We all know the phenomenon: Time and again new attempts are made to finally solve the question of Atlantis. But only rarely media take notice because none of these attempts could convince, so far. Whether an Atlantis skeptic claims to have found proof for the invention of Atlantis by Plato, resp. whether an Atlantis searcher claims to have found Atlantis somewhere – it never does work out for 100 percent, and the search for the solution to the Atlantis riddle seems to go on for eternity. One reason seems to be that certain typical errors and mistakes in reasoning are made again and again by Atlantis researchers. The Atlantis problem unfolds to be a true labyrinth of intertwined paths of thought, peppered with the most subtle thought traps.
The most tempting mistakes of interpretation come into being, if Plato's Atlantis dialogues "Timaios" and "Kritias" are taken literally word by word. Thus, Atlantis would be a civilized small continent in the Atlantic, which existed there around 10000 BC . The impossibility of this assumption can be demonstrated easily: At 10000 BC a civilization such as the Atlantis described by Plato was not conceivable, to say the least, and modern knowledge about geological processes and about the consistence of the sea-bed exclude a sunken small continent categorically.
Yet, already the next mistake in reasoning is lurking: If the Atlantis account cannot be correct in these not unimportant parts, then it must be wrong as a whole, so do many claim . But on this question the tradition of ancient literature teaches us better: The descriptions of Plato stay within the range of the then usual dimensions of mistakes. E.g. Herodotus considers the Egyptian civilization to be older than 11000 years what makes the 9000 years of Atlantis suddenly look very ordinary. So far, nobody had the idea to declare Egypt to be completely inexistent just because Herodotus erred concerning its age.
Plato's geographical descriptions, too, are more sophisticated than it seems at first glance. E.g. Plato does not write literally that Atlantis was situated to the west of the Straits of Gibraltar which were called Pillars of Hercules at his time. Rather, Plato makes the Egyptian priest translating an unknown description of the location from the Egyptian Atlantis account into the Greek phrase "Pillars of Hercules" . But who knows what was the geographical horizon of those, who – according to Plato – once wrote down the Atlantis story in the Egyptian annals?
Surely: This is by far no proof for the existence of Atlantis. Yet we can learn something: The question of Atlantis cannot be solved in a twinkling of an eye. Self-restraint is necessary, the Atlantis question is an open question.
Unfortunately, neither Atlantis searchers nor Atlantis skeptics have sense for such subtleties. Instead, they flee into arguments which support their prejudices. Atlantis searchers mostly add inventions which cannot be found in Plato's original text. They often believe to find support in other sources without giving any proof for the correspondance of these other sources to Plato's Atlantis dialogues. Thus, sky-falling planetoids, flight machines, energy crystals, golden cupolas, pyramids, extraterrestrials and the like are introduced .
But Atlantis skeptics, too, read things in Plato's dialogues which are not there. E.g. they try to demonstrate that Atlantis was described by Plato to be a fabulous land of plenty. They are convinced that Plato exaggerated so immensely that the invention cannot be overlooked. Yet, the comparison to Herodotus is helpful: Plato's Atlantis is indeed a wealthy community, but it does not excel the conditions of its time. What Herodotus writes about other countries of his time surpasses the Atlantis account in many ways .
Other Atlantis skeptics want to see allusions to the Persian empire in Atlantis. They base their opinion e.g. on the army of chariots and the city plan of Atlantis. But again, they have seen things which are not there. Chariots which are really Persian chariots are missing completely in Herodotus' descriptions of the Persian wars. Only the peoples from the borders of the Persian empire, Indians and Libyans, come with certain numbers of chariots, and they are never mentioned in any of Herodotus' battle descriptions. On the other hand, the later scythed chariots mentioned by Xenophon are not mentioned in Plato's Atlantis account, although Plato does not tell of one but even of two types of chariots in Atlantis.
Instead of an army of chariots as an allusion of Plato to the Persian empire, rather a hint to the elite troops of the so-called "Immortals" could have been expected. The notion of the Persian army was then widely defined by these "Immortals". But there is no such hint.
