← All chapters

§8 Verbs

Pine's verbal morphology can be intimidating and rightly so – Pine is a very verb heavy language, meaning much information is encoded in the verb through various means. Most notably, Pine verbs can be polypersonal – information about all participants of the verb is found inside the verb itself. Perhaps the most unusual feature of the polypersonal Pine verb is that object and subject markers are not fixed – the ending -i in the indicative present may refer to a third-person animate, a fourth person, a third-person inanimate or an third-person obviate, depending on what the other argument of the verb is. For example, neg- ‘see’ can be combined with various pronominal prefixes and -i to yield different subject-object alignments:

The suffix -i thus does not mark an object directly but rather tells us about the relationship between the pronominal prefix and the object. If it is instead following by -em-, then it tells us about the relationship between the object pronominal prefix and the subject:

This is a fundamental mechanism in polypersonal verbs – one participant is overtly marked and its relationship to the other participant is expressed through endings.

Verbs are by far the most morphologically complex and rich category of the language, boasting a large array of prefixes, suffixes, modifiers, root elements, extensive derivational morphology and a complex pronoun system. Pine verbs have a fusional morphology, meaning each affix can carry a range of morphemes. In addition, many ideas commonly expressed with nouns, adverbs and adjectives in English instead appear as verbs in Pine, most notably predicative adjectives – while the adjective tammet ‘big’ can stand before a noun as an attributive adjective like in English, as in tammet vukvi ‘a big bird’, it appears as a verb when it is the argument of the verb ‘to be’ in English, as in tammegan vukvi ‘a/the bird is big’, literally ‘a/the bird bigs’. Verbs are an open and very productive class that can accommodate new roots easily. The various elements of Pine verbs are organised around the root into slots, which can or must be filled with various components which give information beyond that found in the basic meaning of the root.

Valency greatly affects the morphology of verbs. Monovalent verbs, those which only have one participant (whether singular or plural) differ markedly from polyvalent verbs, which have more than one participant (transitive and ditransitive verbs). This is because participants are directly and indirectly marked on the verb. The alignment system of polyvalent verbs in Pine is remarkably different from that found in most European languages. Pine indicates polyvalency by means of a system of alignment called coordinate positional alignment. Participants are ordered hierarchically, and only the hierarchically higher participant, called the suprapositional pronoun, is overtly marked. Moreover, the first and second person are grouped together as ‘local’ pronouns while other pronouns, namely the third-person animate (3a), the fourth person (4), the third-person inanimate (3i) and the third-person obviate (3o) are called non-local pronouns.

pine pronominal hierarchy

[2 ⟫ 1] [3a ⟫ 4 ⟫ 3i ⟫ 3o]

All non-suprapositional participants are encoded into the verb as a function of their distance and direction from the suprapositional pronoun. For example, the marker -i indicates one degree of distance from the suprapositional pronoun. Therefore, ga-Ø-i indicates 1⟫3a (I verb her/him), while t-Ø-i indicates 3a⟫4 (s/he verbs someone/something). The distance is measured in degrees of separation from the suprapositional pronoun based on its relative hierarchical position to the suprapositional pronoun. For example, the marker -i is said to be of the 1˚ (first degree). The direction of this distance is measured as preceding from the top of the hierarchy to the bottom. Moving in the opposite direction to the hierarchy (from lower to higher) is the retrograde alignment (which in the examples above is encoded by the marker -em-), while moving with the hierarchy (from higher to lower) is called the prograde alignment. The form ga-Ø-em-i is therefore in the 1˚ retrograde and indicates 3a⟫1 (s/he verbs me).

Thus, information about participants in a verb is given as a sort of grammatical coordinates. As such, polyvalent verbs may have a prograde (from higher to lower hierarchical position) or retrograde alignment (from lower to higher hierarchical position) of anywhere between the zeroeth degree to the fourth degree (0˚, 1˚, 2˚, 3˚ or 4˚). This is true of both participants in transitive verbs, and the position of the direct object in ditransitive verb, which is anti-ergative (meaning the indirect object of a ditransitive verb is treated as the direct object of a transitive verb, which in Pine grammar has been called the antergate, a syntactic indirect object but morphologically a direct object), is relative to the suprapositional pronoun as well. In other words, instead of thinking about who the various participants of a verb are, a Pine speaker thinks of their distance and direction in relation to which ever participant is hierarchically highest. Rather than stating directly who is the subject and who is the object, polyvalent Pine verbs tell us about one participant and how to get from there to the other participants, while also telling us whether the reference point is the subject or the object. This is a very different way of thinking about participants than what most are used to, and requires a good level of familiarity with verbal morphology. As such, polyvalent verbs can represent the biggest hurdle to students of Pine.

Beyond the complex polypersonal nature of verbs, Pine also overtly differentiates agentivity – whether an actor is in control, acts willingly or consciously, or whether the action is experienced by the actor rather than controlled, occurring without their willing or conscious intent. Aspect and translativity are also prominent features of Pine verbal morphology, with stative and translative verbs differing markedly from active verbs in various ways, as will be described in this section.

§8.1 Overview of Verbal Morphology

It is important to separate Pine verbal morphology into two distinct categories. Monovalent verbs behave very differently from polyvalent verbs (transitive verbs with two participants and ditransitive verbs with three participants), and they must be considered separately. However, many elements are common to both. This section aims to detail those elements which form the common core of all verbs in Pine.

The core of the Pine verb is a root, which carries the fundamental meaning of the verb and cannot be analyzed further. The root is what is left when all affixes have been removed, for example migv- ‘eat’. The verbal root can change depending on tense. The past tenses (past and pluperfect) of the realis moods causes the nucleus of the root to change through a type of ablaut called rhizal apophony, either called fronting or backing apophony, for example migv- ‘eat’ ⟫ mỉgv- ‘ate’. A verbal root cannot stand alone, which is why it is written with a hyphen. Affixes added onto the verbal root are either called prorrhizal (coming before the root), such as ga- ‘I, me’ or metarrhizal (coming after the root), for example -va, which indicates a local subject of an intransitive verb in the autonomous indicative present. Combining pro- and metarrhizal affixes yields grammatical, conjugated forms of verbs, for example ga-migv-a ‘I eat’. The most important affix of verbal morphology is called the theme, which is metarrhizal. The theme houses the majority of the information contained in a conjugated form of a verb. Verbs rely heavily on changes to the theme to convey information beyond the bare meaning of the root.

Pine verbs convey a great deal of information through inflection, a combination of various morphemes. Inflections give information about voice (whether the participant acts, acts upon itself, changes state or is the recipient of an action), order (main or subordinate clause), mood set (whether the statement is real or unreal), evidentiality (whether the information is reliable and actual or whether there is a measure of doubt about it), tense (the time at which the action takes place) and participants. The aforementioned categories are obligate, meaning a verb form cannot be grammatical without specifying this information. Other parts of the verb may appear or be left out without causing the verb to be agrammatical. These morphemes are called facultative, and they include aspectual markers. Order, mood sets, evidentiality, moods and tenses combine to give 30 possible combinations. The result of each combination, for example the autonomous realis assertive indicative present, is called an inflection.

Obligate inflectional categories are given below with a description of their function and meaning:

voice - information about the relationship between the action and its participants

active - the participant performs or experiences the action

middle - the subject and the object are one and the same

translative - the participant experiences a change of state

passive - the patient of a transitive verb is promoted to the subject of an intransitive passive verb

order - information about the type of clause

autonomous - main clause

hyponomous - subordinate clause

mood set - information about whether the statement is real or not

realis - the event or statement is factual, real

irrealis - non-factual events, either conditions or potential events

mood - further information about the relationship of the verb to mood set and evidentiality

irrealis moods - the statement is not factual, true

conditional - the statement would be factual (if…)

potential - the statement could be factual (if…)

realis moods - the statement is either inferential or assertive

assertive - the statement is reliable, first-hand information, factual

indicative - the statement is true/declarative

imperative - the statement is a declarative command

inferential - the statement is either dubitative or optative

dubitative - the statement is may or may not be factual

optative - the statement represents a wish, command, hope, desire

tense - information about time

present - the event is concurrent to the frame of reference

past - the event precedes the frame of reference

perfect - the event occurs and ends before the frame of reference

pluperfect - the event had occurred before a past reference point

future - the event will occur at some time beyond the current reference point

dependent - basic dependent form of a non-negative verb

§8.1.1 Voice

Voice gives information about the relationship between the action and its participants. The active voice shows that the subject peforms or experiences an action, which is the default voice of verbs. The middle voice, whose formant is -VttV(g)- is found in verbs whose subject and object and one and the same, meaning the subject acts upon itself, and as such correponds to reflective verbs, for example gaddet ‘I control it’s but gadduttuva ‘I control myself, I have discipline’. The translative voice indicates that the subject undergoes a change of state, and as such all translative verbs are obligate unagentive. Lastly, the passive voice tranforms the recipient or object of a transitive verb into the experiencer of an intransitive verb, and corresponds roughly to the passive of English.

§8.1.2 Order

Pine verbs can appear in one of two orders called the autonomous and hyponomous order. The autonomous order is found in main clauses, while the hyponomous order is found in subclauses, relative clauses and in certain specific constructions. Hyponomous can thus best be described as a conjugation pattern that is not found in main clauses. For example, gamikva ‘I go’ is in the autonomous order (ga-Ø-v-a) but įu gamigga ‘when I go’ is in the hyponomous order (ga-Ø-a), where the pronoun ga- ‘I’ is unchanged but the ending -v-a becomes -a. This ending contains multiple layers of information, such as locality, order, mood, tense and assertivity.

§8.1.3 Mood Sets

All verbs also have two main sets of moods called the realis and the irrealis. Mood sets are a higher order classification than moods, and therefore all moods are part of either the realis or irrealis mood set. Furthermore, the realis mood set distinguishes three moods which are either assertive (the indicative mood) or inferential (the dubitative and optative moods). Thus, each mood is classified according to its evidentiality and mood set.

The irrealis set contains conjugation paradigms that confer to the verb a sense that the information is not realised, not true. For example, gilkka simply means ‘I go’ and is in the realis (more specifically in the indicative mood) because the information is understood to be real or accurate. There are three realis moods: the indicative (gilkka ‘I go’), the dubitative (gilkku ‘I may go, I apparently go’) and the optative (gilgusut ‘I want to go, may I go’). The form gilkkatṡat ‘I would go’ on the other hand is in the irrealis (more specifically the conditional) because the action of going has in fact not occurred and is conditional or imagined. There are two irrealis moods called the conditional (gilkkatṡat ‘I would go’) and the potential (gilkkulut ‘I could go’). The root of a verb in the irrealis mood set is not affected by tense, which sets the irrealis mood set appart from the realis mood set, where past tenses are affected by a change in the nucleus of the root through a process called rhizal apophony. Rhizal apophony is also dependent on order. The realis mood set in the autonomous undergoes fronting, and backing in the hyponomous, while the opposite is true of the irrealis mood, which has backing when autonomous and fronting when hyponomous.

§8.1.4 Evidentiality

The realis mood set further distinguishes evidentiality. All realis verb forms can either be assertive, meaning the utterance is seen as factual, true or that the information comes from first-hand knowledge, or they can be inferential, which indicates some level of doubt, that the information is not confirmed, that the information is perceptual or sensory or that it is second-hand information or reported. As such, the realis mood combines with the assertive to form the indicative mood, and with the inferential to form the dubitative and optative mood. For example, gamidda įa ỏhakittṡu tsesiḷḷuttun ‘I know that you did something’, where tsesiḷḷuttun ‘you did (something)’ has assertive evidentiality (in the indicative mood) contrasts with mittyt nukka, ỏhakittṡu e tsesiḷḷuttun ḷu ‘I don’t know whether you did something’, where tsesiḷḷuttun has inferential evidentiality (in the dubitative mood), because it is a statement of doubt. The dubitative and optative moods have a pluperfect form found only in the hyponomous order, and the dubitative also has a future tense in both orders. The only distinction between gilkka ‘I go’ and gilkku is that the first statement ‘I go’ is seen as factual or gnomic, while the second statement ‘I may/apparently go’ tells us that the information conveyed is not entirely factual, is not first hand information, unconfirmed, speculative or otherwise not guaranteed. Because of their intrinsic unagentive nature, translative verbs lack an optative and instead, the same endings in the translative instead have an inchoative meaning, indicating that the change of state is beginning, for example ḥilgumi ‘it apparently goes’ (unagentive) gains an inchoative meaning with the addition of the translative marker -uƛ-: ḥilguƛumi ‘it starts to go, it gets going’.

