Forty years ago, in 1968, Sebastián Bartina published this Hesiod papyrus[1] in one of the first volumes of Studia Papyrologica, the journal that Josep O'Callaghan, creator of the Palau-Ribes collection, founded in 1962 to host and promote the papyrological studies in Catalonia and Spain. At the 25th International Congress of Papyrologists I had the pleasure of contributing a paper about the "rediscovery" of the Palau Ribes collection. The appearance of the re-edition of this papyrus, which I have been planning for some time, in the proceedings of the Congress is timely, for it offers a good example of the resumption of work on the Palau-Ribes material and the new dynamism we wish to inspire to papyro–logical studies in our country. In fact, the papyrus itself was among those proposed to the students in a recent summer seminar at the present location of the Palau-Ribes collection.[2] By re-editing the text, I do not intend to criticize the work of my predecessors, but to honor their memory by adding my contribution to their scientific efforts.

    Hesiod, Theogony 862–872
    P.PalauRib.Lit. 9 Plate I 4.7 x 7.2 cm
    III AD (beginning) Unknown

    The papyrus contains remains of 11 lines of writing belonging to a single column. The text runs along the fibers and was probably part of a roll. The back is blank, except for a seemingly accidental spot of ink with no definite shape. Only 1 cm of the upper margin has survived.

    The handwriting is roughly bilinear, for letters do no sit regularly on the base-line and the long descenders of ρ, φ, ι and υ, protrude below the line. The mid-sized capitals are in general quite regularly spaced and, with the exception of the middle horizontals of ε and θ, they hardly ever touch each other. Slight variations in size and spacing, as well as the somehow narrow interlinear space, contribute to a general air of informality, although the hand is certainly experienced. It shares many characteristics of the Formal Mixed or Severe Style – following Turner-Parsons' terminology –[3] such as the contrast between a set of broad letters (conspicuously η, μ, υ and ω) and smaller, narrow letters (typically ε, θ, ο and ς), and the angularity of shapes, especially noticeable in the case of the round letters: ε and ς with straight backs and upper parts drawn in a separate movement, lower loop of β with a triangular shape, pointed left-hand elements of α and δ, etc. However, the central parts of μ and ν and the right-hand oblique of α are curved. Shading is also characteristic of this style of writing: thin horizontal strokes contrast with thick vertical strokes, and oblique strokes asceding to the right are thinner than those descending to the right. Other feature often associated to the Formal Mixed style is the slant to the right, even if in our case this is not very conspicuous, and the overall impression is that of an upright writing.

    Ornamentation is not frequent and is restricted to occasional ticks to the left at the foot of long descenders (e.g. ι, ρ, υ) and sometimes at the top of verticals (e. g. in η or ν). Likewise, the left-hand oblique of υ may present a similar tick at its top. The text is provided with accents: acute (ll. 862, 869, 871), grave (l. 869), and circumflex (ll. 863, 868), and has been corrected, apparently by a different hand (see commentary to ll. 863 and 867). The preserved portion of text presented no possible cases for elision in the original paradosis, but it was subsequently effected and marked by a second hand as a correction (l. 867). Line 865 provides the only place in the papyrus where the practice of the scribe regarding iota adscript can be safely ascertained: it is absent.

    Classical exemplars of the Formal Mixed comparable to our manuscript are P.Berol. inv. 9766[4] (Plato, Laws), with a slight curvature in the diagonal of ν, μ with rounded central elements and pointed left-hand parts of α and δ. Long descenders also tend to protrude well below the line, as in our manuscript (ρ, υ, φ and, here, τ more than ι). The hand is, however, more formal than in our papyrus. P.Heid. inv. 1701[5] also compares well with our hand, but is still formal, even if in its general appearance more similar to our somehow thick, not stylized strokes (but modular contrast between narrow and small letters is more pronounced, see, for instance how ω is smaller and suspended high in the line). Both are assigned to the third century AD. P.Oxy. XXXIV 2699 (Apollonius Rhodius, Argonautica),[6] also assigned to the same date, shares the informal air with our script, although ductus and letter shapes, typical for the Formal Mixed style, are not particularly similar to ours. There are more similarities, however, with P.Oxy. LXVIII 4649 (Prose quoting Hesiod, Theogony 6–7): although the material is scanty for comparison of all letters, μ has the curved central elements, ω has the same height and width as the rest of the broad letters, and the triangular set of letters keep their angularity. Similarly, long verticals descend below line-level, but shading, slanting or any other kinds of ornaments are not very pronounced.[7] Since this papyrus has also been assigned to the third century, I would rather suggest the beginning of this century as a probable date for our papyrus. Thus, rather than showing an early stage in the formation of the style, the unperfected appearance of the script shows the informality of the production.

