Hostname: page-component-89b8bd64d-x2lbr Total loading time: 0 Render date: 2026-05-08T10:01:27.003Z Has data issue: false hasContentIssue false
  • English
  • Français

MASSIS AMERINA NON PERVSTIS (STAT. SILV. 1.6.18): ANOTHER ITALIAN PASTRY?

Published online by Cambridge University Press:  07 August 2023

Darcy A. Krasne
Darcy A. Krasne*
Affiliation:
Columbia University
*
dk3009@columbia.edu
Rights & Permissions [Opens in a new window]

Abstract

This article proposes that untethering amerina at Stat. Silv. 1.6.18 from Pliny's mention of varieties of apples and pears called Amerina allows us to read the line as instead referring to a type of pastry originating in Umbrian Ameria, which is within ancient naming practices for pastries and fits better into the context of the catalogue in which the line occurs. In this case, the second half of the catalogue is closely akin to the crustulum et mulsum donative of wealthy Italian patrons in the early empire.

Keywords

SaturnaliaSilvaepastryStatiusDomitianmassaimported goodsgastronomy

Information

Type
Research Article
Information
The Classical Quarterly , Volume 73 , Issue 1 , May 2023 , pp. 293 - 303
Check for updates
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of The Classical Association

Amidst the opening festivities of Domitian's Saturnalian games on the Kalendae Decembres, a bountiful hail showers down on the crowd packing the seats of the Flavian Colosseum (Stat. Silv. 1.6.9–20):Footnote 1

vix Aurora novos movebat ortus,
iam bellaria linea pluebant:
hunc rorem veniens profudit Eurus.
quicquid nobile Ponticis nucetis,
fecundis cadit aut iugis Idymes;
quod ramis pia germinat Damascos,
et quod percoquit †aebosia† Caunos,
largis gratuitum cadit rapinis;
molles gaïoli lucuntulique
et massis Amerina non perustis
et mustaceus et latente palma
praegnantes caryotides cadebant.

9 movebat ς vovebat M | 10 bellaria ς vellaria M | 11 profudit vulg. profudi M | eurus M eous M 1 in marg. | 15 et quo percoquit aebosia caunos M quod ς aestuosa Imhof | 16 cadit ς cadet M | 17 luguntulique M | 18 perustis ς perustus MA | 19 mustaceus ς muscaceus M | 20 pregnantes ς pregnates M | cariotides ς cariatides MM1

Dawn was scarcely starting on her new rising as treats were already raining from the net:Footnote 2 this dew, a rising East Wind poured forth. Whatever falls noble from the fertile Pontic nut groves or hills of Palestine, what pious Damascus grows in her branches and what †Ibizan†Footnote 3 Caunus ripens, falls freely given for bountiful plunder. Soft ‘gaioli’ and ‘lucuntuli’, and ‘Amerians’ of unscorched substance, and must-cake, and bulging nut-dates, their palm out of sight, kept falling.Footnote 4

While the purpose of this article is to question a tacit assumption of all interpretations of line 18, to do so we must consider the catalogue's brief entirety.Footnote 5 The early contents of the edible bounty are easily identifiable, even if the text is still subject to dispute: hazelnuts,Footnote 6 dates, prunes and dried figs. (Not, I would observe, unqualified ‘plums’ and ‘figs’, as English notes often misleadingly indicate;Footnote 7 in December, there are fresh versions of neither to be had.Footnote 8) With the possible exception of the figs,Footnote 9 these all come from the East (cf. also Eurus, 1.6.11); some scholars have argued for imperializing overtones to be read therein.Footnote 10

It is the second half of the short catalogue, consisting mostly of edibles from closer to home, that presents us with a series of problems. Where the nuts and fruits of the first half, periphrastically presented as riddles by the poet, were written to be solved, the very familiarity of the sweets offered up in more straightforward terms to Statius’ audience—gaioli, lucuntuli, amerina, mustaceus, caryotides—threatens to leave them as enigmas for us. Some, fortunately, have solutions. Most easily laid to rest is the mustaceus, as Cato has conveniently left us a recipe for this cake made of wheat flour and grape must mixed with cheese, anise and cumin, and baked with a laurel leaf pressed into the underside.Footnote 11 The caryotides, conversely, raise debate despite a certain familiarity: assuming (as all do) that caryotides and caryotae are the same, we know that they were the fruit borne by a type of palm tree,Footnote 12 possibly distinct from dates,Footnote 13 that may have been common theatre snacks.Footnote 14 But does praegnantes indicate that they were stuffed,Footnote 15 especially plump,Footnote 16 hidden in an outer shell,Footnote 17 or something else? Furthermore, is there in fact no difference between caryotides and caryotae? Apart from Statius and Martial, the only other author to use the term is Dioscorides, who differentiates them from both unripe and ripe regular dates,Footnote 18 but without further discussion. What was clear and familiar to Statius’ contemporaries is opaque to us.

