Writing Seminar 10/14/08

For those of you that missed Peter Sterling's writing seminar yesterday, I've included my notes below. The take-home points are summarized at the bottom. I found his talk to be very straightforward and immensely helpful. Although none of it was earth-shattering, he highlighted a number of common problems people face while writing papers and grants. His website is: Peter Sterling's Website. On the website is a link to his powerpoint presentation which is pretty self explanatory and a link to a paper on writing that he recommends.

Another resource that he recommended is the book Style: Ten Lessons in Clarity and Grace by Williams (not the abridged version).

Notes:

Writing a paper or a grant is, in essence, telling a story.

Parts of a story:
Context (background)
Something happens
Moral

He uses the example of Sleeping Beauty to demonstrate how a story works. You know the story.

Important point:
Length of these sections- something happens is the longest section; make sure the context (introduction) and moral (discussion) are not longer than the results

In science, he recommends to begin writing the paper and creating the figures as soon as you get the first real result. Why? Because it will help you build the story and imagine the appropriate controls and experiments and identify early on what is missing.

When you start writing:
List 3-5 key observations (results)
Sketch the figures
Arrange them in compelling order
Write a strong lead for each legend to tell the story (this will help you avoid repeating it in the text)

Write the title BEFORE the story
-this will remind you of the story from the beginning
-if the story drifts, you either revise the story OR the title

Designing the title
-make the biggest claim that you can support (modified big claims if you have to ie. evidence that vesicles on the synaptic ribbon can be rapidly released; doesn't prove anything, but got published)

Title example:
BAD- Differential expression of chloride transporters in ON and OFF bipolar cells
-end of title is most important to reader, but it's the most specific and least likely to be interesting for most readers
-differential expression- what does this mean? not only does this tell us very little, but this is an example of nominalization (when you make a verb into a noun), action verbs are GOOD
BETTER- ON and OFF bipolar cells express different chloride transporters
-makes a definite claim, uses fewer words, uses a strong verb, appeals to wider audience

Write the context (introduction)
-identify the issue/problem, don't use "little is known" or similar phrases. Clearly, SOMETHING is known and by saying that little is known, you negate the work done in the field and look either stupid or ignorant.
-explain current status
-explain approach
-3 paragraphs is perfect
-Don't summarize results

Outline the story in modules
-a module should match the readers attention span (3-5 paragraphs)
-label each module so you and the reader can keep track
-they require a specific temporal order; DON'T go back and forth in time (ie. anesthetizing the animal after surgery)

Diagrams and figures
-make them intrinsically comprehensible; label everything so that you need the legend as little as possible; no one wants to read any more than they have to

Seek criticism
-don't take it personally but acknowledge that it hurts but is helpful
-it's hard to get constructive criticism (few people can identify and articulate problems)
-goals; maximize your understanding of the problems so that you can revise, minimize time for critic (they are busy too), leave the critic feeling good
-don't explain why you wrote it that way; you are just talking to yourself

Sentence structure
-avoid wordiness
-exploit stress positions (beginning should be context and end should be the important part of the sentence- the result)
-keep subject and predicate together
-avoid weak verbs with nominalization; use strong verbs!

Examples:
Weak sentences: It is our goal to test the hypothesis that the HNF3 proteins have an important function in endoderm development and hepatic specification. We have generated a mutation in the gene encoding HNF3g via gene targeting as a first step towards this goal.

Stronger sentences: Apparently, the HNF3 proteins affect endoderm development. To test his, we mutated …
-why is this stronger? shorter, stress positions utilized, trim wordiness, active verbs

For consecutive sentences:
-the stress position in sentence one introduces the topic position of sentence two
-use parallel construction and words to tie things together

Paragraphs:
-topic sentence is like a title
-one idea per paragraph, ~5 sentences
-last sentence leads into next paragraph

TIPS:
Use the singular, use the positive and avoid double negatives.
Use past tense in results and present tense after the first stage of the discussion.
Use shifts of direction (however, yet, but) at the beginning of sentences to tell the reader where you are going.
Know who is on your study section or who your reviewers might be and give them credit in your grant/paper.
Balance of space: introductiondiscussion
Avoid using "little is known" statements; tell what IS known.
Get to the punchline/moral before the reader loses interest.
When a key piece to your story is missing, take the time to find it or collaborate to get it done; it's easier to do than to defend a story with a missing piece.
Don't use text to describe a figure.
When a paper is not accepted, keep reading; oftentimes it can be resubmitted once the comments are addressed.
Science is already hard to understand; don't complicate things with confusing writing. Keep sentence structure simple!
Avoid excessive use of "we"; you are not the important part of the paper, the results are.
Avoid metadiscourse; TRIM IT!