How exactly do Head-Initial Compounding work?

If you're new to these arts, this is the place to ask "stupid" questions and get directions!
Post Reply
User avatar
Stammalor
cuneiform
cuneiform
Posts: 143
Joined: 23 Apr 2012 16:14
Location: Scania

How exactly do Head-Initial Compounding work?

Post by Stammalor »

Hi guys.
I just want to start by saying that it has been like, seven or eight years since I used this forum so I have very little knowledge of how to use it correctly, but so here I go.
But I've been slowly returning to conlanging, and trying to not just create one that looks like my familiar Germanic ones, but I wanted to implement 2 features and while they initally seemed like no problem, they have now confunded me, and I'm failing at finding any easy-to-access source that explains how they work in natlangs, so here I am.

1. Head-Initial languages.
I get that adjectives comes after the noun. Sort off like they do in English pronouns (I want SOMEONE NICE). BUT, what about compound nouns? Because I assumed they worked they same as in English, that is, a SCHOOL BOY is a kind of boy, and not a kind of school. But the more I read about the subject where I could, it suggested that the opposite is true, that even compoun nouns are head-initial. So a SCHOOL BOY, would be a kind of school, probably one just for boys. I also read somewhere that in Italian (which is Head-initial, "capo familia" meaning "family head" / "head of family") a compound noun only ever consists of two parts, and I've heard other languages use constructions like "A school of boys" instead of compounding, or when compounding has many parts.
Does anyone know well how head-initial languages work with compound nouns, and maybe compounding in general? Because I like compounding and I don't know for sure how else you would create neologisms.

2. Passivity in ergative-absolute languges
So, because of how ergativity works and infering agents and patients, ergative languages often doesn't have a passive voice but instead an anti-passive one, ignoring the patient and spotlighting the agent. What I wonder is how these langauges deal with passivity? I've been thinking about it on my spare time and I can't get it to work. I can understand using a pronoun to use vagueness or ignorance about the agent in a sentence (SOMEONE killed Bob), but the passive is so usefull and there are so many adjectives (or adjective-like words) that hinge on an action peing performed UNTO someone, that I don't know how I, or the speakers I intend, would communicate without it. How would one say things like "We gotta help him, he's been both dumped and shot all in one night" without some way of using a passive. I also don't know exactly how I would implement a passive in a ergative-absolute language, if I chose to do so.
Does anyone have a good idea of how ergative-absolute languages deal with these things?

Thanks in advance for any help.

User avatar
Ser
sinic
sinic
Posts: 238
Joined: 30 Jun 2012 06:13
Location: Vancouver, British Columbia / Colombie Britannique, Canada

Re: How exactly do Head-Initial Compounding work?

Post by Ser »

Stammalor wrote:
01 Jan 2020 17:37
1. Head-Initial languages.
I get that adjectives comes after the noun. Sort off like they do in English pronouns (I want SOMEONE NICE). BUT, what about compound nouns? Because I assumed they worked they same as in English, that is, a SCHOOL BOY is a kind of boy, and not a kind of school. But the more I read about the subject where I could, it suggested that the opposite is true, that even compoun nouns are head-initial. So a SCHOOL BOY, would be a kind of school, probably one just for boys. I also read somewhere that in Italian (which is Head-initial, "capo familia" meaning "family head" / "head of family") a compound noun only ever consists of two parts, and I've heard other languages use constructions like "A school of boys" instead of compounding, or when compounding has many parts.
Does anyone know well how head-initial languages work with compound nouns, and maybe compounding in general?
It is as you say, so in those languages with head-initial compounding, the head goes at the beginning. So "boy school" would mean "schoolboy", and "boy school sport(s)" would mean "sportschool boy", and "boy school sport(s) Minnesota" would mean "Minnesota sportschool boy"...
Because I like compounding and I don't know for sure how else you would create neologisms.
The term "neologism" usually refers to any type of new word. Neologisms can come and be created in all sorts of ways, not just compounding. Some examples in contemporary English would be:

