Wednesday, August 14, 2013

Making science personal (when appropriate)

In a recent Chemistry World blogpost, Philip Ball suggests that scientists should be more proactive in inserting themselves into scientific papers. Ball claims that using a first-person narrative style provides greater clarity in writing as well as an increased individual investment in the science. The enthusiasm for the proposal was evident immediately on social media, but struck a nerve in me about the quality scientific writing generally. There are several caveats to this practice that are not explored in sufficient detail. There is also some ambiguity in the commentary as voice is mixed with narrative style. One does not need to use pronouns to write in an active voice, and passive voice writing can use personal pronouns. Regardless of the first-person narrative issue, an active voice is usually favorable to a passive one in scientific writing.

Having reviewed numerous manuscripts and collaborated on writing both papers and grants, first-person narrative is often used as a stylistic writing crutch, and does very little to achieve the goals outlined by Ball. Training students in scientific writing is an important part of a PI’s job, although this easily can be forgotten in the publish or perish environment of modern academic science. The majority of Ph.D. students will not pursue careers in academia, and they must develop writing skills that reflect professional practices. Journal publications will likely represent the minority of a typical student’s scientific writing, so we need to consider how to use this writing experience as a tool to prepare them for the future.
One of my mantras to students, which relates to first-person style, is economy of words. I commonly encounter practices like:

   “It has been shown that palladium is an effective catalyst for C-C bond formation1
   “Smith has shown that palladium is an effective catalyst for C-C bond formation1

The qualifier “it has been shown” is verbose and should be removed. “Palladium is an effective catalyst for C-C bond formation1” is the useful information. The acknowledgement of an individual’s or group’s contribution to science may have merit, but like the practice of first-person narrative, the impact becomes diluted with overuse. The name inclusion also duplicates information contained in the citation. When reading a review that makes liberal use of names, eventually you will mentally block out the names completely to focus on the scientific content.

Contrast this style to using names for emphasis. Using a name can help denote a particularly impactful discovery or indicate significant contributions to a field by a particular group/individual. In my papers, Roger Tsien and Graham Ellis-Davies are names that appear with some regularity because each can claim to be a pioneer in photocaged complexes, and both have published numerous seminal papers in the field. I am less likely to mention other practitioners by name, nor would I expect them to do so when citing my work.
The use of “we” or “I” can be effective when used appropriately. So when should the first-person narrative be used, and when is it unnecessary (or worse)?

Useful: putting your efforts into context of the field as a whole or the long-term goals of your research program – emphasize where you fit in the picture and what you will contribute.

Borderline: stating “we are interested in…” – close to the previous point, but too specific. You are publishing a paper on the subject, so it’s safe to assume that you’re interested in the topic.

Not useful: any form of “we did this experiment…” – your name is on the paper describing the science anything restating this is redundant.

Not useful: using first person to describe a routine observation or result. Again, this is redundant with the author list.

Borderline: using first person to describe an unusual/unexpected result – the crux of Ball’s assertion is the need a writing style that hedges against individual bias and fallibility. So one must ask the question, to what degree does the data represent a universal truth?

Most useful: providing your explanation and implication of results – this relates to Ball’s fallibility argument too. It emphasizes that this is your best analysis based on the available data, but also leaves room for what is to come. Future research could reinforce or contradict the current interpretation.

The latter seems to be the most important use. In this situation “we” has real meaning. It conveys something about the relationship between the scientist and the science.

1 comment:

  1. This is a balanced and insightful post. I think that judicious use of "we" or, less frequently nowadays in the age of group science, "I" is effective and honest. Amateur writers overuse the passive, and it can be verbose (as you note) and distracting.

    Here is a classic article by the Nobel Laureate Peter Medewar. It is also a critique of the scientific article but consists of a more fundamental argument.

    Is The Scientific Paper a Fraud?