‘The question is,’ said Alice, ‘whether you can make words mean so many different things’.
Lewis Carroll, Through the Looking-Glass and What Alice Found There
I. Introduction
The psychological and mental mechanisms that trigger and govern natural semantic change (metonymy, metaphor, etc.) are fairly straightforward. Usually, words change their meaning gradually through adjacent cognitive domains: for instance, γυνή ‘woman’ > ‘wife’. But, occasionally, languages also attest to unexpected semantic developments in which words take on new meanings that have little or nothing to do with the original. Cross-linguistically, ‘literary’ learned and/or obsolete terms, which are rarely used (if at all) in everyday speech, frequently succumb to the latter kind of change due to the fact that their meaning often remains unclear (if not utterly opaque) to speakers. These words are often more connotative than denotative, serving as high-style markers. Footnote 1
This phenomenon is far from unknown in ancient Greek. In many cases, over-interpretation (as well as its corollary, irrational polysemy) can be traced back to the ancient poets. More often than not, however, the responsibility for such misunderstandings rests with the ancient lexicographers and scholiasts (indeed, in many cases, with contemporary scholars). In dealing with the semantics of rare words found in archaic poets (which had become obsolete in common parlance), ancient grammarians and lexicographers resorted to intuitive guesswork, or simply allowed their imagination to run wild. Modern researchers have demonstrated that the contentions of ancient lexicographers should only be accepted if they are corroborated by further evidence (whether linguistic, archaeological or historical). Footnote 2
In this paper, we are concerned with the ancient Greek nouns λάκος, ῥάκος (with its Aeolic variant βράκος), σπϵῖρον and λαῖφος. These words allegedly attest to an unexpected semantic overlap between ‘rags’ and ‘garments’. However, closer scrutiny of the passages in which they occur tells an entirely different story.
II. Λάκος in an oracular enquiry from Dodona
The quasi-hapax λάκος occurs in a consultation on a lead lamella (now seemingly lost) found at Dodona in 1955 and dated to the late fourth or early third century BC.Footnote 3 We reproduce below Éric Lhôte’s text, followed by his French translation, as well as several English translations by other scholars:
ἔκλϵψϵ Δορκίλος τὸ λάκος;
Est-ce que c’est Dorkilos qui a volé mon vêtement? (tr. Lhôte Reference Lhôte2006)
Did Dorkilos steal the cloth? (tr. Parke Reference Parke1967; Eidinow Reference Eidinow2007)
Did Dorkilos steal the garment? (tr. Chaniotis Reference Chaniotis2017)
The Greek text is straightforward. The translations, however, are problematic in two respects.Footnote 4 First, the enquirer is not necessarily also the possessor of the stolen λάκος, as implied by Lhôte’s possessive mon. Second, and more relevant to the present study, the meaning ‘garment’, which Lhôte and Angelos Chaniotis attribute to this word, is questionable.
Following in the footsteps of Olivier Masson, Lhôte rightly connected our lamella’s λάκος to Hesychius’ gloss λάκη· ῥάκη. Κρῆτϵς (λ 192 Latte and Cunningham; MS ρακκη).Footnote 5 The word must be a cognate of the rare λακίζω ‘to tear’, λακίς (mostly pl. λακίδϵς) ‘shred, tear’ and λάκισμα ‘rags’.Footnote 6 Both the Hesychian gloss and the other words of the family unequivocally point to a meaning ‘rags, tatters’ but, probably owing to the fact that ‘a bundle of tatters’ is (in principle) an unlikely target for a thief, Dimitrios Evangelidis, Masson and Lhôte assume that Dorkilos was suspected of stealing a garment in good condition.Footnote 7 However, neither the semantic promotion ‘tatters’ > ‘garment’ nor a hypothetical demotion ‘garment’ > ‘tatter’ is an obvious semantic change. In the case of ancient Greek at least, the evidence of the would-be parallels is much less compelling than it may seem at first glance, as will be demonstrated in the following sections.
