucilius is supposed to have requested. Lucilius, as a 'man of secure progress' (certi profectus vir), should read books as a whole and not rely on chriae, rendered in Latin as sententiae (7); he should become an active thinker and a teacher in his own right (8–9).81 Letters 34 and 35 bring home the point about Lucilius' improvement: he is now making rapid progress and helping Seneca to improve in turn, as partners in an ideal friendship should. Letter 72 shows Lucilius asking a question, which Seneca cannot answer on the spot. Then, in Letter 75, Seneca describes the different levels of imperfecti and speaks of them both as struggling to improve.

Letter 85 inaugurates the series of late letters in which difficult dialectical questions are often discussed at considerable length. It is Lucilius who is said to have requested the syllogistic proofs for the Stoic tenet that virtue alone is sufficient for happiness, and Seneca is shown in Letter 87 sending him some more on that theme (11). Lucilius is shown requesting a difficult discussion of the parts of philosophy in Letter 89. 1, while Seneca warns him not to be too interested in logic for its own sake (18). At Letter 102. 3 Lucilius wants Seneca to finish a syllogistic discussion from which he had omitted the parts less relevant to conduct. Further difficult questions follow, in Letters 106. 1, 3–4; 108. 1–2; 109. 1, 17; 111; 113. 1; 117. 1; 120. 1, 3. By then, in 121. 1, it is Lucilius who is critical of questions that have no bearing on conduct, while Seneca defends the contribution of doctrines on human nature. Finally, in Letter 124, the last we have, Seneca says that Lucilius has the balance just right in practising and approving of highly subtle argumentation, so long as it contributes to moral progress (1). In Letters 106, 108, and 109, Seneca has mentioned that he is working on a comprehensive work on moral philosophy that will cover these difficult questions. Leeman suggests that it is really this outside project that explains the shift to more technical subject-matter in the later Letters; for, as he pg 148rightly points out, Lucilius's requests are only a fiction within a fictional correspondence.82 Maurach, however, points out that the influence may work the other way and that, in any case, a deeper knowledge of Stoicism entails seeing the system more and more as a whole, so that Lucilius' progress would need to be matched by more systematic teaching.83 In any case, Lucilius is clearly represented as ready for the systematic work on which Seneca is engaged.84

The fiction about Lucilius is central to the pedagogic purpose of the Letters. As Seneca says in Letter 34, just after his correspondent has been weaned off his mainly Epicurean tags: 'I claim you for myself: you are my handiwork. When I saw your talent, I laid my hand on you, I exhorted you, I used the goad and did not let you proceed slowly, but kept on rousing you; and now I am doing the same; but I am urging on someone who is already racing and encouraging me in turn.' For the fiction is not limited to Lucilius' requests and responses: Seneca has created a spiritual development for his addressee, to which these requests add verisimilitude. The parallel with De beneficiis should by now be clear: Lucilius, like Liberalis, becomes capable of a higher level of philosophical understanding, even as the philosophy offered becomes more difficult and technical. The parallel extends to the offering of reassuring remarks to the addressee about what standard is really being demanded. Thus, to Lucilius, Seneca praises the philosopher Sextius, who demonstrated the greatness of the happy life without instilling pessimism about its attainment (64. 5). For Lucilius, like Liberalis, remains an imperfectus. Nonetheless, the reader of the Letters or of De beneficiis, who identifies with the addressee, and indeed with Seneca, who is also presented as a learner, can feel that the extended course he has undergone has set him firmly on the path to virtue and happiness.

pg 1498De beneficiis and Seneca's Other Philosophical Works

In Letter 73, as we have had occasion to notice before (pp. 2, 96), Seneca says that the pre-eminent task of philosophy is to teach us 'to owe and to repay benefits well' (9). Clearly, Seneca does not mean to exclude from philosophy's brief the important activity of giving benefits, but, as the letter concerns benefits received from rulers, he formulates philosophy's task in terms of the receiver. This sentence alone, then, prepares us for the flexible way in which Seneca treats the subject of benefits and gratitude elsewhere. In fact, this same careful tailoring of his teaching to the subject in hand, a general feature of Seneca's writing, can be further illustrated by considering the treatment in this same letter of certain topics, which are also found in De beneficiis.

CONTEXT AND ARGUMENT

Letter 73 aims to show that philosophers are unlikely to be rebellious troublemakers, since they above all enjoy, and know how to use, the benefits of peace and leisure that governments provide for them. In the course of this discussion, Seneca touches on a number of issues found in the treatise, such as the role of envy in generating ingratitude (Ep. 73. 3, cf. Ben. 2. 28); the gratitude owed for communal benefits, including those conferred by the sun and moon (Ep. 73. 5–8, cf. Ben. 6. 19. 2–20); the way that different senses of possession make intelligible the Stoic idea that all things belong to the Sage (Ep. 73. 7, 14, cf. Ben. 7. 4–8. 2); and the fact that grateful receipt of a benefit is equivalent to its repayment (Ep. 73. 9 fin., cf. Ben. 2. 31–5).

The dramatic date of Letter 73 letter is early summer of 64, probably after the publication of De beneficiis, which certainly predates Letter 81 (p. 92, n. 4): in that case Seneca is revisiting topics already dealt with more thoroughly there. This may explain why he feels that he can say cryptically, after noting that philosophy teaches us bene debere beneficia, bene solvere, 'sometimes the pg 150repayment consists in the acknowledgement itself' (interdum autem solutio est ipsa confessio). He alludes briefly here to the Stoic paradox 'He who receives a benefit gladly has returned it', an idea that was fully discussed at the end of Book 2 of the treatise, where it served as an important bridge to Seneca's unorthodox treatment, in Book 3, of his next subject: how to return benefits (see Ch. 6). The other topics revisited in the letter, however, have been completely revamped to suit the author's current subject of concern. Thus, whereas in De beneficiis envy is part of the explanation why benefits are generally not properly received, here the point is the ingratitude towards rulers of those engaged in public life and competing for favours, as compared to the philosopher who has withdrawn from the public gaze to study. Then again, whereas in De beneficiis the treatment of benefits given to one as part of a group is an exploration of the kind of intention that enables a benefit to create obligation, in Letter 73 the point is the superiority over material things of the indivisible and communal goods made available by a divine or human ruler to everyone, such as peace and liberty. A consequence is that, in the Letter, this topic is closely connected to the topic of the Stoic Sage having everything: he alone knows how to use these communal things. In the treatise, however, the Stoic paradox 'all things belong to the Sage' (omnia sapientis sunt) is introduced in connection with a separate issue: namely, an apparent contradiction within Stoicism between the idea that the Sage is an expert in giving and receiving (as in everything else), and the idea that the Sage has everything (and therefore cannot be given anything), see ad 7. 4–13.1

We have had occasion to mention this habit of Seneca, whereby similar ideas are used in different contexts to very different effect, in discussing chronology; and we noted there that thematic parallels are useless for the comparative dating of Seneca's works (pp. 95–6). However, it will still be worth exploring the themes from De beneficiis that are anticipated or repeated most often in Seneca's other philosophical works, in order to see more clearly where the topic of benefits fits in his philosophical scheme of things.

DIVINE BENEFITS

The context in which beneficence appears most frequently in the other philosophical works, is that of divine providence and generosity to humankind. For example, a quick survey of the Letters shows that the beneficia referred to most often are the gifts of fortune (e.g. Epp. 8. 3; 18. 6; 63. 7; 72. 4–8), nature (90. 36; 119. 16), or the gods (15. 10; 74. 11), and in De beneficiis, where this theme is pg 151also prominent (see pp. 26–7, 119–20, 122), Seneca explained that nature, fate, fortune, god are synonymous.2

Already in De ira (1. 5. 3; 3. 5. 6), written some years before De beneficiis handled the theme (Ben. 4. 18, cf. 1. 4. 2), Seneca noted that nature has designed man for mutual help, to confer benefits and live in harmony, and the idea is repeated in De clementia (1. 3. 2) and elsewhere (e.g. Vit. beat 24. 3; Ot. 3. 5). The advice in De beneficiis (4. 25–6; 7. 31. 2–5) that, as rational beings, we should imitate, in this as in other things, the gods who are generous to us, is already present as advice to rulers in De clementia (1. 5. 7; 1. 19. 9). Just as in De beneficiis Seneca defends the gods for bestowing favours on unworthy men (4. 31–2; 7. 31. 2–5), so in De providentia he insists on defending the gods for the apparent hardships they inflict on good men (1. 1). In De beneficiis, Seneca finds the root of ingratitude among men in their ingratitude toward the gods (2. 30. 1). This ingratitude, lamented already in De ira 3. 31. 1, is traced in De providentia and Ep. 74. 10–14 to a misconception of what things are really good and bad, as opposed to being positive or negative indifferents, and it can also involve, according to the Natural Questions, blaming the gods for our own misuse of their blessings (5. 18. 4–15).

Pohlenz therefore was right to stress Seneca's preoccupation with the problems of providence.3 It is clear that the metaphysical basis of social exchange was always at the root of Seneca's belief that the theme of exchanging benefits was of central importance. It is only delayed in the treatise to Book 4, after the basic precepts have been rehearsed, as part of the author's teaching strategy (see Ch. 7B).

DIVERGENT VIEWS

Many more parallels between De beneficiis and Seneca's other prose works could be compiled, but it may be more illuminating to consider cases in which Seneca's views elsewhere diverge significantly from the treatise. One such case concerns the notion that conferring benefits is an investment or insurance policy. In De clementia, Seneca says that the good ruler is loved by his subjects, who are willing to defend him at the risk of their own lives (1.3. 4; 4. 1), and who make it unnecessary for him to have a bodyguard (1.13. 5); and that his generosity ensures that he is venerated like a god (1. 19. 8–9). In pg 152De tranquillitate animi generosity, along with justice and kindness, provides safeguards against later bad fortune (10. 6), and in De vita beata benefits are compared to buried treasure that one can dig up when in need (24. 2). In De beneficiis, however, Seneca is far more guarded. Though, in praising his addressee, he allows that glory can follow beneficence (5. 1. 4), and he acknowledges that one can ask for repayment of benefits if the giver is in need (5. 20. 7), he is adamant that benefits are not investments: 'Let us give benefits rather than lend them' (demus beneficia, non feneremus),4 thus imitating the gods (1. 1. 9; 1. 2. 3). Any advantage we derive from giving a benefit is an accessio (a bonus) of gratitude, not the true fruit of the benefit, which is the virtuous act itself (2. 33.1–2; 4. 1. 3). This we would perform even if it did not yield security and love (4. 22. 3); even if it involves risk and loss (4. 12. 2).

Accessio also features in De vita beata, but in relation to pleasure. The whole discussion there bears a close resemblance to De beneficiis, for in both works the Epicurean insistence on the link between voluptas and virtus, presented in ΚΔ‎ 5, is challenged, and in very similar terms (Vit. beat. 7. 1–3; 11. 3; 13. 5; 14. 1; cf. Ben. 4. 2). Seneca argues that, in making virtue serve pleasure, the Epicureans reverse the proper order of things, for virtue must lead and bear the standard. But that is not to deny that virtue yields joy as an accessio (Vit. beat. 9. 2, cf. 22. 3): it is neither the cause nor the reward of virtue, which we do not practise in order to secure it. Here, in De vita beata, the idea of gaudium is associated with the happy life, which is attained through virtue alone; in the treatise, it is the virtuous act of giving that yields joy to the giver, and the virtuous act of offering gratitude that bestows joy on the recipient (1. 6. 1, cf. 4. 15. 2; 4. 29. 3; 2. 22. 1; 3. 17. 4).

De vita beata is, in fact, the work that is closest in subject matter to De beneficiis, for it shows the most interest in the proper use of wealth. Since the happy life consists in the exercise of virtue, the Sage would prefer to be rich rather than poor, for wealth enables him to be generous to others (21. 4–24). Seneca also allows that the imperfect person needs the indulgence of fortune for his happiness, though the Sage can be happy without it (16. 3). As in De beneficiis, wealth must be acquired honourably (Vit. beat. 23. 1–2, cf. Ben. 2. 17. 1–2; 6. 38. 4), and one must not be spiritually dependent on wealth (Vit. beat. 4. 2; 21. 2), being aware that it is not a good, but a positive indifferent, useful but not necessary to happiness (Vit. beat. 24. 5; 4. 3; 22. 4, cf. Ben. 1. 6. 2; 1. 11. 5; 5. 13). It is difficult to give properly, and discrimination in giving is vital.5 Nonetheless, the attitude to wealth in De vita beata is different from the pg 153austere view that prevails in De beneficiis; it is, in fact, unique among Seneca's works in its positive evaluation of wealth, the reason probably being its function as a defence against the charges of hypocrisy made against Seneca, notably by Suillius Rufus.6 In De beneficiis, by contrast, Demetrius is made to say, with the author's approval, that the Sage would rather not have riches (7. 7. 5–11), and the negative view of wealth as a cause of evil prevails (6. 3. 2; 7. 10). Seneca even has Demetrius say that, if some god wished to commit all our wealth to his hands, 'I would not accept it, even if I were going to give it away, since I can see that there are many things which it would not be appropriate for me to give' (7. 9. 1). Throughout the treatise, Seneca is concerned to show that the things given and the services rendered are not goods and not benefits, but only positive indifferents at best (e.g. 1. 5. 6; 5. 13). The beneficium that give us gaudium is the benevola actio (1. 6. 1). Seneca in fact spends more time in defining precisely what state of mind is required, if a gift or service is to count as a benefit and create an obligation of gratitude, than in itemizing and celebrating the actual gifts and services conferred. Pohlenz may have been right when he suggested that the impulse to treat beneficence came from Seneca's having to defend himself, as he does in De vita beata, but that he only came to a deeper understanding of it when he worked more precisely with the question from a theoretical point of view.7

That deeper understanding of beneficence also guides Seneca's teaching in a more humane direction in De beneficiis. As we have already noted (p. 25), Bellincioni regarded the three treatises, De ira, De clementia, and De beneficiis, as steps on an ethical itinerary, during which Seneca developed a doctrine governing the exercise of power in its political and economic aspects: the first two concentrate on negative lessons about abstaining from anger and injury, while the last moves on to teach the necessity of benefitting humanity.8 De ira treats the passion that most of all threatens the duty of mutual help that nature has ordained for man. De clementia advocates the virtue that is opposed to anger and cruelty (1. 5. 6; 2. 4. 1–3), cruelty being anger when associated with power (Sallust, BC 51. 14). But clementia, in addition to being a species of temperantia or self-control (Cic. Inv. 2. 164, cf. Clem. 2. 3. 1), also has an active aspect within a particular sphere of operation: namely, situations where there is provocation to revenge or occasion for punishment. In this work, especially in Book 2, Seneca concentrates on proper conduct in meting out judicial punishment for offences, a theme to which he gives even more attention than to overlooking personal injuries or sparing enemies in war. In doing so, he moves on, from insisting in De ira that judgement be free of anger but severe, concerned with enforcing the law and with inflicting the deserved punishment pg 154(1. 16. 5; 1. 19. 5–6), to advocating punishment that stops short of what could deservedly be imposed (Clem. 2. 3. 2).

