Lusitanic Scratchpad (Now Playing Elsewhere)

[dhok]

Here is a link to a (for now) frequently updated lexicon.

[dhok]

Grammar

I'll probably just sort of dick around here randomly. The upside here is that there's really nothing in Canarian that you won't find in Spanish or Portuguese. Nouns have gender, usually indicated by a final -o or -a:

fillo m. "son"
filla f. "daughter"

...but this doesn't always hold:

zia m. "day"
mun f. "hand"

They have plurals indicated by -s, and articles:

la filla ~ las fillas
o fillo ~ los fillos

Before vowels, la becomes l', and o becomes el.

A few nouns have irregular plurals...:
o can "the dog" ~ los quins "the dogs"

...though even these usually have rules:

-an -> -ins (o can, los quins)
-on -> ens (l'açón, las acéns)
-daç -> -dazis (la verdaç, las verdazis)

There are a few contractions with prepositions:

en + o = no
pra + o = pro
con + o = có, con + l[a/os/a] = col[a/os/a]

un "one" and dos "two" agree with the gender of their nouns:

un can~ua mun
dos quins~duas muns

So do adjectives, which work exactly as they do in Spanish or Portuguese:

un can nerro "a black dog" ~ dos quins nerras
ua mun nerra ~ duas muns nerras

There are a few irregulars:

un bon can "a good dog" ~ dos bons quins
ua boa filla ~ duas boas fillas

Comparatives are fomed with mas, except for bon, which uses mellé, malo, which produces iné, granç which has maié and equeno which has mé.

Verbs

These more or less work the same way as in Spanish or Portuguese, but because of the loss of intervocalic -r-, the pluperfect and future (except in a few irregular verbs with a leftover synthetic future) are formed through auxiliaries. Here are the synthetic forms of cantai:

Code: Select all

inf.	cantai
	
	indic.			subj.
	canto	cantamos	cante	cantemos
pres.	cantas	cantais		cantes	cantés
	canta	cantan		cante	canten

	canté	cantamos	
pret.	cantaci	cantaçis
	cantó	cantán

	cantava	 cantavamos	cantasse cantassemos
imp.	cantavas cantavais	cantasses cantassés
	cantava	 cantavan	cantasse cantassen

	cantía	cantíamos
cond.	cantías	cantíais
	cantía	cantían

There are several irregular verbs, as well. I'll only run through the highlights.

Code: Select all

inf. ser

so somos         sea etc.
es ecis		 
é  son		 

fui fue etc.

ea etc.          foasse etc.

séa etc.

(plus a synthetic future seá etc.)

Code: Select all

inf. té

tenyo temos    tenya etc.
çans tes
cin  cín

cive etc. (but 3s teve)

cinya etc.     civesse etc.

tendría etc.

synthetic tendré etc.

I need to get to bed now so I am calling that a morphological sketch. I will put up a short story probably tomorrow. In the meantime here are some words for you to peruse:

un, ua
dos, duas
tres
patro
cinco
sés
chaç
vacho
noavi
zaç (note to self: go back into sound changes to account for this and chaç)
vinç
çanto
jo
çu
ele
ela
nós
vós
blanco
nerro
verç
rocho (where Spanish has -j- and Portuguese -ix-, Canarian has -ch-)
azú
amello
madre
adre
irmún
irmán
fillo
filla
primo


Adre Noastro qui estás en los çalos:
Sancificado sea o tu Nome;
Venya o tu Réno;
Sea Fecha la çua Vontaç
En la Terra como no Çalo;
O pan hoje zi cada zia nos dai and oh fuck I am tired night y'all

[Matrix]

the Canaries don't have a language or major dialect of their own

http://en.wikipedia.org/wiki/Guanche_language

[Quantum]

Guanche is an extinct language that was spoken by the Guanches of the Canary Islands until the 16th or 17th century.

[dhok]

My bad! Let's just pretend the Guanches got assimilated when the Romans arrived...or something.

[Thry]

Cool! The p-dropping is weird but original, something like that happened in Proto-Celtic right? Did it happen through a bilabial fricative? p > ɸ > _?

Do the nasals stick in the modern language?

-Initial *pl, *cl, *fl became ʃ as in Portuguese: chové "to rain", but learned loans did not change the *l to *r.

I'm assuming that last remark is a reference to Portuguese?

Also, how does the inchoative conjugation turn out? (-ecer / -escer)? It also becomes a subtype of 2nd conjugation like Sp and Pt?

What happens to medial -pt-? apto

What are the modern words for elbow, city and head? And what's the etymology of iné?

Feel free to answer all, some or none.

[dhok]

Later today I'll go through and do some editing- I realized I don't want to reduce final -oi to -é, among other things.

P-dropping did indeed happen in Celtic. I don't know what the intermediate stage would be, but yeah, probably a bilabial fricative.

The nasals are still around in the modern language, yes.

l -> r in clusters is a Portuguese/Galician thing. It also happens in Sardinian, but that's not important.

I have not thought about the inchoatives, but likely they'll just do whatever Portuguese and Spanish do, and become a 2nd-conjugation subset.

iné is from *penore, if that's the right etymon (I think it is). The -n- sticks around basically because it's the last consonant left; and giving it some more thought it should be enoi.

Medial -pt- became ts: chace "seven".

The modern word for head would likely be cabeça if I follow Spanish and Portuguese, but I might change it to a reflex of *testa, which would I think be çasta?

I don't know what elbow would be. *civitate would likely show up as cidaç.

