The Art of Reading: From Homer to Paul Celan

Citation: Bollack, Jean. 2016. The Art of Reading: From Homer to Paul Celan. Trans. C. Porter and S. Tarrow with B. King. Edited by C. Koenig, L. Muellner, G. Nagy, and S. Pollock. Hellenic Studies Series 73. Washington, DC: Center for Hellenic Studies. http://nrs.harvard.edu/urn-3:hul.ebook:CHS_BollackJ.The_Art_of_Reading.2016.

17. The Heraclitean *Logos**

Historians of Greek thought have tended to begin by trying to read Heraclitus as a systematic thinker; they have tried to decipher a world-system in his work comparable to those of other archaic thinkers, and they have sought to reveal its place in a general cosmology. The repeated failure of these attempts at reconstruction finally convinced me that Heraclitus did not have his own system, and that the unity of his approach did not lie in any positive content but rather in his critical analysis of cosmological theories, nourished by assertions that were current in the learned circles of his day.

By projecting each fragment from Heraclitus onto the specific horizon of its transmission and of its metamorphoses, we gradually manage to clarify the reflexive structure of the aphorisms, by distinguishing, among all the established uses of the language, the tension created by language itself, the tension inherent in words, in the logos as word in its relation to a thing. This distinction, made explicit in the texts, between a reflective use of the logic of discourse and a non-reflective use of language, implies abstraction; it emanates from the content, but at the same time it leads to a return to the content in order to apply to it the laws of abstraction in a complementary fashion. The fragments are concerned alternately with one process or the other, with the analysis of the inherent contradiction in discourse or with the reconfiguration of discourses by means of this internal reference.

This was the focal point of the commentaries I published with Heinz Wismann in 1972. [1] The present essay reconsiders the underlying premises of that work, in the light of certain works published in the interim. [2] Some {217|218} syntactic analyses (and syntax predominates) might well be discussed further; the commentaries on several fragments could be clarified. But the arguments we proposed then still hold true, on the whole, and need only to be reaffirmed.

Our break with the interpretative tradition, in its ancient but especially its modern form, remains unchanged. In the tradition, the logos is usually charged with doctrinal content; it is presented as the site or receptacle of a “truth.” In the modern hermeneutic experience, it shifts away from this role and uses abstraction as the instrument for analyzing what is presented as truth.

If Heraclitus’ thought does not focus on the presence of Being but on the universe of meaning (generally speaking, on everything that seems to be endowed with meaning or that is likely to be subjected to analysis by the law inherent in the logos), then the elimination of distance and the return to hypostasis, from Schleiermacher’s critique up to the present—if we confine ourselves to the modern period of scientific approaches to the texts—have to be understood as an a priori move, manifesting the structure of a prejudice, and should thus be linked to the critics’ own motives. [3]

The definition of the word logos has been central to the construction of a functional system that the critics see as founded on what the logos is thought to express other than itself, that into which it has been assimilated. Hence their approach to the initial aphorism, in fragment 1, [4] which is not presented as an independent unit and is a “fragment” only in relation to the collection:

Now, men always live far from understanding the discourse [logou], which is this one, before they have listened to it, as after they have listened to it the first time. For everything lives according to this discourse, so that one sees them apparently ignorant of what they practice, their words and their actions, such as those that I myself develop thoroughly, dividing each one according to its nature and {218|219} showing how it is made. Other men are unaware of what they do when they are awake and of all they forget during their sleep.

Since discourse (logos) indeed occupies the central position, as the sole reference of the passage, it sufficed to erase the difference between the word and what, according to the Heraclitean corpus, it expressed, that is, the difference between the logos and the utterances to which the structure deduced from it is applied. Thus the logos was seen as a term of pure substitution, the sign of an objective message.

Based on the attributes of war in Fragment 53 (“War is both king of all and father of all”), which were thought to be identical to those of the logos, [5] critics believed they could infer the true nature of the referent, of which the term logos was just an equivalent, presenting things in a didactic order. [6] The thesis is absurd, but instructive, first because of its excess (or its radical nature), and then because appearance, a very obvious similarity, reveals the shift to application. But this principle of pure contradiction can be found nowhere else apart from discourse. {219|220}

This is an extreme position; it is limited to the cosmic dynamic of opposites, drawing from it a principle that controls and explains physical and cosmological evolution, [7] whereas the reduction to this principle, favored over the Platonic flow of Schleiermacher and Hegel, [8] had allowed us to see in the logos not the exposition of opposites but the law of opposites, with its tension. [9]

Starting out from the word, with its possible orientations, scholars sometimes clung to what Heraclitus said with regard to the subject, and to the reasoning that articulates his discourse; [10] sometimes, regarding the content, they preferred meanings that referred to an objective organization, understood as “moderation,” as “relationship” or “reason.” [11]

In any case, it was an objective truth that was to be revealed and could quite easily be fitted in with the translation of “discourse,” if it is in language that this truth can be expressed, as it appears throughout Heraclitus; and, as Heraclitus is alleged to have said, it appeared even before his time. In the nineteenth century, the logos was understood as the “language of nature,” as “nature’s revelation of an intelligible discourse.” [12] According to Heidegger, language speaks Being and spoke Heraclitus before Heraclitus. [13] Being speaks in the language that reveals {220|221} meaning. For F. J. Brecht, a proponent of this line of thinking, man is led by the logos to his own truth, torn away from his alienation. [14]

This shift entailed a move from “listening,” in the strict sense of the word, to “hearing,” that is, knowing how to understand discourse; discourse had thus become a fable of truth, rather than an account or an analysis.

If the more immediate question is that of universal reason, the law that controls becoming, it is more difficult to account for texts like fragment 1, where one had to acknowledge that reference was made to discourse, as it is spoken, produced, and pronounced. The two aspects had to be reconciled. In fact, they could not be made to agree with either option. So certain scholars resorted to the expedient, or the compromise, of explaining the aporia as the result of the deficiency of a thought that was still archaic and unable to distinguish between the two aspects. [15]

In the fragments, words are often used with the meanings that were conventional in Heraclitus’ time, as for instance the word “reputation” (frag. 39). Elsewhere we find “speech” and “discourse,” as such, distinguished by Heraclitus, each isolated in its own function and structure. We remain within the logic of discourse, moving on to the composite formed by the act of saying and what is said, and to the relational value of the “reason” inherent in the relation between contradictory terms. This is neither worldly reason nor the reason of argument in discourse. The internal relationship is projected onto existing utterances. It is thus reproduced, in the restructuring of specific discourses, according to measures and ratios; this “reason” provides an explanatory principle suitable for governing languages, because it is drawn from language.

The crucial problem, which gave rise to the debate over criticism and its aporias, is confined to the rare passages that are directly concerned with the nature of discourse. Secondary uses, where the inferred principle is transferred to meanings and imprints on them a structure that has been rectified and clarified by means of language, have, in the history of criticism, served to organize a system of positive assertions, which was then arbitrarily established as a content, occupying the position of the logos, whose analysis had provided the means to bring these assertions into question. {221|222}

Heraclitus speaks of a reference provided by language; he never tires of talking about situations that show how the people around him miss the point of analysis and fail to grasp the structure of the language that they use and that impinges on their behavior. This presumably very violent struggle, committed to a pure reflexivity that refuses to accept any form at all of a prior external foundation, is something quite simple, yet of capital importance.

This bias is evident right from the first words of the prologue that constitutes fragment 1, an opening that is “parodic” in the broad sense of the word; in a kind of repeat performance, Heraclitus adopts the form of prologues that had preceded his. [16] The content is first of all “this one,” concerning the very situation of the discourse: to show what its status is, what a program is, and what use is made of the genre of the arkhē, the beginning. It is the opposite of an immediate discourse “about everything,” or “of everything,” as is evidenced by the use, so eloquent in its limitative precision and singularity, of the demonstrative, tearing the logos away from any other reference, from any use that would not be self-referential.

Aristotle cites this beginning in a famous passage of his Rhetoric, [17] to show how difficult it is to regroup the words in Heraclitus’ book, because their autonomy, and thus the freedom of construction, is so great. In the first sentence, “Now, men always live far from understanding the discourse [logou] which is here explained,” one can punctuate by separating a second part of the sentence after “always”: “this Logos, which is always [or: is always true], [18] men are [or: become] incapable of understanding it.” It is a distancing from the eternal truth. The adverb “always” goes with “which is.” [19] Or else one makes the {222|223} break before “always,” by taking the demonstrative as the predicate: “this Logos, [which] is as here explained.” [20]

The first of these interpretations may seem more natural, and Aristotle considered it to be so, adding that the text as it is formulated does not oblige us to be confined to this “evidence,” which the ontological reading he refers to suggests. [21] He does not say: it is this or it is that, but: it is not necessarily this. [22]

Aristotle’s comment takes interpretative practices into account. He says nothing of an essential ambiguity; he does not leave the question open, but rather exposes the difficulty. Starting from an already established “natural” reading, Aristotle is interested in the manner of its enunciation, which does not seem clear to him. His account is concerned with an interpretation that {223|224} is situated and dated. [23] So he does not consider the possibility of making “this one” the attribute of “which is,” [24] as one must if it is a question of isolating the discourse as such. The construction of the demonstrative is still corroborated by the iteration of the same phrase, later on in the same fragment 1: “which is this one.” The meaning surely excludes the other construction, which is considered by the supporters of a tradition whose legitimacy had been called into question by Heraclitus’ use of the verb “to be” as a copula, a significant choice; the pronoun toud[e] (“this one”) indicates the discourse, logos, already determined by the article in its raw state as discourse. His use of the pronoun in this context helps to define the logos in its specificity.

