Bryn Mawr Classical Review 95.07.02

C.D.N. Costa, Seneca: Four Dialogues. Warminster: Aris & Phillips, 1994. Pp. 218. £14.95. ISBN 0-85668-560-7.

Reviewed by R. Ferri, University College, London.

Prof. Costa (henceforward C.) had previously edited, for the same series, a selection of Seneca's Letters. The present volume includes a reprint of Reynolds' OCT, although C., in his translation, occasionally disagrees with Reynolds. There is no formal apparatus criticus, but a few conjectures and uariae lectiones of relevant interest are listed at p. 172. Most difficult passages are discussed in the explanatory notes, which follow the text. C.'s translation is lively but accurate and on the whole will provide students wishing to understand the Latin with a very precise guide. The book is aimed at undergraduates, but, as serious, large-scale, not tralatitious commentaries devoted to most of Seneca's dialogues are still much needed, the information contained in the notes will also be of use to more advanced students of Latin.

As C. warns in his preface, his notes are rarely concerned with matters of language and style, except where they are necessary for a fuller understanding of the text. Much useful information is given on philosophical terminology of the Hellenistic schools and its equivalents in Seneca. Readers at all levels will find very helpful the synopsis of each dialogue, a matter which itself, up to a point, is subject to interpretation, as Seneca's argumentative structure in the Dialogues is not always clearly linear. Allusions to historical figures and anecdotes, of which Seneca's treatises are rather full, and which he often expresses in a rather cryptic mode, are aptly and fully explained. E.g. helu. matr. 10.7 (p. 150) 'qui illos inuocauerant ... redibant' (= Regulus).

Yet more perhaps should have been done, even at this level, to clarify the historical context: for example to illustrate the political import of many of Seneca's comments on engagement and detachment, or to situate many anti-tyrannical utterances in the context of Seneca's political career, and of aristocratic and senatorial ideology in the early Principate. The overall picture of Seneca that emerges from reading C.'s notes strikes me as being somewhat conventional, and unlikely to help make these Dialogues more appealing to contemporary undergraduates. Some linguistic problems could also have received more attention.

I shall add only a few remarks on points at which I felt that a student wishing to understand the Latin would need more support or some clearer explanation.1 As the Latin text as given by C. is not sub-divided into paragraphs, I refer to chapter and paragraph numbers as in Reynolds' text (giving 'in parenthesis the pages of C.'s edition).

uit. beat.

4.3 (p. 18): 'cui unum bonum sit honestas, unum malum turpitudo, cetera uilis turba rerum nec detrahens quicquam beatae uitae nec adiciens'; reasons of emphasis and word-order lead one to prefer taking 'cetera' as neutr. plural, i.e. 'cetera sunt uilis turba rerum, quae nec detrahit neque...'.

7.4 (p. 18): 'numquam enim recta mens uertitur nec sibi odio est nec quicquam mutauit optima'; C. is inclined to accept 'mutauit' as a 'gnomic' perfect (p. 176); Grimal interprets 'quicquam' as the subject of the supposed apophthegma. I would think it best to print 'mutat ab' (which follows up Castiglioni's 'mutauit ab', Athenaeum, 9 (1921), 206, with no discussion), an emendation which is palaeographically plausible [mutatab > mutatub > mutauit] and has the advantage of carrying on the sequence of present tenses: 'nec (sc. recta mens) quicquam mutat ab optima' (i.q. distat, differt). mutare is therefore used as an intransitive verb, quicquam is the accusative of the internal object; optima is almost reflexive (= ab optima mente; a sua optima condicione);2 'nor does the recta mens fall short of its perfection, i.e. its true state'. For a similar case, in which an adjective stands for a longer colon or an abstract term referring to the subject, cf. Ov. met. 6. 200 'quae (sc. Latona; quae MSS; qua Bentley) quantum distat ab orba?'; for this use of 'mutare' cf. Aul. Gell. noct. Att. 2. 23, 7 'quantum stupere atque frigere quantumque mutare a Menandro Caecilius uisus est'; Verg. Aen. 2. 274 'quantum mutatus ab illo'.

8.6 (p. 20): 'uirtutes enim ibi esse debebunt ubi consensus atque unitas erit: dissident uitia; C.: 'for virtues are bound to be wherever there are concord and unity, but vices are a symptom of discord'. In fact the Latin says the opposite: 'discord is a symptom of vice', or 'discord accompanies the vices' (Basore).

