Marek Majer
Uniwersytet Łódzki (Univerza v Lodžu)
marek.majer@uni.lodz.pl
Rafał Szeptyński
Instytut Języka Polskiego Polskiej Akademii Nauk
(Inštitut za poljski jezik Poljske akademije znanosti)
rafal.szeptynski@ijp.pan.pl 
Slavistična revija 72/3 (2024): 273–290
UDK 811.16'367.625 
DOI 10.57589/srl.v72i3.4180
Tip 1.01
Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
of the Slavic Nasal Suffix. The Diachronic Background and the 
Rise of the Common Slavic Variation
In a previous study we showed that in some peripheral dialects of Common Slavic, the aorist/
infinitive suffix of Class II verbs was *-nu-. In the present study, we focus on the diachronic 
explanation of the variation *-nǫ- ~ *-ny- ~ *-nu- in this suffix. We propose that its rise was 
facilitated by CSl phonological developments in final syllables, at the juncture of stem mark-
ers and the PIE endings: 2sg *-neh₂-s (or possibly 2sg *-neh₂-s(-s) and 3sg *-neh₂-s-t) > CSl 
*-ny; (possibly) 1sg *-new-m > CSl *-nǫ; 3sg and word-internally *-new-(t) > CSl *-nu(-). The 
resulting allomorphy could be leveled in different directions, whereby the causes traditionally 
invoked to explain *-nǫ- and *-ny- (perseverative nasalization, analogy to other verbal stems) 
played a role too.1
Keywords: Common Slavic, class II verbs, nasal suffix, Proto-Indo-European, Auslautgesetze
Splošnoslovansko *-nǫ-, *-ny- ali *-nu-? Nov pogled na 
rekonstrukcijo pripone glagolov II. vrste. Diahrono ozadje in 
nastanek splošnoslovanskih različic
V prejšnji študiji sva pokazala, da je nedoločniška/aoristna pripona splošnoslovanskih 
glagolov II. vrste na nekaterih območjih imela obliko *-nu-. V pričujočem prispevku se osre-
dotočava na diahrono razlago različic *-nǫ- ~ *-ny- ~ *-nu- v tej priponi. Predlagava, da je bil 
nastanek le-teh pogojen s spsl. fonološkimi procesi v končnih zlogih, kjer so osnovi prvotno 
sledile praindoevropske končnice: 2sg *-neh₂-s (ali mogoče 2sg *-neh₂-s(-s) in 3sg *-neh₂-s-t) > 
spsl. *-ny; (mogoče) 1sg *-new-m > spsl. *-nǫ; 3sg ter znotraj besede *-new-(t) > spsl. *-nu(-). 
Nastala alomorfija je privedla do posplošitev, pomembno vlogo so pa igrali tudi drugi dejavniki 
(progresivna nazalizacija, analogija po drugih glagolskih priponah), na katere se tradicionalno 
sklicujejo pri razlagi *-nǫ- in *-ny-.
Ključne besede: splošna slovanščina, glagoli II. vrste, nazalna pripona, praindoevropščina, 
fonološki razvoj v končnih zlogih 
1 The article has been written under the research project financed by the National Science Centre (Poland) 
decision number: 2019/33/B/HS2/02965. The research discussed here was presented at two conferences: 17th 
Annual Meeting of the Slavic Linguistics Society (SLS17), Sapporo, and 4. Slovenski dialektološki posvet 
(SDP4), Ljubljana, both in September 2022. We are grateful to the audiences of these events for their helpful 
comments and suggestions. Besides, we would like to thank Miguel Villanueva Svensson and Florian Wandl 
for useful observations on an earlier version of the manuscript. All interpretations and opinions are our own.
Slavistična revija, letnik 72/2024, št. 3, julij–september274
1 Basic premises
This article constitutes the continuation of our discussion on the formal evolution 
of the Class II suffix in Slavic published in the preceding volume (Szeptyński, Majer 
2024). Let us restate that we are dealing with the formal evolution of the infinitive/aorist 
suffix of (most) Class II verbs, as in OCS drъznǫti ‘dare.inf’, drъznǫšę ‘dare.aor.3pl’, 
minǫti ‘pass.inf’, minǫ ‘pass.aor.3sg.’ This is the sole entity we have in mind when 
speaking of ‘the nasal suffix’ or ‘*-nV-’. We are not concerned with the formation of 
the present stem or the functional properties of Class II, the morphological distribution 
of its various subtypes, etc. – topics on which plentiful research is available (see the 
literature we list in Szeptyński, Majer 2024: 132).
As noted in our previous study (Szeptyński, Majer 2024: 132-3), research in recent 
years (Andersen 1999, Pronk 2022, and predecessors) has emphasized that the unitary 
reconstruction CSl *-nǫ- is inadequate here, because – although it agrees with the 
majority of the evidence (OCS -nǫ-, Pol -ną-, Plb -nǫ-, etc.) – it does not do justice 
to some varieties of Slavic which continue a preform *-ny- instead (standard Sln -ni-, 
B/C/M/S dial -ni-, possibly USorb -ny-, Pol dial -ny-, Plb -nӗ-, etc.). More information 
on the delimitation of these two variants in the modern material, in light of the present 
state of research, can be found in Szeptyński, Majer (2024: 132-3), where the relevant 
bibliography is also provided.
We claim – as presented in detail in Szeptyński, Majer (2024: 135-46) – that this 
updated picture, incorporating the variation *-nǫ- ~ *-ny-, is still incomplete, as it can-
not explain the dialectal Slovene evidence. We believe to have shown that, for certain 
varieties of Slovene, the only input form that generates the attested shape of the nasal 
suffix is CSl *-nu-. This fact has been hidden from plain sight because of the particularly 
convoluted situation in this language. Most Slovene dialects continue CSl *-ny- (thus 
also standard Sln -ni-), while a substantial part of the north and the north-east has *-nǫ-; 
however, in two peripheral zones – which at least regarding the early isoglosses stand 
out as archaic not only within Slovene, but also more generally in Slavic – the form 
continued is *-nu- (Resia -nü-; northern Jaun Valley and eastern Rosen Valley -nu-). 
This *-nu- has all the markings of a vestige which has been receding under pressure 
from the more expansive variants. It appears that in the eastern and central part of 
Carinthia, the switch from *-nu- to *-nǫ- has progressed all through the central Jaun 
Valley (it is likely that the same process had occurred in Pannonia, proceeding here 
until completion). More to the west – in western Carinthia and Upper Carniola – *-nu- 
has instead given way to *-ny-, but here certain lexicalized exceptions have retained 
*-nu-, most prominently the (effectively synchronically irregular) verb minuti ‘pass’.
To our knowledge, the above material is the totality of the positive evidence for 
CSl *-nu-. OCS spellings with -nu- for usual -nǫ- do not resist closer scrutiny (cf. 
Szeptyński, Majer 2024: 134-5), while the vast areas in which *ǫ falls together with 
*u cannot be considered probative, as *-nǫ- has to be the default assumption here in 
275Marek Majer, Rafał Szeptyński: Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
any case (ESl, Cz, Slk, B/C/M/S etc. -nu- < *-nǫ- a priori more likely than < *-nu-;2 
see Szeptyński, Majer 2024: 145-6). Nevertheless, the Slovene evidence for CSl *-nu- 
needs to be taken completely seriously, given that it occurs dispersed across outlying, 
conservative areas. The identification of this CSl *-nu- may be of paramount importance 
for the diachronic explanation of the suffix, because – as is well-known, and as will be 
reviewed in more detail in §2.1 below – it is in fact far easier to explain diachronically 
than either *-nǫ- or *-ny- (note that *-nu- is expected both on the grounds of internal 
reconstruction, as a synchronically regular alternant of *-nov-, and on the grounds of 
external comparanda, generally subsumed under the PIE archetype *-new-). However, 
its identification in attested material does not by any means solve the puzzle: the task 
remains to determine why this expected CSl *-nu- coexists side by side with the variants 
*-nǫ- and *-ny-, why they (especially the former) have almost completely replaced it, 
and how they arose in the first place.
