Introduction
This paper addresses Parmenides’ highly controversial account of ‘names’. In this regard, two main interpretative approaches are to be distinguished. According to the first, Parmenides affirms that a ‘name’ – indeed, any linguistic expression whatsoever – must refer in order to be significant so that if a referring expression fails to denote, not only is it deceptive, but also meaningless. Briefly, proponents of this reading – such as Cornford, Tarán, Furth and others – emphasise the several warnings by the goddess (the ‘speaker’ in the Poem is not Parmenides himself, but a ‘goddess’ who actually teaches Parmenides the truth) not to speak of what-is-not which is, after all, inconceivable, unspeakable and also unnameable (B 8. 17). In this view, those expressions mortals employ to (attempt to) refer to what-is are both deceptive and meaningless, for they actually fail to denote: in claiming that what-is is changeable, mortals end up speaking of what-is-not (given that reality is not changeable).
The second approach, on the contrary, rejects this conclusion, for its advocates – such as Woodbury, Mourelatos, Vlastos, Palmer and Tor among others – highlight that there is no evidence at all to conclude that ‘mortals’ names’ (i.e. those which mortals erroneously use to refer to what-is) are taken as ‘empty names’ or ‘mere noise’, for such names do refer to what-is, although erroneously (that is, they take reality to be different to how it actually is). Therefore, they argue, whereas it is clear that only those referring expressions that correctly refer to what-is are true, still, those which do not, although false or deceptive, are obviously communicable and meaningful.
In other words, Parmenides either committed himself to a (broadly speaking) Fregean semantics, distinguishing ‘naming’ and ‘meaning’ – so that the meaning of a name is to be distinguished from the object it denotes – or a non-Fregean semantics, i.e. a referential account of meaning, according to which the meaning of a term is its referent.
This makes it clear that the issue of ‘names’, although very specific, is not a minor point. Rather, it is fundamental since – taking into account Parmenides’ semantics – it has a huge impact on both his philosophy of language and epistemology. In fact, ‘falsehood’, ‘not being’ and ‘meaning’ all depend on how this account is understood.
It might be convenient to make explicit that this paper aims to show that both the ‘Fregean’ and the ‘non-Fregean’ readings are not only legitimate but also correct. This is the line of reasoning: the goddess constantly insists on the impossibility, or the illegitimacy, of expressing what-is-not linguistically, thus implying that a sentence referring to what-is-not is not only false, but also meaningless. Nonetheless, it seems clear that, even being erroneously committed to the changeable nature of reality, and speaking accordingly, mortals do say something (something comprehensible and communicable): in other words, although mortals either speak of what-is-not or speak falsely of what-is (8. 38–41), however, i.e. in both cases, they do say something understandableFootnote 1.
Since this ambiguity cannot be ignored or solved, but just explained, my suggestion is that something more – something that does not attempt to solve the puzzle only from a linguistic point of view – is required. More precisely, my point is that taking into account Parmenides’ epistemology, it should become clear that the possibility of identifying in the Poem two different semantics, is not an unpleasant consequence of the Eleatic system, but, on the contrary, its coherent result.
The paper is divided into three main sections: the first part is devoted to an analysis of the status quaestionis concerning mortals’ names in the Poem; in the second half of the paper a disjunctivist reading of Parmenides’ epistemology will be provided; in the final section, the main conclusions will be drawn.
1. B 8. 38–41
1.1 Textual remarks
Even though an explicit account of names is not, of course, developed in the Poem, it is easily possible to identify those passages where the issue is fully addressed. Since the most famous (and controversial) is B 8. 38–41, the first section of this paper is mainly devoted to it.
The main issue is, unsurprisingly, philological. Broadly speaking, two different textual constructs – and, accordingly, two different interpretations – are possible: the first was promoted by Diels-Kranz; the second was first advanced by Leonard Woodbury (Reference Woodbury1958).