[Dewrad]

A couple of things: your soundchanges are lazy, and in some cases go against the common innovations of Ibero-Romance. Obviously, you're setting this up as a scratchpad first, so some sketchiness is to be expected, but in an a posteriori language your sound-changes are your foundation, and should be focused on with rather more attentiveness.

Also, by focusing solely on Spanish and Portugese, I think you've missed a few tricks. Some of the more interesting features of Ibero-Romance (metaphony!) have been lost in Spanish and Portuguese, but are present in the other Ibero-Romance languages.

Furthermore, it's all very well saying "this is Ibero-Romance," but have you given any thought to where precisely Canarian fits in the Iberian dialect continua? This is a missed opportunity: deciding exactly where in a dialect continuum something fits means that a lot of your decision-making is done for you, and it also benefits verisimilitude.

(A thought: Punic and Berber inscriptions have been found in the Canaries: some kind of substratum opportunity here? Notably, late Neo-Punic exhibits an unconditional change of p > f (probably [φ]), which fits in nicely with the most distinct of your soundchanges.)

[dhok]

Thanks for the feedback; I'll wander over to the library later today and see if there's anything on Ibero-Romance I can check out. I think p -> f initially, rather than to zero, would be a better choice (especially since right now the combination of p-deletion, r-deletion and n-deletion is making a lot of basic words very difficult to maintain).

As for the dialect continuum...I'm not quite sure. I could place it in the Azores instead, I suppose, which would make it...Portuguese+, or something. I'll admit that I don't know enough about Ibero-Romance proper (as opposed to Spanish and Portuguese) as I might ought.

[Dewrad]

Ralph Penny's works would be useful: A Historical Grammar of Spanish and Variation and Change in Spanish (the latter particularly for Iberian dialect continua), as might Hualde et al.'s Introducción a la lingüística hispánica, if you read Spanish. And you might find this handy.

[patiku]

Guanche is an extinct language that was spoken by the Guanches of the Canary Islands until the 16th or 17th century.

[dhok]

Putting this here because I can't work on one project at once. This project is an IE language called Lusitanic to be spoken somewhere in the Iberian Peninsula, probably Portugal. I'm aware that there was a language called Lusitanian that was spoken in the region closely related if not a member of Continental Celtic. I could take the inscriptions we have and use them as a jumping-point, but I'm not going to, and the haters can suck it. The idea is that Lusitanic is a sort of "lost third branch" of Italo-Celtic, so it shares a number of common innovations/retentions with Italic and Celtic, insofar as I know about them. Truth be told, my knowledge in that area is somewhat limited, but I have some ideas of what I want to include, such as o-stem genitives and masculine plurals deriving from -ī, a passive in -r (though technically this is a retention), and the assimilation of p to a following kʷ. According to Wikipedia, I'll also want to include a subjunctive in -ā, and- what the heck- I'll suppose that the IE aorist stayed distinct from the perfect in Lusitanic, while they collapsed together in the other two branches, unless there's some really compelling reason to believe that the distinction could not have been retained. Currently I'm working through the nouns; the dual will be retained, with two forms (a nominative/accusative/vocative form and an everything else form), as will all three genders, of course. Cases are trickier; the nominative, genitive, accusative and vocative will be retained; the ablative and instrumental will likely merge; I'm not sure about the locative, although its merging into the dative or genitive is looking tempting. I haven't gotten to the verbs yet.

Sound Changes in Some Sort of Order
Here are the fast and loose sound changes from late PIE to Lusitanic, assuming normal behavior from the laryngeals.


palatovelars merge with the regular velars
Osthoff's Law applies
p -> kʷ if there is a kʷ in the following syllable
stressed ṃ ṇ -> om on/ C_ if X is /w/ or a labiovelar or labial
stressed ṃ ṇ -> em en otherwise
unstressed ṃ ṇ -> a
stressed ḷ ṛ -> la ra
ḷ ṛ -> el er otherwise

p -> f/ #_V
w -> Ø/ #_
sw -> f/ #_
st -> θ/ #_
w -> v
Ø -> y/#_e
w -> u/#_C, #C_, except that tw becomes θ

g -> y/_[ie]
kʷ gʷ gʷʰ -> k g gʰ /_[ie]

Then, intial:
bʰ dʰ gʰ gʷʰ kʷ gʷ -> b θ h v p b
Medial as above, except that gʷ becomes /v/ instead of /b/. Remaining /w/ between vowels becomes /v/. unless between two back vowels, in which case it deletes.

Then:
ŏ -> ă
ō -> ā if unstressed, otherwise -> o. These two only apply if the /o/ is not part of a diphthong.
ĕw -> ŏ
ŏw ōw -> ō
ēw -> iŏ
ey -> i, ēy -> ī
unstressed ăy -> ĕ, āy -> ē, otherwise bother -> ăy
unstressed ŏy -> ĭ, ōy -> ī, stressed ŏy -> ĕ, ōy -> ē
r -> l /C_
m -> n/ _#
ins uns -> ēs ōs, otherwise Vns -> V:s
d -> Ø/_#

Then there is a vowel shift, turning the five-vowel system at this stage into a seven-vowel system:
ā -> ɔ
stressed ŏ -> ɔ, otherwise ->o
stressed ō -> o, otherwise ->u
ŭ -> o
ū -> u
ī -> i
stressed ĭ in closed syllables -> e, otherwise stays
stressed ē -> e, otherwise -> i
stressed ĕ -> ɛ, otherwise -> e

s -> Ø /#_N
sC -> asC /#_

Then, /b/ and /v/ merged into a single phoneme, which was probably /b/ initially and /v/ or /β/intervocalically.
There are also some minor, sporadic changes, like dissimilation of approximants, on occasion, or occasional syncope. At some point *kt and *pt became θ.

