I recently started a new project with a newly registered generic tld domain.

As soon as Google started indexing the page, it displayed a "translate this page" in SERP's, which tries to translate the page to the language of a small Eastern European country from the language that the site actually uses.

I tried everything to prevent this: language meta headers and attributes, localisation through Google Webmaster Tools...all to no avail - nothing helped.

After a couple of weeks I spotted dozens of inbound links popping up in Google Webmaster Tools all coming from that small Eastern European country, from sub-pages that are not active anymore (either sending out 404's or 301's to the main page), and also had been written in that other language. So the domain had been registered before and as it looks, it did got a lot of possibly spam links in that language.

I can't even ask the sites where those links should have been to remove them as they are not active anymore physically, just in Google Webmaster Tools and/or internal data masses...

Now I'm at a loss about what to do?

As my site is pretty new, it does not have many links pointing towards it in my targeted language. So those are probably not enough to convince Google of attaching the right language to it as Google ignores all other signals about the page language.

I'm also unsure if I should use the "disavow" tool, or a reconsideration request...or what else to do about this miserable state. I never used these tools before so I don't have any experience with them.

Somehow I have to convince Google about the right language of the page and also to not count/apply/whatever all those historical links from the previous owner. (The domain had been deleted without any traces in Google before I registered it)

Has anyone here ever dealt with a similar "Translate this page" problem?

(I've also looked at this thread: How can I prevent Google mistakenly offering to translate a page? but didn't find a solution there)