In the following sections, we first review the approaches to this variation that 
predominate in the current literature (with a cursory glance at earlier, now generally 
superseded views); this is the subject of §2. Subsequently, in §3, we suggest a new way 
of accounting for the variation *-nǫ- ~ *-ny- ~ *-nu-, based on the premise that both the 
starting point and the processes involved were less uniform than is standardly assumed.
2 Present state of research
2.1 Overall variation and general diachronic background
Traditionally, it was thought that the nasal suffix in Common Slavic only had the 
form *-nǫ-; consequently, all divergent forms were explained exclusively as due to 
local, dialect-specific innovations, be it of a phonological nature (vowel reduction – 
especially for forms in -ny- in the West Slavic data, cf. Szeptyński, Majer 2024: 132) or 
of a morphological nature (analogical influence of other verbal paradigms, specifically 
that in *-iti), or a combination of both. Surveys of traditional research taking *-nǫ- as a 
unitary point of departure are provided by Andersen (1999: 50) and Pronk (2022: 108-9). 
Some modern accounts of this type – see most recently Šivic-Dular (2021: 17-9) – are 
more sophisticated and ingenious; but irrespective of how well they work for specific 
varieties, they hardly account for the recurrence of forms pointing to *-ny- in so many 
systems at the western periphery of Slavic. 
The idea that *-ny- has to be assumed as an input form for Common Slavic was 
expressed, at least implicitly, already by Trubačev (Трубачев 1968: 374), Szymański 
(1970: 302-3), and Schuster-Šewc (1977: 437-8, 447-8), but it was only developed as a 
full-fledged theory by Andersen (1999). Much research has operated with this finding 
since then, including many of the recent studies on the modern extent of *-ny- in South 
Slavic dialects (referenced in Szeptyński, Majer 2024: 132-3). 
2 On Polabian with its potential -nĕ- < *-nu- see Szeptyński, Majer (2024: 145, fn. 35).
Slavistična revija, letnik 72/2024, št. 3, julij–september276
In spite of this progress, the diachronic aspect remains without a clear solution. 
Neither *-nǫ- nor *-ny- are easily accounted for based on extra-Sl comparanda, much 
less their apparent coexistence. We outline the present state of research below.
The origins of Slavic Class II verbs have been studied extensively and there is no dearth 
of controversy regarding the topic. Still, we may speak of a basic consensus (codified 
by Stang 1942: 56-60, but see also van Wijk 1931: 243-6; Tedesco 1948: 347; Puhvel 
1960: 36; Arumaa 1985: 224; Gorbachov 2007: 30-62; Šekli 2011: 131-3; Ackermann 
2014: 56) that these Slavic verbs ultimately – though via far-reaching morphological 
evolution – developed from PIE present stems in *-new- (PIE *meh₂-new- ‘beckon’ 
>> CSl *mane-, PIE *dʰr̥s-new- ‘dare’ >> CSl *dьrzne-, PIE *h₃r̥-new- ‘push, set in 
motion’ >> CSl *rine-, PIE *stigʰ-new- ‘step’ >> CSl *stigne- ‘reach’) as well as from 
PIE present stems in *-ne-h-, i.e., in PIE terms, stems in which the infix *-ne- preceded 
a root-final *-h3 (PIE *sḱi-ne-h- ‘shine’ >> CSl *sine-, PIE *ǵʰi-ne-h₁- ‘yawn’ >> CSl 
*zine-, *lu-ne-h- ‘cut’ >> CSl *lyne-).4 The development of the Slavic present paradigm 
(CSl *-ne- < post-PIE *-nw-e-, *-nh-e-? See Matasović 2008: 258-9; Gorbachov 2007; 
Shevelov 1964: 196-9; Villanueva Svensson 2025) is a separate matter we shall not 
delve into, as the focus here is on the Slavic suffix of the aorist/infinitive.
It is generally recognized that the use of *-nV- as a CSl aorist/infinitive suffix must 
be secondary, given that the PIE morphemes *-new- and (infixed) *-ne- specifically 
marked present stems. Thus, it has been plausibly assumed that the CSl aorist/infinitive 
*-nV- has its origins in the PIE imperfect. The latter was a preterital formation which 
was built from the present stem, but which, in the course of the restructuring of the 
verbal system, was absorbed into the Sl aorist (Stang 1942: 57; Sadnik 1959: 129-30; 
Vaillant 1966: 230; Gorbachov 2007: 52; Pronk 2013: 127; Andersen 2013; somewhat 
differently Villanueva Svensson 2025).5 Thus, descriptively speaking, a PIE imper-
fect such as *(e)-dʰr̥s-new-t ‘dared’ became the CSl aorist 2/3sg *dьrznǫ ~ *dьrzny 
(Gorbachov 2007: 52); likewise, the PIE imperfect *(e)-sḱi-ne-h-t ‘shone’ surfaces as 
the CSl aorist *sinǫ ~ *siny. Alternatively, the formation in question may have acquired 
the suffix *-s- (> *-x-, *-š-) from the s-aorist at an early stage, which would account 
for the entire paradigm at once (cf. e.g. 1sg *-nǫxъ ~ *-nyxъ); see also below.
3 In Indo-European terms, the sequence with *h₂ (i.e. *-ne-h₂-) is by far the most widespread.
4 This presentation is simplified, tacitly integrating standardly assumed minor analogical changes. To 
name a few of them here: CSl *i, taken from other verbal forms, replaces expected *ь < PIE *i; the stem 
*stigʰ-new- (also in Vedic prá stiṅnoti ‘rises’) is a post-PIE innovation replacing original infixed *sti-ne-gʰ- 
(LIV₂: 594; Gorbachov 2007: 57); CSl *lyne- has a more widespread variant *lune-, with *lu- < PIE *low- 
imported from other related formations. As regards more general matters, some scholars deny, or at any rate 
diminish, the role of *-ne-h- in the rise of Slavic Class II; thus e.g. Aitzetmüller (1991: 208: “unnötig”, though 
without specific arguments), and in a way also Gorbachov (2007) or Villanueva Svensson (2025), where 
quite different scenarios of the IE prehistory of the relevant classes (and their subtypes) are hypothesized. 
Under these interpretations, the *-ne-h- prehistory of many or all of the above-mentioned examples could 
be doubted. This important discussion cannot be further pursued here.
5 More broadly on PIE imperfects yielding Sl aorists cf. Meillet (1934: 252). Dissenting views exist 
(e.g. Aitzetmüller 1991: 209).
277Marek Majer, Rafał Szeptyński: Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
It has generally been assumed that the shape of this suffix must rely on some kind 
of transformation of the reflex of PIE *-new-.6 That the latter is somehow reflected 
here is indeed evident from the fact that, in prevocalic position, the CSl suffix occurs as 
*-nov- – as seen in the past passive participle *-nov-en-ъ (and the verbal noun derived 
from it, cf. OCS drъznovenie ‘courage’) – as well as in the (infrequent7) imperfective 
derivatives in *-ati built to Class II verbs, such as *minovati ‘pass.ipfv’ from *minǫti. 
These forms can be obtained regularly from the post-PIE inputs *-new-en-o- and 
*-new-ā-.8 Therefore, the anteconsonantal and word-final reflex, in accordance with the 
well-known monophthongization rules, should be *-nu-.9 This, as we now know (see §1 
above and Szeptyński, Majer 2024), can be considered attested as a peripheral remnant.
But for the most part we find *-nǫ-, and in addition – as has been clear for some 
time – *-ny-. The question how to explain these variants has been the object of quite 
a number of works. 
