II.2. Laryngeal evolution

II.2.1. Late Proto-Indo-European

In the vocalic inventory of the current Proto-Indo-European reconstruction, the following simplified evolution paradigm is widespread (Beekes 2011; Meier-Brügger 2003; Ringe 2006; Adrados, Bernabé, and Mendoza 2010):






















































































II.2.2. Common Indo-European

A differentiation of Late Indo-European in an early, Common Indo-European (CIE), and a late, Disintegrating Indo-European (DIE) stage is necessary.

After the separation of Proto-Anatolian ca. 4500 BC, Common Indo-European developed probably in the eastern Volga-Don region of the Pontic-Caspian steppes, in the late Khvalynsk and late Repin groups, ca. 4000-3500/3300 BC (Anthony 2007; Quiles 2017). 

In this Common Indo-European phase, trends observed in the last stage of Proto-Indo-Anatolian as shown by Proto-Anatolian might have included the following:

·       Potential uvular-to-pharyngeal shift of *h2, *h3 (Weiss 2016).

·       Early merging and deletion processes (Kloekhorst 2006; Bomhard 2004):

o   PIA *h1R- and *h3R → CIE *hR

o   PIA *VHC → CIE *V̄C

o   PIA *Ho- → CIE *ho-

An auxiliary vowel was probably inserted often in certain positions, which can be reconstructed for all branches alike: *Ch1C*Ch1°C, *Ch2C*Ch2°C, *Ch3C*Ch3°C.

II.2.3. Disintegrating Indo-European

With the Disintegrating Indo-European stage we assume a period of a Northern-Southern dialectal division and internal Southern dialectal split (between Palaeo-Balkan and Pre-Indo-Iranian groups), in which the whole community remained still in contact, allowing for the spread of innovations such as a generalised vocalisation of the auxiliary vowel and the merging of laryngeals (Adrados 1998; Bomhard 2018; Koch 2013).

This linguistic scheme is compatible with the spread of the late Repin/early Yamna culture from ca. 3500/3300 BC westward into the north Pontic steppe, and eastward as a group that would develop the language ancestral to Tocharian (Anthony 2007; Quiles 2017). The time to most recent ancestor of eastern Yamna lineages show that Palaeo-Balkan and Pre-Indo-Iranian groups were already developed in this common early Yamna stage, while the common western European lineages had yet to split.

A generally agreed absence of a common Proto-Indo-European *-a (Lubotsky 1989) contrasts with the unstable vocalic system of this period.

The evolution CIE → DIE can therefore be represented as follows:

·       Colouring of *-e- by laryngeals (but long *ē more stable → uncoloured, “Eichner’s law”), although it has been proposed as a subphonemic feature (Kortlandt 2003-2004).

·       Loss of laryngeals after and before low vowels.

·       Merge of laryngeals *h1, *h2, *h3 → *h (with vocalic allophone *h̥), i.e. probably the voiceless laryngeal fricative /h/ (Szemerényi 1967; Collinge 1970; Bomhard 2004).

·       *HC- → *C- in all dialects but for Palaeo-Balkan languages (Greek, Phrygian, and probably Armenian). In this old branch, they are retained as colourised vowels (Bernabé 1975), but there are exceptions (Hinge 2007).

·       *CH°C *CHəC*ChVC*CVC, with the first phase more common in PIA, and the last one common in the dialectal split phase (see below).

·       Metathesis of *CHIC- to *CIHC-.

·       Eichner’s law.

·       Pinault’s law *-VCH → *VC - (Pinault 1982).

·       *-ERH → *-ĒR. The Saussure effect (Nussbaum 1997; Yamazaki 2009; van Beek 2011) accounts for some irregularities in the outcome of laryngeals (especially with *-h2, but not limited to it) whereby CIE dialects do not show a usual reflection of the inherited sequence. It “reflects something that happened, or failed to happen, already in the proto-language” (Lubotsky 1997):

o   *HRo- → *Rō̌-.

o   *-oRH-C- → *-oRC-.

·       *CIHV- → *CIJV-.

·       *-CR̥/IHV- → *-CR/IV- in compounds.

o   In the group *CR̥HV, a vowel can appear before the resonant, as the laryngeal disappears. That vowel is usually coincident with the vocalic output that a resonant alone would usually give in the different dialects, so it can be assumed that generally *CR̥HV*C(V)RV, although exceptions can indeed be found (Woodhouse 2011). A common example of parallel treatment within the same dialect is Greek pros/paros < *pros/p°ros (Adrados, Bernabé, and Mendoza 2010).

·       *(-)CHV- → *(-)CV- in all branches, but with some showing innovations such as aspiration before *h2, sonorants, gemmination, etc.

·       *CEHI- → *CEI-.

·       *CEHE- → *CEE-.

·       *-EH → *-Ē, with special cases for the group *HEH in Palaeo-Balkan languages (Bernabé 1975).

·       *RHC- → *RVC-, or “Beekes’ law”, with laryngeal in anlaut vocalised in most languages, and the resonant becoming consonantal.

·       *-Hs- probably evolved into geminated *-ss- in Anatolian and Greekand thus potentially Palaeo-Balkan (Ledo 2002).

II.2.4. Late Indo-European dialects

Some laryngeal reflexes reached DIE dialects differently, but still with some apparent contacts. They must have happened after the westward expansion of the Yamna culture, during the existence of a strong network of exchange between Yamna settlers:

·       Loss of word-initial laryngeals *HØ, but for Palaeo-Balkan languages, which appear to show a general output *H°→ *e, a, o (an evolution which had already begun in CIE, see above).

·       *CHC → *CHəC → Western DIE *ChaC → NWIE *CaC, as found in Italo-Celtic (Schrijver 1991; Zair 2012), Germanic (Ringe 2006), and Tocharian, as well as in Armenian (Mondon 2008) and Albanian. Alternative fate was laryngeal loss in certain environments *CC (Byrd 2010).

o   In Proto-Greek, CIE *CHəC evolved into *CaC, *CeC, *CoC depending on the nature of *H.

o   Eastern DIE *ChiC evolved into PII *CiC.

·       DIE *CR̥HiV- → NWIE *CR̥V-, as found in Italo-Celtic *CaRV, cf. Lat. cariēs < *kr̥h2-ē- (Schrijver 1991), also found in Greek and perhaps Sanskrit.

·       DIE *HJV- → NWIE *JV- as found in Italo-Celtic (Schrijver 1991; Zair 2012), Germanic (Ringe 2006), Tocharian, and also in Indo-Iranian, Armenian, and Albanian (Zair 2012).

·       DIE *R̥HC- → NWIE *RǎC-, as found in Italo-Celtic (Zair 2012), cf. Lat. lǎbāre (Schrijver 1991), and Germanic (Beekes 1988).

·       DIE *HIC- → NWIE *IC-, as found in Italo-Celtic (Schrijver 1991; Zair 2012), Germanic (Ringe 2006), and Tocharian, as well as Albanian, Indo-Iranian.

·       DIE *CEHR̥- → NWIE *CER-, with an unclear intermediate development, but necessarily parallel in Italo-Celtic, Germanic, and Indo-Iranian (Zair 2012).

·       DIE *CIHR̥- → NWIE *CIJR̥- in Italo-Celtic, Indo-Iranian (Schrijver 1991; Zair 2012).

·       DIE *-IH → NWIE *-Ī as found in Italo-Celtic and Germanic, as well as Albanian and Indo-Iranian. Vocalisation in Greek-Armenian and Tocharian.

o   CIE *-ih2 ending in auslaut had an alternative form *-°h2, DIE *-ih/-əh, which could produce *-ī, *-ā̌, alternating forms that are found even within the same dialect.

·       Dybo’s rule in North-West Indo-European: short vowels as output of *CHIC-, or *CIHC-, with long vowels remaining when stressed, but shortened in pretonic syllables, as found in Proto-Italic, Proto-Celtic and Proto-Germanic (Zair 2012; Garnier 2015).

The contentious Osthoff’s law, which affected all DIE branches but for the eastern territories (languages ancestral to Tocharian and Indo-Iranian), must have been a general trend after the start of the Yamna expansion to the west, i.e. after ca. 3100/3000 BC.

When *H is in a post-plosive, prevocalic position, the consonantal nature of the laryngeal values is further shown *CHVC → *ChVC. This is more frequent in PII, cf. *pl̥th2ú- → Ved. pr̥thú-, but it appears also in the perfect endings, cf. Gk. oistha. This development might have happened in North-West Indo-European, and later devoiced to *CVC.

The proposed glottal stop *ʔ reconstructed for the merged laryngeal in scattered early proto-language remains (Mayrhofer 2005; Lubotsky 2018) is the most likely phonetic evolution of DIE *h before fully disappearing.

II.2.5. Laryngeal remnants in early Indo-European proto-languages?

II.2.5.1. Glottal stops

Apparently a reflect of consonantal laryngeals is found between non-high vowels as hiatuses (or glottal stops) in the oldest Indo-Iranian languages, in Homeric Greek (Lindeman 1987), and potentially in Germanic (Connolly 1980). However, there is not enough evidence to explain such irregularities by laryngeal remains instead of by the more obvious licence in metric (Kümmel 2014).


In old compositions, some final short vowels are found as heavy syllables, cf. Skt. deví etu, or vocat. vṛki, tanu (Lindeman 1987; Beekes 1982): “The Vedic phrase devyètu, i.e. devì etu is best understandable if we suppose that devī́ ‘goddess’ still contained the laryngeal form *dewíh (with *-ih<*-ih2) at the time of the formulation of the verse in question. In the phase *-íh it was possible for the laryngeal simply to disappear before a vowel” (Meier-Brügger 2003). Other common example used is *u̯ŕ̥kih.

The laryngeal survival in Proto-Indo-Iranian finds limited support for a preservation in intervocalic position in the Gāϑās and in the Vedas (Gippert 1996), which is also controversial (Kümmel 2014; Beguš 2015).

It is not justified, then, why these examples must represent a sort of unwritten laryngeal, and not an effect of it, i.e. a laryngeal hiatus or glottal stop, from older two-word sandhis that behave as a single compound word.

Interesting is also that they are in fact from words already alternating in CIE *-ih2/*-°h2, or DIE *-ih/-əh, which reflect different syllabification in Indo-Iranian vs. Greek and Tocharian, whilst “[t]he source of the difference is not fully understood”(Fortson 2010).

It has long been recognised that the treatment of word-final laryngeals shows a strong tendency to disappear (so e.g. in Hittite), and most of the time it appears associated with morphological elements (Adrados, Bernabé, and Mendoza 2010).

In line with this problem is that the expected case of *-aH stems is missing, which makes it less likely that Indo-Iranian examples come from a common hypothetic PII stage in which a word-final *-H had not still disappeared, and more likely that theseas well as the Homeric Greek and Germanic exampleswere particular cases with certain phonetic rules in action. If they could be traced back to an ancestral stage of each language, they would be frozen remains (probably of a glottal stop) in certain formal expressions.


The sandhi variant in *-aH is found in Greek and Old Church Slavonic (Meier-Brügger 2003; Ringe 2006): In both “clear traces are missing that would confirm a PIE ablaut with full grade *-eh2- and zero grade *-h2- (…)

That is why it appears as if the differentiation between the nominative and vocative singular in this case could be traced to sandhi-influenced double forms that were common at a time when the stems were still composed of *-ah2, and the contraction *-ah2- →*-ā- had not yet occurred.” For an original zero grade alternation in Greek, Latin, Slavic, and Tocharian, see (Kortlandt 2017)

This was rejected by Szemerényi (1999): “The shortening of the original IE ending -ā to -ă is regular, as the voc., if used at the beginning of a sentence or alone, was accented on the first syllable but was otherwise enclitic and unaccented; a derivation from -ah with the assumption of a prevocalic sandhi variant in -a fails therefore to explain the shortening.”

Laryngeal hiatus

The Rig Veda preserves many words that could be interpreted as though some remnant of a laryngeal, probably a glottal stop, was still present between vowels, a phenomenon called laryngeal hiatus. For example, Skt. vā́tas ‘wind’ must sometimes scan trisyllabically as *va’atas, cf. OAv. va.ata-, possibly from PII *áˀata- < Pre-PII *ehn̥tos < CIE *h2eh1-n̥t-o- → DIE *ehn̥tos → NWIE *entos; cf. Lat. ventus, W gwynt,  Gmc. *indaz; but PT *entë < *ēntos.

Compare also potential examples Ved. *ca-kar-ˀa (the *ʔ still preserved in the period of the activity of Brugmann’s law), or Ved. náus < *naˀus. Such finds would support a vocalisation of CIE *n̥, *m̥ PII *a earlier than the loss of laryngeal (or glottal stop) in that environment.

II.2.5.2. *CR̥HC

The group *CR̥HC is explained differently for the individual dialects without a general paradigm, with dialectal outputs explained as (Beekes 2011; Meier-Brügger 2003):

·       *CR°hC into Proto-Tocharian *CRaC, Italic and Celtic *CRāC, Ligurian *CRaC/CRāC (Prósper 2018), Proto-Armenian *CRaC, i.e. an output similar to *CHC in these dialects, which points either to an ancient trend to a-vocalism (i.e. NWIE *CRahC), or to an assimilation of the group to the output of *CHC.

o   PIE *Cwl̥HC into NWIE *Cw°lhC > *CulC > *CulC, as found in Italic and Celtic, in common with the general Germanic output (Prósper 2017).

o   Germanic *CR̥C. There is difficulty reconstructing the potentially old Northern variant *-HC- *-aC- (Müller 2007), among them the scarcity of surviving traces of laryngeals (Fortson 2010).

o   Balto-Slavic *CVRC/CRC, with the same vocalic output as *CR̥C, and distinction by accentuation (Darden 1990), which would mean a merging of the laryngeal posterior to the vocalisation of sonorants.

·       In Proto-Greek, the original laryngeal determined the vocalic output: e.g. *r̥h1*r̥°h1*reh. *RH gives Rāx when the accent follows and axRax when the accent is on the resonant.

A common example of the different dialectal outputs of the *CRHC model in PIE *gn̥h1-- ‘created, born’:

·       Vedic źātá- < PII *ātó- < *ahtó- < *gjˀ-, which means that the laryngeal merged after the evolution DIE * → PII *a.

·       DIE *gnəˀ-; cf. for the same intermediate grade PGk. *gnētó- < *gnəh1-, but Armenian cnaw < *gnahtó-.

·       Pre-NWIE *gn̥ˀ- into PT *gnató- < **gnaˀ-, Ita.-Cel. *gnātó- < **gnaˀ-, Gmc. *kunda-< **gn̥tó-, Bal.-Sla. *gìnta-?< **gìnˀta-? per Hirt’s law, following the *pl̥h1nó- example (Darden 1990).

The palma rule in Latin, which in turn seemed to have distinct developments depending on whether CIE *CRH̥C- sequences were accented or not (Höfler 2017), points more strongly to the unstable nature of compounds including sonorants.

A common explanation of certain alternating forms found even in the same dialect is based on late dialectal morphological and analogical changes (Adrados, Bernabé, and Mendoza 2010): “The different solutions in this case depend solely on two factors: a) if there are one or two auxiliary vowels to facilitate the pronunciation of this group; b) the place where they appear.” So e.g. a group *CR̥hC could be pronounced in DIE with one vowel, *CR°hC or *C°RhC, or with two, *C°R°hC, *C°Rh°C, or *CR°h°C.

Compounds with sonorants like *CR̥C, *RR̥V, *TRV, and *SMV among others are known to behave differently even within the same languages and proto-languages (Adrados, Bernabé, and Mendoza 2010). It is only natural that DIE or NWIE groups that should be traced back to *CRV and *VRC could similarly show unstable outputs that confound any attempt to obtain a stable sound law. That ‘instability’ solution could account for all variants found in the different branches, and within them.

Different outputs are proposed for *CRH groups before certain vowels (Lubotsky 1997): “It is clear that the “short” reflexes are due to laryngeal loss in an unaccented position, but the chronology of this loss is not easy to determine. If the laryngeal loss had already occurred in PIIr., we have to assume that PIIr. *CruV subsequently yielded CurvV in Sanskrit. The major problem we face is that the evidence for the phonetically regular outcome of *CriV and *CruV in Indo-Iranian is meager and partly conflicting.”

II.2.5.3. Cogwill’s law

The contentious Cogwill’s law seems to be a late, independent development reconstructed for three Proto-Germanic forms, whereby *h3 and possibly *h2 would turn into Proto-Germanic *k when directly preceded by a sonorant and followed by *. This would need an evolution CIE *h3w → *gw that remained only in Germanic, and is as such a poor explanation of these few peculiar developments.

II.2.5.4. Language contacts

It should be expected that surviving laryngeals in North-West Indo-European should be seen in borrowings attributed to the language (or its early dialects) or be necessary to reconstruct its proto-languages. However:

·       Merge of laryngeals into a single DIE phonemeprobably the voiceless laryngeal fricative /h/is supported by the Late Proto-Uralic borrowings with a single phoneme (*ś), and by the advanced process of laryngeal loss attested in the borrowings (see above §2.3.2. Late Uralic–Late Indo-European contacts).

·       No laryngeal is necessary for Proto-Italic or Proto-Celtic reconstructions (de Vaan 2008; Matasović 2009).

·       No laryngeal can be reconstructed for Palaeo-Germanic (Pre-Germanic to Proto-Germanic) loanwords (Schrijver 2014), either in Proto-Finno-Samic or in later contacts (see above § Palaeo-Germanic borrowings).

·       Against the reconstruction in Kortlandt (2016), no laryngeal can be reconstructed for loanwords of Proto-Balto-Slavic, Early Proto-Baltic, or Proto-Slavic into western Finno-Volgaic languages (see above §4.18.4. Contacts with Balto-Slavic).

II.2.6. Laryngeal reflexes in North-West Indo-European

Assuming a common North-West Indo-European community and language, we can establish these common developments, from which to derive changes in daughter proto-languages Italic, Celtic, Germanic, and Balto-Slavic.

II.2.6.1. Initially before consonant or resonant

Initially before PIE consonants or resonants laryngeals are lost. This is the result in most historic languages, except in Greek, Armenian and Anatolian, where they are preserved with some limitations in all of them.

·       PIE *h1rudhrós → NWIE *rudhrós, ‘red’; cf. Gk. eruthrós, Lat. ruber, Goth. rauþs.

·       PIE *h1smós(i) → NWIE *smos(i) ‘we are’.

·       PIE *h1imós(i) → NWIE *imós(i) ‘we go’.

·       PIE *h2iu̯-u̯Hens- → NWIE *i̯úōn‘young’, cf. Lat. iuvenis, OInd., i̯úvan-.

·       PIE *h2sters → NWIE *stḗr ‘star’, but cf. Gk. astḗr.

·       PIE *h3lígos → NWIE *lígos ‘little, scarce’, cf. Lith. ligà ‘illness’, Gk. olígos, Arm. aɫk`‘poor’.

·       PIE *h3minéghmi / *h3mínghoh2 → NWIE *minéghmi / mínghō ‘I piss’.

·       PIE *h3r̥néumi → NWIE *r̥néumi ‘I move’.

·       PIE *h1úpo → NWIE *úpo ‘under’.

·       PIE *h2/3upélos → NWIE *upélos ‘evil’.

·       PIE *h2u̯ésoh2 → NWIE *ésō ‘I stay’.

·       PIE *h1pi̯oh2 → NWIE *ápi̯ō, ‘I reach’.

·       PIE *h3bhruHs → NWIE *bhrūs, ‘eyebrow’.

II.2.6.2. Initially before vowel

·       PIE *h1ésmi → NWIE *ésmi ‘I am’.

·       PIE *h1ómHsos → NWIE *ómsos ‘shoulder’.

