This document is the quick and dirty grammar sketch that I developed as part of the Language Creation Society's Conlang Relay 2023 (Song Ring) in the lead-up to that year's Language Creation Conference.
Return to cartweel's Home Page.
Ancish a.k.a. Anca Wita or Saxa is a conlang project I started several years ago and have tinkered with off and on since then. You’ll excuse these notes for being simply that–notes–rather than a formal grammar description. (As I said in an email: “Warts and all!”)
The pitch is this: toki pona but make it Old English.
At least, that’s where it started; it’s evolved since then. This language and its development should be thought of as inspired by Old English, not based on or derived from it. This language is more D&D than Deor.
toki pona: simple phonotactics; lowercase; a “few words, wide meanings” approach; modifiers follow the modified; rigid or even templatic word order; simple clause structures
Old English: the lexicon; pronoun paradigms; TAM marking and participial forms; an appositive style; poetic forms; vibes
Other: Latin and Greek influence a high liturgical register; I’m not picky about stealing a North Germanic word if I like the cut of its jib; I’m not above a few Okrand-style puns; some words are just plain made up
What this leaves us with is a relatively simple, mostly analytic, engineered-ish language great for dungeon riddles and bar fights with barrow-wights.
Don’t think about it too hard. My pronunciation stinks!
If you want to say this stuff out loud, use this handy Old English pronunciation guide: https://oldenglish.info/advpronunciationguide.html
And check out this awesome Beowulf performance by Benjamin Bagby: https://www.youtube.com/watch?v=2WcIK_8f7oQ
All roots in Ancish end in a to start. Elision may cause the deletion of this vowel, as is especially common with the third person pronouns e, es, and er.
hunda er => hunder, ‘her dog’
As seen above, my convention for elision is to append e and similar pronouns to its preceeding word. However, in other circumstances (such as with the genitive particle afa or the accusative particle at), I tend to delete the final a but not join the words orthographically:
hunda afa cilda => hund afa cilda mana cræfta at hūsa e => mana cræft at hūse
This is just a looks thing.
As in toki pona, any basic Ancish sentence follow simple and rigid syntactic patterns.
[! Warning] The following section on sentence structures is dubiously useful at this point–if it ever was useful! The more inflectional Ancish has become, the less templatic it has grown.
(subject) (pre-verb particle) (verb)
(subject) (pre-verb particle) (verb) (post-verb particle) (object)
When both a subject and object are relatively short and unmodified (usually none, absolutely no more than one modifier for each), then the template may be modified as such: (subject) (pre-verb particle) (object) (verb)
(background) (background particle) (subject) (pre-verb particle) (verb) (post-verb particle) (object) (dative marker) (indirect object) (and so on using other particles…)
These categories and things like the various particle types are explained below.
ha - references a preceeding or following clause sete - relative pronoun te…te - establishes clausal disjunction or series (also hwila…te for time) hwa - question
A relative clause may follow a noun phrase and is marked by the relative pronoun sete.
For example:
hūsa, wera cræft at sete ‘a house, which a man crafted that one’ i.e. ‘a house that a man built’ OR, equivalently… hūsa, wera sete cræfta ‘a house, which a man that one crafted’ (n.b. the omission of the pre-verb particle is allowed since the noun phrases are short, sete cannot be a modifier, and there is no ambiguity.)
hūsa, sete breca at mana ‘a house, which crushed the man’
[! Sidebar on Information] I have wondered how the above relative clause construction might serve as a vehicle for topic marking or information introducing… but I haven’t thought through that very far yet.
A man entered a house. The man exited. mana, sete hūsa infared. mana se offared. A man, that one entered a house. That man (the man) exited.
…is that anything?
The pattern te… te… indicates a list or sequence. hwīla… (te)… -> While… then…
The following sections describe noun and verb morphology. Taking inspiration from toki pona (especially toki pona pre-pu!), I am allowing fairly free use of roots as both nouns, verbs, and modifiers. Inflectional morphology, whenever there is any, will give you a clue as to disambiguating parts of speech.
Essentially, I haven’t been able to decide how much inflectional and derivational morphology to steal from Old English as I try to preserve the vibes of toki pona; therefore, I’ve bifurcated the language into two registers:
Plain Register (a.k.a. Common Register), which avoids pesky morphological changes and is the “default” register. This register privileges Germanic roots.
