Derivational Morphology of gjâ-zym-byn

• Overview
• Derivation by compounding
• Derivation by suffixing
  • Suffixes that form nouns
  • Suffixes that form modifiers
  • Suffixes that form verbs
  • Postpositions
  • Conjunctions
  • Suffixes that leave the grammatical category of the stem unaltered
• Suffixes by the types of stem they can apply to
  • Suffixes that apply to numbers
  • Suffixes that apply to family relationship terms
  • Suffixes that apply to concrete entity stems
  • Suffixes that apply to abstract entity stems
  • Suffixes that apply to process stems
  • Suffixes that apply to quality stems
  • Suffixes that can apply to pronouns
  • Suffixes that can apply to modifier and quantifier clitics
  • Clitic that can apply to verbs
  • Suffixes that can apply to reflexive/reciprocal verbs
  • Suffixes that apply to any noun or modifier stem
• Incorporation of pronouns, determiners, and postpositions into compound words
• Numbers
  • Derivation from numbers
  • Counting

Overview

gjâ-zym-byn forms words by three processes: compounding, suffixing, and incorporation. Compounding involves putting nominal root words together to form new stems; suffixing involves adding suffixes to the end of a stem; and incorporation involves combining pronouns, conjunctions, postpositions, and other particles into stems.

In the discussion that follows, a "stem" means either a root word, or a word formed from one of the derivational processes, to which further derivational processes may then be applied.

Content root words in gjâ-zym-byn are, if used in a sentence by themselves without any derivational process being applied, always nouns; all verbs and most modifiers (adjectives/adverbs) are derived from these nominal roots, as are the abstract postpositions and many conjunctions. Root words in gzb which semantically correspond to root verbs or adjectives in English, for instance, tend to correspond syntactically to the "-ing" or "-ness" form of the closest equivalent English word; {mâ} or {râm} can be glossed by the English words "person" or "cat", but the most accurate gloss for {ðru} as a stand-alone word would be "redness" or "the color red", not "red" simply, and for {ruŋ} by itself the best gloss would be "moving" or "motion", not "to move". Suffixes turn those nominal roots into adjectives or verbs: {ðru-bô}, "red"; {ruŋ-zô}, "to move", "I go/come/move", etc.

Derivation by compounding

Compounding of nominal roots is normally head-morpheme-first, modifier-morphemes following.

However, there are a few root words that I call "suffixoid", which logically ought to be the head morpheme and come first, but which actually suffix. E.g.,

Why this irregularity? Early on I was undecided about the order in which morphemes compound, and had some inconsistency. Probably by August 1998 I decided definitively that the order should be head-first, and I fixed most of the compound words in the lexicon to fit this order. However, there were a few words I had already learned well that used these particular morphemes, and I decided I didn't want to lose ground with respect to learning vocabulary; so I made these morphemes "suffixoid". Suffixoid content morphemes are not an open class.

When stems are compounded with the conjunction {ke} incorporated between them, instead of simply put together without intermediary, the compound denotes a balanced, synergetic mixture of the things referred to by the two stems; for instance,

{pwě'lâl-kě'kul} would seem to be ice cream with cola or root beer flavor; {mâ-byj} might be a cowboy or cowherd, or a person who resembles a cow or bull in some unspecified way.

The most common use of {ke} incorporation is with proper names of spouses.

Derivation by suffixing

gjâ-zym-byn has 78 suffixes (as of May 2010), the vast majority of which, or perhaps all of which, are more accurately described as derivational rather than inflectional. Most of them are fairly productive derivational processes, applicable ad-hoc to any stem for whose basic meaning their meaning-transformation would make sense. In the following discussion I'll explicitly note the few suffixes that are not fully productive, whose use forms new words whose meanings are not predictable from the meanings of the stem and the suffix.

I'll classify the suffixes in two ways: (1) by the type of word they produce, and (2) by the type of stem they can be applied to.

In the suffix glosses which follow, a tilde '~' signifies the referent of the stem to which the suffix is applied.

Suffixes that form nouns

In addition, the abstract nominalizer clitic {tǒj} acts much like a suffix, and can be applied to postpositions, postpositional phrases, modifier and quantifier particles, and derived verbs and adjectives to turn them into abstract nouns.

tu-frâ-θaj	muw-i	jǒm-tǒj	tu-i	twâ-zô
AGT-ask-OPP1	subset-at	most.of-NMZ	AGT-at	say-V.ACT

hǒŋ	jyn-lym-fwa-sra	ŋĭn-i	kě'kul	kǒ	mĭ-i.
that	pleasure-taste-CAUS-CMP	CMT-at	cola	this	TOP-at
A majority of the respondents said that this cola was the tastiest.

All of the suffixoid roots also form nouns; some of them will be discussed further below in context with suffixes of similar or related meaning.

The first group of suffixes form words for kinds of person:

ħĭn-kô	ř-ŋ	câ-zô	ruŋ-kě'ĝu-zô	ħĭn-ta-zwa	srǒ.
confine-place	from-inside	try-V.ACT	go-secret-V.ACT	confinement-without-would.be	several
Several would-be escapees tried to sneak out of the prison.

mluj	gjâ-krĭ-za	i	krĭ-gjâ-pja	srǒ
small.convention	language-creation-ADJ	at	create-language-amateur	many

lĭw-o	kuln	jâ-o	hyw-ƥ-môj.
relationship-to	friend	state-to	direct.knowledge-3-V.RECP
At the Language Creation Conference conlangers meet one another and become friends.

italia-wam	i	ler-tǒj-lô	mĭ-i	fraňs-wam	i	dě'dâ-lô	₣um-i	fjǒ.
Italy-NAME.P	at	future-NMZ-member	TOP-at	France-NAME.P	at	Dada-member	similar-at	FUZZY
The Italian Futurists were somewhat similar to the French Dadaists.

These suffixes form terms for relations by marriage:

-mla	spouse of one's relative
-tôl	relative of one's spouse
fru-mla	son-in-law, daughter-in-law
tâ-mla	brother- or sister-in-law (sibling's spouse)
kyn-tôl	mother-in-law, father-in-law
tâ-tôl	brother- or sister-in-law (spouse's sibling)

See other examples under "Kinship terminology".

{-žar} indicates a process or event of becoming the kind of entity denoted by the stem, or beginning to have the quality or be in the state denoted by the stem.

