On the Counterpoint of Rhythm and Meter: Poetics of Dislocation and Anomalous Versification in Parmenides’ Poem

Abstract:

In ancient times, authors such as Plutarch and Proclus questioned Parmenides’ abilities as a poet considering his Poem to be ‘prose in disguise’. Harsh judgments concerning Parmenides’ style are pervasive even in modern scholarship (Diels, 1897; Wilamowitz, 1912; Tarán, 1977; Kirk & Raven, 1977). This paper focuses on specific metrical and rhythmic devices used consistently in the composition of the Poem, that I will refer to, collectively, as ‘poetic dislocation’. This term encompasses the blurring and cancellation of the central caesura, the tendency to break Hermann’s bridge with varying degrees of intensity, the accumulation of enjambment in groups of verses, the use of non-traditional lexicon of the epic genre and the placement of Homeric lexicon in non-Homeric hexameter collocations. My attempt is to prove that these procedures are an integral part of Parmenides’ poetic style. I will also present an analysis of Parmenides’ versification focusing on the rhythmic patterns and the organization of caesura, especially in fr. B8.

Keywords:
Parmenides; Meter; Rhythm; Caesurae; Hermann’s Bridge

Introduction

From antiquity to the present day, the fact that Parmenides composed his philosophy in verse, embodying it in a poem in hexameters, has seemed troublesome to many. To those interested in philosophy, the philosophical argument seemed restrained by the metrical structure; to those concerned with style and poetic quality, the poetry seemed clumsy and poorly executed.¹ Rivers of ink have been poured to answer this question: Why poetry?² The vast majority of these explanations, however, exclude in-depth analysis of the functioning of Parmenides’ verse and the ways in which its structure and internal movements are intrinsic to the shaping of the philosophical message.³

What is the relationship between style-the ‘physiognomy of mind’, as Schopenhauer (1851, II, §282) famously said⁴-and philosophical argument?⁵ In this paper I will try to offer some answers to this question by understanding ‘style’ as the alteration and adaptation of generic norms and expected features within a genre-in terms of subject matter and lexis, frequency and function of certain images and other tropes, but more prominently in terms of rhythm and meter. It is to this specific form of style that I will devote this study.

Parmenides’ hexameter

The most complete study of Parmenides’ verse-making is that of A. Mourelatos (1970, Appendix 1). From a comparison with the statistics made by Porter (1951) in his analysis of early hexameter, Mourelatos identifies 12 verse-types in the 146 complete hexameters that have been preserved. According to Mourelatos (p. 2), the second part of Parmenides’ verse is the weakest and, therefore, the most anomalous, among other reasons, because we find word endings after the first short syllable of the fourth foot much more frequently than in the previous hexametric tradition.

Marlein van Raalte (1986) undertakes an analysis of the dactylic hexameter, taking into consideration Homer, Hesiod, Parmenides, Empedocles, Archestratus, Aratus, Apollonius, Callimachus, Theocritus, and Nonnus. She frequently uses Parmenides’ verses to exemplify some of its recurrent and remarkable characteristics, focusing mainly on the rhythmic effect resulting from the proportion and distribution of dactylic and spondaic realizations.

Based on the information and data collected by these scholars, in what follows I will analyze two metrical phenomena in Parmenides’ versification that are significant not only because of their greater frequency with respect to their occurrence in the antecedents of the epic genre, but also because of the peculiarities they present and their effects in the interpretation of the verses in which they occur. These are:

1. 1. The attenuation or cancellation (bridging) of the central caesura.

2. 2. The functioning of Hermann’s Bridge.

Meter or rhythm?

Before focusing on the phenomena and the cases in which they occur, it is important to differentiate which ones correspond to metrical structure and which to rhythm. According to van Raalte, only a closed set of rhythmic characteristics that have a highly generalized use (obligatoriness) and that have definitional force correspond to meter. These are, for example, principles of alternation (short and long; arsis and thesis), predetermined places where word endings are expected, and places from which word endings are excluded. By contrast, phrasing and syntax are rhythmic, since they intervene in the recitation and determine how the verse sounds particularly, in its individual realization. They are not metrical aspects because they are not necessarily part of a shared pattern.⁶

Jakobson coincides with van Raalte when he circumscribes metrical phenomena to what he calls ‘verse design’. In the traditions of comparative metrics to which he refers, caesura is metrical-because it belongs to the verse and to the audience’s expectations -not rhythmic-because it does not belong to the phrasing contingency of each verse.⁷ Pauses that are not metrical, but rhythmic, are not really caesurae dividing the verse structurally into metrical cola, but only phrasing pauses that separate the verse into semantic and syntactical segments.

Parmenides’ hexameter is an interesting case study for analyzing the tensions between meter and rhythm, since it emphasizes the conflict between syntactical phrasing and metrical norms by subordinating structural caesura to the syntactic rhythm in which he organizes his ideas and arguments. Of course, this is not exclusively or prominently a Parmenidean phenomenon-one might suspect that the same occurs in many Homeric, Hesiodic, and Empedoclean passages-but identifying and analyzing the specific cases in which it occurs helps to better understand our author’s style.

In what follows I will refer to the positions of the verse following O’Neill’s system (1942, p. 113, later adopted by Porter, 1951, p. 16), according to which each syllable of the line is numbered from 1 to 12, the full numbers corresponding to the long syllables in arsis and the short second vowels in thesis of dactylic feet, and half numbers (.5) to the short first vowels of dactyls. According to Fränkel’s (1955) terminology, I will call the segment of the verse that comprehends the first two feet ‘Zone A’, the one located in the third foot ‘Zone B’, and the one that goes from the fourth to the sixth foot ‘Zone C’:

1 1½2⏑⏑ 3 3½4⏑⏑ 5 5½6⏑⏑ 7 7½8⏑⏑ 9 9½10⏑⏑ 11 12 x 2 4 6 8 10

1	1½2⏑⏑	3	3½4⏑⏑	5	5½6⏑⏑	7	7½8⏑⏑	9	9½10⏑⏑	11	12 x
	2		4		6		8		10

A useful way to typify the different verse-types is to include the primordial central caesura in their description (that, in nearly all cases, divides it into two segments), besides the substitutions presented by the verse. By doing this, it is possible to represent not only the rhythmical dimension (substitutions and the places where they occur), but also the metrical shape (caesurae: m = penthemimeral or masculine ; t = trochaic or feminine ; h = hepthemimeral).⁸ According to this, then, we would go from having twelve types of verse in Parmenides (according to Mourelatos, who takes into consideration only the caesurae) or 19 (according to van Raalte, who takes into consideration the proportion and distribution of dactylic and spondaic realizations only)⁹ to 29 types of verses (according to this notation).¹⁰

1. Parmenides’ poetics of bridging

In the introduction to his commentary on the Iliad, G. S. Kirk (1985, p. 20) devotes a chapter to the structural elements of Homeric verse. In the section on word groups and rhythmical cola he coins the term ‘rising threefolders’ to refer to verses that lack a central third foot caesura in Zone B, be it masculine or feminine-or, if they have one, it is syntactically bridged-and show instead a caesura in the fourth foot, so that the verse, due to the pauses in zones A and C, seems to be organized in three cola of increasing length.¹¹ In these types of verse, the caesura of the third foot is bridged by a long and heavy word. Kirk adds that the rhythmic and musical effect of these verses differs greatly from that of verses arranged in four or two cola, for the rising threefolder has a ‘more urgent, progressing or flowing effect’ (Kirk, 1985, p. 21).¹² He states that this type is a rare and infrequent verse in Homer, and is often used to mark a climax or contrast at the end of a speech or a long narrative passage. Much more frequently, according to Kirk, one can find rising threefolder verses in Homer with a semantic bridging of the central caesura (when there is a word ending in the third foot, but this does not constitute a caesura properly, because it would separate two elements intimately linked to one another).¹³

Martin J. Henn (2003, p. 2-6) used this concept of Kirk’s to analyze some verses by Parmenides. He lists in a table the different types of tricola that, according to him, outnumber both Homer and Hesiod in the use of threefold verses by about 5% on average: ‘This statistic would suggest that Parmenides composed his verses with slightly more attention to musical phrasing, rhetorical effect, and dramatic flourish’ (Henn, 2003, p. 5).¹⁴

According to van Raalte, Parmenides’ hexameters show, together with those of Archestratus, the highest percentage of verses without caesura (2.7%, compared to 1.2% of the Iliad and the Odyssey, 2.3% of Hesiod, and 0.85% of the Hymns).¹⁵ Van Raalte gives as examples B1.27, B8.21, and B8.32. All three cases have words ending in position 7 (hepthemimeral), but the author considers that ‘caesura function can only be ascribed to a word boundary occurring at a verse position near the middle of the verse’ (p. 82), that is, a penthemimeral (masculine) or trochaic (feminine).¹⁶

Hexameters without middle caesura-rare but admissible in archaic epic-tended to disappear in Hellenistic poetry.¹⁷ Parmenides, on the contrary, exaggerates this compositional mechanism, in which, therefore, we could see a feature of his style.¹⁸

B1.27

The first case occurs in the Proem. Once the young man has passed through the gates of the paths of night and day, and has arrived at the house of the goddess, she says:

