Language and doctrine in Parmenides’ Way of Reality

Abstract As early as Plato and as recently as current scholarship, readers of Parmenides have diagnosed tensions of one sort or another between his ontological views and the language through which he expresses those views. In the first instance, this article examines earlier claims for such tensions and argues that they are predicated on problematic assumptions concerning Parmenides’ ontological commitments or his strictures regarding the use of language. In the second instance, however, it argues that Parmenides’ Way of Reality does indeed confront us with tensions between language and doctrine, that these tensions are more pointed and sustained than scholars generally recognize and that they can be identified independently of specific or determinate elaboration of Parmenides’ precise ontological views. This analysis discloses a reflective preoccupation with, and a consistent attitude towards, the scope and limitations of human language. Parmenides persistently evinces his awareness that his description of what-is proceeds through expressive measures that are imported with difficulty from a different domain and, consequently, are limited, indirect and often figurative. The article closes by pointing to a meaningful (but partial) affinity between Parmenides and those Platonists who placed their own ultimate philosophical and ontological principle beyond the expressive reach of words.


I. Introduction
As early as Plato, readers of Parmenides have detected tensions of one sort or another between the ontological views they take him to advance and the language and imagery through which he presents those views. 1Attention has also been given to Parmenides' conception of 'naming' (ὀνομάζειν) and overt critiques of human naming, sometimes in connection with his few, deeply obscure remarks on speaking (and thinking) in the Way of Reality. 2 And yet, while the more unmistakable examples of tension between doctrine and language are occasionally marked in passing, scholars seldom ask what consequences such tensions should have for our understanding of Parmenides and the status of the language that makes up his poem. 3This article takes a fresh look at this question.If there is a tension between language and ontology in Parmenides, and I will defend the view that there is, then what precisely is the nature of this tension?What should we, as interpreters of Parmenides, conclude from it?
An important step will be to disentangle clearly different sorts of potential tensions between language and doctrine.In section II, I argue that some oft purported cases of tension are best seen as specious, and that lack of clarity on this point muddies the interpretive waters.Put differently, readers and critics are sometimes too quick to declare a mismatch between Parmenides' language and (as they see it) his doctrine.As we will see, our precise view on which aspects of the poem do or do not occasion these tensions will affect dramatically our evaluation of their interpretive significance.Existing treatments of this issue are often predicated on specific and narrow assumptions about Parmenides' ontology, viz.that only what-is exists while all else is dismissed as non-existent illusion.But, if this used to be the standard or default interpretation of Reality, this is no longer the case and, indeed, a growing majority of scholars rejects this interpretation as unsatisfactory, or even a non-starter. 4In the wake of this disruption of the old orthodoxy concerning Parmenides' ontology, and the current, almost chaotic proliferation of alternative interpretive models, there is a new need to ask what can be gleaned by revisiting the tension between language and doctrine in Parmenides' poem with fresh eyes, not beholden to the old and highly restrictive way of construing his ontological commitments, nor yet to some other, very specific or determinate ontological model.
In section III, I turn to what I consider genuine and under-discussed tensions between language and doctrine in Parmenides' account of what-is in DK28 B8, and I suggest that these are more pointed and persistent than scholars generally recognize.My claim will not be that the expressive limitations of human language were Parmenides' primary philosophical concern, or even that he developed a detailed, worked-out theory on this matter that he elected to convey no more overtly than the evidence we will consider.What this evidence will disclose, however, is a sustained and reflective preoccupation on Parmenides' part with the scope, orientation and limitations of human language, as well as a consistent and critically aware attitude in response to this preoccupation.The tensions we will examine ultimately tell us something about what we can and cannot expect our words to achieve.Parmenides takes himself to be arguing conclusively for a certain view about the nature of Being, and yet it is a conception of Being that we remain unable to express through language in a full or straightforward way.The goddess puts us in a position to think and cognize (νοεῖν) what-is.And yet, when we attempt to go beyond the secure statement that what-is is and put its characteristics into words, we quickly find ourselves straining against the expressive boundaries of our language.
On the view I defend below, the tensions between language and doctrine do not infect the logic of the goddess' arguments.They do not undermine her inferential moves themselves.Consequently, it is not the case that, despite the various programmatic remarks indicating otherwise, Parmenides in fact places a question mark over Reality as much as he does over Opinion. 5The difficulty posed ultimately concerns, not the acceptability of the goddess' claims about the nature of what-is, but the expressibility of what-is itself.6I will also contend (in section IV) that it is textually untenable simply to deny that the goddess offers a description of what-is, even though the text does indeed confront us with challenging questions concerning the status of the language of which this description consists. 7ome readers may find themselves uneasy about allowing daylight between, on the one hand, the intelligibility and acceptability of a philosophical principle or stance and, on the other, its expressibility in language.But, deep-seated and even well-placed as such unease may be, it must not blind us to productive interpretive possibilities in relation to the significance of the marked and striking tensions between language and ontology in Parmenides' poem.There is a similarly entrenched and perhaps not unrelated unease among commentators to permit meaningful philosophical affinities between Parmenides and the later Platonic tradition over which he has clearly exercised (via Plato himself) significant influence.Indeed, it is no exaggeration to say that it is generally considered an interpretive desideratum to distance Parmenides from the Platonists. 8Going against this tendency, I will unapologetically point in my conclusion to a meaningful (though partial) affinity between the Parmenides who emerges below and Plotinus, who places the ultimate philosophical and ontological principle (for him, 'the One') beyond the expressive reach of words.
A final point of clarification.In speaking of tensions between language and doctrine in Parmenides I will not be suggesting that we can somehow isolate the philosophical content of Parmenides' poem from its (poetic) verbal expression and reformulate for him this content in language free of those difficulties.9Rather, my suggestion will be that one aspect of the poem's philosophical content itself is a marked tension or dissonance between Parmenides' ontological doctrine and the language through which he in fact expresses this doctrine. 10

II. Questioning some claims of tension
At one juncture in Plato's Sophist (244b6-d13), the Eleatic Stranger argues that the monists, 'those who say that the all is one' (τῶν ἓν τὸ πᾶν λεγόντων, 244b6), could not coherently express their position, let alone defend it in a dialectical exchange. 11Although Parmenides is not mentioned here by name, Plato earlier (242c4-6, d4-6) and elsewhere (ἓν ... τὸ πᾶν, Prm.128a8-b1) presents his position in these terms. 12The Stranger argues that the monist could not maintain that 'being' and 'one' refer to the same thing, because this would entail admitting at least two names (δύο ὀνόματα) and, therefore, at least two things (244c8-9).Indeed, even a single name would entail distinguishing between signifier (ὄνομα) and referent (πρᾶγμα) and so, again, admitting in our ontology at least two things (244d3-4).Plato cleverly underscores the difficulty through his use of ἀπόκρισις and its cognate verb. 13If we are to give answers, or in general to speak, then we are to impose distinctions, for example between different names, or between signifiers and their referents.Really, the monist was doomed as soon as the Stranger said, 'let them answer' or, literally, 'let them set things apart' (ἀποκρινέσθων, 244b9, c4-6). 14he Stranger's argument targets a strict monism.Since the monist only admits the one Being, there is no logical space in their world for two names, or for both a name and a referent. 15A similar strict conception of Reality appears to underpin the arguments of Rose Cherubin for the view that the goddess' language betrays a problematic reliance on concepts and presuppositions imported from the opinions of mortals. 16To take a representative example, Cherubin maintains that, by charting our progress through the steps of an argument, Parmenides helps himself to the idea of motion and change, a facet of mortal opinions that Reality had ostensibly discarded. 17If, however, Reality never alleged that the unchanging and motionless what-is excludes the existence of everything else, then there is no obvious problem with the idea that something else (for example, humans or the goddess) can move through the premises of an argument. 18hile, thanks in no small measure to Plato's influence, the strict interpretation of Parmenides' ontology was once the standard or default one, this is no longer the case.The question of how what-is might relate to the heterogeneous items that permeate our everyday experiences (as well as Parmenides' Opinion) is more contested than ever.The literature abounds with different permissive ontological models, which allow for the existence of things other than what-is.For example, on one interpretation (of which we find ancient versions in Plutarch and Simplicius), what-is is the only 'real' or 'genuine' (the key terms here are ἀληθείη and ἀληθής) thing, but not thereby the only thing that exists, where this means that it has a mode of being that in no way involves it with notbeing and that exemplifies being perfectly.Thus, for example, there are no times or places in which it is not, and its properties (changelessness, homogeneity, indivisibility, etc.) are all held absolutely and do not implicate it in 'more' or 'less'.Since it is only this eternal, 13 ἀποκρινέσθων (244b9), ἀπόκρισις (244c3), ἀποκρίνασθαι (244c6). 14Cf.Cra.388b13-c1: a name is a tool for distinguishing the essence of different things (διακριτικὸν τῆς οὐσίας), as a shuttle is a tool for dividing warp and woof.