When considering city plans, the comparison of Atlantis with the Persian empire fails entirely, too. Many favour to compare the seven rings of walls of Ecbatana with the three rings of land and water of Atlantis. But seven rings of walls and three rings of land and water are not at all the same. Especially, if the colours of the walls – other then many claim – do not correspond in any way. And the descriptions of the city centers do not fit, too. When assuming an allusion, a certain conceivable correspondance can be expected.
The same is true for Babylon: Here, it's only one ring of walls and the city center is divided in two – if we follow Herodotus – whereas in Atlantis everything is concentrated in the city center. Essential characteristics of Babylon are missing in the Atlantis account, such as the one hundred iron gates or the special look of the Babylonian terraced temple. Also in the case of Babylon a correspondance to Atlantis can safely be excluded.
In order to find correspondances for all features of Plato's Atlantis some Atlantis skeptics search for more and more corresponding cities, until in the end they want to see Atlantis in almost every ancient city: Ecbatana, Babylon, Carthage, Syracuse, Troy, Scheria, Athens, Tartessus, etc.  It is obvious that this is a dead end in the Atlantis labyrinth.
A very popular error on behalf of the Atlantis searchers is the identification of the titan Atlas known from Greek mythology with king Atlas of Atlantis. The same error happens with a people of "Atlantians" mentioned by Herodotus. What seems convincing at first glance is in truth totally wrong. There are no conceivable connections between them. Let us analyze the details.
On the one hand side, there is Atlas, king of an island empire in the West, son of Poseidon, described by Plato as a real and historical person. The historical involvement is additionally underlined by the war of conquest against Egypt and its Saitic tradition. The name "Atlas" is in doubt, because it was transferred (translated?) by Solon from the Egyptian language or even from the language of this island empire, directly into Greek. After king Atlas his island empire with the surrounding sea is named (Atlantis = "of Atlas"). Plato does not call the inhabitants of Atlas' island empire "Atlantians" or "Atlanteans". The real names of island, city and people of king Atlas are entirely unknown to us. This is Plato's king Atlas.[*]
On the other hand, there is Atlas the titan and sky-bearer, situated in the world's extreme West, son of a titan and an oceanide. He is a mythological person, and his myth does not show any connections with history as does e.g. maybe the myths of Hercules. Nowhere is there talk of an "empire" of the titan Atlas. Rather, the titan Atlas got his position as sky-bearer by punishment. He cannot be thougt of as a ruler.
After this titan Atlas later the Atlas mountains as well as the Atlantic Sea was named. The first mentionings of both can be found in Herodotus . It is easy to see with the sea: It is called sea "of Atlas" withouth being more specific about "Atlas"; this can only point to the titan, because only the titan and nothing else was known to the Greek reader. It is the same with the Atlas mountains.
As usual in Herodotus' work, the accounts about peoples become more and more mythical, the more his descriptions approach the ends of earth. This it true for the account of northern African peoples given from East to West, as many attributes clearly show (mountains always in clouds, no dreams), and it is true for the naming by location, i.e. functionally driven naming of the inhabitants of the Atlas mountains. The mountains got their name from titan Atlas, thus the inhabitants of the mountains, too. These "Atlantians" are simply the inhabitants of the Atlas mountains, like alpines are the inhabitants of the Alpes.
Herodotus knows the uncertainty of his information about far-away countries, thus he adds here and there some general warnings. The functional interpretation of the mountains as "sky's column" and the fitting naming "Atlas" are due to the mythological expectations of Greek thinking. Probably it is due to the phenomenon of the flowery chitchat of Herodotus' tour guide: Herodotus got to know a story tailored to fit for a Greek.
While Atlantis searchers waste their time with such superficial language problems, they totally miss a geographical self-contradiction in Plato's Atlantis dialogues: Sometimes, the text reflects an (at Plato's times) outdated two-partite world view, i.e. the world is divided in the two continents of Europe and Asia, where Africa is part of Asia. Sometimes, the text reflects a three-partite world view, where Africa is a continent on its own (Atlantis "was larger than Libya and Asia put together"). So it seems, that the Atlantis account goes back to an older source and thus it is not a mere invention by Plato.
