How to Create a "True Language"
So, I'm pretty sure we all know that the Ancient Language kinda sucks. It's not even a true conlang, just an English cipher using words mostly stolen from Old Norse, with a few features of other languages thrown in randomly to differentiate it once Paolini started receiving criticism. However, my focus here is on why it sucks as a "true language," i.e. one that describes the true nature of things.
Is AL even a "true language"?
There is some disagreement on this point within canon. When Brom first explains the Ancient Lnaguage to Eragon, he says that it "has a name for everything, if you can find it," and that it "describes the true nature of things." However, Paolini later introduced the story that the Grey Folk bound magic to their language in order to make it more easily controllable. It could be argued that these stories are compatible, i.e. that the language of the Grey Folk was a form of "truespeech," but not yet linked to magic. It could also be argued that the second story is entirely bogus, since non-verbal magic is still entirely possible in-series. Thus, whichever way you slice it, the Ancient Language is, on some level, intended to be a kind of "truespeech."
So, why does it fail as such? There are many reasons, but to me, there are three that seem particularly damning.
Problem 1: Synonyms
In the scene referenced above, Brom tells Eragon that brisingr is "not only... a name for fire, it is the name for fire." Yet, the Ancient Language itself would seem to contradict this, for it contains another word for fire: istalrí. If the Ancient Language describes things as they truly are, and every word in it is the definitive name for the thing it describes, why would it contain any synonyms? Brisingr and istalrí can't both be the one true name for fire. It's a simple logical impossibility.
Problem 2: Dual Meanings
Some entires in Paolini's AL glossary list more than one potential meaning for a word, mostly when it comes to verbs. Examples include jierda, "break, hit"; stydja, "rest, remain"; malthinae, "bind, hold in place"; thrysta, "thrust, compress"; and vaetna, "dispel, scatter." While this is a natural feature of real languages, in a "truespeech" language, it would imply that these actions are essentially the same, or at the very least metaphysically connected. If multi-word definitions were inteded to show how difficult it was to translate a word into English, this could be understandable; however, that is not what he is doing here. In fact, I would doubt that he was able to do this if not for the example of the Dwarven thriknzdal, "the temper line on a differentially-tempered blade." Going into the failings of Paolini's Dwarvish would be a digression, but I do feel the need to state that despite its flaws, it is a far better conlang than the Ancient Language. I suspect that some of the added creativity Paolini shows in his Dwarvish is due to not being constrained by stealing words from Old Norse.
Problem 3: Lack of Complexity
This is a far bigger problem than the other two, and one which ties into many of the Ancient Language's technical flaws. One example is the lack of participles. Besides being utterly bizarre, this severely limits the ability to convey certain ideas, such as circumstance. Some of the pitfalls of missing participles could be avoided with an unusually complicated tense system, but since Paolini based AL's grammar on Engilsh, we don't get that.
The issue with this as related to "truespeech" is that one would expect such a language to be incredibly precise. Every sentence would describe exactly what is happening. To take an example from my dad's study of German, you would be able to distinguish between "she walked between the trees (as if through a gate)" and "she walked (back and forth) between the trees."
So how does one construct a "true language"?
Probably the biggest issue with constructing a form of "truespeech" is that the goal is very different from that of a regular conlang. When one builds a conlang, the goal is to create a plausible language. When one attempts to create "truespeech," however, the goal is a perfected language. Because every person's definiton of "perfect" is different, this is incredibly difficult; however, I will attempt to show how I would do it.
"Truespeech" is, by its nature, not the kind of thing that human language evolves into naturally. Though it can be used to communicate, its true purpose is to define and embody. To fulfill this role, it must be extremely complex and exacting. In a way, "truespeech" would be somewhere between a spoken language and an extremely precise programming language.
Phonetics and Phonology
This part of a "true language" may seem rather incidental, but can have unfortunate implications if neglected. A writer who bases the phonetics and phonology of a "true language" solely on their own real-life language could be seen as ethnocentric, or at the very least uncreative. Limiting your sound profile also limits the number of words you can theoretically construct with this language, which, in a "truespeech," can be severely limiting. Thus, when it comes to constructing a "true language," it's probably best to give it a large repertoire of phonemes. It will make the language interesting from the outset, potentially difficult to learn, and much easier to fill with barely-different, subtly-nuanced words.
Grammar
Because of the need for precision, grammar in a "true language" would be necessarily complicated. There are many ways to do this, including a vast number of highly-specific prepositions, heavy agglutination, heavy inflection, or some combination of the above. As a student of Indo-European languages of the Classical era, I would tend to favor inflection; as the adage says, "write what you know." However, I would not stick to the cases present in Ancient Greek, Classical Latin, or even our current reconstruction of Proto-Indo-European. If we want a language to truly describe everything, we would need a vast number of cases, or a moderately large number of cases plus some modifying prepositions. I mentioned earlier that a complicated enough tense system for verbs could eliminate the need for participles, e.g. with a set of circumstantial and adversative cases, but I would not recommend this. For one, it would be extremely difficult; for another, if you wish to implement a system of personal true names, participles might prove invaluable.