§8.1.5 Mood

Pine has five moods: the indicative, the dubitative, the optative, the conditional and the potential. Moods are classified by evidentiality (assertive, inferential) and mood set (realis, irrealis). Irrealis moods do not distinguish evidentiality as evidentiality is a feature of realis moods.

Each mood has its own set of tenses. For example, the dubitative and optative have a pluperfect not found in the indicative (an obligate hyponomous tense), while the dubitative contains the future tense, and irrealis moods only have a present and past, while all realis moods have a perfect tense.

§8.1.6 Tense

Pine distinguishes six tenses. Verbs can be found in the present to indicate that an action is happening now or happens in general, in the past to show that the action occurred before the present, the perfect to indicate that an action occurred and ended before the present, the pluperfect to show that an action occurred and ended before a set reference point in the past (which must appear in the autonomous order, while the pluperfect action must always occur in the hyponomous order), the future to show that an action will or is likely to occur in the future (and is part of the inferential, which confers a certain sense of uncertainty), and finally the imperative to indicate that the action is a command.

§8.1.7 Participants

Participants are the people or things that are involved in the action described by the verb. Intransitive verbs have one participant, transitive verbs have two and ditransitive have three. Participants in intransitive verbs always appear as prorrhizal pronominal affixes, such as ga- ‘I’ or mi- ‘some(one/thing)’. Polyvalent verbs are also found with only one pronominal affix, and the other participant is indicated through metarrhizal elements found mostly in the theme. Pine distinguishes two types of participants based on their agentivity. Agentivity refers to the level of control the subject of a verb has. A subject which is seen to be aware, in control or willing will be found in the agentive, while a subject which is seen to be unaware, not in control, acting by accident, unwillingly or seen as more of an experiencer will be in the unagentive. Pronominal affixes convey this information, meaning there is one set of pronouns called the agentive pronominal set and a second set called the unagentive pronominal set. For example, ga- ‘I’ is agentive, while ṇe- ‘I’ is unagentive. This means that the verb form g-ilg-va ‘I move’ indicates that the first person is acting of its own volition, while its unagentive counterpart ṇ-ilg-va ‘I move’ indicates that the moving is involontary. Agentivity can be obligate (meaning a verb is only considered grammatical if the pronominal set is either agentive or unagentive) or facultative (meaning both agentive and unagentive pronominal sets are grammatical) as in the example of ilg- ‘move’.

The participants in Pine are given here in their hierarchical order:

§8.1.8 Overview of Inflections

The combination of order, mood set, evidentiality, mood and tense combines to give the main inflectional forms of verbs. The table below summarises these combinations. When talking about an inflected verb, it is enough to specify its order, mood and tense, as mood sets and evidentiality combine to form moods.

Pine verbal morphology is complex and rich and goes far beyond these fundamental elements. In order to properly describe how verbs work, intransitive and transitive verbs will be described separately in much greater detail below. First however, components common to both transitive and intransitive verbs will be described here.

§8.2 Structure & Components of Verbs

The Pine verb is structured around its root. Elements assemble around the root in a fixed and predictable way. The various elements of the inflected verb have a fairly rigid internal order. The internal structure of the verb can be described as being composed of slots. A slot is a fixed position relative to the root and other positions, which can be filled or left empty by verbal elements. Elements may appear in a certain slot in one case, but in a different slot if other elements force it to move to that slot. For example, the first-person pronoun ga- always appears in the same slot, but it can be preceded and followed by other elements. Elements that form the core of the inflected verb are called nucleic elements. Other elements are termed non-nucleic elements, and can be conditionally obligate, meaning they must appear in combination with other elements, or they can be facultative, meaning their presence or absence does not determine the grammaticality of an inflected verb. Non-nucleic elements are categorised by their position relative to the root – initials appear first, medials appear after the root, and finals are found as the last elements of a verb.

At the most abstract, elements are classified by their position in relation to the root – prorrhizal elements come before the root while metarrhizal elements follow it. There are three prorrhizzal slots and three metarrhizal slots. Slots may also contain subslots, which also show a rigid internal order. Slots are numbered starting from the root. The furthest slot before the root is numbered -3, the next -2, and the slot immediately preceding the root is numbered -1. Similarly, metarrhizal elements may appear directly after the root (1) or in either of the following slots (2, 3 or 4). Thus, one can say that there are three prorrhizal slots and three metarrhizal slots, for a total of six main slots. Additionally, verbs can be found with clitics, which are not technically part of the verb (although it could be argued that they in fact are), which are hyphenated to the verb.

The general internal structure of the verb can be abstracted in the following way, with nucleic elements in bold:

§8.2.1 The Root

The root is the most fundamental part of the verb – the part which cannot be further analyzed. Only one verb in Pine has no clear root, the copula. Although it has a dependent form in naġan and a connegative form in uġun, it may not be useful to think of the inflected copula as consisting of a root. Some have analyzed it to have a present root in -a- and a past root in -i-, but even that may not be completely accurate in all cases. Unfortunately for students of Pine, the copula is irregular in nearly all aspects and differs enormously from regular verbs.

Regular verbs however always consist of a root with elements arranged around it. The root can never stand alone, but the dictionary included in this book lists verbs as their bare roots. For example, ḷủl- ‘talk, speak’ is the root given in the dictionary. By combining the root with ga-Ø-va to form the first person singular, present indicative form of the autonomous order, one can generate the conjugated form gaḷủlla ‘I speak’.

The vast majority of verb roots in Pine are mono- or bisyllabic, and all end in a consonant.

The root -ṡkv- ‘come/be/proceed in a certain order’ is unique in that it contains no nucleus, and must appear with a kinetic initial, such as s-ava-ṡkv- ‘come first, be first, happen first’ and g-ava-ṡkv- ‘come second, be second, follow, be subsequent’.

§8.2.1.1 Prosthophonic Roots

Most verbal affixes, excluding medials, are weakly prosthothesising, meaning they have no specific front-rounded versions. For example, the intransitive dubitative ending -u in siġġu 'it apparently rains' has no specific front-rounded form, and therefore prosthothesises to -ů, for example in můertṇimů 'it apparently drizzles'. Prosthothesis is therefore only triggered by the presence of prosthophonic vowels in the word, except in a small number of verbs where no prosthophonic vowels exists but which still trigger prosthothesis. These roots are called prosthophonic roots, and all are otherwise amphiphonic. Prosthophonic roots are:

In these verbs for example, the dubitative uses prosthophonic endings: pitserrů 'you apparently have fever' rather than the expected *pitserru.

§8.2.2 Verbal Slots

Verbal slots represent the internal position of inflectional verbal elements in relation to others. Slots can be composed of subslots when there is internal order within a slot. Pine verbs have 8 slots, but only two must always be filled, the pronominal and thematic slot, with the exception of impersonal verbs which have an empty pronominal slot and can appear with only the thematic slot (although many speakers include a final). The pronominal slot may have a zero surface form, which means that the surface form of a verb may only have one slot occupied.

For example, the impersonal verb siġġ- ‘rain’ can have the present form siġġa ‘it rains’ (alongside siġġan or siġġami, all acceptable). The form hiḍḍan ‘s/he dives’ is an example of a minimally filled verb with a non-local person, where -a- is the thematic slot and -n is a final (the pronominal slot has an underlying t- ‘s/he’, which has no surface form before h-). Likewise, gahiḍva ‘I dive’ only has two occupied slots, the pronominal slot (ga- ‘I’) and the thematic slot (-va). It is acceptable to drop the first-person pronoun with monovalent verbs, so the form hiḍva ‘I dive’ may also be heard, where the pronominal slot is empty (or alternatively has a 0 surface form pronoun) and only the thematic slot containing -va.

§8.2.2.1 Slot -3

Slot -3 or the tritomeric prorrhizal slot is the first slot of the verb. It houses initials, and contains two subslots, numbered -3a and -3b. Initials are morphemes that give information about location, motion and manner. Examples of this include kinetic initials such as the translocative initial gi-/ge(g)- ‘from, away, down’ and the cislocative initial si(s)-, which appear in the subslot -3b, as well as adverbial initials such as uṡ(ṡa)- ‘wrong, incorrectly, with difficulty’ or keu-/kem- ‘together, along’, which are found in the subslot -3a. When directly preceding the local pronouns g-i(m)-/ṇ-(i)m- and ts-(i)m-/p-(i)m-, the kinetic initials appear as si-/gi- and cause changes of the pronoun, such that si-g-i ⟫ siski-, si-ṇ-i- ⟫ sitṇi-, si-ts-i- ⟫ sittsi-, si-p-i- ⟫ sippi- and gi-g-i- ⟫ geggi-, gi-ṇ-i- ⟫ getṇi-, gi-ts-i- ⟫ gettsi- and gi-p-i- ⟫ geppi-.

In the examples below, the tritomeric prorrhizal slot is in bold:

Adverbial initials are a type of imbedded adverb, which may be semantically obligate, meaning it may be necessary for an alternative meaning of a verb. For example, tsůḷḷ- alone means ‘carry’ but key-tsůḷḷ- means ‘mean, signify, imply’.

Kinetic initials are also semantically obligate in some cases, and are conditionally obligate with the motion preverb -i-. The two most common kinetic initials are the cislocative si(s)- and the translocative gi-/ge(g)-. The cislocative can mean a variety of things, most commonly that the actions occurs here, now or that there is movement up, closer, towards the subject or inside. Conversely, the translocative indicates that the action occurs there, then or that there is movement down, further or away from the subject of outside. Kinetic initials are very common, and Pine speakers often use them generously. Kinetic initials can appear in combination with each other, for example in(u)- ‘around’ and the trans- and cislocative combine to form iss(u)- and igg(u)-, ‘around here’ and ‘around there’ respectively. Likewise, the inceptive a- ‘go and, start’ can appear as asi(s)- and agi-/age(g)- ‘come (here) and’ and ‘go (there) and’ respectively, or with lengthening as assi(s)- and aggi-/-agge(g)-, although the sequence a-ge-g-i- (inceptive with first-person agentive pronoun g(a)-) is commonly found as ag-ge-g-i or a-g-g-i-.

§8.2.2.2 Slot -2

Slot -2 or the deuteromeric prorrhizal slot houses pronouns. Pine has two distinct sets of pronouns, the agentive and unagentive pronominal sets. The agentivity of the pronominal sets always refers to the agentivity of the subject (objects have no agentivity per se as they do not act), whether it be suprapositional (overtly marked) or infrapositional (inferred). Each set of pronoun has a distinct present and past form. This means that four different sets of pronouns may appear in the deuteromeric prorrhizal slot. The table below shows the four different forms of the second, first and second-person pronouns.

Below are examples of verbal pronouns of the deuteromeric prorrhizal slot -2:

§8.2.2.3 Slot -1

Slot -1 or the proteromeric prorrhizal slot houses preverbs and immediately precedes the verb root. By far the most common preverb is the kinetic preverb -i(m)-. Preverbs are elements that come in between pronouns and the root, and they are conditionally obligate. The kinetic preverb has two main functions: to mark motion and to indicate ditransitivity. A less common preverv is -av(a)-, which indicates (temporal) ordering. Motion preverbs are required when certain kinetic initials are present. Commonly, the kinetic preverb -i(m)- is required to indicate movement, to indicate location (here or there), but it is generally absent with the simple kinetic initial inu-, and to indicate the promotion of an object to the role indirect object (ditransitive verbs), whether a direct pronoun is found or not (hemiditransitive and anthemiditransitive verbs)’.