χαῖρ', ἐπεὶ οὔτι σε μοῖρα κακὴ προὔπεμπε νέεσθαι⁀
— ⏑ ⏑ — ⏑ ⏑ — — — | — — ⏑ ⏑ — — || = 3h4
τήνδ' ὁδόν (ἦ γὰρ ἀπ' ἀνθρώπων ἐκτὸς πάτου ἐστίν) (B1.26-27)

Here, there is no central caesura, as it is bridged by the trisyllabic noun ἀνθρώπων, which extends throughout the entire third foot and half of the fourth. This necessarily implies that the verse must be divided into two metrical cola only up to position 7 (hepthemimeral caesura). The word endings in Zone A, as I argued above, are not caesurae. We only have a rhythmic (syntactic and semantic) diaeresis after the first foot (Fränkel’s A₃ caesura), that generates a brief rhythmic colon. What is more interesting is that this rhythmic colon is strongly linked to the preceding verse, forming a necessary or integral enjambment (προὔπεμπε νέεσθαι / τήνδ' ὁδόν).¹⁹ This suggests not only that the cancellation of the central caesura must be associated with the phenomenon of enjambment-both being resources that develop together-but also that the ending of the preceding line does not find its structural limit at the boundary of the verse ending, but is dislocated into the next metrical structure. This causes the new line, already intervened by the previous one, to become dislocated beyond its central zone, a process in which its elements are freely rearranged.

The word that breaks the middle caesura is not, as sometimes is in Homer’s case, a proper name, but the simple, plain, ordinary noun ἄνθρωπος.²⁰ The path of words that the audience is listening to is not the one that men-poets usually travel, and this is marked by the metrical peculiarity of using a verse without central caesura, and the stylistic singularity of bridging that caesura with this specific word. Of the 96 occasions in which the plural genitive ἀνθρώπων is used in Homer, in 61 (63%) it appears at the end of the verse; in 17 (17.7%) it is used in positions 3-5 (Zones A-B); in 15 (16.6%) it is used in positions 1-3; and only in 3 passages (3.1% Il. 6.202, 18.107, and Od. 9.119) it is used occupying positions 7-9 and functioning as Hermann’s bridge. But under any circumstance it is used in the central positions of the verse (5-7, Zones B-C) and bridging the central caesura. By using a Homeric word in a position where Homer never uses it Parmenides is clearly marking his style.

Thus, the metrical anomaly of the verse seems to be linked to its meaning, and vice versa. Just as the path the goddess describes as outside the steps of men, so the verse expressing this content is composed by means of a deviation from the usual hexametric structure. The horizon of expectations of epic verse is transgressed precisely in the verse where the uniqueness of the path that the narrator of the proem has traveled is emphasized.

The three other cases where this phenomenon occurs are all in fragment 8 (B8.21, B8.32, and B8.40). Its contiguity in a reduced number of verses is, in my opinion, a significant pattern of this author’s style.²¹

B8.21

After enumerating the nine σήματα of Being (B8.3-6), the goddess comes to a conclusion in her argumentation:

— ⏑ ⏑— ⏑ ⏑— — — ⏑ ⏑— ⏑ ⏑— — || = 3hp²²
τὼς γένεσις μὲν ἀπέσβεσται καὶ ἄπυστος ὄλεθρος. (B8.21)

The word that serves to dynamite the central caesura (ἀπέσβεσται) is not a proper noun but a conjugated verb. Apart from this verse, this verbal form is not used in any other passage of hexameter poetry. Likewise, the word choice is remarkable since it is a highly metaphorical expression in which the action expressed by the verb (usually applied to the extinction of an igneous body) is being referred to two abstract concepts (origin and destruction) with which it hardly fits.²³

Unlike the previous example, where the lack of caesura was related to enjambment, here we do not have any enjambment with the preceding line, since it is the conclusion of the argumentative demonstration of the first two signs of Being. The rhythmic colon —⏖—⏑ with which the verse begins, and that presents words ending in positions 3 and 3.5, does not allow us to perceive a significant pause in the syntactic phrasing, since more than closing an idea, it rather opens a proposition, that cancels or blurs the force of that pause.

The goddess is saying something that contradicts common sense: as much as our experience tells us otherwise, there is no genesis or destruction of things, and to express this inconsistency she composes a verse in which surely the ear of her audience must have experienced, at the level of rhythmic and metrical structure, an analogous astonishment.

The semantic notion of igneous extinction expressed by the verb has its metrical correlate in the extinguishing of the central caesura, while it establishes a relationship of iconicity between the meaning of the verse and its structure: just as Being has neither γένεσις nor ὄλεθρος (nouns that are significantly placed at the beginning and at the end of the line), so too the verse annihilates or suffocates the expected pause in the center, mirroring the ontological lack of genesis and destruction.

B8.32

A few verses later, after arguing why Being is not divisible and is continuous, the goddess focuses on immobility (B8.26-28), and on the ideas of permanence (B8-29-30) and containment (B8-31), which are reinforced by the image of Necessity chaining Being. It is here that the goddess declares:

— ⏑ ⏑ — ⏑ ⏑ — — — ⏑ ⏑— ⏑ ⏑ — — || = 3hp
οὕνεκεν οὐκ ἀτελεύτητον τὸ ἐὸν θέμις εἶναι·

This verse, metrically identical to the one just analyzed, presents a bridging of the central caesura by means of the five-syllable heavy word ἀτελεύτητον. In the three cases studied so far, there is a gradation in the syllable length of the multisyllabic word that breaks the caesura, from three syllables (B1.27: ἀνθρώπων), passing through four (B8.21: ἀπέσβεσται), and finally five (B8.32: ἀτελεύτητον).²⁴ It seems to be a stylistic pattern where the heavy word is progressively emphasized as the recitation of the poem proceeds, bringing the attention of the audience to these words, which play an important role in articulating metrical structure and meaning.

It is significant that the word used to bridge the caesura means precisely ‘not unaccomplished’.²⁵ Being is always finished and complete, as well as the verse in which its completeness is expressed, where the lack of caesura integrates it as a continuous (ξυνεχὲς) unit without any parts that would make it divisible (διαιρετόν). Parmenides seems to be playing with the very semantic notions embodied on the σήματα of Being through the structure of his verses: continuity, divisibility, completeness, and unity. The bridging device is the perfect figure to connote cohesion-by eliminating a gap-and self-containment-by having one word, the whole word, in a place where two would be expected.

This device and its correlative connotation were already suggested in the previous verses. In them, although the caesura is not completely broken, there is a clear syntactic blurring of it. There is a distinct metrical anticipation of what happens in B8.32 in a sequence where position 7 is strongly and recurrently emphasized, preparing the audience for the complete cancellation of the central caesura:

— — — — — ⏑ ⏑ — ⏑ ⏑ — ⏑ ⏑ — — || 12t⁀hp²⁶
ταὐτόν τ' ἐν ταὐτῶι τε | μένον | καθ' ἑαυτό τε κεῖται

— — — ⏑ ⏑ — ⏑ ⏑ — ⏑ ⏑ — ⏑ ⏑— — || 1t⁀hp
χοὔτως ἔμπεδον αὖθι | μένει·| κρατερὴ γὰρ Ἀνάγκη

— ⏑ ⏑ — — — ⏑ ⏑ — ⏑ ⏑ — ⏑ ⏑— — || 2t⁀hp
πείρατος ἐν δεσμοῖσιν | ἔχει, | τό μιν ἀμφὶς ἐέργει (…) B8.29-31

In these three lines there is a trochaic caesura. However, in all cases the pause appears softened or blurred, because there is a more important syntagm ending in position 7. The most natural rhythmic pause should be in the hepthemimeral, and not in the penthemimeral position.

This insistence on diffusing the central caesura without dynamiting it through a heavy word, emphasizes the words cloistered in the middle of the two breaks: the participle μένον and the conjugated verbs μένει and ἔχει. It is also noteworthy that the first line (with substitution in the first and second foot) shows a scheme that mixes the rhythmic and metric patterns of the second (substitution of the first foot) and third lines (substitution of the second foot), as if it were announcing the structure and content of the following two lines.

These four lines seem like a small tetrastic stanza, a microstructure within the larger unit that makes up fr. B8, in which the rhythm gradually tends to break the caesura in the last line of the verse cluster. This is supported by the continuous and sustained enjambment in them. From verse 29 to 30 there is an unperiodic or progressive enjambment (since the meaning of the line extends to the next one having already expressed a complete thought);²⁷ from verse 30 to 31 there is a necessary enjambment (the line ends without having uttered the verb of the sentence);²⁸ finally, from verse 31 to 32 we have again an unperiodic or progressive enjambment.²⁹

Ananque’s enchainment is gradually prefigured from B8.29-31 through a relative bridging of the central caesura, which will not be fully consummated until B8.32, with the heavy word ἀτελεύτητον.