It is true that in De ira (1. 19. 5–7) Seneca already thinks of Plato's insistence that punishment should help to reform, and that he also urges making allowances for mistakes and weakness, as in De clementia (2. 7. 2); but Bellincioni points out the contrast between the image in De ira 1. 6, of straightening spear-shafts by fire and force, and the gentler image of punishment as reform, with which our text of De clementia breaks off. Here Seneca compares the way in which the Sage deals with wrongdoers to the way in which good farmers tend trees that need to be straightened, or trimmed, or nourished with fertilizer, or given more access to the sun (2. 7. 4–5). The same contrast appears between the medical simile in the same passage of De ira—where the physician treats patients by fasting or surgery, and no treatment seems harsh if it results in health—and the gentler medical analogy in De clementia 1. 17. 2, where gentle remedies, even placebos, are suggested, and the concern is with restoring health while avoiding shameful scars. Here Seneca is stressing the third ingredient of the Roman concept of clementia, which embraces the Greek praotēs, epieikeia, and philanthrōpia, and he is developing the Stoic notion that man is a social animal, born for the common good (above, p. 151).

De ira treats material belonging to the division of moral philosophy de impetu, De clementia is concerned both with that and with the division de actionibus (see pp. 126–8) . De beneficiis concerns itself principally with the latter,9 and it develops further the positive advice to benefit humanity in a political and economic sense, and thus to maintain the bonds of human society. Moreover, to the offence of ingratitude, which is not, and (according to Seneca) should not be, a punishable legal offence (Ben. 3. 6–17), but which he regards as the root of all crimes and vices (1. 10. 4), he now extends that degree of generosity and understanding offered to offenders against the law in the other two works. And, going further in that direction, he urges us to continue conferring benefits on the ungrateful. The image of the good farmer tending his trees in De clementia has its counterpart at the end of De beneficiis, where Seneca, after urging us at length to imitate the gods, who give benefits even to the malicious and ungrateful, advocates giving the ungrateful a second chance by persisting in conferring benefits on them, like the good farmer who overcomes the sterility of the soil by care and cultivation (7. 31–2).10

pg 155BUILDING ON DE BENEFICIIS

After finishing the treatise, Seneca clearly went on thinking about the subject of benefits, as we have seen in the Letters. In fact, he thought further, not only about the subject itself, but about what he had said on it. For in Letter 81, ostensibly prompted by his addressee Lucilius, he returns to the subject of ingratitude, explicitly mentions De beneficiis, in which the subject is prominent, and takes up a question 'which has not, I think, been sufficiently clarified' (quod non satis, ut existimo, explicatum est). By this he means 'clarified by me', because it is a question he has mentioned in Book 3. 12. 4 of De beneficiis and discussed in Book 6. 4–6.

Letter 81, like so many of the Epistulae Morales, has a complicated structure, which I have tried to render intelligible in the synopsis that forms an Appendix to this chapter. It starts with Lucilius' complaint that he has encountered an ungrateful person. Seneca responds that Lucilius has been lucky not to meet such a person before—there are so many—and exhorts him in the same humane direction, and with the same image, that we have just noted as providing the conclusion to the treatise: 'Even after a poor crop one should sow again, for often what was lost through the stubborn barrenness of poor soil has been made up through one year's fertility' (Ep. 81. 1).11 In this introductory section (1–3), Seneca goes on to repeat several points from the treatise, urging his correspondent not to give up conferring benefits, but to try again, as we do in other activities involving uncertainty and difficulty.12 He then comes to the question that needs further clarification: 'Whether a benefactor who later injured us, has balanced things out and released us from our debt' (an is qui profuit nobis, si postea nocuit, paria fecerit et nos debito solverit). Seneca adds that we may suppose the harm to have greatly exceeded the earlier benefit (Ep. 81. 3).

This question about the subsequently injurious benefactor is brought in as a 'hard case' that will, as the introduction already suggests, lead us to explore the whole issue of showing proper gratitude for benefits and making an appropriate return.13 The context thus guarantees a more ambitious analysis of the question than it had received in De beneficiis. Let us start, however, by reviewing the treatments of the question in the treatise, noting points that will be particularly relevant to its treatment in Letter 81.

pg 156Seneca's previous responses to this question, in De beneficiis, have been given in contexts concerning the possible application of the law to cases of ingratitude—so it is not surprising that, in the Letter, we are going to hear a lot about judges and judgement.14 First, in Book 3, where he has come to the third of his principal topics, that of returning benefits, Seneca considers whether or not ingratitude should be punishable by law (3. 6–17) and concludes that it should not. One reason is that the assessments involved in determining the value of benefits and injuries would be too complex (3. 7–12), and one difficult case he considers is this one:

Aliquis dedit mihi beneficium, sed idem postea fecit iniuriam. Utrum uno munere ad patientiam iniuriarum omnium adstringor, an proinde erit ac si gratiam rettulerim, quia beneficium suum ipse insequenti iniuria rescidit? Quomodo deinde aestimabis, utrum plus sit, quod accepit, an quo laesus est? Dies me deficiet omnes difficultates persequi temptantem. (3. 12. 4)

Someone gave me a benefit, but the same person later inflicted an injury on me. Am I committed by one gift to tolerate any and every injury? Or is it as though I had returned his favour because he cancelled his benefit by the subsequent injury? So how will you assess which is greater, the benefit received or the injury suffered? If I try to work through all the problems, there will not be enough hours in the day.

The unexpected shift in focus between the assessment of the case by the recipient (mihi, adstringor, rettulerim, me) and by a third party (aestimabis), who is neither the recipient nor the injurious benefactor, is something which we shall meet again in Letter 81. The substantive question posed by the recipient in the form of two alternative solutions is not resolved here. The first solution mentioned—total forgiveness of the injury—is not seriously pursued here, because the second solution—counting the injury suffered as a repayment of the benefit—suits the context better, in that it illustrates the problems that a legal settlement would involve, of calculating the relative weights of the benefit and the injury. It is worth noting that the argument given here in support of this second solution, that the subsequent injury has cancelled the benefit, is badly formulated, as Seneca will make clear in Book 6 of the treatise: there he will argue that one cannot cancel a benefit, because the benefit, properly understood, is the act of giving, and that subsequent behaviour can only spoil a benefit, removing the obligation to be grateful and to return the favour (6. 2. 2).15 In Letter 81 that erroneous formulation of the balancing process will be avoided, and Seneca will ultimately come closer to the first solution.

pg 157In Book 6, where the question is raised again of a benefactor who then injures one, the context is precisely this issue of whether a benefit can be cancelled. Seneca there explains:

Non abstulit beneficium, sed opponendo illi parem iniuriam solvit me debito, et, si plus laesit, quam ante profuerat, non tantum gratia extinguitur, sed ulciscendi querendique libertas fit, ubi in comparatione beneficii praeponderavit iniuria. (6. 4. 1)

He has not taken the benefit away but, by setting against it an equal injury, he has freed me from my debt; and, if he wronged me more than he previously benefited me, not only is any gratitude extinguished, but I am free to retaliate and remonstrate, when the injury has outweighed the benefit.

He adds at the end of the discussion:

Dedisti beneficium, iniuriam postea fecisti; et beneficio gratia debebatur et iniuriae ultio; nec ego illi gratiam debeo nec ille mihi poenam: alter ab altero absolvitur. (6. 5. 1)

You conferred a benefit, and afterwards, you inflicted an injury; gratitude was owed for the benefit and revenge for the injury; neither do I owe him gratitude nor does he owe me a penalty: one debt is cancelled by the other.

The similarity in language, especially of the first passage, to the formulation of this question in Letter 81. 3, 15,16 supports the idea that it is this same question that is here answered in the affirmative, while the additional point made in the Letter, about the injury having greater weight, is here said to allow at least a grievance to remain. Next, Liberalis is represented as tired of the distinction between cancelling the benefit and removing the obligation it imposes, but eager to pursue the practical question of how to deal with the injurious benefactor: should I respond separately to the benefit and the injury, with gratitude and revenge, as one does in the courts,17 'or ought I to combine one with the other and do nothing, the benefit being removed by the injury and the injury by the benefit?' (an alterum alteri contribuere et nihil negotii habere, ut beneficium iniuria tollatur, beneficio iniuria?, 6. 5. 3–5). Seneca explains that, since benefits are not subject to the law, the recipient makes the decision: 'Rather, having compared the benefit and the injury, I shall ascertain whether anything further is owed to me' (Potius comparatione facta inter se beneficii et iniuriae videbo, an mihi etiam ultro debeatur, 6. 6. 2). So the same question that will be asked in the Letter is here again, as in Ben. 6. 4. 1, answered in the affirmative, but Seneca does not spell out this time what an imbalance licenses the recipient to do. He ends the discussion by correcting Liberalis, who has deliberately used pg 158one of the expressions ruled out in that previous discussion, i.e. beneficium tollere.18 Seneca insists that the benefit is not removed, only rendered invisible (6. 6. 3).

Letter 81 effectively takes up this question, of how to deal with an injurious benefactor, where De beneficiis left it. It is 'a kind of appendix to the treatise'.19 Indeed, as we have seen, the context in which the discussion of the question arose in the treatise is still apparent in the concern with judges and judgement in the letter. It is not hard to see why Seneca was dissatisfied with his earlier treatments of this difficult question, when we consider the humane attitude to ingrates that he reached by the end of De beneficiis and advocated in its closing chapter, a chapter which he recalls in the opening of the letter. For those earlier treatments rested on the idea of repaying injury by injury, a view which Seneca had already disclaimed in previous works.20

In Letter 81, having posed the question 'whether a benefactor who later injured us, has balanced things out and released us from our debt' in §3, Seneca devotes to it a considerable portion of the letter: but he interrupts this discussion in order to consider the difference between the Sage and the Fool as far as returning benefits goes (8–14), and to defend the paradox 'only the Sage knows how to return a favour' (Ep. 81. 11). Then, marking the latter topic (15) as something already sufficiently examined (presumably in De beneficiis 5. 12–17), Seneca returns to finish the question of the injurious benefactor (15–17). After this, in the last sections of the Letter, he discusses the issue of gratitude more generally: returning a benefit, when rightly understood, as it is by the Sage, yields more joy than receiving one, because it is a virtuous act (17–23). But though the Sage alone has the wisdom to understand this (24–6)—for he alone knows the true value of things (27)—ordinary opinion, though it leads us to wrong valuations and hence to ingratitude (29), also regards gratitude as the most honourable of things (30). In fact, the shame this induces in the ungrateful can make it dangerous to confer great benefits on anyone, 'for, since he then knows it is shameful not to repay, he does not want there to be anyone whom he should repay' (32).21

pg 159This pessimistic close, however, comes as a surprise, after the exhilarating depiction of the attitude of the Sage earlier on. Seneca perhaps wants to bring us back to the level of the imperfectus, with an echo of De vita beata, where the ill effects of public opinion on ordinary people were so vividly described (1. 2–2. 2).

In fact, in Letter 81 problems about moral level confront us already in the discussion of the question concerning the injurious benefactor. Thus we hear of the opinion of a 'strict judge' (rigidus iudex) and of a 'more lenient judge' (remissior iudex), but also of a vir bonus who appears in §6 and again in §15, after the interruption about the Sage. Is any of these identical with the Sage? The point is a vexed one,22 not to be resolved lexicographically.23 As will emerge in what follows, I do not think any of these is identical with the Sage.