- "sashimi" (a borrowing of Japanese 刺身 sashimiꜜ 'sliced meat', a type of dish with raw sliced meat, including fish, optionally on rice)
- "boba tea" (a calque of Taiwanese Mandarin 波霸奶茶 bōbà nǎichá, itself a compound of 波霸 bōbà 'tapioca balls' literally "top_quality-balls" and 奶茶 nǎichá "milk tea". 波霸 bōbà is a compound of 波 bō 'ball', itself a borrowing of English "ball" apparently spread from Cantonese, and 霸 bà 'tyrant', a root commonly used as a pseudo-suffix in contemporary Mandarin to form slang words that refer to top-quality things or people, cf. 學霸 xuébà 'a top student', literally "study-tyrant")
- "bubble tea" (an eggcorn, or in other words a phono-semantic match of "boba tea", due to "boba" being unanalyzable foreign material)
- "to tweet sth" ('to publish sth on Twitter', a transitive semantic expansion of the existing intransitive verb "to tweet", said of birds)
- "to retweet sth" (the prefix re- + "to tweet")
- "router" (from "route" + the suffix -er)
- "bae" (clipped from "babe", itself clipped from "baby")
- "blog" (clipped from "web log")
- "vlog" (a portmanteau of "video" and "blog")
- "RAT" (an initialism from "remote-access Trojan")
- "CAPTCHA" (an initialism which was expanded as "Completely Automated Public Turing Test to Tell Computers and Humans Apart", although obviously the initialism was made to match an intended sound)
- "paywall" (a compound of "to pay" and "wall")
- "devOps" (a compound with clipped forms of "development operations", note that "dev" exists independently but *ops doesn't!)

Only the last two are compounds, but they're all neologisms. I chose these examples while making sure that each of them serves as an example of a different strategy to form neologisms (borrowings, calques, eggcorns, semantic expansions, affixations, clippings, portmanteaus, initialisms, and finally compounds).
2. Passivity in ergative-absolute languges
So, because of how ergativity works and infering agents and patients, ergative languages often doesn't have a passive voice but instead an anti-passive one, ignoring the patient and spotlighting the agent. What I wonder is how these langauges deal with passivity? I've been thinking about it on my spare time and I can't get it to work. I can understand using a pronoun to use vagueness or ignorance about the agent in a sentence (SOMEONE killed Bob), but the passive is so usefull and there are so many adjectives (or adjective-like words) that hinge on an action peing performed UNTO someone, that I don't know how I, or the speakers I intend, would communicate without it. How would one say things like "We gotta help him, he's been both dumped and shot all in one night" without some way of using a passive. I also don't know exactly how I would implement a passive in a ergative-absolute language, if I chose to do so.
Does anyone have a good idea of how ergative-absolute languages deal with these things?
They deal with them by using the absolutive. That is what the absolutive is all about: being able to become the subject of an intransitive verb while still being the semantic patient of the action.

sabrina.ERG shoot.PAST.ACTIVE 3SG.ABS
'Sabrina shot him.'

3SG.ABS shoot.PAST.ACTIVE
'He got shot.'

All you need to do is move the 3SG pronoun 'he' to the normal subject position. The verb "shoot.PAST" remains in the active voice all the same.


Now if you want to say "Sabrina shot (someone)", you could use the antipassive voice:

sabrina.ABS shoot.PAST.ANTIPASS
'Sabrina shot someone. Sabrina shot. Sabrina made a/the shot.'


Note that real-life ergative-absolutive languages often have a passive as well, besides the antipassive voice. This is sometimes because not all verbs can take the ergative-absolutive pattern in all situations (maybe they only do so in past tenses, or there's a subcategory of verbs that must take the nominative-accusative pattern). Sometimes there may be a semantic difference between "3SG.ABS shoot.PAST.ACTIVE" 'He got shot' and "3SG.ERG shoot.PAST.PASSIVE" 'He got shot' in some way.
Last edited by Ser on 02 Jan 2020 00:29, edited 5 times in total.
hīc sunt linguificēs. hēr bēoþ tungemakeras.

