Aldhelm’s Aenigmata, Greek riddles, and the Hisperica famina

Abstract

When Aldhelm came to compose a collection of Latin riddles in the late seventh century, the riddle was already an established literary genre in Greek and Latin. Although Aldhelm’s main source was the Latin Aenigmata of Symphosius, he introduced a number of innovations that transformed the genre. To account for these innovations, it has been suggested that Aldhelm also knew and was influenced by Greek riddles, which are otherwise unattested in Anglo-Saxon England. This article first reviews the evidence for Aldhelm’s knowledge of Greek riddles, especially in his Aenigma 32 about a writing tablet. It then argues that the peculiar features of Aenigma 32 were not derived from Greek riddles but rather from the Hisperica famina, a work that Aldhelm very likely knew. His transformation of the genre therefore can be accounted for by his use of Latin sources available in seventh-century England without appealing to speculative Greek ones.

There is considerable doubt over the extent of Greek learning in early Anglo-Saxon England.¹ Aldhelm (c. 639–709/10), for instance, who was one of the most learned English scholars of his generation,² was a student at the famous Canterbury school under Theodore of Tarsus and Abbot Hadrian, both native speakers of Greek.³ It is likely that Theodore and Hadrian brought Greek books to Canterbury, including copies of the Septuagint, the Greek New Testament and perhaps other Greek texts as well.⁴ A generation later, in 731, Bede could still say that ‘usque hodie supersunt de eorum discipulis, qui Latinam Graecamque linguam aeque ut propriam in qua nati sunt norunt’.⁵ Curiously, Bede never mentioned Aldhelm as one of the students at Canterbury, and no extant writings can be ascribed to the few students that he did name,⁶ although at least one Greek poem was probably translated into Latin at Canterbury, which Aldhelm quoted but did not translate himself.⁷

The main inspiration for Aldhelm’s Aenigmata was the Latin poet Symphosius, whose Aenigmata Aldhelm quoted and whom he mentioned by name.⁸ But Aldhelm may also have known other riddles – notably the Latin Bern Riddles – and it is not impossible that he had encountered Greek riddles at Canterbury before composing his Aenigmata. ⁹ Presumably such Greek riddles would have been translated by Theodore or Hadrian, since there is no evidence elsewhere in Aldhelm’s writings that he had a solid grasp of Greek.¹⁰ But if Aldhelm did know Greek riddles in some form, this would not only transform the study of his works but also be important evidence for Greek learning and the availability of Greek texts in seventh-century England.

And there are reasons to suspect that Aldhelm did in fact draw on Greek riddles. In an article in Anglo-Saxon England, Čecila Milovanović-Barham suggested certain features of Aldhelm’s Aenigmata could be better explained as derived from Greek riddles than from the Aenigmata of Symphosius.¹¹ She noted three general similarities between Aldhelm’s Aenigmata and the Greek riddles of the Anthologia Palatina:¹² both collections share an emphasis on origins and birth; they occasionally use logogriphs that play with words such as corbus/orbus; and they include a few verses that challenge their readers to solve the riddles, such as ‘sciscitor inflatos, fungar quo nomine, sophos’.¹³ But most of her article is concerned with a single riddle by Aldhelm, Aenigma 32 on a writing tablet, which she claimed relied on a particular Greek riddle on the same topic and so reveals the influence of the Greek tradition on Aldhelm.¹⁴ In what follows I should like to examine the evidence for Aldhelm’s use of Greek riddles, and then to suggest what I think are more plausible Latin sources for the peculiar features of Aenigma 32, notably the Hisperica famina. At the end, I shall return to consider the three more general similarities.

Aldhelm’s Aenigma 32 describes the manufacture and use of a pair of writing tablets – pugillares, to give its Latin solution. Following the example of Symphosius, Aldhelm composed this enigmatic poem from the point of view of the writing tablet itself, a rhetorical technique known as prosopopoeia:¹⁵

Melligeris apibus mea prima processit origo,

sed pars exterior crescebat cetera silvis;

calciamenta mihi tradebant tergora dura.

Nunc ferri stimulus faciem proscindit amoenam

5   flexibus et sulcos obliquat adinstar aratri,

sed semen segiti de caelo ducitur almum,

quod largos generat millena fruge maniplos.

