Creating a Logography for a Fusional Language

[snappdragon]

So I am working on a proto-language for my conworld. I haven't name the conworld yet but I know it will be separated into two hemispherical "realms" separated by a magi-physical fence of sorts. Anyways, back to the point, this proto-language is for the Haethic language family, of the Haethion realm. It is a fusional language, and I wanted a way for the people to write it because, let's be honest, they'd probably be scratching writing into tree bark. The problem I'm having is that, historically, only ideographic systems and logographies would be around at this time period.

Proto-Haethic is a FUSIONAL LANGUAGE.

So, I need to find a way to represent a fusional language using a logographic system. I have a few ideas:

• Nouns and verbs would have a top determinative showing the declension or conjugation class, and the rest of the determinatives would be either semantic or phonetic.
• Prefixes and infixes (Proto-Haethic has no suffixes) would be shown using glyphs to the left (for prefixes) and right (for infixes). They would only give the information on which type of prefix to use - which set of prefixes would be determined by the class. Infixes would just be infixes, they wouldn't change from class to class. That'd be tedious.
• Adjectives and Adverbs would work the same way, except they wouldn't need anything other than semantic and phonetic determinatives, as things needed for agreement would be inferred from whatever it is modifying.

I know I'm currently way over my head right now (DARN YOU HISTORICAL ACCURACY, YOU'RE THE REASON I'M IN THIS MESS). Not like I can do anything about it though.

If anyone has any feedback or comments on this, please say something. I could use a bit of help here. Thank you :3 have an awesome day.

[Soap]

If you want to be really perverse you can spell out all of the grammatical inflections with symbols corresponding to the most similar-sounding content word that exists in the language. That way your inflections will take up 90% of the writing space and most people won't even be able to read. Hey, it worked out fine for the Japanese when they did the exact same thing.

Plan #2: What do your ideographs look like? Can you draw lines through them corresponding to the various inflections? Can an ideograph fit inside another ideograph? Both of those ideas are things I've seen in conlangs, and I think they both could work well, given appropriately shaped ideographs. And you could even do both.

[snappdragon]

If you want to be really perverse you can spell out all of the grammatical inflections with symbols corresponding to the most similar-sounding content word that exists in the language. That way your inflections will take up 90% of the writing space and most people won't even be able to read. Hey, it worked out fine for the Japanese when they did the exact same thing.

Plan #2: What do your ideographs look like? Can you draw lines through them corresponding to the various inflections? Can an ideograph fit inside another ideograph? Both of those ideas are things I've seen in conlangs, and I think they both could work well, given appropriately shaped ideographs. And you could even do both.

Option one probably wouldn't work. Then again, I'm still on creating the morphology so who knows.

As for option two, I don't know. I do know that they'd have to be made of straight lines (because tree bark). Lines through them could work, but there are a lot of inflections. I'd have to have a lot of lines.

[clawgrip]

Or you could just partially or completely ignore inflections altogether. It worked for Sumerian. Eventually they started doing rebus writing, i.e. Soap's first suggestion.

[snappdragon]

Or you could just partially or completely ignore inflections altogether. It worked for Sumerian. Eventually they started doing rebus writing, i.e. Soap's first suggestion.

Rebus writing might not work, or only work with certain words, as the lexicon could be very diverse for the language.

Alternatively, since it is a proto-language, I could just say they used ideograms and ignore the problem :p

[Duns Scotus]

• Nouns and verbs would have a top determinative showing the declension or conjugation class, and the rest of the determinatives would be either semantic or phonetic.
• Prefixes and infixes (Proto-Haethic has no suffixes) would be shown using glyphs to the left (for prefixes) and right (for infixes). They would only give the information on which type of prefix to use - which set of prefixes would be determined by the class. Infixes would just be infixes, they wouldn't change from class to class. That'd be tedious.