15 Palmer (1999) 166-73 reads the argument as targeting rather a predicational monist who maintains that each thing can be ascribed only its one proper predicate (for example, 'the person is a person').But why would it then be problematic for the interlocutor that even a single name will involve a distinction between name and referent (244d3-4, with d6-13 elaborating the absurdities that would result from collapsing the distinction between name and referent)?Nonetheless, my use here of this passage from the Sophist neither requires nor precludes agreeing with Palmer (1999) 145, 173 that it does not '[represent] Plato's considered view of Parmenides at this time'.
16 Cherubin (2017); (2018).Plato's Eleatic Stranger and Cherubin concur that the tensions between Reality's stances and the language through which they are communicated undercut those stances.While for the Stranger this is a blunder, however, for Cherubin it is a point that Parmenides wishes us to see.For a self-consciously aporetic and un-dogmatic Parmenides, see also Mackenzie (1982).I address this view below.
17 Cherubin (2017) 257; cf.Owen (1960) 100. 18A similar point applies to the objection from time.Cherubin (2017) 255-57 takes B8.5-6 ('but not ever was it nor yet will it be, since it now is all together, one, continuous') to exclude temporal processes, which inquiry and argumentation necessitate.But, even if what-is is somehow outside the framework of time, it does not follow that so are other things (like humans).Furthermore, Schofield (1970) makes a strong case for taking B8.5-6 as only ruling out that what-is could be subject to perishing or becoming: it is not the case that what-is existed in the past (but not now) or will exist one day (but not yet).The argumentation in the following lines fits with Schofield's deflationary interpretation but offers nothing like a justification for the timelessness of what-is.

Language and doctrine in Parmenides
unchanging and uninterrupted entity that perfectly exemplifies being, it is this that is properly thought of as that to which the term 'Being' (τὸ ἐόν) refers.By contrast, things like humans or trees, which are (here) but also are not (there) or are (now) but also are not (tomorrow), display throughout change, heterogeneity, external divisions and internal articulation.For Parmenides, on this view, such things thus fall short of the rubric of 'reality' (ἀληθείη) and cannot count as a (or the) 'Being' (τὸ ἐόν), but they are not, however, thereby dismissed as non-existent illusions. 19o be sure, this and other permissive interpretations of Reality's ontology all involve some difficulties, and an adequate defence of any particular ontological model will be neither possible nor necessary here.The significant point for us to underline is the negative one.Far from being obvious or the default reading, the strict interpretation that denies existence to anything but what-is, and underpins critiques like those of the Stranger and Cherubin, faces intractable difficulties.After all, if only what-is exists, then it is impossible to see how it could even appear as though other things (such as trees or humans) exist or act.The traditional dismissal of such things among strict interpreters as 'illusions' (a term with no counterpart in the poem) is a non-starter. 20It would be absurd to suggest that it is the unchanging what-is that generates, experiences and is confused by those illusions, but, for strict interpreters, nothing else is there to do so.The strict interpretation asks us to imagine a Parmenides who avowedly excluded the existence of all but what-is while failing to recognize the rudimentary point that he was thereby excluding the possibility of even the appearance of things like humans and human error, or of his own poem, or indeed of a multiplicity of names. 21ne might, however, still think that Parmenides did indeed present in Reality an ontology that excluded the existence of all but what-is, that he was perfectly aware that this ontology conflicted with the language expressing it and with the appearance of the items and processes of everyday experiences, and that his purpose was precisely to leave his readers with this unresolved puzzle.It is just this paradoxical and aporetic Parmenides who emerges from the discussion in Cherubin. 22 As an interpretation of the poem's design and dialectical stance, however, this view is also difficult to accept.The goddess' promise to the kouros (and, by extension, the text's promise to its reader) that he will learn 'all things' (B1.28) gives an air of dogmatic finality.More importantly, the asymmetrical framing of the two parts always presents Reality in thoroughly positive terms.Thus, for example, in the transition from Reality to Opinion Parmenides styles the former, by contrast with the latter, 'a trustworthy account' (πιστὸν λόγον, B8.50).If Parmenides in fact means to place a question mark over Reality no less than Opinion and to leave us uncertain of both, then this and other such programmatic remarks would amount to nothing short of misdirection, and it is difficult to see what the point of such misdirection would be. 23n sum, we need not follow those who identify a tension between language and doctrine in Reality on the assumption that Parmenides expressed there strict numerical monism (whether committing himself to it or as part of an unresolved aporia).Other commentators, however, do not presume this view of Reality yet still advance a version of the Stranger's core contention, that language's reliance on distinctions renders it incapable of coherently articulating the Parmenidean Being.The thought here goes: if Being is unified and undivided, and if language refers and generates sense by means of imposing distinctions, then language cannot articulate Being.This is the attitude underpinning the discussions of Richard Mason and Chiara Robbiano. 24 First, however, I cannot see that our text warrants following Mason and Robbiano in ascribing this very particular view about linguistic reference and meaning to Parmenides.Second, whether we consider this line of thought an interpretation of Parmenides or an objection to him, it lacks bite.It is not clear why language's reliance on distinctions should itself disqualify it from successfully referring to and describing something devoid of real internal divisions.The worry may be that, if our account distinguishes between Being and its predicates ('ungenerated', 'homogeneous', etc.), then it imposes on Being internal divisions that are incompatible with its undivided nature. 25But one could fairly respond that the distinctions between Being and its predicates are conceptual rather than real.The predication that Being 'is imperishable' (ἀνώλεθρόν ἐστιν, B8.3), for example, need not imply that Being and its property of imperishability are discrete items.We may compare a description of Socrates as so many feet tall, which would not imply that Socrates and his height are discrete items, and contrast a description of Socrates as wearing a green shirt, which could fairly be taken to posit Socrates and his shirt as discrete items.Parmenides throughout describes Being only in terms that involve no more than conceptual distinctions of the former sort.Being is one with its characteristic of imperishability.It is true that the account's language itself exhibits real distinctions between discrete components (such as a subject, a copula and a predicate), but this need not preclude it from successfully describing and referring to an entity that exhibits no real distinctions between discrete components.After all, linguistic signifiers that refer successfully to, say, red or square referents need not themselves be red or square. 26herubin sees a different problem with the fact that Reality's arguments are formulated through distinctions and the principle of non-contradiction.She maintains that talk of distinctions and contradictions itself relies on mortal opinions: 'whatever is not Light is Night, and vice versa.Together, Light and Night support distinction [and] division'. 27In this case the poem affords a ready response.It is true that, within Opinion's cosmology, Light and Night function as contrary or even contradictory predicates (B9).It does not follow, however, that any distinction or appeal to the principle of non-contradiction implicitly relies on those two cosmological opposites.Indeed, through his careful use of the vocabulary of krisis, Parmenides takes pains to emphasize that the principle of discrimination that frames Reality's argumentation, between '[it] is' and '[it] is not', is mistaken for accounts of reality, see Nehamas (2002) 59; Johansen (2016) 20; Tor (2017) 199-202.On Opinion's lack of 'real trust', see also Palmer (2009) 92, 167-75.Mason (1988); Robbiano (2016); (2018). 25This concern is prominent in Mason (1988). 26Similarly, Coxon (2009) 21-22 sees the relation between Being and its predicates as one of identity: the different predicates refer to Being under different descriptions.He observes (pp.22, 31-33) that predication and identity statements were not yet clearly demarcated in the Archaic period; see also Kraus (1987) 67; Sattler (2020) 64-65.I do not believe that σήματα at B8.2 (see section III) licenses us to see remarks like ἀνώλεθρόν ἐστιν as somehow not ascribing properties to Being (although, again, such predications need not thereby import nonidentity); contrast Sattler (2020) 102, 115.On the divergence from Plotinus' discomfort with subject-predicate statements about the One, see section IV. 27 Cherubin (2017) 253 and passim.
Language and doctrine in Parmenides qualitatively different from that framing Opinion's cosmology, between 'Light' and 'Night'.
The latter presents us with two opposite principles, each of which is described by contrast with the other (B8.55-59) and both of which are retained in the cosmology. 28Conversely, the goddess exhorts us to retain only one side of Reality's discrimination: to align our notion of Being strictly with '[it] is' and reject as a non-starter the idea of postulating what-is-not. 29o close this section, let us consider two final claims on the basis of which readers have diagnosed a tension between language and doctrine in Reality: first, that Parmenides forbids speaking of what-is-not, yet does so himself; second, that Parmenides considers what-is the sole object of meaningful speech, yet continues to speak of other things as well.
