In both Odyssey 12 and Parmenides’ poem, then, a female divinity with privileged access to knowledge, located in a special Beyond, signs out a hodos that her male mortal charge must travel in order to reach his destination. In both cases this features a choice between two hodoi where one is radically blocked and impassable, and, according to the logic of the exclusive, exhaustive disjunction or krisis, the traveller is therefore forced to proceed by way of the other. In both Odyssey 12 and Parmenides’ poem, the goddess then provides detailed instructions for travel on the remaining route.Footnote ¹ We examined the first part of this parallel in Chapter 5; now it is time to examine the second.

Putting matters this way underscores another benefit of analysing the structure of Parmenides’ ‘Route to Truth’ not in terms of a rigid, one-to-one correlation, but with the greater flexibility afforded by the notion of the ‘rhetorical schema’ governed by the hodos. Rather than being forced (as Mourelatos is) to correlate fragments 2, 6, 7, and 8 with Circe’s hodos as it is ordered in lines Od. 12.55–126, with the analysis of Chapter 3 in hand, we are now in a position to examine the possibility that Parmenides exploits the combinatorial possibilities offered by the entire hodos (Od. 12.39–141) and of the rhetorical schema of the hodos more generally. This points towards a core claim: as the catalogic entries ‘Sirens’, ‘Choice/Krisis’, and ‘Thrinacia’ are linked together in Circe’s hodos according to the relationship we have been calling ‘con-sequence’, so the hodos-units articulated in fragments 2, 6 and 7, and 8.5–49 are linked together in the hodos outlined by Parmenides’ goddess according to the same sequentially ordered pattern.

Before approaching the specifics of this claim, a few preliminary points should be stated at the outset. In what follows, I shall adopt several widely agreed-upon tenets concerning the best way to analyse the constituent elements comprising Fragment 8:Footnote ² that the four sēmata of lines 8.3–4Footnote ³ announce a programme for the remainder of the ‘Route to Truth’;Footnote ⁴ that these sēmata, which name qualities of to eon, fall into four groups: (i) agenēton kai anōlethron, (ii) oulon mounogenes te,Footnote ⁵ (iii) atremes, and (iv) teleston;Footnote ⁶ and that these four qualities of to eon are taken up, and arguments offered in support of them, one by one in the course of lines (i) 8.5/6–21, (ii) 8.22–25, (iii) 8.26–31/33,Footnote ⁷ and (iv) 8.42–49, respectively.Footnote ⁸ Because my interest here lies in the formal principles of arrangement organizing the relationship between Parmenides’ arguments rather than in the substance of the claims they advance, I will not attempt to prove the merits of viewing the structure of argument along these lines, which have been widely accepted since at least Owen’s exegesis undertaken more than sixty years ago.Footnote ⁹ At this stage, we may simply note that the traditional hermeneutic concerns of the poetry critic – attention to the way that repeated words and images help define the structure, and articulate the units, of a poem – are in harmony with analyses that see the repeated use of words like epei as the key to understanding the articulation of the argumentFootnote ¹⁰ (rather than, say, a strategy of combing through the body of Fragment 8 for arguments that seem to line up according to our sense of what makes an argument good).Footnote ¹¹

Before moving on to the body of Fragment 8, it is worth observing three additional ways in which the analysis undertaken in the preceding chapters can shed new light on aspects of the use of the word sēma in the opening movements of the fragment. It begins (Fr. 8.1–3):

The precise meaning of the word here is debated. On one view, the four predicates listed in lines 8.3–4 (or 8.3–6)Footnote ¹⁴ constitute the sēmata;Footnote ¹⁵ on another, it is the arguments (i.e. lines 8.5/6–49) themselves to which the word sēmata refers.Footnote ¹⁶ In the first case, the emphasis falls on the notion of a sēma as a physical object acting as a kind of landmark (as often in Homer);Footnote ¹⁷ in the second, the hermeneutic demands embedded in the word sēmainō – indicating a message neither immediately intelligible nor entirely opaque, but requiring interpretation – come to the fore.Footnote ¹⁸

The first benefit: whichever construal of sēma one favours, we find here yet another benefit of reading Parmenides’ poem against the backdrop of Odyssey 12. Parmenides’ goddess’s choice of words becomes less surprising, and more intelligible, when one recalls that Circe begins her account to Odysseus (Od. 12.25–26):

‘Sign out each thing’ is, in fact, precisely what she does in the course of Od. 12.39–141, just as Parmenides’ goddess will do in the course of Fragment 8.5–49. Had Circe been moved to provide a synoptic overview of ‘each of the things’ she was to ‘sign out’, perhaps she might have provided just such a summary as we find in Fragment 8.3–4; she might even have referred to each of the things to be signed out as a sēma.

Second, the discussion undertaken in Section 1.1 may perhaps help us transcend the division between these two interpretations. Much of this book has proceeded from the premise that one of Parmenides’ main strategies for thinking new thoughts and speaking in new ways is to mobilize and activate the full range of associations between old words – hodos, for example – and their physical referents, their semantic range, and their place in the mesh of discursive, sociocultural, and mythical associations. We will see below how Parmenides exploits the ambivalence between the object-like and activity-like senses of the word hodos. Why should sēma and its word family be any different? Section 1.1 provided several fascinating examples of how both senses of the word sēma – a physical object that can guide, mark, or otherwise act like a road sign, and something whose significance requires interpretation – can intersect, overlap, or be (literally) coextensive. Consider again the inscription on the Altar of the Twelve Gods:Footnote ¹⁹

The physical object – a ‘true record’ or ‘truthful monument’ – itself ‘indicates’ or ‘signs out’ a message, but this message is directed to ‘mortals’ and is presented as meaningful in the course of the process of journeying that these mortals will, or at least may wish to, undertake.Footnote ²⁰