The aporia lies right there, and it basically concerns the comprehension of the text: when one chooses, either directly or by way of a revelatory discourse, to understand the term logos as meaning “law,” the “law” that governs the course of the future, the anthropological break that characterizes life encompasses a before and an after, or at least a time opposed to an anteriority, the very moment of listening.

If it is “discourse,” as it is structured in light of its “object,” then the opposition is clear. The contradiction can only be elucidated for those who are prepared to separate themselves from the community of men, not necessarily through arrogance or disdain, but by the methodical necessity of making distinctions. The thing—that is, the practice of symbolic systems—is always there. Homo loquens. People live with it, next to it, without discerning the meaning that its use holds and that exempts them from introducing any other. Any other reference would be ritual, ritualized, subject to the same analysis, reducible to the status of practices: a poem or a tribune, a tribunal or a religious procession.

If it is “law,” we have to admit that it was formulated before Heraclitus, in the form in which now, it is thought, he reveals it in his book, so that men could have previously been unaware of it or unwilling to accept it. The message, one might say, is not at all subjective: “The art is to listen well, not to me but to reason, to know how to listen to all-things-that-are-one” (Fragment 50). [25] {224|225}

Scholars have often accepted the idea that the voice of an objective law had been heard before, favored by earlier Heraclituses, in a line of prophets of truth, without having been heard by anyone among men, before Heraclitus had repeated it. [26] This is surprising, and no doubt a false logic.

Without doing violence to the text, we begin to assume that men lived without hearing or listening to what Heraclitus focuses on in his discourse, as he listens and analyzes the current discourses around him. This is the only proper subject for critical reflection, to examine and then question the legitimacy of the examination; but instead of postulating this type of thought—however eloquent the distance between fragments, constructed by the reader, may be—one could, in an even more restrictive way, posit the subject as the only possible reference of what is said about it, by exploiting the aporia, which, since Aristotle and even before him, offers the most reliable heuristic method.

On this condition, we can accept the following: we can understand as an intelligible outcome that a thinker might come and say to you, to the Greeks, and then to the West, even to the whole planet, that what he says will not be “heard”—one should say “listened to”—before being heard, because Heraclitus, by invoking this unique object and by defining it thus, constructs an absolute, paradoxical situation, the borderline case of the paradox where the practitioner is invited to turn his reflection away from practice, not toward a theory, which would be just a new (theoretical) practice, with its own technicalities, of course, but rather to draw from it the opposite law, lying beneath what he does, simultaneously within reach (zuhanden) and at a distance.

Any other discourse is secondary, that is, a substitute; the latter negates another and clarifies it with an imitation. Heraclitus enters the world of existing forms, not to make use of it according to its form, but in a different way, according to the law of language, which is the single necessary—not arbitrary—condition.

But in general the not-listening, understood as not-hearing or incomprehension, is linked to the oracular form of Heraclitus’ discourse (is it not “a natural one on the part of an author who has chosen the language of enigma and equivocation”? [27] ) Since the discourse had already failed to be heard before Heraclitus had written it, it is not this discourse but the truth, an objective correspondence called “meaning” that alone, it is thought, deserves to be called {225|226} eternal, like the fire, which for that reason is compared to the logos, accommodating this elevated primitivism through the idea of a rationalization of more primitive beliefs. Depending on the critic’s perspective, the author may be ahead of his time; from other perspectives, he may be archaic. A historical objectivization is created that ratifies the false logic of the exegesis.

Non-perception, in this particular component of the sentence (and only here), changes the scene; we move from the inner voice through which, unbeknownst to human subjects themselves, the truth of the universe is expressed, to the enigmatic revelation of a philosopher, determined to disguise it—but why disguise it, since it lives in everyone? For the pleasure of misleading?

However, this eternity could be problematized, and restored to a meaning. It has been seen as an attribute of the world, as if eternity had been one of Heraclitus’ concerns, which was not the case. If the truth is general, conceived as a revelation that has come down through the ages, one might add that it is foreign and inaccessible to the common man, in the past as in the future, before and after Heraclitus, who was acquainted with the notion, but for nothing—or for no one.

The two moments in time are not really symmetrical in this representation. The intervention of the person, and his speech, are an indispensable element, as we have seen. The moment, and thus the “after,” cannot be reduced to a new utterance of a common truth, proclaimed in ancient times and then later unknown. To understand the reason for the role attributed to the “self” (moi), the content must justify the difference between that person and other men. It must be possible to define the separation at the heart of the relation between the person and what he is listening to. What men, separate from Heraclitus’ person, could have listened to, and have always not listened to (in the literal sense), was language, which makes them, not being, but men, in an eternal world, one might say; and in another world, as well. The ceaseless becoming, on which the attribute is based, is not an affirmation—a doxa—on the part of Heraclitus; it could only become one, as it did for Plato, if it were confronted with the contrary statement, in a discursive organization where opinions are classified—that is, placed in relation, by reference to a system of thought or a world system, whatever it might be.

There is nothing like this in Heraclitus. If we wonder about the meaning in fragment 1 of a phrase like “whereas all things come into being in accordance with …” (ginomenōn kata), we are virtually obliged to include the distance introduced by the reference, which distinguishes from the structure of the logos the decision to apply that structure, as the only available parameter, to {226|227} all constructions and etiological claims. [28] This is where the “always” comes in: “in any case, in the face of the natural phenomena that we see occurring, it will always be this non-speculative principle that prevails,” and it must always have prevailed, ever since thinkers started to explain and order, not to state the truth, which has no place here, since it is only a matter of making a statement, but to come as near to the truth and as perfectly as possible, using the key provided by the fact of talking about it. The agreement (kata) refers to an explicative principle that makes it possible to grasp the shift from listening to the logos to practical inexperience, in the domain of the word and of deeds. So what is at issue is no longer simply discourse as such: in a broad sense, it indicates all the situations where people come together in the exchange of given discourses and in rituals: when these occur according to the same mode of language, discourse applies to their study and analysis. [29]

There is rivalry and predominance, in this sense self-assurance, claimed by a man who expresses a distance that can justly be considered absolute, since this distance refuses to follow customary discourses or to construct others. The fragments are split up according to categories. No assertion is situated outside the norm imposed by the reference drawn from the discourse: the reference does not express the world, but the world can be expressed according to the reference.

According to the fragments, “they,” men, are agents, or interpreters, subject matter, or rivals in the analysis. There is a perceptible parody, perhaps even a sneer: “they” analyze, and yet they have no idea what they are doing, which is “what I am doing there, and they say they are doing, but do not do.” [30] The individual self (moi) is in its place in that place where Heraclitus situates himself, in {227|228} a practice where listening allows him, having chosen this course, to do better than the others. The words imitate one another; imitation eliminates them.

“[This is the order of things] that I myself develop thoroughly, dividing each one according to its nature and showing how it is made” (frag.1). The sentence is in no way oracular. The syntax does not innovate here; the style re-produces, as happens elsewhere; it delineates the domain of cosmologies” (peri phuseōs) with the words of practitioners, and chooses the most sober and unadorned words—to set forth, to divide, to show—always with the basic idea of a guiding structure. “Nature” (phusis) is not the “all,” not the “Universe”; but as it is for doctors, it is the constitution of a particular being, a constitution that can be analyzed according to what it is, according to its nature.

Why would Zeus be presented by the formula “the name Zeus,” if it were a question of his ritual empire, or of another, which the rite allegorizes?

“One, art. It does not allow, and allows, the name of Zeus to be spoken alone” (frag. 32).

At the outset, it might seem plausible that the choice of the term “name” has the function of detaching from the god the word that refers to him under his specific identity, and that the act analyzed by the antithetical figure concerns the fact of “being said,” according to the words of the fragments, thus related to the “name of Zeus” (Zēnos, the genitive of Zeus), distinguished from “Zeus” in the nominative.

The god’s name is etymologized in the formula: it speaks of “life” (Zēnos, from zēn, “to live”) as an essence of the thing to which it refers. [31] The word, in this free and interpretative use, detached from simple denotation, reveals the contradictory structure of language. If it means “life” by superimposition, independently of the word “Zeus,” it would move to its opposite by applying what it names to itself. At the other extreme, on the side of the designated object, it will mean death—the god is his own thunderbolt.