Vocabulary
As previously mentioned, the vocabulary of "truespeech" should be extremely wide. A language that purports to describe every single thing in existence needs nothing less. Of course, if you use your "true language" in a story, you only need to coin words that would actually appear, plus perhaps a few for interested fans. It would, however, be interesting to show the sheer nuance of your language in the text.
For example, a "true language" may have several words that for all practical purposes seem synonymous, but are actually not. To use the example of different words for "fire," perhaps one refers to the flames themselves, while one refers to the chemical reaction of combustion. In a pre-modern society, there would be little to no noticeable difference, though you might hint at their true meanings by having them work differently: the one that meant "fire itself" would create fire instantaneously, while the one that meant "combustion" would cause something to begin combusting, in most cases producing fire. It would be subtle, but it would be an interesting detail.
Semantics
This is perhaps the trickiest part of creating "truespeech." If you have people use it as a legitimate mode of communication, you're going to have a lot of work to do. If you declare that people can't lie in "truespeech," what does that mean? Can you say something if you sincerely believe it to be true (as in Inheritance), or can you not say anything that contradicts reality even if you do believe it (as with the red text in Umineko no Naku koro ni)? Can you use figurative language? These are all complicated questions that Paolini entirely sweeps under the rug with a line from Brom about how elves "say one thing and mean another." That phrasing could be an indication that elves use metaphor, but it could just as easily be a description of bald-faced deception.
Another factor to consider is whether colloquial use of "truespeech" might give rise to a corrupted form of the language. If "truespeech" can be corrupted into a more regular language, does the corrupted version cause unintended side effects by sounding similar to a phrase in the original language? How can the language that accurately describes everything in existence even be "corrupted" in the first place?
AL's Biggest Missed Opportunity
The fact that the Ancient Language sucks as a "true language" actually gives rise to a huge example of lost potential. If we take the Grey Folk binding their language to reality as canon, then the fact that it fails as "truespeech" could be said to have contributed to the continued possibility of wordless magic. This in turn could have led to a message about trying to impose your ideas of the world on others, and how no one language is truly better than any other. The attempt to bind reality to their language could be presented as the ultimate act of hubris, which in turn drove the Grey Folk to extinction. Unfortunately, Paolini seems to have intended them to be wise and self-sacrificing, not hubristic and paying the price for their arrogance. A part of me worries that if he happened upon this post, he would claim that this idea was his intent all along; however, his tendency to portray his protagonists and Sue-species as wonderful and flawless makes me feel a bit more secure.
Is AL even a "true language"?
There is some disagreement on this point within canon. When Brom first explains the Ancient Lnaguage to Eragon, he says that it "has a name for everything, if you can find it," and that it "describes the true nature of things." However, Paolini later introduced the story that the Grey Folk bound magic to their language in order to make it more easily controllable. It could be argued that these stories are compatible, i.e. that the language of the Grey Folk was a form of "truespeech," but not yet linked to magic. It could also be argued that the second story is entirely bogus, since non-verbal magic is still entirely possible in-series. Thus, whichever way you slice it, the Ancient Language is, on some level, intended to be a kind of "truespeech."
So, why does it fail as such? There are many reasons, but to me, there are three that seem particularly damning.
Problem 1: Synonyms
In the scene referenced above, Brom tells Eragon that brisingr is "not only... a name for fire, it is the name for fire." Yet, the Ancient Language itself would seem to contradict this, for it contains another word for fire: istalrí. If the Ancient Language describes things as they truly are, and every word in it is the definitive name for the thing it describes, why would it contain any synonyms? Brisingr and istalrí can't both be the one true name for fire. It's a simple logical impossibility.
Problem 2: Dual Meanings
Some entires in Paolini's AL glossary list more than one potential meaning for a word, mostly when it comes to verbs. Examples include jierda, "break, hit"; stydja, "rest, remain"; malthinae, "bind, hold in place"; thrysta, "thrust, compress"; and vaetna, "dispel, scatter." While this is a natural feature of real languages, in a "truespeech" language, it would imply that these actions are essentially the same, or at the very least metaphysically connected. If multi-word definitions were inteded to show how difficult it was to translate a word into English, this could be understandable; however, that is not what he is doing here. In fact, I would doubt that he was able to do this if not for the example of the Dwarven thriknzdal, "the temper line on a differentially-tempered blade." Going into the failings of Paolini's Dwarvish would be a digression, but I do feel the need to state that despite its flaws, it is a far better conlang than the Ancient Language. I suspect that some of the added creativity Paolini shows in his Dwarvish is due to not being constrained by stealing words from Old Norse.
Problem 3: Lack of Complexity
This is a far bigger problem than the other two, and one which ties into many of the Ancient Language's technical flaws. One example is the lack of participles. Besides being utterly bizarre, this severely limits the ability to convey certain ideas, such as circumstance. Some of the pitfalls of missing participles could be avoided with an unusually complicated tense system, but since Paolini based AL's grammar on Engilsh, we don't get that.