Even more arresting in this respect are Hipparchus’ herms, which literally embody all at once the sēma as road sign, a physical object ‘on the route’ signing out the path and its measure (‘you are halfway between the city and the deme of x or y’); the sēma as interpretans, a maxim verbally communicating an important insight about the world, be it moral (e.g. ‘Do not deceive a friend’) or ontological (e.g. ‘what-is is ungenerable and imperishable’);Footnote ²¹ and the sēma as interpretandum, something to be interpreted in the course of the journey that follows, be it on the road to the astu or the argument supporting the claim about what-is. Here would be one more advantage, then, of reading Parmenides as both a poet and a culturally and physically embedded denizen of the late archaic period, rather than as an analytic philosopher avant la lettre speaking Truth across the void of ages. In the semantic ambiguity of the word sēma, we see Parmenides the poet-thinker, having found only old words and old referents, hammering out new meanings and conceptual connections from the crucible of language upon the anvil of sense and reference.

Third, we may observe the relationship between the programmatic announcement of the sēmata in 8.3–4 and the notion of catalogic discourse discussed above (Section 3.1.4). This inventory of sēmata at lines Fr. 8.3–4 returns us to the characteristics of catalogic speech: the sequential enumeration of a set of items that, were they to form a series (rather than a list), would be ordered according to a specifically determined principle.

This brings us to the substance of Fragment 8 and Parmenides’ argument itself. In brief, my interest lies in examining the types of similarities that obtain between the manner in which the four assertions about the nature of to eon are linked to each other and the kris(e)is in fragments 2 and 6/7, and the manner in which the episode of the Sirens is linked to the krisis between the two hodoi or the trip past Scylla is linked to the sojourn on Thrinacia.Footnote ²²

How might this work? Examining the possible answers to this question will form the bulk of the discussion in Section 6.3 below. A preview of one possibility, however, is as follows. The hodos, as a rhetorical schema, makes possible the linking of what we have been calling hodos-units according to a regular ordering principle: the hodos, that is, would play a decisive role in ordering the items of a catalogue into a series. On this view, in place of episodes dramatizing narrative encounters with mythological creatures (such as we find in Homer), in Fr. 8, Parmenides makes claims about the nature of to eon. Where in Homer episodes are sequenced partly on the basis of the spatial contiguity of the locations where the episodes take place in the story-world of the Odyssey, on this view, the claims about to eon would be sequenced on the basis of their ‘spatial contiguity’ in the underlying ‘logical geography’ of the story-world of fragments 2, 6, 7, and 8 (the physical dimension expressed in part through the sēma qua road sign, grave marker, or other physical object fixed in a particular place).Footnote ²³ And where in Homer the direction of this sequential ordering of episodes in the narrative is fixed by the necessity that Odysseus move in time from location to location within the story-world, in Parmenides the direction of this sequential ordering of claims seems to be dictated by the same consideration in logical space. Narrative time collapses into story time as this hodos of inquiry is explained to the kouros – and to us. On this reading, the rhetorical schema dictated by the figure of the hodos – and the specific mode of discursive organization we have been calling con-sequence – would then provide the basic framework governing the shape of the discursive architecture of fragments 2, 6, 7, and 8.1–49 (see Figures 6.1a–b).

Figure 6.1a One possibility. Con-sequence: Ordered sequential linkage of discursive units (= hodos-units), frs. 2, 6, 7, and 8.5–21Footnote ²⁵

Figure 6.1b Articulation of Fr. 8.5–49 (after Owen = strong reading) according to rhetorical schema of the hodos (con-sequence)

Having thus previewed a ‘strong’ reading of the relationship between Odyssey 12 and Parmenides’ ‘Route to Truth’,Footnote ²⁴ it will be important to distinguish the relationships between Fragment 2, fragments 6 and 7, and Fragment 8.5–21 at the level of hodos-units (two kriseis (or one, if one prefers) and the first sēma down the path ‘IS’) from the relationships between lines 5–21, 22–5, 26–33, and 42–49 of Fragment 8. That the first grouping – fragments 2, 6, and 7 and Fragment 8.5–21 – is organized as a series is not today in serious dispute (see discussion at Section 6.3 below). The specific relationship between each of the different sēmata is, however, somewhat more contentious (again, to be discussed in Section 6.3 below). According to some interpretationsFootnote ²⁶ these, too, form a series; according to othersFootnote ²⁷ they are more list-like (though, as we shall see, even on these interpretations, they do not really comprise a list, strictly speaking). Ultimately, my goal in this book is not to plump for one interpretation or the other. Rather, I want to examine how my overall account of Parmenides’ invention of extended deductive argumentation – with particular emphasis on his mobilization of the associations of the reference of the word hodos, the ambiguities inscribed in its polysemic nature, and, most of all, via the discursive architecture of the hodos – looks when paired with different plausible, internally consistent interpretations of these arguments themselves; it is to these I shall turn in Section 6.3 below. First, however, in sections 6.1 and 6.2, I shall cash out the previous discussions of narration and narrativity, description and descriptivity by examining Parmenides’ tasks and accomplishments in their intellectual and historical context. In Section 6.1, I place Parmenides in his historical and intellectual context and explore particular limitations that his predecessors confronted, thereby revealing the unique set of discursive resources the rhetorical schema of the hodos offered him. In Section 6.2, I consider these questions from the perspective of Parmenides’ seminal ontological and epistemological innovations, and also their relationship to another set of narratologically complex manoeuvres he performs.