III. From ‘rags’ to ‘garments’?
As befits their circumstances, beggars, castaways and the wretched wear tatters in epics and drama. Disguised as a beggar, Odysseus is clad in rags when he finally reaches Ithaca (Od. 13.434–35; 14.342, 349, 512; 17.198; 18.67, 74; 19.507). Other ill-fated heroes in tatters are Xerxes (Aesch. Pers. 125, 834–35), the Danaids (Aesch. Supp. 131, 904), Electra (Eur. El. 185), Menelaos (Eur. Hel. 415–17, 1079) and Telephos (Eur. fr. 697 Kannicht). Ragged kings on stage were a Euripidean staple, which Aristophanes mocked time and again in his comedies (Ach. 415–38; Thesm. 935; Ran. 842, 1063–64).Footnote 8
In the literary passages cited above, ῥάκος and its synonyms, λακίδϵς and τρῦχος (cf. τρύχω ‘wear out’), as well as the adjective ῥωγαλέος (cf. ῥωγάς ‘ragged’, from ῥήγνυμι ‘break’), did not denote items of clothing in a good condition. To all appearances, the ancient Greeks viewed rags and usable clothing in entirely antithetical terms. For instance, in Aristophanes’ Wealth, Chremylos enumerates the many hardships endured by the poor (535–47), one of which is ‘to have rags instead of a cloak’ (ἀνθ’ ἱματίου μὲν ἔχϵιν ῥάκος, 540). In a fragment of Antiphanes’ Soldier or Tychon (fr. 202 PCG transmitted by Ath. 103F), a character complains about the reversals of fortune, mentioning a chorēgos who, having decked out the entire chorus in golden robes, was himself reduced to rags. To assume that the consultant of Lhôte’s lamella no. 120 was a beggar enquiring about the fate of his tattered clothes would be preposterous.Footnote 9
i. Ῥάκος and Lesbian βράκος
Purportedly, a change ‘rag’ > ‘garment’ is on record for Lesbian βράκος, which, according to ancient lexica and modern dictionaries, denotes a ‘(costly) garment’ in Sappho (fr. 57.3 Voigt, transmitted by Ath. 21B) and in Theocritus (Id. 28.11; cf. also βράκος· κάλαμος. ἱμάτιον πολυτϵλές (Hsch. β 1047 Latte and Cunningham)).Footnote 10 Given their seemingly contradictory meanings, modern scholars are reluctant to accept an etymological connection between βράκος ‘garment’ and ῥάκος ‘rag(s)’.Footnote 11 However, as we will soon see, such extreme circumspection is unwarranted, for both words can be safely traced back to the common etymon *ϝράκος.Footnote 12
With regard to the hypothetical semantic change from ‘rags’ to ‘garment’, a closer look at the passages under consideration makes clear that the meaning ‘(costly) garment’ is illusive:Footnote 13
τίς δ’ ἀγροΐωτις θέλγϵι νόον … | ἀγροΐωτιν ἐπϵμμένα σπόλαν … | οὐκ ἐπισταμένα τὰ βράκϵ’ ἔλκην ἐπὶ τῶν σφύρων; (Sappho fr. 57 Voigt)
What country girl beguiles your mind … dressed in country garb … not knowing how to pull her brakea above her ankles?Footnote 14
σὺν τᾷ πόλλα μὲν ἔργ’ ἐκτϵλέσῃς ἀνδρϵΐοις πέπλοις, | πόλλα δ’ οἶα γύναικϵς φορέοισ’ ὐδάτινα βράκη. (Theoc. Id. 28.10–11)
[A distaff] with which you will create many pieces of work for men’s robes, and many brakē as if made of water, such as women wear.
In fr. 57, Sappho addresses a certain Andromeda, a rival for a girl’s affection, whom she taunts for her ‘tattered clothes’ (βράκϵα), which betray her rusticity.Footnote 15 Although Theocritus’ ὐδάτινα βράκη ‘rags made of water (i.e. fluid)’Footnote 16 may describe, half-jokingly, the women’s clothes that Theugenis will be able to produce with the help of the distaff,Footnote 17 we are inclined to think that, by Theocritus’ time, Lesbian βράκος was only a learned γλῶσσα with a distinctly Aeolic flavour, which had already been misinterpreted as ‘garments’ in Sappho’s poem.Footnote 18 This secondary over-interpretation of βράκος does not thereby disprove our contention that a semantic change ‘rags’ > ‘garments’ is unnatural.
