Should the passive voice be avoided?
It’s another very busy day (as well as another lovely one) so I thought instead of sitting indoors this lunchtime writing a typically verbose blog item I’d just pick something out of my back catalogue and give it another airing because it deals with something that’s come up a couple of times recently.
This is the time of year when final-year students are drafting their project reports. Yesterday I was back in Cardiff giving feedback on two such articles. I usually quite enjoy reading these things, in fact. They’re not too long and I’m usually pretty impressed with how the students have set about the (sometimes quite tricky) things I’ve asked them to do for their project work. I think the project report is quite a challenge for UK physics students because they generally haven’t had much practice in putting together a lengthy piece of writing before or during their university course, so haven’t developed a style that they feel comfortable with and are often unfamiliar with various conventions (such as reference style, punctuation of equations, etc). Some of these are explained in quite a lot of detail in the instructions the students are given, of course, but we all know that only girls read instructions….
The thing that strikes me most forcibly about the strange way students write project reports is that they are nearly always phrased entirely in the passive voice, e.g.
The experiment was calibrated using a phlogiston normalisation widget….
I accept that people disagree about whether the passive voice is good style or not. Some journals actively encourage the passive voice while others go the opposite way entirely . I’m not completely opposed to it, in fact, but I think it’s only useful either when the recipient of the action described in the sentence is more important than the agent, or when the agent is unknown or irrelevant. There’s nothing wrong with “My car has been stolen” (passive voice) since you would not be expected to know who stole it. On the other hand “My Hamster has been eaten by Freddy Starr” would not make a very good headline.
The point is that the construction of a statement in the passive voice in English is essentially periphrastic, in that it almost inevitably involves some form of circumlocution – either using more words than necessary to express the meaning or being deliberately evasive by introducing ambiguity. Both of these failings should be avoided in scientific writing.
Apparently, laboratory instructors generally tell students to write their reports in the passive voice as a matter of course. I think this is just wrong. In a laboratory report the student should describe what he or she did. Saying what “was done” often leaves the statement open to the interpretation that somebody else did it. The whole point of a laboratory report is surely for the students to describe their own actions. “We calibrated the experiment..” or “I calibrated the experiment…” are definitely to be preferred to the form I gave above.
That brings me to the choice of pronoun in the active voice. One danger is that it can appear very bombastic, but that’s not necessarily the case. I don’t find anything particularly wrong in saying, e.g.
We improve upon the technique of Jones et al. (1848) by introducing a variable doofer in the MacGuffin control, thereby removing gremlins from the thingummy process.
But the main issue is whether to use the singular or plural form. It can be irritating to keep encountering “I did this..” and “I did that..” all the way through a journal paper, and I certainly would feel uncomfortable writing a piece like that in the first person singular. I think it feels less egotistical to use “we”, even if there is only one author (which is increasingly rare in any case). If it’s good enough for the Queen it’s good enough for me! However, I just looked “we” up in Chambers dictionary and found
..used when speaking patronizingly, esp to children, to mean `you’.
which wasn’t at all what I had in mind!
However in the case of a student project that I’m assessing I actually want to know what the particular student writing the report did, not what was done by person or persons unspecified or by a group of uncertain composition. So I encourage my students to put, e.g.,
I wrote a computer program in 6502 Assembly Language to solve the Humdinger equation using the Dingbat-Schnitzelgruber algorithm.
I also (sometimes) like “we” when there’s, e.g., a complicated mathematical derivation. Going line by line through a lengthy piece or difficult technical argument seems friendlier if you imagine that the reader is trying to do the calculation along with you as you write it:
if we differentiate the right hand side of equation (8), use the expression for x obtained in equation (97), expand y in a power-series and take away the number we first thought of we find…
The “we” isn’t necessarily an author with delusions of grandeur (or schizophrenia), but instead denotes a joint operation between author and reader.
Anyway, to resume the thread, it seems to me that sometimes it is appropriate to use the passive voice because it is the correct grammatical construction in the circumstances. Sometimes also the text just seems to work better that way too. But having to read an entire document written in the passive voice drives me to distraction. It’s clumsy and dull.