6.1 Sēma I: Systematicity and Argumentativeness

The best way to approach the arguments that make up Fragment 8 is to consider them alongside two crucial aspects of the larger intellectual milieu in which Parmenides may be seen to be working.Footnote ²⁸ First is the question of what we might call discursive systematicity, an attempt to create a discursive structure in which claims are linked according to a regular pattern or underlying set of principles; second, the development of argumentation to support claims advanced (as opposed to a mere assertion of the claims themselves). This demands a brief discussion of earlier (or, in the case of Heraclitus, potentially contemporary)Footnote ²⁹ thinkers.

Scholars have found the Milesians to be the most promising place to look for evidence of discursive systematicity among the immediate precursors of Parmenides.Footnote ³⁰ Any evaluation of the discursive structure and argumentation exhibited in the works of the Ionian cosmologists is gravely constrained, of course, by the paucity of ipsissima verba coming down to us from Miletus.Footnote ³¹ A charitable reading, however, would see a certain level of discursive systematicity implied by their apparently systematic cosmological theories. The communis opinio remains that ‘cosmogony is the heir of theogony’, and that Hesiod’s Theogony in particular provides the key model for the Ionians on two levels.Footnote ³² In the first place, it supplies a conceptual framework for understanding the world as one kosmos; in the second, it supplies a discursive framework for expressing this in a discursive unity (viz. a single, unified whole organized by a systemically applied rhetorical schema, the rhetorical schema of the genealogy).Footnote ³³

A genealogical mode of organizing discursive units does not, however, naturally suggest a role for argumentation that justifies the specific cosmological claims advanced.Footnote ³⁴ (Although, again, any assessment of Milesian argumentation remains provisional on account of the lack of original source material.) And although Anaximander is credited with supporting his claims with argumentation rather than merely asserting them in two justly celebrated instances,Footnote ³⁵ the scholarly consensus is that even ‘where there is apparently genuine disagreement with a predecessor [and] we might expect specific arguments against’ views previously espoused, a Milesian theory ‘seems to be a matter of assertions with connecting links, rather than a system whose basis is argued for and in which the various elements are supported by demonstrations of their connections with first principles’.Footnote ³⁶ A generous view of Milesian thought, then, would grant a kind of systematicity (at both conceptual and, potentially, discursive levels) to their cosmogonies and cosmologies, but detects scant interest in indicating why a particular assertion in this system should be accepted over a rival claim.

Xenophanes and Heraclitus cut rather a different pair of profiles. Here, too, we suffer from the patchy, haphazard manner in which their words have come down to us; in what survives we can catch some glimpses of argumentation, but any evaluation of the discursive architecture of these thinkers’ expressions is necessarily speculative. What seems certain is that the argumentative support for individual claims advanced by these two thinkers is unquestionably more developed. Xenophanes uses reductio arguments, notably in Fragment 15;Footnote ³⁷ Heraclitus uses various hypothetical arguments, as in fragments 7 and 23.Footnote ³⁸

Nevertheless, even one of the staunchest defenders of a rationalist Xenophanes admits that, while ‘some fragments contain logical connectives … and take the form of hypothetical argument, on the whole Xenophanes offers little by way of argument in support of specific conclusions’.Footnote ³⁹ Nor do those who would see in his corpus a systematic account of physical phenomena and their causes claim that he supports these daring assertions with much in the way of argumentative justification. Rather, the novelty of the claims lies in their ostensibly systematic nature and scope, not in their being systematically advanced or defended.Footnote ⁴⁰

It is not easy to assess from Heraclitus’ fragments how systematic his argumentation was, or what the report that Heraclitus wrote a ‘book’ might imply.Footnote ⁴¹ The view summarized by Barnes three decades ago remains the generally received wisdom:

Even a leading proponent of the view that Heraclitus’ corpus forms a carefully composed unity envisages this formal ordering of the whole ‘on the analogy of the great choral odes, with their fluid but carefully articulated movement from image to aphorism, from myth to riddle to contemporary allusion’; on this view, supporting a presumed ‘central theme, … hen panta einai’, we find ‘a chain of statements linked together not by logical argument but by interlocking ideas, imagery, and verbal echoes’.Footnote ⁴³ Likewise, one of the most recent attempts to ‘protect … the rationalism of Heraclitus’ concedes ‘a lack of intrinsic order among the fragments of Heraclitus’ which may well ‘stand to one another in no particular order or bear no intrinsic relation to one another, logically or syntactically’.Footnote ⁴⁴

What we find, then, in the case of the Milesians is, most likely, a relatively high degree of discursive systematicity but relatively little argumentation. In Xenophanes and Heraclitus, meanwhile, there are hints of a somewhat more developed level of argumentation, at least at the level of individual claims,Footnote ⁴⁵ but what we do not seem to find is much evidence of discursive systematicity.

By contrast, the rhetorical schema dictated by the figure of the hodos offers a discursive framework that makes possible a single discursive unity that both accommodates a number of different textual units (unlike in Xenophanes and Heraclitus) and the linking together of these units in such a way as to suggest, and build upon, their necessary connection (unlike in the Milesian cosmologies). Studies of Parmenides’ accomplishment emphasize both the systematicity of his discourse and its thoroughly argumentative character;Footnote ⁴⁶ I suggest that it is his use of the figure of the hodos that, by providing a discursive framework that can accommodate both features, makes this combination possible.