With gaioli and lucuntuli, we enter semi-transparent waters. gaioli are mentioned nowhere else, but, as the word appears to be a diminutive of Gaius, the universal assumption of modern commentators is that these are, essentially, the Roman equivalent of gingerbread men (minus the ginger)Footnote 19 or possibly bird-shaped pastries instead (gaius meaning a jay).Footnote 20 We assume that we understand the name; we contextually extrapolate the foodstuff.Footnote 21 Several authors, conversely, mention lucuntuli, or lucunculi, or the non-diminutive lucuns:Footnote 22 they are unquestionably a type of pastry, and we even know from a stray reference preserved in Athenaeus that the recipe included cheese.Footnote 23 Their name, however, is more complicated, and its pairing with gaioli adds a further wrinkle. The word seems to be of Etruscan origin, given its -uns termination;Footnote 24 Gagé floats the tentative hypothesis that it could be connected with the noble or religious title lucumo.Footnote 25 What the Romans themselves thought is, perhaps, a different question: Titinius (fr. 166), for instance, seems to expand the word into luculentaster, suggesting ‘fancy cake’, from luculentus, although this apparent etymology may equally well be a figment of manuscript transmission.Footnote 26 Malamud, however, points out the similarity of gaioli lucuntulique to the generalizing ‘Gaius and Lucius’,Footnote 27 and despite the fanciful etymology of this resemblance, the apparent kinship between an Etruscan-originating lucuntulus and Lucius can be granted a firmer place in antiquity as well: the short fragment that we have of a late antique Liber de praenominibus, which seems to draw its information from Republican sources such as Varro and Verrius Flaccus, preserves an imagined connection between the title Lucumo and the name Lucius.Footnote 28 Furthermore, the pairing of these two pastries in a single line links them in a way that occurs with no other foodstuff in the catalogue; they are meant to be read as a group. At the very least, whether or not we accept Malamud's idea that ‘the food … bears a punning resemblance to those who consume it’,Footnote 29 we can be fairly secure in believing that the second half of the catalogue opens with two pastries whose names and origins are drawn from close to home, Latium and Etruria.Footnote 30

The long-standing interpretation of our remaining line, et massis amerina non perustis (18), is that it refers to apples or pears from the Umbrian town of Ameria (modern Amelia), since Pliny preserves for us mention of a late-ripening variety of apples called Amerina (HN 15.58) and a late-ripening variety of pears called Amerina (HN 15.55). While there are several different subsidiary interpretations deriving from this point, it is the underlying assumption that amerina refers to apples or pears—or, indeed, to fruits at all—that I wish to call into question.Footnote 31 Our ‘decoded’ catalogue so far looks like this:

12 hazelnuts (or walnuts?)Footnote 32 from the Black Sea
13 dates from Palestine
14 dried damsons from Damascus
15 dried figs from Caunus (in southwest Anatolia)
16 intermezzo: ‘a compliment to Domitian's generosity’Footnote 33
17 Roman ‘little Gaius’ pastries and Etruscan cheese-pastries
18 massis amerina non perustis
19 must-cakes
20 stuffed(?) nut-dates

For line 18 to refer to fresh fruits, as is the interpretation of a number of commentators, whether ripe or (odder still) unripe,Footnote 34 makes little sense within the flow of this catalogue. A few have, however, rejected this interpretation, pointing out that massis can hardly be reasonably understood, in its regular usage, to apply to the flesh of fresh fruit:Footnote 35 as Lafaye says, and Liberman re-emphasizes, ‘massa ne peut pas désigner autre chose qu'une agglomération de fruits empilés et serrés en masse compacte’.Footnote 36 If we must indeed take amerina to mean apples or pears, then we should certainly follow those interpretations which understand this line as referring in some fashion to dried fruit.

However, the key fact that seems to have eluded all commentators is that the term amerina should be just as immediately familiar to Statius’ audience as gaioli, lucuntuli, mustaceus and caryotides.Footnote 37 Unless dried Amerian fruit-clusters (vel sim.) were so common a Saturnalian bellarium that the unspecified epithet could evoke the same sense of instantaneous recognition as the other items in the catalogue's second half,Footnote 38 then they have no place in this catalogue. While it is impossible to know whether this is the case, I would propose that the structure of the catalogue is itself a piece of evidence in suggesting that dried Amerian fruit-clusters do not, in fact, belong in line 18.

As lines 17 and 19 both feature pastries, if we did not have Pliny's mention of pira Amerina and mala Amerina, the most obvious way for us (and I stress us, modern readers, without our ability to recognize instantly to what amerina refers) to interpret line 18 would be as referring to a type of pastry as well.Footnote 39 The result would be an unbroken string of local manufactured delicacies filling the second half of the catalogue, at least through line 19 (but possibly including line 20, depending on the implication of praegnantes), just as lines 13–15 proffer an unbroken string of imported dried fruits. Rome's imperial spread imports a bounty of edible wealth,Footnote 40 but traditional Italian rustic (or gourmet) ingenuity creates an edible and familiar bounty,Footnote 41 closely akin to the popular Italian donative of wealthy patrons, crustulum et mulsum.Footnote 42 Is it possible that this is, in fact, the way in which an ancient reader would have understood the line, and thus the catalogue? My goal here is to show that this is possible, and even plausible, although it must remain a speculative exercise; and it is, moreover, a logical way to make sense of all three confusing elements of line 18: the regional epithet amerina, the substance referred to by massis, and the degree of doneness indicated by non perustis. I shall address each of these in turn.

First, amerina. While our state of knowledge of ancient pastry is exceptionally paltry, we do have a reasonable collection of pastry names preserved.Footnote 43 It is therefore clear that calling a pastry after its place of origin has substantial precedent in antiquity: we can, for instance, compare the ‘Canopics’ (Κανωπικά) and ‘Cappadocian’ (Καππαδοκικόν) in Athenaeus’ list of cakes drawn from Chrysippus of Tyana (Ath. Deipn. 14.647c);Footnote 44 conversely, the same list's ‘Sabine pastry’ (Σαβελλικὸν κλοῦστρον) includes the word for pastry, showing that the other terms were not simply abbreviations.Footnote 45 Samiae, too, appear to be cakes of Samian origin.Footnote 46 The same, of course, is equally true of other regional manufactured foodstuffs in antiquity; Lucanian sausages, for instance, are simply Lucanica.Footnote 47

