Today’s Grammaticalizanuary follows closely on the coat tails of yesterday’s.
In additional to a system of obligatory possessive classification, Karyoł also has a system of entirely optional numeral classification.
Three possessive classifiers may be used as numeral classifiers (bicil is particularly common). Numerals classifiers group nouns based on shape and consistency (and to a lesser extent ‘type’)

ilta ‘man’ – people
entoanta ‘sheet’ – flat flexible
twol ‘plate’ – flat rigid
bicil ‘hand’ – a residue ‘catch-all’
uruma ‘animal’s ear’ – some fruits, soft lumpy objects
yorta ‘tooth’
omagkwe ‘tongue’
cwañca ‘person’s chest’ – broad objects, pieces of furniture
lerag ‘arm’ – long objects – including long metaphorical extensions
itula ‘grain/point/dot’
koeya ‘flask’
torra ‘head’
bōaka ‘herd’ – orderly masses
wao ‘tray’ – buildings
kuk ‘burrow’
cica ‘rag, scrap’
kal ‘stone, cobble’
biga ‘scale, fish scale’
rāke ‘root’

I’ll put more together on these later – I’m trying to write a big-ish post on middle voice markers for tomorrow, and I can already tell it’s going to be a very busy week.

Nouns may be classified into three sets according to how they behave as the possessum in possessive phrases. Set 1 is a broad but closed class of inalienably possessa which take pronominal possessor marking similar to the S/O marking on perfective and imperative verbs. Set 2 is a small closed class of alienable possessa which do not require modification by a possessive classifier. Set 3 is a huge open class, to which almost all nouns belong, which do require possessive classifiers.

Set 1 words tend to refer to body parts, kin terms, and some states and emotions. Though their first and second person pronominal possessive suffixes are identical to (or nearly identical to) those found on perfective and imperative verbs (1st person sg: –b(e), 1st plural –(o)n, and 2nd person –I), they indicate third person possessors according to either of two strategies. Just as third person reference on verbs, third person possessors may be Ø-marked, or indicated by a suffix –(o)ño. The word byone is singularly irregular in that it indicates a third person possessor through a prefix ka-. Some Set 1 words are: face, ear, nose, lips, teeth, gums, tongue, throat, shoulder, elbow, arm, wrist, finger, breast, heart, stomach, back, legs, feet, heel; the voice and one’s name; mother, father, son, either grandmother, either grandfather, mother’s brother, mother’s brother’s son.

An inalienable form may be transformed into an alienable form by one of several strategies. This is necessary to indicate differences such as ‘my bones (in my body)’ versus ‘my bones (from an animal I killed)’. In the case of those forms that mark a third person possessor in –(o)ño, the alienable counterpart is identical to the lemma. Those that take a Ø-third person may form the alienable through suffixation (cwol ‘milk (inalien.)’, cwom ‘milk (alien.)’), a system of suppletion in which the alienable form is a Set 2 noun (byore ‘skull (inalien.)’, rākonam ‘skull (alien.)’ – although byorem is also available; łalwa ‘bone (inalien.)’, yārorem ‘bone (alien.)’), or through a quite irregular process of compounding (keao ‘toes, foot (inalien.)’, lerkea ‘foot (of a piece of furniture)’).

Set 2 contains words for more distant kin and some less integral body parts, as well as a set of words which would correspond to quantifiers in other languages. These words tend not to require the support of a classifier (although some speakers may use them with byone). The possessor may optionally take the linking morpheme. Some Set 2 words are one’s daughter and her children, and one’s spouse and the spouse’s family; the word ‘friend’; semen, the kidneys, liver, hair of the head, the one’s smile, shit, vomit, blood and blood vessels, fingernails, and muscles; chronic pain and some diseases; beliefs and words for trust and promise; a few words that refer to one’s language and the town he was born in; a few personal affects: gardening tools, rope, and some clothing.