It has long been held that Parmenides contravenes his own strictures about the use of language by speaking of what-is-not. 30The two types of pertinent cases are Parmenides' use of negative predications, as in his remark that what-is is not divisible (οὐδὲ διαιρετόν ἐστιν, B8.22), and his own uses of the term 'what-is-not'.In neither case, in my view, does Parmenides contravene his strictures. 31We must begin by clarifying just what these do and do not prohibit.In B2, the goddess states that the second road is not at all amenable to inquiry (παναπευθέα), explaining that one cannot know what-is-not (οὔτε γὰρ ἂν γνοίης τό γε μὴ ἐόν) nor 'indicate' it (οὔτε φράσαις, B2.6-8).The verb φράζειν fundamentally carries the sense 'indicate', 'show', 'make evident', rather than 'say' or 'tell'.One might of course 'exhibit' or 'explain' something verbal (such as a μῦθος) or indicate something through the medium of speech or writing, but φράζειν can just as naturally be used when something is 'indicated' or 'shown' precisely without being spoken. 32The goddess, then, does not prohibit all use of the expression τὸ μὴ ἐόν.Rather, she suggests that it would be a non-starter to attempt an inquiry into what-is-not or strive to grasp it or exhibit it through an account.The remark 'it could not be accomplished' (οὐ γὰρ ἀνυστόν, B2.7) colours putative attempts to do so as a fool's errand: a venture that could not be completed or consummated. 33till, some have interpreted this claim, that we could not indicate what-is-not, as the denial that we could make coherent use of negative predications, 34 or informative and heuristically effective use of them when inquiring into the real nature of things. 35It is difficult to accept, however, that the goddess pronounces such a stricture about inquiry (παναπευθέα) and knowledge or understanding (οὔτε γὰρ ἂν γνοίης), while she herself inquires into the nature of what-is precisely by working out systematically what it is not like and urging us to apply our mind to what is not generated, perishable, divisible, changing, etc. 36 Her third and final stricture (οὔτε φράσαις) could not suddenly be deploying 'what-is-not' in a different way.By 'what-is-not', then, the goddess does not have in mind, as is often assumed, negative predication ('x is not F'). 37Her concern seems to be rather with what is not anything at all, that is, what has no properties whatsoever. 38nd indeed we could not begin to inquire into, grasp or indicate what is not anything at all, and one would therefore be searching vainly for this on the forbidden road 'that/how [it] is not and that/how [it] must not be'.This also explains why the goddess can interchangeably use 'nothing' (μηδέν, B6.2) as a functionally equivalent term. 39Her point in B2.6-8, then, is that what-is-not lacks any nature, any properties whatsoever, onto which inquiry, cognition or an account could latch. 40She reiterates this same point when she affirms that the road of inquiry into what-is-not must perforce be left 'unconceived, unnamed' (ἀνόητον ἀνώνυμον, B8.17).
What of the goddess' stipulation, 'not from what-is-not will I allow you to say or to think, for it must not be said or thought that [it]  Parmenides nowhere denies our evident ability to use grammatical negations in relation to items that do have positive properties.He also nowhere pronounces the term 'what-is-not' unspeakable or prohibits making use of it.To begin with, then, we need not saddle Parmenides with an aversion to negative language.In the case of the term 'what-isnot', though, we should acknowledge that he is indeed walking a conceptual tightrope.If pressed to give an account of this term, he would have to respond that no such account could be given.But this is just the point.Parmenides asks us to choose between a viable notion that can be developed and grasped (what-is) and a conceptual non-starter (what-isnot).Suppose we were asked to choose between positing a non-round square and, alternatively, a round square in the Euclidean plane. 42That in both instances the latter concept ('what-is-not', 'round square') is incoherent and must remain counterfactual is precisely the point.Parmenides leverages the incoherent prospect of an item without properties in order to recommend as compelling the alternative and to work out the nature of this alternative.A good example is the goddess' use of the counterfactual prospect of what-is-not when arguing that what-is is indivisible and evenly distributed: 'for neither is it the case that there is not-being (οὔτε γὰρ οὐκ ἐὸν ἔστι), which would halt it [sc.what-is] from coming into the same' (B8.46-47).This and such uses of the term 38 On the absence in Parmenides of a neat distinction between complete (existential and non-predicative) and incomplete (predicative and non-existential) senses of εἶναι, see especially Brown (1986)  54-57, 69; (1994) 216-20. 39 See also the progression in B8.7-12 (ἐκ μὴ ἐόντος . . .τοῦ μηδενὸς . . .ἐκ μὴ ἐόντος).For what-is-not as what is not anything at all, cf.Kahn (1969) 716; Ketchum (1990) 171-73; Palmer (2009) 101-04 (but taking 'what-is-not' as shorthand for 'what-is-not-and-must-not-be'); Sattler (2020) 98-100.Sanders (2002) 104 reads against the grain when he claims 'a fundamental difference in meaning' between 'what-is-not' and 'nothing'.
40 Consequently, when the goddess describes things like Light and Night, humans or locomotion, or when mortals mistakenly present such things as true reality, they are not indicating or displaying what-is-not.They are not doing what in Reality the goddess pronounces impossible (contrast Di Iulio (2021) 205-07 with n.1, 213-14 with n.12, 225).Such things possess properties incompatible with those of what-is, but they do possess properties.Thus, they are not nothing at all and indeed, unlike what-is-not, can be inquired into (B1.28-30;B10), grasped (B10) and displayed through an account (B10-19).This is why such things can be designated neither simply as 'being' nor as 'not-being'.
41 On this point, cf.Sanders (2002)  'what-is-not' are not themselves incoherent nor do they contravene the goddess' stricture that we cannot 'indicate' or 'display' what-is-not. 43et us turn now to the contention that the unchanging and homogeneous what-is exhausts the set of things that can be spoken of meaningfully, a view that some still-influential studies presented as key to Parmenides' thinking. 44Our evidence does not support this view.Whatever we think about the status of Opinion's cosmology, there should be no question that the goddess' account of it is not meaningless. 45Indeed, the goddess urges the subject matter of Doxa as objects of inquiry (πυθέσθαι, B1.28; πεύσῃ, B10.4), by contrast with what-is-not, which resists inquiry altogether (παναπευθέα, B2.6).The goddess' description of wayward mortals as plying 'an aimless eye and echoing ear and tongue' (ἄσκοπον ὄμμα καὶ ἠχήεσσαν ἀκουήν | καὶ γλῶσσαν, B7.4-5) plausibly refers to the attitudes that underpin the mortal world view, with 'tongue' in particular gesturing towards a language structured by categories such as 'cominginto-being', 'perishing' and 'changing place' (cf.B8.38-41). 46It would be a stretch, however, to pin on this reference to 'tongue' the view that all language that describes items other than the unchanging and homogeneous what-is (including, say, the rather specific contention that the moon gets its light from the sun: B14-15) amounts to nothing more than nonsense.
We also find no tenable support for this view in the goddess' deeply obscure remark: χρὴ τὸ λέγειν τὸ νοεῖν τ' ἐὸν ἔμμεναι• ἔστι γὰρ εἶναι, | μηδὲν δ' οὐκ ἔστιν (B6.1-2).To be sure, the syntax is (by design?)underdetermined and amenable to different syntactical construals, including (1) 'it is necessary to say and to think that Being is; for it is to be, but nothing [it?] is not', 47 (2) 'it is necessary to say that this is Being and to think that this is Being, etc.', (3) 'it is necessary that saying and thinking be real-and-true, etc.'. 48Far less plausible, however, is the once-popular rendering 'what can be spoken and thought must be'. 49This construal requires us to take τὸ ... ἐὸν as a potential articular participle ('that which it is possible ...') with the infinitives (λέγειν ... νοεῖν τ') embedded in between, an extremely contorted construction for which it is difficult to find passable parallels.Charles Kahn's harsh verdict is warranted: 'no one would construe the verse this way except under pressure from a previously established view of what Parmenides should be saying'. 50Quite possibly, then, B6.1 recommends to us the truth that what-is is (ἐὸν ἔμμεναι) as a privileged object of both speech and thought; that is, as something (the one thing?)we ought to say and think without reservation (about our mode of expression or anything else). 51This is the other side of the injunction not to say or to think that what-is-not is (B8.8-9,17).Parmenides, however, nowhere expresses the view that what-is is the sole object of meaningful speech (let alone that speaking itself is somehow equivalent to being or to thinking and cognizing), a view which would conflict with the apparent presentation in the poem of a great many other objects of meaningful speech, such as mortals, the goddess and the various topics of Opinion. 52p to this point, I have argued that aspects of Parmenides' poem which were taken by some readers, as early as Plato and as recently as Robbiano and Cherubin, to yield clashes between language and doctrine in Parmenides do not, after all, yield such clashes.Before moving on to analyse what I take to be genuine cases of such tension, I wish to highlight a final, literary weakness in the view that Parmenides contradicts Reality's ontology whenever he deploys distinctions, uses grammatical negations or plurals, or refers to items that involve heterogeneity and change (such as humans).As we are about to see, Reality does indeed bring to our attention more specific, pointed and, in some cases, subtle conflicts between language and doctrine.If, however, Parmenides in fact contradicted Reality's ontology through more or less everything he wrote there, whenever he strung two words together or even just one, then those more nuanced tensions would be rather drowned out.

III. Rehabilitating a case for tension
Let us turn now to passages in Reality that do after all confront us with tensions between language and doctrine.Indeed, we will see that the text repeatedly foregrounds these tensions in a more pointed and sustained manner than is generally recognized.Furthermore, we can identify these tensions independently of presuming some specific or very controversial interpretation of Parmenides' account of Being.The cumulative effect of these passages is an interpretive puzzle that makes a claim on our attention: why does Parmenides repeatedly choose to couch certain ontological views through language that stands in stark tension with those views?
