"Experiencer"

[WeepingElf]

I have spotted a terminological mistake in the grammars of Old Skourene and Tžuro, which has misled other conlangers. In these grammars, the word experiencer is used to mean 'intransitive subject'. This, however, is not how it is used in linguistics. Rather, it refers to a semantic role that denotes a being which experiences something. An experiencer is not always an intransitive subject; there are plenty experiencers that are not, and intransitive subjects that aren't experiencers. In the following sentences, the boy is an agent, the girl is an experiencer, and the ball is a patient:

The boy threw the ball.
The girl saw the ball.
The boy gave the girl the ball.
The boy laughed.
The girl was happy.
The ball rolled away.

[bradrn]

I have spotted a terminological mistake in the grammars of Old Skourene and Tžuro, which has misled other conlangers. In these grammars, the word experiencer is used to mean 'intransitive subject'. This, however, is not how it is used in linguistics. Rather, it refers to a semantic role that denotes a being which experiences something. An experiencer is not always an intransitive subject; there are plenty experiencers that are not, and intransitive subjects that aren't experiencers.

I’m pretty sure I’ve seen it used both ways, though as you note, use as a semantic role is more common. I have my own problems with OS and Tžuro, but this particular choice of terminology is not one of them.

[WeepingElf]

I have spotted a terminological mistake in the grammars of Old Skourene and Tžuro, which has misled other conlangers. In these grammars, the word experiencer is used to mean 'intransitive subject'. This, however, is not how it is used in linguistics. Rather, it refers to a semantic role that denotes a being which experiences something. An experiencer is not always an intransitive subject; there are plenty experiencers that are not, and intransitive subjects that aren't experiencers.

I’m pretty sure I’ve seen it used both ways, though as you note, use as a semantic role is more common. I have my own problems with OS and Tžuro, but this particular choice of terminology is not one of them.

Indeed, there are other weaknesses in these and other languages of Almea. I feel as if zompist doesn't have a particularly good grasp of ergativity, for example. This shows in Old Skourene and Tžuro, and I seem to remember seeing something about an ergative language with OVS word order somewhere on Almeopedia, but I don't remember which language it was. Some of the sound changes of the Eastern languages look quite contrived, and some languages simply don't feel realistic to me, though I am not sure where the wrongness actually lies.

But one should be very careful with saying, "This is not the way it works in natlangs" - there are many natlangs that do things that seem pretty weird. There is no way knowing them all. Also, one has to keep three points in mind:

1. Almea began as a Dungeons & Dragons campaign setting when zompist was in high school, which explains some things. For instance, the player character races of 1970s D&D are still recognizable though they are altered in many points: Elcari are essentially Dwarves, Flaids are essentially Halflings, and Ilii are essentially Elves. Verdurian was pieced together from various modern European languages, and the Eastern family created by working backwards from Verdurian to Cadhinor and from there backwards to Proto-Eastern. And working backwards is always the more difficult direction in making a conlang family.

2. The languages of Almea are, technically, exolangs, and the language faculties of Almean sapients may work differently from ours. Even the Uesti or "Almean humans" are a different species than us. However, I have for a long time felt that "They aren't human!" is a lame excuse for a poorly-designed conlang. Yet, we know nothing about the languages sapient beings on other planets may speak.

3. Almea is, fundamentally, fantasy, not science fiction - zompist one said so, and it ties in with the origin of Almea as a Dungeons & Dragons setting. The standards of realism are somewhat lower in fantasy than in science fiction.

Over all, Almea is a very impressive and lovingly detailed conworld, and making more than 20 well-developed conlangs is quite an achievement. The languages of Almea may not be the most brilliant fictional languages ever made, and there may be weak spots in some of them, but they aren't really bad either. There's no Gnommish nor a Grey Company Elvish among them.

[zompist]

1. Almea began as a Dungeons & Dragons campaign setting when zompist was in high school, which explains some things. For instance, the player character races of 1970s D&D are still recognizable though they are altered in many points: Elcari are essentially Dwarves, Flaids are essentially Halflings, and Ilii are essentially Elves. Verdurian was pieced together from various modern European languages, and the Eastern family created by working backwards from Verdurian to Cadhinor and from there backwards to Proto-Eastern. And working backwards is always the more difficult direction in making a conlang family.

