• 1  See Langella (2019), p. 180–198, and Pirenne-Delforge (2024), with extensive prior bibliography an (...)

1Two thousand years ago, Diodorus of Sicily, doubtless following others, was already attempting to explain the name of Demeter θεσμοφόρος. Some twenty centuries later, linguists, philologists, and historians of Greek religion are still trying to untangle this skein, as shown, among others, by two recent studies by E. Langella and V. Pirenne-Delforge, on which the present paper is based.1 By the number and scale of the broader issues to which it is connected, the problem of the meaning and history of the divine epithet θεσμοφόρος is in fact far from negligible : the Thesmophoria were among the most widespread religious festivals in the Hellenic world in the archaic and classical periods.

• 2  I summarize here the main points of Pirenne-Delforge (2024), to which I refer for further detail.

2Several general aspects of Demeter’s divine power, of the Thesmophoria festival, and of the history of this cult are fundamental for understanding the original sense of the epithet θεσμοφόρος :2

• 8  In a compound such as θεσμοφόρος it is impossible a priori to determine whether θεσμο- is to be un (...)
• 9  See already Farnell (1907) and (1911); Brumfield (1981), p. 70–73, followed by Stallsmith (2008); (...)

3On a more strictly linguistic level, several important points have already been brought to light. Demeter’s epithet θεσμοφόρος, an agentive compound formed from the stem of the noun θεσμός and the root φορ- of the verb φέρω, is equivalent to the phrase “Demeter who φέρει the θεσμός/θεσμοί”.8 To reconstruct the earliest sense this epithet may have had, one must therefore grasp the contextual meaning of θεσμο- and of -φόρος in such a compound, without forgetting that the implicit subject of the phrase crystallized by the compound is Demeter, and no one else. In particular, the epithet does not apply to the women celebrating the Thesmophoria, who are consistently called θεσμοφοριάζουσαι in our sources and never θεσμοφόροι, an appellative reserved for Demeter (and sometimes Persephone).9 Likewise, the epithet θεσμοφόρος should not be explained from the symbolism of this or that element in the unfolding of the Thesmophoria, since it is the festival that takes its name from that of the goddess who is the recipient of the cult, and not the other way round : the epithet must therefore refer, in one way or another, to an aspect of Demeter’s divine power that stood at the heart of the Thesmophoria in general.

• 10  Cf. Callim. Hymn. Dem. 18–21 and Diod. Sic. I, 14, 1–4 and see Langella (2019), p. 181–182.
• 11  See Stallsmith (2008), together with the critiques of Pirenne-Delforge (2024) and Langella (2019), (...)
• 12  I summarize here Langella (2019), p. 180–186 (adding a few occasional personal considerations), to (...)

4On the other hand, there is an old etymological interpretation of θεσμοφόρος, already implicit in Callimachus, later formalized by Diodorus of Sicily,10 and recently taken up afresh,11 according to which the epithet means “she who brings the laws”, i.e. the customs, norms, and rules governing agriculture. This interpretation, however, is unlikely to correspond to the most ancient sense of the compound θεσμοφόρος, for several reasons :12

• 17  e-grade *θη- <PIE *d^heh₁-; ø-grade *θε- <PIE *d^hh₁-. See DELG and EDG s.v. θεσμός, τίθημι, θωμός, (...)
• 18  The following excludes θέμις (obscure formation, cf. gen. θέμιστος; derivation from *θη-/*θε-uncer (...)

6The etymological senses of a word like θεσμός derive from its root—reconstructed from its derivatives—and from its suffixal formation, which adds a specific nuance. These “predictable” senses are general possibilities, and only the examination of texts shows which specific and contextual meanings are actually attested. Thus, the noun θεσμός is derived from the Indo-European verbal root *θη-/θε- (<PIE *d^heh₁-), which in Greek yielded the verb τίθημι, “to set something intended to last” (Chantraine), hence “to set, to place, to establish, to deposit”.17 Stillness and durability, tied to the notions of location, establishment or storage, are essential to grasping the semantics of this root. Among its nominal derivatives are the following primary formations :18

• 28  Chantraine (1933), p. 132–150, groups these under “action nouns” without semantic distinction and (...)
• 29  Lit. “object that has been plaited” (πλέκω); “liquid that has been poured” (χέω); “portion that ha (...)
• 30  Lit. “what serves to irrigate/water” (ἄρδω); “what serves to filter” (ἠθέω); “what serves to haul (...)
• 31  Lit. “place where one hides” (κεύθω); “place where one stops” (ἵστημι).
• 32  Later formations are predominantly abstract (“the action of…”), like -α-σμός and -ι-σμός nouns fro (...)