·       PIE *h1édsi → NWIE *édsi ‘you eat’.

·       PIE *h1óngw-ols → NWIE *óngwōl ‘coal’.

·       PIE *h1eíti → NWIE *eíti ‘goes’.

·       PIE *h1óimos → NWIE *óimos ‘march’.

·       PIE *h1egóh2 → NWIE *egṓ ‘I’.

·       PIE *h1ógwhis → NWIE *ógwhis ‘worm, snake’.

·       PIE *h2énus → NWIE *ánus ‘grandmother’.

·       PIE *h1órsos → NWIE *órsos ‘tail’.

·       PIE *h2égeti → NWIE *ágeti ‘bears’.

·       PIE *h2ógmos → NWIE *ógmos ‘track’.

·       PIE *h3enos → NWIE *ónos ‘load’.

·       PIE *h2eiu̯ós → NWIE *aiu̯ós ‘lifetime-lasting’.

·       PIE *h3éu̯is → NWIE *óu̯is ‘sheep’.

·       PIE *h2óiu̯u → NWIE *óiu̯u ‘vital energy’.

·       PIE *h3épos → NWIE *ópos ‘work’.

·       PIE *h3ólh1neh2 → NWIE *ṓlnā‘elbow’.

·       PIE *h3éidos → NWIE *óidos ‘tumor’.

·       PIE *h3ókwo- → NWIE *ókwos ‘eye’.

II. Special cases: initial vocalisation

·       PIE *h2íh2si̯eh2 NWIE *óisi̯ā ‘rudder’.

·       PIE *h2u̯H1i- → NWIE *áu̯is ‘bird’. 

·       PIE *h2ń̥h2e → NWIE *ána‘on’.

·       PIE *h2ŕ̥gn̥tom → NWIE *árgn̥tom ‘silver’.

·       PIE *(H?)álbhos → NWIE *álbhos, ‘white’, cf. Hitt. alpa-, ‘cloud’.

II.2.6.3. Double initial laryngeals

·       PIE *hºhw2etmos → NWIE *átmos ‘breath’.

·       PIE *h2/3eh3imi → NWIE *ṓimi‘I believe’.

·       PIE *hºhw2etméns → NWIE *ātmḗn ‘spirit’.

·       PIE *h3éhw3smi → NWIE *ṓsmi‘I open’.

·       PIE *h3ºhw3sis → NWIE *óusis ‘ear’.

·       PIE *Héhi2dmi → NWIE *ā́dmi‘I dry’.

·       PIE *H1óH2u̯i̯om → NWIE *ṓu̯ii̯om ‘egg’.

·       PIE *h3ehw3s → NWIE *ōs ‘mouth’.

·       PIE *Hº3eHi2kris → NWIE *ókris ‘summit’.

·       PIE *HoHwo1los → NWIE *áulos ‘tube’.

II.2.6.4. Internally before a vowel

·       PIE *dhh1ent → NWIE *dhent ‘they placed’.

·       PIE *sth2ent → NWIE *stant ‘they stood’.

·       PIE *dh3ent → NWIE *dont ‘they gave’.

·       PIE *h2u̯oh1dhh1éi̯oh2 → NWIE *u̯ōdhéi̯ō ‘I push’.

·       PIE *skwh2ólos → NWIE *skwólos ‘stumbling’ (noun).

·       PIE *h2u̯óh1dhh1onom → NWIE *u̯ṓdhonom‘pushing’.

·       PIE *skélh2onom → NWIE *skélonom ‘splitting’.

·       PIE *somh2ós → NWIE *somós ‘same’.

·       PIE *réth2onti → NWIE *rétonti ‘they run’.

·       PIE *h1ésHos → NWIE *ésos ‘master, lord’.                

II.2.6.5. Internally before vowel, after resonant

·       PIE *mélh2esi → NWIE *mélesi (not xmélasi) ‘you grind’.

·       PIE *sténh2esi → NWIE *sténesi (not xsténasi) ‘you resound’.

II.2.6.6. Second position in compounds

·       PIE *neu̯ognh1ós → NWIE *neu̯ognós ‘newly born’.

·       PIE *kwékwlh1o- → NWIE *kwékwlom ‘wheel’.

II.2.6.7 Internally after a vowel

·       PIE *h1réh1poh2 → NWIE *́‘I creep’.

·       PIE *Hréh3doh2 → NWIE *́‘gnaw’.

·       PIE *meh2térs → NWIE *mātḗr ‘mother’.

·       PIE *meh2is → NWIE *māis ‘more’.

·       PIE *péh2smi → NWIE *́smi‘I heed’.

·       PIE *préh2tis → NWIE *prā́tos‘sale’.

·       PIE *dhidhéh1mi → NWIE *dhidhḗmi‘I put’.

·       PIE *gígnh1H2ei → NWIE *gígnāi ‘I am born’.

·       PIE *stistéh2mi → NWIE *stistā́mi‘I stand’.

·       PIE *didéh3mi → NWIE *didṓmi‘I give’.

·       PIE *h3néhu3mn̥ → NWIE *́mn̥‘name’.

·       PIE *pr̥néh2mi → NWIE *pr̥nā́mi‘I sell’.

·       PIE *soh1déi̯oh2 → NWIE *sōdéi̯ō ‘I settle’.

·       PIE *dhoh1mós → NWIE *dhōmós ‘thesis, opinion’.

·       PIE *stóh2nom → NWIE *stā́nom‘place’.

·       PIE *stóh2los → NWIE *stṓlos‘table’.

II. Special case: Osthoff’s law

·       PIE *h2u̯éh1n̥tos → NWIE *u̯éntos ‘wind’.

·       PIE *meh1msóm → NWIE *̌msóm‘meat’.

II. Special case: Stang’s law

·       PIE *pipéh3imi → NWIE *pipṓmi‘I drink’. Extended to other forms:

·       PIE *pipéh3iti → NWIE *pipṓti‘he drinks’.

II. Special case: laryngal metathesis

·       PIE **spi̯Hutós → *spi̯uHtós → NWIE *spi̯ūtós ‘spat’ (part.). 

·       PIE **bhh2utós → *bhuh2tós → NWIE *bhūtós ‘been’.

·       PIE **siHutós → *si̯uHtós → NWIE *si̯ūtós ‘sewn’.

·       PIE **lh3itós → *lih3tós → NWIE *lītós ‘poured’.

·       PIE **ph3ilós → *pih3lós → NWIE *pīlós ‘having drunk’.

·       PIE *ph3itós → NWIE *pītós ‘drunk’.

·       PIE *pHutós → NWIE *pūtós ‘cleaned’.

·       PIE *liHtós → NWIE *lītós ‘poured’.

·       PIE *gwiHu̯ós → NWIE *gwīu̯ós ‘alive’.

II.2.6.8. Internally between two consonants

·       PIE *ph2térs → NWIE *pár‘father’.

·       PIE *kh3tós → NWIE *katós‘sharp’.

·       PIE *mh2déh1i̯oh2 → NWIE *madḗi̯ō‘I am wet’.

·       PIE *h2u̯oh1dhh1tós → NWIE *u̯odhatós ‘pushed’.

·       PIE *pr̥nh2mós(i) → NWIE *pr̥namós(i)‘we sell’.

·       PIE *dhidhh1mós(i) → NWIE *dhidhamós(i)‘we put’.

·       PIE *stísth2mos(i) → NWIE *stístamos(i) ‘we stand’.

·       PIE *dídh3mos(i) → NWIE *dídamos(i) ‘we give’.

·       PIE *sth2tós → NWIE *statós ‘stood’.

·       PIE *peph3 → NWIE *pepaté ‘keep drinking’ (2nd pl.).

II. Special case: concave syllable between two consonants

·       PIE **sh1déh1i̯oh2 → NWIE *sedḗi̯ō ‘am seated’.

·       PIE **lh1góm → NWIE *legóm ‘I collected’.

·       PIE **lh1bhóm → NWIE *labhóm ‘I caught’.

·       PIE **luh3óm → *lh3u̯óm → NWIE *lou̯óm ‘I washed’.

II.2.6.9. Internally between consonant and resonant or between two resonants

II. Generalised Saussure effect

Some examples are affected by “Pinault’s law” (Byrd 2015).

·       PIE *tórh1mos → NWIE *tórmos ‘hole’.

·       PIE *kl̥mh2-rós → NWIE *klamrós ‘weak’.

·       PIE *gémh1ro- → NWIE *gémros ‘son-in-law’.

·       PIE *(s)porHnós → NWIE *pornós ‘feather’.

·       PIE *pélh1u → NWIE *pélu ‘much’.

·       PIE *bhólh1/2i̯om → NWIE *bhóli̯om ‘leaf’.

·       PIE *h2érh3u̯r̥ → NWIE *áru̯ar‘grain’.

·       PIE *míHi̯etoi → NWIE *mī́i̯etoi ‘decreases’.

·       PIE *kwríh2tor → NWIE *kẃtor‘was bought’.

·       PIE *dhúh2lis → NWIE *dhū́lis‘soot’.

·       PIE *bhh2úi̯etoi → *bhúh2i̯etoi → NWIE *bhū́i̯etoi ‘becomes, begins’.

·       PIE *léu̯h2trom/ *léh2u̯trom → NWIE *lóutrom ‘bath’.

·       PIE *skélh2tis → NWIE *skéltis ‘splitting’.

·       PIE *skl̥h2i̯oh2- → NWIE *skl̥i̯ō ‘I split’.

·       PIE *térh1i̯oh2 → NWIE *téri̯ō ‘I rub’.

·       PIE *sokwh2i̯ós → NWIE *sokwi̯ós ‘allied’.

·       PIE *megh2i̯ós → NWIE *megi̯ós ‘bigger’.

·       PIE *kn̥h1i̯ó- → NWIE *kani̯ós‘recent’.

·       PIE *gń̥h1i̯etoi (=*gígnetoi) → NWIE *gnai̯etoi ‘is born’.