    Apart from l. 863 (see commentary), the text agrees with the manuscript tradition,[8] but an orthographic mistake of a phonetic nature (l. 865), the character of the correction in l. 867, simplifying the text, and the error itself in l. 863 (very probably due to lack of attention) point, together with the already mentioned informal air in the execution, to a non-professional copy of the text and, at the same time, seem to exclude the possibility of a scholar's private copy. However, the fact that the handwriting is an experienced one, and already able to attain the standards of a type such as the Formal Mixed, as well as the presence of accents seem to indicate that the text may have been copied by an advanced student, and probably then corrected by the teacher (who is not responsible, however, for the correction in l. 867; see commentary to the line).

    Bartina also studied the relationship of our papyrus with the rest of the Theogony papyri in an article published one year after its editio princeps, following the appearance of many new Hesiod papyri.[9] He identified four containing passages of the text partially overlapping with ours; I cite them here according to West's list:


    P.Vindob. inv. G 19815; MPER 1.73–83, from the second half of the fourth century AD. Hesiod, Theogony 626–881; Works and Days 210–828; Shield 1–32, 350–470 (lac.)

    Lines 871–872 are preserved in both papyri, but only partially, since Π5 contains only the line-ends, which are absent in our papyrus.


    PSI IX 1086, assigned to the end of the second or beginning of the third century AD.[10] Hesiod, Theogony 839–869. Differences from our text are registered in the commentary to this edition, even though they are not substantial.


    P.Oxy. XXXII 2639 + PSI IX 1191; assigned to the second or third century AD. Hesiod, Theogony 57–75, 84–96, 659–664, 866–876, 913–932. It transmits the end of ll. 866–872. Since our papyrus has not preserved this part of the lines, comparison between the text of the two papyri is not possible.


    P.Ant. II 71; assigned to the fifth or sixth century AD. Hesiod, Theogony 825–853, 868–896; it contains scanty remains of the middle of lines, which prevent any coincidence with our text.[11]

    Since 1969 a new Hesiod papyrus has been published containing portions of text present in ours: P.Oxy. LXVIII 4657, edited in 2003 by P.J. Parsons, and assigned to the second century AD, transmits Theogony 820–831, 859–865. A few letters of ll. 862–865 are present in both papyri, which do not show any textual divergence.[12]

    On palaeographical grounds or otherwise, the possibility that any of these papyri should belong together with our fragment is excluded. Furthermore, I have been unable to identify any other Hesiod papyri[13] as belonging to the same roll as our fragment. The closest parallels found among other Hesiod papyri in palaeographical terms are the two small fragments mentioned in the description of the hand, but it is clear that they were not written by the same hand as ours.

    I have discussed the presence of diacritical signs in papyri elsewhere,[14] arguing that in many cases rather than performing a purely practical task in helping the reader articulate the text, they reflect the rich philological activity which goes back to Alexandrian scholarship. Nonetheless, it is also true that this philological activity partially finds its origin in the explanatory nature of the work of the γραμματικοί, whose function was to explain the literary text and whose work is so graphically reflected in lists of words, glossae or D-Scholia, in the case of Homer.[15] In this respect, what we find in our papyrus is not so much traces of highly learned scholarship as traces of the work of a grammatikos, who, in the educative process, explains rare words, mythological characters, unusual forms and peculiar grammatical phenomena. Signs in the text may serve to mark such passages, which would then be explained in a separate commentary or word list, and accents may be used in precisely this way, regardless of the prosodic characteristics of the word in question. As a matter of fact, of all the known glossed words in ll. 862–872[16] we only have two in the papyrus (θεσπεσίῃ, αἰζηῶν, ll. 862 and 863, respectively), and both carry an accent. Of course there are more instances of accents in the text, which I will try to account for in the commentary.

