Play scratchpad

[Pabappa]

I'd like to return to posting here, perhaps not as frequently as I was during 2017-19 when I had mobile-only Internet and not much else to do. I'm mostly on Discord nowadays, but it seems lately I fill up the conlangs channel with notes that are mostly for myself, so I thought I could put those here. Ive been posting on FrathWiki as well, and for certain things the wiki format is the best way to go, but for other things it might be better for me to post here.

One reason I left is that I can't post http or screenshot images here (only https) and I prefer to use my old-school HTTP-only website rather than a cloud service to host images. I suppose I'll still use Discord for images.

I think the idea of a scratchpad is that it is mostly for my own use, and therefore I'm not expected to explain everything I write. If people are curious about certain things, I'd be glad to answer, but Play's grammar is quite exotic from the standpoint from English and I've noticed that my explanations seem not to make sense. Even so, some of what I write will mix in "lesson" narration with the explain-nothing style I use in my private notes. After all, I tend to forget things myself.

--------------

One feature distinguishing Play from nearby languages is that its 1>2 verb marker is a zero morph, and the 2>1 marker is usually a doubling of the vowel. (All verbs have two arguments, whether stated or not.) Thus one can say

ti deed; what I am doing to you
tii what you are doing to me
Tiitīs? What are you doing to me?
Tiipipu I know what you're doing to me.

These are not true person markers however, as they remain in the same form when non-SAP's are included. These are placed before the verb:

Tanapu tibitīs? What did the rabbit do to you?
Tatup tibitīs? What did I do to the man?
Tanapu tatup tibitīs? What did the rabbit do to the man?

Thus, the verb inflections assign the agent and patient roles to the verb's two arguments. If neither argument is given, the two arguments are the speaker and the listener. If only the first argument is given, the second argument is the listener. If only the second argument is given (distinguished by final -p), the unstated first argument is the speaker. This can be overridden by particles when referring to previously stated entities in another clause or sentence, but if there are no particles given at the beginning of a clause, the default SAP values reappear even within one sentence.

[Pabappa]

Like many other Lava Bed languages, Play not only uses person marking on nouns, but these markers can be transitive, meaning that nouns are marked for the donor and the recipient. These in turn can couple to mood markers, so one can say

šappem hiking boots

Šappiŋup. I have my hiking boots. (Note that this also means "my hiking boots" when used as an argument of a verb)
Šappiŋus. You have your hiking boots. (And this thus means "your hiking boots")
Šappiŋūm. I have your hiking boots.
Šappiŋūs. You have my hiking boots.

Šappiŋupa? May I have my hiking boots?
Šappiŋusa? Do you want your hiking boots?
Šappiŋūmna? May I have your hiking boots?
Šappiŋūsa? Do you want my hiking boots?

Šappiŋuppa. I need my hiking boots.
Šappiŋupa. Please take your hiking boots.
Šappiŋūmpa. I need your hiking boots.
Šappiŋūpa. Please take my hiking boots.

(Note that the two italicized forms are homophonous. Like English, Play uses question intonation, and also can pad these words with additional mood markers. However, on paper somewhere I drew up an alternative form that would remove even the potential for confusion. I just dont have it at hand right now and dont remember which form was replaced.)

The possession markers above are understood loosely. In a store, someone buying boots would refer to those on sale as "your" boots to the listener, and this would be understood without confusion. This is one reason why Play gets by with such long words for buying and selling. If a specific pair of boots is meant, the constructions above still prevail, because as stated in the post above, these are not true person markers; the "you" is really "non-self" and only becomes 2nd person when argument slot 2 is empty. Thus one can say

Fumūmnap šappiŋūmpa. I need the hunter's hiking boots.

The word for hunter, fumūmna, is in the 2nd argument slot, marked with -p, because the hunter is the patient of the verb. The listener is not present in this sentence even though it is understood that the listener is likely to be the one to provide the boots to the customer.

However, it is also possible to specify a specific item within the listener's repertoire. One can also say

Fumūmnas šappiŋūmpa. I need your hunter's hiking boots.

Here, the word for hunter is assigned the observer role (OBS), marked by -s, because the listener is again the patient of the verb, even though the boots belong to a hunter. (This is generally the case for generic patients such that we might translate them into English as "hunters' boots" rather than "hunter's boots".)

[Creyeditor]

One feature distinguishing Play from nearby languages is that its 1>2 verb marker is a zero morph, and the 2>1 marker is usually a doubling of the vowel. (All verbs have two arguments, whether stated or not.) Thus one can say

ti deed; what I am doing to you
tii what you are doing to me
Tiitīs? What are you doing to me?
Tiipipu I know what you're doing to me.

These are not true person markers however, as they remain in the same form when non-SAP's are included. These are placed before the verb:

Tanapu tibitīs? What did the rabbit do to you?
Tatup tibitīs? What did I do to the man?
Tanapu tatup tibitīs? What did the rabbit do to the man?

Ohh, a direct-inverse system with non-concatenative morphology (and pro-drop?). Neat

[Pabappa]

I've partly rescued my time-counting system from a writeup in which I didnt put down most of the etymologies. For the time being, I dont have words for "today" or "yesterday" but I do have a word for "tomorrow" and the other two will be built the same way as this, so there shouldn't be much fuss.

BASIC ROOTS:
Most of these words are normally padded by classifier suffixes, but here they are left off for clarity. They can take more than one classifier suffix.

pip, the sun
vam, the moon
tiu, star(s)
tāpa, the planet's orbital path (Play speakers believe in heliocentrism)
pipata, to play with someone's eyes

ža ~ va, a cycle that recurs (a merger of two unrelated morphemes, both of which surface as just /a/ under some conditions)
papa, a record of time

-fa, in the sky

ku, to approach (which appears as /pu/ due to a regular grammatical operation)

These morphemes must be put together in a rigid order. Thus one can say

tiuvafa, a single night
pipafa, a 24-hour day
pipatavafa, a single daytime period ("when the sky plays with (my) eyes")
tiuvapufa, tonight; this coming night
pipapufa, tomorrow (the whole day); today (if said in early morning)
pipatavapufa, tomorrow (daytime); today (if said in early morning)

tāpavafa, one year, counting from the first day to the last
vamavafa, (possibly) one month. Im not sure I want this to use the word for moon directly since the moon does not take a month to rise and set. This seems like it could instead be another word for day, but perhaps with the meaning "24-hour period" or from tide to tide. Moreover, the global calendar uses irregular months that aren't directly linked to lunar cycles even though they happen to average about 28 days.

There is no equivalent of a week, nor are there strictly defined subdivisions of the days; not even an equivalent of the hour. The Play homeland is on the edge of the tropics, where the day length varies little throughout the year, so work generally follows the cycle of the sun.

When counting time, different words are used. The underlying morpheme structure is similar in meaning to saying "I have ___ years".

tāpapapa, a year (counting)

Tāpapapa putašap. I am five years old. (I haven't gone into numbers beyond ten yet. Getting the number system up to ten was hard work enough and I don't want to add more without being sure I won't need to rework them later.)

A simple serial -s can link time words to the sentences they govern:
Pipapufas, tāpapapa putašap. Tomorrow, I will be five years old. (It is not necessary to use a tense marker here, just as in English some people say "tomorrow I am five years old".)

A note on the infixes: currently Play doesn't have any demonstratives. It's possible that all demonstratives will consist of infixes like this -pu- "which (we) approach". These affixes always appear after the stem of the noun and before the classifier suffix, and importantly, they require that a classifier suffix be present. Most Play nouns, including all animate nouns, have no classifier suffixes. If I go with this system, the only nouns capable of taking demonstratives would be those that take classifier suffixes. However, all handheld objects would be included, along with various other groups of nouns such as clothes, furniture, fruits, and various objects found in the sky, sea, and (to a lesser extent) the land.

Ohh, a direct-inverse system with non-concatenative morphology (and pro-drop?). Neat

Thank you. Play lacks pronouns entirely, and I would say that as best as I can manage it, it lacks person markers too, since the meanings of the verbal markers depend on the arguments passed to the verb, which as explained above include null morphemes. But for convenience I still write them as 1>2 and 2>1, which I'd say mean both "1st and 2nd person" and "argument 1 and argument 2".

[Creyeditor]

Could it be interpreted as marking the more prominent argument acts on the less prominent argument and vice versa?

[Pabappa]

Im honestly not sure. If it can, it's worth noting that the 1st person always occupies the first slot (by being omitted) and the 2nd person always occupies the second slot. Thus, the 1st person is higher on the animacy hierarchy than the 2nd person. This is attested in the wild, but uncommon. I think it makes sense because there will always be people saying "i think", "i know", etc and only less commonly do we speak for the emotions of others, but one major flaw in the Play system is that all verbs with 2nd person agents, even if the patient is 3rd person, must use the double-vowel form of the verb, which is marked, and therefore can be seen as inferior or inverted. This makes it seem as if even animals are higher on the animacy hierarchy than the listener. For example (note that -ib- is a past tense infix; the verb is fūpu):

Māpāpu fūpibu. The wolf chased you.
Māpāpu fūpupibu. You chased the wolf.