B8.40

This case is particularly relevant because it occurs again in a cluster of strongly enjambed verses. From 8.35 to 8.41 a long structure of seven lines is built up whose limits are completely dissolved, as each one extends to the next in a relationship of strong syntactic and semantic interdependence:

οὐ γὰρ ἄνευ τοῦ ἐόντος, ἐν ὧι πεφατισμένον ἐστιν,
εὑρήσεις τὸ νοεῖν· οὐδὲν γὰρ <ἢ>³⁰ ἔστιν ἢ ἔσται
ἄλλο πάρεξ τοῦ ἐόντος, ἐπεὶ τό γε Μοῖρ' ἐπέδησεν
— ⏑ ⏑— - — ⏑ ⏑ — — — ⏑ ⏑ — — || = 2m⁀h4
οὖλον ἀκίνητόν τ' ἔμεναι· τῶι πάντ' ὄνομ(α) ἔσται,
ὅσσα βροτοὶ κατέθεντο πεποιθότες εἶναι ἀληθῆ,
— — — ⏑ ⏑ — — — — — ⏑ ⏑ — — || 13h4
γίγνεσθαί τε καὶ ὄλλυσθαι, εἶναί τε καὶ οὐχί,
καὶ τόπον ἀλλάσσειν διά τε χρόα φανὸν ἀμείβειν. (B8.35-41)

Verse 35 presents a strong necessary enjambment with the following one, marked by the embedded relative clause. This syntactic structure is one of the defining characteristics of Parmenides’ style, and it also appears in combination with a strong enjambment (B1.1-2) in the incipit of the poem.³¹ This must have felt like a syntactic collapse due to the difficulty involved in processing this type of ‘sandwiched’ structure:³²

[οὐ γὰρ ἄνευ τοῦ ἐόντος_i, [ἐν ὧι_i πεφατισμένον ἐστιν,]_rel/ εὑρήσεις τὸ νοεῖν·]_matrix

From verse 36 to 37 we have again a necessary enjambment, as the subject and the predicate of the sentence are divided by the line break, leaving the two conjugated forms of the verb εἶναι at the end of the line waiting for the completion of their copulative function.³³

Up to this point, the verses present a normal metrical structure with central feminine (35 and 37) and masculine caesurae (36). At the metrical level, the most striking phenomenon occurs in line 38 which, again, is enjambed to the previous one-although more subtly, since this is a periodic enjambment (or ‘clausal’, according to Higbie’s nomenclature),³⁴ in which the causal sentence (ἐπεὶ… ἐπέδησεν) is separated from the final sentence (ἔμεναι) by the line break, in a highly hypotactic structure. In verse 38 we have another case of attenuation of the central caesura. There is a word ending in position 5 marking a penthemimeral (m) caesura (ἀκίνητόν, which is not a Homeric word), but the syntactic colon ends until position 7, more forcefully, with a hepthemimeral caesura (h), which is why editors usually put a high dot here. What is important is that the word in this central position is the Aeolic infinitive ἔμεναι, which is thus emphasized.³⁵ This subdued structural and metrical caesura favoring the syntactic pause prepares the listener rhythmically for the bridging in verse 40.

In these lines, we have a strong density of enjambment and hypotaxis. It is noteworthy how, once again, a relative sentence that makes up the whole verse 39 is contained between the second hemistich of verse 38 and verse 40 (all of it filled by the series of ὀνόματα announced in 38).

Here, the central caesura is bridged by the trisyllabic heavy word ὄλλυσθαι, which forces the verse to pause until position 7. It is remarkable how the intensity of the necessary enjambments between lines 35-36 and 36-37 is progressively tempered through two periodic ones (37-38 and 38-39), to be further moderated finally by two unperiodic ones (39-40 and 40-41). It seems as if this cluster of lines were a kind of stanza where the strength of the enjambment only relaxes until it reaches its climatic point with the bridging of the central caesura in line 40.

The verb form ὄλλυσθαι is extremely rare in hexametric poetry: except for this verse of Parmenides and another one (B8.14, where it is situated in positions 2-4), it is not used by any other archaic or classical poet.³⁶ In addition, the expressive negation οὐχί is very rare in Homer (Il. 15.716 and 16.762, in neither of which appears at the end of the verse, as it does in Parmenides), and nonexistent in Hesiod and the Hymns.³⁷ The infinitive γίγνεσθαι is also not attested in archaic epic. The word choice and its use in the hexameter are innovative and allow Parmenides to break the preponderant structure.

As in the case of B8.32, which occurs within the image of Ananque’s enchainment of Being, here the bridging of the caesura also occurs in the context of an analogous enchainment, but now performed by Moira. The metaphorical chaining is reflected metrically and rhythmically through the blurring of the caesura in B8.38, as well as by the concatenation of enjambments. Just as Being is tied together, so too are the verses describing its binding.

It is significant that in the verse where the goddess lists the names with which men construct their false ontology, she uses non-Homeric lexicon and, in addition, the central caesura is broken. It seems as if she wanted to characterize the flatus vocis of mortals by means of a diction evidently foreign to the hexameter tradition. By defamiliarizing the hexameter from its usual metrical structure and lexicon, Parmenides is marking the verse by forcing his audience to engage in a metalinguistic reflection on the true meaning of these verbs. This is particularly necessary with the infinitive εἶναί, denigrated here as a mere noun lacking ontological substance. Defamiliarizing the verse is a way of marking its metalinguistic character.

Finally, in the last verse of this cluster (B8.41) a phenomenon occurs that it should be seen as related to those I just analyzed. That is, the separation of the parts that make up a compound word, also known as tmesis. Tmesis is a morphological phenomenon that appears to be linked to the stylistic choices of poets (vid. D. T. T. Haug, 2012). In B8.41 (διὰ τε χρόα φανὸν ἀμείβειν), it is possible to read a tmesis of διαμείβω.³⁸ This verse presents a clear penthemimeral caesura and word ending in position 7. What is cloistered in between is the preposition διά and the enclitic τε, which places both in a position of emphasis. In view of the frequency with which Parmenides tends to blur the central caesura and thus enhance the elements closing in position 7, it is not surprising that here the preposition at least is emphatically placed, letting the listener know that something is going on with it. The second hemistich presents the preposition and the verb at the beginning and end, so that their symmetrical and emphatic positions suggest reading them together (διαμείβειν).

As I have argued elsewhere, tmesis is an important stylistic device of Parmenides’ poetry, for he uses it in the very first line of his poem (ὅσον τ' ἐπὶ θυμὸς ἱκάνοι), a privileged place for the poet to show off his particular style.³⁹ Tmesis shares with the metrical and rhythmic procedures analyzed so far the fact that both are mechanisms of separation and union, of cutting and bridging. The preposition is separated from the verb to which it belongs, but this separation is only apparent, since both elements are linked to one another, even though they are sequentially distanced within the verse.

What all these procedures seem to have in common is the tendency to dislocate the hexameter by displacing the usual position of its elements. The enjambment disrupts the stichic unity of the verse. In the cluster of lines just analyzed (B8.35-41), the syntactical and semantic closure of the ideas tends to take place in the middle of the verse, so that it is not possible to extract a complete thought from a single line, but rather it extends in irregular cola that open and close in different positions. In addition, the strong tendency to bridging the central caesura and displacing it to position 7 (B8.29-32 and 38) is a phenomenon closely related to this dislocation.

The proximity in which these stylistic anomalies occur (in only twenty lines of fragment 8 there are three full bridgings of the central caesura, at least four cases where it is blurred, and a succession of enjambments that progressively relax their strength, in lines 35-41) must be a sign that these devices are designed to generate a specific effect on the audience. But what is this effect? In formalist terms, defamiliarization, that is, the breaking of the ‘pattern of expectancy’ with which the audience is equipped when listening to hexameters.⁴⁰

2. Parmenides’ Hermann’s Bridge

The next phenomenon I want to analyze is the metrical norm known as Hermann’s bridge, namely ‘the avoidance of polysyllabic word ending at position 7.5’ (Schein, 2016, p. 93), that is, after the trochee or the first short syllable of the fourth foot.⁴¹ As in the case of the bridging of the central caesura, this anomaly constitutes a deviation or dislocation from the usual hexameter tendency.⁴² According to Mourelatos (1970, p. 267), ‘the weakest point in Parmenides’ hexameter line, and the locus of the greatest deviation of the norm, is the first short of the fourth foot’. If we analyze the complete lines that have been preserved from the Poem, there are 14 cases where we find word ending in this position-always monosyllabic words-and ‘if this is raised to the scale of 1000, the frequency figure is twice what it is in other epic texts’ (p. 267). The recurrence of this deviation is quite significant when compared to early hexameter poetry, and therefore requires specific analysis.

Mourelatos, however, considers that of these 14 cases (B2.1, B4.2, B8.3, B8.18, B8.21, B8.25, B8.28, B8.31, B8.32, B8.36, B8.37, B8.44, B10.1, and, B12.5), only one breaks Hermann’s bridge (B8.28), and of the remaining 13, six bridge the bucolic diaeresis (i.e., the word ending in position 8: B4.2, B8.3, B8.18, B8.21, B8.25, and B12.5). According to Mourelatos, ‘the frequency of this combination is six times over what we find in other epic texts. The acoustical effect of this combination is that of a line which does not close with the magnificence of a Homeric cadence but totters’ (p. 267). In the following, I would like to discuss these cases and propose different results from the analysis.

