What annoys me about encyclopedia articles

I'm aware that no article is perfect, and I've got poorly written ideas of my own. Though its frustrating running into encyclopedia and other articles on subjects, that are poorly sourced, poorly written, and reflect the writer's lack of knowledge (while pretending to know more than they do).

Looking through the Encyclopedia Britannica's article on the Baltic Languages, in the section on their developement, we find this:

"By the middle of the 1st millennium BC, the Proto-Baltic area was already sharply split into dialects. From the middle of the 1st millennium AD, the Baltic language area began to shrink considerably; at that time the greater part of Baltic territory, the eastern part, began to be inhabited by Slavs migrating from the south. The Balts there were gradually assimilated by the Slavs; complete assimilation probably occurred around the 14th century. One of these Baltic tribes, the Galindians (Goljadĭ), is mentioned in a chronicle as late as the 12th century. The protolanguage of the so-called Eastern Balts split into Lithuanian and Latvian (Latgalian) around the 7th century. The other languages of the so-called Eastern Balts became separated probably at the same time. Selonian and Semigallian could have been transitional languages between Lithuanian and Latvian. Only Curonian, which some consider to be a transitional language between East and West Baltic, might have developed somewhat earlier. Moreover, the name of the Curonians occurs in historical sources earlier (ad 853: Latin Cori) than the names of the other tribes of the so-called Eastern Balts."

Chunky paragraph aside, this is really sloppy and hard to follow. It looks like someone published their pre-published notes as the completed product.

Another problem, is that they don't state their evidence for ProtoBaltic as splitting from ProtoSlavic (assuming it did, and that its not an areal family) during the 2nd millenium BCE, nor that it was split into 2 dialects by the 1st millenium CE. I'm not saying that it wasn't, but can't you point to the evidence?

I guess this wouldn't be a problem if half the internet didn't parrot the same  wordy substance-light article, making an already obscure subject like historical linguistics, even harder to learn more about.

The way that I would have written this, would be to split it into at least three paragraphs. One that dealt with the language family as a whole, and the period before the split. Then the next paragraphs about each of the  Baltic language family groups individually.

North Picene: An Analysis and a possible interpretation of parts

Starting with the Novilara Stele, let's highlight both possible endings, stems, and affixes. Note, I'm starting from the educated guess that this is, in fact, an IndoEuropean language, although this isn't yet proven.

Suggested roots will be in black, stems/affixes in (mainly) red, and endings in blue.

The Novilara Stele

mimniś erút gaareśtadeś

rotnem úvlin partenúś

polem iśairon tet

śút tratneši krúviś

tenag trút ipiem rotneš

lútúiś θalú iśperion vúl

teś rotem teú aiten tašúr

śoter merpon kalatne

niś vilatoś paten arn

úiś baleśtenag andś et

šút iakút treten teletaú

nem polem tišú śotriś eúś

While I don't particularly think that all of these represent how you could break this down, it does demonstrate the point that, structurally, it does seem IndoEuropeanish. Also, other than -ag and -p-, all of the endings do look like stems and endings that have other IndoEuropean counterparts.

The following fragment, I will skip, due to the difficulties that would occur in analyzing it, at least with what I know of it:

1^st Fragment

úpeś
mresveat

This next fragment is analyzeable:

2^nd Fragment

pa śatigot
kešoteri
amdet : nk
------ k --------

Other than the suffixes -ti- and -ter- (if that's how they're divided), and endings -ot and -i, this looks noteably less indoeuropeanish than the main stele.

And finally, the last text:

3^rd Fragment

tiperašθe raiup bav---
-----ipš---------------------

Counting this as indoeuropean, I've basically gone on the assumption that -a- might have been from e+Laryngeal>a. Though other possibilities exist.

As you can see, the fragments seem noteably less indoeuropean than the maintext. Whether this is due to the fragmentary nature of these short inscriptions, or due to this not being indoeuropean, is something that can be explored further.

Now that we've looked at how this could be analyzed, let's take a second look, shall we?

It seems that the longer words may be verbs.

The Novilara Stele (A closer look)

1. mimniś erút gaareśtadeś

2. rotnem úvlin partenúś

3. polem iśairon tet

4. śút tratneši krúviś

5. tenag trút ipiem rotneš

6. lútúiś θalú iśperion vúl

7. teś rotem teú aiten tašúr

8. śoter merpon kalatne

9. niś vilatoś paten arn

10. úiś baleśtenag andś et

11. šút iakút treten teletaú

12. nem polem tišú śotriś eúś

I find the end of line 2 and beginning of line 3 interesting. Last word in line 2 resembles vulgar latin "(nos) partimos ~ (nos) part(y)amos" or 1^st person present plural indicative~Subjunctive verb form of partire "to depart, leave".

