A normal eclipse at the full moon (Passover) is impossible. Nor did anyone at the time record a universal planetary darkness. Origen suggests that enemies of the Church came up with the eclipse idea and inserted it into Luke, most copies of which have ἐσκοτίσθη:
Dicimus ergo quod Matthaeus et Marcus non dixerunt defectionem solis tune factam fuisse, sed neque Lucas, secundum pleraque exemplaria hebentia sic: Et erat hora fere sexta, et tenebrae factae sunt super omnem terram usque ad horam nonam, et obscuratus est sol. In quibusdam autem exemplaribus non habetur: Tenebrae factae sunt, et obscuratus est sol; sed ita: Tenebrae factae sunt super omnem terram sole deficiente. Et forsitan ausus est aliquis quasi manifestius aliquid dicere volens, pro: Et obscuratus est sol, ponere: Deficiente sole; existimans quod non aliter potuissent fieri tenebrae nisi sole deficiente. Puto autem magis quod insidiatores Ecclesiae Christi mutaverunt hoc verbum, quoniam tenebrae factae sunt sole deficiente, ut verisimiliter Evangelia argui possint secundum adinventiones volentium arguere ea.
Apostle Mark records that there was darkness for threehours ,which is not consistent with an eclipse. If a city is in the “path of totality” then it experiences full darkness during the eclipse, which lasts for a maximum of about two minutes and 40 seconds .The maximum possible duration of a total solar eclipse going from partial to full eclipse to recession is roughly seven minutes and 30 seconds. A far more likely natural explanation is that it was due to some sort of dust cloud activity, which would be consistent with the time span we’re talking about. Here are some pictures of dust clouds blocking out the Sun:
It doesn’t get that dark with a total Eclipse. We had 92% here, hardly noticed it. It doesnt stay dark for several hours. So you can rule out Eclipse. A huge thunderstorm makes for serious darkness.
I think that all commentators are agreed that an eclipse is an impossible explanation here. But don’t Luke’s words τοῦ ἡλίου ἐκλείποντος mean “while/because the sun was eclipsed?” (I’m not sure if it’s genitive of cause, or an absolute, or of time.) Origen thinks that ἐκλείπει must mean eclipse and claims that it cannot be the original text for this reason.
τοῦ ἡλίου ἐκλείποντος certainly refers to a solar eclipse. It’s the standard terminology. Aristarchus of Samos (3rd cent. BCE), the brilliant astronomer who anticipated Copernicus in propounding a heliocentric model of the solar system, credited Thales with recognizing the cause of solar eclipses: ἔφη τε ὁ μὲν Θαλῆς ὅτι ἐκλείπειν τὸν ἥλιον σελήνης ἐπίπροσθεν αὐτῷ γενομένης (P.Oxy. 3710 ii 37-40).
There is significant manuscript variation, and for some reason, I’m being blocked from posting either the UBS apparatus or N-A (perhaps because there are a lot of embedded links).
Metzger:
23:45 τοῦ ἡλίου ἐκλιπόντος {B}
The words καὶ ἐσκοτίσθη ὁ ἥλιος (“the sun was darkened”) appear to be the easier reading, substituted by copyists for τοῦ ἡλίου ἐκλιπόντος [or ἐκλείποντος], which may mean either “the sun’s light failed” or “the sun was eclipsed.”
Metzger, B. M., United Bible Societies. (1994). A textual commentary on the Greek New Testament, second edition a companion volume to the United Bible Societies’ Greek New Testament (4th rev. ed.) (p. 155). London; New York: United Bible Societies.
From a strictly semantic viewpoint, while ἐκλείπω certainly is used to describe an eclipse in the technical sense, can it be used in the more general sense simply of light failing, as Metzger suggests?
As it (the Persian army) was setting out, the sun left his place in the heaven and was invisible, although the sky was without clouds and very clear, and the day turned into night. When Xerxes saw and took note of that, he was concerned and asked the Magi what the vision might signify. They declared to him that the god was showing the Greeks the abandonment of their cities; for the sun (they said) was the prophet of the Greeks, as the moon was their own. Xerxes rejoiced exceedingly to hear that and continued on his march.
(This eclipse, which supposedly happened in the spring of 480, is probably at least partly fiction.)
I should perhaps start a new post for this, but I’ve been wondering about Herodotus 1.74.2:
“They were still warring with equal success, when it happened, at an encounter which occurred in the sixth year, that during the battle the day was suddenly turned to night. Thales of Miletus had foretold this loss of daylight to the Ionians, fixing it within the year in which the change did indeed happen.”
If this account is accurate, it refers to the eclipse of May 28, 585 BC. Critics are generally skeptical about Thales having really been able to predict it. I don’t know enough about astronomy to judge what is exactly at issue here. Does anyone know astronomy here?
Is it possible to predict the year within which an eclipse occurs, without being able to predict the exact date? Shouldn’t you either be able to predict the exact date, or not at all?
What is the hard part about predicting eclipses? Predicting whether one occurs in the first place? Predicting which ones are full eclipses? Predicting the region where the eclipse is full? To my understanding, only total eclipses are impressive. As Stirling said, a 92% eclipse is barely noticeable - 8% daylight is still very bright. I suppose a 99% eclipse is still barely noticeable, especially if the sky is cloudy.
If you know the orbital velocity of the moon and earth, which varies due to eccentricity, you can use this information to predict eclipses in a straightforward way, as the above device seems to do. Everything that we find out about the ancient’s astronomical knowledge is surprising.
Of course, Thales did not live in such an advanced period. Panchenko’s 1994 article on Thales’ prediction is very good, and he gives a plausible account of how Thales might have predicted this particular eclipse.
The death of Jesus called for attendant symbolic prodigies, and the synoptic gospels duly provide them. (Only teleologically fixated John spurns them.) An eclipse of the sun makes a better one than mere darkness.
I think the eclipse must be the original version in Luke. του ηλιου εκλιποντος has to mean an actual eclipse (and a full one, obviously), not just the sunlight somehow failing. The astronomical impossibility of it may have led to its erasure and harmonization with Mk and Mt.
Prediction, as distinct from explanation, was surely impossible for Thales, but it’s just the sort of thing that would have been attributed to him. As the text I quoted shows, Aristarchus of Samos knew better. The thing to consult is Otto Neugebauer’s History of Ancient Mathematical Astronomy.
Incidentally, that quotation occurs in a learned commentary on Odyssey bk.20, in an astronomically and calendarically complex note on αλλα μαλ’ ηρι νεονται, επει και πασιν ἑορτή (156), of the suitors’ expected early arrival for tomorrow’s feast. Apparently Aristarchus the Homeric exegete (from Samothrace, not Samos) took it to be the new-moon festival of Apollo, who was identified with the Sun. And I guess there’s an underlying connexion with Theoclymenus’ famous vision of the suitors’ impending doom later in the book (351-7, … ηελιος δε | ουρανου εξαποπωλεν …), which was understood as indicating an actual eclipse and was related to Odysseus’ return at new moon.
In his final paragraph Gainsford wrongly assigns the papyrus testimony to his (1) rather than his (2), and he misses its significance (see my previous post: this is Aristarchus of Samos, not just any old "report”!). He could have cut to the chase sooner, but I’m glad to see he comes to the right conclusion.
I thought Paul (the Finn, not the preacher man) might be interested in the Odyssey commentary. Hylander evidently doesn’t visit here (which may be wise of him).
I did find it interesting, thank you! It’s long time since I’ve read this part of the Odyssey. I’m not quite sure what you mean by “misses its significance”.
Apollo was identified with the Sun later, but certainly not in Homer. In Demodocus’ tale in Od. book 8 for instance we can clearly see that they are two distinct characters.
Here is the note on Od. 20.356-7 in Rutherford’s Cambridge Green & Yellow commentary (I don’t know how good the book is, since I haven’t read it. The text was scanned with my iPhone - I’m sorry for any remaining mistakes):
356-7 The fatuous notion that these lines refer to a solar eclipse taking place at the time is mentioned in the scholia, and also by Plutarch (On the face of the moon 19, 93 IF); cf. M. W. Haslam on Ρ Oxy 3710.36 (p. 107). Obviously the darkness is symbolic of the suitors’ sins. Further, the idea that the sun should vanish and abandon its function is reminiscent of the threat made by the sun-god at 12.382-3, when Odysseus’ men have killed his cattle: if he does not receive recompense, he will go down to Hades and shine among the corpses. The companions of Odysseus offend against the gods and are duly punished; but in their case there are, as has often been pointed out, extenuating circumstances (cf. Fenik, Studies 208-32): they are starving and can neither escape from the island nor support themselves otherwise, and they intend to honour and build a temple to the sun when they return home, 12.345-7. The suitors are far more straightforwardly guilty. For them, it is as if Helios has fulfilled his threat.
I don’t understand what exactly is meant by the notion being “fatuous”. Is there any way of interpreting these lines other than that they refer to a solar eclipse? (On the other hand, I do think that it’s “fatuous” to pinpoint the exact date of Odysseus’ revenge on the suitors on April 16th 1178 BC, as some have done, just because there was a solar eclipse that day – Homer was composing half a millennium later, after all.).
Is there any way to read that Odyssey note in full?
I found this article on the astronomical dating of the slaying of the suitors interesting, even if I don’t quite agree with the conclusions. Odysseus in book 5 on his σχεδιη watches the Bootes and the Pleiades, which are in the sky simultaneously only in spring and in early autumn - for some reason, the authors take it for granted that it must have been spring, although it seems to me that it’s generally assumed the Odysseus’ homecoming takes place autumn. Also, the theory about Hermes’ journey to Calypso’s island representing the movement of Mercury doesn’t seem tenable to me.
I had another look into this eclipse Thales supposedly predicted. Here’s a reply to Panchenko article, addressing what I find to be its main flaw, i.e. the fact that partial eclipses are not really impressive, and often not even visible.
