Tertullian's De Pallio is in the view of one of its ablest editors, Salmasius, far and away the most difficult and obscure work of a difficult and obscure author. The difficulties and obscurities flow from two sources. First, Tertullian himself, who here devised a telegraphic idiom that often baffles comprehension, especially when couched in his already idiosyncratic Latin. Secondly, the MS tradition, which is not reliable, a point to return to in a moment. Still, as regards Pall. itself the heroic efforts of Salmasius (1622), and Rigaltius (1634), ably furthered by Oehler (Vienna 1853), Gerlo (Wetteren 1940; Turnhout 1954 [Corpus Christianorum series, vol. 2]), and Bulhart (Vienna 1957 [Corpus Scriptorum Ecclesiasticorum Latinorum series = CSEL, vol. 76]) have given us a text we can more or less understand, even though details remain uncertain. As regards Tertullian's style generally, the studies by H. Hoppe (Syntax und Stil des Tertullian [Leipzig 1903], Beiträge zur Sprache und Kritik Tertullians [Lund 1932]), G. Thörnell (four volumes of Studia Tertullianea [Uppsala 1917-26]), and E. Löfstedt (Zur Sprache Tertullians (Lund & Leipzig 1920) have done much to clarify specific usages by this perverse charmer. Finally, the Thesaurus Linguae Latinae is an inestimable boon, given that the Oxford Latin Dictionary ignored the Latinity of Christian writers. The TLL of course relies on what earlier commentators have said (Oehlers's Index uerborum must have been a considerable help to it), but the discussions are also independent and weigh up the many difficulties. So we are not without resources, difficult as the task of understanding this work must remain.
Pall. is something of a sport within the oeuvre of Tertullian. Literary historians rightly see it as belonging to the tradition of epideictic oratory, and so it was an easy transition for Vincent Hunink from the Florida of Apuleius, which he edited with commentary in 2001, to a not dissimilar work by a roughly contemporary Christian. On p. 11 of the Introduction H. sets out his aims: he adopts Gerlo's revised 1954 text, and intends to keep textual questions to a minimum; the commentary is designed to treat the text as a piece of literature rather than an historical document, so the emphasis is on composition, and strategies of communication (what does the author do with words?). The second part of his agenda is unexceptionable, but the reluctance to discuss the establishment of the text in any detail is in my view a bad call. Let me explain why.
This is not a book for tyros, but even the target audience is likely to want considerable help with the Latin, especially since the text is not at all reliable. The three MSS on which Pall. is founded are all fifteenth-century, so very liable to scribal interference; A. Kroymann had specially noted the meddling with the text by the scribe of Florentinus Magliabechianus, Conventi soppressi VI 9 (= N) (see CSEL vol. 47, p. XXI, n. 3). His successor, V. Bulhart, on p. VIII of his introduction to CSEL vol. 76, which contains Pall., was nonetheless enthusiastic about the readings of these three MSS, and particularly liked the often peculiar word order of N. In that same volume, however, there is an edition of the tract De Paenitentia by P. Borleffs. Now N is again one of the MSS he had at his disposal, but he also had two older MSS, the codex Trecensis 523 (= T) of the twelfth century, and a Vaticanus Ottobonianus 25 (= O), of the fourteenth. He noted on p. 131 that T and O are often in agreement with one another and offer the correct reading against the remaining later MSS, including N. More specifically Borleffs repudiated the peculiar word order offered by N (p. 135), which Bulhart endorsed (p. IX). (Did these editors never bother to compare notes?) I at any rate conclude from this that if for Pall. too we had a witness equivalent to T, editors would marginalize N's readings. Given the unreliability of this tradition, discussion of the text is not an option but an imperative, since much depends on the acumen of editors.
As has just been mentioned, H. prints Gerlo's 1954 text, with one lapse, however: somehow or other the square brackets which Gerlo put round the phrase at 3.6.3 uermiculi genus est have been turned into round brackets. Pity: Gerlo followed Ernout in expelling an intrusive gloss. If he was right -- and it seems a manifest intruder -- H.'s note on p. 167 is not very much to the point. Gerlo cautiously put into his apparatus some proposals that might well have been introduced into the text. For instance, at 3.7.3 he proposed deleting 'i. Cato' (which editors turn into idem Cato) as a gloss; H. might have approved of this since he is well aware that often in the treatise Tertullian likes not to name his historical exemplars outright but refers to them by periphrases (see his n. on 4.7.2, p. 218). Or again, at 4.3.1 Gerlo left sortem in the text, but suggested a much better word in the apparatus, seriem.
