I’ve seen this debate going on among Japanese learning models. Most standard resources claim there are two types of adjectives in Japanese, i-adjectives which are conjugated and na-adjectives which are not, but which receive the な before a noun and a copula at end of a sentence.
Cure Dolly’s model suggests that i-adjectives exist (and are one of the three ways to form a predicate along with verbs and the copula) but that na-adjectives are actually adjectival nouns that have the “special ability” to use な that other nouns don’t.
Going further, some models suggested by linguists suggest that Japanese has **no adjectives at all**, like this one which describes them instead as “descriptive verbs” and “descriptive nouns”:
>形容詞: Descriptive verbs
>
>Descriptive verbs are also called stative verbs (verbs of state) or adjectival verbs by linguists, but most of us know these guys as i-adjectives or even true adjectives. Descriptive verbs are real verbs, although they miss some verbal forms that are available to other verbs. In essence, descriptive verbs are just verbs that come before the noun and form a relative clause. There’s really no difference in syntax between 踊る蝶 (the butterfly who is dancing) to 白い蝶 (the butterfly who is white).
>
>形容動詞: Descriptive nouns
>
>Descriptive nouns are also called adjectival nouns by linguists, but most of us know them as na-adjectives or quasi-adjectives. Despite their name, they are no less adjectives than the “true adjectives”, and some of them can actually be connected to nouns by using の rather than な. Historically, descriptive nouns were either nouns followed by the genitive の or nouns followed by な, which is the attributive (before-noun) form of the copula なり. So a の-adjective like 普通の男 started out as basically just saying “a man of ordinary”, and 静かな場所 (which in classical Japanese meant: 静かである場所) started out as saying “a place which is silent”.
>
>It’s instructive to note that descriptive nouns actually take a completely normal copula instead of な when past or negative forms are used (e.g. 普通だった男, 静かじゃない場所), so these adjectives actually still use a copula to this very day, but it’s most common form is replaced by the special な or の (depending on the adjective).
[https://japanese.stackexchange.com/a/1016/1525](https://japanese.stackexchange.com/a/1016/1525)
In this model, 白い is not an adjective but rather a verb meaning “to be white” or “is in the class of white things”. I find the uniformity of this model pretty nice, since it means every predicate is either a verb or copula, and it explains why i-adjectives don’t get a copula (it’s already “bundled in”), and it also unifies relative clauses and use of i-adjectives as prefixes. It also makes it easier for me to understand certain uses of the て-form like this one from a song:
眩しくて逃げた = it was so dazzling that I ran away / it was dazzling so I ran away
It’s hard to see this as a complete sentence without seeing “to be” as being “bundled” in with 眩しい.
On the other hand, this also means less uniformity since (non-adjectival) verbs always end in a -u sound and these adjectival verbs don’t. In general this approach seems a bit mind-bending for people coming from a Western language and may be counterproductive for that reason.
What are your thoughts? What is the best language model here and are there pros and cons that I haven’t considered? Are there other elements of Japanese that are explained really differently by different language models like this? Is it best to know both models and apply them on a case-by-case basis?
8 comments
There is no definitive answer to a question like this.
The best “language model” is whatever helps you accurately understand the Japanese. Any English explanation of Japanese grammar is just a means to an end — a way to get native English speakers to be able to be able to comprehend and process the grammar/syntax of a language that is very different from their own.
You don’t have to “know both models” or anything. It’s best to just find whatever English explanation helps you understand what’s going on in the Japanese, and from then on just understand the Japanese as Japanese. This applies for literally any example of Japanese grammar or sentence structure that differs from English (which is almost all of it).
I think the i-adjective and na-adjective model is fine tbh.
It’s how I was taught and I managed to become fluent in Japanese.
But calling 形容詞 “descriptive verbs” and 形容動詞 “descriptive nouns” when 動詞 means “verb” seems like a very confusing choice of terminology.
I’ve heard multiple takes on adjectives and none of them fundamentally changed the fact that い adjectives basically conjugate the い to 〜い・〜くない・〜かった・〜くなかった・〜く・〜くて, while な adjectives add な before nouns and certain grammatical structures, and otherwise basically conjugate as 〜∅・〜じゃない・〜だった・〜じゃなかった・〜に・〜で. Maybe a simple explanation of “な adjectives are (sometimes) grammatically similar to nouns” is generally all that’s needed for explaining a simplified model.
Arguing “What really *is* an adjective?” is imo getting too far into the minutiae when it doesn’t really matter wrt learning and actually using Japanese, unless you also have an interest in linguistics and just want to know what linguistics are arguing about. Especially the part about what language model is “the best” is not really helpful. Probably the same answer as which romaji system is “the best”. >!None of them are because they all focus on doing / explaining different things well!<.
I personally say い adjectives and な adjectives when speaking here because it is instantly obvious what they mean to almost every learner that has at least glanced at the absolute basics of grammar. If I’m not tailoring my word choices for beginners, I use the Japanese terms because it’s unambiguous what they mean. I don’t see any point in jumping through hoops using 3+ different terms to describe the exact same thing when there’s already an unambiguous word for it.
> Cure Dolly’s model suggests that i-adjectives exist (and are one of the three ways to form a predicate along with verbs and the copula) but that na-adjectives are actually adjectival nouns that have the “special ability” to use な that other nouns don’t.