According to the traditional interpretation (i.e. that follows Diels-Kranz textual construct), B 8. 38–41 states:
While lines 40–41 are quite unproblematic, there is a substantive disagreement concerning B 8. 38–39. In particular, on 8. 38 Diels reads tōi pant'onom’estai, whereas Woodbury (Reference Woodbury1958) claims that an alternative construction must be preferred, namely tōi pant'onomastai.
Coxon, defending Diels’ construal, translates 8. 38 (tōi pant'onom’estai) as follows: ‘therefore all those things will be a name’. His translation followed that previously provided by Cornford ‘therefore all those (names) will be a mere word’ (Reference Cornford1939, p. 34). The main problem with those renderings is clearly recognised by Vlatos who emphasised that Cornford ‘in order to express what he takes to be the sense of the Diels text, has had to force the translation: “mere word”, as though Parmenides has written psilon onoma, or, onoma monon, or something else of this kind’ (Vlastos, Reference Vlatos and Mourelatos2008, p. 370).
Actually, this is the reason behind Woodbury's alternative reconstruction of 8. 38, i.e., as mentioned above, tōi pant'onomastai. Accordingly, the translation is the following: ‘With reference to it (the real world) are all the names given’ (Woodbury, Reference Woodbury1958, p. 149). Since then, many scholars, such as Vlastos, the Mourelatos of the first edition of The Route (1970)Footnote 2, SchofieldFootnote 3 and BurnyeatFootnote 4, actually followed the path opened by Woodbury.
This is not in fact a minor point for, according to Coxon and Cornford, Parmenides would be claiming that mortals’ names are empty names; conversely, according to Woodbury and the others, Parmenides would be explaining that all the names mortals are actually employing cannot but refer to Being, so that they are definitely not empty.
As I mentioned, although both readings have been variously defended both philologically and philosophically, most of scholarsFootnote 5 have eventually followed Woodbury's correction. As for the philological issue, in favour of the tōi pant'onomastai (‘all things have been named’) reading there are three main arguments: 1) as Woodbury remarked (Reference Woodbury1958, p. 147), fragment 8 is quoted in the Simplicius commentary to Aristotle's Physics where both of the alternatives tōi pant'onomastai and tōi pant'onom’estai are attested. In 146, 7–14 Simplicius reported the full text with the variant tōi pant'onomastai, while the construct adopted by Diels, tōi pant'onom’estai, is quoted by Simplicius in a much shorter passage, i.e. 87. 1. According to Woodbury (Reference Woodbury1958, p. 147) and Vlastos (Reference Vlatos and Mourelatos2008, pp. 368-370), the fact that in the longer citation Simplicius wrote tōi pant'onomastai, seems to imply that this is the correct variant. The correction tōi pant'onom’estai in the longer quotation is, in fact, attested to only in the Aldine edition but this is easy to explain, according to Vlastos, by considering that the editor probably corrected the passage at pp. 145–146 of Simplicius’ commentary in accordance with the previous passage (p. 87). 2) According to the advocate of tōi pant'onomastai (‘all things have been named’) that reading is further confirmed by B 9. 1 where it is given the form ‘onomastai’ Footnote 6. 3) As Mourelatos (Reference Mourelatos2008, p. 182) in the first edition of The Route emphasised, the present perfect of the form onomastai would be fully coherent with other present perfect forms in the previous (8.35) and following (8.39 and 8.42) passages.
Since, though not conclusive, those arguments seem to have persuaded most scholars, I will not pursue the matter further, thus confining myself to more strictly philosophical remarks. Indeed, Woodbury opened his article mentioning the unpleasant philosophical consequences to which it seems we are committed in accepting Diels’ reconstruction. The next section will therefore be devoted to the philosophical debate.
1.2 B 8. 38: Parmenides on ‘falsehood’ and ‘meaninglessness’
As I mentioned previously, Vlastos’ main objection to both Diels’ construct and Cornford's translation was that they committed the text to something that is not actually required or even legitimate – i.e. to ascribe an implicit ‘mere’ or ‘only’ (monon) referred to ‘names’ (onomata) – so that the sense would be precisely that all the things men established believing them true are, in fact, ‘just’ or ‘mere’ names.
