Different languages divide up the colour spectrum in different ways. As a result, there often isn't a one-to-one relationship between the colour terms when you translate. In ancient Greek, the word melas covered the range from black to a deep red: it's also the standard colour-term for blood, and occasionally for red wine (though erythros would be the more common label for wine).

In a similar way, the word kyaneos (conventionally translated "blue") covered the range from black to a dark blue: this was the colour term used for what in English would be called "black" hair. Other colours that don't line up with English categories are porphyrios (conventionally translated "purple"), chloros ("green"), and glaukos ("grey"): in reality porphyrios referred to a range of deep vivid colours, and chloros and glaukos divide up pale colours between them. Erythros ("red") and leukos ("white") are just about the only ancient Greek colour terms that mostly line up with their conventional English translations.

Some popular news stories have recently tried to cast this kind of difference between languages as indicating that human senses have somehow changed over the millennia, but given that (a) that's an amazingly extravagant hypothesis, (b) there's no physical evidence to support it, and (c) you can see this even in comparing two modern languages (choose them from different language families for best results), you can safely dismiss it.

Another alternative (which I see another post has suggested) is to opt for the strong Sapir-Whorf hypothesis, which posits that linguistic categories determine cognitive categories, and therefore ancient people supposedly wouldn't be able to perceive colour distinctions if their language didn't have separate words for them. Aside from the fact that it relies on muddling up sensory capability, perception, and cognition as if they were all the same thing, experimental evidence has consistently shown it to be false, minus a small delay in cognitive processing.

That's a very interesting question. I'm not sure specifically about "the black earth," though I have heard a lot about the use of color by Greek poets.

I found the most articulate answer in this segment from Radiolab.

Basically, the Greeks, and other ancient people, simply didn't have the color perception that we do. We don't think it had anything to do with the number of cones in their eyes, though. Rather, they didn't have names for some colors, and lacking labels they the lacked handles necessary for perception. A survey of texts across cultures showed that, for example, red is always the first color to appear in writing, and blue is always the last to be explicitly labeled.

There are several theories as to why. One hypothesis tells us that we only create words for colors we can produce ourselves — creating blue dye seems to come late to most cultures, for example, while humans have never had trouble making red messes wherever they go. But what about the blues in nature — such as water, or the sky? If you look at the works attributed to Homer, the sea is "wine-dark" and the sky is silver or grey, not blue.

There was a book written on the subject, Through the Language Glass: Why the World Looks Different in Other Languages, by Guy Deutscher. The larger Radiolab episode from which the Homer piece is excerpted is also terrific.

I hope this leads you toward a reasonable and more specific answer to your question.

Edit: I neglected to bring it back around and say that lacking more subtle color labels, such as brown, it would make sense that they'd see the earth as simply black. But that's only my hypothesis.