SOURCE: Richard D. McKirahan, Jr., "Parmenides and Elea," in Philosophy before Socrates: An Introduction with Texts and Commentary, Hackett Publishing Company, Inc., 1994, pp. 151-78.

[In the following excerpt, McKirahan clarifies some of the more difficult passages in Parmenides 's work and answers objections to the poem.]

Fragments

Significance and Life

Parmenides' philosophy marks a turning point in the history of thought. Neither his style of argument nor his astonishing conclusions could be overlooked even by those who strongly disagreed with him. Like Heraclitus, Parmenides pushed the limits of his thinking beyond the range of subjects found in the early Ionian philosophers, though his ideas, like those of Heraclitus, have implications for the entities and cosmic processes which his predecessors proposed. If Heraclitus was the first person to approach things philosophically instead of scientifically, Parmenides deserves recognition for introducing deductive arguments to philosophy and for acknowledging their compelling force, and for using this new tool to raise basic philosophical questions: What conditions must exist ing things satisfy? Is reality what our senses tell us it is? How we can tell? He was also the first to undertake explicit philosophical analyses of the concepts: being and coming to be, change, motion, time, and space. And he was the first to use these concepts to analyze the nature of a logical subject, and so in an important sense he is the inventor of metaphysics.

The best piece of information about Parmenides' life¹⁴ indicates that he was born c. 515 and lived until at least 450 or so. He was from the Greek city Elea in southern Italy, as was his follower Zeno, and their distinctive philosophical opinions as well as their philosophical method gave rise to the terms "Eleatic philosophy" and "the Eleatic school." Parmenides was sufficiently respected in Elea to have been asked to draw up a code of laws which were still referred to with respect and were probably still in force over five hundred years later.¹⁵

Parmenides' Poem

Parmenides' philosophy is in ways diametrically opposed to Heraclitus', whom he may have attacked in his writings.¹⁶ He is called a pupil of Xenophanes and said to have had Pythagorean connections, but Parmenides went very much his own way in philosophy. He is best known for a poem written (like some of Xenophanes' poetry and all of Empedocles') in dactylic hexameter, the epic meter of Homer and Hesiod. We are fortunate to possess almost all of the most important section of the work, thanks to Simplicius, who in the sixth century A.D. copied it into his commentary on Aristotle's Physics "on account of the rarity of Parmenides' writings."¹⁷ The poetic value of this philosophical work (except for the prologue) is very limited, but Parmenides may have chosen poetry instead of prose because it is easier to memorize, and also because hexameter verse, as the meter of epic poetry, connotes wisdom and authority, and is the vehicle of divine revelation (cf. 11.1).

The poem falls into three parts. First (11.1), a prologue in which the goddess announces (lines 28-32) that she will tell Parmenides two things: (a) "the unshaken heart of persuasive (or, well rounded) truth," and (b) "the opinions of mortals in which there is no true reliance." These two topics occupy the remaining two parts of the poem, which are known as The Way of Truth and The Way of Mortal Opinions. Parmenides' philosophical importance is due almost entirely to The Way of Truth, of which many think almost all has survived (seventy-eight or seventy-nine lines are extant). The Way of Mortal Opinions, of which the surviving forty-four lines constitute only a few scraps, seems to have contained a cosmogony and cosmology.

The Prologue (11.1)

The prologue proclaims Parmenides a "knowing mortal" and says he received the kind attention of divinities, culminating in a revelation from a goddess. In this he is privileged among mortals. The imagery of light and dark is stressed, but it is not made quite clear whether he moves from darkness into light. The impressive barrier of the great door indicates the difficulty of the journey. Avenging Justice, the gatekeeper, allows only those sent by Right to enter. With the approval of Justice as well, Parmenides is brought to the goddess, who promises to teach him two subjects: the truth, and unreliable human opinions. She summarizes the content of these unreliable opinions (lines 31-32): mortals believe in appearances; that is, they believe that what appears or seems to be really is, and that there is nothing else besides.¹⁸ What this means is as yet obscure, but the indications are clear that the truth is not what we mortals believe, and that our trust in appearances will be called into question.

The content and style of Parmenides' thought go oddly with his portrait of his philosophy as a divine revelation. Revealed truth tends to be truth we would disbelieve except for its unimpeachable source, whereas The Way of Truth is not what we would expect a goddess to reveal to a mortal. It is, rather, a highly structured sequence of careful deductive proofs—the first such arguments in Western philosophy. Such arguments, we feel, should stand on their own, without needing the support of divine authority. The prologue may be just a literary show, but it may also recount a kind of mystical experience Parmenides actually had. I offer as a suggestion that this experience was the discovery of the power of logic, which is perhaps represented by the unnamed goddess. For the Greeks, many things aside from the Olympian gods were considered divine. In general, anything that exists independently of human will or effort, which is everlasting and has effects beyond human control, might be called divine—such things as rivers, love, and other powers in the Universe. Deductive arguments have such power as well. If the premises of a valid deduction are true, the conclusion must also be true, and nothing in human power can make things otherwise. If we accept the premises we must accept the conclusion no matter how little we like it. Likewise for other logical relations. For example, if -A is the negation of A, then either A or -A is true; if one is false, it follows that the other is true. Now this describes Parmenides' arguments in The Way of Truth: arguments from apparently undeniable premises to conclusions unwelcome to common sense. And it is possible that reflection on the nature of such arguments led Parmenides to recognize their inescapable binding force. It is difficult for us to imagine the magnitude of this discovery (which is as profound and important as the Pythagorean discovery of the numerical basis for the concordant musical intervals¹⁹) but if the present suggestion is right, Parmenides considered it worthy of divine honor.

The Two Ways

11.2 characterizes the two Ways of Inquiry the goddess reveals to Parmenides as the only conceivable paths of thought. They are both intelligible (unlike the confused ideas discussed in 11.6), but are of unequal value. The first Way is called the Path of Persuasion, which attends upon Truth; the second is "completely unlearnable" and to be avoided (11.7 1. 2). The two Ways are identified by important claims they make: the first Way is "that it is and that it is not possible for it not to be" (11.2 1. 3); the second Way is "that it is not and that it is necessary for it not to be" (11.2 1. 5).

Two questions arise immediately: (a) what is "it" of which it is declared that it is or is not, and (b) what do "is" and "be" mean? These questions can be answered fully only by those who have understood the doctrine of The Way of Truth. For example, at that point we see that the answer to (a) is that "it" is any possible entity—which turns out to be unique: the single existing thing, the only thing that can be spoken or thought of. Parmenides may have intended these initial accounts of the two Ways to be obscure, leaving (a) and (b) as puzzles for the reader (or hearer) to work out in the course of grasping the message of The Way of Truth.

