@BouzuHarinezumi No, that is not the case

The little arrow next to my message links it to the message I'm responding to.

"To be a predicate it has to be in 連体形" is not the case.

Well, for example, 連用形 in 中止法.

Well, typically in a coordination of predicates the last predicate will be in conclusive form (終止形), yes.

I find it just a little strange referring to the 終止形 as 連体形, since it does not link to a following 体言. I understand that since there is now a widespread syncretism between 終止形 and 連体形 that it is not desirable (for most classes) to posit separate forms for the two,

but I don't really see the argument to use the term 連体形 instead of 終止形 if we're going to use just one, rather than say, 連体終止形 or 終止連体形.

Just because the form now used conclusively is the same form that was historically used adnominally does not mean we have to refer to it as the adnominal form everywhere.

I'd rather just stick to the conclusive label.

Anyway, the label isn't terribly important :-)

@BouzuHarinezumi Really? All? I've only ever seen one.

Which grammar books do that, out of curiosity?

@BouzuHarinezumi They are not necessarily 100% identical.

But yes, there is a widespread syncretism.

It's theoretically simpler not to posit a complete syncretism between forms, and to simply say there is only one form . . .

But in Japanese, you will still see (for example) the literary ～き adnominal form for 形容詞 used productively by native speakers.

@BouzuHarinezumi Most native speakers you'll find on this forum grew up with 学校文法, in which the two are distinguished, so you'll surprise these speakers less if you use the terms the way they expect :-)

But you are free to use the terminology and theories you would like to.

In 日本語教育 they are often referred to as 辞書形 instead, I think.

日本語教育 = Japanese taught as a second language

Move your mouse over the message, and a little arrow should appear at the right side. You can click this arrow to reply.

It prepends the message with a colon and a message id, which the system uses to link the two.

@BouzuHarinezumi 未然形 itself there is a misnomer.

It is in no way irrealis. Actually, the 未然形 as a form is an artifact of analysis rather than a real linguistic phenomenon.

The terms 未然 and 已然 go back to Classical Japanese, when you had the irrealis–realis mood contrast between あらば and あれば, which was lost in Modern Japanese.

There really isn't much motivation to use that term in an analysis of Modern Japanese, except for continuity with analyses of older forms of the language.

Which I think is actually a strong motivation for some people, because in Japan, learning about Japanese grammar in school is a springboard for studying 文語, so it makes sense for the analyses presented in Japanese schools to have some continuity between the two.

We face the same sorts of problems in English, where people still talk about widespread syncretism in forms that, today, are all a single form.

Like, we used to have a mood system in English with an indicative–subjunctive–imperative contrast.

And we'd posit, for example, the inflectional paradigm for take: take, take, take, take, take, take, take, take, takes, take, take, take, took, took, took, took, took, took, take, take, take, take, take, take, take, take, take, taking, taking, and taken.

But isn't it comical to suggest that take, take, take, take, take, take, take, take, take, take, take, take, take, take, take, take, take, take, take, and take are different forms?

But still today a lot of people teach it that way.

@BouzuHarinezumi I tend to use the term 未然形 because it's what people are used to, and if I explain it differently it'll be harder for most people to understand. But

For example:

飲まない nom-anai
飲みます nom-imasu
飲む　　 nom-u
飲め　　 nom-e
飲もう　 nom-oo

If you draw the boundary in morphology at the phoneme level instead of at the kana level (which is after all an artifact of orthography and not a linguistic phenomenon), you naturally conclude that the stem of 飲む ends in /m/, and it only has one form.

But!

We can see that certain forms like 飲み and 飲む and 飲め can all exist independently.

And they tend to have associated usage and semantics, while 飲ま does not exist on its own.

So we might want to divide up 飲み nom-i and say that it is nonetheless a thing (call it the infinitive, as it is the most basic non-finite form), whereas with 飲ま nom-a we have no evidence for calling it a thing on its own.