Importantly, Parmenides’ use of argumentation operates at what we might deem to be two levels. Just as the decision in the krisis in Fragment 2 is supported by (condensed and skeletal) argumentative justification, so each of the four claims advanced in the course of Fragment 8 is defended by argumentative support of varying extensiveness and comprehensiveness (viz. at the level of types of dependence). But these claims – and their supporting argumentation – are also linked to fragments 2, 6, and 7 (viz. at the level of rhetorical schemata) and, on some readings, also to each other, a question to which we shall return in Section 6.3. It is the potential movement along both axes – down the level of dependence and across the level of rhetorical schemata – that helps make Parmenides’ achievement what it is; and it is the hodos – which, unlike the genealogy or the stand-alone argument, accommodates and organizes relationships along both axes – that makes this possible.

6.2 Sēma II: Discursive Architecture and Temporality

What does this mean in terms of the discourse modes associated with the rhetorical schema of the hodos and the types of dependence it dictates? Before examining the specific relationships obtaining between the different fragments and the arguments of Fragment 8, it will be necessary to address aspects of Parmenides’ hodos of inquiry in relation to two other dimensions of import for the history of thought. Against the backdrop of the deep continuities between the discursive architecture of the hodos in Homer and Parmenides, we may also note some changes of extraordinary significance.

We saw that in the Odyssey, the enumeration of an itinerary of a hodos is usually a narrative affair (Section 3.2). This is reflected at the textual level insofar as episodes are linked together by temporal adverbs (e.g. πρῶτα, κεῖθεν, ἔπειτα), and by verbs whose features are closely associated with narration: verbs in the aorist, or in the future or historic present tense; and verbs in the imperative mood and/or second person – the language of time-bound activities that unfold in the course of, and themselves constitute, narrative action. These features suggest that the manner in which the text itself progresses has an irreducibly temporal component: the sequence of items as they appear in the text unfold along temporal lines (i.e. they are related to the passage of time in the story-world). This in turn is connected to the fact that ‘the temporal order in which events happen’ – the underlying events depicted by the narrative, which in turn unfolds along temporal lines according to the passage of time in the story-world – ‘is significant’.Footnote ⁴⁷

Parmenides’ Fragment 8, however, bears little trace of these narrative textual features linking the ‘episodes’ of the sēmata. Instead of the hemistiches πρῶτα μὲν ἐς Πύλον ἐλθὲ (Od. 1.284), κεῖθεν δὲ Σπάρτηνδε (Od. 1.285), and νοστήσας δὴ ἔπειτα (Od. 1.291), or Σειρῆνας μὲν πρῶτον ἀφίξεαι (Od. 12.39), αὐτὰρ ἐπὴν δή (Od. 12.55), and Θρινακίην δ᾽ ἐς νῆσον ἀφίξεαι (Od. 12.127), the opening units of the sēmata in Fragment 8 begin, for example: οὐδέ ποτ’ ἦν οὐδ’ ἔσται (Fr. 8.5), and οὐδὲ διαιρετόν ἐστιν (Fr. 8.22).Footnote ⁴⁸ We do not find the adverbial markers that indicate a temporal progression of text or event, just as we find none of the aorist, imperative, and/or second-person forms of narration or instruction that link the textual units of the hodoi of Odyssey 10 or 12. Although we do find verbs in the past and in future tense in line 8.5, these are both rejected in favour of the third person singular indicative timeless (or even eternal) presentFootnote ⁴⁹ (formally akin to what we find at line 8.22): ἐπεὶ νῦν ἔστιν ὁμοῦ πᾶν | ἕν, συνεχές (Fr. 8.5–6). And at the top level of dependence, we find few actions, and none for which the sequence of events depicted by them is significant. In the hodos detailed by Parmenides’ goddess, the narrative framework that links the various units of the hodos to each other – expressed in Odyssey 12 in the second person future indicative verbs of prophetic utterance – has vanished (a dynamic to be discussed at greater length in Section 6.2.2.1 below; see Figure 6.2).

Figure 6.2 Levels of dependence: Transformation from Homer Od. 12.39–141 to Parmenides Fr. 8

By contrast, verbs in the third person singular indicative omnitemporal present correspond perfectly to the characteristics attributed to ‘description’ given above (see sections 3.1.2 and 3.2.3). Moreover, the opening hemistiches introducing the first and second sēmata (sēma-qua-‘argumentation proper’) also fulfil the very same functions of description – namely, introducing elements of the story-world and attributing qualities to them – that we have identified (see Section 3.1.2). Not only are these opening hemistiches of sēmata 1 and 2 formally similar to the ‘description’ portions of Circe’s hodos but they also perform the same function of attributing qualities.

These observations regarding description approach a larger nexus of topics which will form much of the remainder of the chapter. They can be examined from two perspectives. The first, to be addressed in the remainder of this section, concerns Parmenides’ place in the history of thought: what is at stake in the deployment of the figure of the hodos at this particular phase of Presocratic thought? What possibilities and resources might it afford to one who exploits it, how do these work, and why might they be useful? Second, to be addressed in Section 6.3, ‘Sēma III’: in what ways might this figure actually operate in the sequence spanning fragments 2, 6, 7, and 8 and in Fragment 8 itself? Finally, in Section 6.4, ‘Sēma IV’, I shall attempt to draw some conclusions and assess their implications for our understanding of Parmenides’ poem.

6.3 Sēma III: Hodopoiēsis (the ‘Route to Truth’ and Fragment 8)

We have just seen how the movement of plot, not movement in the story-world, provides the temporal dimension of the spatio-temporal con-sequence that dictates the order in which the rhetorical schema of the hodos catalogues its entries. But what of the spatial side of that equation? Is there such a thing as spatial contiguity with respect to items in the underlying ‘story-world’ that makes up fragments 2, 6, 7, and 8?

Some readers of Parmenides’ Fragment 5 would suggest not. Karsten, for example, understood the fragment to refer to the different hodoi on offer in the course of the poem;Footnote ⁹⁵ as later scholars have pointed out, if one accepts that these number three, or at least that one of them corresponds to the Doxa section, this understanding of Fragment 5 ‘asks us to believe that Parmenides could have altered the order in which he examines these three Ways’.Footnote ⁹⁶ There is no reason, then, that Doxa need be read after the ‘Route to Truth’, and it is not necessarily clear that Fragment 2 need precede fragments 6 and 7, nor that these in turn precede Fragment 8. The items that make up Parmenides’ ‘Route to Truth’ – and indeed the post-proem poem proper – might well form a list, then, plain and simple. On this view, there would be no underlying geography to Parmenides’ story-world at all.

Scholars of Parmenides rarely find time these days to refute this view, much less to hold it.Footnote ⁹⁷ There are at least three reasons for this. Briefly: first, certain elements of the poem would become difficult to explain; were it not the case that all other possible hodoi (whether one or two) had already been ruled out by the time Fragment 8 begins, what grounds could there be to declare (Fr. 8.1–2):Footnote ⁹⁸

Second, parts of Fragment 8 would appear to indicate expressly that they are to come after the krisis announced either in Fragment 2 or a combination of fragments 2, 6, and 7 (8.15–18):

As the perfect tense (κέκριται) suggests, at this stage in poem, the decision between the two hodoi has already been made.

Third, as all commentators agree, the argumentation found in lines 8.5–21 (or 8.6–21), for example, depends entirely on the points established in these earlier fragments: the two arguments offered against coming-to-be, a ‘semantic-epistemological’ rejection of ‘what-is-not’ (Fr. 8.7–8) and the ban on genesis ex nihilo (Fr. 8.6–7, 9–10) both presuppose passage by way of the first (and potentially second) kris(e)is.Footnote ¹⁰⁰ It is clear, then, that Fragment 8 must come after fragments 2, 6, and 7.

What is more, on any interpretation involving a second krisis in fragments 6 and 7, it is crucially important that the second krisis (fragments 6 and 7) comes after the first (Fragment 2).Footnote ¹⁰¹ On many of these interpretations, the mutually implicated revelation of being and not-being in Fragment 2 is a necessary precondition to any consideration of the possibility mooted in Fragment 6; for scholars who advocate such a reading, it is only after having attempted to think or indicate to ge mē eon that it becomes possible to conceive of a path that features both ‘IS’ and ‘IS NOT’.Footnote ¹⁰² On this reading, the three units, Fragment 2, fragments 6 and 7, and Fragment 8.5–21, do proceed according to a regular ordering principle. Put differently, since it seems essential that fragments 2, 6, 7, and 8.1–21 be placed in this order, we may say that there is some kind of a fixed, underlying map of the ‘story-world’ the goddess describes. The catalogue they form, that is, must be deemed a series, not a list.

So far, so good. But what about the relationship between the sēmata themselves? A goddess enumerating a hodos back home to Ithaca is constrained by the geography of the world this hodos traverses. Instead of events tied to places, however, the hodos of Parmenides’ goddess orders claims, predicates that can (or, indeed, must) be predicated of to eon. But what dictates the placement of these claims in adjacent, contiguous locations in a hodos dizēsios? Is there also some underlying, pre-existing logical geography that dictates the sequence according to which these must be ordered? Or is it merely that the figure of the hodos imparts – imposes – the appearance of a reified necessity?

As at so many points of Parmenidean analysis, there is little consensus here. Perhaps the most prudent way to proceed is to examine readings that stake out two extreme positions on this question. Those advanced by G. E. L. Owen and David Sedley come as close to forming just such a pair as perhaps can be found.Footnote ¹⁰³ Furthermore, because these two readings share similar views of several major features of Parmenides’ argumentative structure – each regards the argument of Fragment 8 as made of four distinct arguments corresponding to the four sēmata presented above (lines 8.5/6–21, 8.22–25, 8.26–33, and 8.42–49, respectively) – they are especially easy to compare.

It is worth emphasizing here yet again that my chief aim is not to provide a comprehensive, exhaustive analysis of Parmenides’ specific arguments but to understand the larger shape and structure of the argumentation. Accordingly, the following discussion of Parmenides’ arguments will be undertaken with a view to articulating the possible relationships between each of the different elements that form it – that is, the relationships between each of the four sēmata, and between different sēmata and the arguments of fragments 2, 6, and 7.

Sedley, who would rehabilitate the views that Parmenides is a ‘radical cosmologist’ and that to eon is ‘the sphere that constitutes … the world of mortals’, proposes an ‘unashamedly spatial reading’ of Fragment 8.Footnote ¹⁰⁴ He extracts ‘two Laws’ from fragments 2, 6, and 7. The second of these crystallizes the substance of Fragment 2: ‘No proposition is true if it implies that, for any x, “x is not” is, was or will be true.’Footnote ¹⁰⁵ The first gestures towards a law of non-contradiction, and also seems to encapsulate Fragment 6: ‘There are no half-truths. No proposition is both true and false. No question can be coherently answered “Yes and no”.’Footnote ¹⁰⁶ With these ‘Laws’ in hand, Sedley summarizes his view of the argumentative structure of Fragment 8 thusly:

On Sedley’s interpretation of the arguments in Fragment 8.5–49, then, what we find is a scattering of separate, distinct points – points that, while ‘hard won by argument’, do not necessarily lead onto each other or rely on each other via an intrinsic sequence or pattern. Once one has traversed fragments 2, 6, and 7 in order, the sēmata in 8.3–4 could in theory be visited in any order (provided that sēma 1 is visited before sēma 3).Footnote ¹⁰⁸