Only one tale of a road yet Remains: that [it] is; and on this road there are signs Very many, that what-is is ungenerated and imperishable, Whole, single-born and unshaken [and complete?]. 52The goddess makes one other reference to speech in her equally difficult and uncertain remark: οὐ γὰρ ἄνευ τοῦ ἐόντος, ἐν ᾧ πεφατισμένον ἐστιν, | εὑρήσεις τὸ νοεῖν (B8.35-36).The most plausible construals that have been proposed are: 'for not without what-is, in what has been expressed [sc. the foregoing verses], will you find cognizing [alternatively, 'thinking' or 'understanding']' (cf.Sedley (1999) 120); 'for not without what-is, in that in which cognizing has been expressed [sc. the foregoing verses], will you find cognizing' (cf.Robbiano (2006)  169-70); 'for not without what-is, depending upon which cognizing has been expressed, will you find cognizing' (cf.Mourelatos (2008) 170-72; Palmer (2009) 164 n.40).Plausibly, then, the goddess refers to the preceding verses as an expression of something, quite possibly as an expression of cognizing (πεφατισμένον ἐστιν), and she emphasizes the dependence of cognizing upon Being in the context of this expression.Nothing here suggests, however, that what-is is the sole possible object of meaningful speech, contra Cornford (1939) 34.

Language and doctrine in Parmenides
Parmenides' talk of 'signs' in B8.2 is a first indication that what follows falls short of a straightforward or unproblematic articulation.Heraclitus DK22 B93 offers an instructive comparison: 'the lord whose oracle is the one in Delphi neither says (λέγει) nor conceals (κρύπτει) but gives a sign (σημαίνει)'.Apollo neither speaks the answers to our questions nor takes action to prevent us from acquiring them: instead, his signs offer a starting point from which, with well-directed effort, we can gain insight and understanding concerning our inquiries. 53The goddess' 'signs' here appear to promise 'proofs that something is the case'. 54At the same time, from Parmenides' characterization of what comes next as 'signs', we might already suspect that his discussion of what-is will amount to something other and more demanding than a direct or full spelling-out.This impression will gain further corroboration and resonance when we look back at the goddess' talk of 'signs' by contradistinction with the 'names' that, in turn, underpin and make up the cosmology of mortals, and which are typified by both conventionality and a strong commitment on the part of speakers to the appropriateness of the language used. 55f our suspicion is first roused by the term 'signs', the following words will not allay it.Lines 3-4 offer a catalogue of properties (shown by the 'signs') that, as the goddess will demonstrate, typify what-is.Strikingly, at least three of the terms in question are negations of properties with which we are familiar from our sensory experiences of multiple and heterogeneous things ('un-generated', 'im-perishable', 'un-shaken').Indeed, throughout Reality, the goddess continues to characterize what-is predominantly by working out what is not true of it. 56Clearly, μουνογενές cannot be ascribing birth to whatis, and we can parallel the sense '-kind' for compounds ending in -γενής. 57What-is, then, is 'of single kind'. 58But the literal sense, 'single-born' (as in Hesiod, Theog.426; Op. 376), will be the first to hit the ear, and it is notable that Parmenides resorts to this vocabulary immediately after designating what-is 'ungenerated' (ἀγένητον). 59Our text for the final predicate in B8.4 is uncertain.We may read that Being is 'not incomplete' (οὐδ' ἀτέλεστον) or, alternatively, that it is 'complete' (ἠδὲ τέλειον). 60On the former alternative, Parmenides ends the sequence with yet another alpha-privative adjective.On the latter, we conclude with a positive term, albeit an extremely vague and schematic one.Indeed, whatever Parmenides originally wrote in B8.4 b , this predicate will later be expanded principally through a series of negative remarks, for example that what-is is not incomplete or in need (B8.32-33, 42-49).On the whole, we can say that, even as he introduces the project of deducing its properties, Parmenides speaks of what-is through vocabulary that is manifestly and even emphatically appropriate to other things, things that are born or otherwise come into being, perish and exhibit imperfections and needs.Parmenides' signs already show themselves to be indirect: we begin to talk about the nature of what-is by specifying how, in different respects, it is unlike us and the items of our experiences.We appear to lack a vocabulary designed for speaking directly about whatis itself.The description of what-is as 'single-born' (μουνογενές), following close on its characterization as 'ungenerated' (ἀγένητον), makes us conscious in a particularly pointed way of the fact that Parmenides is deploying here language made for very different items and very different descriptive tasks.
Nor ever from what-is-not will the strength of Trust allow Something to come to be beside it;62 for that reason neither to come to be Nor to perish did Justice permit it by loosening her shackles But she holds it fast in them.And the decision about these matters comes down to this: [It] is or [it] is not; and so the decision has been made, as is necessary, To leave the one road unconceived, unnamed, for it is no true Road, but that the other road is and is genuine.This passage does not deduce new properties for what-is.It underscores the basic principle underpinning the goddess' argumentation: what-is must wholly exhibit being and in no way not-being; it is 'fully' (πάμπαν, B8.11).A first consequence of this, as was demonstrated in B8.6-10, is that it did not first come into being.What is new and striking about these lines, however, is that we are abruptly confronted with a proliferation of new figures and factors not previously encountered along the road of what-is.It is suddenly the force of Persuasion or Trust (Pistis) that will not permit something to come to be from what-isnot. 63If one could wonder whether or not 'trust' is portrayed here as a personified character or an abstract logical principle, there can be no such doubt as we proceed to the anthropomorphized figure of Justice, who prevents what-is from coming into being or perishing by restraining it firmly in her shackles, the first in a series of shackle-wielding goddesses we encounter in B8.Now, I do not doubt that, above all, this recurrent imagery provides indispensable conceptual resources for expressing the ways in which the properties of what-is are appropriate and necessary for it and thereby helps guide readers of the poem towards a proper understanding of what-is. 64And yet, I wish to emphasize that Parmenides also develops this motif in such a way as to keep this form of expression at arm's length and to caution us that we cannot take it at face value.In the present case, we may ask why it is, in the end, that what-is did not come into being and will not perish.Before lines B8.12-14, Parmenides answered this question through an argument that centred on the discrimination between '[it] is' and '[it] is not' as mutually exclusive and rejected the latter.Indeed, immediately after his remarks about Trust and Justice, Parmenides returns to the dialectical necessity imposed by this discrimination (B8.15-18).It is impossible to square this line of thought with the idea of external divine agents who, at their discretion, are acting on what-is and influencing it from the outside.The conceit that it was as a result of the logical considerations previously laid out that Justice decided to restrain what is (τοῦ εἵνεκεν ... ἀνῆκε Δίκη, B8.13-14)only accentuates the impossibility of taking Justice's agency at face value. 65The most pointed aspect of the tension here between doctrine and imagery, however, is the clash between the violent and almost bestializing image of what-is as a shackled animal, or perhaps a Titan, restrained in bonds and the previous and subsequent descriptions of it as whole, unshaken, unmoved and perfected (B8.4-6, 26), as well as 'altogether inviolate' (πᾶν ... ἄσυλον, B8.48). 66he following lines (B8.19-21) foreground another sort of tension: πῶς δ' ἂν ἔπειτα πέλοι τὸ ἐόν; πῶς δ' ἄν κε γένοιτο; εἰ γὰρ ἔγεντ', οὐκ ἔστ', οὐδ' εἴ ποτε μέλλει ἔσεσθαι.τὼς γένεσις μὲν ἀπέσβεσται καὶ ἄπυστος ὄλεθρος.
And how could what-is be hereafter?And how could it become?For if it came to be, it is not, nor is it if it is going to be at some time.
Thus coming-to-be is extinguished and perishing is unheard of.
Parmenides here goes out of his way to frame his conclusions about what-is through terms whose inapplicability to what-is, and unsuitability in an account of reality, he already takes himself to have established.Being was never characterized by a coming-into-being that could be 'extinguished' (nor a perishing that could go unheard).In this context, it is as enticing as it is nonsensical to suggest that coming-into-being perished.Both this particular tension between language and doctrine and others discussed above resurface in the goddess' remarks at B8.26-31: 64 For discussion, see Mourelatos (2008) 25-29, 37-40, 115-20, 148-54, 160-61; Austin (1986) 110-13; Robbiano  (2006) 162-76.It is tempting to consider the various restraining goddesses as different aspects of the same principle; cf.Mourelatos (2008) 160-61 on Constraint-Fate-Justice-Persuasion as representations of the 'modality of necessity' determining the identity of what-is, spanning a spectrum from gentle persuasion (B8.12) to brute force. 65Evans (2021) sees the appeal to Justice at B8.14 as performing a more robust role within the argumentation than is usually thought: Justice is what ensures, crucially, that what-is is reliable and therefore stable (especially pp.24-33).Evans continues, however, to see this Justice as in reality an immanent aspect of what-is (pp.30-33) and her shackling as non-literal imagery (p.3).On his account too, then, one may still wonder about the tension between this imagery itself and the entity in relation to which it is used. 66Victoria Wohl is currently working on a study in which she notes the variance between the (violent) imagery of bonds and the conception of what-is as self-constituted and sees here one instance of metaphysic's essential inability to extricate itself from metaphor.For Being as a chained Titan, cf.Cerri (2018) 229-30, comparing Aeschylus' Prometheus.Morgan (2022) shows how Parmenides appropriates especially Hesiod's descriptions (Theog.521-22, 613-23, 652, 659, 717-21) of Titans and Giants who opposed Zeus and were subsequently chained (in some cases in Tartarus), a particularly resonant echo following Parmenides' appropriations of Hesiod's eschatological imagery in B1 (on this appropriation, see Burkert (1969) 8, 11-13; Pellikaan-Engel (1978) 8-10).On my view, though, Morgan (2022) underplays (especially at p. 233) the tension between the binding imagery and the surrounding account of what-is.On binding in Hesiod as a means of controlling specifically deathless threats to cosmic order (such as the Titans), see Clay (2022).