Still, some further clarification is desirable. Parmenides not only does not state explicitly what "it" is; he does not even say "it." "It is" translates a single word, ESTI (the third person singular form of the present indicative of the verb "to be"). Greek does not require (as English does) subjects of verbs to be expressed when they can be understood from their context. In the present case, it is hard to see how a reader could infer from the context some of the answers that have been given to (a), such as "The One," "The One Being," "What is," "All that exists," "Being," and "Body or Corporeality."²⁰ Another suggestion is that ESTI is used impersonally, with no subject intended (so that the "it" corresponds to the "it" in "it is raining," which means "raining is going on"); but later in The Way of Truth ESTI does get a subject, which is identified in 11.6 lines 1-2 and described in 11.8. The immediate context favors two proposals, "Any subject of inquiry" (inferred from 11.2 1. 2) and "Anything that can be spoken and thought of" (11.6 lines 1-2, cf. closer to hand 11.2 lines 6-7), i.e., anything that can be the subject²¹ of speech or thought. It will turn out that there is only one possible subject of inquiry and only one thing that can be spoken and thought of, and that these are identical.

In question (b) more is at stake. The answers proposed do not come to the same thing, but result in (or from) different interpretations of Parmenides' philosophy. In what follows, I take it for granted that two desirable features of an interpretation are (1) that "is" and "to be" have the same meaning in 11.2 lines 3 and 5 and also in 11.3 1. 1, 11.4 1. 2, 11.6 lines 1-2 and 8, 11.7 1. 1, 11.8 1. 2, and frequently elsewhere in 11.8, and (2) that Parmenides is not speaking nonsense (even if he speaks falsely). The chief contenders are that "is" is (1) existential, attributing existence to a subject, as in "there is a tiger in the zoo"; (2) predicative or copulative, assigning a predicate to a subject, as in "the tiger is hungry"; (3) a fusion (or confusion) of (1) and (2), in which "the tiger is hungry" entails not only that the subject, the tiger, has a certain predicate but also that the tiger exists; (4) veridical, asserting that the subject "is the case" or "is true," where the subject-place is held by a proposition, a description of a state of affairs, a name or a pronoun representing any of these, as in "that is true" or "it is true that tigers have four legs" (a rephrasing of what in Greek would look like "that tigers have four legs is").²²

On interpretation (1) the two Ways assert, respectively, for a suitable subject x, (1.1) x exists and must exist, and (1.2) x does not and cannot exist. On interpretation (2) they assert (2.1) x has and must have at least one predicate, and (2.2) x does not and cannot have any predicates.²³ On interpretation (3) they assert (3.1) x exists and must exist, and x has and must have at least one predicate, and (3.2) x does not and cannot exist, and x does not and cannot have any predicates. On interpretation (4) they assert (4.1) something is true (or is the case) and cannot fail to be true (or to be the case), and (4.2) nothing is or can be true (or the case).

The second Way is "completely unlearnable" for the following reason: it is impossible to know or state that which is not (11.2 lines 7-8). This argument can be given a reasonable sense on all four interpretations. (1) If something does not exist, we cannot know it or express it verbally, since there is no "it" to be known or expressed. (2) If something has no predicates, nothing can be (truly) said about it, and it cannot be known, since there is no way of recognizing it or becoming acquainted with it. Since it makes sense on (1) and (2), it also does on (3). Also on (4): if nothing is true or is the case, nothing can be known or truly stated since only what is true or is the case can be known or truly stated.

The import of Parmenides' thesis can be grasped if we consider an objection to (2) and (4). On (4), if nothing is true, then it is true that nothing is true, and so the argument does not prevent us from knowing this. Similarly for (2), if x has no predicates, we can know of x that it is predicateless. But the argument rules out this move. If "predicateless" is a predicate, then x cannot have even that. The second Way is unintelligible because nothing can be (meaningfully) asserted or denied.

Regarding (1) we might think that we can know and we do think that we can speak of nonexistent things, mermaids for example. But then asserting (or knowing) that mermaids have human heads implies that (some) nonexistent things have human heads, and Parmenides can object that nonexistent things cannot have heads at all. In fact, they cannot have any parts or attributes; no predicates can be true of them, because there is nothing there for them to be true of. Further (as before), if nothing is true of the nonexistent, it is not even true that it does not exist. Again, we have a fundamentally incoherent, inexpressible philosophical position.

The assumption driving the argument is that language and knowledge are grounded in reality. Propositions and knowledge-claims, whether existential ("x exists") or predicative ("x is F," where F is a predicate) presuppose the existence of their subject, i.e., that their subject-term refers to something that exists, and so the statement "x does not exist" is either false (because x exists) or fails to get off the ground (because it is about nothing).²⁴

All four interpretations thus make sense in the context of 11.2, but (4) does not fit 11.8, which infers properties of the subject of ESTI (e.g., lines 3-4) that are extremely implausible on the veridical interpretation, which requires something like a proposition for its subject. 11.8 also tells against (2), and with it, (3), for it separates the claims that IT is without generation²⁵ and destruction (11.8 lines 3, 40, argued at lines 6-21) from the claim that it does not suffer change (11.8 1. 41, argued at lines 22-25), whereas on (2), these claims amount to the same thing. If "x is" means "x is F," then "x came to be" and "x will cease to be" mean "x came to be F" and "x will cease to be F," and these are ways of describing changes which x undergoes—coming to have or ceasing to have predicates.

The following discussion is based on the existential interpretation (1), though most of it can be reworked in terms of interpretations (2) and (3). However, this discussion of the various uses of ESTI is not meant to imply that Parmenides was conscious of these different possibilities. He was not, it is safe to say, since the earliest attempts to analyze different uses of the verb are found a century after Parmenides, in Plato's Sophist and several treatises of Aristotle.²⁶ Not the least of Parmenides' achievements is that The Way of Truth can be read in a way free of equivocation on different meanings of ESTI.²⁷

But the two alternative Ways are not simply "it exists" and "it does not exist." The second clause in the account of each Way makes the alternatives even farther apart:

it exists and must exist
it does not exist and cannot exist

There is no possibility left for something which, say, does not exist but did or might exist, perhaps at a later time. This would allow for coming to be and destruction, which as we shall see are impossible (11.8 lines 6-21).

