So Rachael suggested that we build a robot that provided sentences for linguistic judgment.

This is a great idea, and I’d been discussing stateless twitter bots with Laurie, following an old blog post from my former boss Erin.

So I volunteered us both.

Thus was Cromulatron born. It’s still a baby:

but I suspect it will be developing quickly once I can get the basics working. Several FAQs are TBD below.

Is this good science?

[Aw, heck no. But it’s basically performance art about grammaticality judgments.]

How do I submit new sentences to be posted by @cromulatron?

[Right now it should only accept sentences by DM from someone on its followers list. To get on that list, write or message the bot itself.]

What if my judgments disagree with the crowd?

[Gosh, that sounds like a good question. I don’t know.]

How does this work? How can I get involved?

Well, I’m glad you asked. I [Jeremy] work on the bot github repo here (without credentials, of course) so you can see what’s working.  [nothing, yet, but at least there’s a nice high-quality cookiecutter-pylibrary repo there.]

Here’s the Travis.CI page for the code.

Here’s the ReadTheDocs page for the bot.