Tuesday, 30 November 2021

Unique words in Odyssey 24

Homer scholars have often argued that the last part of the Homeric Odyssey is spurious, added by a different poet. One of the main arguments is that its language is somehow distinctively different from the rest of the Odyssey.

An 1816 book by F. A. W. Spohn made this argument, claiming that the ‘epilogue’ — the last 624 lines of the epic, Odyssey 23.297 to 24.548 — has lots of language in it that doesn’t appear elsewhere in Homer. Spohn’s complaint is that it has lots of unique words.

Odysseus reunited with Penelope (terracotta relief, Melos, ca. 450 BCE; NY Met)

At this point in the Odyssey, Odysseus has slaughtered the suitors and reunited with his wife Penelope. In the epilogue, the suitors’ ghosts arrive in the underworld, Odysseus is reunited with his father Laertes, and his conflict with the suitors’ families is resolved. Spohn’s argument was prompted by two ancient glosses:

Aristophanes [of Byzantium] and Aristarchus make this the end-point of the Odyssey.
sch. M,V, Vind. 133 on Od. 23.296

Aristarchus and Aristophanes say that this is the conclusion of the Odyssey.
sch. H, M, Q on Od. 23.296

Why they said this, and what exactly they meant by it, are open questions. (I have my own ideas: see Gainsford 2015: 123–124.) What we’re looking at here is what Spohn says in support of these glosses (1816: 156–157, my translation):

There are numerous words in the Iliad that are not used in the Odyssey; conversely there are many in the Odyssey which are missing in the Iliad. ... But in this last part of the Odyssey, the number of hapax legomena we find is huge for its length; and it contains many forms which show that the language and style have developed.

‘Hapax legomena’ is Greek for ‘things uttered once’: unique words.

Many later scholars repeated this central claim. Page (1955: 101–136) gives an updated list of supposed anomalies in the language of the epilogue. He identifies 37 features that are unique in Homer, another 19 that are uncommon, and asserts that all such features are by their nature non-Homeric. Others have rejected Page’s argument, dismantling his anomalies one by one (Erbse 1972: 177–229; Wender 1978).

But no one ever questioned the reasoning, or the basic premise:

  • Unique language means it’s spurious.
  • The epilogue has lots of unique words.

Both of these are the purest nonsense.

Unusual language, therefore spurious?

[I]t may be confidently asserted that there is no other passage of comparable length in the Homeric poems in which linguistic evidence of relatively late composition is to be observed in such profusion.
Page 1955: 102

Stylometrics alone will not demonstrate the spuriousness of the Epilogue, though we can point to a selection of oddities which must trouble its defenders.
West 1989: 120

The epilogue has linguistic ‘anomalies’: well and good. That is a single datapoint. A compilation of these anomalies demonstrates that they exist. But that conveys no information at all about how typical or atypical the datapoint is. Anomalies exist everywhere. What matters is how frequent anomalies normally are, and whether the epilogue falls outside the usual parameters.

In other words, Spohn, Page, and West totally ignore the importance of statistical significance. Spohn’s argument was tendentious, but he had a partial excuse: statistics as a field didn’t exist in his time. More recent scholars have no excuse.

When Erbse and Wender demolished Page’s list of anomalies, they missed the point too. They explained the anomalies away, one by one, until there were none left. But in doing so they attacked only the data, not the methodology. The methodology itself was nonsensical. A single datapoint cannot demonstrate a discrepancy.

If a discrepancy were demonstrated, it would also be important to show that the discrepancy was statistically significant. One attempt at a stylometric analysis of the epilogue by Norman Postlethwaite falls down catastrophically here too. His metric is the frequency of certain kinds of noun-epithet phrases in the epilogue. But his control consists of another single datapoint, Odyssey 2.1–434, along with a passage from the Iliad. And that’s totally useless. We need to know not just a typical rate for stylometric features, but also the variance in their rate. Postlethwaite comes so close to confronting variance in the control, but misses: a couple of stats classes at school would have made all the difference.

The scatter figures for the two control passages of Homer are ... sufficiently alike to suggest that a figure in the neighborhood of these represents the normal frequency ...
Postlethwaite 1981: 182 (my emphasis)

Hapax legomena

Hapax legomena are words that appear with a frequency of 1, as I said above. Scientists working on stylometry have sometimes taken an interest in hapaxes. They aren’t a highly-regarded measure of authorial style these days, but they aren’t useless either.

Here’s the really important thing. Every text, of every length, has hapax legomena.

And here’s the really counter-intuitive thing. The ratio of hapaxes to total vocabulary is roughly constant, no matter how large the text corpus is.

Take a moment to digest that. No matter how large. You’d expect that as the size of a text corpus increases, more words get repeated, and so hapaxes as a proportion of the total vocabulary would drop.

You’d expect that. But that isn’t what happens. Hapaxes do get less frequent as corpus size increases — they make up a smaller and smaller proportion of the corpus. But as a proportion of the vocabulary, they stay pretty much constant. Here are some examples from English-language corpuses:

  • Lewis Carroll’s Alice in Wonderland is 26,505 words long, and has a vocabulary of 2651 words. Of those, 1176 appear only once. That is, 44.4% of the vocabulary in Alice is hapaxes.
  • Mark Twain’s Tom Sawyer is 71,370 words long; 49.8% of its vocabulary is hapaxes.
  • Kierkegaard’s collected works have 2 million words; 43.6% of the vocabulary is hapaxes.
  • A ‘representative’ 103 million word corpus of American English studied by Brainerd has a vocabulary with 44.7% hapaxes.
Note. For Alice, see Baayen 2001: 2–12; for Tom Sawyer, Fan 2010: 3; for Kierkegaard and the 103 million word corpus, Brainerd 1988: 14.

Scientific models of vocabulary distribution are forced to assume that the available vocabulary is literally infinite. One study has shown that the proportion of vocabulary that appears twice — dis legomena — hovers around 15.0% to 15.6% for corpuses up to 400,000 words; but above that size, their proportion starts to decline as more and more words are repeated. We can conjecture that for an extremely large corpus, on the order of over a billion words, we might start to see a similar effect for hapaxes ... then again, we might not.

Note. Infinite vocabulary: Baayen 2001: 56 ‘This definition explicitly requires S to be infinite while requiring that the hapax legomena constitute a non-negligible proportion of the vocabulary.’ Dis legomena declining for corpus-size over 400,000 words: Sichel 1986: 53–57.