Contrast Owen’s assessment of Fragment 8: ‘Parmenides’ train of argument breaks into four main stages which are clearly distinguished and correctly ordered in the programme given at the start, and each succeeding movement is introduced by an epei-clause which … shows how the argument depends on a proposition already proved.’Footnote ¹⁰⁹ That is, as Lloyd puts it, ‘the fragment forms a carefully articulated whole in which the later sections build on the conclusions of the earlier in an orderly sequence of argumentation’.Footnote ¹¹⁰

There is in fact less distance between Owen’s view and Sedley’s than may be suggested by Sedley’s characterization of Fragment 8 as consisting of ‘largely independent proofs’, each of which is ‘self-contained’. For Sedley, as for Owen, there is no question that fragments 2, 6, and 7 (captured in his notion of two Parmenidean ‘Laws’) come anywhere but before the four sēmata of Fragment 8. Likewise, if, at least as the argument now stands, sēma 3 would seem to come after sēma 1, this already eliminates a number of the possible sequences in which Parmenides might have ordered his sēmata.Footnote ¹¹¹

For his part, Owen summarizes his views as follows: ‘in the third movement B 8.27 looks back to B 8.6–21 and especially to line 21’; ‘in the fourth B 8.42 looks back to B 8.26–33 and especially to lines 26 and 30–31’.Footnote ¹¹² Reading line 8.22 as ἐπεὶ πᾶν ἔστιν ὁμοῖον (instead of ἐπεὶ πᾶν ἐστιν ὁμοῖον) and taking ὁμοῖον adverbially (viz. ‘exists without intermission’, rather than ‘is all alike’), Owen sees the proof elaborated in lines 8.22–5 as drawing its premise from the claims established at 8.11 and 8.15–18.Footnote ¹¹³ Of lines 8.6–21 he says less, but this is perhaps because the situation is in some respects more clear-cut.Footnote ¹¹⁴ Owen does not address the complexities surrounding the epei clause in lines 8.5–6, but in light of his earlier assertions,Footnote ¹¹⁵ a defender of Owen’s position might say that this is because Parmenides himself so thoroughly stitches the claims of fragments 2, 6, and 7 into the argumentation of lines 8.6–21 (even recapitulating matters at lines 8.15–8.18) that the relationship between the conclusions secured in earlier fragments and the premises of the argument put forward in the first ‘movement’ in Fragment 8 is essentially self-evident.

Owen’s view of the organization of Fragment 8, highly influential over the years but more contested of late, yields a striking vantage on the power the figure of the hodos exerts on the structure of Parmenides’ fragments 2, 6, 7, and 8. This view, that only once one has attained the first sēma – meaning either ‘signpost’ or ‘proof’, or bothFootnote ¹¹⁶ – can one begin to make headway in relation to the second or the third, and only once one has attained the third sēma can one set off on the final stage of the itinerary for the fourth, coincides with what above was described as the ‘strong reading’ of Parmenides’ Fragment 8; notably, it presumes a pre-existing underlying logical geography that defines the map of the ‘story-world’ of Fragment 8 in the way that a pre-existing underlying geography is presumed to define the story-world depicted by Circe in Odyssey 12. On Owen’s reading, we thus see the sēmata concretize, reify, and take root in a domain that claims the same sort of material thickness and free-standing reality as the story-world of the Odyssey, with its Sirens’ meadow, smooth cliffs, hardy fig tree, and so forth; now, however, this substantiality stands in the domain of the hodos dizēsios and the sēmata that mark out its course. Likewise, as the geography of the Odyssey’s story-world possesses a predetermined configuration within the universe of the story (so that Circe can map out the itinerary of Odysseus’ next sequence of adventures, but cannot reconfigure the map), and as the Sirens’ meadow only gives way to the pastures of the Sun’s cattle by way of the Planctae, Scylla, or Charybdis, so on this view one would get to the third point in the itinerary, the third landmark, the third signpost or sēma-object, only by way of the first, and to the fourth only by way of the third.

6.4 Sēma IV: Accomplishments and Completions

It is time to bring this pistos logos to a close. The arrangement of words in Chapter 7 (‘Mortal Opinions’), potentially deceptive in its own way, will offer an invitation to reflect on how our own criteria of knowledge, what we count as a valid contribution to it, and the hodos dizēsios of academic research that leads us there, all retain a fundamentally Parmenidean shape – for better and for worse. If part of this shape is defined by what Karl Popper has called the ‘Parmenidean apology’ of the Doxa and the questions it poses about the status of the ‘Route to Truth’, Part III (Doxai) will explore what this implies for the analysis undertaken in this and preceding the three chapters of Part II (Routes). By testing the limits of reading Parmenides alongside Homer, I hope to call attention to some of our own epistemic presuppositions, which are not always fully articulated or acknowledged, and to underscore their relationship to a Parmenidean, and Homeric, desire for certainty and closure – and to the difficulty of attaining it.

For now, however, it remains to ask what all this – this chapter, this Part (Routes), and the primary line of argument in this book – amounts to.Footnote ¹⁶³ The answer to this question will depend quite considerably on the fields, methods, and aims of the scholar who happens to be reading this book; the analysis undertaken above will likely be valuable for different reasons to, and be used in different ways by, scholars working in different fields, or attempting to answer different questions. One way of organizing the range of possible implications of this book’s claims for our understanding of Parmenides’ arguments would be to discuss matters in terms of ‘priority’.