H. does, to be sure, offer some textual discussion (e.g., at 2.4.1, 2.6.3, 4.10.1), but still one would like to see more. For instance, at 2.3.2 in forma mutauit H. finds forma difficult. Now forma is the conjecture of Salmasius, and the MSS transmit infirma. G. Säflund, De Pallio und die stilistische Entwicklung Tertullians (Lund 1955) 5, n. 3 argued well for dividing this into two words, in firma, taking firma as neuter plural (rather than ablative sing., as Löfstedt had hesitantly suggested); firma undeniably makes a neat contrast with fluitasse in the previous sentence (this is, in other words, as much a literary as a textual matter). Carl Becker approved this solution in Gnomon 28 (1956) 427, and in firma is also read by Bulhart in his CSEL text, though he doesn't notice that Säflund preceded him.
On 2.4.2 mortem uiuit, it is useful to know that Gerlo defended this emendation, but shouldn't we be told who proposed it (Rigaltius), and that Hoppe 1903: 16 too had endorsed it? This is especially important since Bulhart overlooked it entirely.
3.6.1 nec de ouibus dico is especially significant. This is indeed Gerlo's text, but H. doesn't note that in his apparatus he hankered after the reading of N, ouilibus. That is the lectio difficilior, it is certainly not impossible Latin, and it is hard to believe that a scribe faced with ouibus would wilfully change it to ouilibus. Bulhart took the plunge and printed ouilibus in his text, and that reading is now endorsed by TLL 9,2.1190.60. To print ouibus is retrograde from the text-critical point of view, but from the literary point of view it replaces a metonymy, characteristic of this work (cfr sacerdotium in 1.2.1) with a banality.
3.6.2 nam et de mari uellera, qua muscosae lanositatis lautiores conchae comant: H. says the syntax is vague, but if with Bulhart you read quae for qua and take it as an internal accusative with comant, all makes perfect sense (muscosae lanositatis will then be descriptive genitive).
At 3.7.3 H. tries to explain the expression palliato habitu, without noticing that some editors separate the words into different cola (cf. TLL 10,1.127.63-5).
At 5.4.4 he accepts erat without noticing that errat is the MSS reading, and not unlikely, especially after Säflund's defence. I won't dilate upon the textual issues any further, but it should be clear now why I find that the decision to minimize discussion of them detracts from the usefulness of this commentary. It should be axiomatic that textual and literary analysis complement each other, and literary analyses of an unreliable or unacceptable text are a waste of effort.
To return to issues covered in the Introduction, one of the chief problems in interpretation is the date. H. remains agnostic, given the (deliberate) ambiguities (p. 15), but on balance favours an early date, between 209-11. As for the genre, he notes that it is a unique work, similar however to Apuleius' Florida, an epideictic speech. This will form the basis of many of the best observations in his commentary. Finally, the purpose of the work: it is not apologetic (p. 21), nor even specifically Christian (pp. 22-3) -- the aim is pure entertainment, a view as old as Boissier.
Since something has already been said about the problematic text, I turn to consider H.'s translation. Generally it is sound (and that is an achievement!), but there are some lapses, which I here list:
1..3.2 in oscillum penduli impetus 'for the oscillatory work of pending violence'. pendulus is to be taken literally: the swinging beam of the ram is 'suspended' from the fixed beam above it; impetus can thus mean either 'attack' or 'thrust' (cf. Lucan 3.490 suspenso ... ictu with H.'s own n.). The n. on p. 86 perhaps comes closer to the truth if 'lauch' is a misprint for 'launch'.
2.2.2 sol stationibus annuis: H. translates rather vaguely 'yearly positions', but Salmasius said this refers specifically to the solstices: was he right?
3.7.1 ingenia 'talents': more probably quasi-concrete (OLD 6b; the use starts in the first century AD), 'devices', as correctly translated at 1.3.3.
4.8.2 fluxu is not 'luxury'. penes se 'close by'; see rather TLL 10,1.1057.14-20 for the sense in animo suo.