College, not high school; and the flaids are not halflings— indeed, I invented them when I was 11, with no D&D influence at all. I emphasize this because this seems like the sort of criticism that imagines that languages, and worlds, are made by taking real-world things and filing off the serial numbers. We are limited to this world for inspiration, of course, but "Almean group X is just earthly group Y" is very rarely how I operate.

Verdurian was my first published conlang and there's much I don't like about it. It was thoroughly revised once but it's too late to do it again. (Also there is an in-universe, or in-multiverse, reason for the similarities.)

2. The languages of Almea are, technically, exolangs, and the language faculties of Almean sapients may work differently from ours. Even the Uesti or "Almean humans" are a different species than us. However, I have for a long time felt that "They aren't human!" is a lame excuse for a poorly-designed conlang. Yet, we know nothing about the languages sapient beings on other planets may speak.

I'm not sure what things you're referring to. If it's things like linguistic "universals", I hope you realize that these often turn out not to be universal, and that the particular set of features that dominate on Earth are quite arbitrary. There are some excellent linguistic takedowns of proposed universals; it's pretty amusing how often the violator turns out to be English or some other well-known language.

Of course, I want conlangers to learn linguistics, as it broadens their horizons. But some of the most interesting conlangs are those that purposely shy away from naturalism. (The only time I tried this on Almea is with Elkarîl, as it's spoken by non-humans.) Toki Pona, or Fith, or Lojban are not realistic, but they're good precisely because they're not.

3. Almea is, fundamentally, fantasy, not science fiction - zompist one said so, and it ties in with the origin of Almea as a Dungeons & Dragons settings. The standards of realism are somewhat lower in fantasy than in science fiction.

With conlangs it works the other way. What science fiction work does conlangs as well as Tolkien? Many fantasy works at least have a conlang; few science fiction works bother.

[WeepingElf]

1. Almea began as a Dungeons & Dragons campaign setting when zompist was in high school, which explains some things. For instance, the player character races of 1970s D&D are still recognizable though they are altered in many points: Elcari are essentially Dwarves, Flaids are essentially Halflings, and Ilii are essentially Elves. Verdurian was pieced together from various modern European languages, and the Eastern family created by working backwards from Verdurian to Cadhinor and from there backwards to Proto-Eastern. And working backwards is always the more difficult direction in making a conlang family.

College, not high school; and the flaids are not halflings— indeed, I invented them when I was 11, with no D&D influence at all.

I understand. The flaids have always reminded me of Tolkien's hobbits, though.

I emphasize this because this seems like the sort of criticism that imagines that languages, and worlds, are made by taking real-world things and filing off the serial numbers. We are limited to this world for inspiration, of course, but "Almean group X is just earthly group Y" is very rarely how I operate.

Verdurian was my first published conlang and there's much I don't like about it. It was thoroughly revised once but it's too late to do it again. (Also there is an in-universe, or in-multiverse, reason for the similarities.)

Fair. At least, it is one of the best-documented fictional languages I have seen so far, which is what I (and many other conlangers) like it for. I think this was the reason for David J. Peterson to grant it a Smiley Award. Also, I have been feeling for long that it makes sense to have the language of a fictional culture resemble the languages of comparable cultures on Earth, and Verduria always felt pretty European to me - I can't say which country in particular, it has aspects of Britain, Germany and various Romance countries. Cadhinor could be said to be Almea's Latin, and Cuêzi Almea's Greek. The early history of Xengiman reminds me of India, with Ezičimi as "Aryans" and We:dei as "Dravidians", though Xurno is more comparable to Italy as a country that is a second-rank power in terms of military strength and economic performance, but justly proud of its achievements in the fine arts - and its political system, a country ruled by an art academy, is unique!

2. The languages of Almea are, technically, exolangs, and the language faculties of Almean sapients may work differently from ours. Even the Uesti or "Almean humans" are a different species than us. However, I have for a long time felt that "They aren't human!" is a lame excuse for a poorly-designed conlang. Yet, we know nothing about the languages sapient beings on other planets may speak.