“The name of Zēnos,” if one groups the two words together, wavers between the instrumental use of the designation (“unwilling [for the name of Zeus] to be spoken alone”) and the interpretation of the designated thing by the word “life.” [32] If “life” is used separately, it is cut off from the thing and can only be {228|229} applied to it by passing over to the opposite pole. In this case the etymology leads to a contradiction because it is partial. There is no homonymy, as there is for the bow, that might reveal the contradictory structure. [33]

The statement, understood as an analysis of the process of naming, in its dual function, no longer has the predicative significance of revealing the fundamental ambivalence of any hypostasized power. In support of this conclusion, we have the linguistic operation designated by the words “name” and “to be spoken,” and also the formula used to present it: “One, knowledge,” if we consider that semantically “knowledge” (sophon), which is “skill,” is not about an apophthegmatic content [34] but about mastering something through analysis. Without this transfer to the means of expression, what would “knowledge” and the uniqueness of the limitation to this single process be (“one”)? What uniqueness is there other than the single reference? [35] “One, art”—the meaning of “one” is restrictive, exclusive: “it is always that process”; there is no other.

“One” (masculine) will not be the totality “one” (feminine), but a restriction—the exclusivity of a technique or a competence (sophiē). “What real knowledge is [to sophon is determined], is that thing exclusively.” The formula recurs in the attested fragments (it was probably more widely used in the complete {229|230} book), always referring to the same principle of analysis. [36] It is yet another set linguistic expression; the formula pre-existed, or else it parodies another one; the language is re-used to make its structure apparent. What is said is what constitutes knowledge. There is no reason not to assume this, while making it clear that the predication of a speculative assertion—an “all is one” or even “all things, one,” re-introduced into Heraclitus’ work by one of the doxographic traditions [37] —has not been distorted, in the written phrase, so that the speech act can substitute for it. The statements systematically examine this speech act as being the one thing of which one can say, in analyzing it, that there is nothing else by way of knowledge. The speech act refers neither to this One, which would be a Zeus, nor to any unity, but to the uniqueness of a particular structure, inherent in language, which actuates all thoughts as the discourses that they are, without being understood as such.

When language is analyzed in its structure, there can be no relation to a content. Such a relation is, however, what authors assume when, in choosing the meaning of “discourse,” they make it external to the whole; they go on to express the truth of this external whole outside the nature of things in a kind of transcendence, as the law that governs the nature of things or, in other words, as the “unity of opposites” or the “all-one.”

The logos cannot be pressed beyond the contradictory structure that constitutes it in its enunciatory power, being chosen as the unique reference, which introduces no opinion constituted in language, neither “subjective” nor “objective,” no positive assertion. All the analyses of Heraclitus in the surviving fragments (and the same was obviously the case for those that have not survived) delve into the practices, either to identify the structure of the logos that did not appear to those who used it, or to project this structure onto explicative systems, somewhat haphazardly, stripping them of their arbitrary, imaginary, or speculative content, and linking them to this single, logically scrutinized reference.

If the formula of the “all-one” is to be found in Heraclitus’ work, it is not as a doxa, a succinct resume of his own doctrine—he has none—but as the repetition of an utterance, of an earlier doxa, for the purpose of examination, of {230|231} spectroscopic analysis, under the fires of the logos. It is quite true that the assertion was upheld in the past; it condenses, better than any other, the prerequisites of an Ionian physics with which Heraclitus has been doxographically associated since Aristotle.

The “I” is not an assertion of superiority based on a widely held truth, on the professorial model, or on an elective revelation, along prophetic or shamanistic lines; it can be defined only in relation to discourse, to what is said, true or false, about truth or election.

So the listening in question is not directed toward the perception of a message, whatever it may be, but toward what this message communicates through its expression. There is no other knowledge but listening to the mode in which the various forms of knowledge are expressed.

The step taken by listening to the word leads to acceptance of another transcendence, another position of exteriority. One does not enter the system of established references, according to the field of social activity, in order to make use of it; one leaves it in order to analyze its functioning, by going back to the structure of the utterances and by studying what is being expressed in what is said. The “I” defines very precisely the limits of this exclusionary option, which one might legitimately call methodical. It situates itself beyond communion, attracted by the magnetic pole of a singularization that has never been refuted.

The “I” creates a problem in the traditional reading and in the framework of the convictions imported by many scholars. If it is the “I” who is speaking, in the name of the new doctrine that we imagine he is introducing and that he is “publishing” in fragment 1 just as we read it, we may wonder why he presents it as being incomprehensible to everyone, because it is so personal to him. This is the first aporia, which involves communication and the closure of the subject. But then, as fragment 50 signals (these are the same words): “as you are listening, not to me (I don’t count) but to the … logos,” the authority reserved for the speaker is abandoned in favor of the “objective” truth that is always known (by whom?) but never understood, or else that exists, even though unknown, pre-existing like an eternal origin, like the world, awaiting Heraclitus’ revelation. The contradiction is blatant. I am the only one who speaks the truth of all.

The terms of the aporia, rather than being swamped in the incoherence of an inspired prophetic discourse or in the loftiness of melancholy and despair, can be pinned down as such by the observer. In order to put aside our prejudices, as we always should, we must force ourselves to stay within the system of thought, so that we can distinguish between two agencies, that of the “I” and that of the logos. The most natural hypothesis is that the “I” who speaks—and this happens rarely—is not the same as the logos, but refers to the structure of the logos: the “I” does not make the logos appear, but applies it by re-considering {231|232} the systems of organized thought. As a result, the opposition becomes clearer at the heart of the distinction, right from the prologue: such that, depending on the role I choose, I decompose and re-decompose the material without losing the single non-speculative referent that the utterance offers. The dual operation is immediate or applied, which makes it easy to understand fragment 50: the contemplated (and expected) feat is in this case more arduous than in others, for it relies on immediate listening, without recourse to the science mastered by Heraclitus, the competence. The transfer is indicated thus: “By listening to the language in its nature and structure [with the article, as at the beginning, fragments 1 and 2], and not to me.”

The remainder of the sentence is usually (indeed, practically always) translated and understood as if it said: “It is wise to agree that all things are one.” [38] A formula that sums up the truth: “The all is one.” An agreement that, incapable of being applied, according to the presuppositions, to a formerly accepted doctrine, is related to the logos itself, and finds in these paired terms its highest expression; this explains the choice of the word “wisdom” (sophon esti), to designate the recognition of a supreme principle. [39] It is the logos that speaks through Heraclitus, its servant, and who here gives it voice in the place of his own. It speaks in the soul of every man. [40]

The shift from the “I” to the logos is read as a dispossessing of self in favor of an invitation addressed to a “you.” If it is universal, the logos will be dialogical: it will speak to another. At the same time, it is clear that in fragment 1, this other is not listening. An expression of wisdom tossed into the breeze. {232|233}

As for textual criticism, the interpreter notes that the verb “to be” is introduced by a correction: the Hippolytus manuscript has “to know” (eidenei); the content “all-one” is considered an essential, or central, assertion; grammatically it depends (via the infinitive clause) on the verb that this probable correction introduces. [41]

As for the history of the language, we ought to think that the word sophon is not the correct term for the supreme degree of initiation into knowledge of the logos; it does not encompass speculative knowledge so much as competence. This knowledge is closer to the technical virtuosity of an artist or a soothsayer than of a philosophical or moral wisdom. Prior expectations lead to neglect of the linguistic data.

The word logos is reiterated in the verb “to say together” (homologein, “to say in harmony,” to make the logos agree). The achievement, resulting from unmediated listening, presupposing an absolute exteriority, here concerns a formula, “one-all” (hen panta), on the principle of the Milesian speculation about the arkhē; we can acknowledge that it is not repeated to indicate a content known to all those to whom the book was addressed—a book whose interpreters have gone so far as to make it Heraclitus’ credo—but rather to demonstrate the difficulty of justifying its terms within the structure of the logos: to know how to “say together” (homologein) within the inherent and constitutive contradiction.

Heraclitus does not reject any system. To do so would mean putting something else in its place. The reuse of the formula shows that the examination is more critical, that he is wondering about the meaning of what is said, or about the means of justifying its use. The repeated axiom will not be rejected, it will be re-said, bearing new content that transforms the meaning in opposition to its authors and users. What do the words say?

The four fragments (50, 32, 41, and 108) in which the nature of art, of the sophon, can be identified and localized, so to speak, [42] apply the term to this {233|234} external listening, this traversal of discourses—until discourse can be perceived as such. Fragment 108 establishes a strict differentiation:

Of all those whose discourses I have heard, none arrives at the point where it distinguishes that which, separated from all others, makes art.

Thus “art” (or knowledge) can only come from discourses (in the plural) on a topic, before which the “I” of the fragments—through which the separate subject constructed by Heraclitus speaks—arises as a “listener” (ēkousa). [43] The content of knowledge that speech organizes carries it away and distances it from the distancing on which the Archimedean point of listening is constituted, when it produces this art outside the topic, separate from all the topics touched on in discourse. Speaking does not provide access to this point. All aphorisms have as a presupposition this basic mechanism of a distance beyond speech, which determines the choice of words on the basis of absence and negation.

