Yeah, to be honest, when I started seeing all that, I felt like か ought to be called the ‘uncertainty’ or ‘indetermination’ particle.
I think I agree, except… what do you meant by ‘predicate use’? The predicate is everything aside from the subject of the sentence, yes? (Come to think of it though… yeah, it’s like as though the question of ‘do I go now?’ isn’t part of the predicate here. It’s the topic of consideration… Is that what you meant?)
For OP: Particles come at the ends of sentences too. They come right at the end of a sentence and add just a final context, often an emotional or interpersonal one, like ね ne (isn’t it/won’t you) or よ yo (telling someone something you don’t think they know) or か ka (wondering what something is) or の? no? (hard to explain but the kind of thing you’d put at the end of “is something the matter?”) or even っけ kke (rhetorical question).
It’s a sentence final particle, according to my Japanese linguistics primers. SFPs add elocutionary “colour” to a sentence. Some analyses even connect certain SFPs directly to the verb as though they’re inflection but it’s not consistent.
Even intuitions about what constitutes a separate word aren’t always consistent in languages, although they generally are to English speakers because our language is written down and uses spaces.
For OP: This got covered in the megapost. The Japanese grammar article on Wikipedia is amazing but super super dense.
OK, so verb phrases in Japanese can attribute to a noun or predicate for a whole sentence.
“Japanese verb as attribute” comes before a noun - はしる人だ - it’s a running man. The running is an attribute of the man in question.
“Japanese verb as predicate” is where you have a verb living at the end of a sentence or mid-sentence before certain conjunctions like が or けど - they can inflected with -ます or -ません. So for predicate read “the main verb of the sentence”.
In native Japanese linguistic terminology they use the terms shūshikei 終止形 for predicate/terminal form and rentaikei 連体形 for attributive. Source - “terminal” works for Japanese just fine because… well… everything leads up to the head. 
What I meant by か not being a predicate use in your example is that it’s not applied to the predicate verb of the sentence there to create a question - instead it’s creating a little phrase out of 今行く which is hanging out with the actual predicate of the sentence わからない. 今行くか is the “complement” and わからない is the predicate.
Sorry if I wasn’t clear but I’m typing really fast!
Hahaha. I could tell. It’s ok, I understood, and I’m somewhat familiar with the idea of ‘attributive’ and ‘predicative’ even if I hadn’t seen them used the way you just did. Also, thankfully, I’ve studied French, and French grammar is full of “complément” this and that, so most of the technical terms were familiar. I believe I’ve got it.
Yup, I’ve seen those terms before. You make a good point about how appropriate ‘terminal’ is for Japanese though. I’d never thought of it that way. 
Except when compound words don’t 
. Yeah words are less useful a unit than one would think.
Still, what is the head and category of something like 「車で行きたいですか。」, and how do they relate?
Once again going by my naive understanding of things, “phrase” and “phrase + q-particle” are syntactically different, right? And it’s not like the か is acting as a complement for anything here.
All the definitions I’ve seen for head are annoyingly ambiguous for non Indo-European languages.
For OP: Ignore this, sorry for hijacking your thread.
If it’s a full grammatical sentence, it’s fairly safe to analyse it as a verb phrase of some flavour. Usually.
But verb phrases can have noun phrases inside them too, and they can compound, you can add an adverb to it and it’s still a verb phrase… phrases embed inside one another. Like this:
The last few posts are a good example of why the SOV model is much easier to start beginners off with. lol
The sentence is someone asking someone else if they want to go by car.
The “want to go” is essentially the verb いきたい, which is the volitional form of いく(to go).
Thus いきたい is the head of this phrase thereby making it a verb phrase.
The particle か has nothing to do with the type of sentence, it doesnt change anything about the sentence compared to not having it there, the structure remains the same, and so does the head of the sentence.
I get the general concept of breaking stuff down recursively into trees, just not how to apply it in this case.
@Emiloow but is いきたい a verb or an adjective? And what happened to head-final?
I’m not trying to be obtuse or anything, it just doesn’t make sense to me 
@alo I was just thinking that 
I’m so glad I got taught the lie first.
いきたい is the verb 行く conjugated to its volitional form,
(Im not actually sure of this)
it is not an adjective. Well actually it might be, but we’re concerned with its class in its unconjugated form when determining the phrase type.
I can see why you may be confused by the fact that it ends on い but the volitional form basically conjugates a verb to end on -たい, and you could argue that this changes it to an adjective.
It is head-final, but head final doesnt mean that NOTHING can come after it, to simplify, no verb,noun or adjective can come after it. The copula だ/です is neither of those, and the か is a sentence ender particle meaning it is grammatically irellevant.