·       PIE *sth2i̯éh1m → NWIE *stai̯ḗm ‘I would stand’ (aor.).

·       PIE *sth2ih1nt → NWIE *stai̯ī́nt ‘they would stand’.   

·       PIE *dh3i̯éh1m → NWIE *dai̯ḗm ‘I would give’ (aor.).

·       PIE *dh2ih1nt → NWIE *dai̯ī́nt ‘they would give’.

·       PIE *h2u̯oh1dhh1i̯óm → NWIE *odhai̯óm ‘I pushed’ (aor).

II. Special case: Retention of laryngeal

·       PIE *h2énh1mos → NWIE *ánˀmos ‘breath, soul’, cf. Toch. A āñcäm (obl. āñm-), B āñme PT *āñc(ä)me 'self, soul', Lat. animus, Osc. anamúm, OIr. animm, OFris. omma.

·       PIE *kerh2srom → NWIE *kerˀsrom ‘brain’.

·       PIE *temh1sreh2e → NWIE *temˀsrā ‘darkness’, cf. OInd. tamisra, Lat. Tenebrae. Compare also e.g. PIE *temHs- → OHG demar, ‘twilight’. However, there are also reasons to reject such reconstruction in favour of PIE *temHosó-, as OInd. *tamasá-, ‘dark-coloured’ (Müller 2007).

II. Special case: Internal vocalisation

·       PIE *sh2neh2mi NWIE *sánāmi‘I satiate’.

·       PIE *térh1dhrom → NWIE *téredhrom ‘auger’, cf. Lat. terebra, Gk. téretron, OIr. Tarathar.

·       PIE *kr̥tús → NWIE *kartús ‘strong’.

II.2.6.10. Blocked laryngeal with a resonant

The regular reflex of *CR̥HC in Italic and Celtic is *CRāC no matter which laryngeal is involved. The ē found in Italic (in Lat. plēnus, Umb. plener) and partially Celtic (in Corn. luen, Bret. leun) is likely an especial dissimilation not to confuse the word with *plānos. Analogy with the corresponding perfect is the common explanation for other results different from ā, as found in certain participles; cf. nōtus, sprētus, crētus, etc. (Bolotov 2012).

·       PIE *pl̥h1nós → NWIE *pl̥ˀnós‘full’, cf. Ita. *plānos, Cel.*hlēn-, *hlān-, Gmc. *φullaz, Bal. *pîlna-, Sla. *pĭlnŭ.

·       PIE *gr̥h2nóm → NWIE *gr̥ˀnóm‘corn’, cf. Lat. grānum, OIr. grān, Gmc. *kurna-, Bal. ǯirniā̃, Sla. *zĭrno.

·       PIE *gn̥h1tós → NWIE *gn̥ˀtós‘born’, cf. Lat. gnātus, Umb. natine, OIr. cned, Gaul cintu-, Gmc. *kundaz, Bal. *ǯnō̂ta-.

·       PIE *gn̥h3tós → NWIE *gn̥ˀtós‘known’, cf. Toch. A. āknats, aknātsa, Lat nōtus (but Lat. gnāvus<*gn̥h3u̯ós, ‘wise’), OIr. gnāth, Gmc. *kundaz, Bal. *ǯint-, Sla. *žĭn-.

·       PIE *pŕ̥h2u̯os → NWIE *pŕ̥ˀu̯os‘first’, cf. PT *päre, Sla. *pĭrvŭ.

·       PIE *str̥h3tós → NWIE *str̥ˀtós‘strewn’, cf. Lat. strātus, OIr. sreth, Bal. *stir̂tā̂, Sla. *-stĭrtŭ.

·       PIE *gwr̥h2ús → NWIE gwr̥ˀu̯ús‘heavy’, cf. Lat. gravis (brutus), MIr. bair (bruth), Gmc. *kuru-, Bal. *grū̂ta-.

·       PIE *pr̥h2tós → NWIE *pr̥ˀtós‘sold’.

·       PIE *kŕ̥h2tis → NWIE *kŕ̥ˀtis‘wickerwork’.

·       PIE *pépr̥h3th2ei → NWIE *pépr̥ˀtai ‘you got production’.

·       PIE *pépr̥h2dhi → NWIE *pépr̥ˀdhi‘keep selling!’

However, no evidence for laryngeal after * can be traced in:

·       PIE *u̯éu̯r̥th2ei → NWIE *u̯éu̯r̥tai ‘you got found’.

II. Special case: Laryngeal lost by generalised Saussure effect

For example in cases of *Cred.HRC such as:

·       *HRoC → *RoC in Proto-Greek; in NWIE the general rule is laryngeal loss for any vocalism:

o   PIE *h3meigh- ‘to urinate’ → NWIE *méighō, *minghō, *moighós, but cf. Gk. omeíkhō / moikhós.

·       *CoRHC → *CoRC:

o   PIE *kólHnis → *NWIE kólnis ‘hill’.

o   PIE *sólh2o- → *sólo- ‘all, the whole’.

II. Special case: With brief resulting vowel

·       PIE *pŕ̥Htis → NWIE *prátis ‘fern’.