The point of view of exteriority is made autonomous in the recurring formula “one, art [of knowing]” which, each time it appears, presents the result of an absolute listening, the contradiction revealed by the “name of Zeus” (frag. 32), or by the formula of universal government, applied to cosmic principles (frag. 41). [44] The application, that is, the justification, is especially difficult in the case of “all-one,” where the speculative nature may, or should, at first glance, seem too obvious. Getting past it will be quite a feat. The difficulties in translation encounter a serious problem with the logical or philosophical core, when {234|235} one hears: “It is wise to agree [for certain people: within the community of the enlightened [45] —there must be one, after all] that all is one” (and not that “the one is all” [46] ); wisdom is considered the supreme degree of knowledge of the logos, [47] extended to the universal, where the organization of the world appears in the truth of a fundamental unity, discovered by Heraclitus. Even recently, it was possible for one scholar to write: “[T]his is the earliest extant statement of systematic monism, and probably the first such statement ever made in Greece.” [48] This surprising opinion has, in fact, almost unanimous support in the scholarly field.

This proposition has been made the fundamental one, from which all the rest can supposedly be derived. However, its relation to the law of opposites, which the experts moreover still keep at the center, presents a problem, unless one asserts, as some have, that the equation of the “all” and the “one” prove that cosmology “is just a particular case of the law.” [49] In fact, it is the unity of the framework that we see expressed, a framework in which the alternating phases of a cosmic cycle can follow their course, the site of a cosmology. Mutation of all, then mutation of the one, [50] according to the doxographical disposition of fragment 10. [51] One would have expected that it would be more a question of plurality. Thus these scholars imagine the movements of a pluralized totality towards unity, and vice versa, the exchange of fire for all things according to fragment 90: [52]

At the cost of fire everything is exchanged, fire at the cost of all things together, as one exchanges goods for gold and gold for goods. {235|236}

“From all its components a unity emerges, and from this unity all things emerge,” writes Kahn. [53] What has become of all the opposites and the tension that has been made the driving force, [54] along with the Ionian Muses and Plato’s Sophist? [55]

Of course, opposites do not constitute the body of doctrine, as some would have it; they are a construction derived from the inherent contradictory structure of the logos, but even supposing that we might agree hypothetically on a reading that is, in fact, falsely literal—that is, non-reflexive—of the paired terms day-night, fire-earth, and so on, the “all-one” or the “one-all” are irreconcilable.

What has been made the central dogma is a pure aporia of interpretation, which from a scientific and logical standpoint should have led to a new challenge to the hypothesis. These fundamental reading processes did not take place because the readers would have had to deal with the principal questioning formulated by the text. But it is this difficulty that we must tackle—first of all the hermeneutic difficulty and then the objective problem. Heraclitus formulates the problem that the tradition of speculative language poses. “Knowing how to do it” means using other reflections, no less interpretative, of established formulas, as in fragment 41:

To govern all things [torn away and unique] through all things [taken as a whole]

With regard to fragment 30, one wonders what poetic or philosophical work deals with a god who has created the world: {236|237}

The world, the same for all, no one, neither god nor man has created it, but it always was, is, and will be ever-living fire, kindled in measure and quenched in measure.

No such reference can be found. [57] One can get around the problem by taking the phrase as a “polar” expression—“neither god nor man”—meaning “no one.” [58] But why say “no one,” why “the uncreated”? The facts of the problem can be reversed. One can say that neither god nor man is a being that can be considered from this angle, as a creator of worlds [“One does not see which god …”]. Gods do some things, but they do not create. They operate within. The statement is a negation, and by excluding any possible causality, divine or human, one might say what is said about the eternity of the world; we must choose this path in order to take up and transfer, as has been done, the famous formula that characterizes divination in the Iliad. [59] The life of the world is before us, the “always” can also be added to it, the “ever-living” fire. The diversity of imaginable concretions of igneous energy must make us adopt—or prefer—this hypothesis as the site of elementary tensions. The famous opposites would intervene here, once the nature of the substratum is established. We might use the term “blaze”; we might say “extinction” for the fire; and we might say “for one portion” when speaking of the proportion, according to the rules of discourse, that the survival of the world confirms and whose survival presupposes existence. Is there anything oracular, or prophetic, about that?

In this same fragment 30, the structure of the world (kosmos) does not distinguish the world before our eyes—from what would we distinguish it? [60] —but presents it as a subject of the current discourse, which clarifies its structure, in order to produce a fixed reduction: it is “the same for all.” Whatever the {237|238} constructions may be, they can be applied to what is said here, what is identical to all the schemas (and not shared by all men). [61] Whatever the model, we shall always be taken back to this irreducible analysis of all the possible cosmological hypotheses. What can one say?

The freedom to reorder the words, their isolation, with the new regroupings it allows, and the fragmentary use of idioms, according to language and outside of language, to lead toward that freedom, have not been read as marks of mastery and of control in line with the laws of obscuritas, not as signs of an intentional enigmatization, but rather as evidence of an uncontrollable inspiration. [62] The revelation itself is revealed through the irrational mode of the vivid utterance. One could thus stop at the linguistic or semantic aporia, accepting it with its ambiguity, with its mysterious contradictions, which ring “true,” but whose falsity (or devices) can be adopted for all uses, according to the preconditions set (or followed) by each individual. [63]

There is a wide gap between the contradictory structure that the aphorisms extract from a matrix, that they show, and the ambivalences that prophetic expectations assume. “Heraclitus, after all, is a prophētēs [a spokesman] of [that is: for] the logos.” [64] {238|239}

If Heraclitus “demythifies” (without the anachronistic meaning “freeing of opinions” being attached to this word), and if one limits oneself to linking the word to the systematic analysis of the existing discourses, all the interpretative discourses that take what he says back to the philosophical or archeological “mystery”—a mystery that in fact he deconstructs profoundly by an analysis of the underlying assertions—assume the function of mystification (re-mystifying what has been de-mystified).

The idea of a solitary prophet who remains misunderstood when he has spoken, knowing that he speaks for no one, is an arbitrary extrapolation from what Heraclitus says about listening, the correlative of speaking, that can be understood only logically from speech as such.

We find descriptions in the fragments that are familiar in Greek thought and tradition. Obviously the author does not confine himself to them, and his own thought is in no way anticipated. It is the matter that he analyzes, that he does not invent or modify; he examines it to confirm his analysis of language—as if language could and should receive a maximal expansion, language as a symbolic system that expresses itself beyond just its verbal associations—and to apply to it equally systematically the structure resulting from the analysis.

The two points of view necessarily converge. The more one discovers by listening and by testing the extent of the principle’s validity, the more the application, by turning around the assertions of etiological systems, can take it into account and dismantle the artificiality of those systems. The transition, “in the final analysis,” is imperceptible.

Nothing, not even—or especially not—thought, can be compared to the logos, which only provides the schema or the model from which (and by reference to which) the phenomenon can be analyzed, or rather, constructed in order to be illuminated and clarified. Analysis of discourse provides the matrix for analytical discourses. [65]

Philological (and philosophical) truths are challenged as they are formulated; they seem unacceptable. If everything is just language, and does not convey anything else, while conveying everything, nothing refers to any Being or to any substance. No truer Being appears (or is revealed) through language.

Heraclitus has no doctrine of opposites. One never comes to understand a fragment as a stage in a demonstration. Besides, the fragments have a strict autonomy; like lyric poems, they are closed in on themselves and perfectly decipherable, in their limited context, if the aphorisms are not cut out. {239|240}

This is always just an explicative hypothesis, renewed again and again, which only circumscribes the framework in which a single fundamental structure is applied. The word “fire” gets to the heart of the matter by boring into the abyss of darkness; it says “night.” This is a one-way, irreversible relationship; reversion is the result of a symmetrical operation. So one cannot claim, as some do, that Heraclitus proposes opposites as the basis of his system. Strictly speaking, he has no system, nor any “opposites.” The clash results everywhere from a contradictory linkage.

No other author has ever given rise to more divergent discourses. One scholar speaks of his “religious sense,” emphasizing the “worthlessness of human knowledge in comparison with divine,” [66] when there is never any question regarding beliefs or gods, of anything but human practice. In light of this profession of humility, the initiate’s pride masks the truth that he reserves for “those who have no ears.” [67] The “oracular style,” [68] which has been compared to prophetic inspiration, would translate both the loftiness of the person who knows he is chosen—chosen for speech, displaying a justified disdain for the “common man” (even though the logos is said to be “common”)—and the mark of passion, both a chosen protective screen and a sign of election, a necessity imposed on reason for communicating itself through this irrational form. Within the limits of this framework, all interpretation, absolutely all, can come, and has come to stay.

There has been a misunderstanding of the meaning of the obscuritas that concerns the use of elements of language, the gap that another usage inscribes, in relation to conventional forms. There is comparable confusion in the interpretation of the lyrical parts of tragedy, where the unexpected, the most rigorously thought out, is supposed to translate unrestrained ecstasy. Obscurity in Heraclitus, based on a great autonomy in the elements of the discourse, whether he makes or quotes or remakes them, obliges us to listen to the language, to define what is being said and how it can be said. In a necessary re-composition of the subject, it traces a path toward, including the perception of a unique {240|241} referential structure and a return movement—anō, katō (“up,” “down”)—toward its projection, or its re-injection into the figures of established language.