Set 3 is the broad and open class that contains the overwhelming majority of the vocabulary. Strictly speaking, these forms may not themselves be possessed, but instead are in apposition to a directly, inalienably possessed ‘classifier’. There are five weakly grammaticalized ‘classifiers’ which combine with a Set 3 noun based on its semantics. The territory covered by each of these cover is not particularly well-demarcated, and one Set 3 noun may occur with more than one classifier. All of the classifiers except for byone may be used as full, Set 1 nouns, and all these classifiers except omagkwe are open in that they may accept new vocabulary.

1.) omagkwe ‘tongue’ – eating utensils, mass nouns for food, liquid, language, thought, alienable kin relations, some fish, birds, dead animals
2.) byone (no longer used as a noun, used to mean ‘clavicle’) – Wheeled vehicles, boats and ships, steam and steam technology, travel, weapons (swords, knives, axes, spears and poles, bows and arrows, explosives, guns but not ammunition), large pieces of furniture, large domestic animals, fire receptacles, buildings and architecture and rooms in the house, trees, wild animals. – This marker is increasingly used with NP’s that belong to one of the other four domains.
3.) bicil ‘hand’ – Containers for food, countable nouns for food, most clothing (see below), printed and written works (relatively thin), smaller pieces of furniture (notably chairs and hanging artwork), flowers, leaves, small domestic animals
4.) yorta ‘tooth’ – Ammunition, insects and food from insects, jewelry, metal, precious stones, electricity and electric technology, some diseases
5.) wulu ‘liver’ – Thought, perception, written and printed works (relatively thick), some diseases, religious paraphernalia, ‘lumpy things’ (pillows, arthritis), some fish, snakes

This group is classifiers is to an extent, though, an open class. Some forms closely associated with a specific inalienable possessum may select that possessum in lieu of one of the five on this list. This is specifically true of clothing and adornment that is associated with just one body part.

Possessors are generally expressed in their nominative (not absolutive) form. There is a stylistic tendency to place words for female possessors in the oblique in –ał, although this isn’t necessarily correct from a diachronic standpoint. – There is a tendency for byone, as the most common and arguably most grammaticalized classifier, to lose its stress and become an enclitic on the word before it.

Explicit full word possessors are generally immediately before the NP they modify, although this order may be manipulated for pragmatic purposes – when it is reversed, the possessum (in first position within its phrase) will be marked with the linking morpheme. As the classifier-possessum relationship is an appositive relationship and not a matter of dependent marking a head, either phrase is free to come before the other (although we could say there’s a slight preference for the order classifier-possessum). In poetry, the classifier and possessum can be widely separated – this is possible too in spoken language, but exceedingly rare.

Since the beginning of the Karyoł project the language has not had a sense of nominal number. I originally had an idea that verbal number, signaled by an affix or by suppletion, would indirectly signal nominal number, but I’m beginning to feel that this not in keeping with the design goals of the language.

I already have a well-grammaticalized and not really productive collective in –hoal, which is just delightful in that it may be so easily confused with –hoał ‘seat, basis, nest, throne’. I imagine –hoal could be to an extent combined with new vocabulary, but I feel a native speaker would reject it, l for the most part, as being awkward.

Lehmann (p 60) writes, ‘[V]erbs may acquire the category of number by the agglutination of a personal pronoun. This is also a possible origin of nominal number’. – As a matter of fact, toward the beginning of the Karyoł project, when I was reading a grammar of Great Andamanese, I had an idea that a word ɲIb would be used to coordinate both NP’s and VP’s along the lines of ɲIb=P₁ P₂=CONJ, with CONJ most likely being =wa ‘and’. I scrapped the idea, though, when I grammaticalized =wa as a conditional marked, and I turned ɲIb into a pronoun, specifically a third person plural pronoun.