User avatar
Ser
sinic
sinic
Posts: 238
Joined: 30 Jun 2012 06:13
Location: Vancouver, British Columbia / Colombie Britannique, Canada

Re: How exactly do Head-Initial Compounding work?

Post by Ser »

Another thing I'd like to mention is that languages are often inconsistent on whether they're head-initial or head-final. For example, let's look at English and Classical Latin and compare them.

English:
- head-final in adjective + noun word order
- head-final in adverb + adjective word order
- head-initial in determiner + noun word order
- head-initial in noun + prepositional phrase word order
- head-initial in noun + relative clause word order
- head-initial or head-final in possessed + possessor word order ("X of Y" or "Y's X")
- head-final in noun + noun compounds (and these compounds are very productive in this language)
- [no adjective + verb compounds exist]

Classical Latin:
- except for some basic adjectives, normally head-initial in noun + adjective word order
- head-final in adverb + adjective word order
- head-initial or head-final in determiner + noun word order ("this field" or "field this")
- [nouns cannot be modified by a prepositional phrase directly]
- head-initial in noun + relative clause word order
- head-initial or head-final in possessed + possessor word order (e.g. "x.NOM y.GEN" or "y.GEN x.NOM")
- head-final in noun + noun compounds (very uncommon in this language)
- head-final in adjective + verb compounds (fairly uncommon in this language)

There are definitely languages that are fully consistent though. For example, Classical Arabic is consistently head-initial in all the categories above that can be applied to it, and Mandarin Chinese is almost consistently head-final in the same way (the only caveat is that Mandarin has verb+noun compounds that can be interpreted either way, i.e. either the verb or the noun can be the head with different meanings).
Last edited by Ser on 02 Jan 2020 01:45, edited 3 times in total.
hīc sunt linguificēs. hēr bēoþ tungemakeras.

Khemehekis
mayan
mayan
Posts: 2273
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: How exactly do Head-Initial Compounding work?

Post by Khemehekis »

Ser wrote:
01 Jan 2020 22:22
English:
- head-final in adjective + noun word order
- head-final in adverb + adjective word order
- head-initial in determiner + noun word order
- head-initial in noun + relative clause word order
- head-initial or head-final in possessed + possessor word order ("X of Y" or "Y's X")
- head-final in noun + noun compounds (and these compounds are very productive in this language)
- no adjective + verb compounds exist
Isn't English head-final in determiner + noun word order? On his farm he had some pigs, not *pigs some.
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 65,595 words and counting

31,416: The number of the conlanging beast!

User avatar
Ser
sinic
sinic
Posts: 238
Joined: 30 Jun 2012 06:13
Location: Vancouver, British Columbia / Colombie Britannique, Canada

Re: How exactly do Head-Initial Compounding work?

Post by Ser »

Khemehekis wrote:
01 Jan 2020 22:39
Isn't English head-final in determiner + noun word order? On his farm he had some pigs, not *pigs some.
Eh, it is the fashion in current syntax to say the determiner is the head in English. Personally I see some merit to either analysis.
hīc sunt linguificēs. hēr bēoþ tungemakeras.

Khemehekis
mayan
mayan
Posts: 2273
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: How exactly do Head-Initial Compounding work?

Post by Khemehekis »

Wow! What's the logic behind calling the determiner the head?
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 65,595 words and counting

31,416: The number of the conlanging beast!

Salmoneus
MVP
MVP
Posts: 1871
Joined: 19 Sep 2011 19:37

Re: How exactly do Head-Initial Compounding work?

Post by Salmoneus »

Ser wrote:
01 Jan 2020 22:07
- "devOps" (a compound with clipped forms of "development operations", note that "dev" exists independently but *ops doesn't!)
Other way around - "dev" isn't a word (except maybe in some sort of technical computerist jargon?) whereas "ops" is a very common word.

(AIUI, "ops" in situations where it's not just the plural of "op", but is instead an abstract, comes from the military and derives from the plural ("he conducts black ops" > "he works in black ops"). But it's long since spread beyond that. [the two meanings are potentially confusable with healthcare bodies. So I saw a headline talking about "NHS ops chief", meaning the COO ('ops' as abstract 'operational processes') - NOT a person who was in charge of NHS ops in the usual plural sense (medical procedures).)

Khemehekis
mayan
mayan
Posts: 2273
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: How exactly do Head-Initial Compounding work?

Post by Khemehekis »

Salmoneus wrote:
01 Jan 2020 23:34
Other way around - "dev" isn't a word (except maybe in some sort of technical computerist jargon?) whereas "ops" is a very common word.
I have a friend from the youth rights community who likes to talk about video games (especially No Man's Sky), and he often uses, or links to articles that use, the apocopation "dev" (for developer). "The devs at Hello Games have . . ."
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 65,595 words and counting

31,416: The number of the conlanging beast!

User avatar
Ser
sinic
sinic
Posts: 238
Joined: 30 Jun 2012 06:13
Location: Vancouver, British Columbia / Colombie Britannique, Canada

Re: How exactly do Head-Initial Compounding work?

Post by Ser »

Khemehekis wrote:
01 Jan 2020 22:55
Wow! What's the logic behind calling the determiner the head?
It's mainly an attempt to achieve a parallel analysis between traditional NPs on the one hand (where the head is traditionally a content word and determiners, which are function words, are dependents) and VPs, PPs and CPs on the other hand (where heads are function words, namely auxiliary verbs, prepositions and subordinators).

Wikipedia has a somewhat okay article on the topic...

https://en.wikipedia.org/wiki/Determiner_phrase

There is also a controversy going on at the present time on whether this the-determiner-is-the-head analysis can be extended to most or all languages, or only some languages like English. Here's an interesting article discussing how it could or could not apply to Estonian, a language without articles. (The author ends up concluding that non-article determiners show behaviour that favour the the-determiner-is-the-head analysis.)
Salmoneus wrote:
01 Jan 2020 23:34
Other way around - "dev" isn't a word (except maybe in some sort of technical computerist jargon?) whereas "ops" is a very common word.

(AIUI, "ops" in situations where it's not just the plural of "op", but is instead an abstract, comes from the military and derives from the plural ("he conducts black ops" > "he works in black ops"). But it's long since spread beyond that. [the two meanings are potentially confusable with healthcare bodies. So I saw a headline talking about "NHS ops chief", meaning the COO ('ops' as abstract 'operational processes') - NOT a person who was in charge of NHS ops in the usual plural sense (medical procedures).)
Oh, I wasn't familiar with "ops" as a real word. But now that you mentioned "black ops", I remembered a video game called Call of Duty: Black Ops, which had some popularity. I should've thought of that...

And yeah, "dev" is very common in the software development world (also the related world of gaming, as in Khemehekis's anecdote).
hīc sunt linguificēs. hēr bēoþ tungemakeras.

Salmoneus
MVP
MVP
Posts: 1871
Joined: 19 Sep 2011 19:37

Re: How exactly do Head-Initial Compounding work?

Post by Salmoneus »

I'd say the most common use of "ops" in UK English is simply the plural of "op", meaning "medical procedure". It can also be the plural of "op" meaning a short-term project of some sort, particularly in a military context (I think originating in military abbreviations, like "Psy-Ops" for "psychological operations"), or it can be an abstract meaning the operational processes of an organisation (eg in titles like "head of ops"). In the latter senses it's a bit pretentious, but it's very commonplace in the first of those meaning.