18.3 (p. 34): 'quod cum sibi interdixerit habere, interdixit et poscere, negant satis egere. uides enim: non uirtutis scientiam sed egestatis professus est'; 'poscere' = not 'to want', but 'to beg': the whole point of the insinuation lies in the fact that Demetrius appeared suspect as a real, sincere 'pauper' because he would not ask ('même de demander' Waltz). On 'uides enim', C. is probably right to see it as slightly adversative rather than as introducing an ironic comment. The thought probably needs to be expanded a bit more: 'and yet, you see, Demetrius, being a cynic, was quintessentially a teacher of poverty more than a teacher of virtue; indeed, he taught what the true science of poverty was like: to be contented with what one has from nature, and therefore not to beg for anything; but even so they did not believe that he was sincere'.

19.4-5 (p. 36): 'cum refigere se crucibus conentur -- in quas unusquisque uestrum clauos suos ipse adigit -- ad supplicium tamen acti stipitibus singulis pendent: hi qui in se ipsi animum aduertunt quot cupiditatibus tot crucibus distrahuntur'. I have two points.

1) 'in quas ... ipse adigit' is not, I believe, 'to which each one of you nails himself'. Rather, there must be an innuendo to Suillius' accusations ('clauos' = slanders and the like): 'the philistines are only too eager to put philosophers on the cross'.
2) 'hi qui ... distrahuntur'; C. interprets it as 'but those who bring on themselves their own punishment are stretched upon as many crosses as they have desires', which was also Basore's translation. 'animum aduertere' would have here the extended, and relatively uncommon, sense of 'to inflict a punishment', a semantic extension reached by the way of an official 'euphemism', 'the consul will see to it', meaning implicitly 'he will punish him'; as such, it is mostly found in statements concerning official acts, or in descriptions of actions taken by magistrates against criminals.

The additional problem is, it seems to me, that hypocrites do 'inflict a punishment upon themselves' only in the philosopher's opinion, not in reality; in other words, 'in se ipsi animum aduertunt' should be an emphatic conclusion, i.e. the philosopher's judgement about such behaviour,3 not a premise to the argument. In the sentence as it stands at present, the sequence is illogical; the relative pronoun 'qui' seems therefore inappropriate. Lipsius proposed the correction 'in ipsos' [i.e. in illos, 'the philosophers'] for 'in se ipsi', taking 'animum aduertere' in the sense of 'to be indignant with'. Another solution could be found by intervening on 'qui', and taking 'animum aduertere' to mean 'to pay attention to': 'hi, si in se ipsi animum aduertant, quot cupiditatibus tot crucibus distrahuntur'; the sense being that the sinner, if he knew what evils hide inside his breast, should be desperate. Cf. Tac., Ann. 6.6 '[neque frustra praestantissimus sapientiae firmare solitus est,] si recludantur tyrannorum mentes, posse adspici laniatus et ictus'; Boet. Cons. 4. m. 2, 4-5 'detrahat si quis superbis uani tegmina cultus iam uidebit intus artas dominos ferre catenas' (both passages ultimately going back to Plato, Rep. 579 E); interesting also Sen. Her. Oet. 648-50 'o si pateant pectora ditum! quantos intus sublimis agit fortuna metus'. The type si aduertant ... distrahuntur is well-known (cf. Schuster, Wien. Stud. 44 (1924-5), p. 122), but, on the whole, Lipsius' solution is preferable, as altering 'qui' would weaken 'hi' too much (Seneca would have used 'isti' as a deictic without a precise referent in the same sentence); 'in se ipsi', then, must have been substituted for 'in ipsos' by someone who had in mind the Christian (but also Senecan) examination of conscience. For the abrupt passage from second to third person ('unusquisque uestrum ... hi qui') cf. this review, infra, ad const. sap. 9.4.

27.5 (p. 50): for the sake of students' chronology, it would have been well to add a remark on the incongruity of having Socrates as a living person speak of Aristotle's and Epicurus' supposed flaws; since it is obvious from what comes next, 'mihi ipsi Alcibiaden et Phaedrum obiectate', that Socrates carries on speaking to the very end of the remaining text.

