Library : Modern Documents : Richard Carrier : Thallus

Thallus: an Analysis (1999)

Richard Carrier


This is a preliminary essay, outlining some important facts about Thallus, a pagan chronologer of unknown date who is occasionally mentioned in the works of Christian apologists, modern and ancient, as a 1st century pagan witness to the gospel tradition of a "darkness" at the death of Christ: see Mark 15:33; Luke 23:44; and Matthew 27:51-53, whose account includes an earthquake, split rocks, and zombies; John makes no mention of any such events, nor does Paul or any other New Testament author. Such a story has obvious mythic overtones and can easily be doubted. That a solar eclipse should mark the death of a king was common lore among Greeks and other Mediterranean peoples (Herodotus 7.37, Plutarch Pelopidas 31.3 and Aemilius Paulus 17.7-11, Dio Cassius 55.29.3, John Lydus De Ostentis 70.a), and that such events corresponded with earthquakes was also a scientific superstition (Aristotle Meteorology 367.b.2, Pliny Natural History 2.195, Virgil Georgics 2.47.478-80). It was also typical to assimilate eclipses to major historic events, even when they did not originally correspond, or to invent eclipses for this purpose (Préaux claims to have counted 200 examples in extant literature; Boeuffle and Newton have also remarked on this tendency). The gospel stories also make a solar eclipse impossible: the crucifixion passover happened during a full moon, the darkness supposedly lasted three hours, and covered the whole earth. Such an impossible event would not fail to be recorded in the works of Seneca, Pliny, Josephus or other historians, yet it is not mentioned anywhere else outside of Christian rhetoric, so we can entirely dismiss the idea of this being a real event.

Nevertheless, Thallus is cited at least as a witness to the early date of the gospel story of the darkness, if not to the factuality of the darkness itself. But the facts surrounding Thallus are all too often incorrect, or asserted with unjustified boldness, calling for a proper historical treatment of the facts. Being an unfinished work, this essay is brief and does not cite sources in detail--most of the relevant sources are already cited in my translation of two other non-English commentaries on Thallus: for those who can stomach the detailed scholarly text, see Jacoby. A bibliography of all the works I consulted (but not including those mentioned by Jacoby or Müller) is collected at the end of this essay. In the future, a completed version of this essay with full references will be published, but until then anyone desiring to know more about the sources should write to me directly (use the e-mail link above or below).

Introduction to the Problem

We know next to nothing about Thallus or his works. We don't even know if he wrote only one book or several. The only information we have about him, even his name, comes entirely from Christian apologetic sources beginning in the late 2nd century, and that information is plagued with problems. Scholars since the 18th century have even invented facts about him, and some of these groundless notions--like the idea that he was a Samaritan--are repeated even today. Claims are also made, mainly but not exclusively by modern Christian apologists, which make Thallus into the earliest literary witness to the gospel tradition. There are two commentaries which collect and discuss all remnants and attestations, but they are not available in English, and are outdated (I have translated and commented on these: see Jacoby). A new survey of what we know about Thallus needs to be made available in English, and this essay is the first, brief draft toward that aim.

What Thallus Wrote

On various occasions Thallus' work is referred to as the Histories (HISTORIAI in George Syncellus, and possibly in Eusebius, quoting Julius Africanus; HISTORIA in Theophilus, and in Lactantius quoting Theophilus, and possibly in Eusebius quoting Julius Africanus). This is the only work by Thallus of which we have a title. Attempts to reconstruct a title from an Armenian translation of a Greek chronology composed by Eusebius are not trustworthy, since the text seems only to describe the work, not name it, in a list of authors which sometimes names their works and sometimes merely describes them. But whether a discription or a title, the work described in that text does not appear to be the same work quoted by everyone else. This is because it is described as a "brief compendium" (in three volumes, which is indeed exceedingly brief--equivalent to three chapters in a modern book) covering the years from the fall of Troy (1184 BC) to the 167th Olympiad (109 BC), but Thallus is often cited for events long preceding the Fall of Troy, and on one occasion appears to be cited regarding an event at the death of Christ, which comes long after 109 BC (leading several scholars to amend the text to give a later date). In all cases the nature of the facts being drawn from Thallus further suggests a rather detailed work, and not a "brief compendium." It is most likely that the book referenced by Eusebius is one of at least two works by Thallus, and not the work in which he mentions the darkness associated with the death of Christ (if he mentioned this at all). This is an altogether more likely explanation than the many alternatives that have been suggested (for example, that a later, anonymous author expanded Thallus' work).

