Friday, 2 December 2016

Pythagoras and the beans #2: why ban beans?

We continue directly from last time: now it's time to look at the proposed explanations for the Pythagorean bean ban.

3. The favism theory

People with the genetic condition of glucose-6-phosphate dehydrogenase (G6PD) deficiency can sometimes suffer an illness called favism when exposed to broad beans. Favism destroys red blood cells and is dangerous for young children. The favism theory holds that the Pythagoreans were aware of this condition, and banned contact with beans to prevent it.
  • Those for: Arie 1959; Lieber 1973; Brumbaugh and Schwartz 1980; Katz 1987 (a follow-up to Katz and Schall 1979). (See also older bibliography cited in some of these articles.)
  • Those against: Scarborough 1982; Simoons 1998: 216-249; Dye 1999.
  • Undecided: Garnsey 1998: 219-20.
Proponents cite modern genetic studies showing that 5-6% of the world population has G6PD deficiency, and that the type of G6PD deficiency that can lead to favism is concentrated in the eastern Mediterranean, with incidence as high as 8% to 35% on Rhodes. In principle, you can imagine a high incidence leading to bans for medical reasons. (Katz' argument is slightly different: he believes the ban wasn't consciously intended to avoid favism, but a case of biocultural evolution -- the taboo was a favoured behavioural trait because it prevented favism, and G6PD deficiency is a favoured genetic trait because it assists in preventing malaria.)

World distribution of G6PD deficiency
(source: WHO Working Group 1989: 605)

On the 'against' side, Simoons shows that (1) Pythagoras didn't come from Rhodes, he came from Samos, and on modern Samos the incidence of G6PD deficiency is unexceptional; (2) in Calabria, southern Italy, where Pythagoras established his cult, the rate of G6PD deficiency in the modern population is unusually low (between 0% and 2.7%). The incidence of G6PD deficiency is too variable in different parts of Greece for a figure from modern Rhodes to be at all meaningful.

Dye additionally points out that that's G6PD deficiency, not favism. You have to have G6PD deficiency to get favism, but only a fraction of G6PD-deficient people actually get it. The other determining factors for the disease are poorly understood (Kattamis et al. 1969: 34; WHO Working Group 1989: 608, 610).

Moreover, it's a children's disease. According to the epidemiological studies that Brumbaugh and Schwartz themselves cite (Kattamis et al. 1969; Belsey 1973): (1) only 10-20% of people with G6PD deficiency ever suffer a case of favism in their entire lives; (2) of those, 85-95% are children aged six or under; and (3) they report fatalities only at age four or under.

In light of that, favism doesn't sound quite as momentous. For our purposes, that is: obviously favism remains a critical concern for parents in places that genuinely have a high rate of G6PD deficiency, like Sardinia and Cyprus.

Young children had colossally high mortality rates in antiquity, so when we look at ancient witnesses, it's really only illnesses after early childhood that are going to get any attention. If we assume maximal impact -- that 5.5% of ancient Samians had G6PD deficiency (approximately the modern world average), that 20% of them got favism, and of those, 15% were over six -- then at most 17 in every 10,000 people over six would ever suffer a case of favism. Realistically, it'd be more like half that. If we're talking about Calabria and we assume non-maximal figures, the incidence is going to be more like 2 in every 10,000 people over six. And zero fatal cases.

Does that make a blanket ban a 'common sense injunction' as Brumbaugh and Schwartz believe? Hardly! Against the background noise of other unexplained illnesses, ancient observers wouldn't even notice a disease that is (1) not normally life-threatening for over-sixes, and (2) with a 0.02% chance of ever getting the disease.

Now, if ancient sources showed any awareness of the disease, that'd be different. But they don't. The earliest description of the disease dates to 1843 (Davies 1961: 477).* Ancient herbologists and medical writers are far, far better preserved than any documents relating to Orphic-Pythagorean mysticism, and they do discuss beans extensively -- writers like Theophrastus, Dioscorides, and Galen -- but only as a commonplace food. The only medical caution is that they give you gas. When ancient sources bring up the idea of a prohibition, it is only ever in connection with Orphic-Pythagorean mysticism.
* Note: Davies also cites an 1837 poem by Eduard Mörike which he believes was inspired by a case of favism. In fact the poem is pretty clearly alluding to ancient Orphic mysticism. Mörike was heavily influenced by his study of classical literature, and favism is extraordinarily rare in Germany (Mörike spent his whole life in Baden-Württemberg).