I'm not sure what things you're referring to. If it's things like linguistic "universals", I hope you realize that these often turn out not to be universal, and that the particular set of features that dominate on Earth are quite arbitrary. There are some excellent linguistic takedowns of proposed universals; it's pretty amusing how often the violator turns out to be English or some other well-known language.

Concurred - linguistic universals are a dubious business, and may be just as pseudoscientific as glottochronology. This also reminds me of a meaningless quarrel I once had with another list member on CONLANG. It was about the way active-stative alignment worked in my old conlang Nur-ellen, which that guy, a specialist in North American indigenous languages, claimed was wrong - he said that all active-stative languages were head-marking, while Nur-ellen was dependent-marking. I now feel that it was like an Africanist saying that the tones in an East Asian style conlang were done "wrongly" - African tone systems are about as different from East Asian tone systems as either are from European accent systems. Likewise, North American active-stative languages may always be head-marking, but European ones could be a different matter entirely (as Georgian actually is).

Of course, I want conlangers to learn linguistics, as it broadens their horizons. But some of the most interesting conlangs are those that purposely shy away from naturalism. (The only time I tried this on Almea is with Elkarîl, as it's spoken by non-humans.) Toki Pona, or Fith, or Lojban are not realistic, but they're good precisely because they're not.

Yep. Realism is not the one and only quality criterion for conlangs! Exploring patterns not found in natural languages surely is a legitimate and entertaining endeavour.

3. Almea is, fundamentally, fantasy, not science fiction - zompist one said so, and it ties in with the origin of Almea as a Dungeons & Dragons settings. The standards of realism are somewhat lower in fantasy than in science fiction.

With conlangs it works the other way. What science fiction work does conlangs as well as Tolkien? Many fantasy works at least have a conlang; few science fiction works bother.

Indeed, what regards conlangs, fantasy outperforms science fiction! Tolkien's conlangs are among the best ever made (though the question whether languages of immortal beings would change the same way as human languages, is legitimate, and Tolkien reflected on it); science fiction writers who pay any attention to language rather tend to be into the Sapir-Whorf hypothesis - and fail to describe the languages in detail.

[sasasha]

Verdurian was my first published conlang and there's much I don't like about it. It was thoroughly revised once but it's too late to do it again. (Also there is an in-universe, or in-multiverse, reason for the similarities.)

I’ve been doing the nuts and bolts work of learning Verdurian recently ‒ like, learning it with a serious intent to build written and spoken fluency. So I thought I’d offer a few reflections.

One reason I like and respect it is that it unapologetically doesn’t always feel perfectly congruous. The layers of its history have left interesting lacunae, odd lexical moraines, aesthetic departures, and various other cul-de-sacs and anomalies. And not only, by any means, unintentionally: potentially unconscious ‘lumpy’ processes underlying its 40 plus year development seem more than balanced by conscious ones. The more I explore it the more I’m reminded of something ‘real’, biological-feeling ‒ a big, unique structure, holding a smattering of unexplained features inside a law-abiding but complex set of naturalistic frameworks. And all the while, it’s perfectly useable for purposes ranging from gossip to chemistry to poetry to (increasingly) music theory. For these reasons, for me, Verdurian has a qualitative realism far exceeding that of most conlangs I have come across.

As to the similarities with Earth languages. Aside from the in-world explanation, which works as well as any, this aspect of Verdurian gives it this... chimaeric quality to me... Sometimes it sits hard in the uncanny valley; other times it harmonises surprisingly effortlessly, and kinda hangs between, the neighbouring neural frameworks sitting in my brain that house natlangs and auxlangs. It seems to make meta-commentary on itself: by existing in this strange semi-borrowed place it manages to be both of our world and not of it. I find that psychological aspect of it most peculiar and interesting, and it has been creatively fertile for me.