What about the Odyssey? First, be aware that in ancient Greek it’s important to apply ‘stemming’. A regular Greek verb can have over 400 distinct forms, and there’s no value in treating all of them as separate vocabulary items. ‘Stemming’ means we treat λύω, λύεις, and λύει as a single vocabulary item.

With that in mind, here are the figures for Homeric hapaxes:

  Odyssey Iliad both epics combined
size of corpus 87,092 tokens 111,711 198,803
vocabulary 4856 types 5448 6807
hapax legomena 1606 types 1771 1962
hapaxes as % of vocabulary 33.1% 32.5% 28.8%
hapaxes as % of corpus 1.84% 1.59% 0.987%
hapaxes per 100 lines of verse 13.3 per 100 lines 11.3 7.06

These figures are based on the Northwestern University dataset, available through the WordHoard application and The Chicago Homer website. Some classicists found Homeric hapax rates surprisingly high when they first read about them, in Michael Kumpf’s 1984 book Four indices of the Homeric hapax legomena:

In a total of around 9000 lexical items ... used by Homer, [Kumpf] has found 2692 hapaxes, a number that is astonishing at first sight in the context of the formular style, where repetitions are much more numerous than in literary texts of the post-Homeric era.
Ruijgh 1987: 178 (my translation)

As you’ll have noticed, these proportions aren’t high at all: they’re actually lower than in the English corpuses mentioned above.

Note. Kumpf’s figures are different from mine for four reasons. (1) Kumpf includes proper names. (2) Kumpf uses the OCT edition for his text, while the Northwestern dataset uses Murray’s Loeb text. (3) The Northwestern dataset is more generous with stemming: for example, Kumpf counts Od. 1.328 ὑπερωιόθεν and 2.193 ἀσχάλλω as hapaxes, while the Northwestern dataset treats them as forms of ὑπερώιον and ἀσχαλάω. (4) Kumpf includes rejected readings from the OCT critical apparatus.

The linguist George Kingsley Zipf (1935) found that if you order words in a corpus by frequency, and plot their frequency on a graph with a logarithmic scale, you end up with a nearly linear graph. Formally, what that means is that if z = Zipf rank and V(z) = frequency, then Zipf’s Law states

log V(z) ∝ 1/(log z)

Zipf’s Law works just as beautifully for Homer as it does for modern English corpuses, formular style be damned. Here are the top Zipf-ranked words in the Odyssey:

Zipf rank Frequency Word
1 4506 δέ
2 2486 ὁ, ἡ, τό, etc.
3 2435 καί
4 1772 τε
5 1558 ἐγώ, με, etc.
6 1051 εἰμί, εἶ, ἐστί(ν), etc.
7 995 ἐν
8 958 σύ, σε, etc.
9 815 μέν
10 812 ὅς, ἥ, ὅ, etc.

We don’t start getting to nouns until Zipf rank 25 (ἀνήρ, frequency 449). Plot frequency vs. Zipf rank, and you get an absolutely typical Zipf curve:

Zipf graph of Odyssey vocabulary. The most frequent word, δέ, is plotted as the dot at the top left: Zipf rank 1, frequency 4506. The line of dots at the bottom right represents hapax legomena, Zipf ranks 3251 to 4856, all with frequency 1.

Now, remember that any given chunk of text is going to have hapaxes, and there are lots of them. If you want to make a claim like ‘the number of hapaxes we find is huge’, you need to know how many hapaxes you should expect to find.

If you plot the number of hapaxes per book of the Odyssey, it quickly becomes obvious that there’s nothing very special about book 24. There isn’t even anything special about the Laertes scene (Od. 24.205–412), which Page regarded as especially objectionable for its unique language.

Hapaxes in the Odyssey by book, and in the epilogue by episode. ‘Domain: Homer, Hesiod, Hymns’ represents words that appear once in the entire corpus of early hexameter; ‘Domain: Odyssey’ represents words that appear once in the Odyssey. With the larger domain there are fewer hapaxes, because many Odyssey-hapaxes get repeated elsewhere.

It turns out that the bit of the Odyssey that actually has the highest rate of unique vocabulary is in books 5 and 6: specifically, Odyssey 5.402 to 6.128, a passage with a rate of 32.3 hapaxes per 100 lines (71 hapaxes in 219 lines). The hapax-rate in the Laertes scene is half that, a mere 16.3 per 100 lines (31 hapaxes in 208 lines).

Hapaxes are just a small slice of the linguistic pizza, of course. Modern stylometric analysis is much more sophisticated than counting hapaxes. Once upon a time I’d planned to write this discussion up into a journal article; but hapaxes are too trivial, by comparison with what the real experts do.

And we can’t quantify the other kinds of anomalies that Spohn and Page listed: many of them relate to mythological conceptions that don’t appear elsewhere (Hermes as ‘Cyllenian’; ‘nine’ Muses), or unusual morphology (ἠριγένειαν as a substantive; πρου- in the arsis of a foot; contracted genitive Ὀδυσεῦς). These are very diverse, and they can’t really be quantified on the scale of the epic as a whole.

Hapaxes, however, can be dealt with quite straightforwardly. And with respect to hapaxes, Spohn and his followers couldn’t have been more wrong.

References

  • Baayen, R. H. 2001. Word frequency distributions. Dordrecht, Boston, London.
  • Brainerd, B. 1988. ‘Two models for the type-token relationship with time dependant vocabulary reservoir.’ In: Bernet, C.; Thoiron, P.; Labbé, D.; Serant, D. (eds.) Études sur la richesse et la structure lexicales. Paris. 13–22, 165–172.
  • Fan Fengxiang 2010. ‘An asymptotic model for the English hapax/vocabulary ratio.’ Computational Linguistics 36.4: 631–637. [MIT Press link]
  • Gainsford, P. 2015. Early Greek hexameter poetry. Cambridge. [CUP link]
  • Kumpf, M. M. 1984. Four indices of the Homeric hapax legomena. Hildesheim, Zürich, New York.
  • Page, D. L. 1955. The Homeric Odyssey. Cambridge.
  • Postlethwaite, N. 1981. ‘The continuation of the Odyssey: some formulaic evidence.’ Classical Philology 76: 177–187. [JSTOR link]
  • Ruijgh, C. J. 1987. Review of Kumpf 1984. Mnemosyne 40: 178-80. [JSTOR link]
  • Sichel, H. S. 1986. ‘Word frequency distributions and type-token characteristics.’ Mathematical Scientist 11.1: 45–72. [Applied Probability link]
  • Spohn, F. A. W. 1816. Commentatio de extrema Odysseae parte. Leipzig. [Internet Archive link]
  • West, S. 1989. ‘Laertes revisited.’ Proceedings of the Cambridge Philological Society (now Cambridge Classical Journal) 35: 113–143. [JSTOR link]
  • Zipf, G. K. 1935. The psycho-biology of language. Boston.

