@Adam Definite articles in IE languages all kind of developed in that way. None go back very far. It's strange that it arose in Greek, the Romance languages, and the German languages independently of each other (I believe).

Maybe that suggests a naturally evolution in human language in general, or at least those of the PIE family?

@Adam Not sure.

I'm sure someone looked into it before.

The spread of definite articles in European languages is a Sprachbund effect, not truly independent phenomena, and since noun phrases with demonstratives are already definite using demonstratives just to mark definiteness isn't that much of a stretch.

Greek probably got its definite article from Phoenician, and the Phoenician definite article also continues a demonstrative.

(I mean the idea of a definite article, not that ὁ is a loanword.)

@Adam As a native Finnish speaker I find all articles superfluous. I can cope with them and I'm very used to them behaving differently in all my secondary languages.

@Cairnarvon So they all influenced each other over time, then?

@JoonasIlmavirta How does Finnish make the distinction between a definite or indefinite object? Demonstratives?

@Adam We mostly don't make the distinction at all. It's pretty rare that it's genuinely useful, so in those cases we can well use a pronoun or other emphasis. In some situations a similar distinction is communicated by case inflection, as Finnish makes a difference between full and partial objects.

Consider "he shot at me" and "he shot me". In Finnish the difference is in the case of "me".

@JoonasIlmavirta Makes sense. Although with your example the distinction is pretty important if you'd prefer being shot at to actually being shot. :P

What I tried to say is that we do indeed make that distinction.

With some verbs the different objects lead to different verbs in English. The most hilarious tripwire is that that's the only difference between "he married her" and "he fucked her".

@JoonasIlmavirta Hahaha!

@Adam I wonder if Latin has an example like that, making it easy to make the point that a learner needs to learn the cases carefully.

That might actually make a nice question. I'm not sure if the Finnish example should be given for comparison.

@Adam The same thing happened in the Semitic languages earlier. Originally none of them had definite articles, and now (I think) all of them do, and they come from a bunch of different sources. The idea of a definite article spread through language contact and they all came up with a way to express it in their own languages.

And then it spread from Phoenician to Greek, and then from Greek to Latin and the rest of the European IE languages.

Or most of them. The Slavic languages don't have articles at all either, I think.

@Cairnarvon I feel like there's a missing link here. The use of definite articles in Latin is quite late, and doesn't seem to be influenced by Greek, since there arose both definite and indefinite articles. The Phoenician-Greek link is pretty much assured, but the Greek-Latin link less so. Do you know of a good rundown of the evidence?

Also, interesting note from Wikipedia: "Multiple demonstratives can give rise to multiple definite articles. Macedonian, for example, in which the articles are suffixed, has столот (stolot), the chair; столов (stolov), this chair; and столон (stolon), that chair. These derive from the Proto-Slavic demonstratives *tъ "this, that", *ovъ "this here" and *onъ "that over there, yonder" respectively." I don't know much about Macedonian in particular, but I wonder if this too is also from diffusion.

@cmw Not off the top of my head, but I'll be in the library on Monday so I'll have a look around. It may be one of those things that people just repeat as unexamined assumptions.

@Cairnarvon Thanks, I'd be interested in seeing what you find.