In this article I want to clarify the text of one of the short maxims assigned to Pythagoras in the ancient tradition, which are known as symbola or acusmata. Before I turn to the acusma in question, it is important to understand the context in which it appears. It occurs in Chapter 17 of Book 4 of Aelian's Historical Miscellany (ποικίλη ἱστορία). Aelian's work was written in the early third century a.d. in Rome, and is a ‘miscellaneous collection of anecdotes and historical material’.Footnote ¹ It consists of short chapters, usually a page or less long, that are for the most part independent of one another. Chapter 17 of Book 4 is about a page long and is devoted to the sayings and doings of Pythagoras. There is no particular connection between it and the surrounding chapters, and it is clearly meant to stand on its own. The preceding chapter (16) tells us that if we went to Callias for guidance he would turn us into drinkers, … if to Alcibiades, arrogant cheats, if to Demosthenes, orators, … if to Aristides, just men, … if to Socrates, wise men. The following chapter (18) recounts an anecdote about Plato's arrival in Sicily and his reception by the tyrant Dionysius the Younger. The chapter that I am concerned with, Chapter 17, begins by asserting that Pythagoras taught that his ‘lineage was superior to that of ordinary mortals’.Footnote ² This is followed by a list of superhuman acts and traits of Pythagoras, for example his ability to be in Metapontum and Croton at the same time and his golden thigh. The second two-thirds of the chapter are then devoted to the miscellaneous teachings of Pythagoras. Most of these take the form of the brief taboos and maxims known elsewhere as symbola or acusmata, among which is the text on which I want to focus here.

Another famous list of acusmata appears in Iamblichus’ On the Pythagorean Life, where Iamblichus, perhaps following Aristotle,Footnote ³ divides them into three types: the first type indicates what something is, the second what is best in a category, and the third what one must do and not do (VP 82). Aelian does not explicitly refer to this tripartite classification, but the acusmata that he presents are, in fact, grouped according to these three types. He begins by listing three examples that belong to the second category: what is most sacred, the mallow leaf; what is wisest, number; what is the second wisest, he who assigned the names to things. The chapter ends with examples of the third type: do not eat the heart of an animal or a white cock; do not go to the baths or walk on the public roads, since it is not clear that these places are pure. In between these two types of acusmata are what appear to be three examples of the first type, acusmata that indicate what something is. The first example is ‘earthquakes he traced back to nothing other than a meeting of the dead’, while the last is ‘the ringing that often falls upon our ears is the voice of the heroes’ (κρειττόνων).Footnote ⁴ It is the middle example that is the focus of this article. The text of all the manuscripts reads:Footnote ⁵

This seems to be nonsense. Iris has no particular connection to the Nile or to the earth. Accordingly, as early as Gessner's sixteenth-century edition of Aelian,Footnote ⁶ the text was emended to:

The emendation restores order in that Iris, who was identified with the rainbow, was indeed explained in several ancient texts as the reflection of the light of the sun on clouds. Thus, Pythagoras’ contemporary Xenophanes says: ‘that which men call the goddess Iris, this too is by nature cloud, that looks purple and red and yellow’ (DK fr. 32). This emendation has been accepted by all modern editors including Dilts in his edition of Aelian and Wilson in his Loeb edition of Aelian.Footnote ⁷ I want to argue that editors have been too quick to accept this emendation and that we should look at the text transmitted by the manuscripts more closely.

Gessner's emendation is supported by the identification of the rainbow as the light of the sun not only in Xenophanes (fr. 32) but also in the testimonia for Anaximenes (DK 13 A17–18). Moreover, in the doxography (Aetius 3.1.2 = Stob. Ecl. 1.27.2) some Pythagoreans are said to have explained the Milky Way as ‘a reflected image of the sun which bends its rays back to the heavens’ (κατοπτρικὴν … φαντασίαν τοῦ ἡλίου τὰς αὐγὰς πρὸς τὸν οὐρανὸν ἀνακλῶντος), and to this is added the comment ‘the very thing which happens to clouds in the case of rainbows’ (ὅπερ κἀπὶ τῆς ἴριδος ἐπὶ τῶν νεφῶν συμβαίνει). It is not clear whether the additional comment about the rainbow is to be ascribed to the Pythagoreans or whether it is a comment of some doxographer, but the use of the expression ‘rays of the sun’ (τοῦ ἡλίου τὰς αὐγάς) to explain the Milky Way and by extension the rainbow may have seemed to Gessner like a good basis for his emendation. Despite the initial plausibility of Gessner's emendation, however, serious problems remain. Most importantly, the emendation requires a substantial change to the text. ἡ γῆ τοῦ Nείλου has to be replaced with αὐγὴ τοῦ ἡλίου, which supposes that in transmission ἡ replaced αὐ and Νείλ replaced ἡλί, neither of which is a particularly natural corruption. Four letters are changed from the manuscript text; two are dropped and two are replaced by other letters and another two are reversed in order.