Poetic Register (a.k.a. Liturgical or High Register), which incorporates a lot more inflectional and derivational morphology as well as more Latinate roots. It is reserved for–you guessed it–for biblical passages, poems, and the reported speech of one’s betters.
I’ve decided to nest discussions of register mostly within the relevant sections on parts of speech. There are some notes later in this document about more high level or pragmatic considerations vis-a-vis the registers. (Worldbuilding! Exciting!) Put briefly: The King, God, and your betters all speak in Poetic Register all the time, and don’t you forget it, you peasant!
In Reality, I tend to pretty freely move between the two registers as the whim takes me and as poetic meter might allow or dictate.
Any content word may act as a noun, as discussed just above. A noun phrase is made up of the noun and any accompanying particles or other modifiers.
Noun roots exclusively end in a. When one noun follows another, treat it as an adjectival modifier.
cilda ‘child’ hunda ‘dog’ wera ‘man’
cilda lȳta ‘small child’ lȳta cilda ‘a childish small thing’
mana wisa ‘wise man’ wisa mana ‘manly wisdom’
In common register, these basic nouns are unmarked for number. They are pluralized when modified with words like maniga ‘many’, maga ‘mightily’ (used in reference to strong things: bands of warriors or angels, say), swearma ‘a swarm,’ ǣla ‘all’ and so on.
In poetic register, the number of the subject of a sentence is marked on the verb.
The particle afa comes between a possessed noun and the possessing noun.
hūs afa cilda ‘a house of a child’
hunda thrym afa cilda a child’s fierce dog
Omit afa with genitive pronouns.
hūsa er => hūser ‘her house’
Ancish has two systems for personal pronouns, one for each register.
|Person |Nominative |Accusative |Genitive | |———— | ————| ————| ————| |1 |ic |me |me | |2 | tu| tu | tu| |3 |e | e| e|
|Person |Nominative |Accusative |Genitive | |———— | ————| ————| ————| |1 |ic (plural: we) |me (plural: us) |me | |2 |tu |tu (plural: ēо) |tu | |1Dual |wit |unc |unce | |2Dual |git |inc |ince | |3m |e |e |es | |3f |e |er |er | |3inanimate |it |it | its |
When ambiguous, pronouns in both the common and poetic registers can be pluralized with maniga ‘many’ and other quantity or group nouns, as is the case with nouns.
[! What now exactly?] The following might better be described as prepositions or even as case marking. I’m not sure what to call them (and maybe it’s really a mixed list) so I’m calling them particles.
These particles precede a noun and inform its syntactic role. For personal pronouns, these take the accusative form.
at - accusative (n.b. like afa, does not elide: cræft at e ‘created him’)
by - instrument or with someone
tura - through, by way of, via a route, during
for, to - dative meanings
in - locative, dative
Also note the demonstrative pronoun se ‘this/that’ and the relative pronoun sete ‘that which’.
Adverbials can sprinkle a sentence like blood droplets across a breastplate.
They end in either -e (Common Register) or -līc (poetic register)
hrǣde - quickly hrǣdlīc - quickly but fancy
in= - in, inside of, through e.g. infara ‘enter’ of= - off of, from out of, out of e.g. offara ‘exit’ =hād - (cild ‘child’, cildhād ‘childhood’) be= - “about, around, on all sides” =th - intensifies the meaning of a verb, marking it clearly as a “noun of action, state, or quality” (depth, strength, height(h))
[! Note] The following is a list of OE word arts that I stole from some website somewhere and haven’t thoroughly worked through how, if at all, I want to use them.