A second, less common motion preverb is -ava-, which is found in the special case of the verb -ṡkv- ‘come in a specific order’ in combination with the trans- and cislocation: s-ava-ṡkv- ‘come/be first’ and g-ava-ṡk- ‘come/be second/then’, which is also possible to chain directly onto regular verbs instead, for example:

s-ava-ṡkvami kin įa tsƛįeḍḍisitṡis, g-ava-ṡkvami go įa tsatṡẻlġġisitṡis

or ṡkvami kin įa ts-ava-lįeḍḍisitṡis, ṡkvami go įa ts-ava-tṡẻlġġisitṡis

‘first we boil it, then we roast it’

In some instances, the preverb -ava- can be shortened to -av- and precede ditransitive preverbs, for example inu-g-av-im-alk-et-tit ‘first I go around giving it to her/him/them’.

The ditransitive preverb -i(m)- does not give any information about the relationship between the subject and the objects, only that the verb itself has an increased valency. Instead, ditransitive alignment is expressed through two different sets of suffixes. Retrograde markers are found in subslot 2d, indicating that the degree shown by the ditransitive paragogic final proceeds in the retrograde direction from the subject. Ditransitive paragogic finals are found in subslot 3c and show either a relative coordinate of the indirect object (1˚, 2˚ or 3˚), or an absolute indirect object (3o, 2 and 1).

There are two ditransitive retrograde markers (-ƛi-/-ƛen- and -nnu-/-nnon-), the former being the suprapositional retrograde marker (indirect object is suprapositional to the suprapositional pronoun, sometimes called ultrapositional) and the latter being an ipsigrade (retrograde) marker, indicating that the indirect object and subject are at a 0˚ degree position from one another (not the same agent but the same person).

§8.2.2.4 Slot 1

Slot 1 or the proteromeric metarrhizal slot is the first slot immediately following the root, and it houses medials. It is complex, meaning it contains three subslots numbered 1a,1b and 1c.

Medials are semantically obligate or facultative. In other words, they can add aspectual information or be required with certain verbs, altering the root’s meaning. Many medials are sensitive to tense, and they may have a present and past form. There are three categories of medials: voice medials housed in subslot 1a, derivative medials housed in subslot 1b and aspectual medials, housed in subslot 1c.

There are only two common types of voice medials: translative medials such as -ụƛ- or -ụθ-, and the passive voice medial -ϕụt=ụθ- (along with its augments: -ϕ(ụss)=iv=ụθ- ‘-able’ and -ϕ(ụt)=oġ=ụθ- ‘un-Ø-ed, un-Ø-able’, where the initial part of the medial -ụt-/-ụss- can be dropped, but the initial archephonemic -ϕ- is present). The translative voice medials can sometimes be preceded by -(i)s- (which indicates an inchoative translative) or -(i)n- (which indicates an automatic translative), both are also elements of other medials, for example -isƛuθ- and -unguθ- . However, these elements are generally fused with the root, and as such stand somewhere between a vestigial premedial and a root augment. Both are semantically obligate. The passive voice medial -ut=u- can also be found in special augmented forms in -ϕ(uss)=iv=uθ- (‘can be Ø-ed), -ϕ(ut)=oġ=uθ- (‘un-Ø-ed’ or ‘un-Ø-able’) to form verbs in -ϕ(uss)ivuθ- and -ϕ(ut)oġuθ-, which indicate the possibility or impossibility of a passive respectively, for example mikv-utuθ- ‘be eaten’, mikv-(uss)ivuθ- ‘be eadible’ and mikv-(ut)oġuθ- ‘be inedible, be uneaten’. In a few cases, the translative voice medial can also be combined with -iv- and -oġ-.

Derivative medials are used to form new verbs from other roots, be they nouns, adverbs, adjectives or even pronouns, and a such they are semantically obligate. Some derivative medials cause phonological changes in the root, either diaphony, rhizal apophony or both. For example, the derivative medial -ỉg- is a transivitiser for non-verbs (causing diaphony) – from the noun vabna ‘smoke’, the verb root vapṇ-ỉg- ‘smoke (food/skin)’ is formed. The reason these elements are considered medials not an integral part of the root is that some are sensitive to tense, and only the nucleus of a root is affected by rhizal apophony (hence its name) in the past. One of the most common derivative medial is -isƛụθ-, which forms a so-called unagentive automatic verb (often corresponds to English ‘get Ø-ed’). It is closely related to the translative -ụƛ-, and in fact was originally a combination of -ụƛ-, -is-, a vestigial inchoative marker and -ụθ-, a stative medial.

Aspectual medials are a rather large and varied set of affixes, many of which are polysemantic, and can be both semantically obligate or more commonly facultative. Their function is to give information on the aspect of the verb, for example making the verb semelfactive (-iḷḷ-, -iθ-, -iġ-, -isk-, -itṇ-, indicating various manners of short duration, a single quick or sudden occurrence as in aḷḷ-iḷḷ- ‘change quickly’, kauk-isk- ‘strike once’, tảį-itṇ- ‘sneeze’, sask-iθ- ‘wake up quickly’), modulative (-ib-, -iṡ-, -ir/-ur-, indicating endearment, diminutive or augmentative aspects, tảį-ib- ‘sneeze (once, a little, cutely)’, tảį-ir- ‘sneeze loudly’, paį-ur- ‘for something large to fall’, ḷủl-iṡ- ‘speak (in an endearing/diminutive way)’, durative (-ulġ-, -ilġ-, indicating a longer duration as in migv-ilġ- ‘eat a meal, feast’, hįulk-ulġ- ‘fall/snow/rain for a long time’), irresultative (-its-, indicating that the action does not reach an end point as in aḷḷ-its- ‘change over time, slowly, shift’, doll-its- ‘continue, go on’) or frequentative (-aƛƛ-, -add-, -aṡt-, indicating a repeated or recurring action, an activity or that something happen casually, habitually or is distributive, as in tselkv-aƛƛ- ‘camp, tent’, ỉtủl-add- ‘ponder, reflect’, bevved-add- ‘decorate many things’ ).

§8.2.2.5 Slot 2

Slot 2 or the thematic slot is the most central and complex slot of the Pine verb. It houses the theme, which is a combination of obligate inflectional affixes that house morphemes related to order, mood set, evidentiality, mood, tense and alignment and more. The theme is made up of various elements, and although it is said to be housed in a medial slot, thematic elements can and do appear as finals. The theme can be thought of as the core of the verb, whose elements appear in the thematic slot, but which can also appear as finals. As such, thematic elements are not purely defined by their location inside the verb, but by their location and their morphosyntactic role. Under certain conditions, some thematic elements occur in the thematic slot, and under other conditions, they may appear both in the thematic slot of in final position. Thematic elements founds in the thematic slot are called nucleic elements. Thematic elements found outside the thematic slot are called non-nucleic thematic elements.

The thematic slot itself is composed of multiple subslots and exhibits mesotomy in polyvalent verbs, which is the splitting and shifting of elements between the subslots or to a final position. At the core of the thematic slot is the nucleic theme. The nucleic theme is usually a vowel specific to an inflection, and is separate from other parts of the theme such as -v- versioners, plurality markers or alignment markers. A distinguishing feature of the nucleic theme is that it excludes any finals. For example, in the inflected form tsṇegimmuts ‘I want to see you’, the theme is -imm-u-t-s. The nucleic theme is -i-, the retrograde marker is -imm-Ø-t, which itself is composed of the medial -imm- and the final -t. The thematic slot houses both the retrograde marker -imm- and the nucleic theme -i- (and it can be said that its underlying form also houses the optative marker -s-). Likewise, the optative -s- here is normally a medial, but it has migrated to the final position through a process called sigmatropic migration, and as such cannot be considered to be part of the nucleic theme, as it can only appear in medial position.

§8.2.2.6 Slot 3

Slot 3 or the teleutameric metarrhizal slot houses finals. Finals are suffixes which are by definition the last elements of an inflected verb. The teleutameric metarrhizal slot is composed of three subslots (termed 3a, 3b and 3c) belonging to two categories.

The first final subslot houses progogic finals. There are three types of progogic finals – three obviate and one inclusive progogic final. Obviate progogic finals exclusively found in polyvalent verbs and are the obviate markers -mi- and -ni(n)-, which respectively indicate a prograde third-person obviate infrapositional object to a local subject and a retrograde third-person obviate infrapositional subject to a local object. In other words, if a verb has either -mi- or -ni(n) as its progogic final, the third-person obviate and a second or first person are involved. The inclusive progogic final include -tṡi-, which is the first-person plural inclusive pluraliser (you and me, we) of intransitive verbs. The intransitive obviate marker -mi- also appears in the progogic slot.

Paragogic finals those that appear in subslots 3b, 3c and 3d. There are two types of paragogic finals: non-polyvalent (3a and 3b) and polyvalent paragogic finals (3c). Non-polyvalent paragogic finals include the past marker -n, three non-polyvalent pluralisers (which themselves have an internal hierarchy and cannot co-occur), and the sigmatropic optative -s- (the optative element -s- which moves to the paragogic position through a phenomenon called sigmatropic migration). Polyvalent paragogic finals are polyvalent pluralisers, the very last elements of transitive and ditransitive verbs.

§8.2.2.7 Verbal Clitics

Verbal clitics are clitics which directly follow the verb, without being part of it. They are hyphenated to the verb. There is a number of clitics, but perhaps the most common is the particle -(d)di which indicates a probability when combined with the dubitative, for example ƛullun di ‘s/he’s probably leaving’. A word cannot be hyphenated on both ends – instead, only the initial hyphen is used.

§8.2.3 Verbal Pronouns

Verbal pronouns are elements of the verb that indicate one participant. They occupy the deuteromeric prorrhizal slot, and are obligate. Pine only ever overtly marks one person through its pronominal set. In the case of intransitive verb, the only participant of the verb will be housed in this slot. Polyvalent verbs similarly only overtly mark one pronoun in the deuteromeric prorrhizal slot, while the rest of the arguments are encoded through other means. The deuteromeric prorrhizal slot is one of two obligate slots of Pine verbs along with the thematic slot. Some pronouns, namely the third-person animate and inanimate, may have no overt surface form, meaning the slot appears to be unfilled.

Pine verbs have two pairs of pronominal sets: one agentive and one unagentive pronominal set. Each pronominal set has a present (found in the present, future and perfect, the present pronominal set) and past form (found in the past and pluperfect, the past pronominal set), which gives rise to four distinct pronominal sets.

Pine has six verbal pronouns. Pronouns are organised according to a hierarchy, which is extremely important in polyvalent verbs, but which does not affect monovalent verbs, given here in their hierarchical order. Pronouns are either local (the second and first person) or non-local (the third-person animate, fourth person, third-person inanimate and the third-person obviate).

All pronouns except the third-person obviate can be found in the plural, and there is a distinction between an inclusive first person plural (you (and others) and me) and an exclusive first person plural (I and others, excluding you). Pronoun pluralization occurs through different mechanisms in monovalent verbs and in polyvalent verbs. Additionally, the non-local pronouns are sensitive to animacy. Animacy is not a very a salient feature of Pine, but speakers do distinguish animacy when it comes to the choice of a third-person pronoun. People are obligate animates, and many living things such as animals, plants, even natural phenomenon, can be referred to with an animate pronoun. Generally speaking however, the third-person animate mostly refers to people or animals. The third-person inanimate exclusively refers to non-persons, whether living organisms or not.

§8.2.3.1 The Fourth Person

The fourth person (m(i)- irrespective of tense or agentivity) is a pronoun which refers to an indefinite person or pronoun. Generally if the fourth person is agentive, it will refer to ‘someone’ or ‘one’, or can be equivalent to a sort of impersonal or more general ‘we’ (much like ‘one’ can be used for a more impersonal ‘we’ in English). When unagentive, or the object of a polyvalent verb, the fourth person can equally refer to ‘someone’, ‘one’ or ‘something’. In addition, indefinite pronouns (daddas, ỏhakammi ‘someone’, gaggas, ỏhakiṡki ‘something’, gvaiddas, gvaiggas ‘all, everyone’) and interrogative pronouns (kammi ‘who’, kammet ‘who pl.’, kiṡki ‘what’) must be referred to with the fourth person.

§8.2.3.2 Obviation

Pine distinguishes two third person animate by function: the third person animate and the third person obviate. Obviation refers to the ranking of two third person animate arguments, where obviation demotes one third person to a secondary role, among other things.