If we go back to the manuscript text, however, sense can be restored to the passage with the change of a single letter. The text is corrupt not in its second part but in the first part. The goddess in question is not Iris but Isis. It would be extremely easy for a Greek reader to misread the text and replace the more exotic Isis with Iris, who is more to be expected in the context of Presocratic philosophy. The acusma thus correctly reads:

Not only is it plausible to connect Isis, as an Egyptian goddess, with the Nile, this precise connection between Isis and the earth of the Nile is attested by Plutarch in his work On Isis and Osiris. Plutarch reports that the Egyptians ‘believe the earth to be the body of Isis, not all of the earth but so much of it as the Nile covers, fertilizing and uniting with it’ (366A).Footnote ⁸

Therefore, reading ‘Isis’ is much simpler palaeographically than the standard emendation. Which reading works better in the context in Aelian? As we have seen, the text in question is the middle of three examples. The first and third examples deal with what can be described as natural phenomena: an earthquake and the ringing that we sometimes hear in our ears. At first sight the standard emendation seems to fit well into this context, because Iris, understood as the rainbow, is precisely another natural phenomenon. However, closer examination of the context shows a serious problem. In the case of the first and third examples a natural phenomenon is explained not naturalistically but rather in terms of the supernatural. An earthquake is not the result of earth that becomes too wet or too dry, as some early Greeks supposed (Aetius 3.15), and the ringing in the ears is not explained in terms of the structure of the ear or the wind or some other natural factor. Instead, the earthquake is said to be the result of a meeting of the dead, and the ringing in our ears is the voice of some unseen heroes. Thus in both cases the natural phenomena are cited to support the existence of supernatural beings. The standard emendation, which explains the rainbow in purely naturalistic terms as the reflection of the sun on the clouds, does not fit with the other two examples. On the other hand, if we take the simpler emendation, replacing Iris with Isis, the pattern is restored. The natural phenomenon in question is the rich earth deposited by the Nile, whose fecundity is explained in terms of a supernatural figure, the goddess Isis. The proposed reading is thus preferable both on palaeographical grounds, in that it involves a much simpler and more plausible change to the text, and also in terms of the context in Aelian.

Someone might object, however, that it is anachronistic to assign such an assertion about Isis to Pythagoras. Griffiths points out that there is no direct evidence in Egyptian texts equating Isis with the earth and that the earliest text in the Greco-Roman tradition to make the connection is Varro (Ling. 5.57). Griffiths accordingly concludes that ‘Plutarch's statement derives from the assimilated cult of Demeter, a corn-goddess whose name was explained in ancient times as “earth mother” and … the idea of Isis as the earth of the Nile is a Greek superimposition.’Footnote ⁹ If it is a Greek interpretation of Egyptian religion, however, there is no reason why it should not go back to Pythagoras, and several good reasons why it should.

One of the clearest strands in the early evidence for Pythagoras is a connection to Egypt. Thus, Herodotus, writing in the second half of the fifth century, says that Pythagoreans agree with Egyptian practices in forbidding the dead to be buried in wool (2.81).Footnote ¹⁰ Moreover, at 2.123 Herodotus famously assigns the doctrine of metempsychosis to the Egyptians and then comments that ‘some of the Greeks followed this doctrine … as if it were their own’, where he must be referring to Pythagoras. In the first part of the fourth century Isocrates reports that Pythagoras visited Egypt and paid particular attention to the religious practices of the Egyptians (Bus. 28). Heraclitus (fr. 129) describes Pythagoras as practising enquiry beyond all others, and, in light of these passages of Herodotus and Isocrates, that enquiry may well have included study of Egyptian religious practices. Thus it is very likely that Pythagoras knew of Isis and her cult. In fact, Isis would have been one of the Egyptian deities of most interest to Pythagoras, since she was associated with death and rebirth; ‘she is “mother”, “wet-nurse”, of the dead, and brings about rebirth’ (R. Gordon, OCD ⁴ 746, s.v. Isis). Moreover, there is evidence earlier than Varro for the connection between Isis and the earth. The assimilation to Demeter that Griffiths mentions is found already in Herodotus (2.59 ‘in the Greek language Isis is Demeter’), and the identification with Demeter, who is closely tied to the earth, surely would have brought with it the identification of Isis with the earth.Footnote ¹¹ Thus, by the second half of the fifth century, Isis had been equated with the earth. It is quite plausible that Pythagoras with his special connection to Egypt was the key figure in bringing this identification to prominence; even if he was just availing himself of an identification already in existence, it is perfectly plausible that one of the Pythagorean acusmata should have said that Isis was the earth of the Nile.