    αυτμη θεσπ]εσίη[ και ετηκετο κασσιτερος ως
    τεχνη υπ αι]ζ̣ηῶ`ν´ ο ̣[
    θαλ]φθεις ηε σιδηρος[ ο περ κρατερωτατος εστιν
    865ουρ]ε̣ος εμ βησσησι δ[αμαζομενος πυρι κηλεω
    τηκ]εται εν χθονι δ̣[ιη υφ Ηφαιστου παλαμησιν
    ως α]ρ᾿ ⟧α⟧`ε´τηκετο γαι[α σελαι πυρος αιθομενοιο
    ριψε δε μι]ν θυμῶ[ ακαχων ες ταρταρον ευρυν
    εκ δε Τυφω]έος ὲσ[τ ανεμων μενος υγρον αεντων
    870νοσφι Νοτ]ου Βορε[ω τε και αργεστεω Ζεφυροιο
    οι γε μεν εκ ]θεόφ[ιν γενεην θνητοις μεγ ονειαρ
    αι δ αλλαι μα]ψ̣ α̣υ̣[ραι επιπνειουσι θαλασσαν
    – – – – – – – – – – – – – – – – – – – – – – – – – – –

    Unless otherwise stated, I provide West's text[17] in the lacunae, and will only discuss textual variants preserved on the papyrus. I have printed lectional signs only when present in the papyrus, and have omitted iota adscripts, following the scribe's practice.

    862 θεσπ]εσίη: The accent appears displaced to the right; this practice, not unusual in the papyri, has sometimes been attributed to rapid or inaccurate writing.[18] The expression has been the object of an explanation: cf. Glossae in Theogoniam 862.1 ἀϋτμῇ θεσπεσίῃ. τῇ τοῦ πυρὸς ἀτμῇ M2.

    863 αι]ζ̣ηῶ`ν´ ο ̣[ : Our papyrus does not seem to agree with the manuscript tradition: ὑπό τ᾿εὐτρήτου χοάνοιο, nor does it agree with the only papyrus which has preserved the text at this particular point, Π12 in West's list: υπο τ ευτρη[ . Peppmueller had suggested ἐν εὐτρήτοις χοάνοισι for stylistic reasons, but the remains of ink at this point do not confirm his proposal, for they are not consistent with ε, but o. Obviously, the scribe – or corrector – has perceived some problem at this point, as the superscription above the line, to the right of ω, shows. In this respect, attention should be drawn to the fact that this passage still seems troublesome to modern editors,[19] and that P.Oxy. 4657 presents a iota mutum suprascript after η. Though an error, as P.J. Parsons rightly observes, it may be further evidence of confusion at this point.[20] In any case, I do not think, as Bartina and Übel interpreted, that the superscription is a sign for an abbreviation: (υπ)ο.[21] I rather see a ν, of which the second vertical has been almost completely lost in the lacuna. Its shape is somehow different from that in the main text; it looks, in fact, more similar to one of those characteristic of the Severe Style, without the curvature in the diagonal stroke it presents in the main text. I think it likely that the letter was written by a second hand, perhaps that of a teacher, already mastering this writing style in all its formality.

    A hypothesis to explain the second half of the verse might be that our student, having completed the first half of the line (we have masculine caesura after αἰζηῶν), started to copy the second half of the next one: ὅ περ κρατερώτατός ἐστιν, after the feminine caesura: the traces following ο, namely the left-hand end of a high horizontal, are consistent with the high horizontal of π. This lack of attention might have been favored by the somehow parallel constructions used for both metals in the simile: κασσίτερος ... τέχνῃ ὑπ᾿ αἰζηῶν ... θαλφθείς; σίδηρος ... δαμαζόμενος ὑφ᾿ Ἡφαίστου παλάμῃσιν, thus making possible the transposition at this point without ruining completely the sense. If this was the case, however, we cannot know whether the error was subsequently corrected, because the second half of the following line has also not been preserved.