Yet the adherents to the invention hypothesis classify the Atlantis account as a so-called "Platonic Myth" . For them, this is an artificial myth invented by Plato which is intended to illustrate Plato's philosophy allegorically. But the Atlantis account does not at all fit to the idea of an artificial and illustrative myth neither by form nor by content. In addition, the argumentation of the Atlantis dialogues is underpinned too directly on the truth status of the Atlantis account. So it seems to be difficult to classify it as an untrue invention .
Totally amazing is the fact that Atlantis searchers usually seem to have no precise idea of the Egyptian civilization, since the Atlantis account came from Egypt, according to Plato. Instead, Atlantis searchers often have very enthusiastic concepts of Egypt, whereas Atlantis skeptics believe a preoccupation with Egypt to be superfluous since for them, everything is considered to be Plato's invention anyway .
So, what is more obvious than to verify the Atlantis account and the way it was handed down via Solon on the background of egyptological knowledge? The times of the Saitic pharaos is well-investigated, when Solon allegedly got to know the Atlantis account . Herodotus is one of the best historcial sources for this epoch. Indeed, the account of Solon's visit to Sais does fit well to the schemes of this time. Could Plato have invented all this in such an astonishing accuracy?
The date of 9000 years, too, corresponds to Egypt in a special way, as we have already seen above. What Herodotus relates about Egyptian annals and what modern research can confirm of it provides important contributions for the clarification of the Atlantis problem.
On one of these "Egyptian questions" we want to have a closer look: As is known, Herodotus, too, was in the city of Sais. From the Saitic temple of goddess Neith the Atlantis account allegedly was taken. Like Solon, Herodotus talked there to a priest. But he did not get to know anything about Atlantis. Some draw the conclusion that nothing could be get to know about Atlantis in Sais. Yet again, this is an all-too hasty conclusion.
Herodotus did not come to Sais as a guest of state, and his talk partner was an obviously incompetent priest. In addition, Herodotus did not ask this priest for historical things, but for the Sources of the Nile. Moreover we have to consider that Herodotus did not get to know many other events of the Egytpian history, too, such as the Sea Peoples wars. Thus, the power of the argument that Herodotus did not get to know anything about Atlantis in Sais quickly fades away.
We now have shed light into several prominent corners of the labyrinth of the Atlantis problem. We could see that the Atlantis account defies stubbornly all-too straightforward interpretations. What is missing is a real solution of the Atlantis riddle. Yet we made some important progress on the way to such a solution: The more meanderings and dead ends can be discarded, the more the true way through the Atlantis labyrinth to an answer of the Atlantis question could become clear, in the end. There is reason to excitedly expect more!
 Cf. e.g. Otto H. Muck, Atlantis – gefunden – Kritik und Lösung des Atlantis-Problems, 1954; and Charles Berlitz, The Mystery of Atlantis, 1976.
 E.g. Heinz-Günther Nesselrath, Platon und die Erfindung von Atlantis, 2002.
 Timaeus 24e.
 Cf. e.g. Murry Hope, Atlantis – Myth or Reality?, 1991.
 Cf. e.g. Th. C. Franke, Mit Herodot auf den Spuren von Atlantis, 2006.
 Cf. e.g. Pierre Vidal-Naquet, L’Atlantide – Petite histoire d’un mythe platonicien, 2005.
 Herodotus I 202 and IV 184.
 Cf. e.g. Nesselrath and Vidal-Naquet (see above)
 Cf. e.g. Wilhelm Brandenstein, Atlantis – Größe und Untergang eines geheimnisvollen Inselreiches, 1951; and Massimo Pallottino, Atlantide, Archeologia Classica No. 4/1952, pp. 229-240.
 A laudable exception is John V. Luce, cf. e.g. Edwin S. Ramage (ed.), Atlantis – Fact or Ficton?, 1978.
 Cf. e.g. Jan Assmann, Weisheit und Mysterium, 2000.
[*] The original article contained the mistake that Plato did not name the Atlantic sea after (king!) Atlas, but he does, although with a different grammatical form as Herodotus. A statement about the name of the city was ambivalent, too. Corrected Feb 2014.