The main reason for endorsing Diels’ reconstruction is that since the Poem's main point is that what-is-not is unspeakable and unnameable (B 8. 17) – so that only what-is can be named and spoken of –, it follows that the mortals’ names mentioned in B 8. 38 are not – cannot be – in the end, real names. As Furth (Reference Furth1968, p. 121) famously explained:
It is, put one way, that for singular and general terms, meaningfulness is identical with the having of denotation; put another way, that a singular or general term can possess meaning of just one kind, which Parmenides not unnaturally takes to be, that in reality which - if it exists - the term denotes. (…) Because if a singular or general term lacks denotation, “corresponds to (= denotes) nothing in reality”, then it is taken to be an empty name with no meaning at all. This assumption, I believe, is the real root of the matter, the source of the whole doctrine, to which everything else is ancillary.
In this respect, Furth further developed a view that had already been established precisely by Cornford (Reference Cornford1939, p. 34):
all these terms [scil. mortal names] are dismissed as empty names which are meaningless, since they do not apply to what is, and there is nothing else for them to mean.
This kind of reading – very similar to what Mourelatos (Reference Mourelatos1979) called the Standard Interpretation – aims at accounting for what is usually emphasised as the main thesis of the Poem: what-is-not must not be either named or the subject of a declarative sentence, because it is not possible to speak of ‘nothing’. This opens the path to the conclusion that there is no room for some kind of distinction between ‘meaningless’ and ‘false’ nor between ‘meaningful’ and ‘true’Footnote 7. As Mourelatos (Reference Mourelatos1979, p. 4) puts it:
the rejection of the negative route is still based on the assumption that expressions are meaningful only if they refer to an actual entity, in this instance, an actual state-of-affairs, a fact.
Accordingly, Parmenides’ reasoning would be the following: the goddess has claimed (in the previous lines: 8. 36–38) that there is nothing apart from Being, and therefore when mortals speak of ‘becoming’ and ‘perishing’ they actually speak of nothing, i.e. purporting to refer to what-is-not, mortals’ names are meaningless. Conversely, according to those who favour the alternative construal, the correct reading is that, since the goddess claims that there is nothing apart from Being, it must be concluded that mortals’ names cannot but refer to Being itself, even though they do so erroneously (Woodbury, Reference Woodbury1958, p. 149).
Following Woodbury, Vlastos strongly maintained that nowhere in the Poem is it actually suggested that mortals’ names are mere noise: on the contrary, they, even though false or deceptive, still denote something-that-is. After all, as Vlastos (Reference Vlatos and Mourelatos2008, p. 374) claims, if mortals’ names were mere noise it would not be clear how it is that they misdescribe what-is: indeed, ‘a meaningless noise could no more misdescribe than describe’. Along the same line, Mourelatos famously developed Woodbury's basic assumption claiming that
I gave B8.38-41 the paraphrase: Mortals call it Y, but it actually is X. This is normally how we should understand the passage (…). For as we saw, “to name X (as) Y” in Greek does not mean to attach the sound “Y” to X, but rather to describe X in terms of “Y”, which is thus assumed to already have meaning independently of its naming relationship to X. So the relevant categories for interpreting this passage are not meaningfulness versus meaninglessness but truth versus falsehood (Mourelatos Reference Mourelatos2008, p. 189)Footnote 8.
To sum up, according to the latter approach, the mortals’ names mentioned in 8. 38–41 – misdescribing what-is – are false, but the fact that they are false does not straightforwardly entail that they are also meaningless (i.e. empty): they do refer to what-is, although erroneouslyFootnote 9. Conversely, according to the former approach, mortals either fail to describe what-is, and therefore they actually attempt to speak of what-is-not, or, more radically, do speak of what-is, but speaking falsely of it, they actually name what-is-notFootnote 10; because it is clear that it is not possible to speak of what-is-not, it must be concluded that mortal names are impossible, i.e. meaningless.
The main point here is that whatever textual reconstruction we recognise to be the most convincing, it seems plain to affirm that it would be quite eccentric if Parmenides had considered each and every human speech to be ‘mere noise’; after all, is not Parmenides human?Footnote 11 However, if we were committed to the meaninglessness of false names, it is not clear how this conclusion could be escaped.
