Interesting Dissertation on Welsh (but references to some other languages) - "P-stranding"

[Montmorency]

Found this by chance:

http://e.bangor.ac.uk/5122/3/Ryuichiro% ... alised.pdf

About what is apparently referred to as "preposition stranding".

The author is mainly comparing and contrasting the situation in Colloquial Welsh and Literary Welsh and a lot of examples are given.

Although it is primarily about Welsh, there are references to other Celtic languages and other languages.

It is written from an academic linguistics point of view and is therefore full of specialised terms that I don't understand (and nor am I much interested in). But there is plenty in there for the lay-person interested in Celtic languages (and perhaps other languages) to get something out of it.

(I was initially just looking for rules about sentences ending with "with" ("gyda" or "efo") in Welsh. I had seen contrary views expressed elsewhere. It could be that some of the confusion is caused by people applying literary Welsh rules to the colloquial language....but it's probably not quite as simple as that).

[jeff_lindqvist]

Thanks for sharing! It will come in handy the day I choose to add Welsh to my repertoire.

[galaxyrocker]

Hmm. I find it really interesting that Welsh, colloquially, does allow stranded prepositions. If I had time, I'd totally read this. Irish, just for some added information, doesn't allow stranded prepositions, and you either have to inflect it, or shift the preposition; the shifting, however, can only occur in positive clauses.

[Montmorency]

Thanks for sharing! It will come in handy the day I choose to add Welsh to my repertoire.

Well, the more the merrier, and sadly, in spite of it getting quite a lot of official support these days, it needs all the help it can get.

[Montmorency]

Nothing to do with the OP, but an interesting blog article tackling the myth(?) that Welsh is inherently difficult:

http://cneifiwr-emlyn.blogspot.co.uk/20 ... icult.html

Note however some of the interesting comments below, one in particular points out some of the specific problems the blog author slightly glosses over.

[Montmorency]

Another interesting aspect of Colloquial Welsh vs in this case, Middle Welsh or traditional Welsh:

"synthetic" verbs vs "analytic":

http://westernlinguistics.ca/Publicatio ... _Johns.pdf

In modern grammar books, the synthetic forms of the verbs are usually referred to as inflected verbs, or more commonly as "short forms".
The "analytic" forms are usually referred to as "periphrastic", or "long forms".

Learners have traditionally been taught "long forms" first, as they are a lot easier, and "short forms" were reserved for advanced learners.
Even Say Something in Welsh taught them that way in its first generation of courses.
However, their new generation of courses ("Level 1" and "Level 2" (which is still coming out gradually)) take a different approach and introduce the common short forms quite early on.

In fact, it's mostly just a smallish number of common verbs that are usually used in the short form by first language speakers (although you will come across the less common verbs in short form as well). These are verbs like "gweld" (see) "mynd" (go) "dod" (come), "cael" ("to get" or "to get to do" or in some cases "to have").

Periphrastic or long form verbs are formed by using either "bod" (to be) or "gwneud" (to do) as auxiliaries plus the dictionary form (the so-called verb-noun) of the main verb. (Note that Welsh does not have infinitives in the usual sense). In this long or periphrastic form, the main verb does not have to be conjugated, but the auxiliary (bod or gwneud) is conjugated. Bod is highly irregular. Gwneud is somewhat more regular.

[galaxyrocker]

That's actually really interesting, because 'Standard' Irish is taught using a mix of the synthetic and the analytic forms. The 'Standard', which poorly reflects natively-spoken Irish of any dialect (I'm pretty sure they played drunk darts when determining where to get a feature) uses the synthetic in the first person, plural and singular. However, none of the dialects do it that way. Connacht and Donegal sometimes use it in the first person singular, but will never use it in the plural. In Munster, it's used in every instance that has one - so anything except third person singular. It's honestly really weird how they decided to do it in the 'Standard'.

[Montmorency]

Keep finding these slightly obscure, but usually interesting papers on aspects of Welsh (and its relationship to other languages, usually):

http://www.hildegard.tristram.de/media/ ... -07-07.pdf

(Touches on the possible influence of late British - Old Welsh on Old English, among other things).

(Page 213)

...
Most of these aspectual experiments did not enter the English Standard, but
many of them survived in the dialects. In the modern Standard periphrastic
DO has two functions which are clearly distinguished by stress. Stressed DO
expresses emphasis (i.e. marking by ‘contrastive accent’), while unstressed
DO means support of negation and question marking. Non-standard periphrastic
DO expressing habituality is widely used in south-west England,89 Ireland90 and
Newfoundland.91

(Page 214):

...
So, why then don’t the English speak Welsh? My suggestion is that the English
don’t speak Welsh because the native Britons chose to give up their native varieties
of Late British and shift to the emerging Old English dialects first in the
British Lowland Zone and later in the Highland Zone over a period of some 300
years. In doing so they are likely to have Brittonised spoken Old English on the
level of phonology and above all morphosyntax. By shifting they produced OEl,
i.e. vernacular Old English or what we eventually encounter as ‘Middle English’,
which only surfaced in writing after the Norman Conquest. These shift-induced
analyticising tendencies were reinforced by contact with adstratal Old Norse in
the Danelaw areas, particularly in the north. The aspectual tendencies, however,
arose in the south west, where Scandinavian influence was far less pronounced
and substratal influence of Late British therefore likely to have been solely
responsible for grammatical calques

[Montmorency]

https://www.uni-due.de/~lan300/18_Language_Change_in_Early_Britain_(Hickey).pdf

"Language change in early Britain: The convergence account
Raymond Hickey, Essen University"

Touches on some familiar themes, but is mainly talking about the complex factors affecting language change such as contact with other languages, compared to internal change. Makes the point that Britain and Ireland were (probably ) not subject to a sudden "Celtic Invasion" but rather a ‘cumulative Celticity’ arose. This probably meant that the Celtic language(s) was (were) affected by the indigenous languages, and later on, Celtic went on to influence what became English.

[Iversen]

I'm not actively studying any Celtic language right now, but I still have Irish lurking somewhere in the background, and I remember that the division line between analytic and synthetic verbal forms was drawn in different places according to which text book or grammar I was using. My old teach yourself tended to use synthetic forms, while some internet sources recommended analytical forms. I even drew a comparative map to get at least some inkling of an idea about which forms were most likely to be one or the other, and which ones appeared to be fluctuating between the two. There may also be some dialectal differences involved (as far as I remember the TY used a rather conservative variant of the Munster dialect), but as you are dealing with Welsh the dialectal differences within Irish aren't relevant - only that it may be a factor also for Welsh.hem in a lecture during the

As for officially disallowed, but commonly used syntactical construction it would be obvious to point to 'P-stranding' in English, but I would also like to mention that I spoke about one such case in Danish during the Polyglot Gettogether in Berlin (heaps of conjunctions and pronouns like "som at der"). Just days ago I saw a paper that proposed that periphrastic "do" in English had lived long before 1400 where it is first attested, but the claim was that it just had been kept out of official and literary writings until it finally burst through the defenses. We have few chances of ever knowing the details in the spoken languages of our longgone ancestors, so it is very reassuring that at least some research now investigates such phenomena in modern languages as they are and not as some official institution or editorial policies suggest that they should be.