Laryngeal overdose in the Indo-European second person

For the record, I hate the abuse of laryngeals. What I mean by "abuse" is when people, unsatisfied with a protolanguage proven to contain seemingly exotic laryngeals with accompanying vocalic effects, decide to add laryngeals to every stem to account for all long vowels, whether it can be justified or not, and end up succeeding only in muddling the whole grammatical system in the process, obscuring the very thing they attempt to clarify.

A quick and easy example of this is Bhadriraju Krishnamurti's use of laryngeals in the 1st and second pronouns *yān 'I' and *nīn 'you' (or in his view, *yaHn and *niHn^[1]) to account for lengthening in the nominative which opposes oblique stems *yan- and *nin- lacking added vocalic length. Without laryngeals, it should be already clear to a knowledgeable linguist that many languages simplify stems in oblique case forms without the need to appeal to arcane infixation of one 'miracle phoneme' or another. We see this same simplification of pronominal stems between noun cases in PIE when *tu with back high vowel in the default nominative case opposes *twe. What was once *u throughout the entire 2ps pronominal paradigm must have apparently weakened at some point to a lower vowel *e in oblique cases. The phenomena exhibited in these two protolanguages are surely one and the same and therefore do not require laryngeals to explain them.

So to the topic now, there's a succinct reason why the 2nd person singular perfect ending *-th₂e contains a laryngeal that apparently even many IEists, so seduced by their own ideas, forget to take into account. It's given this laryngeal because of the Sanskrit reflex -tha whose added aspiration should otherwise not be present. As a result, we obtain a nice symmetry on the PIE level between 1ps perfect *-h₂e and 2ps perfect *-th₂e.

Yet further, it's precisely that very phonemic symmetry embedded in the perfective system that ultimately helps to explain the emergence of laryngeals in 2ps endings and stems at all! It can be readily seen from this pattern that the true reason for the laryngeal in the 2ps perfect is analogical levelling with the adjacent 1ps ending. While the laryngeal of the 1ps is legitimately ancient and stemming from the earliest stages of Pre-IE, the 2ps laryngeal cannot be. It should also be noted that without a laryngeal in this ending, it would otherwise be identical with the corresponding 2nd person plural ending *-te, adding further motivation to this irregular change^[2].

Coincidently, there is no other evidence for laryngeals elsewhere in second person morphemes, whether it be in *tu (2ps. nom.), *twe (2ps.acc.), *tene (2ps.gen.) or *-s(i) (2ps.dur.). Putting laryngeals in places they needn't be is a grave misanalysis on the part of a comparative linguist who's obligated by Logic to find the simplest solutions possible given the available evidence.


NOTES
^[1] Krishnamurti/Emeneau, Comparative Dravidian linguistics - Current perspectives (2001), p.336.
^[2] I should add that my theories on syllabic structure in Mid IE (ie. the stage before stress-motivated Syncope) forbid onset clustering at all via Occam's Razor. No Semitic loans identified during this period seem to require tautosyllabic consonant clustering in Mid IE either. All clusters in PIE seem to be the result of later Syncope and the changes in legal syllabic structure that ensued. Yet if we're obligated to side with this typological simplicity of Mid IE, this is yet one more reason why the onset cluster seen in *-th₂e cannot be any older than the early Late IE period. We may then presume the singular form to have been *-te before this time.

Laryngeal abuse - Phonemes caught in the reconstructive crossfire

Today I'm here to warn you about the tragedy of laryngeal abuse. This is when you see a long vowel in some proto-language and a devilish thought comes to mind like "Gee, I wonder if that long vowel is underlyingly the result of a vowel-plus-laryngeal combo?" And then before you know it, you've gone and rearranged the entire proto-language according to your laryngeal-obsessed whims. Laryngeals are fun, but we have to keep a level head too.

I remain convinced that Bhadriraju Krishnamurti's version of Proto-Dravidian is one such example of this laryngeal abuse at work, and it seems to me that this can be quickly resolved by examining what a mess he makes out of the pronominal system of this language. As far as I'm concerned it's supposed to look like this (i.e. how most Dravidianists reconstruct it):

	singular	plural
1p	*yān (obl. *yan-)	*yām / *nām (obl. *yam- /*nam-)
2p	*nīn (obl. *nin-)	*nīm (obl. *nim-)
3p (reflexive)	*tān (obl. *tan-)	*tām (obl. *tam-)

However, Krishnamurti has proposed the following^[1]:

	singular	plural
1p	*yaHn	*yaHm / *ñaHm
2p	*niHn	*niHm
3p (reflexive)	*taHn	*taHm

He then suggests that the laryngeals disappear in oblique forms. However the process by which this happens is obscure and left unexplained. In contrast, the idea that long vowels are reduced when used in oblique cases or when preposed to another noun is natural and commonplace. For example, we may take note of French moi "me" versus enclitic m(e) "me, myself", the latter being preposed to verbs as the object (e.g. Elles m'aident. "They help me."; Vous me dérangez "You disturb me."). Since we know where French comes from (i.e. Latin, of course!), we know how absurd and off-track it would be to reconstruct Proto-Latin **me(H) "me" in ignorance of attested Latin, placing a laryngeal in there that appears and disappears conveniently like the Cheshire Cat without rhyme or reason.

I have to say that I just don't buy Krishnamurti's reworking of the pronominal system. Whether Dravidian ultimately has a few laryngeals lurking about is, to be fair, a seperate issue that may still hold true, but these pronouns surely don't contain any. To add them here makes analysis more difficult rather than less.


NOTES
^[1] Krishnamurti, Comparative Dravidian Linguistics (2001), p.336 (see link).

A ramble about the Nostratic pronominal system, part 2

(Continued from A ramble about the Nostratic pronominal system.)

In my view, long-range comparative linguists who think that they have to make a really long list of highly tentative cognates to impress people are a dime a dozen. What seperates the wheat from the chaff is how structured and attentive to detail a theory is. A theory without a structure isn't a theory; it's nothing more than a tale heard at a local pub. Publishing drunken tales still doesn't make them theories.

So this is why I encourage people who are interested in this topic to first explore the Nostratic pronominal system because it's safe to say that if the premise of Nostratic has any truth to it, there should be an underlying pronominal system that explains the interrelationship of pronominal systems of later language groups that Nostratic is said to have begotten. We should be starting with these questions instead of putting the cart before the horse and comparing look-alike words by pure, directionless whim. Knowing how these pronominal systems are related to each other goes a long way to understanding the evolution of Nostratic and to finding more credible sound correspondences.

Now since this is all at the level of entertaining conjecture, I'll just spit out what I think, like the drunken bar patron that I am. Grab yourselves a drink too, my buddies! There is one interesting, recurring feature in Nostratic language groups that I notice: a suppletive system involving two very unrelated forms for each person. So for example, in Indo-European we have two sets of pronominal endings in use: the *mi-set (*-m, *-s, *-t) and the *h₂e-set (*-h₂e, *-th₂e, *-e). The former set was used for imperfective forms and the latter for perfective forms in most IE languages while in Anatolian it seems that verbs were inherently part of either a mi- or hi-conjugation class. We see in Uralic-Yukaghir, Chukchi-Kamchatkan and Eskimo-Aleut languages a shared theme of subjective-objective conjugation and again there are two different sets of endings that seem to be quite ancient (e.g. Hungarian 1ps -m & -k, 2ps -d & -l). Bomhard believes that Dravidian is also a Nostratic language but Dravidian uses quite different pronouns in the first and second person from the other Nostratic languages (*yān, *nīn) forcing him to reconstruct extra Nostratic pronouns just to account for them. In Afro-Asiatic languages, yet again, we apparently have two different sets (Middle Egyptian *anāka "I" > Sahidic Coptic anok versus Middle Egyptian *anāna "we" > Sahidic Coptic anon).

So I figure the best way to explain that is to propose a suppletive absolutive-ergative system for Nostratic as follows (note that my intention is to conjecture for the sake of discussion):

	ergative	absolutive
1ps	*nu (> *mu)	*hu
2ps	*tu	*nu
3ps	*ca	*ʔi

These pronouns might have optionally bore the suffix *-n for uncertain reasons. Absolutive pronouns are used for the agent of intransitive verbs and patients of transitive verbs while ergative pronouns are used for the agents of transitive verbs. As a result, we would expect to see the ergative and absolutive pronouns eventually attached to verbs as affixes in a new subjective-objective conjugation as I believe could have happened in a hypothetical ancestor of Indo-European, Altaic, Uralic-Yukaghir, Chukchi-Kamchatkan, Eskimo-Aleut and Dravidian. In Indo-European, a subjective-objective system could eventually evolve into the mi- and hi-classes of Anatolian IE because some verbs by nature are more apt to be either subjective or objective. It appears that an extra element *-e has been added to this absolutive set at an early stage of PIE, perhaps to use it for transitive verbs by marking it with a dummy object (nb. PIE *i- "he, she, it" and *e "here, there"). The same system can evolve into a contrast of imperfective and perfective as well because there is an implicit nuance of modal differences when dealing with subjective and objective conjugation as in the Siberian language called Nenets. Verbs lacking objects tend to convey a perfective sense^[1]. The PIE 2ps in *-s can be explained by softening of final *-t in word-final position and the 3ps in *-t can be attributed to the attachment of the demonstative stem *to- to an originally vowel-final form long after. In boreal languages like Uralic-Yukaghir, Eskimo-Aleut and Chukchi-Kamchatkan, the subjective-objective system could have evolved into *-m versus *-ɣ and 2ps *-t versus *-n. This system also helps to explain what would have happened in Dravidian. Simply put, Dravidian could have opted to generalize the absolutive pronouns for both agents and patients of actions and thus PDr *yān < *yan < *ʔin < *hun "I (abs.)" and PDr *nīn < *nin < *nun "you (abs.)".