Line 3 starts with what appears to be the Accusative singular form of the greek word polis "city", or in this case pol-em:

2. (...) partenúś

3. polem (...)

Perhaps, this part may be translated as:

2. (...) we left

3. the city (...)

The following word is less clear in meaning, but tentatively, we can view the -on as the genitive plural ending.

Iśairon

So, we have:

"(...) we left the city of Isair (...)"

If this is a proper interpretation of this section, this doesn't mean that this is an Italic nor Greek language that we're dealing with here. These could, in fact, be derived from words from those families, if not being direct loans themselves. In much the same way as terms like "ad hoc" and deja vu" from Latin and French are left untranslated, yet, form meaningful phrases in Modern English.

Lines 9 and 10 seem to have a plural personal pronoun + Verb combination:

niś vilatoś

úiś baleśtenag

Niś appears to be in the first person plural pronoun nes (Nominative nos) in the accusative case, and úiś appears to be the 2^nd person plural pronoun wes (Nominative wos) in the accusative case. The nes~nos and wes~wos coming from ProtoIndoEuropean. The only problem with this, is that they would be enclitic forms.

The verb balestenag can be interpreted as having the elements -est-en-ag in it. The ending -ag appears in multiple places, though what an IndoEuropean equivalent would be is wanting. Though the -este- ~ -est- ~ -ste- could be a realization of -st-, which may have come from the PIE suffix -skʲ-. This same cluster can be found in line one with gaarestades; which could be analyzed as gaare-sta-d-es (among other interpretations).

The above seems to have a perfect suffix -d- and a perfect/past 2^nd person singular ending -es, of which I'm not sure is supposed to occur in IndoEuropean. Or, perhaps, -des is meant, which may have come from the end person plural ending -tes (or perhaps, -ates). The later, though would require -t->-d-, seems more likely.

In line 6, iśperion seems to be conjugated like the greek name hyperion, but I'm not clear on what the root isper would represent.

Soter~sotri seems to be a greek loanword "savior", which can be an epithet for gods, heros, and kings.

For a broader look, to wrap this up for now, would give the following for case endings:

Nominative or Genetive singular, or Nominative and/or Accusative Plural (possibly all of the above) -(e/i)s

Accusative Singular -(e)m

Genetive Plural -(e)n

North Picene: What remains of a 2500 year old language

North Picene is an unclassified language from Northeastern Italy, preserved on stelai fragments from the 800-650 BCE period. A quick look at the best preserved inscription, the Novilara Stele, suggests that the language might be IndoEuropean. Other proposals have also been suggested, but this is the one we're to look at today.

The Novilara Stele (6^th Century BCE)

mimniś erút gaareśtadeś

rotnem úvlin partenúś

polem iśairon tet

śút tratneši krúviś

tenag trút ipiem rotneš

lútúiś θalú iśperion vúl

teś rotem teú aiten tašúr

śoter merpon kalatne

niś vilatoś paten arn

úiś baleśtenag andś et

šút iakút treten teletaú

nem polem tišú śotriś eúś

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

http://www.omniglot.com/writing/olditalic.htm#spicene

For an image of this stele:

http://www.maravot.com/novilara_stele.jpg

And according to this post in this forum:

http://s1.zetaboards.com/anthroscape/topic/1100195/1/

mimniś erút gaareśtadeś
rotneš úvlin parten úś
polem iśairon tet
šút tratneši krúš
tenag trút ipiem rotneš
lútúiś θalú iśperion vúl
teś rotem teú aiten tašúr
śoter merpon kalatne
niś vilatoś paten arn
úiś baleśtenag andś et
šút lakút treten teletaú
nem polem tišú śotriś eúś

possible alternative readings:
line 1: gaariśtadeś
line 2: partenúś
line 4: krúś
line 5: rotnem
line 8: merion
line 11: mút iakút

And as for the other fragmentary texts, also found in Novilara, according to that forum post:

2^nd Stele (left half of two lines)

úpeś
mresveat

The ú in line 1 might also be an L.
In line 2 ve could also be ev or ee.

3^rd fragmentary text

pa śatigot
kešoteri
amdet : nk
------ k --------

The p in line 1 is uncertain.

4^th fragmentary text

tiperašθe raiup bav---
-----ipš---------------------

I like how the user Arnth also added sidenotes:

"There appears to be some flexion at work: rotnem - rotneš, śoter - śotris.

Similarities to Greek words: polem - polis, śoter - soter, iśperion – esperios."

I noticed this as well, and they're (well at least some of them) transparently IndoEuropean! Though they also note how the -sv- element is also found in North Etruscan, as well as several other features also being possibly Etruscan (so it's not a done deal what language family this belongs to).

Now that we've presented the evidence of this language, let's start picking it apart, shall we?

Is or Was ProtoIndoEuropean a Split Ergative Language?

