Thoughts on Passive Voice

NOTE: This article discusses Prosecraft.io, a linguistic analysis website we discontinued in August 2023. To learn more how things work today, check out our follow up article: How Shaxpir Does Linguistics

* * *

I got an email from an author yesterday, asking me a few questions about the new Passive Voice analysis features used by the Prosecraft.io linguistics database and the Shaxpir writing platform.

I thought he raised an interesting question that would be helpful to share with the broader writing community:

Is there a reason you are just measuring helping verbs rather than the true grammatical definition of passive voice: the subject of the sentence being acted upon by the object?

As a matter of fact, yes, there’s a really good reason!

But first, for context, here’s what the passive-voice highlighter looks like in Shaxpir, with a paragraph of text from Nick Hornby’s novel, Juliet Naked:

Arguably, according to the common definition of passive voice, there’s only one instance of passive-voice in the whole paragraph (“The drawing was hung over the fireplace…”). The drawing is the subject of the sentence, and it was acted upon (hung!) by some unknown picture-hanger at some unknown point in the past.

But let’s step back for a moment and ask ourselves why we even care.

The reason we avoid passive voice in the first place is that it describes a state-of-being, rather than the act-of-doing. Eliminating passive voice lets us directly join the action, rather than observing the side-effects of the action.

So, if you have a passive-voice sentence like “John was punched in the face,” you could rephrase it to say “Jane punched John in the face” so that the readers can actually watch the action unfolding before their eyes.

But there are other grammatical constructions that impose the same kind of distancing effect between the author and the reader, and we’d like to get rid of those other weak constructions too.

Predicate Nominative

The predicate nominative construction uses auxiliary verbs (aka “helping verbs”, like be, am, is, was, might, must, etc) to describe an equivalency between two nouns. For example: “Jane is a student.”

Just like in our previous example, this example shows us a state-of-being, rather than showing us how the action unfolded. We could replace this sentence with something like: “Jane stuffed the textbook into her already-overcrowded backpack and started the long walk to school.”

Predicate Adjective

Following the exact same principle, the predicate adjective construction uses helping verbs to ascribe some descriptive property to a noun. For example: “John was exhausted.”

Just like with the other two constructions, we still don’t get to see the real action. We’re still just observing a state-of-being rather than watching a story happening in real-time. In this case, we could replace the sentence with something like “John stumbled into the room, his eyes only half-open, exhausted to the core.”

Predicate Prepositional Phrase

Just like the previous three examples, the predicate prepositional phrase construction uses helping verbs to ascribe a predicated relationship to some noun. For example: “Jane is in the Navy.”

The preposition “in” describes the relationship between “Jane” and “the Navy,” and just like our other examples, this construction describes a state-of-being rather than the act-of-doing.

A Sense of Detachment

All four of these constructions use auxiliary verbs to impose a layer of indirection between the reader and the true action of the story. And when you stack helping verbs into sequences (“Jane might have been in the Navy”), it just intensifies the degree of detachment. In this example, a construction like “might have been” doesn’t just describe a state-of-being. It actually describes a speculative state-of-being, which is two steps removed from the real action of the story. That’s why we count each and every auxiliary verb, and use those total counts to determine the sum-total of passive voice in a body of text.

Incidentally, this is probably the reason why most English-language novels are written in the past-tense. The present-tense — which you’d expect to have more immediacy — uses so many helping verbs (“Jane is running”) that it creates a sense of linguistic detachment. The easiest way to overcome that indirectness is by rewriting in the past-tense (“Jane ran”). So we end up with most mainstream novels preferring past-tense over present-tense, because all those helping verbs make present-tense sound awkward and detached.

Shaxpir’s linguistic tools use the helping-verb ratio (the total number of helping verbs, divided by the total word-count) as a way of measuring the degree of passiveness in any selection of text.

We’ve actually never seen a book that eliminated helping verbs completely. They’re legitimately useful, and your writing would probably be much more awkward if you completely outlawed words like “be” and “is” from your vocabulary.

So don’t think of the Passive Voice metrics on Prosecraft.io (or the linguistic highlighters in Shaxpir) as error-checkers, looking through your prose for mistakes. That’s not the point. This isn’t like spell-check. Writing narrative fiction is an art form. It’s not a math problem or a computer program or a grammar test, so there are no errors.

But your linguistic choices have consequences.

The more you use auxiliary verbs, the more you imbue your writing with a sense of passive detachment, or abstract reflection. As long as you use those constructions mindfully, you can use that sense of detachment to pause the flow of action and make room for contemplation or reflection.

* * *

Sign up for sour creative writing platform, Shaxpir 4: Everyone, which gives any author the power to brainstorm, outline, write, revise, and publish a complete novel or memoir. For anyone, anywhere. For free, forever.