Working on one level, for example, will be scholars whose main approach to philosophical texts begins with an attempt to understand and reconstruct the argumentative moves of a text in relation to what might make a ‘good’ argument by our own standards, regardless of whether these are expressed in ancient Greek, English, or any other language (perhaps including logical notation).Footnote ¹⁶⁴ In this case, what might be called philosophical analysis of Parmenides’ argumentation will likely remain ‘prior to’ the aspects of Parmenides’ poem discussed here. That is, one expects that such a scholar will likely decide first whether he or she finds, say, Owen’s or Sedley’s assessment of the poem’s argumentation persuasive; then, having settled on one or the other, he or she can use the analysis presented here to explore aspects of his or her preferred interpretation in this new light. The questions that will exercise such a scholar will likely concern determining to what extent, and in which distinctive ways, Parmenides was influenced by the pattern of Circe’s description of the hodos, or up to what point he relies on, and at what point he moves beyond, the physical features of Greek rut roads in developing his own arguments.Footnote ¹⁶⁵ Did Parmenides conjure consequence from con-sequence, as a disciple of Owen might feel, as he travelled a hodos along a rut road of argument inscribed into a pre-existing logical terrain? Or was Parmenides a master rhetorician, deploying a discursive architecture with a capacity for a temporally unimpregnated systematicity and argumentativeness, narrativity without narration and description without descriptivity, as a Sedleian interpreter might have it? Or, rather, are the language and imagery used by Parmenides entirely irrelevant, and his arguments fitted together according to some other set of principles entirely – and, if so, what are those?

Working on another level, scholars more focused on Parmenides’ place in the history of thought might approach his poem with a different set of presuppositions and commitments, especially as far as the relationship between language and the ideas it expresses, between signifier and signified, are concerned. Particularly if they are interested in Parmenides’ role as the decisive figure mediating the transition to a conception of knowledge predicated on extended deductive argumentation and the practice of demonstration,Footnote ¹⁶⁶ the semantics of the word hodos, the imagistic force of the rut road, and, especially, the discursive architecture provided by the hodos (and Circe’s hodos in Odyssey 12 in particular) may well maintain some degree of priority in their interpretation of Parmenides’ arguments; this last component would provide the matrix of discursive possibility available to Parmenides within which to undertake his metaphysical or cosmological endeavours.Footnote ¹⁶⁷ For their part, literary critics of the sort who study Pindar, perhaps, or even Homer – with perhaps still other commitments concerning the relationship between words and ideas – might go so far as to advance a form of the stronger claim that in some respects it is Parmenides’ road imagery that plays an active role in driving his discursive structure, just as one might uncontroversially claim the same for either poet.Footnote ¹⁶⁸

Finally, working on yet another level, other scholars of ancient poetry might ‘give priority’ neither to the content of Parmenides’ arguments nor to the role played by his imagery in shaping their form; rather, they might be more interested in the analysis undertaken above as a case study in reception theory, one that departs from the usual strategy of dissecting repeated phrases, or type scenes, or cleverly pointed allusions, and moves towards an approach oriented towards archaeological explorations of discourse. Or, similarly, they might perhaps find the above study more useful as another data point to be woven into a larger story about the diverse modes of engaging with, and reworking, Homer that blossomed in the late archaic era.Footnote ¹⁶⁹ How best to incorporate the analysis undertaken here into one’s understanding of Parmenides’ poem is a choice that each scholar will make depending on his or her own orientations and methods, philosophical commitments, and aims and objectives.

It is also possible, however, that in the final analysis even the dichotomy between the philosophically minded and the history of thought- or poetry-minded analysts of Parmenides will not fully withstand deeper scrutiny. What should a member of the first group who finds McKirahan’s reading of Fragment 8 compelling say to a historian of thought who defends the reading I have advanced here? Surely some account of Parmenides’ invention of extended deductive argumentation and outline of demonstration is required; barring this, we find ourselves back in the Greek Miracle paradigm. And what should future interpreters who attempt to forge their own path, finding satisfactory none of the interpretations of Parmenides’ arguments currently on offer, think of all this? Most crucially: to what factor or set of factors should they give priority as they do so?

This final nexus of questions takes on extra significance in light of the positive reception that McKirahan’s analysis has received.Footnote ¹⁷⁰ I noted above McKirahan’s injunction that our interpretations of Parmenides’ arguments should not be imprisoned by an anachronistic understanding of what makes Parmenides’ arguments ‘good’. Like McKirahan, I, too, wholly subscribe to the notion that one consequence of this is that ‘the interpreter’s job is not to aim for formal validity, but to attempt a reconstruction of Parmenides’ train of thought, showing how he might have supposed that the conclusion follows from the premises he gives’. But needing to remain alert to the risks of binding our interpretation of Parmenides’ arguments within the straightjacket of subsequent canons of argumentation does not imply free licence to interpret them without any consideration for the imagery or discursive architecture in which he chose to express himself. Put differently, that the rules governing their order and structure are not those of Aristotelian or Fregan logic does not mean that we can ignore larger questions concerning the ordering, patterning, and overall structure of Parmenides’ arguments in toto. As McKirahan’s own phrase suggests, just what it means for a conclusion to ‘follow’ from a premise is precisely what is at stake in our different understandings of Parmenides’ poem. That the sense of many words and phrases crucial to Parmenides’ arguments in Fragment 8 (such as eon … eonti pelazei at line 25, for example, or akinēton at line 26) remain obscure and hotly contested is widely acknowledged. And if we peer through so dark a glass at the meaning of so much of Parmenides’ language, one might ask just how comfortable we should be in giving priority to our speculations about the ‘content’ of this language – especially when considering what it meant to Parmenides for a ‘conclusion to follow from a premise’, or how best to reconstruct his ‘train of thought’.

By contrast, what I hope to have shown here is that we have a much better foundation upon which ‘to attempt a reconstruction of Parmenides’ train of thought’ – or, rather, as he himself called it, his hodos dizēsios. This is, of course, to study the nature of the hodos part of the hodos dizēsios. Why might Parmenides have used this term? What resources did it offer him? How might it have exerted its own influenced him in turn? These are the questions I hoped to have answered, or to have begun answering, in this book.