Good scholarly practice calls on us to preserve the text handed down by the manuscripts with as little change as possible while producing good sense. The emendation proposed above changes only a single letter and produces a text that fits better into the context in Aelian than the standard emendation that calls on us to make much more drastic changes in the text. The standard emendation should thus be discarded in favour of the simpler change presented above. It is worth asking, however, why scholars were willing to accept Gessner's emendation with so little question. The answer may be that the standard emendation fits with the interpretation of Pythagoras that dominated in scholarship until the appearance of Burkert's epoch-making book in 1962 (revised English version, 1972). Up to that point Pythagoras had generally been viewed as a mathematician and natural scientist. The standard emendation, which has Pythagoras explaining the rainbow in purely naturalistic terms as the reflection of the sun's light on clouds, fits perfectly with Pythagoras the natural scientist. Since Burkert's work, however, the standard view has become that the early and reliable evidence for Pythagoras reveals that the core of his philosophy was a way of life derived from his expertise on religious ritual and the fate of the soul after death.Footnote ¹²

Since Pythagoras wrote nothing, some of the best evidence for the philosophy of Pythagoras and the way of life that he developed is found precisely in the acusmata or symbola.Footnote ¹³ These maxims were probably originally transmitted orally, as the name acusmata (‘things heard’) suggests. One group of Pythagoreans, known as acusmatikoi, emphasized unquestioning memorization of and adherence to these acusmata (Iambl. Comm. Math. 76.16–77.18). Heraclitus’ reference to ‘these things which have been written up’ as a result of Pythagoras’ wide enquiries in fragment 129 may indicate that written collections of them were already available in Pythagoras’ lifetime.Footnote ¹⁴ At any rate, such collections are likely to have been available by 400 b.c., when Anaximander of Miletus (the Younger) wrote a treatise explaining them (see Suda s.v. Anaximander = DK 58 C 6). In the fourth century, Aristotle reported a great number of them in his lost books on the Pythagoreans, which are the ultimate source of our most detailed surviving accounts of them in Iamblichus’ On the Pythagorean Way of Life (82–6; cf. Protr. 106.19–108.15), Aelian's Historical Miscellany (4.17) and Porphyry's Life of Pythagoras (41–5).Footnote ¹⁵ Consideration of a wider range of acusmata than are found just in Aelian shows that the acusma about Isis fits much better with the philosophy of the acusmata as a whole than the standard reading of it as giving a naturalistic explanation of the rainbow.

A number of the acusmata that fall into the first type identified by Iamblichus (VP 82—see above), the type that says what something is, have cosmological content. Here is a list of ten of them:

1. 1. What are Islands of the Blessed? Sun and moon. (Iambl. VP 82)

2. 2. The planets are the hounds of Persephone. (Porph. VP 41)

3. 3. The Big and the Little Bears are the hands of Rhea. (Porph. VP 41)

4. 4. The Pleiades are the lyre of the Muses. (Porph. VP 41)

5. 5. The sea is a tear of Cronus. (Porph. VP 41)

6. 6. An earthquake is a mass-meeting of the dead. (Ael. VH 4.17)

7. 7. Thunder is to threaten those in Tartarus so that they will be afraid. (Arist. An. post. 94b33)

8. 8. Isis is the earth of the Nile. (Ael. VH 4.17)

9. 9. The noise from struck bronze is the voice of one of the daimones shut up in the bronze. (Porph. VP 41)

10. 10. The ringing heard in the ears is the voice of the heroes (kreittones). (Ael. VH 4.17)