Finally, and to wrap things up, the old accentual system of PIE shifted: any old oxytone was stressed on the first syllable of any inflection, any old barytone was stressed on the antepenult.

That gives us the following phonetic inventory:

Code: Select all

p	t			k
b~v	d			g
m	n
f	θ	s		h
	r		y
	l

i		u
e		o
ɛ		ɔ
	a

I haven't thought much about the writing system, but Greek would be a good alphabet to adapt, and probably the one that would have been adopted, so I'll use it:

Code: Select all

π 	τ			κ
β	δ			γ
μ	ν
ϝ	θ	σ			η
	ρ		ι		
	λ

ι 		υ
ε 		ο
η 		ω 
	α

Though far from ideal, this is probably the adaptation that would have arisen, or is one that could have arisen. NB: digamma, its original value forgotten by the time the Lusitanic speakers met any Greeks, is used as /f/, as indeed it was in Italy. <ɛ> is /e/, not /ɛ/, which is <η>. <η> also serves as an h, as, again, it did in Italy; context will differentiate as /h/ is only found between vowels, where /ɛ/ does not venture. <ι> is used both for /i/ and consonantal /j/. /n/ before /k/ and /g/ is written <γ>. I should get to bed, here, but here are some tasty nibbles of what's to come:

ήνας
δυώ
τρες
κεθάρες
κήγκε
ϝηξ
σεθήν
αθώ
νήβα
δήκα
κατάν

άυις ιήπι κε

βήρω
βήρεσι
βήρετι
βήραμας
βήρετε
βήραντι god those verbal endings look too conservative, but I'm not sure what I'll do with them. anyways, I should get to bed. I'll tackle the verbal system starting after I've finished the nouns, they look to be a bitch because there are so many subconjugations.

[dhok]

Here is an oxytonic o-stem masculine, διβάς, διβί "god":

Code: Select all

	   S		    D	     P
nom	διβάς		διβώ 	 διβί
gen	διβί		 διβώς 	διβών
dat	διβέ		 διβώς 	διβάβας
acc	διβάν		διβώ 	 διβώς
abl	διβώ 		διβώς 	διβάβας
loc	διβή		 διβώς 	διβήσο
voc	δίβε		 διβώ 	 διβί

And a rhizotonic, άγκας, άγκι "hook":

Code: Select all

      S		    D          P
nom	άγκας		άγκο		 άγκι
gen	άγκι		 άγκος		άγκων
dat	άγκι		 άγκος		άγκαβας
acc	άγκαν		άγκο		 άγκως
abl	άγκω		 άγκος		άγκαβας
loc	άγκε		 άγκος		άγκεσο
voc	άγκε		 άγκο		 άγκι

And now I really am getting to bed.

[dhok]

Nouns

Nouns are somewhat more complicated than in Latin or Greek. Lusitanic retains the dual number in full, while Latin lost it and Greek uses it sporadically, as well as retaining seven of late-stage PIE's eight cases to Latin's six (well, OK, seven if you count its half-assed locative) and Greek's five.

The real kicker, however, is the number of declensions. There are basically five declensions in Lusitanic: o-stems, a-stems, i-stems, u-stems, and consonant stems, the last a grab-bag where everything that didn't fit somewhere else got shoved in. However, beware! Each of these declensions split in two depending on whether the stress fell on the stem or the ending in Indo-European, a rhizotonic and an oxytonic declension respectfully. You already saw the o-stem masculines in the post above. It is already helpful before we move on to point out some mergers:

-The vocative is only separate in the singular in o-stem masculines, at least for now. I know Greek has some 3rd-declension vocatives distinguished by accent placement, so there might be a few nouns that aren't o-stems that have a separate vocative, we'll see.

-Nouns only have two forms in the dual: a nominative/accusative/vocative form derived from a PIE form covering the same thing, and an everything else form derived from the PIE genitive/locative dual.

-In the plural, the dative and ablative merge (right now the ending is -βας; Latin and Oscan have descendents of -ois, and the cognate to -βας is the 3rd declension -ibus. Old Irish has -aib, which looks to be from the instrumental plural ending -bʰi, and anyways its dative is a merger of all the other IE cases. This ending may have to change before we're done.)

To round off the o-stems, here is the nominative/accusative/vocative of oxytonic ιογάν, "yoke", and rhizotonic μάραν, "blackberry" (otherwise declined like the masculines):

Code: Select all

S     D    P
ιογάν ιογώ ιογά
μάραν μάρο μάρα

Confusingly, there seems to have been two endings in IE for the nom/acc/voc dual: one ending in i, the other in u, and the book I have nouns that give both, so the etymon of διβώ was *deiwṓw, but the etymon of what I have as ιογώ was *yugṓy, which would have given Lusitanic ιογέ. There doesn't seem to be any way to determine which was which, so I'm going to go with "in o-stems the -u ending generalized and in the a-stems the -i ending generalized". I haven't entirely determined what I'm going to do with the -jo stems, but I suspect they'll just have the ending ιας/ιαν or ιάς/ιάν, with genitive singular/nom. masc. plural -ι/ί.