The first thing to notice is that the vir bonus, on both his appearances, is the recipient himself, as is emphasized in both cases.24 The rigidus iudex is brought in to adjudicate between the two parties, and he gives his just opinion (4) that benefits and injuries can balance out, because, 'although the injuries are greater, credit must be given to the benefits for what remains after the injury' (quamquam iniuriae praeponderent, tamen beneficiis donetur, quod ex iniuria superest). The grounds for this decision are, apparently, that the greater weight of the injury is balanced by the fact that the benefit came first.25 In an aside (5–6), Seneca mentions that, of course, the state of mind of the benefactor (later injurer) must be considered, but, as this is hard to conjecture, he proceeds on the assumption that a true benefit and a true injury are involved, and that the injury is greater. The recipient, if he is a 'reasonable man of good pg 160moral intentions',26 will now apply the assessment of the rigidus iudex to himself: he will set limits to his own calculations and add to the benefit and subtract from the injury, thus balancing them out (6). This gives an affirmative answer to the question posed in §3.27

Seneca, however, prefers a different solution. He now declares a preference for the opinion of the remissior iudex, who orders the injury to be forgotten, and only the service remembered. To Lucilius' objection (7) that this is against justice, which requires gratitude to be given for a benefit and requital for an injury, Seneca replies that this view is only true when separate agents are involved, not the same agent.28 In the latter case, the force of the injury is extinguished by the benefit. As an illustration, he notes that, if someone deserves pardon for an offence without doing good deeds previously, he deserves more than pardon if he conferred prior benefits.29 'For I do not set an equal value on benefits and injuries. I give more weight to a benefit than to an injury'(8).

The passage on the Sage that follows (8–14) apparently explains this advice. 'The Sage will examine all the circumstances in his own mind: how much he has received, from whom, when, where, and how' (§10). By contrast, the ordinary man of good moral intentions does not know how to measure his debt of gratitude, and so he is better off not trying to calculate the relative weight of benefit and injury too exactly, but should incline towards the side of the benefit.

When the discussion of the question resumes (15), the vir bonus (the recipient again) favours the benefit, because it is a benefit. He will calculate the relative weight of benefit and injury, taking into consideration factors such as the persons affected (besides the recipient), but he will overlook a small over-balance on the side of the injury; in fact, he will ignore even a much greater injury, will incline to excuse it, and will be unwilling to regard the benefit as repaid by the injury, but will wish actually to repay it (16–17). He is thus following the assessment of the remissior iudex, given in §6, and answering the question initially posed in §3—whether he is released from his debt—in the negative.

Neither the first (affirmative) answer nor the second (negative) answer is that of the Sage, who, like the vir bonus, considers the question as a recipient of the benefit and of the subsequent injury, but comes up with a more precise answer. For the Sage dismisses the injury done to him and remains well disposed to his benefactor unless the bad deeds far outdistance the good, and in an obvious way (25). When they heavily preponderate, he goes as far as resuming the neutral attitude he had towards the person before he conferred the benefit pg 161and the injury on him: he regards himself as owing the person nothing and resents nothing. That is, his assessment is less severe than that of the rigidus iudex, in that he agrees with his advice to balance out benefit and injury, but only when the injury heavily preponderates, and not when it is only somewhat greater. On the other hand, his assessment is more severe than that of the remissior iudex, in that he does not always dismiss the injury done to him, but does so only when the injury does not far outdistance the good. And it is only when benefit and injury are equal, that he will retain some benevolent feelings, and, while having no obligation, will go on wanting to have one, and will behave like someone who pays his debts even after they have been legally cancelled (25–6). The Sage, then, does not go as far in the direction of strictness as the vir bonus when he is following the rigidus iudex, or as far in the direction of laxity as the vir bonus when he is following the remissior iudex: the Sage can be just without being strict, like the former; he can be humane without being indiscriminately generous, like the latter. The reason is that he is capable of making an accurate calculation as to the weight of the benefit and of the injury, whereas the vir bonus is not. For ordinary people of good moral intention, the hyperbolic generosity advocated by the remissior iudex is the better course. As Seneca says, towards the end of De beneficiis, 'When you cannot quite trust the people to whom you are giving orders you must demand more of them than is necessary, so that the necessary level of compliance is achieved. The point of hyperbole is, in every case, to get to the truth by way of a falsehood' (7. 23. 1).

Letter 81 then continues the forward movement of De beneficiis itself, towards a more humane solution to the problem of ingratitude. In keeping with the generous attitude to ingratitude at the end of the treatise, where we are reminded that we too suffer from the vice and are imperfecti, Seneca here reconsiders the question of a benefit followed by an injury, to which he gave an unsatisfactory answer when he treated it as a subordinate topic in the treatise. Seneca's rethinking here shows that it was not just a rhetorical flourish, when he wrote that the primary job of philosophy is to teach us to owe and to repay benefits well. If we had the comprehensive work on ethics, the Moralis Philosophia that he was working on at the dramatic date of the later Letters (106, 108, 109), i.e. the autumn of 64, we might well have seen a further development of his humane attitudes. In any case, Letter 81 underlines Seneca's belief that the social instinct implanted in humans by providence, and the role of benefits in binding together human society, should take priority over cruder calculations of justice.

pg 162Appendix: Synopsis of Letter 81

1–3

Introduction

 

1

Lucilius has met an ingrate: he is advised not to become ungenerous.

 

1–2

Perseverance will bring success in finding gratitude.

 

3

These points were treated sufficiently in De beneficiis, but not the question that follows.

3–8

Has a benefactor who later injured us balanced things and released us from our debt?

 

 

3

Let us assume that the later injury is much greater than the benefit.

 

4–6

The strict judge gives, as his just opinion, a positive answer: the priority of the benefit balances the greater weight of the injury.

 

 

5

The question of the benefactor's attitude affects the weighting but is hard to assess: so we just assume the injury was greater.

 

 

6

The good man will restrict himself (to balancing the account) by adding to the benefit and subtracting from the injury.

 

6–8

The more lenient judge prescribes forgetting the injury, remembering the service. Seneca prefers this view, as benefits have more weight than injuries, when performed by the same person.

8–14

Only the sage knows how to return a benefit; the fool lacks the knowledge to gauge the amount, the time, the place, the manner, of its return.

 

9–10

The correct expression referre gratiam shows it must be voluntary.

15–17

The question about the benefactor who later injures receives a negative answer.

 

15–16

The good man, judging as the recipient, will ignore the discrepancy in weight, even if it is great.

 

17

He will not let the injury balance out the benefit but will want to make a return.

pg 163

17–23

Returning a benefit brings greater joy than receiving.

 

18

A benefit, like a loan, should be repaid with interest.

 

19

Showing gratitude, a virtuous act, brings happiness, while ingratitude brings unhappiness.

24–6

The Sage gains eternal joy from remembering benefits received and deliberately forgets his injuries.

 

25

When the subsequent injury is far greater than the benefit, he returns to his neutral attitude before the benefit.

 

26

When the subsequent injury is equal to the benefit, he retains benevolence and ceases to owe, but not to want to owe, a return.

27–9

Gratitude requires knowing the true value of things.

 

29

We are misled by public opinion.

30–2

Public opinion is right to advocate gratitude.

 

32

Shame at an inadequate return leads to dangerous ingratitude.

pg 1649The Afterlife of De beneficiis up to the Renaissance

In comparison with some works of Seneca, De clementia and the Epistulae morales in particular, De beneficiis is a self-effacing work. It has rarely been quoted or cited by name by authors who have made use of it, and its presence can be inferred more often than it is attested.

The anonymous author of the Octavia, and the poets Statius, Martial, and Juvenal, show that Seneca was highly regarded after his death, well into the first decades of the second century. Not even the satirists were interested in pursuing the criticisms of his lifestyle that had plagued Seneca in his last years, or in imitating the irony with which Petronius had treated his liberal views on slavery (Ep. 47, cf. Sat. 71).1 Quintilian (10. 1. 125–31), Fronto (De orationibus 1–4), and Aulus Gellius (NA 12. 2), who were themselves critical of his style and literary tastes, felt it necessary to issue warnings aimed at discouraging the young from reading Seneca's prose. Despite the implication in Tacitus' Annals and in Suetonius' Life of Caligula that he was no longer in fashion in their day,2 Seneca continued to be read 'down to the threshold of the Dark Ages', that is the seventh and eighth centuries. Nonetheless, we have no evidence that the poets and stylistic critics mentioned above were acquainted with De beneficiis in particular.3 Even when Aulus Gellius quotes, as one of the few good remarks he can find in Seneca, an epigram about avarice, it is not in the form in which it is found in the treatise.4

pg 165The historian Tacitus, however, as we have suggested (pp. 84–7), did make use of the treatise for the dialogue between Seneca and Nero that he included in the Annals under the year 62, and he clearly expected his readers to recognize the similarity of themes and ideas. Seneca is made to discuss the problems of receiving benefits; Nero is concerned with giving and returning benefits. Thus, between them, they cover the three topics the treatise aims to discuss (Ben. 1. 4. 3). The exchange of benefits that Tacitus treats here is the problematic one between those who are unequal in power, in this case, between ruler and ruled. It is well represented in De beneficiis, which deals with imperial gifts and favours in all but one book, and which shows mostly the negative side of imperial generosity and the difficulties experienced by recipients. All Nero's predecessors as Princeps provide concrete examples of doubtful gifts, and it is through such counter-examples that Seneca teaches how the ruler should give.5 Tacitus does not, however, use Senecan vocabulary: he even eschews the term beneficium. He does not imitate Seneca's favourite prose rhythms either.6 How far he alludes to Seneca's stylistic 'tics' has been actively debated. The presence of military imagery and the reference to studia in umbra educata are hardly peculiar to Seneca. More telling are the abrupt start and the personal address, features with which several of Seneca's dialogues begin, such as De providentia and De brevitate vitae: in De beneficiis, too, Seneca comes straight to the point. There is also the catalogue of questions that Seneca gives, and the use of animus as the subject of concrete activities which would be more suitably attributed to an actual person.7

The Latin church fathers knew Seneca's works well. In the early third century Tertullian, for whom he is 'Seneca, often one of our own' (Seneca saepe noster, De anima 20) actually quotes from De beneficiis (unnamed) from memory (Ben. 4.6.6); Lactantius, in the fourth century, provides many fragments from lost works; while Jerome's little biography of Seneca, written in 393, mentioning for the first time the correspondence with St Paul, ensured that Seneca's philosophy would be read by the Christian world. It is likely that the correspondence was forged precisely between Lactantius' assessment of Seneca as 'the most acute of all Stoics' (Inst. 2. 8. 23) in 325 and Jerome's mention of it.8 St Augustine had doubts about the correspondence (Ep. 153. 14), and regarded Seneca as hypocritical in practising traditional pagan religion, while knowing it was false. However, Jerome's assessment of Seneca, as a man of virtuous life, he did not contest.

pg 166Seneca was known to late antique pagan authors, but we have mostly scattered allusions, such as Sidonius Apollinaris' misunderstanding, in the fifth century, of Martial's celebration at Epig. 1. 61. 7–8 of two Senecas, father and son, as a reference to a tragedian Seneca and a philosopher Seneca (Carm. 9. 232–6): a division that was to linger on into the late seventeenth century.9 Macrobius, however, in his Saturnalia, used, for a discussion of the humane treatment of slaves (1. 11. 23–4), not only Seneca's Letter 47 but the anecdote about the siege of Grumentum in De beneficiis 3. 23. This Senecan theme was to be picked up by writers in the Middle Ages.

In the sixth century Bishop Martin of Braga dedicated to King Miro a work entitled the Formula vitae honestae, which was very popular in the Middle Ages, when it passed for a genuine Senecan work (it was first unmasked by Petrarch, then again by Erasmus10 ). 'Spain rewarded its son by plagiarizing his work', in the words of Leighton Reynolds.11 Since the Bishop had previously composed a work De ira, which is a mosaic of quoted and transmuted passages from Seneca's De ira, it is usually assumed that one particular Senecan work lies behind Martin's later work too (see Ch. 2, p. 21). The usual candidate is Seneca's lost treatise De officiis, attested by the grammarian Diomedes (fr. 57 Vottero; fr. 25 Haase), for Martin's Formula is organized around the four cardinal virtues, like Cicero's De officiis, and it would not be unreasonable to assume that Seneca's work of that title was similarly organized.12

The first of Seneca's works to emerge after the Dark Ages were De beneficiis and De clementia, which were copied in north Italy about the year 800. This manuscript is the archetype of the medieval tradition, and it still exists as the Codex Nazarianus, known as N, now in the Vatican library (Pal. lat. 1547). So does another ninth-century manuscript, known as R (Reg. lat. 1529): the Codex Reginensis, probably a direct copy of N incorporating corrections made in the hand called N2.13 It is from R that all the rest of our extant manuscripts descend. After a dormant period, copies begin to be made in the late eleventh century, spreading out from northern France, where R had ended up in one of the monasteries on the Loire.14 The first quotation from De beneficiis, giving pg 167the name of the author, is in the Chronicle of Hugo of Flavigny in that region, written in the last decade of the eleventh century, and both works are listed in the catalogue, dated to 1093, of the abbey at Pomposa in northern Italy.15 By the twelfth century, the text of the two works was available even in England and Germany; by the thirteenth century, in Spain as well.16

The Middle Ages inherited a picture of Seneca as a virtuous man whose philosophy had many overlaps with Christianity. The supposed friendship between Seneca and St Paul, widely accepted, may help to explain, not only the popularity of Seneca's writings in the Middle Ages, but the positive view taken of them. In transmission, the two treatises De clementia and De beneficiis may have been separated off from the other Senecan works and combined, because of their common theme of leniency and generosity;17 but in any case their combination is symptomatic of the use to which they would be put after the renewed dissemination of Seneca's works. In Germany, in France, in Italy, and in Spain, both treatises were used to provide advice to rulers in the 'Mirror for Princes' tradition, De clementia actually furnishing the image of the mirror and giving political advice, De beneficiis offering exhortations to liberality and cautions against heeding flattery. The idea was to provide an ethical basis for rule, based on nature, not on religion, but compatible with Christianity. It drew on the ancient idea of the virtuous ruler, supported by willing subjects. Notdurft shows how, in the work Moralium dogma philosophorum written by William of Conches for the young Henry of Anjou, who in 1154 became Henry II of England, De beneficiis is used, along with Cicero's De officiis, to provide advice on how and what to give and on showing gratitude, including not complaining of ingratitude.18 The humane view, in Seneca's treatise, of how to treat slaves is also applied to the serfs of his time. John of Salisbury followed Wilhelm, as did Giraldus Cambrensis, who, however, also brought in De clementia. In Italy, the roots of the Mirror of Princes tradition, which Machiavelli was to attack in the sixteenth century, can be traced back to the thirteenth century.19 In Spain too, from the thirteenth century to the sixteenth, De clementia was read as a work of political ethics, a Fürstenspiegel geared to princely education, while De beneficiis was read as a tract about princely liberality, a subject central to that education.20

It would take another book-length study to collect and discuss allusions to De beneficiis in later centuries, and it is unlikely to be as illuminating for our present study of the work itself, as this sketch of the early history of its reception. For what is shown by the use to which De beneficiis was put immediately pg 168after its re-emergence on the European cultural scene is that, even before Tacitus' work was rediscovered in the late fourteenth century, readers saw what the historian had seen: the theme of liberality between ruler and ruled was central to Seneca's concerns. This lends support, first, to Bellincioni's notion of De beneficiis and De clementia as related works, developing a doctrine governing the exercise of power (pp. 153–4). It also lends verisimilitude to the interpretation of De beneficiis given in Ch. 3 from the perspective of social history. For Seneca's essay has many aspects. It is a treatment, from the Stoic point of view, of an important topic in moral philosophy; it is a guide to ethical behaviour in relations of social exchange; and, finally, it is an analysis of the Roman aristocratic code of beneficence, which insists on the inclusion of the Princeps within aristocratic society, and which holds him to the same ideals in giving and receiving gifts as are appropriate for his nominal peers.

