How many of us have read a science paper and wondered what disembodied hand was at work? Literally. Scientists perform surgical procedures on animals and then kill them. Then they analyze the results and report them. They divide people into groups and administer experimental medications or placebos in double-blind controlled experiments. But conventions in science writing would have us believe that no one did any of those things. In Science World, rat brains section themselves, and research subjects suddenly find themselves swallowing tablets, which have sorted themselves into the drug or the placebo. And, in this way, science remains free of the taint of subjectivity. All that is required to keep science pure is to banish the wrong pronouns.
The guy who first started talking like that was convicted of taking bribes. So Francis Bacon should not be an expert on professional objectivity. Meanwhile, the names of the authors are on the front page of the articles, so we know that they were involved. Why do they need to sound like Oliver North?
Experiments are written to be replicated. If you can’t replicate the experiment, it doesn’t matter who drilled the holes. But by adhering to what is stone fiction, writing and thinking becomes unnecessarily complicated.
Here are some examples:
Subsequently, each rat was placed in a stereotaxic apparatus for the brain (SR-6R; Narishige, Tokyo, Japan) until the end of each experiment, as described previously in a published work of our laboratory (Shimizu et al., 2004). The skull was drilled for intracerebroventricular administration of drugs using a stainless-steel cannula (outer diameter of 0.3 mm). The stereotaxic coordinates of the tip of the cannula were as follows (in mm): AP –0.8, L 1.5, V 4.5 (AP, anterior from the bregma; L, lateral from the midline; V, below the surface of the brain), according to the rat brain atlas (Paxinos and Watson, 2005).
When SD rats were used, a catheter was inserted into the bladder from the bladder dome under urethane anesthesia (1.0-1.2 g/kg, i.p.) in order to perform continuous or single CMG.
METHODS: Between December 2009 and December 2010, 16 adult non-randomized patients (ATG group), receiving a liver graft from a deceased donor, received an intraoperative infusion (4-6 h infusion) of thymoglobulin (3 mg/kg, ATG: Thymoglobuline®). These patients were compared (case control approach) with 16 patients who had a liver transplant without ATG treatment (control group) to evaluate the possible effects of intraoperative ATG infusion.
Mice were sacrificed by rapid decapitation after CO2 asphyxiation and the VNOs were dissected out into mouse artificial cerebrospinal fluid (mACSF) that was continuously bubbled with 5 % CO2/95 % O2 and maintained at 4 °C. The tissue was embedded in a gel composed of 4 % low melting point agarose dissolved in mACSF at 37 °C, chilled on ice, mounted on a specimen tray, and secured onto the VF-300 microtome sectioning system (Precisionary Instruments). Tissue samples were sectioned into 100 μm slices, which were then transferred to mACSF solution and continuously bubbled with 5 % CO2/95 % O2 at room temperature. The composition of mACSF is (in mM): NaCl 125, KCl 2.5, CaCl2 2, MgCl2 1, NaHCO3 25, Na2HPO4 1.25, Glucose (Dextrose) 10.
Combination of we and passive voice:
After four hours of stopping nutrition, we administered lipopolysaccharide (lipopolysaccharide from Escherichia coli 0111:B4, Quadratech Diagnostics Ltd.) intraperitoneally, in gradually increasing doses 3, 5, 10 mg/kg, to experimental subjects included in groups 1, 2 and 3, for induction of sepsis and ARDS. In the control group, the rats were treated the same as in LPS groups, except LPS is substituted with an equal volume of normal saline (NS). Lung injury and systemic effects were quantified in terms of hypoxemic, laboratory and histopathological changes at 6 and 24 hours after LPS administration.
Scientists do discuss this among themselves. For example, see Yateendra Joshi and Professor David M. Schultz. Professor Schultz notes that the use of the first person in science appears to be as common among scientific giants as the singular they is in literary history:
In Eloquent Science (pp. 76-77), I advocate that first-person pronouns are acceptable in limited contexts. Avoid their use in rote descriptions of your methodology (“We performed the assay…”). Instead, use them to communicate that an action or a decision that you performed affects the outcome of the research.