Friday, 19 November 2021

The dates of Homer

Datings for Homer that you’ll find in current scholarship range from 800 BCE to the 500s BCE. There’s a popular tendency to quote ‘late 700s’ or ‘around 730’ as a consensus date, but ... it isn’t as simple as that.

As with the dates of Jesus, we have no ancient eyewitnesses that give reliable dates. Unlike Jesus, we have a lot of material attributed to Homer: the Iliad and Odyssey. That means that we’re really talking about the dates of poetic artefacts, rather than the date of a person.

A rhapsode beginning a performance of an epic poem (Attic neck amphora, ca. 490–480 BCE; British Museum 1843,1103.34, CC BY-NC-SA 4.0)

Modern scholars have assigned many different dates to Homer. The upshot of what follows is that that’s mostly because each of them are dating different things. In particular, there’s no consensus putting Homer in the late 700s BCE: you’ll see that date repeated a lot, but as Barbara Graziosi has put it (2002: 91), ‘the eighth-century date is more often stated than argued for.’

No matter what date you think of, there are questions to be asked over what it is, exactly, that’s being assigned to that date. The problem isn’t so much about deciding who’s right, or which evidence is the most important: it’s more that it can be hard to understand what different datings even mean.

Dating models

There are three main methodologies.

  1. Stylometric models look at the development of epic language over time, and date the Homeric epics by where they fit in linguistically in comparison to Hesiod and the Hymns. Janko’s analysis (1982) is the best known to Homer scholars, though his methods are now very outdated.
  2. Terminus models identify a key historical development, at a known date, and then declare that an epic must be earlier or later than that date. For example, Burkert (1976) and others think one passage in the Iliad alludes to Ashurbanipal’s sack of Thebes in Egypt, in 663 BCE, so therefore the entire Iliad is later than that. Conversely Powell (1991) thinks a famous vase dating to the second half of the 700s BCE is inspired by the Iliad; therefore the whole Iliad must be earlier than that. Conversely again, Crespo (2014) thinks the Iliad takes inspiration from the vase ... and so on.
  3. Phase models assume an ongoing tradition, with epics performed orally and subject to great variation, and periodically becoming more fixed. Kirk (1962: 301–334) describes a gradual crystallisation of the story and text of the poems over a period of multiple centuries; Nagy (1996: 41–43, and elsewhere) makes the idea more systematic and re-dates the phases.
Note. I’m not including evidence from ancient mythical-biographical traditions. That’s because they’re ... mythical. Early Greek poetry frequently uses authorial personas that are embedded in the poetry: personas like Homer, Hesiod, Sappho, Orpheus, and others. These are worth studying in their own right (see Graziosi 2002; Gainsford 2015: vi–x). But the ‘Homer’ persona is a stage act, a backstory, a brand-name. To take a parallel: if you like late 90s hip hop, you need to know who Slim Shady is; but you don’t use Slim Shady as evidence for the composition of Eminem’s songs. The same goes for 70s glam rock and Ziggy Stardust. And the same goes for Homer.

Each model has strengths and weaknesses. But if a given methodology produces contradictory results (as terminus models do), something isn’t quite working.

Rather than pitting these models against each other directly, I’ll identify a few points de capiton, ‘quilting points’, in the history of the story, language, and dissemination of the Homeric epics. Dating methods are differentiated not just by what evidence they choose to focus on, but also by how they fix these points de capiton to the ‘quilt’ of Homer’s history.

  • Null Date: the origin of the underlying story, either as poetic or non-poetic legend. The story of the Iliad may have originated in non-poetic storytelling; alternatively, it may have only come into existence when the Iliad was composed. Depending on which you prefer, you’ll organise this date differently relative to the others below.
  • Composition Date: when the story got into epic form, either oral or written. Let’s say the Iliad’s composition date began when the story of Achilles’ wrath at Agamemnon (or Hector?) first appeared in dactylic hexameter. If you think the story evolved over generations of performers, you’ll imagine this as a protracted period. If you think a single romantic genius created it at his writing-desk, then this date will be a single moment in time.
  • Fixation Date: when an epic settled into more-or-less its surviving form, perhaps with wording close to the modern text. You could optionally also have a separate Transcription Date, when the epic was first written down. If you think there was a process of oral recomposition in performance, which continued after the poem’s overall shape had emerged, then you’ll treat the transition from Composition to Fixation as a phase, perhaps with multiple intervening steps. If you think the Iliad was written down straightaway, you’ll treat the Composition, Fixation, and Transcription Dates as the same thing.
  • Dissemination Date. It may have taken time for Homer to hit it big. There’s good evidence that 522 BCE is when the Athenians became seriously obsessed with Homer; we have some evidence of different Dissemination Dates for different parts of the Mediterranean world (see below).
  • Book-division Date. The division of each epic into 24 books may not seem essential to their origins, but it matters. The division can’t have taken place before Fixation and Transcription. Even more: the division can’t have happened before Transcription in the 24-letter Ionic alphabet, each book of each epic labelled with one Ionic letter. Some scholars accept ancient testimony that the division was made in the Hellenistic period; others want to put it in the 700s BCE. For the record, the Athenian alphabet had 20 letters until 404 BCE; also for the record, there are scholars who think that the 24 books and 24 Ionic letters are just a splendid coincidence.
Note. On the division into 24 books, see Jensen et al. 1999: 35–83 for a spread of diverse views; Taplin 1992: 285–293, Richardson 1993: 20–21 for quicker overviews. Diverse as their viewpoints are, hardly any scholars ever question the Ionic doctrine, that the first transcription of the epics used the 24-letter Ionic alphabet. Cf. Cauer 1921: 99–105; Reece 2011; Gainsford 2015: 69–71. Students of ancient Greek are heavily pushed to ignore the existence of non-Ionic alphabets (all textbooks claim to be teaching Attic Greek, but none use Attic letters).

Datings of Homer are differentiated not just by the evidence they choose to focus on, but by how they arrange these points de capiton.