Heu! tam sancta seges diris extinguitur armis.¹⁶

The Greek source suggested by Milovanović-Barham for this text is a riddle on the same topic – δέλτος (‘writing tablet’), which appears anonymously in the Anthologia Palatina and later in a collection ascribed to the Byzantine scholar Michael Psellos (c. 1017–1078):¹⁷

Ὕλη μέν με τέκεν, καινούργησεν δὲ σίδηρος·

εἰμὶ δὲ Μουσάων μυστικὸν ἐκδόχιον·

κλειομένη σιγῶ· λαλέω δ′, ὅταν ἐκπετάσῃς με,

κοινωνὸν τὸν Ἄρη μοῦνον ἔχουσα λόγων.¹⁸

Both riddles take the form of a two-part progression from the past to the present, told from the perspective of the solution itself. In Aldhelm’s first three verses, the writing tablet recounts its origin and construction from various materials: wax, wood, and leather (if that is what line 3 means – see below).¹⁹ The five remaining verses then describe how the finished product is used for writing, here portrayed in an agricultural metaphor: the words are seeds, the page is a field, and the pen is a plough. Since the written word for Aldhelm is Holy Scripture, the metaphor can be extended to include the harvest, for these words produce spiritual fruits. Here Aldhelm was drawing on a widespread topos used by pagan and patristic authors alike, as P. D. Scott and Milovanović-Barham both noted.²⁰ But Milovanović-Barham concentrated her study on the riddle’s first three verses, which form the first half of the riddle’s two-part structure and represent Aldhelm’s interest in origins – an interest that for her was evidence of Aldhelm’s knowledge of the Greek tradition.

The same two-part structure characterizes the Greek riddle on a writing tablet. Its first two verses refer to the object’s origin in nature and its subsequent transformation. The final two verses then portray its use as a writing tablet that speaks to Ares only while open. Ares, the god of war, is an enigmatic reference to the iron stylus. Both riddles thus share a martial image for the stylus, as Milovanović-Barham noted, for Aldhelm referred to the stylus ambiguously as ferrum (‘iron stylus’, ‘weapon made of iron’) and then suggested how it also erases with the phrase diris … armis (‘by cruel weapons’).²¹ She then pointed out that Aldhelm’s use of melliger (‘honey-bearing’) in line 1 recalled a similar compound adjective in a different Greek riddle, γλυκυγόνον (‘sweet-bearing’), where it also refers to the production of wax by bees (here for a candle).²² But Milovanović-Barham’s main contention was that Aenigma 32 exemplified Aldhelm’s interest in origins, which she identified with the Greek tradition.²³ For Aenigma 32 does not merely describe the function of a writing tablet but also its creation, and thus shares the same two-part structure as the Greek riddle on the same topic. But are these similarities evidence of Aldhelm’s dependence on the Greek riddle, or can they be explained through his knowledge of Latin sources?

TRANSFORMATION RIDDLES AND EPIGRAMS

Both riddles are examples of the transformation riddle, a type of riddle that portrays the transformation of its subject (which is also its solution) from one thing into another.²⁴ Such riddles typically consist of two parts, the first describing the subject’s prior existence and the second its current existence, usually progressing from the past to the present tense. Almost all transformation riddles were composed in the voice of their subjects (that is, using prosopopoeia), so that they form a lyrical autobiography of the solution from its past to its present situation.²⁵

Such riddles were ultimately modelled on Greek epigrams, many of which share these same features. The same progressive, autobiographical form can be found in many literary epigrams of the Hellenistic era (c. 323–30 bc), as in the following epigram on the κάλαμος (‘reed pen’), which was probably composed sometime during this period:²⁶

Ἤμην ἀχρεῖον κάλαμος φυτόν· ἐκ γὰρ ἐμεῖο

οὐ σῦκ ̓, οὐ μῆλον φύεται, οὐ σταφυλή·

ἀλλά μ ̓ ἀνὴρ ἐμύησ ̓ Ἑλικωνίδα λεπτὰ τορήσας

χείλεα, καὶ στεινὸν ῥοῦν ὀχετευσάμενος.

