I don’t suppose many of you remember the “Question Man” routine on the old Steve Allen show. Steve would come out with a handful of cards containing “answers,” which he would read aloud, and then, from the depths of his wisdom, he would tell us what questions these were the answers to. For example:
Answer: Go West.
Question: What do wabbits do when they get tired of wunning awound?
Oh, well. The funniest thing about the Question Man was not so much the routine itself as when Steve was so tickled by a joke that he couldn’t stop cackling. The producers of “Jeopardy” have effectively circumvented this problem.
Which brings us to the eternal mystery of wa and ga. If the Japanese are going to insist on using a postposition (or particle) to mark the subjects of their sentences, why can’t they make up their minds and choose one instead of switching between two (not to mention occasionally substituting no for ga)? Which is it, finally—Watashi wa ikimashita or Watashi ga ikimashita? Both of them mean “I went,” don’t they? So which one is right?
Well, that depends upon what question the statement is an answer to. (In fact, for a plain, simple “I went,” both would be wrong, but let me get back to that in a minute. Note here, too, that I am ignoring such strictly conversational forms as Watashi, ikimashita.)
The difference between wa and ga depends entirely on context. Neither is automatically “correct” outside of a context, any more than “a dog” is more correct than “the dog.” Their use depends entirely upon what the author assumes you know already and what he feels you need to know. They function primarily as indicators of emphasis. If at any point in your reading you are unsure where the emphasis lies, one of the best things you can do is ask yourself, “What question is this sentence the answer to?” In the case of Watashi wa ikimashita and Watashi ga ikimashita, each is the answer to a question. But let’s not forget the sentence lkimashita, either. In figuring out what the implied questions are, this could help you in both interpreting texts and deciding which form to use in speech.
1. Ikimashita. “I went.”
2. Watashi wa ikimashita. “Me? I went.”
3. Watashi ga ikimashita. “I went.”
1. Dō shimashita ka. “What did you do?” Or: Ikimashita ka. “Did you go?”
2. Soshite, Yamamura-san wa? Dō shimashita ka.” And now you, Mr. Yamamura. What did you do?”
3. Dare ga ikimashita ka. “Who went?”
I’ve included number 1 here because that is the way to say “I went” in the most neutral, unemphatic way, emphasizing neither who went nor what the person did. That’s why I said above that for a plain, simple “I went,” both the other forms would be wrong, because it is precisely to add emphasis that they would be employed. When we say “I went” in English, we’re assuming that the listener knows who the “I” is. And when we assume that our Japanese listener knows who did the verb, we just say nothing for the subject. Speakers of English are so used to stating their subjects that it takes a lot of practice for them to stop using either form 2 or 3, but perhaps becoming more aware of what they are actually saying could help break them of the habit.
Wa is a problem for English speakers because it is doing two things at once. It differentiates the subject under discussion—or, rather, the “topic” (more later)—from other possible topics, and then it throws the emphasis onto what the sentence has to say about the topic. Let’s deal with the first function first.
Early on, we are usually given “as for” as the closest English equivalent to wa, which it indeed is, but after en-countering wa several thousand times and mechanically equating it with “as for,” we forget the special effect that “as for” has in English, and it simply becomes a crutch for translating Japanese into a quaintly Oriental version of English before turning it into real English. Watashi wa ikimashita = “As for me, I went” “I went.” The last equation in this sequence is wrong.
Sure, we have the expression “as for” in English, but sane people use it much more sparingly than do students of Japanese. Take Patrick Henry, for example: “I know not what course others may take, but as for me, give me liberty or give me death!” Now, there’s a man who knew his as-fors!
The next time you are tempted to say Watashi wa ikimashita, stop and think about whether you really want to proclaim to the world, “I know not what course others may have taken, but as for me, I went!” Your wa differentiates you as a topic of discussion from other possible topics (“I don’t know about those other guys, but as far as I am concerned . . .”) and then, after building up this rhetorical head of steam, it blows it all into the rest of the sentence (“Yes, I did it, I went!”). Notice that wa builds suspense, arousing curiosity in the reader or listener about what is to come. If the speaker were to pause at the wa, the listener’s brain would whisper subliminally, “Yes, yes, and then what?” After having differentiated the named topic from implied other potential topics, wa dumps its emphatic load on what comes after it. This makes it very different from ga, which emphasizes what comes before it.