We can probably all agree that it is somehow on the fantasy side of conlanging, and that some conlanging is yet more obsessed with being described as forensically as natlangs can be described by modern linguistic science. Zompist’s style seems always to have been an in-between one: between the scientist and the artist. (I think it manages both prioroties with aplomb.) It’s in-between-y in another dimension too: between the creator and the other (the reader... the ZBB denizen, even...). With decades of being opened up to readers in various divers ways, Verdurian seems to exists to be observed – and thus, once again, more than most conlangs I know, may more readily be observed to exist. Incidentally, I think this is why zomp’s grammars aren’t even more formidably comprehensive and forensic than they are - it seems to me that his work isn’t so much in service of satisfying his own need to prove that he is a formidable scientist as it is in satisfying a potential onlooker to find both scientific and artistic interest therein.

I remember embarking on conlanging as a kid and thinking “I’m not going to set my sights too high; just... maybe... if I could create something... oh, about as complex as English.”

How fun it was, over time, to learn how absurd that thought was.

Well... I like and respect Verdurian, because while I wouldn’t claim it has achieved this benchmark (as if one could), it feels like it’s hitting that strange place of design which is so ‘feels like it organically grew in a complex, at times chaotic manner over a realistically bumpy history’ that it kinda reminds me of my childhood dream of a conlang “oh, about as complex as English”. Complex, in that it actually feels used.

This was of course just a little subjective detour in this topic; I mainly wanted to bring up that there are different types of realism.

[bradrn]

Indeed, there are other weaknesses in these and other languages of Almea. I feel as if zompist doesn't have a particularly good grasp of ergativity, for example. This shows in Old Skourene and Tžuro, and I seem to remember seeing something about an ergative language with OVS word order somewhere on Almeopedia, but I don't remember which language it was.

I can’t help but agree with this — I’ve always felt OS and Tžuro are the weakest of the Almean languages. All the same, I think you exaggerate here. An OVS ergative language is unusual, but it certainly has precedent, most notably in Päri (Nilotic) and also in Kuikúro (Carib). There are some languages out there with weird alignment systems, and Lenani–Littoral is far from the strangest I’ve seen. Its only really odd quirk is in restricting verbal cross-referencing to overt arguments.

(Hmm, now that I think of it, what is the strangest? Not sure, but good candidates are Georgian (which flips around between head-marking and dependent-marking based on animacy), Manipuri (with purely semantic alignment), Rushani (famously with direct alignment), Nias (marked-absolutive, sort of), Saweru and Choctaw (both with three-way split intransitivity), Enxet Sur (claimed to have only intransitive verbs) and Movima (which is just un-summarisable, go read about it).)

[Ares Land]

I suspect that, to be perfectly naturalistic, a conlang grammar should be as unreadable as possible.

Ideally the author should redefine most linguistic terms, and introduce his own terminology and conventions, which he will force on the reader. These conventions should flout any and all conventions, and generally be as opaque as possible.
I've encountered that sort of thing with plenty of natlang grammars, most notably Nahuatl though I have a soft spot for Egyptian linguistics, where sḏm.t=f is a perfectly reasonable name for a verb form and "pseudo-participle" an excellent way to refer to something that is entirely unlike a participle.

The somewhat idiosyncratic use of "experiencer" pales in comparison to the horrors you actually find in linguistics work

[Raphael]

I'd say there should be a distinction between a naturalistic conlang and a naturalistic conlang grammar.

[WeepingElf]

R. M. W. Dixon once quipped about incomprehensible grammars that in many of them, the theoretical framework needed to understand them will probably be extinct long before the language described is.

[bradrn]

Ideally the author should redefine most linguistic terms, and introduce his own terminology and conventions, which he will force on the reader. These conventions should flout any and all conventions, and generally be as opaque as possible.
I've encountered that sort of thing with plenty of natlang grammars, most notably Nahuatl though I have a soft spot for Egyptian linguistics, where sḏm.t=f is a perfectly reasonable name for a verb form and "pseudo-participle" an excellent way to refer to something that is entirely unlike a participle.

I do feel that widely-spoken and classical languages tend to have some of the least useful terminology, simply because their study began before modern linguistic theory was developed. English traditional grammar is also a mess, for instance.

More modern grammars sometimes work this out, though not always. Fascinatingly, the Egyptian–Coptic volume I happen to have open right now notes that ‘I stick to the very old appellation Pseudo-Participle, because [… it is] completely void of any semantic or syntactic meaning, which has its advantages too’.

[So Haleza Grise]

I suspect that, to be perfectly naturalistic, a conlang grammar should be as unreadable as possible.

Ideally the author should redefine most linguistic terms, and introduce his own terminology and conventions, which he will force on the reader. These conventions should flout any and all conventions, and generally be as opaque as possible.
I've encountered that sort of thing with plenty of natlang grammars, most notably Nahuatl though I have a soft spot for Egyptian linguistics, where sḏm.t=f is a perfectly reasonable name for a verb form and "pseudo-participle" an excellent way to refer to something that is entirely unlike a participle.

The somewhat idiosyncratic use of "experiencer" pales in comparison to the horrors you actually find in linguistics work

^THIS. Good grief Andrews' Nahuatl book makes me roll my eyes.

I've read enough grammars to have developed the view that NO professional linguist uses terminology in exactly the same way as any of their peers.

[fusijui]

Andrew's Nahuatl grammar/textbook is the kind of enthusiastic oddness that just makes me happy, honestly. Unlike Egyptological grammar (or conventional/traditional Japanese, for that matter), which just comes across to me as grim solipsism in old pickle juice.

[WeepingElf]

I think it is the theoretical linguists to blame, and the curricula which make theoretical linguistics mandatory for linguistics students. The theoretical linguists constantly invent new concepts and terms to make people believe they had found out something new about language, but it is all old wine in new skins, and the grammar writers use (or misuse) the terminology.

[zompist]

I've looked at a lot of grammars searching for numbers. The absolute worst (I forget what language, it was from the Americas) was an SIL grammar which used tagmemics, the theory invented by SIL's president, Kenneth Pike. It was absolutely baffling-- the terminology was unrecognizable and simply finding things was a chore.

[bradrn]

I've looked at a lot of grammars searching for numbers. The absolute worst (I forget what language, it was from the Americas) was an SIL grammar which used tagmemics, the theory invented by SIL's president, Kenneth Pike. It was absolutely baffling-- the terminology was unrecognizable and simply finding things was a chore.

The very sad thing about this is the question that immediately came to mind: ‘which SIL tagmemics grammar’? Not only did he invent it, it was made mandatory for all SIL grammars of the time. As a result of which we have a whole number of grammars which might otherwise be excellent, but were spoiled by this obscure fad of a theory.

The funny thing is, tagmemics isn’t even a bad theory, as theories go. I’d say it’s rather more useful than most of them have been. And I still see terms like ‘syntagmeme’ popping up occasionally in modern grammars. But that just goes to show how stupid it is to base your grammar on anything too far beyond Basic Linguistic Theory (as Dixon calls it), because even the better ones can go out of fashion so quickly.

[Creyeditor]

I always wondered if Tagmemics maybe was the Basic Linguistic Theory (or maybe CxG) of its time? Everybody thought they finally found a sensible theory that is not centered on European language and that every structuralist (i.e. probably most linguists of the time) can agree on; this will surely stand the test of time. Except it didn't.
Re OP: I have never seen experiencer used in this way, but I have also not seen any good term for this except for S(ingle argument of an intransitive verb).

[WeepingElf]

Grammars written by conlangers are usually easier to understand than grammars written by academic linguists because of just that: the conlangers have no background in theoretical linguistics and therefore use simpler terminology (though they sometimes misuse terminology).

[bradrn]

Grammars written by conlangers are usually easier to understand than grammars written by academic linguists because of just that: the conlangers have no background in theoretical linguistics and therefore use simpler terminology (though they sometimes misuse terminology).

Rather, I suspect the more important factor there is that most conlangs are — let’s face it — massively simpler than the average natlang. I’ve seen very few conlang grammars which even approach the length of a thorough natlang grammar.

[WeepingElf]

Grammars written by conlangers are usually easier to understand than grammars written by academic linguists because of just that: the conlangers have no background in theoretical linguistics and therefore use simpler terminology (though they sometimes misuse terminology).

Rather, I suspect the more important factor there is that most conlangs are — let’s face it — massively simpler than the average natlang. I’ve seen very few conlang grammars which even approach the length of a thorough natlang grammar.

Yes. Conlangs are models of languages and therefore simpler than the "real thing".