The easiest way is to look for the last Verb (in any of its conjugated forms), Noun or adjective, which ever appears last. That will be the head of the sentence (this method only works for japanese)
This neatly ties in with the fact that japanese is left branching meaning modifications will always appear before what theyre modifying (for example an adjective defining an attribute of a noun)
EDIT:
im also confusing myself with the phrase type now but generally i don’t find it essential to determine the phrase type, im mainly interested in finding the head of the sentence, and thats honestly the essential part of understanding how japanese structurally works.
I think traditionally, ない and たい are considered auxiliary helpers, not part of the conjugation (sort of like how 過ぎる gets added to the end of things). I think they change the verb into a 形容詞, or at least it becomes indistinguishable from one. From a practical language usage perspective, I find this mode of thinking more helpful…
I see. Does this mean they’re adjuncts then?
Yes traditionally its not part of the conjugation, but for learning purposes i think its fine to treat them as such (thats what i do anyway, while i know its not technically correct)
I don’t think there is a “correct” here, so much as two different ways of describing the same squiggles 
Yes in the end its about what works best for learning it unless one is a linguist (im not)
I just happened to stumble on head finality and left branching, and they have been incredibly useful to me. I may have a good tolerance for abstractness tho (i know advanced mathematics), so im not sure how well it will work for others.
Bunpro is completely different from bunpo! Bunpro is much better in my opinion in every way.
Definitely. But it makes sense without thinking about it too much.
So when たい gets stuck to the 連用形 form (ren’youkei - masu stem) or ない gets stuck to the 未然形 form (mizenkei = negative/“irrealis” form), the phrase marking changes. Both are both bound (= sticky) auxiliary adjectives, so syntactically they become the head of the phrase even though they don’t quite carry the meaning. Being the head of the phrase means the verb with the meaning gets “demoted” and the auxiliary takes any inflection - ば form, (く)て form, whatever. For polite speech they need the support of です since grammatically adding ない or たい makes something a verb-like adjective.
過ぎる is one of many bound (= sticky) auxiliary verbs, so when that attaches to the 連用形 form (ren’yōkei - masu stem) it also takes on the inflection and becomes a た form or ば form or whatever, but since it’s also a full verb it can do its own terminal form work too so it can be ~すぎます or ~すぎません or whatever without help from です. There are a TON of auxiliary verbs in Japanese and they are worth looking up to turbocharge your vocab.
いる in ている would be an example of a free (= not-sticky) auxiliary verb that attaches to the て form of the verb to give the verb a progressive/continuative aspect (is -ing). Again, since it’s syntactically the head of the phrase, it picks up the inflection of the sentence so いた、いれば、いたら, even though the meaning comes from the verb it stole the headship from.
Hope that wasn’t too confusing.
For OP: Learning boring conjugation tables pays off. Can highly recommend The Handbook of Japanese Verbs and The Handbook of Japanese Adjectives And Adverbs. They contain all this stuff and lots of little tests.
The thing is, knowing and understanding are so different. lol, I’m able to follow the conversation just fine and you guys are bringing up some fascinating stuff, but even after over a year it’s still hard to wrap mind around the order.
Not sure if you’ve seen Reverse Polish Notation but it’s like taking:
(2 x 3)+(4 x 5)
And writing it:
23 * 45 * +
There’s definitely a mental shift that has to happen.
Tip for OP: It can be super helpful to listen to spoken Japanese (Japanese video with subtitles off for instance) or read printed/on-screen Japanese (e.g. tweets without a translation) and try to see what you can pick out based on what you already know. Get comfortable with understanding fragments of things at a time. Japanese can be a serious stretch for brains only used to Western European languages…
I have an undergrad degree in linguistics but I get where you’re coming from, abstractions are very useful.
You might enjoy the notion of generative grammars if you haven’t come across them yet, the idea of syntax reduced to a set of definable input and output rules which are able to recurse. It’s a pretty neat idea.
I found speaking and writing helped to get my brain in that gear. English starts with the critical detail then fills things in afterwards, whereas Japanese builds up to the important detail. The brainfeel of Japanese is getting information first and then being given a context for it, and English gives you a context first and fills information in afterwards: “box-in” versus “in the box”.
Coming from most European languages you literally have to reverse your order of expectations and that is really hard, especially with longer sentences.
That past/future - front/back metaphor being flipped from English is mildly confusing as well.
I don’t think any of these recent comments are helpful at all for OP, especially as (s)he already said (s)he doesn’t have a background in linguistics or language learning.
So I’ll echo what was said before the thread went off topic to answer OP’s question:
Start with grammar as soon as possible. Whether you use an online resource like Tae Kim/Japanese Ammo/Cure Dolly or an offline resource like Genki.
BunPro is a good way of supplementing those resources (especially the resources it has “paths” for).
I think the hard part is building up that short term memory space. Like, I have to store the objects until I get to the part where the stuff happens.
While they’re not helpful right now, I think they’re good to have here for the OP to come back to later as their Japanese gets better.
