Northian nouns

Northian nominals are the nouns, adjectives, and some adverbs of the Northian language.

History

The Northian noun classes were named by Praetorianus of Duuora (c. 950? – 1027) in his studies on Galic language. Praetorianus was descended from Æþurian settlers who populated the north of the modern Northern States in the course of the 8th and 9th centuries. From a family seeking to build relations with the more southerly cities of Northian heritage, he was sent to take holy orders in the city of Cleiden at the age of 16; he first spent four years acquiring the Galic language before joining the order of training-priests and then graduating to an under-priest in 978, the first firm date of his biography. He celebrated his promotion as plenary priest in 993 with a large banquet for over 100 guests.

Later in life, he published in West Nordic on his priestly career, in which a full third was spent describing the way the language was taught in Cleiden. He engaged in an early form of comparative linguistics and identified many cognate words in Elder Nordic and Galic Northian. For example he said sunuz was the same as huiiuš and furthermore all genitives ending in -auz in Nordic ended in -ō in Northian. He did not discover any evolutionary principles between the two languages, considering them the same language, and instead attributed the difference to speech speed, as according to him, -auz spoken quickly was the same as -ō. In his effort to describe Northian, he summarized the various noun paradigms under the heading of an exemplar, which is still followed in modern education and called elsewhere the Praetorianian caput.

The pre-praetorianian method of language instruction followed the order of words in which they were encountered in the Gales, which intermixed the different parts of speech.

Stem types

	-C	-t	-m	-n	-r	-s	-i	-u	-ī	-ū	-H	-r/n	-nt
OX-B	About 80 root nouns and suffixed nouns in -k, -l, -ns, -d, etc., fos = "foot", xō = "heart"	A few in -ōs, nepōs "child"; -it melit "honey"; -ut xaput "head"	About 20 feminine, in -ō or -am, xay-ō "winter"	Many, in -ō and -ə̄, often binds m-, s-, t-, i-, u-, H-, θxām-ō "human"	haz-r "arm" and personal, feminizing suffix -ez-r	A handful in -ā̊, awšt-ā̊ "dawn"; the perf. act. ptcpl. wayδuš "knowing"	Some in -ō and -iš, hak-ō "friend, ally", ow-iš = "ewe"; neuter var. in -ai, oxϑ-ay "finger"	A few dozen in -ōš or -uš, han-uš "jaw"	Some in -ī, wrk-ī "wolfess"	A handful in -ū, ϑen-ū "body"	fənδōy, "path"	Derived coll. of heteroclitics, in -r ~ n-, ō-grade in nom. > -ā̊	A few in -āṯ, dāṯ = "tooth"
OX-A	No	Productive suffix -tāt-, āβr-tās "immortality" and -tūt-, wəš-tūs "moistness"	No	Productive, in -ṓ and -ā̊	Productive through agentive -ter, duhiϑ-ṓ "daughter"	Productive, comparatives in -yā̊- maz-yā̊ "bigger", perf. act. ptcpl. in -wā̊-, βeβiž-wā̊ "trusting"	Verbal nouns, middle, in -doi, št-oi "an eating"	Certain denominal nouns, in -ṓ, piϑr-ṓ "uncle"	Non-ablauting feminine derivatives in -ī́	No	Certain nouns in -dōy, žβā-dōi "libation-giver"	No	Productive pres. aor. act. ptcpl. of thematic verbs, in -ās hadáʸy-ās "sitting"
PX	No	No	No	Productive, neuter var. of OX n-stems, exhibiting same bound suffixes	yāϑr, "husband's sister"	Productive, neuter var. of OX s-stems, in -ō; neβ-ō "cloud", xrat-ō "power"	Productive, animate in -iš, action nouns -tiš, zāt-iš "a step"; neuter in -i, mər-i "sea"	Productive, animate in -uš, wiš-t-úš "witness"; neuter in -ū, oii-ū "life"	Proliferate suffix deriving feminines in -ī; geniϑr-ī "genitrix"	Same as -ī but more common in adjectives, in -ū; hozr-ū "mother-in-law"	bā̊ "wife"	Neuter, in -r ~ n- suffix, yaH-arə "year"; more in -tr, -mr, -zr	No
PP	About 30 root and suffixed nouns in obstruent stems	nox-ṯ "night", haš-ṯ "bed", xom-d "hand"	tā̊ "house"	Extended with -n-s, maHā "moon"	māϑ-r "mother", βrāϑ-r "brother"	zraw-š "gore"	aŋhi "serpent"	Neuter nouns in -ū, gon-ū "knee", wəšt-ū "settlement"	Feminine pres. aor. act. ptcpl. with static accent, déδāṯīš "of the giveress"	No	A few, particularly neuter in -i, ošti "bone"	A few, in -r ~ n-, f-ō "shrine"	A few nouns; pres. aor. act. ptcpl. with static accent, déδāṯ "giving"

As appears from this schematic, most suffixes are associated with more than one accentual pattern. But even in Galic, the majority of suffixes have only one productive accentual pattern or separate productive patterns associated with masculine-feminine gender and neuter gender (the case of the n-stems and s-stems). Additionally, some suffixes are only productive through petrified compounds, which tended to be hysterokinetic and have invariant stems, such as the comparative in -yā̊- and perfect active participle in -wā̊-; otherwise, the s-stems in amphikinetic is non-productive.

Where there are multiple productive accentuation patterns, neuter nouns are almost always identified with the proterokinetic pattern, and masculine-feminine with the hysterokinetic or amphikinetic. The exception is for i-stems and u-stems, wherein proterokinetic accentuation is standard, and particularly productive through the compound with -t.

As is generally observed across all the Erani-Eracuran daughter languages, the suffixes ending in resonants and semivowels, i.e. n-, i-, and u-stems, tend to be productive over the suffixes in stops. There is some speculation why this is the case, one theory being the resonants were able to be vocalized as syllables and thus did not produce illegal clusters of consonants, given the tendency to delete vowels in unaccent positions. It is argued that the compound suffixes, which often mix one stop and one resonant are the results of transferrals from suffixes in stops to those in resonants; if so, these transferrals must have occurred quite early, as their older siblings in stops generally do not leave visible remnants in attested languages.

Unlike verbal formations, where suffixes often have specific meanings and exist as part of the standard morphology of verbs, the meanings of noun suffixes are barely if at all recoverable. Synchronically, there is no obvious meaning to most of the suffixes present in Galic, even the productive ones, above a rudimentary association with certain parts of speech from which a noun is derived. For example, the -tiš nouns most often are nomina actionis derived from verbs, -tuš having a similar function, -tāt and -tūt deriving states of being, but there is no comparable explanation for their common constituent -t-, which appears as an independent suffix in nouns like haš-t and nep-ot.

Root nouns ending in stops

The category of consonant-stems consist only of those that end in -p and -k, and all are root nouns, as apparently the parent language did not have any suffixes that ended in *-p or *-k, with one exception for the adjective meraxš "soft". There are two ablaut-cum-accent patterns attested in these nouns: one is static on the root syllable, and another shifts to the ending in oblique cases. The static accent is generally (but not universally) seen with the lengthening of the root syllable in direct cases. Phonetically, the obstruents p- and k- are rarely subject to alteration, except before the nom. sing. ending -s where they become fricativized to f- and x-, respectively.