Have you ever stopped to think about why you were taught never to use wa after interrogative words such as dare, nani, and dore? Because ga puts the emphasis on what immediately precedes it, and when you use those interrogative (question-asking) words, they are precisely what you want to know: “Who went?” “What came out of the cave?” “Which one will kill it most effectively?” And just as ga points at exactly what you want to know in the question, ga will always be used in the answer to emphasize the information that is being asked for: Dare ga ikimashita ka / “Who went?” Watashi ga ikimashita / “I went” or Yamamoto-san ga ikimashita / “Miss Yamamoto went.” This is why you don’t want to say Watashi ga ikimashita for a simple “I went,” because what you are really saying is “I went,” to which the proper response is “OK, OK, calm down.”
Notice how the same information can be requested either before ga or after wa: Dare ga ikimashita ka / “Who went?” or Itta no wa dare desu ka / “Who is it that went?” To both of these, the ga-marked answer will be Yamamoto-san ga ikimashita / “Miss Yamamoto went” (she seems to get around a lot).
It is because ga emphasizes the word before it that this subject marker is frequently softened in modifying clauses by replacing it with no, a modifying particle that throws your attention ahead. Shimizu-san no hirotta saifu wa koko ni arimasu / “The wallet that Mr. Shimizu found is here.” Ga can be retained, however, if we want to emphasize the subject: Shimizu-san ga hirotta . . .” gives us “The wallet that Mr. Shimizu found is in here.”
Unless we see the direction in which ga focuses our attention, a Japanese sentence can seem to be belaboring the obvious. Take the definition of “crucifixion” from the Encyclopedia Japonica, for example. After pointing out that the punishment had long been practiced among the Jews, Greeks, and Romans, it goes on, Omo ni Kirisuto-kyō no hakugai ni mochiirare, Iesu Kirisuto no haritsuke ga yūmei de aru, which, without due care given to the ga could be interpreted, “Primarily used in the persecution of Christianity, and the crucifixion of Jesus Christ is famous.” The ga indicates, however, that the point is not that Christ’s crucifixion was famous; rather that the crucifixion of Jesus Christ was famous among crucifixions. Hence, “Primarily used in the persecution of Christianity, the cru-cifixion of Jesus Christ being the best known example.”
Students sometimes get the impression that wa appears in negative sentences and ga in positive. This is simply false. There is a strong tendency for wa to appear in neg-ative sentences, but that is because wa is being used in these cases to do what it always does, and that is to throw the emphasis onto what comes after it—that you’re not going, that it isn’t the one you want, that there aren’t any left, etc. Compare Ikitaku nai / “I don’t want to go” with Ikitaku wa nai / “I don’t want to go (though I might like to hear how it was).” But look at how the wa does exactly the same thing in positive constructions: Nihon-jin ni mo fuman wa aru no da / “Japanese people do have their dis-contents, too.”‘ Mayaku wa tashika ni kutsu o kanwa shi wa shita ga, sono kawari ni kimy5 na genkaku o mo-tarashita / “The narcotics did ease the pain, but they also gave rise to strange hallucinations.”‘ (To emphasize the dif-ferentiating function of wa, we might more wordily para-phrase Ikitaku wa nai like this: “As far as wanting to go is concerned (in distinction to other possible reactions to this situation), I don’t. Fuman wa am can be paraphrased, “As for discontents (in distinction to other sorts of feelings), they exist.”) Our verb “to do” can be another handy tool for con-veying in English translation some of the emphasis that a wa often throws on the verb. Compare Okane ga aru / “I have money” with Okane wa am / “I do have money (but I don’t have time to spend it, or I owe it all to the gov-ernment, or some such implication owing to wa’s usual differentiating function).” The whole question of emphasis in language is involved with the question of what is known information and what is new information. There is no need to accentuate the obvious. It is for this reason that there are often correspondences between wa and ga in Japanese and “the” and “a” in English. “The man” (Otoko wa . . .) is someone we know about and are now going to get new information on, whereas “a man” is someone new who has just entered the scene (. . . otoko ga haitte kita). (That is why “the” is called the “definite article”: we know just what we are re-ferring to, while we use “a”, the “indefinite article”, when we’re not so sure.)
