Lusitanic Scratchpad (Now Playing Elsewhere)

[dhok]

Whoop! I forgot to make the lexicon public. That's fixed, now.

Also, yes, I forgot the prepositions. That is because I posted that late last night and wanted to go to sleep. There's a handful of them in the lexicon, now. Note that the one-syllable ones went through the sound changes as if they carried no stress.

δε- "from, away from", cognate to Latin de, OIr di^H. Takes the ablative.
δα- "to, towards", cognate to OCS, OIr do^H. Takes the dative.
α- also "to, towards", cognate to Latin, OIr ad; takes the dative. I'm not sure how to make this seperate from δα- perhaps δα is temporal or metaphorical while α is spatial?
καν- comitative, cognate to Latin cum, OIr co^N, OCS sŭ; takes the ablative.
άθυι- this is equivalent to French chez, and derives from the locative of *h₂wóstu, "house", not otherwise attested. Takes the genitive.
ιεν- "in, inside, into"- takes the locative when describing a place inside, and the accusative when describing a place into.
ιεντήρ- "between, among". Takes the accusative.
ιεξ- "out of". Takes the ablative.

Comparison of Adjectives

I'm not quite sure how to go about this-PIE had several suffixes, at least for the comparative, because PIE seems to have had several suffixes, and Italic and Celtic seem to disagree. I think the Latin stems in -r- probably come from a PIE -is-, which fits with Gothic -iz-, some Greek adjectives with -i- (a former -s- having dropped out), and OCS -jis-. Therefore:

Code: Select all

positive    comparative    superlative
βρανθάς     βρανθισάς      *βρανθιθάς

This holds for all adjectives, I think. Looking at that βρανθιθάς up there, I need to finalize a synchronic rule for dissmilation of adjacent /θ/- the first one should change to -σ-, but the nasal will change it to -τ-. So there's now a synchronic rule that the first θ in a series of two will become σ, unless followed by a nasal, which changes it to τ: βραντιθάς. Comparatives use the format "X is more Y from Z" (with Z in the ablative):

At this point, here's the present indicative conjugation of whatever the infinitive is for "to be". I haven't worked the verbs out yet, but they're quirky enough in every IE language that I'm not worried about this interfering with other verbs. For now, the final -i in many verb forms is just going to drop off.

Code: Select all

  S     D     P
1 ιήσα  ϝας   ασμάς 
2 ιης   θως   θη
3 ιητ*  θης   σαν†
*From earlier *ιηθ.
†Final -ντ reduces to simple -ν.

This allows us to form our first sentences. For example:
άβιες τενισήβες γώβας σαντ.
sheep-nom.plur thin-comp.m/f.nom.plur cow-abl/dat.plur be-pres.ind.3p
"The sheep are thinner than the cows."

[dhok]

A few more sound changes to note before we move on. I can't be assed to go back through this thread and change too much, because I don't think there are very many morphological changes derived from the following changes, though there are definitely a lot of lexical ones, and these additional changes, I think, make Lusitanic look more like its own branch of IE, and not just something that looks like a close relative of Latin or Greek. They'll definitely be included in the grammar, which I'll start writing after or even as I work the verbs on out (coming soon to a scratchpad near you). I'm going to try and get that done well but quickly, because I want Lusitanic to be a candidate in this year's ZBB Awards.

-Initial /dw/ becomes /θ/ via /ð/. Depending on how unbalanced the lexicon ends up being in favor of /θ/, this may change later. (Right now it's mostly feeling unbalanced in favor of /b~v/, which is unsurprising considering the many sources the phoneme has. Also, I've gone through all of Goatman's lexicon's *bʰ words combing it for nouns.)

-θ's dissimilate in adjacent syllables: the first becomes /s/, or /t/ if after /n/. So σήθι "cheese" was originally *θήθι. The same rule applies with ϝ, which becomes /s/, or /θ/ after /n/. These two rules stay in the grammar as active.

-Lusitanic has no /l/. Initial /l/, or /l/ after a consonant, became /r/ early on (but after the syllabic sonorants did their thing); otherwise, it became /w/. So Greek λευκός, OIr lōch, but Lusitanic ροκάς.