The noun āfš < PEE *h₂ēp-s "river" is often the poster-boy of the Northian nouns due to its straightforward forms and clear ablaut alteration between ā ~ a. Here we also encounter the first diachronic variation: in the Didaskalic Material (noted in small type dm) the long grade is deployed only in the nominative singular, since the long grade was more typically associated with that desinence by way of the much commoner amphikinetic and hysterokinetic nouns. Theoretically, given this tendency, we should also expect the DM to yield nom. du. *apa with the short grade of the stem, but that cannot be confirmed either way as it is not attested there (also noted for interest). In the plural, there is attested a separate collective apō "the divine waters", which has singular number and no separate accusative form.

ap "force" is used for neuter nouns, displaying PEE e ~ Ø ablaut > Galic o ~ Ø. Forms with initial b- are due to vocalization of initial *p-.

	hā āfš, "river"			θaṯ op, "force"
	sing	du	pl	sing	du	pl
nom	āfš	āpi dm ?	āpaŋhā	op	opī	opi
voc	ap
acc	āba dm apa		abā̊
gen	afš	apuš	apõm	bō	bōš	bõm
loc	api	apū	apšū	opi		fštū
dat		afma	afmuš	bā	βmō	βmuš
ins	api		dm aōβiiōḫ	bōi		βžβiiōḫ

t-stems

Ablauting nouns

The stems ending in -t comprise both of root nouns and suffixed nouns. They are underlyingly the same as other obstruent stems but are distinguished in that final -s is preserved in the nominative case, at the expense of *-t-. t-stems not part of a compound suffix such as -tāt- and -tūt- are rare and are descended mainly from the amphikinetic ablaut type.

nēpōs "grandchild" has the stem nep-ot-, where the suffix undergoes ablaut to zero grade in the oblique cases as *nep-t- > nef-θ-.

	ha nēpōs, "grandchild"
	sing	du	pl
nom	nēpōṯ	nepotōi	neptaHā
voc
acc	nepotā		neptā
gen	āptō	neptō	neptõm
loc	nepeti		nefšu
dat	neptā	neptmō	nepšmuš
ins	nepti

Non-ablauting animates

The PEE suffix *-teh₂ts created nouns of states of being. Inherited examples are mainly of hysterokinetic origin, which had the zero grade of the root, but later creations may bind the full grade. The suffix is non-ablauting and takex oxytone endings. The following example has many cognates in EE languages and is from a common root *n̥-mr̥teh₂ts > Galic āmarətāṯ.

	hā āmarətāṯ, "immortality"
	sing	du	pl
nom	āmarətāṯ	āmarətāϑōi	āmarətāϑaHā
voc
acc	āmarətāϑa		āmarətāϑā̊
gen	āmarətāϑō	āmarətāϑō	āmarətāϑõm
loc	āmarətāϑi		āmarətāštū
dat	āmarətāϑā	āmarətāsmō	āmarətāsmuš
ins	āmarətāϑōi

Neuters

This group of neuter nouns combines suffix -t with a semivowel (-it or -ut) or resonant (-n̥t). Neither the stem nor suffix displays ablaut, and accent is persistent over the root.

	hā melit, "honey"
	sing	du	pl
nom	melit	melitī	meliti
voc
acc
gen	melitō	melitō	melitam
loc	melit		melištū
dat	melitā	melismō	melismuš
ins	meliti

	hā kaput, "head"
	sing	du	pl
nom	kaput	kaputī	kaputi
voc
acc
gen	kaputō	kaputō	kaputam
loc	kaput		kapušū
dat	kaputā	kapusmō	kapusmuš
ins	kaputōi

The term ezat has the zero-grade of the stem *-nt. This word is an apparent exception to Cloverdale's law, as the suffix always appears written short, regardless of the quantity of the preceding vowel.

	hā ezat, "food"
	sing	du	pl
nom	ezat	ezatī	kaputi
voc
acc
gen	ezatō	ezatō	ezatam
loc	ezat		ezašū
dat	ezatā	ezasmō	ezasmuš
ins	ezatōi

hašt and non-ablauting items

These are t-stem nouns with acrostatic inflection. The nominative was evidently asigmatic < *ses-t, from *ses- "to rest". The genitive haš must represent *has-t-s, the final consonant cluster being resolved in favour of *s.

	hā hašt, "bed"
	sing	du	pl
nom	hašt	hašti	haštā
voc
acc	hašta
gen	haš	haštuš	haštõm
loc	hašti	haštū	hašū
dat		haštmō	haštmuš
ins

Interestingly, the dual hašti is more frequent than singular hašt and has singular meaning. It is possible that the singular had the sense of a bed-end rather than the entire bed.

The very well-attested word nōxt "evening, night" is usually thought to be a root noun with persistent accent on the initial syllable at the Proto-Erani-Eracuran level, but some authorities consider the root to be *(d)negʷ-, extended by the suffix -t-. Whatever the case in the parent language, the -t- never takes full grade even in the nom. du. and pl. Thus its behaviour is identical to that of hašt. xašt "hunger" is another word which is acrostatically accented; the expected form is *xošt, but the attested vowel is /a/, which may signify a (very) early borrowing from another Erani-Eracuran language that is not attested.

The word for "ten" in compounds of multiple of ten, e.g. twenty, thirty, etc. displays the zero grade and is apparently extended by suffix *-t, -tkaṯ < *-dḱm̥-t.

• θrítkaṯ < *tri-dḱm̥-t, "thirty"
• kóswr̥tkaṯ < *kʷetwr̥-dḱm̥-t, "forty"
• patótkaṯ < *pn̥kʷe-dḱm̥-t, "fifty"
• xšwáxštkaṯ < *ksweks-dḱm̥-t, "sixty"
• háptatkaṯ < *septm̥-dḱm̥-t, "seventy"

nt-stems

The nt-stems are very closely associated with the formation of the present active participle in -nt-. But while participles are adjectives that have distinct gendered forms that agree with their principals, nt-stem nouns have the form of the masculine participle and a lexical gender (that is, the noun can have feminine gender but will always have the same inflectional endings as the masculine participle). In the following example, the oblique stem is from *h₁dn̥t-; as the vocalic *n̥ indeed follows two consonants, it is short by Cloverdale's law. In the acc pl, the -ā ending is for *-ah, which is regular following two consonants, disclosing the strong stem ending in *ent- that has been replaced by the oblique in at-.

	hā dənt, "tooth"
	sing	du	pl
nom	dənt lg dən	dəntōi	dəntaHā
voc
acc	dənta		datā
gen	datō	datō	datām
loc	dənti lg dati		daśtū
dat	datā	daśmō	daśmuš
ins	datōi

m-stems

This class is known from only a few but important nouns, e.g. ϑázā "earth, ground" and xaō "winter". All m-stem nouns in Northian are feminine in gender, though with only a few examples, this may not be a rule in the proto-language. The later the time period, the more the m-stem group tended to become merged with the n-stems, usually thought to be due to assimilation in some case-forms and the common irregular accusative that occurs in both groups and then generates similar results.

As Proto-Erani-Eracuran had a strict rule against geminates, the accusative morpheme *-m causes a geminate to occur in m-stems in both the singular and the plural. In the singular, the long vowel accusative results, while in the plural the sequence of *CVC-em-ms generated -ōi. This also happens to be evidence that the acc pl was originally a case with the strong stem, even though all attested m-stems terms have weak stem: only a full-grade suffix in this position could generate the required *-ē > -ōi.

ziiō

Outside of nouns, it is also known in numerals for 1, 7, and 10—hā̊ hafθa tezā—and the solitary adjective mərā "particulate, ground up". Some authorities believe many m-stems may have, during the time of the proto-language, been either remade to thematic neuters by the insertion of the theme vowel *e ~ o or had an additional *-n appended after *m-, giving rise to the multitude of n-stems of the -mn- type, wherein the -m- is of otherwise unexplained origin. But both hypotheses suffer from the mismatch of grammatical gender—virtually all m-stems in Erani-Eracuran languages are attested as feminine in gender, not neuter.

	hā ziiō, "winter" = hiemps
	sing	du	pl
nom	ziiō	ziiamōi	ziiamaHā
voc	ziiam
acc	xaiiā̊		xaiiōi
gen	ximō	ximō	ximām
loc	ximām		xīsū
dat	ximā		ximuš
ins	ximōi		ximβiyō

ϑázā

Most suffixed nouns with OX inflection have a long o-grade suffix in the nominative singular, cf. xaiiō, akin to the n-stems. Indeed, given phonetic change, *-ōm and *-ōn cannot be told apart, so it is possible this ending has been transported from the n-stems or generalized in the proto-language or later. Yet in the nom. sing. ϑázā xmō "earth" has an ultima reflecting a zero grade as in *dʰégʰ-m̥. Had the second vowel been a typical long o-grade, it would have become *ō and would not elide the guttural consonant, giving *ϑáɣō.