Aristotle points out that the bean 'is destructive' (φθείρει) as one of several possible explanations for the Pythagorean bean ban, and Garnsey thinks this is a point in favour of the favism theory (1999: 88). That's not a strong consideration. The expression appears again in Theophrastus, slightly later than Aristotle --
Let us first deal with beans: applied to the roots and shoots, their pods destroy (φθείρει) not all trees, but only the ones just growing up, since these are weaker. The pods destroy (φθείρει) them by taking the food away by reason of their hardness and dryness, absorbing some of it themselves, and shutting out the rest, for when the trees get no food they perish. The bean pods and the like destroy (φθείρει) the tree by being hostile (as it were) to its sprouting.
-- Theophrastus, On effects in plants 5.15.1-2 (tr. Einarson and Link, corrected)
(Similarly Clement of Alexandria, Stromateis; Geoponica 2.35.1.) Yes, Aristotle calls beans 'destructive'. But apparently he's talking about herbicide, not favism.

The favism theory doesn't amount to a hill of beans. At best it's a mildly interesting speculation. More realistically, it's a piece of speculation that can only be supported by selective treatment of incidence figures, or by ignoring them altogether. The relevant testimony is copious, we would expect to see lots of corroboration there, and yet none exists.

(Katz' form of the favism argument is not completely excluded, though. If the bean ban were an evolved behaviour, as opposed to a medical prohibition -- and that is one hell of an 'if'! -- ancient sources' complete unawareness of favism would be unsurprising. The evolved behaviour idea also sits well with the fact that favism is a children's disease. It's still tenuous: how on earth do we tell when a biocultural explanation is the right kind of explanation for a food taboo? But it's not impossible.)

Herbicide, bringer of sickness, or symbol of human genitals? You choose.

4. Death and reincarnation

The interpretations that are most closely related to genuine Pythagoreanism are those which connect beans with the doctrine of metempsychosis.
Thus quoth the scholar of Greek religion Walter Burkert (1972: 183). To some extent this is borne out by our earliest source on the bean ban, Aristotle:
εἰσὶν ὅμοιοι ἢ ὅτι ᾍδου πύλαις < . . . > ἀγόνατον γὰρ μόνον
(Beans) are like the gates of Hades, for only (the stem of this plant) is unjointed
-- Aristotle fr. 195 Rose
Hypothetically, the unjointed stem might be a symbol for unimpeded progress from the underworld to the living world. However, this is just one of the explanations that Aristotle suggests, and it's not exactly clear.

Reincarnation comes up a lot in 4th century BCE sources, especially Plato, but the terms we throw around for Greek reincarnation -- metempsychōsis 're-ensoulment', metensōmatōsis 're-embodiment' -- are late: neither word is attested before the 1st century. Reincarnation is mainly associated with Epimenides, Pythagoras, Empedocles, and Plato, plus a few more minor names. The jury is out on whether there's any link between Greek and Indian beliefs about reincarnation: I won't touch that today. Herodotus calls it an 'Egyptian' doctrine (2.123), wrongly: that's really just another way of saying it's Pythagorean. An awful lot of Epimenidean-Pythagorean mysticism got spuriously linked to Egypt. Allusions to Egypt were just a way of adding mystique to the mystical. (For another example of faux Egyptianism, see this old post on the 3-4-5 Pythagorean triple.)

In Greek thought, your next life might be determined either by your own choice, or by your moral character. Male and female, human and animal, even plants are potential destinations: one fragment of Empedocles states that in his past lives he had been a boy, a girl, a bush, a bird, and a fish (fr. 31 B.117 Diels-Kranz; also fr. B.127). A substantial discussion in Plato's Republic (book 10, 617e-620d) emphasises that everyone is responsible for choosing their next life. Alternatively, it might depend on whether you're naughty or nice, or on whether you're properly enlightened. Pythagoras supposedly had been a soldier in the Trojan War in a past life, a Trojan named Euphorbus (who appears in the Iliad, naturally): clearly the idea was that he had always been among the great and the good. Empedocles talks about a cycle of reincarnations where the best of the best end up as 'prophets, hymnists, doctors, and front-line fighters' (fr. B.146). This idea shows up in Plato too, in a famous passage casting the mind as the charioteer guiding his horses -- the immortal soul -- through various lives (Phaedrus 246a-254e). Only those with a philosophical inclination get to become human in their next life (249b).