Imposing the principle of self-reflection as a prerequisite for every assertion, Heraclitus’ approach marks a rupture; it is not so easy to measure its impact or assess its significance. It is squarely situated in a speculative tradition, however radical the re-examination of positions may be. A global view, considering the speculative rise of the Ionians initiated by Thales and his force (of which we do not have a perfect representation), suggests that the movement was on its way to realization, and that reflection, appearing at this pivotal moment, found the means to reach its ultimate goal. Heraclitus does not give up on reconsidering totalization. In a way, the reversal, through the definition of art as knowledge of knowledge, creates the possibility of a reflexive speculation in which, whenever the dry is opposed to the wet, what is foregrounded is the unique possibility of producing the relationship.

This is a general rule: it is only when literality is established that the question of meaning, that is, the meaning of this meaning, of the philosophical significance, can be posited.

Heraclitus did not invent truth; he extracts it from the lack of understanding in which it has been held “until the present day.” {241|}

Works Cited

Babut, D. 1976. “Héraclite critique des poètes et des savants.” L’Antiquité classique 45, 2:464–496.

———. 1987. “Une nouvelle édition d’Héraclite.” Revue philosophique 177, 2:201–213.

Barnes, J. 1982. The Pre-Socratic Philosophers. I. Thales to Zeno. 2nd. ed., rev. London. Orig. pub. 1979.

Bechtel, F. 1924. Die griechischen Dialekte. 3 vols. Berlin.

Bollack, J., and H. Wismann. 1995. Héraclite ou la séparation. 2nd ed. with new preface. Paris. Orig. pub. 1972.

Brecht, F. J. 1949. Heraclit. Ein Versuch über den Ursprung der Philosophie. 2nd ed. Wuppertal.

Burnet, J. 1930. Early Greek Philosophy. 4th ed. London. Orig. pub. 1892.

Conche, M. 1986. Héraclite: Fragments. Paris.

De Martino, F., L. Rossetti, and P. Rosati. 1986. Eraclito, Bibliografia, 1970–1984. Perugia.

Diels, H. 1897. Parmenides. Berlin.

———. 1901. Herakleitos von Ephesos. Berlin.

Diels, H., and W. Kranz, eds. 1951. Die Fragmente der Vorsokratiker, 6th ed. Berlin.

Gigon, O. 1935. Untersuchungen zu Heraklit. Leipzig.

———. 1945. Der Ursprung der griechischen Philosophie, von Hesiod bis Parmenides. Basel.

Guthrie, W. K. C. 1962. A History of Greek Philosophy. 6 vols. Cambridge.

Hegel, G. W. F. 1977. Phenomenology of Spirit. Trans. A. V. Miller. Oxford. Orig. pub. 1807.

Heidegger, M. 1979. Heraklit. Gesamtausgabe II (Vorlesungen, 1923–1944), vol. 55. Frankfurt.

Hölscher, U. 1968. Anfängliches Fragen. Göttingen.

———. 1985. “Heraklit zwischen Tradition und ‘Aufklärung.’” Antike und Abendland 31:1–24.

Kahn, C. 1979. The Art and Thought of Heraclitus. Cambridge.

Kassel, R. 1971. Der Text der aristotelischen Rhetorik. Prolegomena zu einer kritischen Ausgabe. Berlin.

———, ed. 1976. Aristotelis Ars Rhetorica. Berlin.

Kirk, G. S. 1954. Heraclitus: The Cosmic Fragments. Cambridge.

Kirk, G. S., and J. E. Raven. 1957. The Pre-Socratic Philosophers: A Critical History with a Selection of Texts. Cambridge.

Kraus, M. 1987. Name und Sache. Ein Problem im frühgriechischen Denken. Amsterdam.

Kurtz, E. 1971. Interpretationen zu den Logos-Fragmenten Heraklits. Hildesheim.

Marcovich, M. 1978. Heraclitus. Florence. Orig. pub. 1967.

Miller, E., ed. 1851. Origenes philosophoumena sive omnium haeresium refutatio. Oxford.

Paquet, L., M. Roussel, and Y. LaFrance. 1988. Les présocratiques. Bibliographie analytique (1879-1980). Montreal.

Reinhardt, K. 1959. Parmenides und die Geschichte der griechischen Philosophie. Frankfurt. Orig. pub. 1916.

Roussos, E. N. 1971. Heraklit-Bibliographie. Darmstadt.

Schäfer, G. 1902. Die Philosophie des Heraklit von Ephesos und die modern Heraklitforschung. Leipzig.

Schuster, P. 1872. Heraklit von Ephesus. Leipzig.

West, M. L. 1971. Early Greek Philosophy and the Orient. Oxford.

Wilamowitz-Moellendorff, U. von. 1895. Euripides, Herakles I. 2nd ed. Berlin.

Zeller, E. 1961. La Filosofia dei Greci. Ed. and trans. R. Mondolfo. 4 vols. Florence.

Zeller, E., and W. Nestle. 1920. Die Philosophie der Griechen in ihrer geschichtlichen Entwicklung, 6th ed. Darmstadt. Orig. pub. 1855. In English as A History of Greek Philosophy: From the Earliest Period to the Time of Socrates, with a General Introduction, trans. S. F. Alleyne, London, 1881.

Fragments

[TN: The English translations are from Ancilla to the Pre-Socratic Philosophers, by Kathleen Freeman (Cambridge, MA: 1962), “Hêracleitus of Ephesus,” 24–34. The traditional interpretations reflected in these translations were often modified by Jean Bollack.]

Footnotes

[ back ] * Originally published as “Réflexions sur les interprétations du logos héraclitéen,” in: Jean Bollack, La Grèce de personne: les mots sous le mythe (Paris, 1997), pp. 288–308.

[ back ] 1. J. Bollack and H. Wismann1995 (hereafter B-W).

[ back ] 2. Roussos 1971, which included titles published up to 1970, has been completed by two collective works: De Martino, Rossetti, and Rosati 1986 (alphabetical catalog of titles, with Complementi 1621–1969 for the earlier periods) and Paquet, Roussel, and LaFrance 1988 (see “Études particulières” IV: “Héraclite” [pp. 444–555, nos. 1775–2436] for brief summaries of the content). The number of titles listed in De Martino, Rossetti, and Rosati is higher than all the publications recorded by Roussos since 1499: “A truly surprising result” (7). It remains to be seen how to generate discussion on the conditions of communication among the authors.

[ back ] 3. Barnes 1979, Chap. 4 (“The Natural Philosophy of Heraclitus”), does not discuss the general thesis of our book [B-W] and thus takes no account of its interpretation of the Fragments. He cites the title in his bibliography (349), setting it in a different sphere of scientific interest from his own (no doubt in another cultural tradition): “Idiosyncratically French accounts may be read in … ”; a list of names follows. Would the author be able to say what he means by “French”? B-W was published in France.

[ back ] 4. [TN: The Greek texts of the numbered Fragments, followed by Kathleen Freeman’s English translations (Cambridge, MA, 1962), are reproduced at the end this chapter, pp. 243–245. The translations incorporated into the text are based on Jean Bollack’s French translations, which differ significantly at many points from Freeman’s.]

[ back ] 5. “Der Streit ist der Vater und Herr aller Dinge, das allgemeine Recht und die Ordnung der Welt” (“Strife is the father and the master of all things, the universal law and ordering of the world”): Zeller-Nestle 1920:655, based on frags. 53 and 80; see Zeller 1961, 1:101–105. This is a common position, taken to the extreme by Gigon (1935:116): “Es ist überall, allgemein gültig, allgemein wahr” (“it is universally established, it is universally true”) (against Hesiod Works and Days 276–279); “Die allgemeine Krieg ist recht” (“universal war is just”). A divinization of war, which is also a known fact: “Wir stehen … auf dem Boden historischer ionischer Wirklichkeit” (“we stand … on the ground of the historical reality of the Ionians”) (119). Gigon goes as far as to write: “Der Kriegertod als solcher entspricht eben allein dem Logos, der ‘Krieg’ heisst im Gegensatz zum ‘Strohtod’ der bequemen allzuvielen” (“only the warrior’s death can respond to the truth of the logos, called ‘war,’ to distinguish it from the ‘death on the straw’ of the lazy masses [according to frag. 24]”) (120). Again (but less violently), in Gigon 1945:210: “Die Wirklichkeit, mit der wir rechnen müssen, ist der Krieg” (“the reality we have to deal with is war”). A universal principle or an empirical truth? Kirk adds to the discussion (1954:248): “All spheres of life.” And likewise, Conche (1986:440), on frag. 80: “[La guerre] est coextensive à toute la nature … et, dès lors qu’elle est le grand phénomène naturel, elle est normale et fatale” (“[War] is coextensive with all of nature … and once it is the great natural phenomenon, it is normal and inevitable”). Kahn (1979:209) limits the significance of frag. 53 to “the destiny of mankind.” (Conche, on the other hand, takes πάντων as a neuter [“of all things”]; however, it is masculine; see B-W:185).