4.10.5 renuntiatorem: H. forgets what the verb renuntio means in classical Latin (Tertullian uses it to mean 'repudiate' at Apol. 38.4). The noun is found again at De Anima 57, where it clearly means 'one who proclaims/declares'. See Lewis & Short, or Forcellini.
5.5.3 ambitio 'desires', but more correctly glossed in the n. on p. 268 'ostentation'.
6.1.3 aemulus is less 'rival', than 'enemy/adversary', a frequent sense in Tertullian, for which see Rigaltius' or Oehler's Index uerborum s.u.
Let us turn now to the core of the work, the Commentary. I'll divide my observations into roughly two groups, 'Realien' and historical context, on the one hand, and linguistic usage on the other.
Tertullian was credited with vast erudition, which he airs in this display piece. H. is pretty reliable in discussing this aspect of the work, but there are some questionable judgments.
At 1.3.2 Tertullian attributes the invention of the battering ram to the Carthaginians, but H. finds the contention unlikely (p. 84), or probably not an historical fact (p. 86), without providing any grounds for his scepticism. He is however aware that Vitruvius too claimed the engine for the Carthaginians, and indeed he gave quite a circumstantial account of its origin and development, 10.13.1-3; it seems to have been based on treatises by engineers of Alexander's, Polyidus and Diades. (I stress the circumstantial quality of Vitruvius' account, because he describes the development of the swinging ram (suspendit), which points to the correct translation of penduli impetus, discussed above). Why should Tertullian not have believed Vitruvius or some other source? I say some other source, for Vitruvius is not alone in his account, which is also found in the Greek treatise of Athenaeus Mechanicus (see J. Kromayer and G. Veith, Heerwesen und Kriegführung der Griechen und Römer [Munich 1928] 224-5 for a translation, a passage actually cited by H. himself in his n. to Lucan 3.490). My aim here is to show that Tertullian, even in a sophistical speech, is not sloppy or inventive; he draws upon the stores of his well-founded learning, with which antiquity readily credited him.
3.6.3 araneorum horoscopis 'spiders' sundials', i.e., webs: it might have helped to draw attention to the spherical shape of some ancient horologia (for illustrations see Daremberg - Saglio, 'horologium', V. 256-60). When you visualize the lines drawn within the conch, the vivid metaphor makes perfect sense.
4.1.3 In his nn. here H. should have avoided using the words shave/shaving; there was nothing unmanly about shaving at Rome. Depilation by plucking was another matter, see Suet. Iul. 45.2.
I now turn to three issues concerning the historical context of the speech.
1.2.1 Aesculapio iam uestro: 'the added iam uestro seems rather too meticulous. By Tertullian's time, Aesculapius could be considered a local god', says H., who goes on to cite Rives Religion and authority in Roman Carthage (Oxford 1995) 181-4; but if he'd looked at p. 156 he'd have seen that Rives believed the cult to be of Antonine date, and so comparatively recent. H.'s note on Apul. Flor. 18.38 was much more circumspect.
2.7.1 On p. 132 H. finds it odd that Tertullian praises the contemporary prosperity of the Roman Empire. His 'take' on the passage and on the following is in my opinion seriously misleading. But first we should see what he has to say about 2.7.2 Deo fauente: H. asks if the 'fanatic' Tertullian, 'the advocate of Christian resistance and martyrdom', would be prepared to say that God favours those on the throne of Rome. I cannot find a passage where he does say exactly that, but there are no fewer than three passages in the Apologeticus (like Pall. addressed to a pagan audience) in which he stresses that the Christian community prays regularly for the well-being of the emperor and the peace, prosperity and stability of the empire (30.1-4, 32.1, 39.2). There is another extended passage at Ad Scapulam 2.6-9 along similar lines. It may therefore be supposed that if the emperor was well and the empire flourished like a well-tended estate (cultissimum rus) then the Christian would see his prayers as answered, and deem that God indeed favoured the Roman throne (which is after all what such prayers were asking for). Tertullian's claims in the Apologeticus echo the noble close of Clement's [First] Epistle to the Corinthians, 60-61, and both insist that the emperors derived their power and position from God. As regards the prosperity of the empire at this time, I miss any notice of De Anima 30. This is odd since H. had just referred to that chapter in his notes on 2.6.4-5, but if he'd read a bit further he'd have found a parallel for Pall.'s praise of the empire's prosperity as well (though without any reference to God's particular favour). I therefore cannot follow H. in finding the tone here ironic. I'd even say that in Pall. it is impossible: Tertullian did not have to begin his review of the empire's current fine condition with the phrase Deo fauente. Its presence however precludes irony to my mind, since one thing the 'fanatic' Tertullian would not allow himself is an ironical tone in referring to God's favour. I conclude that the passage is sincere (in a panegyrical and apologetic sense), since Christians prayed for the continued welfare of the state, and if it continued, then God had shown his favour to Rome.
