| << Previous | Up | Next >> |
Terbian verbs are used to show volitional and non-volitional actions, states and changes of state. Verbs are the most inflected part of the sentence. They encode subject and object, and each has several derived stems. As explained in the cases and semantic roles section, Terbian is a split-S language, which means that there are two basic semantic roles for the arguments of a verb, Agent and Patient, and intransitive verbs are distinguished according to the role of their subject.
Morphologically, the verb phrase consists of an inflected verb and an optional applicative particle or preverb. A finite verb word is made up of an object agreement mark (if transitive), the stem (the root plus the voice mark), the subject agreement mark, and the tense/aspect/mood inflection.
Terbian is moderately satellite-framed, preferring separable particles that encode directional and locational information in the applicated voices (antipassive, unergative, inversive, and applicative).
Note on the examples: The glosses do not necessarily reflect the details of the morphology. That is, not all morphemes are glossed separately, and some morphemes are glossed in a different order from that in which they actually appear. The verbal word is glossed as follows: first the root, then the voice (if not the default, active voice), then the subject agreement, then the object agreement (if transitive), and then the tense/aspect/mood mark (if not the default Realis/Present/Imperfect).
Apart from all other classifications, and orthogonal (independent) of them, verbs are divided in two classes (named I and II). These classes are like noun genders in Latin, Greek, Spanish, German, etc., in that they are completely arbitrary. There is no way to know whether a verb belongs to Class I or to Class II, so this has to be learned for each one. The class determines some differences in the formation of derived stems and in the tense/aspect/mood inflections. Some irregular verbs fall outside of both classes.
Type A verbs are verbs that mark their subjects as agents (A). This usually means that the subject is in control of the action, which is volitional (willful). Examples: pont- ‘swim’, māw- ‘write’. All transitive verbs are Type A. Examples: agy- ‘close’, ddurat- ‘buy’.
The objects of transitive verbs may be deleted, but a transitive verb with no stated object must still agree with its implicit object (see morphology below). Intransitive Type A verbs show no such agreement. Note that deleted third person objects are assumed to be class I.
Sidduratat.
buy-As-3pI-PRF
“I bought them.”
Erdat.
stand-As-PRF
“I stood.”
Type P verbs are intransitive verbs that mark their subjects as patients (P). Examples of P-verbs are aliky- ‘fall’, kkin- ‘know’, null- ‘sleep’. Many P-verbs are in fact mediopassives of A-verbs (aliky- ‘fall’ comes from al- ‘drop’).
Some P-verbs follow a different syntactic pattern. Nonvolitional perception verbs like ukj- ‘see’ are impersonal; the P-argument is the perceived object, but the ‘subject’ appears in an oblique case. These verbs are called inversion verbs. This pattern is special case of topic fronting, so another name could be ‘oblique-topic verbs’.
Ū tta lebr ukjib.
me to house see-3sPII
“I see a house.”
Dā tta babinu ukjĕ.
she to mice see-3pPI
“She sees mice.”
As you can see, the verb agrees with its ‘object’ (the direct object of the English equivalent), in person, number and noun class (lebr ‘house’ is class II; babinu ‘mice’ is class I). This is better understood if one considers all nonvolitional perception verbs as inherently mediopassive (instead of ‘see’, read ‘be seen [as]’; instead of ‘hear’, ‘sound [like]’, etc.).
If one needs to specify that the topic is the P-argument, this argument becomes fronted, and the sentence then looks like one with a regular P-verb. If the oblique is dropped, the verb becomes a true (lexical) mediopassive.
Babinu ukjĕ.
mice see-3pPI
“Mice are seen.” or “Mice are visible.” or “Mice show.”
An inversion verb often has a lexical counterpart that acts as its inversive voice; for example, for ū tta han pezes ‘this is known to me’ we can also have hān dis kkina ‘I know of this’.