But if the stem denotes an action or process, the ordinary root morpheme {gĭn} "beginning" is used to indicate inceptive aspect; compare these uses of {-žar} and {-gĭn} with the entity stem {rě'ĵy}, the quality stem {ĵu}, and the action stem {ruŋ}:

{-žar} was added to the language relatively late (March 2007) as morphological sugar for verbose {tǒj}-nominalizations of {jâ-o} postpositional phrases.

The nominal ordinal suffixes {-gla} and {-ŋla} are discussed in detail in the section on number-derivations, along with the adjectival/adverbial ordinal suffixes {-pa} and {-saw}. I classify them as nominal suffixes because the resulting ordinals can be used as the head of a noun phrase, though they also tend (more often than most other nouns) to act as apposite noun modifiers. E.g.,

ĝu-gla	i	tyn	kǒ	o	ruŋ-ƥ-zô.
thirteen-ORD.T	at	place	this	to	move-3-V.ACT
She arrived at one o'clock p.m.

ĉul	měr'kun-dal	ħy-gla	kâ-i	kâ-ku-ɱ-zô	de.
performance	radio.wave-origin	nineteen-ORD.T	ATT-at	attention-hear-3-V.ACT	HAB
He listens to [a certain] seven p.m. radio program nowadays.

Foreign names and titles not completely assimilated to gzb phonotactics are marked with a suffix indicating the kind of entity named:

The distinct name suffixes allow me to unambiguously put names in the normal order for their native language, i.e. personal name + family name for English, family name + personal name for Hungarian, etc. Mostly these apply to foreign names and titles; however, they could apply to gjâ-zym-byn words which are used as names or titles, e.g. if a foreign name or title is translated rather than merely transliterated. For instance,

?juda-tam	ʝel	θĭ-o	zǒn	ce.
Judah-NAME.E	in.general	help-to	Q.YN	this
Is this good for the Jews?

If {-šam} follows a family name that ends in a fricative, then an epenthetic schwa is inserted in pronunciation (though not in writing).

smĭθ-šam

/'smIθ.ə.çɑm/

If a person has a pen-name or pseudonym, or if they have legally changed their name to a form that has separate parts corresponding to the personal name and family name of a traditional multi-part name, but in which the last name doesn't have a connection to their ancestors (as far as the speaker knows), then the entire name is transliterated with hyphens between its parts, and the {-ram} suffix is appended to the whole thing; the {-šam} family name suffix is not used in such cases. The not uncommon method of forming a pen-name incorporating one's mother's maiden name is an exception; {-šam} could be used with the family name in that case.

Pen name:

mark-twejn-ram	dâm-ř	{hěklberij-ram	fĭn-šam}	kâ-i	lju-zô.
Mark Twain-NAME	authorship-from	Huckleberry-NAME	Finn-NAME.F	ATT-at	read-V.ACT
I'm reading Huckleberry Finn by Mark Twain.

Legally changed name:

źî'gĭ-ĉu-gla	i	sonja-elen-kisa-ram	dâm-ř	toki-pona-lam-la	kâ-i	suŋ-hôw-ca.
2003-2-ORD.T	at	Sonja Elen Kisa-NAME	authorship-from	Toki Pona-NAME.L	ATT-at	know.how-CAUS2-V.REFL
I learned [taught myself] Sonja Elen Kisa's Toki Pona in 2005.

Names of spouses, or collaborators in some artistic work, are often linked together with the synergy-conjunction {ke}, the name-suffix {-ram} being used only once after both names.

Names are sometimes used as apposite modifiers of common nouns, especially if a {-ram} name is used for a nonhuman, and with the generic {-gam} names for brand-name foodstuffs, drugs, vehicles, etc. This rarely if ever occurs with names marked with the more specific suffixes like {-wam} and {-lam}.

pâ-sô-van	râm	merĭn-ram.
restless-tending-V.STATE	cat	Meryn-NAME
Meryn (a cat) is restless.

plĭ	asetamĭnofen-gam	ŝĭw-i	jyn-cô-tâlm	jâ-ř.
pill	acetaminophen-NAME.G	substance-at	pleasure-OPP2-head	state-from
I got rid of my headache using acetaminophen (a pill).

When a foreign name is cited without being transliterated (I may do this if I am unsure of its pronunciation), it goes into quote marks (curly brackets); a name-suffix then follows the quoted name. Sometimes foreign names which are acronyms in the source-language are quoted as all-caps Latin letters, in which case quote marks are not needed because the capital letters thoroughly distinguish them from surrounding gzb text; the name-suffix is still suffixed (with a hyphen).

krĭ-gjâ-pja-ķa	{Javant	Biarujia}-ram	dâm-ř	grâm-cjaj	gǒ.
create-language-amateur-RESPECT	"Javant	Biarujia"-NAME	authorship-from	message-SPEC	behold
Hey, an email from noted conlanger Javant Biarujia!

IHOP-gam	o	ruŋ-zô	Ќ-ƥ.	kujm-šar	vâ-oŋ-zô
IHOP-NAME.G	to	go-V.ACT	1-3	purpose-CONJ	digestion-into-V.ACT
We went to eat at IHOP.

Some few foreign names and titles have fully-assimilated forms that conform fully to gzb phonotactics and morpheme-shape rules. These forms do not require a name suffix when they occur; they occur in the lexicon and block other gzb words from being coined with the same form. E.g.,

If a native gzb word (root or derived) is used as a proper name but has a basic meaning other than as a name, whether it's translating the meaning of a foreign name or originally naming some entity in gzb, the appropriate name suffix must be used.

{-baw} derives a word for a type of phoneme from an example word that contains two or more phonemes that belong to that class. It's idiomatic and not fully productive; given gzb's morphophonology, I haven't been able come up with satisfactory {-baw}-derivations for various kinds of vowel. The stable words derived from {-baw} are as follows:

{-ĉa} forms words for tools, including vehicles, software and other useful things. With process stems, it means a tool for doing so and so; with quality stems, a tool for rendering something such and such:

{-ĉa} is normally used on a nominal stem, but in one case at least it suffixes after a verb suffix:

Contrast {rĭm-ĉa}, "eyeglasses", formed on the bare root {rĭm}.

Of the substance suffixes, {-Φa}, {-kar}, and {-ha} are semantically regular and productive; {-ŋô} is idiomatic and not fully productive. {-Φa} and {-kar} apply only to process stems; {-ha}, to either quality or process stems:

There is no predictable rule for what substance will be selected as the "primary ingredient" of a stem noun by {-ŋô}.