In his encyclopedic Japanese Language Patterns, Alfonso has noted these correspondences and wisely chosen not to dwell on them. The fact remains, however, that there is a good deal of overlap in linguistic function between Japanese wa and ga and English “the” and “a”. Since both have to do with unspoken assumptions concerning how much speaker and listener know, both convey some of the subtlest nuances of their respective languages, and both are extremely difficult for foreigners. Even the most accomplished Japanese speaker of English will continue to make mistakes with “the” and “a”, and native users of English will probably always have some degree of difficulty with wa and ga. This is surely one of those intuitive areas of language that can only be fully mastered in early childhood.
In the days of his youth (though well past his childhood), a sharp-tongued colleague of mine once had a serious falling-out with his Japanese employer over “the” and “a”. He was working in Japan as a translator at the time, and his boss suggested that they were paying him too much because English was so full of these useless little definite and indefinite articles. Since he was being paid by the word, the employer suggested they ought to omit all the the’s and a’s from the word count. The prospect of a pay cut did not set well with my colleague, who somewhat impetuously replied, “Better yet, you do the translations, and you can pay me to put in the the’s and a’s.” For this impolitic thrust at one of the most insecure areas of Japanese knowledge of English, he was fired on the spot.
Ga, we can fairly safely conclude, is a lot simpler than the double-functioning wa. Ga marks the grammatical subject of an upcoming verb or adjective, but wa marks the topic—not the topic of a verb, but the topic of an upcoming discussion. This topic-subject distinction can be more confusing than helpful until you see what a word is the topic of or the subject of. For more on this, pay close attention to the next paragraph.
Ga marks something that is going to have a piece of grammar—a verb or adjective—connected to it, but wa is far less restrictive: it marks something that is going to have a remark made about it, but it gives absolutely no clue as to what kind of remark it’s going to be. Wa merely says, “Hey, I’m going to tell you about this now, so listen.” Ga says “Watch out for the next verb that comes by: I’m most likely the one that will be doing or being that verb.” Ga always marks the subject of a verb or adjective, and if that verb is the main verb, that means ga is marking the subject of the sentence. Wa never does this.
Wait a minute. Did I just say that wa never marks the subject of a sentence? Yes, and I mean it. Wa never ever marks the subject of a verb and so it never marks the subject of a sentence. Wa only marks a topic of discussion, “that about which the speaker is talking.” And, as Anthony Alfonso so sensibly remarks, “Since one might talk about any number of things, the topic might be the subject of the final verb, or time, or the object, or location, etc.”‘
Alfonso gives lots of good examples of each type of topic in a passage that is well worth studying. As a time topic, he gives Aki wa sora ga kirei desu, which can be translated “The sky is clear in autumn” or, more literally, “Autumn, well, the sky is clear,” or “As for autumn (as op-posed to the other seasons), the sky is clear,” etc. One example of an object topic that Alfonso gives is Sono koto wa kyō hajimete kikimashita, “I heard that today for the first time,” or “That matter, well, today for the first time I heard it,” or “As far as that matter goes, I heard about it today for the first time,” etc.
Alfonso’s remark about the possible contents of a topic suggests that a wa topic can be the subject of a sentence, but I am still going to insist that it never is. Let’s expand on those cases in which the wa-marked topic seems to be the thing or person that does the verb. One good example of this is our old Watashi wa ikimashita.