Note that the volitional perception verbs are not like this, but regular agentive intransitive verbs (vi-A), which are often applicated to add a P-argument. A nonvolitional perception verb can thus be derived from a volitional one by applicating a P-argument and then using the mediopassive voice to delete the A-argument.
Er kapos. “He speaks.”
Er pĕlu skappos. “He says words.”
Pĕlu kappikyĕs ē lås. “Words
are spoken by him.”
Pĕlu kappikyĕs. “Words are spoken.”
~ “Spoken words sound.”
Many natural languages employ this kind of pattern with perception verbs and others. English has traces of such constructions: ‘methinks’ is the old dative of the first person singular plus the then-impersonal verb ‘think’: ‘to me it thinks’, which parallels the semantics of ‘to me it seems’, ‘to me it looks’, etc. I believe that Russian does this quite consistently too.
All verbs agree with their grammatical subject. The verb distinguishes between singular and plural numbers, and between non-actual persons and actual persons (third vs. first and second). For the non-actual persons the verb also agrees with the class of the main noun in the subject. Intransitive verbs of type P (vi-P) have different agreement marks for third person subjects (A-intransitives use the same affixes as A-transitives). Subject agreement appears right after the stem, before the TAM mark.
Transitive verbs additionally agree with their object. Object agreement appears as a prefix, which refers to the person and number. Note that the singular third person object (which is by far the most common) is zero-inflected, while the third person plural object distinguishes between class I and class II main nouns.
| Verb argument agreement | ||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||
The subject and object agreement marks of any verb can be replaced by interrogative marks, which in effect works the same as if the speaker had used a dummy agreement mark and the independent pronouns isa ‘what’ or itar ‘who’. The interrogative subject mark has two allomorphs: -uk-, -ku-. The interrogative object mark has two allomorphs as well: ma-, m-.
Atto i hjker ū tta kidindelkurā?
now that cup 1sO for refill-Q-3s-PRO
“Who now shall refill the cup for me?”
Lal nuk?
2s be-Q
“Who are you?” (the pronoun may be omitted)
Madduratat?
buy-2s-Q-PRF
“What did you buy?”
With the use of applicatives, the what/who question can be turned into almost anything else.
Sta majdukkja?
APP:LOC+for look-PLV-2s-Q
“What do you look for?”
Har majŏwwat?
APP:GEN come-PLV-2s-Q-PRF
“Where did you come from?”
Gye makapparā?
APP:DAT speak-PLV-2s-Q-PRO
“Who are you going to tell?”
(lit. “Whom are you going to speak to?”)
The difference between using an interrogative pronoun and an interrogative verb affix is one of register and focus.
Omitting the interrogative pronoun is quite common in short sentences, especially when the topic is already established and the sentence consists of the verbal word alone, or the verb plus one short argument. Deleting the pronoun in longer sentences (where the verb is preceded by a long phrase argument and/or one or more complements) is considered bad style and confusing, though it is admitted in poetry. Nonstandard speech (such as joking and storytelling) often uses interrogative pronoun deletion for surprise endings, whereby a seemingly affirmative statement is suddenly revealed as a question.
Regarding focus, the question word marks a higher level of importance of the questioning; a verb with an interrogative affix alone tends to be considered rhetorical. The question
Itar han got?
who this do-3s-3s-PRF
“Who did this?”
definitely expects an answer in its own terms (especifically the person who did the thing referred to); on the other hand a question like
Han gukut?
this do-Q-3s-PRF
“Who did this?”
is less forceful, admitting another question, an excuse, an explanation, or simply “I don’t know” as alternative answers instead of the name of the responsible person.
Some speakers emphasize questions by using both the pronoun and the interrogative mark, but this is definitely colloquial or humorous.
When an argument is not made explicit but only marked on the verbal word, it can be interpreted as something already mentioned (a topic) or, less commonly, as a non-expressed indefinite argument. By using the indefinite argument marks instead of the regular, topical ones, the speaker can show this distinction clearly.