ἐκ δὲ τοῦ εὖτε πίοιμι μέλαν ποτόν, ἔνθεος οἷα

πᾶν ἔπος ἀφθέγκτῳ τῷδε λαλῶ στόματι.²⁷

Like the Greek riddle on a writing tablet, this epigram consists of elegiac couplets and uses object-personification to recount the history of a reed from its origin as a raw material to its current use as a stylus. Its basic shape is a temporal progression from the past to the present state of the object. Besides writing implements, Greek epigrams portray many different types of objects in this same way, including ships made from trees and weapons taken as spoils.²⁸ The narrative structure of contrasting the past and present states of something is thus found widely in the epigram genre. Some of these epigrams even include a series of temporal adverbs, such as πρίν (‘once’) and νῦν (‘now’), which make the contrast of their two states more explicit.²⁹ The autobiographical structure of these ‘once … now’ epigrams was clearly the model for our Greek riddle on the writing tablet.

All of the riddles in the Anthologia Palatina were composed as epigrams. The Greek riddle on a writing tablet is no exception, taking its metre (the elegiac couplet) and certain rhetorical techniques from this genre.³⁰ In addition to sharing the basic temporal form of the aforementioned epigrams, the writing-tablet riddle even begins with a formula characteristic of epigrams, ‘Ὕλη μέν με τέκεν, καινούργησεν δὲ σίδηρος’, where the speaker identifies ‘woodland’ (ὕλη) as the progenitor who με τέκεν (‘bore me’).³¹ The riddle thus imitates certain biographical epigrams composed in the form of epitaphs, which state the subject’s father or homeland in a similar ‘X begot [or bore] me’ construction.³² Meager of Gadara (c. 135–50 bc), for example, composed an epigram about his own life that began

Νᾶσος ἐμὰ θρέπτειρα Τύρος· πάτρα δέ με τεκνοῖ

᾿Ατθὶς’ ἐν Ἀσσυρίοις ναιομένα Γάδαρ[α].³³

The speaker here memorialized the place where he was born, ‘an Attic fatherland, Gadara, bore me’ (πάτρα με τεκνοῖ / ᾿Ατθὶς’ … Γάδαρ[α]), just as the Greek riddle began ‘woodland bore me’ (ὕλη … με τέκεν). This epigrammatic progenitor formula (‘X begot [or bore] me’) is best known, however, from Virgil’s Latin epitaph:

Mantua me genuit, Calabri rapuere, tenet nunc

Parthenope. Cecini pascua, rura, duces.³⁴

Here Virgil’s life was memorialized in the initial series of statements that progress from his motherland (Mantua) to his final resting place (Parthenope). After Virgil’s epitaph, the progenitor formula was widely imitated in Latin.³⁵ The Greek writing-tablet riddle uses the same formula, further revealing its debt to the epigram genre. It is notable, however, that the progenitor formula is not used at the beginning of Aldhelm’s Aenigma 32, which rather lists the various component parts that combine to make the writing tablet (the wax, wood, and leather).

Such Greek epigrams also influenced the Latin tradition. The same ‘once … now’ structure can be seen, for instance, in many epigrams by the Latin poet Martial (c. 40–103), whose works include two books of short epigrams about inanimate objects.³⁶ These short epigrams sometimes take the form of a first-person account that progresses from the past to the present tense, contrasting the prior and present situations of the subject. Martial’s epigram on Trebulan cheeses, for example, begins with the familiar epigrammatic formula, ‘Trebula nos genuit’, before progressing to the present tense to describe the qualities of the cheeses themselves.³⁷ Martial’s epigram on an oyster has the same progressive structure, portraying its subject as a poor country girl who acquires a taste for luxury after coming to town:

Ebria Baiano ueni modo concha Lucrino:

nobile nunc sitio luxuriosa garum.³⁸

Here the contrast of the past and the present state of the oyster is conveyed not only by the tenses of the verbs ueni (‘I arrived’ – past tense) and sitio (‘I thirst’ – present tense), but also by the occurrence of two temporal adverbs, modo (‘a little while ago’) and nunc (‘now’). These Latin epigrams thus resemble the temporally progressive structure of the ‘once … now’ epigrams in Greek.