I've heard of this theory before, but I never quite gave it much thought. It was seemingly an outdated idea first posited by

I noticed a pattern, one thats relatively easy to see if you lay it out in the form of a table. It seems like animate and inanimate case endings, at least with the nominative case, seem to correspond roughly with what one might expect from a split ergative case pattern. Animate being, or having developed from, an Ergative case; And Inanimate endings being the Absolutive form:

Nominal Endings for the Nominative, Vocative, and Accusative Cases
	Animate			Inanimate/Neuter
	Singular	Dual	Plural	Singular	Dual	Plural
Nominative	-s/-os	-h₁e /-oh₁e	-es	-Ø/-om	-ih₁/-oih₁	-h₂/-eh₂>-ā
Vocative	-Ø /-ĕ	-h₁e -oh₁e	-es	-Ø/-om	-ih₁/-oih₁	-h₂/-eh₂>-ā
Accusative	-m/-om	-h₁e/-oh₁e	-ns	-Ø/-om	-ih₁/-oih₁	-h₂/-eh₂>-ā

Source: http://en.wikipedia.org/wiki/Proto-Indo-European_nominals

See a pattern?

Now, there are other -h2- forms, of which use -h2- followed by case endings, which I believe could have been seperate case endings. I'm not sure if this table is accurate, its based on the wikipedia article, I was under the impression that neuter endings were seperate from the -h2- endings, which gave rise to the feminine nominal conjugation. I could be wrong, but which it derives from is irrelevant.

This pattern is also seen in Hittite, though with a different set of endings than what wikipedia has for ProtoIndoEuropean:

	Animate Nouns		Neuter/Inanimate/Common Nouns
	Singular	Plural	Singular	Plural
Nominative	-as	-ēs	-an	-a
Accusative	-an	-us	-an	-a

Notice, that in both instances, the Nominative and Accusative endings for the Neuter/Inanimate are identical. Also, for Animate endings, the two are not identical. Also, outside of the Nominative and Accusative cases, the Animate and Neuter/Inanimate/Common Noun endings are indistinguishable.

This on its own is nothing special, except, we also know that Hittite Neuter nouns can't function as the subject of a transitive verb, rather, their endings must be replaced by Ergative Endings:

	Animate		Neuter
	Singular	Plural	Singular	Plural
Nominative	-as	-ēs	-an	-a
Ergative	-as	-ēs	-anza	-antēs
Accusative	-an	-us	-an	-a

It's also known that NeoHittite slowly began to confuse Nominative and Accusative Animate plural endings, possibly another indicator of an association between the two. Then again, this easily could be related to the fact that Hittite neighbored and even had a substrate that was an Ergative language. So this need not necessarily be an archaic feature preserved in Hittite.

Its thought that the Ergative case endings for Hittite Neuter nouns originated from a derivationalized form of a suffix -ant-. I'd like to point out that its similar to the Imperfect participle, and 3rd person plural -nt- endings, which are also Active in ProtoIndoEuropean. Perhaps this is the origin of the Ergative case marker?

Active-Nominative Sing. -nt-s>-nz, -nt-es>-entes

The construction seems built upon the animate endings.

A reconstruction of the original paradigm, if ProtoIndoEuropean at some point in its history was a Split Ergative language, it might have looked something like this:

A Possible Reconstruction of ProtoIndoEuropean's Morphosyntactic Case System

	Singular	Dual	Plural
Nominative	-Ø/-om	-ih₁/-oih₁	-h₂/-eh₂>-ā
Ergative	-s/-os	-h₁e /-oh₁e	-es
Accusative	-m/-om	-h₁e/-oh₁e	-ns

The beauty of this system is that, the only thing that is lost are the use of "Nominative Dual and Plural" endings to mark the Accusative in Neuter Nouns. As the Neuter paradigm is lost. Though as can be seen from -h2- reconstructions, -h2m can be reconstructed for the Accusative, so all is not really lost.
Also, the Vocative and the Nominative/Oblique, in my view, would have originally been the same case. This makes the Nominative-Vocative very similar to what we find in ProtoUralic, though I can get to the similarities between IE and U in another post. 

For more thoughts on the Ergative Case in PIE:

http://www.kortlandt.nl/publications/art049e.pdf

https://ojs.lib.byu.edu/spc/index.php/DLLS/article/viewFile/31178/29637

https://repozytorium.amu.edu.pl/jspui/bitstream/10593/7433/1/PSiCL_44_4_Bavant.pdf

I realize that there are critics of this theory, but  I don't quite think that the case is closed on Ergativity in PIE; far from it.

Thoughts on IndoEuropean Nominal Ablaut Patterns Part 2

Now, with different ablaut types, this apparent "stress related" vowel change, seems to be realized slightly differently. Since these ablaut patterns are highly regular, you'll see me focus primarily on the Nonnarten Acrostatic forms, as they're the easiest on the eyes to see. In reality, what I've shown can apply to pretty much all ablaut forms.