-Initial and medial -pt-, -sp- and -pt- now become ϝ, as does PIE *p after a sonorant, medially- this happened before the -l- changes. The labiovelars were unaffected by this change (they merged into other plosives later). So PIE *wobʰsā́ -> Lusitanic αϝώ. -pn- became -wn-. All this means that /f/ is no longer found only initially.

-PIE *ḱ and *k palatalized to *ts -> σ before /i e/ when /g/ became /j/. As a general rule, this generalized in consonant stem nouns and i-stems. Unfortunately, this leaves us with some overlap between the near demonstrative σις, σίω, σι and the 3rd person pronouns σας, σα, σω. They might not get rectified- in a lot of languages with healthy person marking on verbs, pronouns overlap comfortably with demonstratives, anyways. But I'll probably do some tweaking.

-The clusters -nr- and -mr- become -br- and -dr-.

-The clusters /ks/ and /sk/ become /h/ via /x/, except /ks/ survives word-finally. The cluster /sr/ becomes /spr/ -> /fr/.

-Clusters where a nasal comes after a stop or fricative switch places early on.

-Syncope between consonants of unstressed vowels in nouns with a nom. form of over three syllables, so *swéḱuros became ϝήκρας, not **ϝήκορας. This did not apply in words where the extra syllables derived from productive compounding or derivational or inflectional morphology, unless the morphology had previously fossilized- this isn't Old Irish. This happened relatively late, so *dʰugʰatḗr to θοκτέρ via *θογατέρ, not **σοθέρ. (Resulting clusters that disagreed in voicing assimilated to the final consonant, as we see here.) /h/ dropped, but not before lowering preceding /e o/ to /ɛ ɔ/ (though, of course, that /e/ stayed /e/ if it was unstressed, since /ɛ/ does not occur in unstressed environments in Lusitanic).

There are also as of this writing 160 words in the lexicon, none of which, yet, are verbs (though all the pronouns get their own entry). Whee. I'm going to shoot for at least 750 words by the time the grammar is out, maybe more like a thousand (nice round number, and once the verbs start tumbling down it shouldn't be that hard to get a lot of derivations from various roots); work on it should start soon, as I've ordered and checked out some new books that should allow me to start work on the verbal system.

There's one more thing: I'm at university at the moment, and they watch downloads like a hawk. So if you have any PDFs that you think might be useful (or even ones that wouldn't particularly be, but want to donate to my hard drive out of the goodness of your heart), feel free to send my e-mail a copy.

[communistplot]

IElangs for everybody! That said this looks pretty good.

[Chagen]

This looks quite interesting. What books/PDFs are you using as resources on PIE?

[dhok]

Verbs

*ominous drumming noises*

OK, so although none of the books I was looking for came in ILL today, I decided it was time to start work on the verbal system, anyways. This is going to be a pretty complicated process, because the PIE verbal system was, to say the least, kind of baroque. Luckily, there are some things that are already decided for me since I'm making a branch of Italo-Celtic: the passive was formed with an -r suffix, there was one irrealis mood derived from a PIE optative in -ā-, there is a past participle (and derivational process) in -t-, and the aorist and perfect of PIE are not differentiated. The resulting preterite is likely to be formed much of the time with intial reduplication or a suffix in -s-, and on occasion apophony, but not through the addition of an augment, which is a Greco-Aryan innovation. Additionally, certain verbal forms are analogous to those in other IE languages but would have to have been innovated anew, such as the imperfect and future. I am also thinking of innovating a present progressive tense, possibly from "to be", more likely from *stā "to stand". Here's what I think we're going to be looking at in terms of verbal forms:

Code: Select all

             ind. active  subj. active  imper. active  ind. passive  subj. passive  imper. passive 
present      yes          yes           yes            yes           yes            yes
progressive  yes          yes           no             yes           yes            no       
preterite    yes          yes           no             yes           yes            no
imperfect    yes          maybe?        no             yes           maybe?         no
pluperfect   yes          maybe?        no             yes           maybe?         no
future       yes          no            maybe?         yes           no             unlikely