Let’s say that ɲIb is in fact a third person plural pronoun, but that it specifically refers to animate arguments, or even just human arguments. I already have Wackernagel auxiliaries agreeing with A’s in [±human], and although I really enjoyed this being the only corner of the grammar where the distinction mattered, I’m comfortable with spreading the agreement around a bit. ɲIb could then be procliticized onto a human NP to signal number, or perhaps onto a VP if the S or O was human and not explicitly stated. Coordination between human NP’s would be handled with the pattern ɲIb=P₁=LINK P₂, where LINK is the adjectival morphology I talked about the day before yesterday. As the –(a)t linker would no longer apply simply to adjectives, I’ll need some other test for adjective-ness (or just retire the word class entirely). Perhaps an adjective is a para-nominal that cannot exist by itself in the predicate, but requires another NP to support it. (Along the lines of ‘He is good’ being illegal but ‘He is a good boy’ being well-formed.)

If –(a)t is available as a NP to NP marker, it’ll be neither head- nor dependent-marking; it’ll attach to the first word in a phrase, with order in part determined by pragmatics. This (mild) non-configurationality is actually what I had pictured to begin with – and I’ve since moved away from it. I’m interested in what it’ll do to my system of possession – which is broken down into alienable and inalienable categories. It’s also a classifying system, and each alienable noun requires the support of an inalienable noun – so ‘my spoon’ would be rendered omagkwebe loabał ‘my tongue spoon’ and ‘mother’s spoon’ would be ōmace omagkwe loabał. Now that the linker is available to nouns, I’m thinking I’m going to have it join a possessor to an inalienable possessum – the linker wouldn’t attach the omagkwe to loabał, for example, as these are two separate appositive NP’s.

Today I’m going to introduce a new modal quasi-auxiliary and differentiate between weak obligation and strong obligation. – Modals in Karyoł operate according to an accusative scheme – and this makes sense. Modality is far more concerned with what an actor may or may do, and less concerned with whether or not an O is affected or unaffected.

What I’m looking at, then, is intransitives and statives as my lexical sources. Bybee lists a number of both in her list of lexical sources for weak obligation, and two jump out at me – ‘see’, which is used in Maithili, an IE language spoken in Nepal, and ‘fall’, which is used in Baluchi. I don’t know how ‘see’ operates in a Maithili, but in Karyoł it’s a member of a set of ‘extended intransitives’ which have a closely assosiated ri-marked NP. ‘fall’ is also interesting – it looks like we’re looking at a case of a S=O verb (semantics are essentially secondary) carrying over a sense of [-volition]. I like that.

‘The liver is the seat of emotion and mental processes’ is an enduring Karyoł conceptual metaphor. Much like our sense of the four humors, it’s particularly associated with passion and hot tempers. The

tuli ‘flint’ dates back from one of the earliest big vocabulary expansions – maybe from March of 2014. I don’t remember creating the word itself, but the entry is glossed with etymological ideas to make creating a cognate in Gralli as easy as possible (over time I’ve gotten lazy and have stopped doing that!) The word tuli doesn’t seem particularly ‘noun-y’ to me, although there’s also no reason why it couldn’t be a noun. If I were creating the word now, it would honestly probably be tolem, with my quite frequently used –Vm suffix…

To get back to the point at hand – I had made notes that it would correspond to a word in Gralli meaning ‘heat’. Interesting. So let’s recast tuli as a verb for just a second and see where that takes us.

The most obvious related verbal sense dealing with ‘heat’ is probably ‘to spark’. Okay. I already have word imiłi– which means ‘to spark’, which has related senses ‘to fly off’ and ‘to break into chips’. I’m pretty happy with those, and imiłi– is already pretty well integrated into one of my words for ‘amber’. – So ignoring the obvious but also more specific sense of ‘to spark’, I think I’m actually just going to stick with ‘to heat up’.

So let’s say at some point in the past – and pretty far in the past – there was a verb tuli ‘to heat up’. A noun is created through Ø-modification (another noun created this way gowe ‘drinker’), which means ‘fire-starter’ and may be applied either to a person or his tools (cf hāma ‘engraver’ or ‘engraver’s burin’). Over time, though, the word comes only to refer to the tools themselves (as anyone can make a fire), and refers especially to flint. Let’s also have a parallel form with explicit nominal morphology develop – we’ll forget the fairly common –(V)m and use instead the far rarer –yaɲ: tolyaɲ, ‘flint’.