  1. Stylometric models focus on the Composition and Fixation Dates. The Composition Date represents early linguistic strata in an oral tradition (e.g. formulas in the Aeolic dialect, if you believe in the ‘Aeolic phase’), while the Fixation Date represents later forms (e.g. formulas that only work in the Ionic dialect).
  2. Terminus models often assume that an epic was composed directly into a fixed, and often a written, text: that is, the Composition and Fixation Dates are the same thing. If you studied classics at an English-speaking university, you were probably taught that this is an obsolete view, disproved by studies of oral traditions. Like it or not, it isn’t fringe. Many modern datings of Homer assume explicitly that the epics were first composed in writing. (I’m not saying that’s right: but it isn’t fringe.)
  3. Phase models treat Composition and Fixation as strictly separate. Nagy’s phase model also involves the Dissemination Date (Nagy 1996: 44–63).
Rembrandt, Homer dictating his verses (1663; The Hague, Mauritshuis; source: Wikipedia)

Different scholars, different quilting

Let’s look at some examples, and see how they shuffle around the points de capiton in relation to one another.

Wolf’s (1795) oral transmission model was innovative specifically because he treated Composition and Transcription as separate events. That is: for Wolf, the Composition Date and Fixation/Transcription Date were separated by a considerable interval of time: in between, the epics were transmitted orally.

This basic principle led to a common motif in 19th–20th century scholarship: the good bits of Homer are the bits that Homer himself composed; the bad bits were by his later redactors (if you were an Analyst), or actually they’re secretly the best bits (if you were a Unitarian). This nonsense was a plague on 20th century Homeric scholarship. It thoroughly permeated the Analyst and Unitarian schools of thought, and it’s largely responsible for their existence in the first place. If you find the dating of Homer frustrating, because the scholarship is so subjective, this motif is to blame.

That said, advocates of phase models also separate the Composition and Fixation Dates: such as Kirk (1962: 282–287, 1976: 32–39), Jensen (1980, 2011), Seaford (1994: 144–154), Fowler (2004), and Rutherford (2013: 23–35). This doesn’t mean they all come up with the same dates, of course, and there’s lots of wiggle room. Jensen argues that the Fixation Date is necessarily also the Transcription Date (Jensen 2011: 227–230), and she puts Transcription very late: she doesn’t assign a constraint earlier than the Persian Wars. That moves her Fixation Date very late too.

Many modern scholars explicitly argue that the epics were written down at the same time they were composed. This includes for example Parry (1966), West (1995, etc.), Berg and Haug (2000), and Teodorsson (2006). Their datings are quite different in other respects, but what they have in common is that they treat the Composition and Fixation Dates as the same thing, in spite of Wolf’s legacy.

Nagy’s phase model treats text fixation as causally related to diffusion: the more widespread an epic’s circulation, the more popular it is, the more the text becomes fixed. Nagy uses the development of Indian legends and poetry as a paradigm (1996: 44–53). As a result, for Nagy, Fixation and Dissemination occurred simultaneously, in conjunction with one another.

For terminus models, the best example is that of Van Wees (1994). West (2011: 16–19) is also worth mentioning, partly for his prestige, but also because his results agree with Van Wees: they both put the Iliad between 670 and 650 BCE. West explicitly believes that Composition, Fixation, and Transcription took place simultaneously, though he envisages the same poet writing and re-writing an epic over a period of years. Van Wees doesn’t comment on Composition versus Fixation, but since his argument relates to incidental details — military equipment, not plot design — it makes sense to see his argument as pointing to a Fixation Date.

West cites many constraints, but each of them relates to a single passage, or an isolated detail. As a result they can be cast in doubt by a counter-argument: the story and text underwent ongoing evolution, beforehand and/or afterwards, so isolated passages aren’t representative; or they could be interpolations. By contrast, Van Wees’ constraints permeate the entire epic. They relate to every battle-scene. So there’s no question of ongoing evolution or interpolation. Of all attempts that have ever been made to produce a Fixation Date for the Iliad, Van Wees’ case is by far the strongest.

On to stylometry. Janko’s analysis (1982) is the best known: he comes up with a Composition Date in the early 700s BCE, and a Fixation Date of 750–725. Janko has been criticised in several ways, but the thing to highlight here is that stylometric analysis creates a relative timeline. Janko arranges poems along a chronological spectrum: fine. But then to pin that spectrum to a fixed date-range, he has to use a terminus model. And, as so often with terminus models, the choice of constraints makes all the difference. His three constraints (1982: 230–232) have no overlap with West’s; select different constraints, and you come up with different dates.

Out-of-date though his stylometric techniques are, Janko’s is still the best effort. Other attempts at stylometric dating of Homer have been half-baked, often with no awareness of statistical methods. Blößner (2006) argues against stylometry as a valid approach to dating Homer at all. Other than Blößner’s negative findings, there’s been only one new stylometric analysis in the 21st century so far, by Altschuler et al. (2013), and ... it’s really bad. (Their 95% confidence interval for the date of the Iliad ranges from 1157 BCE to 376 BCE. Thanks for nothing!)

Priam begging Achilles for the body of his son Hector, as depicted in Iliad 24 ... oops, except this isn’t as in Iliad 24. Who’s that at the left? Probably Andromache. In which case, this scene may have more in common with the Aethiopis than with the Iliad. (Amphora, 560–540 BCE; Kassel T 674; cf. Gainsford 2015: 99–102, 121 n. 64)

Settling on actual dates

You’ll notice that we haven’t seen much reason to favour specific datings so far, other than my opinion that Van Wees has settled the Fixation Date very firmly.

On the Composition Date, there’s very little to agree on. It’s probably fair to say that most scholars still want to put it before 700 BCE. But the main thing keeping it that early is ‘dogmatic drag’, as West has called it: tradition keeps present-day scholars glued to the early datings that their predecessors preferred. If I’m to express a view of my own, then I’d say that I find it hard to imagine a Composition Date earlier than the Greek resettlement of Ilion in the 700s, the ethnic tensions between Greek settlers and mixed indigenous peoples, the Greek cult of Ilian Athena, and so on. Maybe composition in the 700s is still sustainable, but I don’t see any reason to prefer an early date. The 600s are much easier in every way.

But nearly everyone agrees on the Dissemination Date. And it’s late — later than you were taught in college. You’d have to explore some pretty fringe views to find someone who puts Homer’s rise to popularity before the late 500s BCE in Athens. This is partly thanks to the work of Burkert (1976) and West (1995), who argue for Panathenaic performances of Homer starting in 522 BCE, and partly thanks to increasing awareness that the Iliad and Odyssey had no influence at all on the visual arts until the late 500s (Snodgrass 1998: 67–100; Burgess 2001; Lowenstam 2008: 4–5; Jensen 2011: 237–244). It’s helped by the fact that the only earlier direct mention of Homer that we have, reported in Herodotus 5.67, is referring to the Thebaid, not the Iliad or Odyssey.