·       PIE *kń̥h2meh2 → NWIE *kánmā ‘leg’.

II. Special case: Lost laryngeal in a compound

·       PIE *kompl̥h1nós → NWIE *kompl̥nós ‘extremely full’.

·       PIE *komgnh3tós → MID *komgn̥tós ‘completely known’, cf. Lat. cognitus.

II. Special case: Palma rule

·       PIE *pĺ̥h2meh2 → NWIE *pĺ̥‘palm’.

·       PIE *pĺ̥h2seh2 → NWIE *pĺ̥‘mantle, covering’.

·       PIE *(s)pŕ̥hxseh2 → NWIE pŕ̥sā “winged animal, sparrow’.

Similar cases:

·       PIE *h2/3u̯ĺ̥h1-neh2 → NWIE *u̯ĺ̥ˀ‘wool’, cf. OInd. ū́rṇā-.

·       PIE *pl̥h2́goh2 → NWIE *plń̥‘I beat’.

·       PIE *ghl̥h3tóm → NWIE *ghl̥tóm (not xghlōtóm) ‘gold’.

·       PIE *mĺ̥h2dhh1os → NWIE *mĺ̥dhos‘mild’.

·       PIE *skl̥h2/3tós → NWIE *skl̥tós (not xsklV̄tós) ‘split’ (part.).

·       PIE *pr̥nh2énti → NWIE *pr̥nánti ‘they sell’.

II.2.6.11. Final position before a vowel

·       PIE *u̯óidh2e → NWIE *oida ‘I know’.

·       PIE *u̯óid-th2e → NWIE *oista ‘you know’, but cf. Gk. óistha, OInd. vettha.

II. Special case: vocalisation of a laryngeal appendix

·       PIE *ste-stohw2h2e → NWIE *stéstōu̯a ‘I am standing’.

·       PIE *dhe-dhohi1h2e → NWIE *dhédhōi̯a ‘I have put’.

·       PIE *de-dohu3h2e → NWIE *dédōu̯a ‘I have given’.

II.2.6.12. Final position after a vowel

·       PIE *dígheh2 NWIE *díghā ‘goat’.

·       PIE *h1roh12u̯éh2 NWIE *rōu̯ā́.

·       PIE *héhi2seh2 NWIE *ā́ ‘altar’.

·       PIE *gwéneh2 NWIE *gwénā ‘woman’.

·       PIE *déikoh2 NWIE *déikō ‘I show’.

·       PIE *wĺ̥kwoeh1 → NWIE *u̯ĺ̥kwō, *u̯ĺ̥kwo ‘with (the) wolf’.

II.2.6.13. Final position after a consonant or a resonant

·       PIE *pleh1i̯ósh2 → NWIE *plēi̯ósa ‘more’.

·       PIE *megh2 → NWIE *méga ‘big’.

·       PIE *Hith2 → NWIE *íta ‘so’.

·       PIE *h1́dhh2 → NWIE *́dha‘then’.

·       PIE *h3néhu3monh2 → NWIE *nṓmona ‘names’.

·       PIE *u̯l̥kwíh2 NWIE *kwī́ ‘she wolf’.

·       PIE *u̯ŕ̥h2dih2 → NWIE *u̯ŕ̥ˀ‘root’.

·       PIE *bhh2mésdhh2 NWIE *bhamésdha‘we speak’.

·       PIE *bhh2u̯ésdhh2 → NWIE *bhaésdha‘we two speak’.

II.2.6.14. Kortlandt effect

·       PIE **úddhh1 → *úh1dhh1 → NWIE *ū́dh‘udder’.

·       PIE *dédr(H)is → NWIE *́ris‘separation’, cf. Gk. ́ris ‘dispute’, OInd. veṇu-dāri-.

·       PIE *tr̥dtós → NWIE *tr̥ˀtós (<**tr̥h1tós) ‘pierced’.

·       PIE **médmi → *meh1mi → NWIE *́mi‘I measure’.

·       PIE *h2éhi2dmi → **h2eh12h1mi → NWIE *ā́mi, *ādmi, cf. *aidhō ‘I burn’.

·       PIE *ghdéh2u̯r̥ → **ghh1éh2u̯r̥ → NWIE *ghḗu̯r̥‘emptiness’.

·       PIE *bhidtrós → **bhih1trós → NWIE *bhītrós ‘trunk’.

II. Exceptions

·       PIE *penkwédkm̥th2 → NWIE *penkwédkm̥ta ‘fifty’, but cf. OInd. pañčāśát- <*penkwéh1km̥th2.

·       PIE *h2ed → NWIE *ad ‘at, to’, but cf. OInd. ā < *h2eh1.

·       PIE *Hud → NWIE *ud ‘outside’.

II.2.6.15. Consonantal change

·       PIE *piph3oh2 → NWIE *píbō ‘I drink’.

II.2.6.16. Martinet’s rule

·       PIE *h3ésteh2? → NWIE *kóstā ‘rib’.

·       PIE *dhh1Hoh2 → NWIE *dháki̯ō ‘I do’.