Two suffixes form words for places — typically buildings or rooms, less often outdoor spaces. {-ĵwa} describes a place having many of the thing referred to by the stem; {-kô}, a place where one typically does the process or action referred to by the stem.

Sometimes two nouns formed from {-kô} and {-ĵwa} are used together, one modifying the other, when it seems necessary to reduce ambiguity:

Sometimes {-kô} and {-ĵwa} are used together to derive another place-noun from a more basic one:

{-kô} can also be used with a reflexive verb stem:

I think potentially some other noun-deriving suffixes (maybe {-pja} and {-ha} for instance) could suffix to reflexive verb stems, but this and {rĭm-ca-ĉa} are the only such words attested so far.

The suffixes {-daj} and {-zla} form collective nouns, as do the suffixoid roots {cu} and {kwĭ}:

{-dô} forms a word for a violation of a standard described by the stem. Compare the modifier-forming suffix {-ja} "fitting, in accordance with" discussed below. {-dô} words are basically nouns, but can also act as modifiers of other nouns, or of verbs.

ĝâ-dô	kǒ	mĭ-i	ĥul-gô	ŋĭn-i.
law-violation	this	TOP-at	anger-worthy	CMT-at
This crime is something one ought to be angry about.

vâ-oŋ-tǒj	fĭm-dô	gân-ř	fĭm-cô	jâ-o	ķuj-ť-van.
digestion-into-NMZ	health-violation	cause-from	health-OPP2	state-to	danger-2-V.STATE
Because of unhealthy eating, you are in danger of getting sick.

Ќ	dâm-ř	gju	kâ-i	ku-ƥ-van	kâ-dô.
1	authorship-from	speech	ATT-at	hear-3-V.STATE	attention-violation
She heard inattentively what I was saying.

Above, {ĝâ-dô} is head of a noun phrase; {fĭm-dô} modifies another noun (here a nominalized postpositional phrase) within a noun phrase; {kâ-dô} acts as an adverb, modifying the verb of the sentence.

The suffixes {-hô} and {ĵam} are a bit tricky, and admittedly less frequently useful than most of the other suffixes in gzb. {-hô} derives a word for the qualities a particular kind of thing can have. For instance, {mâ-hô} refers to all the qualities that people can have, or (in a linguistics context) all the modifiers that could appear modifying a noun phrase signifying a person. E.g., intelligent, foolish, pious, angry, female, healthy, etc. {gâ-zuň-hô} refers to the qualities (or modifiers that refer to qualities) that living things in general can have.

Nouns derived from adjectives with {-ĵam} refer to the set of things that can have the quality described by the stem (or nouns that can be modified by such adjectives, in a linguistic context) — those things for which having that quality would not be a meaningless notion or a contradiction in terms, whether they actually have much of those qualities or not. For instance, {ĉâ-ĵam} refers to the set of things which might be described as intelligent — all thinking beings, human or otherwise — even if some particular ones would be more aptly described as stupid ({ĉâ-cô-bô}). From {Φu} "mass" we derive {Φu-ĵam}, which describes all physical things (even those such as photons which actually have zero rest mass), in contrast to information, abstractions and spiritual things for which the concept of "mass" is meaningless.

Suffixes that form modifiers

{-bô} is the basic adjective suffix, used primarily for nominal stems that already signify a quality, state, or description.

It's also used with number stems, to derive cardinal number modifiers from mathematical-object nouns, and with process/action stems, to form gerunds.

{-za} "pertaining to, associated with" is the most general and productive adjectivizing suffix, applicable to practically any stem, though most commonly used with entity or substance stems. In recent years the more specific suffixes {-paj} and {-dal} (see below) have replaced some uses of {-za}.

There is at least one very idiomatic use of -za, however:

{-tan} "like, resembling" is similarly productive as {-za}, applying to almost any kind of stem but especially to entity/substance stems.

{-rô} is explicitly idiomatic in its usage; it selects an arbitrary salient quality of the entity or substance referred to by its stem.

Most words for tastes are derived with {-rô} from a stem denoting a kind of food or drink:

The exception is "salty" from {sâl} "salt" as {sâl-tan}.

gzb has two kinds of causative, {-fwa} "effective causation" and {-hôw} "attempted/intended causation". Words formed from these suffixes are modifiers, and it's very common to form verbs based on stems formed with them, as well as {-tla} or {-pja} agent-nominalizations. Words in {-hôw} do not make any assertion whether the attempted causation is successful or not.

Their effect with different kinds of stems:

E.g.,

{-fwa} is most often used with the mindstate words to form words for subjective qualities, as in {pym} → {pym-fwa} above.

{-ta}, {-ža} and {-da} are used to form modifiers indicating the absence or presence of an entity or substance, usually; they are more rarely used with quality or process stems ({-ta} more often than the others):

{-ja} "according to, fitting" is a counterpart of the noun-deriving suffix {-dô} "violation of"; it applies productively to stems signifying a standard by which actions can be judged, and more idiomatically elsewhere.

With entity stems signifying long objects {-ja} means "along"; if the referent of the stem has a direction of motion, it means "in said direction", while {-dô} means "opposite that direction".

pě'hĭ	road, street
pě'hĭ-ja	along the road
pě'hĭ-dô	the wrong way on a one-way street
sĭ	river
sĭ-ja	along the river, downriver
sĭ-dô	upriver
by-flu	wind (air-flow)
by-flu-ja	downwind
by-flu-dô	upwind

{-faj} and {-gô} correspond to the English suffix -able, -ible:

{-kwa} derives additional color terms from entities which typically have that color (there are also four basic root-word color terms).

Also:

{-na} forms modifiers signifying their head is made the stuff referred to by the stem.

{-paj} and {-dal} were added relatively late (in 2007), to substitute more precisely for vague, polysemous {-za} in some uses. {-paj} is morphological sugar for the benefactive and purposive case postpositions; depending on the meaning of the stem it applies to it can mean "for the benefit of" or "for the purpose of":

ƥ	ŝâj-o	θym	zjâm-paj	ĥy-i	bwĭl-zô.
3	having-to	torus	finger-for	PAT-at	give-V.ACT
I gave her a finger-ring.

ķĭm-zô	sěl'kâ-gôm-paj
exercise-V.ACT	spine-METONYM-for
I'm doing back exercises.

