Some time back I proposed that a Vasconic substrate might be
responsible for the ubiquitous West European (Portuguese, Spanish,
Basque, Italian, Dutch, Northwest German and Danish) 'locativic' ("I
am at <verbal noun>) construction to express progressive tense:
Basque uses a future participle (perfective participle + -en in some
dialects, do. + -ko in others); I proposed that some Vasconic
substrate of Western Germanic (and Danish?) might have an imperative
participle (= perfective participle the final -i or -tu, = the verb
radical), to which was added both -en and -ko for good measure:
-en + -ko > PGerm. *-inga.
Voilà, the Germanic verbal noun, later English gerund.
I was therefore pleased to read that Trask: "History of Non-finite
verb-forms in Basque" proposes that exactly such a participle = the
verb radical once existed in Basque.
Another thing: -en is also the genitive suffix in Basque. Therefore a
putative *-en-ko could also be suffixed to nouns. PGerm. -inga can be
suffixed to nouns too.
Basque makes agents from verbs with the suffix -el. Since agents are
nouns, the putative *-en-ko could also be suffixed to it. *-el-en-ko
> PGerm. *-ling (Lehrling etc).
Basque has an adverb-forming suffix -ik. Is it > PGerm. *-ig-?
Basque has an 'inhabitant-of' suffix -tar, alternating with -ar.
PIE is full of -ter suffixes, alternating with -er.
Trask notes that the e- prefix of the 'old' native Basque verbal
roots, those with perf. part. in -i ('new' ones in -tu, including
Latin loans, don't have it) are used solely in the non-finite forms
of the verb. The prefix ge- in German is used solely with non-finite
forms (ppp., verbal nouns of the type Gerede, Geschrei), except for
Latin loans which don't have it (in Dutch they do, though).
(As to why that progressive construction should have spread to
Southern Europe: Late Latin was a soldier's language; if the
construction was used by the armies, it would have become standard;
veterans became farmers).