Notes

1 The name was restored by Lipsius from the 'C. Aminius rebius' of the Mediceus manuscript in Tacitus' description of the disreputable ex-consul: ex primoribus peritia legum et pecuniae magnitudine cruciatus aegrae senectae misso per venas sanguine effugit, haud creditus sufficere ad constantiam sumendae mortis, ob libidines muliebriter infamis ('among the leading men in his skill at law and the size of his wealth, he escaped the sufferings of an ailing old age by opening his veins, though he was believed incapable of the steadfastness to commit suicide, given his bad reputation for effeminate lusts.').

2 Rebilus is there being likened to Fabius Persicus, who is insulted again later in Book 4 (see ad 4. 30. 2). Clearly, if his death was later than that of Caninius Rebilus, the terminus post for De beneficiis would need to be advanced. In fact, Persicus, who is not attested in the virtually complete Acts of the Arval Brethren after the winter of 57, probably died earlier (PIR2 F 51): Syme 1986, 417 suggests that he died in the reign of Claudius.

3 Parallels to the ideas in §1 of Ep. 81 down to peribunt can be found in Ben. 1. 1. 12–13 and 7. 29. 1; the agricultural metaphor recalled at the end of §1 is in Ben. 7. 32; the notion of trying again in §2 occurs in Ben. 1. 1. 9–10 and 7. 31. 5; the idea of life grinding to a halt comes in Ben. 4. 33. 3. For Mazzoli 2007, 586, n. 3, the clear allusion to Ben. 7. 32 (sterilitatem soli) at Ep. 81. 1 (soli sterilitate) is the conclusive argument that the whole treatise was finished before the letter.

4 Ep. 67. 1 speaks of ver … inclinatum in aestatem ('spring … rounding into summer') while Ep. 86 places itself in the latter part of June: Iunius mensis est, quo tibi scribo, iam proclivis in Iulium ('it is the month of June as I write, well on the way towards July'). See Griffin 1976, 400.

5 Préchac 1961, pp. xiv f., followed explicitly by Guglielmino 1968, p. x; Abel 1967, 165 and 1985, 708. Grimal 1978, 303 and Chaumartin 1989, 1702–3 think that at least the first six books, and probably the seventh, were composed before the letter; Cooper and Procopé 1995, 183 that 'the bulk of the work must have been completed before the composition, in summer ad 64, of Letters 81'.

6 As was already seen by Münscher 1922, 64–5 and Albertini 1923, 33–4. They, however, thought that Books 1–4 were published earlier than 5–7, and that these last were issued singly as they were written: this is to take Seneca's openings of these books, where he claims to be induced by his addressee to go on adding trivial questions, at face value (against which see Ch. 6).

7 Tac. Ann. 15. 45. 3 for the retirement in 64; Dio 62. 25. 3 on handing over money, confirmed by Tac. Ann. 15. 64, where 'then at the peak of his wealth and influence' (tum praedives et praepotens) confirms Seneca's change of circumstances by his death.

8 Tac. Ann. 16. 13. 3; Ep. 91 (dramatic date before the beginning of autumn 64, which is marked in Ep. 122. 1). See also n. 26.

9 De clementia dates itself to between 15 Dec. 55 and 14 Dec. 56 (Griffin 1976, App. A 3). We do not know when in 56 Caninius Rebilus died.

10 If it were certain that Ben. 7. 5. 1 (see Introduction, n. 6 and ad 7. 5. 1) was an allusion to the Libri moralis philosophiae, which S. explicitly says he was working on at the time of the later Letters, it would be reasonable to date the work to the 60s, close to the time of the Letters.

11 See e.g. the remarks of Grimal 1949, 178.

12 Ferri 2003, 251 suggests that the turning-point of 62, celebrated in the pseudo-Senecan tragedy Octavia, was taken over from a Flavian historian by Tacitus. However, the murder of Octavia was already picked out as marking a stage in Nero's decline in Nero's lifetime by Subrius Flavus, according to Tacitus Ann. 15. 67. In Griffin 1976, 226, I suggested that stressing the change of advisers, Burrus being replaced by Tigellinus, was Tacitus' own idea: the praetorian prefect in Octavia is unnamed and colourless.

13 Tac. Ann. 14. 10–11 for the letter; 14. 56. 3 for his restricted activities from 62.

14 This approach, applied to the tragedies by Nisbet 1990, has been well received, but there is evidence that, as regards tragedy, 'Roman audiences were quick to sense contemporary references' (97); one cannot assume that for Seneca's philosophical works (see Griffin 1976, 12–20).

15 By Inwood 1995a, 244 = 2005, 69.

16 Friedrich 1914, 1406; Münscher 1922, 66–7, followed by Pohlenz 1941, 80 n. 82.

17 Grimal 1978, 304–5; Chaumartin 1989, 1703.

18 See the discussion in Gercke 1896, 307–11.

19 Préchac 1961, p. xi; Grimal 1978, 304, followed by Chaumartin 1989, 1708.

20 Fuhrmann 1997, 312–14 stresses that political caution was not the reason. He attributes it to philosophical detachment from the things of this world, since Seneca could easily have mentioned the Rome Fire safely in this letter. Ker 2009, 108 thinks the Roman reader would recognize Seneca's presentation of a 'much-needed perspective on their seemingly singular event', citing Bedon 1991, 53–7 for the idea that the Lyons fire is exaggerated by Seneca to the scale of the Rome fire, which he preferred not to mention, given the rumours surrounding it.

21 The first two are adduced by Letta 1998, 231; the last by Préchac 1961, p. xxiii.

22 On the uselessness for dating purposes of Seneca's thematic repetitions, which occur geared to different questions, see Pohlenz 1941, 64 in particular (concerning the theme of providence), 66–8 in particular (concerning Seneca's attitude to Epicureanism): 'Die Haltung wird im Einzelnen durch das jeweilige Problem bestimmt. Das darf man auch sonst bei Seneca nicht vergessen' (68).

23 On Aebutius Liberalis, see Griffin 1976, 455–6 (Appendix D. 8).

24 On the view of Sonntag 1913, 63, that he was chosen as addressee just because of his name, see p. 142.

25 Compare the case of Annaeus Serenus, who is a real man known from Tacitus and other early imperial authors (PIR2A 618), and to whom Seneca addresses the appropriately titled De tranquillitate animi, among other works.

26 See Griffin 1976, 400, where 'end of July—fire at Lyons' should read 'end of June—fire at Lyons'. Bedon 1991, 48 suggests a date for the Lyons fire of the first half of August, a time when fires are not uncommon in the Rhône valley.

27 If the argument here is accepted, it must count against the otherwise attractive suggestion by Ker 2009, 107–8 that Letter 91's 'real purpose may be partly to present the disaster as an opportunity for Liberalis to live up to his name and contribute beneficia for the rebuilding of his native city'.

28 Tac. Ann. 15. 18; Hist. 1. 14. That his consulship was in 61 or 62 is confirmed by the portoria law of Ephesus (see Eck 1981, 227–30).

29 Dobson 1972; Dobson 1974, 405–12.

1 Sonntag 1913, 12; Gomoll 1933, 76. The same view is taken by Schwarzenberg 1966, 63 n. 2, writing about the Graces.

2 See Ch. 2 nn. 26–7.

3 At SVF i. 422 (D.L. 7. 166), von Arnim suggested separating Dionysius' into two works, the first on wealth, but Ben. 1. 3. 9 and SVF ii. 1083 show how closely the topics of wealth and beneficence were related for the Stoics, being worked together in these passages into the Stoic allegory of the Graces. See below, p. 103.

4 Ben. 3. 22. 3: Inter se contraria sunt beneficium et iniuria; 6. 5. 1: Dedisti beneficium, iniuriam postea fecisti; et beneficio gratia debebatur et iniuriae ultio, cf. Ep. 81. 7: "'But surely", you say, "it is the part of justice to render to each that which is his due—thanks in return for a benefit, and retribution, or at any rate ill-will, in return for an injury" ' ('Hoc certe', inquis, 'iustitiae convenit, suum cuique reddere, beneficio gratiam, iniuriae talionem aut certe malam gratiam'), see Moussy 1966, 295. Compare the Epicurean idea in ΚΔ‎ 1 (D.L. 10. 139) that the deity is affected οὔτε ὀργαῖς οὔτε χάρισι‎ ('neither by feelings of anger nor of favour'), cf. Philodemus, On Piety 1148, being unaffected by human acts: Cicero translates this as neque ira neque gratia (Nat. d. 1. 45; otherwise Lucretius 2. 651).

5 Moussy 1966, 412–14.

6 In support Chaumartin 1985, 34 n. 26 notes the description by Diogenes Laertius of a work of Posidonius, once as Peri kathēkontos (D.L. 7. 129) and another time as Peri kathēkontōn (D.L. 7. 124); the singular without the article, and the indifference as to singular or plural, rather count against Dyck's attempt (1996, 7 n. 18) to explain Cicero's rendering of Panaetius' Peri tou kathēkontos as De officiis by the lack of a Latin definite article. Cooper and Procopé 1995, 187 clearly agree with Chaumartin, as they render both the singular and plural titles of such works as On Favours.

7 Moussy 1966, 412–14.

8 Similarly, Brunt 2013 argues in ch. 5, App. 1 §2 that the plural title of Posidonius' work, i.e. Peri kathēkontōn, is correct (see above, n. 6) and that, whereas Panaetius had concentrated on exposing the very nature of the kathēkon with illustrative examples, Posidonius was more concerned to catalogue the kinds of actions that should be performed. Cf. also §12 of the chapter: 'Cicero's title suggests that his work will register all the kinds of actions that are "appropriate" (kathēkonta) whereas that chosen by Panaetius indicates that his design was to show what is in its very nature appropriate.'

9 Moussy 1966, 442–3; Loew 1908, 82 n. 1.

10 Cic. Nat. d. 1. 39–41.

11 Cooper and Procopé 1995, 197 nn. 5–6 suggest that both versions are Stoic and both inadequate as 'the process of giving, accepting and returning a favour only needs two participants'. Yet another version of the allegory appears in the Peripatetic doxography, probably emanating from Arius Didymus in the late 1st c. bc and preserved by Stob. 2. 143. 18–20 W: 'That " charis " is used in three ways, the doing of a useful favour for its own sake, the return of the useful favour, the remembering of such a favour' (Χάριν δὲ λέγεσθαι τριχῶς, τὴν μὲν ὑπουργίαν ὠφελίμου αὐτοῦ ἐκείνου ἕνεκα, τὴν δ‎ʼ ἄμειψιν ὑπουργίας ὠφελίμου, τὴν δὲ μνήμην ὑπουργίας τοιαύτης‎). The context is identifiably Peripatetic, so the allegory could be an independent development by the Peripatetics from Aristotle's reference to the cult of the Graces (NE 5. 5. 1133a3–5), though the third element is different from his starting the process of giving again (see below). But it could be a development by the Peripatetics of the Stoic idea of allegorizing the Graces. It is notable that Chrysippus' version also included the idea of memory, symbolized by the youth of the goddesses (Ben. 1. 3. 5). For the connection with visual representations, see ad 1. 3. 2–4. 5.

12 Loew 1908, 82 n. 1.

13 Moussy 1966, 411–12, 437; Parker 1998, 108–9. The entry on charis in the Peripatetic doxography (see above, n. 11), just after giving an allegorical interpretation of the goddesses, notes the former sense of charis as 'grace' or 'joy': λέγεσθαι δὲ χάριν καὶ τὴν ἐν ὂψει ἢ ἐν λόγοις, καθ‎ʼ ἣν τὸν μὲν εὔχαριν ὀνομάζεσθαι, τὸν δ‎ʼ ἐπίχαριν‎ ('grace in appearance or in discussion is said to be that in virtue of which one person is called "gracious", another "charming"', Stob. 2. 143. 21–3).