[ back ] 6. The logos is the representative in speech of the truth of War (Gigon 1935:18; war accounts for the style of Heraclitus’ work). The logos is just an abbreviation for the book, that is, “the eternal truth.” A radical position: “eine Lehre, ein System steht am Anfang” (“a doctrine, a system, is there at the beginning”) (5). “Krieg … ist für Heraklit evident, die evidente Wahrheit von Frag, 1. Was er zeigen will, ist die Entwicklung des Gegensatzes in der Welt” (“For Heraclitus, war … is the obvious truth of fragment 1. What he wants to show is the development of the opposites in the universe”) (p. 25); see also 60–61: “Der Logos … ist … eine Abkürzung für die ewige heraklitische Wahrheit” (“logos is an abbreviation for the eternal truth of Heraclitus”), that is, “die Paradoxie des Kosmos,” the paradoxical identity of fire and the universe, and so on.

[ back ] 7. See Kirk and Raven 1957:195: “The total balance in the cosmos can only be maintained … if there is unending ‘strife’ between opposites,” and also: “The logos or proportion remains the same—again it is the measure and regularity of change, this time of large-scale cosmological change, that is stressed” (201); cf. Kirk 1954:402-403: the distinctive aspect of the logos that is fire is by nature kinetic; it ensures the regulation of cosmic movements.

[ back ] 8. Along these lines, Zeller’s introduction, which is more inductive, culminates in a spectacular presentation where everything is swept along by movement (see Zeller-Nestle 1920:806: “everything becomes and is not, caught in the movement of the life of nature”). It is the flux of Cratylus and Plato, and the antithetical Becoming of Heraclitus—Being of Parmenides (see Hegel’s introduction to Phenomenology of Spirit [1977: §12]). One denies the principle that the other conserves.

[ back ] 9. It is the reason for heimarmenè [TN: the personification of fate or destiny] to be considered as the global law, which cannot be directly identified with the physical process it encompasses. See Schäfer 1902, summarizing the definitions of P. Schuster, M. Heinze, and others.

[ back ] 10. Barnes (1982:59, ad frag. 1) is firm in his choice of “account:” “The noun logos picks up, in an ordinary and metaphysically unexciting way, the verb legein; it is wasted labor to seek Heraclitus’ secret in the sense of logos.” He stresses the “clear exposition of the right view” of West (1971:318n7), whose Early Greek Philosophy and the Orient places him in opposition to Guthrie, Hölscher, and Marcovich (see esp. West 1971:124–129). This is what Burnet wrote more than one hundred years ago (1930:133n1): “The logos is primarily the discourse of Herakleitos himself.” But Barnes (p. 59) is obliged to relate the “account” to the law of nature: “everything happens [objectively] in accordance with the account,” which explains how phenomena occur (objective) and what is “the essential nature of each thing.”

[ back ] 11. Kirk (1954:39) proposed “formula of things” as an approximation of logos, where the idea of measure that he considers essential for the organization of things and the world is “implicit.”

[ back ] 12. In opposition to the translation of logos with “reason” in the Hegelian tradition, cf. Schuster (1872:19): “the revelation that nature offers us in a language that we can understand”; he opts for “discourse” (20).

[ back ] 13. See Heidegger (1979:292): “[Logos] ist die ursprüngliche, Ursprung verleihende, im Ursprung einbehaltende Versammlung als das Wesen des Seins selbst” (“the logos … by which Heraclitus names Being itself, the One that unites all that is; the logos is the original structure, that imparts the origin, that upholds in the origin, as the constitutive characteristic of Being”).

[ back ] 14. Through the word of Being, man is torn away from daily distractions, brought back to himself, to the logos within him: Brecht 1949:37–45. These are the tones of a homily, so frequent in books on Heraclitus, up to and including Conche.

[ back ] 15. This is true for Marcovich (1978), who often resorts to this principle to explain or excuse incoherence (see ad frag. 1, for the non-distinction in archaic thought between the subjective aspect, discourse, and what it expresses, the truth); but the same is true of Kahn. It is a way of objectivizing the unresolved contradictions of interpretation, as is done with the principle of ambiguity.

[ back ] 16. The reuse of certain elements of discourse that have been analyzed is often seen as a component of the battle in favor of a new doctrine: “eine offene Parodie der Prädikation des Zeus” (“an obvious parody of the celebration of Zeus”); “die bewusste Verdrängung des Zeus durch das eigene erkannte Weltprinzip” (“a conscious displacement of Zeus, by means of the universal principle he discovered”) (Gigon 1935:119, ad frag. 53). The new science is opposed to popular beliefs. However, the language of science offers the same analyzable matter as the others; it reveals its contradictory structure for the same reasons as do rites. When Heraclitus was linked to the Ionians for his physics, the repetitions were interpreted as proof of his links and the mark of an affiliation.

[ back ] 17. Aristotle’s text is reproduced among the testimonies (Diels and Kranz 1951 [hereafter D-K]), 22 A 4 = Rhetoric 3.5, 1407b 11–18. I do not deal here with Diels’ addition, line 13 (which Kassel 1976 does not retain). Within the narrow context, Heraclitus’ case and, by way of an enlightening example, “the beginning of his book,” illustrate the difficulty of the reading in the absence of a clear-cut division. A feature of the obscuritas: we do not know before analyzing the passage what the words in the sentence go with. In Heraclitus, this is a problem (ergon).

[ back ] 18. Sometimes scholars opt for “which always exists” (past, present, future); sometimes, with the stress on the verb “to be,” meaning “to be true,” they opt for “which is always true” (cf. Kirk 1954:35; Zeller 1961:21, and so on.

[ back ] 19. This is what we might call the dominant opinion: see Zeller 1961:21: “Il participio ἐόντος, che l’interpretazione tradizionale unisce con ἀεί (esistente sempre, ossia: eterno)” (“the participle ἐόντος, which the traditional interpretation links to ἀεί [existing always, or: eternal]”), citing Burnet, Snell 1924, and Gigon; or Kirk 1954:34: “modern scholars have for the most part concurred with the view of Hippolytus and Amelius that ἀεί qualifies ἐόντος” (Zeller, Diels, Capelle, Gigon, Verdenius). Later on, see, among others, Kahn 1979:97: “Although this account (logos) holds ‘forever’,” it is true that ἀεί is translated a second time with ἀξύνετοι: “men ever fail to comprehend” (see Gigon’s proposition of a double construction [1935:2ff.], picked up by Kranz; cf. E. Kurtz 1971:85, contested by Verdenius [Kirk 1954:34n1]); we must make a choice, and not only according to Aristotle, who was well aware of the stakes of meaning.

[ back ] 20. The choice often adheres to formal criteria (ἀεί should go with ἀξύνετοι). The predicative meaning of the demonstrative does not lead to the acceptance of its real semantic significance. It is applied to the content of the book, which the pronoun presents in its specificity: being what it is (“‘which is as I describe it’” [Kirk 1954:36]). The doctrine is assumed to be known through an earlier oral teaching (cf. Kirk), or it will become known in the following part of the work: “Doch für die Rede—den Logos, der dies [Folgende] ist” (“But for the discourse—the Logos which is the following”) (Hölscher 1985:14). To this more qualitative meaning, other scholars have preferred a stronger deictic function, and in support of this stance assumed that what Aristotle and Sextus (22A 16, §132) present as the beginning was, in fact, preceded by something else, without necessarily making the pronoun a predicate (the problem of translation arises with the epithet [τοῦδε]; see, for example, Kurtz 1971:83ff.: “although this discourse may be …”—namely, what will follow; likewise, Barnes 1982:59: “‘Heraclitus of Ephesus says thus’: what he says (logos) is this.”

[ back ] 21. Diels (cf. 22A 4 D-K) with Vettori’s correction of δεῖ στίξαι for Bekker’s διαστίξαι (with the manuscripts) prints <δεῖ> διαστίξαι (Gaisford, Roemer, etc.). Kassel (1971:145) believes that διαστίξαι is superfluous. The brevity of the expression seemed excessive (see Diels in his apparatus, before Kassel). Manuscript A gives us προτέρῳ; the presence of the awkward infinitive could lead us to consider the lesson: it would not be said that we cannot see what to link the word ἀεί with, but that one of the approved divisions is uncertain: “for attaching the word ‘always’ to what precedes (that is, ἐόντος) is not obvious” (τὸ ἀεὶ πρὸς προτέρῳ διαστίξαι).

[ back ] 22. In other words: ἄδηλον does not indicate an uncertainty from the start, as it does three lines earlier. The phrase here, after the quotation, comments on an interpretation considered evident, and ends up saying that it does not emerge obviously from the letter of the text. It merely offers a choice between two possibilities, and this confirms the general observation of the writing style. One cannot infer from the text that Heraclitus did not give an opinion. Aristotle affirms that the syntactical relationship does not appear immediately.

[ back ] 23. I cannot agree with Kirk (1954:34): “Aristotle himself suggested no answer to the problem.” The interpretation of Hippolytus and Amelius (apud Eusebium) is well known to Aristotle; he discusses it.

[ back ] 24. Nor can one attach ἀεί to the verbal group ἀξύνετοι γίνονται in relation to καὶ πρόσθεν ἢ … καὶ … in the next section (cf. Kranz in the apparatus of D-K).