8Among all these nouns, θεσμός is especially close in formation to θωμός, “heap, pile, deposit”. Nouns in -μός (-σ-μός, -θ-μός, etc.) form a particular class of action nouns that differ in part from action nouns in -μα and especially from those in -σις such as θέσις.28 Many action nouns in -μός, several already attested in Homer, in fact have a resultative, passive, and very concrete sense. That is, they designate the object resulting from the accomplishment of the action, rather than the action taken in itself : cf. πλόκαμος, “lock (of hair)”, and πλοχμός, “braid” (Hom.); χυμός, “juice/sap” (Hipp., Plat.) ; ψωμός, “morsel, mouthful” (Hom.); κορμός, “tree trunk” (Hom.) ; τῖμος, “price” (Aesch.) ; etc.29 In some cases, the action noun in -μός is so concrete that, by metonymy, its sense is closer to that of an instrumental noun than to that of an action noun strictly speaking : cf. ἀρδμός, “water for irrigating/watering” (Hom.); ἠθμός, “sieve” (Eur.); ῥυμός, “drawbar” (Hom.).30 Sometimes it can even approach a place-noun : cf. κευθμός, “hiding place” (Hom.), and σταθμός, “stall, fold, dwelling”.31 Here too, among these nouns, there are several more or less technical terms belonging to the vocabulary of agriculture, animal husbandry, or other manual activities. These concrete, even technical senses are sometimes accompanied by a more abstract and expected meaning directly drawn from the verbal root (“the act of…”), a meaning that not all -μός nouns possess, especially in Homer.32

• 33  For a summary of the debates, see Heubeck et al. (1981–1986), ad loc.

10Keeping the etymological senses of θεσμός in view, let us examine two of its earliest attestations. The meaning of this Homeric hapax has been much debated since antiquity—likely without good reason : some ancient and modern commentators have persisted in giving it a sense that is not only doubtful in context, but also anachronistic in light of the attestations and far from its etymological senses.33 Penelope and Odysseus have been reunited after more than ten years apart :

• 35  Cf. with bare accusative: Il. XI, 352 (of a missile: οὐδ’ ἵκετο χρόα καλόν); Od. XIX, 451 (οὐδ’ ὀσ (...)
• 36  Contra the exegetical note of Schol. vet. in Hom. Od. XXIII, 296 (followed by some modern translat (...)

11In λέκτροιο παλαιοῦ θεσμὸν ἵκοντο, the noun θεσμός complements the verb of motion ἱκνέομαι, which, in its very frequent Homeric sense “to arrive at, reach, gain”, takes an accusative of destination with a common noun denoting a place or marker (here θεσμόν), optionally specified by a genitive (here λέκτροιο παλαιοῦ).35 Thus the phrase θεσμὸν ἵκοντο can only mean “they reached the location”, where θεσμός clearly designates the place where the object has been set, while λέκτροιο παλαιοῦ specifies which location is meant (that “of their old bed”).36 The poet’s precision about this site is fully justified narratively by the famous exchange between Penelope and Odysseus at lines 174–204, where she tries to trap him by mentioning the moving of the marriage bed, in fact immovable.

• 37  Some material in Apollonius’ Lexicon may go back to the sources of the D scholia on Homer—possibly (...)

12Let us turn to a testimony from the end of the archaic period. In his Homeric Lexicon, Apollonius the Sophist (first century CE) preserves lexicographical material that mostly goes back to the Alexandrian philologists of the third-second century BCE, and in some cases even earlier :37

• 39  νῦν δ’ οὔτ’ ἄρ πῃ θέσθαι ἐπίσταμαι, οὐδὲ μὲν αὐτοῦ | καλλείψω, μή πώς μοι ἕλωρ ἄλλοισι γένηται, “N (...)
• 40  See also Pirenne-Delforge (2024), p. 199 (“The thesauros is a secure container…”), apt for θησαυρό (...)