The wolf needs to be in the first argument slot in both sentences, because the 2nd argument slot is occupied by the listener (again, occupied by omission). This makes it behave as the agent by default, which can only be changed by flipping the orientation of the verb, which means doubling the vowel (and in this word, also inserting /p/).

I didnt realize this until well along in the design of the language. Play's efficiency relies on its rigid structure, with all the slots occupied. This has always been hard for me to explain, for example on Discord, since I can't just add a slot, add an affix, or do any of the things that most people would do with their languages ... not without spoiling the mechanics that make it work, which rely on having a fixed word order, a fixed number of morphemes in each word, and so on.

So I've decided to just own it, and let Play be the language of verbal aggression. However, there is one thing I can still do: in the first sentence above, the word māpāpu could instead be māpāpūs, which means "the wolf that got you", with a transitive person marker (3>2) like the words in the sentences in the first post. This is a marked form, but using it helps balance the sentences a bit, and since these suffixes only appear when a SAP is the patient, I can at least say that the 2nd person and 3rd person are about equal on the hierarchy, though the 1st person is clearly high above (because if I do this for 2nd person patients, I need to do it for 1st person patients as well: māpāpūm "the wolf that got me" (always in the argument 2 slot).

So, to summarize:
māpāpu, wolf (1st slot; typically but not always the agent; cannot occur if the 1st person is in the sentence)
māpāpup, wolf (2nd slot; typically but not always the patient; cannot occur if the 2nd person is in the sentence)
māpāpūs, wolf (1st slot; indicates a 2nd person patient, and is therefore an agent; can stand alone to mean "the wolf that got you")
māpāpūm, wolf (2nd slot; indicates a 1st person patient, and is therefore an agent; can stand alone to mean "the wolf that got me")

In some Lava Bed languages, such as Galà, which the Players considered the world's most difficult language, ALL NOUNS must be marked for the roles of the 1st and 2nd person. Thus they say the equivalent of "hammer-notme-notyou" for every single noun, unless that noun happens to relate to the listener or the speaker, most commonly as an agent or patient. Some day perhaps I can explain how they manage to wring a workable grammar out of this headache of a system, but since Play is a Lava Bed language, i mention it here to show that the Play system I described above can be seen as a subtype of the typical Lava Bed grammar. In Play, any noun appearing in the 1st argument slot demotes the 1st person to the observer role (OBS), and likewise any noun appearing in the 2nd argument slot does so for the 2nd person.

[Pabappa]

Because all Play verbs MUST have two arguments (stated or unstated; usually agent/patient or identity/observer), another thing I struggle with is handling intransitive verbs like English "I'm sorry" / "I apologize". There is simply no way to translate this into Play; I will need a verb for "to apologize to someone" and a different verb for "to apologize for something" but there can never be a verb that simply means "apologize".

Moreover, due to the lack of pronouns, there is also no means by which to indicate a SAP patient for the "apologize for" verb. Partly I can get around this by making "apologize to someone" be the base form of the verb and "apologize for" the derived form, but I still don't have a way to indicate both on one verb when one argument is a SAP.

But that doesn't mean there's no solution. I may simply need to do something exotic, something a natlang wouldn't need to do, simply because no natlang has a grammar quite like Play's. For example, perhaps there will be a verb that means "to do something to someone", and this will stand in a clause of its own marked with the serial verb ending -s. Thus one might say

Vanuap šatetavibe, vapa tis.
I apologized for the mess, and I did something to you.

Leipzig glossing is difficult and I've been resisting it, but here vapa is the word that primes the listener that a third argument to the verb is coming; tis is a dummy verb that means "to do (something) to someone" but I may need to make a different verb for it.

[Pabappa]

I came up with this a month ago but posted it on Discord only and forgot about it until now.

In Play, objects possessed by SAPs are still considered third-person arguments. This is unlike some related languages, including some of its descendants, where the equivalent of "your key" would take 2nd person agreement and behave just like a second person pronoun. This is important because third-person nouns can appear in either the first or second argument slot in Play.

šešipa, key (appearing as šešikisa when possessed by the listener or by the second argument of a verb)
tus, door (appearing as tufup when used as the second argument of a verb)
paku, to open (appearing as paup when used as an A-stem, as often in compound verbs)

Play has a lot of fusion. What looks insanely difficult is indeed difficult, but rather than look for and memorize dozens of overlapping patterns, Play learners were taught to simply learn two stems for every word in the language, and apply certain other rules on top of these.

I also have a logical contradiction, I realize, since the door is not the owner of the key. It may be that šešikisa is still the historical inflection, marking the noun as being in the OBSERVER theta role, a role that only Lava Bed languages tend to need. All I can say is that šešikisa effectively means "your key", whatever the underlying logic may be.