Right, so now that that's basically taken care of we can move on to the a-stems which, again, have oxytonic and rhizotonic subdeclensions. Thanks to the Lusitanic vowel shift, confusingly, o-stems are marked by a nominative singular with an -α, while a-stems have a recurring -ω. Here's αβικώ "lamb":

Code: Select all

OXYTONIC A-STEM FEMININE "lamb"
      S		     D		     P
nom	αβικώ		 αβικαί		αβικαί
gen	αβικώς		αβικώς		αβικών
dat	αβικαί		αβικώς		αβικώβας
acc	αβικών		αβικαί		αβικώς
abl	αβικώ		 αβικώς		αβικώβας
loc	αβικαί		αβικώς		αβικώσο

A couple of things to note here: the nominative plural has been turned to -αί, as in Latin, by analogy with the masculine plural. The ablative and genitive would otherwise have merged in the singular, so the ablative ending is now -ώ by analogy with the o-stems.

A rhizotonic a-stem, γένω "woman":

Code: Select all

RHIZOTONIC A-STEM FEMININE "woman"
	   S		   D		   P
nom	γένω		γένι		γένι
gen	γένως	  γένος	  γένων	
dat	γένι		γένι		γένωβας
acc	γένων	  γένος	  γένως
abl	γένω		γένος	  γένωβας
loc	γένι		γένι		γένωσο

Again, analogy has applied in similar places to the oxytonics. The distinction between rhizotonic and oxytonic a-stems is not as great as in the o-stems; the only differences are that oxytonic -αί becomes rhizotonic -ι, and the dual oblique is -ος instead of -ώς.

[Dewrad]

Bandwagon-jumper!

I like this, it looks plausible (although I do query *bʰ > /v/ when the reflexes of the other aspirates are voiceless?). The dual was retained in the classical-era Celtic languages, so being retained here is not at all implausible. I think I have some Celtiberian resources if you want them for ideas?

[dhok]

I think the reflex of *bʰ is defensible. Upon further inspection, it may not have been that all the other reflexes were voiceless; they could have been voiced. The aspirates develop as so:

bʰ -> b
dʰ -> θ
gʰ -> h
gʷʰ-> v

It's entirely possible that dʰ developed to ð first, and then became θ, merging with the reflexes of *st, *kt and *pt; as English well attests, interdental fricatives can be a little hazy on voicing distinctions, especially when there is not a well-established voicing distinction in fricatives anyways (/v/ at this stage could have been all sorts of things- [v], [β], [w]?) The same goes for /h/, which could well have been a voiced [ɦ] as in Sanskrit- as I could probably have done a better job of outlining above, /h/ only survives initially and between vowels. I'm not sure what (say) an *ngʰ cluster would have become- probably a deaspirated ng instead.

Also, yeah, shoot whatever Celtiberian stuff you want to donate on over. I'd like to pick your mind on IRC at some point about the verbal system, too.

[dhok]

Some initial work on the u-stems and i-stems before dinner.

The rhizotonic u-stems are characterized by the sudden appearence of a -β- in certain forms (NB: the -β- in the dative/ablative plural is part of the ending proper and does not derive from ablaut). Bottoms up:

Code: Select all

βώηος "elbow", masculine
	   S		     D		     P
nom	βώηος		 βώηυ		  βώηεβες
gen	βώηος 	   βώηυος		βώηεβων
dat	βώηεβι		βώηυος		βώηοβας
acc	βώηον		 βώηυ		  βώηος
abl	βώηυ		  βώηυος		βώηοβας
loc	βώηυι		 βώηυος		βώηοσο
voc	βώηο		  βώηυ		  βώηεβες

Representing the neuters of this class is μήθο, "honey", with its seperate nom/acc/voc forms:

Code: Select all

S	   D	   P
μήθο	μήθυ	μήθοα

Here is an oxytonic u-stem, θινός "bull" (from PIE *dʰeynús):

Code: Select all

      S       D       P
nom   θινός   θινύ    θινήβες
gen   θινώς   θινύος  θινήβων
dat   θινήβι  θινύος  θινóβας
acc   θινόν   θινύ    θινώς
abl   θινύ    θινύος  θινóβας
loc   θινύι   θινύος  θινóσο
voc   θίνο    θινύ    θινήβες

This is similar to the rhizotonic declension, with a couple differences: the accusative plural ending is now -ώς; and the -εβ- that appears in some rhizotonic forms is now -ήβ-. To round this off, here are the nom/acc/voc forms of the neuter oxytone u-stem, ωλό "wild onion". It's cognate to Latin allium "garlic" and Sanskrit ālus and is attested nowhere else, but since I'm having trouble finding neuter oxytone u-stems, I'll throw it in.

Code: Select all

S    D    P
ωλό  ωλύ  ωλόα

On to the i-stems,which do have seperate oxytones and rhiztones, my bad scratch that, because the only difference from the SCA is that the oxytones have -ή/ής in the abl. singular and acc. plural, while the rhizotones have -ε/-ες. Because ή does not seem to do very well on its own if it is unstressed anyways, I'll just leave this as a morphophonological rule that any ending in ή reverts to ε if/when it loses its stress.

