For example

https://www.google.com/?gws_rd=ssl#q=researchers+debunk

https://www.google.com/?gws_rd=ssl#q=researchers+refute

https://www.google.com/?gws_rd=ssl#q=researchers+disprove

https://www.google.com/?gws_rd=ssl#q=study+no+evidence

https://www.google.com/?gws_rd=ssl#q=theory+challenged

Regards,

Joachim

I don't think I understand your question totally. While tha answer of Joachim focusses on keywords indicating a theory being questionable or even refuted, I think you are more after content words. Your examples, however, only mention words which make complete sense: Epicycle is a mathemetical curve: The one a spot on a tyre wall describes in relation to the street surface. IT's just that we now longer believe this is what planets do, the word itself makes perfect sense. The same holds for "stress related".

I do not think that there are many words that exclusively were used in now abandoned theories, even "Ether" has a clearly defined meaning in Chemistry (different from ancient one).

The only example I can think of is "phlogistone"."Hollow world" might be a candidate as well.

Still, I think it is a historical coincidence whether a term

- keeps the meaning if it was used in an obsolete theory ("stress related")

- gets a new meaning afterwards ("Ether"), or is

- discarded ("phlogistone")

What, exactly, are you after?

I think that no, its not possible.

What I understand is you aim to create a "perfect search" that provides only "real scientific proof results". 

Thats impossible as only a few facts are worldwide assumed as true and most, are still being studied.

But the main important thing: a keyword for a search engine has nothing to be with their reality, true or scientific proof.

A keyword is any term a person is loking for and write down in the search engine. When a person writes ulcers, search engine reports all findings for that, but obiously is the reader the one who has to read through and find out if they are caused by H. pilory or by whatever. You can´t just make search engine tell you: "you are outdated man, ulcers, pshh". 

An "anti keyword" is a no keyword itself.

If positive keywords are desired, there might

be the possibility that someone else gathers

lists of these outdated concepts. In order to

find such webpages, do a query for some good

example keywords and hope for results.

Some years ago there was a nice feature from

Google Labs called Google Sets. One could

enter some sample members of a potential set,

and Google would return a more complete list.

http://www.evolvingseo.com/2014/05/20/google-sets-retired-from-new-google-sheets/

As for epicycles: some consider them a

forerunner of fourier synthesis.

https://www.youtube.com/watch?v=QVuU2YCwHjw

Regards,

Joachim

Something to note here, is that Google no longer looks at keywords only, whether it is in your search query, or when it scans your webpage content for indexing. Since the Hummingbird algorithm replaced the old PageRank system, the focus is now on searcher intent and meaning, rather than single keywords. Google tries to find what the meaning is behind your query (i.e. when you include the word "Apple" in your query, it tries to determine whether you are talking about the fruit or the brand name). So, both in your query generation and in your content writing for a webpage, one should keep this in mind - specify longer rather than shorter search queries, and use meaningful words rather than stop-words. For example, rather than typing: who are the american election candidates?, try something like: usa election 2016 republican democrat candidate names. Melius Weideman.

There have been some great comments by everyone so far.

To clarify a little bit,

This is more about Mark Twain's observation

It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so.

 Science is rife with incorrect assumptions that when repeated become fact. It would be great to be able to filter them out from serious consideration and identify their connections.

(Dirk et al) yes out of the center of the universe concept epicycle is not an anti keyword. Also, I beg to differ with you on stress related in a medical venue. While chronic stress, can worsen many serious health problems if you are looking for the cause of America's and now the world's pan-epidemic of obesity then stress just clouds the issue as it did for ulcers.

BTW  my students at the end of last semester called these anti-keywords "nick" or "nique" words which is short for nonsense in French with scatological connotations in slang. Which I think is appropriate. Words and phrases like debunk or "experts did not anticipate" "researchers refute" and so on do bring up some theories or ideas that are false. Thank you  Joachim. They are nickwords in a sense. Perhaps it is the pattern of erroneous theories that needs to be the subject of a deep learning project for AI.

Christophe I wholeheartedly agree with the idea that we feed or train the semantic search engine and be able to change this for a case by case basis even. Five years ago my students created a corpus (collection of documents) from google scholar on medical articles about ulcers before 1980. We had to had handpick the phrases near words like "cause" or "reason for ". Once we had a really extensive collection of phrases like genetic factors, multifactoral, scientists are still unsure why etc. hundreds of phrases that with hindsight showed doubt or were clearly wrong in the case of ulcers, we then took this multitude of nickwords and phrases and searched google scholar with them. Sure often other medical theories were among the results like multifactoral and blood pressure or multiple sclerosis but so were dark matter, dark energy, the Earth's magnetic field, cold fusion and so on. Did this mean that multifactorial was likely to be a cool way of saying = we don't really know. In general I believe so. However does that mean that current theories on quarks and dark matter are necessarily wrong... no of course not.

Melius, I am well aware of google's improvement. Before humiongbird if you had an idea about a weather link to MS and the atmospheric phenomenon the Jet Stream you typed jet stream multiple sclerosis and at best there were five or six results none of which was relevant and you knew you were on to something that no one else was working on

now google furnishes half a million useless results.

If you think that climatic temperature conditions  and MS is barmy compare these two  world maps one on old data on MS and the other on world temps

notice the line going across the US around the 38 parallel this is were the Jet stream often is located.

food for thought

CGY

http://library.med.utah.edu/kw/ms/mml/ms_worldmap.jpg

http://www.climate-charts.com/images/world-temperature-map.png