28. (p. 52): 'turbo quidem animos uestros rotat et inuoluit fugientes petentesque eadem'; C.: 'as they fly around chasing the same things'. It is not clear if C. understands 'fugientes' and 'petentes' as both governing 'eadem' or not. But I believe that 'eadem' loses all point if 'fugientes' is taken otherwise than with 'eadem'; the idea of 'turbo' also implies that the movement of the 'stulti' is circular and without a goal: 'while they flee and pursue the selfsame things' (Basore).

tranq. an.

2.1 (p. 58): I have always found the phrasing of the beginning of Seneca's reply in the dialogue quite interesting: 'quaero iam dudum ... ipse tacitus'. Seneca has been silently listening to Serenus' confession, which is almost a soliloquy. One is reminded of Hor. sat. 2. 7, 1-2 'iamdudum ausculto et cupiens tibi dicere seruos pauca reformido'; Boet. Cons. 1. pr. 1 'haec dum mecum tacitus ipse reputarem'. The reverse in Sen. Ag. 126; 128, where Clytaemnestra's nurse interrupts her mistress' soliloquy which she describes as a silent brooding: 'quid tacita uersas? ... licet ipsa sileas, totus in uultu est dolor'.

2. 15 (p. 64): 'subit illud tabidarum deliciarum'. C.: 'out of their enervating self-indulgence arose the feeling': this must be the sense, but perhaps a word is needed here about the grammatical construction. The same syntactical structure is anticipated, within the tranq. an., by a series of similar phrases: 2.8 'subrepit illa animi iactatio; 2. 10 'hinc illud est taedium (...) inde ille affectus otium suum detestantium'. In 2. 15 the noun-function is supplied by the quotation 'quousque eadem' and something like 'lamentum' has to be understood here; the genitive deliciarum is a proper possessive, or a genitive of reference: 'from all this arose that lament which is always the companion of such self-consuming pleasures: quousque eadem?'

5.3 'et in florenti ac beam †pecuniam† inuidiam, mille alia inertia uitia regnare'. Why is 'pecuniam' so suspect? cf. Hor. epist. 1.6, 27 'regina pecunia donat'; Petr. 14. 2 'quid faciunt leges ubi sola pecunia regnat'; 'pecunia', in 5.3, stands metonymically for 'auaritia'.

5.5 (p. 70) '<vere> [add. Haupt] ut opinor Curius Dentatus aiebat malle se esse mortuum quam uiuere'; C.: 'truly, I believe, C.D. used to say that he preferred real death to living death'. In a 'spoken' or recited text the antithesis 'esse mortuum quam uiuere' could rely on the speaker's or reciter's intonation, but in a text for reading 'mortuum' needs to be repeated, if there is to be a point, and preferably preceding 'uiuere', which would make a perfect chiastic and oxymoronic arrangement: 'malle se esse mortuum quam mortuum uiuere'. 'mortuum uiuere' would also be a very Senecan iunctura. Cf. Castiglioni, Riv. Fil. Istr. Class., 21 (1924), p. 374 for examples of Seneca's love for unexpected conclusions and antithetical brachylogies.

7.5 (p. 192); 'uix tibi esset facultas dilectus felicioris...'. The sense is clear: in the matter of making friends, it would have been best to have lived in the age of the great Athenian philosophers etc.; but the Latin in which this is said is worryingly ambiguous, and it could also mean 'you could not hope for better times than ours'.

8.9 (p. 74): 'quae superfunduntur et undique magnitudo sua uulneribus obiecit'; notice the change of grammatical function of 'quae', first as subject then as object; such constructions are frequent in Senecan prose (cf. J. Müller, Sitzungsb. Wien. Akad., 118 (1889). Here Seneca comes closer to a kind of anakolouthon frequently to be encountered in Vergil (see Vahlen, Opusc. Acad. I. 166, with examples; Wagner, Quaest. Verg. 555).