The indefinite mark is an affix of the form j(e)-, -j, that attaches to the third person agreement mark. In practice:
Singular Plural
Subject -oj -ëj
Object j(e)- jes-
The difference between topical and indefinite (NDF) marks can be shown by examples. Compare the following:
Er kapomi. Kuzl agyomi.
he speak-3s-NAR door close-3s-3s-NAR
"He spoke. He closed the door."
Er kapomi. Kuzl agyojmi.
he speak-3s-NAR door close-NDF-3s-NAR
"He spoke. Someone closed the door."
The indefinite agreement marks are used mainly in subclauses, especially conditional subclauses in the Irrealis mood. When the subject is marked as indefinite, a zero-marked object is assumed to be indefinite.
Eibojzå tugår nuy.
kill-NDF-3s-IRR crime be-SCL
"It is a crime that someone kills someone."
The subject or the object of a verb may refer to whole subclauses instead of noun phrases. These references are marked using other special affixes, -uy- for subjects and ju- for objects.
Er erdorā juēdorana.
3sM stay-3s-PRO hope-1p-SCL
“We hope he will stay.”
Histr ye nas na nam iseibo kodz nuy.
soldier A children and women kill-3s-3p wrong be-SCL
“It is wrong for a soldier to kill women and children.”
Subject and object agreement may, as explained above, be definite (topical) or indefinite; it may refer to noun phrases or to subclauses. Additionally, the agreement marks can be focused or unfocused. A focused argument is marked as such to make a focused statement or question, often one that would be expressed by a ‘cleft sentence’ in English. In Terbian, the focused argument is moved near the verb, and usually marked explicitly with the A or P particles (ye, kos). The focus mark is -ta(t)-, after the agreement affix.
Dar er ye iseibotat.
she he A kill-3s.FOC-3s-PRF
“It was he who killed her.”
Er dar kos istateibot.
he she P kill-3s-3s.FOC-PRF
“It was she whom he killed.”
The non-focused argument may be left out. If the agent is non-focused, the result is an emphatic statement only; if the non-focused deleted argument is the patient, the pragmatic result is a passive-like sentence:
Dar kos istateibot.
she P kill-3s-3s.FOC-PRF
“It was she who got killed.”
If the non-importance of the agent needs to be emphasized, the corresponding agreement mark can be made indefinite:
Dar kos istateibojt.
she P kill-NDF-3s.FOC-PRF
“It was she whom someone killed.”
Verb tense, aspect and mood are fusional (“TAM complexes”). Each TAM complex is shown by a sufix that attaches to the verb, after the subject mark, and varies according to the verb class. There are two moods, Realis (basically what is known as indicative mood) and Irrealis (conditional or subjunctive), and each has several combinations of tense and aspect.
The Realis mood is used for positive and negative assertions about actual events or states. It has five tense/aspect complexes:
Nonfuture/Nonperfect (NPE): Shows habitual actions (“I work at a hospital”) or present actions that last some time and are not finished (“He’s looking at you”). Habituality, when not understood by context, can be marked by using adverbs of time and manner. The main feature of this TAM complex is that the action is not perfect (finished).
Present/Prospective (PRO): Used for actions that one is about to perform, or future actions that begin as a decission in the near present (“She’ll be there next week”, “I’m going home as soon as I finish this”). In practice, extended to all future actions, but requiring some help from context in that case.
Present/Perfect (PRF): Actions that have just ended, especially punctual actions that are not being undertaken in the referenced instant (“I’m done”, “They’ve just surrendered”). In practice, this also covers past actions not included in the Narrative, especially past actions with a result that is visible at the moment of speaking (“Who did this?”).
Past/Narrative (NAR): Used in storytelling and generally of all narrative action (“She came in and I said, ‘What are you doing here’, and she replied...”.)