Now, regarding my original idea before the stress-related idea, I had spotted another possibility. I'd like to explore the idea, that at least in the case of the root-vowel, that the:

o-grade - Marked a sort of "default" case, including the Nominative, Vocative, and Accusative cases.

e-grade - Marked cases that were either Genetival and or Prepositional in meaning.

While not far off from my previous post, its an idea I'd like to explore further as well. It's not entirely exclusive from it either, as this system could reasonably have arisen from the previous.

Though once you get to the secondary places where you'll find vowels, such as in the stems and endings, this gets a little trickier. For instance, half the plural endings have -o- in them (Genetive, Ablative, Dative (alternatively -mus), and Ablative. Though it wouldn't necessarily contradict my hypothesis, it creates problems for my stress-related theory when it comes to certain Ablaut forms, where the ablaut vowel falls in the stem, though these differences are resolveable, it still bothers my thirst for simplicity.

As far as the verbs are concerned, I'll get to them at a later date. I want to try to not quote wikipedia twice if I can't help it. Especially since I'm less familiar with the ablaut when it comes to verbs, I won't so easily spot the error and be able to correct them. At least not at this time.

Thoughts on IE Nominals and Ablaut

 I'm curious about a pattern I noticed with nouns, there seem to be two basic clusters that noun cases fall into; At least when it comes to their basic ablaut patterns. We'll call them Cluster A and Cluster B, and they're defined by their ablaut forms:

Cluster A: Nominative, Vocative, and Accusative
Cluster B: Instrumental, Dative, Ablative, Genetive, and Locative

I wasn't sure what I was seeing in this. I was tempted to compare it to Arabic and Akkadian's case conjugations. Which in the case of Arabic offers up Nominative -u, Accusative -a, and Possessed/Genetive -i. I'm not so sure that correlation is meaningful though, especially since I know too little of the nominal conjugations of the other AfroAsiatic languages to draw such a connection. So I set the idea to the backburner.

It was a curious pattern from a language I created, that developed when -e- became an oblique form that eventually replaced the use of -o- when Acrostatic root nouns were highly productive. Especially when I added an Ergative -a-, so the pattern was freaky similar, but was a coincidence I'd otherwise have missed. 

In all honesty, I knew from the start it was a coincidence. Though it was a curious one, and it got me thinking about it. Obsessing over the ablaut is a better term. I wanted to know its origin. What if it evolved from an earlier consonantal root system? Or maybe it was developing in that direction and stopped?  I didn't have any convincing ideas as of yet.

Though while writing this, I did think about Uralic and its stress patterns at the level of ProtoUralic, and I noticed that it followed this basic formula:
CV=Syllable, and not Consonant+Vowel specifically per se, at least here:
CV́CVCV̀CV(CV̀)(CV),etc...

Due to their geographical proximity towards one another, what if at a stage earlier than what is reconstructed for ProtoIndoEuropean, or possibly at the level of ProtoIndoEuropean itself, it had this same basic pattern? Perhaps, like the above, its a pattern that doesn't really exist, but has seemed to have appeared by coincidence. Though I think its an interesting idea to explore nonetheless.

What I'm thinking is, what if -o- became -e- when preceeded by an adposition of some sort? As if at one point, adpositions influenced the way the noun they modified was pronounced. In the same way that the noun that is modified by the indefinite article in Modern English, influences the way the indefinite article is pronounced:

The indefinite article, when preceeding a Noun that starts with a vowel or consonant, becomes:
Vowel: an
Consonant: a

In other words, perhaps:

Singular Acrostatic Root Noun "night"

sing.	nom.	nókʷt-s
	voc.	nókʷt
	acc.	nókʷt-m̥
	inst.	kóm nékʷt-(e)h₁
	dat.	dó nékʷt-ey
	abl.	h₂ét nékʷt-s
	gen.	h₂ét nékʷt-s
	loc.	ʔén nékʷt-(i)
dual	nom.-voc.-acc.	nókʷt-h₁e
plur.	n.-v.	nókʷt-es
	acc.	nókʷt-n̥s
	inst.	kóm nékʷt-bʰi
	dat.-abl.	dó nékʷt-m̥os
	gen.	ʔén nékʷt-oHom
	loc.	ʔén nékʷt-su

What I immediately liked about this idea, was that I really didn't have to change much. Though I did have to suppose that, as a rule, any noun in Case Cluster B, was required to be preceeded by an adposition, or some sort of particle or prefix. While I'd like to see someone with actual university-level credentials on historical linguistics analyze this (not just my armchair amateurish theorization), I would like to note some potential evidence and flaws for this idea:

In the case of the word "fatherless":

