Among the elementary advice you’d give to someone who wished to learn to write well would be to use the active voice and not the passive.

The reason is that we process sentences “verb first”. In the sixties there was a conflict between those who believed that we actually formulate sentences by building out from the verb and those who believed we use a more structured approach. Although the structured approach “won”, there’s good evidence that we work “verb first”. Active sentences prove quicker to parse and anecdotally we know how hard passive sentences can be to work through.

It’s actually quite hard to precisely define the passive in traditional ways. In the passive the subject experiences the action of the verb, rather than performs it.

He hits the boy.

The boy is hit by him.

You could consider it a “topicaliser” (we know that English loves to topicalise!). This is because English, like many languages, likes to put the most salient information in the sentence, after the verb, first. We call this “fronting” when we’re discussing adverbs or adverbial phrases.

In fact, I’d argue the passive is an “ergativiser”. I know, your head just exploded. I’ll explain.

First of all, let’s talk about two concepts: “marking” and “inflection/agglutination”. Marking is simply a means languages use to show (usually) that words with related meanings have different functions. We generally consider that the words are stored in a sort of “dictionary form” and then marked for function. It’s easy to see this by example.

Dog

Dogs

The plural ending /s z/ is added to words to make them plural. It “marks” the plural.

fire

fired

The simple past ending /t d/ is added to weak verbs to make them past. It “marks” the past. (Before you go off, yes, this is also the marking for the past participle. For historical reasons, it’s the same ending: originally the participle was prefixed (as it is in German) but the prefix was lost.)

Marking does go a little deeper than that.

mark

marked

unmarked

Here we can see that it’s a question of relation. “Marked” is marked from “mark”, but “unmarked” is marked from “marked”. It can also be considered marked from “mark” but it’s generally easier to understand as a binary opposition because this is how marked forms are nearly always used, and how the concept was introduced.

Second, we can look at inflection (or agglutination in agglutinative languages — same idea really but whereas in inflection you change parts of a word to show its function, in agglutination you add bits on).

For example:

amo

amat

Generally, in Latin we consider that the “stem” has “endings” and “amo” and “amat” are marked for first and third person present tense. This is actually a different type of “marking” to that we’ve already discussed, and consequently linguists often use a different word for it.

Let’s compare “amo” with “amavi”. The first means “I love” or “I am loving” and is the present tense. The second means “I loved” or “I have loved” and is the perfect tense.

So verbs can be inflected, or marked, for person and tense.

One of the more interesting things in languages is ergativity. In this, we say the following: subjects of intransitive verbs are marked the same way as objects of transitive verbs.

Woah.

What’s a transitive verb? Well, that’s actually quite easy. It’s one that is done to something or someone else. Contrast “take” with “breathe”. You can’t breathe somebody. And you can’t just take. Breathe is intransitive (never has an “object”); take is transitive (has an object). Some verbs can be both: “eat” for instance.

Look at these two sentences:

I breathe.

I take him.

At a deep level, the verbs can be thought of as “action of breathing” and “action of taking”. Subjects are seen as connected with actions. So in ergativity, these two sentences are seen as “there’s an action of breathing and I undergo it” and “there’s an action of taking caused by me and he undergoes it”.

It’s a bit different from how we conceive of it but it does make sense.

Now it just so happens that the distant ancestor of English, proto-Indo European was a “split ergative” language. In the past tenses, it was ergative. This means verbs were conceived differently when they were in past tenses.

So how do you “ergativise” a verb in English? Let’s look again at an active vs a passive.

I kiss the boy.

The boy is kissed by me.

Usually, a grammarian type will tell you that the passive uses the “past participle” of the verb “kiss”. It’s actually really unfortunate that we use that terminology because if we were taught at school what “kissed” is, we’d understand a lot of English grammar much more readily. Remember what I said: it is a historical coincidence that the “past participle” and the “simple past” are the same word. They have converged with time.

So what is “kissed”? It is a “verbal adjective”. It describes the “action of kissing” as a state rather than a process. There are two types of verbal adjective in English: the present and past participles. The present participle describes actions as continuing; the past participle describes them as complete. Here is somewhere that the technical terms are much more useful for the linguist. They ought to be called the “imperfect participle” and the “perfect participle” because in fact they have nothing at all to do with “tense”. They are marked for “aspect”.

You can perhaps see that when you look at “kissed” as “the action of kissing as one complete action” and “kissing” as “the action of kissing as a continuing action”, you can understand how ergative languages can see verbs as describing actions that don’t “belong” to anyone.

Incidentally, participles really are adjectives pure and simple. The passive is used exactly like a predicative adjective. Compare:

The boy is fat.

The boy is kissed.

Notice that the passive does not need an “agent” any more than “fat” does. You just are fat. No one needs to have made you that way. But you cannot say:

The boy is fat by me.

And you can say:

The boy is kissed by me.

This is fine of course. There are adjectives that can fill different slots in sentences by their nature. You can say:

He was made fat.

But not:

He was made kissed.

So verbal adjectives are not quite identical functionally to what you might call common adjectives. Which is okay. Just as we can have transitive and intransitive verbs, there can be different sets of adjectives and we can still consider them the same thing because functionally they do the same thing in many structures.

I think that it’s probably best to think of “by me” as an “inflection”. Usually, in linguistics, “The boy is kissed by me” is seen as derived from “I kissed the boy” but to be honest, I’m not sure about that. I think we may have been wrong to dismiss generative semantics entirely and English sentences might be better seen as verb first. After all, few enough sentences lack a verb! Seeing “was kissed” as a verb form loses the insight that “kissed” is an adjective. Usually, the passive is seen as a way that English makes a “patient” (someone the action of the verb happens to) a subject (grammatically the “possessor” of the action of the verb). And traditionally we’ve said that is accomplished by using a different verb form because that’s how Latin did it (the passive is an inflected form of the verb in Latin).

But as so often, applying Latin grammar to English doesn’t really work. We’d do better to first of all understand that participles are adjectives of the verb and that we form passives by making verbs de-active, not by making patients into subjects and changing verb form.

I know that last sentence is difficult. What is the difference between:

I kiss.

I am kissed.

In the first, the word “kiss” describes a process. It’s dynamic.

In the second, the word “kissed” describes a state. It’s static.

So to form the passive, we change the verb from a process to a state and then make whatever undergoes that state into the subject of the sentence by using the participle as a predicative adjective.

And if I’m giving the grammar lessons, I tell my students to use verbs to do the work of sentences and avoid adjectives. Sentences with the passive voice feel static because they describe states. The verb “to be” is generally stative and rarely used to describe a process.