In Early Modern English, are there examples of the "a- + gerund" progressive construction where the gerund begins with a vowel?

Question

In Early Modern English, a progressive tense was sometimes constructed as in the example "I was a-hunting". But what happens if the verb begins with a vowel rather than a consonant? Would they say, "I was a-eating", or "I was an-eating", or would they simply have had no way to do this?

Answer 1

Yes, sometimes, at least through the 1700s. After that, not so much. Up through the 1700s in particular it sometimes showed up written an before a word with a vowel or h-. The OED provides one such citation of went an eeling here:

1780 in Narragansett Hist. Reg. (1882) Oct. 104 Went an eeling.

Remember that this is “really”, or originally, the preposition(!) a combined with a gerund. It was not a prefix! It was only in later years it came to be written as a-, although it moved on from on to a/an fairly early.

In particular, the OED gives:

a prep.

Origin: A variant or alteration of another lexical item. Etymon: on prep.
Etymology: Variant of on prep. with loss of the final consonant -n, reflecting an unstressed pronunciation of the word in proclitic use; compare an, variant of on prep., and also o, variant of on prep. Compare a- prefix3, away adv., aright adv.

[...]

II. With a verbal noun or gerund, forming part of a verbal expression. (Now usually written with a hyphen or as one word with the verbal noun.)

11. Expressing action, with a verbal noun or gerund taken actively. Now archaic and regional.

One of the earlier citations of its use with a gerund is Laȝamon’s Brut:

Wreoð nu wel þene king, þat he ligge a swæting.

Which meant something like "Wrap now well the king, that he lie a sweating."

So it didn’t derive from one the way the article a/an did. Rather, it was originally the preposition on, and its a and an forms spelling variants of that original. It never meant one the way the article did. The OED notes:

The loss of final -n in this word occurs early in terms of the developments described at N n. and perhaps began in fixed idioms where the word was felt to be almost a prefix; compare the parallel development represented by a- prefix3. A following consonant favoured the loss of final -n (compare discussion at N n.) and so until the 17th cent. the word was often found in complementary distribution with an , variant of on prep., which was common before vowels (with variation between the two before following h- ).

The separate preposition a ceased to be used in standard English after about 1700, being replaced by the full on , in , or the various prepositions which represent them in modern idiom, surviving only in a few set uses from branch II., such as to go a begging , to set a going (occasionally, before a vowel or h- , in form an, after an, prevocalic variant of a adj.; compare quots. 1759, 1780 at sense 11b), and in temporal distributive phrases, as twice a day , once a year , where it had been early identified with the indefinite article (see a adj. 4). It also survived in a large number of combinations, where it was treated as a prefix to the governed word, and the whole as a compound adverb. Cross-references are given below to a selection of these (see senses 2, 3, 5, 6, 9). Compare a- prefix3.

Yet another a- prefix

English has many prefixes that are now written a- but which are of completely distinct origins. Just to name an easy one, think of the classically derived privative or negating prefix a-, an- seen in acyclic, asymmetric, atomic, apathy, analphabetic, anarchy, anaesthesia.

A comment asked whether this a- prefix before a gerund isn’t the same a- prefix we got from ge-. It is not. The Old English ge- prefix from common Germanic, still overt in the Middle English hymn Adam Lay Ybounden, today survives only:

1. hidden as a-, e-, y-, i- in words like afford, alike, among, enough, handiwork (< OE hand geweorc), some of which actually came to us via Old Norse not Old English;
2. in deliberate archaïcisms like ybent, ybound, ybrought, yclad, yclept, ydight, ydrad, ypent, ypight, ywrought; and
3. in certain rural dialects of England found within a triangle whose three corners are Worcestershire, Surrey, and Cornwall, as well as in Pembrokeshire and Wexford, where it is said that you may yet chance upon such country relics as ameäde (made), azet (set), adood (done).

Answer 2

My regional Appalachian English dialect, famous for retention of EME features, still regularly uses a-gerunding syntax. We do not ever prefix an- so if it was historically done, it was gone before the language acquisition period of my grandparents, and thus very likely my great-grandparents' and probably great-great grandparents', which makes it extinct by the early-mid 19th century at latest.

It is notable that not only is the first syllable of the verb part very stressed, the "a" is also pronounced as a very stressed schwa (can you stress a schwa?), so more like "uh" than an "ay," or "ah," so it's a bit less awkward to say than you might imagine when it precedes a vowel. "A eating," admittedly very rare, sounds more like "UH-(y)EET-un," with just the barest hint of an intrusive y.

The dialect is also famous for intrusive consonants. The letter m is often "ilm" for example.

Answer 3

In Early English Books Online, an early modern English corpus, there is one result for a-eating, from the 1669 book The voyages and travells of the ambassadors sent by Frederick, Duke of Holstein, to the Great Duke of Muscovy and the King of Persia by Adam Olearius:

We were told also that a Bear, finding a Vessel of Herrings, which a Peasant had laid down at an Alehouse door, fell a-eating of them, and went thence into the Stable, whither the Peasants follow'd him; but having wounded some of them, the rest were glad to get away.

In comparison, there are 20 results for a-hunting. an- in early modern English would be unexpected.

Based on the modern function of a-prefixing, as described in Yale Grammatical Diversity Project, the first syllable is almost always stressed in an a-prefixing formation. So a-HUNTing or a-EATing would work, but a-proDUCing wouldn't. (I can attest to this with my personal experience in Appalachian English.) That may be why a-EATing could be found in early modern English but, say, a-exTENDing wouldn't; it would feel redundant to say two separate unstressed vowels in a row.