Moreover, it is evident that nothing in Diels’ construction makes it necessary to read ‘mere’ or ‘just’ names so that, even though we accept the construction tōi pant'onom’estai (‘all things will be its name’), it does not seem that we are automatically committed to this integration as well. This is rightly remarked upon by Shaul Tor who in fact adopts Diels’ text:
It makes no odds to us whether at B8.38 we read τῳ πάντ’ὄνομ(α) ἔσται (to be translated ‘all things will be its name’, with Kingsley Reference Kingsley2003, 190-1; Mourelatos Reference Mourelatos2008, 386-7) or, less likely, τῷ πάντ’ ὀνόμασται (‘to it all things will be named’, Woodbury 1971). On either construal, the point will be that what-is is in fact the only possible referent for categories which purport to get at and capture the fundamental nature of reality (Tor, Reference Tor2015, note 21, p.58, emphasis mine).
All in all, once we reject the assumption that, in order to make sense of the passage, some integration is required, we realize that neither of the texts are actually accounting for the emptiness of names. Therefore, it seems that proponents of the ‘false-but-meaningful’ reading of mortal names are absolutely correct. So far so good. However, many questions still remain unanswered.
Firstly, it still seems necessary to conclude that onomata (‘names’) are always and essentially deceptive. Secondly, and more substantially, it still needs to be explained how it is that the account of ‘mortal names’ is not inconsistent with the assumption that what-is-not is unspeakable and unnameable (B 8. 17). In this regard, it seems that proponents of the traditional reading must be surely right: although Vlastos fails to recognise it, the distinction between ‘falsehood’ and ‘meaninglessness’ leaves us with a strong inconsistency with what is claimed in the Way of truthFootnote 12; in other words: while the goddess evidently seems to be promoting a referentialist account of language (according to which the meaning of a name is the item it picks out in the world), Parmenides’ remarks on mortals speaking of change and generation, on the contrary, seem to suggest that even though false, mortals’ names are nonetheless meaningful. If this is the case, it seems that we are to face a contradiction: while ‘goddess language’ is committed to a referential (i.e. non-Fregean) semantics, ‘mortal names’ take into account a semantics which distinguishes ‘naming’ and ‘meaning’ (i.e Fregean).
As I already highlighted, this dichotomy can hardly be avoided. My further suggestion is that it does not need to be. In other words, the starting point is to reject the assumption that Parmenides’ main concern is, à la Russell, to establish that propositions must denote if they are to be meaningfulFootnote 13. Indeed, it is the whole assumption that Parmenides’ main concern is rooted in something like a ‘philosophy of language’ – that Parmenides is to provide an account of language that is independent from the epistemological concern – that must be abandoned.
To be clear, the question is not: ‘how is the puzzle of the two semantics to be solved?’. Rather, the question is: ‘how is this puzzle to be explained?’, i.e. ‘as goddess language and mortal names are different, why, and in virtue of what, are they essentially different?’.
To provide an answer, as I briefly noted in the introduction, the epistemological background of the Poem must be taken into account.
This strategy should allow us to explain why it seems that ‘mortal naming’ is necessarily deceptive by showing that it is so due to an epistemological error. My proposal is that since Parmenides endorses an epistemological disjunctivism – according to which the truthful and faithful cognitive state is essentially heterogeneous from the deceptive and false cognitive state – he is further committed to a ‘disjunctivist semantics’, according to which the true speech, that expresses truth linguistically, is essentially different from the false one. In other words, just as ‘knowing’ (or ‘thinking veridically’) and ‘mistaking’ (or ‘thinking erroneously’) are two crucially distinct (cognitive) acts, so ‘speaking truly’ and ‘speaking falsely’ are two crucially distinct (linguistic) acts.
In view of this, as I mentioned, the possibility of identifying in the Poem two different semantics, i.e. referentialist for the ‘goddess language’ and Fregean for the ‘mortal language’, is not an unpleasant consequence of the Eleatic system; it is, instead, its coherent result.