The antiphrastic use of βράκος in Sappho and possibly in Theocritus is comparable to similar expressions mostly ironically said of (fancy, expensive) garments in the colloquial registers of modern languages: for example, English rags, French mettre ses loques (literally ‘to put one’s tatters on’ > ‘to get dressed’), Italian stracci (diminutive straccetti) ‘rags’, Spanish trapos (diminutive trapitos) ‘cloths’. This is also attested in a Hellenistic letter on papyrus:
I am naked and we are in the open air. Please, give the order that I receive four drachmas, so that I can buy even a rag.Footnote 19
ii. Λαῖφος and λαίφη
In two passages of the Odyssey, the neuter λαῖφος (a noun of unknown origin) refers to the shabby clothes of the pseudo-beggar Odysseus:
ἀμφὶ δὲ λαῖφος | ἕσσω ὅ κϵ στυγέῃσιν ἰδὼν ἄνθρωπος ἔχοντα. (Od. 13.399–400)
I [Athena] will put a laiphos around you, and whoever sees you wearing it shall hate you.
δϵδάκρυνται δέ μοι ὄσσϵ | μνησαμένῳ Ὀδυσῆος, ἐπϵὶ καὶ κϵῖνον ὀΐω | τοιάδϵ λαίφϵ’ ἔχοντα κατ’ ἀνθρώπους ἀλάλησθαι. (Od. 20.205–06)
And my eyes are filled with tears when I [Philoitios] remember Odysseus, because I imagine him too wandering among men wearing such laiphea.
According to ancient and modern scholars, λαῖφος denotes a ‘tattered garment’ in both passages.Footnote 20 Modern translations usually reflect this viewpoint by rendering the word as ‘rags’.Footnote 21
However, even if Odysseus was clothed in rags, it does not follow that λαῖφος inherently denoted a ragged garment, as evidenced by later uses of the word.Footnote 22 In archaic poetry (Alc.; Hymn. Hom. Ap.) and in Attic tragedy, λαῖφος means ‘sail’. In antiquity, as in more recent times, a sail can hardly have been made of tatters.Footnote 23 Likewise, in the Homeric Hymn to Pan, the god wears a λαῖφος δαφοινὸν λυγκός over his back (23–24). Although the meaning of the adjective δαφοινός is uncertain in this passage, a lynx coat characterizes Pan as a god of the wild, not as a beggar.Footnote 24 Ιn the Hymn to Hermes, the newborn god plays (ἀθύρων) in his cradle with a λαῖφος (152), which seems to be bedding of some kind.Footnote 25
The meaning ‘rags’ probably resulted from over-interpretation of the two passages of the Odyssey quoted above. In Od. 13.399–400, the condition of Odysseus’ λαῖφος is suggested by the context: Athena will render the hero unrecognizable by making him ugly and filthy (13.392–403). In fact, the transformation of Odysseus (13.434–37) includes a more detailed description of the hero’s shabby clothes.Footnote 26 By Od. 20.205–06, Odysseus’ λαίφϵα have been repeatedly mentioned in different passages of the poem, in which they are described as ῥάκϵα ‘tatters’ (Od. 14.512, 18.67), λυγρὰ ϵἵματα ‘pitiful clothing’ (Od. 16.457; 17.203, 338, 573) and κακὰ ϵἵματα ‘wretched clothing’ (Od. 14.506, 17.24, 19.72).
A feminine form, λαίφη, probably coined by the poet himself on the neuter λαῖφος, occurs in a passage of Callimachus’ Hecale, in which Theseus takes refuge from a storm in the humble hut of the eponymous character, an impoverished old widow, and shakes off his sopping wet clothing (διϵρὴν δ᾿ ἀπϵσϵίσατο λαίφην, fr. 239 Pf.).Footnote 27 The Suda glosses λαίφη as χλανίς ‘mantle’ (λ 207 Adler) and, according to Pfeiffer (see the apparatus ad loc.), Callimachus’ verse echoes certain Homeric passages.Footnote 28 There is no reason to believe that Theseus was portrayed as a beggar in his encounter with his hostess. Interestingly, this episode of the myth might have been depicted in a fifth-century BC Attic red-figure hydria, on which a bowed old lady (Hecale?) proffers a flat object (a phiale?) to a young man (Theseus?).Footnote 29 This is wrapped up in a neatly arranged cloak, for which the context betrays no signs of fraying.
We may conclude that the λαῖφος must originally have denoted a piece of fabric whose size varied depending upon the purpose for which it was intended. The word could easily be applied to different objects, such as a garment, a blanket or a sail. Indeed, cloaks in ancient Greece served a dual purpose, being used as outer garments by day and coverings by night. Similarly, although depending upon the context, φᾶρος (Myc. pa-we-a /p h arweha/) can equally refer to a cloak (for instance, Od. 5.230), a shroud (Il. 24.580 and 588), swaddling clothes (Hymn. Hom. Ap. 121) and a bedspread (Soph. Trach. 916). Therefore, the λαῖφος and λαίφη worn by Odysseus and Theseus were simply ‘woollen textiles’ likely used as cloaks.
iii. Σπϵῖρον
As with λαῖφος, the evidence regarding σπϵῖρον in the texts seems to be incompatible with the basic meaning ‘tattered piece of clothing’.Footnote 30 In her famous address to the suitors, Penelope implores them to wait until she has finished the burial shroud she is weaving for Odysseus’ elderly father before she makes a final decision:
μή τίς μοι κατὰ δῆμον Ἀχαιϊάδων νϵμϵσήσῃ, | αἴ κϵν ἄτϵρ σπϵίρου κϵῖται. (Od. 2.101–02 = 19.146–47, 24.106–07)
For fear that any of the Achaean women across the land should blame me, were [Laertes] to lie without a speiron.
Here, σπϵῖρον picks up the aforementioned shroud (φᾶρος … ταφήϊον, Od. 2.97–99). Being the father of the king of Ithaca, and having been a king himself, Laertes was certainly not expected to be buried in a tattered shroud. Furthermore, σπϵῖρον occurs in the description of Odysseus’ shipwreck after his departure from Ogygia:
τηλοῦ δὲ σπϵῖρον καὶ ἐπίκριον ἔμπϵσϵ πόντῳ. (Od. 5.318)
Far into the sea speiron and sailyard fell.
Since the context clarifies that σπϵῖρον is a component of the hero’s raft, ancient and modern scholars agree that it means ‘sail’.Footnote 31
By contrast, all occurrences of σπϵῖρον in later authors point to a garment of some form or another. Ιf the transmitted text is correct, then the diminutive σπϵιρίον in Xenophon denotes some piece of men’s light clothing.Footnote 32 In Euphorion, νυμφίδιον σπϵῖρον can only be understood as a ‘bridal gown’ (fr. 107 Powell), whereas in Nicander it denotes a ‘cloth’ used to apply an ointment (Alex. 460–61).Footnote 33 The σπϵιροφόρος mentioned in an Ephesian inscription is likely to be a young person who carried a sacred robe during a procession, alongside bearers of other sacred objects (σϵλϵινοφόρος [sic] ‘celery bearer’, ἁλοφόρος ‘salt bearer’, κοσμοφόρος ‘bearer of the ornaments’).Footnote 34 The meaning ‘garment’ is confirmed by Pollux, who mentions a market in Athens called ἱματιόπωλις or σπϵιρόπωλις (sc. ἀγορά).Footnote 35
In late lexica, however, σπϵῖρον is not merely a ‘beautiful dress’ (καλὸν ἱμάτιον): it is also glossed as a ‘ragged garment’ (ῥακῶδϵς [ἱμάτιον]).Footnote 36 Once again, the meaning ‘rags’, which has found its way into some modern dictionaries,Footnote 37 was improperly inferred from the context in two Homeric passages. In the first, Helen describes Odysseus’ clothing when he managed to sneak into Troy in disguise:
σπϵῖρα κάκ’ ἀμφ’ ὤμοισι βαλών, οἰκῆι ἐοικώς,
ἀνδρῶν δυσμϵνέων κατέδυ πόλιν {ϵὐρυάγυιαν.
ἄλλῳ δ’ αὐτὸν φωτὶ κατακρύπτων ἤϊσκϵ,
Δέκτῃ, ὃς οὐδὲν τοῖος ἔην ἐπὶ νηυσὶν Ἀχαιῶν·
τῷ ἴκϵλος κατέδυ Τρώων πόλιν}, οἱ δ’ ἀβάκησαν
πάντϵς· ἐγὼ δέ μιν οἴη ἀνέγνων τοῖον ἐόντα. (Od. 4.245–50)Footnote 38
He put some vile speira on his shoulders and, looking like a servant, got into the wide-wayed city of hostile men. He camouflaged himself and took on the appearance of another man, Dektes, he who in no way was of such kind on the ships of the Achaeans. In the guise of that man, he entered the city of the Trojans. They all remained speechless,Footnote 39 and I alone recognized him, even though he had such an appearance.
The Byzantine scholar Eustathius and, after him, some modern translators analyse σπϵῖρα in this passage as denoting a filthy ‘bundle of rags’.Footnote 40 This line of interpretation probably harks back to Aristarchus’ views on the meaning of ΔEKTHI (l. 248). For the Alexandrian scholar, the general sense of lines 247–48 was that Odysseus assumed the disguise of a mendicant (δέκτης) when he entered Troy.Footnote 41 Aristarchus’ interpretation of δέκτης as ‘receiver (of alms)’ (an agent noun of δέκομαι/δέχομαι) and, secondarily, ‘beggar’ has been endorsed by many in both ancient and modern times.Footnote 42 The noun is only documented in a late Christian funerary epigram.Footnote 43 Still, this only proves that, by the end of antiquity, Aristarchus’ reading had gained widespread acceptance. At any rate, δέκτης is at odds with οἰκϵύς: a servant is by no means a beggar.
For all his indubitable talent and skill, the great Aristarchus missed the point here. In the Little Iliad, which also narrates Odysseus’ surreptitious exploration of Troy in the company of Diomedes, the hero is disguised as an inconspicuous man named Δέκτης (fr. 6, Bernabé PEG). According to Aristarchus, the author of the Little Iliad copied this passage, mistaking δέκτῃ for a personal name.Footnote 44 Conversely, some modern scholars are of the opinion that Od. 4.246–49 (ϵὐρυάγυιαν … Τρώων πόλιν) is a later interpolation by some rhapsode who combined the two versions of the episode as narrated in the Odyssey and the Little Iliad.Footnote 45 In some versions of the story, Odysseus manages to slip into Troy by pretending to be an unremarkable commoner named Δέκτης and by wearing humble, undistinctive clothing, not some beggar’s rags. The adjective κακά in the above-quoted passage (Od. 4.245) has social overtones (‘of low quality, vile’) and, contrary to the assumption of some scholars (both ancient and modern), it may imply that σπϵῖρα did not specifically refer to a ragged piece of clothing.Footnote 46
In the second passage of the Odyssey in which σπϵῖρα occurs, the naked Odysseus shipwrecked in the island of Scheria begs Princess Nausicaa for ‘a scrap of cloth’:
δὸς δὲ ῥάκος ἀμφιβαλέσθαι, | ϵἴ τί που ϵἴλυμα σπϵίρων ἔχϵς ἐνθάδ’ ἰοῦσα. (Od. 6.178–79)
Give me a scrap of cloth to throw around myself, if you had a wrapping of the speira when you came here.
Odysseus, who is aware that Nausicaa and the other girls have been washing laundry, does not ask for ordinary clothes, but just for some tatters to cover his shame. The neuter ϵἴλυμα, a derivate of ἐλύω/ϵἰλύω ‘to enfold, enwrap’, denotes a ‘wrap’, and σπϵίρων could be a genitive of material, ‘a wrapping consisting in pieces of cloths’, that is ‘a wrapping-cloth’, which would not necessarily have been tattered.Footnote 47 However, according to the scholia, σπϵῖρα is here synonymous with ἱμάτια or ἐνδύματα,Footnote 48 a sense that finds independent confirmation in the post-Homeric use of the word mentioned above. Consequently, ϵἴλυμα σπϵίρων could be one of the wrappings (ϵἴλυμα) in which the girls carried the bundles of garments (σπϵίρων).Footnote 49 If this interpretation is correct, then σπϵῖρα cannot mean ‘tattered clothes’: the previous passages of book six clearly indicate that Nausicaa and her servants had gone to the river to wash the magnificent clothing of the Phaeacian royal family.Footnote 50 Eventually, Nausicaa’s servants provide Odysseus with a cloak and a tunic (φᾶρός τϵ χιτῶνά τϵ ϵἵματ(α), Od. 6.214). On a mid-fifth-century BC Attic red-figure amphora which portrays the encounter of Odysseus and Nausicaa, the pieces of cloth hanging from a tree, held by four of the princess’ attendants and wrung by another, are not depicted as tatters.Footnote 51
When we consider all the available evidence, we cannot but conclude that σπϵῖρον must originally have designated a kind of ‘cloth’ (like φᾶρος and λαῖφος); that it was a term that could be used in different contexts for different referents in the field of textiles and clothing, like a sail, a shroud, a dress, a cloak, etc., but not specifically for rags.