    For the circumflex accent in αι]ζηῶ`ν´, the Glossae give an equivalent to the expression in the normal speech: Glossae in Theogoniam 863. 1 αἰζηῶν. νέων BM v. Hesych. αἰζηοί. νεανίσκοι. νέων τεχνημάτων (l. τεχνημόνων) M2.

    865 εμ for εν is a common phonetic mistake. Bartina prints ν in his edition at this point, but the breadth of the letter, the fact that the oblique stroke does not touch the lower end of the second vertical, and that there seems to curve up when it joins it – even if not so clearly as in the other instance of μ we have in the text (l. 868) – make me think that we have a μ rather than a ν.

    866 τηκ]εται εν: The papyrus does not support τήκεται δ᾿ ἐν, in vQ, or τήκετο δ᾿ ἐν in S, thus agreeing with the rest of MSS and Π12.

    δ̣[ιη: The two strokes forming the left-hand vertex of δ do not join together; it is certainly not to be seen elsewhere in the papyrus for this letter or α, but I think the lower stroke would be too high in the line to regard it as an accent referring to γαι[α in next line.

    867 α]ρ᾿ ⟦α⟧`ε´τηκετο: On the second α in αρα the augment of the verbal form has been added. The hand, which must also be responsible for the apostrophe after ρ above the line, is not the same as the one correcting the text at l. 863, to judge from the coarse strokes drawn touching the right-hand elements of α. Was this correction due to the student going through the text a second time? Rather than a variant,[22] this should be regarded as a simplification of the text, in accordance with the school related production character of this papyrus (cf. introduction).

    868 θυμῶ[ : The papyrus agrees with the rest of manuscripts against those in group a, offering θυμόν. The word is not glossed, nor is it the object of a scholium at this point. Maybe the accent was stressing the fact that there is a dative here, and not an accusative?

    869 Τυφω]έος ὲσ[τ : there does not seem to be a diacritical purpose for either accent, inasmuch as there seems to be no alternative articulation of the text.[23] Maybe Τυφωέος was glossed as a genitive in -εος, as opposed to the one in -εως, normal for this flexive type in Attic and the koine, or maybe it was glossed as the mythological character. The verse is quoted in a scholion regarding Theogonia 307, δεινόν θ᾿ ὑβριστήν τ᾿ ἄνομόν θ᾿ ἑλικώπιδι κούρῃ, where doubt is expressed as to whether ἄνομος or ἄνεμος should be referred to Τυφωεόνα in the previous line:

    <ἄνεμον:> ἐὰν μετὰ τοῦ <ο> γράφηται, ἄνομον ἀκουσόμεθα, ἐπεὶ τῇ τοῦ Διὸς ἀρχῇ ἐπέθετο· ἐὰν δὲ διὰ τοῦ <ε>, ἄνεμον, ἐπεὶ αἱ πνοαὶ τυφῶνες λέγονται, ὡς καὶ αὐτὸς λέγει (v. 869)·

    ἐκ [δὲ] Τυφωέος ἔστ᾿ ἀνέμων μένος ὑγρὸν ἀέντων. Scholia Vetera (R2WLZX)

    As for the grave accent, also placed one letter further to the right (see commentary to l. 862), it is difficult to see what confusion could be caused by placing a high pitch on the second ε. The construction ἐκ + gen. + ἐστί, however, might have been signaled with the accent characteristic to Hesiodic poetry.[24] It might be interesting to note at this point the following homeric scholion, on the barytonesis of the verbal form in the same construction:

    Scholion to Iliad 1. 63c <<ἐκ Διός ἐστιν:>> τὸ <ἔστιν> ἐγκλιτικόν ἐστιν, εἰ μὴ ἄρχεται ἢ προηγεῖται αὐτοῦ ἡ οὔ ἀπόφασις. b (BC) Tt, attributed to Herodian.[25]

    870 Νοτ]ου: a (and S, according to the apparatus criticus in Solmsen, Merkelbach and West's edition of Hesiod's works) present νότου τε, against the papyrus and the rest of MSS.