13Regarding the aorist middle infinitive θέσθαι, “to store (a treasure) for safekeeping”, in Od. XIII, 207–208,39 Apollonius informs us that there exists a γλῶσσα—that is, a poetic and obsolete term—θεσμός, derived from the same root as θέσθαι and close to θησαυρός, “store”. To support this definition of θεσμός, “store, deposit”, he cites a verse of Anacreon otherwise unknown : ἀπὸ δ’ ἐξείλετο θεσμὸν μέγαν, “and he took out a great deposit”. The antiquity and reliability of this information—both linguistically correct and exegetically pertinent—are not in doubt : in the poetic language of Anacreon (second half of the sixth century BCE), θεσμός could indeed mean “store, deposit”, i.e. “what has been deposited”.40 Given Anacreon’s themes, it is likely a deposit of money, but without context any more precise semantic interpretation remains speculative.

• 41  See Langella (2019), p. 180–181, with bibliography, and particularly Ostwald (1969), already quote (...)
• 42  One may add that the shift from concrete to abstract is also perceptible in the history of the sen (...)

14The two early senses attested in Homer and in Anacreon (via Apollonius) are therefore “site, location” (cf. etym. sense b) and “store, deposit” (cf. etym. sense a), both entirely concrete and very close to the verbal root. The abstract meanings “law”, “rule”, “precept”, “rite”, etc. are attested from Pindar and Aeschylus onward, reflecting a semantic shift from concrete to abstract, a development amply studied and not pursued here :41 starting from “thing set or deposited” (etym. sense a), one can easily posit an abstract extension to “rule”, then “law, order, custom, rite”, etc.42

• 43  Cf. Hom. Od. V, 368 (ἠΐων θημῶνα… καρφαλέων); Arist. Meteor. 344a 26 (θημών ἀχύρων); Op. Hal. IV, (...)
• 44  Cf. Aesch. Ag. 295; Ar. Lys. 973; Thphr. HP, IV, 13 (twice) and VIII, 11 (twice).
• 45  Cf. PRyl. 199.12 (1st c. CE).

16The old agricultural term θημών means “heap, pile”, said of stubble, wheat, or straw in Homer, Aristotle, and Oppian.43 Hesychius also preserves the noun θημωνιαί (θ 501)/θειμωνειαί (θ 187), derived from θημών, glossed as οἱ σωροὶ τῶν δραγμάτων, “heaps of harvested sheaves”. The same noun in the plural, according to Hesychius (θ 504, θημῶνες· θημωνιαί), was a synonym of θημών in the sense “heaps of sheaves of ears”, “heaps of wheat grains”. Soon fallen into disuse and employed as a poetic γλῶσσα by Homer’s imitators, θημών was partly replaced by θωμός, of very similar sense. The noun θωμός, much better attested, means “heap, pile” and is said of heather in Aeschylus, while in Aristophanes and especially Theophrastus it denotes a “heap of (grains of) wheat”; it thus takes over from θημών in the technical agricultural vocabulary of the classical period.44 Finally, θέμα, which likely succeeds its older counterpart θῆμα, is attested with the sense “deposit of wheat” in Roman-period accounting papyri.45

• 49  I had already noted this testimony in Felisi (2023), p. 102 n. 11 (with bibliography), in a synthe (...)

20Moreover, a fairly unique testimony—much earlier than Hesychius and Eustathius—securely attests an agricultural sense of θεσμοί, “stores of wheat-grain”, and links it directly to Demeter.49 The Roman scholar Cornutus (first century CE), in his Compendium of Greek Theology, draws mainly on the lost Περὶ θεῶν by Apollodorus of Athens (second century BCE), itself a compilation of much older sources :

• 50  The synchronic etymology θεσμός/θησαυρίζεσθαι is probably wrong, but diachronical and morphologica (...)

