οὐκ ἐμοῦ, ἀλλὰ τοῦ λόγου ἀκούσαντας ὁμολογεῖν σοφόν ἐστιν ἓν πάντα εἶναι.
Having listened not to me but to the logos, it is wise to agree that all things are one.
01 DO NOT LISTEN TO HERACLITUS
The sentence begins by disturbing the ordinary arrangement between a philosopher and a reader. Here is a voice asking to be heard, and its first move is to redirect our hearing. Do not stop at the person who speaks. Attend to the logos.
This need not be a performance of modesty. Heraclitus still makes an exacting claim about wisdom. Nor can he simply remove himself: someone has chosen these words and placed this demand before us. The distinction is between the person who offers an account and whatever makes that account answerable to the world.
We can therefore bring two different questions to the fragment. What does Heraclitus believe? asks us to identify a position and attach it to its author. What is Heraclitus trying to make us notice? asks us to examine what the position is about. The first question matters for historical understanding. The second prevents the history of a thought from becoming its only interest.
Suppose a reader accepts a claim because an admired philosopher made it. The reader has learned whose claim it is, but has not yet encountered its grounds. Another reader may dismiss the same claim because the philosopher seems obscure or arrogant. Both responses can remain fixed on the speaker. B50 asks whether we can get further than either allegiance or rejection.
The opening does not give us a test that will immediately settle every disputed claim. It places a demand on our attention: the worth of what is said cannot be exhausted by the identity of the one saying it. That demand also applies to the commentary beside the sentence.
02 THE WORD THAT BECOMES SMALLER WHEN TRANSLATED
Logos is the word on which this redirection turns. Its semantic field includes word, speech, account, statement, explanation, reason, relation, measure and reckoning. These are possible directions of sense, not interchangeable labels. An account of an event and a numerical relation do different work even when Greek can use the same word in speaking of them. The LSJ entry records a range across contexts; it does not tell us to pour that entire range into this one sentence.
A translation is already an interpretation. If we choose “account”, we foreground something that can be given and heard. If we choose “reason”, an English reader may imagine either a faculty of thought or a justification. “Word” can make the object seem smaller than an account, while a capitalized “Word” can import a theological history that this fragment alone does not authorize.
It is similarly premature to replace logos automatically with “cosmic law”. That phrase may express an interpretation of what the account concerns. It should not silently decide the interpretation before reading has begun. The task is not to ban these possibilities but to make the choices visible.
Leaving logos untranslated has a cost too. An ordinary Greek word can acquire an air of sacred mystery when it becomes an unfamiliar term in English. Robinson explicitly challenges this practice in the study listed in the apparatus. Here retention serves a narrower purpose: it keeps a question open long enough for us to test possible meanings against the sentence. It is a working decision, not a claim that the word cannot be translated.
Ask what the word must do here. The logos can be heard. It can be distinguished from Heraclitus himself. Listening to it bears on a judgment about all things. Any proposed translation must account for those relations, rather than merely sound philosophical.
03 SPEECH OR STRUCTURE?
One way of reading the fragment takes logos to concern an objective, intelligible order. On this reading, the speaker draws attention to something that holds independently of his preferences. Wisdom requires recognition of that order, not loyalty to its announcer.
Another approach begins closer to speech: logos is an account, and an account becomes significant through what it makes intelligible. We hear words, but what we are asked to understand need not be confined to words. The contrast with Heraclitus would then distinguish the claim’s content and grounds from the personality of the claimant.
These approaches need not disagree about everything. An account can disclose an order, and an order can make an account true. Their disagreement concerns what the word names and how far the fragment takes us beyond the act of speaking. Robinson argues for “account”; Graham’s discussion emphasizes the relation between the message and the world it discloses. Neither should be converted into a unanimous scholarly verdict.
We should also resist making the sentence decide a familiar modern dispute between an entirely subjective language and a completely language-free reality. B50 does not supply that vocabulary or settle every question it would raise. What it does give us is a contrast within the act of listening.
There is something to hear that is not reducible to Heraclitus’ private voice. This is a modest stopping point. It leaves open whether logos is best named as account, structure or a relation between speaking and intelligibility. But it makes the opening instruction philosophically demanding rather than merely self-effacing.