So what happens to the verb? Its semantics are pretty close in keeping with the word ihi, and so let’s say it gets shouldered out and comes to mean ‘to warm up (intr) in the sun or by being kept close to the body, etc’. And this sense of mild change of state and no sense of volition is what forms the basis for the new auxiliary.

By 0 AT tuli has taken on two additional senses: epistemic probability and a kind of future, perhaps immediate future. These senses develop entirely out of the auxiliary and its meanings. I’ll have to do some translation to work out exactly how they function.

Today’s Grammaricalizanuary entry features an example of a noun becoming less-concrete and taking on adjective-like qualities.

The word in question is hurin, which may refer to either the forehead (as a part of the face) or the frontal bone behind the forehead. As the forehead is the thickest part of the skull, the word hurin comes to mean ‘hard’ or ‘the hardest’. This is a fairly early change in meaning, perhaps -1000 to -800 AT, and may happen under areal influence. At this point, however, hurin is still very much a noun. Only about -400 AT, and in Larlaroł only, is hurin compatible with adjectival morphology, specifically the linking morpheme –(V)t, and by this time the meaning has also drifted, and again only in Larlaroł, to mean ‘center-most, most important’.

Common Karyoł has a very old causativizing prefix in te-, which is transparently related to the full verb te ‘to earn’. In addition to raising the valence of intransitive verbs (and it could only function on intransitive verbs), causativizing te– also was able to turn nouns into verbs, e.g. telac ‘to put at ease’ < lac ‘nudity’. The prefix has long ceased to be productive, but it’s far from uncommon – All the same, because of semantic drift speakers tend not to recognize the relationship between an unmarked form and the same form marked with te-, or if they do, the relationship is understood to be derivational and not inflectional.

After causativizing te– ceases to be productive Karyoł speakers develop several periphrastic valence-increasing devices to adhere to their language’s strict transitivity requirements. We’re looking at the year -800 AT or thereabouts – perhaps 100 years later. The verbs amta ‘to give’, heła ‘to send’, and balha ‘to set down, to set up’ are the most commonly used to raise valence, and they should be thought of as serial verb constructions, rather than quasi-auxiliary constructions, in that both the lexical verb and the function verb both act as head of the phrase together; neither takes the

Among those forms that don’t already have a te-marked counterpart, quite a few verb classes are incompatible with the new periphrastic causatives: verbs of posture, stative verbs, copulae, verbs of perception, verbs of bodily function (with the exception of laho ‘to weep’ which may co-occur with any of amta, heła or balha, and which has a te-counterpart telaho ‘to pour water from a container’), some ingestive verbs, and the verb amta itself.

amta is notable in that it is the only form that accepts a transitive main verb. When combining with a transitive lexical verb, the A of the lexical verb becomes the new O, and the O may either be omitted (most usual) or may be expressed in a ri-marked phrase (next most likely) or in a locative phrase (rare). When combined with an ambitransitive, for example tweaba ‘to load wares (O₁) into a wagon (O₂), however, the result is a transitive with no change of A and with the loss of O₂, and with the additional overtone that the action be done with a great deal of energy.

amta is more likely to combine with a verb that assigns a less-forceful role to A, so that (be)cyon ‘to follow is a more likely candidate than akka ‘to take by force, to pillage, to rape’, but this is only a tendency.

amta is also the only one of the three forms able to co-occur with those ingestive verbs that can be causativized: okal amta ‘to feed’ < okal ‘to eat’, wuti amta ‘to nurse a baby’ < wuti ‘to suck on’. Verbs such as gowe ‘to drink’ and im ‘to swallow’ don’t combine with causative quasi-auxiliaries. amta may also occur with a subset of verbs which otherwise combine with balha to form idiomatic expressions: naotba amta ‘to raise a group’s morale in dark times’ < naotba ‘to begin, to start out’.

balha is overwhelming the most productive of the three. Essentially any intransitive action-process verb may combine with balha to suggest that the new A was somehow involved in initiating the action atēce tyohka urka balhama ‘I spooked the rabbit, caused it to run away (but I don’t necessarily know how I managed to do it)’. heła, on the other hand, definitely implies that the new A takes a volitional and even forceful role: atēce tyohka urka hanta ‘I chased the rabbit away’.