The Way of Truth

11.6 begins with an argument for the claim "that which is there to be spoken and thought of must be" (lines 1-2), and takes the argument as further grounds for rejecting the second Way (1. 3). The argument starts from the premises (A): it is possible for what can be spoken and thought of to be, and (B): it is not possible for nothing to be. (A) and (B) imply (C): what can be spoken and thought of is not nothing. (C) implies (D): what can be spoken and thought of is something, i.e., it exists. The "must" in 11.6 1. 1 is the "necessity of consequence": given premises (A) and (B), the conclusion must follow.

Parmenides supposes an ordinary person²⁸ will grant premises (A) and (B). If we can coherently think of something, it is possible for that thing to exist, even if in fact it does not. This holds for mythical beasts, fictional characters, and the contingently nonexistent. If we can coherently conceive of a centaur or James Bond or the King of the United States, then it is possible for those things to exist. Likewise, (B) seems obviously true, although it contains a subtle fallacy, treating "nothing" as if it could intelligibly, even if falsely, be said to exist.²⁹ The two apparently innocuous premises yield a strong conclusion: whatever can be thought or spoken of exists.

At this point we may not realize how strong this conclusion is. It seems simply to grant existence to objects of thought, as if simply discussing or thinking of something guarantees it some kind of existence—not only real things like the Matterhorn which are "there to be spoken and thought of," but anything we can coherently conceive. That is not to say that the Republicans can get a Republican president just by wishful thinking, but that their thinking about a Republican president makes a Republican president exist in their minds, though not for that reason in the White House. Nothing has yet been said to rule out this interpretation, but it turns out that the conditions on what can be spoken and thought of are so strong that they exclude practically everything.

Parmenides next (11.6 lines 4-9) surprisingly introduces a third Way of Inquiry. The description of the two Ways launched in 11.2 as "the only ways of inquiry there are to think" is now seen to mean not that they are the only two Ways, but that they are the only intelligible Ways: the Way presented in 11.6 is not even intelligible, as is shown by the self-contradictory and so unintelligible description "to be and not to be are the same and not the same" (lines 8-9). Also, those who follow this Way seem unable to tell that it is unintelligible (lines 4-7).

The third Way is the Way of mortals, of ordinary people. But it is no ordinary person who goes around saying "to be and not to be are the same and not the same." This must be intended to rephrase what we do think, in a way that brings out implications of our ordinary thoughts. What we do think is that there are many different things in the world, that these same things come into existence and cease to exist, and that they move and undergo other changes. In 11.8 Parmenides will disprove all these beliefs. Here he characterizes them so as to make us see that they are hopelessly confused and represent an incoherent view of reality. In particular they are infected with nonbeing, which has already been proved unthinkable and unutterable, so that any Way that makes any use of it is unintelligible.

11.7 (which may be the immediate continuation of 11.6) says more about the beliefs of mortals: they imply "that things that are not, are," which rephrases the more complex expression of 11.6 lines 8-9. Parmenides fastens on the changes indicated above, in which a thing now is, now is not. In cases of destruction a thing now exists, now does not exist. In changes of quality, it now has, now lacks a given quality (now is hot, now is not hot). In denying that the third Way is a coherent way to think about the world, Parmenides challenges our views of reality, denying the most obvious aspects of the world around us, a world we believe in mainly because of our sense experience ("habit born from much experience" [11.7 1. 3]). The senses (ridiculed in 11.7 1. 4) tell us that there are many things in the world and much change. Also, for the early Greek philosophers the most important features of the natural world are its diversity and change. Consequently Parmenides presents a radical challenge both to unreflective ordinary people and to the presocratic philosophical tradition.

Parmenides' description of our view of the world is so abstract up to now that it is easy to miss the sweeping breadth of his attack, and easy to fail to see where he is headed. However, his attack on the senses (11.7 lines 3-4) is a warning note, as are his instructions (11.7 lines 5-6) to accept the conclusions of his arguments (if we cannot refute them) rather than what sense perception and our habitual ways of thinking tell us. This last claim, it should be noted, involves the first explicit statement of the contrast between reason and the senses, which immediately became and has since remained one of the focal points of philosophical discussion. Moreover, Parmenides' preference of reason over the senses makes him the ancestor of rationalism (in some of its forms) and constitutes an important element in the historical background of Plato's Theory of Forms.

Next comes the account of the true Way. 11.8 is the longest continuous stretch of writing from any presocratic, and lines 1-49, in which Parmenides expounds his theory, contain the first elaborately structured series of deductive arguments in the history of Western philosophy. This is no accident. Parmenides' thesis is entirely contrary to our beliefs and experience, and he rejects the kinds of evidence on which ordinary knowledge-claims are based. Hence, he will not find support for his theories in ordinary mortal beliefs about the world and about how we come to know things; instead, he uses the divine power of logic to prove them and bases his proofs on truths no one could deny, such as premises (A) and (B) in the argument in 11.6 lines 1-2.

He begins by reminding us that there is only one Way left to consider (11.8 1.1), and identifies a number of "signs" along the Way (11.8 lines 2-6). These "signs," which point to the correct interpretation of "it is," are attributes he will discuss. To complete our journey along the Way we must understand all the signs, so that a full understanding of "it is" includes knowing all these features. What is, i.e., any legitimate subject of thought and discourse (from now on I shall call this subject "IT"³⁰) is ungenerated, imperishable, whole, of a single kind, unshaken and complete, timeless, one, and continuous. To interpret these claims in the light of what follows, we see that Parmenides is claiming that anything we can coherently think of exists, but did not come to be and will not cease to be (argued at lines 6-21); it is undivided, unique, and changeless and uniform throughout ("whole," "of a single kind," "one," and "continuous," argued at lines 22-25), motionless ("unshaken," argued at lines 26-33), and possesses all attributes that can coherently be conceived to apply to anything ("complete," argued at lines 22-25, 42-49). The opening lines of 11.8 thus serve as a table of contents for most of the remainder of The Way of Truth.

11.8 lines 6-21. Ungenerated and Imperishable

This section falls into three parts. The first (lines 6-15), which contains three arguments to prove that IT is ungenerated, concludes (1. 14) apparently by asserting that it has eliminated destruction as well (doubtless because parallel arguments hold against the possibility of going out of existence). The second (lines 15-18) strongly reaffirms the dichotomy between what is and what is not, which forms the basis for the preceding arguments. The third (lines 19-21) argues for the timeless nature of IT, which provides a further argument against coming to be and destruction.

11.8 LINES 6-9. ARGUMENT 1

IT did not come into existence out of what is not. What is not cannot give rise to anything or foster the growth of anything. Since (as we know from 11.2 lines 7-8 and 11.3) what is not cannot be intelligibly spoken or thought of, coming into existence out of what is not cannot be coherently conceived.