The first is precisely how I do it in Fyrthir. There are 5 genders, and ~12 "classes" of nouns, making for roughly 60 categories of nouns. Fyrthir is "verbless" in a restricted sense: verbs are a closed class which carry little semantic content, heavily relying on periphrastic constructions a la Jingulu, Basque, Japanese, and what spoken Welsh appears to moving towards, so I won't consider them here. Anyway, nouns inflect for case and number, although Fyrthir does feature suffixaufnahme, so can have more than one ending. Anyhow, there are a bit less than 60 different gender+class determinative glyphs, as not all possible combinations of gender/class are possible in the language and thus aren't provided. Moreover, similar to your Proto-Haethic, Fyrthir features infixes, as well as initial consonant mutation system that derived diachronically from a series of prefixes. In addition, Fyrthir has a system of ablaut, vaguely similar to Germanic languages but used in a very different way. In my grammar, I draw a very important distinction between roots and stems. Each root is an abstract concept, from which concrete stems are derived. That is, the vowel in each root is specified for height (of which there are 4) and otherwise underspecified. Ablaut causes the "abstract vowel" to be instantiated in various manners, differing in roundedness and frontness. Both infixing and ablaut are purely derivational. Thus, ablaut derives a concrete stem from an abstract root, and infixes furthermore derive stems from other stems. I bring this up because both infixing, ablaut, and initial mutation are all collectively referred to as stem modification, and indicated through diacritics on the logogram of the root. That is, the set of stems derived from a particular root share a base glyph. Note that ablaut of the form [-front,-round], and non-mutation (there's just one, unlike the Celtic languages) are unmarked. I haven't decided whether I will allow prefixes in the language.

Anyway, here's a basic template of how the system would work:

Code: Select all

  [mut.]
[ablaut]
  [root][det.][case][number]{[case][number]}*
 [infix]

note: the asterisk represents zero or more, as in regex

The point is that the glyphs used for case and number have no concrete pronunciation on there own, independent of a determinative glyph. I will tell you that in this system, I have made no attempt whatsoever to be naturalistic, so if that's your goal, seeing as how something like this has never really developed, I might avoid it. With that in mind, in-universe, my system was invented by a team of scholars, much like Hangul So perhaps if you want something like this to be justified, perhaps start out like Sumerian, as clawgrip suggests, then have a King (or someone) recognize the inadequacy of the system and assemble of team of scribes to develop a new system. It's also worth nothing that there an attested system vaguely similar to the idea of abstract morpheme glyphs: sometimes in Egyptian hieroglyphs, and in earlier inscriptions generally, plurals are represented by tripling a glyph, and duals by doubling. As I remember, this was eventually replaced with the use of a dual or plural indicator (two or three strokes) but with the addition of a phonetic spelling of the actual ending. (It's been a been awhile, so don't quote me on this). Given that the system moved towards spelling out the endings, it might suggest that abstract glyphs are diachronically unstable, but I don't know. Also shout to WeepingElf who I vaguely recall independently describing a logographic system with diacritics.

[snappdragon]

• Nouns and verbs would have a top determinative showing the declension or conjugation class, and the rest of the determinatives would be either semantic or phonetic.
• Prefixes and infixes (Proto-Haethic has no suffixes) would be shown using glyphs to the left (for prefixes) and right (for infixes). They would only give the information on which type of prefix to use - which set of prefixes would be determined by the class. Infixes would just be infixes, they wouldn't change from class to class. That'd be tedious.

The first is precisely how I do it in Fyrthir. There are 5 genders, and ~12 "classes" of nouns, making for roughly 60 categories of nouns. Fyrthir is "verbless" in a restricted sense: verbs are a closed class which carry little semantic content, heavily relying on periphrastic constructions a la Jingulu, Basque, Japanese, and what spoken Welsh appears to moving towards, so I won't consider them here. Anyway, nouns inflect for case and number, although Fyrthir does feature suffixaufnahme, so can have more than one ending. Anyhow, there are a bit less than 60 different gender+class determinative glyphs, as not all possible combinations of gender/class are possible in the language and thus aren't provided. Moreover, similar to your Proto-Haethic, Fyrthir features infixes, as well as initial consonant mutation system that derived diachronically from a series of prefixes. In addition, Fyrthir has a system of ablaut, vaguely similar to Germanic languages but used in a very different way. In my grammar, I draw a very important distinction between roots and stems. Each root is an abstract concept, from which concrete stems are derived. That is, the vowel in each root is specified for height (of which there are 4) and otherwise underspecified. Ablaut causes the "abstract vowel" to be instantiated in various manners, differing in roundedness and frontness. Both infixing and ablaut are purely derivational. Thus, ablaut derives a concrete stem from an abstract root, and infixes furthermore derive stems from other stems. I bring this up because both infixing, ablaut, and initial mutation are all collectively referred to as stem modification, and indicated through diacritics on the logogram of the root. That is, the set of stems derived from a particular root share a base glyph. Note that ablaut of the form [-front,-round], and non-mutation (there's just one, unlike the Celtic languages) are unmarked. I haven't decided whether I will allow prefixes in the language.

