Where did Israel cross the sea?

What the text gives us

Exodus 12:37 says Israel set out from Rameses for Succoth. Exodus 13:20 has them encamp at Etham 'in the edge of the wilderness.' Then Exodus 14:2 turns them back and locates them at Pi-Hahiroth, 'between Migdol and the sea, over against Baal-Zephon.' Three named places, and the camp sits between them. Pharaoh catches up there. The east wind blows all that night. The waters divide. Israel crosses on dry ground. The waters return at morning. Pharaoh's army is drowned.

Three of the names (Pi-Hahiroth, Migdol, Baal-Zephon) are Egyptian. Migdol is a frontier fortress; Egyptian sources name multiple Migdol forts along the eastern Delta. Pi-Hahiroth has been parsed both as 'mouth of the canals' and as 'house of Hathor.' Baal-Zephon is a Canaanite storm-god whose cult is attested in Egyptian sources at a port called Tahpanhes (Daphnae) and at sites along the Mediterranean coast. None of the three is uniquely located by current Egyptology. That is the central problem for any crossing-site argument.

Two other markers cut differently. Exodus 13:17-18 says God did not lead Israel by 'the way of the land of the Philistines,' which would have been the coastal road (the Via Maris). That rules out a Mediterranean crossing on its face. And 1 Kings 9:26 uses yam suph for the Gulf of Aqaba. The 1 Kings reference is the only unambiguous biblical location for yam suph, and it sits far southeast of Egypt.

The five candidate sites side by side

The four readings above sort the chapter into families. Inside the first three families, five specific crossing sites have been argued. The table below sets them next to each other across the same four columns. The columns are the variables most often used to compare them: where the site sits, how it handles the yam suph translation, what archaeological or textual support it can cite, and the most-cited problems with the proposal.

How the suph debate developed

The translation question is older than modern archaeology. The Septuagint's eruthra thalassa is third-century BCE. Jerome's Vulgate followed with mare Rubrum. The Targums of the rabbinic tradition rendered yam suph in different ways depending on the passage. Medieval Jewish commentators (Rashi, Ibn Ezra, Ramban) treated yam suph as 'reed sea' but did not insist on a specific geographic site; the Talmud's interest was theological more than topographic. The geographic question began to harden in the nineteenth century when Egyptological survey work could attach Egyptian fort names to specific tells.

What each reading still has to account for

The northern-lakes reading has to handle 1 Kings 9:26. The verse uses yam suph for the Gulf of Aqaba in a context (Solomon's fleet at Ezion-geber) where no reed-lake meaning is possible. The northern-lakes defenders answer that yam suph is a flexible geographic phrase whose primary literal meaning is 'reed sea' but whose extended usage covers any body of water Egyptian and Israelite tradition associated with the southeastern frontier. Critics argue this dilutes the lexical case for the literal reed-lake reading at Exodus 14.

The Gulf of Suez reading has to handle the marching distance. The text shows Israel arriving at Etham 'at the edge of the wilderness' (Exod 13:20) and then turning back to camp at Pi-Hahiroth before Pharaoh catches up. The total marching distance from Rameses (Pi-Ramesse) to the head of the Gulf of Suez is roughly 100 miles, which is a long march for a column with elderly and children given the chapter's timeline. Defenders argue Israel was moving fast under threat and that the timeline is compressed.

The Gulf of Aqaba reading has to handle Pi-Hahiroth, Migdol, and Baal-Zephon. None of the three has a plausible Aqaba location, and the Egyptian frontier-fort names sit in the eastern Delta. The Aqaba reading has to either relocate all three names eastward (the Cornuke-Moller approach) or treat the names as later editorial additions disconnected from the original crossing tradition. Both moves have problems.

The literary-construct reading has to handle the precision of the geographic vocabulary. Pi-Hahiroth, Migdol, and Baal-Zephon are detailed eastern-Delta frontier names whose function in Egyptian sources is becoming clearer with continuing excavation. A purely literary construct would not be expected to land on the right Egyptian frontier vocabulary by accident. Defenders argue that exilic and post-exilic writers had access to Egyptian geographic knowledge through diaspora networks and through earlier source materials.

Reading the chapter with the question open

The five candidate sites and the four readings are not the only ways to read Exodus 14. They are the ones that have generated sustained argument. Most readers will not resolve a question that has been live since the Septuagint chose 'Red Sea' for yam suph. The more accessible move is to read the chapter knowing what hangs on each candidate. If the crossing is at a reed lake, the wind setdown is the natural mechanism the text itself flags ('a strong east wind all that night,' Exod 14:21). If it is at the Gulf of Suez, the walls of water (Exod 14:22) become the load-bearing image and the chapter is asking the reader to accept a deepwater miracle. If it is at the Gulf of Aqaba, a much longer route and a different Sinai are in view. If it is no specific site, the chapter is doing theological work without claiming a recoverable event.

The chapter itself names both natural and divine agents. The east wind blows. The waters divide. Israel walks across. The Egyptians pursue. The waters return. The narrative does not say which agent was sufficient to produce which effect. That ambiguity is what gives the four readings their room to operate. Each one foregrounds different verses and reads the others as background.