IV. Dodonaean λάκος revisited
It follows from the foregoing that the evidence in Greek for a semantic change ‘rags’ > ‘clothes’ or ‘clothes’ > ‘rags’, based on the testimony of ancient lexicographers, is extremely shaky and does not warrant the meaning ‘garment’ that epigraphists have attributed to λάκος in the Dodona enquiry. We must search elsewhere for an alternative hypothesis.
Relevant to the elucidation of λάκος is perhaps the evidence of the inventories of garments offered in the sanctuary of Artemis Brauronia on the Athenian Acropolis,Footnote 52 in which the noun ῥάκος is occasionally added as a descriptive appositive to all kinds of garments (ἱμάτιον, χιτών, χλανίς, etc.) in poor condition. For illustrative purposes, a few examples will suffice:
Μέλιττα ἱμάτι|ον λϵυκὸν καὶ χιτωνίσκον ῥάκος. (IG II2 1514.19–20 = 1515.11–12; ca. 350 BC)
Melitta [consecrated] a white himation and a short chiton (tatters).
κάνδυν Διοφάντη Ἱϵρωνύμου γυνὴ Ἀχαρνέ(ως) πασμάτια ἔχοντα χρυσᾶ ῥάκ(ος). (IG II2 1524.180–81 = 1523.8–9; ca. 350 BC)
Diophante, wife of Hieronymos, Acharnian, [consecrated] a kandus with golden spangles (tatters).
In other inventories, the adjectives ῥακώδης (Athens, Delos) and ῥάκινος (Tanagra, Samos) ‘tattered’ serve a similar purpose:Footnote 53
Ἀρχίππη χι[τώνιον σ]τύππινον ῥακῶ(δϵς). (IG II2 1518.66; Athens, ca. 350 BC)
Archippe [consecrated] a short chiton made of flax, tattered.
ἄλλον (sc. χιτῶνα) ἐμ πλαισίωι πϵριπορφυροῦν ῥακώδη κατϵρρυηκότα. (IDélos 1417, face A, col. I.30–31; Delos, 155/4 BC)
Another (chiton) in a frameFootnote 54 purple-edged, tattered, completely ruined.
ἄλλα (sc. προσκϵφάλαια) [ῥ]ακώδη ΙΙΙΙ. (IG XI.2, 147 B.13; Delos, ca. 300 BC)
Four other cushions, tattered.
Φιλοκκώ ταραντῖνον ῥάκινον. (SEG 43.212B.37; Tanagra, ca. 260–250 BC)
Philokko [consecrated] a tarantinon, ragged.
πϵρίζωμα ἁλοργοῦν ῥάκινον ποικίλον. (IG XII.6, 261.22; Heraion of Samos, ca. 346/5 BC)
A purple loincloth, tattered, patterned.
We may tentatively suppose that the λάκος mentioned in Lhôte’s lamella no. 120 might be a sacred object: a ragged piece of clothing dedicated to a goddess at Dodona or another sanctuary. This would justify the otherwise puzzling consultation. The consultant, possibly someone from the sanctuary, does not enquire about the theft of some filthy tatters, but suspects Dorkilos of having committed sacrilege by stealing a sacred object from a shrine.
Alternatively, judging from the meaning of ῥάκος and its diminutive ῥακάδιον in some late Greek papyri, λάκος might denote a tattered piece of cloth tied up for transporting objects (a bundle or parcel):Footnote 55
From the dispatch carrier we received your letter and a ragged piece of cloth containing some gold leaves, which we have handed to Trophimos.