(i) Further, changeless and motionless in the limits of great bonds It is without starting, without stopping, (ii) since coming-to-be and perishing Wandered very far away, (iii) and true Trust pushed them off.(iv) Remaining the same in the same by itself it lies And so it remains there steadfast; (v) for mighty Necessity Holds it in the bonds of a limit, which encloses it round about.
In (i), Parmenides asserts that what-is does not change or move (the language of limits and bonds recurs, but without mention of an external constraining agent), and denies it both a starting point and an end point. 67In (ii) and (iii), he immediately proceeds to frame this latter denial through the sensory and dynamic imagery of motion and change: the fallen figures of Being's Coming-into-Being and Perishing are forced to wander far away; true Trust reappears as the heroic figure who manages to push them off. 68In (iv) Parmenides stresses the still permanence and steadfastly unchanging nature of Being (ταὐτόν ... μένον ... κεῖται ... ἔμπεδον ... μένει) as well as its solitary isolation and lack of outside interactions (ταὐτόν τ' ἐν ταὐτῷ ... καθ' ἑαυτό).In (v), however, he immediately reprises the contrasting image of Being as a sort of shackled beast or titan that must be restrained in bonds (ἐν δεσμοῖσιν), this time held firmly through the discretion and sustained or even violent effort of mighty Necessity.Once again, the logical connective attaching (v) to (iv), 'for' (γάρ), only accentuates the difficulty of squaring these two divergent portrayals of Being.
(ia) The same thing is both for cognizing and that because of which there is cognition.
(ib) For not without what-is, in what has been expressed, Will you find cognizing. 69(ii) For nothing else either is or will be Beside what-is, (iii) since Moira shackled it To be whole, and changeless and motionless.(iv) All things will be its name, 70 As many as mortals established, believing them to be real, Coming-to-be and perishing, to be and not to be, And changing place and exchanging bright colour.
Even by Parmenides' standards, this is an extremely dense passage.His logical connectives give us at least a broad view of the goddess' inferential moves here.(iii) justifies (ii).What-is is a unified whole that does not change (iii); therefore, what-is will never create something other than, or in addition to, itself; given, moreover, that nothing else could come into being from what-is-not (as was established before), what-is is all there is and there will never be anything other than what-is (ii).In turn, (ii) justifies (i).What-is is all there is (ii), and so what-is is the only object of the thinking and apprehension of which the goddess speaks here (i). 71In addition, (ii) can plausibly be taken to justify (iv).What-is is all there is (ii), and so the various misguided labels that mortals try to pin on reality (πεποιθότες εἶναι ἀληθῆ, B8.39) must in fact have what-is as their true referent (iv).
To be sure, this outline raises questions.Some would infer from the restriction of being to what-is (B8.36-37) that only what-is exists, although this would make nonsense of the goddess' subsequent claim that mortals make errors about reality, and indeed of the very possibility of the goddess and the kouros (seemingly) conversing.Alternatively, as outlined above, we may follow Plutarch's lead (Mor.1114d-e) and take it that Parmenides reserves the term 'being' for what-is since it alone exemplifies being perfectly and without qualification.On this line of thought, it is for this reason that, as (i) suggests, what-is alone is the proper object of proper cognition or understanding and amenable to the true, steadfast and reliable apprehension reflected through the arguments of Reality, and that, as (iv) indicates, what-is is de facto the only possible referent for names that purport to disclose the ultimate reality (πεποιθότες εἶναι ἀληθῆ, B8.39). 72he central point of interest for us, however, is the introduction of Moira.The sudden appearance of this final goddess here is a far from obvious or unsurprising choice.Parmenides had already taken himself to have earned, or at least to afford, the points that what-is is a unified whole (οὖλον) that is unmoving and unchanging (ἀκίνητον).He might that cognizing is identical with that because of which there is cognition, that is, Being (Long (1996) 136; Sedley (1999) 120).On alternative construals of B8.35-36, see n.52 above.On any analysis of B8.34-36, however, νοεῖν must be used here as a normative success term and in reference to the successful cognition or understanding of what-is.It is elsewhere clear that our minds can fail to grasp what-is (B6.5-6;B16); cf.Mourelatos (2008) 175-76;  Calvo (1977) 251. 70For this translation of τῷ πάντ' ὄνομ' ἔσται (as opposed to the traditional 'therefore, all things will be a mere name . . .', which postulates a 'mere' that the Greek does not support), see Kingsley (2003) 190-91; Mourelatos  (2008) 386-87; cf.Od. 19.409 (τῷ δ' Ὀδυσεὺς ὄνομ' ἔστω); Hymn.Hom.Ven.198.This translation diminishes the difference in meaning between this text and the alternative reading τῷ πάντ' ὀνόμασται ('to it all things will be named'; Woodbury (1971); Vlastos (2008); Kraus (1987) 92-94), which faces philological difficulties (see Mourelatos  (2008) 386). 71The strength and precise nature of the connection announced in (i) between what-is and cognizing will vary according to different syntactical construals of B8. 34-36; cf.nn.52, 69 above.Minimally, however, these lines  indicate that what-is is a necessary condition for what the goddess here refers to as 'cognizing' (or 'thinking' or 'understanding').It is difficult to know precisely what to make of the γάρ that connects (ib) to (ia).Even if we grant that, in the foregoing lines of Reality (if this is indeed the force of ἐν ᾧ πεφατισμένον ἐστιν), cognizing was never separate from or independent of what-is (ib), how would this justify the claim that cognizing is cognizing of whatis (ia)?We should perhaps interpret the γάρ at B8.35 as something like 'consider how' and take (ib) to offer only illustrative support for (ia): the fact that, throughout Reality, cognizing was never separate from what-is illustrates (even if it does not prove) that the cognizing of which the goddess speaks here is cognizing of what-is.At any rate, (ii) likely grounds Parmenides' claim in (i) as a whole, rather than only (ib) but not (ia).
72 On these issues, see section II above.
have recapitulated those points here without introducing the constrictive influence of yet another ostensibly external divine force. 73Moira is, moreover, at once a precisely appropriate and precisely inappropriate figure for the task of rendering what-is 'whole'.
On the one hand, Moira apportions to each thing its due lot in the world, and it is the suitable lot of what-is to be whole and unchanging.On the other hand, the term μοῖρα itself means 'part', 'division' and 'distribution', and the goddess Moira incarnates division. 74he suggestion that division renders reality whole, a quality that Parmenides elsewhere describes in terms of indivisibility (οὐδὲ διαιρετόν, B8.22), pulls us up short.In (iv), Parmenides next proceeds to consider the various categories or 'names', like 'coming-to-be and perishing', which mortals wrongly think capture truly the structure of reality.And yet, the weakness of mortal names only draws attention to the goddess' own immediately preceding and jarringly dissonant appeal to the figure of Moira.Furthermore, we cannot overemphasize the point that, although Parmenides precisely refuses to accept their conceptual scheme as apt for mapping out the nature of true reality, the terms in which the wayward mortals wrongly place their trust (B8.38-41) are the same terms through which the goddess communicates with Parmenides, and Parmenides with us, when they describe what-is as without generation (γίγνεσθαι) or perishing (ὄλλυσθαι) or motion (τόπον ἀλλάσσειν).In describing what-is, the goddess does not supplant the names that mortals established with some alternative set of names.Rather, she proceeds, in a largely indirect and negative manner, by demonstrating the inapplicability to what-is of the mortal names themselves.Parmenides continues to speak the mortal language, even as he cautions us that its categories cannot directly articulate the nature of what-is. 75In section IV, we will revisit the point that, in speaking this language, Parmenides not only resorts to such rudimentary terms ('ungenerated', 'imperishable', etc.) for expressing the attributes of what-is, but also introduces the elaborate imagery of the external application of force by Moira and the other shackle-wielding goddesses.