Earlier, I translated Watashi wa ikimashita as “Me? I went.” Doesn’t this look suspiciously like those double subjects your first-grade teacher told you never to use? “My uncle, he’s a nice man.” “My family and me, we went to New Jersey.” “Mistah Kurtz—he dead.” In each case, you name the topic of your upcoming remark, and then you go ahead and say a sentence about it. The subject of the verb in each sentence is not “my uncle,” “my family,” or “Mistah Kurtz” but rather the following pronoun. And notice that all the redundant subjects are pronouns. Once you’ve established that it’s your uncle you are talking about, you can demote him to pronoun status when you give him a sentence to do. Likewise, in Japanese, once you’ve established the topic you are going to be talking about, you can use the Japanese zero pronoun when you give it a verb to perform. And that’s just what is happening in Watashi wa ikimashita.
Our old standby “as for” can help clarify this a bit further. “As for me, [I] went.” The “I” is in brackets here be-cause it is present in the Japanese sentence only as an unspoken subject. Watashi is not the subject of ikimashita and is not the subject of the sentence. It is simply the topic of the upcoming discussion. The wa tells us only that the following discussion is going to be about watashi as opposed to other possible people. The subject of the verb ikimashita is not watashi but the silent pronoun that follows it. In other words, when you used to make up sentences with double subjects in the first grade, you were trying, in your childish wisdom, to use wa constructions in English. You could have mastered wa at the age of seven, but that pigheaded Mrs. Hawkins ruined everything!
Take a second and look back at the example of a wa object from Alfonso, Sono koto wa kyō hajimete kikimashita, “I heard that today for the first time,” or “That matter, well, today for the first time I heard it.” Notice that the actual object of the verb kikimashita is not the wa-topic koto but the zero pronoun, which we have to translate as “it” when we start getting literal.
We cannot repeat too often that wa NEVER marks the subject of a verb. It doesn’t mark the object, either. And it certainly doesn’t unpredictably “substitute” for other particles such as ga and o. All wa ever does is tell you, “I know not about others of this category we’ve been talking about, but as for this one .. .” Wa tells you nothing about how its topic is going to relate to the upcoming information: it only tells you that some information is coming up that will be related somehow to the topic. In fact, the only way that you can tell whether wa marks an apparent subject or object (or anything else) in a sentence is in retrospect. But language doesn’t work in retrospect.
When a grammarian tells you that wa can mark the subject of a sentence, he is able to say that only because he has seen the rest of the sentence and knows how it turned out. But when real, live Japanese people read or hear a wa topic at the beginning of a sentence, they have absolutely no idea what’s coming. Look at Alfonso’s time topic example on the clear autumn sky, Aki wa sora ga kirei desu. The only reason Alfonso was able to use this sentence as an illustration of a time topic is because he had read it to the end and could go back and analyze the relationship of aki to the statement made about it after the wa. When a Japanese person hears or sees Aki wa, though, he has no idea what’s coming (aside from any hints he might have picked up from the larger context). It could be daikirai desu / “Autumn—I hate it!” or ichiban ii kisetsu desu / “Autumn—it’s the best season,” making it in both cases an apparent subject (in Japanese, if not in English translation), not a time expression. It could even be an ap-parent object if the sentence went on tanoshiku sugoshita / “The autumn: we passed it pleasantly” or “(The other seasons aside,) the autumn at least we passed pleasantly.” Whatever its various apparent functions, marking subjects or objects or time expressions or locations, these functions can be labeled only after the fact, as the result of analysis. Again, the trouble with wa is that it always performs its double function: it distinguishes known topics from other topics, and it signals you to look for the important information that is about to be imparted in the upcoming discussion. When it does that, it puts no grammatical restrictions on what those discussions can be.