Past/Pluperfect (PPF): Shows past actions that are completed at the referenced time frame, especially in combination with the Narrative (“We gathered at the place where we had agreed to meet that night...”).
The Irrealis mood is used for hypothetical assertions (it may be also called subjunctive), including conditionals. It has only three distinct TAM declensions: the Imperfect, the Perfect, and the Narrative. The Imperfect covers both present and future actions, especially those indicating processes; the Perfect is used for punctual actions and events; the Narrative appears when conditional and hypothetical events are mentioned in storytelling.
| TAM inflections | ||||
| Realis | Irrealis | |||
| Habitual | -0, -s | Imperfect | -zĕ, -zå | |
| Prospective | -rī, -rā | Perfect | -tīt, -tāt | |
| Perfect | -t, -t | Narrative | -y, -y | |
| Narrative | -n, -mi | |||
| Pluperfect | -dī, -dā | |||
Each verb has at least one basic stem, which is part of the citation form. Upon this stem others are formed. In most cases a thematic vowel appears in the new stem; this vowel is i, ī (long or short) for Class I, and a, ā for Class II.
Combining stem: Used as an infinitive in some cases, like subordinate clauses. In most cases where two verbs must be combined, this is the form that should be used. This is also the form that can be followed by a postposition. Class I verbs add -īn; Class II verbs add -ān.
Participial stems: These become Class I noun/adjective stems. The active participle adds -īk for Class I, -āk for Class II. The passive participle adds -im for Class I, -am for Class II.
Voice stems: These stems are further inflected as if they were basic stems of simple verbs. They are described below, in their own section.
| Basic verb stems | ||
| Class I | Class II | |
| Combining | -īn | -ān |
| Active pple | -īk | -āk |
| Passive pple | -im | -am |
Verb voice is the relationship between the verb and its arguments (typically Subject and Object), and the way in which they are marked. As explained below, transitive verbs have A-subjects and P-objects, and intransitive verbs have only a subject that may be marked either as A or P.
A voice can be thought formally as an operation over the argument structure of the verb. The basic voice in any given language is normally termed ‘active’ and shows the original argument structure (in formal terms we would call it an identity operator). Other voices shift argument marks, promoting or demoting arguments (promotion usually means turning objects into subjects, while demotion involves turning subjects into objects or oblique complements).
English only has a periphrastic passive voice (“She eats food” → “Food is eaten [by her]”). In English passive voice, the Object becomes promoted to Subject, and the original Subject is demoted to the position of an optional complement. Terbian has three voice operators besides the active voice, and none of them is this simple passive voice.
In mediopassive voice (MPV) the patient (the argument marked as P) becomes the subject of the verb, while the agent (A) is demoted. Mediopassive voice sometimes acts as a simple passive, others not so. The verb becomes a P-verb, with the patient staying the same (but fronted), and the agent is demoted to an oblique complement.
man-A door-P closes -> door-P is-closed [by man-O]
bardr ye kuzl agyo -> kuzl agyikyo [bardo lås]
the man closes the door -> the door is closed [by the man]
man-A brick-P drops -> brick-P is-dropped [by man-O]
bardr ye unkjun alo -> unkjun alikyo [bardo lås]
the man drops a brick -> a) the brick is dropped [by the man]
b) the brick falls
The mediopassive voice suggests a static verb or a description of a state, or a verb action performed by and for the subject’s benefit. Therefore it shares some of the features of passive voice and of middle voice (as found in ancient Greek and also in the ‘pseudo-reflexive’ verb forms of modern Romance languages). In Terbian, most verb meanings like ‘wash’ (transitive or intransitive-reflexive) and ‘think, reflect’ (intransitive) are in their basic forms transitive A-verbs (‘wash someone, reflect about something’) that become mediopassive (‘wash [oneself], think’). It is possible to use a reflexive pronoun with a basic A-verb such as ‘wash’, if one wants to emphasize the agency of the subject, but not common.