2. The epistemological background
Many scholars, such as Calvo (Reference Calvo1977, p. 259), Coxon (Reference Coxon2009, p. 334) and Cordero (Reference Cordero2010, p. 239), among others, rightly highlighted that the reason behind the falsehood or deceptiveness of names is that they are ‘established’ or ‘instituted by’ men (8. 38–41; 8. 53–55; 9. 1). I would say that the conventionalist nature of mortal naming is of such a significance that it can hardly be overestimated. Indeed, it plainly suggests that it is not the name as linguistic expression, but the act of naming as cognitive act that entails the possibility of attaching names to things erroneously, i.e. the possibility of addressing reality falsely. In other words, it is exactly the conventionalist nature itself that seems to presume – and to be the consequence of – some kind of judgment in virtue of which mortals do name reality.
This seems to arise quite clearly in 8. 53 where mortals are said, according to Rose Cherubin (Reference Cherubin2005, pp. 1–2), to ‘have established (katethento) two viewpoints or opinions (gnomas) in order to (accordingly) name (onomazein) forms (morphas)’ or, according to Coxon and Tor among others, to ‘have decided (katethento… gnomas)Footnote 14 to name (onomazein) two forms (morphas)'Footnote 15.
For present purposes, establishing the correct rendering is not fundamental, for these two readings – as remarked by Mourelatos – agree on the main point: that is, the reason why these two (false) names are actually attached or bestowed is that humans are first and foremost mistaken about that reality they want to discover, i.e. it is the cognitive error that coherently leads to naming falsely Footnote 16.
This is highlighted by Vlastos who explicitly recognises the epistemological error to which the linguistic error is to be reduced (Reference Vlatos and Mourelatos2008, p. 375):
The decision to name the two forms in B8.53 is, of course, a dreadful mistake. But, first of all, the mistake is the decision that these two forms exist – not the use of the name “Light” and “Night” in the expression of this decisionFootnote 17.
Mortals name two things (9. 1) because they (falsely) believe that they had discovered two principles (8. 53–55). That being so, it is the reason why mortals name erroneously that must be investigated and accounted for. In other words, the main question is: what is Parmenides’ account of (epistemological) error?
More specifically and more substantially, the latter question can be thus reformulated: what is Parmenides’ account of the relation between mind and world? That is, what is Parmenides’ account of noein (the meaning of this verb is discussed presently) and of its relation with einai (being)?
Briefly speaking, it seems that regarding the nature of noein and the relation between einai and noein in Parmenides’ Poem two readings are possible. The first – following Owen (Reference Owen1960) more or less explicitly – takes noein as meaning ‘thinking’, and thus maintains that the Poem's concern is stating the coextension of ‘being’ and ‘thinkable’, i.e. stating that thought must be thought of something that exists. The second, originally advocated by Charles Kahn (Reference Kahn1969), basically understands noein as epistemologically connoted and thus as meaning ‘veridical thinking’ or ‘knowing’Footnote 18: accordingly, Parmenides would be arguing that whatever the object of knowledge is, it must exist.
In this latter view, Parmenides is usually (though not necessarily) considered an ‘identity theorist’, inasmuch as the Poem's main statement is that veridical thinking is identical with being; or, in other (perhaps anachronistic) words, that truth-bearers and truth-makers coincide. To be very clear, I would say that such an Identity Thesis is, therefore, to be understood as an Identity Theory of Truth.
In this view, the famous (and controversial) fragments B 3, B 6. 1–2, B 8. 34–36 literally affirm that being and noein are strictly identical and, because noein is understood not as general thinking but as ‘grasping’ or ‘knowing’ or ‘veridical thinking’, this reading is not committed to attributing to Parmenides either some kind of idealism or some kind of neo-platonic theoryFootnote 19. Besides, since ‘identity’ is a symmetrical relation, it is fair to conclude that any account of coincidence between truth-makers and truth-bearers – if it really argues for ‘coincidence’ – actually escapes the realism-idealism dichotomy: Parmenides is, strictly speaking, neither a realist nor an idealistFootnote 20.
