[I wrote the short essay below for undergraduate students but the ideas should also be useful to graduate students, professors, scientists, and other writers.]
When confronted with a writing assignment, some of us mistakenly think that clear sentences and ideas should come easily. But they seldom do. And so we stare at the paper, we stare into space, put on a CD, eat some chips, stare at the paper again, call a friend, change the CD, stare at the paper some more. Because we believe that beautiful sentences and profound ideas should flow complete from our brain to the paper, we are blocked. Our brain cannot seem to do it, and the paper or computer screen remains clean.
Others make the opposite mistake. They assume that whatever appears in their brain is good stuff. They just let it flow to the paper, one idea leads to another and then back again to the original idea and then off to an unrelated idea and then back again to the original idea, back and forth and around and they just keep writing until they have filled enough pages to satisfy the instructor. The paper writes itself. No problem.
You should of course avoid both mistakes. I am more sympathetic with those who make the first mistake, who are blocked because they think they don't have anything to say or because it seems so hard to start, or because the ideas seem complex and contradictory. Time is passing and you have lots of empty paper (or blank computer screen) in front of you.
Those who commit the second mistake, who simply heave whatever comes to mind straight to paper, have an easier time, I suppose, but they also have problems. Consider that their paper's thrust will depend on their first idea. That idea may be good, but it could be bad or scattered, considering that they spend so little time collecting and examining ideas/information. The page or computer screen will be filled with words, but the ideas may be undeveloped or disconnected.
To avoid both mistakes, you should brainstorm and diagram before writing (and while writing). Do not start a paper by trying to write the perfect first sentence or the perfect opening paragraph. Do not start a paper by writing the paper! Start the paper by collecting and examining information and thoughts. We call this brainstorming, and brainstorming is easier done with pencil and paper than with the computer.
When you brainstorm, don't write complete sentences. Don't worry about grammar or spelling. Don't worry about staying within the lines of the paper. Don't judge your ideas. You will revise later. When you brainstorm, your job is to collect all your thoughts without judging them. Just think about the topic and write down whatever comes to mind. Let your mind consider the material without worrying yet about how you will wrap it up in a paper. Because you won't think of everything in one session, you should brainstorm several times. You also may have to find some new information. Some of your ideas may be irrelevant or incorrect, but you can always discard them or modify them later.
Once you have a page of words and phrases that capture your ideas on the subject, look for connections, look for themes (main ideas), look for patterns. Draw circles around related information and connect ideas with arrows. Try numbering which circled groups of information need to come early in a paper and which should appear late. In doing this, you will have produced a cluster diagram. To produce the cluster diagram, you may be able to use your original brainstorming paper or you may have to rewrite the material on a new sheet of paper. Do whatever works. You are not concerned with neatness or correct grammar. You are assembling your thoughts, looking for how the thoughts might be organized and connected. You are staying high and free and not being tied up by whether a sentence sounds correct or whether a particular idea is good or bad. I have added a sample cluster diagram based on predator-prey (lynx-hare) cycles in boreal forests (see Identifying Sentence Components for background).
After you have brainstormed and clustered, begin outlining and writing complete sentences and paragraphs. Of course, as you write you will decide that some patterns require change or that some patterns are more important than othersówe also refer to this as 'thinking'. You will also identify information gapsóget the information.