Questions about inflected prepositions

[Jonlang]

Basically, my question is where do they come from? Or where can they come from? Using Welsh as the example, it has many inflected prepositions which I cannot find any etymological information for, only for the base form, i.e. the preposition i comes from Proto-Celtic *de; o comes from Proto-Celtic *ɸo < PIE *h₂pó. My assumption is that the inflected forms began in the Proto-Celtic seeing as all the Celtic languages feature them.

But how did they form? Why did people start sticking bits onto them? How did i start becoming iddo or iddi, for example? Or did it? If Welsh i comes from Proto-Celtic *de then maybe the inflections were always there and i just got shortened again and again?

I'm interested in including inflected prepositions into a conlang but I want to do it properly.

[Dewrad]

Basically, my question is where do they come from? Or where can they come from? Using Welsh as the example, it has many inflected prepositions which I cannot find any etymological information for, only for the base form, i.e. the preposition i comes from Proto-Celtic *de; o comes from Proto-Celtic *ɸo < PIE *h₂pó. My assumption is that the inflected forms began in the Proto-Celtic seeing as all the Celtic languages feature them.

Unlikely as Gaulish and Celtiberian lack them.

But how did they form? Why did people start sticking bits onto them? How did i start becoming iddo or iddi, for example? Or did it? If Welsh i comes from Proto-Celtic *de then maybe the inflections were always there and i just got shortened again and again?

Literally, they're just from sequences of preposition + object pronoun. Sometimes extra material is added, such as the -dan- in amdano etc. In several cases the prepositional inflections have been remodelled on the basis of verbal inflections as well, but initially it's just preposition + pronoun.

[Jonlang]

But how did they form? Why did people start sticking bits onto them? How did i start becoming iddo or iddi, for example? Or did it? If Welsh i comes from Proto-Celtic *de then maybe the inflections were always there and i just got shortened again and again?

Literally, they're just from sequences of preposition + object pronoun. Sometimes extra material is added, such as the -dan- in amdano etc. In several cases the prepositional inflections have been remodelled on the basis of verbal inflections as well, but initially it's just preposition + pronoun.

So amdano is just am + o with the dan just being something that got thrown in the middle? And iddi is just i + hi with a dd thrown in (and the h lost)?

[Salmoneus]

But how did they form? Why did people start sticking bits onto them? How did i start becoming iddo or iddi, for example? Or did it? If Welsh i comes from Proto-Celtic *de then maybe the inflections were always there and i just got shortened again and again?

Literally, they're just from sequences of preposition + object pronoun. Sometimes extra material is added, such as the -dan- in amdano etc. In several cases the prepositional inflections have been remodelled on the basis of verbal inflections as well, but initially it's just preposition + pronoun.

So amdano is just am + o with the dan just being something that got thrown in the middle? And iddi is just i + hi with a dd thrown in (and the h lost)?

I don't know what the truth of Welsh, but I suspect that "something that got thrown in the middle" is either:
- something otherwise dropped from the end of the preposition
- something otherwise dropped from the end of the pronoun
- something added by analogy to another preposition or to a verbal paradigm

[Jonlang]

But how did they form? Why did people start sticking bits onto them? How did i start becoming iddo or iddi, for example? Or did it? If Welsh i comes from Proto-Celtic *de then maybe the inflections were always there and i just got shortened again and again?

Literally, they're just from sequences of preposition + object pronoun. Sometimes extra material is added, such as the -dan- in amdano etc. In several cases the prepositional inflections have been remodelled on the basis of verbal inflections as well, but initially it's just preposition + pronoun.

So amdano is just am + o with the dan just being something that got thrown in the middle? And iddi is just i + hi with a dd thrown in (and the h lost)?

I don't know what the truth of Welsh, but I suspect that "something that got thrown in the middle" is either:
- something otherwise dropped from the end of the preposition
- something otherwise dropped from the end of the pronoun
- something added by analogy to another preposition or to a verbal paradigm

I was thinking as much. I doubted that it would have been a randomly chosen sound because the patterns are pretty regular.

[Dewrad]

But how did they form? Why did people start sticking bits onto them? How did i start becoming iddo or iddi, for example? Or did it? If Welsh i comes from Proto-Celtic *de then maybe the inflections were always there and i just got shortened again and again?

Literally, they're just from sequences of preposition + object pronoun. Sometimes extra material is added, such as the -dan- in amdano etc. In several cases the prepositional inflections have been remodelled on the basis of verbal inflections as well, but initially it's just preposition + pronoun.

So amdano is just am + o with the dan just being something that got thrown in the middle? And iddi is just i + hi with a dd thrown in (and the h lost)?

I don't know what the truth of Welsh, but I suspect that "something that got thrown in the middle" is either:
- something otherwise dropped from the end of the preposition
- something otherwise dropped from the end of the pronoun
- something added by analogy to another preposition or to a verbal paradigm

Or none of the above. For example, in the personal forms of the preposition o, such as the 1sg ohonof, the -hon- bit is actually from PC *sani- "without" (at least according to Matasović): basically we're looking at a kind of compound preposition here. The Proto-Celtic etymon for ohonof would be something like *aw sani moi.