When scholars with such diverse views as West, Nagy, Jensen, Burgess, and Burkert all agree on a thing, it’s probably fair to say: that’s a pretty strong consensus.

In turn, this means that we have to take it seriously when we’re told that Homer wasn’t performed in Syracuse until 504/1 BCE (Hippostratus, FGrHist 568 F 5). Similarly that Homeric battle scenes were first sung at Delphi by Stesandrus (Timomachus, FGrHist 754 F 1) — we don’t know Stesandrus’ date, but 500s BCE is likely. There are grounds for seeing earlier Dissemination Dates for parts of Homer in parts of the Mediterranean world. West (2001: 10) thinks there is evidence that Iliad book 10, the ‘Doloneia’, was known in the Peloponnesos ca. 600 BCE. Snodgrass (1998: 95–98) interprets an Etruscan vase from ca. 625 BCE, showing the blinding of the Cyclops, as relying on specifics that are peculiar to the surviving text of Odyssey book 9, and which do not appear in Greek depictions of the same theme until over a century later. And, yes: the implication is that the Odyssey as we know it, post-Fixation, may well have been known to Etruscans a whole century before it became popularised in the Greek mainland.

References

  • Altschuler, E. L. et al. 2013. ‘Linguistic evidence supports dates for Homeric epics.’ Bioessays 35: 417–420. [DOI link]
  • Berg, N.; Haug, D. 2000. ‘SO debate. Dividing Homer (continued). Innovation vs. tradition in Homer — an overlooked piece of evidence.’ Symbolae Osloenses 75: 5–23. [DOI link]
  • Blößner, N. 2006. ‘Relative Chronologie im frühgriechischen Epos: eine empirische Methode.’ Zetemata 125: 19–46.
  • Burkert, W. 1976. ‘Die hunderttorige Theben und die Datierung der Ilias.’ Wiener Studien 89: 5–21.
  • Cauer, P. 1923. Grundfragen der Homerkritik, 3rd ed. (1st ed. 1895). Leipzig. [Internet Archive link]
  • Crespo, E. 2014. ‘La copa de Néstor y la datación de la Ilíada.’ In: Bádenas de la Peña, P., et al. (eds.) Homenaje a Ricardo Olmos. Per speculum in aenigmate. Miradas sobre la Antigüedad. Madrid. 73–78. [Academia.edu link]
  • Fowler, R. 2004. ‘The Homeric Question.’ In: Fowler, R. (eds.) The Cambridge companion to Homer. Cambridge. 220–232.
  • Gainsford, P. 2015. Early Greek hexameter poetry. Cambridge. [Cambridge Core link]
  • Graziosi, B. 2002. Inventing Homer. The early reception of epic. Cambridge.
  • Janko, R. 1982. Homer, Hesiod, and the Hymns. Diachronic development in epic diction. Cambridge.
  • Jensen, M. S. 1980. The Homeric Question and the oral-formulaic theory. Copenhagen.
  • —— 2011. Writing Homer. A study based on results from modern fieldwork. Copenhagen.
  • Jensen, M. S. et al. 1999. ‘SO debate. Dividing Homer. When and how were the Iliad and the Odyssey divided into songs?’ Symbolae Osloenses 74: 5–91. [Taylor & Francis link]
  • Kirk, G. S. 1962. The songs of Homer. Cambridge.
  • —— 1976. Homer and the oral tradition. Cambridge.
  • Nagy, G. 1996. Homeric questions. Austin (TX).
  • Parry, A. 1966. ‘Have we Homer’s Iliad?’ Yale Classical Studies 20: 175–216. Reprinted in: Wright, John (ed., 1978) Essays on the Iliad: selected modern criticism. Bloomington, London. 1-27 and 128-34.
  • Powell, B. 1991. Homer and the origin of the Greek alphabet. Cambridge.
  • Reece, S. 2011. ‘Metacharacterism.’ In: Finkelberg, M. (ed.) The Homer encyclopedia. Chichester. 514–515.
  • Richardson, N. 1993. The Iliad: a commentary. Volume VI: books 21–24. Cambridge.
  • Rutherford, R. 2013. Homer, 2nd ed. (1st ed. 1996). Cambridge.
  • Taplin, O. 1992. Homeric soundings. The shaping of the Iliad. Oxford.
  • Teodorsson, S.-T. 2006. ‘Eastern literacy, Greek alphabet, and Homer.’ Mnemosyne 59.2: 161–187. [JSTOR link]
  • Van Wees, H. J. 1994. ‘The Homeric way of war: the Iliad and the hoplite phalanx.’ Greece & Rome 41.1: 1–18, 41.2: 131–155. [JSTOR: link 1, link 2]
  • —— 1999. ‘Homer and early Greece.’ In: De Jong, I. J. F. (ed.) Homer. Critical assessments, vol. 2. London. 1–32.
  • West, M. L. 1995. ‘The date of the Iliad.’ Museum Helveticum 52: 203–219. [ETHzürich link]
  • —— 2001. Studies in the text and transmission of the Iliad. Munich, Leipzig.
  • —— 2011. The making of the Iliad. Oxford.
  • —— 2014. The making of the Odyssey. Oxford.
  • Wolf, F. A. 1795. Prolegomena ad Homerum, sive De operum homericorum prisca et genuina forma variisque mutationibus. Halle. English translation, 1985: Prolegomena to Homer. 1795. Trans. A. Grafton, G. W. Most, J. E. G. Zetzel. Princeton (NJ).

Sunday, 7 November 2021

The dates of Jesus. 4. Sources

§1–§3. The dates | §4–§5. Calendars and prophecy | §6. Christmas and Easter | §7–§8. Sources

§7. Ancient sources on the dates of Jesus

I originally wanted to compile all the testimonia here, in their original languages and in translation, with commentary. But it’s far too large. I try to keep pieces on this site close to 2000 words: the full version of the compilation is currently pushing 25,000 words.

In future I’ll try to publish the full compilation as a mini-monograph. Here, I’ll give a tabulation of the sources, with links, but without my own translations; and just a few select notes below.

The death of Jesus as depicted in the Rabbula gospels, 586 CE (Florence, Bibl. Medicea Laurenziana Plut.01.56, fol. 13v)

My cut-off point is Epiphanius in the late 300s. There’s no authentic information about Jesus after 250 CE, and by that time the dates for liturgical observances of Jesus’ birth and death were close to being settled. But it would be strange to leave out the Chronography of 354, which is often cited as the earliest direct evidence of Christmas on 25 December. (Actually it’s Hippolytus, over a century earlier.) John Chrysostom needs inclusion because he shows that the Syrian church could celebrate Christmas on 25 December just as the Roman church did, and not necessarily on 6 January as attested in connection with Ephraim of Syria (T38). Epiphanius needs to be here because he’s the earliest adherent of the History of Religions Theory (see Episode 3, §6).