The various things that we claim attach to the 未然形 don't have any shared semantics, either. Causative, negative, passive?

@BouzuHarinezumi Yes, that is right.

So we could still potentially divide up the i and masu for example, but dividing up a and nai is harder to justify.

Then you have morphophonemic rules to explain the attachment of -(a)nai to both consonant and vowel stem verbs, and the same goes for -(r)are, -(s)ase, etc.

@BouzuHarinezumi The 連用形 is referred to by the Western structuralists as the infinitive. The 連体形 could never be considered an infinitive form because it is, of course, finite.

@BouzuHarinezumi Hehe, there is no need to refer to me as 先生, I'm just here to learn :-)

A lot of linguists have taken different approaches to the analysis of nouns (名詞) versus noun-like adjectives (so-called 形容動詞, or ナ形容詞).

The approach taught in schools in Japan (学校文法 again) is to teach 形容動詞 as an inflecting type of word, in which the 断定の助動詞「だ」 has been reanalyzed as a kind of inflectional ending rather than as a separate word.

So they teach that 綺麗だ and 綺麗な are both forms of a single word.

Whereas many other analyses would say that 綺麗＋だ is a construction consisting of a word 綺麗 plus the 助動詞「だ」.

And that is exactly the analysis that is used for nouns, isn't it? So why would it be different in the case of ナ容動詞?

In fact, ナ形容詞 behave grammatically almost just like nouns, so some linguists don't recognize them as a grammatical category.

And you can find the editors of 広辞苑 in this camp. They don't list anything in their dictionary as 形容動詞 :-) They just call them 名詞.

But how do we distinguish 名詞 from 形容動詞?

If we analyze them both as non-inflecting words which are used in construction with the 断定の助動詞「だ」, then

in terms of what follows these words, they differ in just one respect: whether they are followed by な or の.

So some argue that there is a strong motivation to consider 形容動詞 a subclass of noun which are semantically like adjectives, and they call them "adjectival nouns".

But we also have to look at how words function when deciding how to divide them up into lexical classes.

And some linguists set up another criterion for noun-hood: can it be a subject?

And if you take this criterion, 形容動詞 don't really seem to fit into the noun category.

So despite appearing in construction with だ like nouns, we could set them up as another category, a type of adjective.

But they are very much like nouns grammatically, so linguists who use this criterion can call them "nominal adjectives".

好く was originally a verb, and so it had the 連用形「好き」. The 連用形 is itself a verbal form, by the way, not a noun, but nouns can be derived from 連用形. Just like how in English, we can have the verb form taming, and we can derive the noun form taming from it.

And 好き was derived a long time ago, so it doesn't make sense to say it still is the 連用形 of 好く. Rather, it was originally the 連用形, but a nominal form was derived from it.

But because nouns are so similar to nominal adjectives or adjectival nouns, whatever you want to call them, the boundaries between the two are pretty fluid. Some words allow both.

Teramura, for example, has written about the continuous nature of these two classes and how there is not always a sharp distinction between them.

So we shouldn't be surprised that a nominal form could end up being a nominal adjective and not just a regular noun.

Anyway, this all happened a long time ago.

@BouzuHarinezumi 未然 was a linguist's attempt, quite a long time ago, to find a unifying semantics of all the forms which attached to the .a form. At the time, Japanese linguists were bound by the limitations of kana. Modern linguists don't have this problem, and today some such as the Russian linguist Alexander Vovin reject the analysis for the reasons I tried to explain earlier.

Still, others stick to the analysis of the 未然形 (but may choose not to use that name) on theoretical grounds, so there isn't really a consensus that it "does not exist" as I put it. Frellesvig is an example of a linguist who analyzes Japanese as having a 未然形, but he calls it the a- stem instead.

The term "a- stem" isn't particularly satisfactory, but it's hard to find a good term that fits the semantics of all the things that attach to it :-(