On the other hand, in iddo we're looking at something like *dū ejū > [δ]ījjū > idd-, with regular development of medial *jj to δ.

So yeah, the extra material can come from a variety of places.

[linguoboy]

I've long wondered. The developments seem much more straightforward in Irish (basically prep + link vowel (if needed) + first letter of the pronoun).

[Zaarin]

Basically, my question is where do they come from? Or where can they come from? Using Welsh as the example, it has many inflected prepositions which I cannot find any etymological information for, only for the base form, i.e. the preposition i comes from Proto-Celtic *de; o comes from Proto-Celtic *ɸo < PIE *h₂pó. My assumption is that the inflected forms began in the Proto-Celtic seeing as all the Celtic languages feature them.

Unlikely as Gaulish and Celtiberian lack them.

I'd say the Gaulish and especially Celtiberian evidence is rather scant to make that assertion too confidently. The available evidence certainly seems that way, but I'm reasonably certain I've read some Celticists suggest Gaulish had them. (I don't recall the particulars or who said it; it was some years ago when I had a keen interest in Gaulish that has since moved on for lack of material.)

[Eiríkr Útlendi]

As another 2p in the pot, German appears to be at an earlier stage of this development. Inflected prepositions are pretty clearly contractions of conjunction + article, where the inflection comes from the article. Examples include zur in zur Tur ("to the door", dative feminine) = zu + dative feminine definite article der, or am in am Tisch ('on the table", dative masculine) = an + dative masculine definite article dem.

[Sumelic]

As another 2p in the pot, German appears to be at an earlier stage of this development. Inflected prepositions are pretty clearly contractions of conjunction + article, where the inflection comes from the article. Examples include zur in zur Tur ("to the door", dative feminine) = zu + dative feminine definite article der, or am in am Tisch ('on the table", dative masculine) = an + dative masculine definite article dem.

The phenomenon of preposition + article contraction in languages like German, French, Spanish and Italian seems fairly different from the phenomenon of "inflected prepositions" in Celtic and Semitic languages. "Inflected prepostions" in these languages are either used in place of preposition + personal pronoun sequences, or used before a personal pronoun.

Prepositions are not inflected into their third-person gender/number-inflected forms before non-pronominal noun phrases (at least, not typically or obligatory: I don't know if there are any special situations where this is possible or required), although prepositions may in addition to their inflected forms have special forms or contractions when they occur before an article.

Wikipedia points out that the use of inflected prepositions has some parallels in Spanish forms like "conmigo" and "contigo".

[linguoboy]

Prepositions are not inflected into their third-person gender/number-inflected forms before non-pronominal noun phrases (at least, not typically or obligatory: I don't know if there are any special situations where this is possible or required), although prepositions may in addition to their inflected forms have special forms or contractions when they occur before an article.

In Irish, there is some overlap between 3.s.m forms of prepositions and the forms used before the article, e.g.:

Bhíos ar scoil leis an mbreitheamh. "I was at school with the judge." (Cf. Bhíos ar scoil le breitheamh. "...with a judge".)
Bhíos ar scoil leis. "I was at school with him." (Cf. Bhíos ar scoil léi. "I was at school with her".)

I'm pretty sure this is due to the definite article and the 3.s.m pronoun both deriving from the same Proto-Indo-European demonstrative, which had initial *s.

[Curlyjimsam]

A plausible development, if you already have agreement marking on adjectives and determiners, is those same endings simply getting extended to the prepositional class.

[Znex]

But how did they form? Why did people start sticking bits onto them? How did i start becoming iddo or iddi, for example? Or did it? If Welsh i comes from Proto-Celtic *de then maybe the inflections were always there and i just got shortened again and again?

Literally, they're just from sequences of preposition + object pronoun. Sometimes extra material is added, such as the -dan- in amdano etc. In several cases the prepositional inflections have been remodelled on the basis of verbal inflections as well, but initially it's just preposition + pronoun.

So amdano is just am + o with the dan just being something that got thrown in the middle? And iddi is just i + hi with a dd thrown in (and the h lost)?

I don't know what the truth of Welsh, but I suspect that "something that got thrown in the middle" is either:
- something otherwise dropped from the end of the preposition
- something otherwise dropped from the end of the pronoun
- something added by analogy to another preposition or to a verbal paradigm

Or none of the above. For example, in the personal forms of the preposition o, such as the 1sg ohonof, the -hon- bit is actually from PC *sani- "without" (at least according to Matasović): basically we're looking at a kind of compound preposition here. The Proto-Celtic etymon for ohonof would be something like *aw sani moi.

On the other hand, in iddo we're looking at something like *dū ejū > [δ]ījjū > idd-, with regular development of medial *jj to δ

Yes, as far as I can recall when I was looking up Insular Celtic stuff for my Gaullang, gemination before unstressed pronominal affixes in the older language does play a big role in shaping the prepositional forms.