Here is an example with ιεγνίς, "fire":

Code: Select all

I-STEM MASCULINE "fire"
      S           D          P
nom	ιεγνίς	   ιεγνί		ιεγνίες
gen	ιεγνίες	  ιεγνίος 	ιεγνίων
dat	ιεγνί		 ιεγνίος 	ιεγνίβας
acc	ιεγνίν		ιεγνί		ιεγνής
abl	ιεγνή		 ιεγνίος 	ιεγνίβας
loc	ιεγνί		 ιεγνίος 	ιεγνίσο
voc	ιέγνι		 ιεγνί		ιεγνίες

An example of a neuter is ώρι, "water". It is, again, different in the nominative, accusative and vocative, which all merge:

Code: Select all

S     D     P
ώρι   ώρι   ώρια

[dhok]

Here are the consonant stems. As in Greek or Latin, these are characterized by a nominative singular form which is different from the stem used for everything else. There are no oxytones or rhizotones; the accent fell on the root in the nominative, vocative and accusative in Pre-Lusitanic, and the ending elsewhere. In some cases, resultingly, the stem is slightly different in the vowel in the accusative, as well. As an example, here is ϝες, ϝεδής, "foot", which is a masculine (feminines inflect the same way):

Code: Select all

      S       D         P
nom   ϝες     ϝέδε      ϝέδες 
gen   ϝεδής   ϝεδώς     ϝεδών
dat   ϝεδί    ϝεδώς     ϝεδήβας
acc   ϝέδα    ϝέδε      ϝέδις
abl   ϝεδή    ϝεδώς     ϝεδήβας
loc   ϝεδί    ϝεδώς     ϝεδήσο

The expected ablative singular would have been ϝεδής, but differentiated itself from the genitive by analogy with the i-stems. There might be one or two other examples of analogical levelling/differentiating here (and of course elsewhere), but I can't remember.

Here is the obligatory neuter νάμα, ναμνής, "name", in the nominative and accusative. (Expected nom/acc singular *νώμα regularized by interference from its other declined forms, where the original *ō shortened due to Osthoff's Law.)

Code: Select all

S       D       P
νάμα    νάμνε   νάμνα

That is about it for the nouns, I think. It's tempting to give *patēr, *matēr etc. their own declension, though. Next up I'll have to work with the verbs, which is going to be a clusterfuck, since it's almost impossible to determine from an IE root what verbs would be derived from it and what they would look like.

(Edit: this is, in fact, what happened; *-tēr nouns form their own declension. Here's the declension of ϝατέρ, ϝατρής, father:)

Code: Select all

        S        D       P
nom     ϝατέρ    ϝατήρε  ϝατήρες 
gen     ϝατρής   ϝατρώς  ϝατρών
dat     ϝατραί   ϝατρώς  ϝατράβας
acc     ϝατήρα   ϝατήρε  ϝατρώς
abl     ϝατρή    ϝατρώς  ϝατράβας
loc     ϝατραί   ϝατρώς  ϝατράσο
voc     ϝάτερ    ϝατήρε  ϝατρώς

Nouns declined like ϝατέρ include μωτέρ, βρωτέρ (the accent has been regularized here; the expected reflex would be **βρώτιρ), ϝήσωρ*, θογατέρ (vocative θογατήρ), and δωτέρ "giver". There are some agent nouns ending in -ώρ instead; these have -ώρ- where ϝατέρ has -ήρ-, but otherwise decline like it.

*ϝήσωρ forms a small subdeclension of its own; in short, the nom. and voc. singular are ϝήσωρ, but the accent follows ϝατέρ for the rest of the forms, so eg acc. singular ϝεσώρα. To top things off, when an -σρ- cluster would be produced, an epenthetic -κ- is inserted, so dative ϝεσκραί.

Finally, there is at least one regular noun, βως, γης "cow". I say "at least one" because I haven't flung myself down the rabbit hole of other nouns that are likely to be irregular, but this one is (I think) some sort of irregular in every IE language I've come across. It has two stems, one (β-) from where the original *gʷ was followed by an o, the other (γ-) from when it was followed by an *e. Some analogical levelling has taken place, particularly in differentiating the genitive from the ablative under influence from the consonant stems, and their change in vowel under influence from the dative and locative; the expected reflex for both would be *γως.

Code: Select all

      S      D     P
nom   βως    βάβε  βάβες
gen   γης    βάβος γήβων
dat   γήβι   βάβος γώβας
acc   βων    βάβε  βως
abl   γη     βάβος γώβας
loc   γήβι   βάβος γώσο
voc   βω     βάβε  βάβες

[dhok]

Moment of restraint! Here are the 1st and 2nd person pronouns:

Code: Select all

      1s      2s     1d     2d     1p      2p      
nom   ιεγώ    τυ     ιε     ιυ     ι       ιυς
gen   μής     τής    αβώς   υβώς   νασκρά  ασκρά
dat   μήια    τήβια  αβώς   υβή    νασμί   υσμί
acc   μη      τη     αβή    υβώς   νως     ως
abl   μη      τη     αβώς   υβώς   νασμί   υσμί
loc   μη      τη     αβώς   υβώς   νασμί   υσμί

Some mergers have taken place. Firstly, note that the locative is always the same as the ablative (and I'm not sure if the locative will ever have a use for these pronouns). 1s and 2s have borrowed their genitive from the consonant declension, as have the duals for their oblique forms. An epenthetic -κ- has appeared in what would otherwise be an -σρ- cluster in the plural genitives (compare Lusitanic νασκρά, Latin nostrum.) There have been some mergers (the ablatives merge with the dative in the plural, just as in nouns, now.) The accusatives will probably have a set of clitic forms με τε αβος υβος νως ως, but I haven't yet determined how these will interact with the verbs. I also haven't figured out where the 3rd person pronouns will come from yet, either.

[Salmoneus]

Seems like everyone's doing IElangs now.

Fortunately for me, none of mine will be on show until the current fad has long since passed...

