In every language in the world, the goal of a clause is to state that [something or someone] [does or is] [something or someone]. The first portion is a noun phrase called the subject, which is the something or someone that does or is something; the second portion is a verb phrase, which is that the subject does or is; the third portion is a noun phrase called the direct object, which is the something or someone that the verb affects directly. At this level of generality, we usually refer to the direct object of the sentence as simply the object.
If we have one or two of these parts of the sentence we can usually figure out the others. For example, consider that we know that
Spot bit John
To get the subject we answer the question:
what/who
To get the verb we answer the question:
To get the object we answer the question:
When we find it difficult to identify one of these parts of the clause, we should go back to these questions.
Languages are often classified by the order in which the subject, verb and object tend to appear in a clause. In English, they tend to appear in the SVO order, while in Japanese they tend to appear in the SOV order. However, this doesn’t mean that they cannot appear in other orders.
Spot bit John
John was bitten by Spot
It is also the case that sometimes a clause genuinely does not have an object, but a clause must always have a subject and a predicate, and a predicate must always have a verb; therefore, a minimal clause will always have a subject and a predicate:
Spot bit
Mary cried
John runs
rain falls
Finally, a clause must always have at least a subject and a verb, but they might be there implicitly instead of explicitly. An implicit subject, verb or object is only possible if it is possible to infer it.
raise your hands
sorry
quiet!
you raise your hands
I am sorry
you be quiet!
If we have the need to write a subject or object explicitly, we do so. If we do not, then we either replace it with a pronoun, or make it implicit. The implication is that a noun phrase that is implicit is standing for a pronoun, not for a noun phrase.
However, noun phrases marked with ‘ga’ are not the same as the subject of the English translation, e.g., following a textbook, ‘mariko ga iku’ would be translated as ‘Mariko goes’, i.e., a matter of fact statement in English, but in Japanese ‘mariko ga iku’ would actually mean ‘It is Mariko! She is the one that goes!’, i.e., a statement with a feeling of novelty or discovery.
the Japanese subject marker ‘ga’
In English, the subject the noun phrase or pronoun that is being or doing something. Likewise, the object of the sentence, who is the person or thing that receives the action, comes after the verb. For example, in ‘the dog bites Bob’, because of the positions of ‘the dog’ and ‘Bob’ in relation to the verb ‘bites’, we know that the subject is ‘the dog’, who does the biting, and the object is ‘John’, who receives it. Just by altering the relative position of the words to ‘John bites the dog’, we turn around the sentence: now ‘Bob’ is the subject and does the biting, while ‘the dog’ is the object and receives it. Languages whose sentences tend to have this Subject-Verb-Object order, like English, are called SVO languages.
In Japanese, the verb is always at the end. The subject and the object, though, don’t have a fixed position although, in general, they tend to have a Subject-Object-Verb order, which loosely qualifies Japanese as a SOV language. However, the point is that this is just a tendency. It is perfectly possible to have the subject preceeding the object or viceversa. This causes a problem because we would end up with sentences ‘John the-dog bites’ and ‘the-dog John bites’ that we would not be able to understand, because we don’t know who performed the action and who received it. To solve this problem, both the subject and the object are followed up with a marker, i.e., a particle that identifies them.
In Japanese, we mark bjects that receive an action with the particle を (wo), and subjects with the particle が (ga). As long as the particles mark their respective noun phrases, we can move them around without loss of clarity:
English
the dog bites Bob
Bob bites the dog
Japanese v.1
[the-dog ga] [Bob wo] bites
[the-dog wo] [Bob ga] bites
Japanese v.2
[Bob wo] [the-dog ga] bites
[Bob ga] [the-dog wo] bites
The particles form a unit with the phrase that they mark, e.g., the phrase [the dog] is the subject phrase in English, then in Japanese the subject phrase would be [the-dog ga].
explicit ‘ga’
We are fairly used to translate English into other western languages without having any problems preserving the feeling of the sentence. For example,
English
Bob eats
Spanish
Bob come
Italian
Bob mangia
French
Bob mange
In all these cases, Bob is the subject, and in all of these cases we are stating the fact that Bob eats. Hence, we would expect that translating ‘Bob eats’ to Japanese would be ‘Bob ga taberu’; however, this is not the case.