There is a lively discussion in the comments section of Professor Schultz’s article, too. For example, a commenter named Bill Lott says:
This argument is approximately correct, but in my opinion off point. The use of first person should always be minimized in scientific writing, but not because it is unacceptable or even uncommon. It should be minimized because it is ineffective, and it is usually badly so. Specifically, the purpose of scientific writing is to create a convincing argument based on data collected during the evaluation of a hypothesis. This is basic scientific method. The strength of this argument depends on the data, not on the person who collected it. Using first person deemphasises the data, which weakens the argument and opens the door for subjective criticism to be used to rebut what should be objective data. For example, suppose I hypothesized that the sun always rises in the east, and I make daily observations over the course of a year to support that hypothesis. I could say, “I have shown that the sun always rises in the east”. A critic might respond by simply saying that I am crazy, and that I got it wrong. In other words, it can easily become an argument about “me”. However, if I said “Daily observations over the course of a year showed that the sun always rises in the east”, then any subsequent argument must rebut the data and not rebut “me”. Actually, I would never say this using either of those formulations. I would say, “Daily observations over the course of a year were consistent with the hypothesis that the sun always rises in the east.” This is basic scientific expository writing.
Finally, if one of my students EVER wrote “it was found that …”, I would hit him or her over the head with a very large stick. That is just as bad as “I found that …”, and importantly, those are NOT the only two options. The correct way to say this in scientific writing is, “the data showed that …”.
And then there is this:
Scientific papers are meant to only include scientific facts. The word “I” often is used to indicate that something is your personal views, rather than what can be proven scientifically.
In scientific papers, the facts are important, not the people. By using the word “I”, you put yourself, instead of your results, in focus.
But science is supposed to be rational. Therefore, it should describe reality accurately regardless of whose ox has been gored. We know that the skull was drilled for intracerebroventricular administration of drugs and we know that Takahiro Shimizu, Shogo Shimizu, Youichirou Higashi, Kumiko Nakamura, Naoki Yoshimura, or Motoaki Saito did the drilling. Or one of their undergraduates did. Why be so precious? Are we going to think that the experiment is more valid because they left out the we?
We shouldn’t. Because in order for that experiment to be valid, someone has to be able to do exactly what they describe in their research and get the same results. So, in fact, by giving in to the idea that the absence of a pronoun makes the heart grow fonder—I mean makes the science more credible—we adhere to a kind of faith. Rationally, we know that all of the scientists in all of the studies I’ve listed in this blog post either performed the acts described in their experiments or their subordinates did. Excising pronouns merely lulls us into a false sense of validity. Are those experiments any less able to be replicated if someone says I or speaks in the active voice?
It is more objective and accurate to acknowledge the truth. People devise, carry out, and report scientific experiments. And other people have to be able to carry out those same experiments and get the same results. It would be best for all of us for those experiments to be described with as much clarity as possible. That includes admitting that there is a hand on that scalpel.
1. Takahiro Shimizu, Shogo Shimizu, Youichirou Higashi, Kumiko Nakamura, Naoki Yoshimura, and Motoaki Saito. “A Stress-related Peptide bombesin centrally induces frequent urination through brain bombesin receptor types 1 and 2 in the rat.” The Journal of Pharmacology and Experimental Therapeutics January 4, 2016 (DOI: 10.1124/jpet.115.230334) (http://jpet.aspetjournals.org/content/early/2016/01/04/jpet.115.230334.abstract)
2. De Pietri L, Serra V, Preziosi G, Rompianesi G, Begliomini B, “Perioperative effects of high doses of intraoperative thymoglobulin induction in liver transplantation,” World J Transplant 5(4) (Dec 24, 2015) 320-328. doi: 10.5500/wjt.v5.i4.320
3. SangSeong Kim, Limei Ma, Jay Unruh, Sean McKinney, and C. Ron Yu, “Intracellular chloride concentration of the mouse vomeronasal neuron,” BMC Neurosci. 16:90 (2015). doi: 10.1186/s12868-015-0230-y, http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4678706/.
4. Raluca-Ştefania Fodor, Anca Meda Georgescu, Adrian-Dan Cioc, Bianca Liana Grigorescu, Ovidiu Simion Cotoi, Pal Fodor, Sanda Maria Copotoiu, Leonard Azamfirei,
“Time- and dose-dependent severity of lung injury in a rat model of sepsis,” Rom J Morphol Embryol 56(4) (2015) 1329–133, http://www.rjme.ro/RJME/resources/files/56041513291337.pdf.