But Panchenko’s article is stimulating. For one, he might well be right as to why Thales only predicted the terms of the year in which the eclipse took place: because the calendar was so rudimentary. Intercalary months were often added on an ad hoc basis - if you wanted to give the exact date for an event that was going to happen in exactly, say, 200 months from now, there was no way to do it, because it depended on the whim of the authority who decided whether an intercalary month was needed or not. This sort of problem is one reason why the Julian calendar reform was made.
I won’t repeat my previous posts, but I will say that I think it’s folly to start from reports that Thales could predict solar eclipses (or one in particular) instead of starting from Aristarchus of Samos’ unequivocal testimony that Thales recognized the cause of solar eclipses (i.e. the moon blocking the sun). That was a notable achievement, however elementary it seems to us today. If Thales had been capable of actually predicting a solar eclipse (most unlikely, to put it mildly, see Neugebauer), Aristarchus would have said so, instead of crediting him with only the lesser feat. Aristarchus was the foremost astronomer of the day, and knew what was what. That Thales came to be credited with actual prediction—by Herodotus (hardly the most trustworthy source) and then by others—was inevitable (I might say perfectly predictable). The predictive astronomical marvel that the Antikythera mechanism seems to have been came many centuries later, when astronomy was very much more advanced.
I had read the previous posts and I agree. Thanks also for having brought the papyrus fragment into our attention. As I tried to say in my previous post, I think Panchenko is ultimately wrong, because he doesn’t take into consideration the fact that partial eclipses are not impressive to the layman.
Still, there are some interesting ideas in the article. For one, it gives a good reason why Herodotus uses such an imprecise formulation for the prediction: οὖρον προθέμενος ἐνιαυτὸν τοῦτον ἐν τῷ δὴ καὶ ἐγένετο ἡ μεταβολή. Why doesn’t he give an exact date for the predicted date? From this site http://www.eclipsewise.com/solar/solar.html, I can find the information that on 5th December 2048 there will be a total eclipse in South America and Africa. Thales could not have given such a precise date, because there was no standardized calendar. If, for the sake of the argument, he had known that there would be an eclipse in exactly 1242 lunar months, he would have been unable to give a precise date for the event, because the date would have depended on how many intercalary months the authorities happened to add to the calendar in the meanwhile, which, as I said, happened more or less on an ad hoc basis. Add to that the fact that this was done independently in every city-state, all which had their own calendar - which, if I understand correctly, where out of sync with each other. Similarly, it would have been difficult to give an exact date for an astronomical event that had already taken place, because one would have to know what adjustments had been made to the calendar in the intervening time.
Could it be that Thales was just lucky? You make a good point that his most important discovery was the cause of solar eclipses. But if he was conscious that there was some regularity to them, he could have made educated guesses about possible eclipse years, and just gotten lucky with 585 BC, which just happened to be visible in Asia Minor. His other predictions would not have turned out so well, so this part of his work would not have been very impressive to posterity and would have been forgotten but for Herodotus, who has a way of aggrandizing things. This is speculative of course.
Gear e* on the Antikythera mechanism precisely tracks the precession of the lunar orbit (8.85 years instead of the correct 8.8826). It’s extremely surprising to see that before the 17th century AD. The eccentricity of the lunar orbit, and the earth’s tilt, are the two observational pieces that you really need to make good eclipse predictions. Everything else is (post-Newtonian) math. Even Paul’s calendar problems fade into the background at that point, because you don’t need hundreds of years of uninterrupted observations if you know about lunar eccentricity. You can make due with a private calendar and use a much shorter period of observation.
Paul, Just lucky, you suggest. But isn’t it wrong-headed from the start to give any credence to Hdt’s report? That Thales understood what was responsible for solar eclipses (the achievement Aristarchus credited him with) would automatically get perverted into an ability to predict them, which any astronomer (certainly Aristarchus!) would have known was a very different and more difficult thing. Hdt knew of an eclipse. The rest follows, as night follows day (and hopefully the reverse).
Joel, The Antikythera mechanism must represent the very best astronomical knowledge of the day (unlikely to be earlier than 1st cent. BCE, I should think; surely post-Hipparchan), and its extraordinary sophistication did indeed come as a big surprise, though perhaps bigger than it should have. Aristarchus of Samos was ahead of Copernicus (anyone could see that everything in the sky revolves around the earth), but this goes way beyond that.
However, it comes a good number of centuries after Thales. There seems no possibility of that anyone was capable of predicting eclipses in his time. Precession was an observed phenomenon that baffled all astronomers before Hipparchus. The Babylonians did in fact have hundreds of years of uninterrupted observations, but that couldn’t help them.
Tonight should be a good night for sky-watching, by the way!
That’s the axial precession of the earth, or precession of the equinox. This is the apsidal precession of the moon. (And the nodal as well?) It’s astonishing to find it modeled in the Antikythera mechanism.