As for what Thallus wrote about, we are told by Eusebius, quoting Julius Africanus, that Thallus recorded Syrian history just as Castor did, which is consistent with other remarks by Tertullian placing Thallus among historians of Eastern events, and with several authors who cite Thallus on details of Assyrian history (Theophilus, Lactantius, George Syncellus) and with another who possibly cites him as an expert on Lydian affairs (Malalas). But Thallus is also listed among those who recorded Greek and Roman history, especially the deeds of Saturn in Italy (Tertullian, Lactantius, Minucius Felix). To confuse matters further, the late forger of a work in the name of Justin Martyr claims Thallus among those who mentioned Moses and the antiquity of the Jews in the context of Athenian history! This last can be dismissed, however, since the forged text is almost a word-for-word adulteration of a quote from Julius Africanus, which we have more reliably preserved in the works of Eusebius, which merely lists Thallus, with Castor, as a reliable historian of Syrian affairs and nothing more.

When Thallus Wrote

The Armenian reference places the end of Thallus' "brief compendium" at the 167th Olympiad (which spans 112-109 BC). This would remain uncontested if it were not for a single reference to Thallus regarding an event long after that time: namely, the darkness at the death of Christ. Since this event must have occurred in the 1st century AD, and no doubt sometime between 28 and 38 AD, there are two possibilities: either the Armenian text is referring to a different work, or the date has been corrupted. Virtually every scholar to date has opted for the latter and made efforts to conjecture the original date--the only two plausible (though still unlikely) options are the 207th Olympiad (which spans 49-52 AD) and the 217th Olympiad (which spans 89-92 AD). The latter in fact is the more likely, judging from palaeography. But as I've already noted, it seems far more likely that the Armenian reference is to a different work. It could even be an excerpted epitome of a longer chronology.

This leaves us with no clue as to when Thallus wrote. Since the 1st-century darkness was probably not mentioned in the "brief compendium," there is no reason to suppose that the date of 109 BC is incorrect--there is nothing physically wrong with the text, nor any other reason to suspect an error (although Mosshammer claims otherwise, his reasoning is hard to justify). However, if Thallus did mention the darkness in another work (probably the Histories), he clearly had to have written after 28 AD. Although the guess of 52 AD as the end-date for the compedium is the one most commonly mentioned, if the date is wrong at all then 92 AD is more likely correct. But all these possible dates--109 BC, 52 AD, 92 AD--only give us the "time after which" he had to have written this "brief compendium." These dates do not tell us when he wrote the Histories or whatever work that mentioned the darkness.

It is also supposed that the final date covered by the compendium should be close to the time the compendium was written, but that also does not follow. Eusebius, for instance, wrote a world chronicle that ended some thirty years before he wrote it. Moreover, when an author writes a compendium there is no telling how much history he intends to cover, or how far back he will end it--and a work as short as three books might very well have been so short because it was unfinished. In other words, the "compendium" could have ended in 109 BC even if it was written in 109 AD, and if the compendium's end-date was 52 AD or 92 AD, it could still have been written in 109 AD, or later. So we have to look elsewhere for a "time before which" Thallus wrote. All we have is this: the first time Thallus is ever mentioned is by Theophilus, writing around 180 AD, which leaves us with over a century of grey area: Thallus could have written any time between 28 and 180 AD. And if he did not mention an eclipse occurring in the first century, then he could have written any time between 109 BC and 180 AD, a span of almost three centuries.