Ablaut grade	PIE (reconstruction)	Greek	(Greek transliterated)	Translation
e-grade or full grade	*ph2-tér-m̥	πα-τέρ-α	pa-tér-a	"father" (noun, accusative)
lengthened e-grade	*ph2-tḗr	πα-τήρ	pa-tḗr	"father" (noun, nominative)
zero-grade	*ph2-tr-és	πα-τρ-ός	pa-tr-ós	"father's" (noun, genitive)
o-grade	*n̥-péh2-tor-m̥	ἀ-πά-τορ-α	a-pá-tor-a	"fatherless" (adjective, accusative)
lengthened o-grade	*n̥-péh2-tōr	ἀ-πά-τωρ	a-pá-tōr	"fatherless" (adjective, nominative)

Now, I'm not sure if the vowel patterns have more to do with having a prefix attached or if its something that happens with the inclusion of an adjective, I'd like to hear someone in the field's take on it. In the meantime, I'll be researching this further. (No, wikipedia isn't my main source for PIE, its simply convenient for copying and pasting.)

It appears like the inclusion of a prefix or particle before the noun, is what changed the vowels within the Noun itself.

However, there is a slight problem with this, at least in the case of the negative prefix n-, It would have been predicted that the prefix would get the stress, not the root. Or perhaps, like with ʔeme>me after the loss of the laryngeal, a prior particle ʔene>ne~n was similarly reduced?


As you can see below in the chart taken from wikipedia, suffixes do sometimes have influence upon how vowel patterns are realized, or so it seems. It does seem to appear to be stress-related in nature. At least from what I can tell. This doesn't deal so much with my theory, and would necessarily be knowledged gleaned before my hypothesis came into existence, so it can't really be used as evidence of any sort:

Taken from Wikipedia:
 http://en.wikipedia.org/wiki/Proto-Indo-European_nominals

		acrostatic root noun	acrostatic lengthened root noun	amphikinetic (?) root noun	hysterokinetic r-stem	amphikinetic n-stem	hysterokinetic n-stem
gloss		night (f.)	moon (m.)	foot (m.)	father (m.)	lake (m.)	bull (m.) (< "ox")
sing.	nom.	*nókʷts	*mḗh₁n̥s	*pṓds	*ph₂tḗr	*léymō	*uksḗn
	voc.	*nókʷt	*mḗh₁n̥s	*pód	*ph₂tér	*léymon	*úksen
	acc.	*nókʷtm̥	*mḗh₁n̥sm̥	*pódm̥	*ph₂térm̥	*léymonm̥	*uksénm̥
	inst.	*nékʷt(e)h₁	*méh₁n̥s(e)h₁	*pedéh₁	*ph₂tr̥éh₁	*limnéh₁	*uksn̥éh₁
	dat.	*nékʷtey	*méh₁n̥sey	*pedéy	*ph₂tr̥éy	*limnéy	*uksn̥éy
	abl.	*nékʷts	*méh₁n̥sos	*pedés	*ph₂tr̥és	*limnés	*uksn̥és
	gen.	*nékʷts	*méh₁n̥sos	*pedés	*ph₂tr̥és	*limnés	*uksn̥és
	loc.	*nékʷt(i)	*méh₁n̥s(i)	*péd(i)	*ph₂tér(i)	*limén(i)	*uksén(i)
dual	nom.-voc.-acc.	*nókʷth₁e	*mḗh₁n̥sh₁e	*pódh₁e	*ph₂térh₁e	*léymonh₁e	*uksénh₁e
plur.	n.-v.	*nókʷtes	*mḗh₁n̥ses	*pódes	*ph₂téres	*léymones	*uksénes
	acc.	*nókʷtn̥s	*mḗh₁n̥sn̥s	*pódn̥s	*ph₂térn̥s	*léymonn̥s	*uksénn̥s
	inst.	*nékʷtbʰi	*méh₁n̥sbʰi	*pedbʰí	*ph₂tr̥bʰí	*limn̥bʰí	*uksn̥bʰí
	dat.-abl.	*nékʷtm̥os	*méh₁n̥smos	*pedmós	*ph₂tr̥mós	*limn̥mós	*uksn̥mós
	gen.	*nékʷtoHom	*méh₁n̥soHom	*pedóHom	*ph₂tr̥óHom	*limn̥óHom	*uksn̥óHom
	loc.	*nékʷtsu	*méh₁n̥su	*pedsú	*ph₂tr̥sú	*limn̥sú	*uksn̥sú
		proterokinetic neuter r/n-stem	amphikinetic collective neuter r/n-stem	amphikinetic m-stem	proterokinetic ti-stem	proterokinetic tu-stem	proterokinetic neuter u-stem
gloss		water (n.)	water(s) (n.)	earth (f.)	thought (f.)	taste (m.)	tree (n.)
sing.	nom.	*wódr̥	*wédōr	*dʰéǵʰōm	*méntis	*ǵéwstus	*dóru
	voc.	*wódr̥	*wédōr	*dʰéǵʰom	*ménti	*ǵéwstu	*dóru
	acc.	*wódr̥	*wédōr	*dʰéǵʰōm	*méntim	*ǵéwstum	*dóru
	inst.	*udénh₁	*udnéh₁	*ǵʰméh₁	*mn̥tíh₁	*ǵustúh₁	*drúh₁
	dat.	*udéney	*udnéy	*ǵʰméy	*mn̥téyey	*ǵustéwey	*dréwey
	abl.	*udéns	*udnés	*ǵʰmés	*mn̥téys	*ǵustéws	*dréws
	gen.	*udéns	*udnés	*ǵʰmés	*mn̥téys	*ǵustéws	*dréws
	loc.	*udén(i)	*udén(i)	*ǵʰdʰsém(i)	*mn̥téy (-ēy)	*ǵustéw(i)	*dréw(i)
dual	nom.-voc.-acc.				*méntih₁	*ǵéwstuh₁	*dórwih₁
plur.	n.-v.				*ménteyes	*ǵéwstewes	*dóruh₂
	acc.				*méntins	*ǵéwstuns	*dóruh₂
	inst.				*mn̥tíbʰi	*ǵustúbʰi	*drúbʰi
	dat.-abl.				*mn̥tímos	*ǵustúmos	*drúmos
	gen.				*mn̥téyoHom	*ǵustéwoHom	*dréwoHom
	loc.				*mn̥tísu	*ǵustúsu	*drúsu
		neuter s-stem	proterokinetic h₂-stem	hysterokinetic h₂-stem	eh₂-stem (ā-stem)	o-stem	neuter o-stem
gloss		cloud (n.)	woman (f.) (> "queen")	tongue (f.)	grain (f.)	nest (m.)	work (n.)
sing.	nom.	*nébʰos	*gʷḗn	*dn̥ǵʰwéh₂s	*dʰoHnéh₂	*nisdós	*wérǵom
	voc.	*nébʰos	*gʷḗn	*dń̥ǵʰweh₂	*dʰoHn[á]	*nisdé	*wérǵom
	acc.	*nébʰos	*gʷénh₂m̥	*dn̥ǵʰwéh₂m (-ām)	*dʰoHnéh₂m (-ā́m)	*nisdóm	*wérǵom
	inst.	*nébʰes(e)h₁	*gʷnéh₂(e)h₁	*dn̥ǵʰuh₂éh₁	*dʰoHnéh₂(e)h₁	*nisdóh₁	*wérǵoh₁
	dat.	*nébʰesey	*gʷnéh₂ey	*dn̥ǵʰuh₂éy	*dʰoHnéh₂ey	*nisdóey	*wérǵoey
	abl.	*nébʰesos	*gʷnéh₂s	*dn̥ǵʰuh₂és	*dʰoHnéh₂s	*nisdéad	*wérǵead
	gen.	*nébʰesos	*gʷnéh₂s	*dn̥ǵʰuh₂és	*dʰoHnéh₂s	*nisdósyo	*wérǵosyo
	loc.	*nébʰes(i)	*gʷnéh₂(i)	*dn̥ǵʰwéh₂(i)	*dʰoHnéh₂(i)	*nisdéy	*wérǵey
dual	nom.-voc.-acc.	*nébʰesih₁	*gʷénh₂h₁e	*dn̥ǵʰwéh₂h₁e	?	*nisdóh₁	*wérǵoy(h₁)
plur.	n.-v.	*nébʰōs	*gʷénh₂es	*dn̥ǵʰwéh₂es	*dʰoHnéh₂es	*nisdóes	*wérǵeh₂
	acc.	*nébʰōs	*gʷénh₂n̥s	*dn̥ǵʰwéh₂ns (-ās)	*dʰoHnéh₂ns (-ās)	*nisdóns	*wérǵeh₂
	inst.	*nébʰesbʰi	*gʷnéh₂bʰi	*dn̥ǵʰuh₂bʰí	*dʰoHnéh₂bʰi	*nisdṓys	*wérǵōys
	dat.-abl.	*nébʰesmos	*gʷnéh₂mos	*dn̥ǵʰuh₂mós	*dʰoHnéh₂mos	*nisdó(y)mos	*wérǵo(y)mos
	gen.	*nébʰesoHom	*gʷnéh₂oHom	*dn̥ǵʰuh₂óHom	*dʰoHnéh₂oHom	*nisdóoHom	*wérǵooHom
	loc.	*nébʰesu	*gʷnéh₂su	*dn̥ǵʰuh₂sú	*dʰoHnéh₂su	*nisdóysu	*wérǵoysu