Anyway, here's a basic template of how the system would work:

Code: Select all

  [mut.]
[ablaut]
  [root][det.][case][number]{[case][number]}*
 [infix]

note: the asterisk represents zero or more, as in regex

The point is that the glyphs used for case and number have no concrete pronunciation on there own, independent of a determinative glyph. I will tell you that in this system, I have made no attempt whatsoever to be naturalistic, so if that's your goal, seeing as how something like this has never really developed, I might avoid it. With that in mind, in-universe, my system was invented by a team of scholars, much like Hangul So perhaps if you want something like this to be justified, perhaps start out like Sumerian, as clawgrip suggests, then have a King (or someone) recognize the inadequacy of the system and assemble of team of scribes to develop a new system. It's also worth nothing that there an attested system vaguely similar to the idea of abstract morpheme glyphs: sometimes in Egyptian hieroglyphs, and in earlier inscriptions generally, plurals are represented by tripling a glyph, and duals by doubling. As I remember, this was eventually replaced with the use of a dual or plural indicator (two or three strokes) but with the addition of a phonetic spelling of the actual ending. (It's been a been awhile, so don't quote me on this). Given that the system moved towards spelling out the endings, it might suggest that abstract glyphs are diachronically unstable, but I don't know. Also shout to WeepingElf who I vaguely recall independently describing a logographic system with diacritics.

... If my brain was a CPU, you just overheated it. But here's some things I picked out along with some responses.

• Proto-Haethic and Fyrthir have a few similarities.
• A semanto-grammatic logography isn't really naturalistic.
• I now know what ablaut is.
• This language was spoken earlier than that time period. As in "no king, just tribe village". Around the time this writing system started, civilization was just getting started as well. But, something similar can still be done. Maybe a tribe elder or equivalent.
• Diacritics? In a logography? WAT???

I believe I have learned a bit now. Thank you duns_scotus (aka "person who is a lot more intelligent than I when it comes to this stuff").

[Porphyrogenitos]

[*]Diacritics? In a logography? WAT???[/list]

Vietnamese Chữ Nôm used a diacritic to indicate a native reading of a character, as opposed to a Chinese reading, which was normally assumed, as described here:

Unmodified Chinese characters were used in chữ Nôm in three different ways.

A large proportion of Vietnamese vocabulary had been borrowed from Chinese from the Tang period. Such Sino-Vietnamese vocabulary could be written with the original Chinese character for each word, for example:[48]

役 dịch ("service", "corvee"), from Early Middle Chinese (EMC) /jwiajk/[49]
本 bản ("root", "foundation"), from EMC /pən'/[50]
頭 đầu ("head"), from EMC /dəw/[51]
To represent a native Vietnamese word, one method was to use a Chinese character for a Chinese word with a similar meaning. For example, 本 may also represent vốn ("capital, funds"). When a character would have two readings, a diacritic may be added to the character to indicate the "indigenous" reading. Thus when 本 is meant to be read as vốn, it is written as 本㆑, with a diacritic at the upper right corner. In this case the word vốn is actually an earlier Chinese loan that has become accepted as Vietnamese; William Hannas claims that all such readings are similar early loans.[48]

Alternatively, a native Vietnamese word could be written using a Chinese character for a Chinese word with a similar sound, regardless of the meaning of the Chinese word. For example, 沒 (Early Middle Chinese /mət/[52]) may represent the Vietnamese word một ("one").[53]

To draw an analogy to the Japanese writing system, the first two categories are similar to the on and kun readings of Japanese kanji respectively. The third is similar to ateji, in which characters are used only for their sound value, or the Man'yōgana script that became the origin of hiragana and katakana.

[WeepingElf]

I once had the idea of a logography in which inflectional categories are marked by diacritics. I never really worked that out, though.

[Duns Scotus]

I once had the idea of a logography in which inflectional categories are marked by diacritics. I never really worked that out, though.

Interesting. If you don't mind, would you be willing to expound on what you think caused it to not work?

[WeepingElf]

I once had the idea of a logography in which inflectional categories are marked by diacritics. I never really worked that out, though.

Interesting. If you don't mind, would you be willing to expound on what you think caused it to not work?

You misunderstood. It did not fail to work - I just never added flesh to the idea.

[Duns Scotus]

You misunderstood. It did not fail to work - I just never added flesh to the idea.

Gotcha. Thanks for the clarification.

[WeepingElf]

The idea was to create such a logography for Old Albic, but then I decided in favour of this featural alphabet. Perhaps, though, there was an earlier period of literacy in the Bronze Age, like Mycenean, with a logography with inflectional diacritics. Don't know yet.