P. 261 'The whole section most probably refers to ... Carthage rather than Rome'. Agreed, but how then to explain the translation Campus Martius for campo?
Before turning to the commentary on linguistic matters, I want to scroll back to a claim H. made on the first page (9) of his Introduction. It was unfair to Bulhart to say that there are no useful tools for approaching the text. Bulhart did two very useful things in his edition. First, in his Praefatio he provided a fifty-page compendium on the style of Tertullian, founded naturally upon the studies of Hoppe and Löfstedt. To this compendium he constantly referred in his apparatus criticus, both to elucidate the meaning of the text and to justify his choice of reading. Secondly, in his apparatus he also appealed regularly to what was by then available of the TLL for an explanation of Tertullian's often unexpected usage. I've been using Bulhart in tandem with H., and have to say that I found the former edition a very useful tool indeed, which often helped either where H. failed to comment or where his comment left something to be desired. I append a list of all the places where I found myself wanting further information, or felt that H. had not provided a correct analysis. Let me say at once that in the main he does provide correct information about Tertullian's usage and there are some attractive accounts of his literary technique. But since this is not a central text by a standard author, the reader can expect to be given maximal support by the commentator, there being few other resources except self-help. It will be clear from what follows that too often I indeed had to do the job myself, and consult the Thesaurus, or one of the monographs on Tertullian's usage referred to above to ensure that I really understood the text.
1.1.1 prosperos temporum: what kind of genitive with the adjective? 'Reference' or 'respect': see Hoppe 1903:24, and cf. 4.5.1 insignes libidinum. It clearly isn't the same as the one cited from Horace.
1.1.2 ab imperio: it cannot possibly be ablative of separation (or if it can, a translation would be helpful). For the use of ab see OLD 25.
1.1.3 Why is subtemen deemed an uncommon word? The n. on p. 88, on the perfect form of stupuere (1.3.3), should be here, on the form fuere.
1.1.4 ab utroque laterum: no n. on the unusual partitive genitive, see Hoppe 1903: 20.
1.2.1 No n. on sacerdotium, which Bulhart takes to be genitive plur. and Oehler to = sacerdotes.
1.2.2 de habitu ... salutasset: the use of de is not perhaps common, but hardly irregular, see OLD 13. Romanum praecoca. Agreed, Romanum is a 'Greek' neuter accusative of respect, but dependent upon an adjective in prose it should be noted as very peculiar (see Hoppe 1903:18). H. doesn't observe that Pall. seems to have a higher incidence of this 'figure' than any other of Tertullian's works; I'm here relying on the admittedly partial 'Index rerum et locutionum' of the Corpus Christianorum series edition, II.1510, s.u. 'accusativus Graecus'. Some fourteen instances are there cited, of which fully seven are from Pall. (H. hasn't noted in his commentary 4.2.3 aurem effeminatus, or 4.7.2 crepidam aeratus). To say moreover that praecocus 'is an old variant form for the normal praecox' is misleading, and H. should have consulted TLL 10,2.512.46-48 to learn that praecox is the older form; further down that column we are told that a grammarian possibly contemporary with Tertullian, Caper, insisted that praecoca (-qua) was the correct form of the nom. sing. fem.; so Tertullian may have been toeing the grammatical line, for a change. Again, the Corpus Christianorum index shows that praecocus (-quus) is the only securely attested form in his writings.
1.2.3 ut senium, non fastigium exemptis: are these really accusative of respect? Hoppe 1903: 17 says that the accustive is a genuine object, retained with the passive. But that passive is nonetheless very odd, and Leumann in his TLL article, 5,2.1502.14-15, offered no parallels for the usage, though he supports H.'s claim that eximo is here used as if it were a verb of 'dressing' ('i. fere exutis', says Leumann). So shouldn't H. stress the weirdness of the syntax, and cross-reference to 2.4.3 Vulsinios deusta and erepta Pompeios? (By the same token, there ought to be a cross-reference at that later passage to this earlier one.)