We should not work too hard to fit every aspect or passage of Reality into the pattern of tension between language and doctrine that we already found to be exemplified amply in B8.One important possible exception is the goddess' repeated insistence that what-is is.With the ascription of 'being' to what-is, if nowhere else, we may find a use of language in relation to what-is that the goddess puts forward without pointed indications that she is holding her expressions at arm's length.Indeed, B6.1 likely enjoins us to say (and to understand) that Being is. 76It is possible, although our evidence nowhere conclusively frames things in this way, that we are invited to think of (τὸ) ἐόν as the name whose denotation is the entity whose properties are worked out in B8. 77 Even here, though, things 73 Cf.'Portion','part': Il. 10.253;16.68;'division','distribution','lot': Il. 9.318;15.195;Od. 8.470;Hes. Theog. 413;Hdt. 2.147;'lot in life','fate': Od. 4.475;19.592;for the goddess or goddesses, see Il. 24.209; Hes.Theog.904-06; Aesch.Cho.306, 910-11; cf.Cherubin (2017) 269 n.19.Cherubin (2017) 258 also emphasizes the tension between division into shares (μοῖρα) and the indivisible what-is.I cannot see, however, how we could justify from the text her further suggestion here that Moira represents a requisite for inquiry that quite specifically renders inquiry dependent on securing continuous explanatory linkages among multiple things and, therefore, undermines the conception of what-is that the goddess prima facie demonstrates through her own inquiry. 75Similarly, Jenny Strauss Clay writes (in a study in preparation): 'the goddess speaks the language of mortals fluently'; cf.Pulpito (2019) 193 on the entire poem as ultimately 'un flujo de onomata'.We must not, then, follow the proposal of Calvo (1977) that Reality and Opinion represent the opposition of two different kinds of language, even if they do represent different kinds of thinking.
76 See section II. 77If 'Cornford's Fragment' is authentic, then Parmenides explicitly referred to 'being' as the 'name' of the whole (τῷ παντὶ ὄνομ' εἶναι, Pl.Tht.180e1); see Cornford (1935).But Plato himself presents these words only as the sort of thing typical of 'Melissuses and Parmenideses' (180e2) and they more likely reflect what he conceives of as a broadly Eleatic approach.Nonetheless, B8.38 does present what-is as the (de facto) referent of names.For τὸ ἐόν as Being's name, see Kraus (1987)  64-69, 96.Robbiano (2018) 38 infers from the inclusion of 'being' and 'not-being' in Language and doctrine in Parmenides are not entirely straightforward.Side by side with this technical use of the verb 'to be' in relation to what-is, 78 we also unsurprisingly find Parmenides using this verb in a more everyday sense, and in relation to everyday (generated, perishable, internally articulated, heterogeneous) things, as for example in his remark that the wayward mortals 'are' in a state of having erred (πεπλανημένοι εἰσίν, B8.54). 79I see no strong reason to think that, through this and other such uses of the verb 'to be', Parmenides is (here too) calling attention to the pitfalls and limitations of human language.At the same time, his everyday and indeed loose or catachrestic uses of the term should make us sympathetic to the worry (a worry that I have been arguing Parmenides evinces throughout Reality) that the exigencies of human language make it an inadequate vehicle for spelling out directly or straightforwardly the goddess' ontological doctrine, a doctrine according to which it is only what-is that can unqualifiedly and properly speaking be said to be. 80lso worthy of comment in this connection is the fact that Parmenides several times uses πέλειν, a term that, elsewhere, often indicates dynamic processes of becoming and connotes a sense of motion, as a functional equivalent of the verb 'to be'. 81While this usage of πέλειν is not uncommon in early Greek poetry, its occurrence within a context that repeatedly draws our attention to the dissonance between the ontological message and the language and imagery used to convey it makes it at least possible to find here another such gesture.Most notable is Parmenides' use of the term as a substitute for εἶναι (and the opposite of οὐκ εἶναι) when expounding the aimless wandering of the wayward mortals: . .. that road on which indeed mortals who know nothing stray double-headed (πλάζονται δίκρανοι), for helplessness guides the wandering mind (πλαγκτὸν νόον) in their breasts, and they are borne along (φορεῦνται) deaf and blind alike in the catalogue of mortal names (B8.39-41) that even the term 'to be' cannot describe what-is.It is preferable, however, to read εἶναί τε καὶ οὐχί in B8.40 as the compound name, 'both being and not-being' (note τε καί), and not as two isolated names, which would contradict Parmenides' earlier assertion that not-being is of necessity 'unconceived, unnamed' (ἀνόητον ἀνώνυμον, B8.17) and also his portrayal of mortals as confused thinkers who lack a clear notion of 'being'; see similarly Vlastos (2008) 368 n.4; Furley (1973) 7 n.23. 78 What Kahn (1988) 260 calls a use of the verb with 'ontological force'. 79De Rijk (1983) 49 classifies every occurrence of the verb 'to be' in Parmenides as technical (for example, B2.3; 8.16) or non-technical (in addition to B8. 54, cf. also B1.11, 27; 6.9; 8.57; 9.3; 16.3-4; 19.1).
80 This is the case whether we think of other things as non-existent illusions or as existing items that display an imperfect mode of being (and so could only be said to be through a compromised and loose use of the term); see section II.
81 πέλειν is etymologically connected to an Indo-European family of words signifying motion.The compounds ἀμφιπέλομαι, ἐπιπέλομαι and περιπέλομαι, and the derivatives πολέω and πολεύω, are all motion verbs, while cognate nominal forms (αἰπόλος, ἱπποπόλος, δικασπόλος, etc.) express some sort of activity.See Chantraine  (1968), s.v.πέλομαι; Beekes (2009), s.v.πέλομαι.This strikes me as sufficient grounds for allowing that πέλειν, especially in the movement-rich context of B6 and in the marked alliterative conjunction with ἰσοπαλές in B8. 44-45 (as discussed anon), has a secondary connotation of motion and dynamic change.We need not go so far as to insist that the bare verb πέλω itself still occasionally means 'to move' (or similar) by the time of our earliest sources.It is not inconceivable that it could, but most lexicographers are sceptical.Chantraine (1968), for example, maintains that we only ever find it in the 'weakened' senses 'se produire, exister, être'.Conversely, Bailly (1935) s.v.πέλω, 1, lists passages in which he considers πέλω a motion verb.In all of Bailly's examples it is unnecessary to follow his reading, and in the first three one wonders whether any sense of motion we detect should be credited to the verb in conjunction with a preposition rather than to the verb alone: Il. 3.3 (πέλει οὐρανόθι πρό); Od. 13.59-60 (ἔλθῃ . . .ἐπ᾽ ἀνθρώποισι πέλονται); Od. 15.407-08 (ἐπὶ . . .πέλεται δειλοῖσι βροτοῖσιν); Od. 19.192 ('but now the tenth or eleventh dawn came/took place (πέλεν)'; cf.περιπλομένων ἐνιαυτῶν, Od. 1.16).At any rate, the sense of becoming is clear in these passages and not infrequently elsewhere (Il.13.103, 632; Od. 1.393).Autenrieth (1891) s.v.πέλω oversimplifies when he remarks: 'perhaps originally containing some idea of motion . . .but in Homer simply to be'.bewilderment, a tribe without discrimination, by whom to be and not to be have been deemed the same and not the same (οἷς τὸ πέλειν τε καὶ οὐκ εἶναι ταὐτὸν νενόμισται κοὐ ταὐτόν).(B6.4-9) In this context, rich with imagery of constant fluctuation and movement, the term πέλειν can plausibly be seen to underscore a linguistic aspect of the error of the mortals.The human confusion about the nature of Being is reflected through both the formal contradiction in lines 8-9 ('the same and not the same') and the readiness of human language (a language that Parmenides too continues to speak) to refer to Being through a term redolent with connotations of becoming and movement.In B8.18, πέλειν, alongside εἶναι, indicates that the first road, the road that [it] is, 'is and is genuine' (πέλειν καὶ ἐτήτυμον εἶναι).In the very next line, πέλειν is used to ask 'how could what-is be hereafter' (πῶς δ' ἂν ἔπειτα πέλοι τὸ ἐόν), a usage that converges on the sense 'become' and, indeed, forms a hendiadys with the question that closes the line: 'how could it become?' (πῶς δ' ἄν κε γένοιτο, B8.19). 82Later in Reality, Parmenides remarks: 'from the middle equally advanced in all directions (μεσσόθεν ἰσοπαλὲς πάντῃ), for it must not be (πελέναι) either at all greater or at all smaller here or there' (B8.44-45).Regarding ἰσοπαλές, Alexander Mourelatos observes 'the dynamic connotation of the root παλ-' (he renders ἰσοπαλές 'pushing out equally', 'equally advanced') and later also marks the alliterative effect with πελέναι. 83We may add that this alliteration brings out, and is itself accentuated by, the dynamic connotations of πελέναι itself, and that in the case of both ἰσοπαλές and πελέναι the dynamic connotation is starkly at odds with the stability and changelessness of the equal distribution of what-is that these terms express here. 84his last couplet comes from Parmenides' well-known description of what-is as 'complete' (τετελεσμένον) and 'like the bulk of a well-rounded ball' (εὐκύκλου σφαίρης ἐναλίγκιον ὄγκῳ, B8.42-49).A proper consideration of this complex passage will take us too far off course, but the central interpretive alternatives can be usefully delineated.In the end, we should keep an open mind concerning the import of this analogy to our question.Some scholars infer from these lines that what-is is a spatially extended sphere.It is reasonably pointed out that, grammatically, the phrase μεσσόθεν ἰσοπαλὲς πάντῃ (B8.44) is said of what-is rather than the ball or its bulk. 85Others read the passage figuratively.Mourelatos, for example, insists that Parmenides does not draw an analogy between what-is and a ball but, rather, between the completeness of what-is and the expanse of a well-rounded ball (τετελεσμένον ἐστί ... σφαίρης ἐναλίγκιον ὄγκῳ).He sees the passage's criteria for the sphere's uniformity and completeness or perfection as standing in for criteria of uniformity and completeness in general.In this light, the remark that what-is is 'from the middle equally advanced in all directions' becomes not the assertion that Being is a literal sphere but an attempt to convey the idea of perfect completeness and uniformity. 86On the literal view, we seem to find in B8.42-49 one respect in which our language and conceptual repertoire allow us to articulate directly one aspect of what-is (its physical shape) and, as such, something of an exception to the general trend we found in Reality.On the figurative view, we find here yet another example of Parmenides' need to resort to inevitably imperfect linguistic and imagistic devices that are drawn from very different conceptual domains in order to construct his description of what-is.