The third-person obviate has three functions. Firstly, it distinguishes between two third-person animate pronouns within the same verb, for example negget ‘s/he saw her/him’. This means that two third-person animate pronouns in a verb cannot co-exist, and one of the two third-person animate arguments must be obviated – the obviated argument is most commonly the object. Its second role is very similar, but extends beyond a single verb and operates on a discourse level, where the third-person obviate refers to less salient animate participants. Obviation in Pine is used to distinguish two third-person animate participants of opposing salience. The obviate third person is non-salient compared to the third-person animate (sometimes called the third person proximate when contrasting the two third animate third persons). For example, in the phrase ‘he told him that he was sick’ in English, it is impossible to know distinguish between the two third persons, and as a result we cannot infer if the person doing the telling is sick, or if the person who is being told is sick. This problem is solved by obviation in Pine, as the distinction between a proximate and an obviate person can give rise to two different readings:

In the first example, without any further context, the person who tells is the person who is sick. Replacing pronouns with names may illustrate this better: John told Peter that he [John] was sick. In the second example, the person who is being told is sick: John told Peter that he [Peter] was sick. Obviation in Pine is context dependent – whatever third-person pronoun is less central to the utterance can appear in the obviate. One and the same argument may be obviate in one context but proximate in a different context.

Thirdly, obviation is used in prerelative clauses, where an antecedent is introduced in a clause and more information is given in a subordinate clause or later in the discourse, whether animate or inanimate. For example, the expression tsakoggimin tsỉn ỉ bỉlįu-t ba, namut tsapuiįįisinin ỉ, literally ‘you only get the berries that you picked’, the antecedent ỉ bỉlįu-t ba is obviated because it functions as a prerelative, which is also reflected in the form of the polypersonal agreements tsa-Ø-imin (2⟫3o). The antecedent must be obviated, but the precedent is not obviated in such prerelative obviate constructions.

As such, obviation serves the following three purposes:

1. To distinguish two third-person animate arguments involved in the same polypersonal verb

2. To distinguish saliency between two third-person animate arguments on a discours level

3. To indicate prerelative antecedents

The third-person obviate is restricted with regard to pluralization. Independent pronouns and suprapositional pronominal markers are only found in the plural. Pronominal pluralization therefore only occurs through infrapositional pluralization markers such as -in and -an. This means that there is a discongruence of number in monotransitive verbs with a plural obviated subject – the referent noun may be pluralised, but the monovalent verb will not, and neither will a suprapositional polyvalent obviate pronoun.

In this example, obviation is found in the mit- restrictive relative pronoun muįįủnit ‘that…from’, which is not pluralised, as well as in the non-pluralised independent pronoun go (muįįủnit…sagvemit go ‘the men that…from/about’), where the preceding noun is pluralised. The negative auxiliary verb tủġekki ‘they do not…them (3a⟫3o)’ has an ambiplural marker -kki (-et-agi ⟫ -ekki), which shows that the third-person obviate is pluralised.

§8.2.4 Pronominal Sets

Pine has two sets of pronouns depending on the agentivity of the subject: the agentive and the unagentive pronominal set. The agentive pronominal sets appear when the subject of a verb has agentivity, regardless of the positionality of the pronoun – agentivity is always encoded in the suprapositional pronoun, and only subjects are affected by agentivity (as all patients are unagentive), which means that the agentivity of the suprapositional pronoun always reflects that of the subject, whether the suprapositional pronoun is the subject or the object. For intransitive verbs, an agentive subject will automatically require the agentive pronominal sets, but this is not true of polyvalent verbs. Likewise, the unagentive pronominal set is used whenever the subject (regardless of whether it is supra- or infrapositional) is unagentive.

The third person obviate and third person animate share the same pronominal prefix but are distinguished from one another through an obviate suffix in intransitive verbs.

The table below shows pronominal prefixes for the agentive set. Pluralization is marked through metarrhizal elements, which are valency specific, such that monovalent pluralisers differ from polyvalent pluralisers.

In its present form, the second-person pronoun tsa- will appear without a vowel (ts-) before certain consonants, for example tsṃihet ‘you like it’ (<*tsa-miḍ-et). The first-person pronouns ga- likewise appears in its short form g- before vowels, but it can also be dropped all together, which is very common in non-formal registers. In the past form, the second person does not cluster with the onset of a verb, but if a verb begins with a vowel or the pronoun is followed by one, then ts- and g- are used. In other words, the second-person pronoun contracts in the present but not the past.

Likewise, the third-person animate and inanimate only appear in certain environments. For the animate pronoun, it only appears before vowels and certain to consonants in the present, and is completely absent in the past. The third-person inanimate pronoun ḥ- only appears before vowels and does not distinguish tense forms. The fourth person is mi- before consonants and m- before vowels irrespective of tense or agentivity, and the third-person obviate is identical to the third person animate.

Here is the agentive set used with the conjugated verb ḷủl- ‘speak’ in the present and past indicative autonomous, and the unagentive set is illustrated with the verb sask- ‘wake up, be woken up’.

§8.2.4.1 Pronominal Mesoclisis

The present agentive pronominal set contains forms with a long and short form – the second and first person have both short and long forms, while the third-person animate and inanimate have has a short form. The third-person inanimate agentive has a short form only found before vowels, while the third-person animate short form can appear before vowels and certain consonants. In other words, short forms appear systematically before vowels, such as ts-oaṛva ‘you drive’, but they may also combine with the onset of verbs in the absence of proteromeric elements, as in ts-ṃigva ‘you eat’. The rules that describe how pronouns are prefixed to verbs are called pronominal mesoclisis.

§8.2.4.1.1 Second-Person Mesoclisis

The second-person agentive present pronoun ts- appears conditionally before consonants with which it can form a legal initial cluster. In the past, it is tse- before consonants and ts- before vowels. The table below shows what initial consonants require the short form of the second-person pronoun tsa-.

The present contracted form tsƛủlla ‘you speak’ therefore becomes tseḷủlin in the past (not *tsƛủlin). Verbs with an onset in v-, s- and h- can also be commonly found with the long form, especially in more formal language, such that tsaḍget ‘you swear’ is commonly found as tsahaḍget. The short form never occurs before hį-.

Some idiolects also allow further contractions, such as ts-k- / ts-g- ⟫ tsk-, ts-r- ⟫ tsġ- and even ts-t- / ts-d- ⟫ tst-, though these contractions are mostly restricted to familiar registers, or specific dialects. Western dialects for example tend favor contractions.

§8.2.4.1.2 First-Person Mesoclisis

The first-person pronouns do not form clusters with verb onsets like the second person. However, it is generally acceptable to drop the first-person pronoun if it is agentive. In some western dialects, the first person can form onset clusters, but only in familiar registers, with the following onsets:

Forms gm- and gn- have become somewhat more acceptable or common in eastern dialects in younger generations. Forms in gv- are acceptable across all dialects but tend to be more common amongst men.

The past agentive form ge- appears before consonants and g- before vowels. Likewise, the present unagentive forms are ṇe-/ṇ- and in the past they are ṇi-/ṇ-.

§8.2.4.1.3 Third-Person Animate & Obviate Mesoclisis

The third-person animate agentive and third person obviate, who both share the same pronouns, has two possible forms in present tenses. Before certain consonants and vowels, it has the form t-. In all other cases, it does not appear at all. In the past, the agentive form is always Ø-. In the unagentive, there is no tense distinction, and the pronoun has the form s- before certain consonants and vowels, otherwise Ø-. Before any element in slot -3b (the initials si(s)-, gi-/geg-, inu-/inug-, issu-/issug-, and iggu-/iggug-), the third-person animate and obviate are -Ø-, and the short forms of the initials are used without their vowel, becoming s-i-/g-i-. The third-person animate and obviate pronoun retains its surface form in the past with the negative auxiliary copula and the copula. For certain verbs, the third-person animate therefore has no surface form at all, regardless of agentivity of tense.

There are six conditions which trigger the appearance of one of the three forms in the present set:

Most speakers have the zero form Ø- before h-, such that sġallunguttan ‘s/he gets lost’ is most commonly heard as hallunguttan. The opposite is true of v-, which many speakers realise s-v- as s-

The table below summarises the mesoclitic rules of the third-person animate:

§8.2.4.1.4 Third-Person Inanimate Mesoclisis

The third-person inanimate has three forms which are not sensitive to tense: ḥ- or Ø- when agentive and h- or Ø- when unagentive. The agentive form ḥ- only occurs before vowels (and į- and v-, though this glides often result in zero-form), whereas Ø- occurs everywhere else. The unagentive form h- also occurs before vowels but also devoices consonants which can be devoiced (m-, v-, į- n-, nį-, l-, lį- and r-), and for most speakers, deletes an onset in s- to form h-. Like its animate counterpart, the third-person inanimate does not have a surface form after kinetic initials in the subslot 3b, and the cis- and translocative elements combine with the preverb -i- in their short form.

The table below summarises the mesoclitic rules of the third-person inanimate:

§8.2.4.1.5 Fourth-Person Mesoclisis

The fourth-person pronouns are similar in that they distinguish neither tense nor agentivity. They do differ in that the fourth person has a plural form but the third-person obviate is normally never pluralised, at least in the standard language. The fourth person appears as mi- before consonants and m- before vowels.

§8.2.4.2 Overview of Pronominal Sets

The table below outlines the unagentive and agentive pronominal sets.

§8.2.4.3 Impersonal Verbal Pronouns

Impersonal verbs monovalent are verbs with no subject or agent. Impersonal verbs have no pronominal prefixed but have a non-local form, generally without any -n in non-past tenses. This is equivalent to a third person inanimate without a pronoun and without the -mi marker.

§8.2.5 Pronominal Pluralization

Pluralization of verbal pronouns in Pine is marked through finals. These elements appear as the last element of a verb and give show that a participant is plural. Because intransitive verbs only have a subject, monovalent pluralisers are fairly uncomplicated. Monovalent pluralization is localised, such that local and non-local pronouns have different pluralisers. In addition, pluralization of intransitive verbs also involves the alternating presence or absence of the locality marker -v-. Polyvalent pluralization differs from monovalent pluralization in that it is not localised, and instead gives information about the position of the pluralised participants.

§8.2.5.1 Monovalent Pluralization

There are two sets of pluralization markers for all monovalent pronouns, both agentive and unagentive. One is found in the autonomous order and the other in the hyponomous order. Pluralization of pronouns is localised, meaning one set of ending occurs with local pronouns (2 and 1), and a second set of ending occurs with non-local pronouns (3a, 4, 3i, the third-person obviate has no plural form). These plural markers are mesotomic. Pluralization will yield for other telethetic markers, such as the past marker -n (the sequence -t-n is not allowed, and has the surface form -n).

In the autonomous order, local pronouns have a plural marker in -(V)n (either -on/-un or -en/-in, harmonizing with the alignment of the previous vowel) while the non-local pronouns gain a -(V)t (either -ot/-ut/-at or -et/-it, harmonizing with the alignment of the previous vowel) or in the case of the third-person inanimate, -a=t. Note that unlike other plural finals, the third person undergoes sigmatropic migration, whereby the form -a-t (progogic -a- and paragogic -t) becomes -at-s (progogic -at- and paragogic -s).

In the hyponomous order, local pronouns have a plural marker in -ot/-ut/-at or -et/-it, while non-local pronouns have a plural in -(V)s (either -us/-as or -is), harmonizing with the alignment of the previous vowel). Animacy is not conserved in the plural of the hyponomous order, which means that the third-person animate and inanimate plural is identical.

The plural marker’s vowel does not appear at all if the theme of the verb a diphthong:

The first person plural inclusive appears with the second-person pronoun and the element -tṡi- inserted between the vowel and the plural marker, while the first person plural exclusive lacks this element and appears with the first-person pronoun.