Understanding the Identity Thesis as an Identity Theory of Truth further implies that the cognitive grasp through which noos (‘mind’) successfully grasps reality is direct and unmediated. As such, the main epistemological thesis is the following: in the true ‘grasping’ of reality there is no distinction between the knowing subject and the object of knowledge, in the sense that there is no room for any representational medium whatsoever. When I grasp the world truly there is nothing between me and the world. In view of this, the Eleatic Identity Thesis entails some form of ‘Direct Naïve Realism’Footnote 21.
Nonetheless, it could be argued, assuming that Parmenides is endorsing such an epistemology – an epistemology according to which noos (‘mind’) is directly and immediately connected to reality and, therefore, truth-bearers and truth-makers coincide – how is falsehood to be understood? And further, how is it that this picture says something relevant concerning false names?
Regarding falsehood, the key expression here is true grasping. ‘Direct Naïve Realism’ is usually associated with – in fact, it seems to entail – ‘Disjunctivism’ (whereas disjunctivists are not necessarily committed to Direct Naïve Realism)Footnote 22.
As is well-known, ‘Disjunctivism’ can be variously understood, however the basic idea is that veridical cognitive states are constitutively different from illusory or hallucinatory states, such that ‘bad cases’ and ‘good cases’ have nothing (metaphysical) in common. Briefly speaking, and somewhat simplifying the matter, one basic way of accounting for the distinction between ‘good cases’ and ‘bad cases’ is emphasising that while the former are both ontologically and phenomenologically object-dependent the latter are not – so that veridical states and deceptive states are both constitutively and qualitatively different. As McDowell famously put it (Reference McDowell1983, p. 472), it is only in ‘good cases’ that the fact itself is disclosed to the experiencer Footnote 23. We might therefore put it as follows: while in veridical states reality is presented to the subject (in the sense that objects are instantiated in, or are the constituents of, veridical experiences), in deceptive states it is merely (and erroneously) re-presented Footnote 24.
Indeed, the fact that when I truly grasp reality – i.e. when I am in direct and unmediated connection with reality – I cannot doubt that what I grasp is true (in other words: I cannot doubt that what I grasp is what is the case) does not mean that I am always successful in grasping reality.
‘Falsehood’ is therefore explained as a cognitive error whose source is explicitly identified by Parmenides with mortals’ noos (‘mind’): indeed, noos is the source of knowledge (B 3; B 6. 1–2; B 8. 34–36) and the source of error as well (B6. 4–6); it is the organ by which the knowing subject attempts to grasp reality and therefore it can go right or wrong. In fact, Parmenides and humans are repeatedly urged by the goddess not to follow the way of not-being, thus meaning that their mind can, in principle, follow both truth and falsehood.
In this regard, I am indebted to Calvo (Reference Calvo1977, p. 251) who highlighted that
infallibility has been also considered by some scholars as one of the specific features of νόος. It becomes by no means easy to bring this feature into line with fr. 6, 5—6, where the goddess refers to a πλακτόν νόον, a “wandering νόος”. K. von Fritz has emphasized this paradox calling it insoluble (…). But this problem does not exist if we accept the above Interpretation according to which νόος in this passage means the principle of cognition in general and not any particular and supposedly infallible faculty of knowledge.
Mortals name two principles because they believe they grasped – they discovered – two principles. In fact, they failed to correctly grasp reality (that is, their illusory experience leads them to take reality as essentially changeable), and therefore they speak of it erroneously. Names (it must be remembered that, according to Parmenides, onomata indicates both ‘name’ and ‘predicates’ and so any linguistic expression whatsoever) are not deceptive on their own, but they become so in the mouth of ‘double-headed’ mortals ‘who do not know anything’ (B 6. 4-6).
It should be now sufficiently clear that, in Calvo's proposal and my own, the main point is the treatment of noein (meaning ‘knowing’ or ‘grasping’ or ‘understanding’) as a success-verb. Mortal noos can successfully grasp reality, but it may happen that it is unable to do so. When human minds grasp reality this ‘grasping’ must be understood in terms of an unmediated contact with the world outside (i.e. as a ‘presentation’ or ‘disclosure’ of world itself); when they do not correctly grasp reality the result is an erroneous re-presentation – i.e. an illusory or hallucinatory re-presentation – of reality.
At this point, it has surely been noted that such an epistemological account quite evidently resembles a perceptual account of knowledge. In fact, as long as the activity expressed by noein is understood as ‘direct and unmediated comprehension’ of true reality, it clearly represents features that are more easily recognised as belonging to ‘perception’ rather than to ‘propositional thinking’ or ‘(intellectual) reasoning’. However, it cannot be (and must not be) reduced to mere sensation, either. It might be perhaps taken as ‘acquaintance’ or ‘awareness’, emphasising its non-inferential nature. Besides, it is not so obvious that Parmenides actually marked the distinction between ‘reasoning’ and ‘perception’, as Mansfeld (Reference Mansfeld1999, p. 343) famously observed:
On ne peut pas nier, naturellement, que ces deux philosophes presocratiques (…) avaient une notion intuitive plus ou moins implicite ou explicite de la difference entre perception et connaissance, ou du moins entre l'usage des yeux et des oreilles et celui de la reflexion et de l'esprit ou la pensee. Mais, dans l'experience de tous les jours, la faculte reflexive et les organes de la perception fonctionnent comme un ensemble. (…) C'est seulement retrospectivement, et a condition d'etre pourvu d'un outillage conceptuel anachronique, qu'on peut supposer, chez Parmenide et Heraclite, un commencement embryonnaire de la distinction entre perception et pensee acceptee aujourd'hui.
So far I mentioned those fragments where the twofold nature of mortal noos is fully addressed. What about B 16Footnote 25? It states the followingFootnote 26:
Overall, the fragment has been variously interpreted and there has been disagreement even on its correct placement in the PoemFootnote 30, however, for present purposes, it is sufficient to highlight that, in whatever interpretation, a physiological account of cognition or knowledge is provided, i.e. Parmenides explains that human minds have a bodily origin and are thus influenced accordinglyFootnote 31.
Indeed, I would say that the twofold nature of noein and noos, as it has been sketched so far, far from being dismissed is actually confirmed by B 16. Parmenides seems to recall here the ‘wandering state of noos’ he had already mentioned in B 6, further explaining that human cognition is strongly influenced by its physiological basis – as the (Homeric) correlation ‘os… tos…’ clearly suggests (Mourelatos, Reference Mourelatos2008, p. 256). That is, it seems that Parmenides is here providing an explanation of why noos is not always and not necessarily rightly connected to reality, i.e. of why, as aforementioned, noos is not infallible.
Overall, it seems that, once we accept that ‘knowing’ is to be understood as ‘grasping’, i.e. in terms of a direct and unmediated contact with reality, the ‘perceptual’ description of noos provided in B 16 is no more problematicFootnote 32. Nor is the skeptical account of human cognition that might be supposed to arise from B 16 problematic: indeed, Tor recently argued that, according to this fragment, human cognition qua human is essentially unable to grasp true realityFootnote 33. In light of this, Tor's resolution is the following:
I argue that Parmenides construes the mortal agent as a complex which comprises not only a human but also a divine part – the mortal's (hot, aethereal) soul. The journey towards grasping Alētheia's krisis, (…) involves, for Parmenides, a divinisation of the knower, who comes to think with – or as – his divine soul, momentarily transcending the uncontrollable, Doxastic deliverances of specifically human cognition (Tor, Reference Tor2017, pp. 8–9)Footnote 34.
I would say that, as fascinating as it is, Tor's solution is demanding. Besides, if we take noos (i.e. human mind) as essentially capable of being both in the right and in the wrong contact with reality, it cannot be a surprise that the single noos is said sometimes to coincide with being and sometimes to be ‘wandering’ and departing from reality.