It was in imitation of such Greek and Latin epigrams that Symphosius composed several transformation riddles with the same temporal progression.³⁹ These riddles rarely use the epigrammatic ‘X begot [or bore] me’ formula, but they nonetheless conform to the ‘once … now’ model of the aforementioned epigrams.⁴⁰ In total, Symphosius composed five riddles in this way, each with one or more temporal adverbs, such as quondam (‘once’) and nunc (‘now’).⁴¹ Aenigma 56 Caliga (Soldier’s Boot), for example, was composed as a mock epitaph on a leather boot, contrasting the life and death of its speaker:

Maior eram longe quondam, dum uita manebat;

at nunc exanimis, lacerata, ligata, reuulsa,

dedita sum terrae, tumulo sed condita non sum.⁴²

The text takes the familiar shape of a progression from the past to the present tense and includes two temporal adverbs, quondam and nunc, in the first two verses. The speaker first recounts its former state as part of a larger animal when it was alive, and then it describes its death and transformation through a series of participles in line two. The final verse openly plays with the conventions of epitaphs, alluding as it does to the burial of the speaker. Since Symphosius used this same ‘once … now’ structure for several other riddles, he apparently recognized it as forming a distinct type of riddle, composed in imitation of epigrams and all portraying the transformation of their subjects.

Aldhelm wrote several transformation riddles after the model of Symphosius, expanding the short form of his predecessor’s riddles into longer compositions. Aldhelm used the ‘X begot [or bore] me’ formula several times,⁴³ perhaps imitating the famous Latin riddle about ice, which he certainly knew: ‘Mater me genuit, eadem mox gignitur ex me.⁴⁴ But Aldhelm also composed transformation riddles that recalled the ‘once…now’ riddles of Symphosius. These riddles all progress from the past to the present state of their subjects and include temporal adverbs. Aldhelm’s Aenigma 99 Camellus (Camel), for example, begins eram quondam (‘once I was’) and continues nunc… nunc (‘now… now’) in the following statements about the speaker’s present existence.⁴⁵ A total of eleven riddles by Aldhelm can be classified as transformation riddles insofar as they portray the transformation of their subjects.⁴⁶ Almost all of these riddles progress from the past to the present tense, and include one or more temporal adverbs, such as quondam and nunc. ⁴⁷ They thus resemble the ‘once … now’ transformations by Symphosius, which were composed in imitation of epigrams. Since the Greek riddle on the writing tablet was influenced by the same sort of epigrams, it naturally resembles these Latin transformation riddles, especially Aldhelm’s riddle on the same topic. The resemblances then are not evidence of Aldhelm’s direct knowledge of the Greek tradition; they are rather the result of the mutual influence of epigrams on Greek riddles and Symphosius, who in turn influenced Aldhelm.

AENIGMA 32 AND THE HISPERICA FAMINA

Aldhelm’s Aenigma 32 Pugillares (Writing Tablets) stands out among his transformation riddles for its unusual opening. Although it shares the same two-part structure as its peers, beginning in the past tense and progressing to the present tense, it does not start with a progenitor formula, ‘X begot [or bore] me’, nor does it state what it once (quondam) was. In fact, Aenigma 32 does not describe its past existence as a single object at all, as the transformations of Symphosius do. Instead it catalogues the various materials from which it was composed: the wax, wood, and leather:

Melligeris apibus mea prima processit origo,

sed pars exterior crescebat cetera siluis;

calciamenta mihi tradebant tergora dura.⁴⁸

This specific approach does not have precedent in the aforementioned epigrams or riddles. But it is noteworthy that Aldhelm began another two riddles in this same way: Aenigma 52 Candela and Aenigma 61 Pugio (Dagger). In Aenigma 52, the opening lines refer to the wax and wick that combine to make the candle:

Materia duplici palmis plasmabar apertis;

interiora mihi candescunt: uiscera lino

seu certe gracili iunco spoliata nitescunt.⁴⁹

And the beginning of Aenigma 61 Pugio describes the metal and leather that compose the dagger:

De terrae gremiis formabar primitus arte;

materia trucibus processit cetera tauris

aut potius putidis constat fabricata capellis.⁵⁰

So Aldhelm began three of his transformation riddles in a similar, formulaic way by cataloguing the various raw materials that combine to form their subjects. This approach does not have any precedent in Greek or Latin epigrams nor in the Aenigmata of Symphosius, but it does bear a striking resemblance to certain passages in a Latin work that Aldhelm likely knew in some form – the Hisperica famina.