What of massis? massa is a loanword from the Greek μᾶζα, which does refer primarily to dough, or even bread, itself deriving from the verb μάσσω, ‘to knead’.Footnote 48 In Latin, however, the specific meaning of flour-and-water dough is limited to ecclesiastical, or at least late, Latin.Footnote 49 In the early days of its Latinate usage,Footnote 50 it has the general meaning of a coagulated and often still-malleable lump of material, and when applied specifically to edible material, it clearly means a homogeneous paste—our surviving references apart from this line are limited to cheeseFootnote 51 or spiced fig paste.Footnote 52 Thus, if the amerina are in fact fruit, they should be fruit that is processed substantially more than commentators have allowed, even those who argue for massis referring to dried and compressed fruit: Columella's fig massa (Rust. 12.15.3–5) consists of figs that have been dried, ground up and mixed with spices.Footnote 53 But there are two points to consider, beyond the plausible fact that many usages may simply not be attested.Footnote 54 First, pastries need not be made of flour-and-water dough (and indeed most were not, in antiquity),Footnote 55 and the coagulated and homogeneous substance that was the product of mixing and kneading should most assuredly be within the word's semantic range, certainly more than any unprocessed fresh or dried fruit. Second, the word's Greek origins clearly do not disappear, or massa would not reclaim its Greek meaning in later antiquity; and we must remember, furthermore, that Statius is the son of a Greek grammarian hailing from the Greekish city of Naples.Footnote 56

As for non perustis, while the verb is rarely preserved in cooking contexts (since one does not tend to intentionally burn food), we do find it once in Scribonius Largus (122) and once in Pliny (HN 24.110); the Plinian usage is especially informative, as it contrasts peruro with torreo, both intentional procedures, but the former implying an even more thorough scorching than torreo provides.Footnote 57 Interpretations thus far of Statius’ non perustis have been forced to a greater or lesser degree, following from the need to understand amerina as referring to fruit. Of those who take amerina as referring to fresh fruit, Vollmer, for instance, imagines that, because Amerian apples and pears ripen with frost (Plin. HN 15.58), non perustis must mean that they are not yet fully ripe (‘sie sind also jetzt noch nicht ganz reif’);Footnote 58 a note in Shackleton Bailey's Loeb volume, conversely, suggests that he takes the phrase to mean that the fruits are fully ripe but have not become overly ripe or damaged by the elements, since they are ‘picked in good time’.Footnote 59 On the other side, we have those commentators who take amerina as referring to dried fruit; in this case, non perustis is generally taken as meaning that the fruit still retains some moisture.Footnote 60 Once again, however, I propose that the context of pastry could give greater intelligibility to the phrase.Footnote 61 Just as Pliny says of lozenges made from acacia gum, which can be toasted or burned still more thoroughly than that (ab aliis torrentur, ab aliis peruruntur, HN 24.110), the same is true of pastries, which are cooked to different degrees of doneness depending on the recipe. Indeed, Varro even claims that the word crustulum is partly derived from uro.Footnote 62

While we must again lament our limited knowledge of ancient pastry techniques, there are a few salutary modern parallels to consider. First is the hyper-local tourteau fromagé of Poitou-Charentes, the method for which requires the top of the cake to be completely carbonized: this would, exempli gratia, clearly be a perustus pastry. Second is the cooking instruction, bis coctus, enshrined in the name of biscuits and biscotti. Although the double baking has been lost from most modern recipes, and although the term itself did not originate in reference to pastries but rather in reference to bread that was baked twice for the sake of longer storage,Footnote 63 nevertheless this mode of preparation produces a hard-baked pastry that might once again be described as perustus, or might simply be tostus, as per Pliny's distinction.Footnote 64 Statius, however, stresses that our hypothetical Amerian pastries are non perustis; we must, therefore, imagine that they are definitely only tostis.Footnote 65 Whether the phrase is simply intended as a litotes or is, rather, a meaningful contrast with a different type of longer-baked amerinum is impossible to say,Footnote 66 but it is certainly important here to note the parallel with the textural epithet molles that is applied to gaioli (and perhaps lucuntuli) in the previous line. The first pastries are soft, while the amerina are harder, just not tooth-breakingly hard.Footnote 67

In light of this interpretation and the textural difference it helps to establish between the (probable) pastries mentioned in the first two lines of the catalogue's second half, it is worth noting a useful parallel with Statius’ longer catalogue of Saturnalian foodstuffs in Silv. 4.9,Footnote 68 a poem that brings the Book 4 addendum to the Silvae to a close just as Silv. 1.6 brought the collection's first book to a close.Footnote 69 Here, we twice find a similar inversion of texture or density between sequentially mentioned related foods: nec lenes alicae, nec asperum far? … non lardum grave debilisve perna (‘no mild-tasting groats, no sharp-tasting emmer wheat? … no heavy bacon or limp ham?’, Stat. Silv. 4.9.31, 4.9.34).Footnote 70 The repeated pairing of foods with antithetical features in both poems is reminiscent of the structural antithesis that defines Martial's Apophoreta, with its alternation of expensive and cheap gifts;Footnote 71 and while Fitzgerald sees this alternation as an expression of Martial's personal predilection for polarity as a rhetorical construct,Footnote 72 Seo argues that questions of reciprocity and asymmetry are embedded in the gift exchange of the Saturnalia;Footnote 73 and Rimell, in turn, sees Martial's poetics as themselves deriving from ‘the mundus inversus of carnival’.Footnote 74 It may be, therefore, that we should stretch that idea of asymmetricality (whether in gift exchanges or in poetic pairings) slightly further to reflect the principle of inversion and reversal that underlies the Saturnalia as a whole. Likewise, the larger structural contrast, in the mini-catalogue of Silv. 1.6, of foreign vs local and my proposed dissection of the catalogue's second half into hyper-local origins resemble Roman's argument that Martial's Xenia and Apophoreta juxtapose ‘food-items and gift-objects [that] are designated by geographical province’ to create a reflection of Rome's status as the centre of global consumption; he likewise proposes that ‘denominazione d'origine controllata was very much part of the idiom of the urban consumer in Rome’.Footnote 75 While Roman's arguments are directed at Martial's own poetics, the similarity with what we find here in Statius suggests the possibility either that such an overt juxtaposition and blending of cultures had become a mark of the Saturnalia (just as the blending of social classes was at its roots)Footnote 76 or that Statius was picking up on Martial's Saturnalian programmatics in his own two Saturnalian compositions,Footnote 77 although without diminishing the force of the complementary local reading that I proposed above.

