Today, we're pretty sure we know what truth is: it is something that can be proved. Failing proof, we'll take logic. Of course, if we don't have the raw information, we can't test or reason through it and we're left with gut reaction.
Back when all our authors were writing, people understood truth quite differently. It was just as systematic, but the rules were different. For them, truth was something claimed by a high-status authority figure. High-status came from social rank, and from being the first to make the claim. It also came from fame and repetition. So the way you tested for truth back then was to ask questions:
Challenging a claim was very bad form. Anyone who did so would need all the backing of high social rank and being the first in their field. And that in turn would create problems for the people who came after, who would now have to accommodate two truths, since choosing between them was not an option. So let's see where the challenger Berossus fits, so that we can decode the various claims into modern terms.
Herodotus was a Greek nobleman and long acknowledged as the world's first historian. At any rate, he was the first to write his history in prose rather than poetry, to do so with an intellectual rather than an inspirational purpose, and to base it on interviews with real people and personal observation rather than on tradition. He sets the field, c.450 BC. No surviving claims about the gardens or their builder.
Ctesias was a Carian nobleman, and a doctor at the Persian court. He is the first in his field regarding this particular truth-hierarchy, and his sources are utterly high-ranking and quite different from Herodotus'. He may also claim personal observation, like Herodotus. But he does not challenge Herodotus. Rather, he expands the information, c.400 BC. Semiramis apparently made the gardens (see Ctesias and Diodorus for more detail). So far so good.
Clitarchus was a Greek nobleman and historian who may or may not have been in Alexander the Great's retinue (note the status-raising purpose of this vague possibility raised by Pliny, several hundred years after the fact). Clitarchus' eye witnesses were hoplites—that is minor aristocrats performing national service as infantrymen. The resultant inexpert but very visual description adds charm to the tradition c.310-301 BC, but almost certainly did not suggest another builder (see Clitarchus for more detail). Semiramis still apparently built the gardens. Great.
Megasthenes was a Greek ambassador, geographer and traveller, about whom we know very little. His work is only known through later writers and it is usually overlooked with regard to the Hanging Gardens. Megasthenes mentioned that Nebuchadnezzar built some hanging gardens (possibly several such) in the city of Teredon, on the Gulf Coast (200 miles from Babylon). This does not challenge any truth tradition, it merely documents a perfectly harmonious additional tradition. At this stage, c.310 BC, we are still fine: Semiramis apparently built the gardens in Babylon, and Nebuchadnezzar certainly built the ones in Teredon.
Berossus was a Babylonian nobleman, priest and historian. He used Babylonian records from the temple and court. He challenges all the Greeks, charging them with inaccuracy. This is not merely an insult, it is a full assault and he knows it will cause problems for ever afterwards. He needs his full authority as the first non-Greek nobleman and first priest in the truth-hierarchy, and the full support of the Babylonian court and temple. Now we have problems, c.250 BC. Nebuchadnezzar made the Gardens in Babylon and the Greeks are liars.
Diodorus and Strabo (both Greek) apparently follow the original tradition. Diodorus seems to have followed Ctesias, although someone later added a marginal note attributing the gardens to an unnamed Syrian king who made them for his homesick concubine (see Diodorus for more detail). Most translations do not incorporate this into his original text. Strabo seems to have avoided committing himself about the creator of the gardens. Being noncommittal is another valid way of dealing with conflicts in this truth-hierarchy system. Probably no new issues c.22 AD. Anxious silence prevails.
Quintus Curtius was a Roman commoner, military commander and self-made administrator, whose history of Alexander is a disguised analysis of Caligula. He based his description on Clitarchus and adds an unexpected comment that the gardens are "celebrated" among the Greeks (see Quintus Curtius for more detail). In this way, he reminds us about the strength of the Semiramis attribution without overtly taking sides. In some editions he (or some later copyist) also incorporated the marginal comment about the homesick concubine from the Diodorus tradition, which confuses the issue for modern audiences who miss the hidden message about the Greeks. A careful middle course steered, 31 AD and 41 AD.
Josephus transmits Berossus for us. He was a Jewish aristocrat of priestly descent and a historian, landowner and Roman Citizen in the circle of the Roman Emperors. Josephus cleverly reconciles Berossus with the earlier record. He does this by reporting Berossus' claims at the end of chapter 19, and then commenting somewhat ambiguously in chapter 20 that although Berossus must have the most credibility, there is no conflict with the Greeks. By leaving the reader to infer that Nebuchadnezzar's rebuilding of the gardens is perfectly compatible with Semiramis' creation of the original ones, Josephus avoids overtly contradicting Berossus. But he goes further by citing some sources that ostensibly "confirm" the attribution—which don't. In fact, they fail to mention the gardens or any other architecture (if you read them, you'll find they merely confirm Nebuchadnezzar's greatness in war). Problem tenuously solved c.80 AD (about the same time that Pliny—himself a Roman—sought to raise Clitarchus' authority).
Eusebius also transmits Berossus for us, and his treatment is different again. He approaches our modern historical techniques by citing his sources carefully and comparing them against each other. Thus he quotes Josephus, and also Josephus' source, Alexander Polyhistor, so we know he approves so far. But (and this is so slick you could miss it), he suddenly abandons both these sources without further comment, and turns instead to one we haven't met before: Abydenus. Abydenus cites Megasthenes—who says that Nebuchadnezzar's gardens were in Teredon. Clever. No one is called a liar, but Josephus' hint is taken and developed. Now we're back in the clear: Nebuchadnezzar built the gardens in Teredon, which leaves the ones in Babylon for Semiramis. c.300 AD and problem solved!
Today, however, we assume that our sceptical scientific view of truth is the only possible one, and in making that assumption we have unwittingly re-opened the conflict. That conflict cannot be resolved in a scientific way because we only have a series of claims and counter-claims and not a shred of proof in any direction. But now that we understand the rules of truth-hierarchies, we can see that the original writers "dealt" with Berossus but did not agree with him. We still have no proof, but we now have a strong indication that Berossus had a political agenda and that his claim was, shall we say, disputable.
. . . We are free to consider the eye-witnesses, Ctesias and Clitarchus, on their own merits because: