29 Oct 2008

The trouble with the PIE 1st & 2nd person plural endings (2)

(Continued from The trouble with the PIE 1st & 2nd person plural endings.)

My goal in this latest blog rant is to find a pleasing reconstruction of the Old IE (OIE) objective conjugation, that is, the antecedent of Proto-Indo-European's (PIE) *mi-conjugation. I'll be referring back to Pre-IE sound changes as I've defined them in my online pdf in order to find my solution. Before we can arrive at a solution, however, we must piece together what happened with the first and second persons plural endings. I guess to keep this topic together, we first might peck off a few options of possible ancient protoforms. Here we go.

The first, most obvious post-IE innovation is *-mosi/*-mos/*-mo as might be based on Latin -mus or OIr. -m, for example. This must be a new phenomenon that, at the very least, postdates Schwa Diffusion since in order for *o to have developed in an unstressed syllable, *s would have to be voiced at the time. Yet, this is not possible because we know that the plural had unvoiced *s throughout the Late IE period, that it was unaccented and that its preceding vocalism was *e as a result. Likewise, just in case some may think that *-mes was instead the product of 1ps *-m and 2ps *-s “you” (i.e. “I and you” → “we (inclusive plural)”), the same problem arises and *e is yet again the expected vocalism since it is afterall what is found in the 2ps thematic presentive ending, *-e-si (never **-o-si). So no matter how you slice and dice the 1pp ending, forms like *-mo[s(i)] must be patterned on something else, something later. Indeed, it is no doubt affected by the independent pronoun *nos “we” (in turn a late derivative of unaccented PIE *n̥s < MIE *mas). Surely this is a post-IE form and has no bearing on PIE, let alone Pre-IE. So let's toss these endings in the trashcan straight away!

Next up are the 1pp variants *-més/*-mé and 2pp *-té. In this grouping, we observe an accent that should not be there since etymologically speaking, the 1pp ending is surely nothing more than *-me- [first person pronominal stem] and *-(e)s [plural]. Judging by QAR, the plural marker was unaccented because it contained no word-final vowel in the stages preceding Syncope . In other words, even before Syncope, *-as was the MIE plural ending, not **-ésa. Further, this word-final *s is testimony to a most ancient event which I call Indo-Aegean Sibilantization. Ultimately this plural is related to Uralic plural *-t. The sibilantization of early Indo-Aegean *t would not have happened unless *t was word-final at the time (n.b. Sibilantization also explains the origin of *s/*t-heteroclitic stems in PIE like participles in *-wos). Subsequently as a result of a lack of word-final vowel in the Mid IE (MIE) plural, the plural could never have been voiced to **-ez in the Late IE period. Yet... we find accented athematic presentive 1pp and 2pp endings nonetheless. Why??? How??? And if the primary form were *-mési and secondary were *-més, then despite unexpected accent, the source of the pattern might be a little more transparent. However, the primary form was apparently *-més without *-i and the secondary was *-mé. The absence of *-i in the primary marker adds to the strangeness but it might suggest an archaicism. The loss of *-s in the secondary is no doubt by analogy with other late innovations such as the secondary middle endings with its similar loss of *-r ultimately inspired by the inherited loss of final *-i in secondary active forms from pre-IE times. At first glance, it's impossible to tell what exactly was going on with this particular group of endings because of a few things obscuring the problem, so let's move on and come back to this group of endings later on.

Next up: the *-mén(i)/*-tén(i) group. More later.

(Continue reading The trouble with the PIE 1st & 2nd person plural endings (3).)


  1. glen, can you recommend a couple of textbooks or sourcebooks in this area (PIE, pre-IE, etc. historical linguistics) for those of us who allus wannuhd to be lingwists but couldn't take the classes for whatever reason? your posts are fascinating and filled with tantalizing sentences that begin "we know that ..." and they always make me go "wait..wait..we DO? Shit, *I* sure don't!" So help us ignoramuses out here please - list a few basic books to read where we can get up to snuff on the latest in this field.


  2. I recommend reading any book on PIE, and then read another, and then question everything you read so that you form your own informed opinions. There are plenty of sources both in your local university library and online on PIE. Google is your friend, as is your librarian. Indo-European is one of the most studied language families on the planet.

    Pre-IE and Nostratic resources however are much more scant so this is why I'm exploring the topic on my own and arriving at my own thoughts on the subject. I appreciate Allan Bomhard's thoughts on the prehistoric movement of Pre-IE into the Pontic and some of his ideas on reconstruction, but I'm trying to push it all further.

    Now, the "we-know-that" sentence you're referring to seems to be: "[...] we know that the plural had unvoiced *s throughout the Late IE period, that it was unaccented and that its preceding vocalism was *e as a result." You didn't know that?? Egad. Yes, the PIE plural marker is *-es and, no offense, but that's basic knowledge. Sometimes the plural marker is simply *-s however, as in the accusative plural *-ns but I believe that this is because of Syncope occasionally acting on MIE *-as when it otherwise preserves its vocalism (check my pdf).

  3. I have both Szemerenyi's and Beekes' books on PIE and recommend them both. They provide a good counterbalance to each other because they don't agree on all points of reconstruction. Where they disagree with each other they both tend to overstate their cases so the counterbalance is necessary. Also, while they cover many of the same topics they don't go into the same level of detail for each topic. IMO, Szemerenyi has better info on morphology but presents an antiquated phonological system. Beekes' treatment of morphology is more shallow but has a much better treatment of phonology and reconstructed lexicon.

  4. Thanks, Mordrigar. Good insight. Also, like I said earlier, even if you can't afford books (and let's face it, linguistics books are an arm and a leg), Google can give you plenty of information on anything you want to know on Indo-European if you're resourceful enough. Just check out for example what a search through Google Books under "indo-european" gives you: click here. And if you type in "indo-european pdf" in plain ol' generic Google, you get the following tasty resources: click here.

    See?? Amazing stuff! So there's no excuse to not know about Indo-European. Get cracking, mon ami :)