11.8 LINES 10-11. ARGUMENT 2

IT did not come into existence out of what is not, because if it did, it came into existence at a certain time. But there is no reason (necessity) for it to be generated at any one time rather than at any other, since that would mean that what is not has different attributes at different times. In particular, it would mean that what is not supplies a condition (necessity) for coming to be at one time but not at others. But what is not has no attributes at any time. Since there is no time at which IT should come into existence rather than at any other time, and since IT cannot come into existence at all times, there is no reason to suppose that IT came into existence at all. This is the second application of the Principle of Sufficient Reason we have seen.³¹

11.8 LINES 12-14. ARGUMENT 3

IT did not come into existence out of what is not, because what is not cannot give rise to anything except itself. This is the first enunciation of the principle "Out of nothing, nothing comes to be,"³² which was implicit in earlier Greek thought, even as far back as Hesiod, and which afterwards, because of Parmenides, became a touchstone for subsequent Greek cosmogonies.

These arguments show that coming to be out of the sheer nonexistent is impossible, which is most obviously relevant in the first stage of a cosmogony. The arguments say nothing of more familiar cases of coming to be, which can be described in terms of changes among already existing things. The table comes to be when already existing wood, screws, and glue are treated and arranged in the right way. The baby comes to be when an existing sperm and an existing egg join in the right way in the right circumstances, and already existing food is transformed in the mother's body and passed on to the embryo in the appropriate way, etc. Nothing Parmenides has said so far excludes the possibility of the familiar world, just as long as we follow Xenophanes and Heraclitus in declaring it to be eternal.³³

11.8 LINES 19-21. ARGUMENT 4: TIMELESS EXISTENCE

Another claim about IT is suggested in 11.8 lines 5 and 19-20. Since IT is now, IT never was nor will IT be (1. 5). Conversely, if IT came into existence or if IT is ever going to be, IT is not. Parmenides excludes past and future tenses in describing IT. Just as it is timelessly true that 2 + 2 = 4 (it is nonsense to say that 2 + 2 used to be 4), so it is timelessly true to say that IT exists. No argument is offered for this interesting point, though the following Parmenidean considerations point to it. Parmenides holds that IT is eternal (a conclusion that follows from the facts that IT exists (1. 3), and that IT is without generation or destruction (lines 6-18) and is unchanging (lines 22-25). So at any given time IT is identical to the way IT is at any other time. All things that are ever true to say or think of IT are always true. Therefore, there are no time distinctions relevant to IT. But past and future tenses only make sense by contrast to each other and to the present, and if there is no possibility of any differences, there is no basis for any such contrast. "2 + 2 was 4" only makes sense if there is a possibility that at some time it is not then true that 2 + 2 = 4. In the absence of such a contrast, there is no use for past and future, or even for the (tensed) present; all that is left is the timeless present. Further, the conception that all that is exists timelessly does away with time itself. There are no temporal distinctions, because there is temporal uniformity, IT exists fully, always, IT is held in shackles, bound fast (lines 14-15), unable to escape, or even to move. Moreover, parallel reasoning would imply that there are no spatial distinctions either, so that the concepts of space, place, and spatially distinct parts of IT do not make sense, IT (and therefore the universe) proves to be without time or space.

In its context, timelessness is presented as another obstacle to coming to be and destruction (cf. 1. 21). The point will be that coming to be and destruction take place in time: at one time a thing exists and at another it does not. Timeless existence thus precludes the possibility of the different times needed for coming to be and destruction to take place.

11.8 lines 22-25. Undivided and Continuous

11.8 lines 6-21 show that IT is not generated from or destroyed into what is not. Lines 22-25 go on to prove some consequences: IT is undivided, IT is not any more or any less in any way, IT is all full of what is, and IT is all continuous. The premise from which this stretch of argument begins, "IT is all alike," restates the results of lines 6-21, that there is no generation or destruction. It is equivalent to "IT must fully be" (1. 11). Since IT all is fully or alike, IT is undivided: there are no gaps. The clearest application of this consequence has to do with existence in time: lines 6-21 show that IT has no absolute beginning or end; lines 22-25 show that IT has no intermediate beginnings or ends either, IT is temporally undivided and continuous, with no temporally separate stretches occurring one after the other. Further, since lines 6-21 rule out all generation and destruction, not only can IT not come to be or cease to be as a whole, but the same is true for any parts or attributes IT may have. Consequently, IT cannot change, in the sense of acquiring new parts or attributes or losing ones IT has. That would make IT more or less at different times in those respects ("more" in the sense that something more belongs to IT or is true of IT), "IT is all full of what is," in that at any time IT is complete, with all possible parts and attributes.

A further consequence seems to be that IT is unique, that there is only one existing thing. At any rate this is the best candidate in the fragments for an argument for monism.³⁴ IT is not more or less in any way, therefore IT holds together (equivalently, IT is all full of what is, and equivalently again, what is draws near to what is). Therefore, IT is all continuous.³⁵ I think Parmenides' idea is that since what is clings together, there cannot be more than one distinct thing. Whatever is, is uniform and has the same (i.e., all possible) attributes. It is not possible to say that one thing, or attribute, leaves off and another starts here and/or now.

The point can be argued as follows (though Parmenides does not put it this way).³⁶ If there is more than one existing thing, say there are two. (The point will hold no matter how many there are.) Call them A and B. A and B are different from one another. (If they were the same thing, we would have one thing, not two.) Therefore, each has at least one thing true of it that is not true of the other. (At the very least, it is true of A and false of B that it is identical with A). Further if A and B exist in space, then A and B must occupy distinct regions of space, so that it is true of A and false of B that it is here, and false of A and true of B that it is there (where "here" and "there" designate the spatial locations of A and B respectively). But if anything is true of B and false of A, then A is lacking in that respect, and so is not "full." Once we have plurality, we violate the "full" and "not any more or any less" requirements as discussed above. Since "not full" and "more and less" are unintelligible notions, plurality is ruled out as well.

11.8 lines 26-33. Motionless

Since motion is one kind of change, change in place, lines 22-25, which claim that IT is changeless, already imply that IT is motionless. In fact, motionlessness is mentioned only at lines 29-30. The remainder of this section provides material to bolster the too brief arguments of the previous section. Lines 26-29 say that changelessness is a consequence of no coming to be or ceasing to be, which was an important element in the reconstruction of the argument for changelessness offered above. Lines 32-33 contain material used in the proof that IT is unique. In fact, these lines go farther: "IT is not lacking; if IT were lacking, IT would lack everything." This follows from the radical dichotomy between what is and what is not, strongly stated at 1. 11 : "it must fully either be or not (be)." If IT is lacking, then IT is infected with not-being. But if IT has some not-being, IT must completely not be.

