A river of difference

You can't read the same fragment twice

This week I’ve been reading Kuhn’s Last Writings. One thing he proposes, in line with comparatively few in the history of Western philosophy,¹ is that differentiation precedes identity. In other words, we recognize faces, sounds, and so on, on the basis of a set of differences and not a set of similarities.

In Kuhn’s understanding, difference provides recognition through distinguishing features. We distinguish faces from other faces, and distinguish sounds from other sounds. Identity comes first from this act of distinguishing between things, and not by the features of the things themselves. Thus difference allows recognition and recognition provides identity — the constellation of what we regard as real — after which inference and reason can operate.

This undermines any philosophical position that makes static identity “basic” or “innate” (e.g., Plato’s forms, Aristotle’s logic, and so on). It is not being that is the basis of recognition. Instead, it is difference — or change — or becoming which are the basis of recognition, and therefore (latterly) of cognition.

“Being” and “Becoming”

“Hang on,” you might be saying. “This sounds familiar. Weren’t there pre-Socratic debates about this, between Heraclitus, who advocated for flux or becoming, and Parmenides, who advocated for being?”

Yes. Well-remembered. Let’s have a look at Heraclitus in particular, since his viewpoint supports Kuhn’s.

If you know of Heraclitus (~500 BC), you most likely know that he asserted that “you cannot step twice into the same river twice.” This is usually interpreted to mean that, because the river changes constantly, it is a different river each time you step in. In other words, he questions the stability of the identification of any given river.

But did Heraclitus actually say this?

It turns out to be impossible to tell. The most famous statement about flux in history has itself been in flux for the whole of its history.

This textual uncertainty ironically embodies the very principle Heraclitus, so far as we can establish, advocated. The flow of this river of thought seems to tell us something about how meaning and identity evolve over time, so let’s take a brief look at the history of difference.

Phrased in the famous way, i.e., that “you cannot step into the same river twice,” Plato (~428–347 BC) attributes the words to Heraclitus, in Cratylus, 402a (~388 BC). So if you know that quote, really you know Plato’s description of Cratylus, not Heraclitus himself. Cratylus (mid 5th Century) was a radical follower of Heraclitus, who took Heraclitean flux to new levels, and who probably influenced the young Plato.

Plato, Cratylus 402a, C.D.C. Reeve translation:

SOCRATES: Heraclitus says somewhere that “everything gives way and nothing stands fast,” and, likening the things that are to the flowing of a river, he says that “you cannot step into the same river twice.”

HERMOGENES: So he does.

(I’ve emphasized flow because it’s the same word as in another of Heraclitus’ famous but possibly apocryphal sayings, made popular by Plato: “Everything flows.”)

But what can we get from the paltry fragments that remain from Heraclitus’ own writing?

A waterfall near Lake Mavrolimna, Polymnio, Greece.

First Fragment

Here’s Heraclitus’ fragment, fragment 12 in Robinson’s translation (1987):

As they step into the same rivers, different and <still> different waters flow upon them.

The same fragment appears as fragment L in Kahn’s translation:

As they step into the same rivers, other and still other waters flow upon them.

This sounds sort of similar, doesn’t it? But what happened to the bit about being unable to step in twice?

Note also that this quote is disputed. From Kahn’s commentary, we learn that the above is quoted in Cleanthes (~330–230 BC), who is contrasting Heraclitus’ view of the psyche with Zeno’s (~334–262 BC). Zeno and Cleanthes were the first two heads of the Stoic school. Oddly, given Cleanthes’ usage, it contains no reference to the psyche.

Kahn comments on its lexical ambiguity:

The wording offers several oddities. There is the plural form of ‘rivers’: why is one river not enough? And there are four consecutive dative forms […], which can in principle be construed in either of two ways: (1) ‘into the same rivers, as they step’, as in my translation, or (2) ‘into rivers, as the same [men] step’. This ambiguity would have been avoided if either bathers or rivers had been referred to in the singular. Thus the two oddities of the sentence in fact coincide.