21Cornutus, explaining Demeter’s epithet θεσμο-θέτις, reports a debate over the sense of θεσμός : according to ancient sources he consulted, θεσμός is in fact the καρπός, the “harvested fruit”—most often “wheat”, the καρπός par excellence—so called because it is “put in deposit” (ἀπο-τίθεσθαι) and “stored” (θησαυρίζεσθαι), recalling Apollonius’ lemma with the Anacreontic fragment (θεσμός = θησαυρός).50 This may well be a genuine lexical fact concerning a usage otherwise lost, due to the development of θεσμός, “deposit, store”, into a predominantly abstract sense (“institution, norm, custom”) and/or a legal one (“law”) from the end of the archaic period onward.

• 51  The noun ἄγγος, “vessel”, is not a technical term. I follow West (1988) in rendering it “storage-j (...)

22On the other hand, already in Hesiod’s Works and Days (esp. v. 473–478, 597–603) we find a fairly precise description of the stages after the harvest : the ears of wheat are gathered on a threshing floor, the grains removed and separated from their husk ; the grains not used for food or trade are then stored in purpose-made jars—Hesiod calls them ἄγγη (v. 475, 600), to be understood as “storage-jars” (West)51—to prevent spoilage over the summer. In this description the noun θεσμός is absent, but the act of storing the grain is expressed by the verb κατα-τίθημι (v. 600–601) : αὐτὰρ ἐπὴν δὴ | πάντα βίον κατάθηαι ἐπάρμενον ἔνδοθι οἴκου, “When you have stored away all your substance under lock inside the house…” (West). Thus, θεσμός may indeed have been the name of the object resulting from the action expressed in this verse by κατα-τίθημι.

• 52  I owe this information to Langella (2019), p. 186–187 (with earlier bibliography, n. 26).

23These valuable data from ancient lexicographical and antiquarian sources warrant positing a technical agricultural usage of θεσμός, “heap, deposit, store of grain”, in Greek (cf. etym. sense a). But how early can this semantic usage be ? Indo-Iranian, Baltic, and Hittite have all preserved traces of an old Indo-European noun built on the o-grade base *d^hoh₁- of the root *d^heh₁- (Greek *θη-/θε-), from which τίθημι, θεσμός, etc., derive and which is identical with the base of θωμός. In these Indo-European languages the noun is suffixed *-no-/-neh₂- and is attested with the senses “grain, cereal” (cf. Vedic dhāná-, Avestan dāno-, Tocharian B tāno-) and “bread, piece of bread” (Hittite -dannaš, Lithuanian dúona, Latvian duõna).52 If these data are correct, the Indo-European comparison provides striking and very ancient parallels for the use of nominal derivatives of PIE *d^heh₁- (Greek *θη-/θε-) in an agricultural sense specifically referring to cereal grains. In Greek, the old agricultural use of θεσμός must have fallen very early into disuse— like the nouns θημών/θημῶνες and θημωνιαί (and perhaps the phantom *θημοί)— and been replaced by a single term : θωμός, “heap” (of grain, straw, heather, etc.). The semantic link between the root *θη-/*θε- and agricultural vocabulary was thus likely lost at a quite early date, all the more since the ω-vocalism of θωμός did not favor its association with the family of τίθημι. In this, the influence of nouns such as θέσις, “the act of placing”, and θέμις, “divine law”, etc., must have played a decisive role in distancing θεσμός still further from its ancient agricultural uses.

• 53  What follows is based on Langella (2019), p. 193–194, with a few additions from DELG and LSJ, s.v.(...)
• 54  Cf. also Aesch. fr. 196.5; Pind. N. XI, 41; Hdt. V, 82.3; VI, 139; Eur. Hec. 593; Xen. Vect. 1.5; (...)