The argument against motion does not depend on properties of space, only of time. The whole series of arguments so far has depended on temporal considerations, and subsequent proofs are based on the arguments against coming to be and ceasing to be.

11.8 lines 34-41. Linguistic Difficulties

At this point Parmenides goes back to his principle that it is impossible to think (or know or express) what is not (lines 35-36, cf. 11.2 lines 7-8, 11.3). Since IT is unique (lines 36-37, cf. lines 22-25), there is only one legitimate subject of thought, speech, and knowledge, so all thinking has IT as its subject. But 1. 34 says more. Not only is there only one intelligible subject, IT, there is also only one intelligible thought, that IT is. If any other statements could intelligibly be made about IT, they might conceivably be true or false, and that would lead to the difficulties noted previously. Moreover, what Parmenides has shown implies that many statements are unintelligible—those involving time, change, plurality, and motion. Strictly speaking, in referring to IT, we cannot use past or future, but only the timeless present. Even the tensed present is ruled out because it depends on the illegitimate contrast with past and future. In the same way we cannot strictly refer to IT in ways that depend on illegitimate contrasts with other illegitimate predicates. Strictly speaking, it is wrong to call IT changeless, motionless, or one, since "motionless," "changeless," and "one" are coordinate with the unintelligible notions motion, change, and plurality.

Parmenides now confronts a question raised by his doctrine. If, strictly speaking, the only intelligible thought or statement is "IT is," what happens when we think or utter other sentences? "The kids are running" is certainly a different sentence from "IT is," and it has a plural subject and a predicate that involves motion. What does Parmenides propose to do with that sentence, and for that matter with all the rest of language except for the one legitimate sentence? Further, we might try to turn his reasoning upside down. If there is an exact correspondence between what can be spoken or thought of and what is, don't sentences like "The kids are running" guarantee the existence of their subjects? Instead of restricting the range of legitimate subjects to IT, why not extend the range of existing things to anything we can think of, including Julius Caesar, the Wizard of Oz, my first billion dollars, and square circles? Simply by thinking of them or expressing them in an intelligible utterance we guarantee that they exist.

Parmenides has a straightforward reply to this last attack: anything infected with any nonbeing cannot exist at all; IT is the only subject not so infected, so IT is the only legitimate subject. But what, then, of the first question? How are we to deal with all but one of the vast number of apparently well-formed sentences that can be formed in an ordinary language? The answer is given in lines 38-41: all words name IT, even if the people who use them do not realize it. This thesis can be supported by a familiar line of reasoning: each word must name either IT or what is not; but IT is the only thing that is; what is not cannot be named; therefore, each word (despite appearances) names IT.

As the examples (lines 40-41) show,³⁷ the names in question are predicates. "It has been named all names" means that all (the indicated) names ("coming to be," etc.) are predicated ofIT.³⁸ Since in 11.18 mortals are said to have given names to the subjects of these predicates, Parmenides seems to view language as consisting of names that can be formed into subject-predicate propositions. The "correct" language has only one subject, "IT," and one predicate, "is," which form one proposition.³⁹ The "incorrect" languages of mortals contain many names, but insofar as mortals use these names in significant speech, they form them into propositions that amount to the single correct proposition "IT is." Apparently all subject terms amount to "IT" and all predicate terms to "is." Although humans, misled by their senses, have created languages to describe the world the senses reveal to them, their languages (like the senses) mislead, and in that they relate to reality at all, what they say must express the only meaningful utterance, "IT is."

Strictly speaking, only one thought is intelligible, "IT is." Mortals are mistaken in using language to say other things. They cannot be doing the impossible, referring to what is not, so they are (despite what they think) referring to IT. This solution may not satisfy, but it invites a further question. Occasionally Parmenides recognizes what can be spoken or thought of "strictly speaking." But even in describing IT, he frequently says things which, strictly speaking, he cannot say. In calling IT unchanging, for instance, he describes IT in terms of change, which is nonexistent and therefore inexpressible. Worse, he frequently speaks of what is not. In effect, he uses words to express concepts which he analyzes, showing that they are incoherent and so unintelligible, and then he forbids their use. This approach is dictated by the peculiar nature of his project. To eliminate virtually all our language and concepts he must begin in the world of our language and concepts and finish in another world where practically none of our language and concepts apply. In moving from the one world to the other, he uses and then discards words and concepts. At the end, he is left with only one expressible thought, but on the way there, and even afterwards in communicating to mortals who are not yet there, he must use many technically illegitimate words and concepts.

11.8 lines 42-49. Spatial Uniformity

The proof that IT is motionless depends on properties of time, not space, but it naturally brings up the question of the spatial extension of IT, the final topic discussed in The Way of Truth. The actual argument (lines 44-49) closely parallels the argument for temporal invariance (lines 22-25) and the consequences are similar: because IT is inviolate (i.e., complete), IT cannot be in any way (here, in any place) more or less; therefore, any spatial stretch of IT is identical with any other stretch. But there are no separate stretches, since that would require there to be stretches of what is not in between, and there is no such thing as what is not. Consequently, IT is spatially uniform ("not at all greater or smaller here than there" [lines 44-45], later rephrased as "equal to itself on all sides, it meets with its limits uniformly" [1. 49]). Spatial uniformity and temporal uniformity imply analogous results: IT is not only timeless, but also "spaceless" (there are no spatial distinctions); and timelessness and spacelessness too have analogous results: there is no time or space.

There are two obstacles to this interpretation: the prominent mention of limit at the beginning and end of this passage and the comparison of IT with a ball have led many to believe that IT is spherical. But if IT is spherical, and its (spatial) limit is the surface of the sphere, we are entitled to ask what lies outside (or inside⁴⁰)—and there is no answer, since it can be neither IT (which is all found in the sphere) nor what is not (which does not exist).⁴¹ Also, though a spherical surface is uniform, it is possible to distinguish different areas and points on it, and so there are spatial distinctions. And once there are distinct points, there are things true of each point that are not true of any other, so that IT is not "full" (cf. the argument sketched above in the discussion of lines 22-25). These are strong reasons against believing that IT is spherical. And there is no need to follow that interpretation. For Parmenides uses the word 'limit' not of limits in space or time, but to express fixity and invariance, also expressed by the image of shackles (lines 13-14) and bonds (1. 26). And his comparison with the sphere stresses not its shape but its "completeness," i.e., being "evenly balanced," which in the context of the treatment of the spatial properties of IT corresponds to the property of being "not at all greater or smaller here than there." Like IT, a sphere is spatially uniform, but it is so in a different way. Parmenides emphasizes the similarity, but we should not be misled into thinking that the similarity goes further than it does or can.