[ back ] 25. Law and discourse (Gigon 1935:6ff.), if we wish to avoid the absurd (or again, Babut 1976:496: “the very voice of the logos,” and n. 110: “indissolubly the word [parole] … and the supreme Principle”). Resorting to the “principle” offers one of the answers, when the problem is raised, which is not always the case. But does the text not suppose a real listening (ἀκοῦσαι), that is, to speech? The aporia offers one of the many signs that persuade us to give logos the sense of “discourse,” apart from the content.

[ back ] 26. Comparison is sometimes made with fragment 72; the “strangeness” of men is found in the sentence (Fremdheit, often with a Heideggerian tint). See also Kurtz 1971:87ff. But do they listen as strangers (and superficial minds) to the word of the Being that they do not hear?

[ back ] 27. According to Kahn (1979:98), “once they have heard,” it is too obscure and too difficult for the public; but “even before” remains disconcerting. Whence the passage from actualization in speech (which is the single meaning for Barnes, but he does not discuss the problem; one wonders why) to the content of the discourse, and to the hypostasized unity. The aporia remains intact as long as Heraclitus is supposed to have called eternal a truth that is new and his alone.

[ back ] 28. The demonstrative, in its richest sense (see above) is repeated, κατὰ τὸν λόγον τόνδε, to be isolated and separated from any utterance that might fix its meaning. Obviously the phrase γίγνεσθαι κατά, as what follows, with κατὰ φύσιν, shifts the claim of a unitary speculative explanation that interpreters extend to Heraclitus, with the same words, as they are used by the authors of systems. The point of reference appears to be even more different.

[ back ] 29. It is the whole of human activities: καὶ ἐπέων καὶ ἔργων; see Reinhardt 1959:218; Kurtz 1971:90, and so on.

[ back ] 30. The relation between τοιουτέων and ὁκοίων reveals the identity of the practices and experience, expressions of a language that has not been analyzed because it is applied. It is erased with the contrast (which is a break): “Whereas I, for my part …” (Marcovich 1978:6 and 9: “perhaps an Ionian idiomatic expression”). When it is formally maintained, the difference is related to the truth, sought for unconsciously but unknown (see the translation of κατὰ φύσιν: “according to its true nature” [Reinhardt 1959:218]); the opposition is then introduced in the participle πειρώμενοι, Hölscher 1985:14 “ob sie gleich ihre Erfahrungen machen …” [“if they immediately have the experiences …”]), as well as in Kirk 1954:33 (“‘even when they experience …’”).

[ back ] 31. Interpreters of Heraclitus do this when they see in Zeus a figure of “fire-life” (see Guthrie 1962, 1:463), and many others; the etymology is contested by Marcovich (1978), who takes Ζηνός to mean Διός.

[ back ] 32. The “name” uncouples the appellation from the god through “life” (Ζηνός). It is a restriction (μοῦνον), which, as such, is contradicted by the usual designation of Zeus, who is the righter of wrongs. Kraus’s analysis (1987:32ff.) makes the name of Zeus, according to a theological interpretation, the principle of the coincidence of opposites (life and death); but the “name” is just one of the poles. The understandings intersect: “to be named alone”—does this refer to one of the poles (life as opposed to death), or to the two together (life and death)?

[ back ] 33. [TN: Bollack is referring here to Heraclitus’ fragment 48 on the bow: “the bow is called life, but its work is death.” This is a pun, where the word at issue is bios, which, depending on the accent, can mean either bow or life.]

[ back ] 34. It is not easy for the apostles of transcendence to justify use of the term that means “knowledge” or even “competence” in the language, for this knowledge that is divine in nature, outside the world or the “all”; nor is it even easy, if the aspect is neglected, to justify resorting to the word “wisdom” (das Weise, sagesse); the meaning is doubtless anachronistic. Does “wisdom” (for humankind) consist in being aware that a “wisdom” exists beyond human reach, according to the proposed interpretations of fragment 50? See below, note 46.

[ back ] 35. In the order of ontological hypostases, some scholars have constructed, in defiance of syntactical possibilities, in my view, ἕν as the subject of the clause, and μοῦνον as a determination of the apposition τὸ σοφόν: see D-K or Marcovich 1978:445: “one (being), the only (truly) wise, is … ,” all the categories are combined; Kahn (1979:267) ties μοῦνον to ἕν in his translation: “The wise is one alone, unwilling … ,” but weakens the impact of his choice by resorting to ambiguity (268); Hölscher 1985:16: “Eines ist das Weise: allein (pro aduerbio) es will nicht …” is not easy to understand, compared to “es will nicht allein—und will doch … ,” (Hölscher 1968:132ff.; see n10: “μοῦνον concerning λέγεσθαι, required by the meaning as well as the rhythm of the sentence”; and Hölscher 1968:n47. μοῦνον necessarily leads us to consider the “name” in its apodeictic capacity. If the One (and the Wise One) refers to a metaphysical principle, the contradiction must take note of an arbitrary passage from one order to another: philosophy (a transcendent metaphysical principle) against religious belief (Zeus), according to Diels 1901:10, or Marcovich 1978:446; similarly, Kirk 1954:392–393: “unrivalled wisdom” versus “traditional religion” (cf. D-K); the logos, as the unity of opposites (life and death) versus Zeus = life, according to Guthrie 1962:463; the cosmic principle is life, but it is more than that: “life and death are two sides of the same coin,” according to Kahn 1979:271 (more or less like Guthrie).

[ back ] 36. Apart from fragment 32, see the use of the same formula applied to the analysis of an opinion on physics, fragment 41, and the definition of knowledge, indicated by the neuter σοφόν, as absolutely separate from the learned discourses (λόγους) in fragment 108, also with this variant without ἕν (σοφόν ἐστιν), the application to analysis of the constituent discourses.

[ back ] 37. The analysis cannot be condensed; it remains specific. Doxography, whatever its origin, is disarmed when faced with Heraclitus, who cannot be summarized or translated, in contrast to other pre-Socratic systems of thought, which can be re-constituted with great precision by the doxographers.

[ back ] 38. See Gigon (1935:44): “Das ἕν πάντα εἶναι ist der Kern der Heraklitischen Kosmologie in seiner abstraktesten Form” (“The ἕν πάντα εἶναι is the nub of Heraclitean cosmology in its most abstract form”); the One is just a special case, an “abstraction” of the theory of opposites: “The unity of opposites and the community of the logos provide the initial clues for interpreting this extraordinary claim … The rest of our commentary will be an exegesis of this proposition” (Kahn 1979:131).

[ back ] 39. A transition to the metaphysical plane for Hölscher (1985:20): unity transcends plurality, as god does (Hölscher, following Reinhardt, believes that through a reversal of the succession of the two thinkers, the concept of unity was prefigured for Heraclitus in Parmenides’ Being); the formula expresses the content of the logos; and Babut 1987, in accordance with Conche, concludes (ad frag.108, Conche 1986:329) that: “the unique god … is transcendental [to the other gods], as well as to men.” Elsewhere (Hölscher 1985:28), the unique transcendental Wisdom is that of “nature” (as opposed to man) etc.

[ back ] 40. For the antithesis of the “I” (who speaks) the word itself, Kahn (1979:130) substitutes: I-you others (as well as me), listening “to the discourse within your soul,” which contradicts the interpretation of fragment 1, even if there one gives λόγος an objective meaning (98): how can anyone perceive in his soul what no one else has ever perceived? The message has not been put forward: he is addressing those who have not yet heard it. Universality remained to be revealed. We go round in circles.

[ back ] 41. εἶναι: Miller, in the Hippolytus edition (1851; “universally accepted” according to Kirk and Marcovich, who cite, however, several references to defenses of the text of Parisinus: Bernays, Gomperz, and others) The understanding of εἰδέναι does not, in fact, depend necessarily on the construction of the sentence; I refer here to the defense of the transmitted text in B-W:175–177.

[ back ] 42. Recurrence (see n. 36) obviously does not provide clues for the classification of the aphorisms in the book (contra Hölscher 1985:16). I should note in passing that the observations I develop here express my disapproval of the attempts at reconstitution according to the contents or the progression of the work (Marcovich, Kahn, Conche, or Hölscher, following Bywater, Schuster, and others; many authors have accepted the principle, giving up attempts at application; see Gigon 1935:11, or Guthrie 1962:427). Diels (1901:viii) was more correct in admitting that the aphorisms did not have the coherence of a system. He underestimated the development and the work of reflection; the principle he upheld was good, with the distance he nevertheless gave to the observation. The groups closed ranks, and obscurity is linked to this closure.

[ back ] 43. See B-W: 305-307, for fragment 108. Of all the fragments, this is the one where interpreters have found transcendence to be most clearly enunciated, with σοφόν and πάντων κεχωρισμένον (understanding of one is deduced from the other), and where it is the least clear, if the least makes any sense, with separation applying to language as such, in all types of discourse. In the usual interpretation, Heraclitus claims to go beyond the intuition of the divine of which men may have had an idea before him (in the ambiguity raised in principle by Kahn 1979:115, his principle would be at the same time, with πάντων masculine, unknown to men, and separated from all things, in the neuter). The conditions of understanding are more concerned with the gender of ὁκόσων, which should be discussed first (”none of the discourses, οὐδείς, whatever their object”); we should return to the construction that links οὐδείς to ὁκόσων , and make it masculine.