If Heraclitus has one defining feature, it is — ironically — deep ambiguity, even over and above the whole flux thing. And it turns out that we don’t even know if the quote, if it is authentic, means. It could either mean “As the same people step into rivers, other and still other waters flow upon them” or “As they step into the same rivers, other and still other waters flow upon them,” because of grammatical ambiguity. Both contain the idea of change in “other and still other waters” — but not the claim that it’s not the same river.

Second Fragment

Maybe Robinson’s fragment 12 was the wrong one. Here’s Robinson’s rendering of fragment 49a:

We step and do not step into the same rivers; we are and are not.

That seems right! Although I don’t remember that second bit.

But wait, it’s also disputed. Kahn, for example, rejects this fragment altogether:

Many editors have accepted D. 49a as a genuine quotation, but I can only see it as a thinly disguised paraphrase of the river fragments (L and LI), modelled on the contradictory form of CXVIII (D. 32), and influenced by the thought of XCII (D. 62): we are and are not alive. The text of D. 4 is in a class by itself: preserved in Latin by Albertus Magnus in the thirteenth century A.D., with no hint of a Greek source, it seems nonetheless to preserve a Heraclitean kernel; and I have used it with some hesitation in the commentary on LXX—LXXII.

So that one, too, may be an interpolation, and you can see that possibly some medieval confusion may have introduced something inauthentic. Or — perhaps stranger — introduced something authentic?

Third Fragment

Robinson’s Fragment 91a:

(a) [For, according to Heraclitus, it is not possible to step twice into the same river, nor is it possible to touch a mortal substance twice in so far as its state (hexis) is concerned. But, thanks to the swiftness and speed of change,] (b) it scatters things? and brings them? together again, [(or, rather, it brings together and lets go neither ‘again’ nor ‘later’ but simultaneously)], it forms and it dissolves, and it approaches and departs.

Kahn’s Fragment LI:

One cannot step twice into the same river, nor can one grasp any mortal substance in a stable condition, but it scatters and again gathers; it forms and dissolves, and approaches and departs.

From Kahn’s commentary on Fragment LI (D. 91):

Plutarch: [According to Heraclitus one cannot step twice into the same river, nor can one grasp any mortal substance in a stable condition, but by the intensity and the rapidity of change it scatters and again gathers. Or rather, not again nor later but at the same time it forms and dissolves, and approaches and departs.]

OK, so we’ve got it! Via Plutarch (46–120). He must be an authority, right?

Here’s Robinson’s commentary:

Whether this famous statement is meant to be read as Heraclitus’ own words or simply as Plutarch’s own summary of a longer statement we cannot know. Whatever the case, the question whether Heraclitus ever said anything like ‘you cannot step twice into the same river’ is a matter of deep dispute (see, eg, Reinhart 1916, 165, 207 nl, Gigon 106ff, Guthrie 441ff, Marcovich, Kahn).

If fragment 12 is genuinely Heraclitean, then, as several have argued (eg, Vlastos [338ff ], Kahn), the original thought was presumably something like: ‘One cannot step twice into the same river (fragment 91); (for) upon those who (do) step into the same rivers different and different waters flow’ (fragment 12); or, possibly: ‘Upon those who step into the same rivers different and different waters flow. (For that reason) one cannot step twice into the same river.’ This has a prima facie plausibility, but is ultimately unlikely, in that the causal clause is not in fact an explanation for the main clause constituted by fragment 12 at all. Fragment 12, spelled out, would presumably run: ‘Upon those stepping into the same rivers different and different waters flow; <but the rivers are none the less the same rivers, however many times, and however many different people, step into them>.’ Fragment 91a, spelled out, would presumably run: ‘You could not step twice into the same river; <for, given the constantly changing water, no river can ever be said to be the same at times one and two>.’ Now, while the two statements are admittedly not self-contradictory, since fragment 12 stresses the unity amid diversity that constitutes the universe (on the assumption, accepted by most modern editors, contra Plutarch, that the river-image is an analogical description of the universe) while fragment 91a stresses perhaps the sheer diversity of the universe, without reference to any putative unity that subsumes it, they do pull in such contrary directions that the use of one in explanation of the other risks begging the question.*

Here’s Kahn:

It is curious that the most celebrated and in a sense the most profound saying of Heraclitus, that you cannot step twice into the same river, is not unmistakably attested in his own words. It was already a famous saying in Plato’s time; and even before Plato, Cratylus must have been familiar with the paradox, since he tried to cap it with one of his own. Cratylus denied that you could even step in the river once, since you are changing too.