What is most striking, besides the frequency, is the proximity in which this phenomenon is repeated in contiguous verses (almost all in fragment B8 and, significantly, the same ones I analyzed before as presenting other peculiar metrical phenomena). Anomalies seem to concentrate and agglomerate in singular verses, giving them a pronounced stylistic density.

Contrary to what Mourelatos argues, I am convinced that, out of these fourteen lines, five do indeed violate Hermann’s bridge-or at least, partially break it with a certain violence-while the other nine preserve it or violate it to a lesser degree.

I propose the following typology to analyze the cases in which we find word endings in position 7.5 from a lexical point of view, but not necessarily from a prosodic perspective:

Cases in which Hermann’s zeugma is not broken —⏑⏑ —⏑⏑ —⏑⏑ —͡⏑⏑—⏑⏑ — — ||

I. A prepositive preserves Hermann’s zeugma ⁴³

I.1. After central caesura and hepthemimeral

… |^hep [prepositive⁀host word]…
οὐ γὰρ ἀποτµήξει | τὸ ἐὸν | τοῦ⁀ἐόντος ἔχεσθαι (B4.2) = 2m⁀hp
πολλὰ µάλ᾽, ὡς ἀγένητον | ἐὸν | καὶ⁀ἀνώλεθρόν ἐστιν (B8.3) = pt⁀hp
ἔστιν ὁδός, τὴν δ᾽ ὥστε | πέλειν | καὶ⁀ἐτήτυµον εἶναι. (B8.18) = 2t⁀hp

In these three cases there is no real break in Hermann’s zeugma, since the monosyllable that occupies position 7.5 is a prepositive that intonationally and syntactically belongs to the following word and, therefore, does not imply a true word ending by itself. In these three lines there is again a very recognizable blurring of the central caesura, since the colon closes with greater force syntactically in position 7, so that the word that is cloistered between the two cuts appears emphasized. In all three lines the highlighted word belongs to Parmenides’ ontological vocabulary: τὸ ἐὸν, ἐὸν, and πέλειν.⁴⁴ In the three verses where Homer uses the neuter singular participle ἐόν (Il. 4.426, 5.903, and 11.367), this is placed between the first and second foot. Parmenides shifts the position of the participle beyond the central caesura, marking his stylistic innovation through this displacement.

The preservation of the zeugma in B4.2 fits perfectly with the meaning of the verse: using a double bridging of the caesurae to convey that Being cannot be separated from Being is an effective way to emphasize its unity and continuity (see Mourelatos, 1970, p. 267). Similarly, in B8.3, the fact that Being is ungenerated and indestructible seems to be reproduced through the bridging of the two caesurae, which creates tension between the places where words begin and end and the normative expectations of authentical metrical caesurae.

I.2. After hepthemimeral, without central caesura

…—͡⏑⏑— |^hep [prepositive⁀host word]…
Τὼς γένεσις µὲν ἀπέσϐεσται | καὶ⁀ἄπυστος ὄλεθρος (B8.21) =3hp
οὕνεκεν οὐκ ἀτελεύτητον | τὸ⁀ἐὸν θέµις εἶναι· (B8.32) =3hp

These two lines, analyzed above as examples of bridging of the central caesura, present this second peculiarity (word ending in position 7.5) which, however, is attenuated because the word in question is a prepositive and, therefore, must be integrated intonationally and syntactically with the host word that follows. The occurrence of both phenomena in the same verse is highly significant. Although Hermann’s zeugma seems partially broken, it is syntactically bridged, repeating the rupture of the central caesura. These verses, thus, display a double bridging that frustrates the audience’s expectation of a pause breaking the verse into identifiable structural units.

II. A postpositive preserves Hermann’s zeugma

… [host word ⁀postpositive] —⏑⏑ — —||
τῆλε µάλ΄ ἐπλάχθησαν, ἀπῶσε⁀δὲ πίστις ἀληθής (B8.28) =2tp

According to Mourelatos (1970, p. 267), this is the sole case where the zeugma is broken, but because the word ending in position 7.5 is followed by the particle δέ, which falls within the group of the so-called appositives, there is a clear bridging of the two words.⁴⁵ The conflict and tension between the metrical norm (the bridge) and the syntactical rhythm is meaningful.

III. The combination of prepositive + enclitic/postpositive preserves Hermann’s zeugma.

... [prepositive⁀enclitic] —⏑⏑ — — ||
πείρατος ἐν δεσμοῖσιν ἔχει, τό⁀μιν ἀμφὶς ἐέργει (B8.31) =2t⁀hp
ἄλλο πάρεξ τοῦ ἐόντος, ἐπεὶ τό⁀γε Μοῖρ΄ ἐπέδησεν (B8.37) =pt⁀hp
µεσσόθεν ἰσοπαλὲς πάντῃ· τὸ⁀γὰρ οὔτε τι µεῖζον (B8.44) = 3m⁀hp

These are interesting instances. In B8.31 and B8.44 there is again a syntactic dilution of the central caesura in favor of the hepthemimeral (in both cases, marked by the strong punctuation in position 7 ἔχει, πάντῃ·). In these verses, there is a syntactical bridging of the central caesura and, at the same time, the apparent rupture of Hermann’s zeugma through two monosyllables that are not placed contiguously very often.⁴⁶ In both verses, the intention of bridging caesurae to make the syntactic rhythm prevail without breaking the metrical norms completely is evident. In B8.37, the bridging of Hermann’s zeugma through the sequence of appositives (τό⁀γε) occurs before the periodic or clausal enjambment⁴⁷ that again anticipates the syntactic bridging of the central caesura in B8.38. Since enjambment is the most radical form of bridging, Parmenides uses it as a transitional mechanism to introduce verses that tend to break their caesurae.

Cases in which Hermann’s zeugma seems violated rather than bridged⁴⁸ —⏑⏑ —⏑⏑ —⏑⏑ —⏑|⏑ —⏑⏑ — — ||

I. Potential violation of the zeugma through appositives.

... [host word + pospositive] |⏑ —⏑⏑ — —||
Εἰ δ΄ ἄγ΄ ἐγὼν ἐρέω, κόµισαι δὲ | σὺ µῦθον ἀκούσας. (B2.1) =pmVp⁴⁹
εὑρήσεις τὸ νοεῖν· οὐδ΄ ἦν γὰρ | [ἢ] ἔστιν ἢ ἔσται (Β8.36) =13mVp⁵⁰

In these two cases, the zeugma can be reasonably considered to be broken. The appositive particles δὲ and γὰρ are linked to the host word that precedes them. Such appositives produce a very clear rhythmic closure at position 7.5.⁵¹ These would be slight, but certain violations of Hermann’s Bridge.⁵² In B.36 the particle causing this violation is γάρ which is one of Parmenides’ favorite words (Mourelatos, 1970, p. 3).⁵³ This particle that marks the relations of logical dependence between sentences is one of the clearest indicators of the argumentative character of the poem. Of its 32 uses in the fragments, 14 are in fragment 8, which is clearly the section where the poetic discourse is most insistently filled with argumentative resources. That the zeugma is broken by this particle in B8.36 is relevant, since here the argumentative density hinders the metrical norm. On the other hand, as seen above, this line belongs to a sequence of strong contiguous necessary enjambments. In addition, this group of lines includes one that breaks the central caesura (B8.40, see above). The anomalies tend to occur in close proximity.

In the case of B2.1 we have, again, an appositive (δὲ) embedded in the preceding host word (the verb κόμισαι, a form, by the way, not attested in Homer). We find a similar word order in B1.28 (χρεὼ δέ σε παντα πυθέσθαι), where two appositives preserve Hermann’s bridge, but in B2.1 the second monosyllable is σύ, the orthotonic pronominal form in nominative (not an oblique clitic form, nor even a postpositive). A review of all the cases in the Iliad where the particle δέ is placed in position 7.5 reveals that, in the vast majority, this appositive is followed by a clitic pronoun or particle (see note 51), so that, as the latter is integrated into its host word, this cannot be considered an authentic word ending and, by extension, a breaking of the bridge. The sequence host word + postpositive + clitic forms a metrical, syntactic, and intonational unity. But the placement of an orthotonic pronominal form here (σύ) compromises the bridge by threatening the intonational unit formed by δὲ σὺ, and thus reinforces the possibility of a cut.⁵⁴

As far as the interpretation of the verse is concerned, the rupture of the zeugma is meaningful, for here there is an exhortation in which the goddess commands her apprentice to welcome, pay attention to, and appropriate her words. What better way to get the attention of her audience than by violating the expected structure of the hexameter? The violation occurs at a strategic point in the goddess’s speech.

A case where the verse also feels violated to a more pronounced degree is B8.25:

... [word + postpositive] |⏑ —⏑|⏑ — —||
Τῷ ξυνεχὲς πᾶν ἐστιν· ἐὸν γὰρ | ἐόντι | πελάζει. (B8.25) =2tVp

Again, the postpositive particle γάρ, by establishing a rhythmic and syntactical union with the participle ἐόν, breaks the zeugma. The violation of the bridge is mimicking the content of the verse: the continuous quality of Being is explained through the image of ‘what is’ touching or being adjacent (πελάζει) to ‘what is’. This linguistic splitting of Being into two entities (one in contact with the other) is reproduced in the metrical and rhythmic structure, where there is a word ending in the ‘forbidden’ position and the bridge that would unite Being with itself seems broken. For there to be contact between two things, they have to be separated beforehand, although they are ultimately connected. The verse imitates through its structure the conflict between separation and contact.

It is useful to compare this verse with B4.2, the other fragment where the participle forms of the verb εἶναι are used in polyptoton and where the zeugma was not broken (τὸ ἐὸν | τοῦ⁀ἐόντος ἔχεσθαι). There, the goddess says that it is impossible to separate what is in such a way that it would be apart from what is. She conveys this idea using a verse with a double bridging of caesurae, using this fracture as an effective rhythmic mechanism to structurally imitate the continuity of Being. On the contrary, in B8.25, the goddess says that Being is in contact with Being by breaking more intensely the zeugma through the appositive γάρ. This emphasizes the very notion of contact between two things, which presupposes their separation necessarily. The denied separation is represented by an effective union. The affirmed union is represented by a forced and artificial separation. Bridging and caesurae are clearly two resources with which Parmenides relates the content of his verses with their structure.

Finally, I would like to focus on the last two cases. I believe that there are sufficient arguments to sustain that they tend more to rupture than to bridging:

εἴσῃ δ΄ αἰθερίαν τε φύσιν τά τ΄ | ἐν αἰθέρι πάντα (B10.1) =1tVp
πέµπουσ΄ ἄρσενι θῆλυ µιγῆν τό τ΄ | ἐναντίον αὖτις (B12.5) = 1tVp

Both verses present trochaic caesura and hepthemimeral word ending. Although there are word endings in the middle of nominal syntagms, it is very frequent, in the epic genre, to have caesurae after elided τε or δὲ.⁵⁵ In both cases we have a prepositive in position 7.5 and an enclitic in 8 (as in B8.37). If the enclitic covers position 8 and forms a prosodic domain with the preceding word, there is no real word ending in 7.5. The problem here is that the enclitic is elided, and therefore does not really fill position 8, so not only do we lack the extra syllable in the verse that would avoid the bridge violation, but we are also forced to read the preceding word (the prepositives τά and τό) with the following (the enclitic τ᾽), which adds to the feeling that there is a word ending in 7.5, due to this bidirectional loop generated by the combination of prepositive and enclitic.

In the cases where the zeugma appeared more clearly bridged (B8.31, B8.37 and B8.44), the prepositive τὸ did not function as an article but as a pronoun, while in these two cases the prepositives τὰ and τὸ clearly go with nouns that, in both cases, are deferred until the next verse (σήματα and ἄρσεν) in violent enjambments.⁵⁶

In B10.1 the potential violation of the bridge occurs in a verse that presents a violent enjambment with the following one (separation of adjective and noun by the end of line: πάντα/ σήματα). This form of enjambment, which is very infrequent and rare in Homer, here occurs in concomitance with the possible violation of the zeugma, which makes this a very marked verse that, moreover, is found in a rather peculiar group of verses. Verse B10.2, like the previous one, is also violently enjambed with the one that follows it, for, again, an adjective is separated from its noun by the end of verse (καθαρᾶς... / λαμπάδος). The potential breaking of the bridge would occur here as a phenomenon adjacent to that of violent enjambment.

The word collocation clearly thwarts the audience’s expectation, which fits well with the claim that there is a violation of the bridge here. When hearing τά τε the listener can expect two things:

1. 1. A relative clause, either definitional or restrictive (the so-called epic τε).

2. 2. A noun, a substantivized adjective, or an adverbial paraphrase acting as the nucleus of a nominal syntagm.

But, above all, he would expect either of these two possibilities to be fulfilled before the natural prosodic boundary: the end of the verse. The frustration of the expectancy stems from the introduction of the noun not at the end of the line, but until the following one.⁵⁷ These two cases are misleading and constitute two good examples of how the deceptive order of words is composed by verses whose interpretation is ambiguous. This in turn frustrates the expectation of the listener, accustomed to more usual structures of epic verse.

If all the cases discussed so far are counted as actual violations of Hermann’s bridge, we would have that, in the 148 preserved verses of the Poem there are five cases where the zeugma is perceived to be broken to a more pronounced degree, which gives a total of 3.37%. If this data is compared with the frequency of this phenomenon in Homeric poetry, it is highly significant. Maas (1962, p. 60) says that the breaking of the bridge occurs once every 390 Homeric verses; van Leeuwen (1890) that it occurs once every 550 verses: that is, a total of 281 cases in the entire Iliad and Odyssey, of which 93% involve either a prepositive in position 7.5 or a postpositive in position 8 (Schein, 2016, p. 99). Cantilena (1995) reports it every 422 verses (a total of 66 instances in the entire Iliad and Odyssey). Gentili-Lomiento (2003 p. 270, n. 17) say that this phenomenon occurs once in every thousand Homeric verses.⁵⁸ Finally, Schein (2016) recently proposed that there are only 19 or 20 instances of violation of the bridge in the 15,693 verses of the Iliad, whereas in the 12.110 verses of the Odyssey, it is violated only 24 times, which means that the bridge would be broken once in about every 631 Homeric verses approx.

It is evident thus that, given the profound differences of criteria concerning what qualifies (or not) as a rupture of the zeugma that the various scholars follow, it is difficult to establish a common ground. For the same reason, the number of possible violations in Parmenides would have to be compared to the number of ruptures in Homer, both considered under the same criteria. While this work is carried out, it can be said provisionally that whatever number is assumed, these five instances in 148 verses represent an important and significant proportion, and this should be interpreted as an identifiable stylistic feature of Parmenides’ hexameter. Obviously, being a fragmentary poem, it is only possible to reach a relative and approximate degree of certainty. But if we take into account the estimations about the poem’s length, which oscillate between 500 (Diels, 1897, p. 25-26) and 300 verses (M. L. West, 1971, p. 221, n. 3; Laks-Most, 2016, p. 4), and if there were, hypothetically, no further instances of bridge breaking in the lines we do not preserve from the poem, still the five cases analyzed above would represent a significative number (more than four times compared to the highest estimate in Homer). We can reasonably conclude that the tendency to break Hermann’s zeugma-or at least to problematize the possibilities of its rupture-is part of Parmenides’ stylistic devices.

Conclusions

I conclude that Parmenides’ verse presents a series of anomalies, recurrences, and insistences that are distinctive of his style. These are:

1. 1. Blurring and cancellation of the central caesura: the strong tendency to bridge the central caesura and displace it to position 7, and the subordination of the structural caesura to the syntactic-rhythmic phrasing.

2. 2. A significative recurrence of partial or total violation of Hermann’s bridge: the recurrent placement of appositives in position 7.5 and the interplay they establish with clitics and other appositives to allow or prevent the bridging.

3. 3. Accumulation of enjambment in groups of verses or stanzas that are dynamically combined with the phenomena previously described.

4. 4. Use of non-traditional lexicon of the epic genre in bridging positions and placement of Homeric lexicon in non-Homeric places of the hexameter.⁵⁹

5. 5. Tendency to hypotaxis and syntactic embedding.

6. 6. The use of the particle γὰρ in the breaking position of Hermann’s bridge.

All these stylistic marks usually occur in lines whose meaning is related to what each one provokes in the verse itself: separation (caesurae) or union (bridging and enjambment) of elements. The verses thus marked establish a mimetic relationship with their content.

It is important to emphasize that the bridging of central caesurae occurs mostly in fr. B8 and, unsurprisingly, it does not occur in the doxai section, since it is precisely in B8 where the poetic diction and style clearly departs more from the Homeric epic models. Moreover, it is in fr. B8 where the arguments about the continuity and inseparability of being are elaborated, so the bridging of caesurae is an effective stylistic device for linking structure, form, and content. The deceptive order of words of the doxai, on the other hand, presents, according to our analysis, two cases out of five where it can be argued that there is at least a partial violation of Hermann’s bridge. It makes sense that the rupture of the bridge takes place in the doxa, since a cosmogony and a cosmology are presented there, and both assume the discontinuity of being-namely, the process through which being is separated from being-as a principle. By breaking the bridge, the structure of the verse mirrors the radical discontinuity that reigns in cosmogonic and cosmological discourse.

Beyond blaming these traits on poor poetic skills and stylistic incompetence, what reasons are behind the will to compose in this kind of anomalous verse? I think there is no other reason than the desire to distance himself from the model and in this way defamiliarize the audience and awaken their rational aptitudes through structures that break with their expectations and force them to reflect on the relationship between language, poetic structure, and meaning.⁶⁰ By defamiliarizing the hexameter from its more usual behavior in traditional models, Parmenides’ verse forces its addressee to make some kind of adjustment,⁶¹ that is, to reflect metalinguistically on the meaning encoded in rhythm and meter. And this reflection is precisely the philosophical exercise that the goddess intends to provoke in her apprentice. More than aesthetic or artistic goals, the deautomatization would have pedagogical and parenetical purposes: to awaken her audience to the philosophical revelation by defamiliarizing them from their accustomed beliefs and patterns of expectancy.