[dhok]

Since I can't be assed to get on to the adjectives right now (they're long, and mostly repetitive), here are the forms of the definite article, which is always unstressed (probably, maybe there'll be a stressed version too, but it won't look significantly different):

Code: Select all

        MASCULINE:         FEMININE:         NEUTER: 
      S    D     P       S    D    P       S    D     P
nom  τις  τιο   τι    | τω   ται  ται   | τι   τιο   τια
gen  τι   τιος  τιων  | τως  τος  των   | τι   τιος  τιων
dat  τιε  τιος  τιβας | ται  ται  τωβας | τιε  τιος  τιβας
acc  τιν  τιο   τιως  | των  τος  τως   | τι   τιο   τια
abl  τιω  τιος  τιβας | τω   τος  τωβας | τιω  τιος  τιβας
loc  τιε  τιος  τισο  | ται  τος  τωσο  | τιε  τιος  τισο

Again, some analogy has applied- the expected reflex of feminine ται would be τι, but this changed by analogy to the a-stem oxytones. The article will be used in the lexicon (now the first post in the thread) to denote gender.

Here are the third-person pronouns. Although originally these had a stem of either σ- or τ-, the shorter forms have all migrated to σ- to differentiate them from the article.

Code: Select all

     MS     FS      NS     M/ND  FD   MP      FP      NP
nom  σας    σα      σω     σω    σαι  ση      σαι     σα
gen  σια    ιεσώς   σια    σως   σως  τασμών  τασμών  τασμών
dat  τασμέ  τεσιαί  τασμέ  σως   σως  τήβας   σώβας   τήβας 
acc  σαν    σων     σω     σω    σαι  σως     σως     σα
abl  τασμώ  τεσιώ   τασμώ  σως   σως  τήβας   σώβας   τήβας 
loc  τασμί  τεσιαί  τασμί  σως   σως  τήσο    σώσο    τήσο

Demonstratives are diachronically tricky. The first, "this" derives from PIE *ḱis, and is declined pretty much the same way as the definite article, but for the insertion of an -ί- into the feminines, and the restoration of stress:

Code: Select all

        MASCULINE:         FEMININE:         NEUTER: 
      S    D     P       S    D    P       S    D     P
nom  κίς  κίο   κί    | κίω   κίαι  κίαι    | κί   κίο   κία
gen  κί   κίος  κίων  | κίως  κίος  κίων    | κί   κίος  κίων
dat  κίε  κίος  κίβας | κίαι  κίαι  κίωβας  | κίε  κίος  κίβας
acc  κίν  κίο   κίως  | κίων  κίος  κίως    | κί   κίο   κία
abl  κίω  κίος  κίβας | κίω   κίος  κίωβας  | κίω  κίος  κίβας
loc  κίε  κίος  κίσο  | κίαι  κίος  κίωσο   | κίε  κίος  κίσο

"That" derives from *ḱis tāmo, basically "this there"; note that often the -ι- drops out. It is unusual in that the declension all happens inside the noun (because the *tām did not decline):

Code: Select all

          MASCULINE:           FEMININE:                 NEUTER: 
      S      D      P        S      D       P         S      D      P
nom   κιθών  κοτών  κιτών    κωτών  καιτών  καιτών    κιτών  κοτών  κατών
gen   κιτών  κοθών  κωντών   κωθών  κοθών   κωντών    κιτών  κοθών  κωντών
dat   κετών  κοθών  παθών*   καιτών κοθών   παθών‡    κετών  κοθών  παθών
acc   κιντών κοτών  κωθών    κωντών καιτών  κωθών     κιτών  κοτών  κατών
abl   κωτών  κοθών  παθών*   κωτών  κοθών   παθών‡    κωτών  κοθών  παθών
loc   κετών  κοθών  ασκοτών† καιτών κοθών   ασκοτών‡  κετών  κοθών  ασκοτών
* From *ḱiBos tāmo -> kivostām -> kvoθām -> kʷoθām -> poθām.
† From *ḱisu tāmo -> ksutām -> skutām.
‡ By analogy.

This one is screwy enough that I'm not sure I've done enough levelling, although the forms are fairly easily derivable from the other demonstrative, with a few rules (no initial -ι- unless it's the only vowel, -στ- becomes -θ-), and the really weird forms (παθών, ασκοτών) are consistent across all three genders. Maybe in the daughters, if I make any, one or two stems will win out, the final -ών will be reanalyzed as an inflectional ending, and they'll become more "regular".

The relative pronoun πη, πάιας (from *kʷoj) is declined as follows:

Code: Select all

          MASCULINE:       FEMININE:          NEUTER: 
      S     D    P      S    D      P      S      D    P
nom  πη     πω   πι     πω   παι    παι    πα     πω   πάια
gen  πάιας  πως  πάιων  πως  πώιος  πων    πάιας  πως  πάιων
dat  παι    πως  πάβας  παι  πώιος  πώβας  παι    πως  πάβας
acc  παν    πω   πως    πων  παι    πως    πα     πω   πάια
abl  πω     πως  πάβας  πω   πώιος  πώβας  πω     πως  πάβας
loc  πη     πως  πάσο   παι  πώιος  πώσο   πη     πως  πάσο

The interrogative pronoun originally derived from a phrase *kʷis kʷoy, or "who that..." Over the centuries, this has morphed into a prefixed form of the relative pronoun which moved its stress to the prefix: the result is similar to but not quite the same as the relative:

Code: Select all

         MASCULINE:               FEMININE:                 NEUTER: 
     S        D      P        S      D        P        S        D      P
nom  κίπι     κίπο   κίπι     κίπω   κίπαι    κίπαι    κίπα     κίπο   κίπαια
gen  κίπαιας  κίπος  κίπαιων  κίπως  κίπωιος  κίπων    κίπαιας  κίπος  κίπαιων
dat  κίπαι    κίπος  κίπαβας  κίπαι  κίπωιος  κίπωβας  κίπαι    κίπος  κίπαβας
acc  κίπαν    κίπο   κίπως    κίπων  κίπαι    κίπως    κίπα     κίπο   κίπαια
abl  κίπω     κίπος  κίπαβας  κίπω   κίπωιος  κίπωβας  κίπω     κίπος  κίπαβας
loc  κίπι     κίπος  κίπασο   κίπαι  κίπωιος  κίπωσο   κίπι     κίπος  κίπασο

Here is the reflexive pronoun, from Ø-grade ablaut on *sewe-, leading to a stem *sw-. This pronoun does not differentiate by gender.