The particle ‘ga’ marks the subject of the sentence but, the feeling of the sentence ‘Bob ga taberu’ is one of disclosing new information. Hence ‘Bob ga taberu’ doesn’t mean ‘Bob eats’ but ‘It is Bob who eats’. Here are some examples:
romaji
bob ga taberu
onegai ga arimasu
anata ga suki
watashi ga
Tanaka san ga
wrong translation
bob eats
I have a favor to ask
I like you
I
Mr. Tanaka
correct translation
it is bob who eats!
it is a favor that exists!
it is you who is likeable to me!
it is me!
it is Mr. Tanaka!
Although the sentences above convey the feeling associated with using ‘ga’, a translation using ‘It is ___ who/that…’ is wordy in English. Here are some options:
romaji
bob ga taberu
equivalent correct translations
It is Bob! He is the one who eats.
Bob! He is the one who eats.
Bob is the one who eats!
It is Bob who eats!
Bob! He eats.
Meaning-wise all the translations above appear to convey a similar feeling of discovery. Hence, we will translate the subject phrase ‘[subject ga]’ as ‘subject! pronoun…’. Let’s try it out with our previous examples:
romaji
bob ga taberu
onegai ga arimasu
anata ga suki
watashi ga
tanaka san ga
translation
Bob! He eats
A favor! It exists
You! You are likeable to me
I! I…
Mr. Tanaka! He…
implicit ‘ga’
Since ‘Bob ga taberu’ actually means ‘Bob! he eats’, with a sense of discovery, we might we wondering how do we simply say ‘Bob eats’. Simple, we have to make it so that the subject is implicit, i.e., we need a zero subject.
japanese
taberu
implicit subject
{∅ ga} taberu
explicit subject
{bob ga} taberu
English
Bob eats
For the subject to be implicit, it has to have been mentioned before, or be a default:
taberu
taberu?
bob wa, taberu
{watashi ga} taberu
{anata ga} taberu?
Bob wa, {kare ga} taberu
I eat
You eat?
Bob eats
Japanese default subjects
Let’s consider again the would-be Japanese version of the first lines of ‘Moby Dick’:
We can infer in all the sentences that follow the first one that the speaker is Ishmael, so any time he says ‘I’ or ‘me’ or ‘my’, which we have replaced with the ∅ pronoun, he is refering to himself.
Now, let’s consider just the first sentence in isolation, i.e., in English – “call me ‘Ishmael'” – and in Japanese “call ∅ ‘Ishmael'”. In English we can infer that the speaker is talking about himself because he uses the pronoun ‘me’; however, in Japanese we have removed that clue. Hence, from the Japanese sentence, how do we know the speaker is talking about himself? Might he not mean ‘Call Ishmael’ in the sense of ‘give Ishmael a call’? Or maybe he is introducing us to someone and he is saying ‘Call him Ishmael?’ How do we know that he, Ishmael, is refering to himself? How do we figure out the value of the ∅ pronoun in “Call ∅ ‘Ishmael‘”? Simple: in Japanese, if the clause is a statement and we cannot infer its subject from context then, by default, this implicit subject is ∅ = ‘I’; for example:
implicit pronoun
oso-i
{∅ ga} oso-i
It is ∅ who is late
ima iki-masu
{∅ ga} ima iki-masu
It is ∅ who will go now
ima tabe-ru
{∅ ga} ima tabe-ru
It is ∅ who eats now
explicit pronoun
{watashi ga} oso-i
It is I who is late
{watashi ga} ima iki-masu
It is I who will go now
{watashi ga} ima tabe-ru
It is I who eats now
zero pronoun
∅ = watashi
∅ = I
∅ = watashi
∅ = I
∅ = watashi
∅ = I
If the clause is a question to someone else, or if it is an imperative, i.e., a request or order, then the implicit subject defaults to ∅ = ‘you’:
implicit pronoun
oso-i?
{∅ ga} oso-i?
you are late?
ima iki-masu?
{∅ ga} ima iki-masu?
you go now?
ima tabe-ru?
{∅ ga} ima tabe-ru?
you eat now?
ima itte
{∅ ga} ima itte
you go now
ima tabete
{∅ ga} ima tabete
you eat now
explicit pronoun
anata ga oso-i?
Is it you who is late?
anata ga ima iki-masu?
Is it you who goes now?
{anata ga} ima tabe-ru?