Pluralization can thus be summarised as follows:

§8.2.5.1.1 Monovalent Locality & Pluralization

Monovalent verbs have comarkers of both locality, pluralization and tense. The -v- versioner is a thematic element which sometimes co-occur depending on the locality, number and tense of a specific inflected form. For example, in the autonomous indicative, the -v- versioner is found in local singular pronouns in the present, but in non-local singular and plural pronouns in the past. As such, simply knowing which monovalent pluraliser to use is not enough to grammatically encode plurality or tense in monovalent verbs – one also has to know in which case the -v- versioner is required for each tense and for which pronouns. In polyvalent verbs, -v- versioners do not comark plurality, but instead mostly depend on tense.

but

§8.2.5.2 Polyvalent or Positional Pluralization

Polyvalent verbs (verbs with a valency of two or more) have so-called positional pluralization, where the pluraliser gives informational about whether it applies to a suprapositional pronoun, infrapositional pronoun or both (called ambigrade). Unlike monovalent pluralisers, polyvalent pluralisers are not localised, instead they only give information about the number of the participants and their position. There is a total of six positional pluralisers, though it can also be described as three pluralisers that can also appear with the first-person inclusive pluraliser -tṡi-: the suprapositional pluraliser -s, the infrapositional pluraliser -n and the ambigrade pluraliser -gi. Like pronouns, positional pluralisers have a present form (found in the present, future and perfect) and a past form (found in the past and pluperfect). The present form appears with an -a- if the precending element ends in a consonant and the past form appears with an -i- if preceded by a consonant:

The six polyvalent pluralisers are:

suprapositional plural

Only the suprapositional pronoun is plural.

inclusive suprapositional plural

The first person inclusive plural is suprapositional and the infrapositional pronoun is singular.

infrapositional plural

Only the infrapositional pronoun is plural.

inclusive infrapositional plural

The first person inclusive plural is infrapositional and the suprapositional pronoun is singular.

ambigrade plural

Both supra- and infrapositional pronouns are plural.

inclusive ambigrade plural

Both participants are plural and there is a first person inclusive plural involved.

Polyvalent pluralization can be summarised in the table below:

§8.2.6 The Theme

The theme is the core inflectional element of a verb. Thematic elements are combined to inflect a verb, in addition to pronominal sets, plurality markers and other markers. The theme is very sensitive to valency, and for that reason, monovalent and polyvalent themes must be considered separately.

§8.2.6.1 The Monovalent Theme

The theme of intransitive and impersonal verbs is significantly less complex than that of transitive and ditransitive verbs, which is not to say that it is simple. The monovalent theme is composed of elements found in various subslots of the thematic and/or teleutameric metarrhizal slot, whose elements can shift and move between slots. The exact configuration fo the theme is dependent on many factors. One element may appear in a specific slot under certain conditions, but the presence of a second element may shift the first element to a different position. For that reason, the monovalent theme is best described according to its order. Elements common to both autonomous and hyponomous intransitive verbs are described below.

§8.2.6.1.1 -v- Versioner

The -v- versioner in monovalent verbs is involved in contrasting locality and number across mood, tense in the autonomous order for transitive verbs and in both orders in intransitive verbs. It is called a locality and number staggered comarker, as it does not have a meaning in and of itself, but is dependent on the co-occurrence or absence of other markers, and conditionally appears or lacks in opposing patterns. In the indicative autonomous present, it indicates a singular local pronoun, but in the indicative autonomous past it indicates a non-local person, both plural and singular. It also appears in transitive and ditransitive verbs but serves a different purpose with a similarly staggered or contrasting pattern. The -v- versioner does not always appear as is, but instead may coalesce with a presending consonant resulting in a geminate.

§8.2.6.1.1.1 -v- Versioner Coalescence

One characteristics of the -v- versioner is that it coalesces with a number of preceding consonants or consonant clusters (either the root onset or a medial) in predictable ways. The result may be a geminate (-n-v- ⟫ -nn-), a labialised geminate (-ƛƛ-v- ⟫ -ƛv-), a labialised lenis cluster or a delabialised fortis cluster (-ld-v- ⟫ -ldv- or -ltt-) or a labialised fortis cluster (-mb-v- ⟫ -mpp-). Labials form a special category as they only ever become geminate (-bb-) or fortis when part of a cluster (-cpp-). A number of clusters are unaffected by the v-versioner, and other clusters cannot be labialised.

Lenis clusters in l-, r- and n- excluding those containing a labial (-ld-, -lg-, -rd-, -rg-, -nd-) have two alternative form, either labialised and lenis or delabialised and fortis (-ltt-, -lkk-, -rtt-, -rkk-, -ntt-), which are more formal. In some contexts, fortis forms are preferred, such as in medials (the affective medial -ir-, the opinionative medial -end- and the iterative medial -ild-) or when the verb root already contains -v-.

The consonants that coalesce with the v-versioner are listed below and grouped by the outcome of the coalescence.

Clusters that are unaffected by the v-versioner are:

All other clusters delete the v-versioner all together. This includes all clusters containing a labial, nasal, -s- or -ṡ-, for example lidn- ‘say’ ⟫ galidna ‘I say’.

These rules apply equally to verbs as they do to possessive suffixes of local pronouns on nouns and postpositions, with some exceptions, namely that case endings in -t combine with -v- to form -vv-, as in bakvet ‘home (illat)’ but ṇibakvedda ‘to my home’, not *ṇibakvetva as would be expected if it was a verb.

§8.2.6.1.2 -ϕ- Versioner

The -ϕ- versioner is similar in function to the -v- versioner, it functions as a staggered comarker and occurs in various paradigms of verbs. For example, the connegatives -ϕit/-ϕat/-ϕụt, the passive -ϕutuθ- and all endings of the conditional and potential regardless of order or tense. It is also found in the agent marker -ϕi, for example telg- 'write' and telkki 'writer'. The -ϕ- versioner causes fortition of the preceding consonant or consonant cluster.

§8.2.6.1.3 Perfectives

Perfectives are infixes that appear within the theme of the monovalent verb and indicate that the verb is in the perfect or pluperfect tense. For example, the autonomous indicative perfective is -ung-/-ing-, while the optative intransitive perfective is -ut-. There is also the pluperfective -an- of transitive and intransitive verbs.

Note that the perfective marker for the optative autonomous can combine with optative -s-, but the latter is subject to sigmatropic migration in the third-person inanimate plural:

§8.2.6.1.4 Nucleic Theme

The nucleic theme carries most of the inflectional information, and as such is sensitive to order, mood set, evidentiality, mood, tense and locality. For example, the nucleic theme is -a- for the indicative autonomous present, -i- for the indicative autonomous past, and -us- for the indicative optative autonomous:

§8.2.6.1.5 Locality Markers

Locality markers are fusional elements that are required by the of a singular local subject in the autonomous optative and potential. They are a type of local augment found on the nucleic theme, and have five forms:

The reason for locality markers being considered separately from the theme is that the optative perfect theme -ut- causes the appearance of the optative -s- before the locality marker (-ut-s-ut) in singular. Their absence in the hyponomous likewise causes digammatropic migration of the -v- versioner from subslot 2a to 2d, fusing with the nucleic theme resulting in gemination.

§8.2.6.1.6 Teleutameric Finals

Teleutameric finals are the last elements of the monovalent verb. Many teleutameric finals have multiple functions, such as the final -n, which can indicate a local plural (autonomous local pluraliser), or a local singular (as in the autonomous indicative present). It is also found in the past tenses of singular non-local pronouns and plural local pronouns. Many of the teleutameric finals have an alternating distribution, being found in one case and absent in the opposite case. The varied function and distribution of teleutameric finals makes a detailed description in isolation from other elements of the intransitive verb futile. However, a rough overview of their form and function is given below.

§8.2.6.2 Polyvalent Theme

Like its monovalent counterpart, the polyvalent theme is made up of various elements with internal order. However, because polyvalent verbs must intrinsically encode more information due to their higher valency, the polyvalent theme is significantly more complex. Although they both encode the same type of information (order, mood set, mood, tense, valency and alignment as well as tense), the morphology of polyvalent themes is markedly different from that of monovalent themes. Transitive and ditransitive verbs do not differ greatly in terms of their theme – ditransitive verbs have extra grade markers, but otherwise their theme is identical.

One difference between mono- and polyvalent themes is that finals in polyvalent verbs are exclusively pluralization markers, where as finals and theme elements are less well delimited in monovalent verbs.

Two important phenomena affect the polyvalent theme, which are termed mesotomy and sigmatropic migration. Mesotomy describes the splitting up of thematic elements, while sigmatropic migration describes the movement of an element -s- from within the theme to a final position.

§8.2.6.2.1 Polyvalent Thematic Elements & Order: Autonomous Order

The polyvalent theme has here seven subslots in the autonomous thematic slot, numbered 2a through 2g. At minimum, only one subslot is occupied (the nucleic theme in subslot 2c), for example neg-i ‘it₁ sees it₂’, and at most five are occupied, for example m-im-alk-imm-u-ssi-t-s ‘someone will want to give something to someone. Transitive verbs lack the subslot 2d all together, and prograde verbs also lack subslot 2b.

Because the irrealis mood set has mesotomy, the thematic slots in the autonomous order are structurally more like the hyponomous theme.

§8.2.6.2.2.1 Versioners

Like monovalent themes, the polyvalent autonomous theme can contain a -v- or -ϕ- versioner, but their role differs markedly from the role of the monovalent -v-/-ϕ- versioner. In the autonomous, the versioners takes part in comarking the past tense in the realis mood set, and as such is not sensitive to number, as number in polyvalent verbs is exclusively marked through pluraliser finals.

§8.2.6.2.2.2 Perfectives

As in monovalent verbs, autonomous polyvalent verbs also have a perfective subslot. There are three perfective markers in the autonomous, which are only found in the prograde alignment, as the retrograde alignment has had their retrograde markers and perfective marker fuse together. The indicative prograde perfective marker is -ung-/-ing- as in intransitive verbs, but the retrograde perfective is -umm-. The pluperfective -an- is found in the dubitative exclusively, as the optative uses the -atts- (originally -an-s- ⟫ -antts- ⟫ -atts-). The table below summarises the different perfectives of polyvalent verbs. For the autonomous retrograde forms, the first form listed is for the 0˚, 1˚, 2˚ and 4˚ while the second form is in the 3˚. For the hyponomous retrograde forms, the first perfective is for the 1˚, 2˚ and 4˚ while the second form below is for the 0˚ and 3˚.

§8.2.6.2.2.3 Retrograde Markers

Retrograde markers of polyvalent autonomous verbs indicate that the direction of the subject-object relation is opposite to that of the pronominal hierarchy. There are two types of retrograde markers in polyvalent verbs, which do not co-occur and as such are mutually exclusive. Transitive retrograde markers exclusively indicate that the suprapositional pronoun is not the subject but rather the subject. These transitive retrograde markers are sensitive to inflection. The second type of retrograde markers is only found in ditransitive verbs, and their function is to indicate that the coordinate alignment between the indirect object and the subject occurs in the opposite direction to that of the pronominal hierarchy. Whereas there are many forms of each retrograde marker depending on the specific inflection of the verb, ditransitive retrograde markers are not sensitive to inflection, and remain the same throughout the conjugation pattern of ditransitive verbs. There are only two sets of ditransitive retrograde markers. The ditransitive retrograde marker is -kin-, and the ditransitive ipsigrade marker is -nnun-. The ditransitive ipsigrade marker indicates that the indirect and subject are one and the same gramamtical person (but not necessarily referent), and as such are used for ditransitively reflexive verbs.

§8.2.6.2.2.4 Nucleic Theme

The nucleic theme in polyvalent verbs serves much the same purpose as that of monovalent verbs – it serves as the core inflectional element of the theme. The polyvalent nucleic theme, like its monovalent equivalent, is composed of numerous elements that are sensitive to tense, order and mood, but most importantly to degree and grade. It is the nucleic theme that carries most of the coordinate information of polyvalent verbs – the distance and direction of the distance to from the supra- to the infrapositional pronoun. In the irrealis mood set, the nucleic theme is fused with an irrealis mood marker, for example -i- (1˚ prograde) becomes -ϕattṡ=i- in the conditional. The two irrealis moods are always found with the -ϕ- versioner in the autonomous order, whereas the realis moods has no versioner in the present, and the -v- versioner in the past autonomous. All moods and tenses have a -ϕ- versioner in the hyponomous. Therefore, all hyponomous forms of verbs are fortified before the theme. Likewise, hyponomous past forms end in a paragogic -n, whereas hyponomous present forms end in a paragogic -t (which attracts the optative -s and future -ṡ). However, because a paragogic final in -n precludes the optative and future markers (-s-/-ṡ-) from appearing, the optative and dubitative hyponomous past forms are not distinguished formally, but forms in -s do exist and are common in informal registers. The same is true of the third degree hyponomous marker -isi-, which does not distinguish any mood in the present within the realis mood set: the past hyponomous indicative, dubitative and optative are identical, while the present indicative and dubitative are also identical (the optative is distinguished by the presence of -s-). Some speakers combine the optative -s- with a past final -n to form -s instead of the normal -n, but this is considered a form of hypercorrection, but it may be considered acceptable to distinguish the dubitative/indicative from the optative if necessary for disambiguation.