To return to the amerina themselves, while there is, of course, no evidence for such a pastry existing, we must remember that there is no evidence other than this poem for gaioli, either. For those who find my reasoning too fanciful, let my arguments at least help us lay to rest the idea that massis amerina non perustis refers to fresh fruit. In either case, too, as I have argued above, the familiarity of the names in the second half of the catalogue is critical, and to be contrasted with the imported imperial bounty of the first half:Footnote 78 local snacks from Latium, Etruria and Umbria; a cake that at least forms part of every Roman wedding (Juv. 6.202–3; cf. Anth. Lat. 190.49 Shackleton Bailey), if not other banquets as well; and a snack that, while of imported origin, is so common in Rome as to be notas … theatris (Mart. 11.31.10). The catalogue ends where it began, with imported goods, but it has digested this one into that which is thoroughly Roman.Footnote 79 Without recognizing the importance inherent in the regional denomination amerina, and perhaps even the element of local human labour necessary to produce these treats, we cannot fully grasp the implications of the catalogue's second half. These items are not, as Malamud proposed, representative of those who consume them; rather, I would argue, they serve as a reminder of the sweet rewards to be reaped from peacetime labor when they do not fall free from heaven,Footnote 80 set in distinct contrast with the largis gratuitum … rapinis (Silv. 1.6.16) that can be reaped as the fruits of empire.

References

1 I print here a text and apparatus criticus derived from that of Marastoni's 1961 Teubner edition. I have condensed his reporting of individual manuscripts other than our earliest manuscript, M (Matritensis 3678), to the siglum ς, following common practice (see B. Gibson, Statius, Silvae 5 [Oxford and New York, 2006], lii). Works repeatedly cited throughout are as follows: André, J., L'alimentation et la cuisine à Rome (Paris, 1961)Google Scholar; Kreuz, G.E., Besonderer Ort, poetischer Blick: Untersuchungen zu Räumen und Bildern in Statius’ Silvae (Göttingen, 2016)CrossRefGoogle Scholar; Lafaye, G., Quelques notes sur les Silvae de Stace: premier livre (Paris, 1896)Google Scholar; G. Liberman, Stace: Silves. Édition et commentaire critiques (Paris, 2010); Malamud, M., ‘That's entertainment! Dining with Domitian in Statius’ Silvae’, Ramus 30 (2001)10.1017/S0048671X00001569CrossRefGoogle Scholar, 23–45; C.E. Newlands, Statius’ Silvae and the Poetics of Empire (Cambridge, 2002); Thomas, F., ‘Autour d'un passage de Plaute: Mén. 141 sqq.’, in Hommages à Léon Herrmann (Brussels and Berchem, 1960), 705–14Google Scholar; Vollmer, F., P. Papinii Statii Silvarum libri (Leipzig, 1898)Google Scholar.

2 On the appearance and operation of the linea dives, see Killeen, J.F., ‘What was the linea dives (Martial, VIII, 78, 7)?’, AJPh 80 (1959), 185–8Google Scholar.

3 Translating the frequently proposed Ebosea. I do not assume that it is necessarily correct.

4 Several of the translations in this last sentence involve debated interpretations that I discuss below.

5 See Malamud (n. 1), 23–7 and Newmyer, S.T., The Silvae of Statius: Structure and Theme (Leiden, 1979), 111CrossRefGoogle Scholar, briefly, on its structure.

6 For nuces Ponticae as one name for hazelnuts, see Plin. HN 15.88. Is it possible, however, that nobile hints at the Greek βασιλικόν, thus playfully suggesting walnuts (κάρυα βασιλικά) that simply have a Pontic origin? Pliny, again: optimum quippe genus earum Persicum atque basilicon vocant, et haec fuere prima nomina (‘In fact, they call the best kind of these nuts [i.e. walnuts] “Persian” and “royal”, and these were their original names’, HN 15.87).

7 e.g. D.R. Shackleton Bailey and C.A. Parrott, Statius: Silvae (Cambridge, MA and London, 20152), 69 nn. 3–4; Newlands (n. 1), 241; A. Wasserstein, ‘An edition of Book I. of the Silvae by Publius Papinius Statius with notes, critical and explanatory’ (Diss., University of London, 1951), 222.

8 On ancient preservation methods, see André (n. 1), 87–93; for dried damson plums at the Saturnalia, see e.g. Mart. 5.18.3, 13.29; see also Coleman, K.M., Statius: Silvae IV (Oxford, 1988), 231–2Google Scholar.

9 Lafaye (n. 1), 65 argues for reading cannos in place of Caunos and accepting the emendation Ebosia; in this case, they would come from Spain.

10 Malamud (n. 1), 25 and Newlands (n. 1), 241.

11 Cato, Agr. 121; for the laurel leaf, evidently a defining feature of the mustaceus, see also Cic. Att. 5.20.4 and Plin. HN 15.127.

12 Varro, Rust. 2.1.27 non scitis palmulas careotas Syrias parere in Iudaea, in Italia non posse? (‘Are you unaware that Syrian palms bear careotae [or: that palms bear Syrian careotae] in Judaea, but cannot in Italy?’); Plin. HN 6.205 palmetis caryotas ferentibus (‘caryotae-bearing palm tree groves’), 13.44–5.