11.4, whose original position in the poem is uncertain,⁴² plays some role in the discussion of spatial properties. It seems to argue for spatial indivisibility (as 11.8 1. 22 argues for temporal indivisibility): IT is not spatially divided into bits separated by stretches of what is not.

Concluding Remarks on the Way of Truth

Parmenides' chief contributions to philosophy occur in The Way of Truth. The present account has been generous in its assessment of this section. It would be easy to fault Parmenides for expressing his ideas and arguments in difficult and obscure language, as well as for leaving large gaps in his thought. An obvious case occurs at 11.8 1. 14, where he concludes that IT can neither come to be nor perish, whereas the arguments up to that point attack only coming to be. He also fails to argue for his view that IT has timeless existence (11.8 lines 5, 19-20). More importantly, there seem to be errors in the arguments. At 11.8 lines 22-25 he infers that since generation and destruction are impossible, so is change in qualities. But even if we agree that generation and destruction are inadmissible, since they involve what is not, it is not clear why something cannot change from red to blue. The same comment holds for the argument against motion (11.8 lines 26-33). Also, if, as suggested above, 11.8 lines 22-25 are meant to argue against plurality, the reasoning turns on moves from plurality to difference, from difference to lacking some possible attribute, and from lacking some possible attribute to being infected with what is not. But this last move is highly dubious: why should something's being red (or here) and so not blue (or not there) imply that it is infected with the unintelligible what is not? These are some of the most important flaws in Parmenides' argumentation. Other objections can be made against the arrangement of the arguments, since it is not always clear where one topic leaves off and another begins.

The reason for interpreting Parmenides charitably is that only in this way can we fully appreciate the interest, the potential, and the challenge of his ideas and arguments. Only if we make the effort to unravel his tortuous reasoning and fill in the gaps in ways congenial to his point of view can we hope to understand his enormous contribution to philosophy.⁴³

The Way of Mortal Opinions

11.8 lines 50-52. Transition to The Way of Mortal Opinions

The goddess now proceeds to fulfill her promise (11.1 lines 28-32) to expound the opinions of mortals. She makes it clear that this way is unreliable (11.1 1. 30) and that her account is deceitful (11.8 1. 52). Why does she deal in deceit in this way? The only clue in the text is found in the final two lines of 11.8. She will expound the world "as it appears," i.e., the world as the senses present it to us, "so that no mortal opinion may ever overtake you." The Way of Mortal Opinions gives an account of the phenomenal world along familiar Milesian lines. However, it differs significantly from earlier theories, nonphilosophical as well as philosophical. Thus the goddess does not present Mortal Opinions in a straightforward way. Her starting point is the characterization she offers of Mortal Opinions: mortals believe "that the things that appear must genuinely be, being always, indeed, all things" (11.1 lines 31-32). And her project in The Way of Mortal Opinions is to give the best possible account of the KOSMOS which appears to us and which mortals wrongly believe to constitute reality.

Earlier philosophers had attempted this task but their accounts were flawed. The goddess has presented what she asserts to be the truth. Now she presents a theory about the world of appearances that, as we shall see, is based on an error. For this reason, The Way of Mortal Opinions cannot win out over The Way of Truth. But if someone else proposed a different account of the world of appearances, it might be judged superior to The Way of Truth. Hence the need to give Parmenides the best possible account, "so that no mortal opinion (i.e., no other theory about the phenomenal world) may ever overtake you." If Parmenides can show that The Way of Mortal Opinions, which he knows to be false, is superior to all possible rival theories, then no such theory can threaten The Way of Truth.

11.8 lines 53-55. Foundations of The Way of Mortal Opinions

The Way of Mortal Opinions is based on a radical error. Humans posit, or rather, the best possible reconstruction of the world of appearance posits, a basic duality, and all other things in the KOSMOS and the history and development of the KOSMOS depend on these two basic things, or are reduced to these two things and their interactions, and "in this way they have gone astray."

In what precisely did they go wrong? They named two forms "of which it is not right to name one." This may mean (a) that their basic error was being dualists instead of monists, (Since only one thing exists, it is wrong to posit two principles. In fact, since both principles are different from the one existing thing, neither exists and neither of the two can be the basis of a correct account, but this is not Parmenides' point. Here he only calls attention to the incorrect number of principles.), (b) that they went wrong in naming even one of the two principles they named (Neither exists, so neither can be a principle.), and (c) that it was correct to name one, but not the other (One of the two principles, namely Fire, really exists and is identical with what The Way of Truth calls what is, and the other one, Night, does not exist, and is identical with what The Way of Truth calls what is not.).⁴⁴

The Way of Mortal Opinions is importantly different from Milesian accounts, which are monistic in deriving the KOSMOS from a single principle. The Way of Truth has shown that true monism is incompatible with the world of appearances. No single principle can give rise to plurality and opposition;⁴⁵ at least two principles are needed, and Parmenides' dualistic cosmology puts opposition in the world from the start (11.8 1. 55). This minimal amount of error is thus ineradicable from any plausible account, even the best one, of the phenomenal world (interpretation [a] above) so even though the two forms the goddess reveals are not arbitrary, they remain fabrications designed to account for what does not really exist (interpretation [b] above).

11.8 lines 55-61, 11.9. The Basic Duality, Fire and Night

Just as what is has attributes called "signs" (11.8 1. 2), so the two principles of The Way of Mortal Opinions have opposing signs. Fire is mild, light (as opposed to dense), bright (so that it can be called light as opposed to dark), whereas Night is dark, dense, and heavy. No doubt other attributes hold as well. Fire will also be hot and dry, and Night cold and wet. Each of the two elements is distinct from the other and identical with itself,⁴⁶ its identity and difference being marked by the indicated "signs." Now for the first time we have true elements, distinct basic forms of matter that always preserve their own identity; they may intermingle with each other and form other substances, but under no conditions can they be transformed into each other.

The phenomenal world of Mortal Opinions is full of things composed ultimately of Fire and Night and endowed with properties that stem from the opposed "signs" of the two elements. 11.9 lines 3-4 can be taken as saying either that everything (that is, everything in the phenomenal world other than Fire and Night in their purest form) is a compound of the two elements or that the KOSMOS as a whole is full of the two elements (counting both their pure and their mixed forms). These interpretations both accurately characterize the world of Mortal Opinions.