14 For the significance of this way of defining the exchange obligation, see p. 45 with n. 66.

15 SEG 39 (1989), 1243 v 45; 41 (1991), 1003 C/D 29–44.

16 Habicht 1982, 85–93; Parker 1996, 272–3.

17 For the abundant evidence for public honouring of benefactors in this period, see Gauthier 1985.

18 Mikalson 1998, 173–7 argues that for Athenians the cult had the traditional significance of agricultural prosperity and peace linked to democracy.

19 SEG 43 (1993), 704, vv. 16–18, cf. e.g. SEG 41 (1991), 1003, C/D 40–1; cf. Ma 1999, no. 10. 17–18; 11. 10–12. Dio Chrysostom says: ὁ νόμος πᾶσιν ὧν ἂν εὐεργετήσωσιν ἑτέρους ἐκτίνει τὰς χάριτας‎ ('the law renders thanks to all for the kindnesses they show to others', Or. 75. 6).

20 Plut. Flam. 1; OGI. 669, 44; P. Oxy 273/14; P.Grenf. 2.70, 5; Ael. Arist. Funeral Speech for Alexander 15; Marcus Aurelius 5.6.1. uses it of a favour given. See MacMullen 1986, 523 n. 35.

21 Cornutus' formulation is οἱ δ‎ʼ ἐξ Εὐρονόμης‎, καὶ τούτου παριστάντος ὅτι χαριστικώτεροί πώς εἰσις ἢ ὀφείλουσιν εἶναι οἱ μεγάλους νεμόμενοι‎ ('some say they descend from Eurynome, and this represents that those who enjoy great possessions are more bounteous in some way than they are obliged to be'); Seneca's is Eurynomen enim dictam, quia late patentis patrimonii sit beneficia dividere ('she was called Eurynome because the sharing out of benefits requires an inheritance that spreads far and wide').

22 SVF ii. 1081: φησὶν εἶναι καὶ τὰς Χάριτας τὰς ἡμετέ‎[ρ‎] ας καταρχὰς κα‎[ὶ‎] τὰς ἀνταπ‎[ο‎]δόσεις τῶν εὐε‎[ργ‎]εσιῶ‎[ν‎]; Ben 1. 4. 4: ad hanc honestissimam contentionem beneficiis beneficia vincendi sic nos adhortatur Chrysippus.

23 Moussy 1966, 409.

24 The best discussion of the various meanings of gratia and the expressions in which it occurs is still Moussy 1966, especially 249–302.

25 Gomoll 1933, 75–6.

26 Moussy 1966, 286–7.

27 Cic. Fam. 2. 6. 2, which actually reads nullam esse gratiam tuam quam non vel capere animus meus in accipiendo vel in remunerando cumulare atque illustrare posset ('whatever favour you may bestow on me my mind would have the capacity to accept it and the ability to enhance and add lustre to it in repaying it'); Amic. 31.

28 e.g. II Verr. 3. 115: benefici gratiaeque causa, 189: in benefici loco et gratiae.

29 Leg. ag. 2. 7.

30 See n. 1.

31 e.g. Cic. Sest. 70: ad amplissimi beneficii gratiam magis pertinere videret ('he saw it would bring him greater gratitude for so generous a benefit'); Fin. 2. 117: tollitur beneficium, tollitur gratia, quae sunt vincula concordiae ('generosity is abolished, gratitude is abolished, which are the bonds of mutual harmony'); Seneca, Ben. 2. 33. 3: cum benigne acceptum est beneficium, qui dedit, gratiam quidem iam recepit ('when a benefit is received graciously, the giver has already received gratitude'); 5. 11. 1: beneficium et relatio gratiae ultro citro ire debent ('a benefit and the return of a favour ought to go reciprocally'); 4. 1. 1: an beneficium dare et in vicem gratiam referre per se res expetendae sint ('whether conferring a benefit and doing a favour in return are things worthy of choice in themselves'); 6. 11. 4: beneficium dedero, non gratiam rettulero ('I shall have given a benefit, not returned a favour'); Ep. 81. 7: suum cuique reddere, beneficio gratia ('to render to each his due, gratitude for a benefit').

32 Moussy 1966, 153 n. 1, 249–50: the active sense accounts for 75% of his uses (89% in De beneficiis; 65% in Epistulae Morales) vs. 45% of Cicero's and 48.75% of Livy's: these are the authors who show the most frequent use of gratia.

33 Chaumartin 1989, 36–7 n. 35 attempts, unconvincingly, to reinterpret 4. 18. 1 so as to avoid attributing to Seneca the reciprocal use of beneficium: the passages he adduces (Ira 1. 5. 3; Ben. 1. 4. 2) do not indeed imply that the word is being used of both sides of an exchange, but they are not true parallels.

34 See Ch. 3 n. 32. The common use of beneficium reddere (e.g. Plaut. Pers. 762; Ter. Phorm. 336; Cic. Sen. 2; Off. 1. 48) as well as beneficium referre, beneficium remunerari, and beneficia rependere are not comparable (pace Dyck 1996, 162–3) because they do not strictly imply that the thing returned is also called a beneficium, any more than to 'repay a loan' implies that what we give the lender is itself a 'loan'. These expressions use the term appropriate to the original transaction, though what is returned is not a beneficium, any more than the actual thing returned is identical with the thing given, as Seneca indicates at Ben. 6. 5. 2: Cum dicimus 'Beneficium illi reddidi', non hoc dicimus illud nos, quod acceperamus, reddidisse, sed aliud pro illo. As the same discussion makes clear, beneficium reddere is just an alternative locution to beneficio gratiam reddere.

35 As Armisen-Marchetti 2004, 7 remarks, 'De fait le De beneficiis est en grande partie un traité De gratia, la question du don étant considerée d'emblée dans la perspective de la reconnaissance, l'ingratitude du bénéficiaire tenant largement au fait qui la bienfaisance s'est mal exercée'.

36 Saller 1982, 21. Schwarzenberg 1966, 63 n. 2 points out that neither Latin word conveys the sense of charm or grace in charis, but that beneficium conveys it less than gratia.

37 Hellegouarc'h 1963, 205.

38 See ad 6. 2. 1–2. Cf. 1. 5. 2 where the distinction is put as non potest beneficium manu tangi ('a benefit cannot be touched with the hand').

39 As with officium, where, as Dyck 1996, 5 points out, 'the etymology from facere/efficere long continued to be felt'.

40 This is the level of Phidias' artificium at Ben. 2. 33. 2 in which there is an expected return for something actually produced.

41 See Ch. 2A.

42 Moussy 1966, 300; e.g. Cic. QF 1. 1. 20; Livy 3. 36. 7; Elder Seneca, Controv. 9. 2. 11.

43 Moussy 1966, 376.

44 Ibid. 371.

45 Ibid. 273.

46 See Ch. 3C. The quotation is from Inwood 1995a, 263 = 2005, 91.

47 See Ch. 3C.

48 See p. 11.

49 TLL ii, cols. 1885–6, s.v. beneficium 2.C.a and b.

50 Hellegouarc'h 1963, 202–8.

51 Bellincioni 1984a, 122 regards the theme that holds the whole complex work together as the process of moving away from the usual meaning of beneficium to the meaning that captures its ethical content.

52 Hyg. 2 agrim. 202 L = 158.31–2 Campbell.

53 CIL vi. 1884: see Diz. epig. i. 998; Plin. Ep. 10. 95, cf. 10. 105 (a grant of ius Quiritium).

54 Diz. epig. i. 996: e.g. beneficium respondendi; beneficium adsidendi; beneficium anulorum.

55 Caesar, BC 1. 75. 2, 3. 88. 5; Dise 1997, 274. According to the Historia Augusta, a beneficarius serving Julius Servianus, governor of Upper Germany, was sent as a messenger to Rome to inform Trajan of Nerva's death in January 98 (Hadr. 2. 5–6).

56 Compare also the discussions of paradoxical language at 7. 5. 1, and 5. 12.3–15. 2.

57 As Inwood 1995a notes.

58 A similar move by Seneca is pointed out by Moussy 1966, 264, that at 6. 43. 2 fin. and 2. 17. 6 Seneca tries to elide gratias agere (to give thanks verbally) and gratiam referre (to return a favour), but that this does not accord with ordinary usage, or indeed with his own usage elsewhere.

59 As Wilson Greatbatch, the inventor of the cardiac pacemaker said, during a ceremony to mark a Lifetime Achievement Award, 'the true reward is not in the results but in the doing' (The Independent for 30 Sept. 2011, p. 18 of 'Viewspaper'). The results here are both the thing invented and the honour received for it.

1 Thomas 1918, 145.

2 Albertini 1923, 92. The idea goes back to Caligula, who described Seneca's writing as 'sand without lime' (harena sine calce, Suet. Calig. 38).

3 Sandbach 1975, 157.

4 Bellincioni 1984a, 122. She finds the unity of the work in its central theme of revising the utilitarian conception of beneficium.

5 Sonntag 1913.

6 Ibid. 38, 60–3. A notable example for him (19) is the placing of 2. 26–30 in Book 2, rather than in 3, although its subject-matter is ingratitude, the subject of Book 3. For another explanation, i.e. that Seneca intended the passage to function as a bridge from one book to another, see pp. 116–18 below.

7 Abel 1987a = Abel 1995, 42–73. His analysis is accepted by Maurach 1996, 101–10.

8 Abel 1967, where he treated five of the twelve dialogi listed in the Codex Ambrosianus. He discussed the structure of the rest of the twelve dialogi, plus De beneficiis and the Quaestiones naturales, in Abel 1991 = Abel 1995, 166–87.

9 Introd. n. 1. The label may go back to Seneca himself. It presumably reflects the importance in all these works of interchange, sometimes with a generalized interlocutor, sometimes with the addressee (Griffin 1976, 412–15).

10 Abel 1987a, 19–23 = 1995, 59–65.

11 As Erasmus also complained, Trillitzsch 1971, ii. 435.

12 See S. Braund 2009, 45–7.

13 Recently, attempts have been made to defend the overall coherence of De ira. Fillion-Lahille 1984, 283–94 stressed Seneca's concern with reviewing philosophical opinions in chronological order; Nussbaum 1994, 405–9 countered this with the idea that the structure of Seneca's therapeutic argument, addressed to a non-Stoic interlocutor, is dictated by the latter's resistance; and Ramondetti 1996, 9–10, 81–3 explained the apparent repetitions as re-elaborations of specific themes, rather as Abel 1987a stressed the development of themes within De beneficiis. The latter is an approach developed later in this chapter and in Ch. 7.

14 Fuhrmann 1997, 236–7.

15 For this analysis of the structure, see Griffin 1976, 306–9.

16 'In lingering further, I am not serving my subject, but indulging it, since it has to be followed where it leads, not where it entices me' (Quidquid ultra moror, non servio materiae, sed indulgeo, quae quo ducit sequenda est, non quo invitat, 5. 1. 1); 'Let us go on, now that we have completed what belonged to the subject, to study also what is, to tell the truth, connected to it, not integral to it' (perseveremus peractis quae rem continebant, scrutari etiam ea quae, si vis verum, conexa sunt, non cohaerentia', 5. 1. 2); 'This book collects the remnants; with the subject exhausted, I am looking around to see, not what I shall say, but what I have not said' (reliqua hic liber cogit, et exhausta materia circumspicio, non quid dicam, sed quid non dixerim, 7. 1. 1); 'Now that I am rounding up anything that has escaped me' (nunc, si quid effugit, recolligo, 7. 1. 2). No connection with the preceding books is claimed at the start of Book 6, which proclaims that some of the topics that will be treated 'are investigated only to exercise the intellect and lie invariably outside the scope of real life' (exercendi tantum ingenii causa quaeruntur et semper extra vitam iacent, 6. 1. 1).

17 'Non-superfluous rather than necessary' (magis non supervacuum quam necessarium, 5. 1. 1); 'Whoever inspects these things carefully is not repaid for his efforts, but does not waste them either' (quae quisquis diligenter inspicit, non facit operae pretium nec tamen perdit operam, 5. 1. 2); 'Other matters are both enjoyable while under scrutiny and, once investigated, useful' (quaedam et, dum quaeruntur oblectamento sunt et quaesita usui, 6. 1.); 'Instead, I piled up at the start all the most important themes' (sed quidquid maxime necessarium erat, in primum congessi, 7. 1. 2).

18 Sonntag 1913, 15 and Abel 1987a, 25 = 1995, 67 think that the first topic after the definition is 'to whom we should give' (quibus demus). See also ad 1. 9. 2.

19 The four last chapters (2. 14–17) could be taken as reverting to the first topic, and indeed filling out the first division of the first topic by discussing what not to give. But the distinction of 'what' and 'how' is not always easy to make, and one can consider these chapters as treating how to give, with regard to the true interests of both benefactor and beneficiary, and how to make sure that gifts are appropriate to both (see the Synopsis of Book 2).