9. 1 (p. 76). 'placebit autem haec nobis mensura, si prius parsimonia placuerit, sine qua nec ullae opes sufficiunt, nec ullae non satis patent, praesertim cum in uicino remedium sit et possit ipsa paupertas in diuitias se aduocata frugalitate conuertere'. The passage is obviously corrupt, and C. 's translation does not make much sense: 'without which no amount of wealth is enough and no amount is not ample enough'. 'praesertim cum in uicino...' implies a positive, if modest, alternative, such as 'poverty is not altogether an evil' or 'a small substance is even too much for the wise man'. 'Nec ullae non satis patent' on its own would do, but it is an arduous enterprise to disentangle it from its correlative 'sine qua nec ullae opes sufficiunt'. Editors focus on the second colon, assuming something to have dropped by homoeoteleuton: 'nec <cum illa> ullae non Haase; (nec ullae <cum illa> non Wesenberg. There are difficulties about these solutions (especially Haase's), as the demonstrative pronoun 'illa' would not be consistent with the relative 'sine qua'. I wonder if it could not be assumed that something is wrong in the first colon instead; I would therefore write: 'sine qua nullae opes sufficiunt; nec ullae non (= at nullae non) satis patent, praesertim...' ['yet no wealth is not ample enough...']; for the syntactical sequence (a negative pronoun contrasted by a 'positive' double negative) cf. Sen. tranq. an. 1.10 'quorum tamen nemo ad rem publicam accessit, et nemo non misit'; 'nec' here would have an adversative-contrastive force, cf. Krebs, Antibarbarus, 19628, ii, p. 134, with examples; Nägelsbach, Latein. Stilist., 19059, p. 771; the corruption would seem due to an attempted linking by polysyndeton of colon a and colon b, with 'nullae' changed to 'nec ullae'. Cases of non-correlative 'nec' introducing a new period are frequent; cf. Tac. Agr. 8; Luc. 1. 129, and elsewhere.

9.3 (p. 76): 'cogendae in artum res sunt ut tela in uanum cadant ideoque exilia interim calamitatesque in remedium cessere et leuioribus incommodis grauiora sanata sunt'; ideo translated, with C., 'for that reason', is slightly illogical. Better, 'which having been done', or the like (e.g. 'de là vient que' Waltz).

10. 5 (p. 80): 'quae excelsa uidebantur praerupta sunt'. C.: 'what look like towering heights are precipices'. I wonder why the imperfect is used; the sentence has almost a biblical sound, which may appear suspect. With the imperfect one would then expect a dramatic consequence, an event breaking out ('what stood high has been broken down'), but 'praerupta' here is an adjective, and 'sunt' a copula. Perhaps yet another case of 'Christian' phrase-structure?

11.3 (p. 82): 'appellauerit natura quae prior ... credidit'; 'prior' here is said with regard to fortune and its gifts, which come second.

11.12 (p. 84): 'das in te uires rebus aduersis quas infregit quisquis prior uidit'; C.: 'you give adversity a power over you which the man who sees it first can crush'. The relative pronoun 'quas' is a kind of connective relative and agrees with 'res aduersae', not with 'uires': 'but he who saw the 'res aduersae' in time, before they leapt upon him, always managed to vanquish them'.

14.4 (p. 88): 'Canus Iulius uir in primis magnus'. C. translates: 'Julius Canus, an outstandingly fine man'; cf. also const. sap. 18.2 'Asiaticum Valerium' which is rendered with 'Valerius Asiaticus', whereas ibid. 17.1, 'Fidus Cornelius' for no apparent reason remains 'Fidus Cornelius'. A note could well have been supplied here, though I do not really know that any definitive explanation has yet been found for such inversions of nomen and cognomen. It occurs already in Horace, epist. 1.8, 1 'Celso ... Albinouano', and elsewhere in poetry, where it can be explained with reference to stylistic considerations; in prose, this phenomenon becomes increasingly frequent in authors of the early Imperial age. B. Salway, Journ. Rom. Stud. 84 (1994), 124 ff., esp. 130, asserts that Tacitus' practice of reversing nomen and cognomen is probably to be considered no more than a mere stylistic device (although one would think that a stylistic device shared by such people as Seneca and Tacitus would have deserved more attention). In the case of Julii not belonging to the Imperial family, but deriving their name from the granting of citizenship either by Julius Caesar or one of his successors, and mainly indicating Roman citizens of Gaulish origin (like Tacitus' father-in-law, Agricola), one could argue that a reason for the inversion may reside in the lack of specificity which pertains to such nomina. So also Kraus on Livy 6. 18, 4 (Cambridge, 1994, p. 199), following Ogilvie.