I had overlooked the video game community, which I guess is probably enough to push 'dev' into 'real word' territory.

User avatar
Stammalor
cuneiform
cuneiform
Posts: 143
Joined: 23 Apr 2012 16:14
Location: Scania

Re: How exactly do Head-Initial Compounding work?

Post by Stammalor »

Ser wrote:
01 Jan 2020 22:07
Stammalor wrote:
01 Jan 2020 17:37
1. Head-Initial languages.
I get that adjectives comes after the noun. Sort off like they do in English pronouns (I want SOMEONE NICE). BUT, what about compound nouns? Because I assumed they worked they same as in English, that is, a SCHOOL BOY is a kind of boy, and not a kind of school. But the more I read about the subject where I could, it suggested that the opposite is true, that even compoun nouns are head-initial. So a SCHOOL BOY, would be a kind of school, probably one just for boys. I also read somewhere that in Italian (which is Head-initial, "capo familia" meaning "family head" / "head of family") a compound noun only ever consists of two parts, and I've heard other languages use constructions like "A school of boys" instead of compounding, or when compounding has many parts.
Does anyone know well how head-initial languages work with compound nouns, and maybe compounding in general?
It is as you say, so in those languages with head-initial compounding, the head goes at the beginning. So "boy school" would mean "schoolboy", and "boy school sport(s)" would mean "sportschool boy", and "boy school sport(s) Minnesota" would mean "Minnesota sportschool boy"...
Because I like compounding and I don't know for sure how else you would create neologisms.
The term "neologism" usually refers to any type of new word. Neologisms can come and be created in all sorts of ways, not just compounding. Some examples in contemporary English would be:

- "sashimi" (a borrowing of Japanese 刺身 sashimiꜜ 'sliced meat', a type of dish with raw sliced meat, including fish, optionally on rice)
- "boba tea" (a calque of Taiwanese Mandarin 波霸奶茶 bōbà nǎichá, itself a compound of 波霸 bōbà 'tapioca balls' literally "top_quality-balls" and 奶茶 nǎichá "milk tea". 波霸 bōbà is a compound of 波 bō 'ball', itself a borrowing of English "ball" apparently spread from Cantonese, and 霸 bà 'tyrant', a root commonly used as a pseudo-suffix in contemporary Mandarin to form slang words that refer to top-quality things or people, cf. 學霸 xuébà 'a top student', literally "study-tyrant")
- "bubble tea" (an eggcorn, or in other words a phono-semantic match of "boba tea", due to "boba" being unanalyzable foreign material)
- "to tweet sth" ('to publish sth on Twitter', a transitive semantic expansion of the existing intransitive verb "to tweet", said of birds)
- "to retweet sth" (the prefix re- + "to tweet")
- "router" (from "route" + the suffix -er)
- "bae" (clipped from "babe", itself clipped from "baby")
- "blog" (clipped from "web log")
- "vlog" (a portmanteau of "video" and "blog")
- "RAT" (an initialism from "remote-access Trojan")
- "CAPTCHA" (an initialism which was expanded as "Completely Automated Public Turing Test to Tell Computers and Humans Apart", although obviously the initialism was made to match an intended sound)
- "paywall" (a compound of "to pay" and "wall")
- "devOps" (a compound with clipped forms of "development operations", note that "dev" exists independently but *ops doesn't!)