{-dal} creates modifiers signifying origin from the entity referred to by the stem. I created it while trying to find ways to express various senses of polysemous English "natural" in gzb:

ħun-tôn-daj	iŋ	luw-cu	šĭl-dal	kâ-i	tru-zô.
pine-GNR-COLL	inside	bone-system	snake-origin	ATT-at	find-V.ACT
I found a snake skeleton in the woods.

The suffixes {-ĝa} and {-ðwa} correspond roughly to English "pro-" and "anti-".

{-ðwa} can be used with a person's name to form an adjective describing those who support them, e.g., believe them innocent when they are accused of something, or {-ĝa} to form an adjective describing those who oppose them or believe them guilty:

tĭm-fî'suň-bly	ħy-pa	šiŋ,	fraňs-wam	iŋ
hundred-Earth-orbit	19-ORD	after-at-inside	France-NAME.P	inside

mâ	draj'fîs-šam-ðwa	pe	draj'fîs-šam-ĝa	tu-i	sî'ðyr-ga-źa-môj.
people	Dreyfus-NAME.F-pro	and	Dreyfus-NAME.F-anti	AGT-at	fight-METAPH-AUG-V.RECP
In the late nineteenth century in France, Dreyfusards and anti-Dreyfusards fought intensely.

The suffixes {-sô} and {-ŝra} are prototypically used with process/action stems, forming modifiers meaning "tending to do ~ often" or "likely to do ~ soon". With quality or state stems, their meaning is similar: "tending to have this quality often", "likely to enter this state soon".

With entity or substance stems, {-sô} modifiers tend to mean "having a partiality toward ~", "preferring ~ to some implicit alternative":

{-ŝra} with such stems means "likely to become ~ soon" (cf. {-zwa} among the nominalizing suffixes above):

{-ŝa} applies mostly to entity/substance stems, signifying "being a fruitful source of ~, producing many ~":

It's unattested with quality or process stems; it seems the resulting modifiers would be near-synonymous with {-fwa}-causatives. Maybe:

bî'lym	sî'ðyr-ŝa	rjâ-i	šî'fy-źa-ŝy	dâ-bô.
apple	fighting-productive	quest-at	spirit-AUG-female	three-ADJ
Three goddesses sought the Apple of Discord.

The ordinal suffixes {-pa} and {-saw} are discussed in the section on derivation from numbers.

Depending on the semantics of the stem {-pôm} is applied to, it can derive evidentiality, validationality, or attitudinality adverbs; in any case the syntax of said adverbs is the same.

With stems that indicate a potential source of information — a sense, a communication process, a person, a book, etc. — it forms evidentiality adverbs:

With stems that indicate a degree or kind of certainty, it forms validationality adverbs:

With other mindstate stems, emotional rather than intellectual, it forms attitudinality adverbs:

These adverbs, unlike most (I think {-saw} ordinals are the only other exception), apply at the sentence level, and can either follow the main verb or come at the beginning of the sentence. For more on the syntax of these adverbs and example sentences for their use, see the relevant section of the syntax document.

Suffixes that form verbs

The use of verbs derived with these suffixes is discussed in the syntax document.

With process/action stems, verb derivations work as follows:

See numerous examples in the syntax document.

With quality/state stems:

The causative suffix {-fwa} is not always needed in going from a quality stem to a causative verb, though it's often used redundantly.

{šyj-fwa-zô} is also attested; it can be synonymous with {šyj-zô}, but it has another sense where one is causing e.g. clothes or dishes to be washed by a machine rather than washing them directly by hand.

With entity/substance stems, {-fwa} is needed to explicitly form a causative active verb meaning "to make something into an entity of this kind". Verb derivations from entity/substance stems are:

For instance,

Postpositions

Derived postpositions are formed by suffixing one of the core spacetime postpositions — {i}, {o}, or {ř} — to a stem, usually a bare root word. Such postpositions are discussed in the syntax document.

Conjunctions

Exactly one suffix forms conjunctions, {-šar}. For example:

In retrospect, this aspect of gzb does not seem especially well designed. I should probably have had multiple conjunction-forming suffixes for making phrase-scope and clause-scope conjunctions, and perhaps different ones for applying the meaning of the stem leftward or rightward. As it is, there's just {-šar}, and you can't deduce the usage of a {-šar} conjunction just from the meaning of the stem (although you probably can from the meaning of the stem plus the context the first two or three times you see it used).

For instance, seeing the conjunction {mĭ-šar} and knowing the meaning of the root {mĭ}, you can figure out that the words on one side of it are somehow the topic or theme of the words on the other side; but you have to memorize the fact that it introduces a subordinate clause to its right, which is a topic of (typically) a verb of thinking or communciation in the matrix clause to its left. E.g.,

gju-môj	θě'mâ	pe	sě'râ	mĭ-šar	tě'θru-ga-van	mâ-cu	kâj-ha-sô.
talk-V.RECP	Thomas	and	Sarah	topic-CONJ	fall-METAPH-V.STATE	person-system	exchange-stuff-seeking
Thomas and Sarah talked about how the company was going bankrupt.

Ќ	ŋâw-o	grâm-zô	âleks-ram	fiŋk-šam	mĭ-šar	Ќ	dâm-ř
1	call-to	message-V.ACT	Alex-NAME	Fink-NAME.F	topic-CONJ	1	authorship-from

pě'pâ-ga	sǒ	im	ðâ-dô	ĉu-bô.
page-METAPH	certain	part.of	logic-violation	two-ADJ
Alex Fink sent me a message about how there were two mistakes in a certain essay of mine.

The manner-conjunction {ru-šar}, introducing a subordinate clause describing the way the action of the main clause is done, works similarly:

gjâ

kâ-i

suŋ

jâ-o

Ł

tu-i

ru-šar

te

ĥy-i

syj-zô

reŋ

kiň

gjâ-dô-zô

reŋ.

language

ATT-at

know.how

state-to

3.GEN

AGT-at

manner-CONJ

3.INAN

PAT-at

use-V.ACT

much

and

language-violation-V.ACT

much

One learns a language by using it a lot and making a lot of mistakes.

{muw-šar}, the subset conjunction, however, doesn't make sense as a clausal conjunction; it's used between nouns within a noun phrase, and works pretty much like the subset postposition {muw-i} in reverse (though it's much rarer in the corpus than {muw-i}):

rî'mâ	vyŋ-lwa-bô	muw-i	rî'mâ	ny-sra-lwa-bô	iŋ	ty-van	zě'ĥâr.
house	ten.thousand-APPROX-ADJ	subset-at	house	small-COMP-ADJ	inside	lodge-V.STATE	Zachary
In the smallest of the ten thousand houses lived Zachary.

rî'mâ-ŋul	gem	muw-šar	rî'mâ-tôn	kâ-i	prym-van	zě'ĥâr.
house-protect	especially	subset-CONJ	house-GNR	ATT-at	appreciation.of.beauty.V.STATE	Zachary
Among types of building, Zachary especially loves castles.

