How often do good ideas already exist?

[thegreenlaser]

For context, I'm working as an undergraduate researcher for the summer. Last week I felt like I was on a roll. Everything was going my way and I ended up deriving an interesting method which was different from anything anyone in my research group had heard of. With my method hashed out, it was a little easier to search for something similar in the literature, and lo and behold there's a somewhat obscure paper from like 20 years ago with my idea in it. This is the first time I've had a research idea actually work out, and also the first time someone beat me to the punch.

I'm not really upset about this (it was just a few days' worth of summer work, not my entire PhD thesis), but it made me curious: how often does this kind of thing happen in the research world? Does it happen less as you spend more time getting familiar with the landscape of literature in your field?

 

[TeethWhitener]

how often does this kind of thing happen in the research world?

Probably once a week to once a month for me, but I'm working in a very competitive field (graphene chemistry and physics). Chances are, if you've come up with a new idea in a few days or a few weeks, someone else has thought about it, and if they haven't published on it, they're probably actively working on it. This is especially true if you're not an established expert in the field. A Ph.D. thesis is different, namely because you'll spend years of research getting to the bleeding edge of whatever topic you're working on. From this vantage point, it's much easier to see the way forward, and you are much more familiar with what has and hasn't been tried.

 

[atyy]

It happens a lot. Sometimes what one does is truly revolutionary, like Max Planck's proposal for black body radiation, or Lynn Margulis's proposal of endosymbiosis. Other times, even very important work will have clearer precedent. For example, the Higgs mechanism was first envisaged by Anderson, then shown to work in the relativistic context by Brout and Englert, Higgs, and Guralnik, Hagen and Kibble, which was an important piece that went into Weinberg's work.

 

[Choppy]

It happens fairly regularly for me as well. That's why starting with a literature search is so important. I try to take is as a sign that I have good ideas.

 

[JorisL]

A quick question related to searching in literature, how do you guys search in literature?
Often I'm searching for something to clarify textbook claims and find research that's close to what I want but not exactly.
I end up having to track back through references until I sometimes get what I actually needed.
For example recently I needed the continuity of the von Neumann entropy which gave me recent results (which is good) but weren't applicable with the objects I had at my disposal. In the first article I opened there was a reference to the inequality I needed/could use.

Regular Google displays (famous) articles multiple times which makes the overview too restricted.
Also some of the obvious results seem to be omitted.

 

[thegreenlaser]

A quick question related to searching in literature, how do you guys search in literature?
Often I'm searching for something to clarify textbook claims and find research that's close to what I want but not exactly.
I end up having to track back through references until I sometimes get what I actually needed.
For example recently I needed the continuity of the von Neumann entropy which gave me recent results (which is good) but weren't applicable with the objects I had at my disposal. In the first article I opened there was a reference to the inequality I needed/could use.

Regular Google displays (famous) articles multiple times which makes the overview too restricted.
Also some of the obvious results seem to be omitted.

I'm obviously not an expert in literature reviews, but have you tried Google Scholar? It's way better than regular Google for this kind of thing, since that's what it's designed for. It's certainly not the only search engine you'll ever need, but it's not a bad place to start.

 

[JorisL]

I'm obviously not an expert in literature reviews, but have you tried Google Scholar? It's way better than regular Google for this kind of thing, since that's what it's designed for. It's certainly not the only search engine you'll ever need, but it's not a bad place to start.

Doh, I keep forgetting about scholar...

 

[A. Neumaier]

It's the usual thing - especially if you are working on something of interest to more than a few people. Still, it is important and it gives you a good feeling to have done something that was new given your current horizon. The time lag between your rediscoveries is an indication of how close you are to knowing the state of the art on this particular question.

Often one needs to discover something for oneself before one knows how to search for it in the literature; though it is now much easier through scholar.google.com than it used to be still 15 years ago. And of course, one then usually finds it (or one tries to publish it and a referee points it out). You need to know a field (or an aspect of it) very well before you are likely to do something really new. This is why research usually begins with doing something the supervisor suggests - s/he is supposed to know the literature much better, thus decreasing the likelihood that something done is not new.

 

[TeethWhitener]

A quick question related to searching in literature, how do you guys search in literature?

Web of Knowledge and Scifinder are pretty good too.

 

[ZapperZ]

For context, I'm working as an undergraduate researcher for the summer. Last week I felt like I was on a roll. Everything was going my way and I ended up deriving an interesting method which was different from anything anyone in my research group had heard of. With my method hashed out, it was a little easier to search for something similar in the literature, and lo and behold there's a somewhat obscure paper from like 20 years ago with my idea in it. This is the first time I've had a research idea actually work out, and also the first time someone beat me to the punch.

I'm not really upset about this (it was just a few days' worth of summer work, not my entire PhD thesis), but it made me curious: how often does this kind of thing happen in the research world? Does it happen less as you spend more time getting familiar with the landscape of literature in your field?

First of all, I want to congratulate you and say "well done" for being diligent in your literature search. This is what many of us have to do each time we think we discover something new, or thinking that we might be on to something. If only several of our delusional members on this forum have the same perspective as you do, we won't have to lock or delete so many threads on various claims of the next best thing since sliced bread!

Secondly, when doing any kind of research, especially when you're staring out, this is where having a really good mentor or supervisor can help. He/she should know the state of knowledge of that particular field, and would have an inkling if something looks familiar. So as someone just starting out in any field, try to rely a bit more on senior researchers or your supervisor.

Still, no one is infallible, and things may slip through, and the onus on doing the dirty work of searching if something has been done before falls onto you. While many things may already have been done before, you also need to look at whether there are details that might be new and might be important enough, especially for specific applications. Again, this is where your supervisor or mentor might be of help.

If not, then chalk it up to being able to derive something without "seeing the answer". It should give you some confidence in your ability to correctly do something.

Zz.