16.2 (p. 92); it could perhaps be suggested that 'ipsorum illos animo desidera' does not mean (C.) 'long with your spirit for a spirit like theirs', but desidera illos animo ipsorum, i.e. with their strength of mind (cf. also what follows next: 'quid enim est turpius quam si maximi uiri timidos fortiter moriendo faciunt?'); in other words, si fortes fuerunt, fortiter desidera, sin autem ignavi, nihil periit, ergo nihil desiderandum.

const. sap.

6.2 (p. 108); a few words could perhaps have been spared for the word-order, which makes it difficult for the student to separate 'in hominem' from 'tantam animi magnitudinem': it is the familiar word-pattern named after Hammelrath, (Gramm.-stil. Beiträge ..., Emmerich, 1895), in which the common element of a dicolon is usually placed in the middle position between the two cola.

6.7 (p. 108): 'isti diuites'. C. wrongly omits the deictic in his translation; cf. Sen. uit. beat. 1.4 'ista tanta coaceruatio'; 2. 4 'uides istos qui eloquentiam laudant', two examples in which the demonstrative pronoun does not refer to anything the speaker has already mentioned, but accompanies an imaginary eloquent gesture by the moral preacher. This pragmatic function of the demonstrative (adject. and pron.) is a recurrent feature in Senecan enargeia. The sense is: 'these rich men, we all know of, and we live surrounded by'; cf. Basore: 'the losers are yonder rich men who have lost their estates'.

9. 1 (p. 114): 'omne autem fortuitum circa nos saeuit et in uilia': it could have been made more clear in the translation that 'circa nos' means 'round us, i.e. without touching us': 'circa' is emphatic; the fury of fortune exhausts itself before reaching the wise man. F. Pincianus proposed 'citra nos', in order to make this less ambiguous.

9.4 (p. 114): notice the strange transition from 'faueamus, obsecro uos, huic proposito aequisque et animis et auribus adsimus' (an inclusive plural addressed to a friendly audience) to the chastising 'nec quicquam ideo petulantiae uestrae ... detrahitur' which immediately follows (some early editions, including Lipsius, have 'nostrae'). One is under the impression that the speaker is addressing two different groups (one favourably inclined to him, the other a throng of aduersarii), and that he would make himself clear in recitation by first lowering his voice and then raising it and making gestures.

9.5 (p. 115): 'in certaminibus sacris': C.'s reader is left to wonder if there is here a specific allusion or not; in fact, there could be a reference to a special kind of games, introduced by Augustus in imitation of the Greek Olympic games, and then renewed on special occasions under Gaius, Claudius and Nero; who, in fact, celebrated these games in the year 60 A.D. This could give us a clue for the dating of this dialogue, as Seneca seems to refer to things actually seen. Cf. Friedländer, Sittengesch., Leipzig, 192210, II 149 and n. 6; cf. also Wiel (Utrecht, 1950), ad 1.

10.4 (p. 116): 'nulla uirtus est quae non sentias perpeti'. C.: 'there is no virtue which you would not perceive having to endure something'. This is a clear slip. The passage means instead: 'it is no virtue to endure evils one does not feel'.

12.2 (p. 118): 'ergo par pueris longiusque progressis, sed in alia maioraque error est'. 'in alia maioraque error est' explains only the second component of the preceding phrase, 'longiusque progressis', but this abruptness is not so obvious, and perhaps needed explanation.

14. 2 (p. 122): 'ut uincat, par fuit'; C.: 'for the sake of winning he puts himself at the same level'. But, although the succession of perf. and pres. subj. would not necessarily be an obstacle (cf. Hofmann-Szantyr 550-1), it is probably better to understand the ut-clause as a concessive, taking full account of the order of words: 'true, he wins; by which he has just managed to put himself at one level with them'. Cf. Lipsius' paraphrasis ad l.: 'et quamvis superet, in aequo cum suo stetit'. But even if C. were right, a note illustrating the 'exception' to the sequence of tenses would have been very welcome.

18.2 (p. 126): 'in conuiuio, id est in contione': I think C. has got the point with his translation ('in a banquet, which means a public assembly'), but a student needs more (at least I did when I read it for this review); the point is that a royal banquet is so crowded with important people that it is tantamount to a public occasion.

helu. matr.

9.6 (p. 146): 'ne exules essent si sine illo fuissent'. C. :'in case they would be exiled if deprived of him'; I am not sure if this is accurate; better 'in order not to be like exiles themselves, once they had been deprived of him'; cf. Basore: 'lest they should be like exiles if they should be left without him'.