This is in addition to a number (not sure how many) of non-finite verb forms, including participles and at least one infinitive (probably...) There are already some suffixes that suggest themselves to derivational use as well; for example, PIE *-tós with zero-grade ablaut in the stem was used to form many a noun in the sister languages, as it does in Lusitanic. It also forms the past participle with a stress and ablaut shift (the ablaut shift being an innovation), thus creating doublets like οπτάς "speech" and ιεπάτας "said" from *wekʷ "to say", and...I'm completely getting ahead of myself and starting to ramble.

Now comes the tricky bit. PIE had a fuckton of verb classes, and shoehorning them into anything like a nice conjugational system like Latin has will require a good bit of shuffling around here. I'm going to bop around with a couple ideas here for a few days. Luckily I've finally got a few books that should allow me to start in-depth work, specifically Szereményi's Introduction to Indo-European Linguistics, and, διβήσο βεβεντήσο, the ones I ordered in specially (Kortlandt, Ringe, Stifter, and Baldi) should appear later this week. I'm also drawing on Dewrad's sketch of Gallo-Brittonic. As he says himself, it's not perfect, but frankly if Lusitanian passes Dew's approval, then I'll probably be happy. As I said, though, I'm starting to ramble and it's now becoming clear that I'm actually kind of out of my depth until more books arrive, and also it's midnight, so I'll leave, make some derivative nouns from -tós, and call it a night. It's also clear that I've been distinctly sloppy dealing with the laryngeals; that wasn't much of a problem with the nouns, but it is a big one with the verbs. So I'll be back tomorrow.

[dhok]

Here are the first few posts on the verbs. I have introduced one final sound change early in Lusitanic's history, namely that nasals before an aspirate become length on the preceding vowel, in a change akin to the Ingvaeonic Spirant Law, but with aspirates. I'll update the sound changes accordingly and push the lexicon through again later. Meanwhile, here's what I'm looking to do with the verbal system.

Basically, the verb has two principle parts, at this point, a present system form and a perfect system form. Of the tenses named, the present, progressive, and imperfect will derive from the present stem, while the preterite and pluperfect will derive from the perfect stem. It is very likely that I will have a future and probably a future perfect, too, although the jury is still out on how that will be formed The idea is that there was a good bit of analogic shuffling early on in Lusitanic's history that allowed this sort of regularity. There may be other principle parts, too, I don't know. (EDIT: the future stem, at least, will comprise a principle part.) There will definitely be at least two thematic-vowel conjugations, one where the present indicative thematic vowel derived from PIE *e or *o, and one, cognate to Latin 1st conjugation verbs like amāre and weak Old Irish verbs like caraid, where the thematic vowel was uniformly *a/ā. I've got most of the system more or less sketched out, with the exception of the infinitive, which I may just conjure out of thin air- the other IE languages did that, so why can't I?- and the future. Expect some reshuffling of this page as new books arrive and new developments come to light.

So let's look at a few paradigms. For now, I can only work with the present system, but luckily there's a good bit of diversity here. Right now, I am only worrying about three classes of verb: denominatives in *ā, oxytones in *e/o, and rhizotones in *e/o. There are likely to be several irregulars (the only one I have yet worked out is δήθω, δάσω, perfect unknown, "to give".

So first we'll look at a denominative verb, because they're the most regular ones. As in Latin and Old Irish, the thematic vowel is -ā throughout. δήθω acts like this, but, diachronically, for the wrong reasons, so I'll choose something else. (Lusitanic has a principle part system like Greek, rather than a cut-and-dried conjugation system like Latin). These all follow the same stress pattern. Our example will be ϝονώ, σίψασω, "to dream", from *swopniom. (I might decide to "regularize" its future into something more like σίϝυσω...) Here's the present indicative active:

Code: Select all

  S      D       P
1 ϝονώ   ϝονώβες ϝονώμες
2 ϝονώς  ϝονώτας ϝονώτε
3 ϝονώτ  ϝονώτας ϝονών

The stem is ϝον-, as shown, and the endings all have a uniform thematic -ώ-, from -ā-.