Is it you who eats now?
zero pronoun
∅ = anata
∅ = you
∅ = anata
∅ = you
∅ = anata
∅ = you
In the example we used the word ‘oso-i’, which means ‘late’; we are writing it with a dash that separates the root of the word, ‘oso’, from the suffix ‘-i’; this suffix ‘-i’ replaces any present conjugation of the verb ‘to be’, i.e., ‘-i’ can mean ‘am’, ‘is’, or ‘are’. Hence, ‘oso-i’ can mean ‘am-late’, ‘is-late’ or ‘are-late’.
In Japanese, the subject is always the noun, noun phrase, or pronoun that precedes the particle ‘ga’; the subject and ‘ga’ form a phrase, i.e., a subject phrase, that we cannot break, so we indicate the subject and its marker in curly braces, as a single group, i.e., ‘{subject ga}’.
As the examples above show us, it is not the case that the Japanese pronouns have simply vanished and we no longer know who we are talking about, but simply that one of two things happen: either the subject defaults to ‘I’ or ‘you’, or that the subject has already been introduced and we know it from the context of the conversation.
infering ∅ by default and context
Let’s think of an example where the implicit ∅ pronoun, which is also the ∅ subject, changes depending on context.
Suppose John is going on a date with Mary, he looks at his watch, and realizes he is late. Then he says to himself ‘oso-i’, which explicitly is ‘{∅ ga} oso-i’; since this is a statement and the subject is implicit, then the ∅ pronoun defaults to ‘I’ and the explicit sentence is ‘{watashi ga} oso-i’, loosely translated as ‘{It is I who} is-late’; in English John would have said to himself “I am late”.
As John hurries to the meeting place of the date, he runs across his friend Bob. John is panting and Bob figures out John is late, probably to a date with his girlfriend Mary. Bob looks at John as asks ‘oso-i?’, which explicitly is ‘{∅ ga} oso-i?’; since this is a question and the subject is implicit, then the ∅ pronoun defaults to ‘you’ and the explicit sentence is ‘{anata ga} oso-i?’, loosely translated as ‘{Are you the one who} is-late?’; in English, Bob would have asked John “Are you late?”.
Eventually, John arrives, out of breath, to the meeting place; Mary looks at him and says ‘oso-i’. Well… this is exactly the sentence that John told himself earlier, with an implicit subject, i.e., ‘{∅ ga} oso-i’. However, in this case the subject is not the default ‘I’ in spite of being a statement with an implicit subject; certainly, Mary is not going to say ‘I am late’ to herself when she is the one that has been waiting. The difference is that both of John and Mary can infer the subject from the context: both of them know that the one that is late is John, and therefore, from context, they can both infer that the implicit subject represented by the ∅ pronoun has to be ‘you’, i.e., the explicit version of Mary’s ‘oso-i’ is ‘{anata ga} oso-i’ or, ‘{It is you who} is-late’; in English she would have said ‘You are late.’
Let’s take notice of how different English and Japanese are: English favors making the subjects explicit, while Japanese favors making the subjects implicit; in English we seldom use the ∅ pronoun as a subject, while in Japanese we use it often:
Japanese
implicit Japanese
explicit Japanese
explicit translation
English
John
osoi
{∅ ga} oso-i
{watashi ga} oso-i
{It is I who} is-late
I am late
Bob
osoi?
{∅ ga} oso-i?
{anata ga} oso-i?
{It is you who} is-late?
Are you late?
Mary
osoi
{∅ ga} oso-i
{anata ga} oso-i
{It is you who} is-late
You are late
As a caveat, the above labeled ‘Japanese’ and ‘English’ sound natural in Japanese and English, respectively. However, all the other ones do not, but to improve our translation we need to go over the ‘ga’ and ‘wa’ particles first. For now, we are focusing on the different ways in which we treat pronouns in English and Japanese; an English speaker uses explicit pronouns as subjects – ‘I am late’, ‘You are late’ – as naturally as a Japanese speaker uses implicit ones – ‘{∅ ga} oso-i’. Likewise, given the dating scenario we just described, to an English speaker, the sentence ‘<am/is/are> late’, by itself, is confusing and incomplete, while to a Japanese speaker, the sentence ‘watashi ga oso-i’ is unnecesarily redundant and self-centered.
Summary
Given how accustomed we are to explicit pronouns as subjects in English, the ability to fathom a subject represented with an implicit pronoun in Japanese appears supernatural, but that isn’t actually the case: