Object Marking: が VS を
In previous lessons, we learned that the particle が marks the subject, that the particle を marks the direct object, but that が may also mark the object of stative-transitive predicates which manifest in Japanese as no longer being transitive, therefore が marks what still semantically fills the role of the object. Then, when we learned about the auxiliary ～たい, we learned that this logic stems from a phenomenon known as ergativity, in which objects are treated as the subjects of intransitive verbs, with this only occurring with certain predicate types in Japanese, all of which exhibit this issue between が vs を that must be sorted out.
The purpose of this lesson will be to reexamine the differences between が・を from other angles, and in doing so, we will better understand how these particles function.
Established Interchangeability
Before we dive into this lesson’s analysis, let’s recap the instances we have learned が・を are absolutely NOT interchangeable and when they are interchangeable.
■NOT Interchangeable: In Typical Intransitive/Transitive Predicates
Typical intransitive predicates do not possess objects, and so the particle が cannot be interpreted as an object marker in said clauses. In turn, with a standard transitive predicate, が marks the subject and を marks the object.
1. 災害が起きた。(w/ Intransitive Predicate)
A natural disaster occurred.
2. 神々が災害を起こした。
The gods caused a natural disaster.
■Limited to Full-Fledged Interchangeability
In ergative-like stative-transitive predicates, we come across the notion that が is the most linguistically pure choice for the object marker wherever its use appears, and that it is a growing consequence of said predicates exhibiting heightened control on the part of their agents that allows for を to appear in those contexts, despite not having appeared in those contexts traditionally.
Incidentally, in modern speech, this has resulted in a spectrum of naturalness between が and を, with agency playing a major role. So far, we have delved into the following scenarios:
①With the Potential
In many potential statements, が・を may be interchanged with no apparent change in nuance in the simplest of statements in the spoken language.
3a. 韓国語が話せます。
3b. 韓国語を話せます。
I can speak Korean.
If one were to ponder how 3a and 3b would differ in nuance, the particle を would be akin to the simple exercise of freewill, whereas the particle が would highlight the object of capability as being a natural association with the ability manifesting.
The more typical an action is when the potential form is used, the more likely を is seen. Likewise, when the object’s importance in the situation cannot be ignored and places a quintessential role in the state at hand, が is used, to the point that を becomes unnatural.
4. 美恵子は自然に美しい歌詞｛が 〇・を △｝書けた。
Mieko was naturally able to write beautiful lyrics.
5. ホームラン｛が 〇・を △｝打てた。
I was able to hit a home run.
6. この漢字｛が ◎・を 〇｝書ける人はあまりいないでしょうね。
There probably isn’t a lot of people who can write this Kanji.
②With the Auxiliary ～たい
The same dynamic is found with the auxiliary ～たい, and in modern speech, there is a 50% likelihood of either が・を appearing as the object marker. Aside from が being preferred to highlight the object and を appearing with typical statements of desired action, が is preferred in simpler predicates, whereas を is found most frequently when other things are interjected between the object and the predicate.
7. SFが読みたい！
I want to read science fiction!
8. 自分のやることすべてを完璧にやりたいから。
Because I want to do everything I do perfectly.
9. 私はポケットモンスター新作｛が・を｝買いたいです。
I want to buy the latest Pokémon game.
③With 好きだ & 嫌いだ
These adjectival nouns derive from transitive verbs, similarly to how ～たい also ultimately derive from transitive verb forms—好く and 嫌う respectively. This provides a basis for how their predicates became ergative-like predicates which correspond to stative-transitive predicates in English.
Currently, the use of を is not quite accepted as other instances of it being interchangeable with が, only appearing ironically in simpler predicates void of other complex sentence structures.
10. 私は自分のこと｛が ◎・を △｝好きになった。
I’ve gotten to like myself.
④With 分かる
～を分かる has traditionally been incorrect, being the poster child for stative-transitive predicates manifesting as intransitive verbs in Japanese grammar, but the use of を marking the object with it has been observed since the late 1800s, and it is here to stay. The trigger for why を is being acceptable is a change in perceptible control implied by 分かる over an emotional circumstance.
11. 誰もあたしの気持ちを分かるはずなんてない！
There’s no way anyone understands my feelings!
⑤With 欲しい
In a rather limited capacity, we have also seen how ～をほしい is becoming possible, at least in very colloquial speech. The reasoning for the appearance of を here can be explained by what has already been occurring to ～たい.
12. ちゅーる{をほしくて △・がほしくて 〇・をほしがって ◎}待ち構えてるのに無反応な猫も超可愛い！
Cats with no reaction even though they’re waiting so much for that churu (kind of cat food treat) are so cute!
Sentence Note: Ex. 12 exhibits other relatively newer grammatical phenomena as well, most notably the use of な instead of の after 無反応.
⑥With ～やすい & ～にくい
Although both these endings are adjectival, when attached to transitive verbs, we find that を appears regardless of whether the agent is willfully acting or not as the subject-transitive predicate still has an object.
13. お腹を壊しやすい普段食べているものと腸内細菌の関係を知りたいです。
I want to know the relationship between the foods that I normally eat which often upset my stomach and my gut bacteria.
14. 体が柔らかいと怪我をしにくい。
You’re less likely to get injured if your body is flexible.
Particle Distribution w/ The Potential
Japanese phraseology generally frowns upon the same particle showing up more than once in a single clause, and this tendency becomes outright ungrammatical if the particle is being used to serve the same function.
In the context of a potential sentence, in which the object is marked by が by default, the subject, in theory, ought to be capable of being marked by が out of grammatical necessity. In practice, this is not what ends up happening. This then begs question: what can naturally follow the subject here? Also, how do those options pair up with the option between が and を for the object marker?
To get down to the bottom of things, let’s posit that there are four types of outcomes regarding particle choice with potential verb predicates. These types manifest as follows:
15a. 百合さんが英語が話せます。X (Type 1a)
15b. 百合さんは英語が話せます (Type 1b)
15c. 百合さんは英語を話せます。(Type 2a)
15d. 百合さんが英語を話せます。(Type 2b)
15e. 百合さんに英語が話せます。△ (Type 3)
15f. 百合さんに英語を話せます。X (Type 4)
The archetypal structure seen in 15a (XがYがZ) in which Z is a potential verb predicate1 is ungrammatical when spoken; however, it is nonetheless the underlining (deep sentence) structure for Type 1b. As we know, the particle は, when following a subject, prevents the subject marker が from manifesting.
Aside from が, of course, the particles that do overtly follow the grammatical subject varies between が, は, に, には, and にも in potential sentences, but to provide full coverage of its distribution, we must also keep in mind what is marking the object: が or を.
The chart illustrates the percentage each combination of case markings found for potential verb predicates as can be found in natural speech. Combinations that are marked as being 0-1% are indicative of ungrammatical statements to native speakers.
| Particle w/ Subject | Object Marker が | Object Marker を |
| が | ≧0% | 50% |
| は | 25% | 49% |
| に | 8% | 0% |
| には | 51% | ≧1% |
| にも | 16% | ≧0% |
Chart Note: These statistics reflect neither the subject nor the object being dropped from a potential verb predicate as well as the sentence pattern [X (subject) が Y (object) が Z] in relative isolation. As such, when the object is dropped, the likelihood of が marking the subject is just as likely as when an overt object is marked with を.
From these statistics, we also see that on seldom occasion, [subject + に] occurs, albeit usually with は or も following. However, because this use of に derives from its role of marking the agent (doer) in passive as well as spontaneity grammar, and because the potential form derives from said passive grammar, the object marker does not manifest as を here when に takes on this role.
16a. 俺は涙を流せない。◎
16b. 俺には涙を流せない。△・X
I can’t shed a tear.
17. [人間には涙が流せる]し、[自分は涙を流せない]が、[涙を流す理由は理解できた]。
Man can cry, and so though I myself cannot shed tears, I have now understood the reason for crying.
In the final clause [涙を流す理由は理解できた], we see that the particle after the grammatical direct object is は, but は’s inherently adverbial nature trumps the use of が and を2.
Stative Predicates 状態的述語
The concept of the stative-(transitive) predicate cannot be overlooked. So far, we have seen how が is intrinsically tied to these predicates with various grammar points – the potential and the auxiliary ～たい being exemplary of this.
The biggest take away from the interchangeability seen between が・を is undoubtedly the shift in emphasis. This can be visualized as follows:
18a. 友里さんは英語が話せます。
18b.友里さんは英語を話せます。
Yuri can speak English.
19a. 僕はパンを食べたい。
19b. 僕はパンが食べたい。
I want to eat bread.
20a. その文型が説明できません。
20b. その文型を説明できません。
I can’t explain that grammar structure.
21. 北朝鮮はいつでもミサイルを発射できる状態だ。
North Korea is in the state of being able to fire a missile whenever.
We will see again this concept of whether the ending is modifying the verb only or the entire phrase itself. Also, you may be wondering why を is used in Ex. 21 even though the subject is clearly in the state of being able to do the action. That is because of the next factor: control.
Ultimate Control 最大限のコントロール
We have touched briefly on how を is associated with typical willpower, but what if control over what the speaker/agent does is truly within the hands of the speaker/agent? In Ex. 18a-b, we find that を is the natural choice.
22a. あいつは他人｛を ◎・が △/X｝偽れる男だ。
22b. あいつは自分｛を ◎・が △/X｝偽れる男だ。
22a. That guy is a man who can deceive others.
22b. That guy is a man who can deceive himself.
With を, the focus is on the action, and just from hearing the verb 偽る, how the agent goes about deceiving people is immediately on the mind. Deception is also inflammatory, meaning if the particle が were to be used before it, it would be hard to not view it as the subject marker. If, though, that ambiguity were to be alleviated in context, that would at least elevate the naturalness of が to a degree.
23. 俺に自分が偽れるだろうか？
Do I have it in me to be able to deceive myself?
～ようになる・～ようにする
～ようになる is inherently intransitive. When the subordinate clause that modifies ～よう is also intransitive, を marking the object with potential verbs that may happen to compose said subordinate clause becomes far less likely.
24. 毎日やっているうちに[自然に漢字｛が ◎・を ?｝書けるように]なりました。
I became able to naturally write Kanji while I was studying every day.
25. 乃理子は独学で[ハングル｛が 〇・を ?｝書けるように]なった。
Noriko became able to write Hangul through self-study.
That does not mean を is not possible. So long as the object does not bear the majority of the emphasis, を will at best sound colloquial.
26. 文法だけをひたすら勉強してても、フランス語を話せるようになるわけじゃないですよ。
Even if you’re wholeheartedly studying grammar alone, that doesn’t mean you’ll become able to speak French.
In contrast, ～ようにする is transitive, and we find that because する is the main verb of the overall sentence, [potential verb + よう]に truly behaves as an adverbial clause modifying する. Meaning, する’s direct object must be found elsewhere, and that direct object, must be marked by を. That direct object may happen to be placed right before the [potential verb + よう] clause, or it could appear earlier in the sentence. That much is irrelevant as to why the direct object must take を.
27. 4年生が終わるまでに常用漢字｛が X・を 〇｝[書けるように]します。
I will have (the students) able to write the Joyo Kanji by the time the fourth year students end.
28. 乃理子はハングル｛が X・を 〇｝独学で[読めるように]した。
Noriko had herself able to read Hangul through self-study.
The necessity for the transitive verb する to have its direct object marked by を is not obscured whenever する happens to be used in another complex structure. In Exs. 29-30, the presence of a potential verb in the clause containing ～ように is irrelevant as to why を is present.
29. 運転中は、絶対にスマホ｛が X・を 〇｝使えないようにしなくてはいけない。
You must make it that you can’t use your smartphone ever while you drive.
30. 引っ越しの前に、不要なもの｛が X・を 〇｝捨てられるようにしておけ。
Before moving, have it that you are able to throw away unneeded things.
Agency 動作主性
Control is directly tied to a sense of agency. An agent is the grammatical entity which carries out an action. In non-passive constructs, the agent is synonymous with the subject. When the agent has no control over an outcome at all, we find that を becomes unnatural. Conversely, when freewill is exercised and the object of desire/potential is not dwelled upon, を becomes the most natural choice – although が cannot be ruled out due to aforementioned etymological reasons.
31. 中国に長く住んでいて、英語を話す機会がほとんどなくなったので、英語｛が 〇・を ??｝話せなくなった。
I lived in China for a long time, and because I basically lost my chances to speak English, I became unable to speak English.
32. いつでも野良猫｛を 〇・が ？｝殺せるわけではない。
You aren’t able to just kill the stray cats whenever.
33. 思ったより早くエッセイを書けてよかった。
I’m so glad I was able to write the essay earlier than I thought.
34. 冬の間はあの洞窟にある氷柱を溶かせない。
You can’t melt the icicles in that cave during the winter.
35. 量が多すぎて、ビール｛を 〇・が ？｝｛飲み干せなかった・飲み切れなかった｝。
I was unable to {douse down/completely drink} all of the beer because there was just too much.
In Ex. 36a, [potential verb + よう] is a part of a larger subordinate clause preventing the subject of that clause, 山口さん, from being topicalized with は. This alleviates the unnaturalness of [subject + が object + が] somewhat.
36a. [山口さんがフランス語｛を 〇・が △｝話せる]ように、私はフランス人も招待しました。
So that Yamaguchi-san could speak French, I also invited French people.
Another possible route to alleviate this issue is by adding adverbial phrases to the clause containing the subject and object of the potential verb. Here, we see that this in tandem with both being in a subordinate clause makes [subject + が object + が] valid, but this ought not be extrapolated to independent clauses.
36b. 山口さんが、フランス語｛が 〇・を ◎｝好きなだけに話せる場を作るために、私はフランス人も招待しました。
When a supplementary verb that has a high level of volition behind it is used in combination with the main verb but the auxiliary ～たい still concludes the predicate, we find that the supplementary verb has just enough autonomy to cause the ergative-like nature of ～たい to not prevail in the sentence, allowing for を to appear in the previous clause ending in ～て.
37. [神経科学試験の前に韓国語の宿題｛を 〇・が X｝して]しまいたい。
I want to get my Korean homework over before my neurology exam.
38. [子供たちが来る前に、テーブル｛を 〇・が X｝片づけて]おきたい。
I want to have the table cleared off before the children come.
Confusing Intransitive & Potential Forms Together
There are instances in which the potential form of a transitive form looks just like the intransitive form in a transitive-intransitive pair. This renders Ex. 35 ambiguous.
39a. パンが焼ける。
For bread to bake.
I can bake bread.
This ambiguity is often left unnoticed depending on whether an agent is implied. Here, the difference would be negligible because for bread to bake, there must be an agent baking it and the ability to make it happen. Potential verbs also tend to be spontaneous in nature just like intransitive verbs. In their case, the agent can so happen to perform the action, whereas with intransitive verbs, spontaneity implies that the predicate just happens to occur. In either case, が is the best fit.
Of course, 39b is also possible but without this happenstance ambiguity. However, the use of を greatly diminishes the focus on bread and shifts it to being able to get the act of “baking” done.
39b. パンを焼ける。
I can bake bread.
Spontaneity 自発性
There is a subset of verbs in Japanese known as spontaneous verbs (自発動詞) in which the event/action described naturally happens on its own. All such verbs have transitive forms, with the spontaneous form being the most derived. These verbs, unsurprisingly are all viewed as intransitive with the focus of attention being marked with が.
This sense of spontaneity (自発性) bleeds into how が may be chosen over を whenever there is a total lack of control over the action. Of course, in actual practice, so long as no other critical grammatical restraint is present, the particle を may be seen used nonetheless, with spontaneity being beside the point.
40. セス先生は日本語の天才だから、次から次にとめどなく新しいレッスン｛が・を ?/△｝作れたんだ。
Because Seth Sensei is a genius at Japanese, he was able create new lessons nonstop one after the other.
41. 俺はお前｛が 〇・を 〇 ?/△｝好きで好きでたまらねーんだよ。
I like you so, so much I can’t stand it.
42. 僕はきのうから豚骨スープ｛が・を ?/△｝食べたくて、食べたくて仕方なかったので、あの屋台に行ってみた。
I wanted to eat tonkotsu soup so bad since yesterday, and so I tried that stand over there out.
Lines drawn in grey zones such as these truly rests in the individual speaker: how does each person internalizes spontaneity versus control and how both factors interrelate with potential （可能）and desire（願望）.
Control ≈ Will
Though this connection has already been touched on to a degree, it goes without saying that when one has control over something, one usually willfully exercises said control.
Because of this sense of control, を is never possible with true spontaneous verbs. Meaning, ～を聞こえる or ～を見える are not grammatical; not only do these verbs express completely spontaneous actions, they fundamentally lack willpower (有意志性) in both a literal and a grammatical sense.
43. 山が見える。
I can see the mountain(s).
44. 音（が）聞こえてない？
Can’t you hear a sound?
Grey zones such as those seen in Exs. 40-42, although touching on the idea of spontaneity, show where separate semantic domains overlap. When no overlap exists and we are looking at the extreme case of a phenomenon being true, we cannot expect interchangeability that is only found in said overlap to hold true for that extremity.
The Ungrammaticality of ～をできる
This point of spontaneity is being brought up so heavily because of the tendency of mainstream textbooks to introduce 聞こえる and 見える as potential verbs, when in fact, they are not.
What, then, about 出来る? 出来る is often described as the potential form of する. This relationship, however, is a relatively recent development. Its primary means are still, in fact, describing something that used to not exist coming into existence. The idea of potential was borne from the idea of one’s ability to do something innately coming into being.
Meaning, 出来る has a naturally higher degree of spontaneity than other potential verbs, resulting in ～を出来る still being ungrammatical – although instances are starting to appear in highly colloquial speech.
45. 知也君は英語｛が 〇・を X}出来る。
Tomoya can speak English.
46. 知也君は英語｛が 〇・を X}出来るようになった。
Tomoya became able to speak English.
We saw earlier how ～ようになる and ～ようにする influence the interchangeability between が and を by how phrase boundaries function.
47. 知也君は英語｛が 〇・を X｝出来るようになりたいと思っている。
Tomoya wants to become able to speak English.
48. 知也君は英語｛が ◎・を 〇｝出来るようにしたいと思っている。
Tomoya wants to have himself able to speak English.
Yet, in Ex. 48, we find that が is far more common than を, which can be explained by 出来る’s high level of spontaneity. Using ～ようにする with it is not problematic, as the speaker is expressing the want to make themselves get to that level of capacity.
Next, let’s see examples of ～ておく used in conjunction with 出来る.
49. これまでに出来なかった問題｛が・を｝できるようにしておくことを願います。
We ask that you prepare yourself to be able to answer the problems that you haven’t been able to up to now.
50. 休み時間になったら、噂話｛が・を｝できるようにしておいた。
We had that we cold gossip once we were in free time.
51. 学校で使っている問題集の問題が出来るようにしておくこと。
Have it that you can do the problems in the problem set used at school.
52. お金の管理｛が・を｝できるようにしておくことが、自分の生活を守る上でも大切になってくる。
Being able to manage one’s money before will become even more important than protecting one’s way of life.
53. みだりに青年が埋立地に立ち入るのを防止すること｛が 〇・を X｝できるようにしておくこと。
Have it that you can prevent youths from trespassing recklessly into the land reclamation site.
Under normal circumstances, we see that が and を appear interchangeable, but the internalized clause boundaries will still differ. Whereas ～ができるよう functions as a subordinate clause, [～を][できるように～] constitutes as two separate clauses. Even so, ～が出来る will still usually prevail.
As for Ex. 53, we see that under no condition does ～することができる, an extended means of creating the potential form for する verbs, get altered to allow for を to appear inside it.
Speaking of する verbs, when 出来る creates their potential form, が・を interchangeability may be observed with transitive verbs so long as they are before the noun composing the する verb. This is allowed because できる is rendered as a supplementary verb, which means it loses its literal interpretation to simply serve the grammatical role of marking potential.
54. 汚染｛が・を｝防止できる。
(We) can prevent pollution.
When a する verb is in isolation, particles may be interjected in between the noun and する, but when turned into the potential, only が can mark the object. This is because 出来る regains its autonomy.
55. これで再発防止｛が 〇・を X｝できるのか？
Can (we) prevent it from happening again with this?
Returning to 見える・聞こえる briefly, the phenomenon seen with ～ておく and 出来る holds true for them. Either が or を may be used, but the clause boundaries will still differ.
56. 通学路の近くでは、どこからでも交通標識｛が・を｝見えるようにしておきなさい。
Please make sure that the traffic signs are visible from any direction near the school zone.
W/ 分かる
Traditionally, 分かる is viewed as an intransitive verb with all the hallmarks of being a stative predicate. Because its object is marked with が, it can also be viewed as being an ergative-like predicate.
The use of ～を分かる, though, has been on the rise, but for it to be made grammatical, a bending of its definition had to take place. When we see examples of this, the intended meaning is instead “emotionally understanding someone out of one’s own volition,” not intrinsically understanding something.
Through this redefining, we find 分かる used with a range of endings which pertain to one’s volition that it traditionally had no access to. Because of this redefining, though, 分かる is no longer being used as an intransitive verb, making が ironically ungrammatical.
57. 相手が自分のことを分かってくれないと気にするよりも、自分が相手のことを分かろうとしないことを気にしなさい。
Rather than worry about the other person not understanding you, worry about you yourself not trying to understand the other person.
58. 俺の気持ち｛を 〇・が X｝分かってくれ！
Understand my feelings!
参照
「～が～できる」と「～を～できる」について by 田村泰男.
- While [X (subject) が Y (object) が Z] is possible for various predicate types, even for others featured in this lesson such as stative-transitive predicates, the pattern does not manifest with potential verb predicates due to the ergative-like nature of the potential affix(es) themselves. Ergo, the intransitive-like nature of the grammar structure is incompatibility with the exclusivity nuance of the exhaustive focus marking function of X which is inherent to the aforementioned sentence structure, which would direct focus onto the agent; yet, for potential grammar in the traditional sense, innate ability is spontaneous in nature and not reflective of a willful agent exercising said ability. If that were to be the case, を would be invoked as the object marker, at which point が could resume its focus marker function. ↩︎
- Either particle would still have to be in the underlining, deep structure of the potential verb clauses. ↩︎