In the North, and later in Larlaroł, amta is appears in the same role that heła does in the rest of the Karyoł-speaking community, in part because the perfective of heła, hanta, sound so much like amta in rapid speech; we can see evidence of this confusion in that the lemma form amta and never its perfective amma is used in this construction.

In the Larlaroł of 0 AT amta may combine as an affix of the form –anta, in part influenced by the relative participle in –Vnta: tweabanta ‘to heave’, okalanta ‘to feed’, urkanta ‘to chase’. (I picked these three forms specifically because they all have irregular relative participles: tweonta, okanta, and suppletive oganta, respectively. Don’t want to confuse myself just quite yet.) Their specific meanings are conditioned by the original sense of the serial construction: Note that the first and the third example definitely have undertones of force, but that the second does not. These –anta forms are defective in that they have no specific perfective form; ergative alignment alone is used with these to signal the perfective.

balha remains in all dialects the only productive form able to combine with new vocabulary, although it still may not combine with intransitives. In order to suggest that someone made someone else perform a transitive action, some sort of biclausal construction would need to be employed. That I’ll have to work out later.

Moving forward, I’ll be referring to variety of Karyoł spoken in Pagria as Larlaroł, a self-designation Twāogowe settlers in Pagria use for themselves and for their language after the Paleo-Pagric term larularul, ‘Southerners’. Karyoł as spoken in Twāo will still be called just plain Karyoł. It’s getting a bit cumbersome to refer to each dialect’s locality when I want to highlight differences. When I needed to refer to the language continuum as a whole, I’ll use ‘Common Karyoł’. – It may be fun to note that in terms of internal language history the –ł in Karyoł is etymologically an l, but has been influenced by oblique/more-marked nominative ending –(a)ł. In terms of external language history, what actually happened when I was creating the darn thing, is simply that I failed to pay close enough attention to my etymological notes when I wrote in a distinction between l and ł, and accidentally confused the ł at the end of ‘Karyoł’ as a case-marker. Ahem. I figure if I can make the mistake, the Twāogowe, even with their rather self-aware philological tradition, can too.

Today’s Grammaticalizanuary post deals with another purely Larlaroł instance of grammaticalization. wao began its life as a term for any permanent structure. The earliest ancestors Twāogowe were desert nomads, and permanent structures were rare and the exclusive property of tribal monarchs: fortified dwellings, towering granaries, and cliff-side astronominal observatories nested in the extreme east of the Aurès Mountains. Up until 0 AT, the building used for Twāogowe religious activities are still referred to as Nāebānaɲyaɲ Wao, ‘Astronomical Colleges’.

In the South, wao refers to a kind of tray with architectural references used for holding cosmetics. In the North of Twāo Proper, and later in Pagria, the word comes to refer to any mudbrick stucture of any size.

Up until today, Karyoł has only had a no-longer-productive nominalizing suffix in –(ō)am; most of the forms marked by this suffix have non-combinatorial meanings. The goal in creating this new suffix (or enclitic, albeit a tightly bound one, as it can modify an entire verbal phrase) is to develop a simple and productive means of nominalizing VP’s. In northern Karyoł the participle is less-used than in the Southern dialect, and the lemma form may be used to modify a N directly, which makes for sleek, efficient V + N univerbation. N + N univerbation, on the other hand, is rather clunky, owing to the periphrastic nature of the Karyoł system of linking and possession.