Code: Select all

    S  D   P
gen ϝι ϝως ϝων
dat ϝι ϝως ϝήβας
acc ϝη ϝα  ϝες
abl ϝε ϝως ϝήβας
loc ϝι ϝως ϝήσο

And, uh. I think that's it for the pronouns. Wow. Maybe I should have an indefinite, too...and some other interrogatives and relatives, like "why" and "how", although most of those will probably be taken up by an associated case.

[dhok]

To wrap up the nouns, which I didn't quite do, here's how to decline a noun from its dictionary form. As in Greek or Latin, happily, the listed forms are the nominative and genitive singular, plus the article to mark gender.

-If the noun is masculine and ends in -άς, -ί, it declines like διβάς. If the stress is on the root, it declines like άγκας. If instead you find are -άν or -αν, it is neuter, and declines like ιογάν or μάραν respectively. All of these are -o stems.

-If the noun is feminine (maybe some masculines, too?) and ends in -ώ, ώς it declines like αβικώ; if it ends in -ω, ως and the stress is on the root, it declines like γένω. Those are the ā-stems.

-If the noun ends in -ός, ώς it declines like θινός; if -ος, -ος with a stressed root, it declines like βώηος; if -ό, -ώς it is neuter and declines like ωλό; if -ο, -ος with a stressed root, like μήθο. All these are u-stems.

-If the noun ends in -ις, -ιες, it declines like ιεγνίς; if it ends in -ι, -ιες it is neuter and declines like ώρι. These are i-stems. Accent placement is inconsequential in i-stems, except in the abl. singular and masc/fem acc. plur, where oxytonic -ή/ής becomes -ε/-ες in rhizotones.

-Finally, if the genitive singular ends in -ής, you are dealing with a consonant stem, declined like ϝεδής, unless the nominative singular ends in -τέρ or -τώρ, in which case you have an r-stem, declined like ϝατέρ. Neuters of this class will decline like νάμα instead. Here, you must use the genitive stem in other cases and numbers; the nominative stem is for citation only and cannot reliably be used to decline the noun in other forms.

After I get back from class, I'll post about the adjectives...

Adjectives

These come in four tasty classes: o/ā-stems, u-stems, i-stems and consonant stems.

o/ā stems are declined like an o-stem noun in the masculine and neuter, and an ā-stem noun in the feminine. You can therefore derive their declension from the nouns already given, but just for completeness, here is the declension of the rhizotone βρήνθας, βρήνθω, βρήνθαν, "blind" (there should be no need to show an oxytone as well). Code dump ahead.

Code: Select all

        MASCULINE                     FEMININE                        NEUTER
    S        D        P          S        D        P          S        D        P
nom βρήνθας  βρήνθο   βρήνθι     βρήνθω   βρήνθι   βρήνθι     βρήνθαν  βρήνθο   βρήνθα
gen βρήνθι   βρήνθος  βρήνθων    βρήνθως  βρήνθος  βρήνθων    βρήνθι   βρήνθος  βρήνθων
dat βρήνθι   βρήνθος  βρήνθαβας  βρήνθι   βρήνθι   βρήνθωβας  βρήνθι   βρήνθος  βρήνθαβας
acc βρήνθαν  βρήνθο   βρήνθως    βρήνθων  βρήνθος  βρήνθως    βρήνθαν  βρήνθο   βρήνθα
abl βρήνθω   βρήνθος  βρήνθαβας  βρήνθω   βρήνθος  βρήνθωβας  βρήνθω   βρήνθος  βρήνθαβας
loc βρήνθε   βρήνθος  βρήνθεσο   βρήνθι   βρήνθι   βρήνθωσο   βρήνθε   βρήνθος  βρήνθεσο
voc βρήνθε   βρήνθο   βρήνθι     βρήνθω   βρήνθι   βρήνθι     βρήνθαν  βρήνθο   βρήνθα

Here is an oxytone u-stem, τενός, τενό "thin". These decline the same way in the masculine and feminine.

Code: Select all

        MASCULINE/FEMININE           NEUTER
      S       D       P        S       D       P
nom   τενός   τενύ    τενήβες  τενό    τενύ    τενόα
gen   τενώς   τενύος  τενήβων  τενώς   τενύος  τενήβων
dat   τενήβι  τενύος  τενóβας  τενήβι  τενύος  τενóβας
acc   τενόν   τενύ    τενώς    τενό    τενύ    τενόα
abl   τενύ    τενύος  τενóβας  τενύ    τενύος  τενóβας
loc   τενύι   τενύος  τενóσο   τενύι   τενύος  τενóσο
voc   τήνο*   τενύ    τενήβες  τενό    τενύ    τενόα
*The stress change in PIE leads to an irregular vocative, as in many u-stems. Maybe this should be τένο by analogy?