The phenomena analyzed in this paper provide a solid basis for understanding what Plutarch could find objectionable in Parmenides’ versification (de audiendo 45a -b) and why Proclus says that his form of exposition was unadorned, dry, and simple (In Parmenidem, p. 665, 17). Both judgments (to which one more passage from Plutarch 16c = A15 and two passages from Philo, de providentia 2.39 and 2.42, can be added) provide unquestionable evidence that, to the ears of two Greek scholars (one of them, also a hexameter versifier), Parmenides’ verse was strange. And, apart from the fact that these might be subjective stylistic judgments, in agreement with an anachronistic poetic expectation, the repeated statement that Parmenides’ poetry is in fact prosaic speech awkwardly embodied in verse must have responded to objective elements of Parmenidean hexameter that find in the phenomena analyzed here a firm basis.

The patterns of dislocation I have exposed here are ultimately resources through which the processes of philosophical rationalization intervened and colonized poetic expression. This connection between meter, rhythm, and thought cannot be a simple coincidence or a clumsy way of ornamenting philosophical expression with bad poetry. Meter, structure, and rhythm have something to say to philosophical reasoning.

Bibliography

• ANNÉE, M. (2012). Parménide. Fragments Poème. Précédé de Enoncer le verbe être Paris: Vrin.
• BARNES, H. R. (1986). The Colometric Structure of the Homeric Hexameter. Greek, Roman and Byzantine Studies, v. 28, p. 125-150.
• BERRUECOS FRANK, B. (2020). Parmenides and Heraclitus Revisited. Palintropic Metaphysics, Polymathy and Multiple Experience. Anais de Filosofia Clássica, v 14, n. 27, p. 37-70. (R. Cherubin & N. Galgano (ed.). Parmênides Eon 1)
• BERRUECOS FRANK, B. (forthcoming). Paying Heed to the Deceitful Order of Words. Palintropic Composition and the Role of Ambiguity in Parmenides’ Poem.
• CANTILENA, M. (1995). Il ponte di Nicanore. In: FANTUZZI, M.; PRETAGOSTINI, R. (ed.). Struttura e storia dell’ esametro greco Roma: Grupo Editoriale Internationale, p. 9-67.
• COXON, A. H. (1983). The Fragments of Parmenides. A Critical Text with Introduction, Translation, the Ancient Testimonia and a Commentary Assen: Van Gorcum. (Phronesis, Supplementary Volume III). (2^nd ed.: Revised and expanded edition edited with new translations by R. McKirahan and a new Preface by M. Schofield. Las Vegas: Parmenides Publishing, 2009.)
• DEVINE, A. M.; STEPHENS, L. D. (1994). The Prosody of Greek Speech Oxford: Oxford University Press.
• DIELS, H. (1897). Parmenides Lehrgedicht Berlin: Druck und Verlag Georg Reimer. (2^nd ed. with a new preface by Walter Burkert and a revised bibliography by Daniela De Cecco. Sankt Augustin: Academia Verlag, 2003).
• EDWARDS, M. W. (1966). Some Features of Homeric Craftmanship. Transactions of the American Philological Association, v. 97, p. 115-179.
• FRÄNKEL, H. (1955). Der homerische und der kallimasche Hexameter. In: FRÄNKEL, H. Wege und Formen frühgriechischen Denkens Ed. F. TietzeMunich: Beck. p. 100-156.
• FRIEDRICH, R. (2000). Homeric Enjambement and Orality. Hermes, v. 128, p. 1-19.
• GENTILI, B.; LOMIENTO, L. (2003). Metrica e rítmica. Storia delle forme poetiche nella Grecia antica Milan: Mandadori Università.
• GHEERBRANT, X. (2017). Empédocle, une poétique philosophique Paris: Garnier.
• HENN, M. J. (2003). Parmenides of Elea. A Verse Translation with Interpretative Essays and Commntary to the Text Westport CG-London: Praeger.
• HERMANN, G. (1801). De emendandae ratione graecae grammaticae Leipzig: Gerhardvm Fleischervh.
• HERMANN, G. (1805). Orphica Leipzig: C. Fritsch.
• HEUBECK, A.; WEST, S.; HAINSWORTH, J. B. (1988). A Commentary on Homer’s Odyssey Vol. I: Introduction and Books I-VIII Oxford: Oxford Clarendon Press.
• HIGBIE, C. (1990). Measure and Music. Enjambement and Sentence Structure in the Iliad Oxford: Oxford University Press .
• JAECKEL, F. (1902). De poetarum siculorum hexâmetro Diss. Leipzig.
• JAKOBSON, R. (1960). Linguistics and Poetics. In: RUDY, S. (ed.). Roman Jakobson. Selected Writings III. Poetry of Grammar and Grammar of Poetry Hague/Paris/New York: Mouton, p. 18-51.
• KIRK, G. S. (1976). Homer and the Oral Tradition Cambridge: Cambridge University Press.
• KIRK, G. S. (1985). The Iliad. A Commentary Vol. I: Books 1-4 Cambridge: Cambridge University Pres.
• KIRK, G. S.; RAVEN, J. E. (1977). The Presocratic Philosophers. A Critical History with a Selection of Texts Cambridge: Cambridge University Press.
• LAKS, A.; MOST, G. (2016). Early Greek Philosophy V. Western Greek Thinkers. Part 2 Cambridge, MA: Harvard University Press. (Loeb Classical Library, 528).
• MAAS, P. (1962). Greek Metre Trad. Hugh-Lloyd Jones. Oxford: Oxford University Press.
• MACKENZIE, T. (2021). Poetry and Poetics in the Presocratic Philosophers. Reading Xenophanes, Parmenides and Empedocles as Literature Cambridge: Cambridge University Press.
• MONRO, D. B. (1891). A Grammar of the Homeric Dialect Oxford: Clarendon Press.
• MOURELATOS, A. (1970). The Route of Parmenides: A Study of Word, Image, and Argument in the Fragments New Haven: Yale University Press.
• OSBORNE, C. (Rowett). (1998). Was Verse the Default Form for Presocratic Philosophy? In: ATHERTON, C. (ed.). Form and Content in Didactic Poetry Bari: Levante Editori, p. 23-35.
• PARRY, M. (1929). The Distinctive Character of Enjambement in Homeric Verse. Transactions of the American Philological Association, v. 57, p. 200-220. (rpt. In A. Parry (ed.). The making of Homeric Verse Oxford: Clarendon Press , 1971).
• PORTER, H. N. (1951). The Early Greek Hexameter. Yale Classical Studies, v. 12, p. 3-63.
• PROBERT, P. (2015). Early Greek Relative Clauses Oxford: Oxford University Press .
• SCHEIN, S. L. (2016). A Cognitive Approach to Greek Metre: Hermann’s Bridge in the Homeric Hexameter and the Interpretation of Iliad 24. In: SCHEIN, S. L. Homeric Epic and its Reception. Interpretive Essays Oxford: Oxford University Press, p. 93-115.
• SCHOPENHAUER, A. (1851). Parerga und Paralipomena: Kleine philosophische Schriften Berlin: Zweiter Band.
• TARÁN, L. (1965). Parmenides. A Text with Translation, Commentary and Critical Essays Princeton: Princeton University Press.
• TARÁN, L. (1977). Concerning a New Interpretation of Parmenides. Gnomon, v. 49, p. 651-666. (Review-Article of A. P. D. Mourelatos, The Route of Parmenides) = in (2001)
• TARÁN, L. (2001). Collected Papers (1962-1999) Leiden/Boston/Köln: Brill, p. 171-192.
• VAN LEEUWEN, J. (1890). Homerica (contin.). Mnemosyne, v. 18, n. 3, p. 265-299.
• VAN RAALTE, M. (1986). Rhythm and Metre. Towards a Systematic Description of Greek Stichic Verse Assen/Maastricht: Van Gorcum.
• WEST, M. L. (1971). Early Greek Philosophy and the Orient Oxford: Oxford University Press .
• WEST, M. L. (1978). Hesiod. Works and Days Oxford: Oxford University Press .
• WEST, M. L. (1982a). Greek Metre Oxford: Oxford Clarendon Press .
• WEST, M. L. (1982b). Three Topics in Greek Meter. Classical Quarterly, v. 32, p. 281-297.
• WILAMOWITZ, Möllendorf U. (1912). Die Griechische Literatur des Altertums Stuttgart/Leipzig: Teubner.