13 Lafaye (n. 1), 66–7, relying in part on a description of the modern Caryota genus of palms, asserts that caryotides were a fruit that grew on neither the date palm nor the coconut palm but on a distinct species. He proposes, based primarily on a recipe in Scribonius Largus (74) that calls for the pulp of fresh caryotae with the outer skin and inner veins removed, that palma refers to the inner pulp and caryotis to the outer hull. Vollmer (n. 1), 306 ad loc. disagrees. Pliny tells us that, of the various fruits produced by palm trees, caryotae are the fifth most famous, thanks to their exceptional honey-like juice (HN 13.44). He also mentions several slightly inferior fruits that are in the same category (genus) as caryotae but have different names (the dryer but very large nicolaus, the less attractive but sweet adelphis and the exceptionally juicy pateta, HN 13.45). Pesando, F., ‘Tutti frutti su qualche periodo di raccolta e su qualche frutto esotico nell'antichità’, ArchClass 67 (2016), 629–44Google Scholar, at 639–40 has recently argued that these fruits are included as subtypes of caryota owing to a Hellenistic extension of the species name's original meaning; Athenaeus (14.652a–b) and Plutarch (Quaest. conv. 723A–D), however, simply refer to the nicolaus as a type of date.

14 They are notas … theatris (Mart. 11.31.10, with N.M. Kay, Martial Book XI: A Commentary [London, 1985], 139 ad loc.; André [n. 1], 84); this is the only other Latin use of caryotis, as opposed to caryota. Wrapped in gold leaf, caryotae seem also to have been a cheap gift given at the Kalends of January (Mart. 8.33.11–12, with T.J. Leary, Martial: Book XIII. The Xenia [London, 2001], 77–8 on Mart. 13.27; André [n. 1], 84).

15 Malamud (n. 1), 26; Briand-Ponsart, C., ‘Les «lancers de cadeaux» (missilia) en Afrique du Nord romaine’, AntAfr 43 (2007), 79–9710.3406/antaf.2007.1419CrossRefGoogle Scholar, at 81.

16 Vollmer (n. 1), 306.

17 Lafaye (n. 1), 67.

18 At Diosc. 1.109.2, καρυώτιδες come at the end of his discussion of uses for the fruit of date palms; at Diosc. Eup. 2.51.6, they are given as an alternative ingredient to φοινικοβάλανοι (‘dates’). However, at Diosc. Eup. 2.31.7, he also refers to καρυωτοί, the Greek equivalent of caryotae.

19 This theory of human-shaped pastry goes back at least as far as Calderini's annotations from 1476.

20 TLL 6.2.1669.16, 6.2.1669.32–45. For discussion of our limited knowledge of Roman shaped or moulded pastries, with bibliography, see Simon, I., ‘Un aspect des largesses impériales: les sparsiones de missilia à Rome (Ier siècle avant J.-C. - IIIe siècle après J.-C.)’, RH 310 (2008), 736–88Google Scholar, at 768 n. 30.

21 A. Souter, ‘Vollmer's Statius’ Silvae’, CR 12 (1898), 314–15 proposed reviving an old reading caseoli instead, but this disrupts the pattern of precise types given in the second half of the catalogue. Some early commentators, such as Poliziano (who read here caioli lagunculique), proposed that this line referred to coins.

22 Afran. com. 162 (although the spelling here is uncertain; see Thomas [n. 1], 708); Varro, Sat. Men. 417, and again at Sat. Men. 508, possibly parodying Lucilius (see F. Bücheler, ‘Bemerkungen über die varronischen Satiren’, RhM 14 [1859], 419–52, at 426); Apul. Met. 10.13; Paul. Fest. 106.27 Lindsay; Thomas (n. 1), 706–7 collects several additional probable references. See further below, with nn. 25–6.

23 Ath. Deipn. 14.657d, who gives the ‘pastry scholar’ (πεμματολόγος) Chrysippus of Tyana as his source. For a further discussion of the probable recipe and its defining features, see Thomas (n. 1), 707–8.

24 T. Georgescu, ‘Le latin à la cuisine: lagana, qui vulgo dicitur lasagna… Quelques noms de pâtisseries en latin vulgaire’, in P. Molinelli, P. Cuzzolin and C. Fedriani (edd.), Latin vulgaire - Latin tardif X. Tome II: Sémantique - Lexique - Textes et contextes (Bergamo, 2014), 557–66, at 559; Thomas (n. 1), 706.

25 Gagé, J., ‘Arruns de Clusium et l'appel aux Gaulois (?): à propos d'une tradition haruspicinale sur la vigne et l'olivier’, RHR 143 (1953), 170–208CrossRefGoogle Scholar, at 206 n. 1. Lindsay, W.M., ‘Lucuns. Lucuntulus’, ALL 11 (1900), 332Google Scholar, however, thinks that the word is of Greek origin and argues for a derivation from ‘λυκόεις (in the sense of λυκοειδής)’; this is by contrast with the favoured derivation at the time of his article, from γλυκοῦς. The quantity of the first syllable is in dispute: it sometimes is definitively short in Latin and sometimes is ambiguous, leading to Lindsay's proposed etymology, but the Greek spelling (which Lindsay does not adduce) is λούκουντλοι. (It is unknowable whether there is any relevance to the word λοῦκα, preserved by Hesychius [Λ1269], meaning a porridge made of barley and attributed to the Kaukones.) On the syllable length, see also Thomas (n. 1), 708.

26 On the hypothetical connection, see Heraeus, W., ‘Die Sprache des Petronius und die Glossen’, in W. Heraeus, Kleine Schriften von Wilhelm Heraeus (Heidelberg, 1937), 52–150Google Scholar, at 150 n. 1; Lindsay (n. 25), however, argues that the transmitted form luculentaster is a mistake for lucuentaster, in which case no such derivation is implied. But see Thomas (n. 1), 705–9 on the preference for keeping lucul- in this and similar passages.