04 TO AGREE WITH THE LOGOS
The next movement is from hearing to ὁμολογεῖν, homologein. “Agree” is a defensible translation, but it is worth delaying our usual picture of agreement. We often imagine two people comparing opinions and arriving at a shared answer. The fragment gives us a more unusual scene: listening to the logos, then acknowledging that all things are one.
The lexical range of homologein includes agreeing, saying the same and corresponding or coming into accordance. There is a resonance with logos that an English rendering will struggle to retain. Yet a resonance is not proof that the whole sentence is organized around a demonstrable pun. We should distinguish what the words allow us to hear from what we can establish about deliberate literary construction.
What would it mean to come into accordance here? Perhaps the listener’s account must answer to something beyond the listener’s conviction. Repetition is insufficient. I can say the same words as another person while failing to grasp what they say. Conversely, acknowledging a relation between things may require revising the words with which I approached them.
B51, with its bow and lyre, offers a nearby interpretive possibility: something can belong together through a tension that has not vanished. This is a connection between fragments, not an additional line hidden inside B50. The comparison invites us to ask whether agreement must mean the elimination of difference. It does not establish that B50 already contains a complete theory of harmony.
05 ALL THINGS ARE ONE
The final words, ἓν πάντα εἶναι, bring the sentence to its most compressed claim: all things are one. The plural “all things” remains in view even as the sentence asserts unity. What sort of oneness can hold without making that plurality unintelligible?
A quick answer would say that differences are only appearances. But B50 does not explain how appearances relate to reality, specify a single underlying substance or give an argument proving that plurality is an illusion. Such theories would have to be developed and defended; they cannot be treated as translations of these few words.
Another possibility is that unity names an order to which differentiated things belong. Consider, as an illustration of the distinction rather than evidence about Heraclitus, the parts of an argument. They form one argument because they have relations to one another. Making every premise identical would not make the argument more unified. It would destroy the differences through which it works.
That illustration does not explain the cosmos. It only loosens the assumption that “one” must mean “all the same”. B50 asks for a larger recognition, but leaves us to investigate what could warrant it. The fragment’s brevity should sharpen that investigation rather than license us to fill every gap with a preferred metaphysics.
Unity may therefore be a problem to understand rather than a feeling to cultivate. What connects the things being gathered? Which differences matter to that connection? What would count against the proposed unity? Without such questions, “all things are one” becomes a slogan that seems to include everything because it specifies almost nothing.
06 WHY ARE PEOPLE UNABLE TO HEAR IT?
The claim that the logos is common belongs to the context supplied by B2, not to the wording of B50 itself. B1 describes failures of understanding in relation to hearing. Read together, these fragments raise a difficulty: availability and understanding are not the same. Graham’s account of Heraclitus’ method discusses this contrast. Their modern numbering does not establish the sequence of Heraclitus’ lost work.
If something is common, how can understanding it remain rare? A common object does not guarantee a common quality of attention. Two people can face the same event while noticing different relations, accepting different assumptions and refusing different corrections. Possessing information is not yet knowing how its parts belong together.
This gives us an editorial way of approaching the problem, not a diagnosis of every person Heraclitus criticized. Perhaps the obstacle is less that the account is secret than that a listener continually returns it to private expectations. We hear the familiar in a difficult sentence and stop before the unfamiliar can disturb our categories.
There is a danger in this suggestion. A commentator can make disagreement look like mere failure to listen, protecting an interpretation from criticism. B50’s opening should make us suspicious of that manoeuvre. If the speaker’s authority is insufficient, the commentator’s authority is insufficient too. A claim to have heard better must remain answerable to the words and to reasons that others can examine.
Attention is therefore not passive submission. It includes the ability to pause, compare, object and return. An account that survives such attention earns a different kind of agreement from a sentence repeated because its author is famous.
07 A PHILOSOPHY OF LISTENING
We can make this fragment easier in a few moves. Replace logos with a familiar abstraction. Treat homologein as assent to a proposition. Turn the ending into an uplifting statement about universal unity. The resulting sentence may be smooth, but its smoothness has been purchased by removing the questions through which it becomes worth reading.