20 Seneca uses beneficium reddere, beneficium referre, gratiam referre, and less commonly gratiam reddere (but see his reservations about the last in Ep. 81. 9) as equivalents (Moussy 1966, 267–9). Shortly before the beginning of Book 3, in 2. 35. 1, the Stoic paradox itself, which had been given at 2. 31 as eum qui libenter accipit beneficium reddidisse, is rendered rettulisse illum gratiam dicamus, qui beneficium libenter accipit. Despite Ep. 81. 9, Seneca does not always eschew gratiam reddere. We find it at Ben. 3. 2. 2 and 5. 16. 4, and indeed at Ep. 81. 7 (in the mouth of an objector) iustitiae convenit suum cuique reddere, beneficio gratiam, iniuriae talionem aut certe malam gratiam ('It is in keeping with justice to render to each what is his due, a favour for a benefit, requital for an injury or, in any case, ill-will'). That the aversion was, however, common is suggested by Cicero's rendering of the familiar epigram, pecuniam qui habeat, non reddidisse, qui reddiderit, non habere, gratiam autem et qui rettulerit habere et qui habeat rettulisse ('that the person who has his money has not repaid it, and the person who has repaid it does not have it, but the person who has requited gratitude still has it, and if he has it, he has requited it'), where he uses reddere with pecuniam, but referre with gratiam, not only here in De officiis 2. 69, but in speeches: Red. pop. 23 (if that is the right reading) and Planc. 68.

21 'Now that we have completed what belonged to the subject' (peractis quae rem continebant, 5. 1. 2); cf. 'for these define the limits of this responsibility' (hi enim sunt huius officii fines, 5. 1. 1).

22 Chaumartin 1985, 13.

23 Fuhrmann 1997, 285–6.

24 Sonntag 1913, 21–3, in trying to reconstruct how Hecato, whom he regards as Seneca's single source, would have dealt with beneficium reddere, points to a practical difficulty: it would be difficult to use the same divisions of the topic as for giving and receiving, because what we should give in return, and how we should give it, can be inferred from the discussion of beneficium dare, while there is no point in discussing to whom we should return a benefit, as it is obviously our benefactor. On the other hand, Sonntag thinks Hecato treated difficulties concerning the right opportunity, or a change in character of the benefactor elsewhere.

25 Abel 1987a, 18, 22, 25 = 1995, 58, 63, 66–7. Book 5 itself is much concerned with the question of returning benefits.

26 Abel 1987a, 31 = 1995, 73.

27 Abel 1987a, 22 = 1995, 63. For the importance of memory in the treatise, see Armisen-Marchetti 2004.

28 Inwood 1995a, 250 = 2005, 76.

29 See ad 1. 1–3. 1. Sonntag 1913, 11, 59 thought that 1. 1. 9–13 was echoed by 7. 26–32; Abel 1987a, 18 = 1995, 59 gives the parallel as with 1. 1–4, which I regard as preferable. Abel 1991, 186 finds similar symmetry in De brevitate vitae and the Consolatio ad Helviam matrem.

30 The theme is also treated in 1. 1. 9 and 7. 31. 2–4, apart from the incidental mention at 3. 15.

31 Even Albertini, so hostile to the idea of conscious design in Seneca, recognized (1923, 93–4) that 4. 1–25 is important and raises a fundamental issue of morality: the conflict of idealism and utilitarianism.

32 6. 20–24 returns to the question of these benefits, exploring divine motivation.

33 Chaumartin 1985, 12; Abel 1987a, 17 = 1995, 58.

34 Abel 1987a, 23–4 = 1995, 65 regards all four books, 4–7, as being concerned at their close with the theme of repaying a benefit (beneficium reddere), the excessive sense of obligation by the beneficiary being damped down at the end of 4 and of 6; the insufficient sense of obligation being stimulated at the end of 5 and and of 7. But he does not note that the perspective alternates between recipient and donor.

35 Albertini 1923, 93–4 notes that much of Book 3 is casuistical, as well as the last chapters of Book 4; Chaumartin 1985, 12; Fuhrmann 1997, 289.

36 Jonsen and Toulmin 1988, 8–13. They find the inspiration of casuistry in ancient philosophy and rhetoric, notably the writings of Aristotle and Cicero; in the judicial practice of Roman law; and in the traditions of Jewish rabbinical debate (47–87).

37 Sonntag 1913, 60–1, 63.

38 For the perspective of De officiis as practical ethics, and the omission of any treatment of the nature of ethics or of the first principles of morality, see Griffin and Atkins 1991a, pp. xxii f.

39 Ross 1925 calls 8. 13 (1162a34–1163a23) 'Casuistry of Friendship'. Aristotle's word, to judge from 9. 2. 1164b22, is aporia.

40 Off. 3. 7 tertio, si id quod speciem haberet honesti pugnaret cum eo quod utile videretur, quomodo ea discerni oporteret; 3. 29–33 on Phalaris.

41 Pohlenz 1935, 107, apparently regarded as plausible by Dyck 1996, 612.

42 See the discussion of the Letters in Ch. 7C and in Griffin 2007, 91–5. Cooper and Procopé 1995, 127 suggest that De clementia 'could well have gone into the casuistry of mercy' at the end of Book 2, which, as we have it, is incomplete.

43 Mazzoli 2007, 589, 592–3 treats some examples of apparent repetitions in Book 5 of the same themes in the early books.

44 Ibid. 589.

45 See Armisen-Marchetti 2004, 16.

46 'So that the donors forget, and the debtors retain a persistent memory' (ut, qui praestiterunt obliviscantur, pertinax sit memoria debentium, 1. 4. 5); 'The law that governs benefits between two people is this: one of them should immediately forget that the benefit was given; the other should never forget that it was received' (Haec enim beneficii inter duos lex est: alter statim oblivisci debet dati, alter accepti numquam, 2. 10. 4).

* An earlier version of the material in this chapter appeared as ch. 3 in the first volume of R. Sorabji and R.W. Sharples (eds.), Greek and Roman Philosophy 100 BC–200 AD (BICS Supplement 94; London, 2007).

1 Cicero at Fin. 2. 43; 3. 12; 3. 50 helps us to make sense of Aristo's position on praecepta, for he tells us that he recognized no distinctions within adiaphora or indifferentia and regarded the katorthōma or perfectum officium, as done by the Sage, as the only officium: praecepta help us to make the right choices among the indifferents, and hence to perform kathēkonta or officia. The Sage, of course, needs no guidance in behaving virtuously. Elsewhere, in Ep. 89. 13, Seneca tells us that Aristo also regarded the parts of philosophy other than ethics, that is, the pars rationalis and the pars naturalis, as unnecessary. See also D.L. 7. 160 and Sextus Empiricus, Adv. Math. 7. 12, who says that Aristo rejected the preceptive and hypothetical subject matter (parainetikos topos and hypothetikos topos) of ethics.

2 Schafer 2009, 25–32, 80 suggests that Seneca, by choosing Aristo as his opponent in Letter 94 and supporting Cleanthes against him, is stressing that his own broad-minded practice is standard within the Stoic school. D.L. 7. 163 tells us that Aristo wrote four books of letters to Cleanthes.

3 Nussbaum 1994, 338 points out that Seneca moves from the concrete to the general and back again. Inwood 1999, esp. 114–16, notes various uses of precepts at different stages. Sellars 2003, 77 concentrates on the role of precepts as a form of training after the doctrines have been learned, but his focus is on askēsis, which he admits is narrower than the notion of precepts (76). Schafer 2009 largely ignores the practical function Seneca attributes to praecepta after we have accepted decreta and says (56) that Seneca, in giving praecepta a role, has in mind only 'the moral beginner' and (63) that they are 'a ladder to be removed once they have been climbed', but at 57 n. 33 they are 'an educative and protreptic tool for beginners and backsliders' and at 94 'the rational component of the way praecepta work grows as the agent's rationality grows'. He has to admit (58) that at Ep. 94. 50 Seneca is clearly thinking of non-Sages, rather than of moral beginners.

4 Giusta 1964–7, i. 151–3 regards Eudorus as the original source of the doxographic passages systematizing the teaching of various schools, and of the general schema for moral philosophy found in Seneca and Clement of Alexandria (Strom. 6. 8. 69). He reconstructs Eudorus' system at pp. 155–60. His attempt to date Eudorus before Cicero's philosophical works is contested: see Chaumartin 1985, 26 n. 13. I agree with Dobbin 1998, 94, that Seneca's division of ethical topics is not the same as that in Epictetus 1. 4. 11 or 3. 2. 1–5, despite L–S 1987–8, i. 346, who, however, were right to regard Epictetus' also as a division of ethics rather than of philosophy, the latter being the view of P. Hadot 1992, 98–115; P. Hadot 1995, 192–4, now followed by Long 2002, 117–18, 126.

5 The identification of Posidonius' Greek terms behind Seneca's Latin rendering of them is very uncertain, given the fluidity of such Greek terms: see Kidd 1988, 647–51 ad F 176.

6 In Ep. 95. 65 Seneca also mentions aetiologia which he renders as causarum inquisitio and ethologia (called by some characterismos) or the use of exempla. Giusta 1964–7, i. 165–8 equates the former with Eudorus' 'assigning of the causes that bring about certain states of movements' (ὁ δὲ τῶν αἰτιῶν ἀποδοτικὸς τῶν ἐπιτελούντων τινὰς σχέσεις ἢ κινήαεις‎) and takes it to refer to the practical consequences of virtue: that is, states and movements caused by virtue (or by vice).

7 It is generally agreed that it was the through the work of the philosopher Arius Didymus that Eudorus' division reached Stobaeus, though scholars differ on which works of his were involved and whether the 'Epitome of Didymus', cited by Stobaeus at 4. 39. 28, is a work by Didymus or a work summarizing him. For a discussion of the question, see Hahm 1990. Eudorus' division might have reached Seneca through Arius Didymus (Dillon 1977, 121 n. 1; Brittain 2001, 277 n. 45, 279 n. 47). If he is identical with Augustus' friend (Suet. Aug. 89), Seneca knew at least one work by him (Cons. Marc. 4. 2–5. 6).

8 Inwood 1999, 110. He adduces Ben. 4. 9–11 as an explicit statement of the connection between a general principle and the procedures of choice.

9 e.g. Ep. 94. 33: 'Duties are set in order by precepts' (officia praeceptis disponuntur); Ep. 95. 45: 'Marcus Brutus in that book which he entitled Concerning Proper Functions gives many precepts to parents, children, and brothers' (M. Brutus in eo libro, quem περὶ καθήκοντος‎ inscripsit, dat multa praecepta et parentibus et liberis et fratribus), cf. Priscian, Inst. gramm. 6.7.

10 Mazzoli 2007, 590 remarks with some justice that De beneficiis is a conceptual laboratory where Seneca applies on a vast scale what he will theorize about in Epp. 94 and 95. But his idea that Books 1–4 teach dogmatically by decreta, whereas 5–7 more realistically use techniques of the pars praeceptiva, seems to reverse what is suggested by the linguistic forms that predominate in Books 1–3.

11 Ben. 1. 15. 2: 'So, if anyone thinks that we, in laying down these precepts … he has misunderstood our advice' (quare si quis existimat nos, cum ista praecipimus … ne perperam monitiones nostras exaudivit); 1. 12. 3: 'No one is so foolish that he needs to be warned that …' (Nemo tam stultus est ut monendus sit …)—an example of praeteritio. In Epp. 94 and 95, monitio (94. 12, 21, 24, 39, 55; 95. 63) and admonitio (94. 25, 29, 31, 32, 33, 36, 44, 45; 95. 61) are used to characterize the work of praecepta.

12 Imperative: 'walk in this way, eat in this way' (sic incede, sic cena, Ep. 94. 8), 'be good' (bonus esto, 95. 50); gerund/gerundive: 'one should never lose one's temper' (irascendum non esse, 94. 9), 'how we should make use of things' (quomodo rebus sit utendum, 95. 54), 'towards these things alone our study should be directed (in haec sola studium conferendum, 95. 60); jussive subjunctive: 'you must expect (exspectes, 94. 43), 'let us forbid' (prohibeamus, 95. 47); future: 'live thus with your father' (sic vives cum patre, 94. 5), 'avoid this, do that' (hoc vitabis, hoc facies, 94. 50), 'do that, if you would have self-control' (illa facies, si voles temperans esse, 95. 66). It is clear, however, from Ep. 94 that certain maxims in the indicative (in prose or in verse) perform the work of admonitio and count as praecepta (Ep. 94. 27–30, 43). Morgan 2007, 26–7, 84 n. 2, 261 concentrates on those praecepta and sententiae mentioned in Ep. 94 that are proverbs and gnomic quotations (note the reference to Dicta Catonis in 94. 27) and discusses the combination of commands and statements used in the teaching of popular morality (290–2).

13 'Moreover, those who reject doctrines do not realise that doctrines are established in the very act of rejection. What do they say? That life is sufficiently developed by precepts, that the doctrines of wisdom [that is dogmas] are superfluous. But the very thing they say is as much a doctrine, for heaven's sake, as if I were now to say that we must give up precepts as superfluous, must use doctrines, must pay attention to them alone; it would be a precept I uttered in the very act of denying that precepts were to be taken seriously' (Praeterea non intellegunt hi qui decreta tollunt, eo ipso confirmari illa, quo tolluntur. Quid enim dicunt? Praeceptis vitam satis explicari, supervacua esse decreta sapientiae, [id est dogmata]. Atqui hoc ipsum, quod dicunt, decretum est tam me hercules quam si nunc ego dicerem recedendum a praeceptis velut supervacuis, utendum esse decretis, in haec sola studium conferendum; hoc ipso, quo negarem curanda esse praecepta, praeciperem).

14 Ben. 1. 9. 1: 'not the magnitude of each gift, but the quality of the giver, should be considered' (non quanta quaeque sint, sed a quali profecta, prospiciendum); 2. 4. 2: 'benefits should be bestowed' (repraesentanda sunt beneficia); 2. 9. 1: 'some benefits should be given openly, some in secret' (quaedam beneficia palam danda, quaedam secreto); 2. 11. 2: 'we should not mention … we should not press the point, we should not revive the memory' (non est dicendum … non est instandum, non est memoria renovanda).