mis- defective (dǣd ‘deed’, misdǣd ‘misdeed’; faran ‘to go’, misfaran ‘to go astray’) ofer- excess (ǣt ‘eating’, oferǣt ‘gluttony’; fyllan ‘to fill’, oferfyllan ‘to fill to overflowing’) un- negative (cūþ ‘known’, uncūþ ‘unknown’; riht ‘right’, unriht ‘wrong’) al- wholly, fully, infinitely
common adjective suffixes include: -ful (cearu ‘care, sorrow’, cearful ‘sorrowful’) -ig (blōd ‘blood’, blōdig ‘bloody’) -isc (cild ‘child’, cildisc ‘childish’) -lēas (hlāford ‘lord’, hlāfordlēas ‘lordless’) -sum (childsome)
Abstract nouns often end in: -dōm (wīs ‘wise’, wīsdōm ‘wisdom’) -hād (cild ‘child’, cildhād ‘childhood’) -nes (beorht ‘bright’, beorhtnes ‘brightness’) -scipe (frēond ‘friend’, frēondscipe ‘friendship’)
[! Note] Currently I am of two minds as to how to have Ancish verbs work. Essentially there is the strategy more closely aligned with how toki pona works, and the strategy more like Old English. I’m going to take the time to write out both in the below so that I can better imagine what it might mean to choose between them. Maybe I’ll come up with some compromise.
The idea here is that, as in toki pona, there must be a marker that separates the subject from the predicate (i.e., the verb and everything that follows it). My idea for a long tie has been to do this with the various forms of am/be/was–a complicated morass of verb stems, as it is in modern English!
Pros: toki-pona inspired; simplest, most analytic Cons: Makes phrases longer and seemingly more repetative. Bad for poetry?
Present tense: be Past tense: wasa
Let’s assume for now that this follows the same logic as in toki pona; i.e., it only appears in the third person.
Therefore…
ic be scot at singa
This is the least worked out and I’m already tempted to drop it. The idea is that (like tp: li) everything would get a word separating the subject and the predicate, and that for that I’d used conjugated forms of be/wasa.
Pros: toki-pona inspired; can come up with paradigms for only a few stems if I do want to have some verbal conjugation.
beo bist bith beoth
|person|form| |–|–| |1s |beo| |2s|bist| |3s|bith| |3p|beoth|
Okay, I broke down and tried to make some decisions about what Ancish verbs would look like if they conjugated “fully” in the style of Old English. Because it will be absolutely clear what is the verb, given all of this morphology, there is no predicate marker necessary. Do I love this? No. Right now I’m considering moving it to the “sometimes food” category of the liturgical register or something similar. Maybe nobles conjugate their verbs in reported speech, lol.
|singular | plural | |— | — | |ic singa |we singad | |tu singast | tu singadast| |e singath |e singon |
If we follow Strategy 3, then we have an infinitives, imperatives, and past and present participles! But do we enjoy that? Suddenly it seems like a lot of verb forms for something that is supposed to cleve to toki pona…
(to) singan - infinitive to is not required here; only use it if it’ll make prosidy nicer.
(be) sing(a) - imperative singular uton singan - imperative 3rd person inclusive (jussive, ’let’s) singath - imperative plural be is not required
Use it here not to mark a continuous aspect; but only as an adjective.
=enda singenda - singing wera sigenda - ‘singing man’
Note that wera singa se would mean something more like ‘that singer.’
In OE the past participle could be used to form a perfect aspect with habban (i.e. “have’”). I don’t want this, I want to mark aspect in other ways (how?). So both participles are just adjectival.
ge=V=ed
singa - sing gesinged - sung singa maniga se gesinged - those sung songs
Use by after a past participle to mark its subject.
by mana Beowulfa gefared scopa singed at fara maniga se The bard sung (about) many deeds which Beowulf accomplished.
Use for to mark a past participle’s object.
For both past and present participles I haven’t made any firm decisions about word order. I’m leaving it up to the “context” gods.
Subjunctive
OE has this, but I don’t want it. We’ll do it in a more toki pona-isc way.
Investigate the etymologies for “ought” and “should” and “shall” and “would” etc. These seem right to have as full content words, not as endings.
These can come before verbs.
maga (ic maga crafta e se.) able, can
wīla - future auxiliary as well as expressing desire
scula - future aux as well as expressing duty
There are also two small verbal particles:
ana - specific, pointed, narrow
su - generic, broad
END
Stuff that if I delete it, the whole file breaks for some infuriating reason:
scula - future aux as well as expressing duty
Phrasal Verb Bits ** Particles follow ve ana - specific, pointed, narrow su - generic, broad, pointed, narrow