The first uses of –wao refer explicitly to physical structures: harwao ‘shit-house = latrine’ and cāmawao ‘roast-house = brick stove’. But the meaning comes to be extended to abstract states, following the prevalent states-are-places metaphor.

Of fairly recent origin, –wao is a disjunct marker and does not cause lowering-harmony.

Words to add to the Larlaroł vocabulary are:
gāhyawao ‘nation, one’s countrymen (as opposed to the majority Kepper population)’
narbwao ‘hunger, privation, shortage, deficiency’
tiɲcuwao ‘pleasure, recreation, idle pasttime’
tatwao ‘beauty (esp. of a woman)’
netehoalwao ‘collection of books, library’

daugod and ittaugodmah both have reflexes in contemporary Karyoł. daugod surfaces 2000 years later as łōgo, a hypercorrection from earlier łōgoł, with the meaning ‘to read, to study a topic, to make out, to discern’. ittaugodmah gives us etōgona in the South and ettogona in the North, where both forms have rather general semantics ‘savoir’ and ‘connaître’ as well as ‘to understand’. In the North ettogona also covers ‘to experience, to undergo’. Around -500 AT ettogona comes to be used in a type of serial verb or auxiliary verb construction I refer to as a ‘quasi-auxiliary’ construction – the quasi-auxiliary, in this case ettogona, comes clause-finally and acts as the head, taking the lexical verb as an ri-marked object. (Remember from Grammaticalizanuary the 1st that Karyoł’s ‘true auxiliaries’ function very much like modifiers.) The earliest uses of this construction follow the semantics of ettogona fairly closely – one may say gōam ettogona ‘to experience hardships’; where gōam ‘hardships’ is a noun, but one may also say itumba ettogona ‘to feel nervous’, where itumba is a stative verb (marked with the same it(a)= form we saw yesterday.

The full verb ettogona is one of class of ‘extended intransitives’ which are syntactically intransitive but are closely associated with a NP marked with the dative. And as the syntax of quasi-auxiliary ettogona closely follow that of full verb ettogona, these quasi-auxiliary constructions lower the valence of a transitive verb. As ettogona presupposes a human A argument, the valence change results in A=S with a O > oblique. This valence-decreasing effect, which really develops as an Iaccident from the semantics of the original full verb, comes to be as important as the sense ‘to experience’. At this point, around -300 AT, both senses occur with equal frequency.

Over the next one hundred years, quasi-auxiliary ettogona develops an habitual sense, but this habitual sense retains the detransitivizing grammar of the original construction. At 0 AT, ettogona may come either clause-finally or in Wackernagel position, but it hasn’t undergone any phonetic reductions that would lead it on the path to cliticization.

I mentioned on Grammaticalizanuary the 1st that Karyoł already has an imperfective, and that alignment is conditioned by transitivity requirements – and so one may surmise, and surmise correctly, that Karyoł already has a detransitivizing habitual mechanism implicit in its imperfective. What’s the difference? – Pragmatics. The imperfective is used when the aspect is expected by the speaker’s audience – or at least when aspect isn’t specifically inportant. But the periphrastic construction gets dragged in if the aspect itself is in focus – in keeping with the theme that new information tends to be placed in a privileged position, in this case Wackernagel position.

I’ll be coming back to edit this entry with some translations. I’ll be taking the day off tomorrow to cope with this cold that isn’t getting any better, and I’m hoping to get some good work in then 🙂

(Today is Part 1 of a 2-Parter…)

Porto-Krelyic or Pre-Karyoł (at any rate ‘PK’) exists right now not so much as sketchlang as a collection of typological notions. One of these is that at the morphological level the primary opposition is not one of ‘stem’ and ‘affix’ but of ‘phonological word’ and ‘grammatical word’. Critics abound. One of the over-arching themes behind the Karyoł project is phonological attrition and cliticization – the reason being, in order to accomplish all I want to do historically, I’d have to roll out and decommission morphology at what I would consider an unnaturalistic rate. Consider the English third person singular present in -s, a suffix with at least 7000 years of history behind it. The English copula clitic -’s is of far more recent origin, and in some dialects, one of which I group hearing everyday, it’s already been retired from active duty. In North Georgia you can hear quite a few people say He goin for He’s going. Some linguists would of course argue that in these dialects the -’s hasn’t disappeared but in fact was actually never there – but that’s really beside the point. Or perhaps it is the point.