This is where proper historical method turns the tables on Christian apologists. The usual argument is that Thallus is the earliest witness to the gospel tradition, proving that the story was circulating, and taken seriously enough by pagans to debunk it, before the 2nd century. But the opposite reasoning applies: since we do not know that Thallus wrote in the 1st century, but know that he could have written in the 2nd, and since no other sources attest to any gospel tradition earlier than the 2nd century, it follows that Thallus most likely wrote in the 2nd century--or at the earliest, the 90's AD, since there is some evidence that Josephus referred to Luke in that decade, although that same evidence just as easily suggests that Luke used Josephus, dating that gospel after 96 AD. Otherwise, since all other sources which mention any gospel tradition appear only in the 2nd century, and Thallus may easily have written in that period, it follows that Thallus most likely wrote in the 2nd century. This conclusion would change if any further data were rescued from the sands of time which made an earlier date more plausible, but odds are, Thallus is not the earliest witness to the gospel tradition. Even at best, there is at present no reason to assume he is.

The Invented Evidence

One of the key pieces of "evidence" used to prove a 1st-century date for Thallus was actually innocently invented in the 18th century. The item in question is a supposed reference in Josephus to a Samaritan freedman of Tiberius, which places a man by the name of Thallus from Samaria, a region in the East (from where a historian of Syrian affairs might come), in a position which would produce historians in later years (Phlegon, a freedman of Hadrian, also wrote a chronicle), in a definite first-century date (Tiberius reigned from 14 AD to 37 AD). Too good to be true? Indeed. First of all, it has long been noticed that Josephus says nothing about this "freedman" composing any literary work, and thus it is already a leap to suppose it would be the same man. Thallus, as it turns out, is a common name, appearing regularly in inscriptions throughout antiquity.

But most importantly, the name does not in fact appear in any extant text of Josephus. The passage in question (Antiquities of the Jews 18.167) does not have the word THALLOS in any extant manuscript or translation, but ALLOS. The addition of the letter theta (TH) was conjectured by a scholar named Hudson in 1720, on the argument that ALLOS didn't make sense, and that Thallus was the attested name of an imperial freedman of Tiberius in inscriptions: in his own words, "I put 'Thallos' in place of 'allos' by conjecture, as he is attested to have been among the freedmen of Tiberius, going by the inscriptions of Gruter" (p. 810, translated from Hudson's Latin). But there is no good basis for this conjecture. First, the Greek actually does make sense without the added letter (it means "another"), and all extant early translations confirm this very reading. Second, an epitome of this passage does not give a name but instead the generic "someone," which suggests that no name was mentioned in the epitomizer's copy.

Finally, the most likely name, if one were needed here at all, would be HALLOS, requiring no added letters (there is no letter for the H-sound in Greek), since an imperial freedman by this name is also known in the time of Tiberius from inscriptions. Thus, Hudson's conjecture is groundless and is to be rejected. Although we still have an inscription recording a man named Thallus as an imperial freedman, this name we know is common, and appears often in inscriptions. And the inscription in question says nothing about the man being a Samaritan, much less an author. Therefore, this attempt to place Thallus in the 1st century fails. It also fails to establish him as a Samaritan, a detail which is still cited as if it were a fact, even by good scholars.

What Thallus Said About Jesus

What exactly is Thallus supposed to have said about Jesus? We don't really know. We can only guess, based on an obscure passage passed down to us second-hand which already shows signs of at least one interpolation. George Syncellus, a 9th-century monk, composed a world chronicle, quoting verbatim from numerous previous chroniclers, one of whom being the 3rd-century Christian chronicler Julius Africanus. In one such case, Africanus is quoted regarding "what followed the savior's passion and life-giving resurrection" as follows:

This is all we get. It isn't clear what Thallus actually said, or whether he even mentioned Jesus at all. Africanus is merely criticising the possibility that the darkness at the death of Christ was a solar eclipse, and thus a natural rather than a supernatural event--an attack addressed in the Apology of Tertullian, and voiced by the Jews in the Gospel of Nicodemus, which may have been written in the time of Africanus. Although this implies that Thallus mentioned the death of Christ in some way, it does not entail it. For Thallus may have simply recorded an eclipse that occurred around the time that Christ was believed to have died, with Africanus connecting the events on his own. We do not have the context of this quote, and we do not know what else Africanus said about this event or about Thallus. Of course, even if Thallus did mention the death of Jesus, we have already shown that he then probably wrote in the 2nd century, when we know this gospel story was already circulating nearly a century after the event. In such a case, Thallus is not an independent witness to the story, but is merely responding to Christian literature. This makes him of practically no use to apologists.