In scientific papers, things are a little bit different but I still think using the active voice makes them easier to read and less likely to be ambiguous. In the introduction to a journal paper it’s quite acceptable to discuss the background to your work in the passive voice, e.g. “it is now generally accepted that…” but when describing what you and your co-authors have done it’s much better to use the active voice. “We observed ABC1234 using the Unfeasibly Large Telescope..” is, to my mind, much better than “Observations of ABC1234 were made using..”.
Reading back over this post I notice that I have jumped fairly freely between active and passive voice, thus demonstrating that I don’t have a dogmatic objection to its use. What I’m arguing is that it shouldn’t be the default, that’s all.
My guess is that a majority of experimental scientists won’t agree with this opinion, but a majority of astronomers and theoreticians will.
This guess will now be tested by reactivating an old poll..
In the Dark 
May 1, 2013 at 1:55 pm
Gremlin Removal from the Thingummy Process via the Introduction of a Variable Doofer in the MacGuffin Control
– I think I’ve seen this on http://snarxiv.org/
May 1, 2013 at 4:04 pm
I think that your view is exactly right. The problem with the passive voice is that it often (but by no means always) leads to indirect, vague sentences. One should generally avoid indirect, vague sentences. “Avoid the passive” is a not-terribly-bad heuristic for “avoid indirect, vague sentences,” but it’s much better to teach people to develop an ear for strong writing and to write what sounds best.
Here’s what I tell my students. Writing teachers often tell you never to use the passive voice. Scientists sometimes tell you to use it all the time. Once you realize this, it’s oddly liberating: someone will be annoyed with you no matter what you do, so do whatever sounds best.
A few scientific journals (not any that I regularly deal with) have official policies mandating the passive. That is, of course, absurd. As long as you’re boycotting Elsevier anyway, boycott these journals too. ( http://allendowney.blogspot.com/2012/04/passive-voice-is-hoax.html )
Geoffrey Pullum has a good piece on the passive voice in Lingua Franca: http://chronicle.com/blogs/linguafranca/2011/10/01/mistakes-are-made/ . Pullum also has an entertaining takedown of Strunk and White’s Elements of Style, which remains a mainstay of writing instruction in the US: http://chronicle.com/article/50-Years-of-Stupid-Grammar/25497 . He devotes a lot of firepower to S&W’s treatment of the passive voice. I don’t agree with everything he says ( http://blog.richmond.edu/physicsbunn/2009/04/24/strunk-and-white/ ), but it’s a good piece nonetheless.
May 1, 2013 at 4:48 pm
I was also told at school never to say that anything was ‘taken’ as a preliminary. If not warned then people typically write, “A test tube was taken and half-filled with molar aqueous solution of sulphuric acid”, rather than simply “A test tube was half-filled with molar aqueous solution of sulphuric acid”.
May 1, 2013 at 5:18 pm
I think it’s a rather archaic usage, but “take” can mean “have sexual intercourse with”, which is probably best avoided in the case of a test tube.
May 1, 2013 at 5:33 pm
It could also be taken (geddit?) to mean “stolen”.
May 1, 2013 at 5:04 pm
There’s the tale (sic) of Hetherington’s Cat. Jack Hetherington was a nuclear physicist who wrote a preprint as sole author, using “we” throughout. He was warned that this would be unacceptable to his target journal (as was “I”) and, as it would be a hassle to rewrite it in the passive in the days before word processors, he simply added a fictitious co-author: a cat – related to his own – called Willard (named after Willard Gibbs). Hetherington recounted the tale in a letter to Physics Today (April 1997, p94), and I see that FDC Willard now has his own page on Wikipedia.
May 2, 2013 at 10:55 am
I think that women may have more difficulty using an active voice. It was in grad school that one of my professors (a woman) pointed this out to me – based on her 30+ years of reviewing students’ papers – and thereafter, it was a trait that I noticed (in myself, using the passive quite unconsciously, as well as in other women writers and researchers).
May 31, 2017 at 4:02 pm
> punctuation of equations
I recall a colleague whose PhD you examined mentioning that you didn’t like her lack of equation punctuation. And rather than just write this once you laboriously corrected every single equation in the thesis!
May 31, 2017 at 4:31 pm
I tend to do that, but I find that if I’m reading something thoroughly I can’t help noticing things like that so I tend write corrections as I go.