Let’s consider, for a moment, that clitics are in fact the phrase-level equivalent to the affix. Not only do they tend to carry similar kinds of meanings, they also tend to exhibit similar morphosyntax. Among affixes, we have prefixes, suffixes, and infixes (among others we won’t mention because they don’t help my argument), and among clitics we have proclitics, which come before a phrase; enclitics, which follow a phrase; and that special sort that fall in Wackernagel position. There aren’t any clitics we know of that come in penultimate position, and, at least according to Stephen R. Anderson, there aren’t any infixes that fill in a slot right before the last segment in a word.

It seems to me that the correct use of clitic morphology may be just a bit more elusive to speakers than inflectional morphology. Affixes have rules and structure, and though the meanings they produce may be non-compositional, their deployment is rigidly predictable. Any system of clitic deployment, however, will be more complicated than an affix system by a power of n, where n is the number of parts of speech that will accept the clitic. (When that occurred to me, I actually thought it was a bit of a joke, but the more I think about it, the more it seems to make sense.). So it seems to me that the speakers who say He goin have just decided to hell with a semantically-meaningless form with a grotesquely complex syntax.

I also feel that the opposite may happen too. Once a clitic becomes significantly bleached, it can appear in all positions available – speakers overcompensate by way of hypercorrection. I’m thinking about the first person enclitic in Sandawe that appears on every word in the sentence but the verb (See Conlangery, Episode 54.)

Today’s Grammaticalizanuary post begins in the distant past with just such an overused clitic, it(a) ‘already’. I’m not sure how long in the past we’re discussing, but probably between -3000 and -2500 AT. Let’s call it -2750. At this point in PK resultative senses may be productively formed from a dynamic verb stem through the addition of the proclitic adverbial element it(a) ‘already’; with daugod this produces it=daugod or it=taugod. Over the next 400 years, this construction is grammaticalized to mark a broader range of stative expressions; when attached to a stative verb it= forms the present, and the unmarked form takes on inchoative meanings. it(a)= may be used to form stative verbs from abstract (or less-concrete) nouns: one such example would be: it=nau ‘to be blind’ from nau ‘blindness, a blind man’. – If all of this sounds familiar, that’s probably because this is straight out of The Evolution of Grammar. Now it gets a little weird: stress-bearing it(a) ‘already’ still exists in the language, and the full and reduced forms are similar enough that speakers are aware of its historical connection to each other. As full and reduced it(a) are both capable of existing in the same sentence together, there’s nothing stopping multiple permutations of stative-present-marking reduced it(a)= from leaping off and connecting to the beginning of every word after the verb. Even though PK is predominantly SOV, it places oblique phrases and focus information after the verb – there’s plenty of material to mark gratuitously.

So not long after its inception as a resultative, it(a)= becomes thoroughly regularized as a present tense marker. The distinction between ittaugod and daugod has been obliterated. However, a new wave of enclitic and suffixal morphology is developing, the result of language contact. One of these new pieces of morphology is -mah ‘anterior ’ from mah ‘yesterday’, which is appended once more to process verbs to produce a resultative.