The Curious Case of Phlegon of Tralles

The story gets more curious, however. For the quoted Africanus passage continues:

There is a lot of interesting material here, but I only wish to discuss what relates to Thallus. Most scholars have assumed that Africanus is here quoting Phlegon, too, as a witness to the darkness story--although we know for a fact that Phlegon wrote in the 140's AD, and was fond of fantastic stories, so it would not be surprising to find him borrowing this one from Christian literature. But Martin Routh noticed some telling details: the sentence mentioning Phlegon is grammatically and logically out of place.

In Greek, new sentences are marked by certain special words, usually left untranslated, such as MEN or DE or OUN, etc. The Phlegon sentence is not marked. That is like not leaving a period at the end of the preceding sentence. Also, Africanus has just finished attacking Thallus' idea of a solar eclipse, yet here cites Phlegon favorably, who calls it the same exact thing. Moreover, the flow of thought is broken by this sentence. Africanus has begun a rhetorical argument with the phrase "let it be so," which is otherwise interrupted by interjecting a historical note about Phlegon. Remove that sentence (and the added "Clearly this is our eclipse!") and we have a continuous stream of thought that makes more sense. The Phlegon sentence, for all of these reasons, does not belong here.

In fact, the phrase "Clearly this is our eclipse" (literally "clearly this is it") is a telltale sign of an interlinear note by some other scribe. It appears that some copyist was copying or reading this passage in either Africanus or Syncellus and remembered the Phlegon connection, writing it as a note to the side or in between the lines. A later copyist then mistook this marginal note as text to be re-inserted, since, not having erasers, scribes who forgot a line would add it in the margins or between the lines (if they noticed the error at all). This was very common in the transmission of ancient and medieval books. There was no standard rule for distinguishing between added notes and re-inserted text. Both were marked and written the same way, leading to many marginal notes being read as re-inserted text and many lines of re-inserted text being mistook for marginal notes. Without further data, we might say that Syncellus mistook the marginal note of a previous owner of his copy of Africanus, or made the note himself while a later copier of Syncellus mistook it as text, or that the note and the mistake happened entirely before or after the involvement of Syncellus. But since Syncellus immediately follows the Africanus quote with a passage from Eusebius which quotes Phlegon correctly, it is almost certainly the case that the Phlegon passage here was inserted after Syncellus. This is further supported by the extent of the insertion's inaccuracy, which looks more like something that appears in the work of Agapius in the 10th century, or in Michael the Syrian in the 12th century.

This leads us to the most important reason for supposing this line to be an insertion by someone other than Africanus (or Syncellus): Phlegon almost certainly said no such thing. Eusebius quotes Phlegon verbatim (the only one to do so), and what Phlegon actually said is telling--the text is attested in Syncellus in the original Greek, but also in the Latin of Jerome, the Syrian epitome, and the Armenian:

This quotation shows that Phlegon did not mention Jesus in this context at all (he may still have mentioned him in some other obscure context, if we believe Origen). Rather, Phlegon merely recorded a great earthquake in Bithynia, which is on the coast of the Black Sea, more than 500 miles away from Jerusalem--so there is no way this quake would have been felt near the crucifixion--and a magnificent noontime eclipse, whose location is not clear. If the eclipse was also in Bithynia, as the Phlegon quote implies but does not entail, it also could not have been seen in Jerusalem, any more than partially, since the track of a total eclipse spans only 100 miles and runs from west to east (Jerusalem is due south).

