Writing #4--Organizing a paper
Bearing in mind that
this blog is mostly directed at grad students, postdocs, and earlier-career
professionals, I pick up here on some of my thoughts about writing.
Writing a scientific paper is one of those things that we should have
been taught systematically, but often were not. Often it's kind of hit or
miss; we pick things up along the way, during thesis revision, etc. So if
you are feeling a bit uncertain about some aspects of writing, this blog will help, I hope.
One of the things you have almost certainly been taught, or already picked up on, is that scientific papers are expected to have a certain form of organization. The basic elements are Abstract, Introduction, Methods, Results, and Discussion. In geology, if the length of the paper makes it easier, we commonly have a section titled "Previous Work" between the Introduction and the Methods. We also may have a separate section for "Conclusions", which follows the Discussion.
I think this form is recognizable to most readers and is pretty self-explanatory, so I will make just a couple of comments. I can elaborate on the basics if readers request it. The first comment is, write the abstract last. The abstract should contain just the findings. A one-sentence introduction is helpful, but it should be no longer than that. If warranted, include your conclusions, if they are not obvious from the findings. The "Conclusions" should be similarly brief and do not need to include the justification; you've just provided that in the "Discussion".
Three common problems arise in the organization of papers. One is repetition. As you write, you should decide exactly where a particular point you want to raise should go. Sometimes it's not obvious. If you find yourself repeating a point--you will clue in on this if you find yourself using the words "as previously mentioned"--stop and question yourself. "Am I making the same point twice?Do I really need to make this point twice? Where is the most appropriate place to make this point?"
The second problem is not using parallel construction. It is hard to overemphasize how important parallel construction is. It is a major technique for ensuring transparency in your writing (see Writing #2a). If you present information in a particular order, maintain that order as you develop your paper. A simple example would be a statement such as, "Three aspects of this work are critical to understanding the phenomenon: X, Y, and Z." Then you go on to develop your arguments (or provide information) about X, Y, and Z. Do it in that order: X, Y, and Z. Don't do it as Y, X, and Z or any other variation. Parallel construction can affect single sentences to sections to entire papers.
The third, and by far most common, type of organizational error is mixing up the introduction and methods, the methods and results, and the results and discussion. The latter two pairs are particularly common. Before submission, you should scrutinize every sentence and make sure it's in the right place.
"But", you say, "my work has several levels of results and interpretations! If I am ruthless about breaking these out completely, it will be confusing!"
That is entirely possible. Say, for example, you have done some petrographic work on two different environments. You have to interpret the petrography and then you have to interpret the environments. And then you have to place your results in a larger context.
This is how I've handled this problem. I start by asking myself if I really do have those levels of interpretation. Sometimes I don't, not really. But if I do, I present the results and interpretations at two levels. I'll give an example from a recent paper. We had petrographic results, which we used to define facies. This was a major component itself because we were the first to attempt to define facies on these kinds of rocks. So, we had a "Results and Interpretation" section (level 1) that was divided into three parts. The first part was "Carbonate facies". That section was divided into the facies descriptions (i.e., "results") and other sections discussing interpretations of the biota, evidence for weathering, etc. (i.e., "discussion"). The second part was "Field relationships, facies, and interpretation--Site 1". That was divided into "description" (i.e., results) and "interpretation". The third part was the same thing but for Site 2. Although technically "carbonate facies" is a separate topic from the "field relationships, etc." for the two sites, placing them under the same heading hierarchy worked.
Then we discussed the entirety of the study, the overall interpretation of our findings, i.e., a comparison of the two sites and the broader implications of the study.
It is not common to mix results into the discussion, but it is all too easy to discuss details of the results when presenting them. You must resist the temptation to do so. There was an old TV show called "Dragnet", and the lead role was a detective, Det. Joe Friday. He is purported to have said, "Just the facts, ma'am."* That's how you should write the results--just the facts.
If you don't have two levels of results and discussion, as in the example, you can handle discussion of the details by dividing the Discussion section into two parts. These are, roughly, "discussion of the details of the findings" and "broader interpretations/implications". You have a lot of freedom in using different headings in the discussion. Don't be shy about using them, so long as they aren't cute or confusing.
The main thing is to convey the information as succinctly and transparently (see Writing #2a) as possible.