These mystics taught that the cycle of reincarnation is a punishment for some crime committed when your soul was a divinity (daimōn). If you signed up to the parish newsletter, so to speak, you could break free and go back to being a divinity. In Empedocles, salvation takes 30,000 years (fr. B.115.3-8); in Plato, 3000 years (Phaedrus 249a).

We don't know for sure that this has anything to do with the bean ban -- that's why people look for alternate explanations, like the favism theory. The Aristotle fragment quoted above helps, but it doesn't really settle anything. There's also the Orphic fragment we looked at last time:
I tell you, eating beans is the same as eating your parents' heads ...
(the bean) is a path and stairway out of Hades' house
for the souls of the strong, whenever they ascend into the light
-- Orphica fr. 648 ed. Bernabé
Ancient writers tell us that the first line comes from 'Orpheus', but it's attributed to Pythagoras nearly as often. (Bernabé 2013: 123 n. 34; Plutarch Q. conv. ii.635e, 'either Orphic or Pythagorean'.) Some ancient writers explained the bean ban by pointing out that bean pulp under certain conditions changes to resemble human blood or parts of the anatomy, including heads. That suggests someone was taking 'eating beans is the same as eating your parents' heads' very literally. Pliny reports that according to some, 'the souls of the dead are in (beans)' -- but, inconveniently, he explicitly distinguishes this from Pythagorean beliefs (Nat. hist. 18.118).

Roman religion, too, tended to treat beans as a holy symbol and sometimes made a connection between beans and the transition between life and death. In the Lemuria festival, beans were used in a ritual to expel dead spirits from the house (Varro, reported in Nonius Marcellus 135.15 M). Beans played a role in the Parentalia, a festival designed to honour ancestors (Ovid Fasti 2.576); Pliny Pliny NH 18.118). And Roman priests weren't allowed to eat beans (Varro, reported in Pliny NH 18.119); one priest, Jupiter's flamen Dialis, wasn't even allowed to mention them (Aulus Gellius 10.15.12). Some of this is strikingly similar to some things we hear from the Neo-Pythagoreans.

The exact nature of the religious meaning of beans is never made clear, but there does seem to be something here. But if the bean ban was ever framed in terms of some definite purpose, with cause and effect in mind, the purpose was theological, not mundane.

Accordingly, the great theorist of ancient religion Marcel Detienne interprets the Pythagorean treatment of beans as a term in a set of symbols with structured links to one another -- a structuralist interpretation, in other words -- where the taboo on beans is purely religious (1977: 49-59). Beans are a key symbolic term in the Pythagorean system of food classification, Detienne thinks, linked to notions of beast-like behaviour. Beans are quintessentially symbolic of meat-eating: hence the links that some sources draw between beans and raw meat, beans and blood, beans and animal urges. In one story we mentioned last time, where Pythagoras persuaded an ox never to eat beans again, that represented a de-animalising of the ox.

Beanfield in bloom

5. Other theories

You didn't think these two theories were the only two ones, did you? Ancient sources were as confused by the Pythagoreans as we are today. Our earliest source, Aristotle, is among the most confused:
Aristotle says in On the Pythagoreans that (Pythagoras) told them to abstain from beans, either because they are like genitals; or because they are like the gates of Hades, for only (the stem of this plant) is unjointed; or because it is destructive; or because it is like the nature of the whole; or it is oligarchic -- at any rate, (people) use them for drawing lots.
-- Aristotle, reported in Diogenes Laertius 8.34 (≈ Arist. fr. 195 Rose)
So without further ado let's make a complete list of ancient theories on the bean ban.
  1. Because beans are like genitals.
    Aristotle fr. 195; Antonius Diogenes Wonders beyond Thule (reported in Porphyry Life of Pythagoras 44, John Lydus De mensibus 4.42: the bean blossom after being buried in soil for 90 days looks like a baby's head or like female genitalia); Lucian Sale of lives 6 (unripe beans inside the pod look like male genitalia). Cf. Empedocles fr. B.141, with explanation in Aulus Gellius 4.11.9-10.
  2. Because the bean pod has a herbicidal effect.
    Aristotle fr. 195. Cf. Theophrastus Effects in plants 5.15.1-2; Clement of Alexandria Stromateis; Geoponica 2.35.1.
  3. Because the bean 'is like the nature of the whole'.
    Aristotle fr. 195, ὅτι τῇ τοῦ ὅλου φύσει ὅμοιον. Cf. Refutation of all heresies 1.2.14: the bean arose 'at the beginning and composition of all things, when the earth was still combined and sifted together'. Porphyry Life of Pythagoras 44 is parallel, and casts doubt on the reading 'sifted together' in the Refutation; but the Porphyry passage is obscurely phrased. The parallel suggests that the Refutation author, like Porphyry, is paraphrasing Antonius Diogenes' Wonders beyond Thule.
  4. Because beans are a symbol of political engagement, since they are used for drawing lots.
    Aristotle fr. 195; Lucian Sale of lives 6; pseudo-Plutarch On educating children 12f. Cf. Andocides De mysteriis 1.96 ἡ βουλὴ οἱ πεντακόσιοι <οἱ> λαχόντες τῷ κυάμῳ, 'the Council of the Five Hundred allotted by the bean'. (On bean-eating as a social symbol in antiquity see Garnsey 1998: 220-4.)
  5. Because the bean or its blossom transforms into substances or shapes reminiscent of human body parts under various conditions.
    Antonius Diogenes, Wonders beyond Thule (reported in Porphyry Life of Pythagoras 44, John Lydus De mensibus 4.42: chewed bean pulp left in the sun smells like blood, the blossom after being buried in soil for 90 days looks like a baby's head or like female genitalia); Refutation 1.2.14-15 (also based on Antonius Diogenes? see 3. above; chewed bean pulp left in the sun smells like semen, the bean and blossom buried in soil for a few days look like a womb with a baby's head inside it); Lucian Sale of lives 6 (cooked beans left in moonlight turn into blood). Cf. Heracleides fr. 41 (beans buried in dung for 40 days change to look like human flesh).
  6. Because the bean is 'thought to dull the senses and cause insomnia'.
    Cicero On divination 1.(§30).62; Pliny Natural history 18.118. For beans' influence on dreams cf. Plutarch Q. conv. viii.734e-f, Geoponica 2.35.8.
  7. Because the word 'bean' (kyamos) sounds like the word 'pregnancy' (kyēsis).
    Plutarch Q. conv. ii.635e.
  8. Because beans cause sterility in women.
    Clement of Alexandria Stromateis
  9. Because the pattern on the bean blossom contains ill-omened letters.
    Pliny Natural history 18.119; Geoponica 2.35.6.
All these sources are explicit that they are talking about Pythagoras' ban. Obviously there's a lot of overlap between numbers 1 and 5. (Note that some of the sources are less than serious: Lucian is satyrical, and Antonius Diogenes' lost Wonders beyond Thule was a fantasy novel. In item 5, the Lucian passage is almost certainly a satire of Antonius Diogenes.)

It should be pretty obvious that most of these must be purely speculative. No one had any real clue of why the Pythagoreans banned beans, so they were just rummaging around to see what they could dig up. (That's also pretty much how I see the favism theory.)

6. Yes, it's mysterious -- that's the point

I'd better close by repeating that even our ancient sources don't know the reason for the taboo. Aristotle, our earliest witness, and one of the most reliable, lists off five random speculations for want of anything better.

Even in the early period, Pythagoreanism was more a cult than a school. Its teachings were more religious mysticism than anything rationalistic. Alberto Bernabé (2013) makes a thorough catalogue of places where our sources mix up Orphic, Pythagorean, and Empedoclean doctrines. I mentioned above that the line about 'eating beans is the same thing as eating ancestors' heads' is attributed to Pythagoras almost as often as Orpheus.

And the bean ban isn't an isolated bit of weirdness. The Pythagoreans venerated other plants too, like the mallow. Last time we saw that the Eleusinian cult had some kind of link to the Pythagorean bean ban; well, at the Haloa, another Attic festival in honour of Demeter, the list of things you weren't allowed to eat included eggs, fowl, pomegranates, apples, and various other things, as well as beans. The Roman flamen Dialis wasn't allowed to say the word 'bean': well, he wasn't allowed to mention female goats, raw meat, or ivy either.