1.3.1 tabulata 'layered'; there might be a cross-reference to 5.1.4, or better, the matter there should have been set out here.
2.2.4 memor uiridem cum conspicis flauam, mox uisurus et canam: of memor uiridem H. says on p. 102, 'a somewhat elliptic construction; i.e. memor eam [terram] uiridem fuisse'. Alternatively, H. might have followed up the references in Bulhart and told us that Löfstedt 1920: 12 had noted that in later Latin memor commonly governs a direct accusative, and this passage he reckoned to be a further example of that syntax. He didn't help research by failing to put it into his Stellenregister, but still it didn't escape the notice of O. Prinz in TLL 8.659.67-9. Löfstedt further observed in n. 3 on that page that elsewhere in his writings Tertullian used the more regular genitive with memor, so once again we are confronted with a deliberate oddity in the style of Pall. H.'s note banalizes the usage.
2.2.5 'Three colours are specified [green, yellow, white], where four might perhaps be expected': quite right, and for the fourth, the colour of autumn cf. Spect. 9: uenetus 'sea-blue', which doesn't suit the earth, hence its omission here.
2.3.3 'angustis: i.e. angustiis' says H. Salmasius pointed out that the word is actually the plur. of angustum, and that, it should be noted, it is highfalutin in prose (we find it in Tacitus). nouum uitii = nouum uitium: true, but we could be told, following Hoppe 1903: 112, that the construction of the genitive is unusual.
2.4.2 For Tertullian's use of hactenus 'no further/longer' see Oehler on De Exhort. Cast. I. 738; Hoppe 1903: 111, both of whom list more examples.
2.5.1 non minus quam et: H. notes the redundant et as characteristic of Tertullian, but Bulhart pp. XLVII-XLVIII gives a much fuller account.
primus regnare: the n. on the syntax, taken over from Hoppe, is both misleading, since virtually all such infinitives dependent upon adjectives are owed to Greek not just those that appear in later Latin, and incomplete: Hofmann-Szantyr's Lateinische Grammatik (Munich 1975) also point out that Tertullian introduces this particular poeticism into prose (p. 351).
autumant ... profanitas: the number of the verb is correctly explained by 'constructio ad sensum', but a cross-reference to 2.6.5 posteritas ... producunt (and vice versa) would surely help.
2.5.2 ultra adverb used of extension back into time might have deserved a note.
2.6.1 on the form detotondit: there is no point in the reference to Ennius, since it is a palpable error (it doesn't scan). The reason for the choice of the form is possibly the rhythm of the clausula (cf. Hoppe 1903: 154-5).
2.6.4 educatur 'was brought' (i.e., = educitur), but no n. Salmasius established this bizarre sense by appealing to De Virg. Vel. 12; TLL 5,2.116.61-5 adds a passage from the Itala.
2.6.5 producunt is very obscure, and H. might have referred to the discussion in TLL 10,2.1641.60-4.
2.7.1 curricula nostra 'our days': granted the noun refers to time, it should be noted that without a defining genitive (e.g., annorum) it is pretty unusual, and this passage could be added to the handful of instances in TLL 4.1507.54-9.
3.1.1 On Tertullian's use of the archaic pauus rather than pauo H. says he 'may' be imitating Enn. Ann. 11Sk. He definitely is, when it is recalled that he refers to that passage explicitly in De Anima 33, and implicitly in De Resurrectione Carnis 1. It tells us something of Tertullian's influence that the form then crops up in Christian writers (see TLL 10,1.835.70-5).
3.3.1 Pacuvius: H. suggests that his plays might have been well known in Roman Africa. But since the line Tertullian has just quoted is found in Cicero, isn't it more likely that the only Pacuvius anyone knew at this date was what was to be found quoted in classical authors?
3.3.2 licet: 'Licere is used, as often in Tertullian, in a weakened sense'. No evidence is offered to support this claim. Cf. however 2.2.1 for liquet in a very similar expression; the words are for scribal purposes the same.
3.3.3 Chamaeleon pellicula uiuit: 'Pellicula is surely ablative'. Perhaps, but shouldn't the alternative be briefly discussed? For an explanation of the nominative see Bulhart p. LV.