24Before we can employ the reconstructed agricultural use of θεσμός to interpret θεσμοφόρος, we must clarify what the second member -φόρος could mean in such a compound : the two elements interact both syntactically (with θεσμο- as the object of -φόρος) and semantically.53 This ground is well known : the Greek base φερ-/φορ- goes back to the Indo-European root *b^her-, “to carry”, which in Greek yields the lexical family of φέρω. The verb means “to carry” (hence “to bear, endure”), “to transport”, “to bring” (hence “to offer, deliver”), etc., but also “to produce, to bring forth, to bear”, especially in agricultural contexts or with reference to the earth and the plant world. Thus, “the island would produce all fruits in their season” (Hom. Od. IX, 131, [νῆσος] φέροι ὥρια πάντα); “the grain-giving ploughland produces the most remedies” (Od. IV, 229, πλεῖστα φέρει ζείδωρος ἄρουρα φάρμακα) ; “the grain-giving ploughland produced much fruit” (Hes. Op. 117, καρπὸν δ’ἔφερεν ζείδωρος ἄρουρα) ; “the vines produce wine” (Hom. Od. IX, 110, ἄμπελοι φέρουσιν οἶνον).54 But above all, what the earth “produces” (φέρει) is “Demeter’s grain”, its fruit par excellence, as Hesiod says (Op. 31–32) : ᾧτινι μὴ βίος ἔνδον ἐπηετανὸς κατάκειται | ὡραῖος, τὸν γαῖα φέρει, Δημήτερος ἀκτήν, “[a man] who has not food for the year laid up at home in its ripeness, produce of the earth, Demeter’s grain” (West).

• 55  Cf. Hom. Il. XII, 314; XIV, 123; Simon. fr. 15.
• 56  Cf. Aesch. Pers. 618; Hdt. 7.8 αʹ; Hipp. Coac. 502; Pl. Leg. 704c; Thphr. HP III, 2, 6.
• 57  Cf. Hdt. I, 193; II, 156; Xen. Cyr. VI, 2, 22, etc. (trees); Pind. P. IV, 6; Eur. Hel. 1485, etc. (...)
• 58  Cf. Ar. Ra. 384; Paus. VIII, 53, 7; also epigraphically: CIG 4082 (Pessinus), IG XII(5).226 (Paros (...)

25Likewise, several -φόρος compounds already in the earliest Greek sources mean “that produces…”, notably of the earth or Demeter herself on the one hand, and of cereals on the other : thus, the earth is πυροφόρος,55 “wheat-bearing”, and πάμφορος,56 “all-bearing”; trees or lands are καρποφόροι,57 “fruit-bearing”, and Demeter herself is καρποφόρος,58 “who produces the fruits (of the harvests)”. Therefore, the sense of -φόρος to retain in interpreting θεσμο-φόρος, in line with the agricultural sense established for θεσμο-, is to produce, to bear, to bring forth”.

• 59  On the link between πλοῦτος and the god of the dead’s euphemistic by-name Πλούτων, see Felisi (202 (...)

26We can now put together the different pieces of this puzzle. We have seen that θεσμός, like other words in its family meaning “heap, deposit”, could have an agricultural use designating a “deposit or store of grain”. This deposit would be nothing other than the “stores of wheat” with which every household was supplied, as Hesiod and Xenophon attest for different periods (see below). As the ancient interpretation that Cornutus rejects suggests, the θεσμός was indeed the result of the May-June harvests, whose surplus was laid in deposit and carefully preserved. This surplus was also called πλοῦτος, “wealth”, since it enabled the agricultural cycle to begin again with the sowings in October and November, at the time of the Thesmophoria.59 If the θεσμός was abundant and constituted a genuine πλοῦτος, this was because the previous year’s harvests had been fruitful. The intervention or cooperation of Demeter was evidently essential for the fields to produce many ears of wheat, and thus potentially much grain to be stored with a view to sowing. This is a wish clearly attested in Hesiod, who directly links Demeter with the storage of wheat-grain in the granary :

• 61  In her excellent study, Langella (2019) reaches a result akin to mine by a wholly different route (...)

27Thus, Demeter truly appears as the goddess θεσμο-φόρος, “she who produces (-φόρος) the stores of grain (θεσμο-)”, filling the granary thanks to the fields’ fertility. On this reconstruction, then, θεσμο-φόρος would be a near-synonym of καρπο-φόρος, differing only in the object of φέρω in the compound : καρπο-φόρος emphasizes the harvested grain (καρποί), whereas θεσμο-φόρος would emphasize the stored grain (θεσμοί)—that is, the very grain stores brought out at the time of the Thesmophoria to begin sowing.61 These stores were likely central to the celebrations, since they made possible the activity the festival celebrated : sowing itself.

• 62  The following ancient references are drawn from Langella (2019), p. 189.