27 Malamud (n. 1), 25.

28 De praen. 5 (Lucii coeperunt adpellari … ut quidam arbitrantur, a Lucumonibus Etruscis, ‘Luciuses began to be called, as some think, from the Etruscan Lucumos’); see also Farney, G.D., ‘The name-changes of legendary Romans and the Etruscan-Latin bilingual inscriptions: strategies for Romanization’, EtrStud 13 (2010), 149–57Google Scholar.

29 Malamud (n. 1), 25.

30 Gaius is ‘quintessentially Roman’ (Farney [n. 28], 151).

31 With the exception of those who emend massis to passis and understand the line as referring to dried grapes rather than to apples or pears (a proposal of Heinsius, for which see Liberman [n. 1], 147), no one appears to have broken away from the Plinian route. However, Kreuz (n. 1), 186 n. 517 comes close to my argument while retaining a traditional interpretation of amerina; see below.

32 See n. 6 above.

33 Newmyer (n. 5), 111.

34 See nn. 58–9 below.

35 As Liberman (n. 1), 147 observes, the TLL's definition for this citation, pulpa pomorum, is in line with neither the heading materia coacta rerum esculentarum nor any of the word's other usages.

36 Lafaye (n. 1), 66; Liberman (n. 1), 147; see also Wasserstein (n. 7), 224.

37 The approach of Kreuz (n. 1), 185–7 to amerina is particularly egregious: as he wants the entire catalogue to be a series of goods ex oriente—which is manifestly inaccurate given the clearly rustic Italian origins of the mustaceus if nothing else—he thus proposes that the local meaning of amerina must fade behind a transferred meaning of ‘late-ripening’.

38 I am not disputing fruit itself as a common Saturnalia gift, nor its common function as a secunda mensa, but rather dried Amerian apples or pears in particular.

39 Kreuz (n. 1), 186–7 n. 517 draws the same conclusion about the progression of the catalogue, as well as also taking massa to mean ‘dough’ (see below), but he assumes that amerina must still refer to apples or pears and thus conjectures something along the lines of an apple dumpling, or ‘eine derart primitive Form der Obstbäckerei’.

40 Cf. Malamud (n. 1), 25: ‘The audience, as they snatch their plunder, re-enact in the symbolic arena of the amphitheatre Rome's acquisition of the world's wealth.’

41 Contra J.F. Donahue, The Roman Community at Table during the Principate (Ann Arbor, 20172), 19, who assumes that ‘the small fancy cakes (gaioli)’ and the treats that follow in the catalogue ‘further underscore the theme of luxury’, with emphasis throughout on ‘the Eastern origins of these foods’.

42 For the crustulum et mulsum, which has Republican origins but was most popular in the first century c.e., see Mrozek, S., ‘Crustulum et mulsum dans les villes italiennes’, Athenaeum 50 (1972), 294–300Google Scholar and Donahue (n. 41), 11, 97–8, 102–3, 123; Donahue also gathers all the inscriptions that mention the crustulum donative.

43 Solomon, J., ‘Tracta: a versatile Roman pastry’, Hermes 106 (1978), 539–56Google Scholar, at 539. Many of the names come from lists at Ath. Deipn. 3.109a–116a, 14.641e–649b and Poll. Onom. 6.11.72–82; André (n. 1), 213–18 collects a fairly comprehensive list of names and ingredients from other sources as well. Georgescu (n. 24), 558–9 sees three types of names for Roman pastries specifically: those deriving directly from Indo-European, those with dialectical/Italian origin and those borrowed from other languages, generally Greek (these are divided into early and late borrowings). Pliny discusses the derivation of various names given to bread at HN 18.105; these include gustatory context, quality, production method, ingredients and national origin.

44 Perhaps also the ‘Terentine’ (Τερεντῖνον), although André (n. 1), 216 assumes that this is named after an inventor Terentius rather than after a location. Canopica are also mentioned at Anth. Lat. 190.48 Shackleton Bailey and in at least two papyri (P.Oxy. 1774, SB 8.9746; perhaps also P.Oxy. 738, if Κανωπικὸν does not modify ἧπαρ). At Ath. Deipn. 4.130d, cakes are divided into Κρητικῶν καὶ … Σαμιακῶν καὶ Ἀττικῶν (‘Cretan and Samian-style and Attic’), each with their own type of container.

45 A pun is created in the other direction at Anth. Lat. 190.32 Shackleton Bailey, where the pastry chef claims to be urbe Placentinus, as though the city of Placentia were named after placenta, ‘cake’; Bailey, D.R. Shackleton, ‘Three pieces from the Latin Anthology’, HSPh 84 (1980), 177–217Google Scholar, at 213 n. 13 notes the parallel jest at Plaut. Capt. 162.

46 André (n. 1), 215; the usage is at Tert. Adv. Marc. 3.5. King, A., ‘A graffito from La Graufesenque and “samia vasa”’, Britannia 11 (1980), 139–4310.2307/525677CrossRefGoogle Scholar, at 140–1 n. 4 argues for samias in Tertullian's text as referring to Samian ware pottery instead, but the context (et terram audimus lacte et melle manantem, non tamen ut de glebis credas te unquam placentas et samias coacturum) strongly implies that samiae, like placentae, are something made with milk and honey. Similarly, Poll. Onom. 6.78 refers to οἱ Σάμιοι πλακοῦντες as famous; that this is their name is suggested both by the context of Pollux's list and by their pairing with οἱ Φιλοξένιοι [sc. πλακοῦντες] (cf. Ath. Deipn. 1.5e).

47 Dalby, A., ‘Goût d'empire’, Kentron 35 (2019), 155–66CrossRefGoogle Scholar argues for a ‘geography of the senses’ evoked in literature through the place name epithets applied to regional foodstuffs from across the empire; while he largely focusses on agricultural products, he does give some attention to manufactured regional specialties (for Lucanian sausages, see 159–60).