10.3 (p. 148): 'inuitus enim sanatur [sc. pauper a luxuria] et, si remedia ne coactus quidem recipit, interim certe, dum non potest, illa nolenti similis est'. C.'s rendering of the passage is 'his inability to have those things looks like unwillingness', but the Latin phrasing is hardly intelligible to an undergraduate -- and indeed, without a note explaining the undeniable grammatical oddity, to more advanced students of Seneca as well -- 'illa', in the punctuation of Reynolds' text, can only be understood as referring to poverty, but even so it is quite suspect, and the recc. omit it. If 'illa' is sound, the problem lies rather with the pres. part. 'nolenti'. 'similis' plus a pres. part. is acceptable, and it occurs frequently even in prose (discussion in Traina, Maia, 21 (1969), p. 71 ff.), but if 'illa' refers to 'paupertas' the Latin becomes quite strange. There seems to be a contamination of two different constructions ( = 'paupertas eum facit similem nolenti'; and 'non posse aeque ac nolle est'). Certainly, it is hard to see how the subject of 'similis' can be other than 'pauper'. According to Gertz, Studia critica, Hauniae, 1888, 160, A offers some punctuation after 'illa': 'dum non potest illa (= 'dum non capit remedia'), nolenti similis est', which, although rejected by Gertz himself, still seems a reasonable solution, doubtlessly much better than having 'illa' as a subject of 'similis est'. For 'ille' in a similar context, cf. tranq. an. 2. 2 'opus est itaque non illis durioribus quam iam transcucurrimus'. Madvig's 'non potest uelle' is also quite attractive.

16.6 (p. 162): 'filius magno aestimauit Gracchorum natales, mater et funera'. What is the point here? C., following Duff, translates 'natales' with 'parentage'; but this seems misguided, if the sentence is to be taken along with Gaius' utterance 'tu matri meae male dicas quae me peperit?' The opposition 'natales' and 'funera' is a quintessentially Senecan, if somewhat obscure, concettismo: the pair here stands metonymically for 'day of one's birth', 'day of one's funeral'. 'funus' for funeral does not need any defence; 'aestimare' is here employed in a weighted sense: almost 'to honour', 'to celebrate a given day': Gaius was proud of the day of his and his brother's birth [as it was, in his opinion, a glorious occasion]; Cornelia, strong-minded lady as she was, considered that it was not meet for her to lament on the day of her son's funerals, but rather to celebrate her pride and almost to thank fortune for having allowed her to give birth to such men: cf. also Gertz, cit., 164 n.

19.6-7 (p. 168): 'neminem prouincialem domum suam admisit'. What exactly is Helvia's sister being praised for in not having admitted into her house any provincials? A historical note would have been welcome. Cf. Viansino, Milano 1990, ad l., quoting Tac. Ann. 2.55; 3.33; Iuv. 8. 129. A pedantic point: 'non metus mortis ... deterruit quominus ... non quaereret quemadmodum inde exiret sed quemadmodum efferret': the second 'non' specifies only 'q. i. exiret', and this is the reason why it was postponed by Pincianus; in similar cases, Seneca normally postpones the negative particle: cf. uit. beat. 18.1 'dicebant non quemadmodum ipsis uiuerent sed quemadmodum esset et ipsis uiuendum; 18.2 'ne uirus istud ... me impediet quominus perseuerem laudare uitam non quam ago sed quam agendam scio'.


  • [1] Both the Latin and the English text contain a number of misprints, mostly, to be fair, self-correcting even to the least expert of students [p. 14 'corunua' for 'cornua'; p. 42 'abscondent' for 'abscondet'; p. 58 'uera similtudine'; p. 70 'neo' for 'nemo'; p. 80 'in longinquae' for 'in longinqua'; p. 144 'ige' for 'igne'; p. 150 'tanc’ for 'tunc'; p. 156 'signominia', then again 'igominia' for 'ignominia'.
  • [2] A subtle distinction between reflexive 'se mutare' and intransitive 'mutare' seems to be made by Seneca at de ira 2.36, 3 'qui ad speculum uenerat ut se mutaret, iam mutauerat', unless 'se' is apo koinou.
  • [3] One would expect it to run as follows: 'but they, by doing what they do, cannot see that they are only inflicting the same punishment upon themselves'.