28Can we find parallels for θεσμο-φόρος, “she who produces the stores of grain”, in Demeter’s onomastic repertoire ? The indications, once again, are not lacking.62 Book 6 of the Greek Anthology contains a dedicatory epigram of a sacrifice to Demeter, attributed to Adaeus, an epigrammatist of the fourth century BCE, to whom we owe several poems evoking rustic life :

• 64  Cf. Athenaeus, XIV, 10 (= Semos, fr. 19): “The handfuls of barley, taken each by itself, they used (...)
• 65  Cf. Eust. Thess. Comm. in Hom. Il. vol. IV, p. 252 van der Valk: “[…] from this hamallē, too, [com (...)

29After listing the offerings he has set out on the threshing-floor for the goddess, Krethon invokes Demeter under the name πολύσωρος, literally “rich in heaps (of grain)”. The σωροί were precisely the heaps of wheat-grain from the harvested sheaves, sorted on the threshing-floor (see above Hesiod), which were then stored in the granary. In other words, on our semantic reconstruction, once stored the σωροί constituted the θεσμοί. Likewise, in an Orphic Hymn (XL, 5) Demeter is called σωρῖτις, “giver of heaps (of grain)”. Other onomastic attributes, by contrast, stress heaps of sheaves rather than heaps of grain : so Ἰουλώ, “goddess of sheaves”, a ritual epithet of Demeter mentioned by Semos of Delos (3rd–2nd c. BCE) in his lost On Paeans,64 and ἀμαλλοφόρος, “bearing sheaves”, a name under which, according to Eustathius of Thessalonica, the Athenians sacrificed to Demeter.65

• 66  On the mud-brick granary at Eleusis used to store first-fruits for Demeter and Persephone (cf. the (...)
• 67  See Spatafora (2004) and Laudani et al. (2016), respectively.
• 68  See Clinton (2009).

30The epithet θεσμοφόρος, “she who produces the stores of grain”, giving its name to the Thesmophoria, would thus designate an aspect of Demeter’s divine power that was venerated by the Greeks year-round, in sanctuaries likewise called θεσμοφόρια. It was common practice to store grain provisions in rural temples dedicated to the two goddesses, as archaeology shows—above all for Eleusis and for Sicily.66 The Thesmophorion of Entella in Sicily (sixth–fifth century BCE) is particularly noteworthy : the cult of Demeter Thesmophoros is well attested there by several statuettes found in the excavations, while the entire building abutted a granary that formed an integral part of the sanctuary.67 As for the temple of the two goddesses at Eleusis—whose links, stronger or weaker, with the Mysteries or with the Thesmophoria remain debated—we know of offerings of κέρνη (whence the journal title Kernos ! ), special vessels with compartments to hold cereal grains.68

• 69  Cf. Hom. Od. II, 374–380 (Eurycleia supplying Telemachus with meal for his departure) and XX, 105– (...)

31As recalled in the introduction, throughout the Greek world the Thesmophoria were a women’s festival connected with the sowing. Can we then draw a meaningful link between the θεσμοί, “stores of grain”, and women ? Already in Homer, female figures have a privileged relation both to provisions and to grain.69 Most explicitly, the already quoted passage from Hesiod’s Works and Days (v. 600–603) states : “When you have stored away all your substance under lock inside the house, I suggest you set about engaging a man with no household, and seek a woman without a child—a working woman with calf at her is a nuisance” (West). It is easy to suppose that the male servant the farmer was to hire, according to Hesiod, was chiefly tasked with guarding against outside thieves. The female servant, by contrast, was surely charged with managing those same stores, in connection with preparing meals and wheat-based products (flour, cakes, etc.).

• 70  Cf. Xen. Oec. VII, 21–22: “when these goods are brought under the roof, someone must keep them saf (...)

32For the classical period we have much more explicit testimonies linking women to the management of grain provisions stored in the household. In the Oeconomicus Xenophon has the wealthy landowner Ischomachus say that his house contains a room set aside entirely for storing grain (IX, 3), and he illustrates the distinction between “outdoor tasks” (harvesting, sowing, planting, etc.), assigned to men, and “indoor tasks”, assigned to women : among the latter are preserving the produce of the harvests, managing and arranging the grain stores, and transforming the produce into food.70