[Verginius Rufus]

Latin has a more flexible word order because nouns are declined, so usually the ending will tell you which word is the subject and which word is the object.

A Roman looking at the sentence "monkey eats banana" may wonder whether the monkey is eating or being eaten (not having any idea what a "banana" is). But if he was told a banana is a fruit, he could say Simius comedit bananam or Bananam comedit simius and it would be clear who was the eater and what was the eatee.

[FredZarguna]

I think it is generally agreed that the Romans did not eat, or even know about bananas. Which makes it kind of strange that they had a word for it in your sentence...

[donna]

Simius comedit bananam or Bananam comedit simius

What if . . . that Roman didn't know what a monkey or a banana was!

[nonsporting]

A Roman looking at the sentence "monkey eats banana" may wonder whether the monkey is eating or being eaten (not having any idea what a "banana" is). But if he was told a banana is a fruit, he could say Simius comedit bananam or Bananam comedit simius and it would be clear who was the eater and what was the eatee.

It would be "simius bananam comedit", since Latin is and SOV language.

I reject the premise of the article. SVO langauges do not as a rule allow for more complex sentence structures. On the contrary, some of the most complex sentence structures I encountered were by Cicero in his Catilinarian Orations. Latin is a highly inflected SVO language.

The same is true for Attic Greek. Some of the sentences in Plato's Apology and Symposium, as I recall, were so complex, that the author lost the grammatical thread and produced sentences that simply did not parse.

Inflected langauges are different, not more difficult for native speakers.