Again, a rhizotone example should not be necessary. Here is an i-stem, ώκρις, ώκρι "sharp" ; like the u-stems, the masculine and feminine decline the same way:

Code: Select all

       MASCULINE/FEMININE           NEUTER
     S       D       P        S       D       P
nom  ώκρις   ώκρι    ώκριες   ώκρι    ώκρι    ώκρια
gen  ώκριες  ώκριος  ώκριων   ώκριες  ώκριος  ώκριων
dat  ώκρι    ώκριος  ώκριβας  ώκρι    ώκριος  ώκριβας
acc  ώκριν   ώκρι    ώκρες    ώκρι    ώκρι    ώκρια
abl  ώκρε    ώκριος  ώκριβας  ώκρε    ώκριος  ώκριβας
loc  ώκρι    ώκριος  ώκρισο   ώκρι    ώκριος  ώκρισο
voc  ώκρι    ώκρι    ώκριες   ώκρι    ώκρι    ώκρια

Finally, the consonant stem adjectives. Again, the masculine and feminine decline the same way, and this is, again, identical to the nominal declension; example ϝωρ, ϝωρής "equal".

Code: Select all

        MASCULINE/FEMININE            NEUTER
      S       D         P        S      D      P
nom   ϝωρ     ϝώρε      ϝώρες    ϝωρ    ϝώρε   ϝώρα
gen   ϝωρής   ϝωρώς     ϝωρών    ϝωρής  ϝωρώς  ϝωρών
dat   ϝωρί    ϝωρώς     ϝωρήβας  ϝωρί   ϝωρώς  ϝωρήβας
acc   ϝώρα    ϝώρε      ϝώρις    ϝωρ    ϝώρε   ϝώρα
abl   ϝωρή    ϝωρώς     ϝωρήβας  ϝωρή   ϝωρώς  ϝωρήβας
loc   ϝωρί    ϝωρώς     ϝωρήσο   ϝωρί   ϝωρώς  ϝωρήσο

And I think that's it for the adjectives. I think. The numbers will decline too, at least the first four, but I'll get to those later once I actually know how numbers interact with the case system.

[dhok]

I'm waiting on several books to show up at the university library before I can tackle the verbal system. I'm looking forward to having done the verbal system, of course, but I'm not exactly looking forward to doing it, since the PIE verbal system was enormously complicated and "fixed" by each daughter in its own individual way. So in the meantime, here are some prepositions, and Dewrad kindly sent me PDFs of a pair of otherwise three-hundred-dollar etymological dictionaries of Italic and Celtic, as well as a PDF on Celtiberian, so I can also work on increasing the vocabulary. I'm thinking about importing a few loans from Punic as well; for example, the Lusitanic word for "reader" is καρτώρ, from Punic *kar-. (The word for "to read" is from the same root, but since I have yet to work the verbs out, I don't have a standard citation form for my verbs. I'm now updating the lexicon into a Google doc with etymology.

[Drydic]

I'm waiting on several books to show up at the university library before I can tackle the verbal system. I'm looking forward to having done the verbal system, of course, but I'm not exactly looking forward to doing it, since the PIE verbal system was enormously complicated and "fixed" by each daughter in its own individual way. So in the meantime, here are some prepositions, and Dewrad kindly sent me PDFs of a pair of otherwise three-hundred-dollar etymological dictionaries of Italic and Celtic, as well as a PDF on Celtiberian, so I can also work on increasing the vocabulary. I'm thinking about importing a few loans from Punic as well; for example, the Lusitanic word for "reader" is καρτώρ, from Punic *kar-. (The word for "to read" is from the same root, but since I have yet to work the verbs out, I don't have a standard citation form for my verbs. I'm now updating the lexicon into a Google doc with etymology.

You forgot the prepositions...

[Salmoneus]

I'm waiting on several books to show up at the university library before I can tackle the verbal system. I'm looking forward to having done the verbal system, of course, but I'm not exactly looking forward to doing it, since the PIE verbal system was enormously complicated and "fixed" by each daughter in its own individual way. So in the meantime, here are some prepositions, and Dewrad kindly sent me PDFs of a pair of otherwise three-hundred-dollar etymological dictionaries of Italic and Celtic, as well as a PDF on Celtiberian, so I can also work on increasing the vocabulary. I'm thinking about importing a few loans from Punic as well; for example, the Lusitanic word for "reader" is καρτώρ, from Punic *kar-. (The word for "to read" is from the same root, but since I have yet to work the verbs out, I don't have a standard citation form for my verbs. I'm now updating the lexicon into a Google doc with etymology.

De Vaan and Matasovic? De Vaan is fantastic.

I don't suppose Dewrad also has the rest of the set, by any chance? I've got Hittite and Armenian, but not Germanic, Balto-Slavic, Tocharian (is there a Tocharian one?), Albanian, Indo-Iranian, or, more importantly, Greek.

(Kloekhorst's Hittite is considerably less concise and readable than De Vaan or Matasovic (though partly I'm biased against him due to his less appealing fonts and formats (when every word in the lexicon is followed by a page of "is-ki-ia-iz-zi (KUB 9.31 ii 36 (MH/NS))" and so forth it's a little harder on the eyes), and the Armenian one is really annoying. Though partly that's because it's about Armenian, so it's all "hec - probably reflects 'kmth2-2k'-ous-ie-mh1n-os', or possibly 'h2ot'". All PIE words seem to end up as 'hec', 'azr' or 'egl' in Armenian.

De Vaan and Matasovic also both have clearly laid-out sound-changes, which helps. They're contradictory sound-changes, admittedly, at the Italo-Celtic stage, but it's better than nothing. The Armenian dictionary instead goes through the letters one by one, basically saying that anything can turn into anything.