The eccentric Latin texts known collectively as the Hisperica famina were composed, according to Michael Herren, during the mid-seventh century in Ireland before circulating in England.⁵¹ Aldhelm’s firsthand experience with Irish education is revealed by several early sources, including a letter to him from an unknown Irishman (Scottus ignori nominis) that said that Aldhelm ‘was nourished by a certain holy man of our race’.⁵² It is therefore likely that Aldhelm encountered the Hisperica famina in some form, although its influence on him has sometimes been overstated in the past.⁵³ As Andy Orchard has shown, there are many structural, topical and stylistic similarities between the Hisperica famina and Aldhelm’s Aenigmata. ⁵⁴ In its two most complete versions, the Hisperica famina dramatizes the rhetorical exercises of a group of students, who are challenged by a master to compose in Latin on a wide range of topics. The version known as the A-text includes two passages that describe in detail some of the tools used by scholars, De tabula (On the Writing Tablet) and De taberna (On the Book-container).⁵⁵

These two descriptive passages share a similar structure, and it is likely that their formulaic elements made it easier to compose for as long as possible on the given topics.⁵⁶ The recurring elements include an introductory account of the various raw materials that constitute the finished products – in the case of the writing tablet, wood and wax:

Haec arborea lectis plasmata est tabula fomentis,

quae ex altero climate caeream copulat lituram.

Defidas lignifero intercessu nectit colomellas,

in quis compta lusit caellatura.

A535   Aliud iam latus arboreum maiusculo ductu stipat situm,

uaria scemicatur pictura,

ac comptas oras artat.

Haec olim frondea glaucicomi creuit inter robora fundi,

ferrialique crescentem amputauit opifex scuri stipitem,

A540   quadrigonum ligneo dolauit incrementum neruo,

micram eruit ascia margeriam,

ornatamque perfecit tabulam,

quae dexterali historium gestatur iduma,

ac sophica caereis glomerat misteria planetis.

A545   Nunc loquelarem celeri flexu retraho tramitem,

ne ingeniosas rhetorum grauauero domescas.⁵⁷

The speaker begins by describing the principal materials of the writing tablet, wood and wax. These are referred to in the extravagant language typical of the Hisperica famina; the wood is lectis … fomentis (‘from select kindling’), and the beeswax is caeream … lituram (‘a waxy smearing’).⁵⁸ The passage then describes the physical appearance of the object, as if the speaker were holding a writing tablet in his hand (lines 533–37). When this description has been exhausted, the speaker imagines how the object was once transformed from its raw materials into the finished product, deviating into an account of the object’s origin at line 538: ‘Haec olim frondea glaucicomi creuit inter robora fundi’.⁵⁹ After recounting the process of manufacture (lines 539–42), the speaker concludes with the customary formulaic ending, beginning with nunc in line 545.⁶⁰ The other passage, De taberna (On the Book-container), has the same structure: the raw materials are mentioned first, then the physical appearance of the object is described, and finally the origin and transformation of the object are recounted. The opening lines of these two passages are thus similar to each other, and they also recall the description of raw materials in the opening lines of Aldhelm’s Aenigma 32:

Melligeris apibus mea prima processit origo,

sed pars exterior crescebat cetera siluis;

calciamenta mihi tradebant tergora dura.⁶¹

Since Aldhelm likely knew the Hisperica famina in some form, such formulaic passages may well have inspired the beginning of his three unusual transformation riddles – Aenigma 32 Pugillares (Writing Tablets), Aenigma 52 Candela (Candle), and Aenigma 61 Pugio (Dagger).

There thus is no need to posit Aldhelm’s knowledge of Greek riddles from the origin of the writing tablet that begins Aenigma 32. The two-part structure of Aldhelm’s transformation riddles, including Aenigma 32, was clearly modelled after the ‘once … now’ riddles of Symphosius. Their general resemblance to the Greek riddle on a writing tablet is due to the shared influence of epigrams on all these texts. Symphosius and the authors of the Greek riddles in the Anthologia Palatina all intentionally imitated the form of epigrams. The opening lines of Aenigma 32, which describe the various raw materials that compose the writing tablet, do not resemble the Greek riddle on the same topic; they rather look like the formulaic passages about physical objects in the Hisperica famina, a work that Aldhelm likely knew in some form. This peculiar feature of Aenigma 32 therefore can be explained using Latin texts known in seventh-century England without appealing to Greek sources.