15 Jussive subjunctives at 2. 1. 1, 'let us give in the way we would wish to receive them' (sic demus quomodo vellemus accipere); 2. 4. 3, 'let no one intervene, let no one retard them' (nemo illa intercipiat, nemo detineat); 2. 11. 6, 'let us spare their ears' (parcamus auribus). Futures: 1. 11. 6, 'we shall be careful not to send gifts that are superfluous' (cavebimus ne munera supervacua mittamus); 2. 14. 1, 'so we shall consider the interests rather than the wishes of those making the request' (aestimabimus itaque utilitatem potius quam voluntatem petentium).

16 Ben. 2. 10. 4: 'but if he should be helped' (sin adiuvari illum oportet); 2. 14. 2: 'it is appropriate to keep in view not merely the first effects of one's benefits but also the eventual outcome' (initia beneficiorum suorum spectare tum etiam exitus decet).

17 Ben. 2. 6. 2, 'choose a different time' (aliud tempus eligito); 2. 15. 3, 'therefore compare the role of each and assess in that context the very gift you are going to give' (utriusque itaque personam confer et ipsum inter illas quod donabis examina).

18 Epp. 94. 39; 95. 34 and 65. Cancik 1967, 16–25 showed in her study of the Letters to Lucilius how the theoretical and doxographical style of teaching is characterized by descriptive grammatical forms, while the paraenetic passages use prescriptive ones. She points out that, in addition to sentences that have a prescriptive form, Seneca also uses for paraenetic purposes exempla, instances, and comparisons (23 n. 42). Codoñer 2000 rightly emphasizes that paraenetic teaching is not confined to certain grammatical forms, and that theoretical and paraenetic teaching are combined in single letters.

19 Philo of Larissa, in his division of the ethical part of philosophy, justifies the hypothetical statement (hypothetikos logos) on the grounds that those who are not Sages need preceptive statements (parainetikoi logoi) to help them act correctly in particular situations (Stob. 2. 41. 18–25 W). Clearly, like Seneca's Letters 94 and 95, this reflects controversy over the usefulness of precepts (cf. also Musonius Rufus at Stob. 4. 50. 94). It also suggests that it is wrong to think praecepta are only for the moral beginner (see above, n. 3).

20 Seneca switches from the level of 'correctly performed functions' (katorthōmata) to that of 'proper functions' (kathēkonta). At 1. 5. 3 he defines a beneficium as a recte factum in which the intention is everything, and Book 4 is much concerned with the sapiens. But since a katorthōma is, in fact, a kathēkon that has all its numbers, as done by the Sage (Ben. 7. 17. 1), Seneca naturally concentrates on the deliberations that make an action appropriate, providing it with a 'plausible defence' (eulogos apologismos, D.L. 7. 107), and giving it the quality of rationality required of virtue (Ben. 4. 9. 3), though, strictly speaking, the token actions of this type done by anyone but a Sage are 'errors' (hamartēmata). Similarly, Eudorus in the practical division of ethics is concerned both with kathēkonta and katorthōmata, including those concerned with charis and other activities, like marriage, that can be practised by the Sage and by others: there is no reason to think he treated the two levels of action separately, as Giusta 1964–7, ii. 455 suggests.

21 Chaumartin 1985, 292–3. At Ep. 27. 9 Seneca quotes with approval Epicurus' repetition of a valuable piece of advice: 'For some people, remedies need to be pointed out; for others, ground into them.'

22 Veyne 1990, 351–2 (translation modified).

23 Cancik 1967.

24 Maurach 1970, 199–206. See also André 1970, 201–2, pointing to the way in which the evolution of Lucilius matches the spiritual maturation of the pedagogue.

25 Maurach 1996, 109 n. 115, arguing against Chaumartin 1985, 297–302.

26 Off. 1. 7: 'The duties for which precepts are laid down are indeed relevant to the end of good things, though that is less obvious because they seem to apply rather to the instruction of ordinary life' (Quorum autem officiorum praecepta traduntur, ea quamquam pertinent ad finem bonorum, tamen minus id apparet, quia magis ad institutionem vitae communis spectare videntur). The vita communis is, as is made clear at 3. 14–15, that aspect of human activity which is shared between Sages and ordinary men. Precepts teach the type of conduct of which a 'reasonable defence' can be given, the officia media or communia, as Cicero calls them at 3.14. See Inwood 1999, 102.

27 It is defined at 1. 6. 1 as 'an act of benevolence bestowing joy and deriving joy from the action, performed from inclination and the spontaneous readiness to do so' (benevola actio tribuens gaudium capiensque tribuendo in id quod facit prona et sponte sua parata).

28 Cancik 1967, 23 n. 41.

29 supervacua: Ben. 5. 1. 1, cf. Ep. 45. 4; 49. 5. Uselessness decried in Ep. 45. 5; 48. 6–7; 113. 26; described as positively harmful in Epp. 48. 9–10; 49. 5–6; 82. 19.

30 Enjoyment: Ben. 5. 1. 1; 5. 12. 2; 6. 1. 1, cf. Ep. 111. 5; games: Ben. 6. 1. 1, cf. Epp. 71. 6; 102. 20; 124. 21; refreshment to the soul: Ben. 5. 12. 2, cf. Ep. 65. 15. Fuhrmann 1997, 290 suggests: 'Wahrscheinlich gibt diese heitere ethische, der Lebenspraxis nicht eben nahestehende Rabulistik das Muster wieder, nach dem man in den Philosophenschulen Geschicklichkeit im Argumentieren zu erwerben suchte.'

31 Trillitzsch 1971, ii. 436. Erasmus writes, ' "an sapiens sapienti possit prodesse"; "an sapienti dari possit beneficium a diis?" … Huiusmodi naenias quum frequenter inculcet usque ad tedium, tamen subinde damnat. Quid enim opus est facere quippiam, in hoc ipsum ut factum reprehendas?' The two examples resemble Ben. 7. 4. 1 'quemadmodum potest aliquis donare sapienti, si omnia sapientis sunt?' and 5. 4. 3 'quibus nihil potest praestare ipsa fortuna'. The first is treated in Ep. 109, which looks (17) to its appearance in the lost libri moralis philosophiae.

32 Barnes 1997, 14.

33 'At that time he was almost the only author in the hands of the young' (tum autem solus hic fere in manibus adulescentium fuit, 10. 1. 126); 'if he had not been in love with his own ideas … he would have won the general approval of the learned rather than the adoration of boys' (si non omnia sua amasset … consensu potius eruditorum quam puerorum amore comprobaretur, 10. 1. 130).

34 Barnes 1997, 8–10 points out that dialektikē included the study of sound and voice (which we would regard as a matter for psychology) and the study of sense impressions (which we would regard as epistemology).

35 Ep. 102. 4. The examples concern the value of posthumous glory. Seneca gives examples of three questions concerned with mores (4), followed by three purely dialectical questions.

36 Barnes 1997, 13–23 discusses Seneca's attitude to logic and logical puzzles. Cicero too recognized the value of dialectic to the orator and the jurist (Reinhardt 2003, 306–7).

37 Another two paradoxes—all things belong to the Sage; friends have all things in common—give rise to questions at 7. 4–13.

38 Topica 1. 11. 104b1–105a9.

39 Topica 2. 1. 108b34–5. On Cicero's claim in Or. 46 that Aristotle used the thesis as a rhetorical device, see Brittain 2001, 335–6.

40 Quintilian 3. 5. 5 tells us that, instead of Cicero's term proposita (Top. 79), others used quaestiones universales civiles or quaestiones philosopho convenientis. Cicero also uses consultatio (De orat. 3. 111; Part. Or. 4, 67). See Bonner 1969, 2.

41 Cic. Inv. 1. 8. The principal discussions are in Cicero, De oratore 2. 78; 3. 111–18; Or. 44–6; Topica 79–86; Part. Or. 61–7, and Quintilian 3. 5. 5–18, but both authors discuss them elsewhere. All this is well analysed by Riposati 1947, 162–203.

42 See the excellent discussion by Clarke 1951, who points out that Seneca's father wrongly regarded the thesis as confined to the period before Cicero. See also Bonner 1949, 10 and Fairweather 1981, 104–6, 117–19.

43 De inventione 1. 8. Quintilian discusses Cicero's change of heart at 3. 5. 14–15.

44 Brittain 2001, 339–43; Reinhardt 2000, 535–40; Reinhardt 2003, 14–17; 346–7. For the problem of relating Cicero's change of heart to Philo's teaching, see Reinhardt 2000, 535; 547 n. 57.

45 De orat. 3. 107; Or. 46 (Aristotle himself).

46 Part. Or. 139, cf. 61–7.

47 On Quintilian's views, see Viano 1995.

48 Parad. pref. 4–5.

49 Cicero points out that theseis too, like hypotheseis, can be determinate with regard to one or more of the specifications (peristaseis), but not in the most important respects (Top. 80, on which see Reinhardt 2003, 351–2).

50 Quint.3.5.8–13.

51 The sort of case envisaged by Aristotle at Top. 1. 11. 104b31–5.

52 According to Quintilian 2. 4. 25, theseis like the ones mentioned above, with the addition of named persons, become the suasoriae beloved of declaimers. Seneca's Ep. 14. 12–13 gives a good example of the thesis 'whether one ought to enter politics' (an accedendum ad rem publicam), adapted to the situation of Cato at the time of the civil war. As Bonner 1969, 8–9 points out, the series of rhetorical questions and arguments addressed to Cato in this letter suggests that this is an example of a suasoria formed by Seneca as a concrete case to illustrate and give practice in handling the general question.

53 Sen. Controv. 2. 5 and 9. 1. See Bonner 1969, 6–11; 87. Sen. Controv. 1. pref. 12 identifies controversiae with what Cicero called causae.

54 Reinhardt 2000, 540–1; Brittain 2001, 289–90.

55 These subdivisions are staseis in Greek and status in Latin; their use in classifying lawsuits or political debates is associated with Hermagoras, Philo being credited with extending their application to the theseis (see references in n. 44).

56 See Brittain, 2001 309, n. 25 and Reinhardt 2003, 348 n. 3 for the criticism of a certain Archedemos as reported by Quintilian 3. 6. 31, 33: he thought the third sub-division, qualitas, was already covered by the second, definitio.

57 The question turns on whether everything a slave does for his master is a ministerium (something he cannot choose not to do).

58 The negative answer involves exploring the essentially social character of benefits (5. 8–11).

59 The solution turns on whether doing good to a person against his will counts as a benefit.

60 The solution turns on whether a benefit is the act of giving or the thing given.

61 Quintilian divides the quaestiones infinitae relating to actio into two divisions, 'how to acquire; how to use' (quomodo adipiscamur; quomodo utamur). Cicero in Part. or. 63 gives (1) 'to pursue and to avoid something' (ad persequendum aliquid aut declinandum) and (2) 'what relates to some convenience or use' (quod ad aliquam commoditatem usumque referatur). Both divisions clearly belong to the officium half of actio (De orat. 3. 118).

62 The quale sit subdivision of cognitio is the most obvious to be involved in such ambiguity. Theon in his Progymnasmata (p. 84 Patillon, 16–19 = ii. 121 Spengel, 14–17), notes 'It makes no difference whether someone says "Should one marry or not?" or again "Is marriage something to be sought or avoided", for it is one and the same thing that is indicated by all such questions'. Riposati 1947, 198 marks the difficulty particularly with reference to this subdivision, pointing out that e.g. suscipiendine liberi could belong with de expetendo fugiendoque under quale sit.

63 Et haec quaestio facile expedietur et si qua similis huic moveri potest, si totiens illo cogitationem nostram converterimus beneficium nullum esse, nisi quod ad nos primum aliqua cogitatio defert, deinde amica et benigna.

64 Giusta 1964–7, ii. 451 suggests that the questions in the later part of Book 4 introduced peristaseis ('complicating circumstances'), which lead us to qualify the positive answers to the two questions posed at the start of the book. However, his analysis of the quaestiones in the later books, purely in terms of Aristotle's Categories (453), ignores the substantial development of Aristotle's classifications by the Peripatos, the Stoa, and the rhetoricians, in the interval between Aristotle and Cicero.

65 See Inwood 1999, 109 on praecepta as providing instructions for the kinds of actions that are normally appropriate and why. 'The distinction between normally and exceptionally appropriate actions operates at the level of types; it is a distinct point about token actions that the concrete particulars of each situation, including the character of the agent and the place of the action in his life as a whole, determine the final moral evaluation of that particular action.'

66 Inwood 1999, 120 n. 74, 124 points out how slippery are the terms formula, praecepta, lex, regula. His own suggestion (126) that Stoic theory used general rules of thumb as a reference point in moral reasoning, enabling the moral reasoner to find the balance between abstract theory and the demands of a particular context, is helpful. He notes (120), as have others (e.g. Gomoll 1933, 48–52), the use by Cicero of the resources of Roman legal reasoning in his formula in De officiis 3. 19–20. But regula too has such credentials, as many jurists wrote books of Regulae, which seem to have been handbooks giving short statements of the basic principles of Roman law, without going into all the subtleties and qualifications.

67 See Schafer 2009, 106, pointing out that the distinction between praecepta and decreta in Letters 94 and 95 belongs to Seneca's opponent, whereas Seneca is concerned to bring them together in his notion of philosophical training.

