Saturday, October 15, 2011

Proto-Human and Yodic syntax

I first read earlier this week about new work on the word order of the earliest human speech (first in a footnote to a post on the Log on an entirely different subject, see here). The work is by Murray Gell-Mann and Merritt Ruhlen, two proponents of the notion that many characteristics of the earliest human patterns of speech are well within our reach, and it was published in — wait for it — the Proceedings of the National Academy of Sciences of the United States. The abstract opens with this:
Recent work in comparative linguistics suggests that all, or almost all, attested human languages may derive from a single earlier language.
The blood pressure of almost every trained, working historical linguist in North America skyrockets on reading that. Just ask one. But here's the big take-home from the story:
The word order in the ancestral language was SOV.  Except for cases of diffusion, the direction of syntactic change, when it occurs, has been for the most part SOV > SVO and, beyond that, SVO > VSO/VOS with a subsequent reversion to SVO occurring occasionally. Reversion to SOV occurs only through diffusion.
I had been pondering whether it was worthwhile to write about the story. The evidence for Proto-Human, as it's often called, is insanely controversial, to give it the most positive spin, and whether syntax can be reconstructed in a significant way is also still controversial. (There's reasonable evidence that some elements of it can be, but that gets us as far off the ground as a good high-jumper, while Proto-Human word order is a trip to Mars.) I had decided that it wasn't worth it. But then …

Alert Linguistics major Carla Oppenheimer passed on word of a couple of news stories on this to a member of Team Verb:
  • Life's Little Mysteries (again, under the rubric 'Weird'): "The Original Human Language Like Yoda Sounded"
  • HuffPost Weird News: "Yoda Language Study: New Research Shows Human Ancestors Spoke Like Star Wars Character"
YES!!!! Finally, a level of seriousness that Proto-Human so richly deserves!

So, I'm thinking, is this how Yoda speaks? I tried entering some simple English SVO sentences into the Yoda-Speak Generator (yup, it's real: here). A bunch came out SVO with 'yes, hmmm' added at the end. But in many cases the object was topicalized: 'Yoda has odd syntax' came out as 'Odd syntax, Yoda has. Yes, hmmm.' That's actually OSV, one of the rarest word order patterns known.

But that's not the key pattern if you put in more complex sentences. Try this:
I will tell you about the earliest human language.
and you get this:
Tell you about the earliest human language, I will.  Yeesssssss.
So, the Yoda generalization here is that he topicalizes like mad, a generalization among many explored years ago on Language Log, like here, in the aptly titled "Unclear of Yoda's syntax the principles are, if any". Certainly, whatever generalizations are possible, Yodic is not SOV.

At the same time, comparing Proto-Human word order to the speech of a Star Wars character seems just right.

I(mage from the Life's Little Mysteries story.)


Anonymous said...

The funnest take on Yodic is Weird Al's ... listen until very late in the song: No comment on the word order, though.

Adam Ussishkin said...

I think we also need to take into account the fact that Yoda lived a long time ago in a galaxy far, far away.

Mr. Verb said...

Ha. So long ago and far away that it was before the emergence of stable SOV syntax?

Anonymous said...

The PNAS paper is stunning. Sure, let's posit proto-word-order patterns for genetic grouping that aren't accepted by much of anybody (Amerind).

It's a galaxy far, far away all right.

Anonymous said...

Wow, talk about COMPARATIVE syntax ... Proto-Human PLUS Yodic!

Alon said...

My favourite take on this is John McWhorter's, who simply states "Merritt Ruhlen is insane".

Gell-Mann, FWIW, lacks any kind of linguistic training. His claims on the topic are roughly as solid as mine on physics would be.

Mr. Verb said...

Gell-Mann once said something to the effect that he didn't need training in linguistics. Most historical linguists are tired enough of this stuff that they aren't wasting energy on it. That's a real shame, since it is the most attention that subfield ever gets.

nemaveze said...

FWIW, Yodic word order is pretty easy to describe using Hallidayan functional grammar (Halliday & Matthiessen 2004): it basically reverses the Mood and Residue elements of the clause. (Quick simplification: Mood is what gets reproduced in a tag question.) So:

I can | help you.

Help you | I can.

Mr. Verb said...

Oh, I had never thought in those terms. Does it work for the whole set of patterns? Thanks.

schoolforlinguists said...

Close, but no. Take this example:

Actual Yoda dialogue

You require no more training.
Require no more training, you do.
No more training do you require.

You already know that which you need.
Already know that which you need, you do. (or) Know that which you need, you do already.
Already know you, that which you need.

You must confront Vader.
Confront Vader you must.
You must confront Vader.

You will be a Jedi.
Be a Jedi, you will.
A Jedi will you be.

You will confront him.
Confront him you will.
Confront him you will.

This trick (which I got from Martin & White 2005) seems to work best in clauses with an auxiliary verb (or, in Hallidayan terms, where Finite and Predicator are not conflated). I'd also argue that "Confront Vader, you must" sounds more Yodic than what Yoda actually said.

Mr. Verb said...

Nice, thanks!