		Athematic						Thematic
		acrostatic root noun	hysterokinetic r-stem	proterokinetic ti-stem	proterokinetic neuter u-stem	eh₂-stem (ā-stem)		o-stem	neuter o-stem
						PIE	Post-PIE1
gloss		night (f.)	father (m.)	thought (f.)	tree (n.)	grain (f.)		nest (m.)	work (n.)
sing.	nom.	*nókʷt-s	*ph₂t-ḗr	*mént-i-s	*dór-u	*dʰoHn-éh₂	*dʰō̬n-ā̬́	*nisd-ó-s	*wérǵ-o-m
	voc.	*nókʷt	*ph₂t-ér	*mént-i		*dʰoHn-[á]2	*dʰō̬n-á	*nisd-é
	acc.	*nókʷt-m̥	*ph₂t-ér-m̥	*mént-i-m		*dʰoHn-éh₂-m (-ā́-m)	*dʰō̬n-ā̬́-m	*nisd-ó-m
	inst.	*nékʷt-(e)h₁	*ph₂t-r̥-éh₁	*mn̥t-í-h₁	*dr-ú-h₁	*dʰoHn-éh₂-(e)h₁	*dʰō̬n-ā̬́?	*nisd-ó-h₁	*wérǵ-o-h₁
	dat.	*nékʷt-ey	*ph₂t-r̥-éy	*mn̥t-éy-ey	*dr-éw-ey	*dʰoHn-éh₂-ey	*dʰō̬n-ā̬́-y	*nisd-ó-ey	*wérǵ-o-ey
	abl.	*nékʷt-s	*ph₂t-r̥-és	*mn̥t-éy-s	*dr-éw-s	*dʰoHn-éh₂-s	*dʰō̬n-ā̬́-s	*nisd-é-ad	*wérǵ-e-ad
	gen.	*nékʷt-s	*ph₂t-r̥-és	*mn̥t-éy-s	*dr-éw-s	*dʰoHn-éh₂-s	*dʰō̬n-ā̬́-s	*nisd-ó-syo	*wérǵ-o-syo
	loc.	*nékʷt-(i)	*ph₂t-ér-(i)	*mn̥t-éy (-ēy)	*dr-éw-(i)	*dʰoHn-éh₂-(i)	*dʰō̬n-á̬-y	*nisd-é-y	*wérǵ-e-y
dual	nom.-voc.-acc.	*nókʷt-h₁e	*ph₂t-ér-h₁e	*mént-i-h₁	*dór-w-ih₁	?	?	*nisd-ó-h₁	*wérǵ-o-y(h₁)
plur.	n.-v.	*nókʷt-es	*ph₂t-ér-es	*mént-ey-es	*dór-u-h₂	*dʰoHn-éh₂-es	*dʰō̬n-ā̬́-s	*nisd-ó-es	*wérǵ-e-h₂
	acc.	*nókʷt-n̥s	*ph₂t-ér-n̥s	*mént-i-ns		*dʰoHn-éh₂-ns (-ās)	*dʰō̬n-ā̬́-s	*nisd-ó-ns
	inst.	*nékʷt-bʰi	*ph₂t-r̥-bʰí	*mn̥t-í-bʰi	*dr-ú-bʰi	*dʰoHn-éh₂-bʰi	*dʰō̬n-ā̬́-bʰi	*nisd-ṓ-ys	*wérǵ-ō-ys
	dat.-abl.	*nékʷt-m̥os	*ph₂t-r̥-mós	*mn̥t-í-mos	*dr-ú-mos	*dʰoHn-éh₂-mos	*dʰō̬n-ā̬́-mos	*nisd-ó-(y)mos	*wérǵ-o-(y)mos
	gen.	*nékʷt-oHom	*ph₂t-r̥-óHom	*mn̥t-éy-oHom	*dr-éw-oHom	*dʰoHn-éh₂-oHom	*dʰō̬n-á̬-ō̬m	*nisd-ó-oHom	*wérǵ-o-oHom
	loc.	*nékʷt-su	*ph₂t-r̥-sú	*mn̥t-í-su	*dr-ú-su	*dʰoHn-éh₂-su	*dʰō̬n-ā̬́-su	*nisd-ó-ysu	*wérǵ-o-ysu

A Thought Experiment on Linguistic Change (1)

I was basically doing a little thought experiment, and noticed something interesting. I was thinking about how much english has changed over the span of 250 years, and then how much it has changed over the span of 500 years, and how after 1000 years its unrecognizeable. Then I wondered, is there a pattern behind this?

I'm not a fan of glottochronology, nor glottochronological methodology, but I did notice a fairly consistent pattern. I noticed that phonological change seems to occur generationally in steps, and that, it becomes noticeable after about 4-5 generations. So if I were to divide that 250 years up into 125 years.

The assumption made here, is that isolated or not, two areas where a language is spoken, will undergo slightly different changes. Those changes may or may not be regular, that is, if language A underwent 3 sound changes, language B might not also undergo 3 sound changes, but they would at least be noticeably different as a requirement after 125 years in some way. I'd say that it'd be the latest time span before language B would resemble exactly the ancestral form of both language A and B, if it hadn't already gone through some change.

Essentially, my point is, if we were able to hear a voice recording of Abraham Lincoln speaking while still living in Illinois, he'd have a slightly different accent than someone today living in that same part of Illinois.

Now, since the settlement of North America, and even the spread westward in the 1860's onward, we actually do see dialectical differences. The map posted by the Washington Post below could be subdivided further, but it shows the basic outline:
http://www.washingtonpost.com/blogs/govbeat/wp/2013/12/02/what-dialect-to-do-you-speak-a-map-of-american-english/

Now, at the borders of these dialects, there tends to be merging of differing dialects. For instance, as someone from Upstate New York, my own speech seems to obey some sound changes for both New England Western and the Inland Northern dialects, and sounds quite similar to the south eastern parts of Ontario and Quebec.

The basic idea is, if you trace your way backward, from dialect to language to language family, to language superfamily, etc.... you can get a somewhat reliable most recent possible date a given language was spoken. While far from finished. The basic idea, before I run through in more depth, is that based on the internal relationships between dialects, you can work your way backward to get an idea as to how long ago a language would have been spoken:

2000-1875 CE Dialectical differences between American English - 125 years
1875-1750 CE Dialectical differences between American and British English - 250 years
1750-1625 CE Dialectical differences in Early Modern English in England - 375 years
1625-1500 CE Modern English developes from Middle English -  500 years

Now, for our purposes, we're counting Middle English as a seperate language. I think that it is sufficiently different from modern english to be counted as such, in the same way that as similar as Portuguese and Spanish are, they are different languages still. Or to make a more time similar comparison, Dutch and Afrikaans.

1500-1375 CE Dialectical differences occur within Middle English, compare Yola and the dialect that gave us Modern English, that are subclassifiable (as say American and British English are subclassifiable).
1375-1250 CE Dialectical differences occur within Middle English.
1250-1125 CE Middle English evolves from the clash between Old English and Norman French. Old English being a seperate language for the same reason Modern and Middle English are being considered seperate.
1125-1000 CE Start of that developement of Middle English from Old English.

1000-875 CE Old English shows multilevel dialectical differentiation
875-750 CE Old English dialects show dialectical differentiation
750-625 CE Old English shows dialectical differentiation
625-500 CE Old English developes from West Germanic. (450 CE is given for the start of Prehistoric Old English according to wikipedia.)
http://en.wikipedia.org/wiki/Old_English

It's important to note, that the dates given above are approximate, and they don't coincide perfectly with that given by linguists. A language can be divided up arbitrarily at any point, the point here is utilitarianly.

The basic idea that I came up with, is that after:

125 years dialects begin to emerge from a language. (at this stage, the language itself is still a dialect of a previous language).

250 years dialects form within dialects of a given language. Forming a cluster of related dialects.

375 years a full blown family of dialects emerges within a language, as it begins to split into different languages.

500 years, dialects that began to differentiate 125 years ago, begin to form into seperate languages.

For instance, Latin was brought to Sardinia in 238 BCE.

http://www.britannica.com/EBchecked/topic/524166/Sardinian-language

Almost immediately, dialectical differences between Vulgar Latin and Sardinian began to occur, as the two began to develope along different trajectories. By about 600 CE, Vulgar (Nonclassical) Latin, had split into nonmutually intelligeable dialects, or languages.

If we were to measure this time distance based on the present, we'd measure our way backwards in intervals of 125 years. Vulgar Latin would exist in Sardinia from about 250 BCE to 125 BCE when it became the Sardinian Dialect of Vulgar Latin.

125-1 BCE would see the rise of dialects within Sardinian Vulgar Latin on the island.

1-125 CE would see multiple level dialectical differentiations in Sardinia.

125-250 CE would then have Sardinian develope into a seperate language.

Now, whether a group of dialects developes into one or multiple languages, depends on whether those dialects end up forming a coherent group. The same process that brought English to being similar to Norman French, is the same process that would've developed Sardinian and Icelandic into languages. The speakers of the dialects in those groups commonly communicated with one another, as they lived on the same island or same country, so there was a high degree of linguistic exchange between their communities.

You could do the same by comparing the spread of words like tea~chai into and into and through Europe from East Asia. You have communities of speakers with a method of communication (between Europe and East Asia, the oceanic and silk road trade routes).

Basically, in a nutshell, the process by which one language would develope into another, seems to take about 500 years. We'll explore this more in further posts.