[ back ] 44. See B-W:154–156. Kahn (1979:321n204) believes we should eliminate the form ὁτέη for linguistic reasons; he refers to Bechtel 1924, 3:171. Diels (1901:11) had mentioned a “conscious archaism,” with reference to Parmenides’ form: Bechtel notes that there is no trace of a feminine –tea (“I have some difficulties”). Kahn’s conclusion (1979) that there is simply no such thing as a feminine stem for τις (= Latin quis) might not apply to this secondary formation, fitting Attic ἥτις and Attic ὅτου, next to οὖτινος. Diels (1897:90, ad 28 B 8, 46) had proposed τεος as an analogical formation that would justify οὔ τεον (= οὔ τι). Bechtel (1924:169) accepts this as a possible invention of the poet (because of the antithesis οὔτεον-τὸ ἐόν) on τέου, τέῳ. Since one has ἥτις next to ὅστις, Heraclitus may have used an Ionian form, by analogy with the oblique forms (τέο or τέῳ), ὅτεος, ὁτέη (with the first element undeclinable). Nothing permits us to eliminate the form, even if it justifies the meaning in the context. (See the commentary in B-W:155, 2, 4.)

[ back ] 45. If the consensus is based on the quite uneven penetration by the Spirit, the members of this community end up having the pleasure of saying what the logos says “in them” (Kurtz 1971:106n97). The thought is familiar. To avoid the idea of general consent (the listeners’ agreement among themselves, e.g., Conche 1986:27—in manifest contradiction to frag. 108, however), the agreement has often been limited to just listening to the logos (D-K; Kirk 1954:68; Kahn 1979:131; Hölscher 1985:17, and so on).

[ back ] 46. Marcovich (1978:116) finds it useful to make clear that ἕν is the predicate.

[ back ] 47. See above, note 34. τὸ σοφόν denotes “the highest degree of knowledge” (Kurtz 1971:108); thus the formula stands out from all the others, explaining the order of the universe.

[ back ] 48. See Kahn’s discussion (1979:131: he discusses a “monism,” distinguished from the assimilation of fire into air proposed by Anaximenes, according to Aristotle’s Metaphysics 1.3, and that one might, following this point of view, qualify as Milesian. It would be another path leading to the key to its meaning.

[ back ] 49. See Gigon 1935:44.

[ back ] 50. Gigon 1935:43.

[ back ] 51. [TN: Freeman’s translation of fragment 10: “Joints: whole and not whole, connected-separate, consonant-dissonant.”]

[ back ] 52. See Kahn 1979:285.

[ back ] 53. See the paraphrase he proposes for frag. 10 (Kahn 1979:286).

[ back ] 54. 242D–E (= 22A10 D-K).

[ back ] 55. The significance of Plato’s doxographical evidence is strongly undermined by Kahn (1979:316, n. 156); the unity of opposites can be illustrated by Plato without this reference to the cycle: “Heraclitus does not need this periodic pattern.” So what differentiates Heraclitus from Empedocles? We should point out that: 1. the passage in the Sophist is an ordering of the fragments under discussion; 2. there is no periodicity to be found in Heraclitus; 3. the aphorisms connected with cosmology examine the Milesians’ system.

[ back ] 56. See Barnes (1982:61), who argues from fragments 30 and 90 that Heraclitus constructed with the prime stuff of fire a “monism” that represented “a physical science of a standard Milesian type.” That would be true if this tradition had not been attributed to Heraclitus too, as it was to Anaximenes. If it is not, then fire no longer has the function of a universal substratum; see B–W:154–156, 264–267.

[ back ] 57. Cf. Kurtz 1971:200; why does he raise the problem of the creator?

[ back ] 58. Reinhardt 1959:176n.1, based on Wilamowitz 1895, ad line 1106; Gigon (1935:55), who however eliminates the problem of creation (κόσμον ἐποίησε = διεκόσμησε); Kirk 1954:311, and so on. Kahn dismisses the stylistic expedient, which he relates to the perfection of the order that we have seen expressed in κόσμον: the natural world, neither god nor man has made it; it made itself. “Thus the cosmological idea begins to emerge …” (1979:134; see also 135).

[ back ] 59. Iliad 1.70.

[ back ] 60. Clement of Alexandria, who alone gives the fragment in its entirety, omits the demonstrative (Stromata, V, 9, 59, 5); it is there in Simplicius and in Plutarch, where τὸν αὐτὸν ἁπάντων is missing, often eliminated by critics, as what follows after ἐποίησεν (summarized with ἀλλ’ ἦν ἀεί in Simplicius); see B-W:131. The demonstrative, if it were included, would not say: “such as it is before our eyes” (Kurtz 1971:200; and see Kirk 1954:314: “the addition of τόνδε is important, since it obviously limits the κόσμος to that which we experience”) nor the constitution or the state in which it is found (Reinhardt 1959:175ff; cf. Hölscher 1985:18: “Diese Welt [wie sie ist] …” [“This world (as it is)”]); it would be rather “such as I say it is” (in line with the structure of the discourse).The determination by τὸν αὐτὸν seems to exclude one of the two citations and leads us to opt for Clement’s coherent version.

[ back ] 61. By ἁπάντων we understand men living in this world or everything that populates the world (cf. D-K), to accentuate the concept of unity: what all beings have in common; the reason for this world unity, in the context, does not appear (it is easy to understand why that part of the sentence has often been rejected). We should understand it as: τὸν αὐτὸν ἁπάντων τῶν κόσμων. In any case, it is still the same structure (cf. B-W: 132).

[ back ] 62. Whatever Hölscher has to say (1985:1), in keeping with Kahn, the thought is not more obscure than the style, as it is understood. Heraclitus does not abandon logic, he does not resort to enigma to make the unutterable comprehensible (“his listeners cannot follow a plain tale,” Kahn 1979: 270). The paradoxes are precise and significant. “Obscurity” concerns a freedom of composition, and the difference in the language compared to set-piece parlance. It serves as a guide, avoiding confusion. Clarity lies at the end.

[ back ] 63. Barnes (1982:58, 80–81) rejects the characterization of Heraclitus as oracular and enigmatic, but he accepts the metaphysics of flux, of opposites, and of unity (60), which fits in with science (“he offered a philosophy of science which exhibits an admirable articulation,” 81). Rejecting the “enigma,” which is perhaps oracular in the writing process, he rejects the deciphering of the fragments that support abstraction and a logical transposition that is not established by the reading of the text. Obscurity, left in place, then becomes really obscure; it is perceived as incoherence, indicating a relative primitivism (“his account of the world is fundamentally inconsistent,” 80). This “enlightened” arrogance masks the weaknesses of the critical approach. The thought is historically determined; it is thus “archaic,” obviously; but it has the coherence that the author gave it, and it requires analysis before it is judged, and then compared to Spinoza and other philosophers. Otherwise, we are groping in the dark.

[ back ] 64. “Heraclitus is after all a ‘spokesman’ for the logos,” (Kahn 1979:130): it is the logos that speaks, and Heraclitus at the same time, in its place, “an attempt to make this larger discourse audible to a few, at least.” Since the aporia is not formulated as a hermeneutic impasse, or is not accepted, we are obliged to accept the compromise.

[ back ] 65. If the logos has the same structure as the law that rules Becoming, as is often asserted, it is because the analysis of assertions about the laws is made through the structure of the discourse. The two levels hang together thanks to this radical distinction (saying, and what is said).

[ back ] 66. Guthrie (1962:413) discovers yet another side of Heraclitus, his religiosity; see also 1962:414: “Many things in the fragments suggest the religious rather than the philosophic teacher.” Babut (1976:496): “Le porte-parole ou l’instrument privilégié d’une révélation qui s’apparente à la parole d’un dieu” (“The spokesperson or the privileged instrument of a revelation that is like the word of a god”).

[ back ] 67. “He that hath ears to hear, let him hear”: Guthrie (1962:413) attributes this idea to Heraclitus, citing Clement (Stromata 5.14.718), “who actually compares Fragment 34 with this saying of Jesus” (Guthrie 1962:n. 2).

[ back ] 68. Guthrie 1982:414; Hölscher 1985:1, and others.

About

Fellowships

The Library

Programs

The Art of Reading: From Homer to Paul Celan

17. The Heraclitean *Logos**

Works Cited

Fragments

Fragment 1

Fragment 2

Fragment 30

Fragment 32

Fragment 41

Fragment 50

Fragment 53

Fragment 90

Fragment 108

Footnotes

The Center for Hellenic Studies	Contact Us
	Phone: 202-745-4400
3100 Whitehaven St NW	Fax: 202-797-8650
Washington, DC 20008	outreach@chs.harvard.edu

The Art of Reading: From Homer to Paul Celan

17. The Heraclitean Logos*

Works Cited

Fragments

Fragment 1

Fragment 2

Fragment 30

Fragment 32

Fragment 41

Fragment 50

Fragment 53

Fragment 90

Fragment 108

Footnotes

17. The Heraclitean *Logos**