  Source Source date Notes
T1 Mark 14.1–2, Matthew 26.1–5, Luke 22.1–2 ca. 70–100 one Passover in Jesus’ ministry (Short Chronology) (see note below)
T2 Matthew 2.1–3, 19–22 ca. 70–90? birth just before Herod’s death (4 BCE)
T3 John 2.13, 6.4, 11.55 ca. 80–110? three Passovers in Jesus’ ministry (Long Chronology)
T4 Josephus, Jewish antiquities 17.35518.2 90s Quirinius’ census (see note below)
T5 Josephus, Jewish antiquities 18.63–64 90s testimonium flavianum; death during Tiberius’ reign (14–37 CE), Pilate’s governorship
T6 Luke 1.5, 26–31 ca. 100? conception during Herod’s reign (4 BCE or earlier)
T7 Luke 2.1–2 ca. 100? birth at time of Quirinius’ census (6 CE)
T8 Luke 3.1–3, 3.21–23 ca. 100? baptism in Tiberius 15 (29 CE) at age 30
T9 Luke 23.44–46 ca. 100? solar eclipse at death (see Episode 1, §3)
T10 Tacitus, Annals 15.44 110s death during Tiberius’ reign, Pilate’s governorship
T11 Justin Martyr, First apology 13 150s death during Tiberius’ reign, Pilate’s governorship
T12 Irenaeus, Against heresies 2.20.1, 2.22.1 mid-100s (180s) Valentinian beliefs: death at age 30; ministry lasting 12 months; numerological symbolism
T13 Irenaeus, Against heresies 1.27.2, 4.6.2, 4.22.2 180s baptism during Tiberius’ reign, Pilate’s governorship
T14 Irenaeus, Against heresies 2.22.3 180s argument for Long Chronology; Passovers in John 2.23, 5.1, 11.55 (but not 6.4!)
T15 Irenaeus, Against heresies 2.22.5–6 180s death aged 40–49
T16 Irenaeus, Against heresies 3.21.3 180s birth in Augustus 41 (4 BCE?)
T17 Irenaeus, Demonstration 74 180s–190s death during Claudius’ reign (41–54 CE)
T18 Clement of Alexandria, Stromateis 1.21.144–146 ca. 200 count of regnal years; various exact dates for genesis, baptism, and death; discussion of prophecies in Daniel
T19 Tertullian, Against the Jews 8 ca. 200–220 count of regnal years; Daniel’s 70 weeks; birth in Augustus 41; birth 28 years after Cleopatra’s death; death on 25 March, consulship of Gemini
T20 Tertullian, Apologetic 21.19 ca. 200–220 rejects solar eclipse at death
T21 Julius Africanus, Chronographiae T92 Wallraff 220s birth in AM 5500 = Augustus 42 = 2 BCE
T22 Julius Africanus, Chronographiae F93 Wallraff 220s rejects solar eclipse at death; parousia in AM 5531 (implying Short Chronology)
T23 Julius Africanus, Chronographiae T93c Wallraff 220s birth in AM 5500; death in AM 5531 (Africanus T93d: AM 5532)
T24 Hippolytus, paschal table 222–224 genesis on 2 April 2 BCE; death on 25 March 29 CE
T25 Hippolytus, Commentary on Daniel fr. 4 1st half of 200s birth on 25 Dec., Augustus 42; death on 25 March, Tiberius 18
T26 Origen, Homilies on Luke fr. 108 Rauer ca. 240s birth in Augustus 41; baptism in Tiberius 15
T27 Origen, Against Celsus 2.33 ca. 240s death during Tiberius’ reign; solar eclipse at death
T28 De Pascha computus (pseudo-Cyprian) 18–20 243 birth on 28 March 2 BCE; death in spring 30 CE; sun symbolism; numerological symbolism
T29 Lactantius, Institutiones divinae 4.10.18 304–311 death on 23 March 29 CE (or 31 CE?)
T30 Lactantius, De mortibus persecutorum 2 313–316 death on 23 March, consulship of Gemini
T31 Eusebius, Chronicle, Ol. 194,4 (Latin, Armenian) ca. 310s–330s birth in Ol. 194,4 (1 BCE/1 CE)
T32 Eusebius, Chronicle, Ol. 201,4–202,4 (Latin, Armenian) ca 310s–330s years of baptism, ministry, and death (different years in Latin and Armenian versions); solar eclipse at death
T33 Chronography of 354, Fasti consulares, 1 CE 336–354 birth on 25 Dec. 1 CE
T34 Chronography of 354, Fasti consulares, 29 CE 336–354 death in 29 CE on 14 Nisan
T35 John Chrysostom, Oration on the birthday of our saviour Jesus Christ 5 2nd half of 300s birth in December
T36 Epiphanius, Panarion, on the incarnation 1.4–2.7 377 birth in Augustus 42 = Herod 33; baptism in Herod Agrippa 18 (= 31 years after birth); death in Herod Agrippa 20
T37 Epiphanius, Panarion 51.16.1–2 377 birth on 6 Jan.; baptism on 8 Nov.
T38 Epiphanius, Panarion 51.22.3–13 377 Ephraim of Syria’s teaching; birth on 6 Jan., Augustus 42; birth in consulship of Augustus and Silvanus (2 BCE); numerological symbolism; correspondences in pagan festivals

Notes

Short and Long Chronologies

The synoptic gospels mention only one Passover in the course of Jesus’ ministry, namely the one at which he died (T1). In contrast John mentions three (T3). As a result ancient sources adopt either a Short Chronology or a Long Chronology of Jesus’ ministry. As Christian chronography developed, and as the date of Easter was scrutinised more carefully beginning with the Quartodeciman synods in the 190s, the existence of two chronologies caused confusion.

In the Short Chronology Jesus’ death takes place the same year as his baptism: that is, as per Luke 3, Tiberius 15, or the consulship of the Gemini, which was in 29 CE. In this scheme Jesus’ ministry is contained within the period of a year (T19, T22–23, T24, T34, T36). The most common date that emerged was 25 March 29 CE, which had the further benefits of (a) being the traditional day of the equinox, tying in nicely with sun symbolism associated with the Messiah (T28); (b) falling on a Friday, the day before the Sabbath, since the gospels depict Jesus dying on that day of the week.