The antipassive voice (APV) works on transitive verbs only. Both arguments are demoted and the verb becomes an intransitive P-verb. The subject is marked as P and the patient is dropped (it may appear as an oblique complement). The verb becomes the description of a state or condition. The original function of the object appears as an applicative. The locative applicative can appear on its own, or in a reduced form attached to a postposition of location (‘on’, ‘under’) or motion path (‘towards’, ‘from’), which should be the head of the original complement (see below). The object applicative is zero.
dog-A cat-P bites -> dog-P (APP:OBJ) bites [cat-O] kimĕd ye zemsr hhulos -> kimĕd kos hhulōhos [zemso] the dog bites the cat -> the dog is a biter [of the cat]
Attempted actions can be shown by substituting the dative for the object applicative:
dog-A cat-P bites -> dog-P APP:DAT bites [cat-O] kimĕd ye zemsr hhulos -> kimĕd kos gye hhulōhos [zemso] the dog bites the cat -> the dog takes a bite [at the cat]
This alternative form should be read as “The dog bites at the cat”, i. e. it attempts to bite the cat, but does not necessarily reach it. Some verbs allow a wide range of meanings using different applicatives (hhul- ‘bite’ can be used with the locative applicative plus ar ‘in’, too).
The name ‘antipassive’ for this voice comes from ergative/absolutive languages; I adopted it following encouragement from fellow members of CONLANG. It’s not used only for syntactic purposes (detransitivization), but also to focus on the static nature of the verb. With regards to its pure syntactic use, the antipassive tends to appear (in educated speech) when coordination is needed between an A-verb and a P-verb in a subclause with a common subject.
Some speakers use the antipassive stem for some or all negated A-verbs. This apparently rests on the assumption that a negated action is in fact a state (non-doing); therefore ‘dog does not bite cat’ becomes ‘dog is not biter of cat’.
The unergative voice (UEV) works on transitive verbs only. The subject stays the same and the object is dropped. The verb thus becomes intransitive, with a subject marked as A, or ‘unergative’. The original function of the object appears as an applicative. This voice is similar to the antipassive, the difference being that the antipassive creates a P-verb, while the unergative creates an A-verb. Technically speaking, an unergative verb is an intransitive verb with an agentive subject.
this dog-A cat-P bites -> this dog-A (APP:OBJ) bites ah kimĕd ye zemsr hhulos -> ah kimĕd ye hhulējos this dog bites the cat -> this dog bites
The inversive voice (INV) works on P-intransitive verbs with a complement. The P-subject is dropped (or demoted to oblique), while the complement is promoted to subject (i. e. subject and complement are exchanged). The original function of the complement is marked with an applicative.
cat-P sleeps on mat-O -> mat-P APP:LOC-on sleeps [cat-O] zemsr nullo sīgebo ek -> sīgebr sek nulinto [zemso] the cat sleeps on the mat -> the mat on-sleeps [the cat]
This may be read as “The mat on-it-sleeps the cat”, but also and more accurately from the semantic point of view, “The mat is a sleeping-location [for the cat]”, or “The mat is slept on [by the cat]”.
man-P lives at that place-O -> that place-P APP:LOC-at lives man-O bardr yōto i kwarō ar -> i kwar sar yōtinto bardo the man lives at that place -> that place at-lives [of the man]
This should read as “This place is inhabited by the man”, or “This place houses the man” or “This place is a living-location for the man”. Without the oblique complement, this reads as a simple passive (or pseudo-passive state), “This place is inhabited”, i. e. “Someone lives in this place”.