Šešikisa tufup paku. Your key opens the door.

This is taken literally, not as a statement of general truth; it means the key is opening the door at that very moment. Now some logical modifiers about need and sufficiency:

Šešikisa tufup pauppapa. Your key can open the door. (Your key is all you need.)
Šešikisa tufup paupava. Your key can open the door. (You need both your key and something else, perhaps a card.)
Šešikisa tufup paupapa. Your key can open the door. (Your key is all you need, but if you don't have it, something else will also work, perhaps a card.)
Šešikisa tufup paupame. Your key cannot open the door. (But something else will, perhaps a card.)

These sentences all have positive polarity even though the last is translated into English with a negative word. In English, it would be like answering the sentence "Can my key open the door?" with "Your card can." without telling the listener "no".

Logical modifiers like these are common on Play verbs, and Play speakers often provide more information than speakers of other languages would. The set above, and many others like it, can be thought of as a 2X2 matrix corresponding to the four truth values "X and only X", "both X and Y", "X or Y", and "not X but Y". (I dont know how to express this more succinctly).

These are distinct from the verbal moods which modify the entire sentence. For example, using the potential mood, one says

Šešikisa tufup pakutau. It's possible that your key is opening the door.

Which says nothing about whether a key alone is sufficient or necessary, but does bind the action to the present tense. Although moods are written as suffixes, they occupy a different slot and could be considered independent verbs. This means that the potential mood can be combined with the habilitative content words above:

Šešikisa tufup paupapatau. Your key might be able to open the door (and if you don't have it, a card might work.)

[Pabappa]

Although I say all Play verbs have two arguments and thus are transitive, I admit there is a large class of verbs that I've been using essentially intransitively. This is not an intractable problem, as English has equivalent constructions like "this one gets you" where a generic you is meant. In Play one can say

Tum namekāsa, pumi mippāsa, taā tavaŋa.
The baby acts on instinct, the elder acts on wisdom, the child acts on thoughts.

This perhaps could be the motto of a school. Here, all three verbs end in -ŋa, which both means "act on" and effectively turns the targeting of the verb's patient off. (Note that /sŋ/ always surfaces as /s/.)

I may be able to say that the patient of these verbs is the whole stem except for the /-ŋa/ suffix. In any case, for all verbs that I want to function as effectively intransitive, I will need to make sure they end in suffixes like these, so that ordinary verbs like tavanite "think about" will not be confused with them.

EDIT: I'm also adding -va, another suffix that turns off targeting, making a verb effectively intransitive. The difference here is that /va/ is semantically empty, just making a verb intransitive, whereas /ŋa/ and some others will both do that and also add meaning to the verb. So for exmaple tavava "think", etc. This suffix is not used often because much of what is intransitive in English will be transitive in Play ... Play is more likely to use the equivalent of expressions like "give it a thought" when not using traditional transitive verbs.

[Pabappa]

fivāpia "dirt"
vapias "war"

Possibly fivāpia be vapiafu "dirt and war".

Thus the expression for "and; with; accompanying" is X be Y-fu. I don't see a way for me to do it without a suffix at the end, since otherwise any inflections on the phrase would only apply to the second element. This "capping" or "padding" requirement is common to my other languages too, even ones that are very different from Play. This is in part because there is no copying of inflections to same-state nouns in such clauses. Thus the sentence

Fivāpia be vapiafup fušīpibe!

Can only mean "I say no to dirt and war!" even though only the word for war is marked with the patient suffix. Put another way, Play's word for "and; with" is -fu, and be was a helper word (originally meaning something like "with hold on; holding") that became mandatory as longer phrases came into common use. It's possible I may allow constructions without be for idiomatic usage or for certain very common phrases. For example tata tamafu sounds like it would be a nice way to say "girls and boys".

Alongside this construction, I might have yet another, where the two words are compounded in their B-stems, as though an equative compound, and then capped by -bu, which is not cognate to the -fu suffix above (but is actually cognate to the /be/). Thus I could say that the construction X-Y-bu means "X that is with Y" and X-Y-Ø means "X that is Y". This represents an evolution from the parent language where only the zero-marked equative compound existed. This would yield

tatatamabu "girls and boys"
tatatama "girls that are boys"

I should note, since I havent yet, that while Play has a plural marker bem, attached to the end of any noun (but before the classifier suffix if there is one), it is not often used, and number is often determined by context.
edit: the full sentecne above was originally an imperative sentence, but i changed it to 1st person indicative because i remembered that if the 2nd person is the agent, even in an imperative sentence, the -p that i call the patient suffix does not actually appear.