One of the reasons why I feel conlangers are attracted to the theory of grammaticalization is that it rescues our morphology from the cold tables of inflection we all used to post on the internet in the 90’s and it breathes into it the same sort of life and personality we love about our favorite words. For those of who are interested in how morphology develops from an incomplete bleaching of a word’s semantics, and then of course the subsequent reduction of its phonological material, we’re particularly interested in the fact that bleaching is incomplete, and that it leaves a measure of semantic residue behind. Rather than inflectional material applying perforce to the entire vocabulary with exactly the same outcomes, we can imagine specific relationships developing between individual words and individual affixes.

it=taugod is one such close relationship. Even though it(a)= is proclitic in most cases, it forms a tight bond in this particular instance. (I believe, but am not certain, this is an instance of what Lehman would call ‘univerbation’.) This more marked form continues to be used for the more abstract ‘to know’ (alongside more concrete daugod ‘to learn’), and it’s to this more marked form -mah is attached; the result is ittaugodmah, which results in Karyoł etōgona.

We’ll talk about weird things to do with etōgona tomorrow.

This follows a general lingusitic trend, and the logic seems to be: Instances in which an O may be non-referential refer are instances in which the entire action is familiar enough to speaker and hearer both the verb doesn’t need to be specified by a separate, distinct N which may be scrutinized and pulled apart from its V. The N is understood to be a part of the V; it’s shared background information, it’s a cliché. And clichéd turns of phrase tend to be unstressed and reduced. By the year -1200 AT, non-referential O and V form a tight enough bond that incorporating a non-referential O may actually lower the valence of a transitive verb. Transitivity-obsessed Twāogowe are aware of the ambivalence of this construction, and come to exploit to feed their languages S=O pivot. In the year -1200 AT, VP’s with non-referential O’s are at the surface level labile.

Once this system of noun-incorporation becomes self-conscious, the Twāogowe begin using it to other ends – most importantly, to narrow the semantics of a verb with a rather general meaning. This may be applied to both a transitive and an intransitive verb; in the case of intransitives, the noun incorporated is a related but non-core N – this may be a locative (as in the case of me yu ‘to go back to one’s homeland’, which still exists in the current language as meyo ‘to approach, draw near’), an instrumental, or even a comparative, as in the case of ambak-yu ‘to go/walk like a dog’.

-1200 AT to -750 AT is the Golden Age of Karyoł noun-incorporation. Those familiar with the Mithun paper in the December 1984 issue of Language will recognize the sort of noun-incorporation I just described as her Type I noun-incorporation. During this Golden Age Karyoł also develops Type II incorporation: in some instances, an N may incorporate into a V, but the V remains transitive. This is used to promote topical N’s from an oblique role to O, and generally involved verbs of grooming and putting on clothing. But Type II incorporation doesn’t last long, and in the current language the descendent of the incorporated form (e.g. temmabelo ‘to wash the hands (trans)’ is considered a separate from the descendent of the unincorporated form (bilu ‘to bathe someone else)’. By -400 AT noun-incorporation on a grand scale is only moderately productive, and is no longer available to detransitivize a verb.

So most the incorporated forms we have in the current language are relics of a far earlier time. As such, semantic drift has pulled the semantics of these forms away from their original combinatorial senses. One such example is -yu, for example as in the me-yu and ambak-yu examples above. Originally ambak-yu meant only ‘to go around like a dog’, but its descendent form ambacyo has come to mean ‘to act sycophantically.’ Other examples are:
earoyo ‘to be sure of oneself in one’s surrounding, to “walk around like one owns the place”’ < ‘ibex’-yu
umruyu ‘to occur as aspected’ < ‘to move from east to west’ < ‘sun’-yu
hebaoyo ‘to be carefree, to have no worries’ < ‘sparrow’-yu
haŋkoyo ‘to be a sort of social parasite, to scavenge around for kindnesses and then to speak badly of those who have given them to you’ < ‘gull’-yu

Such colorful vocabulary is quick to be used to modify other verbs, and as modifiers these forms are most commonly found as nonfinite and adverbial converbs. The converb is formed with the comitative marker in –(a)ł, which is also used to form the more-marked nominative. As such, these forms may rather painlessly form verbal nouns, and so in the year 0 AT ambacyoł ‘a sycophant’ is more common than ambacyo ‘to act like a sycophant’.