In the acc, the sequence *-em-m̥ in the proto-language resolves prehistorically to *-ēm via Stang's law and appears in Galic as -ā̊, while the equivalent sequence for n-stems *-en-m̥ yielded Northian -əmā.

	hā ϑázā, "ground, earth" = χθών
	sing	du	pl
nom	ϑázā
voc
acc	xōi		lg ϑámō
gen	xmō
loc	xšam
dat	xmā
ins	xmōi

To this table above must be subjoined that xā̊ < *ǵʰḗm < *dʰǵʰém-m̥ is only found in a handful of instances in Galic Period I, and even there it is not exclusive. The proto-form of the accusative is uncertain, since many authorities predict this word should have a full-grade root syllable, which should result in dysyllabic *θaɣā̊. But the form actually found is monosyllabic, which could only come from a zero-grade root. If the oblique stem *dʰǵʰm- was later introduced to replace the nominative stem, the resulting form would have been expected xam, with the *m of the stem probably vocalized as short by position.

n-stems

This group of nouns, all with a stem ending in -n, had numerous subtypes. Some of them were proliferate in Northian, while others were vestigial. Their phonetic outcomes can be quite different due to the paradigmatically irregular vocalization of resonants, which are particularly common because the resonant suffix is often attached to another resonant. It appears the suffix -n could be attached to nearly all other suffixes at some point in the prehistory of the language, with only the combination of *-r-n- being unattested. The common feature of all n-stem nouns is that the primary ablauting element is the syllable containing -n-, with or without the root syllable also participating; all other suffixes do not ablaut when followed by the -n suffix.

As a general observation, most n-stems can be divided into those of amphikinetic origin and those of hysterokinetic origin. Only the latter are productive in the recent prehistory of Northian. Amphikinetic nouns, e.g. laēmō "lake", always have an accented root in the nominative singular and more often than not lose the final *-n there; the kʷetwóres rule applies to them in the accusative singular, as the shape of the stem is always *CéC-on- > *CeC-ón, e.g. limōi. Hysterokinetic nouns have accented suffixes and endings, while the root is never accented. Amphikinetic nouns are actively transferred into the hysterokinetic paradigm during the historical period and thereby acquire an invariant root, which may retain the full grade or receive the zero grade instead.

mn-stems

A common subtype of the n-stems is through the suffix *-mn-, which is common in Northian. The OX pattern was normal for animate nouns, and the PX in neuter nouns. Items with this inflection can further be divided into sub-classes: some are of amphikinetic origin like asmō in the following example, while others are of hysterokinetic origin.

In ā́smō "life", the nom. singular ending -ō (which occludes the stem consonant) arises due to the effects of a sound law that deleted final *-n in the environment of *-ōn, though some related languages have either retained or restored this stem final *-n. A few ablauting nouns in this class have an accented root in the direct cases and accented ending in oblique cases, except in the locative singular where the accented suffix is attested (note that "ablauting nouns" in Northian usually refers to root-ablauting nouns, rather than suffix-ablauting nouns, as these are regular).

In the accusative singular and plural the deletion of geminate consonants operates after adjacent nasals assimilate in early Northian history, causing compensatory lengthening of the suffix syllable *-en- in each case the ending *-m disappears.

There are two phonologically-regular oblique stems arising conditionally: where the ending began with a vowel, the *m part of the suffix -mn- was vocalized, and where it began with a consonant, *n was vocalized instead. Thus, the gen sing is tanō and gen du smaHō, given the presence of a laryngeal in the latter. Yet it should be noted that such a regular application of the vocalization rule is exceptional. Most nouns have generalized the vocalization pattern of *-m̥n- rather than *-mn̥-, possibly because the commonly-used cases tended to begin with vowels and trigger *-m̥n-.

Owing to the presence of an o-grade following an e-grade in a three-syllable word, the kʷetwóres rule intervenes in the nom pl and shifts the accent to the suffix syllable. The apparent e-grade is considered an analogy to the en-stems. The rule does not operate regularly in case the final syllable is long, but here the final syllable actually reflects short *-eH, which entertains the rule and also shifts the accent to the o-grade.

	ha ā́smō, "life"
	sing	du	pl
nom	ā́smō	āsmә́nōi	āsmә́naHā
voc	ā́smən
acc	ā́smōi		smə̄ŋh
gen	tanō	smaHō	tanõm
loc	smən		smahū
dat	tanā	smamō	tanuš
ins	tanōi		smaβiiō

fnaōmin "breath" is an ablauting mn-stem noun with PX inflection. Owing to its semantic connection with fraōmō "lung" as well as the shape of the oblique stem under the influence of u, the alternate spelling fraōmin is also seen, for example G.Nr. 1477 mōi βā xšəm apū praōmənō praōmən dite "By the Earth, let therefore not [his] breaths be given away from his two lungs". But these two words are from etymologically unrelated roots.

	ϑaṯ ϑərma, "edge"
	sing	du	coll
nom	ϑərma	ϑә́ranī	ϑərmā̊
voc
acc
gen	ϑarəmáŋh	ϑarəmә́nuš	ϑarəmә́nõm
loc	ϑarəmә́ni	ϑarəmənū	ϑarəmāhū
dat		ϑarəmә́nō	ϑarəmāmuš
ins			ϑarəmāβiiō

apṓ

apṓ "holy waters" < *h₂ep-Hō(n) has the same suffix, but preceded by a laryngeal that colours the vowel. As a result, the rule that deletes final resonants following lengthened *-ō comes into play and produces the same ending as the animate mō-stems. But here in the gen. and loc. du. it is the suffix *-n- that becomes vocalized when followed by a consonant. The resulting vowel is short on account of the preceding laryngeal, with regular alterations of quality and quantity to surrounding vowels.

voc sing ápan has regular recessive accent for vocatives; otherwise accent is regularly on the suffix syllable. acc pl requires the weak stem of *h₂ep-Hn-m̥s. Under Cloverdale's law, interconsonantal laryngeals count as consonants, which means the word is underlyingly scanned as *h₂epH-nm̥s; thus the vocalized *m̥ will be short and produce *-ah > -āʰ. In some Early Galic phrases, this word was indeed only dysyllabic and not trisyllabic as the orthography suggests, recovering the extreme antiquity of the scansion of -i- as a laryngeal consonant.

	hā apṓ, "holy waters"
	sing	du	pl
nom	apṓ	aponōi	apə̄naHā
voc	ápən
acc	apə̄nā		apnā
gen	apnō	apə̄nō	apnõm
loc	apə̄n		apšū
dat	apnā	apmō	apmuš
ins

xaēñin

This group of nouns had the compound suffix *-i-n- and showed no ablaut, irrespective of gender.

	ha xaēnin, "girl"
	sing	du	pl
nom	xaēnin	xaēninōi	xaēninaHā
voc
acc	xaēninā		xaēninā̊
gen	xaēnī	xaēninō	xaēninõm
loc	xaēnini		xaēnaŋhū
dat	xaēninā	xaēninmō	xaēninmuš
ins	xaēninōi		xaēnimβiyō

xrauuštō, anžiiō

	ha xrauuštō, "reputation, honour"
	sing	du	pl
nom	xrauuštō	xrauuštenōi	xrauuštónaŋhāʰ
voc	xrauuštan
acc	xrauuštṓnəm		xrauužnā̊
gen	xrauužnōḫ	xrauužnōš	xrauužnõm
loc	xrauuštan	xrauužnō	xrauužnū
dat	xrauužnai	xrauužnmō	xrauužnmuš
ins	xrauužnōi		xrauužāβiiōḫ

	ha yaoxštō, "team of draft animals"
	sing	du	pl
nom	yaoxštō	yaoxštenōi	yaoxštonaŋhāʰ
voc	yaoxštan
acc	yaoxštinąm		yaoxštanā̊
gen	yāoxnōḫ	yāoxnōš	yāoxnõm
loc	yaoxštan	yāoxnō	yāoxnū
dat	yāoxnai	yaoxštmō	yāoxāmuš
ins	yāoxnōi		yāoxzāmβiiōḫ

hmərun

There is some debate over the origin of this class of nouns that seemed to be rare otherwise in the other daughter languages, as there are few comparanda. Such a conspicuous absence of comparanda could very well indicate this type was young or even arose after Northian had split off from the parent language. In support of that chronologization, all members of this class had completely non-ablauting roots with accented full grade, and an invariant suffix in zero-grade. The stem nevertheless exhibits inverse vocalization patterns depending on the vocalism of the ending, which means it must at least predate the vocalization of syllabic resonants.