Cosmology

The total amount preserved from The Way of Mortal Opinions is too small to enable us to form anything like a complete picture of the views it contained. The programmatic remarks in 11.10 and 11.11 indicate some of its contents: a cosmogony and cosmology treating the astronomical subject matter obligatory in any presocratic treatise. 11.14 and 11.15, which refer to the moon, form part of this section, as does 11.12, which is part of a description of the "wreaths" or rings which are the courses of the celestial bodies.

An interesting feature of the cosmology is the presence of "the goddess who governs all things" (11.12 1. 3), who is presumably identical with Necessity (11.10 1. 6), and who generates Love first of all gods (11.13). We seem to have a theogony as well as, or as part of, the cosmogony, and this reading is supported by ancient references to other gods in Parmenides (War and Discord) and to stories Parmenides told about the gods.⁴⁷ Unfortunately we do not know more.

The prominence Parmenides gives to Love ties in with his interest in sex (11.12 lines 4-6) and embryology (11.17).

The final one of Parmenides' beliefs we will consider is his account of perception and thought, which attempts to find a physical basis for these psychological phenomena.

11.20 Most general theories of sensation are of two kinds. Some make sensation occur by like and other by the opposite. Parmenides, Empedocles, and Plato by like…. Parmenides made no general definition, but said only that there are two elements and knowledge is in accordance with the one which exceeds. For if the hot or the cold surpasses, the thought becomes different. Thought that is due to the hot is better and purer, but even this requires a certain balance. [Here Theophrastus quotes 11.16.] For he speaks of sensation and thought as the same. This is why both memory and forgetfulness are produced from these through their mixture. But he did not at all determine whether or not there will be thought or what will be its arrangement if they are equal in the mixture. But that he makes sensation occur also by the opposite in its own right is clear from the passage where he says that a corpse does not perceive light, heat, and sound because the fire has left it, but it does perceive their opposites cold and silence. And in general everything that exists has some knowledge. (Theophrastus, On Sensation 1, 3 = DK 28A46)

This extract is given at length principally to show the level of Parmenides' thinking on these subjects and because it is the first detailed account we have of any treatment of these phenomena. Two implications of the discussion are that the soul, whose presence makes things alive, is fiery, and that human (and other) thought depends on the presence of Fire and Night. Thus, this account of thought makes sense only in the unreal world of Mortal Opinions and cannot be taken as a clue to how it is possible to have knowledge of the real world of Truth.

This last question is an important one to be sure. If the only thing that exists is the unique, unchanging, spaceless, timeless IT, then what is the status of Parmenides himself, the goddess, and the mortals whose opinions he disbelieves? Parmenides can dismiss the last group as part of the unreal world of opinions. And he can dismiss his own body in the same way. But how about his mind which is having the thoughts? Some have claimed that IT is identical with thought, translating 11.3 as "For thinking is the same as being," and 11.8 1. 34 as "thinking is the same as the object of thought" (where the only possible object of thought is IT). But apart from these two lines, which can easily be translated differently, there is no suggestion that IT thinks, and the whole approach Parmenides takes in The Way of Truth, as well as the absence in the earlier philosophical tradition of any tendency to consider thought as an entity, tell against this interpretation. I find it more plausible that he simply did not raise or answer this question, which may simply not have occurred to him.

11.18 appeared in Parmenides' poem "after he had related the ordering of perceptible things"⁴⁸ and serves as an appropriate conclusion to The Way of Mortal Opinions, stressing that the account covers what occurs "according to opinion," emphasizing change (growth), the time distinctions of past, present, and future, the names applied to all these illusory phenomena, and the fact that the names are the product of human decisions (not due to the nature of reality).

Notes

¹ The numbers in parentheses are the numbers of the fragments in H. Diels and W. Kranz, Die Fragmente der Vorsokratiker, 6th ed., Berlin, 1951 and later editions. (1) = DK 28B1.

² The pronoun can refer to either the road or the goddess.

³ The text is difficult here. The word translated "keys" can also mean "locks," and "that fit them" is a loose rendering of a word that means more literally "alternating" or "in exchange for."

⁴ The manuscript text of this word varies.

⁵ The last two lines of 11.1 are controversial. I follow Owen's text and interpretation ("Eleatic Questions," Classical Quarterly 10 [1960]:84-102). Other possibilities: "how what is believed would have to be assuredly, pervading all things" (G. S. Kirk, J. E. Raven and M. Schofield, The Presocratic Philosophers, 2nd ed., Cambridge, 1983) "… all of them passing through all [the tests]" (J. Lesher, "Parmenides' Critique of Thinking," Oxford Studies in Ancient Philosophy 2 [1984]:1-30).

⁶ The Greek is ambiguous between "all things" and "all mortals."

⁷ Alternatively, "born from much experience" can modify "this way."

⁸ I follow Owen's suggestion (reference in n. 5). In DK, the text of 1.4 reads "for it is complete, unmovable and without end."

⁹ Some accept a different manuscript reading that would be translated as "wherefore all things are a name."

¹⁰ Other possible translations: "it is the same thing which the constitution of the limbs thinks" (lines 2-3); "the full is thought" (1. 4), which can be understood to mean that the content of what people think is "the full," i.e., the reality described in The Way of Truth. The translation given in the text accords better with Theophrastus' account of Parmenides' views on the nature of thought (11.20), which quotes 11.16.

¹¹ Context in Galen, Commentary on Book 6 of Hippocrates' Epidemics, 11.46 (= DK 28B17).

¹² I follow the text given in Gallop's edition (Toronto, 1984).

¹³ 11.19 is quoted by Plato (Theaetetus 180e) and Simplicius. Most editors, including DK, believe that it is a misquotation of 11.8 1. 38. Its authenticity was defended by Cornford ("A New Fragment of Parmenides," Classical Review 49 [1935]: 122-23) and has found influential support. 11.19 complements the claim in 11.8 lines 38-39 that all words really name the one existing thing, and mortals invented the many names in their mistaken belief in a world of plurality and change. The point of 11.19 is that the correct name for the one existing thing is "to be."

¹⁴ Plato, Parmenides 127b-128d (= 12.1).

¹⁵ Plutarch, Against Colotes 72, 1126A = (DK 28A12).

¹⁶ Parmenides' unmoving one existing thing forms a natu ral contrast with Heraclitus' world full of plurality and change (especially on the doctrine of Heraclitean flux—see pp. 142-44 [of my Philosophy before Socrates]). There seem to be some verbal echoes of Heraclitus, for example, compare 11.5 with 10.67.

¹⁷ Simplicius, Commentary on Aristotle's Physics 144.25-28 (= DK 28A21).

