On defence:
Looked at the save. How about switching production to archer? If a plains forest is worked for 1 turn it could be built in 3. The turn the archer completes, Delhi should grow into unhappiness. Revolt to slavery and whip settler for 2 pop the same turn. Then depending on the circumstance the overflow could go into another archer or a worker. If the warrior threatens the pasture next turn, promote the archer to combat 1, and defend it. Also send the warrior down from the forested hill to the plains forest, ready to mop up the barb warrior just in case he defeats the archer. The barb may also attack our forest warrior instead (probably better). Alternatively, if the barb heads straight for Delhi we will at least have another archer built the turn the barb can attack.
On settling:
If we are serious about getting the gold, I think we should settle it inside the inner culture ring. Monty is close, and I wouldn't want him to get any ideas about stealing it. If we settle 1W of the gold we would still be on the river (unpillagable roads as obsolete mentions). We'd miss the cow, but after 1 border pop we would cut off Monty from our northern territory. I also think there could be merit to Dirk's thought of settling 1E of gold on desert hill, sacrifices the river, but superior defensively, would also cut off the north after 1 border pop. Though as Duckweed says, an early dow from Monty would probably be futile anyway. I prefer the look of obsolete's proposed southern site on a plains hill. Also, I think a city near the copper should rank as a priority soon, it's a great location, and axes will help in the likely event of a dow from Toku/Joao, and may even catch us a barb city.
On tech:
Like Duckweed, I'd be inclined to go for wheel next, regardless of rivers as substitute roads. It leads us to granaries, connected happy resources, potential foreign trade routes, faster movement, libraries and pretty much everything else.