2.2 Origin of *-nǫ-
The dominant approach to *-nǫ- since Èndzelins (Эндзелинъ 1912: 370-2; on 
followers, notably including Stang 1942: 56-57; cf. Andersen 1999: 49) has been to 
assume that the expected CSl *-nu- underwent secondary perseverative nasalization 
of *u following a nasal consonant. As already mentioned in Szeptyński, Majer (2024: 
133-4), Slavic attests doublets of the type *gnusъ ~ *gnǫsъ ‘filth’, *vъnukъ ~ *vъnǫkъ 
‘grandson’, or *nuditi ~ *nǫditi ‘coerce’, and in many of them, including the above 
pairs,10 it can be demonstrated on etymological grounds that the form with *nu is origi-
nal and the one with *nǫ secondary. It should also be noted, however, that this process 
is never observed as a neogrammarian sound law, but rather as a sporadic tendency 
yielding inconsistent results, sometimes even within a single written tradition (OCS 
gnušati sę ~ gnǫšati sę ‛be disgusted’, OPol wnuk ~ wnęk ‘grandson’).
6 The role of PIE verbs in *-ne-h- has hardly been mentioned in this particular context – which, as we 
shall argue later below (§3), may have been an important oversight.
7 The imperfectivization *-nǫti → *-novati is rare and often competes with other types, some of which 
may in fact be older (e.g. *mijati ‘pass’, which omits the nasal suffix). The type in *-novati is nevertheless 
clearly residual and may hardly be a recent creation (Arumaa 1985: 227; Tedesco 1948: 348).
8 Given that the conditions for the change CSl *e > *o (or rather PSl *e > a) before *v (PSl *w) are not 
fully understood, the development to CSl *-nov- may be considered unproblematic. For what it is worth, 
most examples with CSl *-evV- sequences can in fact be accounted for by assuming late restructuring, cf. 
e.g. *kleveta ‘slander’ (< *klebeta?, see Варбот 2010), *nevěsta ‘bride’ (due to the restitution of the negative 
prefix ne-, cf. ЭССЯ XXV: 71-3), and prs.1du *-e-vě (due to the secondary fronting in *-je-vě and/or the 
*-e- in most of the other forms in the paradigm, cf. 2sg *-e-ši, 3sg *-e-tь, etc.).
9 Or possibly *-nju-, see preceding footnote. However, while *-CewC- > *-CjawC- may have been a 
Balto-Slavic innovation, its reflexes have probably been ousted in the case of alternating suffixes: cf. CSl 
-u < *-ew-(C) in desinences following non-palatalized root-final consonants in the u-stems, e.g. CSl *syn-u 
‘son’, not †synj-u.
10 Beside standard etymological references, see now also Olsen (2020) on *vъnukъ. The process of 
this perseverative nasalization is well-known; cf. Shevelov (1964: 319-20) or Sławski (1947; some of the 
individual etymologies here have to be treated with caution). On variation in OCS texts see Haralampiev 
(Харалампиев 2001: 61).
Slavistična revija, letnik 72/2024, št. 3, julij–september278
In and of itself, the development of CSl *-nu- (< PIE *-new-) to *-nǫ- is entirely 
credible. What has traditionally been considered problematic in invoking this process 
for the history of the Class II suffix is the apparent lack of preserved reflexes of the 
original *-nu-, much unlike in root morphemes with sporadic perseverative nasalization. 
This paradox has been widely noted,11 but no compelling explanation has ever been 
offered.12 In the light of our findings presented in Szeptyński, Majer (2024) – with CSl 
*-nu- now directly documented ‒ this has to some extent been alleviated, but the picture 
is nevertheless still quite different from what results from the perseverative *Nu > *Nǫ 
process operating on its own.
When the necessity of the reconstruction of the variant CSl *-ny- became evident, 
it was soon recognized (Andersen 1999: 53-4) that CSl *-nǫ- may also represent a 
secondarily nasalized version of this *-ny-. Andersen himself dates the nasalization to 
an earlier period, predating the development PSl *ū > CSl *y; thus, the bifurcation is 
schematized as PSl *-nū- (> CSl *-ny-) vs PSl *-nūn- (> CSl *-nǫ-). This would rep-
resent an earlier process than the *gnusъ ~ *gnǫsъ phenomenon and would therefore 
explain its completely different patterning. However, the reason for the nasalization 
is not obvious to Andersen; he considers a phonetic, perseverative explanation (there-
fore typologically similar to the *gnusъ ~ *gnǫsъ phenomenon, though perseverative 
nasalization of CSl *y lacks direct parallels13) as well as a morphological one (stacking 
of the present stem marker PSl/CSl *-n- onto the aorist/infinitive stem PSl *-nū- > CSl 
*-ny-, yielding PSl *-nū-n- > CSl *-nǫ-). The latter explanation, in particular, is neatly 
specific and well-defined. Still, such contamination of the present and aorist/infinitive 
stem would be entirely unparalleled in early Slavic. Competing (older) theories on the 
origin of *-nǫ- are even more problematic.14
2.3 Origin of *-ny-
Far less has been written on the potential origins of the newly discovered *-ny-. 
There are only two explanations on record. Andersen (1999: 53) considers PSl *-nū- 
> CSl *-ny- as an analogically lengthened apophonic zero-grade of PIE *-new- (the 
expected zero-grade, PIE *-nu-, would have yielded PSl †-nu- > CSl †-nъ-). He points 
to the fact that an identical pre-form *-nū- is also known from Greek (1sg deík-nū-mi 
11 E.g. Arumaa (1985: 226). Shevelov, referenced in the previous footnote, singles this out too (1964: 320).
12 Wandl (2020: 39) operates with a distant assimilation preceding the monophthongization (PIE *-now- 
> PSl *-nan- > CSl *-nǫ-). Admissible in and of itself, the approach makes it difficult to account for the 
material in its entirety, as we shall see below.
13 There are also potential examples for the inverse process – e.g. CSl *lyko ‘bast’ vs. Lith lùnkas – though 
essentially all of them can also be interpreted differently.
14 Vaillant (1951: 161) speaks of a “fusion de deux types d’infinitif, en *-neu-tei et en *-in-tei”, which 
is in effect similar to the contamination solution; Machek (1938) and Fraenkel (1950: 271-2) believe the 
suffix *-nǫ- was abstracted from a pariticipial form built to the present stem of verbs in *-neh- (CSl *-nǫt- 
<< PIE *-nh-ont-); Arumaa (1985: 225) attempts a direct link to the Hittite infix -nin-, but this is known 
to be a local restructuring of PIE *-ne-; Jacobsson (1951: 7-20) believes the infinitive *-nǫti derives from 
a gerundive in *-nt-. An overview of older, even less acceptable solutions is provided by Machek (1938: 
85-6) and Jacobsson (1951: 3-7).
279Marek Majer, Rafał Szeptyński: Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
~ 1pl deík-nŭ-men ‘show’). In the history of Greek, however, the analogical rise of 
such an alternation is far easier to motivate, given productive ablaut schemes involving 
other vowels in parallel verbal classes, such as -nā- ~ -nă-.15 An alternative scenario is 
proposed by Pronk, who reconstructs the pre-form of *-nyti as “*-nuHtei, with analog-
ical length and acute intonation from infinitives in -ati < *-aH-tei and -ěti < *-eH-tei” 
(2022: 111; similarly already 2013: 128).
The idea that *-ny- originated at least inter alia due to the influence of other 
productive verbal stems with acute *-V̰̄- is certainly plausible.16 However, Pronk also 
believes (2022: 112-3) that the original locus of *-ny- in the infinitive – as opposed 
to the aorist, l-participle, and the clearly secondary past passive participle – is still 
reflected in the modern distribution of -ni- in certain West South Slavic dialects. We 
fail to see evidence for this in the dialectal material reviewed by Pronk. The modern 
dialectal systems display a significant departure from the Common Slavic situation 
in all respects, with growing impact of Class IV (*-i-) and mutual influence among 
the stems of former Class II itself. The prerequisites for abandoning the infinitive in 
*-nyti specified by Pronk (2022: 113) would not explain the rampant spread of *-nǫti 
in almost all of West and East Slavic. Besides, given the complete parallelism of 
*-a-ti, *-ě-ti, *-i-ti with l-participles in *-a-lъ, *-ě-lъ, *-i-lъ, an analogically created 
class in *-ny-ti could have hardly inherited a different l-participle than *-ny-lъ in 
the first place.
2.4 Interim conclusion
We conclude that the question of the origin and distribution of *-nǫ- and *-ny- re-
mains open. A fresh analysis of the issue appears welcome, especially in the light of 
the identification of the reflexes of the third variant *-nu- in Slovene. In what follows, 
we offer a novel approach to the issue in its entirety.
3 A new theory of the *-nǫ- ~ *-ny- ~ *-nu- split
3.1 Exposition
On the one hand, the updated dataset – now with no fewer than three variants (*-nu-, 
*-ny-, and *-nǫ-) – is even more messy. On the other hand, the multifacetedness may 
mean that we should aim for a different kind of explanation than endeavored so far.
15 Such patterns of neo-apophonic length in this suffix are also attested in Indo-Iranian, where similar 
– though not as favorable – prerequisites for innovation existed as in Greek. Since PIE did not know the 
apophonic alternation *u : *ū, common inheritance is out of the question. A possible connection between 
Slavic and Indo-Iranian is, however, discussed by Dybo (Дыбо 1981: 241-2), although solely on accento-
logical grounds, i.e. without invoking the CSl variant *-ny- itself.
16 But note that *-ā̰- and *-ḛ̄- were accentually dominant (in the sense as used by Dybo; cf. Дыбо 1981: 
10), unlike the nasal suffix; *-ḭ̄- provides the closest parallel.
Slavistična revija, letnik 72/2024, št. 3, julij–september280
In our view, the seemingly erratic distribution makes the impression of phonologi-
cally conditioned allomorphy, recently levelled out in different directions. In particular, 
the fact that the variation is consistently found at the right edge of the morpheme, as 
well as that the suffix directly preceded inflectional endings, makes one think of the 
Slavic laws of final syllables.
Note that the clash of the suffixes *-new- and *-neh- with the original inflectional 
endings inherited from PIE created complexes such as *-new-(s)(C)# and *-neh(s)(C)#,17 
providing the conditions for diverse Slavic Auslautgesetze. The operation of the latter, 
if uninhibited by analogical developments, would have led to a paradigm in which the 
surface shape of the stem-forming morpheme would have differed in the individual 
slots – assuming, inter alia, the forms *-nu-, *-ny-, and *-nǫ- – while the primordial 
inflectional endings would have been absorbed into it.18 The potential developments 
can be envisaged in many different ways depending on the assumed point of departure 
and the specifics concerning the individual laws of final syllables. We limit ourselves to 
laying out one representative possibility. The projected phonological developments from 
the post-PIE situation all the way up to Common Slavic are presented in Table 1 below.
Table 1: Phonologically expected Common Slavic reflexes in the aorist of verbs in 
*-new- and *-neh₂/₃-.
verbs in *-new- verbs in *-neh₂/₃-
post-PIE expected CSl post-PIE expected CSl
1sg *-new-m > *-nǫ *-neh₂/₃-m > *-nǫ
2sg *-new-s > *-nu *-neh₂/₃-s > *-ny
3sg *-new-t(*-new-s-t)? >
*-nu 
(*-nu)
*-neh₂/₃-t
(*-neh₂/₃-s-t)? >
*-na
(*-ny)
pre-CV *-new- > *-nu- *-neh₂/₃- > *-na-
We hypothesize that the above pattern – or at any rate something resembling it (as 
already mentioned, exact results will depend on the assumed input and phonological 
framework) – was the original spawning ground of the *-nu- ~ *-ny- ~ *-nǫ- variation. 
Analogical levelings of the individual allomorphs, no doubt aided by the traditionally 
assumed factors (perseverative nasalization supporting *-nǫ-, preference for *-ny- 
due to the structural similarity to other verbal suffixes in -V̄-), led to the blurry image 
reconstructible for Common Slavic.
17 With *(s) due to secondary sigmatization, which is uncertain.
18 For some Slavic parallels for the absorption of Auslautgesetze-driven allomorphs (containing ‘sub-
merged’ inflectional endings) into the stem, cf. derivatives of the type *kam-y-kъ, *kam-ę-kъ ‘stone’ (Boryś 
1981: 43-4; see also Polański 1976 and Мајер 2021: 264-71), where *-y(-) and *-ę(-) are the results of 
developments specific to final syllables leading to the phonological coalescence of the etymological stem 
and ending (regardless of which precise scenarios are accepted for explaining them; see Olander 2015: 83-6). 
281Marek Majer, Rafał Szeptyński: Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
3.2 Commentary
A few explanatory remarks are necessary. We limit the scope to verbal stems in 
*-new- and *-neh₂-/*-neh₃- (the latter two developing identically in Slavic); we leave 
out *-neh₁-, for which no role in this scenario is detectable, although some such verbs 
are probably inherited (CSl *zine- << *ǵʰi-ne-h₁- ‘yawn’, see above). We assume that 
the PIE apophony between the ‘strong stem’ (*-new-, *-neh₂/₃-) and the ‘weak stem’ 
(*-nu-, *-nh₂/₃-) was levelled in favor of the former in the aorist paradigm; note that 
no trace of quantitative paradigmatic suffixal ablaut is found Proto-Slavic in the indic-
ative of any other verbal class except for some portmanteau morphemes. We restrict 
our presentation to four environments: the 1sg, 2sg, the 3sg (where the PIE endings 
of the imperfect comprised a single consonant: *-m#, *-s#, and *-t# respectively, thus 
positioning the whole complex in the word-final syllable19) and a single preconsonantal 
slot covering the remaining endings (all of which were syllabic in PIE and thus would 
not provide a word-final context).20
As regards the phonological development involving word-final *-s# (PSl *-ās > 
CSl *-y), we consider it secured based on the independent evidence from different 
morphological subsystems, cf. the pronominal clitics 1pl.acc *ny, 2pl.acc *vy < PIE 
*nōs, *wōs as well as the gen.sg *-y < *-ās < PIE *-eh₂-es (the matter is surrounded 
by some dose of controversy, but the evidence appears overwhelming – see most 
recent discussion in Kim 2019: 5-7, Majer 2020: 83-4, and Szeptyński, Majer 2021: 
51; competing view and summary of earlier research in Olander 2015: 56-7, 129-33, 
230-1). As regards the two alternative options in the 3sg, there is no a priori way to 
determine whether the *-s- of the s-aorist had been analogically introduced here; if so, 
the development would have been such as indicated in parentheses.
Of course, the outline in Table 1 compresses several chronological layers. For 
example, the loss of final *-t may have been a Proto-Balto-Slavic change already; the 
development of PIE *-eh₂- and *-eh₃- proceeded via PBSl *-ā-, *-ō- to PSl *-ā- to 
CSl *-a-, and so on. However, the crucial Auslautgesetze involving *-s# and the final 
nasal clearly operated between the PSl and the CSl era. We consider it superfluous to 
reconstruct the forms at the intermediate stages, as the details here are immaterial to 
our problem as long as the final outcome holds. The one issue that bears mentioning is 
the autonomous development of the 2sg and the 3sg forms. Given that these are always 
19 Unless sigmatization is assumed.
20 The 3pl form (with PIE *-nt#) might have followed an interesting path of its own here too, but since 
projecting its evolution would require extreme speculation, we refrain from including it here systematical-
ly. Likewise difficult to trace is the development of the 1sg form *-new-m. We consider it likely that the 
non-syllabic nasal marker of the 1sg was analogically kept stable across all paradigms (i.e., not *-newm̥ 
> CSl †-novь or similar) and assume the rise of the CSl back nasal vowel following monophthongization 
(*-newm > *-nūn > *-nǫ), based on general tendencies, though no exact parallel can be provided. (We leave 
aside the potential development of PIE *-ewm# > *-ēm#, as it is not observed in the reflexes of this class in 
the ancient IE languages either. We also do not take into account the possibility that the 1sg form retained 
a specific ablaut grade, *-nu-m, roughly in parallel to the respective present form in *-ǫ < *-о̄-m, whose 
quantity was, incidentally, as different as could be.)
Slavistična revija, letnik 72/2024, št. 3, julij–september282
identical in aorist paradigms in attested Sl (and thus also in reconstructible Common 
Slavic), there is a tendency in the literature to treat their development jointly even at 
earlier stages (e.g. Aitzetmüller 1991: 179-80). This may well be justified for most of 
the other stem classes, where inputs such as PIE *-e-s ~ *-e-t or *-eye-s ~ *-eye-t indeed 
ultimately yielded the same outcomes: here, even if the endings remained distinct for 
a long time (until the PSl period), the difference was eventually obliterated. But this 
need not have been the case at least with the type in *-neh₂/₃-, where the identity of the 
2sg and the 3sg can be analogical. The phonological development would have yielded 
2sg *-neh₂/₃-s > CSl *-ny and 3sg *-neh₂/₃-t > CSl *-na (unless both 2sg *-neh₂/₃-s(-s) 
and 3sg *-neh₂/₃-s-t > CSl *-ny21), and there is no reason not to believe that such results 
indeed emerged before being levelled out (see further below on CSl *-na-). When one 
considers this in conjunction with the general blending of the classes in *-neh₂/₃- and 
*-new- into a single conjugational type, it is only to be expected that the resulting Slavic 
nasal class displays a less uniform structure than the remaining ones.22
It is not known when the unification of the classes in *-new- and *-neh- into a single 
Slavic conjugation occurred. The key treatments of the issue (such as Stang 1942: 60) 
do not mention a fixed chronology. However, just like in the case of the 2sg and the 
3sg endings discussed above, it is likely that the distinction persisted until not long 
before the CSl era, as is indeed generally surmised by those few scholars who have 
dealt with this question directly (Tedesco 1948: 348: “That nā-presents reached down 
into Slavic is very likely”). The stems in *-neh- were presumably thematized early on 
(see Gorbachov 2007 for context), while *-new- may have remained a distinct class 
until the phonological confluence of pivotal forms such as the reflexes of the 3pl.
prs *-nu-nti (*-new-nti?) and *-nh₂-onti (*-nh₂-enti, *-neh₂-nti?); cf. Tedesco (1948: 
21 As suggested above, it is not clear whether or not the sigmatization of the aorist had reached these 
forms; the choice would depend on one’s particular views on the prehistory of the category. In Table 1 and 
the accompanying discussion, we treat the non-sigmatized *-neh₂/₃-t and *-new-t as the default assumption, 
but we also refer to the possibility of reconstructing *-neh₂/₃-s-t and *-new-s-t (which actually yields results 
that are easier to interpret, see below).
22 It is also not difficult to see why the CSl verbal class in -a-, which ultimately goes back to a stem 
in *-eh₂-, had no chance to retain the *-y ~ *-a allomorphy based on 2sg *-eh₂-s vs 3sg *-eh₂-t. Here, the 
present stem *-a-je- transparently consisted of the aorist stem extended with *-je-, providing clear support 
for *-a- as the sole aorist stem. (Cf. also Zucha 1986: 135-6, one of the few scholars to discuss directly the 
potential influence of Auslautgesetze on verbal stem allomorphy in Slavic; he likewise observes that forms 
like 2sg *dy ‘gave’ and *děly ‘worked’, from original *-ā-s, must have been levelled to match the 3sg *-a < 
*-ā-t). In the nasal class, the situation was different: the present stem *-ne- was not connected in any parti-
cular way with *-nu-, *-ny-, or *-nǫ- in the aorist. As regards the development of the past passive participle, 
which likewise had a unitary CSl form *-novenъ (< *-new-eno-), the abandoning of the expected reflex 
*-nanъ < *-neh₂/₃-eno- was probably motivated by its ambiguous segmentation and potential conflict with 
related CSl a-stems (on which see below), cf. *-na-n- vs. *-n-a-n- (on laryngeal contractions in this type of 
participles see Kortlandt 1975: 8). The same obtains, a fortiori, for the imperfectivizing type *minovati. It 
must be borne in mind that neither formation is overly ancient. In PIE, past participles (or verbal adjectives 
that provided the basis for them) were not derived from present stems, so that the creation of *-nov-enъ has 
often been considered as late as Sl-internal (Arumaa 1985: 227; Tedesco 1948: 348; under a theory like the 
one put forth in Villanueva Svensson 2025, the formation could be projected back to Balto-Slavic times 
at most). The iterative morpheme *-a- < *-eh₂- (see Villanueva Svensson 2020 on its origins and context) 
could not have been stacked on top of a present stem either.
283Marek Majer, Rafał Szeptyński: Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
347-8). Since the issue rests on the development of the present stems, we shall not treat 
it here in detail. The key premise for us is that nothing speaks against the possibility 
of distinct paradigms of new-verbs and neh-verbs feeding the Slavic Auslautgesetze. 
Effectively, our scenario shares with Pronk’s (2022: 110) the view that “*-ny- orig-
inally coexisted with […] *-nǫ- within the same paradigm” (cf.1sg *-nǫ alongside 2sg 
*-ny – and possibly 3sg *-ny – in the *-neh₂/₃- class). At the same time, the mechanism 
is markedly different. We believe that the rapid onset of the allomorphy as presented 
in Table 1, caused by the operation of the Auslautgesetze, was soon met by levelings 
and unificatory tendencies. In our view, late CSl varieties had most likely already lost 
the etymological association of the given variants with the original loci of their rise. 
We would attribute distributions such as that currently found in West South Slavic to 
contact between dialects that had previously leveled different variants (in this case *-ny- 
and *-nǫ-/*-nu-), negotiating the use of a given allomorph in particular grammatical 
forms – a process which can still be observed there at work today. For the same reason, 
the fact that some subtypes of Class II verbs lack the nasal suffix in the aorist – more 
widely than infinitives – does not, in our opinion, preclude the possibility that the 
aorist paradigm was the key locus in which the variation arose (all the more because 
the historical spread of the nasal suffix probably began from the aorist; cf. above on 
the origin in the PIE imperfect).
To our knowledge, such a line of reasoning, involving the laws of final syllables, has 
scarcely been explored before – certainly not for the variants *-nu- (hitherto considered 
non-existent) and *-ny- (discovered recently and only subject to very few diachronic 
interpretations so far, all of them covered in §2.3 above). As regards *-nǫ-, the notion 
that the unexpected nasalization may result from the absorption of the 1sg and the 
3pl endings has already been proposed by Vaillant23 (1966: 230): “La 1re personne 
du singulier *-neun […] a pu de son côté donner *-nǫ, puis -nǫ-xŭ”, similarly 3pl 
*-nw-ont >> *-nu-nt > CSl *-nǫ(šę). This approach was criticized e.g. by Aitzetmüller 
(1991: 208), but note that it was considered in isolation and not as part of a broader 
Auslautgesetze-based scenario such as presented here. While we would not concur with 
all of Vaillant’s exact preforms, we believe that the process as outlined above indeed 
may have contributed to the establishment of *-nǫ-. Once in play, the variant could be 
interpreted as the perseverative nasalization of *-nu- (and perhaps *-ny-), as generally 
assumed.24 This latter factor is no doubt responsible for the considerable expansion 
of this variant at the cost of all others; but the inception within the finite inflectional 
paradigm seems likely.
23 Elsewhere, Vaillant also suggested other origins for *-nǫ- (fn. 14).
24  If aor.1sg *-nǫ >> *-nǫ-xъ, as possibly common to both paradigms, contributed to their contamination 
after the irregular change *-nu- > *-nǫ- (alongside with the more important generalization of the present 
type in *-ne-), the process should have been impeded in the western peripheral zones of South Slavic, which 
resisted the perseverative nasalization (cf. §2.2). It is, in fact, indeed thinkable that the distinction of the 
separate classes survived longer in this area – cf. the residual Sln dial. minuti against the generalized -ni-. 
The variation observed in West Slavic, however – if real (see fn. 26 and Szeptyński, Majer 2024: 132) ‒ is 
not quite in line with such an interpretation, as no local ‘immunity’ to the nasalization of the suffix can be 
observed here.
Slavistična revija, letnik 72/2024, št. 3, julij–september284
Another advantage of the Auslautgesetze-based approach is that it provides a 
direct and clear basis for the variant *-ny-. At the surface level, the assumption that 
the entire paradigm of the new nasal aorist aquired the suffix *-s- early would make 
our explanation almost trivial (cf. the pivotal role of 3sg forms as discussed below); 
however, as we have mentioned throughout the paper, this solution is not necessarily 
justified. For that reason, in what follows we pursue the alternative scenario, in which 
only the 2sg form *-neh₂/₃-s yields the sequence *-ny phonologically in final position. 
Importantly, this *-ny could be levelled into the 3sg based on the identity of 2/3sg aorist 
in all other verbal classes and, from here, potentially generalized as the aorist/infinitive 
stem marker of the unified nasal class. Again, it is likely that this process was aided 
by an additional dynamic. Even if the absolute, foundational dominance of 3rd-person 
forms in analogical changes in paradigms cannot be substantiated (cf. Hill 2016), the 
hierarchical superiority of the 3sg for the indicative can hardly be doubted, and the 
chances of the entire paradigm being remodeled after the prs.ind.2sg form, in particular, 
are low (cf. Grossman 2016: 12-5). Therefore, we believe that the rise of *-ny- is best 
explained as due to the combined pressure of the 2sg form *-neh₂/₃-s, of other aorist/
infinitive stems in long, acute monophthongs25 (*-a-, *-ě-, and most probably *-i- – a 
factor that we have already mentioned before), and of the inherited alternation CSl *u 
~ *y elsewhere in the system (cf. derivationally related items such as *duxъ ‘spirit’ ~ 
*dyxati ‘breathe’ or *kujǫ ‘forge’ ~ *kyjь ‘club, hammer’ as well as doublets of the 
type *pluti ~ *plyti ‘swim, sail’).26
Finally, one more aspect of the situation as schematized in Table 1 must be discussed. 
As may be seen, two non-negligible slots in the original neh₂/₃-verbs are expected to 
yield CSl *-na-; this may include the crucial 3sg form (unless *-neh₂/₃-s-t > *-ny) and 
certainly includes the preconsonantal environment, which actually represents several 
forms. Although analogical processes are famously unpredictable – so that the complete 
obliteration of this variant is unquestionably thinkable – one might, in principle, expect 
the direct27 survival of attested CSl *-na- at least in some remains. Now, interestingly, 
forms consistent with CSl *-na- are found in certain areas in which a is not the general 
25 Note that the *u of *-nu(-) was, before the CSl era, a diphthong. See below on the possible reason for 
the dispreference for *-na(-),  whose *a was undoubtedly a long, acute monophthong and competed with 
*y of *-ny in the first place.
26 If the West Slavic evidence for *-ny- were to be disregarded (see Szeptyński, Majer 2024: 132), *-ny 
could be treated as a strictly West South Slavic entity and brought into association with the nom.sg.m of the 
present participle, which, as is well-known, displays SSl *-y (vis-à-vis NSl *-a). But this SSl *-y competes 
with an early analogical variant *-ǫ, taken from the oblique cases. The variation of *-ny and *-nǫ in the nom.
sg.m of these participles built to Class II verbs could have been analogically extended, ultimately leading 
to the expansion of *-ny-. The mechanism seems rather convolute, however; besides, Class II verbs were 
productive predominantly in semelfactive/punctual, perfective verbs, where the prominence of the present 
participle was low. As regards the possible parallel development of the nasalized variant of the *-nV- suffix 
with the nom.sg.m of the present participle (cf. aor.2/3sg *-nVN-C# > SSl *-ny ~ NSl *-na?), this is con-
tradicted by the retention of non-nasalized *-nu- in Slovene dialects as well as by the traces of *-ny- rather 
than *-na- in North Slavic.
27 As for indirect evidence, we believe that the phonologically expected acute on *-neh₂/₃- > *-nā- (CSl 
*-na-̋, when stressed) is the straightforward source of the consistent acuteness of the nasal suffix, irrespective 
of its segmental structure: CSl (when stressed) *-nǫ̋-, *-ny̋-, *-nű-. Cf. Дыбо (1981: 240-1).
285Marek Majer, Rafał Szeptyński: Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
reflex of *ǫ28 (or *y, *u). Such reflexes with surface -na- occur in certain Slovene and 
Croatian dialects (a general overview is provided by Celinić, Menac-Mihalić 2017; see 
also Pronk 2022: 105-8). Still, the hypothesis that these dialects indeed conserve a CSl 
variant *-na- is not easy to substantiate. In most instances, the -na- occurring here can 
be accounted for via vowel reduction in unstressed syllables, ultimately reflecting either 
-ni- < *-ny- (cf. Jakop 2003: 7-8) or *-nǫ- (Lundberg 2013: 47).29 For at least some 
varieties, however, previous research admits that forms in -na- have to be analogical 
after verbs in *-ati (e.g. Pronk 2022: 107 on the type krénati ‘move’ in Ozalj). Given 
that non-trivial transformations of this kind are considered indispensable in any case, 
one might entertain the theoretical possibility of inherited *-na- instead. Still, given 
the extremely slim basis and the presence of easily explainable -na- in neighboring 
varieties, the idea is best abandoned. Besides, one has to reckon with the possibility 
that the CSl allomorph *-na- was ousted due to the risk of ambiguous segmentation, 
possibly disturbing the aspectual oppositions between perfective *-na- and imperfective 
*-a- (this may also be the reason why the similarity to the acute suffixes of other verbal 
classes promoted *-ny- rather than *-na-, cf. above).30
4. Conclusions
The analysis presented above was based on two premises: firstly, on the fact that the 
existing explanations of both *-nǫ- and *-ny- are in our view unconvincing or at least 
incomplete, and secondly on our identification of the CSl variant *-nu- in the Slovene 
evidence, as documented in detail in our earlier study (Szeptyński, Majer 2024). We 
proposed a speculative but principled and comprehensive account of the variation 
*-nǫ- ~ *-ny- ~ *-nu- in its entirety. We believe the seminal split may have begun as 
allomorphy conditioned by the Slavic phonological laws of final syllables, coupled 
with the fact that the different classes of nasal presents (PIE *-new-, *-neh₂/₃-) were in 
the process of merging into a single conjugational type. The model entails interesting 
predictions; first and foremost, one would also expect the existence of a variant CSl 
*-na- (< *-neh₂/₃-C-). The Croatian and Slovene material that could theoretically be 
interpreted as evidence for CSl *-na- can (in an overwhelming majority) be accounted 
for in more straightforward ways, so that CSl *-na- has to remain a heuristic construct; 
concrete reasons for its elimination can be cited, too. Our model does generate *-nǫ- 
and *-ny-, although their distribution remains rather remarkable. This is especially true 
of the stunning ‘career’ of *-nǫ-, suggesting that its areal (rather than intrasystemic) 
expansion was clearly driven by other factors, first and foremost the traditionally as-
sumed sporadic nasalization.
28 We disregard modern East South Slavic, where *ǫ is generally reflected as a in suffixal syllables.
29 Note that some Croatian dialects regularly reflect *-nǫ- as -na- also when accented. The Istrian 
Čakavian material studied by Vranić (2017) contains interesting cases where the nasal suffix is not fully in 
keeping with the reflexes of *ǫ in lexical morphemes, but it does not reveal likely candidates for inherited 
*-na- (or *-nu-) either.