    871 θ]εόφ[ιν: The word does not appear in the Scholia or Glossae to Hesiod, but it does in the Homeric Scholia (e.g. scholia to Iliad 14. 318 Til, ex,[26] and to 17. 101b b(BCE3)Til , ex.,[27] etc.). The accent[28] is likely to be drawing the reader's attention to this unsual form, rather than preventing confusion by warning the reader against the barytonesis that the final syllable of θεός would experience in normal circumstances; for, as a matter of fact, θεὸ in itself would be no form of the substantive.

    Plate I: P.PalauRib.Lit. 9
    Plate I
    P.PalauRib.Lit. 9


      1. S. Bartina, "Un nuevo papiro de Hesíodo: Teogonía 862–872 (PPalauRib. inv. 24)" in Stud.Pap. 7 (1968) 23–30. It was subsequently republished by Josep O'Callaghan, Papiros Literarios Griegos del Fondo Palau-Ribes (PLit.Palau Rib.) (Barcelona 1993) no. 9.return to text

      2. And thus I wish to thank my colleagues and friends present at that seminar, Cesar Hernández, Raquel Martín and Marco Antonio Santamaría, for their comments and reflections regarding this text. return to text

      3. E.G. Turner and P.J. Parsons, Greek Manuscripts of the Ancient World. BICS Suppl., 46 (London 19872) 22.return to text

      4. BKT II, p. 53 ff.; reproduced in R. Seider, Paläographie der Griechischen Papyri (Stuttgart 1970) II no. 33.return to text

      5. Reproduced in Seider, op.cit. (above, n. 4) no. 36.return to text

      6. Ibid., no. 49.return to text

      7. For a similar appearance of another Hesiodic papyrus, compare P.Oxy. LXVIII 4641 (Works and Days 563–567), but the fragment is even smaller to allow comparison of individual letters. It has also been assigned to the third century.return to text

      8. For the purposes of collation I have used M.L. West, Hesiod Theogony (Oxford 1966). I have also considered A. Rzach, Hesiodi Carmina (Stuttgart-Leipzig) 19133 and F. Solmsen, R. Merkelbach and M.L. West, Hesiodi Theogonia, Opera et Dies, Scutum, Fragmenta Selecta (Oxford 19903).return to text

      9. S. Bartina, "Teogonía 862–872, de Hesíodo. Nuevas consideraciones (PPalau Rib. inv. 24)," Stud.Pap. 8 (1969) 101–109.return to text

      10. In G. Cavallo et al., Scrivere libri e documenti nel mondo antico. Mostra di papiri della Biblioteca Medicea Laurenziana, Firenze, 25 agosto – 25 settembre 1998. Pap.Flor. XXX (Florence 1998) 39.return to text

      11. "Ambos (both PLit. Palau Rib. 9 and PAnt 2.71) a su vez contienen reliquias mediales de los versos, tan reducidas que tampoco coinciden entre sí en ninguna letra, verso por verso," Bartina, op.cit. (above, n. 9) 106.return to text

      12. See, however, commentary to line 863.return to text

      13. For a survey of the Hesiod papyri, I have used the online Base de données expérimentale Mertens-Pack3, at the University of Liège, last updated in February 2008 ( Other lists are those in Solmsen-Merkelbach-West's edition of 1990 (including the papyri listed in West's edition of the Theogony [Oxford 1966] and Works and Days [Oxford 1978]) and Bartina's list of the Theogony papyri (1969). Before, Arrighetti's work should be mentioned: G. Arrighetti, "Il testo de la Teogonia di Esiodo," Athenaeum, NS 39 (1961) 211–284. return to text