NOTES
^[1] See Pedersen, Zur Frage nach der Urverwandtschaft des Indoeuropäischen mit dem Ugrofinnischen, Memoires de la Société Finnoougrienne 67, pp.311-315, Helsinki. He discusses the derivation of PIE's so-called "perfective" endings from a pre-IE intransitive conjugation. Also Abraham/Kulikov (eds.), Tense-Aspect, Transitivity and Causativity (1999), pp.21-42, Amsterdam. Kulikov shows a relationship between imperfectives and transitives (and conversely between perfectives and intransitivity) using similar data from Yukaghir and Aleut.

A ramble about the Nostratic pronominal system

I'm sure I must have hinted before that I hate when some treat long-range theories (like Nostratic, North Caucasian, Dene-Caucasian, or whatever far-away proto-language) as if they're written in stone. A person with a level head recognizes these ideas for what they are, idle conjectures requiring many ammendments before something more substantial can be made of them. However, I'm not against conjecture as long as it's fully differentiated from facts or well-substantiated theories. I also think there is an important difference between sharing conjectures for discussion on a blog or in a forum versus wasting trees to write a manifesto of your pseudolinguistic doctrine for you to enforce on disbelievers.

As much as I sound like a conservative fart for downplaying long-range comparison, I'm actually quite interested in it. It's just that I haven't read anything serious enough for me to go "wow!" yet and as I learn more, the errors in books start to become more apparent. Overall, I'm the most impressed (in a very moderate sense) by the Nostratic hypothesis as presented by Allan Bomhard who proposes that Indo-European, Uralic-Yukaghir, Altaic, Eskimo-Aleut, Elamite, Dravidian, Sumerian, Kartvelian and Afro-Asiatic language families come from a parent language dated to about 15 000 BCE in a period following the last ice age. He wasn't the first to come up with this century-old theory but he had a few different takes on it. For now, Nostratic is not an established theory because it doesn't present enough evidence to prove its claims, but it doesn't hurt to suggest further improvements that may help to inspire discussion and, just maybe, progress.

When looking through Allan Bomhard's Indo-European and the Nostratic Hypothesis (1996) or The Nostratic Macrofamily: A Study in Distant Linguistic Relationship (1994) co-authored by Allan Bomhard and John Kerns, one thing that I noticed was how many pronouns are being reconstructed without a clear structure. This is but one of a number of serious gaps in this theory just waiting to be resolved. The reconstructions presented by Bomhard and Kerns are always cited ad nauseum in ablaut pairs (e.g. *ma-/mə-) which of course serves no other purpose than to make the book twice as long. Since the ablaut patterns are said to be regular, there is no need to cite the second pair of each reconstruction any more than it is necessary to cite the Indo-European root *bʰer- as *bʰer-/*bʰor-/*bʰēr/*bʰr̥- each and every time. So I will dispense with irrelevancies and cite only the first pair of each of their reconstructions below.

First off, Bomhard and Kerns, on page 3 of The Nostratic Macrofamily: A Study in Distant Linguistic Relationship, show us this list of pronouns in the 1st and 2nd persons: *mi "I" [1ps], *tʰi "you" [2ps], *ma "we" [1pp.inclusive], *wa "we" [1pp] and *na "we" [1pp]. Immediately after are "notes" which are hampered either by irrelevancies or false information. For example, it suffices to say that Indo-European (IE) has a 1ps enclitic pronoun *me, 1ps genitive *mene, verbal 1ps thematic secondary ending *-m and verbal 1pp ending *-mes, the last being nothing more than a 1ps element with the plural ending *-es. So indeed there is ample evidence of an underlying 1ps pronominal root *me- in the deepest recesses of IE's prehistory. It's development in IE's Celtic branch however is wasteful rambling since it's obviously immaterial to Nostratic reconstruction and *me is well established in all other branches of IE even without the consideration of Celtic. Basing an Afro-Asiatic reconstruction solely on Chadic is bad practice known as "reaching". The so-called Etruscan imperative endings cited (-ti, -θ, -θi) are without substantiation, if not provably false altogether, despite ad hoc claims made by some prominent Etruscologists such as Giuliano and Larissa Bonfante. The belief that these endings are imperatives are based on ad hoc comparisons with Indo-European imperatives in *-dʰí (e.g. *h₁sdʰí /ʔəsdí/ "be!").

These aren't all the first and second person pronouns that are suggested by Bomhard and Kerns (see here). False comparisons are made between an underlying Uralic and Eskimo-Aleut 1ps ending in a velar stop on the one hand and Indo-European *h₁eǵoh₂ (cited as *ʔekʼ-) on the other^[1]. Some fun pronoun splicing of random data from the Afro-Asiatic family and presto changeo, yet another 1ps pronoun, *ʔa-. Then don't forget Bomhard's 1st person pronoun *ʔiya, supposedly proved by evidence from Chadic.

So in the 1st person alone, we now have five claims: *ʔa-, *ʔiya-, *ma-, *na- and *wa-. I'll discuss this more later.

(Continue reading the sequel: A ramble about the Nostratic pronominal system, part 2.)

NOTES
^[1] Read my views on the etymology of PIE's nominative 1ps pronoun in The origin of Indo-European ego.