Lubotsky dictionary gives the word for "road" in PIE as

NOM *pónt-h₁-s

ACC *pont-éh₁-m

GEN *pnt-h₁-ós

This looks strange to me because I would expect a thematic vowel instead of the laryngeal. Can this reconstruction be called "overlaryngealism" or it is a standard reconstruction? Is -eh₁- a suffix with fixed meaning?

  • @Alex B. Latin etymological.
    – Anixx
    Sep 20, 2012 at 20:12
  • @Alex B. I like this form better. In fact every time I consulted this dictionary it left me in doubt. For example their rejection of aqua as a borrowing and suggesting h2ep- as a native PIE root instead.
    – Anixx
    Sep 20, 2012 at 21:04
  • @Alex B. BTW can you give a reference?
    – Anixx
    Sep 20, 2012 at 21:37
  • @Alex B. so they all three give pent-oh2-s?
    – Anixx
    Sep 21, 2012 at 1:36
  • @Alex B. it seems you removed your comment that other sources give other form. Why you did so? Do not you mind to post it as an answer?
    – Anixx
    Sep 21, 2012 at 20:47

2 Answers 2


The word for 'path' certainly contained an ablauting laryngeal suffix, as the Avestan paradigm of the word demonstrates:

  • acc.sg. paṇtąm < *pént-oH-m̥ (cf. Skt. pántham) [NB Av. ą < ā / _m$]
  • gen.sg. paθō < *pn̥t-H-ós (cf. Skt. patháḥ)

The PIE accent/ablaut type here is amphikinetic (e-grade root and o-grade suffix in the strong stem, zero grade of both root and suffix in the weak steam, with accent on the root in the strong stem and the ending in the weak stem).

A laryngeal is necessary in order to account for the Avestan θ [interdental fricative] and Sanskrit th [voicless aspirate], continuing Proto-Indo-Iranian *th. Only when the laryngeal is directly adjacent to the stop does θ result (Skt. has evidently imported the voiceless aspirate throughout the whole paradigm: pántham << *pántam). Proto-Indo-Iranian *th exists only in forms that are reconstructable to Proto-Indo-Iranian, or to Proto-Indo-European voiceless stop + laryngeal sequences (thus, the same applies to PIIr. *ph [> Av. f and Skt. ph] and *kh [> Av. x and Skt. kh]). For instance, the 2.sg. ending of the perfect, Av. -θa, Skt. -tha, Greek -θa, Hitt. -ta is to be reconstructed as PIE *-th2e, thus accounting at one stroke for the a-vocalism in the Greek ending and the voiceless aspirate in the Indo-Iranian forms.

Whether the laryngeal is to be reconstructed as *h1 or *h2 not wholly certain, though *h2 seems more likely. In all instances in which the quality of a laryngeal that produces a voiceless aspirate in Indo-Iranian is clearly discernible, the laryngeal is *h2. For instance: Skt. rátha- 'chariot' < *rot-h2-o, cf. Lat. rota; Skt. pr̥thú- 'broad', cf. Grk. πλατύς. Hence, Mayrhofer (1986), for instance, holds that only *h2 produces aspiration on stops in Indo-Iranian. The best possible example for aspiration by *h1 is perhaps the 2.pl.pres. ending of Vedic, -tha, which corresponds to Greek -τε, thus potentially implying a form *-th1e. Vedic 2.pl.pres. -tha otherwise lacks an explanation.

The derivational suffix *-eh1 would certainly be uncommon; I cannot call to mind any other words that decisively show it. I believe that scholars in Leiden follow Pedersen (La cinquieme declenasion latine) in accepting a derivational suffix *-eh1 as the origin for Lithuanian nouns in -ė (as well as perhaps contributing to the Latin fifth declension). Others (e.g., Weiss) take the Latin fifth declension as a Latin-internal development from certain nouns that contained for other reasons (e.g., rēs and diēs), and reject the potential Baltic connection, in which case little basis would exist for the assumption of a derivational suffix *-eh1. *-eh2, in this specific instance, is also peculiar, since it primarily derives feminine nouns, and 'path' is a masculine in both Skt. and Avestan. Alternatively, *-h2 here could be a collective derivational suffix (compare Skt. rátha 'chariot' < 'set of wheels' above), in which case a 'path' would be 'a group/set of pont-' (whatever exactly a *pont- might have been -- we have no such root noun).

Further inflection off of another inflectional form (as Mark Beadles suggests) is highly unlikely -- such "decasuative" derivation is typologically unsubstantiated.

If truly pressed, I'd reconstruct *pónt-eh2- / *pn̥t-h2-, but we cannot be wholly certain.

  • 1
    So in Greek pf. 2sg. -θa < *-th2e, is there a one-off aspiration due to the laryngeal? This doesn't usually happen in Greek, does it?
    – TKR
    Oct 10, 2013 at 17:17
  • 1
    It is possible that the Greek aspiration is due to the *h2. Without assuming that at least *h2 can produce aspiration on stops, when parsed into the same syllable onset, then we lack an explanation for why Greek has -θα rather than *-τα. Weiss (2009: 53) accepts a similar formulation. In fact, we can contrast *-th2e > -θα with *pl̥th2-ú-s > πλατύς. In the former case, a perfect CóC-th2e will have a syllabification [CoC.th2e] ([CoCt.h2e] instead will give an overlong syllable and, if C = a stop, violate the Obligatory Contour Principle); in the latter, [pl̥t.h2us], and so not *πλαθύς.
    – rsandell
    Oct 10, 2013 at 21:11

This looks like what is commonly reconstructed for the instrumental case in *-Vh₁- (see e.g. Kim, 2010 lecture and even Wikipedia).

Perhaps this could be from road-INSTR, i.e. "by road"?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.