Here's the present progressive indicative active:

Code: Select all

  S        D         P
1 ϝονώθω   ϝονώθαβες ϝονώθαμες
2 ϝονώθας  ϝονώθατας ϝονώθατε
3 ϝονώθατ  ϝονώθατας ϝονώθαν

What this appears to be is a suffix -θα- between the vowel and the ending. In point of fact this derives from the zero-grade derivative of *steh₂ attached to the verb.

Here's the imperfect. It was clear that it had to be re-derived, since the reflexses PIE secondary endings would be largely indistinguishable from those of the primary endings. Latin derives its imperfect from a reflex of *bhew-; it is unclear where the Old Irish imperfect comes from. Since this verb fossilized in several other IE languages, that's what will be used. (In this case, we derive from an unstressed, 0-grade form *bhu-, with the secondary endings):

Code: Select all

  S        D         P
1 ϝονώβον  ϝονώβυς   ϝονώβομες
2 ϝονώβος  ϝονώβοτας ϝονώβοτε
3 ϝονώβοτ  ϝονώβοτας ϝονώβον

Note that the 1st singular and 3rd plural are identical. The 1st dual is a slightly irregular derivation from *bhuwes -> bhūes -> bhūs.

The future is derived, usually, similarly as in Celtic: a reduplicated stem with an -s- suffix. I'm not sure whether or not all verbs will have the reduplication, though all will have the -s-. The endings are in -e/o, which will reappear when I get to the non-denominative verbs. Because the future stem is so often underivable from the present, or irregularly so, it is given as a principle part. The stress is always on the reduplicated vowel. I'm not sure whether that's entirely accurate, but that's what Sanskrit does, so I'm going to run with it unless there's a reason not to.

Code: Select all

  S        D            P
1 σίψασω   σίψασαβες    σίψασαμες 
2 σίψασες  σίψασετας    σίψασετε
3 σίψασετ  σίψασετας    σίψασαν

Should I make a future progressive? OK, I'll make one. The PIE root *sth₂ (zero-grade, remember) reduplicates to *sisth₂s, which becomes an ending -σιθασ-, attached to the present (not future!) stem. This later shortens, and becomes simple -σιθ-, with the aforementioned endings:

Code: Select all

  S         D           P
1 ϝονώσιθω  ϝονώσιθαβες ϝονώσιθαμες 
2 ϝονώσιθες ϝονώσιθετας ϝονώσιθετε
3 ϝονώσιθετ ϝονώσιθετας ϝονώσιθαν

I might make that ending shorten even more to something like -στ- due to the syncope rule...or, the syncope doesn't apply twice, but applies to the first vowel, and becomes -στασ-. Not sure which.

That's it for now. There's still loads and loads left to do: the perfect system, the passive voice, the subjunctive, and verbs that aren't nice and neat like this one is.

[dhok]

Now onto the everything else club. There are two conjugations here, the rhizotonic and the ozytonic ones, and they differ slightly. (I think. I'm honestly not as knowledgeable about PIE accent placement as I might be.)

Here's the present indicative active for a rhizotonic, βήρω, "to bring, bear, carry":

Code: Select all

  S      D        P
1 βήρω   βήραβες  βήραμες
2 βήρες  βήρετας  βήρετε
3 βήρετ  βήρετας  βήραν

The present progressive indicative active is formed more or less the same way as in the ā-stems, with the thematic vowel being uniformly ε:

Code: Select all

  S        D        P
1 βήρεθω  βήρεθαβες βήρεθαμες
2 βήρεθας βήρεθατας βήρεθατε
3 βήρεθατ βήρεθατας βήρεθαν

The imperfect is identical to the ā-stem imperfect in the same way as in the present progressive: the thematic vowel is ε, but the endings are the same.

The future is derived the same way as in the ā-stems, so I shan't repeat it here. Likewise with the future progressive.