If you stop and think about it, “as for” works in the same way. After Patrick Henry set up his topic with “as for me,” he had to mention the “me” again to make grammatical sense: “… give me liberty or give me death.” The subject of the main clause here is an understood “you” or “King George” or whoever it is that is supposed to give “me” either liberty or death. And “me” is not even an object: it’s what we call an “indirect object.” The direct objects of “give” are “liberty” and “death.” In other words, “as-for” topics in English are as grammatically flexible as wa topics in Japanese: “As for the men, we paid them and sent them home.” “As for the time, she arrived around two o’clock.” “As for her mothers future, Mary Wang still wonders what lies ahead.”‘ “Madame Boyary, c’est moi.”
Notice how, in the English examples, the degree of distinction that “as for” sets up between the topics it marks and other implied topics is quite variable. The same is true for wa. Depending on the situation, the amount of contrast can vary from quite a lot to nearly none.
Here is a sobering anecdote to illustrate how potent a little wa can be in differentiating a topic from implied others. The topic in question happens to be a time expression, not an apparent sentence subject, but the differentiating function is the same.
I and a few other American scholars were at a party and one of us tried to compliment our Japanese host by saying, Konban wa oishii mono ga takusan arimasu ne. By this he intended to say, “What a lot of tasty dishes you’re serving us tonight.” The host laughed and remarked, You mean I’m usually stingy on other nights?” By putting wa after “tonight,” my colleague had in effect said, “Tonight, for a change, you’re serving us a lot of tasty dishes.” Although our host seemed to take this in good humor, he unobtrusively committed seppuku later as the rest of us were drinking cognac.
On the other hand, as we shall see below, wa can appear to have virtually none of its differentiating or contrastive function when we encounter it at the beginning of a text, especially in fictional narratives.
Whoever first realized, in those early murky meetings of English and Japanese, that wa is like “as for,’ had a brilliant insight. As nearly as I can tell, the credit for that particular phrase should go to Basil Hall Chamberlain, the great nineteenth-century Japanologist to whom so much of our knowledge about Japan and Japanese can be traced. Profiting from some earlier remarks by W. G. Aston that drew parallels between wa and certain Greek and French constructions, Chamberlain went on to note the usefulness of “as for” perhaps as early as 1888. The only problem with “as for” nowadays, as I mentioned earlier, is that we tend to stop interpreting it properly in English when we encounter so many wa‘s in Japanese. Understood correctly, “as for” is an excellent device for helping us analyze a Japanese sentence, but when it comes to translating Japanese into real, bearable English, it is usually best disposed of. So much for the general principles of wa and ga. Now let’s look at a famous sentence in which we find both a wa and a ga:
Zō wa hana ga nagai.
As literally as possible, we can render this: “As for elephants, (their) noses [i.e., trunks: the Japanese don’t happen to have a special word for trunk; it’s nothing to laugh about] are long.” That is to say, we first note that our topic is elephants, and concerning this topic we formulate the grammatical construction “trunks are long,” in which “trunks” is the subject and “are long” is the predicate.
So now we have “As for elephants, their trunks are long.” What do we do with it? What does it mean? How do we make it real, live English that someone other than a language student could love? Does it simply mean “Elephants have long trunks?”
Maybe we should look at the Japanese. When would anyone ever really say Zō wa hana ga nagai except to make a point about how odd Japanese is? Isn’t this sentence about elephants really just a red herring? Its only conceivable real-life use is for teaching a small child the distinguishing characteristics of various animals. It would have to come in a list, probably while the speaker was turning the pages of a picture book: Giraffes have long necks, lemurs have big eyes, minks have nice fur, tapirs have huge rumps, and as for elephants, well, they have long noses.
This is not to say there are not genuine Japanese sentences of the Zō wa hana ga nagai pattern. They are, in fact, quite common. Here are a couple more:
Aitsu wa atama ga amari yoku nai nee. / “That guy’s not too bright, is he?”
Oyaji wa atama ga hagete kita. / “The old man’s lost a lot of hair.”