48 R.S.P. Beekes, ‘Μᾶζα’, in R.S.P. Beekes, Etymological Dictionary of Greek (online, 2010; consulted 12/12/2020, https://dictionaries.brillonline.com/search#dictionary=greek&id=gr4310) proposes the Indo-European root *meh2ǵ-.

49 TLL 8.430.9–20. Kreuz (n. 1), 186 n. 517 does not appear to be aware that the straightforward meaning of ‘dough’ is only attested as a late usage for Latin; his assumption, therefore, that the massa-substance could be entirely separate from the substance of the amerina themselves is highly unlikely. massis … non perustis is best construed as an ablative of quality.

50 Georgescu (n. 24), 559 dates the word's borrowing from Greek to the first century b.c.e.

51 Of our four surviving references, two specify that the massa is composed of coagulated milk (lactis massa coacti, ‘a mass of congealed milk’, Ov. Met. 8.666; massam modo lactis alligati, ‘a mass of just-curdled milk’, Mart. 8.64.9), while Martial's Velabrensi massa coacta foco (‘a mass congealed on a Velabran hearth’, 11.52.10), referring to smoked cheese, helps to clarify that, at least here, massa does not simply mean something like ‘large portion’ but most likely refers to the actual substance (similarly, C. Schöffel, Martial, Buch 8: Einleitung, Text, Übersetzung, Kommentar [Stuttgart, 2002], 537 on Mart. 8.64.9). The final example, Vestino de grege massa (‘a mass from a Vestine flock’, Mart. 13.31.2), is more ambiguous (Shackleton Bailey and Parrott [n. 7], 185, for instance, translate massa here as ‘hunk’), but the usage is not so far from the other examples as to preclude referring to the substance of the cheese itself.

52 TLL 8.430.1–7. Columella's usage at Rust. 12.34 (sed mustum desub massa et limpidum sit), included by the TLL in this list of edibles, is mysterious; André, J., Columelle: De l'agriculture. Livre XII (De l'intendente) (Paris, 1988), 68 n. 3Google Scholar explains desub massa as defining the type of must used to produce embamma as ‘mèregoutte’, or the must that is produced before pressure is applied to the grapes. In this case, massa would refer to the mass of grapes that is pulped to make wine.

53 Thus massa would be quite distinct from the globus of prunes and little figs packed into a collapsing cone at Stat. Silv. 4.9.27–8.

54 Cf. e.g. the probable meaning of fossa at Stat. Silv. 5.3.54, with Gibson (n. 1), 287 ad loc., elsewhere unattested.

55 In parallel terms, the Cambridge Greek Lexicon (2021) notes that μᾶζα was ‘made of roasted barley-meal, mixed w. water, milk, wine or oil’.

56 quasi Graecam urbem (Tac. Ann. 15.33). On Statius’ father and Naples’ Hellenism, see McNelis, C., ‘Greek grammarians and Roman society during the early empire: Statius’ father and his contemporaries’, ClAnt 21 (2002), 67–94Google Scholar, at 73–4, with further bibliography. Gibson (n. 1), 346–7 points to a possible Graecism by Statius at Silv. 5.3.210; cf. also his predilection for the Greek chelys (A. Keith, ‘Lyric resonances in Statius’ Achilleid’, in F. Bessone and M. Fucecchi [edd.], The Literary Genres of the Flavian Age [Berlin and Boston, 2017], 283–95, at 291). Leiwo, M., Neapolitana: A Study of Population and Language in Graeco-Roman Naples (Helsinki, 1995), 51–7Google Scholar discusses linguistic variation in Neapolitan Greek owing to dialect and bilingualism.

57 Scribonius Largus, conversely, uses peruro to mean ‘become overly burned’.

58 Vollmer (n. 1), 306. Newlands (n. 1), 241 points out the strangeness of unripe apples and pears even while accepting the explanation; Kreuz (n. 1), 186 n. 517, conversely, argues that unripe fruit is both improbable for the time of year and inappropriate for the panegyric context.

59 Shackleton Bailey and Parrott (n. 7), 69 n. 6. J. Markland, P. Papinii Statii libri quinque Silvarum (Dresden and London, 1827 [1728]), 211 likewise follows the early commentator Gronovius in understanding perustus as meaning ripe but not dried.

60 Lafaye (n. 1), 66; Frère, H., Stace: Silves. Tome I (Livres I–III) (Paris, 1944), 47Google Scholar n. 3. Alternatively, Liberman (n. 1), 147 proposes emending non perustis to sole tostis, which seems quite a palaeographical stretch and in contravention of the lectio difficilior rule.

61 Kreuz (n. 1), 187 n. 517 proposes emending to nuper ustis, meaning ‘freshly baked’; this would still be a possible emendation and meaning if we follow my slightly different line of thinking, but, as I discuss below, the emendation is unnecessary and may even obscure a structural feature of the catalogue.

62 Varro, Ling. 5.107 crustulum … quod ut corium et uritur (‘little crust … because it is like a hard shell and is burned’).

63 OED s.v. ‘biscuit’; Vocabolario degli Accademici della Crusca (Florence, 16913), 225.

64 It is worth noting that the cantucci and tozzetti of Etruria and Umbria (home of Ameria) do still require a double baking. This style of biscuit, while likely not of ancient origin, was certainly already in existence prior to the late seventeenth century, as it is given as one definition of cantuccio in the third edition of the Vocabolario degli Accademici della Crusca (n. 63), 276.

65 Kreuz (n. 1), 186–7 n. 517 aptly points out that it would be inappropriate to a panegyric context for non perustis simply to mean that the bakers had successfully avoided burning the pastry.