In fact, the only coincidence with the gospel story is the year (although some modern scholars calculate the eclipse in question to have actually occurred in 29 AD) and time: it began at the sixth hour. Prigent suspects this last detail is a corruption by another scribe drawing from the gospel stories, although a noon eclipse is particularly startling and might get special mention (although the total eclipse would only occur at noon in one location--are we to suppose it was in Nicaea?). What is most important, however, is that Phlegon says nothing about the eclipse occuring during a full moon or lasting three hours (both physical impossibilities), yet these details are attributed to him in the lines added to Africanus. Clearly the quote has been altered over time.

Africanus wrote in the early 3rd century. His contemporary, Origen, also cites Phlegon's mention of an earthquake and eclipse but does not repeat the exaggerations. Indeed, he expressly denies one of them in his commentary on Matthew, stating that "Phlegon, who mentioned an eclipse during the reign of Tiberius Ceasar, did not say that it happened during the full moon." This suggests that the exaggerated quote, which would surely have been seized upon as a valuable testimony, did not yet exist, in Africanus or anywhere else. But it appears in Agapius in the 10th century. And by the time of Michael the Syrian, in the 12th century, the Phlegon quote had already gone way over the top, to include the astonishing sentence: "the dead were resurrected, entered Jerusalem and said 'Woe to the Jews!'" Syncellus wrote in the 9th century. So a copier of his work who had also read Agapius probably put two and two together and gullibly added the note, which was eventually pulled into the text as copies continued to be made by other scribes.

This is significant for the Thallus passage because it shows that another chronologer who did not mention Jesus was distorted and later believed to have mentioned him or events surrounding him. The same thing could have happened to Thallus.


Eusebius, in the passage quoted above, cites "another" Greek historian as reporting the eclipse and earthquake in Bithynia in the year of the crucifixion. Was this Thallus? If so, then Thallus did not actually mention Jesus, and Africanus was clearly drawing his own conclusions. Indeed, if Thallus had mentioned Jesus, why wouldn't Eusebius quote so precious a source? Possibly because his Histories were lost, explaining why Eusebius only had a "brief compendium" to work from, which probably did not include this event (although he did have Africanus to consult). But if Eusebius had the right text at hand, another explanation for why he did not use it is that he did use it: the sentence quoted, after all, is exceptionally concise--exactly what we would expect from a "brief compendium."

Although this would entail a corrupted date for the conclusion of that work, this is a tempting theory, especially since there are two other tantalizing details that might support the notion: first, the quoted passage of Africanus identifies the reference as the third book of the histories of Thallus, which, as many scholars have noted, nicely corresponds to the "three books" of the "brief compendium" listed by Eusebius. Second, his name might in fact have been written by Eusebius after all: the Greek now reads EN ALLOIC MEN ELLHNIKOIC UPOMNHMACIN, "in another Greek compendium," and the Latin and other versions say essentially the same thing, so if this was corrupted, the error had to have happened very early--but this is still possible. If so, it could have originally read EN THALLOY MEN ELLHNIKOIC UPOMNHMACIN, "in the Greek compendium of Thallus." The only changes required here are the loss of a single letter (theta), just as Hudson had supposed in the text of Josephus, and the mistaking of IC as Y, which is not impossible.

It seems likely that even if the text as we have it is correct, Eusebius was thinking of Thallus, since he lists him as a source in the introduction which survives in the Armenian translation. If we accept this, then we must conclude that Thallus did not mention Jesus or the gospel tradition at all, since the quote is clearly devoid of any such references. On the other hand, it is also possible that Africanus was thinking of Phlegon, and simply wrote Thallus by mistake, confusing the two chronologers. We know from Eusebius (and from Origen) that Phlegon recorded the Bithynian earthquake and eclipse in the thirteenth book of his own work. This opens the possibility that the nice correspondence of "third book" and "three books" is a red herring, and in fact an error made by Africanus, or a later scribe who was fated to vex us. Thirteen would have been written TRITHI KAI DEKATHI, while three would have been written simply TRITHI. Africanus may have simply confused himself, or a scribe may have skipped over the KAI DEKATHI, as commonly happens: looking away to write and then returning to the text, seeing the second THI and mistaking it for the first before continuing to copy.