A few members of this class are also neuter in gender, which are declined with the same apophonic pattern, except with accusatives identical to the nominative; some of them are associated with the acrostatic u-stem, e.g. hánun "knee". It has been noted this entire declensional pattern reminiscs of the old acrostatic neuter pattern ending in *-u, only enlarged with a suffix *-n of unclear function. That connection is mooted by a unique, newly-identified (2020) instance of gen sg ending -ū in G.Nr. 1127 hánū, which by that analysis represent the zero-grade genitive ending in *-un-s. Such a form would be difficult to ascribe as a recent innovation. As yet, it is not agreed which type was the older or if they share a common origin; the more prevalent view regards the neuter type as the original, and the non-neuter type as secondary to other nouns.

The Nordic scholar Havarsson has connected this suffixal *-n to the Nordic weak noun declension, which were n-stems in the ancestral language, saying "this mysterious Galic word could bear witness to the ancient origin of the weak nouns, hitherto believed to be a young and deviant noun type." Yet he also provides that since hanun is a hapax legomenon and is glossed as an equivalent to nom sg hánū "knee", it would be most difficult to explain functionally how these formations are alike in origin.

	ho hmərun, "bone marrow"
	sing	du	pl
nom	hmərun	hmərunōi	hmərunaHā
voc
acc	hmərum		hmərū
gen	hmərunō	hməruuaHō	hmərunam
loc	hməruni		hməruuāhū
dat	hmərunā	hməruuāmō	hməruuāmuš
ins	hmərū		hməruuāβiiō

bare n-stems

There are also n-stems not part of a compound suffix like *-mn-. They are rarer and sometimes emphasized as "bare" n-stems. They are of two types, distinguished in the nom. sing., those with -ō and those with -ā̊. Though less common, they are typical of adjectival origin and often family names; the Venetian name Cato is a cognate example, < *kat-ōn "the smart one".

In uə̄rštā̊ "male of an animal", of the type ending in -ō, PNN stems are ablauting *uors-on- and *uərəs-n-. In the nom. sing. the root vowel is lengthened, occasioning the loss of the root-final resonant prehistorically. As it will appear, the stem-final -n- is vocalized if the ending begins with a consonant; the gen. and loc. du. begin with the prehistoric laryngeal, which still trigger vocalization and are reflected as Galic hiatus and quantitative and qualitative alteration. After *r and its vocalized allophone, *s obligatorily becomes š, which cannot precede a vowel directly and to which a t is added. However, this addition is chronologically late and therefore applicable to the loc. and gen. du. forms, even though earlier they were considered to begin with consonants and trigger vocalization.

	ho uxšō, "bull"
	sing	du	pl
nom	uxšō	uxšənōi	uxšənā
voc	uxšən
acc	uxšənā		uxnā
gen	uxnō	uxšənō	uxšənõm
loc	uxšəni		uxšahū
dat	uxnā uxšənā	uxšamō	uxšamuš
ins	uxnōi		uxšaβyō

r-stems

duxiϑṓ

duxiϑṓ "daughter" represents the oxytone group of the r-stems, which includes many agentive nouns that terminate in *-tēr.

Similarly inflected: fϑō piϑrō "father", nomina agentis in -ā̊ and -ō.

	hā tuxiϑṓ, "daughter"
	sing	du	pl
nom	dukiϑṓ	dukitərōi	dukitəraHā
voc
acc	duxiϑərā		dúxiϑrā̊
loc	duxiϑəri	duxiϑarHṓ	duxiϑrštū
gen	duxiϑrō		duxiϑrõm
dat	duktrā	duxiϑrmṓ	duxiϑrmō
ins	duxiϑrōi		duxiϑrβiyōḫ

mā̊tar

mā̊tar "mother" represents the group with inherited acrostatic pattern, whose accent persists on the root syllable and always takes suffix and ending in zero-grade. βrātar "brother" is declined in like manner.

gen.sg mā̊tarəš from *meh₂tr̥s. nom.pl, aside from more frequent mā̊traHā, has alloforms mā̊traH ~ mā̊traHaH preserved in Old Material, in spondee and bacchius positions; these attested to the more archaic simplex nom.pl ending *-es as well as complex *-eses. The bacchius reading precedes another word that beings with *s-, whereby degemination between words caused the ending to lose its final consonant.

The acc.pl mā̊tərā seems to reflect PEE *meh₂tr̥n̥s (for expected *meh₂tern̥s), while most daughters reflect *meh₂tern̥s. The form is a prima facie deviation from the general rule of resonant vocalization, but the vocalization of *r̥ could have been carried over from the gen. sing., where it is the only resonant that could be vocalized; moreover, it is a perfect match with Xevdenite mātərąs, hinting that the vocalization of -n̥s was not a firm rule in the proto-language. Perhaps the lack of vocalization here could also be explained by analogy to the acc. sing., where the -m is never vocalized.

In Galic, mā̊tar "two mothers" is often an ellipsis for "mother and father".

	hā mā̊tar, "mother"
	sing	du	pl
nom	mā̊tar	mā̊tar	mā̊traHā ~ mā̊traHaH ~ mā̊traH
voc
acc	mā̊tra		mā̊tərā
gen	mā̊tarəš	mā̊taruš	mā̊tram
loc	mā̊tar	mā̊tarū	mā̊tərəšū
dat	mā̊tri	mā̊tərəmō	mā̊tərəmuš
ins	mā̊trōi		mā̊tərəβiiō

yāϑr

	hā yāϑr, "husband's brother's wife"
	sing	du	pl
nom	yāϑr	yaϑəra	yaϑeraŋhāʰ
voc
acc	yaϑram		yāϑrā̊
gen	yaϑōs	yaϑəruš	yaϑrõm
loc	yaϑéri	yaϑərū	yaϑrštū
dat		yaϑərma	yaϑərmuš
ins	yaϑəra		yaϑərβyōḫ

Feminines in -zarə

Certain Northian nouns formed feminine counterparts with the suffix *-sr.

	hā hwózr, "sister"
	sing	du	pl
nom	hwózr	hwozórōy	hwózoraŋhāʰ
voc
acc	hwozóram		huzrā̊
gen	huzrōʰ	huzrōʷ	huzrõm
loc	huzeri		huzrštū
dat		huzrma	huzrmuš
ins	huzra		huzrβyōḫ

l-stems

There are a few nouns which anciently had an ablauting stem ending in -l, the most important being āuuarə "Sun", nəmβar "navel", naōgar "nail", əṇgar "coal-fire", and abar "apple". Aside from nəmβar, a feminine noun, all attested l-stems are neuter heteroclitics that attest a stem in -n in oblique cases.

For the most part, l-stems are indistinguishable from r-stems since the two sounds merge completely in Northian, but traces of their former existence can be seen. Since the parent language prohibited geminates, the sequence of nom.sg *CVr-r̥ would simplify to *CVr-, which appears endingless and is indeed found in that shape in Northian. But if the suffix had been *-l̥, this sound does not degeminate and surfaces in Northian as CVr-r instead, as in pərarə "a type of food-offering", reflecting *per-l̥; this forms, indeed, a minimal pair with pər < *per(-r̥) "mountain".

s-stems

The s-stem nouns can be divided into two classes, more numerous, neuter nouns like xrétō and animate nouns like awšā̊. The former suffix exhibits qualitative ablaut from *-os ~ es-, while the second class has a more conventional, quantitative -os ~ s- pattern.

xrétō

The os/es-stems are a class of very common neuter nouns in Northian and are directly cognate to those found in Nordic languages, where the ablaut surfaces as -az ~ iz-.

In PX the nom sing ended in -ō < *-os, a bare stem. The suffix ablauted to -es- for all oblique cases, but here the ending once again acts prominently on the suffix. If the ending did not begin with a voiceless stop, the -s- of the suffix disappeared or become -H- intervocalically and coloured any neighbouring /e/ to /a/. As a result, the gen. ended in -aHā < *-es-es. If the ending began with a voiced consonant, the -s- elided and caused compensatory lengthening to the preceding *e-, which surfaced as -aē- instead. The gen loc du endings began with a laryngeal consonant, which assimilated with the preceding -s- and was deleted.

	θaṯ xrétō, "intellect, power"
	sing	du	pl
nom	xrétō	xrétaHā e xrétā	xrétā̊
voc
acc
gen	xrétaHā e xrétā	xrétaHuš e xrétōš	xrétaHam e xrétām
loc	xrétā	xrétaHū e xrétō	xrétahū
dat	xrétaHā e xrétā	xrétaēmō	xrétaēmuš
ins			xrétaēβiyō

aōšā̊

The noun aōšā̊ "dawn" has feminine gender and therefore a distinct accusative. The noun had three stems: the strong stem which appears in the sing. nom. and voc. comes from PEE *h₂eu̯s-os-, with regular lengthening of the final syllable anticipating a zero nominative ending; the middle stem appears in the other direct cases and is from *h₂us-es-; the weak stem appears in all other cases and is from *h₂us-s-, which dissimilated to ušt- in Northian, with full-grade endings as expected in OX nouns. The acc. sing. shows elision of *z and compensatory lengthening. The *z was generated before a syllabic nasal *m̥.

	hā aōšā̊, "dawn"
	sing	du	pl
nom	aōšā̊	ušaHā	ušáHā
voc	aōšō
acc	ušézā		uštā
gen	uštō	uštō	uštām
loc	ušā		ūštū
dat	uštā	ušmō	ušmuš
ins	uštōi		užβiyō

i-stems

mātiš

The i-stems were a prolific class of nouns in Northian during the Galic period. In the proto-language, the i-stems were completely parallel to the u-stems in virtually all ways, but due to sound changes their surface forms in Northian are quite different. Accordingly, they are considered separate classes in Northian tradition.

The PX pattern of the i-stems gained primacy early in Northian history, and these nouns proliferated, being built to a variety of roots. These nouns usually show no root ablaut, but the example provided here does. Where the root did not show ablaut, the accent was presistent on the suffix, even in zero-grade. The other forms require little comment, other than the gen. and loc. du. and gen. pl. where the suffix *-i is consonantal before endings beginning with a vowel, which here triggers the palatalization of preceding *-t.

Of the neuter nouns in -i, only one example can be considered original to the parent language, that being məʸri "sea". Most other neuter nouns can be shown to be petrified neuter forms of adjectives in i-stem. It is assumed that the neuter i-stems would have formed a collective in the same way as the u-stems, but such is not attested.

	hā mātiš, "thought"			ϑaṯ məri, "sea"
	sing	du	pl	sing	du	coll
nom	mātiš	mātī	mātaiiā	məri
voc	māti
acc			mātī
loc	mātō	mātaiiō	mātišū
gen		mātaiiuš	mātaiiõm
dat	mātaiiā	mātimō	mātimuš
ins	mātī		mātiβiiō

ākō

The OX pattern of the i-stems was moderately productive in Galic times and less so by later ages. In the nominative du. and pl., the suffix is in long o-grade; the suffix consonantal *-i is lost. The same deletion corresponds exactly with forms found in archaic grammars in Syaran and Tennite languages, so the dropping of final -i is likely to have been old.

Perhaps owing to the phonetic similarity between the dual and plural nom. forms, *-ē > -ōi, which is not expected after i-, often displaces the expected -ā. However, there is also a hapax of short -e found in Galic and in final position, which suggests the -ōi may be a late alteration, inserted after du *-aH had ceased to be distinguishable in speech from pl *-ah.

This pattern has a neuter equivalent that ended in -ā < *-ei̯ in the nominative; otherwise, oblique cases inflect identically. This neuter pattern is very rare and only present in a few examples like óxϑā "palm", whose archaic dual form óxϑō means "eight".

	ho ākō, "friend, ally" = socius			θaṯ óxϑā, "palm"
	sing	du	pl	sing	du	pl
nom	ākō	hokoiiā dm hokoyōi	hokoiiā	óxϑā	óxϑō
voc
acc			hokiiā
loc	hokā	škiHō	škišū
gen	hiiō		hiiõm
dat	hiiā	škimō	škimuš
ins			škiβiiō

u-stems

Ablauting u-stems in PX

	etó hūiiuš, "child" = υἱύς			etóṯ oiiū, "life, generation"
	sing	du	pl	sing	du	coll
nom	hūiiúš	hūiiū	hūiiáuuā	oiiū		oiiō
voc	hūiiū
acc	hūiiúm		hū́iiū var. həruuā
gen	hūiiō	hūiiáuuō	hūiiauuam
loc		hūiiū	hūiiušū
dat		hūiiumō	hūiiumuš
ins	hūiiū		hūiiumβiiō

The PX pattern in u-stem nouns is very prolific in both the Galic and Epic languages. Many are abstract nouns with the ending -tuš, though combinations with -m- and -s- are also seen.

The nom. ended in -uš. While many Northian nouns showed root ablaut, none of the nouns in the pattern of hūiiúš do and therefore have all components in zero grade in the nom sing; the hypothetical forms *haōiiuš < *sewHyus or *sowHyus might be expected based on general ablaut tendencies, but it is in fact the phonological pre-form of hūiiúš that is reflected in all Erani-Eracuran languages, showing, if once these nouns had a full-grade root, that the spread of the zero-grade root probably occurred in the proto-language. Yet it is also suspected this pattern only became productive after root ablaut ceased to be productive, as suffix ablaut is aberrant compared to other stem types (viz. below).

Note the voc.sing form hūiiū which contrasts with acc.pl hū́iiū only by accent.

The gen., dat., and loc. forms are all hūiiō in the oldest texts, but they have different sources. The gen. ended in *-ou̯-s, where *-s has dropped after the final resonant, giving *-ōu̯, which changed to -ō in Northian in final position. In the dat. and loc. the final *-i is lost after the stem resonant. But the stem-final -u̯ lingers as a spectre in sandhi: a following *-kʷe delabializes and becomes -xa < *-ke. In the later Gales, a final -s is often restored to the genitive unless prohibited by metre. In this instance, Northian presents the most archaic forms and demonstrates that widespread compensatory lengthening had indeed occurred prior to analogical restoration, which in all other branchs obscured the outcome of the sound change. The Northian restoration of -s must be later than the period when -s is retroflexed after *-w, as it is written -s.

The nom. du. has the ending -ū, which is from *-u-h₁, and then the gen. has -áuuō < *-ew-ow; both have unexpected ablaut grades. It has been suggested the full grade in the gen. could have been introduced from the singular; if so, it must have been before the final *-ow > -ō. The loc. du. has -ū, which must reflect *-u-u, but this an irregular syncopation acorss a laryngeal boundary, which is otherwise well-preserved in Northian.

The nom. pl. has the expected forms, while the acc. pl. shows the suffix in the weak grade, *-u-ns > Galic -ū. The weak suffix in the acc. pl. is often taken as a sign of lateness, but there is a common alternative form associated with certain words that has -uuā as an ending. There is debate which one is the more original, as the sequence -uuā is a more regular outcome of the general ruler of resonant vocalization from the sequence *-wn̥s, while the more common variant -ū would reflect something like *-w̥ns. The latter can be seen as a tendency of thematization, that is the regular appearance of a certain sound between stem and ending. On the other hand, this "thematized" form appears in the earlier G1 texts more often and is better attested across cognate languages, leading some to assert that the -uuā ending is a result of subsequent re-analysis and that the more regular vocalization was not in use in the parent language.

The gen. has the strong suffix again and shows *-ew-om. All other forms in the plural are straightforward and have the zero-grade suffix -u-, which departure from the normal rules of proterokinetic declension has also been linked with the attempt to thematize the u- and i-stems.

The neuter version of the PX u-stems, in contradistinction to the animate ones, often displays root ablaut, which alternates strong and weak stems. The strong has o-grade from PEE *h₂ói̯-u > unchanged Galic oiiū, save the regular lengthening of final *-u. The oblique stem is from *h₂i̯-ou̯-s > also unaltered Galic yō. For this word, of great canonical significance, final *-s is never restored in Galic.

Ablauting u-stems in OX

There are also u-stem nouns with an OX pattern, of which there are several sub-types that are subsumed under the heading of hənuš.

	hā gənuš, "jaw" = jaw			hā haiiō, "horizon"
	sing	du	pl	sing	du	pl
nom	gənō	hənauuā	hənauuā	haiiō
voc	hənū
acc	hənā̊		hənnā̊
gen	jñuuōḫ	gnuōš	jñuuõm
loc	jñaō dm jñauui	gnuō	gnuštū
dat	jñuuoi	gnumō	gnumuš
ins	jñuua

The OX nouns had a nom. sg. end that vacillated amongst the lexical items between -uš and -ō / -ōs < *-ou̯-s. These reflect the zero and full grades respectively. The gen. and dat. reflects PEE *-u-os and *-u-ei̯ as expected. The dual behaves as expected. The acc. is on account of the deletion of resonants before a syllabic nasal, genā̊ < *gen-ēm < *gen-eu̯-m̥.

The acc. pl. gennā̊ < *gen-w-ns has the suffix assimilated into the root.

Note in this word some oblique forms have the stem jñuu-, which is an altered form of *gnw-. Northian generally permits up to three consonants (or four, if beginning with s-) in anlaut, but only if they contain no stops; if they do, stops are altered to fricative counterparts. This rule is not in effect for sequences of only two members in anlaut, hence the conserved gn- when the following *w is vocalized. Note also, the genitive and dative dual endings begin with a (lost) laryngeal, triggering the vocalization of *w.

The other EE languages attest a neuter version of this OX patern as in Tennite madhu gen. madhvas < *medʰ-w-es that is not found in Northian. This word in particular is found as an acrostatic meδū gen. meδuš. Instead, there is a solitary neuter noun haiiaō "horizon" that seems to have a neuter ending in *-ew. It is quite uncertain whether this is an inherited pattern or a petrified Northian adjective on the pattern of the s-stems adjectives (though it is equally unclear what original adjective would give the meaning of "horizon"), or it could be somehow related to ei-stem nouns attested in Xevdenite.

ϑórū

Some neuter nouns in -ū, such as ϑórū "tree, wood", hanū "knee", uuəštū "settlement" descend from an old acrostatic ablaut pattern that subsequently gained proterokinetic forms in the oblique cases. The original oblique stem *deru- is still visible as a compounding form. These changes result in a locative singular that is identical to the nominative. These special nouns need to be distinguished from those that also end in -ū like oiiū "life", which have a distinct locative.

	θoṯ ϑórū, "wood" = δόρυ, "tree"
	sing	du	coll
nom	ϑərū	ϑərū	ϑərō
voc
acc
loc
gen	drā
dat			drumō
ins	drū

The word ϑərū gives rise to several intriguing concepts in related languages. In Nordic, an important derivative is triwwiz "true, faithful", evidently comparing a reliable person's character to the durability of wood. Northian drəmβuuōḫ, reflecting the same *drew-wos, means "firm, difficult", from the same sense that wood is resilient and hence difficult to manipulate. But the word carries more of a neutral tone than the usually-positive one in the Nordic cognate: a person described as drəmβuuō in the Epics is inflexible and ill-disposed to the speaker, but not hostile. Compare another possible cognate, duros in Venetian, which means "hard, harsh" and is more overtly critical and pejorative applied to persons, cp. dura lex sed lex.

ī-stems

orī

The ī-stems in paroxytone has remained productive down to the Epic period as a feminizing suffix for athematic nouns. This suffix showed ablaut from full grade *-ieh₂ ~ -ih₂ > Northian -iiā ~ ī-. Some inherited nouns have a root that participated in ablaut, so the root was in full grade and suffix in zero grade in the direct cases, and vice versa in oblique cases.

Notice that, in the gen. oriiā̊, the laryngeal is syllabified with the preceding vowel and causes it to lengthen, but in the dat. oriiaē it syllabifies with the ending -i and does not lengthen the suffix; still, in both cases, the suffix vowel is coloured. The regular syllabification is only attested in Galic; by Didaskalic and Epic times, the dative was remade according to the genitive and has the ending -iiā-yi. The nom. pl. attests the ending -īš < *ih₂-s, which features an unexpected zero-grade suffix and a zero-grade variant of the nom. pl. ending, which is unique here and *-es everywhere else. This zero-grade variation has been explained by some as rather the nom. sing. ending but not to general acceptance.

The acc. pl. ending is the same as the nom., which indicates the final resonant failed to vocalize, thus from *-ih₂-ns rather than *-ih₂-n̥s. This renders it the same as the acc. pl. of the i-stems of the type mātiš. Such a syllabification is attested also in the 3 pl. of the athematic optative for stems with persistent accent, which has -inṯ < *-ih₂-nt rather than *-iyat < *-ih₂-n̥t.

	hā orī, "eagless, hen eagle"
	sing	du	pl
nom	orī		orīš
voc
acc
gen	oryā̊	oriiāouš	oriiaõm
loc	oriiayi	oriiaoū	oriiāštū
dat	oriiaē	oriiāma	oriiāmuš
ins	oriiā		oriiāβiyōḫ

hənitrī

In the example below, Sievers's law operates in oblique form as the agentive *-tr- invariably creates a heavy syllable before the yod of the full-grade feminizing suffix *-yeh₂. Thus, in old texts, the surface forms are trisyllabic hātriyā̊ and hāϑriyay, but the disyllabic forms hāϑəriyā̊ and hāϑəriyaHē are often seen in texts of an only slightly younger vintage. Jeweller indicates that metrical considerations can be discounted from this variance. Minimally, these could be analogical replacements from words in like stem where Sievers's law does not operate, or maximally they could indicate Sievers's law governed a kind of free variation but not a definite shift. The shortcoming of the former is that stems obtaining Sievers's are far more numerous than those not, owing to the preponderating usage of the feminizing suffix after another suffix.

The -o- in the gen. and dat. du. forms is a Runic orthographic insertion done to prevent the incorrect scansion of dysyllabic /a.us/ as monosyllabic /aws/; the -o- is spurious and silent. It is not present in all texts: monumental inscriptions usually omit it.

	hā hənitrī, "genitrix" = genitrix
	sing	du	pl
nom	hənitrī
voc
acc
gen	hatriyā̊	hatriyaHō	hatriyaHõm
loc	hatriyā	hatriyaHā	hatriyāšū
dat		hatriHāmō	hatriyāmuš
ins	hatriHā		hatrīβiyō

štərī

This type, consisting mostly of feminine derivatives, was much more common in the Gales than in later texts.

	hā štərī, "heifer"
	sing	du	pl
nom	štərī	štəriiā	štəriiā
voc
acc			štəriiā̊
loc		štrīHō	štrīšū
gen	štriHō		štriHõm
dat	štriHā	štrīmō	štrīmuš
ins	štrī		štrīβiyō

ū-stems

ϑənūˀ

The ū-stems were originally parallel to the ī-stems, but phonetic changes have rendered them distinct. The nom. sing. always had accented root syllable and an ending in PEE *-uh₂ > Galic *-ūˀ. The acc. requires a proto-form *-ueh₂-m, but already has simplified to *-u̯ām > -uā̊. In the oblique cases, the stem alternated between -ū preconsonantally and -uH prevocalically, to which full-grade endings were added. Other disfigurements are attributable to the shape of this stem. The stem in zn- occurs whenever the *u of the stem is vocalized in the environment of *tnuh₂-, and the stem in dā-, when *u borders a vowel and *n is vocalized instead, i.e. *tn̥weh₂-. The irregularity in this paradigm was preserved certainly owing to its frequent usage.

	hā ϑənū, "body"
	sing	du	pl
nom	ϑənū	ϑənnaHā	ϑənnaHā
voc
acc	ϑənnā̊		dāuuā̊
gen	znuHō	znuHō	znuHõm
loc	dāuuā		znuŋhū
dat	znuHā	znūmō	znūmuš
ins	znuHōi		znūβiyō

hōxrū

The ū-stems also includes one member with paroxytone accent, namely hōxrūḥ "mother-in-law".

	hā hōxrū, "mother-in-law" = socrus
	sing	du	pl
nom	hōxrū	hōxrō	hōxrwā
voc
acc	hōxrwō
gen	hōxrwā̊	hōxrwāHuš	hōxrwaHõm
loc	hōxrwā	hōxrwāHū	hōxrwāhu
dat		hōxrwāmi	hōxrwāmuš
ins

H-stems

The H-stems or laryngeal-stems have a suffix consisting of an ablauting vowel followed by a laryngeal and were similar in structure to the ī- and ū-stems. But, there, semivowels *i- and *u- preceded the ablauting vowel, and so in oblique cases, they become vowels in the absence of the ablauting vowel and are lengthened by the laryngeal. Here, however, the ablauting vowel is not preceded by a semivowel and disappears completely in the oblique cases, causing the laryngeal to scan as part of the ending syllable.

pəntā̊

The word pəntā̊ "path" had a stem ending in PEE *póntoh₁- ~ pn̥th₁-. Fundamentally, it is a regular laryngeal-stem noun of amphikinetic origin of the same type that has full-grade root and o-grade suffix. In the accusative singular, the ending was *oh₁-m̥ > *ōm, with regular operation of Stang's law.

This word is a direct cognate to the Shalumite word "path", which is a borrowing from the oblique stem of the reflex in another Erani-Eracuran language.

	ho pəntā̊, "path, way" = "path"
	sing	du	pl
nom	pəntā̊	pəntōi	pəntoHaHā
voc
acc	pāϑā̊		pəntoHā
gen	pāϑō		pāϑām
loc			pāšū
dat	pāϑā	pāsmō	pāśmuš
ins	pāϑōi

bā̊

bā̊ "woman" continues the proterodynamic declension ending in *-h₂, and the nom. was asigmatic. This suffix also underlies the ī- and ū-stem declensions in Northian but is otherwise rarely seen alone. The direct stem is from *gʷénh₂ > *gʷḗn > gā̊, and the oblique *gʷnéh₂- > gnā-. Both nominative and accusative forms are affected by Stang's law.

	hā gā̊, "wife" = γυνή
	sing	du	pl
nom	gā̊	gənā̊	gənā
voc
acc	gənā̊		bnā
gen	bnā	bnaHuš	βna·ā̊m
loc	bnaēii	βnaHū	βnāhū
dat		bna·ā̊	βnāmuš
ins	βnā

r/n-stems

The r/n-stems, or heteroclitics, continue a class of EE nouns that had different suffixes for direct and oblique stems. Excepting productive derivative suffixes *-tr̥ and *-mr̥, they are all neuter nouns with basic meanings and show P/PX inflection in the singular and dual. Many, but not all, formed a collective with an o-grade root with OX ablaut pattern. The collective was endingless in the nom, like a singular, but took plural endings in the oblique cases.

In the Epic language, many heteroclitics also formed ordinary plurals from the zero-grade stem, which agree with singular verbs and have meanings different from those of their collective forms.

yuHarə "year" is a case where the spelling of the Gales leaves open a number of interpretations as to their phonological forms. The nom yuHarə < *yoh₁-r̥ had the o-grade of the root. Proto-form of gen *yoh₁-n̥-s or *yeh₁-n̥-s should anticipate a PNR form like *yaH-ā̊, but nowhere is this found or metrically allowed; instead, one finds monosyllabic yā̊, which according to Krueger may be an ad hoc replacement for *i-āh, logically assumed to be from < *ih₁-n̥-s. However, the contraction of syllables across laryngeal borders is very rare in Northian, so this explanation has not achieved agreement by authorities. du loc yaŋhaˀū is surely for a rather awkward *yaˀāˀū < *yéh₁n̥Hu.

	θaṯ yuHarə, "year" = ὥρα → "hour"
	sing	du	coll
nom	yuHarə	yōrī	yōrō
voc
acc
gen	yuHā̊	yaHāHuš
loc	yūni
dat
ins

pāwr̥ "fire" has PX inflection, with accented suffix in the oblique cases. In the nom. sing. the proto-form was from *péh₂wr̥; the final -ərə is paedagogically taught as /ara/, but as it is a single long syllable and not two short as the orthography implies, authorities concur it probably represents a preserved, tautosyllabic -R-r̥ sequence. This ending also occurs with the *-mr̥ compound suffix, but not *-tr̥. Oblique stem is from *ph₂wén- > *fiwen > Galic fūvon-. The collective form is fūvə̄, which is presumably from *ph₂wṓ, showing the replacement of full-grade root by zero-grade that is common in derived OX nouns, cp. direct cognate in Elder Nordic fōr.

	θaṯ pāwr̥, "fire" = πῦρ → pyro-
	sing	du	coll
nom	pāwr̥		pāuuō
voc
acc
gen	puuanh
loc	puuən
dat	puuənā
ins	puuā̊

Other common heteroclitic nouns include:

• PX azr̥ azā̊ "day"
• PX fetar fθā̊ "feather"
• PP woδar woθā̊ "water" < PEE *wodr̥ *wodn̥s
• PP ferərə ferūš "mountain"
• PP fō ferā̊ "house"

ā-stems

The main point of interest in this paradigm is in the genitive and locative dual forms, where the form suggests the e-grade *-ew- instead of the anticipated zero grade -u-. Nowhere else in Northian or other Erani-Eracuran languages does this ending take the e-grade.

	hā mīrā, "mist"
	sing	du	pl
nom	mīrā	mīraHī	mīraHā
voc
acc	mīrō		mīrā
gen	mīraHā	mīraHā	mīraHõm
loc	mīrā		mīraHū
dat	mīraHā	mīrāmō	mīrāmuš
ins

o-stems

The thematic declension in o-stem could be divided into two subclasses that arose on account of the position of the accent, without which the short *o became /a/.

	ha θūmōḫ, "smoke"			θaṯ yuɣõm, "yoke"
	sing	du	pl	sing	du	pl
nom	θūmōḫ	θūmṓḥ	θūmā̊	yuɣõm	yuɣṓ(yīḥ)	yuɣāḥ
voc	θūmi
acc	θūmõm		θūmə̄ṇġ
gen	θūmōiiō	θūmōš	θūmõm
loc	θūmōi	θūmō	θūmohū
dat		θūmoma	θūmomuš
ins	θūmō		θūmomβiiōḫ

Acrean declensions

Many words in later Northian are Early Elder Nordic loanwords from speakers of Acrean, which was used as a lingua franca in western Eracura for centuries under the influence of the Acrean Empire. These words were so numerous that many were not nativized but declined according to an approximation of the thematic declension in ELder Old Nordic. Naturally, Old Nordic vocabulary would not occur within the Galic and Epic corpora, but as the Epic language survived in literary and liturgical usages, later material did incorporate a considerable number of Acrean words.

Hybridized paradigms

The displacement of native vocabulary in favour of Acrean words was widespread and penetrating to a basic level, unlike previous assumptions that it was only introduced for topics of exotic political and commercial interest. Baker notes that the Old Nordic word sōwulą sōwulas "Sun" has hybridized with native Northian hāuuərə hūvə̄ṇġ and displaced the latter's oblique forms. Displacement concentrated in Epic forms made opaque by regular sound change. While all these forms are now analyzed as regular, at least in this noun they were no longer so understood after the Epic age.

The Northian inscriptions from the Epic age are revelatory of the extent of Nordic substitutions from an early time. Nordicisms were regarded as "vulgar substitutions" by antiquiarians, but a mid-6th-century BCE stone discovered in 2002 has the word <sōwulas> in genitive usage for the solar goddess, found within her very own temple precinct. Such finds have cast doubt whether Nordicisms were really regarded as "vulgar" forms by Northians of the Epic age, and some scholars have come to think the "correct" forms like hūvə̄ṇġ were actually poetic forms that had not been used regularly for some generations by that time.

	hā hāuuərə, "Sun"
	sing	du & pl
nom	hāuuərə	sōwulōs
voc
acc
gen	sōwulas	sōwulōm
dat	sōwulai̯	sōwulamas
ins	sōwulō	sōwulamiš

a-stems

For the most part, these endings are cognates with the o-stems (2nd declension) ones in Northian, but because of intervening sound changes, they are taught separately. Where there was a distinct vocative, the nominative form has displaced it. There are also no dual forms for these nouns, for which neologisms have sometimes been suggested, to no general acceptance. Since Elder Nordic had no locative case, this form is always identical to the dative where Northian syntax demands the locative.

	ha wulfas, "lupine" = lupus		θaṯ ētą̄, "food"
	sing	du & pl	sing	du & pl
nom	wulfas	wulfōs	ētą̄	ētō
voc
acc	wulfą̄	wulfāŋš
gen	wulfas	wulfõm	ētas	ētõm
dat	wulfai̯	wulfamas	ētai̯	ētamas
ins	wulfō	wulfamis	ētō	ētamis

ō-stems

These are cognates of the ā-stems in Northian.

	sā erδō, "Earth"
	sing	du & pl
nom	erδō	erδōs
voc
acc	erδōm	erδōs
gen	erδōs	erδōm
dat	erδōi̯	erδōmas
ins	erδō	erδōmiš

Venetian declensions

Old Venetian was another Erani-Eracuran language whose terms have been borrowed into Northian during the Late Canon Period. The first and second declensions of Old Venetian, in -ā and -os respectively, were directly cognate to the Northian ā- and o-stems.