I opened this study by discussing the heavy price scholars have paid for anachronistically treating Parmenides’ poem as if it were nothing more nor less than a sequence of extended deductive arguments as we understand that term.Footnote ¹⁷¹ Doing so not only cast aspects of Parmenides’ argumentation in an unjustly unflattering light, but also obscured the seminal role he played in forging from the discursive forms he inherited a new and powerful way of speaking persuasively – one that shares decisive features with what Aristotle would later call apodeixis or demonstration (and, indeed, defines and establishes them). But detaching Parmenides from the story of what came after him for (well-intentioned) fear of anachronism is arguably no less dangerous, no less distorting – and no less anachronistic. Demonstration does have a direct progenitor and distinguished pedigree in the road-thought and road-speech that Parmenides explicitly invokes. And, much more to the point, as I have tried to establish in this book, Parmenides’ road-thought and road-speech is in turn integrally related to the road-thought and road-speech of his predecessors, specifically Homer, especially what we find in Odyssey 12.37–141. It is precisely this inherited discursive infrastructure that Parmenides reuses and reworks to craft his own radical new way of thinking and speaking persuasively – and thus precisely what can offer us such a promising basis upon which to reconstruct his ‘train of thought’ and grapple with what it might mean to him for ‘a conclusion to follow from a premise’ in the movement of his hodos dizēsios.

It is, however, just this road-thought and road-speech, so definitive for the shape and texture of the design of the ‘Route to Truth’, that McKirahan must jettison to get his interpretation of its arguments to stick. One could say – no doubt somewhat idiosyncratically – that it is as if for McKirahan, Parmenides’ arguments are a kind of jigsaw puzzle-baby that must be rescued from the bathwater of their argumentation in order to be assembled properly outside it; by contrast, I would contend that Parmenides’ argument-baby has in fact been developed exactly to fit the bath.Footnote ¹⁷² It does not follow from this, incidentally, that the philosopher’s, or historian of philosophy’s, concerns must be rigorously secondary to those of the historian of thought or the literary critic. Rather, adequate attention to the structure of Parmenides’ argumentation (thanks to the efforts of the latter) can be an invaluable guide in helping the former grapple with his or her quandaries. Likewise, insights divined by the former can help the latter to refine and improve his or her analysis – which can in turn help guide further study by the philosophers, and so on. By considering questions of form and content as deeply – inextricably – interrelated, we can better understand the shape of this bath and the nature of the philosophy-child that it holds, which is both the scion of Homer’s line and the founding dynast of Western philosophical and scientific thinking.

Taking several steps back, we may also observe that trying to square the historical account offered here with the interpretive accounts offered by Owen, Sedley, Austin, and especially McKirahan is a valuable exercise in its own right. This enterprise highlights just how complex is the web of hermeneutic assumptions and interpretative priorities that any reader of Parmenides’ poem brings to bear on his or her reading. When it comes to the Presocratics, to whom we are so indebted for the modes of thought with which we investigate themFootnote ¹⁷³ and yet whom we still so little understand, the truism that what we get out of the hermeneutic circle depends on where we enter it is even more vertiginously true than usual. Are we invested in locating Parmenides in his physical time and linguistic context, or was his brilliance such that this is unimportant, that whatever the nature of his intellectual or discursive milieu might have been, he would not have been constrained – or perhaps even influenced – by it? If we do want to discuss language and imagery, is this to be done in relation to the Homer (or Hesiod) of Parmenides’ past, say, or to the Plato (or Democritus, or Empedocles) standing in his future, or to Orphic or other religious – or legal, or what have you – language that may have been current in the Elea of his present? If we want to gain purchase on just what, precisely, Parmenides was arguing for, how much should we emphasize those against whom he might have been arguing (and should that be an Ionian cosmologist, or Heraclitus, or members of a competing mystery cult or religious sect, or some other under explored or still-unexplored possibility?), the specific language of the arguments themselves, their form, the way that Parmenides’ different successors understood them – or the degree to which any of these factors might still have a bearing on our own contemporary issues, in philosophy or elsewhere? How important is it that Parmenides be understood to argue as we do today? If it is important, how powerful is our commitment to the soundness or validity of Parmenides’ arguments? How much do we feel the need to ‘salvage’ them if we wish to preserve Parmenides’ standing among the giants in the history of thought?

These are important questions, each of which can be answered in a number of legitimate ways – and in each case we are likely to see a subtly or profoundly different Parmenides emerge. Ultimately, of course, how we answer will likely tell us more about our own theories of language, of the history of conceptual change, and of the process by which new modes of thought emerge than about Parmenides himself. For my part, I would urge that we spend at least some time viewing Parmenides as we would any other archaic Greek poet, taking care to historicize his use of language, its sense and reference; to re-embed him not only within his intellectual tradition, but also, especially, his poetic tradition; and to attend to the manner in which the form, imagery, and content of his poem are interrelated. Even for those interpreters who insist on giving hermeneutic priority strictly to content independent of form (on the premise that the one could be strictly independent of the other), these considerations must remain a powerful criterion in assessing the strength, persuasiveness, and credibility of philosophically oriented interpretations of Parmenides’ arguments. Ideally, however, the historical question of how Parmenides came to argue as he did will become a top-tier consideration in its own right, assuming a well-earned place alongside questions such as against whom, or in favour of what, he might have been arguing. It should ascend, that is, to the status of a premier consideration orienting our hermeneutic stance to Parmenides’ poem, and especially the arguments he advances in the ‘Route to Truth’.