σπέρματα σικυδίων | σπουδαῖα ἔπϵμψα ὑμϵῖν (= ὑμῖν) διὰ | Διογένους τοῦ φίλου Χαιρέου τοῦ πο|λϵιτικοῦ (= πολιτικοῦ), ῥάκη δύο κατασϵσημημ|{μ}ένα [τ]ῇ σφραγϵῖδι (= σφραγῖδι) μου, ἐξ ὧν δώσϵις | τοῖς παιδίοις σου ἓν ἐξ αὐτῶν. (P.Oxy. 117.9–16, second or third century AD)
I sent you some quality melon seeds through Diogenes, the friend of Chaireas the citizen, [and] two ragged pieces of cloth (that is, bundles) sealed with my seal, one of which you shall give to your children.Footnote 56
As in many other languages, in ancient Greek, bags and wallets frequently took their name from the material of which they were made: for example, ἀσκός ‘animal skin, hide’ > ‘bag’, ‘wineskin’; δέρμα ‘skin’ > ‘wineskin’; διφθέρα ‘leather’ > ‘haversack’; σάκκος ‘rough cloth made of animal-hair’ > ‘sack’; σκύτος ‘tanned hide, leather’ > ‘bag’, etc.
One could counter-argue that the chronological and geographical gaps between the Egyptian letters and the Dodonaean enquiry is too great. Nevertheless, a diminutive ῥάκιον already occurs in a Delian inventory of the temple of Apollo, where different types of coins were wrapped with a ragged cloth in a bundle, stored in a jar (στάμνιον) with other objects:
[πλινθοφόρους καὶ ἡμιρόδι]α τρία καὶ διώβολα τέτταρα καὶ χαλκὸν ἀδόκιμον ἐν ῥακίωι καὶ λϵπίδια χρυσᾶ τὰ ἐκ τοῦ θησαυροῦ Δ 
Ι· ταῦτα ἔνϵστιν ἐν στ[αμνίωι]. (IDélos 1450.103; 140/39 BC).Footnote
57
plinthophoroi, and three Rhodian hemidrachmas, four diobols and some bronze money (not current) in a small, ragged cloth, and some small pieces of gold plate from the treasure: (weight) 15 drachmas and 1 obol. These are contained in a stamnos.
This use of cloth is far from unparalleled in Greek sanctuaries. In the temples of Artemis Brauronia and Apollo at Delos, a piece of linen (ὀθόνιον, which was not necessarily ragged) was used to wrap different objects and materials, such as soft wool (ἔρια μαλακά), anklets (πϵρισκϵλίδϵς), coins (νόμισμα), etc.Footnote 58
Consequently, it is not inconceivable that old, ragged cloths were used in Classical times for bundles. Dorkilos may have been suspected by his master(s) of stealing a piece of ragged cloth (λάκος) in which valuable objects were wrapped.
V. Conclusions
If our analysis is correct, a natural semantic change from ‘garments’ to ‘rags’ is nowhere to be found in classical Greek. The obsolete nouns λαῖφος and σπϵῖρον were misinterpreted as synonyms of ‘rags’ by late lexicographers only because both words occurred in some Homeric passages which deal with beggars and outcasts clad in tatters. However, a careful analysis of the texts unequivocally proves that λαῖφος and σπϵῖρον denoted pieces of clothing of various types. Needless to say, all those garments could be in perfect condition or in tatters.
Evidence for a hypothetical semantic evolution ‘rags’ > ‘garment’ is hardly provided by some passages in which ῥάκος and its Aeolic correlate βράκος ironically allude to garments. As far as we can see, this antiphrastic use did not bring about an actual change in the meaning of these words.
Finally, the linguistic analysis of this paper contributes to a more reasonable interpretation of the enigmatic theft of a λάκος, a rare synonym of ῥάκος, in a Dodonaean enquiry. Instead of a garment, we consider two possibilities: the λάκος was either a consecrated piece of cloth worn by the passage of time, or a tattered cloth used to bundle and carry various unnamed goods. Of course, we cannot boast of having solved the case: only omniscient Zeus knew whether Dorkilos actually stole the λάκος!
Acknowledgements
We would like to thank I. Katsadima (Ephorate of Antiquities of Ioannina) for her help in trying to find Dorkilos’ lamella in the Archaeological Museum of Ioannina and the anonymous reviewers for their valuable remarks and corrections. Angela M. Andrisano kindly sent us a copy of her paper on Sappho fr. 57 Voigt.
Funding Statement
This paper is part of the research programme ‘Phonology and Dialectal Contact in Ancient Greek’ (FFI2017-82590-C2-2-P) funded by the Spanish Ministry of Economic Affairs and Digital Transformation (MINECO).