Some religious doctrines are just not meant to be explained. Sometimes they're symbolic without symbolising anything specific. They exist solely so that they can be revealed to catechumens. That way, advanced initiates get to have access to secret knowledge, but the knowledge doesn't have to be meaningful outside that context.

Remember the story of how the Pythagoreans were caught and killed by soldiers, because they weren't willing to escape through a beanfield? Listen to what happens next:
'Teach me one thing,' (Dionysios) said, 'and you shall go free with a suitable escort.' Myllias asked what he was so eager to learn. 'This,' said Dionysios, 'the reason why your companions chose to die rather than tread on beans.' Myllias promptly replied 'They were prepared to die rather than tread on beans, and I would rather tread on beans than tell you the reason'.
-- Iamblichus, Life of Pythagoras §31.193 (tr. Clark)
Here the entire point of the bean ban is that it's a secret. No one really cares about the reason for the ban: they just want to uncover the secret. Again and again, Iamblichus tells us that everything in Neo-Pythagoreanism has to be hush-hush. Not because of any reasoned danger if the information gets out: it's just that that's what being a Neo-Pythagorean is all about.

You can see this elsewhere too. There's a report on a cult of Demeter at Pheneus, in Arcadia, in the travel writer Pausanias (8.15.3-4). Apparently they had a 'sacred story' (hieros logos) to explain why the locals could eat any kind of pulse except broad beans, which were 'impure'. Will Pausanias tell us this story? No, of course not. It's a secret.


  • Arie, T. H. D. 1959. 'Pythagoras and beans.' Oxford Medical School gazette 11: 75-81.
  • Belsey, M. A. 1973. 'The epidemiology of favism.' Bulletin of the World Health Organization 48: 1-13.
  • Bernabé, A. 2013. 'Orphics and Pythagoreans: the Greek perspective.' In: Cornelli, G., et al. (eds.) On Pythagoreanism. Berlin: De Gruyter. 117-151.
  • Brumbaugh, Robert; Schwartz, Jessica 1980. 'Pythagoreans and beans: a medical explanation.' Classical world 73: 421-422.
  • Burkert, W. 1972. Lore and science in ancient Pythagoreanism. Cambridge, Mass.: Harvard UP. (NB: 180-185 on Pythagorean food taboos)
  • Davies, P. 1961. 'Favism.' Postgraduate medical journal 37: 477-80.
  • Detienne, M. 1977. The gardens of Adonis. Spices in Greek mythology. Princeton: Princeton UP. (Orig. Les jardins d'Adonis, Paris: Gallimard, 1972.)
  • Dye, J. 1999. 'Explaining Pythagorean abstinence from beans.' The Internet Archive (original web publication now deleted).
  • Garnsey, P. 1998. Cities, peasants and food in classical antiquity. Cambridge: CUP. (NB: 214-25 on beans)
  • Garnsey, P. 1999. Food and society in classical antiquity. Cambridge: CUP. (NB: 85-91 on Pythagorean food taboos)
  • Kattamis, C. A.; Kyriazakou, M.; Chaidas, S. 1969. 'Favism. Clinical and biochemical data.' Journal of medical genetics 6: 34-41.
  • Katz, S. H.; Schall, J. 1979. 'Fava bean consumption and biocultural evolution.' Medical anthropology 3.4: 459-476.
  • Katz, S. H. 1987. 'Fava bean consumption: a case for the coevolution of genes and culture.' In: Harris, M.; Ross, E. B. (eds.) Food and evolution. Philadelphia: Temple UP. 133-59.
  • Lieber, E. 1973. 'The Pythagorean community as a sheltered environment for the handicapped.' In: Karplus, H. (ed.) International symposium on society, medicine and law. Amsterdam: Elsevier. 33-41.
  • Scarborough, J. 1982. 'Beans, Pythagoras, taboos, and ancient dietetics.' Classical world 75: 355-358.
  • Simoons, F. J. 1998. Plants of life, plants of death. Madison: U. Wisconsin Press.
  • WHO Working Group 1989. 'Glucose-6-phosphate dehydrogenase deficiency.' Bulletin of the World Health Organization 67: 601-11.


  1. Fusako's theory: "The ancient Greeks did athletics in the nude, didn't they? Maybe they wanted to avoid farting."

    1. Most of the medical writers do mention that property of beans, actually!