DICTION, METAPHOR AND ETYMOLOGY IN AENIGMA 32

Although the structure of Aenigma 32 does not reveal Aldhelm’s knowledge of Greek riddles, perhaps some of its diction does. As we have seen, Milovanović-Barham proposed that Aldhelm’s use of melliger (‘honey-bearing’) in line 1 was inspired by a similar compound adjective in a Byzantine Greek riddle about a candle, γλυκυγόνον (‘sweet-bearing’).⁶² Both words are unattested before their occurrence in their respective texts. But Aldhelm’s use of melliger should be viewed in the wider context of his unusual diction and its place in the Latin literary tradition. As many scholars have noted, Aldhelm was especially fond of compound adjectives ending -fer and -ger. ⁶³ Such compounds had long been a part of the Latin poetic tradition,⁶⁴ and were popular among some of Aldhelm’s favourite poets, such as Juvencus, Sedulius and Arator.⁶⁵ Their initial use by ancient Roman poets may have been inspired by similar compounds in Greek poetry, but they had been fully absorbed into the Latin poetic lexicon by Aldhelm’s time.⁶⁶ Most of the compound adjectives used by Aldhelm can be found in earlier texts, although he does seem to have coined a few himself, such as melliger in Aenigma 32.⁶⁷ It is nonetheless likely that these few neologisms were modelled after existing Latin combinations. In the case of melliger (‘honey-bearing’), Aldhelm was probably thinking of Ovid’s description of bees as mellifer (also ‘honey-bearing’).⁶⁸ Perhaps Aldhelm indulged here in the substitution of -fer for -ger in the spirit of the Hisperica famina, where many more such neologisms occur;⁶⁹ note, for example, the rare word lignifer (‘wood-bearing’) and the neologism glaucicomus (‘glaucous-coloured’) in the above quotation at lines 533 and 538. So here again an apparent parallel between Aenigma 32 and a Greek riddle can be best explained by Aldhelm’s use of his immediate Latin sources.

Another similarity noted by Milovanović-Barham is the martial metaphor. Aldhelm referred to the stylus as ferrum (‘iron stylus’, ‘iron weapon’), and described its erasing power with the phrase diris … armis (‘by cruel arms’). Milovanović-Barham suggested that these warlike images might have been modelled after the figurative use of Ares in the Greek riddle on a writing tablet.⁷⁰ But Aldhelm’s representation of the stylus as a weapon was probably inspired yet again by the Hisperica famina. The image of scholars as warriors characterizes the whole of the Hisperica famina, as Herren and Orchard have both noted.⁷¹ In the two most complete versions, the Hisperica famina portrays a rhetorical contest that begins with the speaker extending an open challenge to a group of newly arrived students; the A-text reads ‘huic lectorum sollertem inuito obello certatorem’.⁷² After boasting of his previous victories, the speaker then describes his weapons and armour in the manner of a heroic arming scene, including his writing tablet and stylus:⁷³

Dum truculenta me uellicant spicula,

30   protinus uersatilem euagino spatham,

quae almas trucidat statuas;

arboream capto iduma peltam,

quae carneas cluit tutamine pernas;

ferralem uibro pugionem,

35   cuius pitheum assiles macerat rostrum cidones;

ob hoc cunctos lastro in agonem coaeuos.⁷⁴

Here the scholar is portrayed as an armed warrior, with his wooden writing tablet for a shield and his iron stylus for a dagger. In the Hisperica famina, this martial metaphor even exists at the level of individual words, such as the Hisperic term arcator, which Orchard compared to both arca (‘book-chest’) and arcus (‘bow’).⁷⁵ Although it is clear in the text that the word arcator refers to a scholar – that is, to someone who uses an arca (‘book-chest’) – the potential connection to arcus (‘bow’) nonetheless encourages the association of the scholar to a warrior.

