When I see or hear about something exciting--an unpublished result that backs up what I've long thought to be true, for example--my first instinct is to tell someone else. After all, not only does it show how acute my scientific intuition is, but, if the information comes from another scientist, it also demonstrates how successfully I'm developing my network of contacts. What do you do if you find something surprising during your PhD? You may well want to share your news to impress people or, if it's particularly hard to explain, to verify that it means what you think it means.
BEWARE! Yes, you need to learn what is and is not worth getting excited about by checking what you've seen or heard with others. However, there comes a point, quite early in your career, when you have to learn to bite your lip and break this habit. It's about developing enough self-confidence to make it as an independent scientist or, to put it another way, developing trust in your own ability to reason.
Part of becoming a successful scientist is building up your own portfolio of ideas: experiments you'd love to do if you ever get the time (and money). And if you're heading for your first postdoc position, the more irons you can get in your research fire the better. So allow yourself the luxury of keeping some selected information secret. Then, one rainy Tuesday afternoon, you can play around with an idea and see if it's worth following up. This is what successful scientists do. They get ideas (often not entirely their own), play around with them through the wonderful medium that is the pilot experiment, and then follow up the worthwhile stuff.
The weak link in the chain is keeping quiet about the information in your head, possibly for years. If you ask yourself honestly, you will probably have at least some instinct to keep useful nuggets of information to yourself. In practice, not speaking about these nuggets may prove much harder than the act of deciding not to. This is especially true if you're a blabbermouth like me. OK, so it's unlikely that anyone in the pub will want to beat you to it when you tell about the killer experiment that you think might turn your field on its head. However, you do need to curb your natural enthusiasm for sharing information when it comes to potential rivals.
I'm talking about consciously not mentioning sensitive information in e-mails--for instance, e-mails to that mate of yours who works in a lab that collaborates with your main competitors. I'm talking about not disclosing everything you know at the poster session and not gushing information during question time at the end of your presentation. I mean learning to play a little cagey. No, not sneaky or dishonest: I would never advocate lying to deliberately throw someone off the scent. But you can certainly accelerate your career by opting not to share all your information with others.
Hold on a minute, what about your PhD boss? Isn't it naughty to keep things from him or her? You are supposed to tell your supervisor everything, right? Let's say you're writing up and soon to move to another lab. That hopefully close relationship you've developed  with your boss over the past 3 years is unlikely to last much longer. Of course you'll stay in touch, but from now on you must consider what not to tell. This is especially true if you are moving to another group working in the same field. In my experience, the vast majority of group leaders are motivated primarily by one thought--prestige. And as we all know, prestige in the eyes of one's peers derives largely from publications. Your boss is hungry for prestigious publications. Your results and good ideas and your contacts' information are excellent food to help satisfy this hunger.
Once you've moved on, it's your boss who'll be presenting your old work, not you--but at least your contribution should be acknowledged. However, if you leave sensitive information and good ideas behind when you depart, expect them to be snapped up and used without a moment's thought. What you need to realise is that by talking about your ideas, you are denying yourself an unhindered opportunity to follow them up at your leisure. Some people might even misinterpret your openness as disinterest in pursuing these ideas on your part--a sort of invitation to others to experiment. I'm afraid this is just as true of PhD bosses as it is of any other type of scientist.
Sometimes you might be able to second-guess that your boss won't have the time to follow up a particular lead. In such cases it might be worth disclosing what's on your mind with the explicit intention of getting a right of first refusal. Personally, I feel much happier with this kind of arrangement. As I said, I'm a naturally open person not given to holding secrets long term.
Choosing whether to discuss what you know with others is a key step in your push for early independence in your research career. Distasteful as it may seem, unless you learn to be selective when communicating with your peers, you may find yourself left behind in the relentless hunt for success.