[hwhatting]

Quite interesting. I personally don't believe in Italo-Celtic as a node, but I like what you're doing here.
Question - you mention a "PIE optative in -ā-" - did you find anyone who reconstructs something like that for PIE? I always thought that kind of thing was a post-PIE development in some European IE languages.

[dhok]

Quite interesting. I personally don't believe in Italo-Celtic as a node, but I like what you're doing here.
Question - you mention a "PIE optative in -ā-" - did you find anyone who reconstructs something like that for PIE? I always thought that kind of thing was a post-PIE development in some European IE languages.

It's a post-PIE form found in Italic and Celtic, and is the source of Latin's subjunctives in all but the first conjugation (habeat, regat, audiat) and most of Old Irish's subjunctives (which had some wacky intersections with the perfect I neither completely understand nor will attempt to replicate). Latin's denominative conjugation (the ā-stems) used e in the subjunctive instead, and I don't know why. Old Irish's denominatives cognate to Latin's ā-stems can use the a-subjunctive, too, so this wasn't a development at the Italo-Celtic level. What I really need here is a copy of The Foundations of Latin by Baldi, but I don't have it because whatever library mine requested it from is being a little bitch and doesn't want to ship it. (Upon thinking, it might come from the same -oy- that the Greek and Sanskrit optatives come from...unless I find out otherwise, I think that's the assumption I'm going to make, for now. Let's see if I can double-check this assumption...JSTOR, ho!)

[hwhatting]

Quite interesting. I personally don't believe in Italo-Celtic as a node, but I like what you're doing here.
Question - you mention a "PIE optative in -ā-" - did you find anyone who reconstructs something like that for PIE? I always thought that kind of thing was a post-PIE development in some European IE languages.

It's a post-PIE form found in Italic and Celtic, and is the source of Latin's subjunctives in all but the first conjugation (habeat, regat, audiat) and most of Old Irish's subjunctives (which had some wacky intersections with the perfect I neither completely understand nor will attempt to replicate).

That's what I know, so I was astonished at your calling it a "PIE optative" - it's clearly a later development.

Latin's denominative conjugation (the ā-stems) used e in the subjunctive instead, and I don't know why.

Cause otherwise, with the deletion of intervocalic -y- and the abolition of the distinction between primary and secondary endings, the subjunctive and the indicative would have become indistinguishable, except for the 1st Sg?

Old Irish's denominatives cognate to Latin's ā-stems can use the a-subjunctive, too, so this wasn't a development at the Italo-Celtic level.

Well, as I said, I don't think that ever was a node and all developments common between both branches are areal features IMO. So no problem here.

(Upon thinking, it might come from the same -oy- that the Greek and Sanskrit optatives come from...unless I find out otherwise, I think that's the assumption I'm going to make, for now. Let's see if I can double-check this assumption...JSTOR, ho!)

One explanation I've seen is that the optatives in /e:/ and /a:/ are imperfects (comparable to the /e:/ and /a:/ imperfects in Baltic) that became used as subjunctives. So the assumed development is that forms secondarily derived with "stative" /e:/ and (originally) denominal /a:/ were re-interpreted first as imperfects, and then these imperfects were repurposed as subjunctives. I think I read that explanation in Pisani's "Manuale storico della lingua latina", but I read that 25 years ago and maybe it was elsewhere. Other possible origins of the subjunctive in /e:/ are:
1) the subjunctive of thematic stems (originally with lengthened thematic vowel, i.e. /e:/-/o:/), which is also often named as the origin of the future tense in /e:/ (e.g. faciem, legem), although I'm sceptical with regards to that
2) the athematic forms of the PIE optative in *-yeH₁- / -iH₁-, where a form like Proto-Latin *-aye:m would quite naturally have been contracted into -e:m. Thematic -oyH₁- cannot be the origin, as it wouldn't have resulted in Latin -e:- (/oi/ gives /u:/ or /i:/, a contracted form would probably have given /ae/).