In the autonomous indicative, the nucleic theme forms are -i- (1˚), -a- (2˚) and -e=t (3˚), with one additional form found in the zeroeth degree, which in the autonomous is the 2˚ nucleic theme augmented by a marker final (-t), which in the retrograde co-occurs with the 1˚ retrograde marker (-emm-a=t), but in the hyponomous it is the 2˚ nucleic theme with a 2˚ retrograde marker (-i=dd=u=t).

If only true nucleic themes are considered (1˚ and 2˚), there are five possible autonomous nucleic themes: -i- or-e- and -a- or -o- or -u-.

The table below summarises the nucleic theme of polyvalent verbs by mood, tense and degree in the prograde.

The table below summarises the nucleic theme of polyvalent verbs by mood, tense and degree in the retrograde:

§8.2.6.2.2.5 Obviatives & Marker Finals

Obviatives are two elements (-mi and -ni=n) that indicate a third-person obviate participant, which exhibit splitting. The final -n of the obviatives appears conditionally based on the type and presence of absence of marker finals.

Marker finals are elements of the theme that are involved in comarking coordinate position. There are two such marker finals in polyvalent verbs: -n, and -t. The first marker -n is involved in comarking the past tense, and appears conditionally in the absence of a -t marker in the autonomous, but always in the hyponomous. The second marker final is called the distal marker final (or simply distal marker) is involved in indicating a 0˚ and 3˚ alignment, which can sometimes cause a change in the nucleic theme – combining 1˚ nucleic theme with the -t marker final does not yield *-i-t in the indicative autonomous but instead -e=t (this is not true of the inferential however, where it is -i-t).

The first obviate marker -mi is used with a local suprapositional subject and a third-person obviate object (meaning it is a prograde comarker) and is added onto the 1˚ nucleic theme, for example ga-neg-i-mi ‘I see her/him₂’. The second obviate marker -ni=n is found in the retrograde, and indicates the opposite alignment to -mi, meaning a local suprapositional object and a third-person obviate subject, and also requires the 1˚ nucleic theme and 1˚ retrograde marker (-em-i-), for example ga-neg-em-i-nin ’s/he₂ sees me’. The obviate markers’ final -n does not always appear, if for example a non-past optative marker -s- appears, it supersedes the final -n to give -mi-s and -ni-s.

§8.2.6.2.2.6 Future & Optative Markers

The last element of the autonomous polyvalent theme are the future (-ṡ-) and optative (-s-) markers, which as their names indicate are markers of the future and optative respectively. They are sensitive to sigmatropic migration, which is a phenomenon whereby these markers migrate from their normal position in polyvalent verbs (2g) to the paragogic final position in the presence of a paragogic -t. For example, g-ỉlis-es ‘I want to meet her/him’ its optative -s- in the thematic subslot 2g, but should a hyponomous marker final -t appear in the hyponomous, the sequence becomes reversed (-s-t ⟫ -t-s) to įa g-ỉliss-e-t-s ‘that I want to meet her/him’, in which case the marker final is now in the thematic subslot 2g and the optative -s- occupies the paragogic slot.

§8.2.6.2.2.7 Marker Final & Optative Marker Hierarchy

The optative marker -s- does not always have a surface form. There is a three-tier hierarchy of theme-final elements (which include marker finals in -t and -n and the optative marker -s in the case of autonomous polyvalent verbs), which dictates that -t supersedes -n, which in turn supersedes -s:

-t ⟫ -n ⟫ -s

This means that whenever the past marker final -n should occur, but a marker final in -t would also be present, then the -n marker does not appear. However, the presence of a marker final in -t can co-occur with the optative marker -s-. This means that in the absence of a marker final in -t and in the presence of a past marker final -n, the optative -s- simply does not appear. For past optative verbs, this precludes the optative marker -s- from appearing in the 1˚, 2˚ and 4˚ retrograde, but it does appear in the 0˚ and 3˚. It is common for the optative marker to appear instead of an -n, but not in standard language.

Note that the pluraliser final -s is extremely commonly found to delete a final -n. The distinction between the behavior of the optative -s and the plural -s is mostly a issue of register. In normal and standard registers, the pluraliser -s can supersede -n, but the optative only does so in informal registers.

§8.2.7 Initials

Initials are the very first elements of the Pine verb, and are housed in the tritomeric prorrhizal slot -3, which has two subslots (-3a and -3b). There are two types of initials, called adverbial and kinetic initials. Kinetic initials are fewer and may be conditionally obligate in many cases, and include true kinetic initials (which indicate movement or location) and ditransitive preverbs (which indicate the direct object of a ditransitive verb). In contrast, adverbial initials are more varied, and constitute a more open class, and are never conditionally obligate, but may me semantically obligate, meaning the presence of an adverbial initial may change the meaning of a verb entirely, and to maintain that meaning, the adverbial initial is obligate.

§8.2.7.1 Adverbial Initials

Adverbial initials can be thought of a adverbs that have been incorporated into the verb, and are found in subslot -3a. They mostly add information about the manner in which the action takes place. Unlike kinetic initials, adverbial initials do not trigger the presence of preverbs. Adverbials do not affect the form of the third-person pronouns, while kinetic initials do. For example, tsṇegemat 'I see you’ changes its meaning when found with the adverbial initial iss(i)- ‘well, right’: issitsṇegemat ‘I like/understand you’. The most common adverbial initials are as follows:

The preverb uṡ(ṡa)- (uṡ- before consonants with which it can form a legal cluster, often causing devoicing of initial voiced stops and uṡṡa- before other consonants) or uṡṡ- (before vowels) indicates that an action is not performed properly (uṡ-ƛủbm- ‘speak incorrectly, mispronounce, lie’), correctly or successfully or that it happens poorly or somehow has a negative outcome (uṡ-nįikk- ‘steal’ cf nįikk- ‘take’). Note that the sequence uṡ-g(a/e)- is realised as uṡk(a/e)-.

This preverb is opposite to iss(i)/issi(t-)- in that it indicates that an action is performed correctly (issi-tem- ‘go well’ cf. tem- ‘flow’ or issi-siġv- ‘understand, confirm’ or ‘hear well’ cf siġv- ‘hear’), in the right or proper way (issi-lủbm- ‘speak the truth/correctly’ or ‘pronounce’ cf lủbm- ‘speak’), easily or well (iss-alk- ‘look good/promising’ cf alk- ‘give’). The preverb issi- combines with the second-person pronoun as issit-ts(a/e)- and -issip-p(i)- and with the first person as issik-k(a/e)- and issit-ṇ(e/i)-.

The preverb keu-/kem- indicates that an action is performed accordingly (keu-min- ‘work, function correctly’ cf. min- ‘walk’) or is consequencial (keu-nen- ‘follow, agree’ cf. nen- ‘sit’), and can be added to plural forms to show that participants perform the action together (keu-lủbm-its- ‘discuss, debate’ cf. lủbm-its- ‘discuss, small talk, debate’).

The preverb iḷḷu(l)- indicates that an action is performed separately (iḷḷul-aḷḷ- ‘become different from’), not accordingly (iḷḷu-nen- ‘disagree), or in a way that results in discrete, separate or scattered objects (iḷḷu-marƛ-add- ‘cut into pieces’). It is also found in some cases where it means ‘of others’

The preverb ỉ-, found as ỉts- before vowels, shows that an action was succesfully (ỉ-lukk- ‘succeed’ cf. lukk- ‘pierce, puncture’), happened all the way (ỉ-tsůḷḷ- ‘finish, complete’ cf. tsůḷḷ- ‘carry’), despite struggles (ỉ-tann- ‘remain, stay’ cf. tann- ‘stand, wait’) or may simply emphasise the degree at which the action happens (ỉ-giḍḍ- ‘completely forget’ cf. giḍḍ- ‘forget’), and as such is often found where English may use adverbs such as ‘very’ or ‘a lot’ (ỉ-siġġ- ‘rain a lot’ cf. siġġ- ‘rain’). It is one of the most ubiquitous adverbial initials.

The preverb ỉl-, found as ỉlda- before vowels, shows that an action proceeds forward (ỉl-did- ‘step forward, volunteer’ cf. did- ‘step’), ahead (ỉl-neguƛ- ‘is visible ahead’ cf. neguƛ- ‘is visible’), ahead of time (ỉlda-radn- ‘finish early’ cf. radn- ‘come to an end’), and it is found in many verbs that relate to offering or giving (ỉl-mieg- ‘present, offer’ and ỉl-g-i-mieg- ‘explain’ cf. mieg- ‘tell, inform’). It can appear as ỉlt- before ṇ-, s-, ṡ- and s-ṇ- and ỉlp- before ṃ-, and the combination ỉl-h- is often found as ỉḷḷ-

Verbs found with the preverb kỉl(a)- denote that an action happens again (kỉl-tadniθ- ‘stand up again’ cf. tadniθ- ‘stand up’) or still (kỉl-neguƛ- ‘still visible’ cf neguƛ- ‘visible’) but it may also modify the meaning of the verb as well (kỉl-iks- ‘try again’ cf. iks- ‘shoot an arrow’). It can appear as kỉlt- before ṇ-, s-, ṡ- and kỉlp- before ṃ-.

The preverb giḷḷi(t)- usually indicates that an action occurs before (giḷḷit-tem- ‘precede’ cf tem- ‘flow’), in advance (giḷḷin-neg- ‘predict’ cf. neg- ‘see’) or first (giḷḷim-min- ‘lead, walk first’ cf. min- ‘walk’ or giḷḷin-naġv- ‘first born’ cf. naġv- ‘be born’). It causes the following consonant to be geminate, and in the case of the first person ga-/ge-, also voiceless (giḷḷikka-/giḷḷikke-), but with ṇe-/ṇi-, the usual form is giltṇe-/giltṇi-, as well as with si(s)- and gi-/geg- as giltsi(s)- and gilkki-/gilkkeg-.

The preverb eġo-/eġ- (also found as eġġo-/eġġ-) indicates that an action occurs after something (eġ(o)-tem- ‘follow, come from’ cf. tem- ‘flow’) or last (eġ-aḷḷ- ‘be left, survive’ cf. aḷḷ- ‘become’) or behind something (eġ-ỏk- ‘regret’ cf. ỏk- ‘watch, look’).

The preverb ůḷḷ(ů)-/uḷḷ(u)- indicates that the action does not quite happen but very nearly so (ůḷḷů-paitṇ- ‘nearly fall’ cf. paitṇ- ‘fall’), barely so (when negated, u-Ø ůḷḷů-kvỉ- ‘barely feel’ cf. kvỉ- ‘feel’) or that the action is just about to happen (uḷḷu-nakṇuƛ- ‘about to break’). In some cases it appears as ůl- (usually only before t- and k-).

The preverb akn(a)- indicates that the subject assists an object in performing the action of the verb. It makes transitive verbs from intransitive verbs. Transitive verbs are rarely found with this preverb, but they may be found in some cases, where the object of the original verb is found as the direct object while the subject becomes the indirect object. Such assistive verbs generally require an animate or agentive subject.

The preverb gaġ(va)- indicates that the action occurs cyclically, circularly, ends and begins again, or may also indicate natural cycles. It is preferred to kỉl- with many verbs depicting natural cycles. For example, gaġ-ảb- ‘born again, be reincarnated’, gaġ-nįav- ‘reuse, recycle, cycle through use’, gaġva-tem- ‘reoccur (cyclically), come around’, gaġva-min- ‘patrol, guard’, gaġ-ell- ‘regrow, be perennial’, gaġ-tủl-aƛƛ- ‘obsess’, gaġ-ỉlpp-eḍ- ‘bloom again’ and gaġva-sask- ‘reawaken, come back to life’. It is found as gaġ- before v-, n-/ṇ-, m-/ṃ-, p-, l-/ḷ-/ƛ- and t-/ts-/tṡ-.