The applicative voice (PLV) works on A-verbs with a complement (i. e. either transitive or A-intransitive). The complement is promoted to the P-role, and an applicative is used to marked its original function. If the verb was transitive, the original P-object is demoted.
cat-A runs towards mouse-O -> cat-A APP:LOC-towards runs mouse-P zemsr åkwo babinō tīg -> zemsr ye stīg åkkwo babin the cat runs towards the mouse -> the cat towards-runs the mouse priest-A preaches to crowd-O -> priest-A APP:DAT preaches crowd-P dĕkkin bagderos jissō tta -> dĕkkin ye gye bagderros jiss the priest preaches to the crowd -> the priest to-preaches the crowd priest-A preaches sermon-O -> priest-A APP:OBJ preaches sermon-P dĕkkin bagderos kēddermō ma -> dĕkkin ye bagderros kēdderm the priest preaches a sermon -> the priest preaches a sermon
Note that bagder- ‘preach’ is an intransitive verb always. Unlike English, where verbs can function as intransitive, transitive or ditransitive without change, Terbian tends to express this kind of concepts (‘give’, ‘offer’, ‘tell’, ‘dictate’, etc.) as intransitive verbs, leaving room for valency increases using the applicative voice. The applicative for beneficiaries (the people given or told something) is the dative gye, while the applicative for the conveyed things (given) or topics (spoken about) is the object applicative (zero). The ablative har is used (with discourse verbs) for sources of information.
The translation of ‘applicated’ verbs is not always simple, since English has no means to render this construction except lexically. One can make some approximations, though. ‘Run’ with a ‘towards’ locative applicative suggests ‘chase’; it can be substituted for a dative applicative on its own. Discourse verbs like ‘preach’ or ‘say’ with the dative applicative can be rendered as ‘address’, ‘lecture’ or simply ‘tell’.
The causative voice (CAV) works on P-intransitive verbs that can take an agentive or causative complement. The P-argument is left as is, and the complement is promoted to the A-role. (The P-argument is now the object and it must be marked in the verb, as shown in the example.)
we-P know -> he-A us-causes-to-know lluks kkinana -> ery yakkiniddo we know -> he teaches us
| Name | Abbr | Mark | Appl | Val | Base | Makes | What it does |
| Mediopassive | MPV | -iky, -akw | no | -1 | vt | vi-P | demotes A to O |
| Antipassive | APV | -āz, -ōh | yes | -1 | vt | vi-P | demotes A and P |
| Unergative | UEV | -ēy, -ēj | yes | -1 | vt | vi-A | demotes P |
| Inversive | INV | -int, -ant | yes | 0 | vi-P | vi-P | exchanges P and O |
| Applicative | PLV | -:, -: | yes | +1 | vt, vi-A | vt | demotes P if present, promotes O |
| Causative | CAV | -idd, -add | no | +1 | vi-P | vt | promotes O to A |
Appl = uses an applicative; Val = valency change.
| Name | Abbr | Other | Lex | Description |
| Object | APP:OBJ | Patient | 0 | object or patient of a transitive verb |
| Locative | APP:LOC | Essive | ŏs, s- | location or physical state, ‘in’, ‘on’, ‘at’ |
| Dative | APP:DAT | Allative | gye | destination, goal, direction, ‘to’, ‘directed at’ |
| Ablative | APP:ABL | Genitive | har | source, basis, topic, ‘about’, ‘associated to’ |
| Instrumental | APP:INS | Causative | jut | means, instrument, physical cause, ‘by’ |
Other = other names and functions; Lex = lexical form.
The bound form of the locative s- is followed by a postposition.
| Object | ROOT | Voice | Subject |
TAM |
|
ū- |
-0 -iky -āz -ēy -int -: -idd |
-a -(a)so, -o -o, -ib(i) -ana -(a)sĕ, -ĕ -ĕ, -ubĕ |
-0 -rī -t -n -dī |
|
| -zĕ -tīt -y |
| Object | ROOT | Voice | Subject |
TAM |
|
ū- |
-0 |
-a -(a)so, -o -o, -ib(i) -ana -(a)sĕ, -ĕ -ĕ, -ubĕ |
-s -rā -t -mi -dā |
|
| -zå -tāt -y |