First declension

	hā azā, "altar"
	sing	du & pl
nom	azā	azās
voc
acc	azām	azas
gen	azās	azām
dat	azāi	azāis
abl	azād

Second declension

	ha cabos, "measure of corn"		θaṯ alom, "comfrey"
	sing	du & pl	sing	du & pl
nom	cabos	caboi	alom	alā
voc	cabe
acc	cabom	cabōs
gen	cabōiio	cabozom
dat	caboi	cabois
abl	cabod

Third declension

The third declension in Venetian was cognate with the athematic nouns in Northian, though in Venetian the number of sub-patterns was smaller. As learned borrowings, they generally retained their original endings, though substitutions were only too common.

	hā loux, "light"
	sing	du & pl
nom	loux	loukeiies
voc
acc	loukem	loukēs
gen	loukes	loukom
abl		loukibos
dat	loukei

Fourth declension

	hā manus, "hand"		θaṯ cornu, "horn"
	sing	du & pl	sing	du & pl
nom	manus	manūs	cornu	cornuā
voc	manu
acc	manum
gen	manuos	manuom
abl		manubos
dat	manui

Irregular nouns

ā̊ "mouth"

ā̊ is an acrostatic neuter root noun in s-stem. Like other acrostatic nouns, its inflectional pattern can show irregularities under the influence of other, more productive ablaut patterns. After the Late Canon period, all the oblique forms of this noun were replaced by thematized forms based on the stem ā̊-, e.g. gen. ā̊ōiiō, but it is the athematic originals that are discussed below. Nom. sing. ā̊ is for *ō-h < *HoH-s-Ø.

	θaṯ ā̊, "mouth"
	sing	du	pl
nom	ā̊	ōhīḥ	ōha
voc
acc
gen	ōs	ōhuš	ōhõm
loc	āḫ	ōhū	ōššū
dat	ōhi	ōma	ōmuš
ins	ōha

mā̊ "meat"

mā̊ < *mēm-s must be distinguished from the word for "moon", as outside of paedagogical texts they are both printed and written as mā. This is either a root noun or a reduplication of a root *me-ms-. The word has no dual or collective forms.

	θaṯ mā̊, "meat"
	sing
nom	mā̊
voc
acc
gen
loc
dat	meṃśi
ins	meṃśa

potiš "master" in compounds

The word *potiš "lord, master" has OX-B inflection. As a word it is not seen alone but does serve as the second element in the terms déṃpśpatiš "master of the house" and uuei̯xšpatiš "lord of the settlement". The word can sometimes appear with the o-grade suffix, e.g. déṃpśpatōiš.

While the root *potiš is not seen independently, the feminine -īḥ derivative potnīḥ is used as part of certain goddesses' titles; there, the full grade root is invariant.

	ha déṃpśpatiš, "master of the house" = δεσπότης
	sing	du	pl
nom	déṃpśpatiš	déṃpśpačiiāḥ	déṃpśpačiiāḫ
voc	déṃpśpatai
acc	déṃpśpatim		déṃpśpatī
gen	déṃpśpatiiāḫ	déṃpśpatiyōš	déṃpśpačiiõm
loc	déṃpśpati	déṃpśpatiyō	déṃpśpatištū
dat	déṃpśpačiiai	déṃpśpatimō	déṃpśpatimuš
ins	déṃpśpačiiāḥ

šiiō "god, sky"

This word šiiō is a direct cognate with several theonyms across the Erani-Eracuran family. The original *ew was transformed to *aw (written aō) after yod. Under the influence of that phone, the initial obstruent was palatalized to z, where it is otherwise preserved before vocalic i. There is an alternate nom. form žiiōš, which may either be a contraction or a reflex of *dyou̯-s, the o-grade of the same root. The acc. sing. žiiā̊ is a product of Stang's law, which requires the sequence of *dyeu̯-m to be simplified to *dyēm; this sequence is attested in virtually all branches and is assumed to be old. šyā̊ is cognate to Syaran Ζῆν.

šyō often co-occurs with the epithet ufšištō "highest" as ufšištoz-dyō "Heaven Most High", in much the same way as fərətištā-taɣam "Earth Most Broad". These compounds, other than being appellations of their titular deities, were also used of their agents. Aithar, the god of numina, is almost always accompanied by the epithet ufšištṓiia-diuuō "of Heaven Most High". After Aithar, the pantheon of Valstígr was also called āhaṓuuāḫ põm ufšištṓiia diuuō, "Lords of Powers and Heaven-Most-High", defining them as celestial, rather than chthonic, deities.

It is to note that the epithet ufšištōḫ "highest" does not carry the implication that the god is highest in rank or power, at least in the Galic and Didaskalic corpora. The idea of a deity that is supreme over other deities was not endorsed by the earlier Northians, and instead a god's supremacy was envisioned more as "excellence" or "extremity", or simply the quality or domain assigned to that god in a superlative, peerless state. Thus in Northian theology, Ziiaōš was the highest, and θaɣam the broadest, and the two are both peerless in the qualities recognized in them. But, being differently or oppositely characterized, they partake in nothing in common and thus could not compare with each other. There thus could not exist a hierarchy between them.

	ha šiiō, "sky, Sky God"
	sing	du	pl
nom	šiiō		ziiauuō
voc	ziiā
acc	ziiā̊		𝛿uwā̊
gen	𝛿uwō		𝛿ūwam
loc	ziiā		𝛿ūwšū
dat	𝛿uwō		𝛿ūmuš
ins	dī

hazərə "hand"

hazərə is an OX noun in r-stem with the inverted declension, with a zero-grade in the nominative singular. -ərə here is treated as a single, long syllable, representing Erani-Eracuran *-r̥. That -ərə is a consonantal is elucidated by the reflex of *s as -z- and not -h-, which would be regular had -ərə been a vowel sound. As with all nouns with inverted declension, the accusative is regarded as a weak case as to its appropriate stem, thus xšrə̄m over the expected *xšerəm; however, the syllable weight of the original is preserved in the long vowel of the ending.

	hā hazərə, "hand"
	sing	du	pl
nom	hazərə	xšerōi̯	xširiš
voc
acc	xšrə̄m		xšrā̊
gen	xšrōḫ	xšarōš	xšrõm
loc	xšrei̯	xšarō	xšərəšū
dat		xšərəmō	xšərəmuš
ins	xšroi̯

monuš "man, human"

monuš is the basic word for "human" in Northian, and it frequently appears as the principal to the adposition θxāmō "earthling". It defines the human being as "the thinking one" or "one who remembers" and is formally very similar to the perfect active participle ménuš "remembering, recalling", from móna "to remember". monuš defines mankind in opposition to beasts, and θxāmō in opposition to gods.

The word monuš has two competing etymologies: some stress its connection to the o-grade of the PEE root *men- "to think", thus characterizing the human being as a characteristically intelligent being, and others provide a more mundane connection to the very basic Nordic word mannuz "male, man". These etymologies are not necessarily incompatible. mannuz is obligatorily male in Elder Nordic and all its descendants, but monuš can be female in Northian, though the feminine form mnuuozərə < putative *mn-u-és-r̥ is more commonly seen with women. The feminine participle form of the verb mona is ménuuī.

monuš takes the oxytone declension in u-stem, gen. mŋʷuuōḫ, e.g. GNr. yōrō mnuuōḫ "a person's lifetime".

See also

• Northian language
• Northian grammar
  • Northian verbs
  • Northian adjectives