A variant of the Short Chronology puts the death date at Passover the following year (T28). This belief was most explicitly held by the Valentinians in the 2nd century CE, who held that Jesus’ ministry lasted exactly 12 months, numerologically representing the 12 apostles (T12).

For Christians who preferred the Long Chronology, Jesus’ ministry had to last at least two years: Passover to Passover to Passover (T25, T32, T36), or even longer (Irenaeus: T14–15, T17). As a result Sunday 25 March 31 CE appears as a competing date (T25) — but for Jesus’ resurrection, not his death, since 25 March fell on a Sunday that year. But the situation was confused. Hippolytus put Jesus’ death 30 years after his genesis (T24), which ought to be 29 CE in the chronology he gives; but elsewhere he assigns it to Tiberius 18 (T25), which ought to be 3 years later. Lactantius puts Jesus’ resurrection on 25 March (T29–30), but he assigns it to the year of the consulship of the Gemini: in that year, 29 CE, 25 March fell on a Friday, not a Sunday. In these cases we must be looking at a Long Chronology contaminated with elements of the Short Chronology, or vice versa.

Irenaeus explicitly adheres to an exceptionally Long Chronology. He concocts an argument against the Valentinians to assert that Jesus’ ministry continued well into his 40s (T14–15), and consequently that he died during Claudius’ reign (T17). This argument must have been formulated before the Quartodeciman synods of the 190s, and was roundly ignored in later centuries.

Conversely Clement, a couple of decades later, spurns the Long Chronology altogether, and refuses even to acknowledge John’s three Passovers, instead claiming that Isaiah’s ‘acceptable year’ (Isaiah 61.2; also cited by the Valentinians, T12) was the duration of Jesus’ ministry.

Quirinius’ census

When Archelaus was ousted from his position as ethnarch of Judaea, Syria was extended to include Judaea, and Quirinius was appointed to assay the province’s wealth and sell off Archelaus’ estate. Both Josephus (T4) and Luke (T7) place the census immediately after Quirinius’ appointment (Luke 1.2: αὕτη ἀπογραφὴ πρώτη ἐγένετο ἡγεμονεύοντος τῆς Συρίας Κυρηνίου, ‘This was the census that took place at the very start of Quirinius’ governorship of Syria’). It would be incoherent to imagine Quirinius assaying Archelaus’ property while Archelaus was still in power, so Luke’s phrasing corroborates Josephus’ report that the census was immediately after Archelaus’ departure: that is, in 6 CE.

Luke’s portrayal of the census draws on two precedents:

  • A Hebrew model: the first census of the Israelites depicted in Numbers 1.1–4. This model provides the motif of counting people according to their ancestral house, and motivates Joseph's temporary move to Bethlehem.
  • A Roman model: Augustus’ censuses of Roman citizens in 28 BCE, 8 BCE, and 14 CE, reported in Res gestae 8. These provide the motif of counting ‘the entire world’, even though virtually no residents of the Roman empire outside Italy were Roman citizens.

Quirinius’ census was not just an assay of property, as Josephus states, but also a head-count of people as in Luke. This is shown by the funerary inscription of Q. Aemilius Secundus, who conducted a census of Apamea at Quirinius’ order (Corpus inscriptionum latinarum iii 6687). However, Luke is badly mistaken in omitting the reign of Archelaus, in imagining that a census of Roman citizens might have included Judaeans, and in imagining that such a census might have followed a model set by the Hebrew Bible.

Hippolytus

Hippolytus needs special comment because (a) his paschal table is unintelligible without explanation, and (b) the reference in his Commentary on Daniel (T25) has at times been thought a mediaeval interpolation. First, the paschal table, here in translation:

In year 1 of emperor (Severus) Alexander’s reign [222 CE], the 14th day of the paschal moon took place on Saturday 13 April, after an intercalary month (embolismos). In future years it will take place as laid out in the table. In past years it took place as shown. The (lenten) fast must be broken when Sunday falls.
em. 13 April Sat. Fri.
Esdras acc. to Daniel; in the desert
Thu. Wed. Tue. Mon. Sun.
  2 Apr. Wed.
genesis of Christ
Tue. Mon. Sun. Sat. Fri. Thu.
B 21, 22 Mar. Sun.
Hezekiah
Sat. Fri.
Joshua
Thu. Wed. Tue. Mon.
em. 9 Apr. Sat.
Josiah
Fri. Thu. Wed. Tue. Mon. Sun.
  29 Mar. Wed. Tue. Mon. Sun. Sat. Fri. Thu.
  18 Mar. Sun. Sat.
Hezekiah acc. to Daniel; Josiah
Fri. Thu. Wed. Tue. Mon.
B
em.
13 Apr. Sat. Fri. Thu. Wed.
Joshua acc. to Daniel
Tue. Mon. Sun.
  25 Mar. Wed. Tue. Mon. Sun. Sat. Fri. Thu.
em. 13 Apr. Tue. Mon. Sun. Sat. Fri. Thu. Wed.
  2 Apr. Sat. Fri. Thu. Wed. Tue. Mon. Sun.
exodus
B 21, 22 Mar. Wed. Tue. Mon. Sun. Sat. Fri. Thu.
em. 9 Apr. Tue. Mon. Sun. Sat. Fri. Thu. Wed.
in the desert
  29 Mar. Sat. Fri. Thu. Wed. Tue. Mon. Sun.
  18 Mar. Wed. Tue. Mon. Sun. Sat. Fri. Thu.
Esdras
B
em.
5 Apr. Tue.
exodus acc. to Daniel
Mon. Sun. Sat. Fri. Thu. Wed.
  25 Mar. Sat. Fri.
suffering of Christ
Thu. Wed. Tue. Mon. Sun.

This computus is inscribed on the side of a statue of Hippolytus seated in a chair. It is generally agreed to be based on one of his lost works (the back of the statue gives a list of his writings).

Column 1 indicates leap years in the Julian calendar (‘B’ in my translation), and years where the lunar year, of 12 29½-day months, requires an intercalary lunar month to compensate for the shortfall of 11¼ days each solar year (‘em.’ in my translation). Note that these figures are based on ancient reckoning and are not nearly accurate enough for the purpose of this table. Column 2 purports to give the Julian date of the equinoctial full moon in a 16-year cycle. I say ‘purports’, because moon phases do not in reality follow a 16-year cycle. Columns 3 to 9 indicate the weekday for the date in column 2, in seven cycles of 16 years. Weekday matters for liturgical purposes because the Roman church celebrated Easter on the following Sunday, as the inscription’s header indicates. Hippolytus’ full moon dates are not accurate, but his weekdays are.

222 CE corresponds to the first row, third column. Each year you move down one row, and after 16 years you move to the top of the next column. The end of the 7th column corresponds to the 112th year (= 333 CE), and you return to the beginning of the table.

Hippolytus retrojects his 112-year cycle with earlier iterations running 110 CE to 221 CE, 3 BCE to 109 CE, and so on. The table attaches specific biblical events to equinoctial full moons in earlier iterations. The relevant ones for our purposes are the ones relating to Jesus in the iteration running 3 BCE to 109 CE: ‘genesis of Christ’ on Wednesday 2 April in the second year of the cycle, that is, 2 BCE; and ‘suffering of Christ’ on Friday 25 March in the 32nd year of the cycle, that is, 29 CE.

The other events marked in the table refer to Passovers described in the Hebrew Bible, in Exodus 12, Numbers 9, Joshua 5, and 2 Chronicles 30 and 35; and in the Septuagint in 1 Esdras 1.

The table was compiled between 222 and 224 CE. The 222 CE cut-off is given by the table header, which refers to the start of Severus Alexander’s reign; the 224 CE cut-off is given by the fact that Hippolytus’ full moon dates are only accurate for the years 216 to 224 CE. (For accurate full moon dates see Espenak 2014.)

Next, the Commentary on Daniel fragment. Kellner (1901: 94–95) argued that this must be an interpolation dating to the 9th–10th centuries, on the grounds that (a) the regnal year and consular year quoted in T25 disagree with the year implied by T24; (b) T25 cites multiple era systems and regnal year systems, and this is characteristic of 9th–10th century computistic research, and not possible in a 3rd century writer. Roll agrees, but without close inspection of the evidence (1995: 79-81).

On the second point, Kellner is simply wrong: ancient figures like Julius Africanus and Eusebius clearly illustrate the careful (and erroneous) work done in antiquity on synchronising era and regnal year systems. On the first point, Kellner’s point is undercut by the confusion caused by variation between the Short and Long Chronologies. I mentioned above that Lactantius’ dates for Jesus’ death and resurrection display the same confusion of years; and the dates in T25 reappear in almost identical form in the 5th century Acts of Pilate (prologue), and that means that the error is an ancient one.

In the past I have followed Roll in rejecting T25 as a mediaeval interpolation. But after closer inspection of the evidence I am now persuaded that there is no good reason to see anything spurious about the fragment.

Other topics

Ideally I would like to add notes on the numerological symbolism adopted by the Valentinians, the De Pascha computus, and Ephraim of Syria (T12, T28, T38); Irenaeus’ ultra-Long Chronology; the relationship between the Armenian celebration of Jesus’ birth on 6 January, and the 25 December date used by everyone else; and on the counts of regnal years in Clement and Tertullian (T18, T19), and how they differ from reality. Even without these notes, this episode is running to 3000 words. Just addressing Tertullian’s regnal year count would require a lengthy discussion of textual corruptions.

I hope to publish this material in a more formal venue, so in the meantime I shall just have to say that these themes are important; they need to be dealt with as part of the history of early Christianity; and, in case it needs saying explicitly, they tell us nothing whatsoever about the historical Jesus.

I’ll settle for the main narrative I’ve put forward here. The gospels are the only primary sources that anyone ever used for Jesus’ dates; 1st–2nd century witnesses gave only vague and symbolic dates, until liturgical practices demanded more specific dates because of the Quartodeciman synods of the 190s; the solar eclipse of 29 CE was, for a while, an important tool for pinning down more precise dates; and most of the dating discrepancies that we see from Hippolytus onwards can be explained in terms of numerological symbolism, liturgical practices, and conflation of the Short and Long Chronologies.

§8. References

  • Bickerman, E. J. 1980. Chronology of the ancient world, 2nd edition (1st edition 1968). Ithaca (NY).
  • Eshel, H. 2005. ‘4Q390, the 490-year prophecy, and the calendrical history.’ In: Boccaccini, G. (ed.) Enoch and Qumran origins. Grand Rapids (MI). 102–110.
  • Espenak, F. 2014. ‘Six millennium catalog of phases of the moon. Moon phases from –1999 to +4000 (2000 BCE to 4000 CE).’ Astropixels.com. (Retrieved Oct. 2021)
  • Förster, H. 2007. Die Anfänge von Weihnachten und Epiphanias. Eine Anfrage an die Entstehungshypothesen. Tübingen.
  • Hijmans, S. E. 2009. Sol. The sun in the art and religions of Rome. Diss. Groningen. [ResearchGate link | Rijksuniversiteit Groningen link]
  • Kellner, K. A. H. 1901. Heortologie, oder das Kirchenjahr und die Heiligenfeste in ihrer geschichtlichen Entwicklung. Freiburg. [Internet Archive link]
  • Mosshammer, A. A. 2008. The Easter computus and the origins of the Christian era. Oxford.
  • Nothaft, C. P. E. 2011. Dating the Passion. The life of Jesus and the emergence of scientific chronography (200–1600). Leiden.
  • —— 2012. ‘The origins of the Christmas date: some recent trends in historical research.’ Church History 81: 903–911. [DOI link]
  • —— 2013. ‘Early Christian chronology and the origins of the Christmas date. In defense of the “calculation theory”.’ Questions Liturgiques 94: 247–265. [DOI link]
  • Olson, K. A. 1999. ‘Eusebius and the testimonium Flavianum.’ Catholic Biblical Quarterly 61: 305–322. [JSTOR link]
  • Roll, S. K. 1995. Towards the origins of Christmas. Kampen.
  • —— 2000. ‘The origins of Christmas: the state of the question.’ In: Johnson, M. E. (ed.) Between memory and hope. Readings on the liturgical year. Collegeville (MN). 273–290.
  • Samuel, A. E. 1972. Greek and Roman chronology. Munich.
  • Schmidt, T. C. 2015. ‘Calculating December 25 as the birth of Jesus in Hippolytus’ Canon and Chronicon.’ Vigiliae Christianae 69: 542–563. [JSTOR link]
  • Talley, T. J. 1991. The origins of the liturgical year, 2nd edition (1st edition 1986). Collegeville (MN).
  • Wallraff, M. 2007. Iulius Africanus. Chronographiae. The extant fragments. Berlin.
§1–§3. The dates | §4–§5. Calendars and prophecy | §6. Christmas and Easter | §7–§8. Sources