¹⁸ Lines 31-32 have been taken differently, with "these" (1. 31) referring not backwards, to "the opinions of mortals" (1. 30), but forward, to the final line and a half of the fragment, which then do not contain just a thumbnail sketch of the false opinions of mortals, but the Goddess's endorsement of the cosmology presented in The Way of Mortal Opinions as true.

¹⁹ See pp. 91-92 [of my Philosophy before Socrates].

²⁰ Some of these proposals have further drawbacks. What, for example, could "Being is" mean? And "What is, is" (and perhaps "All that exists, is") is a tautology, which some have considered an unsuitable starting point for a metaphysical theory.

²¹ By "subject" I mean, roughly, "topic" or "subject matter"—what we speak or think of.

²² For more possible meanings of "is," see Mourelatos, The Route of Parmenides (New Haven, 1970). Mourelatos favors what he calls the "is of speculative predication," which assigns to a subject a predicate that reveals the nature, reality, or essence of the subject (p. 56 ff.). On this interpretation, Parmenides claims that the essence of any subject is to be ungenerated, indestructable, unchanging, etc. I find the suggestion interesting, but difficult to work out in detail. (See D. Furley, "Notes on Parmenides," in Exegesis and Argument, ed. E. Lee, A. Mourelatos, and R. Rorty [Assen, 1973], pp. 1-15; and D. Furley, The Greek Cosmologists, vol. 1 [Cambridge, 1987], p. 37 n. 10.) I find the evidence presented for the occurrence of this use of "is" prior to Parmenides (Mourelatos, pp. 60-61) unconvincing.

²³ An alternative interpretation—(2.1′) x has and must have all predicates and (2.2′) x lacks and must lack at least one predicate—is ruled out by 11.8, where the subject of ESTI is said to lack several predicates (e.g. lines 22-23). However, (2.1) and (2.1′) are identical (as are (2.2) and (2.2′) if (as I argue later) there is only one possible predicate (see on 11.8 lines 34-41). In that case, the apparently negative assertions are "strictly speaking" wrong and (like apparently true affirmative assertions like "IT is one") to be understood as expressed in ordinary, misleading human language, not the philosophically correct language for which The Way of Truth prepares the way.

²⁴ Referential theories of language have a powerful intuitive appeal, but lead to paradox: to say what is not is to say nothing, but to say nothing is not to speak at all (cf. Plato, Theaetetus 189a, Sophist 263b). Or, "to 'mean' something is to spear it with a spoken (winged?) word. Then to speak of what is not is to hurl a term at—what? It isn't there" (M. Furth, "Elements of Eleatic Ontology," in The Pre-Socratics, ed. A. P. D. Mourelatos [Garden City, N.Y., 1974], p. 255 n. 27).

²⁵ I use this word as synonymous with "coming to be"

²⁶ Especially Categories, De Interpretatione, Posterior Analytics, and Metaphysics.

²⁷ For example, there is no need to see him moving from "x is not F" to "x is not," i.e., "x does not exist."

²⁸ His philosophical forebears ought to agree too.

²⁹ "A, which can exist, is distinguished from B, which (poor thing) cannot: invalid, for to say 'nothing cannot exist' is not to ascribe compulsory non-existence to anything but to say that it is necessarily (truistically) true that what doesn't exist doesn't exist, and this unexciting reformulation disables the argument. The fallacy is the so-called de re interpretation of modal statements" (Owen, op. cit. [p. 152 n. 5], p. 94 n. 2). I assume that Parmenides was unaware of the fallacy.

³⁰ At present "IT" is only shorthand for "whatever can be legitimately spoken or thought of," with no suggestion intended that IT is unique or even a "thing" (as opposed, say, to a concept or mental idea). It will emerge from the argument that IT is unique, but otherwise unlike other familiar (real or mental) entities.

³¹ Anaximander provides the first, see p. 40 [of my Philosophy before Socrates].

³² This principle is usually known in its Latin form, "ex nihilo nihil fit."

³³ See pp. 64, 139 [of my Philosphy before Socrates].

³⁴ Despairing of finding an argument for monism, J. Barnes (The Presocratic Philosophers, rev. ed. [London, 1982], p. 207) has maintained against a unanimous tradition that Parmenides was not a monist. He criticizes efforts to see 11.8 lines 36-37 as an argument for monism, but does not consider lines 22-25.

³⁵ SUNECHES, the word for "continuous," is closely related in form and meaning to SUNECHESTHAI, the word translated "holds together"; notions of mathematical continuity are not here present.

³⁶ See 11.8 lines 57-58 … for evidence that Parmenides conceived of IT along these lines.

³⁷ The examples are the predicates treated and eliminated so far ("alternating bright color" stands for the kinds of change ruled out in lines 22-25).

³⁸ On the alternative translation "it has been named all things," the point will be the same, though not so well expressed. Since coming to be and the other "things" mentioned in lines 40-41 have been shown to involve what is not, they cannot be spoken or, therefore, predicated of anything, including IT.

³⁹ In fact, this language consists of only one word, ESTI, a verb which implies its subject….

⁴⁰ On one interpretation of a spherical IT, IT is the spherical surface, not the entire bulk of the sphere enclosed by the surface. This interpretation has the advantage that a spherical surface is uniform, which cannot be said for the other case, where, for example, the center has different properties from the other points.

⁴¹ The idea that Parmenides had a conception of "Einsteinian," "curved" space (so that the question of what is outside does not arise) has been proposed (F. Cornford, "On the Invention of Space," in Essays in Honor of Gilbert Murray [Oxford, 1936], pp. 215-35) but is historically absurd.

⁴² We would expect it to have occurred somewhere in the middle of 11.8, perhaps just before 1. 42, but Simplicius, our source for 11.8, seems to claim that he is quoting a continuous passage.

⁴³ One of Melissus' virtues is that he presents his version of Eleatic monism in a clearer and more systematic way….

⁴⁴ Interpretation (c), which seems incompatible with The Way of Truth, is Aristotle's interpretation (Aristotle, On Generation and Corruption 1.3 318b6-7 [= DK 68A42], Metaphysics 1.5 986b31-987a2 [= DK 28A24]).

⁴⁵ Hence Milesian monism is mistaken. Anaximander illegitimately and obscurely derives hot and cold from the APEIRON … and Anaximenes' monism involves air in a plurality of states….

⁴⁶ Fire and Night thus are "lacking" or "incomplete" in a way incompatible with what fully is. See the discussion of uniqueness, p. 169 [of my Philosophy before Socrates].

⁴⁷ Cicero, On the Nature of the Gods 1.11.28 (= DK 28A37), Plato, Symposium 195c (not in DK).

⁴⁸ Simplicius, Commentary on Aristotle's On the Heaven 558.8 (= DK 28B19).