All of these ten fit a pattern in that they equate a feature of the natural world (for example sun and moon, the sea, an earthquake, ringing in the ears, the earth of the Nile) with some aspect of Greek religion and myth (for example Islands of the Blessed, a tear of Cronus, a meeting of the dead, the voice of the heroes, Isis). Some have seen in these acusmata evidence that Pythagoras provided a naturalistic cosmology, in contrast to the view common since Burkert's revolutionary study that Pythagoras was not a natural philosopher.Footnote ¹⁶ This might be suggested by the first example, since the item from religion comes first (the Islands of the Blessed) and is followed by the part of the natural world with which it is equated (the sun and the moon). One could argue that all of the cosmological acusmata were originally put in the question-form found in the first example, so that the second may have been formulated ‘What are the hounds of Persephone? The planets.’ It is not clear, however, that this was so or, more importantly, that the item from religion or myth always came first. Indeed, in most of the other examples the aspect of the natural world comes first. Even if it were true that the item from religion came first, no naturalistic explanation of phenomena is, in fact, given. To say that the Islands of the Blessed are the sun and the moon explains neither the subject nor the predicate in naturalistic terms. We are not told of what material the sun and the moon are composed or the processes by which they were formed or the rules which govern their behaviour. Instead, the sun and the moon are connected to Greek views of the afterlife by being identified as the Islands of the Blessed, where humans can hope to go after their death. Similarly in each of the other cases, some aspect of Greek religious life and myth is equated with a part of the natural world, but no naturalistic explanation is given.

M.L. West has shown that a convincing religious/moral cosmology can be constructed from these acusmata:Footnote ¹⁷

Aristotle's discussion of the seventh acusma listed above makes clear the distinction between a naturalistic explanation of thunder and a moral explanation such as is found in these acusmata (An. post. 94b27–34). He argues that thunder can be explained both as the result of natural necessity and also as serving a purpose. As an example of a naturalistic explanation, Aristotle says that thunder can be explained as the sound that arises when fire is put out by water, that is, when lightning is put out by rain. Here a natural phenomenon is explained simply in terms of the necessary results of the properties and behaviour of natural stuffs such as fire and water. The Pythagorean acusma, on the other hand, is cited by Aristotle as an example of an explanation in terms of purpose. It does not explain thunder in terms of the necessary properties of natural substances, but instead explains it in religious and moral terms as serving the purpose of frightening the malefactors imprisoned in Tartarus in order to keep them in line.

Number 8, the acusma that is the focus of this article, on the emendation that I have proposed, fits well with the rest of the list of ten cosmogonical acusmata given above. A natural phenomenon, the rich earth laid down by the Nile, is equated with an aspect of religious life. Just as the Big Bear and the Little Bear are equated with the hands of the goddess Rhea in number 3, so here the earth of the Nile is equated with Isis. On the standard interpretation of number 8 (the rainbow [Iris] is the light of the sun), it alone is an anomaly on the list. It alone gives a naturalistic explanation of a phenomenon. Although Delatte noted that this attempt at naturalistic explanation was ‘assez rare’ among the acusmata,Footnote ¹⁸ all scholars, including those who have worked most closely with the acusmata such as Burkert and Delatte, evidently beguiled by the parallel with Xenophanes and Anaximenes, accepted the standard emendation without question.Footnote ¹⁹ In some cases it has been used as a crucial piece of evidence to show that Pythagoras, despite the evidence of the rest of the acusmata, was none the less engaged in naturalistic cosmology like the Ionians such as Xenophanes and Anaximenes. Thus Riedweg argues that it illustrates Pythagoras’ interest in a scientific cosmology.Footnote ²⁰

As we have seen, however, rather than providing evidence for the old view of Pythagoras as a natural scientist, the acusma, when correctly emended, is yet another strong piece of evidence that he was, most of all, an expert in religious practices and in particular on those associated with the fate of the soul in the next life. We do not know exactly how the equation of Isis with the earth of the Nile was to be connected with the rest of Pythagoras’ moral cosmology, in which the sun and the moon are the Isles of the Blessed, but Isis’ connection to the dead and rebirth secures the coherence of this acusma with the other cosmological acusmata. Most striking of all, we have for the first time clear evidence for an interest in and connection to Egypt in words that might have come from the mouth, if not the pen, of Pythagoras.