A similar word-play may even lurk behind the solution to Aldhelm’s Aenigma 32 Pugillares (Writing Tablets). As Nicholas Howe explained, Aldhelm composed many of his riddles around the etymology of the solutions, believing that the words for his solutions were not arbitrary signifiers but rather revealed some essential truth about the things they signified, often drawing on Isidore’s Etymologiae. ⁷⁶ It is interesting then that Aldhelm chose an unusual word for his writing tablet, pugillares, where one might expect the common word tabula (‘tablet’), as in the passage in the Hisperica famina. Although the word pugillares does not occur in Isidore’s Etymologiae, it clearly came from pugillus (‘fistful’, itself related to pugil, ‘boxer’, and pugnus, ‘fist’). Aldhelm therefore would have thought that the word pugillares expressed the image of the warrior-scholar, as Orchard has suggested.⁷⁷ The pugillares are literally ‘what is held in the fist’ – they are a scholar’s arma (‘weapons’, ‘arms’). The martial metaphor in Aldhelm’s Aenigma 32 is thus an etymological clue that reveals the connection between pugillares (‘writing tablets’) and pugillus (‘fistful’), pugil (‘boxer’), and the rest. That Aldhelm should develop a martial image that describes the use of a stylus as diris … armis (‘by cruel arms’) should come as no surprise then; it does not imply a knowledge of the Greek riddle on the same topic.

There may even be a reference to this etymology in the text of Aenigma 32 itself. Line 7 reads ‘quod largos generat millena fruge maniplos’, where the noun manip(u)lus literally means ‘a handful’ (related to manus ‘hand’), and by extension can refer to ‘a sheaf of wheat’ among other things (such as ‘a company of soldiers’).⁷⁸ The word manip(u)lus therefore mirrors the etymology of the solution pugillares, as well as participating in the agricultural metaphor developed in the second half of Aenigma 32, which Milovanović-Barham considered largely irrelevant to a discussion of its sources.⁷⁹ The writing tablets (pugillares) are thus ‘hand-held’ things, where the largi manipuli (‘plentiful handfuls’) of the Holy Word can be harvested. If a source is needed for this word-play, there is an intriguing analogue in a passage by one of Aldhelm’s favourite Latin poets, Arator – a name that means ‘ploughman’, incidentally. Here, Arator portrayed the Apostles as holy harvesters in the same agricultural metaphor developed by Aenigma 32.⁸⁰ The many double-meaning words in the passage, including manip(u)lus, are given in parentheses in the translation below:

365   Da semina uerbi

per tua dona coli, signisque noualibus usa,

colligat ista manus, te fructificante, maniplos,

de quibus ipse tui componas horrea caeli.⁸¹

Arator thus described how the Apostles should disseminate the Word of God and reap new followers of Christ, with maniplos in line 367 referring to these new Christians as both ‘companies of soldiers’ and ‘sheaves of wheat’ for God’s heavenly granary. It is not implausible that this punning passage by Arator inspired Aldhelm to play with the polysemy of the word manip(u)lus in Aenigma 32. Aldhelm’s riddle about the writing tablet is thus fully immersed in the Latin literary tradition, making artful and inventive use of his Latin sources, especially the Hisperica famina. There is little reason to infer a knowledge of Greek sources from this riddle alone.

GENERAL SIMILARITIES

What about the general similarities between Aldhelm’s Aenigmata and Greek riddles? Do these similarities reveal Aldhelm’s knowledge of the Greek tradition or do they have other explanations? As I mentioned above, Milovanović-Barham noted three characteristics of Aldhelm’s collection that were better attested by Greek riddles than by Symphosius: the use of logogriphs; an emphasis on origins and birth; and the inclusion of verbal challenges to solve the riddles.⁸² With respect to logogriphs, Aldhelm’s Aenigmata and the riddles in the Anthologia Palatina certainly include a few verses that play with words, such as corbus/orbus and paries/aries. ⁸³ Aldhelm’s Aenigma 63 Corbus (Raven), for example, contains this clue: ‘littera tollatur: post haec sine prole manebo’, which refers to the word orbus (‘bereft of children’) contained within the solution corbus (‘raven’).⁸⁴ But as Milovanović-Barham herself acknowledged, Symphosius also included such logogriphs, as in the final line of Aenigma 36 Porcus (Pig), ‘nomine numen habens si littera prima periret’, which alludes to the word Orcus (‘god of the underworld’) contained within the solution porcus (‘pig’).⁸⁵ Another two Latin riddles in the Anthologia Latina use similar logogriphs, including one on paries/aries, a plausible source for Aldhelm’s own logogriph on the same pair of words.⁸⁶ It is likely then that Aldhelm’s logogriphs were inspired by these Latin sources rather than by Greek ones.