3.5.1 ad illum ex Libya Ammon facit: no n. on the absence of iter, see TLL 6,1.122.5, a reference Bulhart provided.
3.5.2 H. has notes on both of the first words, denique and cum, but he doesn't explain why he disagrees with Bulhart about their meaning. I make a point of this since Bulhart's interpretation strikes me as superior: 'For among the Egyptians they say ... '.
3.6.1, 4.9.3 quis = quibus, no n., but the form is characteristic of the highfalutin style of the piece (Apuleius uses it twice in Met., but not in Flor.). Neue - Wagener vol. II p. 470 cites these as the sole examples in all of Tertullian, and there are no more added in the Corpus Christianorum's Index rerum et locutionum, II.1597.
3.7.2 passiuitus: the adverb is indeed unusual, but it would help to say that Tertullian is very keen on the related words passiuus and passiuitas (4.8.3n.).
3.7.3 ipse qui Graecos praeter urbem censebat: H. tells us what TLL's view of censeo is, but not its doctrine on praeter: see 10,2.992.10-12. It is also worth glancing at what Löfstedt has to say about this sort of impressionistic expression in his Kommentar on the Peregrinatio Aetheriae, p. 212.
3.7.3 iuridicinae suae in tempore: Salmasius says crisply that this refers to his praetorship.
4.1.1 nec needs a n. = ne ... quidem 'not ... either'.
4.3.5 sobriae, glossed by Bulhart ualidae (unnoticed by H.).
4.3.6 tota oris contumelia: correctly explained by H., but amply illustrated by Bulhart section 108.
4.6.3 sarabara: H. should have followed up Bulhart's hint, and read G. N. Knauer's essay on this word in Glotta 33 (1954) 100-18. On p. 104, Knauer observed that Tertullian picked up the rare word in all probability from the Book of Daniel, which he referred to in Orat. 15.2 (it is cited by H.).
4.7.1 The note on the type of shoe called baxa is unexceptionable, but it would have helped to quote in full the reference to Apuleius, Met. 11.8.3, since it too refers the footwear to a (would-be) philosopher.
5.1.4, p. 248 good observations on the complexity of the sentence as a mirror of the complexity of the toga, but add, with Säflund 1955: 60n., that it is also a caricature.
5.3.2 inhumano: no notice taken of Kroymann's interpretation apud Bulhart, or of TLL 7,1.1605.82-1606.2.
cruci: this is the emendation of Salmasius, who explained that when the tongs (forcipes) are opened they form a cross, or x. This might have been mentioned, for TLL 4.1259.67 didn't miss the point.
5.4.2 nihil = non, but no n. See Bulhart.
I conclude these observations with a curious feature of the 'speech', of which H. is aware: the number of the second person shifts between plural and singular. The speech begins with the plural exclusively. Leaving aside instances of the indefinite second person at 2.2.4, 3.2.3, and 4.2.3, we find a singular imperative at 2.4.1 aspice. but revert to a plural at 2.5.2 apud uos. At 2.6.3 another singular, intelligis (hardly indefinite), followed by denotas at 2.7.4, putas at 3.3.1, and indeed a cluster of singulars in that context. Then suddenly at 3.5.1 and 3 back to the plural, de uestro, uos. Back to singular at 3.6.3. At 4.1.1-2, 4.3.1, 4.5.2 the plural prevails. At 4.6.2 the singular pops up, habes, and continues at 4.9.1 and 5. If I'm not mistaken, H.'s first comment on this shift is at 4.10.5, p. 241, and it is manifestly wrong, because the singular had been used before this. He notes that it will occur again at 5.2.1. He suggests no reason for the fluctuation, nor does he say whether it is found in any other of Tertullian's supposed speeches. (There is a further n. on 6.1.1, p. 284.) I begin to wonder if Pall. was ever delivered at all. Possibly suprascriptis at 5.5.4 gives the game away. It might have been useful if a part of the Introduction had been dedicated to this whole issue.
It will be gathered from what has been written above that this commentary often leaves the reader in the lurch, by inadequate or misleading discussion. Hunink has a solid track record as a commentator on out-of-the-way literary works, which nonetheless deserve attention, and I approached this book with high expectations. On balance, however, I now deem it to be a missed opportunity.