But such sentences don’t exist in a vacuum (except in classrooms and grammar books). There is always a larger context implied. This is true primarily because of the function of wa in differentiating the known topic from other topics and directing the attention of the listener to the important information that follows. “The man? Well, he’s in Washington.” “The woman? She disappeared.” Notice the use of “the” here, implying a certain amount of understanding already established between speaker and listener—a context. You wouldn’t say Otoko wa Washinton ni iru except as the continuation of a discussion that has already established the existence of the man and now imparts more information about him. The same principle is at work in news reports. A story about a new appointment made by the American president may begin, Busshu Bei–Daitōryō wa …, going on the assumption that everyone knows about him and the office he holds. A close equivalent of the Japanese phrase would be “US President Bush …,” which makes the same assumptions about what the reader knows as does “George Bush, THE President of THE United States …” A report on doings in the Diet will start out, Kokkai wa … / “THE Diet …” Where the existence of a less well-known entity must be established, though, we will often find a ga at work: Habado-dai no sotsugyo-ronbun ni ‘Fuji Santaro’ nado Nihon no sarari-iman manga o toriageta Beijin josei Risa Rosefu-san ga, Tokyo no terebi-kyoku de bangumi-seisaku no kenshu-chu da / “Lisa Rosef, an American co-ed who did a study of ‘Fuji Santaro’ and other such salaryman comics for her Harvard graduation thesis, is presently on an internship for program production at a Tokyo television station.”
Another famous grammatical red herring involves eels: Boku wa unagi da. Literally (no, not “literally,” but perversely), this would seem to mean “I am an eel.” But it’s just a sentence that Japanese with some consciousness of their own language like to chuckle over. If Sore wa pen da means “That is a pen” and Are wa kuruma da means “That is a car,” how can Boku wa unagi da not mean “I am an eel”? Before we answer that, it’s important to note that “That is a pen” is not the same as “It’s a pen.” When, aside from some kind of grammar drill in an ESL class, would we actually say, “That is a pen” in English? The customer, pointing through the glass, mistakenly asks to see “this mechanical pencil, please,” and we, the clerk, must point out to her that “That is a pen.” The real a-swer to “What is this I’m holding?” is the non-sentence, “A pen,” or, for those abnormally addicted to speaking in complete sentences, “It’s a pen,” but certainly not “That is a pen.”
Likewise, Sore wa pen da (or desu, since we are polite in the classroom) is mainly an obedient language student’s answer to the teacher’s question Kore wa nan desu ka. A natural answer to the question would be Pen desu. The full Sore wa pen desu means “That one (as opposed to another object the teacher is holding) is a pen.” But notice that, even here, while pen may be the topic of the sentence, it is not the grammatical subject of desu. The subject of desu is, as noted earlier, the unspoken “it”: “As for that, (it) is a pen.” All the wa does is hold up the topic and distinguish it from other possible topics, and then it tells you that the important information on the topic is about to follow. If the context has established that we are talking about long, slender objects or objects that people happen to be holding, the unspoken subject is easily and automatically equated with the thing that sore refers to.
If, however, the context has established that we are talking about what the various individuals in a group want to eat, the slippery unspoken subject can easily adapt to that: “(I know not what others may take for this course, but) as for me, (what I want to eat) is eel.” The topic of Boku wa unagi da is boku, but the subject of the verb da is “what I want to eat.”
The one place where a wa topic might seem to materialize out of a vacuum is the opening sentence of a fictional narrative, but in fact what is going on here is that the wa is being exploited by the author to give the fictive impression of a known context.
Natsume Soseki’s novel Mon (The Gate), for example, starts out, Sosuke wa sakki kara engawa e zabuton o mochidashite …” A reasonably readable translation of this might go: “Sosuke had brought a cushion onto the veranda and …” This looks so unexceptionable both in Japanese and in English that we can easily forget how much literary history lies behind our being able to begin a third-person fictional narrative with the narrator establishing such apparent instantaneous intimacy between himself and his character on the one hand and himself and the reader on the other. A nineteenth-century reader might ask, “Who is this Sosuke fellow? When was he born? Who were his parents? What does he look like? Where does he live? When did this happen? This can’t be the beginning of the story. What happened to the introduction? It seems to start in the middle of things.”