30 Cf. East Slavic imperfectives of the secondary type ipfv *min-a-ti ‘pass’ (ЭССЯ XIX: 51) ← pfv 
(?min-u-ti <<) mi-nu-ti with which potential perfectives of the type *mi-na-ti would have conflicted.
Slavistična revija, letnik 72/2024, št. 3, julij–september286
This is also in line with the status of *-nu- as the most archaic and recessive vari-
ant (at least judging by the material where it can be distinguished from *-nǫ-). As the 
phonological reflex of PIE *-new- in several key positions, it must have originally been 
much more prominent, but later receded under pressure of phonological and morpho-
logical innovations favoring *-ny- and especially *-nǫ-. These did not fully reach some 
of the outlier varieties of Slovene, also known for several other conservative traits.
list of abbreviations and symbols
B/C/M/S – Bosnian/Croatian/Montenegrin/Serbian; BSl – Balto-Slavic; CSl – 
Common Slavic; Cz – Czech; dial – dialectal; ESl – East Slavic; IE – Indo-European; 
Lith – Lithuanian; NSl – North Slavic; O – Old; OCS – Old Church Slavic; Plb – Polabian; 
Pol – Polish; P – Proto-; Sl – Slavic; Slk – Slovak; Sln – Slovene; SSl – South Slavic; 
USorb – Upper Sorbian. Grammatical glosses in small caps follow the Leipzig Glossing 
Rules (online), extended by: aor – aorist. The symbols < and > denote phonological 
change; the symbols << and >> denote additional morphological modification.
references
Katsiaryna ackermann, 2014: Die Vorgeschichte des slavischen Aoristsystems. Leiden: 
Brill. Also online.
Rudolf aitzetmüller, 1991: Altbulgarische Grammatik als Einführung in die slavische 
Sprachwissenschaft. Freiburg i. Br.: Weiher.
Henning andersen, 1999: The Western South Slavic contrast Sn. sah-ni-ti // SC sah-
nu-ti. Slovenski jezik – Slovene Linguistic Studies 2. 47-62. Also online.
Henning andersen, 2013: On the origin of the Slavic aspects: Aorist and imperfect. 
Journal of Slavic Linguistics 21. 17-43. Also online.
Peeter arumaa, 1985: Urslavische Grammatik. Vol. III. Formenlehre. Heidelberg: Winter.
Władysław Boryś, 1981: Geneza formacji typu sch. kamek, kamečak, bułg. kamek. 
Zbornik za filologiju i lingvistiku 23/2. 37-45.
Anita Celinić, Mira MenaC-Mihalić, 2017: Poveznice i razdjelnice u srodnim jezičnim 
sustavima. Jezikoslovni zapiski 23/1. 93-105. Also online.
Ernst fraenkel, 1950: Zum baltischen und slavischen Verbum. Zeitschrift für Slavische 
Philologie 20. 236-320. Also online.
Yaroslav gorbachov, 2007: Indo-European Origins of the Nasal Inchoative Class 
in Germanic, Baltic and Slavic: Doctoral dissertation. Cambridge, MA: Harvard 
University. Online.
Eitan grossman, 2016: From rarum to rarissimum: An unexpected zero person marker. 
Linguistic Typology 20. 1-23. Also online.
Eugen hill, 2016: Synchronic vs. diachronic roots of Watkins’s Law. Paper from: 
49th Annual Conference of the Societas Linguistica Europaea, Naples, 31.08–3.09 
2016. Online.
287Marek Majer, Rafał Szeptyński: Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
Gunnar Jacobsson, 1951: Le suffixe verbal -nǫ- en slave. Göteborg: Elanders Boktryckeri 
(Göteborgs Kungliga Vetenskaps- och Vitterhets-Samhälles handlingar A, 6/4,4).
Tjaša Jakop, 2003: Tipologija narečnih glagolskih oblik na primeru govora Ložnice 
pri Žalcu. Slavistična revija 51/1. 1-25. Also online.
Ronald I. kim, 2019: North Slavic -ě vs. South Slavic -ę: A problem of forward recon-
struction. Journal of Slavic Linguistics 27/1. 1-26.
Frederik H. H. kortlandt, 1975: Slavic Accentuation. A Study in Relative Chronology. 
Lisse: Peter de Ridder.
LIV₂ = Helmut rix et al. (eds.), 2001: Lexikon der indogermanischen Verben. 2nd ed. 
Wiesbaden: Reichert.
Grant H. lundberg, 2013: Dialect Leveling in Haloze, Slovenia. Maribor: Zora. Also 
online.
Václav machek, 1938: Die Herkunft des nǫ-Stammes in der slavischen II. Verbalklasse. 
Zeitschrift für Slavische Philologie 15. 85-92. Also online.
Marek maJer, 2020: Slavic *ljuby and the heterogeneity of the inflectional class in 
*-y. Indogermanische Forschungen 125. 79-104. Also online.
Ranko Matasović, 2008: Poredbenopovijesna gramatika hrvatskoga jezika. Zagreb: 
Matica hrvatska.
Antoine meillet, 1934: Le slave commun. 2nd ed. Paris: Champion.
Thomas olander, 2015: Proto-Slavic Inflectional Morphology. A Comparative Handbook. 
Leiden, Boston: Brill.
Birgit A. olsen, 2020: Insiders and outsiders. The etymology of Slavic *vъnukъ ‘grand-
son’. Vácāmsi miśrā krṇavāmahai. Proceedings of the International Conference of 
the Society for Indo-European Studies and IWoBA XII, Ljubljana 4–7 June 2019. 
Eds. Luka Repanšek, Harald Bichlmeier, Velizar Sadovski. Hamburg: baar. 527-38.
Kazimierz Polański, 1976: Uwagi o rozwoju prasłowiańskiej deklinacji atematycznej. 
Opuscula polono-slavica. Munera linguistica Stanislao Urbańczyk dedicata. Eds.
Jan Safarewicz et al. Wrocław: Zakład Narodowy im. Ossolińskich. 291-5.
Tijmen pronk, 2013: On the development of *in, *im, *un and *um in Slavic. Miklošičeva 
monografija. Ob dvestoletnici rojstva Franca Miklošiča. Ed. Marko Jesenšek. 
Ljutomer: Gimnazija Franca Miklošiča. 117-40.
Tijmen pronk, 2022: The verbal suffix *-nǫ-/-ny- in Western South Slavic dialects. 
Hrvatski dijalektološki zbornik 26. 103-17. Also online.
Jaan puhvel, 1960: Laryngeals and the Indo-European Verb. Berkeley: University of 
California Press.
Linda sadnik, 1959: Slavische Akzentuation. Wiesbaden: Harrassowitz.
Heinz schuster-Šewc, 1977: Zur Bedeutung des Sorbischen und Slowenischen für 
die slawische historisch-vergleichende Sprachforschung. Slovansko jezikoslovje. 
Nahtigalov zbornik. Ed. Franc Jakopin. Ljubljana: Filozofska fakulteta. 433-51.
George Y. shevelov, 1964: A Prehistory of Slavic. Heidelberg: Winter.
Franciszek sławski, 1947: Oboczność ǫ : u w językach słowiańskich. Slavia Occidentalis 
18. 246-90.
Christian S. stang, 1942: Das slavische und baltische Verbum. Oslo: Jacob Dybwad.
Rafał szePtyński, Marek maJer, 2021: With all due respect, on Slavic abstracts in 
-y: The history of Proto-Slavic cěty ‘respect’ and some comparative notes on its 
Slavistična revija, letnik 72/2024, št. 3, julij–september288
congeners (ljuby ‘love’, cěly ‘healing, greeting’). Journal of Slavic Linguistics 29. 
45-100. Also online.
Rafał szePtyński, Marek maJer, 2024: Common Slavic *-nǫ-, *-ny-, or *-nu-? A new 
look at the history of the Slavic nasal suffix. Slovene evidence for *-nu-. Slavistična 
revija 72/2. 131-52. Also online.