The time-conjunctions formed with {-šar} are pretty straightforward; they correspond to English "then" when used as a conjunction, i.e., the clause to their right happens after the clause to their left:

ty	o	ruŋ-zô	nu-šar	vâ-oŋ-zô.
home	to	go-V.ACT	moment-CONJ	digestion-into-V.ACT
I went home, then-immediately ate. = The moment I got home, I ate.

ty	o	ruŋ-zô	vĭj-šar	vâ-oŋ-zô.
home	to	go-V.ACT	period-CONJ	digestion-into-V.ACT
I went home, then-eventually ate. = A while after I got home, I ate.

ty	o	ruŋ-zô	kiň	vâ-oŋ-zô,	sun-šar	tĭw-mwĭl	son	zyŋ-zô.
home	to	go-V.ACT	and	digestion-into-V.ACT	end-CONJ	chair-sleep	onto	crawl-V.ACT
I went home and ate, then-finally went to bed.

Most if not all of the clausal conjunctions can occur at the beginning of a sentence, implicitly linking it to a previous sentence or some element thereof. The time-conjunctions {nu-šar}, {vĭj-šar} and {sun-šar} work that way — {sun-šar} probably occurs more often at the beginning of a sentence than in the middle of a sentence. The example-conjunction {žy-šar} is also likely to start a sentence, but can occur in the middle:

grĭ-cô-sô-van	de	šĭm-ĉa	kǒ;	žy-šar
function-OPP2-tending-V.STATE	HAB	algorithm-tool	DEM1	point-CONJ

ĵwy-cô-van	sĭŋ-flu-kô	o	lĭn	mĭ-i.
fast-OPP2-V.STATE	information-flow-place	to	link	TOP-at
This computer is malfunctioning lately; for instance, its network connection is slow.

Contrast this clausal {žy-šar}, introducing a subordinate clause which is an example of what the main clause is talking about, with the postposition {žy-i}, which marks a simple noun phrase as an example of the sentence's topic.

gjâ-krĭ

reŋ

im

tyn-van

fĭ-θy

nâ-cô-bô

mĭ-i,

g,z,b

im

{θ,

ƥ,

ř}

žy-i.

language-create

many

part.of

place-V.STATE

syllable-element

common-OPP2-ADJ

TOP-at

gzb

part.of

θ

ʘ

ɹ̣

point-at

Many conlangs have rare [among natlangs] phonemes, for instance gzb's /θ ʘ ɹ̣/.

Suffixes that leave the grammatical category of the stem unaltered

The use of the comparative suffixes {-sra} "more/most" and {-ĵar} "less/least" with quality and process stems is discussed in comparison section of the syntax document. Here I'll just note that when they apply to number or unit of measure stems, they mean more than or less than the amount specified by the stem:

The augmentative and diminutive suffixes {-źa} and {-θô} are fully productive and their meaning is transparent with quality stems and most process stems. With entity stems, and with some process stems, their use is idiomatic and nonproductive. For instance,

but:

With entity roots, the augmentative and diminutive suffixes don't necessarily mean "large" or "small" (the quality roots {ĝĭ} and {ny} can be compounded onto stems for those meanings), but take a salient aspect of the word's meaning and augment or diminish it. With unit of measure stems, they are metric multipliers kilo- and milli- respectively:

The "meta" suffix {-ma} is somewhat idiomatic; its use with kinship terms is regular and predictable, but in other areas it's less so. In general it refers to recursive or indirect patterns and relationships. It's often used as a lexical dissimilation method where otherwise a suffix (or suffixoid root) would be applied twice in a row, e.g. with a system of systems or an indirect causative:

Some other uses of {-ma}:

The suffix {-ʝa} take a stem denoting a dimension or a physical object with a particular orientation, and derives a word for another dimension or object with orientation rotated 90 degrees around an axis perpendicular to the direction of gravity. That is, it takes words for horizontal things and derives words for similar vertical things, or vice versa.

The suffixes {-tôn} and {-cjaj} are idiomatic and not productive; using them involves a conscious decision to expand the lexicon. The design of gzb in this area is based on the idea that things that one talks about more often should have shorter names. For some semantic categories, the specifics are talked about more often; for others, the genera. For instance,

If one lives with a cat, or has friends who do, one talks/thinks more about cats than about animals in general. On the other hand, one often talks about sending a message to someone; less often one has to specifically state that one sent an email. (This was the theory I was going by when I coined those words back in 1998. In practice, I find that in my journal I use {grâm-cjaj} more often than simple {grâm}, mostly because of the idiomatic sense its verb form has developed: {grâm-zô}, to send a message (probably an email) to someone; {grâm-cjaj-zô}, to read and perhaps reply to various emails.)

I suspect that idiomatic compounds such as these are easier to learn than completely separate, unrelated root words. I'm not sure if I can verify this suspicion without a properly controlled experiment, though.

The repetitive/intermittent suffix {-ra} is discussed and some examples given in the aspect section of the syntax document. It's normally used with process/action stems, of course, but it has at least one use with an entity stem:

The "approximately" suffix {-lwa} is mostly used with number stems; see the section on derivation from numbers for examples. It's also somewhat interchangeable with the diminutive suffix {-θô} with quality stems, e.g.:

The fuzzy-logic clitic {fjǒ} has a similar meaning.

The opposite-derivation suffixes are {-θaj}, {-cô}, and {-fja}. {-θaj} derives words signifying the complement of a pair of which the stem denotes the other member; for instance, members of asymmetrical relationships, or complementary actions (where one action typically if not necessarily implies another):

{-cô} derives opposites for quality-terms, and reversives of action or process stems.

{-fja} signifies a minimal degree of the quality denoted by the stem. It can also be used with entity stems, deriving modifiers meaning "having a small but nonzero amount of ~"; in this sense it's in contrast with {-ta}, {-ža}, and {-da}.

The metaphorical use suffix {-ga} is productive (you can use it to explicitly mark any metaphorical use of a word), but with certain words it has a conventional, defined meaning.

Similar to but distinct from {-ga} is the metonymic suffix {-gôm}:

{-gôm} is also used to derive a word for an author's work (their sole or best-known work, or a contextually salient one) from their name.

herǒdǒtǒs-ram-gôm	kâ-i	lju-zô	jǒj	de.
Herodotus-NAME-METONYM	ATT-at	read-V.ACT	again	HAB
I'm reading Herodotus [i.e. his Histories] again lately.

One kind of metonymy typical in natural languages also occurs in gzb, unmarked: using the same word (or zero-derivation) to refer to a plant as a whole or to its edible part, in agricultural and culinary contexts respectively:

šun	pǒ	i	bâwŋ-van	θĭl.
region	DEM3	at	grow-V.STATE	potato
Potatoes grow there.

šun	pǒ	i	θĭl	ĥy-i	vâ-oŋ-Ł-zô	ƴyr-bô.
region	DEM3	at	potato	PAT-at	digestion-into-3.GEN-V.ACT	custom-ADJ
People eat potatoes there.

The attitudinal suffixes typically mark the speaker's attitude toward the entity (or sometimes the process or quality) denoted by the stem. (In fiction with a closely-focused viewpoint, they may mark the viewpoint character's attitude, not necessarily the writer's.) They're most commonly used with proper names, especially the first time a person or other named entity is mentioned in a discourse, and fairly often with second and third-person pronouns (sometimes even with the first-person pronoun). They bind more weakly than other suffixes except the verb endings, so they would follow any other suffixes applied.

If a person is mentioned by their full name, there is a nuance to whether the attitudinal suffix is applied to their personal name or their family name. The latter suggests that the speaker feels a certain way toward their family as a whole; the former, that they feel thus toward this individual specifically. If a person is mentioned only by their family name, however, the application of an attitudinal suffix doesn't necessarily imply said attitude toward other family members who are not relevant to the discussion.

gym-tla-ķa
honorable leader

rě'ĵy-la
my dear wife

źy-ba
a dream both good and bad

mâ-ħa
scoundrel, rascal

θuň-ŋa
a surprising story

The attitudinal suffixes are fairly commonly used with pronouns, especially in the third person:

tyn	kǒ	o	ruŋ-zô	ƥ-la.
place	this	to	go-V.ACT	3-affectionate
The dear one is coming hither.

re-ħa	i	te-ŋa	kâ-i	tru-zô.
there-contempt	at	3.INAN-surprise	ATT-at	find-V.ACT
I found it (the surprising thing) there (in that horrible place).

They can also be used in verb and modifier stems:

mâ	ĵlân-ķa-bô	ŋâw-o	ce	mĭ-i	frâ-zô.
person	wise-respectful-ADJ	call-to	that	TOP-at	ask-V.ACT
I asked a wise person about that.

₣âl-ŋa-van	fu-ĉa.
sudden.change-surprise-V.STATE	light-tool
The light unexpectedly changed.

Finally, the attitudinal suffixes can occur on their own as interjections:

*la.
Yay!

*ķa.
Wow!

*ba.
Huh.

*ħa.
Gah! Yuck!

*ŋa.
WTF?

The speaker's (or, in a story, the viewpoint character's) attitude toward a situation as a whole, rather than one of its elements, can be denoted more precisely by a sentence-scope attitudinal adverb derived from a mindstate stem with the suffix {-pôm}. See the discussion in the syntax document for examples.

Suffixes by the types of stem they can apply to

Here I'll just list the suffixes in each category, since I've already treated the use of each in some detail in the sections above. To avoid multiplying combinatorial categories indefinitely, I'm listing some suffixes in multiple categories.

These categories refer mostly to the types of stems for which suffixes form productive derivations; some may have improductive, idiomatic uses with stems in other categories.

Suffixes that apply to numbers

See the discussion on derivation of numbers below.

Suffixes that apply to family relationship terms

See the section on kinship terms in the semantics document.

Suffixes that apply to concrete entity stems

Suffixes that apply to abstract entity stems

Suffixes that apply to process stems

Suffixes that apply to quality stems

Suffixes that can apply to pronouns

ƥ-la	ĥy-i	trâw-zô	ƥ-ħa.
3-ATD1	PAT-at	hit-V.ACT	3-ATD4
The previously mentioned person whom I don't like hit the previously mentioned person whom I like.

Here, the attitudinal affixes {-la} and {-ħa} disambiguate the compound pronouns, where simple {ƥ} would be ambiguous. This is an alternative to various other ways of disambiguating the referents: by using a full name or noun phrase instead of a pronoun for one of the referents, or by using other modifiers to make different compound pronouns, or by using pronouns with various modifier adjectives following them (depending on whatever attributes of the persons mentioned are most useful to distinguish them).

ƥ-pôm	lju-gô	ŋĭn-i	twâ-cu	pǒ.
3-EVD	read-worthy	CMT-at	sentence-system	DEM3
According to her, that book is worth reading.

{-pôm} following any personal pronoun, as when it follows any noun referring to a person, makes an adverb meaning "according to so-and-so", that is, the person who's the referent of the pronoun is the source of the information on which the rest of the sentence is based.

ť-ŋa	kâ-i	rĭm-van	wlâ-bô.
2-ATD5	ATT-at	see-V.STATE	shock-ADJ
I am surprised to see you.

The {-ŋa} attitudinal suffix emphatically reinforces the surprise expressed more explicitly by the adverb {wlâ-bô}. You could delete either and the sentence would mean pretty much the same thing, but be less emphatic.

gâ	Ќ-za	ĥy-i	byn-ť-zô	źǒ.
thing	1-pertaining	PAT-at	tinker-2-V.ACT	IMP.NEG
Don't mess with my stuff.

Here, postpositive {Ќ-za} is something like informal syntactic sugar for a heavier genitive postpositional phrase like {Ќ wuŋ-i} or {Ќ ŝâj-i} that would precede the noun {gâ}.

Suffixes that can apply to modifier and quantifier clitics

These apply to quantifier clitics, but their use with them is rare:

For instance:

The attitudinal suffixes can apply to many quantifier and qualifier clitics:

mwĭl-van	ķe-ŋa.
sleep-V.STATE	too.much-ATD5
Whoa, I've overslept!

ĝyl-van	le-ba	gâm-ʝĭl-kwĭ	pǒ.
abort-V.STATE	probably-ATD3	picture-motion-series	DEM3
That TV show will probably be canceled, a fact about which I'm ambivalent.

And to interjections:

Clitic that can apply to verbs

Root words are nouns to begin with; the roots of most verbs are nouns referring to a process or action in the abstract. To get the nominalization of, for instance, {ruŋ-zô} "to come/go", just remove the {-zô} verb suffix to get {ruŋ} "motion, coming, going". However, when a verb suffix is added to a noun which in its basic meaning refers to something other than a process or action, you can add the nominalizer clitic {tǒj} to the derived verb to get a noun meaning something different than the root noun. For instance:

Suffixes that can apply to reflexive/reciprocal verbs

See the earlier discussions of {rĭm-ca-ĉa} and {šyj-ca-kô}.

Suffixes that apply to any noun or modifier stem

Incorporation of pronouns, determiners, and postpositions into compound words

Incorporation of pronouns into verbs is treated in the syntax document. Pronouns can also be incorporated into compound nouns and modifiers, however: in the simplest case, suffixing a pronoun to a noun stem, they mark a general, vague kind of genitive, hypernymic to the various highly specific genitive case tags — for instance,

They can also be incorporated into modifiers, most often with {-ja} "according to" and {-dô} "violation of":

Certain determiners and other modifier particles can also be incorporated into compound words; for instance:

The conjunctions {ke} and {pe} can sometimes incorporate into compound nouns. The conjunctions {se}, {ðe} and {me} incorporate only into number-words.

Numbers

Root words for numbers are nouns; {fy} refers to the mathematical object "seven", the set of all sets with seven members. By extension, it can refer to any specific set with seven members; so {ĉu} for instance could mean "the number two" or "a pair/duet/duo" depending on context. Number-modifiers, used to express how many of something there are, are formed with the adjectival suffix {-bô}; number-verbs can be formed with the stative verb suffix {-van}. There are four different ordinal suffixes, discussed further on.

Originally, I had root words only for some prime numbers (all of those up to 113 and a few larger ones for recent and future Gregorian years, e.g. 1973 and 1999) and other interesting constants (e.g. 0, 1, pi); all composites were expressed by compound words or phrases. After about four years, finding that this system was fun but not very practical, I added root words for powers of 10 and of 16. There are multiple ways to express other composite numbers — with math conjunctions, or simply by compounding the core number morphemes together. If number-morphemes are compounded in lesser to greater order, they multiply. If in greater to lesser order, they add.

Base-ten numeration:

Powers of 10:

Powers of 16:

Hexadecimal numbers:

0xFFFF = dâ-ðy-ĝyŋ-vâl-dâ-ðy-vâl-dâ-ðy-ĝyŋ-dâ-ðy

Words for non-integral constants:

Derivation from numbers

In compounds of numbers with non-number substantive roots, the number can be the head or the modifier, depending on emphasis; in general the number coming first as the head indicates a stronger connection between the entities referred to by the modifier; and either kind of compound indicates a stronger connection than if the root noun were followed by a separate number-modfier. E.g., contrast:

The first example could be glossed as "a septet consisting of days"; the second, "day/days characterized by seven-ness".

Numbers modifying pronouns always compound with them, and follow the pronoun-head:

Fractions are formed with the conjunction {ðe} (divided by). If the initial number is omitted, {cĭ} (one) is assumed.

Before {se} (minus), the default number is {bâ} (zero).
Before {me} (raised to the power of) the default number is {ĉu} (two).

Precedence of the math conjunctions within compounded number-words is about what you'd expect; concatenation (which can mean either multiplication or addition depending on the relative magnitude of the number-roots concatenated, remember) has the highest precedence, followed by:

I.e., a construction such as {se-ĉu-me-gâr} means "minus (two to the tenth power)" = -1024, not "(minus two) to the tenth power" = 1024.

Numbers can form quantifying adjectives or stative verbs with the basic adjectival suffix {-bô} or the verb suffix {-van}:

byn-pja	ðy-bô	o	grâm-zô.
hack-amateur	five-ADJ	to	message-V.ACT
I sent a message to the five hackers.

 

Ќ	mĭ-i	cĭ-van	žǒŋ.
1	TOP-at	one-V.STATE	merely
There's only one of me.

{-bô} number-adjectives used after verbs signify "to do said action N times". The same is more or less true of quantifier clitics, though they tend to be more ambiguous (e.g., {reŋ} can signify "many" or "much" depending on the kind of noun it applies to, so with verbs it's ambiguous between "do several times" and "do for a long time").

Ќ	ĥy-i	ƴum-Ł-zô	cĭ-bô	še,	mǒj	ĉu-bô	źǒ.
1	PAT-at	defraud-3.GEN-V.ACT	one-ADJ	maybe	but	two-ADJ	IMP.NEG
One might fool me once, but not twice.

gâm-ʝĭl-ba	kâ-i	kâ-rĭm-zô	cě'ku-bô	tǒlm.
picture-motion-ATD3	ATT-at	attention-seeing-V.ACT	aleph.null-ADJ	HYPERBOLE
I've seen that movie an infinite number of times.

Basic ordinals are derived with {-pa}.

etc.

twâ-cu-kwĭ	kǒ	im	twâ-cu	cĭ-pa
sentence-system-sequence	DEM1	part.of	sentence-system	one-ORD

zen	kâ-i	lju-zô	gwe.
only	ATT-at	read-V.ACT	already
I've only read the first book of this series.

{-pa} after a verb signifies "to do said action for the Nth time" (ever, or more likely within a salient time-context such as a day). In this context it contrasts with the ordinal suffix {-saw}, "as the Nth in a series of heterogeneous but perhaps related actions". E.g.,

ty	o	ruŋ	ši,	mrân-zô	dâ-pa.
home	to	going	after	eat-V.ACT	three-ORD
After going home, I ate for the third time [that day].

ty	o	ruŋ	ši,	mrân-zô	cĭ-saw.
home	to	going	after	eat-V.ACT	one-ORD2
After going home, I ate first [= the first thing I did was eat].

twâ-cu	ĥy-i	kě'ĝu-zô	cĭ-saw.
sentence-system	PAT-at	hide-V.ACT	one-ORD
I hid the book first.

{-saw} ordinals are not used to modify nouns, as far as I know.

Time-period ordinals are formed with {-gla}. What time-period such a word refers to depends on context.

!hâ-gla	i	Ќ	o	ruŋ-zô	mwe	ť	tu-i.
seventeen-ORD.T	at	1	to	go-V.ACT	IMP	2	AGT-at
Please come at seventeen (= 5pm); or, Please come on the seventeenth (of this month)

Dates are written as "year{-gla} month{-gla} day-of-month{-gla}". E.g.

