Here.

They seem to have found evidence that some form of written language was being attempted by our Stone Age ancestors, an idea that – if substantiated – would push back the recognised birth of writing from about 6,000 years ago, as produced by the first agrarian societies, to an incredible 30,000 years ago.

The operative words being “seem,” “if substantiated,” and “incredible.” This seems like the usual excitable journalistic reporting of more cautious scientific reports.

Von Petzinger and Nowell remain cautious, however. “We cannot use the ‘L’ or ‘W’ words yet,” says Nowell. “This is not writing as we know it or language as we understand it. However, in these caves we are looking at the patterning of symbols, and if we can unravel that, we can get to their meaning.”

And yet, not cautious enough.  No, we can’t get to their meaning.  Not ever.  Unless we invent time travel, all we can do is hypothesize. And if we’re smart, we won’t do much of that.

The headline says it all:

Did Stone Age cavemen talk to each other in symbols?

Of course not. They may have written to each other in symbols, but they didn’t talk using those symbols.

More seriously, there is a big step between written symbols and written “language.” Language is a lot more than just a couple dozen signs. Humans 15,000 or 30,000 years ago certainly had language, that’s not really disputed. The question is whether they had written language. It may be that these symbols are one of the first steps in creating some kind of written language—not written language itself, just the first steps toward it.

That is a mistake, according to von Petzinger. For the symbols provide clear evidence of the way our ancestors moved from representing ideas realistically—as in those beautiful images of bisons and mammoths—to the stage where they began to represent concepts symbolically.

That statement seems pretty reasonable to me, although “clear evidence” may be overstating the case a bit. And I agree with Languagehat, if this is a language, we’ll never crack it. Linear A is definitely a written language and has a helluva lot more evidence behind it than these scattered symbols, and we’ve gotten bupkis trying to figure that one out.