This is simply objectively, distributionally wrong.
I don’t understand why this idea that na-adjectives are “actually nouns” has any shred of currency behind it. It is both objectively wrong, and subjectively not very helpful in teachig students as a simplification.
Na-adjectives are objectively grammatically different from nouns, and while it’s true that a great many stems are both nouns and na-adjectives, much as in English the word “human” is both a noun and an adjective, and “love” is both a noun and a verb, many are also not.
– “穏やか” is a na-adjective, but not a noun. One can’t say “*穏やかと犬はココだ*” for instance as one can say “*人と犬はココだ*”
– “人” is a noun, but not a na-adjective, one can’t say “*すごく人だ*”, as one can say “*すごく穏やかだ*”.
– “地味” is both, and be used in both ways.
So there are some words that are both na-adjectives and nouns, yes, here are also some words that are both na adjectives and i-adjectives such as “エロ” which can both be declined as “エロな人” and “エロい人”, which doesn’t mean na-adjectives and i-adjectives the same class.
The argument seems to stem from that both nouns and na-adjectives need a copula to link with a noun and i-adjectives do not, but that’s how adjectives work in most European languages. This argument would mean that adjectives are nouns in English, Latin or Finnish because one has to say “*The room is quiet.*” and “*The room quiet.*” is not good enough, this on first glance makes “quiet” seem a noun, but as in Japanese:
– It can’t be modified by determiners
– It can’t be modified by another adjective and needs an adverb
– It can’t simply be used as the subject of a sentence
They’re simply two different grammatical categories.
As for why I find it unhelpful for language learners: I once believed it: they told me that na-adjectives were actually nouns and that “〜な” is simply the attributive form of “〜だ”. So taking what they said for truth, I was happily making constructs such as “子供な人” for “a person who is a child” and “きれいと穏やかな場所” for “a place which is beautiful and calm”, after all, nouns and na-adjectives are one and the same so one should be able to do that and one can’t.
Now, as for i-adjectives actually being verbs. Yes, they distribute in the same way as yet another conjugation class for verbs and one could argue that Japanese has three classes of verbs: u-verbs, ru-verbs, and i-verbs and I see no solid argument against this analysis down to the fact that i-verbs in Japanese can carry “nominative-objects” and in some cases even accusative objects though that’s a recent innovation.
This seems interesting to look at from a linguistics perspective, but when it comes to learning the language for conversational purposes, explanations like this tend to overcomplicate things (and learning a new language, especially one as grammatically different as Japanese, is already plenty complicated!)
I learnt い and な adjectives simply as adjectives with different forms/conjugations. All I needed to learn was how to properly use them in sentences. It was then through study, media, and conversation with native speakers that I got a grip on their various purposes, meanings, and how to fluently use them. I didn’t exactly study it by the book, it was much more of a “monkey-see-monkey-do” kind of venture.
If I had studied adjectives the way you are describing, I likely (a) would have taken a *lot* longer to understand them, and (b) would probably still be speaking Japanese formulaically rather than conversationally, you know?
By comparison, Korean treats the equivalent of na-adjectives as noun+copula, but Korean verbs (including the copula) have clearly-marked attributive forms for past, present, retrospective, and future/irrealis so there’s no ambiguity over what the “na” part means. But also, Korean uses the same helper verb *hada* (“to do”, equivalent of *”suru”*) to form fully conjugable adjectives from Chinese roots, the same as it does for action verbs (verbs and adjectives have almost identical conjugation patterns and both can predicate a sentence by themself at all levels of speech), so the whole noun-copula structure isn’t used quite as much as it is in Japanese.
If you want to take a look at a linguistics point of view of Japanese word classes [this ](https://en.wikipedia.org/wiki/Adjectival_noun_(Japanese)) wiki article, and especially the section “Internal properties” where there is an analysis of Japanese lexical properties using Chomsky’s lexical feature system might be a good start.
> Japanese Lexical Categories
Verb [+V, -N]
Noun [-V, +N]
Verbal Noun [-V, +N]
Adjectival Verb [+V]
Adjectival Noun [+V, +N]
Postposition [-V, -N]
Under this system, Japanese adjectival nouns are classed similarly to English pure adjectives [+V, +N]. However, because Japanese also has adjectival verbs with the lexical property [+V], it is observed that Japanese adjectival nouns and English pure adjectives are distinct.[5]
Linguistic categories like nouns, verbs, adjectives are still a matter of scholarly debate. Do they have psychological or neurological reality in some way or are they just neat little boxes that we artificially construct to talk about how we speak? Does every language have its own, completely separate categories or are some of them shared? Does every language even have a distinctly identifiable class such as “adjective”, and if so, how do we define it?
IMHO, we shouldn’t expect that language organises everything into clear categories. Depending on whether you consider semantics, morphology, syntax or any aspect, a particular word seem to be one thing, but considered in a different way, it might seem to be a different thing.
Semantically, there’s a spectrum between completely stative descriptions (to be tall), to rather passive activities (to sleep), to actions performed willingly (to hit someone). If a language does distinguish nouns, verbs and maybe also adjectives, how exactly this spectrum is subdivided can be somewhat arbitrary and change from one language to the next (and usually it’s not terribly consistent within one language either).