Of course, that is exactly the point. Many modern novels and stories purposely try to give the impression of being direct observations of real life—events and people that existed before the narrator started telling us about them. The effect is even clearer when the first character we encounter doesn’t have a name, as in the opening sentence of Soseki’s earlier novel, Sanshino: Uto-uto to spite me ga sameru to onna wa / “He drifted off, and when he opened his eyes, THE woman …”
Jack London opens The Call of the Wild (1900) with the observation that “Buck did not read the newspapers.” We know better than to ask, “Buck who?” Hemingway’s “Indian Camp” begins, “At the lake shore there was another rowboat drawn up,” and his “Cat in the Rain” starts out, “There were only two Americans stopping at the hotel.” As modem readers, we have learned not to ask “Which lake shore?” or “What hotel?” It’s the hotel, the one we and the narrator know about. We enjoy the impression of journalistic immediacy conveyed by this clipped style. And perhaps we get impatient when Henry James begins the 1880 Portrait of a Lady: “Under certain circumstances there are few hours in life more agreeable than the hour dedicated to the ceremony known as afternoon tea,” etc. etc.
James’ garrulous narrator, who even refers to himself as “I” and tells us that he is “beginning to unfold this simple history,” is but the most subtle permutation of the traditional storyteller, who might inform us that “Once upon a time, in a certain kingdom, there lived a girl with long, golden hair.” The Japanese formula for opening a fairy tale is Mukashi, aru tokoro ni, ojiisan to obaasan ga sunde imashita / “Long ago, in a certain place, there lived an old man and an old woman.”
We can almost hear the storyteller clearing his throat as he stands before us and invites us to imagine the existence of a self-contained, make-believe world inhabited by an old man and an old woman, whose existence must first be established in the form of ga-marked subjects before the tale can unfold. The implied question to which this is the answer is “Who lived in a certain place once upon a time?”
The modern author, by contrast, more often wants to give a strong impression of the pre-existence of the elements in his fictive world rather than calling attention to the voice of the narrator and the mere existence of his characters. In English, he does this with “the,” and in Japanese, wa serves the purpose. Murakami Haruki, for example, begins a 1985 novel, Erebeta wa kiwamete kanman na sokudo de josho o tsuzukete ita / “The elevator continued its ascent at an extremely sluggish pace.” The same thing is going on in the Soseki novel cited earlier: “(I know not about other people, but) as for Sosuke [the one we all know about], he had brought a cushion onto the veranda and …” The implied question behind this opening sentence is “What was Sosuke doing?” Translated into the corresponding English medium, we get nothing more complicated than, “Sosuke had brought a cushion onto the veranda and …” It would be laughable to imagine a modern, introspective novel like Mon starting out “In Tokyo, there lived a man named Sosuke,” which would, of course, have a ga-marked subject in Japanese. The implication of the wa marker is that we know Sosuke—at least as well as we knew President Bush in the news story mentioned above.
First-person narrators will always refer to themselves at the outset with wa since, of course, they do not have to establish their own existences (“Once there was a me”). Indeed, part of what makes such narrators feel so powerfully real and present is their implied existence, diarist-like, outside their texts.
Now, don’t go out and exult over finding a ga-marked subject in the opening sentence of a piece of modern fiction. More than likely it’s the subject of a wa-marked subordinate clause like this: Ueda Toyokichi ga sono furusato o deta no wa ima yori oyoso niju nen bakari mae no koto de atta / “It was some twenty years ago that Ueda Toyokichi left his native village.” Or: Tomimori ga sono onna o roji no yama no waki ni aru ie ni tsurete kita no wa, hachigatsu mo haitte kara no koto datta / “It was already after the beginning of August when Tomimori brought the woman to the house by the ghetto hill.”
All of this business about narrators is meant to illustrate that you do not have to learn a lot of different functions for wa. It is completely consistent in its double function, differentiating the known topic it marks from others and throwing the emphasis on ahead in the sentence to what really matters.