Occasionally I’ve heard modern commentators make the “you can’t step in the same river once” expansion, and assume that this is a new idea. But in fact, it’s almost as old as the quote itself, dating to probably 450 BC.

Kahn continues:

Thus the statement seems to go back to Heraclitus himself. But Plato does not give a verbatim quotation: ‘Heraclitus says, doesn't he, that all things move on and nothing stands still, and comparing things to the stream of a river he said that you cannot step twice into the same river’ (Cratylus 402A). Like the formula panta rhei ‘all things flow’, which occurs later in the dialogue, the remark about the river seems to be paraphrase rather than quotation.199 The citation from Plutarch in LI is similar to Plato’s except for the impersonal form. This is probably as close as we can get to Heraclitus’ own wording, unless we assume with several recent editors that the famous statement is simply a free rendering of L (D. 12). (Kirk (pp. 374f., 381), Marcovich (p. 206). Cf. Reinhardt, Par men ides, p. 207n.)

But if it is strange that Heraclitus should have expressed himself twice in such similar terms, it seems even stranger that Cratylus or some anonymous predecessor should have invented this formula, which would have been enough to assure Heraclitus’ immortality even if all his other works were lost. Hence I prefer to regard ‘One cannot step twice’ as an independent fragment, perhaps designed to complete L (12) by drawing an even more radical conclusion: since new waters are ever flowing in, it is in fact not possible to step into the same river twice.

This river of change has, ironically, led to Heraclitus’ static immortality, for a quote which we will probably never know whether he actually said or not.

The quote flows through centuries of paraphrases, interpretations, interpolations, and translations. The philosopher who allegedly claimed we can never encounter the same river twice can himself never be encountered directly — not even once.

To briefly bring back Kuhn, we don’t recognise Heraclitus by some “essence” or some list of defining features. Instead, we understand him as part of what Kuhn calls a “contrast set.” Kuhn notes that we typically learn “dog” in contrast to “cat,” even though they overlap in most ways that are easy to articulate in language. (If you want to experience this yourself, please play the game in last week’s post, A Word Game.)

We also normally learn Heraclitus’ thought as a part of a “contrast set.” His flux or “becoming” view is learned alongside Parmenides’ (~475 BC) opposing “being” view. Parmenides contrasts categorical states of “being” and “non-being” and proceeds to reason using these categories — precisely the way in which Kuhn claims (on the basis of psychological experiments) that perception does not operate, but often the way we naïvely assume that it might.

As we wade into this river of differences, we do not access any fixed entity, but an eternally evolving contrast set. For most of Western philosophical history, the Parmenidean view prevailed. Pythagoras, Plato, Aristotle, Plotinus, Aquinas, Descartes, Kant, Russell, and Chomsky are all, in different ways, attempting to operate within stably reified “categories.” But recent evidence from sociology, neuroscience, psychology, and even AI may undermine this perspective, suggesting that difference is more fundamental than identity, and that identity is built on the basis of difference rather than similarity.

If you felt this was a good post, please share it to someone else who might like it, let me know in the comments, subscribe, or support me on Patreon or Ko-Fi!

Sources

• Heraclitus. Fragments: A Text and Translation. Translated by Thomas M. Robinson (1991)

• Heraclitus. The Art and Thought of Heraclitus: An Edition of the Fragments with Translation and Commentary. Translated by Charles H. Kahn (1989)

See also:

1. The ones that come to mind from the past few centuries are Nietzsche, Saussure, James, Whitehead, Korzybski, Deleuze, Derrida, and possibly Wittgenstein, but if you know more please comment.