The issue with this as related to "truespeech" is that one would expect such a language to be incredibly precise. Every sentence would describe exactly what is happening. To take an example from my dad's study of German, you would be able to distinguish between "she walked between the trees (as if through a gate)" and "she walked (back and forth) between the trees."
So how does one construct a "true language"?
Probably the biggest issue with constructing a form of "truespeech" is that the goal is very different from that of a regular conlang. When one builds a conlang, the goal is to create a plausible language. When one attempts to create "truespeech," however, the goal is a perfected language. Because every person's definiton of "perfect" is different, this is incredibly difficult; however, I will attempt to show how I would do it.
"Truespeech" is, by its nature, not the kind of thing that human language evolves into naturally. Though it can be used to communicate, its true purpose is to define and embody. To fulfill this role, it must be extremely complex and exacting. In a way, "truespeech" would be somewhere between a spoken language and an extremely precise programming language.
Phonetics and Phonology
This part of a "true language" may seem rather incidental, but can have unfortunate implications if neglected. A writer who bases the phonetics and phonology of a "true language" solely on their own real-life language could be seen as ethnocentric, or at the very least uncreative. Limiting your sound profile also limits the number of words you can theoretically construct with this language, which, in a "truespeech," can be severely limiting. Thus, when it comes to constructing a "true language," it's probably best to give it a large repertoire of phonemes. It will make the language interesting from the outset, potentially difficult to learn, and much easier to fill with barely-different, subtly-nuanced words.
Grammar
Because of the need for precision, grammar in a "true language" would be necessarily complicated. There are many ways to do this, including a vast number of highly-specific prepositions, heavy agglutination, heavy inflection, or some combination of the above. As a student of Indo-European languages of the Classical era, I would tend to favor inflection; as the adage says, "write what you know." However, I would not stick to the cases present in Ancient Greek, Classical Latin, or even our current reconstruction of Proto-Indo-European. If we want a language to truly describe everything, we would need a vast number of cases, or a moderately large number of cases plus some modifying prepositions. I mentioned earlier that a complicated enough tense system for verbs could eliminate the need for participles, e.g. with a set of circumstantial and adversative cases, but I would not recommend this. For one, it would be extremely difficult; for another, if you wish to implement a system of personal true names, participles might prove invaluable.
Vocabulary
As previously mentioned, the vocabulary of "truespeech" should be extremely wide. A language that purports to describe every single thing in existence needs nothing less. Of course, if you use your "true language" in a story, you only need to coin words that would actually appear, plus perhaps a few for interested fans. It would, however, be interesting to show the sheer nuance of your language in the text.
For example, a "true language" may have several words that for all practical purposes seem synonymous, but are actually not. To use the example of different words for "fire," perhaps one refers to the flames themselves, while one refers to the chemical reaction of combustion. In a pre-modern society, there would be little to no noticeable difference, though you might hint at their true meanings by having them work differently: the one that meant "fire itself" would create fire instantaneously, while the one that meant "combustion" would cause something to begin combusting, in most cases producing fire. It would be subtle, but it would be an interesting detail.
Semantics
This is perhaps the trickiest part of creating "truespeech." If you have people use it as a legitimate mode of communication, you're going to have a lot of work to do. If you declare that people can't lie in "truespeech," what does that mean? Can you say something if you sincerely believe it to be true (as in Inheritance), or can you not say anything that contradicts reality even if you do believe it (as with the red text in Umineko no Naku koro ni)? Can you use figurative language? These are all complicated questions that Paolini entirely sweeps under the rug with a line from Brom about how elves "say one thing and mean another." That phrasing could be an indication that elves use metaphor, but it could just as easily be a description of bald-faced deception.
Another factor to consider is whether colloquial use of "truespeech" might give rise to a corrupted form of the language. If "truespeech" can be corrupted into a more regular language, does the corrupted version cause unintended side effects by sounding similar to a phrase in the original language? How can the language that accurately describes everything in existence even be "corrupted" in the first place?
AL's Biggest Missed Opportunity
The fact that the Ancient Language sucks as a "true language" actually gives rise to a huge example of lost potential. If we take the Grey Folk binding their language to reality as canon, then the fact that it fails as "truespeech" could be said to have contributed to the continued possibility of wordless magic. This in turn could have led to a message about trying to impose your ideas of the world on others, and how no one language is truly better than any other. The attempt to bind reality to their language could be presented as the ultimate act of hubris, which in turn drove the Grey Folk to extinction. Unfortunately, Paolini seems to have intended them to be wise and self-sacrificing, not hubristic and paying the price for their arrogance. A part of me worries that if he happened upon this post, he would claim that this idea was his intent all along; however, his tendency to portray his protagonists and Sue-species as wonderful and flawless makes me feel a bit more secure.