*Joe Friday actually never said this. Just like "Elementary, my dear Watson", which wasn't said by Sherlock Holmes, and my bête noire, "Judy, Judy, Judy", which Cary Grant never said, this is one of those phrases that was distorted then shortened in the popular vernacular. But it is particularly apt for describing what the "Results" section should contain.
One of the things you have almost certainly been taught, or already picked up on, is that scientific papers are expected to have a certain form of organization. The basic elements are Abstract, Introduction, Methods, Results, and Discussion. In geology, if the length of the paper makes it easier, we commonly have a section titled "Previous Work" between the Introduction and the Methods. We also may have a separate section for "Conclusions", which follows the Discussion.
I think this form is recognizable to most readers and is pretty self-explanatory, so I will make just a couple of comments. I can elaborate on the basics if readers request it. The first comment is, write the abstract last. The abstract should contain just the findings. A one-sentence introduction is helpful, but it should be no longer than that. If warranted, include your conclusions, if they are not obvious from the findings. The "Conclusions" should be similarly brief and do not need to include the justification; you've just provided that in the "Discussion".
Three common problems arise in the organization of papers. One is repetition. As you write, you should decide exactly where a particular point you want to raise should go. Sometimes it's not obvious. If you find yourself repeating a point--you will clue in on this if you find yourself using the words "as previously mentioned"--stop and question yourself. "Am I making the same point twice?Do I really need to make this point twice? Where is the most appropriate place to make this point?"
The second problem is not using parallel construction. It is hard to overemphasize how important parallel construction is. It is a major technique for ensuring transparency in your writing (see Writing #2a). If you present information in a particular order, maintain that order as you develop your paper. A simple example would be a statement such as, "Three aspects of this work are critical to understanding the phenomenon: X, Y, and Z." Then you go on to develop your arguments (or provide information) about X, Y, and Z. Do it in that order: X, Y, and Z. Don't do it as Y, X, and Z or any other variation. Parallel construction can affect single sentences to sections to entire papers.
The third, and by far most common, type of organizational error is mixing up the introduction and methods, the methods and results, and the results and discussion. The latter two pairs are particularly common. Before submission, you should scrutinize every sentence and make sure it's in the right place.
"But", you say, "my work has several levels of results and interpretations! If I am ruthless about breaking these out completely, it will be confusing!"
That is entirely possible. Say, for example, you have done some petrographic work on two different environments. You have to interpret the petrography and then you have to interpret the environments. And then you have to place your results in a larger context.
This is how I've handled this problem. I start by asking myself if I really do have those levels of interpretation. Sometimes I don't, not really. But if I do, I present the results and interpretations at two levels. I'll give an example from a recent paper. We had petrographic results, which we used to define facies. This was a major component itself because we were the first to attempt to define facies on these kinds of rocks. So, we had a "Results and Interpretation" section (level 1) that was divided into three parts. The first part was "Carbonate facies". That section was divided into the facies descriptions (i.e., "results") and other sections discussing interpretations of the biota, evidence for weathering, etc. (i.e., "discussion"). The second part was "Field relationships, facies, and interpretation--Site 1". That was divided into "description" (i.e., results) and "interpretation". The third part was the same thing but for Site 2. Although technically "carbonate facies" is a separate topic from the "field relationships, etc." for the two sites, placing them under the same heading hierarchy worked.
Then we discussed the entirety of the study, the overall interpretation of our findings, i.e., a comparison of the two sites and the broader implications of the study.
It is not common to mix results into the discussion, but it is all too easy to discuss details of the results when presenting them. You must resist the temptation to do so. There was an old TV show called "Dragnet", and the lead role was a detective, Det. Joe Friday. He is purported to have said, "Just the facts, ma'am."* That's how you should write the results--just the facts.
If you don't have two levels of results and discussion, as in the example, you can handle discussion of the details by dividing the Discussion section into two parts. These are, roughly, "discussion of the details of the findings" and "broader interpretations/implications". You have a lot of freedom in using different headings in the discussion. Don't be shy about using them, so long as they aren't cute or confusing.
The main thing is to convey the information as succinctly and transparently (see Writing #2a) as possible.
*Joe Friday actually never said this. Just like "Elementary, my dear Watson", which wasn't said by Sherlock Holmes, and my bête noire, "Judy, Judy, Judy", which Cary Grant never said, this is one of those phrases that was distorted then shortened in the popular vernacular. But it is particularly apt for describing what the "Results" section should contain.
Comments
Post a Comment