66 For two different regional products, we may compare the hard-baked copta Rhodiaca (Mart. 14.70) with the soft ἐσχαρίτης Ῥοδιακός, μεθόριος ἄρτου καὶ πλακοῦντος (‘Rhodian hearth-bread, between bread and cake’, Poll. Onom. 6.78; cf. Ath. Deipn. 3.109d–e). Rhodes also produced a variety of echinus (Ath. Deipn. 14.647a–b), which was a cake endemic to the islands (νησιωτικὸς πλακοῦς, Poll. Onom. 6.78).

67 Vollmer (n. 1), 306 supposes that molles is meant to draw a contrast between edible gaioli and the clay sigillata (‘Thonfigürchen’) that are common Saturnalia gifts, rather than that it indicates anything about the actual texture of the gaioli. However, cf. Mart. 14.69–70, where a Priapus of soft wheat flour (siligineus) is juxtaposed with a hard-baked copta Rhodiaca; thanks to Ana Lóio for drawing my attention to the parallel.

68 I owe the inspiration for this paragraph to the anonymous reviewer for CQ. As Nauta, R., Poetry for Patrons: Literary Communication in the Age of Domitian (Leiden / Boston / Köln, 2002), 399CrossRefGoogle Scholar notes, Statius’ list of foodstuffs in our present passage ‘reads like a list of typical Saturnalian presents’.

69 For the respective publication dates of the books of the Silvae and the concept that Books 1–3 are a unit, see Coleman (n. 8), xvi–xix and, differently, Nauta (n. 68), 285–90, 443–4; Book 5 of the Silvae is typically considered to have been published posthumously (see Coleman [n. 8], xxxi; Gibson [n. 1], xxviii–xxx). On the shared closural nature of Silvae 1.6 and 4.9, see Newlands (n. 1), 257; Seo, J.M., ‘Statius Silvae 4.9 and the poetics of Saturnalian exchange’, MD 61 (2009), 243–56Google Scholar, at 253–4; Gowers, E., ‘Lucan's (G)natal poem: Statius’ Silvae 2.7, the Culex, and the aesthetics of miniaturization’, ClAnt 40 (2021), 45–75Google Scholar, at 47.

70 The manuscripts read leves rather than lenes at 4.9.31, which would make the comparison one of actual texture rather than mouthfeel; graves occurs again at 4.9.35 in the manuscripts, but Coleman (n. 8), 234 emends to breves (citing the apparently short and plump shape of Faliscan sausages).

71 Including the hard copta Rhodiaca and the soft Priapus siligineus; see n. 67 above.

72 Fitzgerald, W., Martial: The World of the Epigram (Chicago, 2007), 110CrossRefGoogle Scholar.

73 Seo (n. 69). B.L. Ullman, ‘Apophoreta in Petronius and Martial’, CPh 36 (1941), 345–55, at 348 likewise notes the prevalence of pairing in Petronius’ apophoreta game at Sat. 56.

74 Rimell, V., Martial's Rome: Empire and the Ideology of Epigram (Cambridge, 2008), 141Google Scholar.

75 Roman, L., ‘Martial and the city of Rome’, JRS 100 (2010), 88–117Google Scholar, at 94–5. On Rome's global consumerism in Martial's Saturnalian poetry, see also Blake, S., ‘Martial's natural history: the Xenia and Apophoreta and Pliny's encyclopedia’, Arethusa 44 (2011), 353–77CrossRefGoogle Scholar; Rimell (n. 74), 144–5. For Statius’ own showcasing of regional foods in a Saturnalian context, we may observe his inclusion at Silv. 4.9.35 of Lucanian sausages (see page 299 above) and a second type of sausage, Falisci (elsewhere called Falisci ventres, Mart. 4.46.8, Varro, Lang. 5.111), which likewise took their name from their Falerian origin (Varro, Lang. 5.111). Newlands (n. 1), 244–5 also makes the case for an ethos of global consumerism in Silvae 1.6.

76 The riddling first half vs transparent second half is also Saturnalian: on the connection of riddles with the Saturnalia, see W.R. Johnson, ‘Small wonders: the poetics of Martial, Book Fourteen’, in W.W. Batstone and G. Tissol (edd.), Defining Genre and Gender in Latin Literature: Essays Presented to William S. Anderson on his Seventy-Fifth Birthday (New York, 2005), 139–50, at 141–2; Blake (n. 75); U. Egelhaaf-Gaiser, ‘Saturnalian riddles for Attic nights: intratextual feasting with Aulus Gellius’, in S. Harrison, S. Frangoulidis and T.D. Papanghelis (edd.), Intratextuality and Latin Literature (Berlin and Boston, 2018), 431–47. On ludic poetry at the Saturnalia more generally, see Citroni, M., ‘Marziale e la letteratura per i Saturnali (poetica dell'intrattenimento e cronologia della pubblicazione dei libri)’, ICS 14 (1989), 201–26Google Scholar.

77 Martial's Xenia and Apophoreta were published between 83 and 85 c.e.; for the dating, see Citroni, M., ‘Pubblicazione e dediche dei libri in Marziale’, Maia 40 (1988), 3–39Google Scholar, at 11–12; Leary, T.J., Martial Book IV: The Apophoreta (London, 1996), 9–12Google Scholar; Leary (n. 14), 12–13; Nauta (n. 68), 441.

78 Cf. Newlands (n. 1), 241.

79 It may be that we should compare the fictionality of Roman acquisition and assimilation visible in Martial's epigram on the British bascauda, which Rome has evidently adopted as a local product: barbara de pictis veni bascauda Britannis, | sed me iam mavult dicere Roma suam (‘A barbarian basket, I came from the tattooed Britons, but now Rome prefers to say that I am her own’, Mart. 13.99). Cf. Mart. 13.42–3.

80 For pastry—or dough involving butter—as the product of peacetime, see Gowers, E., The Loaded Table (Oxford, 1993), 17Google Scholar; Plin. HN 18.105–6.