The preverb tsỉ(t)- or tsỉd(a)- indicates that the action starts from a centre and moves away from it,, or converges to the centre, proceeds in or from different or all directions or radially. It is found as tsỉt- before consonants that can be geminated, tsỉda- or tsỉ- before consonants or clusters and tsỉd- before vowels. Examples include tsỉ-tem ‘flow out from’, tsỉṇ-ṇakṇ- ‘break from the centre, break (such that pieces fly everywhere)’, tsỉd-ell- ‘grow out (radially), spread into all directions’, tsỉḷ-ḷabb- ‘splatter around, in all directions’, tsỉd-ilg- ‘radiate, permeate’, tsỉṛ-ṛilg- ‘radiate’, tsỉd-ilpp-eḍ- ‘have umbel inflorescence’ and tsỉd-iġ- ‘scatter into all directions’, tsỉg-gigg- ‘converge, gather, contract’

The preverb iggi(t)- or iggin- indicates that an action occurs or goes well, pleasantly or nicely. It is found as iggit- before consonants that can be geminated, iggi- before consonants or clusters and iggin- before vowels. Examples of verbs with this preverb include iggiƛ-ƛigg-im-muutu- ‘be likeable, be personable’, iggiƛ-ƛisƛ- ‘sing well, sound good, be melodic, is harmonious’, iggit-tem- ‘go well, go smoothly’, iggis-siḷḷ- ‘be odorant, perfumed’, iggin-utṇ- ‘look good, seem good’, iggiv-vapṇoḍḍ- ‘burn or smoke such that it produces a pleasant smell’, iggin-ell- ‘grow nicely, prosper’, iggiṛ-ṛiḍḍį- ‘cut precisely, cut well’ and iggin-nįal- ‘hang, fit nicely (of clothes)’.

The preverb hủ(t)- or hủd(a)- indicates that an action occurs or goes unwell, unpleasantly or clashes. It is found as hủt- before consonants that can be geminated, hủda- or hủ- before consonants or clusters and hủn- before vowels. Examples of verbs with this preverb include hủƛ-ƛigg-im-mutu- ‘is disagreeable, is unpleasant’, hủƛ-ƛisƛ- ‘sing poorly, sound bad, be cacophonic’, hủt-tem- ‘go poorly, regress’, hủs-siḷḷ- ‘be malodorant, stink’, hủn-utṇ- ‘look bad, seem bad’, hủv-vapṇoḍḍ- ‘burn or smoke such that it produces an acrid smell’, hủn-ell- ‘grow poorly, fail, wither, collapse’, hủṛ-ṛiḍḍį- ‘cut poorly’ and hủn-nįal- ‘not fit (of clothes)’.

Other and rarer preverbs include mita(θ)- 'beyond, more than, further' and aḷḷ(a)- 'variously, in a changing way'.

§8.2.7.2 Kinetic Initials

Kinetic initials are housed in the subslot -3b, only precded by adverbial initials. Kinetic initials show movement (the most common of which being the cislocative s-i- and the translocative g-i-) and are sometimes optional but in many cases are obligate elements of verbs.

Kinetic initials must co-occur with the motion preverb -i(m)- (or other motion preverbs in a few rare cases) in conjugated forms of the verb, but the motion preverb is generally not found with dependent forms of verbs or copulas, for example sis·uktat nẻ na ‘to put (in)’ but sis·k·im·ukt·et ‘I put it in’. The two most common kinetic initials are the cislocative s(i)-/si(s)- and the translocative g(i)-/ge(g)-, which are found abundantly and have a wide variety of functions and meaning. Many verbs of movement have obligate kinetic initials (along with their motion preverb), and in many cases both cis- and translocative occur according to the direction of the motion. Kinetic initials are also obligate in non-copular infinitives found in relative clauses, and may be strongly favoured on the copula of copular infinitives found in main clauses. In the case of a zero surface form pronoun, the locatives generally appear in their short form s- and g-. If a pronoun does have a surface form, the sequence*gi-Ø-i is not allowed, and instead forms in ge(g)-Ø-i instead occur. In other words, locatives do not have an -i- if they are separated from the motion preverb -i(m)- by a pronoun (-gi-Ø-i- becomes -ge-Ø-i-).

The cislocative, when it precedes certain initials, is reduced to s-, as is the case with the sequence si-keu-/kem- ⟫ skeu-/skem-. Likewise, when the cislocative is found on a dependent infinitive before a verb that normally takes a translocative prefix, the result is s-ki-/s-keg- instead of the expected si-gi-/si-geg-

With dependent infinitives, the cis- and translocatives required by the governing verb come before any adverbial initials:

skeumindat gỉƛiggedin

'I decided to come along'

(not *keu-si-mindat)

gagga skegatṇat tsƛukka

'you must choose something'

(not *si-geg-atṇat)

§8.2.7.2.1 Cislocative

This kinetic initial is called the cislocative and indicates a motion which is by or towards the speaker or addressee (if together), upward motion, motion into something or motion forward. It is extremely common, and can replace the locative adverb deįįa, deḥḥet ‘here, to here’ completely, especially when motion is involved. It can also be used to emphasise a telic or conclusive, often future action, for example gaskagga ‘I can, I am able’ but sis-k-i-skaget ‘I can do it, I can make it’. The cislocative is a very common feature of dependent phrases where the action is set to happen, will happen, or it can indicate a goal (with hỉt ‘in order to’), an intention, a future time point (with agỉt ‘until’) or even a need or necessity.

The cislocative can be augmented by a- (a-s(i)-/a-si(s)-), which often translates to ‘come and’ or ‘come to’, indicating moving to a cislocative position before or while performing the action, which is especially common in commands. This is called the venitive kinetic initial:

§8.2.7.2.2 Translocative

This kinetic initial is called the translocative and indicates a motion which is away from the speaker or addressee (if together), downward motion, motion out of something or motion backwards – it is opposite to the cislocative. It is extremely common, and can replace the locative adverb nįiįįa/oįįa, nįiįįet/oįįot and deįįỉn ‘there, to there’ and ‘from here’ completely, especially when motion is involved. Like the cislocative, the translocative is a very common feature of subordinated copular phrases where the action already happened, or it can indicate a cause, effect, reason or result (with (a)hỉn ‘because’), a lack (with uġġit ‘without’) or a past time point (with agỉn ‘since’).

The translocative can also be augmented by a- (a-g(i)-/a-ge(g)-), which often translates to ‘go and’ or ‘go to’, indicating moving to a translocative position before or while performing the action, which is especially common in commands. This is called the andative kinetic initial. This andative kinetic initial has a short form that has fused with the first-person pronoun -g-, such that the sequence a-ge-g-i- is found as a=g=g-i-.

§8.2.7.2.3 Circumlocative

This kinetic initial is called the circumlocative and indicates a motion around a general location. Whereas the cislocative indicates movement to the speaker and translocative away from the speaker, the circumlocative does not have a specific destination but rather an abstract radius around a location. It can also have a meaning of distributive location, especially in combination with a distributive medial such as -aƛƛ-/-add-, for example inugalkaƛƛ- ‘distribute’ from alk- ‘give’. The circumlocative kinetic initial is the only kinetic that is found without the motion preverb, though many speakers still use it. Both are considered grammatical.

Like the cis- and translocative, the circumlocative also has a venitive and andative form, in which case the motion preverb is obligate. The venitive circumlocative has the form s-inu(g)- ‘come around and’ while the andative circumlocative has the form igg(u)-/igge(g)- ‘go around and’, although forms in issu(g)- and iggu(g)- also exist. With the kinetic preverb -i-, issug- and iggug- cause gemination of local prefixes just as me- and ge- do.

§8.2.7.2.4 Inceptive

The inceptive (ince) a(n)- kinetic initial shows that an action begins. It is nearly exclusively used with agentive subjects. Within Pine grammar, the inceptive has an inherently agentive meaning while the inchoative is an obligate unagentive aspect, which is not marked through kinetic initials. It is often found without the preverb -i-.

§8.2.8 Preverbs

Preverbs are elements housed in slot -1 or the proteromeric prorrhizal slot, immediately preceding the root. With very few exceptions, preverbs are exclusive elements, meaning only one can occur at the same time. There are two types of preverbs: movement preverbs and ditransitive preverbs, ditransitive preverbs, which function to indicate the presence of an indirect object, either a true indirect object or the antergate – an indirect object promoted to direct object. The kinetic preverb (-kin-) -i(m)- is required on ditransitive and hemiditransitive but not on anthemiditransitive verbs.There are two motion preverbs, -i(m)- which is found when kinetic initials are present, and much less commonly there is also the so-called precedative preverb (-prec-) -av(a)-, which comes instead of the kinetic preverb -i(m)- and despite the name does not truly indicate motion in space but order in time, and can be combined with the cislocative to form m-ava- ‘first’ and g-ava- ‘then’, as in mavamindats ‘come here first’.

In finite verbs, the kinetic preverb -i- appears as -iϕ- if a non-local pronoun is dropped – this is true of the third person animate/obviate and third person inanimate. The sequence h-i-Ø and ḥ-i-Ø (for the third person inanimate) and Ø-i-Ø (for a zero-form third person animate/obviate) are replaced by iϕ-Ø.

Likewise, in these same conditions (Ø-im-, ḥ-im- or h-im- )if the kinetic preverb has the form -im-, such as in verbs beginning with a vowel, the initial -i- is dropped with the third-person inanimate and the third persona nimate agentive past form:

§8.2.9 Medials

Medials are verbal elements housed in slot 1 which immediately follow the root. Medials are a rather large set of infixes that give additional information about aspect or voice, or may function as derivative infixes to create new verbs. There are three classes of medials, each housed in a specific subslot of slot 1.

The first class of medials which are housed in subslot 1a are voice medials, which are involved in marking the four overtly marked voices of Pine – the translative, the stative, the middle and the passive voice. With the exception of the passive, voice medials generally form a semantically and syntactically distinct verb and therefore cannot be freely added or removed like many other aspectual medials. The passive medial is found on verbs that are originally transitive and in some rarer cases ditransitive.

The second class of medials are called derivative medials and are found in subslot 1c. Derivative medials are often verbalizing formants whose purpose is to form verbs from a number of different roots. Some derivative medials can be found on verbal roots, modifying the function of the verb, such as -isƛuθ- which generate a new verb stem that has an automatic translative meaning such as ‘get Ø-ed’. Nominal incorporative medials are used as nominal verbalisers, such as -(i)neġ- ‘have object as one’s noun’ in tsṇenneġemat ‘I have you as my friend’ or ‘you are my friend’.

The third class of medials are aspectual medials housed in subslot 1c, which modify the aspect of a verb, such as -iġ- which confers a subitive-intensive aspect. Aspectual medials are perhaps the most flexible class of medials as many verbs can accept aspectual medials when the speaker wishes to emphasise a specific aspect.

One difference between voice and derivative medials and aspectual medials is that both voice and derivative medials have specific nominalizative forms while aspectual medials are found as-is in deverbalised nouns, with the medial inserting before the last vowel of the result or more commonly action noun of a verb. For example, the causative medials -ed-/-od- regularly derive their deverbalised noun forms by -iddun/-uddun, while for example -ask- in oallį-ask- ‘procrastinate’ is inserted into oallįa ‘tardiness, being late’ as oallį-ask-a ‘procrastination’, which does not otherwise affect the sematic form of the noun (oallįa ⟫ ảlįen and oallįaska ⟫ ảlįasken)

The main types of medials and their subcategories are:

voice medials

translative medials

stative medials

passive medials

derivative medials

diathetic medials

denominative medials

adjectival medials

causative medials

aspectual medials

Some medials are sensitive to tense, and together with the theme and personal pronouns as well as changes in the root through rhizal apophony are the only verbal elements that participate in tense marking. However, whole classes of medials are unaffected by tense, such as voice, adjectival and causative medials. Slot 1 has three subslots, each housing one of the three types of medials: 1a houses voice medials, 1b houses derivative medials and 1c houses aspectual medials.

Medials can be semantically obligate, meaning their presence can be required to change the meaning of a root verb, or they can be fully facultative, as is the case with many aspectual medials.