Tadeusz szyMański, 1970: Review of: Piotr Bąk, 1968: Gwara okolic Kramska w powie-
cie konińskim. Wrocław, Warszawa, Kraków: Zakład Narodowy im. Ossolińskich. 
Język Polski 50. 302-7.
Matej Šekli, 2011: Besedotvorje praslovanskega glagola v primerjalnojezikoslovni 
luči. Philological Studies 9/1. 129-47. Also online.
Alenka Šivic-Dular, 2021: Analogija in metodologija jezikoslovnega raziskovan-
ja. Slavistična prepletanja 3. Eds. Gjoko Nikolovski, Natalija Ulčnik. Maribor: 
Univerzitetna založba. 3-34. Also online.
Paul tedesco, 1948: Slavic ne-presents from older je-presents. Language 24/4. 346-
87. Also online.
André vaillant, 1951: Chronique: Généralités. Revue des études slaves 28. 154-62. 
Also online.
André vaillant, 1966: Grammaire comparée des langues slaves. Vol. III. Le verbe. 
Paris: IAC.
Miguel villanueva svensson, 2020: The Balto-Slavic ā-aorist. Transactions of the 
Philological Society 118. 376-400. Also online.
Miguel villanueva svensson, 2025: PIE nu-factitives in Balto-Slavic: Slavic nǫ-perfec-
tives and Lithuanian verbs in -inti. Indogermanische Forschungen 130. Forthcoming.
Silvana vranić, 2017: Refleksi *ǫ u istarskim čakavskim govorima zastupljenim u 
dijalektološkim atlasima. Hrvatski dijalektološki zbornik 21. 283-97. Also online.
Florian wandl, 2020: Vom Urindogermanischen zum Russischen: Eine Studie zum 
Lautwandel und seiner relativen Chronologie: Doctoral dissertation. Universität 
Zürich. Online.
Nicolaas van wiJk, 1931: Geschichte der altkirchenslavischen sprache. Berlin: De 
Gruyter.
Ivo zucha, 1986: Zu den Endungen -y, -ę, -ě im G. Sg. und N. Pl. der slawischen 
Deklination und zu den Nasalvokalen. Wiener Slavistisches Jahrbuch 32. 133-7. 
Also online.
***
Владимир А. Дыбо, 1981: Славянская акцентология. Москва: Наука.
[Vladimir A. dybo, 1981: Slavjanskaja akcentologija. Moskva: Nauka.]
Жанна Ж. ВАрбот, 2010: Нерегулярные изменения в славянских языках и славян-
ская этимология. Этимология 2006–2008. 37-51.
[Žanna Ž. varbot, 2010: Nereguljarnye izmenenija v slavjanskich jazykach i slavjan-
skaja ètimologija. Ètimologija 2006–2008. 37-51.]
Марек МАјер, 2021: Потеклото на глаголските прилози од типот -ник. Прилози. 
Македонска академија на науките и уметностите. Оделение за лингвистика 
и литературна наука 46/1. 251-76.
289Marek Majer, Rafał Szeptyński: Common Slavic *-nǫ-, *-ny-, or *-nu-? A New Look at the History 
[Marek maJer, 2021: Potekloto na glagolskite prilozi od tipot -nik. Prilozi. Makedonska 
akademija na naukite i umetnostite. Odelenie za lingvistika i literaturna nauka 
46/1. 251-76.]
Олег Н. трубАчеВ, 1968: О составе праславянского словаря (проблемы и резуль-
таты). Славянское языкознание. VI Международный съезд славистов (Прага, 
август, 1968 г.). Доклады советской делегации. Eds. Виктор В. Виноградов 
et al. Москва: Наука. 366-78.
[Oleg N. truBačev, 1968: O sostave praslavjanskogo slovarja (problemy i rezul’taty). 
Slavjanskoe jazykoznanie. VI Meždunarodnyj s”ezd slavistov (Praga, avgust, 
1968 g.). Doklady sovetskoj delegacii. Eds. Viktor V. Vinogradov et al. Moskva: 
Nauka. 366-78.]
Иван ХАрАлАМпиеВ, 2001: Историческа граматика на българския език. Велико 
Търново: Фабер.
[Ivan haralampiJe, 2001: Istoričeska gramatika na bălgarskija ezik. Veliko Tărnovo: 
Faber.]
Иванъ М. ЭнДзелинъ, 1912: О происхожденіи праславянскихъ инфинитивовъ 
на -nǫti. Русскій филологическій вѣстникъ 68/4. 370-2.
[Ivan M. ÈnDzelin, 1912: O proischoždenii praslavjanskich infinitivov na -nǫti. Russkij 
filologičeskij věstnik 68/4. 370-2.]
ЭССЯ = Олег Н. Трубачев et al. (eds.), 1974-2022: Этимологический словарь 
славянских языков. Праславянский лексический фонд. Vols. I-XLII. Москва: 
Наука.
[ÈSSJa = Oleg N. truBačev et al. (eds.), 1974-2022: Ètimologičeskij slovar’ slavjanskich 
jazykov. Praslavjanskij leksičeskij fond. Vols. I-XLII. Moskva: Nauka.]
povzetek
V ločeno objavljenem prispevku sva pokazala, da je imela splošnoslovanska aoristna/infi-
nitivna pripona glagolov II. vrste na nekaterih obrobnih območjih obliko *-nu-, kar dokazuje 
nekaj arhaičnih slovenskih narečij: prim. rezijansko -nü-, (severno)podjunsko -nu-, (vzhodno)
rožansko -nu- ter leksikalizirane ostanke širše znotraj slovenščine. 
V pričujoči študiji obravnavava posledice te ugotovitve za splošnoslovansko geolingvistiko 
kot tudi njene globlje diahrone (indoevropske) vidike. Vzpostavitev različice *-nu- pomeni, 
da je bila formalna variabilnost te spsl. pripone širša, kot se je doslej domnevalo: ne le *-nǫ- ~ 
*-ny-, temveč *-nǫ- ~ *-ny- ~ *-nu-. Poleg tega je prav zadnja od teh različic tista, ki jo dia-
hrono najlažje razložimo: pravzaprav je bil njen obstoj že prej domnevan na podlagi notranje 
rekonstrukcije (trpni deležnik na *-nov-en-) ter primerjave z drugimi indoevropskimi jeziki 
(pie. *-new-).
Predlagava nov pristop k celoviti zadevi. Izhajava iz dejstva, da je spsl. II. vrsta etimološko 
gledano konglomerat nekaj indoevropskih glagolskih tipov (predvsem *-new- in *-neh2-) in da 
je njihovo združevanje lahko vključevalo obdobji razpada in reintegracije. Natančneje, pred-
lagava, da se upošteva spsl. fonološki razvoj v končnih zlogih, kjer so v prvotnih preteritalnih 
paradigmah osnovama *-new- in *-neh2- neposredno sledile osebne končnice: 2sg *-neh2-s (ali 
mogoče 2sg *-neh2-s(-s) in 3sg *-neh2-s-t) > spsl. *-ny; (mogoče) 1sg *-new-m > spsl. *-nǫ; 
Slavistična revija, letnik 72/2024, št. 3, julij–september290
3sg ter znotraj besede *-new-(t) > spsl. *-nu(-). Prvotni rezultat je bila alomorfija, ki je – z 
napredovanjem oblikovanja enotnega glagolskega razreda – v posameznih narečjih lahko bila 
predmet analoških posplošitev v različnih smereh. Vendar pa priznava, da so pri oblikovanju 
končne slike igrali vlogo tudi dejavniki, na katere se tradicionalno sklicujejo pri razlagi variant 
*-nǫ- in *-ny- (progresivna nazalizacija, analogija po drugih glagolskih priponah itn.); le-ti so 
lahko prispevali k bolj intenzivnemu širjenju nekaterih različic.