      14. A. Nodar, "Ancient Homeric Scholarship and the Medieval Tradition: Evidence from the Diacritics in the Papyri" in B. Palme (ed.), Akten des 23. Internationalen Papyrologen-Kongresses (Vienna 2007) 469–481. return to text

      15. Cf. H. van Thiel, "Der Homertext in Alexandria," ZPE 115 (1997) 22–24.return to text

      16. See H. Flach (ed.), Glossen und Scholien zur hesiodischen Theogonie (repr. 1970, Osnabrück) which I will use to quote the corresponding glossae in the commentary. For exegetical work on Hesiod's poems, see M. Stroppa, "Esegesi a Esiodo nei papiri" in G. Bastianini and A. Casanova (eds.), Esiodo Cent' Anni di Papiri. Atti del convegno internazionale di studi di Firenze, 7–8 giugno 2007 (Florence 2008) 83–95.return to text

      17. West, op.cit. (above, n. 8). I will also follow this edition and adopt West's sigla when quoting readings from different manuscripts and papyri. return to text

      18. See B. Laum, "Alexandrinisches und byzantinisches Akzentuationssystem," RhM 73 (1920–4) 28–30 and C.M. Mazzucchi, "Sul Sistema di Accentazione dei Testi Greci in Età Romana e Bizantina," Aegyptus 59 (1979) 148–149, regarding, above all, the study of grave accents, such as the one we have in our text in l. 869.return to text

      19. See West's discussion of the passage, including Peppmueller's suggestion, op.cit. (above, n. 8) 394–395. return to text

      20. Cf. Parson's note at the edition of the line: "... some ancient scholars believed ... that αἰζῃός should be written by analogy with the four-syllable αἰζήϊος."return to text

      21. They also interpreted the sign on ω not as a circumflex accent, but as a mark of abbreviation, in this case for final ν in αἰζηῶ(ν), cf. Bartina's editio princeps and F. Übel, "Literarische Texte unter Ausschluss der christlichen," APF 22–23 (1974) 361. return to text

      22. It is not registered in West's apparatus as a variant reading.return to text

      23. It is unlikely that the scribe might have wished to prevent the reader from placing the accent on the ω, thus reading the accusative form Τυφώ.return to text

      24. Cf. West, op.cit. (above, n. 8) 330 on ll. 590–591.return to text

      25. I cite according to H. Erbse's edition Scholia Graeca in Homeri Iliadem. 7 vols. (Berlin 1969–1988). See also, though not so clearly suited to our case, the scholion to Iliad 6. 152a, also attributed to Herodian:

        <ἔστι> {πόλις}:> τὸ <ἔστι> ἐπὶ τῆς πρώτης συλλαβῆς ἔχει τὴν ὀξεῖαν, ὅταν ἄρχηται, "ἔστι δέ τις προπάροιθε πόλιος" (Β 811), καὶ ὅταν ὑποτάσσηται τῇ οὔ ἀποφάσει, "ὡς οὐκ ἔσθ᾿ ὅδε μῦθος" (Ψ 62). εἰ μέντοι γε μὴ εἴη τὸ τοιοῦτο, οὐκέτι ἐπὶ τῆς πρώτης συλλαβῆς ἔχει τὴν ὀξεῖαν, οἷον "ὃ δὴ τετελεσμένον ἐστίν" (Ξ 196. Σ 427. ε 90) (A).return to text

      26. <<θεόφιν:>> θεῷ ἢ θεοῖς· τὸ γὰρ ἴσος ὅμοιος ἀλίγκιος ἀτάλαντος δοτικῇ συντάττεται.return to text

      27. <<ἐκ θεόφιν:>> ἐκ θεῶν. περισσὴ οὖν ἡ ἔξ.return to text

      28. Even if we find an acute accent printed at this point in the volume edited by O'Callaghan, the note to the line draws the attention to its strange shape; already Bartina in the editio princeps noted that from its lower end a short stroke seems to go to the right, and printed accordingly a triangular sign on ο. I think, however, that the blobby shape of such a stroke may have something to do with the fact that the accent touches the lower elements of ρ in the preceding line, and thus it might have been caused accidentally. return to text