§8.2.9.1 Voice Medials

Voice medials are housed in subslot 1a and are the first medials to follow the root. There are four voice medials: the translative, the passive and the middle medials.

The translative and passive medials can be combined with further elements, such as the habilitive -iv- which can preceed a passive medial (-ϕ(uss)ivuθ-) to indicate that the action can be performed, as in tsint-tussivuθ- ‘be drinkable’.

§8.2.9.1.1 Translative Medial -uƛ-

Translative verbs are verbs denoting a change of state. All translative verbs are exclusively unagentive. The most common translative medial is -uƛ-, which can be combined with additional translative elements such as -s- (inchoative, automatic) and -n- (automatic).

The translative medial -uƛ- can replace other medials, such as the causative medials -ed-/-od- or -eg-/-og-, as they cannot co-occur. For example, maġl-od- means ‘untie, take seams apart’, which can be transformed into a translative verb by replacing -od- by -uƛ-, yielding maġl-uƛ- ‘become untied, come undone (of seams)’.

Two important elements of translativity often co-occur with the translative medial, called translative augments – the automatic augmented translative -(i)n-uƛ- and the inchoative-automatic augmented translative -(i)s-uƛ-. In many cases, these augments have fused with the root and are no longer considered to be infixes, but in the cases where the -i- is present, it is sensitive to tense and appears as -e- in past tenses. The automatic translative indicates a change of state which occurs by itself (without outside forces), while the inchoative-automatic translative indicates a gradual change of state, or the start of a change of state which happens by itself. Another augment of the translative medial is -elk-uƛ- which indicates an accidental change of state. A significant number of augmented translative verbs exist in pairs contrasting the automatic and inchoative nature of the action. Although the element -(i)s- has a mostly inchoative meaning, when it fused with other elements it often functioned as an automatic marker.

nominalised form of -uƛ-

§8.2.9.1.2 Stative Medials -uθ-/-is-

There are two stative medials, -uθ- and -is-. Stative verbs containing the medial -uθ- indicate the result of a change of state, or in other words, a changed state, usually without any reference to an agent and therefore have an automatic connotation. Stative verbs are nearly always unagentive.

The stative medial -is- often implies a resultative sensory experience or some sort of mental state, but in certain verbs it can imply some sort of reflective action and in a handful of cases, -is- can have both a stative and translative meaning, as in the verbs ỉl-is- ‘be/become acquainted with’, mỉlį-is- ‘approach/be close to’, apṃ-is- ‘curl up/be curled up’, tuv-is- ‘turn/be turned towards’, loįį-is- ‘slow down/be slowed down’ and kįasƛ-is- ‘speed up/be sped up’ . In other words, -is- generally refers to states experienced by a conscious agent. Both can be combined into -isuθ- which is often in free variation with -is-, and sometimes conveys a sense of an external agent:

The medial -uθ- is also called the perfective translative medial. It is found in a small number of verbs by itself, which indicate a state resulting from a change, but more commonly -uθ- is found combined with other elements, such as -(i)s-uv-, which also have a stative meaning. Both -uƛ- and -uθ- are also found in derivative medials, such as -isƛuθ- (originally -is-uƛ-uθ-). Etymologically, -uθ- is related to the perfective formant found in transitive prograde verbs, such as ganeguvi ‘I have seen her/him’. It is therefore not surprising that stative medials confer a perfective aspect to the verbs in which they are found. Translative forms that have corresponding stative forms often are defective in the perfect tense, for example ƛunƛ-unan ‘s/he has become burried’ is often replaced by ƛun-uttan ‘s/he is buried’, although there may be a slight nuance – the translative perfect form literally means that the action of becoming or getting buried is now completed, while the stative form indicates that the subject has undergone a change of state to being buried and is currently in that state. Substitutive perfect statives does occur, especially when the nuance between the translative perfect verb and the stative verb is very slight. Translative forms are also generally associated with movement locative cases (illative and elative) while stative use mostly the static inessive case.

translative perfect

ƛunƛunami pįuoįot vảlpṃimyt mi boagve na

‘the house has become buried under snow during the storm’

stative present

ƛunuttami pįuoįumit mi boagve na

‘the house is buried in snow’

A number of verbs which have incorporated the translative -uƛ- into their stem have a separate stative form, for example esƛ- ‘become loose’ and ess-uθ- ‘be loose’, ysƛ- ‘freeze’ and yss-yθ- ‘be frozen’ or ƛunƛ- ‘bury’ and ƛunuθ- ‘be buried’. There are also verbs which have augmented translative medial in -(s-/n-)uƛ- which lose their augment but gain the stative medial -uθ-, such as nakṇ-uƛ- ‘break’ and nakk-uθ- ‘be broken’, tỏltṇ-uƛ- ‘rip’ and tỏlt-uθ- ‘be ripped’, vatṇ-uƛ- ‘open’ and vatt-uθ- ‘be open’ or rekṇ-uƛ- ‘close’ and rekk-uθ- ‘be closed’. Stative verbs do not occur with translative augments. Note that in ulbn-uƛ- ‘become swollen’ and ulbn-uθ- ‘be swollen’, the -n- is part of the root and is not a translative augment.

There is a small number of stative verbs in -iġ-uθ-, which denote a state resulting from a semelfactive, subitive, momentane or otherwise rapid or sudden change of state. Nearly all such verbs have to do with a mental state. All have a corresponding translative form in -iġƛuθ-. The most notable of these verbs is slightly irregular, namely nỉ-ġƛuθ- ‘die (suddenly)’ and nỉ-ġ-uθ- ‘be dead’, although originally it meant ‘be dead after dying suddenly’ but the semelfactive aspect has been lost in modern Pine.

Abstract nouns in -ehit/-ohut/-ůhyt or some other nouns denoting measurable or quantifiable qualities may have derivative forms in -uθ- that indicate that the subject of the verb is of such and such a quality, for example tammohut 'size' and tammuθ- 'be of such and such a size, to measure', or tunkvohut 'weight' and tunkvuθ- 'be of such and such a weight, weigh'.

stative medials in -uθ-

nominalised form of -uθ-

stative medials in -is-

nominalised form of -is-

stative medials in -iġ-uθ-

nominalised form of -iġ-uθ-

§8.2.9.1.3 Passive Medial -ϕutuθ-

The passive voice in Pine requires the presence of the main passive medial -ϕutuθ- (or -ϕytyθ- following a front-rounded vowel), which like the connegative markers has a fortifying archephoneme called phi <ϕ>, which causes the preceding consonant to undergo fortification. It also contains the archephoneme <θ>, whose surface form depends on the following sound:

All passive verbs, like translative verbs, are obligate unagentive. The passive voice indicates promotion of the object of a transitive verb to the subject of an intransitive verb, thereby reducing the valency by -1. And like translative medials, the passive medial -ϕutuθ- can be augmented by one or two augments – the habilitative -ϕuss=iv=uθ-, which indicates that something can be done, and the inabilitative -ϕut=oġ=uθ-, which indicates that something cannot be or isn’t done.

Not all passive verbs have corresponding non-passive transitive forms, for example the form siġġussivuθ- ‘is interesting’ is allowed by *siġġ- does not exist (but siġt- ‘interest’ does).

The demoted subject of a passive verb can be expressed in Pine through the use of the elative. This is true of true passives and passive-like diathetic medials.

Unlike the translative medials, which can be added onto non-verbal stems such as adjectives, the passive medials are exclusively added onto transitive verbs, with the exception of the causative passives.

negative augmented passives

ninnaskuts, kvayvyt ohat iggiƛƛikk(uss)ivuppin keggillit sagvema ỉldot gilembat gehalin

ninnaskuc-, kfaa--fa-t u-hat iggiTTikkqussqifuppin ki-ggillit sagfi-ma iilt-u-t

gili-mp-at gi-hali

[ˈnɪnːasːkʊt͡s ˈkᶣæy̯ɥʏh ˈɔːhah ɪkːɪˈt͡ɬːɪʰk(ʊsː)ɪwʊsːɪn ˈkɛkːɪlːɪt ˈsakᶣːɛːmɑ ˈiːlːdɔt gɪˈlɛmːbat gɛˈhɑːlɪn]

ninnas-kut-ts kvayv=yt ohat iggit_-Ø-_ƛikk=ϕussivuθ-ϕin keggillit sagve-ma ỉldot gi-lemb-at ge-hal-in

young-comp-ess in.the.summer very 3a.unag.pst-be.likeable-nloc.hyp.pst western man-gen for

transl-work-dep 1.ag.pst-use.to-loc.pst

‘I used to work for a very likeable man from the west in the summers when I was younger’

§8.2.9.1.4 Middle Medials -VttV(g)-/-av-

The middle voice in Pine shows that the subject of a verb acts upon itself. There are two middle voice medials, -VttV(g)- and -av-. Much like the passive -ϕutuθ-, middle voice medials reduce the valency of a verb by one, such that most middle verbs are usually monovalent. Some middle verbs can be transitive however, and generally have a corresponding ditransitive form. The middle medials -uttu(g)-/-ůttů(g)-/-atta(g)- indicates that the subject acts upon itself, while the medial -av- indicates that the subject causes itself to undergo some sort of change, and is therefore a middle causative or translative medial. In addition, middle verbs in -VttV(g)- occasionally have a passive meaning, usually where there is no clear or logical cause, for example alts- ‘to cause’ and alts-uttug- ‘to be caused, to be brought about, to grow, to arise’, or ints-uƛ- ‘to split, to diverge’ and ints-attag- ‘to be divided, to split up, to be constituted of’

The -v- versioner appears directly following -uttu-/-ůttů-/-atta- (before -v-) while any vowel appears following -uttug- (or -uttog-~-ottog-)/-ůttůg-/-attag-. The form -uttu(g)- appears following a back vowel (including front-vowel initial diphthongs that end in a back vowel such as -eu-), -ůttů(g)- after front-rounded vowels and -atta(g)- after front vowels, though some back vowel verbs use -atta(g)- as well, especially in western dialects. A few examples exist of the middle being augmented by the habilitative or inabilitive medial augment (verbs in Ø-iv-atta(g)- or Ø-oġutt-atta(g)-). Some verbs can have both -VttV(g)- or -av-, such as tṡoalk-uttu(g)- and tṡoaḷḷ-av- which both mean ‘dry oneself’, although -uttu(g)- is found on the transitive stem tṡoalk- while -av- lacks the causative marker -k-. The verb starr- ‘injur’ can have either -uttu(g)- or -av- middle forms with no clear distinction between the two. The middle medials -VttV(g)- can be nominalised through the suffix -utuvat/-ůtůvat/-atavat, but it is also common to find it as a prefix (iṡ(ki)-), for example nuṡkv-uttu(g)- ‘kill oneself’ ⟫ iṡkinuṡkva ‘suicide’ (rather than nuṡkvutuvat, an alternative and less common form).

middle verbs in -VtU(g)-

nominalised form of -VttV(g)-

middle verbs in -av-

nominalised form of -av-

§8.2.9.1.5 Overview of Voice Medials

The table below summarises the form and function of voice medials.

§8.2.9.2 Derivative Medials

Derivative medials are elements housed in the subslot 1b which modify or augment the root of a verb, and which are sensitive to tense like most medials. Derivative medials are the main mechanism to create new verbs from roots, verbal or otherwise. Some derivative medials were originally voice medials which were instead incorporated into and fused with other derivative medials, for example -isƛuθ- which contains a vistigial translative medial -uƛ- and the automatic augment -(i)s-. Generally speaking, elements of subslot 1a and 1b rarely if ever co-occur.

There are four categories of derivative medials. Diathetic medials are closely related to voice medials and mostly form verbs with an inherently translative meaning. Causative medials form causative verbs, denominative medials form verbs from mostly non-verbal stems such as nouns or adjectives, but in some cases from other verb roots, and adjectival medials are found on verbal adjective stems.

§8.2.9.2.1 Diathetic Medials

Diathetic medials are derivative medials which are derived from the basic voice medials but which have usually gained some more specific function. They most commonly indicate automatic or involuntary changes of state, and therefore form unagentive verbs. Another distinguishing feature of diathetic medials is that they have regularly derived causative, and in some cases corresponding stative forms.

§8.2.9.2.1.1 Automatic Translative Medial -isƛuθ-