Whatever view one takes of the sphere passage, it is clear by now that this project of describing what-is is one for which, by and large, Parmenides does not have quite the right words.He repeatedly brings this predicament to our attention both by underscoring the predominantly negative and indirect means by which his description of what-is proceeds and by foregrounding other terminology and elaborate imagery that sit ill at ease, sometimes pointedly so, with the very views this language is used to convey.The question, then, is why Parmenides does so.What is the point of the tensions between language and doctrine that pervade the Way of Reality?

IV. Conclusions
In the proem, the goddess remarks that the youth has travelled 'far from the track of humans' (B1.27).Later she encourages him to resist letting 'habit born of much experience' (B7.3) force him onto the road of wayward mortals.And yet, though what-is represents an unfamiliar and strange object of inquiry and knowledge, the goddess continues throughout to communicate to the youth about it in the language of mortals.It is appropriate, then, that, time and again and in a variety of ways, Parmenides reflects his awareness that the language he uses to give an account of what-is, and to guide our minds towards an apprehension and understanding of it, has been imported with difficulty from its natural domain, in which it is put to very different uses and in relation to very different objects.As highlighted in the introduction, my suggestion is not that we can tease out from Parmenides' fragments an elaborated theory on the workings and limits of human language, or that this was a primary focus of his efforts.What we can more modestly glean in Parmenides, however, is a reflective preoccupation with, and a consistent and selfaware attitude towards, the scope, orientation and limitations of human language, a preoccupation and attitude that have important philosophical consequences. 87armenides takes himself to be arguing conclusively for a particular view about the nature of Being, and to place us in a position to ascertain for ourselves the truth of this view.And yet, this idea of Being is one that cannot be articulated in a full, direct or straightforward way.We can (perhaps) say that what-is is without evincing any reservation concerning our form of expression.But as soon as we try to go beyond this bare point, and to put into words the nature of what-is, we find ourselves straining against the expressive limits of our language.The description of what-is in Reality is ultimately restricted to a small number of schematic characterizations.This constrained aspect of Reality is especially marked by contrast with the richer and detailed descriptions in the Way of Opinion (as even our meagre extant evidence allows us to say) of the multiple and heterogeneous items of our everyday experience (B8.53-19).
It might be objected that the predicates deduced in Reality for what-is exhaust the nature of what-is.That is, it might be objected that, perhaps once we say that what-is is ungenerated, imperishable, unchanging and immobile, evenly distributed (that is, not more or less here or there: B8. 44-45), all alike and indivisible, etc. (possibly including: spatially extended in the shape of a sphere), we have given a full account of its nature.If so, then our conceptual scheme can even be said to do a perfect job of capturing and articulating the nature of what-is.As we have seen, however, even these few characteristics are by and large expressed negatively and indirectly, by denying to 87 That Parmenides should have been exercised by these concerns is not surprising from a historical perspective.We find different responses to comparable preoccupations with the limitations, partiality and shiftiness of human language and human names in, for example, Hesiod (see Vergados 2020), Pherecydes (Granger (2007)  144-47) and  Heraclitus (Kirk (1962) 116-22; Graeser (1977b) 365-66; Kahn (1979) 270-71; Dilcher (1995) 124-25).what-is qualities that characterize multiple and heterogeneous things. 88What-is, we learn, is un-generated, im-perishable, in-divisible, un-shaken, un-moving, and so on.As we build our understanding of what-is, we repeatedly find ourselves forced to discuss it through a language that is clearly designed for expressing altogether different things, the generated, perishable, divisible, changing, mobile, multiple and heterogeneous items of everyday experience.Indeed, this is the very same language in whose aptness for articulating reality wayward mortals wrongly place their trust (B8.38-41), and through which Parmenides and his goddess continue to communicate about what-is.We may also note how, by contrast with this prevalent tendency of Reality, in Opinion Parmenides is better able to call on positive characteristics, familiar from lived experience, so as to introduce and describe directly the properties of the two cosmological elements: 'the aethereal fire of flame, being gentle, extremely light .. .night unknowing, 89 dense and heavy in body' (φλογὸς αἰθέριον πῦρ | ἤπιον ὄν, μέγ' ἐλαφρόν ... νύκτ' ἀδαῆ, πυκινὸν δέμας ἐμβριθές τε, B8.56-59).It makes little difference (and it is impossible to determine) if it is intrinsic to what-is that it largely resists direct expression or if it merely resists direct expression through human language, for we possess no other.Parmenides' proclivity for negative formulations especially claims our attention in the wake of Xenophanes, whose account of his Greatest God is echoed in Parmenides' account of what-is and whose theological remarks, in recognition of the way in which our perspective is constrained by our experiences (DK21 B34), tend to stipulate negatively what the gods are not like. 90t is in this context that we should view the especially marked cases of tension between language and doctrine in Reality.Parmenides not only describes what-is in a largely negative and indirect manner, but also repeatedly introduces elaborate language and imagery that is pointedly and precisely dissonant with its nature (as, for example, with 'Moira' and the imagery of shackles applied from the outside by anthropomorphized goddesses).In doing so, Parmenides alerts us to the general inappropriateness of our language for an articulation of what-is. 91Put differently, the problematic and elaborate mythological imagery flags up what the fundamental and perhaps more inevitable negative descriptions also repeatedly convey in their own way: that, time and again, we are straining to describe the nature of what-is through a language that is ill-suited to the task of expressing this nature and that was developed for dealing with other things.By alerting us to the limitations of his language and imagery, Parmenides prescribes for us general vigilance when it comes to the business of using language as a tool with which to work our way towards an understanding of what-is.By deploying some descriptions of what-is that are very precisely dissonant with its nature and could not be taken literally, Parmenides 88 And, in the special case of the sphere analogy, by comparing what-is to an item of everyday sensory experience: the bulk of a ball.
89 This is the attested meaning of the word ἀδαής (for example, Pind.Pae.4.27) and would have been prominent for a contemporary audience.It fits with Theophrastus' testimony regarding the cognitive superiority of the other element (A46), cf.Coxon (2009) 348; Tor (2017) 249 with n.69.In context, it is tempting to allow 'without light' (vel sim.) as a further possible sense, but I know of no parallels (cf.DK ad loc., who hypothesize a connection with δάος, 'torch': 'lichtlose . . .nur hier in dieser Bedeutung (?)').In Tor (2017) 185-86 with n.55,I rendered ἀδαῆ 'dark' and noted that the term also indicates ignorance, but I would now translate it as 'unknowing' or 'unintelligent', while allowing darkness as a possible further connotation. 90For the echoes of Xenophanes' Greatest God in Parmenides, see Bryan (2012) 97-100.For negative theological characterizations in Xenophanes, see B1. 21-24; 11-12; 14-16; 18; 23; 25-26.  9I come close here to the view of Morgan (2000) 81-87, that Parmenides' mythological imagery is emblematic of language's general inability to express reality.Morgan, however, relies on some general assumptions about philosophical hostility to myth (pp.47, 84; cf.16-17, 34-35, 290-91) and does not pursue the particular tensions between Reality's imagery and doctrine explored above.Since Morgan assumes that, for Parmenides, 'only being exists' (p.82), she takes it, in a way reminiscent of Plato's Eleatic Stranger, that any mention of anything other than Being already violates Reality's ontology (especially pp.80-84).impels us to ask what we can take away from any discursive account of what-is unreservedly and what we must keep at arm's length. 92one of these conclusions denies that the goddess offers a description of what-is.It is difficult to see how else we might classify remarks like 'Being is ungenerated and imperishable' (ἀγένητον ἐὸν καὶ ἀνώλεθρόν ἐστιν, B8.3) or 'it is all alike' (πᾶν ἐστιν ὁμοῖον, B8.22) and, in general, the goddess' extended delineation of the characteristics of Being in B8. 93 Parmenides does, however, evince and encourage an awareness that this description proceeds through expressive measures that are imported with difficulty from a very different domain and are, consequently, limited, indirect and often figurative.We can usefully contrast here the unreserved and committed attitude that wayward mortals take towards their 'names'.They establish names like 'coming-to-be and perishing', 'changing place' and 'exchanging bright colour' and see these as categories that capture the nature of reality truly and faithfully (πεποιθότες εἶναι ἀληθῆ, B8.38-41).That mortals' names represent something stronger and more committal than the mere labelling of an item is again clear when the goddess diagnoses their decision to 'name' Light and Night as their cardinal error (B8.53-54; cf.B19). 94Parmenides' attitude towards the vocabulary that underpins his account in Reality, and which includes prominently and throughout terms like 'becoming', 'perishing', 'motion' and 'change', must be viewed in relation to his overt critiques of mortal names and especially of mortals' blithe acquiescence and misplaced trust in the presumption that these names map onto the nature of reality faithfully.The goddess, by marked contrast, nowhere expresses comparable commitment to the different linguistic terms and categories that make up her account of what-is in Reality.On the contrary, the attitude of the mortals contrasts with her more guarded, elusive and indirect framing of what she proffers in Reality as 'signs' (B8.2-3). 9592 A note of caution is in order.By recognizing that we should not interpret literally some or all of what we are inclined to describe as mythological imagery in Parmenides, we need not thereby exclude religious import in such imagery or, for example, rule out that Parmenides makes real claims for divine revelation.On this point, cf.Gernet  (1981) 354; Tor (2017) 162; Bernabé (2019a) 68-69. 93For the denial that B8 offers a description of what-is, see Mason (1988); Robbiano (2016) 275-82; (2018).Robbiano interprets being as 'the fact of being' rather than an object or entity.Whatever view we hold of the subjectless 'is' in B2, however, by B8 we encounter 'what-is' or 'being' ((τὸ) ἐόν) as a grammatical subject to which the goddess ascribes properties.Even if, as Robbiano (2016) 280-81 emphasizes, B8 is preoccupied with rejecting would-be ascriptions of certain predicates to what-is ('generated', etc.), the goddess shows no hesitation about consequently affirming for what-is the contradictories of those predicates ('not-generated').Robbiano's interpretation could work in principle, then, only if we could coherently construe 'the fact of being' as the bearer of all those properties (ungenerated, imperishable, all alike, indivisible, unmoving, spherical or spherical-like, etc.).The textual strain of her denial that B8 offers a description of what-is is visible when, at one juncture, Robbiano (2016) 288, cf.298, seems conversely to pursue just this route and to recognize the goddess' predications, writing in passing that the fact of being 'is without birth and death, continuous and without lack, as becomes clear in B8'.I leave aside Robbiano's other fascinating central thesis, that what-is is identical with our experience and awareness of it and is not 'distinct from a subject observing it' or 'searching' for it (Robbiano (2016) 263, 270, 291 and passim).I note only that this further claim faces the difficulty that some things are always true of what-is (for example, it does not move or change) but sometimes false of subjects who search for it (such as the kouros or his mind).