The other two similarities are common to many riddle traditions, and thus cannot prove Aldhelm’s knowledge of Greek riddles in particular. It is true that many of Aldhelm’s Aenigmata mention the origin of their subjects (thirty-four by my count), including Aenigma 32 discussed above.⁸⁷ But this is not a particularly striking characteristic of his collection, since the riddles of many traditions refer to the birth or origin of their subjects.⁸⁸ Although Milovanović-Barham claimed that Symphosius only included ‘about a dozen’ riddles ‘concerned with the provenance of the subject in question’, I count twenty-one references to a subject’s birth or origin, a comparable number to that of Aldhelm’s much larger collection.⁸⁹ Aldhelm’s interest in origins then does not obviously reveal the influence of Greek riddles. And finally, his inclusion of verbal challenges to solve his riddles is not uncommon in riddle traditions.⁹⁰ As Milovanović-Barham rightly noted, Aldhelm’s few challenges to the reader, such as ‘sciscitor inflatos, fungar quo nomine, sophos’, have no precedent in the Aenigmata of Symphosius, whereas similar challenges do occur in some Byzantine Greek riddles.⁹¹ But the earlier Greek riddles in the Anthologia Palatina do not include such challenges to the reader, so it is hard to say to what extent they were in fact characteristic of the earlier Greek tradition. Verbal challenges arise naturally from the competitive nature of the riddle genre itself, and they often appear in literary representations of riddle contests. After each riddle in the Old Norse contest in Hervarar saga ok Heiðreks ins vitra, for example, the command Heiðrek konungr, hyggðu at gátu is repeated.⁹² Even outside riddle contests, such challenges became formulaic in the Old English riddles of the Exeter Book, as in the famous challenge saga hwæt ic hatte and its variations.⁹³ It is plausible then that the few challenges to the reader in Aldhelm’s Aenigmata arose from the inherently competitive nature of the riddle genre, rather than being imitations of such challenges in Greek riddles. Their occurrence in Aldhelm’s Aenigmata is not likely evidence of his knowledge of Greek riddles.

As we have seen, there are indeed some general features of Aldhelm’s Aenigmata that are more noticeable in Greek riddles than in his main source Symphosius, and in particular there are many similarities between Aldhelm’s Aenigma 32 Pugillares (Writing Tablets) and the Greek riddle on the same topic. Both riddles take the form of a two-part progression from the past to the present, recounting the origin of their subjects, and both are told from the perspective of the solution itself. But these resemblances do not necessarily imply that the Greek riddle was Aldhelm’s source. Aldhelm’s Aenigma 32 is a transformation riddle, a type of riddle that he modelled after similar transformation riddles by Symphosius, who in turn modelled them after epigrams. The Greek riddle on a writing tablet was also modelled after similar epigrams, so it naturally resembles Aldhelm’s riddle on the same topic. Many of the details of Aldhelm’s Aenigma 32 have precedents in the Hisperica famina, which he is more likely to have known than Greek riddles; although he could have encountered Greek riddles at Canterbury, we have no evidence that he did. Like Aenigma 32, the Hisperica famina mentions the raw materials of the writing tablet and the martial metaphor of the stylus, so it is not at all implausible that Aldhelm was directly inspired by some form of this text when he composed Aenigma 32. If it could be shown that Aldhelm did in fact know Greek riddles, it would be an exciting development in the understanding of his works, and it would provide important evidence for the reception of Greek texts in seventh-century England. But it is in fact more likely that Aldhelm’s Aenigmata were more directly inspired by the Hisperica famina than has generally been recognized, suggesting that further study of the connections between these two seventh-century Latin texts would be very much worthwhile.