Difficulty is not automatically depth. An obscure sentence can be confused, and an ingenious commentary can supply connections that the text does not support. Close reading needs ways to test itself: an exact passage, explicit translation choices, alternative interpretations and a willingness to name what remains uncertain.
The purpose of preserving difficulty is to let the sentence change the questions we bring to it. At first we may ask for Heraclitus’ doctrine. We leave asking what makes an account worthy of agreement; whether intelligibility belongs to speech, the world or their relation; and how a unity can include differences that remain real.
A commentary should not replace the text. It should return us to it with better questions. Read the two lines again, watching the movement from the speaker, to the logos, to listening, to agreement. Then ask whether the last words have become clearer, or whether you can now say more precisely what you do not yet understand.
The purpose of commentary is not to stand between the reader and the text. It is to teach the reader where to listen.
Sources / Apparatus
Primary text
Heraclitus, Fragments. DK 22 B50; transmitted in Hippolytus, Refutation of All Heresies IX.9.1.
Supplied Greek compared word by word with Hermann Diels, Die Fragmente der Vorsokratiker, third edition, volume I (Berlin: Weidmann, 1912), p. 87, Herakleitos B50. The twelve Greek words, accents and breathings agree with the printed quotation. The display uses editorial line breaks and a final full stop outside the witness's quotation frame.
Original Paragraphos working translation supplied for this publication, not a quotation from a historical published translation.
B50 is a modern fragment identifier. Diels prints λόγου and credits Bergk in the apparatus against the abbreviated Paris manuscript reading associated with δόγματος. He also prints εἶναι, crediting Miller against the manuscript's εἰδέναι. The Folio follows that identified printed text; it does not claim manuscript unanimity. The surrounding Hippolytus framing and adjacent B51 were inspected on p. 87, not treated as Heraclitus' recovered original sequence. This is a comparison with a historical critical edition, not fresh manuscript collation or a survey of all later editions.
Publication authorization covers the Greek passage and original working translation supplied by the commissioning editor for this folio. No modern edition's apparatus or third-party translation is reproduced. This clearance concerns the supplied material, not a claim about rights in every edition.
Secondary reading & lexical references
- Hermann Diels, Die Fragmente der Vorsokratiker, third edition, volume I (1912), p. 87, B50 and apparatus.
Digitized title page and printed p. 87 inspected. The supplied Greek agrees with the quotation; its lineation and terminal full stop are editorial. The apparatus records the λόγου and εἶναι editorial readings. The historical chapter number is 12; the modern identifier is DK 22 B50. This check establishes an identified textual basis without removing the variants.
- Daniel W. Graham, ‘Heraclitus’, Stanford Encyclopedia of Philosophy (revised 2023), §§2, 3.2 and 5.
Consulted for the B1/B2 context, B50 and the unity of opposites. An overview of interpretive problems, not a substitute for a critical text.
- Liddell, Scott and Jones, A Greek-English Lexicon, s.v. λόγος.
Scaife ATLAS lexical entry consulted. Its range of uses across Greek literature does not establish that every sense is active in B50.
- Liddell, Scott and Jones, A Greek-English Lexicon, s.v. ὁμολογέω.
Entry consulted for agreement and correspondence, including its references to Heraclitus B50 and B51. Lexical resonance does not by itself prove deliberate wordplay.
- Thomas M. Robinson, ‘Heraclitus and Logos: Again’, ΣΧΟΛΗ 7.2 (2013).
Article consulted, especially its argument for account in B1, B2 and B50 and its objections to importing later meanings. Robinson also objects to leaving logos untranslated; our temporary retention is an editorial choice, not his recommendation.
- Jonathan Barnes, ‘The beliefs of a Pyrrhonist’, The Cambridge Classical Journal, note 38.
The online note was checked for the Hippolytus IX.9.1 witness and the dogmatos/logou textual issue. It does not constitute independent manuscript collation.
- Charles H. Kahn, The Art and Thought of Heraclitus: An Edition of the Fragments with Translation and Commentary (Cambridge University Press, 1979).
Further reading. Publisher bibliographic record checked; the book's individual passage discussions and textual apparatus have not been verified for this folio. No wording or page citation is attributed to it here.