Or, if the original text were alphanumeric, thirteen would have been IG' and three would have been G'. A mistake then would be even easier: letters were run together in ancient texts, and it would be easy to see (or think) ENIG'TWN and write ENG'TWN by mistake. If anything like this happened, then Africanus was thinking of the same reference to an eclipse that everyone else thought of--Origen, Philopon, Eusebius, Agapius, Michael, and the anonymous interpolator of Africanus or Syncellus--and, again, drawing his own conclusions about the correspondence with the death of Jesus, a conclusion that was also easily drawn by his contemporary, Origin. In support of this theory is the fact that even though Thallus is well known by Christian apologists, being cited by Eusebius, Theophilus, Lactantius, Tertullian, Minucius Felix, Pseudo Justin, and Malalas, no one ever mentions his reference to Jesus or to any events of any kind after 109 BC. This is a very strange state of affairs--certainly such a juicy reference would have been quoted repeatedly and gleefully, not ignored.

This leaves us with four options: Africanus meant Phlegon, not Thallus; or Eusebius quoted Thallus verbatim, revealing that Thallus did not mention Jesus; or Thallus mentioned Jesus, but wrote in the 2nd century, when we know the gospels were already in circulation; or Thallus mentioned Jesus and wrote in the 1st century, and is the earliest witness to the gospel tradition. Although all of these are possible, it is clear that any of the first three are more likely than the last one, since there are several facts which support each of them, but none which support the last one--in other words, it is a "mere" possibility, whereas the others actually have some arguments in their favor.


To this list must be added all additional works cited by Jacoby or Müller.

André le Boeuffle, Le Ciel des Romains Boccard (Paris) 1989

Richard Carrier, Cultural History of the Lunar and Solar Eclipse in the Early Roman Empire 1998

M. Eisler, "Un Nouveau Témoignage Non-Chrétien sur la Tradition Évangelique," Revue de l'Histoire des Religions, vol. 98 (1928)

M. Goguel, The Life of Jesus 1933

Guignebert, Jesus 1935

Murray Harris, "References to Jesus in Early Classical Authors," Gospel Perspectives: the Jesus Tradition Outside the Gospels vol. 5, 1985, pp. 343-68

F. Jacoby, Fragmente der griechischen Historiker 1923

Ida Miévis, "A Propos de la Correction 'Thallos' dans les 'Antiquités Judaïques' de Flavius Josèphe," Revue Belge de Philologie et d'Histoire, vol. 13 (1934), pp. 733-740.

Alden Mosshammer, The Chronicle of Eusebius and Greek Chronographic Tradition, 1979

Carolus Müller, Fragmenta Historicorum Graecorum, 1840

Robert Newton, Ancient Astronomical Observations and the Accelerations of the Earth and Moon Johns Hopkins (Baltimore) 1970

B. G. Niebuhr and W. Dindorf, "Georgius Syncellus et Nicephorus," Corpus Scriptorum Historiae Byzantinae 1828

Pauly's Real-Encyclopädie der Classischen Altertumwissenschaft s.v. "Thallos"

Claire Préaux, "La Lune dans la Pensée Grecque," Memoires de L'Académie des Sciences de Belgique. 2. Ser. Classe des Lettres 61.4 (1973)

P. Prigent, "Thallos, Phlégon et le Testimonium Flavianum Témoins de Jésus?" Paganisme, Judaïsme, Christianisme: Influences et Affrontements dans le Monde Antique, ed. Frederick Bruce, 1978.

Horace Rigg, "Thallus: The Samaritan?" Harvard Theological Review, vol. 34 (1941), pp. 111-9.

Martin Routh, Reliquiae Sacrae, 2nd ed. (vol. 2) 1846

W. N. Stearns, Fragments from Graeco-Jewish Writers (U. of Chicago) 1908

Home Search Feedback Forum Library Shop Support What's New

Internet InfidelsSupport Us!Join SETI@homeFreethought RingChurch-State Ring


Copyright © Internet Infidels 1995-2003. All rights reserved.
« disclaimer »

Last updated: Saturday, 31-Aug-2002 16:04:31 MDT