źî'fu-gla	ðy-gla	ju-gla
1999	May	29

That day at 7:37 AM:

źî'fu-gla	ðy-gla	ju-gla	fy-gla	ky-gla
1999	May	29	7:	37

In most contexts, a {-gla} ordinal by itself, or two {-gla} ordinals, refers to an hour or hour and minute. To disambiguate (if the context and the scale of the numbers involved isn't sufficient) one can begin the sequence with {měn'θu} "month" or {hyr} "hour".

fy-gla	ĝu-gla
seven-ORD.T	thirteen-ORD.T
7:13 am or July 13

měn'θu	fy-gla	ĝu-gla
month	seven-ORD.T	thirteen-ORD.T
July 13

hyr	fy-gla	ĝu-gla
hour	seven-ORD.T	thirteen-ORD.T
7:13 am

{-gla} ordinals with negative numbers usually occur as the second of a pair of hour-minute ordinals, indicating N minutes before the hour:

gĭ-gla	se-gâr-gla	i
eleven-ORD.T	minus-ten-ORD.T	at
ten minutes till eleven, i.e. 10:50 a.m.

In theory such negative time-ordinals could be used elsewhere in date expressions, e.g., to indicate the Nth day before the beginning of the next month, or the Nth month before the beginning of the next year, but in practice I've never used such.

Names of days of the week can be formed with {-ŋla}. That suffix doesn't apply only to numbers, though.

Relative time is expressed with {-pa} ordinals, including negative ordinals.

Time-ordinals formed with {-gla} or {-ŋla} can be used as modifiers of event-nouns, meaning "the [event] that happens at [number] o'clock" or "on the [number]th day of the month", etc., by context.

ĥun	ĝu-gla	o	ðu-ƥ-van	heŋ	ruŋ-zô.
meeting	thirteen-ORD.T	to	able-3-V.STATE	not	go-V.ACT
She won't be able to come to the one p.m. meeting.

When a single {-gla} or {-ŋla} stem is involved, one can say what day or time it is by appending the stative verb suffix {-van} to the time word:

gĭ-gla-van.
It's eleven o'clock. or perhaps It's November, or the eleventh of the month.

tě'θru-ga-van	pân;	ʝǒn	ĉu-ŋla-van.
falling-METAPH-V.STATE	everything	therefore.inference	two-ORD.D-V.STATE
Everything is going wrong; it must be Monday.

The default subject rule is not violated here. The implied subject of the first example above is the first-person pronoun {Ќ}; literally "I (am at) eleven o'clock", — I along with everything else around me. In the second, the "everything" referred to in the first clause with {pân} is the subject of the second verb {ĉu-ŋla-van}, and again the semantics make sense. There are some contexts where the default subject is some timeless entity and, changing the subject to say what time it is, you would need to explicitly supply a subject — any noun or pronoun referring to something that is currently existing would do, but I suppose I would usually use {Ќ} "I".

prym-źa-fwa-van	ŋî'bĭ	kǒ	~~~	hwǒ,
aesthetic.pleasure-AUG-CAUS-V.STATE	number	this	...	whoa

fy-gla-Ќ-van	gwe.	su	jâ-o	mwe.
seven-ORD.T-1-V.STATE	already	standing	state-to	IMP
[lying awake and thinking about number theory] This sure is a beautiful number... whoa, it's seven already, I have to get up.

When a more complex time expression is used, one must put the time into a postpositional phrase with {i} and use the verb {nu-van}:

gĭ-gla	hâ-gla	i	nu-van.
eleven-ORD.T	seventeen-ORD.T	at	moment-V.STATE

The semantics of default or explicit subjects with {nu-van} is the same as with {-gla-van} or {-ŋla-van} verbs.

The suffix {-lwa} means "approximately" and is used primarily with number words, often with one of the other suffixes appropriate to numbers.

ĝu-lwa-gla	i	vâ-oŋ-zô	de	Ќ-ƥ.
thirteen-approximate-ORD.T	at	digestion-into-V.ACT	extended.now	1-3
We're eating [lunch] at about 1:00 pm these days.

mluj	fĭw-câŋ	om	ruŋ-zô
small.convention	fiction-experimental.science	to-part.of	go-V.ACT

mâ	dâ-tĭm-lwa-bô	tu-i.
person	three-hundred-approximate-ADJ	AGT-at
About 300 people came to the science fiction convention.

At least the suffixes {-bô}, {-gla}, and {-lwa} can occur with fractional numbers; I'm not sure if {-pa} or {-saw} would ever make sense in such a context.

fy-pe-ðe-ĉu-bô	seven and a half
fy-pe-ðe-ĉu-gla	seven thirty a.m.
fy-pe-ðe-ĉu-lwa	approximately seven and a half
fy-pe-ðe-ĉu-gla	approximately seven thirty a.m.

At least two of the proper name suffixes can be applied to numbers. {-wam}, the place-name suffix, is used to designate highway numbers, e.g.,

pě'hĭ-źa	ðy-hâ-wam	ol	fyn-zô.
road-AUG	five-seventeen-NAME.PLACE	through	drive-V.ACT
I was driving along Interstate 85.

{-gam}, the miscellaneous name suffix, is used with room numbers and the like.

du	ĉu-tĭm-cĭ-gam	i	râ-van	ĥun.
room	two-hundred-one-NAME.GENERAL	at	happen-V.STATE	meeting
The meeting will occur in room 201.

Counting

When counting discrete things, not in a strongly linear order, one says {cĭ ~~~ ĉu ~~~ dâ ~~~ ĉu-ĉu ~~~} and so forth, using bare number words; or perhaps, if counting by threes for instance, {dâ ~~~ ĉu-dâ ~~~ dâ-dâ ~~~) etc. If one is counting a series of similar actions or events, e.g., while lifting weights or some such iterated exercise, one uses ordinal numbers: {cĭ-pa ~~~ ĉu-pa ~~~ dâ-pa ~~~ ĉu-ĉu-pa ~~~}. I reckon that in counting heterogeneous actions or events, one would use {-saw} ordinals, but I don't think this has actually come up in practice. Counting discrete items in a linear sequence (e.g., sitting in one's car at a railroad crossing and counting the cars of the passing train), one could use either cardinal or ordinal numbers.

Main {gjâ-zym-byn} page
Phonology
Syntax
Semantics
My conlang page
My home page
Last updated February 2022