Indeed – this is a major point – if noos were not ‘wandering’ and ‘unstable’, how could error be accounted for? The fact that noos is fallible – so that it may happen that it fails to grasp reality – far from being problematic, is the key to understanding why not all of our beliefs, impressions (and so on and so forth) are the same. In other words, while B 16 clearly explains that human noos is (essentially or in each and every occasionFootnote 35) such that it can go wrong, it does not seem to necessarily entail that it is unable to go rightFootnote 36.
Moreover, if we read B 16 in light of B 6. 4–10 where the goddess speaks of the ‘double-headed mortals who do not know anything’, we see that the reason why humans erroneously bring together ‘it is’ and ‘it is not’ is not generally their ‘wandering mind’. More precisely, they are said to be ‘akrita phyla’, i.e. they have not distinguished (yet) ‘being’ and ‘not being’, so that the right contact with reality is, or rests on, the correct distinction (or on the correct act of distinguishing) between ‘it is’ and ‘it is not’ or, in other words, between ‘good cases’ and ‘bad cases’ (B 8. 15).
Does this capacity of distinguishing ‘being’ and ‘not being’ truly require a divinization, as both Tor and Curd (although differently) maintain? Does the fact that the ‘Way of Truth’ and ‘Opinion's account’ seem to presume two essentially different – and mutually exclusive – cognitive states, automatically imply such a demanding psycho-ontological dualism? I would say that the answer is ‘not necessarily’. That is, Tor's proposal is in fact compatible with my ownFootnote 37; my aim is just to underline that although we can further endorse such a view, we do not have to in order to account for what is, after all, an epistemological issue. I suggest a less radical and less demanding strategy than Tor's – a strategy that embraces a form of ‘epistemological disjunctivism’, according to which the truthful cognitive state is essentially different from the deceptive one.
3. Concluding remarks
To sum up, at the end of the first section I suggested that Parmenides seems to be committed to two different semantics: in maintaining that what-is-not is unspeakable and unnameable (B 8. 17), he argues for a referentialist semantics; in ascribing to mortals the capacity of erroneously naming reality, he makes room for some kind of distinction between ‘naming’ and ‘meaning’, thus taking a Fregean semantics into account.
In order to account for this ambiguity, which is actually unavoidable, I further suggested that the Poem's epistemological background had to be taken into account. More specifically, I proposed understanding Parmenides’ epistemology in terms of a ‘Direct Naïve Realism’, committed to ‘Disjunctivism’. The idea is that, for Parmenides, knowledge must be understood as a direct, unmediated (i.e. presentational) contact that allows for a strict identity between truth-bearer and truth-maker: on the face of it, ‘falsehood’ arises as a cognitive error which leads to an illusory re-presentation of reality.
In view of this, the fact that mortals do name erroneously has been defined as the linguistic consequence of a cognitive act. Since mortals failed in truly grasping reality, they failed in naming and speaking of it, i.e. since mortals falsely believed they could account for reality by means of Light and Night, they named them, accordingly.
At this point, this picture should allow us to understand the question of the ‘twofold semantics’.
Indeed, the reason why, as Furth put it, ‘for singular and general terms, meaningfulness is identical with the having of denotation’ is thus explicable: if grasping reality is understood as a presentation or disclosure of reality itself (‘Direct Naïve Realism’) – according to which there is no ontological gap between mind and world or between language and reality – it follows that a Fregean semantics is simply impossible, i.e. since it is necessary that the meaning of a name and what the name denotes are identical, a referential semantics must be endorsed. This explains why in the Way of Truth, specifically in B 8. 17, what-is-not is not only unspeakable, but also unnameable (referential semantics) even though in the account of Opinion mortals are repeatedly said to erroneously name reality (Fregean semantics).
As I mentioned, this does not seem to be just an unpleasant and unwanted consequence of Parmenides’ reasoning. On the contrary, Parmenides seems to be perfectly aware of the fact that ‘goddess language’ and ‘mortal language’ are essentially – semantically and syntacticallyFootnote 38 – different. Indeed, it seems that Parmenides’ main point is precisely to highlight that they cannot but be different.