Only the last two are compounds, but they're all neologisms. I chose these examples while making sure that each of them serves as an example of a different strategy to form neologisms (borrowings, calques, eggcorns, semantic expansions, affixations, clippings, portmanteaus, initialisms, and finally compounds).
Well, that's gonna be a rollercoaster for my native head-final perspective.
I was inspired by this video called Grammar From Nothing by the YT channel Worldbuilding notes (https://www.youtube.com/watch?v=5OGDlp2XIdg ) which basically didn't want a language created fully from nothing but noticably have origin in earlier "cave man speech" as she put it. Essentially, I want to use methods like her in creating a proto-language from which I later could create descendants from. And I imagined the speakers, for all intents and purposes, to be alone and just have left their hunter-gathering state. So any neologism that stems from foreign languages or spelling conventions is not an option. I have so far found it rewarding trying to create a new word from stuff I already have rather than to just say that this collection of sounds now means this. For example, I wanted a word that could capture all variants of a win state (success, victory etc) and so I thought about what could work as a metaphore or picture-simile for hunter-gatherers. I eventually came up with red hand because red could infer blood, thus "bloody hand", which in turn could infer successful kill in a hunt, which in turn is a sort of win. I then filed that compound word down to something that word could be if it was said very often and for a long time, which eventually became my final word for success/victory. I've then tried to do something similar with complex words wherever I can. Even simple stuff like tent is made using this method.
I hope that clarifies my situation a bit. I'm using a lot of compounding right now to make new vocabulary, and I thought to, at least partially, use the "cave man method" to also create some of the morphological derivation.

2. Passivity in ergative-absolute languges
So, because of how ergativity works and infering agents and patients, ergative languages often doesn't have a passive voice but instead an anti-passive one, ignoring the patient and spotlighting the agent. What I wonder is how these langauges deal with passivity? I've been thinking about it on my spare time and I can't get it to work. I can understand using a pronoun to use vagueness or ignorance about the agent in a sentence (SOMEONE killed Bob), but the passive is so usefull and there are so many adjectives (or adjective-like words) that hinge on an action peing performed UNTO someone, that I don't know how I, or the speakers I intend, would communicate without it. How would one say things like "We gotta help him, he's been both dumped and shot all in one night" without some way of using a passive. I also don't know exactly how I would implement a passive in a ergative-absolute language, if I chose to do so.
Does anyone have a good idea of how ergative-absolute languages deal with these things?
They deal with them by using the absolutive. That is what the absolutive is all about: being able to become the subject of an intransitive verb while still being the semantic patient of the action.

sabrina.ERG shoot.PAST.ACTIVE 3SG.ABS
'Sabrina shot him.'

3SG.ABS shoot.PAST.ACTIVE
'He got shot.'

All you need to do is move the 3SG pronoun 'he' to the normal subject position. The verb "shoot.PAST" remains in the active voice all the same.


Now if you want to say "Sabrina shot (someone)", you could use the antipassive voice:

sabrina.ABS shoot.PAST.ANTIPASS
'Sabrina shot someone. Sabrina shot. Sabrina made a/the shot.'


Note that real-life ergative-absolutive languages often have a passive as well, besides the antipassive voice. This is sometimes because not all verbs can take the ergative-absolutive pattern in all situations (maybe they only do so in past tenses, or there's a subcategory of verbs that must take the nominative-accusative pattern). Sometimes there may be a semantic difference between "3SG.ABS shoot.PAST.ACTIVE" 'He got shot' and "3SG.ERG shoot.PAST.PASSIVE" 'He got shot' in some way.
Now, I see where I made my mistake, I thought that the absolutive case was to be used in all contexts where a singular verb was used, regardless of valency. That is "Bob(ERGATIVE) sees a rabbit(Absolutive)"
Bob sees a rabbit.
But also "A rabbit(absolutive) sees"
A rabbit sees (in general. Static verb)

But the absolutive is only for if the singular verb is intransitive, as in "Bob(absolutive) falls". But then what about verbs that have more than one possible valency? And adjectives that imply passivity? Would you just need to make several sentences where English would have one? Like, "I cried for help and said I was looking for my dumped friend". Would "I" be absolutive or ergative? Since "cry" can be either transative or intransitive. And what about "dumped" which here works just like an adjective, but clearly implies passivity?
I don't currently intend to have nominative-accusativity mixed with ergativity in this language, if that matters.

Thanks for all the help so far!

Post Reply