[dhok]

Quite interesting. I personally don't believe in Italo-Celtic as a node, but I like what you're doing here.
Question - you mention a "PIE optative in -ā-" - did you find anyone who reconstructs something like that for PIE? I always thought that kind of thing was a post-PIE development in some European IE languages.

It's a post-PIE form found in Italic and Celtic, and is the source of Latin's subjunctives in all but the first conjugation (habeat, regat, audiat) and most of Old Irish's subjunctives (which had some wacky intersections with the perfect I neither completely understand nor will attempt to replicate).

That's what I know, so I was astonished at your calling it a "PIE optative" - it's clearly a later development.

Yeah, I think I goofed here. It's an Italo-Celtic form (Italic and Celtic shared feature as per you), and it's cognate to certain optatives elsewhere, I think I may have read (hence my calling it a PIE form...) but I can't recall which ones and I may be misremembering.

Latin's denominative conjugation (the ā-stems) used e in the subjunctive instead, and I don't know why.

Cause otherwise, with the deletion of intervocalic -y- and the abolition of the distinction between primary and secondary endings, the subjunctive and the indicative would have become indistinguishable, except for the 1st Sg?

Right, that's clear, but the question is where they came from.

(Upon thinking, it might come from the same -oy- that the Greek and Sanskrit optatives come from...unless I find out otherwise, I think that's the assumption I'm going to make, for now. Let's see if I can double-check this assumption...JSTOR, ho!)

One explanation I've seen is that the optatives in /e:/ and /a:/ are imperfects (comparable to the /e:/ and /a:/ imperfects in Baltic) that became used as subjunctives. So the assumed development is that forms secondarily derived with "stative" /e:/ and (originally) denominal /a:/ were re-interpreted first as imperfects, and then these imperfects were repurposed as subjunctives. I think I read that explanation in Pisani's "Manuale storico della lingua latina", but I read that 25 years ago and maybe it was elsewhere. Other possible origins of the subjunctive in /e:/ are:
1) the subjunctive of thematic stems (originally with lengthened thematic vowel, i.e. /e:/-/o:/), which is also often named as the origin of the future tense in /e:/ (e.g. faciem, legem), although I'm sceptical with regards to that
2) the athematic forms of the PIE optative in *-yeH₁- / -iH₁-, where a form like Proto-Latin *-aye:m would quite naturally have been contracted into -e:m. Thematic -oyH₁- cannot be the origin, as it wouldn't have resulted in Latin -e:- (/oi/ gives /u:/ or /i:/, a contracted form would probably have given /ae/).

OK. Yeah, maybe treating the a-stems as athematics in the subjunctive and adding the athematic subjunctive -yeh₁- or -ih₁- would be a workable solution. That looks like it would give a Pre-Proto-Lusitanic thematic vowel a:y.

One thing that continues to bug me is accent placement...it's very difficult to determine where the stress went in which sorts of verbs, and how, if at all, the stress changed throughout the paradigm. I need more sources

[dhok]

Here's the subjunctive for ā-stems. A present, present progressive and imperfect form are extant; maybe I'll add an innovated future form later, too... the vowel in Pre-Proto-Lusitanic was -āy-, which was the thematic vowel in the present and imperfect; in the progressive it migrated to the infixed -steh₂-.
Present:

Code: Select all

  S      D        P
1 ϝοναίν ϝοναίβες ϝοναίμες
2 ϝοναίς ϝοναίτας ϝοναίτε
3 ϝοναίτ ϝοναίτας ϝοναίν

Present Progressive:

Code: Select all

  S        D         P
1 ϝονώθε   ϝονώθεβες ϝονώθεμες
2 ϝονώθες  ϝονώθετας ϝονώθετε
3 ϝονώθετ  ϝονώθετας ϝονώθεν

Imperfect:

Code: Select all

  S         D         P
1 ϝοναίβον ϝοναίβυς   ϝοναίβομες
2 ϝοναίβος ϝοναίβοτας ϝοναίβοτε
3 ϝοναίβοτ ϝοναίβοτας ϝοναίβον

Next up I'll try to get the non-ā-stem verbs. I might have to rework that paradigm of βήρω I have up there...