Are there languages without non-finite verb forms at all?

Question

This is inspired by the comments to this answer: Are there languages without any non-finite verb forms, or almost without any non-finite verb forms?

Examples of such languages are welcome!

Answer 1

Inuit (Greenlandic)

My Greenlandic is rudimentary at best, but as far as I can recall from my uni classes many years back, Greenlandic (and I believe other Inuit) verbs have only finite forms. The typologically most common non-finite forms, infinitives and participles, don’t exist, and I don’t believe there are others either. The Wikipedia article on Greenlandic doesn’t mention any either, nor do the words infinitive or non-finite appear anywhere in this extensive introduction to Greenlandic.

There is a participial mood, but that’s an inflected, finite mood like the indicative, interrogative, imperative, optative, etc.

It is of course possible to derive nouns from verbs using the plethora of derivational suffixes available, but though this is a very integral part of the grammar, it goes beyond the actual verbal system itself.

One possible exception

I just looked at the answer linked to in the question and realised that I’d made this same suggestion two years ago, but had looked it up and partially disproved myself. It seems Greenlandic does have one single entity that could be classified as a non-finite verb form (though, having read the relevant literature again, it’s not entirely clear to me whether it’s really a verb form or a derived noun). This is the suffix -neq, which creates abstract nouns from verbal phrases (note: verbal phrases, not just verbal roots, like most nominalising suffixes).

According to Michael Fortescue’s West Greenlandic (Croom-Helm, 1984, p. 44ff; spelling in Greenlandic forms modified to Standard Greenlandic orthography; glosses modified to better match Leipzig rules), -neq can

convert a clause of any degree of complexity to an NP acting as object or subject of a superordinate verb. […]

• nunaqarfim-mi savaateqarfi-u-su-mi nukappiara-a-lluni sava-leri-neq
  settlement-LOC sheep.herding.place-be-INTR.PART-LOC young.boy-be-4s.CONT sheep-be.occupied.with-NEQ

[…]

The first example savalerineq ‘sheep-herding’ on its own is siffucient as noninal subject. Note the impersonal use of the 4th person contemporative mood in the full sentence […]. The contemporative in these constructions can be said to be non-finite in so far as it de-specifies the subject.

-neq forms do not include markers for mood, person and number from, all of which would be present in the corresponding finite clause, but may be inflected for case and possession as any other nominal form:

The latter marking [for possession] can reintroduce number and person referring to the subject of the clause. If the subject is overtly expressed it will be in the relative case:

• nalu-aa qinnuta-ata qanoq naammassi-neqar-ni-ssa-a
  not.know-3s>3s.INDIC request-3sPOSS.REL how implement-PASS-NEQ-FUT-3sPOSS

There’s a whole lot of fairly complex stuff about how various objects and other constituents are deleted or able to be expressed in roundabout ways using half-transitive constructions, which may or may not be relevant to the categorisation; but including it here would be overkill.

At any rate, it’s clear that, while this is perhaps not a straightforward participle or run-of-the-mill non-finite verb form, it is at least a type of derivation that is unique in Greenlandic in allowing the resultant nominal form to retain certain aspects of its verbal base. As such, it is at least arguably a kind of non-finite verb form.

 

---

 

Goidelic (Irish)

You could also argue that the Goidelic languages (Irish, Scottish and Manx) do not have non-finite verbal forms. Since I know Irish best, I’ll base the following on Irish (see Irish conjugation on Wikipedia for details), but I think it should work for Scottish and Manx as well. There are two slight copouts you’d have to accept for this to hold up:

• The autonomous (or impersonal) forms used to form generic ‘one does’ statements and passives are actually synthetic, finite forms
• The verbal noun and the form usually called the ‘past participle’ are actually derived nouns/adjectives, rather than an infinitive and true participle

The first of these is, I think, a fairly standard view.

Most synthetic verb forms also have an optional analytic forms (whose usage varies by dialect), but this is not the case for the autonomous form (e.g., táim ‘I am’ is equivalent to tá mé, but there is no analytic form of táthar ‘one is, people are’). This is true of a few other synthetic forms as well, though, so I don’t think it’s a deal-breaker (e.g., bheifeá ‘you would be’ also has no analytic counterpart).

The second is more controversial.

It’s generally accepted that the verbal noun is not an infinitive – although English influence has meant it is behaving increasingly like an infinitive, it is still in the standard language (and to most native speakers) fully a noun and inflects as such. Unlike integrated infinitives, there is also no single way to form the verbal noun – every verb (except the copula) has one, but you have to learn it individually for each verb.

The ‘past participle’ is more like a regular participle, especially in that it is (nearly) always regularly formed. However, it is not used to form compound forms in the same way that participles are in most other Indo-European languages, such as perfective forms in Germanic and Romance language. Perfectives in Irish are formed in one of three ways (using ‘I have read the book’ as an example):

Structure	Example	Literal translation
simple past	Léigh mé an leabhar	read.PAST I the book
AdvP (e.g., i ndiaidh ‘after’) [+ object + a] + verbal noun	Tá mé i ndiaidh an leabhar a léamh	is I after the book to reading.VN
patient as subject + past participle as attributive adjective + agentive PP with ag ‘by’	Tá an leabhar léite agam	is the book read.VAdj by-me

In other words, it’s used similarly to the past participle in Classical Latin (in active verbs), before the development of the compound tenses. Whether this is sufficiently distinct from a true participle is probably impossible to say objectively. But a few facts point to this being at least a plausible analysis:

• the participle cannot take objects like finite forms can
• the participle is only used adjectivally with a passive sense
• there is no corresponding active/present participle (the preposition ag ‘at’ + verbal noun is used instead)
• the participle and the verbal noun have a sort of symbiotic relationship, in that the genitive of the verbal noun is always the same as the participle; so for example, ‘people who read books’ or ‘book-reading people would be lucht léite leabhar (lit. ‘class-of-people reading.VN.GEN.SG book.GEN.PL’, so ‘people of the reading of books’)