• 1
  See Diels (1897, p. 7-8), Wilamowitz (1912, p. 62), Tarán (1977, p. 173), and Van Raalte (1986, p. 44-45 and 52), who calls Parmenides’ hexameters ‘demonstrative’, as opposed to ‘poetry of a purely literary nature’, and ‘less poetical’.
• 2
  A recent and valuable discussion on Parmenides’ poetics can be found in Mackenzie (2021, p. 20-23, 65-67).
• 3
  In Mackenzie’s book I only found a single mention of one metrical aspect of Parmenides’ Poem (2021, p. 73).
• Schopenhauer (1851, II, §282).
• 5
  Valuable remarks on how the structure of the poem conveys its message and how content and form language are inextricably bound together can be found in Osborne (= Rowett) (1998, p. 31-35).
• 6
  In stichic verse, the caesurae and the end of line, although they are ‘metrical’ phenomena, exert a kind of gravity with respect to the rhythmic phenomena, since the caesurae often coincide with the pauses of meaning (see West, 1982a, p. 36). However, from the abundant sense-pauses in early epic reported by West-represented by the trihemimeral (7%) and the bucolic (11%)-it should not be inferred that these are caesurae in a structuralist sense, because they have no obligatoriness. Of course, this problem is debatable (Fränkel and Kirk, for example, do not share this criterion).
• 7
  Jakobson (1960, p. 12) talks about: “(…) the erroneous identification of a break with a syntactic pause.”
• 8
  Our notation marks the feet in which there is substitution of dactyls with full numbers and with these letters the types of caesura in each verse. I take this notation from the database my colleague Santiago Reza is preparing to record the various verse-types that occur in Presocratic philosophical poetry. I am very grateful to him for allowing me to use it.
• 9
  Although the author proposes only 19 verse types, I have found 20, according to her criteria, three of which (sdsds[s], dddss[s], and dssds[s]) are used only once and can easily go unnoticed.
• 10
  Mourelatos works with a sample of 146 complete verses (not counting as complete the union of B7-B8 and taking Cornford’s fragment into account). Van Raalte presents some inconsistencies in her statements about the size of her sample of Parmenides verses (1986, p. 32 = 148 verses and p. 79 = 150; she declares to be based on Jaeckel’s (1902), to whom she attributes this inconsistency). Our database considers 148 lines of the Poem (excluding the Latin hexameters of B18), of which B2.8-B3, B4.4-B5.1, B7.6-B8.1 form complete verses, and B10.7, B11.1, B11.4, and B12.6 are incomplete.
• 11
  Kirk’s theory that phrasing is the truly determining element of colometric structure above word ending has been criticized by Harry R. Barnes (1986, p. 137).
• 12
  Kirk gives Il. 2.173 as an example.
• 13
  The examples given are Il. 1.48 and 1.61.
• 14
  Henn makes the scansion of B8.32 as follows: — ⏑ ⏑ — |⏑ ⏑ — — — | ⏑ ⏑— ⏑ ⏑ — — || οὕνεκεν οὐκ ἀτελεύτητον τὸ ἐὸν θέμις εἶναι· However, it does not seem that we should place neither caesura nor syntactic pause after οὐκ, but before it, since οὐκ is either a modifier of ἀτελεύτητον or of εἶναι.
• 15
  Schein (2016, p. 102), says that, approximately, only 1.5% of Homer’s verses lack a central caesura. In his chapter on the adaptation of the hexameter to philosophical expression, Gheerbrant (2017, p. 245-270) states that the proportion of lines with hepthemimeral and bridging of the central caesura is 3.31% in Parmenides. He counts a total of 157 verses. In my count (see above, note 9) there are 148 verses (excluding the six Latin hexameters of B18). This implies a slightly higher percentage of verses with bridging of the central caesura (5 cases, representing 3.37%).
• 16
  Although van Raalte’s thesis resembles that of West and Gentili in her tendency to recognize as caesurae only the central ones, in the radicality with which she states the hepthemimeral is not a caesura even when the central one is missing, she is, as far as I know, completely alone. On West’s position in this regard, see West (1982b, p. 249).
• 17
  The Hellenistic poets use less and less heavy words that break the central caesura (Aratus 0.7%, Apollonius 0.03%, Theocritus bucolicus 0.3% and epicus 0.2%, and Nonnus 0% (Van Raalte, 1986, p. 79).
• 18
  Empedocles, who is normally stylistically closer to the Hellenistic poets, moves away from them completely in this respect (according to van Raalte his hexameters present a 2.5% of cases without caesura). Gheerbrant (2017, p. 245-270) states that Empedocles has 2.98% of verses without central caesura (3.34% in the physical poem and 1.87% in the Purifications).
• 19
  Parry (1929) distinguished two types of necessary enjambment; on the one hand, when the poet ends the verse at the end of a word group. G. S. Kirk (1976) calls this type periodic enjambment (Parry doesn’t assign a name to it). On the other hand, when the word group is divided in two verses. Parry called it prosaic enjambment and Kirk integral enjambment. For the terminological differences see R. Friedrich (2000, p. 4).
• 20
  See van Raalte (1986, p. 82), and Maas (1962, § 85).
• 21
  It is remarkable that we do not preserve cases where this phenomenon occurs in the δόξαι section of the poem. I propose a hypothesis about this below. The only case we will not analyze here is B6.8, where the central caesura is bridged not by a heavy word but by the monosyllable οὐκ: οἷς τὸ πέλειν τε καὶ οὐκ εἶναι ταὐτὸν νενόμισται. Here it is not possible to consider that there is a penthemimeral caesura after the negation οὐκ (see Gheerbrant, 2017, p. 258, n. 48).
• 22
  By ‘p’ (plenum) we mean that the hemistich has no substitutions.
• 23
  On the meaning of ἀπέσβεσται, see Berruecos Frank (2020, p. 42-43).
• 24
  Such a gradation is striking to the ear because all words end with a molossus, located in the space of the third arsis, the third thesis (always substituted), and the fourth arsis. Given a predefined scheme (— — —|^hep), the heavy words, metrically, grow backwards.
• 25
  The adjective ἀτελεύτητον is used only twice in Homer (Il. 1.527 and 4.175), and in both cases in different positions than in Parmenides (located in positions 1.5-5 (Zones A-B), in the first case, and in position 5.5-9 9 [Zones B-C], in the second). On the meaning of ἀτελεύτητον and its function in the Poem, see M. Année (2012, p. 29-32, 54-55, 67-69, and 98-99).
• 26
  t⁀h represents the blurring phenomenon of the trochaic caesura.
• 27
  See M. Parry (1929). ‘Progressive’ is Kirk’s term (1976), roughly equivalent to Parry’s ‘unperiodic’. On the use and function of enjambment in Parmenides’ Proem, see Berruecos Frank (forthcoming).
• 28
  See note 18 above.
• 29
  The whole fragment 8 seems to be organized in groups of verses that are articulated through enjambment. The first four lines (B8.1-B8.4) also seem to form a small tetrastic stanza.
• 30
  The text transmitted by Simplicius in Phys. p. 146.9 reads οὐδ᾽ εἰ χρόνος, while on p. 86 it reads οὐδὲν γὰρ. Diels adopted this second reading by adding Preller’s supplement ἢ. Coxon (1983, p. 333) considered that the second reading was not really a variant but an adaptation by Simplicius. He thus corrected the text slightly and printed οὐδὲ χρόνος ἐστὶν ἢ ἔσται [...]. Against this reading, see Tarán (1965, p. 128-129).
• 31
  [ἵπποιi [ταίi με φέρουσιν, [ὅσον τ' ἐπὶ θυμὸς ἱκάνοι,] free.rel] rel / πέμπον]matrix (…)
• 32
  On the difficulty for the listener’s comprehension process this type of structure poses, see Probert (2015, p. 399). The subject of the periphrasis πεφατισμένον ἐστίν is the infinitive νοεῖν, which is employed as the object of the verb εὑρήσεις and as the subject of πεφατισμένον ἐστίν (zeugma). Coxon (1983, p. 332) constructs the syntax differently.
• 33
  Something similar occurs in Il. 21.81-82. C. Higbie (1990, p. 50) catalogues it within the necessary enjambments of her typology.
• 34
  See note 18 above. The name ‘periodic’ is from Kirk.
• 35
  This Aeolic form is undoubtedly marked in this verse, for this is not only the sole place where it is used in the poem (as are the other Aeolic forms ἔμμεναι in 6.1 and ἔμμεν in 2.6), but it is, within the infinitives of εἶναι attested in Homer, one of the least used. Vid. M. Année (2012, p. 35, n. 1).
• 36
  With the sole exception of Soph. OT 799. See Gheerbrant (2017, p. 265, n. 67).
• 37
  This is the only Parmenidean verse with this structure (13h4). A structurally similar Homeric verse is Od. 12.93. Van Raalte (1986, p. 72) calls analogous cases “suboptimal realization of caesura”.
• 38
  See Empedocles’ imitation (Fr. B21.18: τόσον διὰ κρῆσις ἀμείβει) and Il. 19.858; διὰ δὲ χρόα καλὸν ἔδαψεν. See A. Coxon (1983, p 334) and H. Diels (1897, p. 87).
• 39
  Other instances of tmesis in the poem are B8.2 (ταύτῃ δ᾽ ἐπὶ σήματ᾽ ἔασι), B8.49 (ὁμῶς ἐν πείρασι κύρει), B12.2 (μετὰ δὲ φλογὸς ἵεται αἶσα), and possibly B1.2 and B1.3 (see Berruecos, forthcoming).
• 40
  The effect of defamiliarization caused by Hermann’s bridge has been masterfully approached by Schein (2016) (see below). An interesting issue to discuss is when the insistence on defamiliarizing begins to create a new expectation in the audience (I thank C. Vieira for having drawn my attention to this). To answer this, it would be necessary to study the continuity of this phenomenon in the later epic tradition, particularly in the case of Empedocles. Gheerbrant (2017, p. 245-270) shows that the frequency of lines with hepthemimeral and bridging of the central caesura is a Parmenidean stylistic feature that clearly has continuity in Empedocles (from Parmenides 3.31% to Empedocles’ 2.98%. See note 14 above).
• 41
  Schein’s definition of Hermann’s bridge breaking only with polysyllables does not seem to be generalized. It does not belong to Hermann’s own approach, nor does it seem to be shared by Maas (1962), Heubeck (1988) or West (1978 and 1982a).
• 42
  This metrical norm was first formulated by Gottfried Hermann (1805, p. 609 ff.) who considered it to be an incisio a melioribus poetis improbata. Hermann says that the break is more tolerable if the word in this position is a monosyllabic proclitic. Hermann does not call them here proclitics, but voces monosyllabae, quae ad sequentia trahuntur [...]. The term ‘proclitic’ was coined by himself in 1801 (p. 96 ss.). Vid Schein (2016, p. 98, n. 10).
• 43
  West (1982a, p. 25-26) considers that caesura is not usually found dividing an appositive (prepositive or postpositive) from the word with which it is joined by the rhythm of the phrase. Among the prepositives he counts articles, prepositions (except anastrophes), ὦ, καὶ, ἀλλὰ, ἤ, the relative pronoun, the demonstrative pronouns ὅς/ὅ, the particles ἦ, τοίγαρ and ἆρα (when interrogative), and the adverbs of negation οὐ and μή. Among the postpositives he counts all enclitics proper and also some non-enclitics: ἄρα, γὰρ, γοῦν, δαὶ, δὲ, δὴ, δῆθα, δῆθεν, μὲν, μὴν, οὖν, τοίνυν, ἄν, αὖ. Van Raalte (1986, p. 162-165) and Devine and Stephens (1994, p. 340) give similar and useful lists.
• 44
  The form πέλειν is not attested in Homer.
• 45
  According to West, in epic seven appositives never appear in these positions with respect to caesura: ...καὶ ⁝, ...ἀλλὰ ⁝, ⁝ δὲ..., ⁝ μὲν..., ⁝ γὰρ..., ⁝ κεν..., ⁝ (the only true enclitic in this list), and ἄν.
• 46
  The combination τό μιν is not common at all in Homer; I found it in in two passages of the Iliad only (5.795 and 17.404), and in neither is located in the two short syllables of the fourth foot. The combination τό γε is employed in Homer in 21 verses, and it only occurs in the same position as in Parmenides in two of them (Il. 8.7, and 16.302). The combination τὸ γὰρ is only used in 12 Homeric verses, and in none is it placed in the breaking position of Hermann’s bridge.
• 47
  See notes 18 and 33 above.
• 48
  In his treatment of Hermann’s bridge in Parmenides and Empedocles, Gheerbrant (2017, p. 262) claims that there are no cases of bridge violation in Parmenides, and that the verses where the bridge is apparently broken are either cases where the word ending in the bridge position (position 7. 5 or, according to his notation, position 15) is an article (B4.2 and B8.32, B10.1), a negation (B8.7), a preposition, or a relative pronoun; or cases where there is a proclitic in position 7.5 (B8.3, B8.18, B8.21) and an enclitic in 8 (B1.28, B2.1, B8.28, B8.31, B8.37 and B8.44). The following analysis differs from his in that here the emphasis is on the prepositive or postpositive function of appositives; in addition, a distinction is made between enclitics and orthotonic postpositives. For example, in note 61, Gheerbrant says that there is an enclitic in position 8 in B2.1, but the pronoun σύ in the nominative is not an enclitic (see below p. 14-15 and note 53).
• 49
  I represent the breaking of the bridge with a capital V.
• 50
  For the text of this verse, see n. 30 above. See also Gheerbrant (2017, p. 262).
• 51
  In cases like these, some scholars have advocated in favor of a possible secondary hepthemimeral cut, which would reduce the harshness of the breach at the forbidden position (West, 1978, ad v.427, p. 266 and Heubeck, 1988, vol. 1, p. 123). Monro (1891, p. 340, §368) says that an enclitic or a light monosyllable (μὲν, δὲ) is admissible in 7.5, but calls this sometimes condition, sometimes alleviation, making it impossible to distinguish whether in these cases he thinks the zeugma is broken-albeit in an attenuated form-or not. Monro does not identify any distinction between the monosyllables that produce the breaking of the bridge. Specifically, he does not identify appositive clusters (of the form ἐγὼ⁀δέ⁀σε|).
• 52
  With regard to Hes. Op. 427, West (1978) suggests that δὲ (i.e., what we would consider here as the bridge destroyer) is actually in a normal place. Perhaps West has in mind that, in this position, δὲ is usually followed by appositives and, therefore, it is not a word ending. Of the 164 cases where δὲ occupies position 7.5 in the Iliad, in 144 cases (87.80%) it is followed by proper clitics, either particles or pronouns, while in only 3 cases (1.8%: 9.245, 18.20, and 24.398) it is followed by the postpositive δὴ. Only 5 cases (3.04%: 5.285, 6.55, 11.288, 15.244, and 16.515) are represented by other monosyllables (2 μέγ᾽ and 3 σὺ). In light of this, the norm is not really finding two consecutive monosyllables at 7.5 and 8 (pace Monro, 1891 and Schein, 2016), but that they form the same prosodic domain, at least if δὲ is the first of these monosyllables. Only 4 cases place words that are neither pronouns nor clitic or orthotonic particles, or a combination of two clitics (5.285 and 11.728: δὲ μέγ᾽; 5.178: δὲ θεοῦ; and 11.728: δὲ Ποσειδάωνι).
• 53
  Of the 32 uses of this word in 31 lines of the Poem, the majority (12) are located in the first foot and in the fifth (7), while the distribution in the remaining places of the hexameter is more homogeneous (4 in the second foot and 3 in the third). Of the 6 cases where this postpositive particle is located in the fourth foot, it is found in the zeugma-breaking position in three verses (8.25, 8.36 and 8.44)!
• 54
  Only in Il. 6.55, 15.244, and 16.515 the same combination appears as in Parmenides: δὲ σὺ. Devine-Stephens (1994, p. 540) give a list of what they consider the appositives, dividing them into clitics, on the one hand, and orthotonics bridging zeugmas, on the other. In the first category are σοι and σε, but neither σύ nor σέθεν are in the second: σύ is also missing in the analogous list made by van Raalte (1986, p.164 ss.), and in West’s (1982a, p. 21-22).
• 55
  See van Raalte (1986, p. 72) and van Leeuwen (1890, p. 269-271).
• 56
  Kirk coined the term violent enjambment, which is a reformulation of M. W. Edwards’ (1966)harsh enjambment. C. Higbie (1990, p. 51 ff.) distinguishes three forms of violent enjambment (see Berruecos Frank, forthcoming). In B10.1 the line break separates the adjective (πάντα) from its corresponding noun (σήματα). I consider 12.5-6 a violent (or at least somewhere between necessary and violent) enjambment, because the verse boundary interrupts the link between τὸ (clearly functioning as an article here) and the noun it introduces (ἄρσεν).
• 57
  See the cases cited by van Leeuwen (1890, p. 269-271). Interestingly, in 10.1, the audience could have understood πάντα as the nominal nucleus of τὰ τ᾽ before hearing the following (σήματα).
• 58
  They rely on Cantilena (1995), but the numbers do not match.
• 59
  On the placement of words with determined metrical structures in rare positions of the hexameter, see Gheerbrant (2017, p. 264-267), who concludes that 15.29% of the verses of Parmenides involve words placed in unusual positions of the verse.
• 60
  I share this idea with Mackenzie, who encourages dealing with the difficulties of Parmenidean verse ‘as functional, that is, as designed to elicit certain responses from the audience’ (2021, p. 65). According to Mackenzie (2021, p. 99-100), Parmenides ‘defamiliarizes traditional epic language by adapting and altering formulaic lines and (...) defamiliarizes the world presenting familiar entities differently from how they are usually regarded’. This paper humbly aims to provide elements for strengthening these theses at the level of meter and rhythm.
• 61
  This is Schein’s thesis (2016) in his study on the violation of Hermann’s bridge in Homer. He links it to the ‘event-related brain potentials’ (stimuli generating brain activity that can be measured through the waves shown in an electroencephalography), and in particular to the brain’s perception of certain deviations in the usual functioning of syntactic, semantic, and prosodic processes. Stylistic irregularities, in primis the breaking of the bridge, require the listener to make adjustments to mitigate the anomalous functioning. I think that, through these stylistic maladjustments, Parmenides’ goddess awakens in her auditorium processes of rationalization and philosophical reflection.
• FRANK, B. B. A. (2024). On the Counterpoint of Rhythm and Meter: Poetics of Dislocation and Anomalous Versification in Parmenides’ Poem. Archai 34, e03408

Publication Dates

• Publication in this collection
  02 Sept 2024
• Date of issue
  2024

History

• Received
  31 May 2023
• Accepted
  04 Sept 2023