94 See further Barrett (2004) 279-87;Gianvittorio (2013) 17-19;Macé (2019) 49-51. 95 As discussed in section III: text with nn.53-55.See there also on the possible exception of the term (or 'name'?) τὸ ἐόν: text with n.77.Although mortals too employ their own 'signs' in their cosmology, these are differentiated from the goddess' signs in Reality.Most pertinently, the cosmological signs of the mortals are always subordinated to those committal 'names' that they trust unreservedly: . . .ὀνομάζειν . . .καὶ σήματ' ἔθεντο . . .(B8.53-55), ὄνομ' . . .ἐπίσημον (B19.3).Furthermore, while the goddess' signs are presented as waiting there to be discovered on the road of what-is (ταύτῃ δ᾽ ἐπὶ σήματ' ἔασι, B8.2), Parmenides emphasizes the conventional status of mortal names and their attendant signs: κατέθεντο (B8.38-39, 53-55), κατέθεντ' ἐπίσημον (B19.3); on this point, see Coxon (2009) 334, 343-44, 387 (on ἐπίσημον as 'a metaphor from coinage'); cf.Nussbaum (1979) 73-75;  Kraus (1987)  Can we, then, frame those aspects of Parmenides' language that conflict with his ontology as a ladder he climbs and then throws away? 96There are two divergent problems with this metaphor.On the one hand, if the view defended here concerning what should and should not count as genuine conflict between language and doctrine in Reality is on the right lines, then the ladder metaphor overstates Parmenides' argumentative reliance on dissonant language.Moira and the other anthropomorphized shackling goddesses play an important role in conceptualizing the appropriateness and necessity of the nature of whatis; but their role is throughout no more or less than that of a figurative conceptual resource: the argumentation at no stage requires us to treat this imagery of external shackling (or the other discordant imagery we have identified) literally or at face value. 97e also saw that this imagery is more than once introduced to offer a new way of representing and thinking about properties of what-is that Parmenides had already established or affirmed without it.On the other hand, the ladder metaphor understates our inexorable dependence on a conceptual framework that can only make up a description of what-is indirectly and with difficulty and that we could never discard.The goddess enables the youth, and Parmenides enables us, to grasp and understand (νοεῖν) what what-is is like only by deploying, and denying to what-is, such concepts as generation, perishing, internal divisions, change and mobility, and indeed by resorting further to mythological language that cannot be taken at face value.There is no suggestion in the poem of some further stage of initiation in which we will be able simply to dispense with this conceptual scheme and review or retrace the nature of Being without it.Where the ladder metaphor does point us in the right direction is in its suggestion that the value of human language to ontological inquiry is strictly heuristic.Our language is useful to us in this context, not insofar as it can directly articulate the nature of Being as it does with such things as stars or humans (it cannot), but as a tool with which to work towards an apprehension of the nature of Being and to place ourselves in the appropriate frame of mind to contemplate it. 98he Parmenides who emerged above anticipates to a meaningful but still partial extent the later Platonic idea that the One is ineffable.Plotinus writes that we cannot ascribe any predicates to the One, since this would turn it into a multiplicity of subject and predicates (Enn.5.3.10).Consequently, we cannot in fact say anything of the One but only use language as an imperfect heuristic device in relation to it, or: 'make signs to ourselves' about it (ἡμῖν αὐτοῖς σημαίνειν, Enn. 5. 3.13.5).We largely develop some awareness of the One by differentiating it from the world of complexity and multiplicity that, in Plotinus' scheme, comes after the One and is causally reliant on it; that is, we develop an awareness of the One through negation (Enn.5.3.14).Plotinus emphasizes that Platonists speak about the One as they do because it is impossible to speak about it as they wish.He cautions that the qualification 'as if' should be assumed to apply to every statement about the One (Enn.6.8.13; 6.8.18.52-53).He repeatedly helps himself to metaphors and images (such as the famous comparison of the One and the world order dependent on it to a king's triumphal procession), which inevitably involve both illuminating points of contact and 96 So Owen (1960) 100; (1966) 321-22, repurposing the image of the ladder from Sextus (Math.8.481) and Wittgenstein (Tractatus 6.54); followed by, for example, Graeser (1977b) 365; Mackenzie (2016)  43-44.Owen had in mind Parmenides' use of negations and of temporal and spatial distinctions (relying on interpretive assumptions criticized in section II above).
97 Mourelatos (2008) 161 underscores that the shackling goddesses have the status of figurative imagery and are not 'an element of the ontology'; cf.Bryan (2020b) 88-90. 98 For different elaborations of the view that, for Parmenides, language has only heuristic value in the context of ontological inquiry, see Owen (1960)  100; Mackenzie (2016) 43; (2021) 69-70; Robbiano (2016) 275-82; (2018);  Sattler (2020) 108-09.Language and doctrine in Parmenides misleading discontinuities between the source domain of the metaphor and the target domain of the One (Enn.5.5.3;6.8.9). 99o be sure, there are notable divergences between Parmenides and Plotinus on this score.First, Parmenides does not seem to regard what-is as wholly ineffable.If nothing else, we can assert seemingly without reservation that what-is is.Second, Parmenides remains adamant that what-is can be the object of intellectual apprehension (νοεῖν), whereas for Plotinus the One cannot even be an object of thought and understanding but only somehow received through a kind of 'touching', free of speech or thought (Enn.5.3.10.41-42).A third and related difference is that Parmenides nowhere evinces a similarly exacting notion of radical simplicity that precludes even the having of a conceptual structure nor, therefore, any discomfort with thinking of Being as having properties.Indeed, it is Plotinus' distinctive refusal to think of the One as a propertybearing entity (a 'this') that leads him to describe it as 'beyond being' (Enn.5.5.6;cf.5.1.10.2;6.8.9.27-28).
These divergences notwithstanding, Parmenides too evinces the awareness that what-is resists unproblematic or direct expression through language and that, in the context of ontological inquiry, language is thus useful to us only as an imperfect heuristic device with which to work our way towards an apprehension of what-is and its properties, indirectly and with difficulty.Parmenides can fairly be described as a forerunner to the later Platonic idea that the key philosophical and metaphysical principle is something we cannot quite put into the only words at our disposal.
is not' (οὔτ' ἐκ μὴ ἐόντος ἐάσσω | φάσθαι σ᾽ οὐδὲ νοεῖν· οὐ γὰρ φατὸν οὐδὲ νοητόν | ἔστιν ὅπως οὐκ ἔστι, B8.7-9)?This remark must be read in the context of what immediately precedes, where the goddess asks from what we might maintain that what-is emerged into being (πῇ πόθεν, B8.6-7). 41What the goddess prohibits here is not using the expression 'what-is-not', but denying the existence of whatis, asserting the existence of what-is-not or, consequently, entertaining what-is-not as a cosmological origin for what-is.