Language typology

Question

I think that there are at least 2 types of languages: those that make derivates/inflections predominantly from nouns or protonominals, and those that make them from verbs or protoverbs, plus a 3rd category that make them direct from abstract meaning). Anybody knows what exactly term for this 2-way or 3-way typology? Or where can I read something close to this?

I want to have something like World Atlas of Language Structures where will be something like this: English predominantly verbal in its morphology: verb is that morpheme that brings meaning to all possible following derivates/inflections (I mean a live rules of the language: I understand that historically it was a complex process (deverbalization, denominalizations and vise versa), but what language prefers now?), or Swahili is reported to be nominal: all its derivatives, especially verbs that are formed from nouns. (I don't have direct experience with Swahili.) Something like this: There was a reason for my interest: in Finnish, the verb 'to write' is formed from the word 'a book', as the noun 'a writer' is formed from the word 'a book', i. e. Finnish is denominal language? But in many other there the opposite situation: to write >a writer (a person) >> *a writer (a notebook), or Turkish: yazmak (verb) > yazcı >> yazma, etc.

P.s. in the course of this theme I want to ask additionally about Proto-Indo-European roots: it seems to be to all of them a verbal roots, or not?

What is a priori state of Indo-European lexicon?

Where or what I can read about all of these?

P.p.s. I don't know why my question is downvoted ( it becomes standard situation on this site: I have a total non-understanding why sane questions/answers are downvoted.) but it is a valid question, at first, and many years years ago I was reading description like 'language X is predominantly verbal' but I don't remember what description it was.

Thanks to the answer about Finnish word 'kirja', I know that it is complicated question, and at many cases there is 'situation vice versa'. But it still interesting, intriguing question: what is the first point of the compilation? If say about the history of the languages. Or what is the productivity type, what language prefers now? And maybe somebody knows what background reading can be useful to read for this.

Answer 1

I will address one aspect of the question, contained in the assertion that English is predominantly verbal in morphology and Swahili is nominal. This is entirely incorrect. English morphology is highly fossilized, and the productive morphology is rather limited (not as limited as Chinese, but limited nonetheless). Any claims about verbal vs. nominal being "predominant" in English are hair-splitting at best.

The situation in Swahili, and Bantu in general, is vastly clearer, and may help to understand what the underlying issues are. The main question that has to be addressed is whether roots have an intrinsic specification for part of speech. There are some roots that don't appear in any verb form (/ŋombe/ 'cow', /kamusi/ 'dictionary'). There are almost no roots that can only appear in verbs, the exception being the copula root li. Any supposedly verbal root has many verbal realizations (past, present, future, subjunctive, imperative etc.) and a number of nominal ones ("the act of", "one who", "thing to V with", plus the infinitive which is syntactically a higher-level N dominating a lower level V). We can either say that Swahili has nouns and verbs, or it has nouns and "neutral" roots which can be treated as verbs with one set of affixes, and nouns with a different set of affixes. The clearly verbally-inflected possibilities combine to allow thousands of word forms (and there are let's say a dozen verb-to-noun processes), but there are only a handful of noun-to-noun processes. By any metric, Swahili is more "verbal" than it is "nominal".

This is a general fact about the fast majority of human languages: there are usually very many more verbal morphological forms than there are nominal ones.