68 It is worth noting that, in the functional ordering of ethical teaching, devised by the Academic Philo of Larissa in the 90s bc and paraphrased by Stobaeus (Ecl. 2. 7. 39–41 W), the third division concerned with safeguarding happiness puts in parallel the parangelmata of the doctor safeguarding health and the theōrēmata peri biōn, of which the examples given are theseis: should the sensible man participate in politics or consort with members of the ruling class; should the sage marry? In an appendix, it is said that men who are not Wise but 'in an intermediate condition' may need the hypothetikos logos, so that they will have in handbooks maxims for handling each matter securely and correctly. Thus bare precepts are pedagogically a step below the theseis and are said to be derivable from them by the Wise. This third topos then is operating at the level of kathēkonta in Eudorus' scheme (Brittain 2001, 288), but clearly at a more advanced and demanding level than particular praecepta, 'teaching the individual to cope with problematic situations by analysing them and generalizing his own situation' (Reinhardt 2000, 540). For the Stoics, of course, only the non-Wise would need either.

69 In Ben. 7. 2. 1 Demetrius wants the proficiens to rehearse the necessary tenets, until they come to him spontaneously, and he can distinguish right and wrong without hesitation. At the end of his prescribed meditation, Seneca speaks of the pleasure experienced by the Sage. The slide from one to the other is clearly supposed to suggest the metamorphosis.

70 Sonntag 1913, 63.

71 Griffin 1976, 319 n. 5. Though there would be nothing unusual in Seneca's addressing Liberalis by his cognomen alone (as with Serenus), Seneca does use his nomen as well. The fact that he delays doing this until Book 3 suggests that it is the cognomen that he wants to emphasize. See Dickey 2002, 50, 53, 68–70.

72 Mazzoli 2000, 251 marks this as a characteristic of the twelve dialogi in the Codex Ambrosianus, but it is true of De clementia and De beneficiis as well (see below, n. 75). The Naturales Quaestiones forms a partial exception, though Lucilius is addressed in the prefaces of those books that have them (1, 3, 4A), at the beginning and end of Book 6, and at the end of Books 2 and 5. Book 4B is incomplete at the start, so only Book 7 lacks any invocation of the addressee. But the incompleteness of parts of the work and the confused order of the books make it difficult to be clear about the pattern.

73 Seneca's 'we', I', and 'you' are often ambiguous as between the personal and impersonal. Here they appear to be specifically himself and Liberalis, whereas in 7. 28. 2 (mentioning senatorial office) and 30. 1 (stulte) we are clearly in the territory of the impersonal 'you'.

74 Mazzoli 2000.

75 On the likelihood that the category of Senecan works that Quintilian termed dialogi includes these longer works, Naturales quaestiones, De clementia, and De beneficiis, see pp. 1, n. 1; 112, n. 9; Griffin 1976, App. B2. See also n. 72 for another similarity.

76 Not incompatible with this analysis is the idea, adduced as an explanation of the contrast in Griffin 1976, 153–4, that the two books belong to different genres.

77 Ibid. 316–17; 353–5.

78 Ibid. 351–3. At 417 the wavering of the image of Lucilius resulting from Seneca's inconsistent tone of voice in addressing him, now from above, now as a fellow-learner, is discussed. See on the complexities, the discussion by Edwards 1997. See also Griffin 2007, 90–5.

79 Foucault 1988, 49, 53. See also Nussbaum 1994, 337–8.

80 The order here is that described as standard in Cons. Marc. 2. 1.

81 Wilson 2001, 183–5 suggests that Letter 33 redefines retrospectively Books 1–3 as preliminary protreptic and marks a turning-point in the collection and in the 'epistolary narrative' of moral and intellectual progress.

82 Leeman 1953. Lana 1991, 285–6, 289 defends this view, pointing out that Seneca begins to label his theoretical problems quaestiunculae after his mention of the comprehensive work.

83 Maurach 1970, 205.

84 Schafer 2009, 75.

1 Only incidentally, in the treatise, does Seneca mention that the communal things are the only possessions the Sage would think worth having (Ben. 7. 7. 5).

2 Ben. 4. 8. 3 (with notes ad loc.). 'Fortune' tends to be used when the gifts are being viewed negatively, to remind us that they are only positive indifferents.

3 Pohlenz 1941, 104–7. However, his interpretation of Ben. 2. 26–30 on ingratitude, as an interruption of the plan of the work in order to accommodate this concern, is questionable: see pp. 115–18 for an explanation of this passage in terms of the difficulty of distinguishing receiving from returning, when beneficium is conceived as the act of giving, not the thing given.

4 As Pohlenz 1941, 73–4 notes, the difference is all the more striking because the context in De vita beata is one where things very similar to Seneca's teaching in the treatise are said (Vit. beat. 24. 1 on difficulty of giving properly, cf. Ben. 2. 18. 2; Vit. beat. 24. 3 on need for discrimination in giving, cf. Ben. 1. 1. 2; 2. 15; 4. 10. 2–3).

5 See references in n. 4 above.

6 See Griffin 1976, 302–10.

7 Pohlenz 1941, 74.

8 Bellincioni 1984a, 10, 104–5. See also Maurach 1991, 110–11 and Griffin 2003a, 178–82.

9 All these works are also concerned with the first division, inspectio, assigning the right value to things: in De beneficiis, as we saw, there is a lot of discussion of the distinction between goods and indifferents.

10 Cf. Ben. 1. 1. 9–11 for the sentiment, and 2. 11. 4–5 for the image.

11 Though, in the Letter, the perseverance depicted in the image is directed at finding other people who are grateful, whereas in the treatise Seneca extends his perseverance to the conversion of the ingrates themselves.

12 For the parallels, see Ch. 4 n. 3. My analysis of Letter 81 owes an incalculable debt to the two essays of Brad Inwood, Inwood 1995a, and Inwood 2004.

13 On the quaestiones in the treatise that serve as 'hard cases' used to clarify philosophical points, see pp. 120–2, 135–42.

14 Inwood 2004 discusses Seneca's frequent use of the model of a judge in treating the moral agent, including the passages in Letter 81 (pp. 212–18). For a sketch of the use of legal terminology by Cicero and Seneca, see Griffin 2012.

15 The verb here for undoing the benefit, rescindere, appears in 6. 2. 2, though the verb most often used in 6. 2–4 is eripere.

16 an is, qui profuit nobis, si postea nocuit, paria fecerit et nos debito solverit (Ep. 81. 3); in hac comparatione beneficii et iniuriae (Ep. 81. 15).

17 See ad 6. 4. 4 debitori suo creditor on the apparent contradiction with the legal cases adduced at 6. 4. 2 fin. and 6. 4. 5, where injury does seem to be weighed against a debt.

18 Cf. 6. 4. 2 non beneficium tollitur, sed beneficii gratia, et efficitur, non ne habeam, sed ne debeam.

19 The phrase used by Inwood 1995a, 249 n. 44 = 2005, 75 n. 43.

20 Ira 2. 32. 1–3; Clem. 1. 20. 2; 1. 7. 3 where ignoscite and videtur show that Seneca dissents from the idea that private citizens are right to feel they need to take revenge on wrongs to them.

21 It is hard to resist the temptation to see in this curiously negative ending (a much stronger version of what is said in De beneficiis 3. 1. 1 fin.) Seneca's anticipation of Nero's ingratitude, which would cause his own death: he actually died about a year after the dramatic date of Letter 81, and near in time to its publication (on which see Griffin 1976, 349, 418 n. 4). There is perhaps an echo of the passage in the sentence Tacitus gives the philosopher, 'Who did not know that Nero was cruel? Nothing was left, after he killed his mother and father, than that he should add the killing of his teacher and instructor' (educatoris praeceptorisque, Ann. 15. 62. 2): the last leaves a strong hint of ingratitude, given Ben. 3. 17. 4 (non educatoris non praeceptorum); 6. 15. 2; 6. 16–17.

22 Inwood appears to have changed his mind on this point between his earlier essay, where he thought none of them identical with the Sage 'who only comes into the discussion as sole possessor of relevant moral certainty first at 81. 8' (Inwood 1995a, 251 n. 54, cf. 252 nn. 55, 56 = 2005, 77 n. 53, 78 nn. 54, 55) and the later one, where he identifies the vir bonus (and apparently the rigidus iudex) with the Sage (Inwood 2004, 86 = 2005, 216).

23 Vir bonus, which has a legal resonance from formulae containing clauses related to bona fides and from the provision in some contracts for arbitration by a vir bonus, is used by Cicero to denote 'good men' in common parlance, as opposed to Sages, and particularly men known for justice (Off. 1. 20; 3. 17; 1. 31; 2. 33; 2. 38). In Ep. 42. 1 Seneca points up the ambiguity of the phrase, as between the Sage (who occurs rarely if at all) and the vir bonus secundae notae, who is making progress. In Ep. 11. 8–10 we are told to adopt a vir bonus as a guardian: either one who is rigidus like Cato, or one who is remissioris animi like Laelius.

24 'The good man so makes his calculations of both (benefit and injury), that he himself limits himself' (utrosque calculos sic ponit, ut se ipse circumscribat) in §6, and 'he will overlook it, that is if the injury exclusively affects himself' (remittet; id est, si ad ipsum tota pertinebit iniuria) in §16.

25 I agree here with the interpretation of Bellincioni 1984b, 174 (= Bellincioni 1986, 113). Inwood 1995a, 251 n. 52 = 2005, 77 n. 51 thinks a new point is introduced with plus nocuit; sed prius profuit, and suggests that the ex iniuria superest is literally something that is still enjoyed after the injury. But Ben. 6. 6. 1 (if Seneca is making a similar point) is against this: ultra debeatur here cannot be a part of the benefit still enjoyed, because the benefit is no longer visible beneath the injury. Again in Ben. 6. 4. 6 we are told that the endurance of the benefit does not guarantee that it imposes an obligation, if the giver spoils it by his subsequent grudging attitude.

26 The phrase used for vir bonus by Inwood 1995a, 251 n. 54 = 2005, 77 n. 53.

27 I follow here the interpretation of Bellincioni 1984b, 173–5, as does Inwood 2004, 86 = 2005, 215.

28 Bellincioni 1984b, 181 (= Bellincioni 1986, 122) rightly regards as a rhetorical strategy this concession to the stern Stoic view of justice as practical wisdom in distributing what is deserved (Griffin 1976, 158); see above, n. 20.

29 On this passage, see Inwood 1995a, 252 n. 55 = 2005, 78 n. 54.

1 Trillitzsch 1971, i. 44–8, 53–4, 56–61. The historians Tacitus, at Ann. 13. 42, and later Cassius Dio, at 61. 10, do report Suillius Rufus' charges of hypocrisy.

2 Tac. Ann. 13. 3. 1: Seneca's funeral speech for Claudius showed 'an attractive talent and one suited to contemporary ears' (ingenium amoenum et temporis eius auribus accommodatum); Suet. Cal. 53. 2: Caligula's criticisms of 'Seneca, then very much in fashion' (Seneca tum maxime placentem).

3 Holford-Strevens 2003, 277 n. 75, however, suggests that Marcus Aurelius disobeyed Fronto and finds an echo of Ben. 5. 6. 2, 6 in Med. 11. 25. The praise of Seneca's generosity by Martial (Epig. 12. 36. 8–10) and Juvenal (5. 108–10) contains no hint of his literary celebration of generosity.

4 'What does it matter how much you have? There is far more that you don't have' (quid enim refert quantum habeas? multo illud plus est, quod non habes, NA 12. 2. 13), cf. 'Let him compare what he has with what he desires: he is still a poor man' (quidquid habet, ei, quod cupit, comparet: pauper est, Ben. 7. 10. 6), cited by Trillitzsch 1971, i. 74.

5 Brinkmann 2002, 27–31. The preponderance in De beneficiis of exempla showing the way emperors confer gifts was noted by Mayer 1991, 159, 162 (= 2008, 307, 310).

6 Woodman 2010.

7 As pointed out by Grimal 1967. On the last, see ad 7. 1. 7 si animum virtuti consecravit. Syme 1958a, i. 335 found distinctive Senecan expressions and imagery in the dialogue.

8 Momigliano 1955, 20–1.

9 Trillitzsch 1971, i. 190; Ker 2009, 197–8. The implicit combining of the Elder Seneca, author of the declamations, with his son was to be repeated by Boccaccio in the 14th c. and only began to be sorted out at the end of the 15th, by Paolo Pompilio (Panizza 1984, 67–8, 70).

10 Blüher1969, 27.

11 L. D. Reynolds 1983, 358.

12 See Ch. 2 n. 32. Martin's four virtues are prudentia, magnanimitas, continentia, and iustitia. It is worth noting that when Seneca lists the four cardinal virtues, he uses fortitudo, not magnanimitas, and that he usually prefers temperantia to continentia.

13 Busonero 2000. Gertz in his edition realized that N was the most authoritative evidence for the text of De beneficiis and De clementia, and he made it the basis of his edition of 1876. The Hosius text of 1914 that I generally follow, and on which the Loeb edition of 1935 by Basore and the translation by Cooper and Procopé 1995 are generally based, unfortunately rests on the idea that R is independent testimony, alongside N, to the lost archetype.

14 L. D. Reynolds 1983, 363–4.

15 Mazzoli 1978, 92–7; L. D. Reynolds 1983, 364–5.

16 Blüher1969, 32–41.

17 As suggested by Mazzoli 1978, 86.

18 Nothdurft 1963, 93–122.

19 Stacey 2007.

20 Blüher 1969, 48, 75–78, 203. He points out (77) that, in the 13th and 14th centuries, De beneficiis was mostly known at second hand, particularly from William of Conches, whose Moralium dogma philosophorum was the source used by the author of the Libro del Caballero Zifar, and by Juan García de Castrojeriz, author of the Glosa Castellana al Regimiento de Príncipes de Egidio Romano.

logo-footer Copyright © 2018. All rights reserved.
Access is brought to you by Log out